Atmel 11121 32 Bit Cortex A5 Microcontroller SAMA5D3 Datasheet

ARM-based Embedded MPU
SAMA5D3 Series
DATASHEET
Description
The Atmel SAMA5D3 series is a high-performance, power-efficient embedded MPU based on the ARM Cortex -A5 processor, achieving 536 MHz with power consumption levels below 0.5 mW in low-power mode. The device features a floating point unit for high-precision computing and accelerated data processing, and a high data bandwidth architecture. It integrates advanced user interface and connectivity peripherals and security features. The SAMA5D3 series features an internal multi-layer bus architecture associated with 39 DMA channels to sustain the high bandwidth required by the processor and the high-speed peripherals. The device offers support for DDR2/LPDDR/LPDDR2 and MLC NAND Flash memory with 24-bit ECC. The comprehensive peripheral set includes an LCD controller with overlays for hardware-accelerated image composition, a touchscreen interface and a CMOS sensor interface. Connectivity peripherals include Gigabit EMAC with IEEE1588, 10/100 EMAC, multiple CAN, UART, SPI and I2C. With its secure boot mechanism, hardware accelerated engines for encryption (AES, TDES) and hash function (SHA), the SAMA5D3 ensures anti-cloning, code protection and secure external data transfers. The SAMA5D3 series is optimized for control panel/HMI applications and applications that require high levels of connectivity in the industrial and consumer markets. Its lowpower consumption levels make the SAMA5D3 particularly suited for battery-powered devices. There are four SAMA5D3 devices in this series. Table 1-1 SAMA5D3 Devices shows the differences in the embedded features. All other features are available on all derivatives; this includes the three USB ports as well as the encryption engine and secure boot features.
11121BATARM08-Mar-13
1.
Features
z
Core
z
ARM Cortex-A5 Processor with ARM v7-A Thumb2 Instruction Set

z
CPU Frequency up to 536 MHz
z z
32 Kbyte Data Cache, 32 Kbyte Instruction Cache, Virtual Memory System Architecture (VMSA) Fully Integrated MMU and Floating Point Unit (VFPv4) One 160 Kbyte Internal ROM Single-cycle Access at System Speed, Embedded Boot Loader: Boot on NAND Flash, SDCard, eMMC, serial DataFlash, selectable Order One 128 Kbyte Internal SRAM, Single-cycle Access at System Speed High Bandwidth 32-bit Multi-port Dynamic Ram Controller supporting 512 Mbyte 8 bank DDR2/LPDDR/LPDDR2 Independent Static Memory Controller with SLC/MLC NAND Support with up to 24-bit Error Correcting Code (PMECC) Power-on Reset Cells, Reset Controller, Shut Down Controller, Periodic Interval Timer, Watchdog Timer and Real-time Clock Boot Mode Select Option, Remap Command Internal Low-power 32 kHz RC Oscillator and Fast 12 MHz RC Oscillators Selectable 32768 Hz Low-power Oscillator and 12 MHz Oscillator One 400 to 1000 MHz PLL for the System and one PLL at 480 MHz optimized for USB High Speed 39 DMA Channels including two 8-channel 64-bit Central DMA Controllers 64-bit Advanced Interrupt Controller Three Programmable External Clock Signals Programmable Fuse Box with 256 fuse bits, 192 of them available for Customer Shut Down Controller Battery Backup Registers Clock Generator and Power Management Controller Very Slow Clock Operating Mode, Software Programmable Power Optimization Capabilities LCD TFT Controller with Overlay, Alpha-blending, Rotation, Scaling and Color Space Conversion ITU-R BT. 601/656 Image Sensor Interface Three HS/FS/LS USB Ports with On-Chip Transceivers
z z
Memories
z z z z
System running up to 166 MHz

z z z z z z z z z
Low Power Management

z z z z
Peripherals
z z z
One Device Controller One Host Controller with Integrated Root Hub (3 Downstream Ports)
z z z z z z z z z z
One 10/100/1000 Mbps Gigabit Ethernet Mac Controller (GMAC) with IEEE1588 support One 10/100 Mbps Ethernet Mac Controller (EMAC) Two CAN Controllers with 8 Mailboxes, fully Compliant with CAN 2.0 Part A and 2.0 Part B Softmodem Interface Three High Speed Memory Card Hosts (eMMC 4.3 and SD 2.0) Two Master/Slave Serial Peripheral Interface Two Synchronous Serial Controllers Three Two-wire Interface up to 400 Kbits supporting I2C Protocol and SMBUS Four USARTs, two UARTs, one DBGU Two Three-channel 32-bit Timer/Counters
SAMA5D3 Series [DATASHEET]

z z z
One Four-channel 16-bit PWM Controller One 12-channel 12-bit Analog-to-Digital Converter with Resistive Touch-Screen function Write Protected Registers TRNG: True Random Number Generator Encryption Engine
z z z
Security
z z
AES: 256-bit, 192-bit, 128-bit Key Algorithm, Compliant with FIPS PUB 197 Specifications TDES: Two-key or Three-key Algorithms, Compliant with FIPS PUB 46-3 Specifications SHA: Supports Secure Hash Algorithm (SHA1, SHA224, SHA256, SHA384, SHA512)
Atmel Secure Boot Solution Five 32-bit Parallel Input/Output Controllers 160 I/Os Input Change Interrupt Capability on Each I/O Line, Selectable Schmitt Trigger Input Individually Programmable Open-drain, Pull-up and Pull-down Resistor, Synchronous Output, Filtering Slew Rate Control on High Speed I/Os Impedance Control on DDR I/Os 324-ball LFBGA, pitch 0.8 mm
I/O
z z z z z z
Package
z
Table 1-1.
SAMA5D3 Devices SAMA5D31 SAMA5D33 X X X X X X X X SAMA5D34 X X X X X X X X X SAMA5D35
LCDC GMAC EMAC CAN0, CAN1 HSMCI2 UART0 UART1 TC1

2.
Block Diagram
Figure 2-1. SAMA5D3 Block Diagram

DH S DH DP SD /HH M/ SD HH PA G1 SD 25 MA GT CK -G X GT CK 12 5 X GC EN GRX CKO -G C GRRS, TX K X G E GR ER CO R - L GT X0- GRX X GR D GM 0-G X7 V D TX7 ER C, G M ET EFC DI O X K EC EN R ER SD V, X E T 0- E E R X R X EM 0-ET X1 ER DC X1 ,E LC MD LCD-D D_ AT0 IO LC V -L D_ SY CD LC PC NC _ D_ K , L DA DE , LC CD T23 N, D_ _H LC D S IS I_D D_ ISP YN C PW IS OI_P IS M I _ C IS I_H K D11 SY NC , IS I_V SY NC DD R_ CA LP DD R_ CA LN
NT RS T TD I TD O TM TC S/SW K/S D WC IO LK JTA GS EL
TST
BMS
HH S HH DPC SD H H MC HH SDP SD B MB VB G
SysC
FIQ IRQ
PIO
JTAG / SWD
AIC
In-Circuit Emulator
HS Trans
HS Trans
HS Trans
PIO
DDR_VREF DDR_A0-DDR_A13 DDR_D0-DDR_D31 DDR_DQM[3..0] DDR_DQS[3..0]
DRXD DTXD PCK0-PCK2
DBGU
PC PB
PA HS USB Device DMA GMAC 10/100/1000 EMAC 10/100 LCD ISI
Cortex-A5
PLLA PLLUTMI
ICache 32 KB MMU BIU
VFP DCache 32 KB
HS EHCI USB HOST DMA
DDR2 LPDDR2 512 MB
XIN XOUT
Osc12 MHz
PMC
DMA
DMA
DMA
DMA
12 MHZ RC Osc WDT RC

XIN32 XOUT32 SHDN WKUP VDDBU NRST OSC 32K
I/D
PIT
4 GPBR EBI
DDR_DQSN[3..0] DDR_CS DDR_CLK,DDR_CLKN DDR_CKE DDR_RAS, DDR_CAS DDR_WE DDR_BA[2..0] D0-D15 A21/NANDALE A22/NANDCLE NRD/NANDOE NWE/NWR0/NANDWE NCS3/NANDCS NANDRDY A0/NBS0 A1-A20 A23-A25 NWR1/NBS1 NCS0,NCS1,NCS2 NWAIT
Multi-Layer Matrix
RTC RSTC
TRNG SHA AES TDES DMA
SHDC POR POR PIOA PIOC PIOE
NAND Flash Controller MCL/SLC ECC (4 KB SRAM)
PIOB PIOD
ROM 160 KB
SRAM0 64 KB
SRAM1 64 KB
8-CH DMA0
8-CH DMA1
Peripheral Bridge
Reduced Static Memory Controller
PIO
DMA CAN0 CAN1 TWI0 TWI1 TWI2
DMA USART0 USART1 USART2 USART3
DMA
DMA SPI0 SPI1
DMA SSC0 SSC1
DMA MCI0/MCI1/MCI2 SD/SDIO eMMC TC0, TC1 TC2, TC3 TC4, TC5
DMA 4-CH PWM
UART0 UART1
Real-time Events
DMA 12-CH 12-bit ADC TouchScreen SMD
PIO
P
N CA RX NT 0-C X0 A -C NR A X TW NT 1 TW D0 X1 CK -TW 0TW D2 CK CT 2 S RT 03 SCS03 RDK0X 3 TX 0-3 UR D0 D -3 UT X0NP UR X CS D0 D X 1, NP -UT 1 XD CS 1 2, NP C NP S3 CS SP 0 C M K O M SI TK ISO 0 TF -TK TD 0-T 1 F 0 RD -T 1 0 D RF -RD1 0 1 RK -RF 0 1 M -RK CI 1 M 0_C CI D M 1_ A CI C D 2 M _C A C D M I0_ A CI C M MC 1_CK CI I K M 0_D 2_C CI A K M 1_D [7..0 CI A ] 2 [ TI _D 3..0 O A ] TI A0 [3.. O -T 0] B TC 0 IO -T A PW LK0 IOB5 M -TC 5 LK PW H0 5 PW ML PW M 0-P MH FI W 3 0- M PW L3 M TS FI AD 3 TR AD IG 0U AD L 1 AD UR 2 AD LL GP 3 AD A LR 5- D4 G P TS PAD I AD 11 VR EF
DIB
SPI0_, SPI1_
CA

DI
BN
3.
Signal Description
Table 3-1 gives details on the signal names classified by peripheral.
Signal Description List Function Type Active Level Frequency (MHz) Comments
Table 3-1.
Signal Name Clocks, Oscillators and PLLs XIN XOUT XIN32 XOUT32 VBG PCK0 - PCK2 Main Oscillator Input Main Oscillator Output Slow Clock Oscillator Input Slow Clock Oscillator Output Bias Voltage Reference for USB Programmable Clock Output Input Output Input Output Analog Output
Shutdown, Wake-up Logic SHDN WKUP Shut-Down Control Wake-Up Input ICE and JTAG TCK/SWCLK TDI TDO TMS/SWDIO JTAGSEL Test Clock/Serial Wire Clock Test Data In Test Data Out Test Mode Select/Serial Wire Input/Output JTAG Selection Reset/Test NRST TST NTRST BMS Microcontroller Reset Test Mode Select Test Reset Signal Boot Mode Select I/O Input Input Input Debug Unit - DBGU DRXD DTXD Debug Receive Data Debug Transmit Data Input Output Advanced Interrupt Controller - AIC IRQ FIQ External Interrupt Input Fast Interrupt Input Input Input Low Input Input Output I/O Input Output Input
PIO Controller - PIOA - PIOB - PIOC - PIOD - PIOE PA0 - PAxx PB0 - PBxx PC0 - PCxx PD0 - PDxx PE0 - PExx Parallel IO Controller A Parallel IO Controller B Parallel IO Controller C Parallel IO Controller D Parallel IO Controller E I/O I/O I/O I/O I/O

Table 3-1.
Signal Description List (Continued) Function Type Active Level Frequency (MHz) Comments
Signal Name External Bus Interface - EBI D0 - D15 A0 - A25 NWAIT Data Bus Address Bus External Wait Signal I/O Output Input Static Memory Controller - HSMC NCS0 - NCS3 NWR0 - NWR1 NRD NWE NBS0 - NBS1 NANDOE NANDWE Chip Select Lines Write Signal Read Signal Write Enable Byte Mask Signal NAND Flash Output Enable NAND Flash Write Enable Output Output Output Output Output Output Output DDR2/LPDDR Controller DDR_VREF DDR_CALP DDR_CALN DDR_CK, DDR_CKN DDR_CKE DDR_CS DDR_BA[2..0] DDR_WE DDR_RAS, DDR_CAS DDR_A[13..0] DDR_D[31..0] DQS[3..0] DQSN[3..0] DQM[3..0] Reference Voltage Positive Calibration Reference Negative Calibration Reference DDR2 differential clock DDR2 Clock Enable DDR2 Controller Chip Select Bank Select DDR2 Write Enable Row and Column Signal DDR2 Address Bus DDR2 Data Bus Differential Data Strobe DQSN must be connected to DDR_VREF for DDR2 memories Write Data Mask Input Input Input Output Output Output Output Output Output Output I/O/-PD I/O- PD I/O- PD Output
Low
Low Low Low Low Low Low Low
High Low Low Low Low
High Speed Multimedia Card Interface - HSMCI0-2 MCI0_CK, MCI1_CK, MCI2_CK MCI0_CDA,MCI1_C DA, MCI2_CDA MCI0_DA[7..0] MCI1_DA[3..0] MCI2_DA[3..0) Multimedia Card Clock Multimedia Card Command Multimedia Card 0 Data Multimedia Card 1 Data Multimedia Card 2 Data I/O I/O I/O I/O I/O

Table 3-1.
Signal Name
Universal Synchronous Asynchronous Receiver Transmitter- USART0-3 SCKx TXDx RXDx RTSx CTSx USARTx Serial Clock USARTx Transmit Data USARTx Receive Data USARTx Request To Send USARTx Clear To Send I/O Output Input Output Input
Universal Asynchronous Receiver Transmitter - UARTx [1..0] UTXDx URXDx UARTx Transmit Data UARTx Receive Data Output Input
Synchronous Serial Controller - SSCx [1..0] TDx RDx TKx RKx TFx RFx SSC Transmit Data SSC Receive Data SSC Transmit Clock SSC Receive Clock SSC Transmit Frame Sync SSC Receive Frame Sync Timer/Counter - TCx [5..0] TCLKx TIOAx TIOBx TC Channel x External Clock Input TC Channel x I/O Line A TC Channel x I/O Line B Input I/O I/O Output Input I/O I/O I/O I/O
Serial Peripheral Interface - SPIx [1..0] SPIx_MISO SPIx_MOSI SPIx_SPCK SPIx_NPCS0 SPIx_NPCS[3..1] Master In Slave Out Master Out Slave In SPI Serial Clock SPI Peripheral Chip Select 0 SPI Peripheral Chip Select I/O I/O I/O I/O Output Low Low
Two-Wire Interface -TWIx [2..0] TWDx TWCKx Two-wire Serial Data Two-wire Serial Clock I/O I/O CAN controller - CANx CANRXx CANTXx CAN input CAN output Input Output Soft Modem - SMD DIBN DIBP Soft Modem Signal Soft Modem Signal I/O I/O

Table 3-1.
Signal Name
Pulse Width Modulation Controller- PWMC PWMH[3..0] PWML[3..0] PWMFIx PWM Waveform Output High PWM Waveform Output LOW PWM Fault Input Output Output Input USB Host High Speed Port - UHPHS HHSDPA HHSDMA HHSDPB HHSDMB HHSDPC HHSDMC USB Host Port A High Speed Data + USB Host Port A High Speed Data USB Host Port B High Speed Data + USB Host Port B High Speed Data USB Host Port C High Speed Data + USB Host Port C High Speed Data Analog Analog Analog Analog Analog Analog
USB Device High Speed Port - UDPHS DHSDP DHSDM USB Device High Speed Data + USB Device High Speed Data Analog Analog
GIgabit Ethernet 10/100/1000 - GMAC GTXCK G125CK G125CKO GTXEN GTX[7..0] GTXER GRXCK GRXDV GRX[7..0] GRXER GCRS GCOL GMDC GMDIO Transmit Clock or Reference Clock 125 MHz input Clock 125 MHz output Clock Transmit Enable Transmit Data Transmit Coding Error Receive Clock Receive Data Valid Receive Data Receive Error Carrier Sense and Data Valid Collision Detect Management Data Clock Management Data Input/Output Input Input Output Output Output Output Input Input Input Input Input Input Output I/O
RMII Ethernet 10/100 - EMAC EREFCK ETXEN ETX[1..0] ECRSDV ERX[1..0] ERXER Transmit Clock or Reference Clock Transmit Enable Transmit Data Carrier Sense/Data Valid Receive Data Receive Error Input Output Output Input Input Input

Table 3-1.
Signal Description List (Continued) Function Type Output I/O Active Level Frequency (MHz) Comments
Signal Name EMDC EMDIO Management Data Clock Management Data Input/Output
LCD Controller - LCDC LCDDAT[23..0] LCDVSYNC LCDHSYNC LCDPCK LCDDEN LCDPWM LCDDISP LCD Data Bus LCD Vertical Synchronization LCD Horizontal Synchronization LCD pixel Clock LCD Data Enable LCDPWM for Contrast Control LCD Display ON/OFF Output Output Output Output Output Output Output Image Sensor Interface - ISI ISI_D[11..0] ISI_HSYNC ISI_VSYNC ISI_PCK Image Sensor Data Image Sensor Horizontal Synchro Image Sensor Vertical Synchro Image Sensor Data clock Input input input input
Touch Screen Analog-to-Digital Converter - ADC AD0UL AD1UR AD2LL AD3LR AD4PI AD5-AD11 ADTRG ADVREF Upper Left Touch Panel Upper Right Touch Panel Lower Left Touch Panel Lower Right Touch Panel Panel Input 7 Analog Inputs ADC Trigger ADC Reference Analog Analog Analog Analog Analog Analog Input Analog

4.
Package and Pinout

The SAMA5D3 product is available in a 324-ball LFBGA package.
4.1
Mechanical Overview of the 324-ball LFBGA Package

Figure 4-1 shows the orientation of the 324-ball BGA Package.
Figure 4-1. Orientation of the 324-ball LFBGA Package

Bottom VIEW
V U T R P N M L K J H G F E D C B A
1 2
3 4 5 6 7 8 9 10 11 12 13 14 15
16 17 18

10
4.2
324-ball LFBGA Package Pinout

SAMA5D3 Pinout for 324-ball LFBGA Package
Primary Alternate PIO Peripheral A PIO Peripheral B PIO Peripheral C Reset State Signal, Dir, PU, PD, HiZ, ST
PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST ISI_D0 ISI_D1 TWD2 TWCK2 PWMH0 PWML0 PWMH1 PWML1 I/O O O O O O ISI_D2 ISI_D3 ISI_D4 ISI_D5 ISI_D6 ISI_D7 I I I I I I I I PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST URXD1 UTXD1 PWMH0 PWML0 TK1 TF1 PWMH1 PWML1 TD1 RK1 I O O O I/O I/O O O O I ISI_VSYNC ISI_HSYNC I I PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST
Table 4-1.
Pin
E3 F5 D2 F4 D1 J10 G4 J9 F3 J8 E2 K8 F2 G6 E1 H5 H3 H6 H4 H7 H2 J6 G2 J5 F1 J4 G3 J3 G1 K4 H1 K3 T2 N7 T3 N6 P5 T4 R4 U1
Power Rail
VDDIOP0 VDDIOP0 VDDIOP0 VDDIOP0 VDDIOP0 VDDIOP0 VDDIOP0 VDDIOP0 VDDIOP0 VDDIOP0 VDDIOP0 VDDIOP0 VDDIOP0 VDDIOP0 VDDIOP0 VDDIOP0 VDDIOP0 VDDIOP0 VDDIOP0 VDDIOP0 VDDIOP0 VDDIOP0 VDDIOP0 VDDIOP0 VDDIOP0 VDDIOP0 VDDIOP0 VDDIOP0 VDDIOP0 VDDIOP0 VDDIOP0 VDDIOP0 VDDIOP1 VDDIOP1 VDDIOP1 VDDIOP1 VDDIOP1 VDDIOP1 VDDIOP1 VDDIOP1
I/O Type
GPIO GPIO GPIO GPIO GPIO GPIO GPIO GPIO GPIO GPIO GPIO GPIO GPIO GPIO GPIO GPIO GPIO GPIO GPIO GPIO GPIO GPIO GPIO GPIO GPIO GPIO GPIO GPIO GPIO_CLK2 GPIO GPIO GPIO GMAC GMAC GMAC GMAC GMAC GMAC GMAC GMAC
Signal
PA0 PA1 PA2 PA3 PA4 PA5 PA6 PA7 PA8 PA9 PA10 PA11 PA12 PA13 PA14 PA15 PA16 PA17 PA18 PA19 PA20 PA21 PA22 PA23 PA24 PA25 PA26 PA27 PA28 PA29 PA30 PA31 PB0 PB1 PB2 PB3 PB4 PB5 PB6 PB7
Dir
I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O
Signal
Dir
Signal
LCDDAT0 LCDDAT1 LCDDAT2 LCDDAT3 LCDDAT4 LCDDAT5 LCDDAT6 LCDDAT7 LCDDAT8 LCDDAT9 LCDDAT10 LCDDAT11 LCDDAT12 LCDDAT13 LCDDAT14 LCDDAT15 LCDDAT16 LCDDAT17 LCDDAT18 LCDDAT19 LCDDAT20 LCDDAT21 LCDDAT22 LCDDAT23 LCDPWM LCDDISP LCDVSYNC LCDHSYNC LCDPCK LCDDEN TWD0 TWCK0 GTX0 GTX1 GTX2 GTX3 GRX0 GRX1 GRX2 GRX3
Dir
O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O I/O O O O O O I I I I
Signal
Dir
Signal
Dir

11
Table 4-1.
SAMA5D3 Pinout for 324-ball LFBGA Package (Continued)

PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST GTX4 GTX5 GTX6 GTX7 GRX4 GRX5 GRX6 GRX7 G125CKO O O O O I I I I O PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST TIOA3 TIOB3 TCLK3 TIOA4 TIOB4 TCLK4 TIOA5 TIOB5 TCLK5 I/O I/O I I/O I/O I I/O I/O I PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST LCDDAT20 LCDDAT19 TIOA1 TIOB1 TCLK1 PCK2 I/O I/O I O LCDDAT18 LCDDAT17 LCDDAT16 LCDDAT21 O O O O O O PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST
Pin
R5 P3 R6 V3 P6 V1 R7 U3 P7 V2 V5 T6 N8 U4 M7 U5 M8 T5 N9 V4 M9 P8 M10 R9 D8 A4 E8 A3 A2 F8 B3 G8 B4 F7 A1 D7 C6 E7 B2 F6 B1 E6 C3
Power Rail
VDDIOP1 VDDIOP1 VDDIOP1 VDDIOP1 VDDIOP1 VDDIOP1 VDDIOP1 VDDIOP1 VDDIOP1 VDDIOP1 VDDIOP1 VDDIOP1 VDDIOP1 VDDIOP1 VDDIOP1 VDDIOP1 VDDIOP1 VDDIOP1 VDDIOP1 VDDIOP1 VDDIOP1 VDDIOP1 VDDIOP0 VDDIOP0 VDDIOP0 VDDIOP0 VDDIOP0 VDDIOP0 VDDIOP0 VDDIOP0 VDDIOP0 VDDIOP0 VDDIOP0 VDDIOP0 VDDIOP0 VDDIOP0 VDDIOP0 VDDIOP0 VDDIOP0 VDDIOP0 VDDIOP0 VDDIOP0 VDDIOP0
I/O Type
GMAC GMAC GMAC GMAC GMAC GMAC GMAC GMAC GMAC GMAC GMAC GMAC GMAC GMAC GMAC GMAC GMAC GMAC GMAC GPIO GPIO GPIO GPIO GPIO GPIO GPIO GPIO GPIO GPIO GPIO GPIO GPIO GPIO GPIO GPIO GPIO GPIO GPIO GPIO MCI_CLK GPIO GPIO GPIO
Signal
PB8 PB9 PB10 PB11 PB12 PB13 PB14 PB15 PB16 PB17 PB18 PB19 PB20 PB21 PB22 PB23 PB24 PB25 PB26 PB27 PB28 PB29 PB30 PB31 PC0 PC1 PC2 PC3 PC4 PC5 PC6 PC7 PC8 PC9 PC10 PC11 PC12 PC13 PC14 PC15 PC16 PC17 PC18
Dir
I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O
Signal
Dir
Signal
GTXCK GTXEN GTXER GRXCK GRXDV GRXER GCRS GCOL GMDC GMDIO G125CK MCI1_CDA MCI1_DA0 MCI1_DA1 MCI1_DA2 MCI1_DA3 MCI1_CK SCK1 CTS1 RTS1 RXD1 TXD1 DRXD DTXD ETX0 ETX1 ERX0 ERX1 ETXEN ECRSDV ERXER EREFCK EMDC EMDIO MCI2_CDA MCI2_DA0 MCI2_DA1 MCI2_DA2 MCI2_DA3 MCI2_CK TK0 TF0 TD0
Dir
I O O I I I I I O I/O I I/O I/O I/O I/O I/O I/O I/O I O I O I O O O I I O I I I O I/O I/O I/O I/O I/O I/O I/O I/O I/O O
Signal
PWMH2 PWML2 RF1 RD1 PWMH3 PWML3 CANRX1 CANTX1
Dir
O O I/O I O O I O
Signal
Dir

12
Table 4-1.

PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST TWD1 TWCK1 PWMFI0 PWMFI2 I/O O I I ISI_D11 ISI_D10 ISI_D9 ISI_D8 ISI_PCK PWMFI1 I I I I I O PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST TIOA0 TIOB0 TCLK0 I/O I/O I PWMH2 PWML2 PWMH3 PWML3 O O O O PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST SPI0_NPCS1 SPI0_NPCS2 SPI0_NPCS3 O O O CANRX0 CANTX0 PWMFI3 I O I PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST PIO, I, PU, ST
Pin
D6 C4 D5 C2 G9 C1 H10 H9 D4 H8 G5 D3 E4 K5 P1 K6 R1 L7 P2 L8 R2 K7 U2 K9 M5 K10 N4 L9 N3 L10 N5 M6 T1 N2 M3 M2 L3 M1 N1 L1 L2 K1 K2
Power Rail
VDDIOP0 VDDIOP0 VDDIOP0 VDDIOP0 VDDIOP0 VDDIOP0 VDDIOP0 VDDIOP0 VDDIOP0 VDDIOP0 VDDIOP0 VDDIOP0 VDDIOP0 VDDIOP1 VDDIOP1 VDDIOP1 VDDIOP1 VDDIOP1 VDDIOP1 VDDIOP1 VDDIOP1 VDDIOP1 VDDIOP1 VDDIOP1 VDDIOP1 VDDIOP1 VDDIOP1 VDDIOP1 VDDIOP1 VDDIOP1 VDDIOP1 VDDIOP1 VDDIOP1 VDDANA VDDANA VDDANA VDDANA VDDANA VDDANA VDDANA VDDANA VDDANA VDDANA
I/O Type
GPIO GPIO GPIO GPIO GPIO GPIO_CLK GPIO GPIO GPIO GPIO GPIO GPIO GPIO GPIO GPIO GPIO GPIO GPIO GPIO GPIO GPIO GPIO MCI_CLK GPIO GPIO GPIO_CLK GPIO GPIO GPIO GPIO GPIO GPIO GPIO GPIO_ANA GPIO_ANA GPIO_ANA GPIO_ANA GPIO_ANA GPIO_ANA GPIO_ANA GPIO_ANA GPIO_ANA GPIO_ANA
Signal
PC19 PC20 PC21 PC22 PC23 PC24 PC25 PC26 PC27 PC28 PC29 PC30 PC31 PD0 PD1 PD2 PD3 PD4 PD5 PD6 PD7 PD8 PD9 PD10 PD11 PD12 PD13 PD14 PD15 PD16 PD17 PD18 PD19 PD20 PD21 PD22 PD23 PD24 PD25 PD26 PD27 PD28 PD29
Dir
I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O
Signal
Dir
Signal
RK0 RF0 RD0 SPI1_MISO SPI1_MOSI SPI1_SPCK SPI1_NPCS0 SPI1_NPCS1 SPI1_NPCS2 SPI1_NPCS3 URXD0 UTXD0 FIQ MCI0_CDA MCI0_DA0 MCI0_DA1 MCI0_DA2 MCI0_DA3 MCI0_DA4 MCI0_DA5 MCI0_DA6 MCI0_DA7 MCI0_CK SPI0_MISO SPI0_MOSI SPI0_SPCK SPI0_NPCS0 SCK0 CTS0 RTS0 RXD0 TXD0 ADTRG AD0 AD1 AD2 AD3 AD4 AD5 AD6 AD7 AD8 AD9
Dir
I/O I/O I I/O I/O I/O I/O O O O I O I I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I O I O I I I I I I I I I I I
Signal
Dir
Signal
Dir

13
Table 4-1.

PIO, I, PU, ST PIO, I, PU, ST A,I, PD, ST A,I, PD, ST A,I, PD, ST A,I, PD, ST A,I, PD, ST A,I, PD, ST A,I, PD, ST A,I, PD, ST A,I, PD, ST A,I, PD, ST A,I, PD, ST A,I, PD, ST A,I, PD, ST A,I, PD, ST A,I, PD, ST SCK3 CTS3 RTS3 RXD3 TXD3 SCK2 I/O I O I O I/O A,I, PD, ST A,I, PD, ST A,I, PD, ST A,I, PD, ST A,I, PD, ST A,I, PD, ST A,I, PD, ST A,I, PD, ST CTS2 RTS2 RXD2 TXD2 TIOA2 TIOB2 TCLK2 I O I O I/O I/O I LCDDAT22 LCDDAT23 O O A,I, PD, ST A,I, PD, ST A,I, PD, ST A,I, PD, ST PIO,I, PD, ST PIO, I, PD, ST PIO, I, PD, ST PIO, I, PD, ST PWML1 O PIO,I, PD, ST I, PD, I I O I O O I, ST I, PU, ST
Pin
J1 J2 P13 R14 R13 V18 P14 U18 T18 R15 P17 P15 P18 R16 N16 R17 N17 R18 N18 P16 M18 N15 M15 N14 M17 M13 M16 N12 M14 M12 L13 L15 L14 L16 U15 U9 U8 V8 U16 V16 T12 T10 V9
Power Rail
VDDANA VDDANA VDDIOM VDDIOM VDDIOM VDDIOM VDDIOM VDDIOM VDDIOM VDDIOM VDDIOM VDDIOM VDDIOM VDDIOM VDDIOM VDDIOM VDDIOM VDDIOM VDDIOM VDDIOM VDDIOM VDDIOM VDDIOM VDDIOM VDDIOM VDDIOM VDDIOM VDDIOM VDDIOM VDDIOM VDDIOM VDDIOM VDDIOM VDDIOM VDDBU VDDIOP0 VDDIOP0 VDDIOP0 VDDBU VDDBU VDDBU VDDBU VDDIOP0
I/O Type
GPIO_ANA GPIO_ANA EBI EBI EBI EBI EBI EBI EBI EBI EBI EBI EBI EBI EBI EBI EBI EBI EBI EBI EBI EBI EBI EBI EBI EBI EBI EBI EBI EBI EBI EBI EBI EBI SYSC SYSC CLOCK CLOCK CLOCK CLOCK SYSC SYSC RSTJTAG
Signal
PD30 PD31 PE0 PE1 PE2 PE3 PE4 PE5 PE6 PE7 PE8 PE9 PE10 PE11 PE12 PE13 PE14 PE15 PE16 PE17 PE18 PE19 PE20 PE21 PE22 PE23 PE24 PE25 PE26 PE27 PE28 PE29 PE30 PE31 TST BMS XIN XOUT XIN32 XOUT32 SHDN WKUP NRST
Dir
I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I I I O I O O I I/O
Signal
Dir
Signal
AD10 AD11 A0/NBS0 A1 A2 A3 A4 A5 A6 A7 A8 A9 A10 A11 A12 A13 A14 A15 A16 A17 A18 A19 A20 A21/NANDALE A22/NANDCLE A23 A24 A25 NCS0 NCS1 NCS2 NWR1/NBS1 NWAIT IRQ
Dir
I I O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O I I
Signal
PCK0 PCK1
Dir
O O
Signal
Dir

14
Table 4-1.

I, PU, ST I, ST O SWDIO SWCLK I/O I I, ST I, ST I, PD O, PU O, PU I, PD I, PD I, PD I, PD I, PD I, PD I, PD I, PD I, PD I, PD I, PD I, PD I, PD I, PD I, PD I, PD O, PU I, PU O, PU O, PU I O O O O O O O O O O O O O O
Pin
P11 R8 M11 N10 P9 T9 V6 U6 K12 K15 K14 K16 K13 K17 J12 K18 J14 J16 J13 J17 J15 J18 H16 H18 L12 L18 L17 K11 C13 B10 C11 A9 D11 B9 E10 D10 A8 C10 B8 F11 A7 D9 A6
Power Rail
VDDIOP0 VDDIOP0 VDDIOP0 VDDIOP0 VDDIOP0 VDDBU VDDIOP0 VDDIOP0 VDDIOM VDDIOM VDDIOM VDDIOM VDDIOM VDDIOM VDDIOM VDDIOM VDDIOM VDDIOM VDDIOM VDDIOM VDDIOM VDDIOM VDDIOM VDDIOM VDDIOM VDDIOM VDDIOM VDDIOM VDDIODDR VDDIODDR VDDIODDR VDDIODDR VDDIODDR VDDIODDR VDDIODDR VDDIODDR VDDIODDR VDDIODDR VDDIODDR VDDIODDR VDDIODDR VDDIODDR VDDIODDR
I/O Type
RSTJTAG RSTJTAG RSTJTAG RSTJTAG RSTJTAG SYSC DIB DIB EBI EBI EBI EBI EBI EBI EBI EBI EBI EBI EBI EBI EBI EBI EBI EBI EBI EBI EBI EBI
Signal
NTRST TDI TDO TMS TCK JTAGSEL DIBP DIBN D0 D1 D2 D3 D4 D5 D6 D7 D8 D9 D10 D11 D12 D13 D14 D15 NCS3/NANDCS NANDRDY NRD/NANDOE NWE/NANDWE DDR_VREF
Dir
I I O I I I O O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O O I O O I O O O O O O O O O O O O O O
Signal
Dir
Signal
Dir
Signal
Dir
Signal
Dir
DDR_IO DDR_IO DDR_IO DDR_IO DDR_IO DDR_IO DDR_IO DDR_IO DDR_IO DDR_IO DDR_IO DDR_IO DDR_IO DDR_IO
DDR_A0 DDR_A1 DDR_A2 DDR_A3 DDR_A4 DDR_A5 DDR_A6 DDR_A7 DDR_A8 DDR_A9 DDR_A10 DDR_A11 DDR_A12 DDR_A13

15
Table 4-1.

HiZ HiZ HiZ HiZ HiZ HiZ HiZ HiZ HiZ HiZ HiZ HiZ HiZ HiZ HiZ HiZ HiZ HiZ HiZ HiZ HiZ HiZ HiZ HiZ HiZ HiZ HiZ HiZ HiZ HiZ HiZ HiZ O O O O I, PD I, PD I, PD I, PD I, PU I, PU I, PU
Pin
H12 H17 H13 G17 G16 H15 F17 G15 F16 E17 G14 E16 D17 C18 D16 C17 B16 B18 C15 A18 C16 C14 D15 B14 A15 A14 E12 A11 B11 F12 A10 E11 G12 E15 B15 D12 E18 G18 B17 B13 D18 F18 A17
Power Rail
VDDIODDR VDDIODDR VDDIODDR VDDIODDR VDDIODDR VDDIODDR VDDIODDR VDDIODDR VDDIODDR VDDIODDR VDDIODDR VDDIODDR VDDIODDR VDDIODDR VDDIODDR VDDIODDR VDDIODDR VDDIODDR VDDIODDR VDDIODDR VDDIODDR VDDIODDR VDDIODDR VDDIODDR VDDIODDR VDDIODDR VDDIODDR VDDIODDR VDDIODDR VDDIODDR VDDIODDR VDDIODDR VDDIODDR VDDIODDR VDDIODDR VDDIODDR VDDIODDR VDDIODDR VDDIODDR VDDIODDR VDDIODDR VDDIODDR VDDIODDR
I/O Type
DDR_IO DDR_IO DDR_IO DDR_IO DDR_IO DDR_IO DDR_IO DDR_IO DDR_IO DDR_IO DDR_IO DDR_IO DDR_IO DDR_IO DDR_IO DDR_IO DDR_IO DDR_IO DDR_IO DDR_IO DDR_IO DDR_IO DDR_IO DDR_IO DDR_IO DDR_IO DDR_IO DDR_IO DDR_IO DDR_IO DDR_IO DDR_IO DDR_IO DDR_IO DDR_IO DDR_IO DDR_IO DDR_IO DDR_IO DDR_IO DDR_IO DDR_IO DDR_IO
Signal
DDR_D0 DDR_D1 DDR_D2 DDR_D3 DDR_D4 DDR_D5 DDR_D6 DDR_D7 DDR_D8 DDR_D9 DDR_D10 DDR_D11 DDR_D12 DDR_D13 DDR_D14 DDR_D15 DDR_D16 DDR_D17 DDR_D18 DDR_D19 DDR_D20 DDR_D21 DDR_D22 DDR_D23 DDR_D24 DDR_D25 DDR_D26 DDR_D27 DDR_D28 DDR_D29 DDR_D30 DDR_D31 DDR_DQM0 DDR_DQM1 DDR_DQM2 DDR_DQM3 DDR_DQS0 DDR_DQS1 DDR_DQS2 DDR_DQS3 DDR_DQSN0 DDR_DQSN1 DDR_DQSN2
Dir
I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O I/O O O O O I/O I/O I/O I/O I/O I/O I/O
Signal
Dir
Signal
Dir
Signal
Dir
Signal
Dir

16
Table 4-1.

I, PU O O O O O O O O O O O O I O, PD O, PD O, PD O, PD DHSDP DHSDM O, PD O, PD I I
Pin
A13 C8 B12 A12 B7 C12 E13 G11 A5 B5 E9 B6 F9 R11 U14 V14 U12 V12 U10 V10 V15 T13 C5, C7, D14, T15, T7, U17, V7 A16, C9, N13, T14, T8, V17 D13, F14, G10, G13, H11 E14, F10, F13, F15, H14 P12, T16 J11, T17 G7, V11 L11, M4
Power Rail
VDDIODDR VDDIODDR VDDIODDR VDDIODDR VDDIODDR VDDIODDR VDDIODDR VDDIODDR VDDIODDR VDDIODDR VDDIODDR VDDIODDR VDDIODDR VBG VDDUTMII VDDUTMII VDDUTMII VDDUTMII VDDUTMII VDDUTMII VDDBU GNDBU
I/O Type
DDR_IO DDR_IO DDR_IO DDR_IO DDR_IO DDR_IO DDR_IO DDR_IO DDR_IO DDR_IO DDR_IO DDR_IO DDR_IO VBG USBHS USBHS USBHS USBHS USBHS USBHS power supply ground
Signal
DDR_DQSN3 DDR_CS DDR_CLK DDR_CLKN DDR_CKE DDR_CALN DDR_CALP DDR_RAS DDR_CAS DDR_WE DDR_BA0 DDR_BA1 DDR_BA2 VBG HHSDPC HHSDMC HHSDPB HHSDMB HHSDPA HHSDMA VDDBU GNDBU
Dir
I/O O O O O I I O O O O O O I I/O I/O I/O I/O I/O I/O I I
Signal
Dir
Signal
Dir
Signal
Dir
Signal
Dir
VDDCORE
power supply
VDDCORE
GNDCORE
ground
GNDCORE
VDDIODDR
power supply
VDDIODDR
GNDIODDR
ground
GNDIODDR
VDDIOM
power supply
VDDIOM
GNDIOM
ground
GNDIOM
VDDIOP0
power supply
VDDIOP0
VDDIOP1
power supply
VDDIOP1

17
Table 4-1.

I
Pin
E5, J7, N11, U7 V13 U13 R12 R10 P10 U11 T11 L6 L4 L5 R3 P4
Power Rail
GNDIOP
I/O Type
Ground
Signal
GNDIOP
Dir
I
Signal
Dir
Signal
Dir
Signal
Dir
Signal
Dir
VDDUTMIC VDDUTMII GNDUTMI VDDPLLA GNDPLL VDDOSC GNDOSC VDDANA GNDANA VDDANA VDDFUSE GNDFUSE
Power supply Power supply Ground Power supply Ground Power supply Ground Power supply Ground Power supply Power supply Ground
VDDUTMIC VDDUTMII GNDUTMI VDDPLLA GNDPLL VDDOSC GNDOSC VDDANA GNDANA ADVREF VDDFUSE GNDFUSE
I I I I I I I I I I I I
I I I I I I I I I I I I

18
4.3
Input/Output Description
SAMA5 I/O Types Description
Voltage Range 1.65-3.6V 1.65-3.6V 1.65-3.6V 3.0-3.6V 1.65-1.95V, 3.0-3.6V 3.0-3.6V 1.65-3.6V 3.0-3.6V 1.65-3.6V 3.0-3.6V I/O I/O I/O I Pull-up Analog Pull-up Switchable Switchable Switchable Switchable Switchable Reset State Typ Value (Ohm) 100K 100K 100K 100K 100K 100K Switchable Reset State Reset State 100K 100K 15K Reset State Reset State Pull-down Switchable Switchable Switchable Pull-down Typ Value (Ohm) 100K 100K 100K Schmitt Trigger Switchable Switchable Switchable Switchable
Table 4-2.
I/O Type GPIO GPIO_CLK GPIO_CLK2 GPIO_ANA EBI RSTJTAG SYSC USBHS CLOCK DIB
When Reset State is indicated, the configuration is defined by the Reset State column of the Pin Description table (see Table 4-1 on page 11).
Table 4-3.
I/O Type GPIO MCI_CLK GPIO_CLK GPIO_CLK2 GPIO_ANA
SAMA5 I/O Type Assignation and Frequency

I/O Frequency (MHz) 33 52 66 75 25 Load (pF) 40 20 20 20 20 16 mA, 40 mA (peak) Fan-out Drive Control High/Medium/Low High/Medium/Low High/Medium/Low High/Medium/Low Fixed to Medium High/Medium/Low 1.8V/3.3V High/Medium/Low Fixed to Low Fixed to Medium No No No No High/Medium/Low Signal Name All PIO lines except the lines indicated further on in this table MCI0CK, MCI1CK,MCI2CK SPI0CK, SPI1CK, ETXCLK,ERXCLK LCDDOTCK ADx
EBI DDR_IO RST JTAG SYSC VBG USBHS CLOCK GMAC
66 166 3 10 0.25 0.25 480 50 125
50 20 10 10 10 10 20 50 15
All EBI signals All DDR signals NRST, NTRST, BMS TCK, TDI, TMS, TDO WKUP, SHDN, JTAGSEL, TST, SHDN VBG HHSDPC, HHSDPB, HHSDPA/DHSDP, HHSDMC, HHSDMB, HHSDMA/DHSDM XIN, XOUT, XIN32, XOUT32 Gigabit Ethernet I/Os

19
5.
5.1
Power Considerations
Power Supplies
Table 5-1 defines the power supply rails and the estimated power consumption at typical voltage.
Table 5-1. Name SAMA5D3 Power supplies Voltage Range, Nominal 1.1-1.32V, 1.2V 1.7-1.9V, 1.8V 1.14-1.30, 1.2V 1.65-1.95V, 1.8V 3.0-3.6V, 3.3V 1.65-3.6V 1.65-3.6V 1.65-3.6V 1.1-1.32V, 1.2V 3.0-3.6V, 3.3V 1.1-1.32V, 1.2V 1.65-3.6V 3.0-3.6V, 3.3V 2.25-2.75V, 2.5V Associated Ground Powers The core, including the processor, the embedded memories and the peripherals LPDDR/DDR2 Interface I/O lines LPDDR2 Interface I/O lines NAND and HSMC Interface I/O lines Peripheral I/O lines Peripheral I/O lines The Slow Clock oscillator, the internal 32 kHz RC Oscillator and a part of the System Controller The USB device and host UTMI+ core The UTMI PLL The USB device and host UTMI+ interface The PLLA cell Main Oscillator cell and VBG if 3.0V<VDDOSC<= 3.6V The Analog to Digital Converter Fuse box for programming. VDDFUSE GNDFUSE It can be tied to ground with a 100 resistor for fuse reading only.
VDDCORE VDDIODDR
GNDCORE GNDIODDR
VDDIOM VDDIOP0 VDDIOP1 VDDBU VDDUTMIC VDDUTMII VDDPLLA VDDOSC VDDANA
GNDIOM GNDIOP GNDIOP GNDBU GNDUTMI GNDUTMI GNDPLL GNDOSC GNDANA
5.2
Power-up Consideration
The user must first activate VDDIOP and VDDIOM, then VDDPLL and VDDCORE with the constraint that VDDPLL is established no later than 1 ms after VDDCORE. The VDDCORE and VDDBU power supplies rising time must be defined according to the Core and Backup Power-OnReset characteristics to ensure VDDCORE or VDDBU has reached VIH after the POR reset time. Please refer to the Core Power Supply POR Characteristics and Backup Power Supply POR Characteristics sections of the product datasheet for power-up constraints.
5.3
Power-down Consideration
The user must remove VDDPLL first, then VDDCORE, and at last VDDIOP and VDDIOM, to ensure a reliable operation of the device.

20
6.
Memories
Figure 6-1. Memory Mapping

Internal Memory Mapping Addressl Memory Space
0x0000 0000
Notes: (1) Can be ROM, EBI_NCS0 or SRAM depending on BMS and REMAP
0x0000 0000
Boot Memory
0x0010 0000
(1)
ROM
0x0020 0000
Internal Memories
0x0FFF FFFF 0x1000 0000
256 MBytes
NFC SRAM
0x0030 0000
EBI Chip Select 0

0x1FFF FFFF 0x2000 0000
Peripheral Mapping
256 MBytes
0xF000 0000
0x0031 0000 0x0040 0000 0x0050 0000 0x0060 0000
SRAM0 SRAM1 SMD
HSMCI0
0xF000 4000 0xF000 8000
SPI0 SSC0 CAN0
UDPH SRAM UHP OHCI
DDRCS
0xF000 C000
512 MBytes
0xF001 0000
0x0070 0000
UHP EHCI
0x0080 0000 0x0090 0000
TC0, TC1, TC2

0x3FFF FFFF 0x4000 0000 0xF001 4000
AXI Matrix DAP
TWI0 EBI Chip Select 1

0xF001 8000
0x00A0 0000
256 MBytes
0xF001 C000
TWI1
0x0FFF FFFF
Undefined (Abort)
0x4FFF FFFF 0x5000 0000
USART0 EBI Chip Select 2

0xF002 0000
256 MBytes
0xF002 4000
USART1 UART0
0x5FFF FFFF 0x6000 0000
EBI Chip Select 3

0x6FFF FFFF 0x7000 0000
0xF002 8000
256 MBytes
GMAC
0xF002 C000
System Controller Mapping

PWMC
0xFFFF C000
0xF003 0000
HSMC
NFC Command Registers

0x7FFF FFFF
256 MBytes
LCDC
0xF003 4000
0xFFFF D000
ISI
0xF003 8000
Reserved
0xFFFF E400
SFR
0xF003 C000
Reserved
0xF800 0000
FUSE
0xFFFF E600
DMAC0 HSMCI1
0xFFFF E800
0xF800 4000
DMAC1 HSMCI2
0xFFFF EA00
0xF800 8000
SPI1
0xF800 C000
MPDDRC
0xFFFF EC00
SSC1
0xF801 0000
MATRIX
0xFFFF EE00
CAN1
0xF801 4000
DBGU
0xFFFF F000
TC3, TC4, TC5 Undefined (Abort)

0xF801 8000
AIC
0xFFFF F200
TSADC
0xF801 C000
PIOA
0xFFFF F400
TWI2
0xF802 0000
PIOB
0xFFFF F600
PIOC USART2
0xFFFF F800
0xF802 4000
USART3
0xF802 8000 0xFFFF FA00
PIOD PIOE
0xFFFF FC00
UART1
0xF802 C000
EMAC
0xF803 0000
PMC
0xFFFF FE00
RSTC UDPHS
0xFFFF FE10
0xF803 4000 0xEFFF FFFF 0xF000 0000
SHDC SHA
0xFFFF FE20 0xFFFF FE30
Reserved PITC WDT
0xF803 8000
Internal Peripherals
0xFFFF FFFF
256 MBytes
0xF803 C000
AES TDES
0xF804 0000
0xFFFF FE40 0xFFFF FE50 0xFFFF FE54
SCKCR BSC
TRNG
0xF804 4000 0xFFFF FE60
GPBR Reserved RTCC
Reserved
0xFFFF C000
0xFFFF FE70 0xFFFF FEB0
SYSC
0xFFFF FFFF
0xFFFF FEE0 0xFFFF FFFF
Reserved

21
6.1
6.1.1
Embedded Memories
Internal SRAM
The SAMA5D3 product embeds a total of 128 Kbytes high-speed SRAM0 and SRAM1. After Remap the SRAM is accessible at address 0 but also at address 0x00300000. Only the ARM core has access to the SRAM at address 0. The others masters (DMA, peripherals, etc.) always access the SRAM at address 0x00300000. SRAM0 and SRAM1 can be accessed in parallel to improve the overall bandwidth of the system.
6.1.2
Internal ROM
The SAMA5D3 product embeds one 160-Kbyte internal ROM containing a standard and a secure bootloader. The secure bootloader is described in a separate document, under NDA. The standard bootloader supports booting from:
z z z z z
8-bit NAND Flash with ECC management SPI Serial Flash SDCARD EMMC TWI EEPROM
The boot sequence can be selected using the boot order facility (Boot Select Control Register). The internal ROM embeds Galois field tables that are used to compute NandFlash ECC. Please refer to Figure 12-9 Galois Field Table Mapping in the Boot Strategies section of this datasheet.
6.1.3
Boot Strategies
For standard boot strategies, please refer to the Boot Strategies section of this datasheet. For secure boot strategies, please refer to the Application Note Secure Boot on SAMA5D3 Series (NDA required).
6.2
External Memory
The SAMA5D3 features interfaces to offer connexion to a wide range of external memories or to parallel peripherals.
6.2.1
DDR2/LPDDR/LPDDR2 Interface
z z z z z z z
32-bit external interface 512 Mbytes address space on CS1 Supports DDR2, LPDDR and LPDDR2 memories Drive level control I/O impedance control embedded Supports 4-banks and 8-banks and up to 512 Mbytes Multi-port
6.2.2
Static Memories and NAND Flash

The static memory controller is dedicated to interfacing external memory devices: The static memory controller is able to drive up to 4 chip select. NCS3 is dedicated to the NAND Flash control.
z z
Asynchronous SRAM-like memories and parallel peripherals NAND Flash (8-Bit MLC and SLC)
The HSMC embeds a NAND Flash Controller (NFC). The NFC can handle automatic transfers, sending the commands and address cycles to the NAND Flash and transferring the contents of the page (for read and write) to the NFC SRAM. It minimizes the CPU overhead.

22
In order to improve overall system performance the DATA phase of the transfer can be DMA assisted. The static memory embeds a NAND Flash Error Correcting Code controller with the features as follows:
z z z
Algorithm based on BCH codes. Supports also SLC 1-bit (BCH 2-bit), SLC 4-bit (BCH 4-bit) Programmable Error Correcting Capability:
z z
2-bit, 4-bit, 8-bit and 16-bit errors for 512 Bytes/sector (4 Kbyte page) 24-bit error for 1024 Bytes/sector (8 Kbyte page)
z z z z z z z z z
Programmable sector size: 512 bytes or 1024 bytes Programmable number of sector per page: 1, 2, 4 or 8 blocks of data per page Programmable spare area size Supports spare area ECC protection Supports 8 Kbyte page size using 1024 Bytes/sector and 4 Kbyte page size using 512 Bytes/sector Error detection is interrupt driven Provides hardware acceleration for error location Finds roots of error-locator polynomial Programmable number of roots

23
7.
Real-time Event Management

The events generated by peripherals are designed to be directly routed to peripherals managing/using these events without processor intervention. Peripherals receiving events contain logic by which to select the one required.
7.1
Embedded Characteristics
z
Peripherals generate event triggers which are directly routed to event managers such as ADC, for example, to start measurement/conversion without processor intervention.
7.2
Real-time Event Mapping List

Real-time Event Mapping List Event Generator PMC Event Manager Pulse Width Modulation (PWM) PWM Function Safety / Puts the PWM Outputs in Safe Mode (Main Crystal Clock Failure Detection) Safety / Puts the PWM Outputs in Safe Mode (Overspeed, Overcurrent detection, etc.)
Table 7-1.
Analog-to-Digital Converter (ADC)

24
8.
System Controller
The System Controller is a set of peripherals that allows handling of key elements of the system, such as power, resets, clocks, time, interrupts, watchdog, etc. The System Controller User Interface also embeds the registers that configure the Matrix and a set of registers for the chip configuration. The chip configuration registers configure the EBI chip select assignment and voltage range for external memories. The System Controllers peripherals are all mapped within the highest 16 KB of address space, between addresses 0xFFFF D000 and 0xFFFF FFFF. However, all the registers of System Controller are mapped on the top of the address space. All the registers of the System Controller can be addressed from a single pointer by using the standard ARM instruction set, as the Load/Store instruction have an indexing mode of 4 KB. Figure 8-1 on page 26 shows the System Controller block diagram.

25
Figure 8-1. SAMA5D3 System Controller Block Diagram

System Controller irq fiq periph_irq[2..42] pit_irq wdt_irq dbgu_irq pmc_irq rstc_irq MCK periph_nreset dbgu_rxd MCK debug periph_nreset SLCK debug idle proc_nreset Periodic Interval Timer Watchdog Timer wdt_fault WDRPROC NRST VDDCORE POR por_ntrst jtag_nreset periph_nreset proc_nreset backup_nreset VDDBU Powered SLCK Real-Time Clock rtc_irq rtc_alarm UPLLCK UHP48M UHP12M periph_nreset periph_irq[32] USB High Speed Host Port Debug Unit dbgu_irq dbgu_txd pit_irq jtag_nreset Boundary Scan TAP Controller VDDCORE Powered nirq irq_vect Advanced Interrupt Controller nfiq fiq_vect Cortex A5 por_ntrst ntrst proc_nreset PCK debug
wdt_irq
MCK periph_nreset Bus Matrix
Reset Controller
VDDBU
VDDBU POR
SLCK backup_nreset SLCK SHDN WKUP backup_nreset rtc_alarm 32 kHz RC OSC SCKCR SLCK int MAINCK 12 MHz MAIN OSC UPLL PLLA periph_nreset SMDCK UPLLCK PLLACK
Shut-Down Controller
UPLLCK USB High Speed Device Port
periph_nreset 4 General-purpose Backup Registers periph_irq[33]
XIN32 XOUT32
SLOW CLOCK OSC
12 MHz RC OSC XIN XOUT
Power Management Controller
periph_clk[2..49] pck[0-1] UHP48M UHP12M PCK MCK DDR sysclk LCD Pixel clock pmc_irq idle SMDCK = periph_clk[11]
SMDCK periph_nreset periph_irq[11] SMD Software Modem
periph_clk[2..49] periph_nreset
periph_nreset periph_clk[5.9] dbgu_rxd PA0-PA31 PB0-PB31 PC0-PC31 PD0-PD31 PE0-PE31
PIO Controllers
periph_irq[5..9] irq fiq dbgu_txd
Embedded Peripherals periph_irq[2..49] in out enable
Fuse Box

26
8.1
Chip Identification
z z
Chip ID: 0x8A5C07C2 Extended ID:

SAMA5D33 0x00414300 SAMA5D34 0x00414301 SAMA5D35 0x00584300 SAMA5D31 0x00444300
Device EXT ID
z z z
Boundary JTAG ID: 0x05B3103F Cortex A5 JTAG IDCODE: 0x4ba00477 Cortex A5 Serial Wire IDCODE: 0x2ba01477
8.2
Backup Section
The SAMA5D3 features a Backup Section that embeds:
z z z z z z z z
RC Oscillator Slow Clock Oscillator SCKR register Real-time Clock (RTC) Shutdown Controller 4 Backup registers Part of the Reset Controller (RSTC) Boot Select Control Register
This section is powered by the VDDBU rail.

27
9.
9.1
Peripherals
Peripheral Mapping
As shown in Section 6. Memories the Peripherals are mapped in the upper 256 Mbytes of the address space between the addresses 0xFFF7 8000 and 0xFFFC FFFF. Each User Peripheral is allocated 16 Kbytes of address space.
9.2
Peripheral Identifiers
Peripheral Identifiers Instance Name AIC SYS DBGU PIT WDT HSMC PIOA PIOB PIOC PIOD PIOE SMD USART0 USART1 USART2 USART3 UART0 UART1 TWI0 TWI1 TWI2 HSMCI0 HSMCI1 HSMCI2 SPI0 SPI1 Instance Description Advanced Interrupt Controller System Controller Interrupt Debug Unit Interrupt Periodic Interval Timer Interrupt Watchdog timer Interrupt Multi-bit ECC Interrupt Parallel I/O Controller A Parallel I/O Controller B Parallel I/O Controller C Parallel I/O Controller D Parallel I/O Controller E SMD Soft Modem USART 0 USART 1 USART 2 USART 3 UART 0 UART 1 Two-Wire Interface 0 Two-Wire Interface 1 Two-Wire Interface 2 High Speed Multimedia Card Interface 0 High Speed Multimedia Card Interface 1 High Speed Multimedia Card Interface 2 Serial Peripheral Interface 0 Serial Peripheral Interface 1 External Interrupt FIQ PMC, RSTC, RTC Wired-OR Interrupt Clock Type SYS_CLK SYS_CLK PCLOCK SYS_CLK SYS_CLK HCLOCK PCLOCK PCLOCK PCLOCK PCLOCK PCLOCK HCLOCK PCLOCK PCLOCK PCLOCK PCLOCK PCLOCK PCLOCK PCLOCK PCLOCK PCLOCK PCLOCK PCLOCK PCLOCK PCLOCK PCLOCK
Table 9-1. Instance ID 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

28
Table 9-1. Instance ID 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50-63
Peripheral Identifiers (Continued) Instance Name TC0 TC1 PWM ADC DMAC0 DMAC1 UHPHS UDPHS GMAC EMAC LCDC ISI SSC0 SSC1 CAN0 CAN1 SHA AES TDES TRNG ARM AIC FUSE MPDDRC Reserved Instance Description Timer Counter 0 (ch. 0, 1, 2) Timer Counter 1 (ch. 3, 4, 5) Pulse Width Modulation Controller Touch Screen ADC Controller DMA Controller 0 DMA Controller 1 USB Host High Speed USB Device High Speed Gigabit Ethernet MAC Ethernet MAC LCD Controller Image Sensor Interface Synchronous Serial Controller 0 Synchronous Serial Controller 1 CAN controller 0 CAN controller 1 Secure Hash Algorithm Advanced Encryption Standard Triple Data Encryption Standard True Random Number Generator Performance Monitor Unit Advanced Interrupt Controller Fuse Controller MPDDR controller IRQ External Interrupt Wired-OR Interrupt Clock Type PCLOCK PCLOCK PCLOCK PCLOCK HCLOCK HCLOCK HCLOCK HCLOCK HCLOCK + PCLOCK HCLOCK + PCLOCK HCLOCK HCLOCK PCLOCK PCLOCK PCLOCK PCLOCK PCLOCK PCLOCK PCLOCK PCLOCK PROC_CLOCK SYS_CLK PCLOCK HCLOCK -

29
9.3
Peripheral Signal Multiplexing on I/O Lines

The SAMA5D3 product features 5 PIO controllers, PIOA, PIOB, PIOC, PIOD and PIOE, which multiplex the I/O lines of the peripheral set. Each PIO Controller controls 32 lines. Each line can be assigned to one of three peripheral functions, A, B or C. The multiplexing table (Table 4-1) in the pin description section Section 4.2 324-ball LFBGA Package Pinout defines how the I/O lines of the peripherals A, B and C are multiplexed on the PIO Controllers. Note that some peripheral functions which are output only, might be duplicated within the table. The column Reset State indicates whether the PIO Line resets in I/O mode or in peripheral mode. If I/O is mentioned, the PIO Line resets in input with the pull-up enabled, so that the device is maintained in a static state as soon as the reset is released. As a result, the bit corresponding to the PIO Line in the register PIO_PSR (Peripheral Status Register) resets low. If a signal name is mentioned in the Reset State column, the PIO Line is assigned to this function and the corresponding bit in PIO_PSR resets high. This is the case of pins controlling memories, in particular the address lines, which require the pin to be driven as soon as the reset is released. Note that the pull-up resistor is also enabled in this case.
9.4
Peripheral Clock Type

The SAMA5D3 Series embeds peripherals with five different clock types:
z z z z z
HCLOCK: AHB Clock, managed with the PMC_SCER, PMC_SCDR and PMC_SCSR registers of PMC System Clock PCLOCK: APB Clock, managed with the PMC_PCER, PMC_PCDR, PMC_PCSR and PMC_PCR registers of Peripheral Clock HCLOCK+PCLOCK: Both clock types coexist. The clock is managed with the PMC_PCER, PMC_PCDR, PMC_PCSR and PMC_PCR registers of Peripheral Clock SYS_CLOCK: This clock cannot be disabled. PROC_CLOCK: The clock related to Processor Clock (PCK) and managed with the PMC_SCDR and PMC_SCSR registers of PMC System Clock
Please refer to Table 9-1 Peripheral Identifiers for details.

30
10.
10.1
Cortex-A5 ARM
Description
The ARM Cortex-A5 processor is a high-performance, low-power, ARM macrocell with an L1 cache subsystem that provides full virtual memory capabilities. The Cortex-A5 processor implements the ARMv7 architecture and runs 32-bit ARM instructions, 16-bit and 32-bit Thumb instructions, and 8-bit Java byte codes in Jazelle state. The Floating-Point Unit (FPU) supports the ARMv7 VFPv4-D16 architecture without Advanced SIMD extensions (NEON). It is tightly integrated to the Cortex-A5 processor pipeline. It provides trapless execution and is optimized for scalar operation. It can generate an Undefined instruction exception on vector instructions that enables the programmer to emulate vector capability in software. The design can include the FPU only, in which case the Media Processing Engine (MPE) is not included. See the Cortex-A5 Floating-Point Unit Technical Reference Manual.
10.1.1 Power management

The Cortex-A5 design supports four main levels of power management: 10.1.1.1 Run Mode Run mode is the normal mode of operation where all of the processor functionality is available. Everything, including core logic and embedded RAM arrays, is clocked and powered up. 10.1.1.2 Standby Mode Standby mode disables most of the clocks of the processor, while keeping it powered up. This reduces the power drawn to the static leakage current, plus a small clock power overhead required to enable the processor to wake up from Standby mode. The transition from Standby mode to Run mode is caused by one of the following:
z z z z
The arrival of an interrupt, either masked or unmasked The arrival of an event, if standby mode was initiated by a Wait for Event (WFE) instruction A debug request, when either debug is enabled or disabled A reset
10.2
z z z z z z z z z
In-order pipeline with dynamic branch prediction ARM, Thumb, and ThumbEE instruction set support Harvard level 1 memory system with a Memory Management Unit (MMU) 32 Kbytes Data Cache 32 Kbytes Instruction Cache 64-bit AXI master interface ARM v7 debug architecture VFPv4-D16 FPU with trapless execution Jazelle hardware acceleration.

31
10.3
Block Diagram
Figure 10-1. Cortex-A5 Processor Top-level Diagram
Embedded trace macrocell (ETM) interface Cortex-A5 processor APB interface
Debug
Data processing unit (DPU) Data micro-TLB
Prefetch unit and branch predictor (PFU) Instruction micro-TLB
Data store buffer (STB)
Data cache unit (DCU) CP15
Main translation lookaside buffer (TLB)
Instruction cache unit (ICU)
Bus interface unit (BIU)
AXI interface

32
10.4
Programmer Model
10.4.1 Processor Operating Modes

In all states, there are seven operation modes:
z z z z z z z
User mode (USR) is the usual ARM program execution state. It is used for executing most application programs Fast Interrupt (FIQ) mode is used for handling fast interrupts. It is suitable for high-speed data transfer or channel process Interrupt (IRQ) mode is used for general-purpose interrupt handling Supervisor mode (SVC) is a protected mode for the operating system Abort mode (ABT) is entered after a data or instruction prefetch abort System mode (SYS) is a privileged user mode for the operating system Undefined mode (UND) is entered when an undefined instruction exception occurs
Mode changes may be made under software control, or may be brought about by external interrupts or exception processing. Most application programs execute in User Mode. The non-user modes, known as privileged modes, are entered in order to service interrupts or exceptions or to access protected resources.
10.4.2 Processor Operating States

The processor has the following instruction set states controlled by the T bit and J bit in the CPSR.
z
ARM state: The processor executes 32-bit, word-aligned ARM instructions. Thumb state: The processor executes 16-bit and 32-bit, halfword-aligned Thumb instructions. ThumbEE state: The processor executes a variant of the Thumb instruction set designed as a target for dynamically generated code. This is code compiled on the device either shortly before or during execution from a portable bytecode or other intermediate or native representation.
Jazelle state: The processor executes variable length, byte-aligned Java bytecodes.
The J bit and the T bit determine the instruction set used by the processor. Table 3-1 shows the encoding of these bits.
Table 10-1. CPSR J and T Bit Encoding J 0 0 1 1 T 0 1 0 1 Instruction Set State ARM Thumb Jazelle ThumbEE
Changing between ARM and Thumb states does not affect the processor mode or the register contents. See the ARM Architecture Reference Manual for information on entering and exiting ThumbEE state. 10.4.2.1 Switching State It is possible to change the instruction set state of the processor between:
z z z z
ARM state and Thumb state using the BX and BLX instructions. Thumb state and ThumbEE state using the ENTERX and LEAVEX instructions. ARM and Jazelle state using the BXJ instruction. Thumb and Jazelle state using the BXJ instruction.
See the ARM Architecture Reference Manual for more information about changing instruction set state.

33
10.4.3 Cortex-A5 Registers

This view provides 16 ARM core registers, R0 to R15, that include the Stack Pointer (SP), Link Register (LR), and Program Counter (PC). These registers are selected from a total set of either 31 or 33 registers, depending on whether or not the Security Extensions are implemented. The current execution mode determines the selected set of registers, as shown below in Table 10-2. This shows that the arrangement of the registers provides duplicate copies of some registers, with the current register selected by the execution mode. This arrangement is described as banking of the registers, and the duplicated copies of registers are referred to as banked registers.
Table 10-2. Cortex-A5 Modes and Registers Layout User and System R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 PC Monitor R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13_MON R14_MON PC Supervisor R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13_SVC R14_SVC PC Abort R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13_ABT R14_ABT PC Undefined R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13_UND R14_UND PC Interrupt R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13_IRQ R14_IRQ PC Fast Interrupt R0 R1 R2 R3 R4 R5 R6 R7 R8_FIQ R9_FIQ R10_FIQ R11_FIQ R12_FIQ R13_FIQ R14_FIQ PC
CPSR
CPSR SPSR_MON
CPSR SPSR_SVC
CPSR SPSR_ABT
CPSR SPSR_UND
CPSR SPSR_IRQ
CPSR SPSR_FIQ
Mode-specific banked registers
The core contains one CPSR, and six SPSRs for exception handlers to use. The program status registers:
z z z
Hold information about the most recently performed ALU operation Control the enabling and disabling of interrupts Set the processor operation mode
Figure 10-2. Status Register Format

31 30 29 28 27 24 23 20 19 16 15 10 9 8 7 6 5 4 0
N Z C V Q
IT J Reserved [1:0]
GE[3:0]
IT[7:2]
E A I F T
Mode

34
z z z z z z
N: Negative, Z: Zero, C: Carry, and V: Overflow are the four ALU flags Q: cumulative saturation flag IT: If-Then execution state bits for the Thumb IT (If-Then) instruction J: Jazelle bit, see the description of the T bit GE: Greater than or Equal flags, for SIMD instructions E: Endianness execution state bit. Controls the load and store endianness for data accesses. This bit is ignored by instruction fetches.
z z
E = 0: Little endian operation E = 1: Big endian operation
z z z z z
A: Asynchronous abort disable bit. Used to mask asynchronous aborts. I: Interrupt disable bit. Used to mask IRQ interrupts. F: Fast interrupt disable bit. Used to mask FIQ interrupts. T: Thumb execution state bit. This bit and the J execution state bit, bit [24], determine the instruction set state of the processor, ARM, Thumb, Jazelle, or ThumbEE. Mode: Five bits to encode the current processor mode. The effect of setting M[4:0] to a reserved value is UNPREDICTABLE.
Table 10-3. Processor Mode vs. Mode Field Mode USR FIQ IRQ SVC M[4:0] 10000 10001 10010 10011
ABT UND SYS Reserved
10111 11011 11111 Other
10.4.3.1 CP15 Coprocessor Coprocessor 15, or System Control Coprocessor CP15, is used to configure and control all the items in the list below:
z z z z z
Cortex A5 Caches (ICache, DCache and write buffer) MMU Security Other system options
To control these features, CP15 provides 16 additional registers. See Table 10-4 below.
Table 10-4. CP15 Registers Register 0 0 1 1 Name ID Code
(1)
Read/Write Read/Unpredictable Read/Unpredictable Read/Write Read/Write
Cache type(1) Control

(1) (1)
Security

35
Table 10-4. CP15 Registers Register 2 3 4 5 5 6 7 8 9 10 11 Name Translation Table Base Domain Access Control Reserved Data fault Status
(1)
Read/Write Read/Write Read/Write None Read/Write Read/Write Read/Write

(1)
Instruction fault status Fault Address Cache and MMU Operations TLB operations cache lockdown(1) TLB lockdown Reserved
Read/Write Unpredictable/Write Read/Write Read/Write None
13 13 14 15
FCSE PID(1) Context ID Reserved Test configuration

(1)
Read/Write Read/Write None Read/Write
Note:
1.
This register provides access to more than one register. The register accessed depends on the value of the CRm field or Opcode_2 field.

36
10.4.4 CP 15 Register Access

CP15 registers can only be accessed in privileged mode by:
z z
MCR (Move to Coprocessor from ARM Register) instruction is used to write an ARM register to CP15. MRC (Move to ARM Register from Coprocessor) instruction is used to read the value of CP15 to an ARM register.
Other instructions such as CDP, LDC, STC can cause an undefined instruction exception. The assembler code for these instructions is: MCR/MRC{cond} p15, opcode_1, Rd, CRn, CRm, opcode_2. The MCR/MRC instructions bit pattern is shown below:
31
30 cond
29
28
27 1 19
26 1 18 CRn
25 1 17
24 0 16
23
22 opcode_1 14 Rd
21
20 L 12
15
13
11 1 3
10 1 2 CRm
9 1 1
8 1 0
6 opcode_2
4 1
CRm[3:0]: Specified Coprocessor Action

Determines specific coprocessor action. Its value is dependent on the CP15 register used. For details, refer to CP15 specific register behavior.
opcode_2[7:5]
Determines specific coprocessor operation code. By default, set to 0.
Rd[15:12]: ARM Register

Defines the ARM register whose value is transferred to the coprocessor. If R15 is chosen, the result is unpredictable.
CRn[19:16]: Coprocessor Register

Determines the destination coprocessor register.
L: Instruction Bit
0 = MCR instruction 1 = MRC instruction
opcode_1[23:20]: Coprocessor Code

Defines the coprocessor specific code. Value is c15 for CP15.
cond [31:28]: Condition

37
10.4.5 Addresses in the Cortex-A5 processor

The Cortex-A5 processor operates using virtual addresses (VAs). The Memory Management Unit (MMU) translates these VAs into the physical addresses (PAs) used to access the memory system. Translation tables hold the mappings between VAs and PAs. See the ARM Architecture Reference Manual for more information. When the Cortex-A5 processor is executing in Non-secure state, the processor performs translation table look-ups using the Non-secure versions of the Translation Table Base Registers. In this situation, any VA can only translate into a Nonsecure PA. When it is in Secure state, the Cortex-A5 processor performs translation table look-ups using the Secure versions of the Translation Table Base Registers. In this situation, the security state of any VA is determined by the NS bit of the translation table descriptors for that address. Following is an example of the address manipulation that occurs when the Cortex-A5 processor requests an instruction. 1. 2. The Cortex-A5 processor issues the VA of the instruction as Secure or Non-secure VA accesses according to the state the processor is in. The instruction cache is indexed by the bits of the VA. The MMU performs the translation table look-up in parallel with the cache access. If the processor is in the Secure state it uses the Secure translation tables, otherwise it uses the Non-secure translation tables. If the protection check carried out by the MMU on the VA does not abort and the PA tag is in the instruction cache, the instruction data is returned to the processor. If there is a cache miss, the MMU passes the PA to the AXI bus interface to perform an external access. The external access is always Non-secure when the core is in the Non-secure state. In the Secure state, the external access is Secure or Non-secure according to the NS attribute value in the selected translation table entry. In Secure state, both L1 and L2 translation table walk accesses are marked as Secure, even if the first level descriptor is marked as NS.
3. 4.

38
10.5
Memory Management Unit
10.5.1 About the MMU

The MMU works with the L1 memory system to translate virtual addresses to physical addresses. It also controls accesses to and from external memory. The ARM v7 Virtual Memory System Architecture (VMSA) features include the following:
z
Page table entries that support:

z z z z
16-Mbyte supersections. The processor supports supersections that consist of 16-Mbyte blocks of memory. 1-Mbyte sections 64-Kbyte large pages 4-Kbyte small pages
z z z
16 access domains Global and application-specific identifiers to remove the requirement for context switch TLB flushes. Extended permissions checking capability.
TLB maintenance and configuration operations are controlled through a dedicated coprocessor, CP15, integrated with the core. This coprocessor provides a standard mechanism for configuring the L1 memory system. See the ARM Architecture Reference Manual for a full architectural description of the ARMv7 VMSA.
10.5.2 Memory Management System

The Cortex-A5 processor supports the ARM v7 VMSA. The translation of a Virtual Address (VA) used by the instruction set architecture to a Physical Address (PA) used in the memory system and the management of the associated attributes and permissions is carried out using a two-level MMU. The first level MMU uses a Harvard design with separate micro TLB structures in the PFU for instruction fetches (IuTLB) and in the DPU for data read and write requests (DuTLB). A miss in the micro TLB results in a request to the main unified TLB shared between the data and instruction sides of the memory system. The TLB consists of a 128-entry two-way set-associative RAM based structure. The TLB page-walk mechanism supports page descriptors held in the L1 data cache. The caching of page descriptors is configured globally for each translation table base register, TTBRx, in the system coprocessor, CP15. The TLB contains a hitmap cache of the page types which have already been stored in the TLB. When the processor is put into dormant mode as described in Power control on page 2-9, the TLB RAM data is retained. To save and restore the hitmap, the register can be accessed directly from software using CP15 as described in C15, TLB access and attributes on page 4-12 of the Cortex-A5 Technical Reference Manual, DDI0433. 10.5.2.1 Memory types Although various different memory types can be specified in the page tables, the Cortex-A5 processor does not implement all possible combinations:
z z z
Write-through caches are not supported. Any memory marked as write-through is treated as Non-cacheable. The outer shareable attribute is not supported. Anything marked as outer shareable is treated in the same way as inner shareable. Write-back no write-allocate is not supported. It is treated as write-back write-allocate.

39
Table 10-5 shows the treatment of each different memory type in the Cortex-A5 processor in addition to the architectural requirements.
Table 10-5. Treatment of Memory Attributes Memory Type Attribute Strongly Ordered Device Other attributes Shareability Non-shareable Shareable Non-cacheable Write-through cacheable Non-shareable Write-back cacheable, write allocate Write-back cacheable, no write allocate Non-cacheable Write-through cacheable Normal Inner shareable Write-back cacheable, write allocate Write-back cacheable, no write allocate Non-cacheable Write-through cacheable Outer shareable Write-back cacheable, write allocate Write-back cacheable, no write allocate Treated as inner shareable non-cacheable. Treated as inner shareable non-cacheable unless the SMP bit in the Auxiliary Control Register is set (ACTLR[6] = b1). If this bit is set the area is treated as Write-back cacheable write allocate. Does not access L1 caches. Treated as non-cacheable. Can dynamically switch to no write allocate, if more than three full cache lines are written in succession. Treated as non-shareable write-back cacheable, write allocate Treated as inner shareable non-cacheable. Treated as inner shareable non-cacheable unless the SMP bit in the Auxiliary Control Register is set (ACTLR[6] = b1). If this bit is set the area is treated as Write-back cacheable write allocate. Notes
10.5.3 TLB Organization

TLB Organization is described in the sections that follow:
z z
Micro TLB Main TLB
10.5.3.1 Micro TLB The first level of caching for the page table information is a micro TLB of 10 entries that is implemented on each of the instruction and data sides. These blocks provide a lookup of the virtual addresses in a single cycle. The micro TLB returns the physical address to the cache for the address comparison, and also checks the access permissions to signal either a Prefetch Abort or a Data Abort. All main TLB related maintenance operations affect both the instruction and data micro TLBs, causing them to be flushed. In the same way, any change of the following registers causes the micro TLBs to be flushed:
z z z z z
Context ID Register (CONTEXTIDR) Domain Access Control Register (DACR) Primary Region Remap Register (PRRR) Normal Memory Remap Register (NMRR) Translation Table Base Registers (TTBR0 and TTBR1).

40
10.5.3.2 Main TLB Misses from the instruction and data micro TLBs are handled by a unified main TLB. Accesses to the main TLB take a variable number of cycles, according to competing requests from each of the micro TLBs and other implementationdependent factors. The main TLB is 128-entry two-way set-associative. TLB match process Each TLB entry contains a virtual address, a page size, a physical address, and a set of memory properties. Each is marked as being associated with a particular application space (ASID), or as global for all application spaces. The CONTEXTIDR determines the currently selected application space. A TLB entry matches when these conditions are true:
z z z
Its virtual address matches that of the requested address Its Non-secure TLB ID (NSTID) matches the Secure or Non-secure state of the MMU request Its ASID matches the current ASID in the CONTEXTIDR or is global.
The operating system must ensure that, at most, one TLB entry matches at any time. The TLB can store entries based on the following block sizes:
Supersections Sections Large pages Small pages Describe 16-Mbyte blocks of memory. Describe 1-Mbyte blocks of memory. Describe 64-Kbyte blocks of memory. Describe 4-Kbyte blocks of memory
Supersections, sections and large pages are supported to permit mapping of a large region of memory while using only a single entry in the TLB. If no mapping for an address is found within the TLB, then the translation table is automatically read by hardware and a mapping is placed in the TLB.
10.5.4 Memory Access Sequence

When the processor generates a memory access, the MMU: 1. 2. 3. Performs a lookup for the requested virtual address and current ASID and security state in the relevant instruction or data micro TLB. If there is a miss in the micro TLB, performs a lookup for the requested virtual address and current ASID and security state in the main TLB. If there is a miss in main TLB, performs a hardware translation table walk.
The MMU can be configured to perform hardware translation table walks in cacheable regions by setting the IRGN bits in Translation Table Base Register 0 and Translation Table Base Register 1. If the encoding of the IRGN bits is write-back, an L1 data cache lookup is performed and data is read from the data cache. If the encoding of the IRGN bits is writethrough or non-cacheable, an access to external memory is performed. For more information refer to: Cortex-A5 Technical Reference Manual. The MMU might not find a global mapping, or a mapping for the currently selected ASID, with a matching Non-secure TLB ID (NSTID) for the virtual address in the TLB. In this case, the hardware does a translation table walk if the translation table walk is enabled by the PD0 or PD1 bit in the Translation Table Base Control Register. If translation table walks are disabled, the processor returns a Section Translation fault. For more information refer to: Cortex-A5 Technical Reference Manual. If the TLB finds a matching entry, it uses the information in the entry as follows: 1. The access permission bits and the domain determine if the access is enabled. If the matching entry does not pass the permission checks, the MMU signals a memory abort. See the ARM Architecture Reference Manual for a description of access permission bits, abort types and priorities, and for a description of the Instruction Fault Status Register (IFSR) and Data Fault Status Register (DFSR).

41
2.
z z z
The memory region attributes specified in both the TLB entry and the CP15 c10 remap registers determine if the access is: Secure or Non-secure Shared or not Normal memory, Device, or Strongly-ordered. The TLB translates the virtual address to a physical address for the memory access.
For more information refer to: Cortex-A5 Technical Reference Manual, Memory region remap. 3.
10.5.5 Interaction with memory system

The MMU can be enabled or disabled as described in the ARM Architecture Reference Manual.
10.5.6 External Aborts

External memory errors are defined as those that occur in the memory system rather than those that are detected by the MMU. External memory errors are expected to be extremely rare. External aborts are caused by errors flagged by the AXI interfaces when the request goes external to the Cortex-A5 processor. External aborts can be configured to trap to Monitor mode by setting the EA bit in the Secure Configuration Register. For more information refer to: Cortex-A5 Technical Reference Manual. 10.5.6.1 External Aborts on Data Write Externally generated errors during a data write can be asynchronous. This means that the r14_abt on entry into the abort handler on such an abort might not hold the address of the instruction that caused the exception. The DFAR is Unpredictable when an asynchronous abort occurs. Externally generated errors during data read are always synchronous. The address captured in the DFAR matches the address which generated the external abort. 10.5.6.2 Synchronous and Asynchronous Aborts Chapter 4, System Control in the Cortex-A5 Technical Reference Manual describes synchronous and asynchronous aborts, their priorities, and the IFSR and DFSR. To determine a fault type, read the DFSR for a data abort or the IFSR for an instruction abort. The processor supports an Auxiliary Fault Status Register for software compatibility reasons only. The processor does not modify this register because of any generated abort.
10.5.7 MMU Software Accessible Registers

The system control coprocessor registers, CP15, in conjunction with page table descriptors stored in memory, control the MMU. Access all the registers with instructions of the form: MRC p15, 0, <Rd>, <CRn>, <CRm>, <Opcode_2> MCR p15, 0, <Rd>, <CRn>, <CRm>, <Opcode_2> CRn is the system control coprocessor register. Unless specified otherwise, CRm and Opcode_2 Should Be Zero.

42
11.
11.1
SAMA5D3 Series Debug and Test

Description
The device features a number of complementary debug and test capabilities. A common JTAG/ICE (In-Circuit Emulator) port is used for standard debugging functions, such as downloading code and single-stepping through programs. A 2-pin debug port Serial Wire Debug (SWD). SWD replaces the 5-pin JTAG port and provides an easy and risk free alternative to JTAG as the two signals SWDIO and SWCLK are overlaid on the TMS and TCK pins, allowing for bi-modal devices that provide the other JTAG signals. These extra JTAG pins can be switched to other uses when in SWD mode. The Debug Unit provides a two-pin UART that can be used to upload an application into internal SRAM. It manages the interrupt handling of the internal COMMTX and COMMRX signals that trace the activity of the Debug Communication Channel. A set of dedicated debug and test input/output pins gives direct access to these capabilities from a PC-based test environment.
11.2
z
Cortex-A5 Real-time In-circuit Emulator

z z z z z
Two real-time Watchpoint Units Two Independent Registers: Debug Control Register and Debug Status Register Test Access Port Accessible through JTAG Protocol Debug Communications Channel Serial Wire Debug Two-pin UART Debug Communication Channel Interrupt Handling Chip ID Register
Debug Unit
z z z
IEEE1149.1 JTAG Boundary-scan on All Digital Pins.

43
11.3
Block Diagram
Figure 11-1. Debug and Test Block Diagram
TMS / SWDIO TCK / SWCLK TDI
NTRST Boundary Port SWD/ICE/JTAG SELECT JTAGSEL
TDO SWD DEBUG PORT ICE/JTAG DEBUG PORT POR TST
RTCK
Reset and Test
Cortex-A5
DTXD DMA DBGU PIO DRXD

44
11.4
Application Examples
11.4.1 Debug Environment

Figure 11-2 on page 45 shows a complete debug environment example. The ICE/JTAG interface is used for standard debugging functions, such as downloading code and single-stepping through the program. A software debugger running on a personal computer provides the user interface for configuring a Trace Port interface utilizing the ICE/JTAG interface.
Figure 11-2. Application Debug and Trace Environment Example
Host Debugger PC ICE/JTAG Interface
ICE/JTAG Connector
SAM device
RS232 Connector
Terminal
SAM-based Application Board

45
11.4.2 Test Environment

Figure 11-3 on page 46 shows a test environment example. Test vectors are sent and interpreted by the tester. In this example, the board in test is designed using a number of JTAG-compliant devices. These devices can be connected to form a single scan chain.
Figure 11-3. Application Test Environment Example
Test Adaptor
Tester
JTAG Interface
ICE/JTAG
Chip n
Chip 2
SAM device
Chip 1
SAM-based Application Board In Test

46
11.5
Debug and Test Pin Description

Table 11-1. Debug and Test Pin List Pin Name Function Reset/Test NRST TST Microcontroller Reset Test Mode Select ICE and JTAG NTRST TCK TDI TDO TMS RTCK JTAGSEL Test Reset Signal Test Clock Test Data In Test Data Out Test Mode Select Returned Test Clock JTAG Selection SWD SWCLK SWDIO Serial Debug Clock Serial Debug IO Debug Unit DRXD DTXD Debug Receive Data Debug Transmit Data Input Output Input Input/Output Input Input Input Output Input Output Input Low Input/Output Input Low High Type Active Level

47
11.6
Functional Description
11.6.1 Test Pin

One dedicated pin, TST, is used to define the device operating mode. The user must make sure that this pin is tied at low level to ensure normal operating conditions. Other values associated with this pin are reserved for manufacturing test.
11.6.2 EmbeddedICE
The Cortex-A5 EmbeddedICE-RT is supported via the ICE/JTAG port. It is connected to a host computer via an ICE interface.The internal state of the Cortex-A5 is examined through an ICE/JTAG port which allows instructions to be serially inserted into the pipeline of the core without using the external data bus. Therefore, when in debug state, a storemultiple (STM) can be inserted into the instruction pipeline. This exports the contents of the Cortex-A5 registers. This data can be serially shifted out without affecting the rest of the system. There are two scan chains inside the Cortex-A5 processor which support testing, debugging, and programming of the EmbeddedICE-RT. The scan chains are controlled by the ICE/JTAG port. EmbeddedICE mode is selected when JTAGSEL is low. It is not possible to switch directly between ICE and JTAG operations. A chip reset must be performed after JTAGSEL is changed. For further details on the EmbeddedICE-RT, see the ARM document: ARM IHI 0031A_ARM_debug_interface_v5.pdf
11.6.3 JTAG Signal Description

TMS is the Test Mode Select input which controls the transitions of the test interface state machine. TDI is the Test Data Input line which supplies the data to the JTAG registers (Boundary Scan Register, Instruction Register, or other data registers). TDO is the Test Data Output line which is used to serially output the data from the JTAG registers to the equipment controlling the test. It carries the sampled values from the boundary scan chain (or other JTAG registers) and propagates them to the next chip in the serial test circuit. NTRST (optional in IEEE Standard 1149.1) is a Test-ReSeT input which is mandatory in ARM cores and used to reset the debug logic. On Atmel Cortex-A5-based cores, NTRST is a Power On Reset output. It is asserted on power on. If necessary, the user can also reset the debug logic with the NTRST pin assertion during 2.5 MCK periods. TCK is the Test ClocK input which enables the test interface. TCK is pulsed by the equipment controlling the test and not by the tested device. It can be pulsed at any frequency. RTCK is the Return Test Clock. Not an IEEE Standard 1149.1 signal added for a better clock handling by emulators. From some ICE Interface probes, this return signal can be used to synchronize the TCK clock. This signal is only available in JTAG ICE Mode and not in boundary scan mode.
11.6.4 Chip Access Using JTAG Connection

In some cases, the JTAG connection is not allowed on this chip (JMem, SAM-BA, etc.) due to the Secure ROM Code implementation. By default, the SAMA5D3 devices boot in Standard mode and not in Secure mode. When the Secure ROM Code starts, it disables the JTAG access for the whole boot sequence. If the Secure ROM Code does not find any program in the external memory, it enables the USB connection and waits for a dedicated command to switch the chip into Secure mode. If any other character is received, the Secure ROM Code starts the Standard SAM-BA Monitor, locks access to the ROM memory, and enables the JTAG. Then you can access to the chip using the JTAG connection. If the Secure ROM Code finds a bootable program, it disables automatically ROM access and enables JTAG just before launching the program.

48
The procedure to enable JTAG access is as follows:

z z z z z
Connect your computer to the board with JTAG and USB (J20 USB-A) Power on the chip Open a terminal console (TeraTerm or HyperTerminal, etc.) on your computer and connect to the USB CDC Serial COM port related to the J20 connector on the board Send the '#' character. You will see then the prompt '>' character sent by the device (indicating that the Standard SAM-BA Monitor is running) Use the Standard SAM-BA Monitor to connect to the chip with JTAG
Note that you don't need to follow this sequence in order to connect the Standard SAM-BA Monitor with USB.
11.6.5 Debug Unit

The Debug Unit provides a two-pin (DXRD and TXRD) USART that can be used for several debug and trace purposes and offers an ideal means for in-situ programming solutions and debug monitor communication. Moreover, the association with two peripheral data controller channels permits packet handling of these tasks with processor time reduced to a minimum. The Debug Unit also manages the interrupt handling of the COMMTX and COMMRX signals that come from the ICE and that trace the activity of the Debug Communication Channel.The Debug Unit allows blockage of access to the system through the ICE interface. A specific register, the Debug Unit Chip ID Register, gives information about the product version and its internal configuration. For further details on the Debug Unit, see the Debug Unit section.
11.6.6 IEEE 1149.1 JTAG Boundary Scan

IEEE 1149.1 JTAG Boundary Scan allows pin-level access independent of the device packaging technology. IEEE 1149.1 JTAG Boundary Scan is enabled when JTAGSEL is high. The SAMPLE, EXTEST and BYPASS functions are implemented. In ICE debug mode, the ARM processor responds with a non-JTAG chip ID that identifies the processor to the ICE system. This is not IEEE 1149.1 JTAG-compliant. It is not possible to switch directly between JTAG and ICE operations. A chip reset must be performed after JTAGSEL is changed. A Boundary-scan Descriptor Language (BSDL) file is provided to set up test.

49
11.7
Access:
The Boundary JTAG ID Register

Read-only
30 29 28 27 26 25 24 31
VERSION
23 22 21 20 19
PART NUMBER
18 17 16
PART NUMBER
15 14 13 12 11 10 9 8
PART NUMBER
7 6 5 4 3
MANUFACTURER IDENTITY
2 1 0
MANUFACTURER IDENTITY
VERSION[31:28]: Product Version Number

Set to 0x0.
PART NUMBER[27:12]: Product Part Number

Product part Number is 0x5B31
MANUFACTURER IDENTITY[11:1]
Set to 0x01F. Bit[0] required by IEEE Std. 1149.1. Set to 0x1. JTAG ID Code value is 0x05B3_103F.

50
11.8
The Cortex-A5 DP Identification Code Register IDCODE

The Identification Code Register is always present on all DP implementations. It provides identification information about the ARM Debug Interface.
11.8.1 JTAG Debug Port (JTAG-DP)

It is accessed using its own scan chain, the JTAG-DP Device ID Code Register Access:
31
Read-only
30 29 28 27 26 25 24
VERSION
23 22 21 20 19
PART NUMBER
18 17 16
PART NUMBER
15 14 13 12 11 10 9 8
PART NUMBER
7 6 5 4 3 2
DESIGNER
1 0
DESIGNER

Set to 0x0.

Product part Number is 0xBA00
DESIGNER[11:1]
Set to 0x23B. Bit[0] required by IEEE Std. 1149.1. Set to 0x1. Cortex-A5 JTAG-DP IDCODE value is 0x0BA0_0477

51
11.8.2 Serial Wire Debug Port (SW-DP)

It is at address 0x0 on read operations when the APnDP bit = 0. Access to the Identification Code Register is not affected by the value of the CTRLSEL bit in the Select Register Access:
31
Read-only
30 29 28 27 26 25 24
VERSION
23 22 21 20 19
PART NUMBER
18 17 16
PART NUMBER
15 14 13 12 11 10 9 8
PART NUMBER
7 6 5 4 3 2
DESIGNER
1 0
DESIGNER

Set to 0x0.

Product part Number is 0xBA01
DESIGNER[11:1]
Set to 0x23B. Bit[0] required by IEEE Std. 1149.1. Set to 0x1. Cortex-A5 SW-DP IDCODE is 0x0BA0_1477

52
12.
Standard Boot Strategies

The system always boots from the ROM memory at address 0x0. The ROM Code is a boot program contained in the embedded ROM. It is also called First level bootloader. This microcontroller can be configured to run a Standard Boot Mode or a Secure Boot Mode. More information on how the Secure Boot Mode can be enabled, and how the chip operates in this mode, is provided in the application note Secure Boot on SAMAD3 Series, literature number 11165A. Please refer to the Atmel web site at www.atmel.com. By default, the chip starts in Standard Boot Mode. Note: JTAG access is disabled during the execution of ROM Code Sequence. It is re-enabled when jumping into SRAM when a valid code has been found on an external NVM, or after reception of a character through USB connection when entering SAM-BA Monitor.
The user can choose to boot from an external NOR Flash memory with the help of the BMS pin. The sampling of the BMS pin is done by hardware at reset, and the result is available in the BMS_EBI bit of the SFR_EBICFG register. The first steps of the ROM Code program is to check the state of this pin by reading this register. If BMS signal is tied to 0, BMS_BIT is read at 1 The ROM Code allows execution of the code contained into the memory connected to Chip Select 0 of the External Bus Interface. To achieve that, the following sequence is preformed by the ROM Code:
z z z z
The main clock is the on-chip 12 MHz RC oscillator, The Static Memory Controller is configured with timing allowing code execution inCS0 external memory at 12 MHz AXI matrix is configured to remap EBI CS0 address at 0x0 0x0 is loaded in the Program Counter register
The user software in the external memory must perform the next operation in order to complete the clocks and SMC timings configuration to run at a higher clock frequency:
z z z z z
Enable the 32768 Hz oscillator if best accuracy is needed Reprogram the SMC setup, cycle, hold, mode timing registers for EBI CS0, to adapt them to the new clock Program the PMC (Main Oscillator Enable or Bypass mode) Program and Start the PLL Switch the system clock to the new value
If BMS signal is tied to 1, BMS_BIT is read at 0 The ROM Code standard sequence is executed as follows:
z z z
Basic chip initialization: XTal or external clock frequency detection Attempt to retrieve a valid code from external non-volatile memories (NVM) Execution of a monitor called SAM-BA Monitor, in case no valid application has been found on any NVM

53
12.1
Flow Diagram
The ROM Code implements the algorithm shown in Figure 12-1.
Figure 12-1. ROM Code Algorithm Flow Diagram
Chip Setup
Valid boot code found in one NVM
Yes
Copy and run it in internal SRAM
No
SAM-BA Monitor
12.2
Chip Setup
At boot start-up, the processor clock (PCK) and the master clock (MCK) source is the 12 MHz Fast RC Oscillator. Initialization follows the steps described below: 1. 2. Stack Setup for ARM supervisor mode Main Oscillator Detection: The Main Clock is switched to the 32 kHz RC oscillator to allow external clock frequency to be measured. Then the Main Oscillator is enabled and set in the bypass mode. If the MOSCSELS bit rises, an external clock is connected, and the next step is Main Clock Selection (3). If not, the bypass mode is cleared to attempt external quartz detection. This detection is successful when the MOSCXTS and MOSCSELS bits rise, else the 12 MHz Fast RC internal oscillator is used as the Main Clock. Main Clock Selection: The Master Clock source is switched from the Slow Clock to the Main Oscillator without prescaler. The PMC Status Register is polled to wait for MCK Ready. PCK and MCK are now the Main Clock. C Variable Initialization: Non zero-initialized data is initialized in the RAM (copy from ROM to RAM). Zero-initialized data is set to 0 in the RAM. PLLA Initialization: PLLA is configured to get a PCK at 96 MHz and an MCK at 48 MHz. If an external clock or crystal frequency running at 12 MHz is found, then the PLLA is configured to allow communication on the USB link for the SAM-BA Monitor; else the Main Clock is switched to the internal 12 MHz Fast RC, but USB will not be activated.
3. 4. 5.

54
12.3
NVM Boot
12.3.1 NVM Boot Sequence

The boot sequence on external memory devices can be controlled using the Boot Sequence Configuration Register (BSC_CR). The user can then choose to bypass some steps shown in Figure 12-2 NVM Bootloader Sequence Diagram according to the BOOT value in the BSC_CR register.
Table 12-1. Values of the Boot Sequence Configuration Register BOOT Value SPI0 NPCS0 0 1 2 3 4 5 6 7 Y Y Y Y Y SD Card / eMMC (MCI0) Y SD Card / eMMC (MCI1) Y Y NAND Flash Y Y Y SAM-BA Monitor Y Y Y Y Y Y Y Y
SPI0 NPCS1 Y Y Y Y Y -
TWI EEPROM Y Y Y Y Y -

55
Figure 12-2. NVM Bootloader Sequence Diagram

Device Setup
SPI0 CS0 Flash Boot
Yes
Copy from SPI Flash to SRAM
Run
SPI Flash Bootloader
No
SD Card Boot
Yes
Copy from SD Card to SRAM
Run
SD Card Bootloader
No
NAND Flash Boot
Yes
Copy from NAND Flash to SRAM
Run
NAND Flash Bootloader
No
SPI0 CS1 Flash Boot
Yes
Copy from SPI Flash to SRAM
Run
SPI Flash Bootloader
No
TWI EEPROM Boot
Yes
Copy from TWI EEPROM to SRAM
Run
TWI EEPROM Bootloader
No
SAM-BA Monitor

56
12.3.2 NVM Bootloader Program Description

Figure 12-3. NVM Bootloader Program Diagram
Start
Initialize NVM
Initialization OK ?
No
Restore the reset values for the peripherals and Jump to next boot solution
Yes Valid code detection in NVM
NVM contains valid code
No
Yes Copy the valid code from external NVM to internal SRAM.
Restore the reset values for the peripherals. Perform the REMAP and set the PC to 0 to jump to the downloaded application
End

57
The NVM bootloader program first initializes the PIOs related to the NVM device. Then it configures the right peripheral depending on the NVM and tries to access this memory. If the initialization fails, it restores the reset values for the PIO and the peripheral, and then tries to fulfill the same operations on the next NVM of the sequence. If the initialization is successful, the NVM bootloader program reads the beginning of the NVM and determines if the NVM contains a valid code. If the NVM does not contain a valid code, the NVM bootloader program restores the reset value for the peripherals and then tries to fulfill the same operations on the next NVM of the sequence. If a valid code is found, this code is loaded from the NVM into the internal SRAM and executed by branching at address 0x0000_0000 after remap. This code may be the application code or a second-level bootloader. All the calls to functions are PC relative and do not use absolute addresses.
Figure 12-4. Remap Action after Download Completion
0x0000_0000 Internal ROM 0x0010_0000 Internal ROM 0x0030_0000 Internal SRAM Internal SRAM Internal ROM 0x0030_0000 REMAP Internal SRAM 0x0010_0000 0x0000_0000
12.3.3 Valid Code Detection

There are two kinds of valid code detection. 12.3.3.1 ARM Exception Vectors Check The NVM bootloader program reads and analyzes the first 28 bytes corresponding to the first seven ARM exception vectors. Except for the sixth vector, these bytes must implement the ARM instructions for either branch or load PC with PC relative addressing.
Figure 12-5. LDR Opcode
31 1 1 1 28 27 0 0 1 I 24 23 P U 1 W 20 19 0 Rn 16 15 Rd 12 11 Offset 0
Figure 12-6. B Opcode

31 1 1 1 28 27 0 1 0 1 24 23 0 Offset (24 bits) 0

58
Unconditional instruction: 0xE for bits 31 to 28. Load PC with the PC relative addressing instruction:
z z z z z
Rn = Rd = PC = 0xF I==0 (12-bit immediate value) P==1 (pre-indexed) U offset added (U==1) or subtracted (U==0) W==1
The sixth vector, at the offset 0x14, contains the size of the image to download. The user must replace this vector with the users own vector. This procedure is described below.
Figure 12-7. Structure of the ARM Vector 6
31 Size of the code to download in bytes 0
The value has to be smaller than 64 Kbytes.
Example
An example of valid vectors: 00 04 08 0c 10 14 18 12.3.3.2 boot.bin File Check This method is the one used on FAT formatted SD Card and eMMC. The boot program must be a file named boot.bin written in the root directory of the file system. Its size must not exceed the maximum size allowed: 64 Kbytes (0x10000). ea000006 eafffffe ea00002f eafffffe eafffffe 00001234 eafffffe B0x20 B0x04 B_main B0x0c B0x10 B0x14<- Code size = 4660 bytes B0x18
12.3.4 Detailed Memory Boot Procedures

12.3.4.1 NAND Flash Boot: NAND Flash Detection After the NAND Flash interface configuration, a reset command is sent to the memory. Hardware ECC detection and correction are provided by the PMECC peripheral. Please refer to the PMECC Controller Functional Description section of this datasheet for more details. The Boot Program is able to retrieve NAND Flash parameters and ECC requirements using two methods as follows:
z
The detection of a specific header written at the beginning of the first page of the NAND Flash, Through the ONFI parameters for the ONFI compliant memories
or
z

59
Figure 12-8. Boot NAND Flash Download

Start
Initialize NAND Flash interface
Send Reset command
No First page contains valid header NAND Flash is ONFI Compliant
No
Yes
Yes
Read NAND Flash and PMECC parameters from the header
Read NAND Flash and PMECC parameters from the ONFI
Copy the valid code from external NVM to internal SRAM.
Restore the reset values for the peripherals. Perform the REMAP and set the PC to 0 to jump to the downloaded application
End
Restore the reset values for the peripherals and Jump to next bootable memory

60
NAND Flash Specific Header Detection

This is the first method used to determine NAND Flash parameters. After Initialization and Reset command, the Boot Program reads the first page without an ECC check, to determine if the NAND parameter header is present. The header is made of 52 times the same 32-bit word (for redundancy reasons) which must contain NAND and PMECC parameters used to correctly perform the read of the rest of the data in the NAND. This 32-bit word is described below:
31 30 29 28 27 26 25 24
key
23 22 21 20
19 18
eccOffset
17 16
eccOffset
15 14 13 12 11 10 9
sectorSize
8
eccBitReq
7 6 5 4 3
spareSize
2 1 0
spareSize
nbSectorPerPage
usePmecc
usePmecc: Use PMECC

0 = Do not use PMECC to detect and correct the data 1 = Use PMECC to detect and correct the data
nbSectorPerPage: Number of Sectors per Page spareSize: Size of the Spare Zone in Bytes eccBitReq: Number of ECC Bits Required
0 = 2-bit ECC 1 = 4-bit ECC 2 = 8-bit ECC 3 = 12-bit ECC 4 = 24-bit ECC
sectorSize: Size of the ECC Sector

0 = For 512 bytes 1 = For 1024 bytes per sector Other value for future use.
eccOffset: Offset of the First ECC Byte in the Spare Zone

A value below 2 is not allowed and will be considered as 2.
key: Value 0xC Must be Written here to Validate the Content of the Whole Word.
If the header is valid, the Boot Program continues with the detection of a valid code.

61
ONFI 2.2 Parameters

In case no valid header is found, the Boot Program checks if the NAND Flash is ONFI compliant, sending a Read Id command (0x90) with 0x20 as parameter for the address. If the NAND Flash is ONFI compliant, the Boot Program retrieves the following parameters with the help of the Get Parameter Page command:
z z z z
Number of bytes per page (byte 80) Number of bytes in spare zone (byte 84) Number of ECC bit correction required (byte 112) ECC sector size: by default, set to 512 bytes; or to 1024 bytes if the ECC bit capability above is 0xFF
By default, the ONFI NAND Flash detection will turn ON the usePmecc parameter, and the ECC correction algorithm is automatically activated. Once the Boot Program retrieves the parameter, using one of the two methods described above, it reads the first page again, with or without ECC, depending on the usePmecc parameter. Then it looks for a valid code programmed just after the header offset 0xD0. If the code is valid, the program is copied at the beginning of the internal SRAM. Note: Booting on 16-bit NAND Flash is not possible, only 8-bit NAND Flash memories are supported.
12.3.4.2 NAND Flash Boot: PMECC Error Detection and Correction NAND Flash boot procedure uses PMECC to detect and correct errors during NAND Flash read operations in two cases:
z z
When the usePmecc flag is set in a specific NAND header. If the flag is not set, no ECC correction is performed during the NAND Flash page read. When the NAND Flash has been detected using ONFI parameters.
The ROM memory embeds the Galois field tables. The user does not need to embed them in his own software. The Galois field tables are mapped in the ROM just after the ROM code, as described in Figure 12-9.
Figure 12-9. Galois Field Table Mapping
0x0010_0000 0x0010_0000
ROM Code Code ROM 0x0010_8000 0x0010_8000 Galois field field Galois tables for for tables 512-byte 512-byte sectors sectors correction correction
0x0011_0000 0x0011_0000
Galois field field Galois tables for for tables 1024-byte 1024-byte sectors sectors correction correction
For a full description and an example of how to use the PMECC detection and correction feature, refer to the software package dedicated to this device on Atmels web site.

62
12.3.4.3 SD Card / eMMC Boot The SD Card / eMMC bootloader looks for a boot.bin file in the root directory of a FAT12/16/32 file system.
Supported SD Card Devices

SD Card Boot supports all SD Card memories compliant with the SD Memory Card Specification V2.0. This includes SDHC cards. 12.3.4.4 SPI Flash Boot Two types of SPI Flash are supported: SPI Serial Flash and SPI DataFlash. The SPI Flash bootloader tries to boot on SPI0, first looking for SPI Serial Flash, and then for SPI DataFlash. It uses only one valid code detection: analysis of ARM exception vectors. The SPI Flash read is done by means of a Continuous Read command from the address 0x0. This command is 0xE8 for DataFlash and 0x0B for Serial Flash devices.
Supported DataFlash Devices

The SPI Flash Boot program supports all Atmel DataFlash devices.
Table 12-2. DataFlash Device Device AT45DB011 AT45DB021 AT45DB041 AT45DB081 AT45DB161 AT45DB321 AT45DB642 Density 1 Mbit 2 Mbits 4 Mbits 8 Mbits 16 Mbits 32 Mbits 64 Mbits Page Size (bytes) 264 264 264 264 528 528 1056 Number of Pages 512 1024 2048 4096 4096 8192 8192
Supported Serial Flash Devices

The SPI Flash Boot program supports all SPI Serial Flash devices responding correctly at both Get Status and Continuous Read commands. 12.3.4.5 TWI EEPROM Boot The TWI EEPROM Bootloader uses the TWI0. It uses only one valid code detection. It analyzes the ARM exception vectors.
Supported TWI EEPROM Devices

TWI EEPROM Boot supports all I2C-compatible TWI EEPROM memories using the 7-bit device address 0x50.
12.3.5 Hardware and Software Constraints

The NVM drivers use several PIOs in peripheral mode to communicate with external memory devices. Care must be taken when these PIOs are used by the application. The connected devices could be unintentionally driven at boot time, and thus electrical conflicts between output pins used by the NVM drivers and the connected devices could occur. To assure the correct functionality, it is recommended to plug in critical devices to other pins, not used by the NVM. Table 12-3 contains a list of pins that are driven during the boot program execution. These pins are driven during the boot sequence for a period of less than 1 second if no correct boot program is found.

63
Before performing the jump to the application in the internal SRAM, all the PIOs and peripherals used in the boot program are set to their reset state.
Table 12-3. PIO Driven during Boot Program Execution NVM Bootloader Peripheral EBI CS3 SMC EBI CS3 SMC EBI CS3 SMC NAND EBI CS3 SMC EBI CS3 SMC EBI CS3 SMC MCI0 MCI0 MCI0 MCI0 MCI0 MCI0 SD Card / eMMC MCI1 MCI1 MCI1 MCI1 MCI1 MCI1 SPI0 SPI0 SPI Flash SPI0 SPI0 SPI0 TWI0 TWI0 EEPROM TWI0 DBGU SAM-BA Monitor DBGU DTXD PIOB31 TWCK0 DRXD PIOA31 PIOB30 MCI1_CK MCI1_CDA MCI1_D0 MCI1_D1 MCI1_D2 MCI1_D3 MOSI MISO SPCK NPCS0 NPCS1 TWD0 PIOB24 PIOB19 PIOB20 PIOB21 PIOB22 PIOB23 PIOD11 PIOD10 PIOD12 PIOD13 PIOD14 PIOA30 NAND ALE NAND CLE Cmd/Addr/Data MCI0_CK MCI0_CDA MCI0_D0 MCI0_D1 MCI0_D2 MCI0_D3 PIOD9 PIOD0 PIOD1 PIOD2 PIOD3 PIOD4 Pin NANDOE NANDWE NANDCS PIO Line -

64
12.4
SAM-BA Monitor
If no valid code is found in the NVM during the NVM bootloader sequence, the SAM-BA Monitor program is launched. The SAM-BA Monitor principle is to:
z z z
Initialize DBGU and USB Check if USB Device enumeration occurred Check if characters are received on the DBGU
Once the communication interface is identified, the application runs in an infinite loop waiting for different commands as listed in Table 12-4.
Figure 12-10. SAM-BA Monitor Diagram
No valid code in NVM
Init DBGU and USB No
USB Enumeration Successful ?
No
Character(s) received on DBGU ?
Yes Run monitor Wait for command on the USB link
Yes Run monitor Wait for command on the DBGU link
12.4.1 Command List
Table 12-4. Commands Available through the SAM-BA Monitor Command N T O o H h W w S R G V Action Set Normal Mode Set Terminal Mode Write a byte Read a byte Write a half word Read a half word Write a word Read a word Send a file Receive a file Go Display version Argument(s) No argument No argument Address, Value# Address,# Address, Value# Address,# Address, Value# Address,# Address,# Address, NbOfBytes# Address# No argument Example N# T# O200001,CA# o200001,# H200002,CAFE# h200002,# W200000,CAFEDECA# w200000,# S200000,# R200000,1234# G200200# V#

65
Mode commands:
z z
Normal mode configures SAM-BA Monitor to send / receive data in binary format, Terminal mode configures SAM-BA Monitor to send / receive data in ASCII format. Address: Address in hexadecimal Value: Byte, halfword or word to write in hexadecimal Output: > Address: Address in hexadecimal Output: The byte, halfword or word read in hexadecimal followed by > Address: Address in hexadecimal Output: >
Write commands: Write a byte (O), a halfword (H) or a word (W) to the target
z z z
Read commands: Read a byte (o), a halfword (h) or a word (w) from the target
z z
Send a file (S): Send a file to a specified address

z z
Note:
z
There is a time-out on this command which is reached when the prompt > appears before the end of the command execution. Receive a file (R): Receive data into a file from a specified address
z z z
Address: Address in hexadecimal

NbOfBytes: Number of bytes in hexadecimal to receive
Output: > Address: Address to jump in hexadecimal Output: > once returned from the program execution. If the executed program does not handle the link register at its entry and does not return, the prompt will not be displayed Output: version, date and time of ROM code followed by >
Go (G): Jump to a specified address and execute the code

z z
Get Version (V): Return the Boot Program version

z
12.4.2 DBGU Serial Port

Communication is performed through the DBGU serial port initialized to 115,200 Baud, 8 bits of data, no parity, 1 stop bit. 12.4.2.1 Supported External Crystal/External Clocks The SAM-BA Monitor supports a frequency of 12, 16, 24 or 48 MHz to allow DBGU communication for both external crystal and external clock. 12.4.2.2 Xmodem Protocol The Send and Receive File commands use the Xmodem protocol to communicate. Any terminal performing this protocol can be used to send the application file to the target. The size of the binary file to send depends on the SRAM size embedded in the product. In all cases, the size of the binary file must be lower than the SRAM size because the Xmodem protocol requires some SRAM memory in order to work. The Xmodem protocol supported is the 128-byte length block. This protocol uses a two-character CRC16 to guarantee detection of maximum bit errors. Xmodem protocol with CRC is supported by successful transmission reports provided both by a sender and by a receiver. Each transfer block is as follows: <SOH><blk #><255-blk #><--128 data bytes--><checksum> in which:
z z z z
<SOH> = 01 hex <blk #> = binary number, starts at 01, increments by 1, and wraps 0FFH to 00H (not to 01) <255-blk #> = 1s complement of the blk#. <checksum> = 2 bytes CRC16

66
Figure 12-11 shows a transmission using this protocol.

Figure 12-11. Xmodem Transfer Example
Host C SOH 01 FE Data[128] CRC CRC ACK SOH 02 FD Data[128] CRC CRC ACK SOH 03 FC Data[100] CRC CRC ACK EOT ACK Device
12.4.3 USB Device Port

12.4.3.1 Supported External Crystal / External Clocks The SAM-BA Monitor supports a frequency of 12, 16, 24 or 48 MHz to allow USB communication for both external crystal and external clock. 12.4.3.2 USB Class The device uses the USB Communication Device Class (CDC) drivers to take advantage of the installed PC Serial Communication software to talk over the USB. The CDC class is implemented in all releases of Windows , from Windows 98SE to Windows 7. The CDC document, available at www.usb.org, describes how to implement devices such as ISDN modems and virtual COM ports. The Vendor ID is the Atmels vendor ID 0x03EB. The product ID is 0x6124. These references are used by the host operating system to mount the correct driver. On Windows systems, INF files contain the correspondence between vendor ID and product ID. 12.4.3.3 Enumeration Process The USB protocol is a master/slave protocol. The host starts the enumeration, sending requests to the device through the control endpoint. The device handles standard requests as defined in the USB Specification.
Table 12-5. Handled Standard Requests Request GET_DESCRIPTOR SET_ADDRESS SET_CONFIGURATION GET_CONFIGURATION GET_STATUS SET_FEATURE CLEAR_FEATURE Definition Returns the current device configuration value Sets the device address for all future device access Sets the device configuration Returns the current device configuration value Returns status for the specified recipient Used to set or enable a specific feature Used to clear or disable a specific feature

67
The device also handles some class requests defined in the CDC class.
Table 12-6. Handled Class Requests Request SET_LINE_CODING GET_LINE_CODING SET_CONTROL_LINE_STATE Definition Configures DTE rate, stop bits, parity and number of character bits Requests current DTE rate, stop bits, parity and number of character bits RS-232 signal used to indicate to the DCE device that the DTE device is now present
Unhandled requests are STALLed. 12.4.3.4 Communication Endpoints Endpoint 0 is used for the enumeration process. Endpoint 1 (64-byte Bulk OUT) and endpoint 2 (64-byte Bulk IN) are used as communication endpoints. SAM-BA Boot commands are sent by the host through Endpoint 1. If required, the message is split into several data payloads by the host driver. If the command requires a response, the host sends IN transactions to pick up the response.

68
13.
13.1
Boot Sequence Controller (BSC)

Description
The System Controller embeds a Boot Sequence Configuration Register to save timeout delays on boot. The boot sequence is programmable through the Boot Sequence Configuration Register (BSC_CR). This register is powered by VDDBU, the modification is saved and applied after the next reset. The register is taking Factory Value in case of battery removing. This register is programmable with user programs or SAM-BA and key-protected.
13.2
z
VDDBU powered register
13.3
Product Dependencies
z
Product-dependent order

69
13.4
Boot Sequence Controller (BSC) User Interface
Table 13-1. Register Mapping Offset 0x0 Register Boot Sequence Configuration Register Name BSC_CR Access Read-write Reset
13.4.1 Boot Sequence Configuration Register

Name: Address: Access: Factory Value:
31
BSC_CR 0xFFFFFE54 Read-write 0x0000_0000

30 29 28 BOOTKEY 27 26 25 24
23
22
21
20 BOOTKEY
19
18
17
16
15 7
14 6
13 5
12 4 BOOT
11 3
10 2
9 1
8 0
BOOT: Boot Media Sequence

This value is defined in the product-dependent ROM code. It is only written if BOOTKEY carries the valid value. Please refer to the NVM Boot Sequence section of this datasheet for details on the BOOT value.
BOOTKEY
0x6683 (BSC_KEY): valid key to write BSC_CR register; it needs to be written at the same time as the BOOT field. Other values disable the write access. This key field is write-only.

70
14.
14.1
AXI Bus Matrix (AXIMX)

Description
The AXI Bus Matrix (AXIMX) comprises the embedded Advanced E X tensible Interface (AXI) bus protocol which supports separate address/control and data phases, unaligned data transfers using byte strobes, burst-based transactions with only start address issued, separate read and write data channels to enable low-cost DMA, ability to issue multiple outstanding addresses, out-of-order transaction completion, and easy addition of register stages to provide timing closure.
14.2
z z z z z z z z
High Performance AXI Network Interconnect 1 AXI Slave Interface 1 AHB-Lite Slave Interface 3 AXI Master Interfaces 1 APB3 Slave Interface Single-cycle Arbitration Full Pipelining to prevent Master Stalls 2 Remap States

71
14.3
Operation
14.3.1 Remap
There are two remap states using bits 0 and 1 in the Remap Register (AXIMX_REMAP)
z z
Bit 1 is used to remap EBI @ addr 0x00000000 for external boot. Bit 0 is used to remap RAM @ addr 0x00000000
Refer to Section 14.4 AXI Matrix (AXIMX) User Interface" and Table 14-1, Register Mapping. The number of remap states can be defined using eight bits of the remap register, and a bit in the remap register controls each remap state. Each remap state can be used to control the address decoding for one or more slave interfaces. If a slave interface is affected by two remap states that are both asserted, the remap state with the lowest remap bit number takes precedence. Each slave interface can be configured independently so that a remap state can perform different functions for different masters. A remap state can:
z z z
Alias a memory region into two different address ranges Move an address region Remove an address region
Because of the nature of the distributed register sub-system, the masters receive the updated remap bit states in sequence, and not simultaneously. A slave interface does not update to the latest remap bit setting until:
z z
The address completion handshake accepts any transaction that is pending Any current lock sequence completes
The BRESP from a GPV after a remap update guarantees that the next transaction issued to each slave interface, or the first one after the completion of a locked sequence, uses the updated value. The AXI Matrix uses two remap bits. At powerup, ROM is seen at address 0 After powerup, ahbslave can be moved down to address 0 by means of the remap bits. Figure 14-1 shows the memory map when remap is set to 000, representing no remap,

72
Figure 14-1. No Remap

0x00000000
ROM
0x000FFFFF 0x00100000 0x001FFFFF 0x00200000 0x002FFFFF 0x00300000 0x0007FFFFF 0x00800000 0x008FFFFF 0x00900000 0x09FFFFF 0x0A00000 0x001FFFFF 0x01000000 0x01FFFFFF 0x20000000 0x3FFFFFFF 0x40000000 0x7FFFFFFF 0x80000000 0xEFFFFFFF 0xF0000000 0xFFFFFFFF
ROM ahbslave ahbslave gpv_0 dap [apb3bridge] reserved ahbslave MPDDR ahbslave reserved ahbslave
Figure 14-2 shows mapping when remap state is 01 or 11. This state is used for RAM boot. RAM is seen at address 0 through ahbslave.
Figure 14-2. Remap state is 01 or 11
0x00000000
ahbslave (RAM)
ROM ahbslave ahbslave gpv_0 dap [apb3bridge] reserved ahbslave MPDDR ahbslave reserved ahbslave
Figure 14-3 shows mapping when remap state is 10. This state is used for external boot. EBI is seen at address 0 through ahbslave.

73
Figure 14-3. Remap State is 10

0x00000000
ahbslave (EBI)
ROM ahbslave ahbslave gpv_0 dap[apb3bridge] reserved ahbslave MPDDR ahbslave reserved ahbslave

74
14.4
AXI Matrix (AXIMX) User Interface
Table 14-1. Register Mapping Offset 0x00 0x04 - 0x43108 Register Remap Register Reserved Name AXIMX_REMAP Access Write-only Reset 0x00000000 0x00000000

75
14.4.1 AXI Matrix Remap Register

Name: Address: Access: Reset:
31
AXIMX_REMAP 0x00800000 Read-write 0x00000000

30 29 28 27 26 25 24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
1 REMAP1
0 REMAP0
REMAP0 has higher priority than REMAP1, i.e., if both REMAP0 & REMAP1 are asserted, the matrix is in remap state 0.
REMAP0: Remap State 0

SRAM is seen at address 0x00000000 (through AHB slave interface) instead of ROM.
REMAP1: Remap State 1

HEBI is seen at address 0x00000000 (through AHB slave interface) instead of ROM for external boot.

76
15.
15.1
Bus Matrix (MATRIX)

Description
The Bus Matrix (MATRIX) implements a multi-layer AHB, based on the AHB-Lite protocol, that enables parallel access paths between multiple AHB masters and slaves in a system, thus increasing the overall bandwidth. The Bus Matrix interconnects up to 16 AHB masters to up to 16 AHB slaves. The normal latency to connect a master to a slave is one cycle except for the default master of the accessed slave which is connected directly (zero cycle latency). The Bus Matrix user interface is compliant with ARM Advanced Peripheral Bus.
15.1.1 Matrix Masters

The Bus Matrix of the SAMA5D3 product manages 15 masters, which means that each master can perform an access concurrently with others, to an available slave. Each master has its own decoder, which is defined specifically for each master. In order to simplify the addressing, all the masters have the same decodings. List of Bus Matrix Masters
Master 0 Master 1, 2, 3 Master 4, 5, 6 Master 7 Master 8, 9 Master 10 Master 11 Master 12 Master 13 Master 14 Cortex A5 DMA Controller 0 DMA Controller 1 GMAC DMA LCDC DMA UHP EHCI DMA UHP OHCI DMA UDPHS DMA EMAC DMA ISI DMA
15.1.2 Matrix Slaves

The Bus Matrix of the SAMA5 product manages 13 slaves. Each slave has its own arbiter, allowing a different arbitration per slave.
Table 15-1. List of Bus Matrix Slaves Slave 0 Slave 1 Slave 2 Slave 3 Slave 4 Internal SRAM0 Internal SRAM1 NFC SRAM Internal ROM Soft Modem (SMD) USB Device High Speed Dual Port RAM (DPR) Slave 5 USB Host OHCI registers USB Host EHCI registers Slave 6 Slave 7 Slave 8 Slave 9 External Bus Interface/NFC DDR2 Port0 DDR2 Port1 DDR2 Port2

77
Table 15-1. List of Bus Matrix Slaves Slave 10 Slave 11 Slave 12 DDR2 Port3 Peripheral Bridge 0 Peripheral Bridge 1
15.1.3 Master to Slave Access

All the Masters can normally access all the Slaves. However, some paths do not make sense, for example allowing access from the USB Device High speed DMA to the Internal Peripherals. Thus, these paths are forbidden or simply not wired, and shown as - in the following table.
Table 15-2. SAMA5 Master to Slave Access Masters 0 1 2 3 4 5 6 7 GMAC DMA X X 8 9 10 UHPHS EHCI DMA X X 11 UHPHS OHCI DMA X X 12 UDPHS DMA X X 13 EMAC DMA X X 14 ISI DMA
Slaves 0 1 2 3 4 Internal SRAM0 Internal SRAM1 NFC SRAM Internal ROM SMD UDPHS RAM 5 UHP OHCI Reg UHP EHCI Reg EBI CS0..CS3 6 NFC Command Register DDR2 Port 0 DDR2 port1 DDR2 port2 DDR2 port3 APB 0 APB 1
A5 X X X X X X X X X X
DMAC0
DMAC1 X X
LCDC DMA
X X X
7 8 9 10 11 12
X X
X X
X X X X X X X X
X X
X X

78
15.2
z z z z z z z z z z
AMBA Advanced High-performance Bus (AHB Lite) Compliant Interfaces 32-bit or 64-bit Data Bus APB Compliant User Interface Configurable Number of Masters (Up to sixteen) Configurable Number of Slaves (Up to sixteen) One Decoder for Each Master Several Possible Boot Memories for Each Master before Remap One Remap Function for Each Master Support for Long Bursts of 32, 64, 128 and Up to the 256-beat Word Burst AHB Limit Enhanced Programmable Mixed Arbitration for Each Slave
z z
Round-Robin Fixed Priority No Default Master Last Accessed Default Master Fixed Default Master
Programmable Default Master for Each Slave

z z z
z z z z z z
Deterministic Maximum Access Latency for Masters Zero or One Cycle Arbitration Latency for the First Access of a Burst Bus Lock Forwarding to Slaves Master Number Forwarding to Slaves One Special Function Register for Each Slave (Not dedicated) Write Protection of User Interface Registers

79
15.3
Memory Mapping
The Bus Matrix provides one decoder for every AHB master interface. The decoder offers each AHB master several memory mappings. Each memory area may be assigned to several slaves. Booting at the same address while using different AHB slaves (i.e. external RAM, internal ROM or internal Flash, etc.) becomes possible. The Bus Matrix user interface provides the Master Remap Control Register (MATRIX_MRCR), that performs remap action for every master independently. The Bus Matrix user interface provides Master Remap Control Register (MATRIX_MRCR) that performs remap action for every master independently.
15.4
Special Bus Granting Mechanism

The Bus Matrix provides some speculative bus granting techniques in order to anticipate access requests from masters. This mechanism reduces latency at first access of a burst, or for a single transfer, as long as the slave is free from any other master access. It does not provide any benefit if the slave is continuously accessed by more than one master, since arbitration is pipelined and has no negative effect on the slave bandwidth or access latency. This bus granting mechanism sets a different default master for every slave. At the end of the current access, if no other request is pending, the slave remains connected to its associated default master. A slave can be associated with three kinds of default masters:
z z z
No default master Last access master Fixed default master
To change from one type of default master to another, the Bus Matrix user interface provides the Slave Configuration Registers, one for every slave, that set a default master for each slave. The Slave Configuration Register contains two fields: DEFMSTR_TYPE and FIXED_DEFMSTR. The 2-bit DEFMSTR_TYPE field selects the default master type (no default, last access master, fixed default master), whereas the 4-bit FIXED_DEFMSTR field selects a fixed default master provided that DEFMSTR_TYPE is set to fixed default master. Refer to Section 15.10.2 Bus Matrix Slave Configuration Registers on page 87.
15.5
No Default Master
After the end of the current access, if no other request is pending, the slave is disconnected from all masters. This configuration incurs one latency clock cycle for the first access of a burst after bus Idle. Arbitration without default master may be used for masters that perform significant bursts or several transfers with no Idle in between, or if the slave bus bandwidth is widely used by one or more masters. This configuration provides no benefit on access latency or bandwidth when reaching maximum slave bus throughput whatever the number of requesting masters.
15.6
Last Access Master

After the end of the current access, if no other request is pending, the slave remains connected to the last master that performed an access request. This allows the Bus Matrix to remove the one latency cycle for the last master that accessed the slave. Other non privileged masters still get one latency clock cycle if they want to access the same slave. This technique is useful for masters that mainly perform single accesses or short bursts with some Idle cycles in between. This configuration provides no benefit on access latency or bandwidth when reaching maximum slave bus throughput whatever is the number of requesting masters.

80
15.7
Fixed Default Master

After the end of the current access, if no other request is pending, the slave connects to its fixed default master. Unlike the last access master, the fixed default master does not change unless the user modifies it by software (FIXED_DEFMSTR field of the related MATRIX_SCFG). This allows the Bus Matrix arbiters to remove the one latency clock cycle for the fixed default master of the slave. All requests attempted by the fixed default master do not cause any arbitration latency, whereas other non-privileged masters will get one latency cycle. This technique is useful for a master that mainly performs single accesses or short bursts with Idle cycles in between. This configuration provides no benefit on access latency or bandwidth when reaching maximum slave bus throughput, regardless of the number of requesting masters.
15.8
Arbitration
The Bus Matrix provides an arbitration mechanism that reduces latency when conflict cases occur, i.e. when two or more masters try to access the same slave at the same time. One arbiter per AHB slave is provided, thus arbitrating each slave specifically. The Bus Matrix provides the user with the possibility of choosing between 2 arbitration types or mixing them for each slave: 1. 2. Round-robin Arbitration (default) Fixed Priority Arbitration
The resulting algorithm may be complemented by selecting a default master configuration for each slave. When re-arbitration must be done, specific conditions apply. See Section 15.8.1 Arbitration Scheduling on page 81.
15.8.1 Arbitration Scheduling

Each arbiter has the ability to arbitrate between two or more different master requests. In order to avoid burst breaking and also to provide the maximum throughput for slave interfaces, arbitration may only take place during the following cycles: 1. 2. 3. Idle Cycles: When a slave is not connected to any master or is connected to a master which is not currently accessing it. Single Cycles: When a slave is currently doing a single access. End of Burst Cycles: When the current cycle is the last cycle of a burst transfer. For defined length burst, predicted end of burst matches the size of the transfer but is managed differently for undefined length burst. See Undefined Length Burst Arbitration on page 81 Slot Cycle Limit: When the slot cycle counter has reached the limit value indicating that the current master access is too long and must be broken. See Slot Cycle Limit Arbitration on page 82
4.
15.8.1.1 Undefined Length Burst Arbitration In order to prevent long AHB burst lengths that can lock the access to the slave for an excessive period of time, the user can trigger the re-arbitration before the end of the incremental bursts. The re-arbitration period can be selected from the following Undefined Length Burst Type (ULBT) possibilities: 1. 2. 3. 4. 5. 6. Unlimited: no predetermined end of burst is generated. This value enables 1-kbyte burst lengths. 1-beat bursts: predetermined end of burst is generated at each single transfer during the INCR transfer. 4-beat bursts: predetermined end of burst is generated at the end of each 4-beat boundary during INCR transfer. 8-beat bursts: predetermined end of burst is generated at the end of each 8-beat boundary during INCR transfer. 16-beat bursts: predetermined end of burst is generated at the end of each 16-beat boundary during INCR transfer. 32-beat bursts: predetermined end of burst is generated at the end of each 32-beat boundary during INCR transfer.

81
7. 8.
64-beat bursts: predetermined end of burst is generated at the end of each 64-beat boundary during INCR transfer. 128-beat bursts: predetermined end of burst is generated at the end of each 128-beat boundary during INCR transfer.
The use of undefined length16-beat bursts, or less, is discouraged since this generally decreases significantly the overall bus bandwidth due to arbitration and slave latencies at each first access of a burst. If the master does not permanently and continuously request the same slave or has an intrinsically limited average throughput, the ULBT should be left at its default unlimited value, knowing that the AHB specification natively limits all word bursts to 256 beats and double-word bursts to 128 beats because of its 1 Kbyte address boundaries. Unless duly needed, the ULBT should be left at its default value of 0 for power saving. This selection can be done through the ULBT field of the Master Configuration Registers (MATRIX_MCFG). 15.8.1.2 Slot Cycle Limit Arbitration The Bus Matrix contains specific logic to break long accesses, such as very long bursts on a very slow slave (e.g., an external low speed memory). At each arbitration time, a counter is loaded with the value previously written in the SLOT_CYCLE field of the related Slave Configuration Register (MATRIX_SCFG) and decreased at each clock cycle. When the counter elapses, the arbiter has the ability to re-arbitrate at the end of the current AHB bus access cycle. Unless a master has a very tight access latency constraint, which could lead to data overflow or underflow due to a badly undersized internal FIFO with respect to its throughput, the Slot Cycle Limit should be disabled (SLOT_CYCLE = 0) or set to its default maximum value in order not to inefficiently break long bursts performed by some Atmel masters. In most cases, this feature is not needed and should be disabled for power saving. Warning: This feature cannot prevent any slave from locking its access indefinitely.
15.8.2 Arbitration Priority Scheme

The bus Matrix arbitration scheme is organized in priority pools. Round-robin priority is used in the highest and lowest priority pools, whereas fixed level priority is used between priority pools and in the intermediate priority pools. For each slave, each master is assigned to one of the slave priority pools through the priority registers for slaves (MxPR fields of MATRIX_PRAS and MATRIX_PRBS). When evaluating master requests, this programmed priority level always takes precedence. After reset, all the masters belong to the lowest priority pool (MxPR = 0) and are therefore granted bus access in a true round-robin order. The highest priority pool must be specifically reserved for masters requiring very low access latency. If more than one master belongs to this pool, they will be granted bus access in a biased round-robin manner which allows tight and deterministic maximum access latency from AHB bus requests. In the worst case, any currently occurring high-priority master request will be granted after the current bus master access has ended and other high priority pool master requests, if any, have been granted once each. The lowest priority pool shares the remaining bus bandwidth between AHB Masters. Intermediate priority pools allow fine priority tuning. Typically, a moderately latency-critical master or a bandwidth-only critical master will use such a priority level. The higher the priority level (MxPR value), the higher the master priority. All combinations of MxPR values are allowed for all masters and slaves. For example, some masters might be assigned the highest priority pool (round-robin), and remaining masters the lowest priority pool (round-robin), with no master for intermediate fix priority levels. If more than one master requests the slave bus, regardless of the respective masters priorities, no master will be granted the slave bus for two consecutive runs. A master can only get back-to-back grants so long as it is the only requesting master.

82
15.8.2.1 Fixed Priority Arbitration Fixed priority arbitration algorithm is the first and only arbitration algorithm applied between masters from distinct priority pools. It is also used in priority pools other than the highest and lowest priority pools (intermediate priority pools). Fixed priority arbitration allows the Bus Matrix arbiters to dispatch the requests from different masters to the same slave by using the fixed priority defined by the user in the MxPR field for each master in the Priority Registers, MATRIX_PRAS and MATRIX_PRBS. If two or more master requests are active at the same time, the master with the highest priority MxPR number is serviced first. In intermediate priority pools, if two or more master requests with the same priority are active at the same time, the master with the highest number is serviced first. 15.8.2.2 Round-Robin Arbitration This algorithm is only used in the highest and lowest priority pools. It allows the Bus Matrix arbiters to properly dispatch requests from different masters to the same slave. If two or more master requests are active at the same time in the priority pool, they are serviced in a round-robin increasing master number order.
15.9
Write Protect Registers

To prevent any single software error that may corrupt the Bus Matrix behavior, the entire Bus Matrix address space can be write-protected by setting the WPEN bit in the Bus Matrix Write Protect Mode Register (MATRIX_WPMR). If WPEN is at one and a write access in the Bus Matrix address space is detected, then the WPVS flag in the Bus Matrix Write Protect Status Register (MATRIX_WPSR) is set and the field WPVSRC indicates in which register the write access has been attempted. The WPVS flag is reset by writing the Bus Matrix Write Protect Mode Register (MATRIX_WPMR) with the appropriate access key WPKEY. The protected registers are: Bus Matrix Master Configuration Registers Bus Matrix Slave Configuration Registers Bus Matrix Priority Registers A For Slaves Bus Matrix Priority Registers B For Slaves Bus Matrix Master Remap Control Register

83
15.10 Bus Matrix (MATRIX) User Interface

Table 15-3. Register Mapping Offset 0x0000 0x0004 0x0008 0x000C 0x0010 0x0014 0x0018 0x001C 0x0020 0x0024 0x0028 0x002C 0x0030 0x0034 0x0038 0x003C 0x0040 0x0044 0x0048 0x004C 0x0050 0x0054 0x0058 0x005C 0x0060 0x0064 0x0068 0x006C 0x0070 0x0074 0x0078 0x007C 0x0080 0x0084 0x0088 Register Master Configuration Register 0 Master Configuration Register 1 Master Configuration Register 2 Master Configuration Register 3 Master Configuration Register 4 Master Configuration Register 5 Master Configuration Register 6 Master Configuration Register 7 Master Configuration Register 8 Master Configuration Register 9 Master Configuration Register 10 Master Configuration Register 11 Master Configuration Register 12 Master Configuration Register 13 Master Configuration Register 14 Master Configuration Register 15 Slave Configuration Register 0 Slave Configuration Register 1 Slave Configuration Register 2 Slave Configuration Register 3 Slave Configuration Register 4 Slave Configuration Register 5 Slave Configuration Register 6 Slave Configuration Register 7 Slave Configuration Register 8 Slave Configuration Register 9 Slave Configuration Register 10 Slave Configuration Register 11 Slave Configuration Register 12 Slave Configuration Register 13 Slave Configuration Register 14 Slave Configuration Register 15 Priority Register A for Slave 0 Priority Register B for Slave 0 Priority Register A for Slave 1 Name MATRIX_MCFG0 MATRIX_MCFG1 MATRIX_MCFG2 MATRIX_MCFG3 MATRIX_MCFG4 MATRIX_MCFG5 MATRIX_MCFG6 MATRIX_MCFG7 MATRIX_MCFG8 MATRIX_MCFG9 MATRIX_MCFG10 MATRIX_MCFG11 MATRIX_MCFG12 MATRIX_MCFG13 MATRIX_MCFG14 MATRIX_MCFG15 MATRIX_SCFG0 MATRIX_SCFG1 MATRIX_SCFG2 MATRIX_SCFG3 MATRIX_SCFG4 MATRIX_SCFG5 MATRIX_SCFG6 MATRIX_SCFG7 MATRIX_SCFG8 MATRIX_SCFG9 MATRIX_SCFG10 MATRIX_SCFG11 MATRIX_SCFG12 MATRIX_SCFG13 MATRIX_SCFG14 MATRIX_SCFG15 MATRIX_PRAS0 MATRIX_PRBS0 MATRIX_PRAS1 Access Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Reset 0x00000001 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x000001FF 0x000001FF 0x000001FF 0x000001FF 0x000001FF 0x000001FF 0x000001FF 0x000001FF 0x000001FF 0x000001FF 0x000001FF 0x000001FF 0x000001FF 0x000001FF 0x000001FF 0x000001FF 0x33333333(1) 0x33333333(1) 0x33333333(1)

84
Table 15-3. Register Mapping (Continued) Offset 0x008C 0x0090 0x0094 0x0098 0x009C 0x00A0 0x00A4 0x00A8 0x00AC 0x00B0 0x00B4 0x00B8 0x00BC 0x00C0 0x00C4 0x00C8 0x00CC 0x00D0 0x00D4 0x00D8 0x00DC 0x00E0 0x00E4 0x00E8 0x00EC 0x00F0 0x00F4 0x00F8 0x00FC 0x0100 0x0104 - 0x010C 0x01A0 - 0x01E0 0x01E4 0x01E8 Register Priority Register B for Slave 1 Priority Register A for Slave 2 Priority Register B for Slave 2 Priority Register A for Slave 3 Priority Register B for Slave 3 Priority Register A for Slave 4 Priority Register B for Slave 4 Priority Register A for Slave 5 Priority Register B for Slave 5 Priority Register A for Slave 6 Priority Register B for Slave 6 Priority Register A for Slave 7 Priority Register B for Slave 7 Priority Register A for Slave 8 Priority Register B for Slave 8 Priority Register A for Slave 9 Priority Register B for Slave 9 Priority Register A for Slave 10 Priority Register B for Slave 10 Priority Register A for Slave 11 Priority Register B for Slave 11 Priority Register A for Slave 12 Priority Register B for Slave 12 Priority Register A for Slave 13 Priority Register B for Slave 13 Priority Register A for Slave 14 Priority Register B for Slave 14 Priority Register A for Slave 15 Priority Register B for Slave 15 Master Remap Control Register Reserved Reserved Write Protect Mode Register Write Protect Status Register Name MATRIX_PRBS1 MATRIX_PRAS2 MATRIX_PRBS2 MATRIX_PRAS3 MATRIX_PRBS3 MATRIX_PRAS4 MATRIX_PRBS4 MATRIX_PRAS5 MATRIX_PRBS5 MATRIX_PRAS6 MATRIX_PRBS6 MATRIX_PRAS7 MATRIX_PRBS7 MATRIX_PRAS8 MATRIX_PRBS8 MATRIX_PRAS9 MATRIX_PRBS9 MATRIX_PRAS10 MATRIX_PRBS10 MATRIX_PRAS11 MATRIX_PRBS11 MATRIX_PRAS12 MATRIX_PRBS12 MATRIX_PRAS13 MATRIX_PRBS13 MATRIX_PRAS14 MATRIX_PRBS14 MATRIX_PRAS15 MATRIX_PRBS15 MATRIX_MRCR MATRIX_WPMR MATRIX_WPSR Access Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-only Reset 0x33333333(1) 0x33333333(1) 0x33333333(1) 0x33333333(1) 0x33333333(1) 0x33333333(1) 0x33333333(1) 0x33333333(1) 0x33333333(1) 0x33333333(1) 0x33333333(1) 0x33333333(1) 0x33333333(1) 0x33333333(1) 0x33333333(1) 0x33333333(1) 0x33333333(1) 0x33333333(1) 0x33333333(1) 0x33333333(1) 0x33333333(1) 0x33333333(1) 0x33333333(1) 0x33333333(1) 0x33333333(1) 0x33333333(1) 0x33333333(1) 0x33333333(1) 0x33333333(1) 0x00000000 0x00000000 0x00000000
Notes: 1. Values in the Bus Matrix Priority Registers are product dependent.

85
15.10.1 Bus Matrix Master Configuration Registers

Name: Address: Access:
31 23 15 7
MATRIX_MCFG0...MATRIX_MCFG15 0xFFFFEC00 Read-write

30 22 14 6 29 21 13 5 28 20 12 4 27 19 11 3 26 18 10 2 25 17 9 1 ULBT 24 16 8 0
This register can only be written if the WPEN bit is cleared in the Write Protect Mode Register .
ULBT: Undefined Length Burst Type

0: Unlimited Length Burst No predicted end of burst is generated, therefore INCR bursts coming from this master can only be broken if the Slave Slot Cycle Limit is reached. If the Slot Cycle Limit is not reached, the burst is normally completed by the master, at the latest, on the next AHB 1 Kbyte address boundary, allowing up to 256-beat word bursts or 128-beat double-word bursts. This value should not be used in the very particular case of a master capable of performing back-to-back undefined length bursts on a single slave, since this could indefinitely freeze the slave arbitration and thus prevent another master from accessing this slave. 1: Single Access The undefined length burst is treated as a succession of single accesses, allowing re-arbitration at each beat of the INCR burst or bursts sequence. 2: 4-beat Burst The undefined length burst or bursts sequence is split into 4-beat bursts or less, allowing re-arbitration every 4 beats. 3: 8-beat Burst The undefined length burst or bursts sequence is split into 8-beat bursts or less, allowing re-arbitration every 8 beats. 4: 16-beat Burst The undefined length burst or bursts sequence is split into 16-beat bursts or less, allowing re-arbitration every 16 beats. 5: 32-beat Burst The undefined length burst or bursts sequence is split into 32-beat bursts or less, allowing re-arbitration every 32 beats. 6: 64-beat Burst The undefined length burst or bursts sequence is split into 64-beat bursts or less, allowing re-arbitration every 64 beats. 7: 128-beat Burst The undefined length burst or bursts sequence is split into 128-beat bursts or less, allowing re-arbitration every 128 beats. Unless duly needed, the ULBT should be left at its default 0 value for power saving.

86
15.10.2 Bus Matrix Slave Configuration Registers

31 23 15 7
MATRIX_SCFG0...MATRIX_SCFG15 0xFFFFEC40 Read-write

30 22 14 6 13 5 29 21 28 20 27 19 26 18 25 17 24 16
FIXED_DEFMSTR 12 4 SLOT_CYCLE 11 3 10 2
DEFMSTR_TYPE 9 1 8 SLOT_CYCLE 0
SLOT_CYCLE: Maximum Bus Grant Duration for Masters

When SLOT_CYCLE AHB clock cycles have elapsed since the last arbitration, a new arbitration takes place to let another master access this slave. If another master is requesting the slave bus, then the current master burst is broken. If SLOT_CYCLE = 0, the Slot Cycle Limit feature is disabled and bursts always complete unless broken according to the ULBT. This limit has been placed in order to enforce arbitration so as to meet potential latency constraints of masters waiting for slave access. This limit must not be too small. Unreasonably small values break every burst and the Bus Matrix arbitrates without performing any data transfer. The default maximum value is usually an optimal conservative choice. In most cases, this feature is not needed and should be disabled for power saving. See Slot Cycle Limit Arbitration on page 82 for details.
DEFMSTR_TYPE: Default Master Type

0: No Default Master At the end of the current slave access, if no other master request is pending, the slave is disconnected from all masters. This results in a one clock cycle latency for the first access of a burst transfer or for a single access. 1: Last Default Master At the end of the current slave access, if no other master request is pending, the slave stays connected to the last master having accessed it. This results in not having one clock cycle latency when the last master tries to access the slave again. 2: Fixed Default Master At the end of the current slave access, if no other master request is pending, the slave connects to the fixed master the number that has been written in the FIXED_DEFMSTR field. This results in not having one clock cycle latency when the fixed master tries to access the slave again.
FIXED_DEFMSTR: Fixed Default Master

This is the number of the Default Master for this slave. Only used if DEFMSTR_TYPE is 2. Specifying the number of a master which is not connected to the selected slave is equivalent to setting DEFMSTR_TYPE to 0.

87
15.10.3 Bus Matrix Priority Registers A For Slaves

Name: Address: MATRIX_PRAS0...MATRIX_PRAS15 0xFFFFEC80 [0], 0xFFFFEC88 [1], 0xFFFFEC90 [2], 0xFFFFEC98 [3], 0xFFFFECA0 [4], 0xFFFFECA8 [5], 0xFFFFECB0 [6], 0xFFFFECB8 [7], 0xFFFFECC0 [8], 0xFFFFECC8 [9], 0xFFFFECD0 [10], 0xFFFFECD8 [11], 0xFFFFECE0 [12], 0xFFFFECE8 [13], 0xFFFFECF0 [14], 0xFFFFECF8 [15] Access:
31 23 15 7
Read-write
30 22 14 6 5 M1PR 13 M3PR 4 21 M5PR 12 29 M7PR 20 28 27 19 11 3 26 18 10 2 1 M0PR 9 M2PR 0 17 M4PR 8 25 M6PR 16 24
This register can only be written if the WPE bit is cleared in the Write Protect Mode Register .
MxPR: Master x Priority

Fixed priority of Master x for accessing the selected slave. The higher the number, the higher the priority. All the masters programmed with the same MxPR value for the slave make up a priority pool. Round-robin arbitration is used in the lowest (MxPR = 0) and highest (MxPR = 3) priority pools. Fixed priority is used in intermediate priority pools (MxPR = 1) and (MxPR = 2). See Arbitration Priority Scheme on page 82 for details.

88
15.10.4 Bus Matrix Priority Registers B For Slaves

Name: Address: MATRIX_PRBS0...MATRIX_PRBS15 0xFFFFEC84 [0], 0xFFFFEC8C [1], 0xFFFFEC94 [2], 0xFFFFEC9C [3], 0xFFFFECA4 [4], 0xFFFFECAC [5], 0xFFFFECB4 [6], 0xFFFFECBC [7], 0xFFFFECC4 [8], 0xFFFFECCC [9], 0xFFFFECD4 [10], 0xFFFFECDC [11], 0xFFFFECE4 [12], 0xFFFFECEC [13], 0xFFFFECF4 [14], 0xFFFFECFC [15] Access:
31 23 15 7
Read-write
30 22 14 6 5 M9PR 13 M11PR 4 21 M13PR 12 29 M15PR 20 28 27 19 11 3 26 18 10 2 1 M8PR 9 M10PR 0 17 M12PR 8 25 M14PR 16 24
MxPR: Master x Priority

Fixed priority of Master x for accessing the selected slave. The higher the number, the higher the priority. All the masters programmed with the same MxPR value for the slave make up a priority pool. Round-robin arbitration is used in the lowest (MxPR = 0) and highest (MxPR = 3) priority pools. Fixed priority is used in intermediate priority pools (MxPR = 1) and (MxPR = 2). See Arbitration Priority Scheme on page 82 for details.

89
15.10.5 Bus Matrix Master Remap Control Register

31 23 15 RCB15 7 RCB7
MATRIX_MRCR 0xFFFFED00 Read-write

30 22 14 RCB14 6 RCB6 29 21 13 RCB13 5 RCB5 28 20 12 RCB12 4 RCB4 27 19 11 RCB11 3 RCB3 26 18 10 RCB10 2 RCB2 25 17 9 RCB9 1 RCB1 24 16 8 RCB8 0 RCB0
RCB: Remap Command Bit for Master x

0: Disable remapped address decoding for the selected Master 1: Enable remapped address decoding for the selected Master

90
15.10.6 Write Protect Mode Register

31
MATRIX_WPMR 0xFFFFEDE4 Read-write

30 29 28 WPKEY 23 22 21 20 WPKEY 15 14 13 12 WPKEY 7 6 5 4 3 2 1 0 WPEN 11 10 9 8 19 18 17 16 27 26 25 24
For more details on MATRIX_WPMR, refer to Section 15.9 Write Protect Registers on page 83. The protected registers are: Bus Matrix Master Configuration Registers Bus Matrix Slave Configuration Registers Bus Matrix Priority Registers A For Slaves Bus Matrix Priority Registers B For Slaves Bus Matrix Master Remap Control Register
WPEN: Write Protect Enable

0: Disables the Write Protect if WPKEY corresponds to 0x4D4154 (MAT in ASCII). 1: Enables the Write Protect if WPKEY corresponds to 0x4D4154 (MAT in ASCII). Protects the entire Bus Matrix address space from address offset 0x000 to 0x1FC.
WPKEY: Write Protect KEY (Write-only)

Should be written at value 0x4D4154 (MAT in ASCII). Writing any other value in this field aborts the write operation of the WPEN bit. Always reads as 0.

91
15.10.7 Write Protect Status Register

31 23
MATRIX_WPSR 0xFFFFEDE8 Read-only

30 22 29 21 28 20 WPVSRC 15 14 13 12 WPVSRC 7 6 5 4 3 2 1 0 WPVS 11 10 9 8 27 19 26 18 25 17 24 16
For more details on MATRIX_WPSR, refer to Section 15.9 Write Protect Registers on page 83.
WPVS: Write Protect Violation Status

0: No Write Protect Violation has occurred since the last write of the MATRIX_WPMR. 1: At least one Write Protect Violation has occurred since the last write of the MATRIX_WPMR.
WPVSRC: Write Protect Violation Source

When WPVS is active, this field indicates the register address offset in which a write access has been attempted. Otherwise it reads as 0.

92
16.
16.1
Special Function Registers (SFR)

Description
Special Function Registers (SFR) manage specific aspects of the integrated memory, bridge implementations, processor and other functionality not controlled elsewhere.
16.2
z
32-bit Special Function Registers control specific behavior of the product
16.3
Special Function Registers (SFR) User Interface
Table 16-1. Register Mapping Offset 0x00 -0x04 0x08-0x0C 0x10 0x14 0x18 0x1C 0x28 0x2C 0x30 0x40 0x44 0x48-0x3FFC Register Reserved Reserved OHCI Interrupt Configuration Register OHCI Interrupt Status Register Reserved Reserved Security Configuration Register Reserved UTMI Clock Trimming Register EBI Configuration Register Reserved Reserved Name SFR_OHCIICR SFR_OHCIISR SFR_SECURE SFR_UTMICKTRIM SFR_EBICFG Access Read-write Read-only Read-write Read-write Read-write Reset 0x0 0x0 0x00010000

93
16.3.1 OHCI Interrupt Configuration Register

31
SFR_OHCIICR 0xF0038010 Read-write

30 29 28 27 26 25 24
23
22
21
20
19
18
17
16
UDPPUDIS
15
14
13
12
11
10
APPSTART
ARIE
RES2
RES1
RES0
RESx: USB PORTx RESET

0: Reset USB PORT. 1: Usable USB PORT.
ARIE: OHCI Asynchronous Resume Interrupt Enable

0: Interrupt disabled. 1: Interrupt enabled.
APPSTART: Reserved
0: Must write 0.
UDPPUDIS: USB DEVICE PULL-UP DISABLE

0: USB device Pull-up connection is enabled. 1: USB device Pull-up connection is disabled.

94
16.3.2 OHCI Interrupt Status Register

31
SFR_OHCIISR 0xF0038014 Read-write

30 29 28 27 26 25 24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
RIS2
RIS1
RIS0
RISx: OHCI Resume Interrupt Status Port x

0: OHCI Port resume not detected. 1: OHCI Port resume detected.

95
16.3.3 Security Configuration Register

31
SFR_SECURE 0xF0038028 Read-write

30 29 28 27 26 25 24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
FUSE
0
ROM
ROM: Disable Access to ROM Code

This bit is writable once only. When the ROM is secured, only reset signal can clear this bit. 0: ROM is enabled. 1: ROM is disabled.
FUSE: Disable Access to Fuse Controller

This bit is writable once only. When the Fuse Controller is secured, only reset signal can clear this bit. 0: Fuse Controller is enabled. 1: Fuse Controller is disabled.

96
16.3.4 UTMI Clock Trimming Register

31
SFR_UTMICKTRIM 0xF0038030 Read-write

30 29 28 27 26 25 24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
FREQ
FREQ: UTMI Reference Clock Frequency

Value 0 1 2 3 Name 12 16 24 48 Description 12 MHz reference clock 16 MHz reference clock 24 MHz reference clock 48 MHz reference clock

97
16.3.5 EBI Configuration Register

31
SFR_EBICFG
0xF0038040 Read-write
30 29 28 27 26 25 24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
BMS
8
SCH1
4 3
PULL1
2 1
DRIVE1
0
SCH0
PULL0
DRIVE0
This register controls EBI pins which are not multiplexed with PIO controller lines.
DRIVE0, PULL0, SCH0 control EBI Data pins when applicable. DRIVE1, PULL1, SCH1 control other EBI pins when applicable. DRIVEx: EBI Pins Drive Level
Drive level should be programmed depending on target frequency and board characteristics. Refer to pad characteristics to set correct drive level.
Value 0 1 2 3 Name LOW RESERVED MEDIUM HIGH Description Low drive level Low drive level Medium drive level High drive level
PULLx: EBI Pins Pull Value

Value 0 1 2 3 Name UP NONE Reserved DOWN Description Pull-up No Pull No Change (forbidden write value) Pull-down
SCHx: EBI Pins Schmitt Trigger

0: Schmitt Trigger off. 1: Schmitt Trigger on.
BMS: BMS Sampled Value (Read Only)

This bit examines whether boot is on EBI or ROM. 0 (ROM): Boot on ROM. 1 (EBI): Boot on EBI.

98
17.
17.1
Advanced Interrupt Controller (AIC)

Description
The Advanced Interrupt Controller (AIC) is an 8-level priority, individually maskable, vectored interrupt controller, providing handling of up to hundred and twenty-eight interrupt sources. It is designed to substantially reduce the software and real-time overhead in handling internal and external interrupts. The AIC drives the nFIQ (fast interrupt request) and the nIRQ (standard interrupt request) inputs of an ARM processor. Inputs of the AIC are either internal peripheral interrupts or external interrupts coming from the product's pins. The 8-level Priority Controller allows the user to define the priority for each interrupt source, thus permitting higher priority interrupts to be serviced even if a lower priority interrupt is being treated. Internal interrupt sources can be programmed to be level sensitive or edge triggered. External interrupt sources can be programmed to be positive-edge or negative-edge triggered or high-level or low-level sensitive. The fast forcing feature redirects any internal or external interrupt source to provide a fast interrupt rather than a normal interrupt.
17.2
z z
Controls the Interrupt Lines (nIRQ and nFIQ) of an ARM Processor 128 Individually Maskable and Vectored Interrupt Sources
z z z z z
Source 0 is Reserved for the Fast Interrupt Input (FIQ) Source 1 is Reserved for System Peripheral Interrupts Source 2 to Source 127, Control up to 126 Embedded Peripheral Interrupts or External Interrupts Programmable Edge-triggered or Level-sensitive Internal Sources Programmable Positive/Negative Edge-triggered or High/Low Level-sensitive External Sources Drives the Normal Interrupt of the Processor Handles Priority of the Interrupt Sources 1 to 127 Higher Priority Interrupts Can Be Served During Service of Lower Priority Interrupt Optimizes Interrupt Service Routine Branch and Execution One 32-bit Vector Register for all Interrupt Sources Interrupt Vector Register Reads the Corresponding Current Interrupt Vector Easy Debugging by Preventing Automatic Operations when Protect Models are Enabled Permits Redirecting any Normal Interrupt Source to the Fast Interrupt of the Processor Provides Processor Synchronization on Events Without Triggering an Interrupt
8-level Priority Controller

z z z
Vectoring
z z z
Protect Mode
z
Fast Forcing
z
General Interrupt Mask

z
Write Protected Registers

99
17.3
Block Diagram
Figure 17-1. Block Diagram
FIQ IRQ0-IRQn AIC ARM Processor Up to 128 Sources nFIQ nIRQ
Embedded Peripher alEE Embedded

Peripher al Embedded
Peripheral
APB
17.4
Application Block Diagram

Figure 17-2. Description of the Application Block
OS-based Applications Standalone Applications OS Drivers RTOS Drivers Hard Real Time Tasks General OS Interrupt Handler Advanced Interrupt Controller Embedded Peripherals External Peripherals (External Interrupts)
17.5
AIC Detailed Block Diagram

Figure 17-3. AIC Detailed Block Diagram
Advanced Interrupt Controller FIQ PIO Controller External Source Input Stage Fast Interrupt Controller ARM Processor nFIQ
nIRQ IRQ0-IRQn PIOIRQ Internal Source Input Stage Fast Forcing Interrupt Priority Controller Processor Clock Power Management Controller User Interface Wake Up
Embedded Peripherals
APB

100
17.6
I/O Line Description
Table 17-1. I/O Line Description Pin Name FIQ IRQ0 - IRQn Pin Description Fast Interrupt Interrupt 0 - Interrupt n Type Input Input
17.7
17.7.1 I/O Lines

The interrupt signals FIQ and IRQ0 to IRQn are normally multiplexed through the PIO controllers. Depending on the features of the PIO controller used in the product, the pins must be programmed in accordance with their assigned interrupt function. This is not applicable when the PIO controller used in the product is transparent on the input path.
Table 17-2. I/O Lines Instance AIC AIC Signal FIQ IRQ I/O Line PC31 PE31 Peripheral A A
17.7.2 Power Management

The Advanced Interrupt Controller is continuously clocked. The Power Management Controller has no effect on the Advanced Interrupt Controller behavior. The assertion of the Advanced Interrupt Controller outputs, either nIRQ or nFIQ, wakes up the ARM processor while it is in Idle Mode. The General Interrupt Mask feature enables the AIC to wake up the processor without asserting the interrupt line of the processor, thus providing synchronization of the processor on an event.
17.7.3 Interrupt Sources

The Interrupt Source 0 is always located at FIQ. If the product does not feature an FIQ pin, the Interrupt Source 0 cannot be used. The Interrupt Source 1 is always located at System Interrupt. This is the result of the OR-wiring of the system peripheral interrupt lines. When a system interrupt occurs, the service routine must first distinguish the cause of the interrupt. This is performed by reading successively the status registers of the above mentioned system peripherals. The interrupt sources 2 to 127 can either be connected to the interrupt outputs of an embedded user peripheral or to external interrupt lines. The external interrupt lines can be connected directly, or through the PIO Controller. The PIO Controllers are considered as user peripherals in the scope of interrupt handling. Accordingly, the PIO Controller interrupt lines are connected to the Interrupt Sources 2 to 127. The peripheral identification defined at the product level corresponds to the interrupt source number (as well as the bit number controlling the clock of the peripheral). Consequently, to simplify the description of the functional operations and the user interface, the interrupt sources are named FIQ, SYS, and PID2 to PID127.

101
17.8
17.8.1 Interrupt Source Control

17.8.1.1 Interrupt Source Mode The Advanced Interrupt Controller independently programs each interrupt source. The SRCTYPE field of the AIC_SMR (Source Mode Register) selects the interrupt condition of the interrupt source selected by the INTSEL field of the AIC Source Select Register. Note: Configuration registers such as AIC_SMR, AIC_SSR, return the values corresponding to the interrupt source selected by INTSEL. The internal interrupt sources wired on the interrupt outputs of the embedded peripherals can be programmed either in level-sensitive mode or in edge-triggered mode. The active level of the internal interrupts is not important for the user. The external interrupt sources can be programmed either in high level-sensitive or low level-sensitive modes, or in positive edge-triggered or negative edge-triggered modes. 17.8.1.2 Interrupt Source Enabling Each interrupt source, including the FIQ in source 0, can be enabled or disabled by using the command registers; AIC Interrupt Enable Command Register on page 125 and AIC Interrupt Disable Command Register on page 125. The interrupt mask of the selected interrupt source can be read in the AIC_IMR register. A disabled interrupt does not affect servicing of other interrupts. 17.8.1.3 Interrupt Clearing and Setting All interrupt sources programmed to be edge-triggered (including the FIQ in source 0) can be individually set or cleared by writing respectively the AIC_ISCR and AIC_ICCR registers. Clearing or setting interrupt sources programmed in levelsensitive mode has no effect. The clear operation is perfunctory, as the software must perform an action to reinitialize the memorization circuitry activated when the source is programmed in edge-triggered mode. However, the set operation is available for auto-test or software debug purposes. It can also be used to execute an AIC-implementation of a software interrupt. The AIC features an automatic clear of the current interrupt when the AIC_IVR (Interrupt Vector Register) is read. Only the interrupt source being detected by the AIC as the current interrupt is affected by this operation. (See Priority Controller on page 105.) The automatic clear reduces the operations required by the interrupt service routine entry code to reading the AIC_IVR. Note that the automatic interrupt clear is disabled if the interrupt source has the Fast Forcing feature enabled as it is considered uniquely as a FIQ source. (For further details, See Fast Forcing on page 109.) The automatic clear of the interrupt source 0 is performed when AIC_FVR is read. 17.8.1.4 Interrupt Status AIC_IPR registers represent the state of the interrupt lines, whether they are masked or not. The AIC_IMR register permits to define the mask of the interrupt lines. The AIC_ISR register reads the number of the current interrupt (see Priority Controller on page 105) and the register AIC_CISR gives an image of the signals nIRQ and nFIQ driven on the processor. Each status referred to above can be used to optimize the interrupt handling of the systems.

102
17.8.1.5 Internal Interrupt Source Input Stage

Figure 17-4. Internal Interrupt Source Input Stage
AIC_SMRI (SRCTYPE) Source i Level/ Edge AIC_IPR AIC_IMR Fast Interrupt Controller or Priority Controller AIC_IECR
Edge
Detector Set Clear AIC_ISCR AIC_ICCR FF
AIC_IDCR
17.8.1.6 External Interrupt Source Input Stage

Figure 17-5. External Interrupt Source Input Stage
High/Low AIC_SMRi SRCTYPE Level/ Edge Source i AIC_IPR AIC_IMR Fast Interrupt Controller or Priority Controller Pos./Neg. Edge Detector Set AIC_ISCR AIC_ICCR Clear AIC_IDCR AIC_IECR
FF

103
17.8.2 Interrupt Latencies

Global interrupt latencies depend on several parameters, including:
z z z z
The time the software masks the interrupts. Occurrence, either at the processor level or at the AIC level. The execution time of the instruction in progress when the interrupt occurs. The treatment of higher priority interrupts and the resynchronization of the hardware signals.
This section addresses only the hardware resynchronizations. It gives details of the latency times between the event on an external interrupt leading in a valid interrupt (edge or level) or the assertion of an internal interrupt source and the assertion of the nIRQ or nFIQ line on the processor. The resynchronization time depends on the programming of the interrupt source and on its type (internal or external). For the standard interrupt, resynchronization times are given assuming there is no higher priority in progress. The PIO Controller multiplexing has no effect on the interrupt latencies of the external interrupt sources. 17.8.2.1 External Interrupt Edge Triggered Source
Figure 17-6. External Interrupt Edge Triggered Source
MCK IRQ or FIQ (Positive Edge) IRQ or FIQ (Negative Edge)
nIRQ Maximum IRQ Latency = 4 Cycles
nFIQ Maximum FIQ Latency = 4 Cycles
17.8.2.2 External Interrupt Level Sensitive Source

Figure 17-7. External Interrupt Level Sensitive Source
MCK IRQ or FIQ (High Level) IRQ or FIQ (Low Level) nIRQ Maximum IRQ Latency = 3 Cycles
nFIQ Maximum FIQ Latency = 3 cycles

104
17.8.2.3 Internal Interrupt Edge Triggered Source

Figure 17-8. Internal Interrupt Edge Triggered Source
MCK IRQ or FIQ (High Level) IRQ or FIQ (Low Level) nIRQ Maximum IRQ Latency = 3 Cycles
nFIQ Maximum FIQ Latency = 3 cycles
17.8.2.4 Internal Interrupt Level Sensitive Source

Figure 17-9. Internal Interrupt Level Sensitive Source
MCK
nIRQ
Maximum IRQ Latency = 3.5 Cycles Peripheral Interrupt Becomes Active
17.8.3 Normal Interrupt

17.8.3.1 Priority Controller An 8-level priority controller drives the nIRQ line of the processor, depending on the interrupt conditions occurring on the interrupt sources 1 to 127 (except for those programmed in Fast Forcing). Each interrupt source has a programmable priority level of 7 to 0, which is user-definable by writing the PRIOR field of the AIC_SMR (Source Mode Register). Level 7 is the highest priority and level 0 the lowest. As soon as an interrupt condition occurs, as defined by the SRCTYPE field of the AIC_SMR (Source Mode Register), the nIRQ line is asserted. As a new interrupt condition might have happened on other interrupt sources since the nIRQ has been asserted, the priority controller determines the current interrupt at the time the AIC_IVR (Interrupt Vector Register) is read. The read of AIC_IVR is the entry point of the interrupt handling which allows the AIC to consider that the interrupt has been taken into account by the software. The current priority level is defined as the priority level of the current interrupt. If several interrupt sources of equal priority are pending and enabled when the AIC_IVR is read, the interrupt with the lowest interrupt source number is serviced first. The nIRQ line can be asserted only if an interrupt condition occurs on an interrupt source with a higher priority. If an interrupt condition happens (or is pending) during the interrupt treatment in progress, it is delayed until the software indicates to the AIC the end of the current service by writing the AIC_EOICR (End of Interrupt Command Register). The write of AIC_EOICR is the exit point of the interrupt handling.

105
17.8.3.2 Interrupt Nesting The priority controller utilizes interrupt nesting in order for the high priority interrupt to be handled during the service of lower priority interrupts. This requires the interrupt service routines of the lower interrupts to re-enable the interrupt at the processor level. When an interrupt of a higher priority happens during an already occurring interrupt service routine, the nIRQ line is reasserted. If the interrupt is enabled at the core level, the current execution is interrupted and the new interrupt service routine should read the AIC_IVR. At this time, the current interrupt number and its priority level are pushed into an embedded hardware stack, so that they are saved and restored when the higher priority interrupt servicing is finished and the AIC_EOICR is written. The AIC is equipped with an 8-level wide hardware stack in order to support up to eight interrupt nestings pursuant to having eight priority levels. 17.8.3.3 Interrupt Vectoring The interrupt handler address corresponding to the interrupt source selected by the INTSEL field can be stored in the registers AIC_SVR (Source Vector Register). When the processor reads AIC_IVR (Interrupt Vector Register), the value written into AIC_SVR corresponding to the current interrupt is returned. This feature offers a way to branch in one single instruction to the handler corresponding to the current interrupt, as AIC_IVR is mapped at the absolute address 0xFFFF F100 and thus accessible from the ARM interrupt vector at address 0x0000 0018 through the following instruction: LDR PC,[PC,# -&F20] When the processor executes this instruction, it loads the read value in AIC_IVR in its program counter, thus branching the execution on the correct interrupt handler. This feature is often not used when the application is based on an operating system (either real time or not). Operating systems often have a single entry point for all the interrupts and the first task performed is to discern the source of the interrupt. However, it is strongly recommended to port the operating system on SAMA5D3 Series devices by supporting the interrupt vectoring. This can be performed by defining the AIC_SVR of the interrupt sources to be handled by the operating system at the address of its interrupt handler. When doing so, the interrupt vectoring permits a critical interrupt to transfer the execution on a specific very fast handler and not onto the operating systems general interrupt handler. This facilitates the support of hard real-time tasks (input/outputs of voice/audio buffers and software peripheral handling) to be handled efficiently and independently of the application running under an operating system. 17.8.3.4 Interrupt Handlers This section gives an overview of the fast interrupt handling sequence when using the AIC. It is assumed that the programmer understands the architecture of the ARM processor, and especially the processor interrupt modes and the associated status bits. It is assumed that: 1. 2. The Advanced Interrupt Controller has been programmed, AIC_SVR registers are loaded with corresponding interrupt service routine addresses and interrupts are enabled. The instruction at the ARM interrupt exception vector address is required to work with the vectoring LDR PC, [PC, # -&F20] When nIRQ is asserted, if the bit I of CPSR is 0, the sequence is as follows: 1. The CPSR is stored in SPSR_irq, the current value of the Program Counter is loaded in the Interrupt link register (R14_irq) and the Program Counter (R15) is loaded with 0x18. In the following cycle during fetch at address 0x1C, the ARM core adjusts R14_irq, decrementing it by four. The ARM core enters Interrupt mode, if it has not already done so.
2.

106
3.
When the instruction loaded at address 0x18 is executed, the program counter is loaded with the value read in AIC_IVR. Reading the AIC_IVR has the following effects:
z z z z z
Sets the current interrupt to be the pending and enabled interrupt with the highest priority. The current level is the priority level of the current interrupt. De-asserts the nIRQ line on the processor. Even if vectoring is not used, AIC_IVR must be read in order to de-assert nIRQ. Automatically clears the interrupt, if it has been programmed to be edge-triggered. Pushes the current level and the current interrupt number on to the stack. Returns the value written in the AIC_SVR corresponding to the current interrupt.
4.
The previous step has the effect of branching to the corresponding interrupt service routine. This should start by saving the link register (R14_irq) and SPSR_IRQ. The link register must be decremented by four when it is saved if it is to be restored directly into the program counter at the end of the interrupt. For example, the instruction SUB PC, LR, #4 may be used. Further interrupts can then be unmasked by clearing the I bit in CPSR, allowing re-assertion of the nIRQ to be taken into account by the core. This can happen if an interrupt with a higher priority than the current interrupt occurs. The interrupt handler can then proceed as required, saving the registers that will be used and restoring them at the end. During this phase, an interrupt of higher priority than the current level will restart the sequence from step 1. If the interrupt is programmed to be level sensitive, the source of the interrupt must be cleared during this phase. The I bit in CPSR must be set in order to mask interrupts before exiting to ensure that the interrupt is completed in an orderly manner. The End of Interrupt Command Register (AIC_EOICR) must be written in order to indicate to the AIC that the current interrupt is finished. This causes the current level to be popped from the stack, restoring the previous current level if one exists on the stack. If another interrupt is pending, with lower or equal priority than the old current level but with higher priority than the new current level, the nIRQ line is re-asserted, but the interrupt sequence does not immediately start because the I bit is set in the core. SPSR_irq is restored. Finally, the saved value of the link register is restored directly into the PC. This has the effect of returning from the interrupt to whatever was being executed before, and of loading the CPSR with the stored SPSR, masking or unmasking the interrupts depending on the state saved in SPSR_irq. The I bit in SPSR is significant. If it is set, it indicates that the ARM core was on the verge of masking an interrupt when the mask instruction was interrupted. Hence, when SPSR is restored, the mask instruction is completed (interrupt is masked).
5.
6. Note: 7. 8.
Note:
17.8.4 Fast Interrupt

17.8.4.1 Fast Interrupt Source The interrupt source 0 is the only source which can raise a fast interrupt request to the processor except if fast forcing is used. The interrupt source 0 is generally connected to a FIQ pin of the product, either directly or through a PIO Controller. 17.8.4.2 Fast Interrupt Control The fast interrupt logic of the AIC has no priority controller. The mode of interrupt source 0 is programmed with the AIC_SMR and INTSEL = 0, the field PRIOR of this register is not used even if it reads what has been written. The field SRCTYPE of AIC_SMR enables programming the fast interrupt source to be positive-edge triggered or negative-edge triggered or high-level sensitive or low-level sensitive Writing 0x1 in the AIC_IECR (Interrupt Enable Command Register) and AIC_IDCR (Interrupt Disable Command Register) respectively enables and disables the fast interrupt when INTSEL = 0. The bit 0 of AIC_IMR (Interrupt Mask Register) indicates whether the fast interrupt is enabled or disabled.

107
17.8.4.3 Fast Interrupt Vectoring The fast interrupt handler address can be stored through the AIC_SVR (Source Vector Register). The value written into this register when INTSEL = 0 is returned when the processor reads AIC_FVR (Fast Vector Register). This offers a way to branch in one single instruction to the interrupt handler, as AIC_FVR is mapped at the absolute address 0xFFFF F104 and thus accessible from the ARM fast interrupt vector at address 0x0000 001C through the following instruction: LDR PC,[PC,# -&F20] When the processor executes this instruction it loads the value read in AIC_FVR in its program counter, thus branching the execution on the fast interrupt handler. It also automatically performs the clear of the fast interrupt source if it is programmed in edge-triggered mode. 17.8.4.4 Fast Interrupt Handlers This section gives an overview of the fast interrupt handling sequence when using the AIC. It is assumed that the programmer understands the architecture of the ARM processor, and especially the processor interrupt modes and associated status bits. Assuming that: 1. 2. 3. 1. The Advanced Interrupt Controller has been programmed, AIC_SVR is loaded with the fast interrupt service routine address, and the interrupt source 0 is enabled. The Instruction at address 0x1C (FIQ exception vector address) is required to vector the fast interrupt: LDR PC, [PC, # -&F20] The user does not need nested fast interrupts. The CPSR is stored in SPSR_fiq, the current value of the program counter is loaded in the FIQ link register (R14_FIQ) and the program counter (R15) is loaded with 0x1C. In the following cycle, during fetch at address 0x20, the ARM core adjusts R14_fiq, decrementing it by four. The ARM core enters FIQ mode. When the instruction loaded at address 0x1C is executed, the program counter is loaded with the value read in AIC_FVR. Reading the AIC_FVR has effect of automatically clearing the fast interrupt, if it has been programmed to be edge triggered. In this case only, it de-asserts the nFIQ line on the processor. The previous step enables branching to the corresponding interrupt service routine. It is not necessary to save the link register R14_fiq and SPSR_fiq if nested fast interrupts are not needed. The Interrupt Handler can then proceed as required. It is not necessary to save registers R8 to R13 because FIQ mode has its own dedicated registers and the user R8 to R13 are banked. The other registers, R0 to R7, must be saved before being used, and restored at the end (before the next step). Note that if the fast interrupt is programmed to be level sensitive, the source of the interrupt must be cleared during this phase in order to de-assert the interrupt source 0. Finally, the Link Register R14_fiq is restored into the PC after decrementing it by four (with instruction SUB PC, LR, #4 for example). This has the effect of returning from the interrupt to whatever was being executed before, loading the CPSR with the SPSR and masking or unmasking the fast interrupt depending on the state saved in the SPSR. The F bit in SPSR is significant. If it is set, it indicates that the ARM core was just about to mask FIQ interrupts when the mask instruction was interrupted. Hence when the SPSR is restored, the interrupted instruction is completed (FIQ is masked).
When nFIQ is asserted, if the bit F of CPSR is 0, the sequence is:
2. 3.
4. 5.
6.
Note:
Another way to handle the fast interrupt is to map the interrupt service routine at the address of the ARM vector 0x1C. This method does not use the vectoring, so that reading AIC_FVR must be performed at the very beginning of the handler operation. However, this method saves the execution of a branch instruction.

108
17.8.4.5 Fast Forcing The Fast Forcing feature of the advanced interrupt controller provides redirection of any normal Interrupt source on the fast interrupt controller. Fast Forcing is enabled or disabled by writing to the Fast Forcing Enable Register (AIC_FFER) and the Fast Forcing Disable Register (AIC_FFDR). Writing to these registers results in an update of the Fast Forcing Status Register (AIC_FFSR) that controls the feature for each internal or external interrupt source. When Fast Forcing is disabled, the interrupt sources are handled as described in the previous pages. When Fast Forcing is enabled, the edge/level programming and, in certain cases, edge detection of the interrupt source is still active but the source cannot trigger a normal interrupt to the processor and is not seen by the priority handler. If the interrupt source is programmed in level-sensitive mode and an active level is sampled, Fast Forcing results in the assertion of the nFIQ line to the core. If the interrupt source is programmed in edge-triggered mode and an active edge is detected, Fast Forcing results in the assertion of the nFIQ line to the core. The Fast Forcing feature does not affect the Source 0 pending bit in the Interrupt Pending Register (AIC_IPR). The FIQ Vector Register (AIC_FVR) reads the contents of the Source Vector Register 0 (AIC_SVR0), whatever the source of the fast interrupt may be. The read of the FVR does not clear the Source 0 when the fast forcing feature is used and the interrupt source should be cleared by writing to the Interrupt Clear Command Register (AIC_ICCR). All enabled and pending interrupt sources that have the fast forcing feature enabled and that are programmed in edgetriggered mode must be cleared by writing to the Interrupt Clear Command Register. In doing so, they are cleared independently and thus lost interrupts are prevented. The read of AIC_IVR does not clear the source that has the fast forcing feature enabled. The source 0, reserved to the fast interrupt, continues operating normally and becomes one of the Fast Interrupt sources.
Figure 17-10. Fast Forcing
Source 0 _ FIQ Input Stage AIC_IMR AIC_IPR
Automatic Clear
nFIQ
Read FVR if Fast Forcing is disabled on Sources 1 to 127. AIC_FFSR Source n Input Stage Automatic Clear AIC_IMR AIC_IPR Priority Manager nIRQ
Read IVR if Source n is the current interrupt and if Fast Forcing is disabled on Source n.

109
17.8.5 Protect Mode

The Protect Mode permits reading the Interrupt Vector Register without performing the associated automatic operations. This is necessary when working with a debug system. When a debugger, working either with a Debug Monitor or the ARM processor's ICE, stops the applications and updates the opened windows, it might read the AIC User Interface and thus the IVR. This has undesirable consequences:
z z
If an enabled interrupt with a higher priority than the current one is pending, it is stacked. If there is no enabled pending interrupt, the spurious vector is returned.
In either case, an End of Interrupt command is necessary to acknowledge and to restore the context of the AIC. This operation is generally not performed by the debug system as the debug system would become strongly intrusive and cause the application to enter an undesired state. This is avoided by using the Protect Mode. Writing PROT in AIC_DCR (Debug Control Register) at 0x1 enables the Protect Mode. When the Protect Mode is enabled, the AIC performs interrupt stacking only when a write access is performed on the AIC_IVR. Therefore, the Interrupt Service Routines must write (arbitrary data) to the AIC_IVR just after reading it. The new context of the AIC, including the value of the Interrupt Status Register (AIC_ISR), is updated with the current interrupt only when AIC_IVR is written. An AIC_IVR read on its own (e.g., by a debugger), modifies neither the AIC context nor the AIC_ISR. Extra AIC_IVR reads perform the same operations. However, it is recommended to not stop the processor between the read and the write of AIC_IVR of the interrupt service routine to make sure the debugger does not modify the AIC context. To summarize, in normal operating mode, the read of AIC_IVR performs the following operations within the AIC: 1. 2. 3. 4. 5. Calculates active interrupt (higher than current or spurious). Determines and returns the vector of the active interrupt. Memorizes the interrupt. Pushes the current priority level onto the internal stack. Acknowledges the interrupt.
However, while the Protect Mode is activated, only operations 1 to 3 are performed when AIC_IVR is read. Operations 4 and 5 are only performed by the AIC when AIC_IVR is written. Software that has been written and debugged using the Protect Mode runs correctly in Normal Mode without modification. However, in Normal Mode the AIC_IVR write has no effect and can be removed to optimize the code.
17.8.6 Spurious Interrupt

The Advanced Interrupt Controller features protection against spurious interrupts. A spurious interrupt is defined as being the assertion of an interrupt source long enough for the AIC to assert the nIRQ, but no longer present when AIC_IVR is read. This is most prone to occur when:
z z z
An external interrupt source is programmed in level-sensitive mode and an active level occurs for only a short time. An internal interrupt source is programmed in level sensitive and the output signal of the corresponding embedded peripheral is activated for a short time. (As in the case for the Watchdog.) An interrupt occurs just a few cycles before the software begins to mask it, thus resulting in a pulse on the interrupt source.
The AIC detects a spurious interrupt at the time the AIC_IVR is read while no enabled interrupt source is pending. When this happens, the AIC returns the value stored by the programmer in AIC_SPU (Spurious Vector Register). The programmer must store the address of a spurious interrupt handler in AIC_SPU as part of the application, to enable an as fast as possible return to the normal execution flow. This handler writes in AIC_EOICR and performs a return from interrupt.

110
17.8.7 General Interrupt Mask

The AIC features a General Interrupt Mask bit to prevent interrupts from reaching the processor. Both the nIRQ and the nFIQ lines are driven to their inactive state if the bit GMSK in AIC_DCR (Debug Control Register) is set. However, this mask does not prevent waking up the processor if it has entered Idle Mode. This function facilitates synchronizing the processor on a next event and, as soon as the event occurs, performs subsequent operations without having to handle an interrupt. It is strongly recommended to use this mask with caution.
17.8.8 Write Protected Registers

To prevent any single software error that may corrupt AIC behavior, the registers listed below can be write-protected by setting the WPEN bit in the AIC Write Protect Mode Register (AIC_WPMR). If a write access in a write-protected register is detected, then the WPVS flag in the AIC Write Protect Status Register(AIC_WPSR) is set and the WPVSRC field indicates in which register the write access has been attempted. The WPVS flag is automatically reset after reading the AIC Write Protect Status Register. List of the write-protected registers:
z z z z
AIC Source Mode Register AIC Source Vector Register AIC Spurious Interrupt Vector Register AIC Debug Control Register

111
17.9
Offset 0x00 0x04 0x08 0x0C 0x10 0x14 0x18 0x1C 0x20 0x24 0x28 0x2C 0x30 0x34 0x38 0x3C 0x40 0x44 0x48 0x4C 0x50 0x54 0x58 0x5C 0x6C 0xE4 0xE8
Advanced Interrupt Controller (AIC) User Interface

Register Source Select Register Source Mode Register Source Vector Register Reserved Interrupt Vector Register FIQ Interrupt Vector Register Interrupt Status Register Reserved Interrupt Pending Register 0
(2) (2)
Table 17-3. Register Mapping Name AIC_SSR AIC_SMR AIC_SVR AIC_IVR AIC_FVR AIC_ISR AIC_IPR0 AIC_IPR1 AIC_IPR2 AIC_IPR3 AIC_IMR AIC_CISR AIC_EOICR AIC_SPU AIC_IECR AIC_IDCR AIC_ICCR AIC_ISCR AIC_FFER AIC_FFDR AIC_FFSR AIC_DCR AIC_WPMR AIC_WPSR Access Read-write Read-write Read-write Read-only Read-only Read-only Read-only Read-only Read-only Read-only Read-only Read-only Write-only Read-write Write-only Write-only Write-only Write-only Write-only Write-only Read-only Read-write Read-write Read-only Reset 0x0 0x0 0x0 0x0 0x0 0x0 0x0(1) 0x0(1) 0x0(1) 0x0(1) 0x0 0x0 0x0 0x0 0x0 0x0 0x0
Interrupt Pending Register 1
Interrupt Pending Register 2(2) Interrupt Pending Register 3(2) Interrupt Mask Register Core Interrupt Status Register End of Interrupt Command Register Spurious Interrupt Vector Register Interrupt Enable Command Register Interrupt Disable Command Register Interrupt Clear Command Register Interrupt Set Command Register Fast Forcing Enable Register Fast Forcing Disable Register Fast Forcing Status Register Reserved Debug Control Register Write Protect Mode Register Write Protect Status Register Reserved
0xEC - 0xFC
Notes: 1. The reset value of this register depends on the level of the external interrupt source. All other sources are cleared at reset, thus not pending. 2. PID2...PID127 bit fields refer to the identifiers as defined in the Peripheral Identifiers Section of the product datasheet.

112
17.9.1 AIC Source Select Register

31 23 15 7
AIC_SSR 0xFFFFF000 Read-write 0x0

30 22 14 6 29 21 13 5 28 20 12 4 27 19 11 3 INTSEL 26 18 10 2 25 17 9 1 24 16 8 0
INTSEL: Interrupt line Selection

0-127 = Select the interrupt line to handle. See: Section 17.8.1.1 Interrupt Source Mode.

113
17.9.2 AIC Source Mode Register

31 23 15 7
AIC_SMR 0xFFFFF004 Read-write 0x0

30 22 14 6 SRCTYPE 29 21 13 5 28 20 12 4 27 19 11 3 26 18 10 2 25 17 9 1 PRIOR 24 16 8 0
PRIOR: Priority Level

Programs the priority level of the source selected by INTSEL in except FIQ source (source 0). The priority level can be between 0 (lowest) and 7 (highest). The priority level is not used for the FIQ.
SRCTYPE: Interrupt Source Type

The active level or edge is not programmable for the internal interrupt source selected by INTSEL.
Value 0x0 Name INT_LEVEL_SENSITIVE Description High level Sensitive for internal source Low level Sensitive for external source Positive edge triggered for internal source Negative edge triggered for external source High level Sensitive for internal source High level Sensitive for external source Positive edge triggered for internal source Positive edge triggered for external source
0x1
INT_EDGE_TRIGGERED
0x2
EXT_HIGH_LEVEL
0x3
EXT_POSITIVE_EDGE
Value 0 0 1 1 0 1 0 1
Internal Interrupt High level sensitive Positive edge triggered High level sensitive Positive edge triggered
External Interrupt Low level sensitive Negative edge triggered High level sensitive Positive edge triggered

114
17.9.3 AIC Source Vector Register

Name: Address: Access: AIC_SVR 0xFFFFF008 Read-write
Reset:
31
0x0
30 29 28 VECTOR 23 22 21 20 VECTOR 15 14 13 12 VECTOR 7 6 5 4 VECTOR 3 2 1 0 11 10 9 8 19 18 17 16 27 26 25 24
VECTOR: Source Vector

The user may store in this register the address of the corresponding handler for the interrupt source selected by INTSEL.

115
17.9.4 AIC Interrupt Vector Register

Name: Address: Access: AIC_IVR 0xFFFFF010 Read-only
Reset:
31
0x0
30 29 28 IRQV 23 22 21 20 IRQV 15 14 13 12 IRQV 7 6 5 4 IRQV 3 2 1 0 11 10 9 8 19 18 17 16 27 26 25 24
IRQV: Interrupt Vector Register

The Interrupt Vector Register contains the vector programmed by the user in the Source Vector Register corresponding to the current interrupt. The Source Vector Register is indexed using the current interrupt number when the Interrupt Vector Register is read. When there is no current interrupt, the Interrupt Vector Register reads the value stored in AIC_SPU.

116
17.9.5 AIC FIQ Vector Register

Name: Address: Access: AIC_FVR 0xFFFFF014 Read-only
Reset:
31
0x0
30 29 28 FIQV 23 22 21 20 FIQV 15 14 13 12 FIQV 7 6 5 4 FIQV 3 2 1 0 11 10 9 8 19 18 17 16 27 26 25 24
FIQV: FIQ Vector Register

The FIQ Vector Register contains the vector programmed by the user in the Source Vector Register when INTSEL = 0. When there is no fast interrupt, the FIQ Vector Register reads the value stored in AIC_SPU.

117
17.9.6 AIC Interrupt Status Register

Name: Address: Access: AIC_ISR 0xFFFFF018 Read-only
Reset:
31 23 15 7
0x0
30 22 14 6 29 21 13 5 28 20 12 4 27 19 11 3 IRQID 26 18 10 2 25 17 9 1 24 16 8 0
IRQID: Current Interrupt Identifier

The Interrupt Status Register returns the current interrupt source number.

118
17.9.7 AIC Interrupt Pending Register 0

Name: Address: Access: AIC_IPR0 0xFFFFF020 Read-only
Reset:
31 PID31 23 PID23 15 PID15 7 PID7
0x0
30 PID30 22 PID22 14 PID14 6 PID6 29 PID29 21 PID21 13 PID13 5 PID5 28 PID28 20 PID20 12 PID12 4 PID4 27 PID27 19 PID19 11 PID11 3 PID3 26 PID26 18 PID18 10 PID10 2 PID2 25 PID25 17 PID17 9 PID9 1 SYS 24 PID24 16 PID16 8 PID8 0 FIQ
FIQ, SYS, PIDx: Interrupt Pending

0 = The corresponding interrupt is not pending. 1 = The corresponding interrupt is pending.

Reset:
31 PID63 23 PID55 15 PID47 7 PID39
0x0
30 PID62 22 PID54 14 PID46 6 PID38 29 PID61 21 PID53 13 PID45 5 PID37 28 PID60 20 PID52 12 PID44 4 PID36 27 PID59 19 PID51 11 PID43 3 PID35 26 PID58 18 PID50 10 PID42 2 PID34 25 PID57 17 PID49 9 PID41 1 PID33 24 PID56 16 PID48 8 PID40 0 PID32
PIDx: Interrupt Pending


119

Reset:
31 PID95 23 PID87 15 PID79 7 PID71
0x0


Name: Address: Access: AIC_IPR3 0xFFFFF02C Read-only
Reset:
31 PID127 23 PID119 15 PID111 7 PID103
0x0


120
17.9.11 AIC Interrupt Mask Register

Name: Address: Access: AIC_IMR 0xFFFFF030 Read-only
Reset:
31 23 15 7
0x0
30 22 14 6 29 21 13 5 28 20 12 4 27 19 11 3 26 18 10 2 25 17 9 1 24 16 8 0 INTM
INTM: Interrupt Mask

0 = The interrupt source selected by INTSEL is disabled. 1 = The interrupt source selected by INTSEL is enabled.

121
17.9.12 AIC Core Interrupt Status Register

Name: Address: Access: AIC_CISR 0xFFFFF034 Read-only
Reset:
31 23 15 7
0x0
30 22 14 6 29 21 13 5 28 20 12 4 27 19 11 3 26 18 10 2 25 17 9 1 NIRQ 24 16 8 0 NFIQ
NFIQ: NFIQ Status

0 = nFIQ line is deactivated. 1 = nFIQ line is active.
NIRQ: NIRQ Status

0 = nIRQ line is deactivated. 1 = nIRQ line is active.

122
17.9.13 AIC End of Interrupt Command Register

Name: Address: AIC_EOICR 0xFFFFF038
Access:
31 23 15 7
Write-only
30 22 14 6 29 21 13 5 28 20 12 4 27 19 11 3 26 18 10 2 25 17 9 1 24 16 8 0 ENDIT
ENDIT: Interrupt Processing Complete Command

The End of Interrupt Command Register is used by the interrupt routine to indicate that the interrupt treatment is complete. Any value can be written because it is only necessary to make a write to this register location to signal the end of interrupt treatment.

123
17.9.14 AIC Spurious Interrupt Vector Register

Name: Address: AIC_SPU 0xFFFFF03C
Access: Reset:
31
Read-write 0x0
30 29 28 SIVR 27 26 25 24
23
22
21
20 SIVR
19
18
17
16
15
14
13
12 SIVR
11
10
4 SIVR
SIVR: Spurious Interrupt Vector Register

The user may store the address of a spurious interrupt handler in this register. The written value is returned in AIC_IVR in case of a spurious interrupt and in AIC_FVR in case of a spurious fast interrupt.

124
17.9.15 AIC Interrupt Enable Command Register

Name: Address: AIC_IECR 0xFFFFF040
Access:
31 23 15 7
Write-only
30 22 14 6 29 21 13 5 28 20 12 4 27 19 11 3 26 18 10 2 25 17 9 1 24 16 8 0 INTEN
INTEN: Interrupt Enable

0 = No effect. 1 = Enables the interrupt source selected by INTSEL.
17.9.16 AIC Interrupt Disable Command Register

Name: Address: AIC_IDCR 0xFFFFF044
Access:
31 23 15 7
Write-only
30 22 14 6 29 21 13 5 28 20 12 4 27 19 11 3 26 18 10 2 25 17 9 1 24 16 8 0 INTD
INTD: Interrupt Disable

0 = No effect. 1 = Disables the interrupt source selected by INTSEL.

125
17.9.17 AIC Interrupt Clear Command Register

Name: Address: AIC_ICCR 0xFFFFF048
Access:
31 23 15 7
Write-only
30 22 14 6 29 21 13 5 28 20 12 4 27 19 11 3 26 18 10 2 25 17 9 1 24 16 8 0 INTCLR
INTCLR: Interrupt Clear

Clears one the following depending on the setting of the INTSEL bit FIQ, SYS, PID2-PID127 0 = No effect. 1 = Clears the interrupt source selected by INTSEL.

126
17.9.18 AIC Interrupt Set Command Register

Name: Address: AIC_ISCR 0xFFFFF04C
Access:
31 23 15 7
Write-only
30 22 14 6 29 21 13 5 28 20 12 4 27 19 11 3 26 18 10 2 25 17 9 1 24 16 8 0 INTSET
INTSET: Interrupt Set

0 = No effect. 1 = Sets the interrupt source selected by INTSEL.

127
17.9.19 AIC Fast Forcing Enable Register

Name: Address: AIC_FFER 0xFFFFF050
Access:
31 23 15 7
Write-only
30 22 14 6 29 21 13 5 28 20 12 4 27 19 11 3 26 18 10 2 25 17 9 1 24 16 8 0 FFEN
FFEN: Fast Forcing Enable

0 = No effect. 1 = Enables the fast forcing feature on the interrupt source selected by INTSEL.
17.9.20 AIC Fast Forcing Disable Register

Name: Address: AIC_FFDR 0xFFFFF054
Access:
31 23 15 7
Write-only
30 22 14 6 29 21 13 5 28 20 12 4 27 19 11 3 26 18 10 2 25 17 9 1 24 16 8 0 FFDIS
FFDIS: Fast Forcing Disable

0 = No effect. 1 = Disables the Fast Forcing feature on the interrupt source selected by INTSEL.

128
17.9.21 AIC Fast Forcing Status Register

Name: Address: AIC_FFSR 0xFFFFF058
Access:
31 23 15 7
Read-only
30 22 14 6 29 21 13 5 28 20 12 4 27 19 11 3 26 18 10 2 25 17 9 1 24 16 8 0 FFS
FFS: Fast Forcing Status

0 = The Fast Forcing feature is disabled on the interrupt source selected by INTSEL. 1 = The Fast Forcing feature is enabled on the interrupt source selected by INTSEL.

129
17.9.22 AIC Debug Control Register

Name: Address: AIC_DCR 0xFFFFF06C
Access: Reset:
31 23 15 7
Read-write 0x0
30 22 14 6 29 21 13 5 28 20 12 4 27 19 11 3 26 18 10 2 25 17 9 1 GMSK 24 16 8 0 PROT
PROT: Protection Mode

0 = The Protection Mode is disabled. 1 = The Protection Mode is enabled.
GMSK: General Mask

0 = The nIRQ and nFIQ lines are normally controlled by the AIC. 1 = The nIRQ and nFIQ lines are tied to their inactive state.

130
17.9.23 AIC Write Protect Mode Register

31
AIC_WPMR 0xFFFFF0E4 Read-write See Table 17-3


0 = Disables the Write Protect if WPKEY corresponds to 0x414943 (AIC in ASCII). 1 = Enables the Write Protect if WPKEY corresponds to 0x414943 (AIC in ASCII). Protects the registers:
AIC Source Mode Register AIC Source Vector Register AIC Spurious Interrupt Vector Register AIC Debug Control Register WPKEY: Write Protect KEY
Should be written at value 0x414943 (AIC in ASCII). Writing any other value in this field aborts the write operation of the WPEN bit. Always reads as 0.

131
17.9.24 AIC Write Protect Status Register

31 23
AIC_WPSR 0xFFFFF0E8 Read-only See Table 17-3


0 = No Write Protect Violation has occurred since the last read of the AIC_WPSR register. 1 = A Write Protect Violation has occurred since the last read of the AIC_WPSR register. If this violation is an unauthorized attempt to write a protected register, the associated violation is reported into field WPVSRC.

When WPVS is active, this field indicates the write-protected register (through address offset or code) in which a write access has been attempted. Note: Reading AIC_WPSR automatically clears all fields.

132
18.
18.1
Watchdog Timer (WDT)

Description
The Watchdog Timer can be used to prevent system lock-up if the software becomes trapped in a deadlock. It features a 12-bit down counter that allows a watchdog period of up to 16 seconds (slow clock at 32768 kHz). It can generate a general reset or a processor reset only. In addition, it can be stopped while the processor is in debug mode or idle mode.
18.2
z z z
12-bit Key-protected Programmable Counter Provides Reset or Interrupt Signals to the System Counter May Be Stopped While the Processor is in Debug State or in Idle Mode

133
18.3
Block Diagram
Figure 18-1. Watchdog Timer Block Diagram

write WDT_MR WDT_MR WDT_CR WDRSTT reload 1 0 WDV
12-bit Down Counter WDT_MR WDD Current Value reload 1/128 SLCK
<= WDD WDT_MR WDRSTEN = 0 wdt_fault (to Reset Controller) wdt_int
set set read WDT_SR or reset WDERR reset WDUNF reset WDFIEN WDT_MR

134
18.4
The Watchdog Timer can be used to prevent system lock-up if the software becomes trapped in a deadlock. It is supplied with VDDCORE. It restarts with initial values on processor reset. The Watchdog is built around a 12-bit down counter, which is loaded with the value defined in the field WDV of the Mode Register (WDT_MR). The Watchdog Timer uses the Slow Clock divided by 128 to establish the maximum Watchdog period to be 16 seconds (with a typical Slow Clock of 32768 kHz). After a Processor Reset, the value of WDV is 0xFFF, corresponding to the maximum value of the counter with the external reset generation enabled (field WDRSTEN at 1 after a Backup Reset). This means that a default Watchdog is running at reset, i.e., at power-up. The user must either disable it (by setting the WDDIS bit in WDT_MR) if he does not expect to use it or must reprogram it to meet the maximum Watchdog period the application requires. If the watchdog is restarted by writing into the WDT_CR register, the WDT_MR register must not be programmed during a period of time of 3 slow clock periods following the WDT_CR write access. In any case, programming a new value in the WDT_MR register automatically initiates a restart instruction. The Watchdog Mode Register (WDT_MR) can be written only once. Only a processor reset resets it. Writing the WDT_MR register reloads the timer with the newly programmed mode parameters. In normal operation, the user reloads the Watchdog at regular intervals before the timer underflow occurs, by writing the Control Register (WDT_CR) with the bit WDRSTT to 1. The Watchdog counter is then immediately reloaded from WDT_MR and restarted, and the Slow Clock 128 divider is reset and restarted. The WDT_CR register is write-protected. As a result, writing WDT_CR without the correct hard-coded key has no effect. If an underflow does occur, the wdt_fault signal to the Reset Controller is asserted if the bit WDRSTEN is set in the Mode Register (WDT_MR). Moreover, the bit WDUNF is set in the Watchdog Status Register (WDT_SR). To prevent a software deadlock that continuously triggers the Watchdog, the reload of the Watchdog must occur while the Watchdog counter is within a window between 0 and WDD, WDD is defined in the WatchDog Mode Register WDT_MR. Any attempt to restart the Watchdog while the Watchdog counter is between WDV and WDD results in a Watchdog error, even if the Watchdog is disabled. The bit WDERR is updated in the WDT_SR and the wdt_fault signal to the Reset Controller is asserted. Note that this feature can be disabled by programming a WDD value greater than or equal to the WDV value. In such a configuration, restarting the Watchdog Timer is permitted in the whole range [0; WDV] and does not generate an error. This is the default configuration on reset (the WDD and WDV values are equal). The status bits WDUNF (Watchdog Underflow) and WDERR (Watchdog Error) trigger an interrupt, provided the bit WDFIEN is set in the mode register. The signal wdt_fault to the reset controller causes a Watchdog reset if the WDRSTEN bit is set as already explained in the reset controller programmer Datasheet. In that case, the processor and the Watchdog Timer are reset, and the WDERR and WDUNF flags are reset. If a reset is generated or if WDT_SR is read, the status bits are reset, the interrupt is cleared, and the wdt_fault signal to the reset controller is deasserted. Writing the WDT_MR reloads and restarts the down counter. While the processor is in debug state or in idle mode, the counter may be stopped depending on the value programmed for the bits WDIDLEHLT and WDDBGHLT in the WDT_MR.

135
Figure 18-2. Watchdog Behavior

Watchdog Error Watchdog Underflow if WDRSTEN is 1 FFF Normal behavior WDV Forbidden Window WDD Permitted Window 0 WDT_CR = WDRSTT if WDRSTEN is 0
Watchdog Fault

136
18.5
Watchdog Timer (WDT) User Interface
Table 18-1. Register Mapping Offset 0x00 0x04 0x08 Register Control Register Mode Register Status Register Name WDT_CR WDT_MR WDT_SR Access Write-only Read-write Once Read-only Reset 0x3FFF_2FFF 0x0000_0000

137
18.5.1 Watchdog Timer Control Register

31
WDT_CR 0xFFFFFE40 Write-only

30 29 28 KEY 27 26 25 24
23 15 7
22 14 6
21 13 5
20 12 4
19 11 3
18 10 2
17 9 1
16 8 0 WDRSTT
WDRSTT: Watchdog Restart

0: No effect. 1: Restarts the Watchdog.
KEY: Password
Should be written at value 0xA5. Writing any other value in this field aborts the write operation.

138
18.5.2 Watchdog Timer Mode Register

31
WDT_MR 0xFFFFFE44 Read-write Once

30 29 WDIDLEHLT 21 28 WDDBGHLT 20 WDD 27 26 WDD 19 18 17 16 25 24
23
22
15 WDDIS 7
14 WDRPROC 6
13 WDRSTEN 5
12 WDFIEN 4 WDV
11
10 WDV
WDV: Watchdog Counter Value

Defines the value loaded in the 12-bit Watchdog Counter.
WDFIEN: Watchdog Fault Interrupt Enable

0: A Watchdog fault (underflow or error) has no effect on interrupt. 1: A Watchdog fault (underflow or error) asserts interrupt.
WDRSTEN: Watchdog Reset Enable

0: A Watchdog fault (underflow or error) has no effect on the resets. 1: A Watchdog fault (underflow or error) triggers a Watchdog reset.
WDRPROC: Watchdog Reset Processor

0: If WDRSTEN is 1, a Watchdog fault (underflow or error) activates all resets. 1: If WDRSTEN is 1, a Watchdog fault (underflow or error) activates the processor reset.
WDD: Watchdog Delta Value

Defines the permitted range for reloading the Watchdog Timer. If the Watchdog Timer value is less than or equal to WDD, writing WDT_CR with WDRSTT = 1 restarts the timer. If the Watchdog Timer value is greater than WDD, writing WDT_CR with WDRSTT = 1 causes a Watchdog error.
WDDBGHLT: Watchdog Debug Halt

0: The Watchdog runs when the processor is in debug state. 1: The Watchdog stops when the processor is in debug state.
WDIDLEHLT: Watchdog Idle Halt

0: The Watchdog runs when the system is in idle mode. 1: The Watchdog stops when the system is in idle state.
WDDIS: Watchdog Disable

0: Enables the Watchdog Timer. 1: Disables the Watchdog Timer.

139
18.5.3 Watchdog Timer Status Register

31 23 15 7
WDT_SR 0xFFFFFE48 Read-only

30 22 14 6 29 21 13 5 28 20 12 4 27 19 11 3 26 18 10 2 25 17 9 1 WDERR 24 16 8 0 WDUNF
WDUNF: Watchdog Underflow

0: No Watchdog underflow occurred since the last read of WDT_SR. 1: At least one Watchdog underflow occurred since the last read of WDT_SR.
WDERR: Watchdog Error

0: No Watchdog error occurred since the last read of WDT_SR. 1: At least one Watchdog error occurred since the last read of WDT_SR. Note: The WDD and WDV values must not be modified within a period of time of 3 slow clock periods following a restart of the watchdog performed by means of a write access in the WDT_CR register, else the watchdog may trigger an end of period earlier than expected.

140
19.
19.1
Reset Controller (RSTC)

Description
The Reset Controller (RSTC), based on power-on reset cells, handles all the resets of the system without any external components. It reports which reset occurred last. The Reset Controller also drives independently or simultaneously the external reset and the peripheral and processor resets.
19.2
z
Manages All Resets of the System, Including

z z z z
External Devices Through the NRST Pin Processor Reset Peripheral Set Reset Backed-up Peripheral Reset
z z
Based on 2 Embedded Power-on Reset Cells Reset Source Status

z z
Status of the Last Reset Either General Reset, Wake-up Reset, Software Reset, User Reset, Watchdog Reset
External Reset Signal Shaping

141
19.3
Block Diagram
Figure 19-1. Reset Controller Block Diagram
Reset Controller
Main Supply POR Backup Supply POR Startup Counter rstc_irq Reset State Manager proc_nreset
user_reset
NRST
nrst_out
NRST Manager
exter_nreset
periph_nreset
backup_neset WDRPROC wd_fault
SLCK

142
19.4
19.4.1 Reset Controller Overview

The Reset Controller is made up of an NRST Manager, a Startup Counter and a Reset State Manager. It runs at Slow Clock and generates the following reset signals:
z z z z
proc_nreset: Processor reset line. It also resets the Watchdog Timer. backup_nreset: Affects all the peripherals powered by VDDBU. periph_nreset: Affects the whole set of embedded peripherals. nrst_out: Drives the NRST pin.
These reset signals are asserted by the Reset Controller, either on external events or on software action. The Reset State Manager controls the generation of reset signals and provides a signal to the NRST Manager when an assertion of the NRST pin is required. The NRST Manager shapes the NRST assertion during a programmable time, thus controlling external device resets. The startup counter waits for the complete crystal oscillator startup. The wait delay is given by the crystal oscillator startup time maximum value that can be found in the section Crystal Oscillator Characteristics in the Electrical Characteristics section of the product documentation. The Reset Controller Mode Register (RSTC_MR), allowing the configuration of the Reset Controller, is powered with VDDBU, so that its configuration is saved as long as VDDBU is on.
19.4.2 NRST Manager

After power-up, NRST is an output during the ERSTL time defined in the RSTC. When ERSTL elapsed, the pin behaves as an input and all the system is held in reset if NRST is tied to GND by an external signal. The NRST Manager samples the NRST input pin and drives this pin low when required by the Reset State Manager. Figure 19-2 shows the block diagram of the NRST Manager.
Figure 19-2. NRST Manager
RSTC_SR
URSTS NRSTL
user_reset
NRST
RSTC_MR
ERSTL nrst_out External Reset Timer exter_nreset
NRST Signal The NRST Manager handles the NRST input line asynchronously. When the line is low, a User Reset is immediately reported to the Reset State Manager. When the NRST goes from low to high, the internal reset is synchronized with the Slow Clock to provide a safe internal de-assertion of reset. The level of the pin NRST can be read at any time in the bit NRSTL (NRST level) in RSTC_SR. As soon as the pin NRST is asserted, the bit URSTS in RSTC_SR is set. This bit clears only when RSTC_SR is read.

143
19.4.2.1 NRST External Reset Control The Reset State Manager asserts the signal ext_nreset to assert the NRST pin. When this occurs, the nrst_out signal is driven low by the NRST Manager for a time programmed by the field ERSTL in RSTC_MR. This assertion duration, named EXTERNAL_RESET_LENGTH, lasts 2(ERSTL+1) Slow Clock cycles. This gives the approximate duration of an assertion between 60 s and 2 seconds. Note that ERSTL at 0 defines a two-cycle duration for the NRST pulse. This feature allows the Reset Controller to shape the NRST pin level, and thus to guarantee that the NRST line is driven low for a time compliant with potential external devices connected on the system reset. As the field is within RSTC_MR, which is backed-up, this field can be used to shape the system power-up reset for devices requiring a longer startup time than the Slow Clock Oscillator.
19.4.3 BMS Sampling

The product matrix manages a boot memory that depends on the level on the BMS pin at reset. The BMS signal is sampled three slow clock cycles after the Core Power-On-Reset output rising edge.
Figure 19-3. BMS Sampling
SLCK
Core Supply POR output XXX

BMS sampling delay = 3 cycles
BMS Signal
H or L
proc_nreset

144
19.4.4 Reset States

The Reset State Manager handles the different reset sources and generates the internal reset signals. It reports the reset status in the field RSTTYP of the Status Register (RSTC_SR). The update of the field RSTTYP is performed when the processor reset is released. 19.4.4.1 General Reset A general reset occurs when VDDBU and VDDCORE are powered on. The backup supply POR cell output rises and is filtered with a Startup Counter, which operates at Slow Clock. The purpose of this counter is to make sure the Slow Clock oscillator is stable before starting up the device. The length of startup time is hardcoded to comply with the Slow Clock Oscillator startup time. After this time, the processor clock is released at Slow Clock and all the other signals remain valid for Y cycles for proper processor and logic reset. Then, all the reset signals are released and the field RSTTYP in RSTC_SR reports a General Reset. As the RSTC_MR is reset, the NRST line rises 2 cycles after the backup_nreset, as ERSTL defaults at value 0x0. When VDDBU is detected low by the Backup Supply POR Cell, all resets signals are immediately asserted, even if the Main Supply POR Cell does not report a Main Supply shutdown. VDDBU only activates the backup_nreset signal. The backup_nreset must be released so that any other reset can be generated by VDDCORE (Main Supply POR output). Figure 19-4 shows how the General Reset affects the reset signals.
Figure 19-4. General Reset State
SLCK MCK Backup Supply POR output

Any Freq.
Startup Time
Main Supply POR output backup_nreset

Processor Startup
proc_nreset RSTTYP periph_nreset
XXX
0x0 = General Reset
XXX
NRST (nrst_out)
EXTERNAL RESET LENGTH BMS Sampling = 2 cycles

145
19.4.4.2 Wake-up Reset The Wake-up Reset occurs when the Main Supply is down. When the Main Supply POR output is active, all the reset signals are asserted except backup_nreset. When the Main Supply powers up, the POR output is resynchronized on Slow Clock. The processor clock is then re-enabled during Y Slow Clock cycles, depending on the requirements of the ARM processor. At the end of this delay, the processor and other reset signals rise. The field RSTTYP in RSTC_SR is updated to report a Wake-up Reset. The nrst_out remains asserted for EXTERNAL_RESET_LENGTH cycles. As RSTC_MR is backed-up, the programmed number of cycles is applicable. When the Main Supply is detected falling, the reset signals are immediately asserted. This transition is synchronous with the output of the Main Supply POR.
Figure 19-5. Wake-up Reset
SLCK MCK Main Supply POR output

Any Freq.
backup_nreset
Resynch. 2 cycles Processor Startup
proc_nreset
RSTTYP
XXX
0x1 = WakeUp Reset
XXX
periph_nreset
NRST (nrst_out)
EXTERNAL RESET LENGTH = 4 cycles (ERSTL = 1)

146
19.4.4.3 User Reset The User Reset is entered when a low level is detected on the NRST pin. When a falling edge occurs on NRST (reset activation), internal reset lines are immediately asserted. The Processor Reset and the Peripheral Reset are asserted. The User Reset is left when NRST rises, after a two-cycle resynchronization time and a Y-cycle processor startup. The processor clock is re-enabled as soon as NRST is confirmed high. When the processor reset signal is released, the RSTTYP field of the Status Register (RSTC_SR) is loaded with the value 0x4, indicating a User Reset. The NRST Manager guarantees that the NRST line is asserted for EXTERNAL_RESET_LENGTH Slow Clock cycles, as programmed in the field ERSTL. However, if NRST does not rise after EXTERNAL_RESET_LENGTH because it is driven low externally, the internal reset lines remain asserted until NRST actually rises.
Figure 19-6. User Reset State
SLCK MCK
Any Freq.
NRST
Resynch. 2 cycles Processor Startup
proc_nreset RSTTYP periph_nreset Any XXX 0x4 = User Reset
NRST (nrst_out)
>= EXTERNAL RESET LENGTH

147
19.4.4.4 Software Reset The Reset Controller offers several commands used to assert the different reset signals. These commands are performed by writing the Control Register (RSTC_CR) with the following bits at 1:
z z
PROCRST: Writing PROCRST at 1 resets the processor and the watchdog timer. PERRST: Writing PERRST at 1 resets all the embedded peripherals, including the memory system, and, in particular, the Remap Command. The Peripheral Reset is generally used for debug purposes. Except for Debug purposes, PERRST must always be used in conjunction with PROCRST (PERRST and PROCRST set both at 1 simultaneously.) EXTRST: Writing EXTRST at 1 asserts low the NRST pin during a time defined by the field ERSTL in the Mode Register (RSTC_MR).
The software reset is entered if at least one of these bits is set by the software. All these commands can be performed independently or simultaneously. The software reset lasts Y Slow Clock cycles. The internal reset signals are asserted as soon as the register write is performed. This is detected on the Master Clock (MCK). They are released when the software reset is left, i.e.; synchronously to SLCK. If EXTRST is set, the nrst_out signal is asserted depending on the programming of the field ERSTL. However, the resulting falling edge on NRST does not lead to a User Reset. If and only if the PROCRST bit is set, the Reset Controller reports the software status in the field RSTTYP of the Status Register (RSTC_SR). Other Software Resets are not reported in RSTTYP. As soon as a software operation is detected, the bit SRCMP (Software Reset Command in Progress) is set in the Status Register (RSTC_SR). It is cleared as soon as the software reset is left. No other software reset can be performed while the SRCMP bit is set, and writing any value in RSTC_CR has no effect.
Figure 19-7. Software Reset
SLCK MCK
Any Freq.
Write RSTC_CR
Resynch. 1 to 2 cycles Processor Startup = 3 cycles
proc_nreset if PROCRST=1 RSTTYP periph_nreset if PERRST=1 NRST (nrst_out) if EXTRST=1

EXTERNAL RESET LENGTH 8 cycles (ERSTL=2)
Any
XXX
0x3 = Software Reset
SRCMP in RSTC_SR

148
19.4.4.5 Watchdog Reset The Watchdog Reset is entered when a watchdog fault occurs. This state lasts Y Slow Clock cycles. When in Watchdog Reset, assertion of the reset signals depends on the WDRPROC bit in WDT_MR:
z
If WDRPROC is 0, the Processor Reset and the Peripheral Reset are asserted. The NRST line is also asserted, depending on the programming of the field ERSTL. However, the resulting low level on NRST does not result in a User Reset state. If WDRPROC = 1, only the processor reset is asserted.
The Watchdog Timer is reset by the proc_nreset signal. As the watchdog fault always causes a processor reset if WDRSTEN is set, the Watchdog Timer is always reset after a Watchdog Reset and the Watchdog is enabled by default and with a period set to a maximum. When the WDRSTEN in WDT_MR bit is reset, the watchdog fault has no impact on the reset controller.
Figure 19-8. Watchdog Reset
SLCK MCK
Any Freq.
wd_fault
Processor Startup = 3 cycles
proc_nreset RSTTYP periph_nreset Only if WDRPROC = 0 Any XXX 0x2 = Watchdog Reset
NRST (nrst_out)
EXTERNAL RESET LENGTH 8 cycles (ERSTL=2)

149
19.4.5 Reset State Priorities

The Reset State Manager manages the following priorities between the different reset sources, given in descending order:
z z z z z
Backup Reset Wake-up Reset User Reset Watchdog Reset Software Reset When in User Reset:
z z
Particular cases are listed below:

z
A watchdog event is impossible because the Watchdog Timer is being reset by the proc_nreset signal. A software reset is impossible, since the processor reset is being activated. A watchdog event has priority over the current state. The NRST has no effect. The processor reset is active and so a Software Reset cannot be programmed. A User Reset cannot be entered.
When in Software Reset:

z z
When in Watchdog Reset:

z z
19.4.6 Reset Controller Status Register

The Reset Controller status register (RSTC_SR) provides several status fields:
z z
RSTTYP field: This field gives the type of the last reset, as explained in previous sections. SRCMP bit: This field indicates that a Software Reset Command is in progress and that no further software reset should be performed until the end of the current one. This bit is automatically cleared at the end of the current software reset. NRSTL bit: The NRSTL bit of the Status Register gives the level of the NRST pin sampled on each MCK rising edge. URSTS bit: A high-to-low transition of the NRST pin sets the URSTS bit of the RSTC_SR register. This transition is also detected on the Master Clock (MCK) rising edge (see Figure 19-9). Reading the RSTC_SR status register resets the URSTS bit.
z z
Figure 19-9. Reset Controller Status and Interrupt

MCK read RSTC_SR
Peripheral Access
2 cycle resynchronization NRST NRSTL
2 cycle resynchronization
URSTS rstc_irq if (URSTEN = 0) and (URSTIEN = 1)

150
19.5
Reset Controller (RSTC) User Interface
Table 19-1. Register Mapping Offset 0x00 0x04 0x08 Register Control Register Status Register Mode Register Name RSTC_CR RSTC_SR RSTC_MR Access Write-only Read-only Read-write Reset 0x0000_0001(1) 0x0000_0000 0x0000_0000 Back-up Reset
Note:
1. The reset value of RSTC_SR either reports a General Reset or a Wake-up Reset depending on the last rising power supply.

151
19.5.1 Reset Controller Control Register

31
RSTC_CR 0xFFFFFE00 Write-only

30 29 28 KEY 27 26 25 24
23 15 7
22 14 6
21 13 5
20 12 4
19 11 3 EXTRST
18 10 2 PERRST
17 9
16 8 0 PROCRST
PROCRST: Processor Reset

0 = No effect. 1 = If KEY is correct, resets the processor.
PERRST: Peripheral Reset

0 = No effect. 1 = If KEY is correct, resets the peripherals.
EXTRST: External Reset

0 = No effect. 1 = If KEY is correct, asserts the NRST pin and resets the processor and the peripherals.
KEY: Password

152
19.5.2 Reset Controller Status Register

31 23 15 7
RSTC_SR 0xFFFFFE04 Read-only

30 22 14 6 29 21 13 5 28 20 12 4 27 19 11 3 26 18 10 25 17 SRCMP 9 RSTTYP 1 24 16 NRSTL 8
0 URSTS
URSTS: User Reset Status

0 = No high-to-low edge on NRST happened since the last read of RSTC_SR. 1 = At least one high-to-low transition of NRST has been detected since the last read of RSTC_SR.
RSTTYP: Reset Type

Reports the cause of the last processor reset. Reading this RSTC_SR does not reset this field.
RSTTYP 0 0 0 0 1 0 0 1 1 0 0 1 0 1 0 Reset Type General Reset Wake Up Reset Watchdog Reset Software Reset User Reset Comments Both VDDCORE and VDDBU rising VDDCORE rising Watchdog fault occurred Processor reset required by the software NRST pin detected low
NRSTL: NRST Pin Level

Registers the NRST Pin Level at Master Clock (MCK).
SRCMP: Software Reset Command in Progress

0 = No software command is being performed by the reset controller. The reset controller is ready for a software command. 1 = A software reset command is being performed by the reset controller. The reset controller is busy.

153
19.5.3 Reset Controller Mode Register

31
RSTC_MR 0xFFFFFE08 Read-write

30 29 28 KEY 27 26 25 24
23 15 7
22 14 6
21 13 5
20 12 4
19 11
18 10 ERSTL
17 9
16 8
ERSTL: External Reset Length

This field defines the external reset length. The external reset is asserted during a time of 2(ERSTL+1) Slow Clock cycles. This allows assertion duration to be programmed between 60 s and 2 seconds.
KEY: Password

154
20.
20.1
Shutdown Controller (SHDWC)

Description
The Shutdown Controller controls the power supplies VDDIO and VDDCORE and the wake-up detection on debounced input lines.
20.2
z
Shutdown and Wake-up Logic

z z
Software Assertion of the SHDW Output Pin Programmable De-assertion from the WKUP Input Pins
20.3
Block Diagram
Figure 20-1. Shutdown Controller Block Diagram

SLCK Shutdown Controller
SHDW_MR read SHDW_SR reset
CPTWK0 WKMODE0 WKUP0
WAKEUP0 SHDW_SR
set
read SHDW_SR
Wake-up
reset
RTCWKEN RTC Alarm
SHDW_MR
RTCWK
set
SHDW_SR SHDW_CR
Shutdown Output Controller Shutdown
SHDN
SHDW

155
20.4
I/O Lines Description
Table 20-1. I/O Lines Description Name WKUP0 SHDN Description Wake-up 0 input Shutdown output Type Input Output
20.5

The Shutdown Controller is continuously clocked by Slow Clock. The Power Management Controller has no effect on the behavior of the Shutdown Controller.
20.6
The Shutdown Controller manages the main power supply. To do so, it is supplied with VDDBU and manages wake-up input pins and one output pin, SHDN. A typical application connects the pin SHDN to the shutdown input of the DC/DC Converter providing the main power supplies of the system, and especially VDDCORE and/or VDDIO. The wake-up inputs (WKUP0) connect to any pushbuttons or signal that wake up the system. The software is able to control the pin SHDN by writing the Shutdown Control Register (SHDW_CR) with the bit SHDW at 1. The shutdown is taken into account only 2 slow clock cycles after the write of SHDW_CR. This register is passwordprotected and so the value written should contain the correct key for the command to be taken into account. As a result, the system should be powered down. A level change on WKUP0 is used as wake-up. Wake-up is configured in the Shutdown Mode Register (SHDW_MR). The transition detector can be programmed to detect either a positive or negative transition or any level change on WKUP0. The detection can also be disabled. Programming is performed by defining WKMODE0. Moreover, a debouncing circuit can be programmed for WKUP0. The debouncing circuit filters pulses on WKUP0 shorter than the programmed number of 16 SLCK cycles in CPTWK0 of the SHDW_MR register. If the programmed level change is detected on a pin, a counter starts. When the counter reaches the value programmed in the corresponding field, CPTWK0, the SHDN pin is released. If a new input change is detected before the counter reaches the corresponding value, the counter is stopped and cleared. WAKEUP0 of the Status Register (SHDW_SR) reports the detection of the programmed events on WKUP0 with a reset after the read of SHDW_SR. The Shutdown Controller can be programmed so as to activate the wake-up using the RTC alarm (the detection of the rising edge of the RTC alarm is synchronized with SLCK). This is done by writing the SHDW_MR register using the RTCWKEN field. When enabled, the detection of the RTC alarm is reported in the RTCWK bit of the SHDW_SR Status register. It is reset after the read of SHDW_SR. When using the RTC alarm to wake up the system, the user must ensure that the RTC alarm status flag is cleared before shutting down the system.Otherwise, no rising edge of the status flag may be detected and the wake-up fails fail.

156
20.7
Shutdown Controller (SHDWC) User Interface
Table 20-2. Register Mapping Offset 0x00 0x04 0x08 Register Shutdown Control Register Shutdown Mode Register Shutdown Status Register Name SHDW_CR SHDW_MR SHDW_SR Access Write-only Read-write Read-only Reset 0x0000_0000

157
20.7.1 Shutdown Control Register

31
SHDW_CR 0xFFFFFE10 Write-only

30 29 28 KEY 27 26 25 24
23 15 7
22 14 6
21 13 5
20 12 4
19 11 3
18 10 2
17 9 1
16 8 0 SHDW
SHDW: Shutdown Command

0 = No effect. 1 = If KEY is correct, asserts the SHDN pin.
KEY: Password

158
20.7.2 Shutdown Mode Register

31 23 15
SHDW_MR 0xFFFFFE14 Read/Write

30 22 14 7 6 CPTWK0 5 4 29 21 13 28 20 12 27 19 11 3 26 18 10 2 25 17 RTCWKEN 9 1 WKMODE0 0 24 16 8
WKMODE0: Wake-up Mode 0
WKMODE[1:0] 0 0 1 1 0 1 0 1
Wake-up Input Transition Selection None. No detection is performed on the wake-up input Low to high level High to low level Both levels change
CPTWK0: Counter on Wake-up 0

Defines the number of 16 Slow Clock cycles, the level detection on the corresponding input pin shall last before the wake-up event occurs. Because of the internal synchronization of WKUP0, the SHDN pin is released (CPTWK x 16 + 1) Slow Clock cycles after the event on WKUP.
RTCWKEN: Real-time Clock Wake-up Enable

0 = The RTC Alarm signal has no effect on the Shutdown Controller. 1 = The RTC Alarm signal forces the de-assertion of the SHDN pin.

159
20.7.3 Shutdown Status Register

31 23 15 7
SHDW_SR 0xFFFFFE18 Read-only

30 22 14 6 29 21 13 5 28 20 12 4 27 19 11 3 26 18 10 2 25 17 RTCWK 9 1 24 16 8 0 WAKEUP0
WAKEUP0: Wake-up 0 Status

0 = No wake-up event occurred on the corresponding wake-up input since the last read of SHDW_SR. 1 = At least one wake-up event occurred on the corresponding wake-up input since the last read of SHDW_SR.
RTCWK: Real-time Clock Wake-up

0 = No wake-up alarm from the RTC occurred since the last read of SHDW_SR. 1 = At least one wake-up alarm from the RTC occurred since the last read of SHDW_SR.

160
21.
21.1
General Purpose Backup Registers (GPBR)

Description
The System Controller embeds 4 general-purpose backup registers.
21.2
z
Four 32-bit General Purpose Backup Registers

161
21.3
General Purpose Backup Registers (GPBR) User Interface
Table 21-1. Register Mapping Offset 0x0 ... 0x6C Register General Purpose Backup Register 0 ... General Purpose Backup Register 3 Name SYS_GPBR0 ... SYS_GPBR3 Access Read-write ... Read-write Reset ...
21.3.1 General Purpose Backup Register x

31
SYS_GPBRx 0xFFFFFE60 Read-write

30 29 28 27 26 25 24
GPBR_VALUE
23 22 21 20 19 18 17 16
GPBR_VALUE
15 14 13 12 11 10 9 8
GPBR_VALUE
7 6 5 4 3 2 1 0
GPBR_VALUE
GPBR_VALUE: Value of GPBR x

162
22.
22.1
Real-time Clock (RTC)

Description
The Real-time Clock (RTC) peripheral is designed for very low power consumption. It combines a complete time-of-day clock with alarm and a two-hundred-year Gregorian calendar, complemented by a programmable periodic interrupt. The alarm and calendar registers are accessed by a 32-bit data bus. The time and calendar values are coded in binary-coded decimal (BCD) format. The time format can be 24-hour mode or 12-hour mode with an AM/PM indicator. Updating time and calendar fields and configuring the alarm fields are performed by a parallel capture on the 32-bit data bus. An entry control is performed to avoid loading registers with incompatible BCD format data or with an incompatible date according to the current month/year/century.
22.2
z z z z z
Ultra Low -power Consumption Full Asynchronous Design Gregorian Calendar up to 2099 Programmable Periodic Interrupt Valid Time and Date Programmation Check

163
22.3
Block Diagram
Figure 22-1. RTC Block Diagram
Slow Clock: SLCK
32768 Divider
Time
Date
Bus Interface
Bus Interface
Entry Control
Interrupt Control
RTC Interrupt

164
22.4

The Real-time Clock is continuously clocked at 32768 Hz. The Power Management Controller has no effect on RTC behavior.
22.4.2 Interrupt
Within the System Controller, the RTC interrupt is OR-wired with all the other module interrupts. Only one System Controller interrupt line is connected on one of the internal sources of the interrupt controller. RTC interrupt requires the interrupt controller to be programmed first. When a System Controller interrupt occurs, the service routine must first determine the cause of the interrupt. This is done by reading each status register of the System Controller peripherals successively.
Table 22-1. Peripheral IDs Instance RTC ID 1
22.5
The RTC provides a full binary-coded decimal (BCD) clock that includes century (19/20), year (with leap years), month, date, day, hours, minutes and seconds. The valid year range is 1900 to 2099 in Gregorian mode, a two-hundred-year calendar. The RTC can operate in 24-hour mode or in 12-hour mode with an AM/PM indicator. Corrections for leap years are included (all years divisible by 4 being leap years). This is correct up to the year 2099.
22.5.1 Reference Clock

The reference clock is Slow Clock (SLCK). It can be driven internally or by an external 32.768 kHz crystal. During low-power modes of the processor, the oscillator runs and power consumption is critical. The crystal selection has to take into account the current consumption for power saving and the frequency drift due to temperature effect on the circuit for time accuracy.
22.5.2 Timing
The RTC is updated in real time at one-second intervals in normal mode for the counters of seconds, at one-minute intervals for the counter of minutes and so on. Due to the asynchronous operation of the RTC with respect to the rest of the chip, to be certain that the value read in the RTC registers (century, year, month, date, day, hours, minutes, seconds) are valid and stable, it is necessary to read these registers twice. If the data is the same both times, then it is valid. Therefore, a minimum of two and a maximum of three accesses are required.
22.5.3 Alarm
The RTC has five programmable fields: month, date, hours, minutes and seconds. Each of these fields can be enabled or disabled to match the alarm condition:
z z
If all the fields are enabled, an alarm flag is generated (the corresponding flag is asserted and an interrupt generated if enabled) at a given month, date, hour/minute/second. If only the seconds field is enabled, then an alarm is generated every minute.
Depending on the combination of fields enabled, a large number of possibilities are available to the user ranging from minutes to 365/366 days.

165
22.5.4 Error Checking when Programming

Verification on user interface data is performed when accessing the century, year, month, date, day, hours, minutes, seconds and alarms. A check is performed on illegal BCD entries such as illegal date of the month with regard to the year and century configured. If one of the time fields is not correct, the data is not loaded into the register/counter and a flag is set in the validity register. The user can not reset this flag. It is reset as soon as an acceptable value is programmed. This avoids any further side effects in the hardware. The same procedure is done for the alarm. The following checks are performed: 1. 2. 3. 4. 5. 6. 7. 8. Note: Century (check if it is in range 19 - 20) Year (BCD entry check) Date (check range 01 - 31) Month (check if it is in BCD range 01 - 12, check validity regarding date) Day (check range 1 - 7) Hour (BCD checks: in 24-hour mode, check range 00 - 23 and check that AM/PM flag is not set if RTC is set in 24hour mode; in 12-hour mode check range 01 - 12) Minute (check BCD and range 00 - 59) Second (check BCD and range 00 - 59) If the 12-hour mode is selected by means of the RTC_MR register, a 12-hour value can be programmed and the returned value on RTC_TIMR will be the corresponding 24-hour value. The entry control checks the value of the AM/PM indicator (bit 22 of RTC_TIMR register) to determine the range to be checked.
22.5.5 Updating Time/Calendar

To update any of the time/calendar fields, the user must first stop the RTC by setting the corresponding field in the Control Register. Bit UPDTIM must be set to update time fields (hour, minute, second) and bit UPDCAL must be set to update calendar fields (century, year, month, date, day). Then the user must poll or wait for the interrupt (if enabled) of bit ACKUPD in the Status Register. Once the bit reads 1, it is mandatory to clear this flag by writing the corresponding bit in RTC_SCCR. The user can now write to the appropriate Time and Calendar register. Once the update is finished, the user must reset (0) UPDTIM and/or UPDCAL in the Control When entering programming mode of the calendar fields, the time fields remain enabled. When entering the programming mode of the time fields, both time and calendar fields are stopped. This is due to the location of the calendar logic circuity (downstream for low-power considerations). It is highly recommended to prepare all the fields to be updated before entering programming mode. In successive update operations, the user must wait at least one second after resetting the UPDTIM/UPDCAL bit in the RTC_CR (Control Register) before setting these bits again. This is done by waiting for the SEC flag in the Status Register before setting UPDTIM/UPDCAL bit. After resetting UPDTIM/UPDCAL, the SEC flag must also be cleared.

166
Figure 22-2. Update Sequence
Begin
Prepare TIme or Calendar Fields
Set UPDTIM and/or UPDCAL bit(s) in RTC_CR
Read RTC_SR Polling or IRQ (if enabled)
ACKUPD =1?
No
Yes
Clear ACKUPD bit in RTC_SCCR
Update Time and/or Calendar values in RTC_TIMR/RTC_CALR
Clear UPDTIM and/or UPDCAL bit in RTC_CR
End

167
22.6
Real-time Clock (RTC) User Interface

Register Mapping Register Control Register Mode Register Time Register Calendar Register Time Alarm Register Calendar Alarm Register Status Register Status Clear Command Register Interrupt Enable Register Interrupt Disable Register Interrupt Mask Register Valid Entry Register Reserved Register Reserved Register Reserved Register Name RTC_CR RTC_MR RTC_TIMR RTC_CALR RTC_TIMALR RTC_CALALR RTC_SR RTC_SCCR RTC_IER RTC_IDR RTC_IMR RTC_VER Access Read-write Read-write Read-write Read-write Read-write Read-write Read-only Write-only Write-only Write-only Read-only Read-only Reset 0x0 0x0 0x0 0x01810720 0x0 0x01010000 0x0 0x0 0x0
Table 22-2. Offset 0x00 0x04 0x08 0x0C 0x10 0x14 0x18 0x1C 0x20 0x24 0x28 0x2C
0x300xC4 0xC80xF8 0xFC
Note:
If an offset is not listed in the table it must be considered as reserved.

168
22.6.1 RTC Control Register

31
RTC_CR 0xFFFFFEB0 Read-write

30 29 28 27 26 25 24
23
22
21
20
19
18
17
16
15
14
13
12
11
10 9
CALEVSEL
8
2 1
TIMEVSEL
0
UPDCAL
UPDTIM
UPDTIM: Update Request Time Register

0 = No effect. 1 = Stops the RTC time counting. Time counting consists of second, minute and hour counters. Time counters can be programmed once this bit is set and acknowledged by the bit ACKUPD of the Status Register.
UPDCAL: Update Request Calendar Register

0 = No effect. 1 = Stops the RTC calendar counting. Calendar counting consists of day, date, month, year and century counters. Calendar counters can be programmed once this bit is set.
TIMEVSEL: Time Event Selection

The event that generates the flag TIMEV in RTC_SR (Status Register) depends on the value of TIMEVSEL.
Value 0 1 2 3 Name MINUTE HOUR MIDNIGHT NOON Description Minute change Hour change Every day at midnight Every day at noon
CALEVSEL: Calendar Event Selection

The event that generates the flag CALEV in RTC_SR depends on the value of CALEVSEL
Value 0 1 2 3 Name WEEK MONTH YEAR Description Week change (every Monday at time 00:00:00) Month change (every 01 of each month at time 00:00:00) Year change (every January 1 at time 00:00:00)

169
22.6.2 RTC Mode Register

31
RTC_MR 0xFFFFFEB4 Read-write

30 29 28 27 26 25 24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
HRMOD
HRMOD: 12-/24-hour Mode

0 = 24-hour mode is selected. 1 = 12-hour mode is selected.

170
22.6.3 RTC Time Register

31
RTC_TIMR 0xFFFFFEB8 Read-write

30 29 28 27 26 25 24
23
22
21
20
19
18
17
16
15
AMPM
14 13 12 11
HOUR
10 9 8
7 6 5 4
MIN
3 2 1 0
SEC
SEC: Current Second

The range that can be set is 0 - 59 (BCD). The lowest four bits encode the units. The higher bits encode the tens.
MIN: Current Minute

HOUR: Current Hour

The range that can be set is 1 - 12 (BCD) in 12-hour mode or 0 - 23 (BCD) in 24-hour mode.
AMPM: Ante Meridiem Post Meridiem Indicator

This bit is the AM/PM indicator in 12-hour mode. 0 = AM 1 = PM
All non-significant bits read zero.

171
22.6.4 RTC Calendar Register

31
RTC_CALR 0xFFFFFEBC Read-write

30 29 28 27 26 25 24
23
22 21 20 19
DATE
18 17 16
DAY
15 14 13 12 11
MONTH
10 9 8
YEAR
7 6 5 4 3 2 1 0
CENT
CENT: Current Century

YEAR: Current Year

MONTH: Current Month

DAY: Current Day in Current Week

The range that can be set is 1 - 7 (BCD). The coding of the number (which number represents which day) is user-defined as it has no effect on the date counter.
DATE: Current Day in Current Month

All non-significant bits read zero.

172
22.6.5 RTC Time Alarm Register

31
RTC_TIMALR 0xFFFFFEC0 Read-write

30 29 28 27 26 25 24
23
22
21
20
19
18
17
16
HOUREN
15
AMPM
14 13 12 11
HOUR
10 9 8
MINEN
7 6 5 4
MIN
3 2 1 0
SECEN
SEC
SEC: Second Alarm

This field is the alarm field corresponding to the BCD-coded second counter.
SECEN: Second Alarm Enable

0 = The second-matching alarm is disabled. 1 = The second-matching alarm is enabled.
MIN: Minute Alarm

This field is the alarm field corresponding to the BCD-coded minute counter.
MINEN: Minute Alarm Enable

0 = The minute-matching alarm is disabled. 1 = The minute-matching alarm is enabled.
HOUR: Hour Alarm

This field is the alarm field corresponding to the BCD-coded hour counter.
AMPM: AM/PM Indicator

This field is the alarm field corresponding to the BCD-coded hour counter.
HOUREN: Hour Alarm Enable

0 = The hour-matching alarm is disabled. 1 = The hour-matching alarm is enabled.

173
22.6.6 RTC Calendar Alarm Register

31
RTC_CALALR 0xFFFFFEC4 Read-write

30 29 28 27 26 25 24
DATEEN
23
22 21 20 19
DATE
18 17 16
MTHEN
15
14
13 12 11
MONTH
10 9 8
MONTH: Month Alarm

This field is the alarm field corresponding to the BCD-coded month counter.
MTHEN: Month Alarm Enable

0 = The month-matching alarm is disabled. 1 = The month-matching alarm is enabled.
DATE: Date Alarm

This field is the alarm field corresponding to the BCD-coded date counter.
DATEEN: Date Alarm Enable

0 = The date-matching alarm is disabled. 1 = The date-matching alarm is enabled.

174
22.6.7 RTC Status Register

31
RTC_SR 0xFFFFFEC8 Read-only

30 29 28 27 26 25 24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
CALEV
TIMEV
SEC
ALARM
ACKUPD
ACKUPD: Acknowledge for Update

0 (FREERUN) = Time and calendar registers cannot be updated. 1 (UPDATE) = Time and calendar registers can be updated.
ALARM: Alarm Flag

0 (NO_ALARMEVENT) = No alarm matching condition occurred. 1 (ALARMEVENT) = An alarm matching condition has occurred.
SEC: Second Event

0 (NO_SECEVENT) = No second event has occurred since the last clear. 1 (SECEVENT) = At least one second event has occurred since the last clear.
TIMEV: Time Event

0 (NO_TIMEVENT) = No time event has occurred since the last clear. 1 (TIMEVENT) = At least one time event has occurred since the last clear. The time event is selected in the TIMEVSEL field in RTC_CR (Control Register) and can be any one of the following events: minute change, hour change, noon, midnight (day change).
CALEV: Calendar Event

0 (NO_CALEVENT) = No calendar event has occurred since the last clear. 1 (CALEVENT) = At least one calendar event has occurred since the last clear. The calendar event is selected in the CALEVSEL field in RTC_CR and can be any one of the following events: week change, month change and year change.

175
22.6.8 RTC Status Clear Command Register

31
RTC_SCCR 0xFFFFFECC Write-only

30 29 28 27 26 25 24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
CALCLR
TIMCLR
SECCLR
ALRCLR
ACKCLR
ACKCLR: Acknowledge Clear

0 = No effect. 1 = Clears corresponding status flag in the Status Register (RTC_SR).
ALRCLR: Alarm Clear

SECCLR: Second Clear

TIMCLR: Time Clear

CALCLR: Calendar Clear


176
22.6.9 RTC Interrupt Enable Register

31
RTC_IER 0xFFFFFED0 Write-only

30 29 28 27 26 25 24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
CALEN
TIMEN
SECEN
ALREN
ACKEN
ACKEN: Acknowledge Update Interrupt Enable

0 = No effect. 1 = The acknowledge for update interrupt is enabled.
ALREN: Alarm Interrupt Enable

0 = No effect. 1 = The alarm interrupt is enabled.
SECEN: Second Event Interrupt Enable

0 = No effect. 1 = The second periodic interrupt is enabled.
TIMEN: Time Event Interrupt Enable

0 = No effect. 1 = The selected time event interrupt is enabled.
CALEN: Calendar Event Interrupt Enable

0 = No effect. 1 = The selected calendar event interrupt is enabled.

177
22.6.10 RTC Interrupt Disable Register

31
RTC_IDR 0xFFFFFED4 Write-only

30 29 28 27 26 25 24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
CALDIS
TIMDIS
SECDIS
ALRDIS
ACKDIS
ACKDIS: Acknowledge Update Interrupt Disable

0 = No effect. 1 = The acknowledge for update interrupt is disabled.
ALRDIS: Alarm Interrupt Disable

0 = No effect. 1 = The alarm interrupt is disabled.
SECDIS: Second Event Interrupt Disable

0 = No effect. 1 = The second periodic interrupt is disabled.
TIMDIS: Time Event Interrupt Disable

0 = No effect. 1 = The selected time event interrupt is disabled.
CALDIS: Calendar Event Interrupt Disable

0 = No effect. 1 = The selected calendar event interrupt is disabled.

178
22.6.11 RTC Interrupt Mask Register

31
RTC_IMR 0xFFFFFED8 Read-only

30 29 28 27 26 25 24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
CAL
TIM
SEC
ALR
ACK
ACK: Acknowledge Update Interrupt Mask

0 = The acknowledge for update interrupt is disabled. 1 = The acknowledge for update interrupt is enabled.
ALR: Alarm Interrupt Mask

0 = The alarm interrupt is disabled. 1 = The alarm interrupt is enabled.
SEC: Second Event Interrupt Mask

0 = The second periodic interrupt is disabled. 1 = The second periodic interrupt is enabled.
TIM: Time Event Interrupt Mask

0 = The selected time event interrupt is disabled. 1 = The selected time event interrupt is enabled.
CAL: Calendar Event Interrupt Mask

0 = The selected calendar event interrupt is disabled. 1 = The selected calendar event interrupt is enabled.

179
22.6.12 RTC Valid Entry Register

31
RTC_VER 0xFFFFFEDC Read-only

30 29 28 27 26 25 24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
NVCALALR
NVTIMALR
NVCAL
NVTIM
NVTIM: Non-valid Time

0 = No invalid data has been detected in RTC_TIMR (Time Register). 1 = RTC_TIMR has contained invalid data since it was last programmed.
NVCAL: Non-valid Calendar

0 = No invalid data has been detected in RTC_CALR (Calendar Register). 1 = RTC_CALR has contained invalid data since it was last programmed.
NVTIMALR: Non-valid Time Alarm

0 = No invalid data has been detected in RTC_TIMALR (Time Alarm Register). 1 = RTC_TIMALR has contained invalid data since it was last programmed.
NVCALALR: Non-valid Calendar Alarm

0 = No invalid data has been detected in RTC_CALALR (Calendar Alarm Register). 1 = RTC_CALALR has contained invalid data since it was last programmed.

180
23.
23.1
Slow Clock Controller (SCKC)

Description
The System Controller embeds a Slow Clock Controller. The slow clock can be generated either by an external 32768 Hz crystal oscillator or by the on-chip 32 kHz RC oscillator. The 32768 Hz crystal oscillator can be bypassed by setting the bit OSC32BYP to accept an external slow clock on XIN32. The internal 32 kHz RC oscillator and the 32768 Hz oscillator can be enabled by setting to 1, respectively, RCEN bit and OSC32EN bit in the System Controller user interface. The OSCSEL command selects the slow clock source.
23.2
z z
32 kHz RC Oscillator or 32768 Hz Crystal Oscillator Selector VDDBU Powered

181
23.3
Block Diagram
RCEN
On Chip RC OSC Slow Clock SLCK XIN32 XOUT32 Slow Clock Oscillator OSCSEL OSC32EN OSC32BYP
RCEN, OSC32EN, OSCSEL and OSC32BYP bits are located in the Slow Clock Configuration Register (SCKC_CR) located at the address 0xFFFFFE50 in the backed up part of the System Controller and, thus, they are preserved while VDDBU is present. After a VDDBU power on reset, the default configuration is RCEN = 1, OSC32EN = 0 and OSCSEL = 0, allowing the system to start on the internal 32 kHz RC oscillator. The programmer controls the slow clock switching by software and so must take precautions during the switching phase.

182
23.3.1 Switch from Internal 32 kHz RC Oscillator to 32768 Hz Crystal Oscillator

To switch from the internal 32 kHz RC oscillator to the 32768 Hz crystal oscillator, the programmer must execute the following sequence:
z z z z z z
Switch the master clock to a source different from slow clock (PLL or Main Oscillator) through the Power Management Controller. Enable the 32768 Hz oscillator by setting the bit OSC32EN to 1. Wait 32768 Hz Startup Time for clock stabilization (software loop). Switch from internal 32 kHz RC oscillator to 32768 Hz oscillator by setting the bit OSCSEL to 1. Wait 5 slow clock cycles for internal resynchronization. Disable the 32 kHz RC oscillator by setting the bit RCEN to 0.
23.3.2 Bypass the 32768 Hz Oscillator

The following steps must be added to bypass the 32768 Hz oscillator:
z z z
An external clock must be connected on XIN32. Enable the bypass path OSC32BYP bit set to 1. Disable the 32768 Hz oscillator by setting the bit OSC32EN to 0.
23.3.3 Switch from 32768 Hz Crystal Oscillator to Internal 32 kHz RC Oscillator

The same procedure must be followed to switch from the 32768 Hz crystal oscillator to the internal 32 kHz RC oscillator:
z z z z z z
Switch the master clock to a source different from slow clock (PLL or Main Oscillator). Enable the internal 32 kHz RC oscillator for low power by setting the bit RCEN to 1 Wait internal 32 kHz RC Startup Time for clock stabilization (software loop). Switch from 32768 Hz oscillator to internal RC by setting the bit OSCSEL to 0. Wait 5 slow clock cycles for internal resynchronization. Disable the 32768 Hz oscillator by setting the bit OSC32EN to 0.

183
23.4
Slow Clock Configuration (SCKC) User Interface
Table 23-1. Register Mapping Offset 0x0 Register Slow Clock Configuration Register Name SCKC_CR Access Read-write Reset 0x0000_0001
23.4.1 Slow Clock Configuration Register

31 23 15 7
SCKC_CR 0xFFFFFE50 Read-write

30 22 14 6 29 21 13 5 28 20 12 4 27 19 11 3 OSCSEL 26 18 10 2 OSC32BYP 25 17 9 1 OSC32EN 24 16 8 0 RCEN
RCEN: Internal 32 kHz RC Oscillator

0: 32 kHz RC oscillator is disabled. 1: 32 kHz RC oscillator is enabled.
OSC32EN: 32768 Hz Oscillator

0: 32768 Hz oscillator is disabled. 1: 32768 Hz oscillator is enabled.
OSC32BYP: 32768 Hz Oscillator Bypass

0: 32768 Hz oscillator is not bypassed. 1: 32768 Hz oscillator is bypassed, accept an external slow clock on XIN32.
OSCSEL: Slow Clock Selector

0 (RC): Slow clock is internal 32 kHz RC oscillator. 1 (XTAL): Slow clock is 32768 Hz oscillator.

184
24.
24.1
Clock Generator
Description
The Clock Generator User Interface is embedded within the Power Management Controller and is described in Section 25.13 Power Management Controller (PMC) User Interface. However, the Clock Generator registers are named CKGR_.
24.2
Embedded characteristics
The Clock Generator is made up of:
z z z z z z
Low-power 32768 Hz Slow Clock Oscillator with bypass mode Low-power RC Oscillator 8 to 48 MHz Crystal Oscillator or a 24/48 MHz XRCGB Crystal Resonator, which can be bypassed (12 MHz, 24 MHz (preferred) or 48 MHz must be used in case of USB operations) Fast RC Oscillator, at 12 MHz 480 MHz UTMI PLL providing a clock for the USB High Speed Device Controller 400 to 1000 MHz programmable PLL (input from 8 to 50 MHz), capable of providing the clock MCK to the processor and to the peripherals SLCK, the Slow Clock, which is the only permanent clock within the system MAINCK is the output of the Main Clock Oscillator selection: either Crystal Oscillator or 12 MHz Fast RC Oscillator PLLACK is the output of the Divider and 400 to 1000 MHz programmable PLL (PLLA) UPLLCK is the output of the 480 MHz UTMI PLL (UPLL) SMDCK is the Software Modem Clock Main crystal oscillator clock failure detector Frequency counter on main clock and an on-the-fly adjustable main RC oscillator frequency
It provides the following clocks:

z z z z z
The Power Management Controller also provides the following operations on clocks:
z z

185
24.3
Block Diagram
Figure 24-1. Clock Generator Block Diagram
Clock Generator RCEN On Chip 32K RC OSC XIN32 XOUT32 Slow Clock Oscillator
Slow Clock SLCK OSCSEL OSC32EN OSC32BYP MOSCRCEN MOSCSEL
On Chip 12M RC OSC XIN XOUT

8 to 48 MHz Main Oscillator
Main Clock MAINCK
UPLL
UPLLCK
PLLA and Divider
PLLA Clock PLLACK
Status
Control
Power Management Controller

186
24.4
Main Clock Selection

The SAMA5D3 main clock can be generated either by an external 12 MHz crystal, an XRCGB crystal resonator or by the on-chip 12 MHz RC oscillator. This fast RC allows the processor to start or restart in a few microseconds when 12 MHz internal RC oscillator is selected. The 12 MHz crystal oscillator can be bypassed by setting the MOSCXTBY bit to accept an external main clock on XIN.
Figure 24-2. Main Clock Selection
MOSCRCEN
On Chip 12 MHz RC Oscillator Main Clock
XIN XOUT
Main Clock Oscillator
MOSCSEL SMDCK MOSCXTEN MOSCXTBY
MOSCRCEN, MOSCXTEN, MOSCSEL and MOSCXTBY bits are located in the PMC Clock Generator Main Oscillator Register (CKGR_MOR). After a VDDBU power on reset, the default configuration is MOSCRCEN = 1, MOSCXTEN = 0 and MOSCSEL = 0, the 12 MHz RC is started as Main clock.
24.4.1 Fast wake-up

To speed up the wake-up phase, the user can switch the system clock from 32 kHz RC (SLCK) to 12 MHz RC (Main Clock). This enables the user to perform system configuration (PLL, DDR, etc.) at 12 MHz instead of 32 kHz during 12 MHz Oscillator start-up.
Figure 24-3. PMC Startup
12 MHz RC
External Main clock
Main Supply 12 MHz RC Startup Time POR output

Crystal Startup Time System starts on 32 kHz RC Wait MOSCRCS = 1 RCEN = 1 User switch on Main Clock OSC32EN = 0 to speed-up the boot OSCSEL = 0 PMC_MCKR =1 MOSCRCEN = 1 System is running at 12 MHz MOSCXTEN = 0 MOSCSEL = 0 External oscillator PMC_MCKR = 0 is started for better accuracy MOSCXTEN = 1 MOSCSEL = 0 Wait MOSCXTS = 1 User switches on external oscillator MOSCSEL = 1 Wait while MOSCSELS = 1 System is runnning on 12 MHz Crystal PLL can be used

187
24.4.2 Switch from Internal 12 MHz RC Oscillator to the 12 MHz Crystal

The programmer controls the main clock switching by software and must take precautions during the switching phase. To switch from internal 12 MHz RC oscillator to the 12 MHz crystal, the programmer must execute the following sequence:
z z z z z
Enable the 12 MHz oscillator by setting the bit MOSCXTEN to 1. Wait for 12 MHz oscillator status MAINRDY is 1. Switch from internal 12 MHz RC to the 12 MHz oscillator by setting the bit MOSCSEL to 1. If not the bit MOSCSEL is set to 0 by the PMC. Disable the 12 MHz RC oscillator by setting the bit MOSCRCEN to 0.
24.4.3 Bypass the 12 MHz Crystal Oscillator

The following step must be added to bypass the 12 MHz crystal oscillator.
z z z
An external clock must be connected on XIN. Enable the bypass path MOSCXTBY bit set to 1. Disable the 12 MHz oscillator by setting the bit MOSCXTEN to 0.
24.4.4 Switch from the 12 MHz Crystal Oscillator to internal 12 MHz RC Oscillator
The same procedure must be followed to switch from a 12 MHz crystal oscillator to the internal 12 MHz RC oscillator.
z z z z
Enable the internal 12 MHz RC oscillator for low power by setting the bit MOSCRCEN to 1 Wait internal 12 MHz RC Startup Time for clock stabilization (software loop). Switch from 12 MHz oscillator to internal 12 MHz RC by setting the bit MOSCSEL to 0. Disable the 12 MHz oscillator by setting the bit MOSCXTEN to 0.

188
24.5
Main Clock
Figure 24-4. Main Clock Block Diagram
MOSCRCEN MOSCRCF MOSCRCS 12 MHz Fast RC Oscillator MOSCSEL MOSCSELS
1
MOSCXTEN
MAINCK Main Clock 0
XIN XOUT
8 to 48 MHz Crystal Oscillator
MOSCXTCNT
SLCK Slow Clock
8 to 48 MHz Crystal Oscillator Counter
MOSCXTS
MOSCRCEN MOSCXTEN MOSCSEL
Main Clock Frequency Counter
MAINF MAINRDY
The Main Clock has two sources:

z z
12 MHz Fast RC Oscillator which starts very quickly and is used at startup 8 to 48 MHz Crystal Oscillator, which can be bypassed
24.5.1 12 MHz Fast RC Oscillator

After reset, the 12 MHz Fast RC Oscillator is enabled and it is selected as the source of MCK. MCK is the default clock selected to start up the system. Please refer to the DC Characteristics section of the product datasheet. The software can disable or enable the 12 MHz Fast RC Oscillator with the MOSCRCEN bit in the Clock Generator Main Oscillator Register (CKGR_MOR). When disabling the Main Clock by clearing the MOSCRCEN bit in CKGR_MOR, the MOSCRCS bit in the Power Management Controller Status Register (PMC_SR) is automatically cleared, indicating the Main Clock is off. Setting the MOSCRCS bit in the Power Management Controller Interrupt Enable Register (PMC_IER) can trigger an interrupt to the processor.

189
24.5.2 12 MHz Fast RC Oscillator Clock Frequency Adjustment

It is possible for the user to adjust the main RC oscillator frequency through PMC_OCR. By default, SEL is low, so the RC oscillator will be driven with Fuse calibration bits which are programmed during chip production. The user can adjust the trimming of the 12 MHz Fast RC oscillator through this register in order to obtain more accurate frequency (to compensate derating factors such as temperature and voltage). In order to calibrate the 12 MHz oscillator frequency, SEL must be set to 1 and a good frequency value must be configured in CAL. It is possible to restart, at anytime, a measurement of the main frequency by means of the RCMEAS bit in the Main Clock Frequency Register (CKGR_MCFR). Thus, when the MAINFRDY flag is set, the MAINF field returns the frequency of the main clock and software can calculate the error with an expected frequency and correct the CAL field accordingly. This may be used to compensate frequency drift due to derating factors such as temperature and/or voltage.
24.5.3 8 to 48 MHz Crystal Oscillator

After reset, the 8 to 48 MHz Crystal Oscillator is disabled and it is not selected as the source of MAINCK. The user can select the 8 to 48 MHz crystal oscillator to be the source of MAINCK, as it provides a more accurate frequency. The software enables or disables the main oscillator so as to reduce power consumption by clearing the MOSCXTEN bit in the Main Oscillator Register (CKGR_MOR). When disabling the main oscillator by clearing the MOSCXTEN bit in CKGR_MOR, the MOSCXTS bit in PMC_SR is automatically cleared, indicating the Main Clock is off. When enabling the main oscillator, the user must initiate the main oscillator counter with a value corresponding to the startup time of the oscillator. This startup time depends on the crystal frequency connected to the oscillator. When the MOSCXTEN bit and the MOSCXTCNT are written in CKGR_MOR to enable the main oscillator, the MOSCXTS bit in the Power Management Controller Status Register (PMC_SR) is cleared and the counter starts counting down on the slow clock divided by 8 from the MOSCXTCNT value. Since the MOSCXTCNT value is coded with 8 bits, the maximum startup time is about 62 ms. When the counter reaches 0, the MOSCXTS bit is set, indicating that the main clock is valid. Setting the MOSCXTS bit in PMC_IMR can trigger an interrupt to the processor.
24.5.4 Main Clock Oscillator Selection

The user can select either the 12 MHz Fast RC Oscillator or the Crystal Oscillator to be the source of Main Clock. The advantage of the 12 MHz Fast RC Oscillator is to have fast startup time, this is why it is selected by default (to start up the system) and when entering in Wait Mode. The advantage of the Crystal Oscillator is that it is very accurate. The selection is made by writing the MOSCSEL bit in the Main Oscillator Register (CKGR_MOR). The switch of the Main Clock source is glitch free, so there is no need to run out of SLCK, PLLACK or PLLBCK in order to change the selection. The MOSCSELS bit of the Power Management Controller Status Register (PMC_SR) allows knowing when the switch sequence is done. Setting the MOSCSELS bit in PMC_IMR can trigger an interrupt to the processor.
24.5.5 Software Sequence to Detect the Presence of Fast Crystal

The frequency meter carried on the CKGR_MCFR register is operating on the selected main clock and not on the fast crystal clock nor on the fast RC Oscillator clock. Therefore, to check for the presence of the fast crystal clock, it is necessary to switch the main clock on the fast crystal clock.

190
The following software sequence order must be followed:

z z z z z z z
MCK must select the slow clock (CSS=0 in PLL_MCKR register). Wait for the MCKRDY flag in the PLL_SR register to be 1. The fast crystal must be enabled by programming 1 in the MOSCXTEN field in the CKGR_MOR register, with the MOSCXTST field being programmed to the appropriate value (see the electrical characteristics chapter). Wait for the MOSCXTS flag to be 1 in the PLL_SR register to get the end of startup period of the fast crystal oscillator. Then, MOSCSEL must be programmed to 1 in the CKGR_MOR register to select the fast main crystal oscillator for the main clock. The MOSCSEL must be read until its value equals 1. Then the MOSCSELS status flag must be checked in the PLL_SR register. If MOSCSELS = 1, there is a valid crystal connected and its frequency can be determined by initiating a frequency measure by programming RCMEAS in the CKGR_MCFR register. If MOSCSELS = 0, there is no fast crystal clock (either no crystal connected or a crystal clock out of specification). A frequency measure can reinforce this status by initiating a frequency measure by programming RCMEAS in the CKGR_MCFR register. If MOSCSELS=0, the selection of the main clock must be programmed back to the main RC oscillator by writing MOSCSEL to 0 prior to disabling the fast crystal oscillator. If MOSCSELS=0, the crystal oscillator can be disabled (MOSCXTEN=0 in the CKGR_MOR register).
At this point, 2 cases may occur (either MOSCSELS = 0 or MOSCSELS = 1).

z z
z z
24.5.6 Main Clock Frequency Counter

The device features a Main Clock frequency counter that provides the frequency of the Main Clock. The Main Clock frequency counter is reset and starts incrementing at the Main Clock speed after the next rising edge of the Slow Clock in the following cases:
z z z
When the 12 MHz Fast RC Oscillator clock is selected as the source of Main Clock and when this oscillator becomes stable (i.e., when the MOSCRCS bit is set) When the Crystal Oscillator is selected as the source of Main Clock and when this oscillator becomes stable (i.e., when the MOSCXTS bit is set) When the Main Clock Oscillator selection is modified
Then, at the 16th falling edge of Slow Clock, the MAINFRDY bit in the Clock Generator Main Clock Frequency Register (CKGR_MCFR) is set and the counter stops counting. Its value can be read in the MAINF field of CKGR_MCFR and gives the number of Main Clock cycles during 16 periods of Slow Clock, so that the frequency of the 12 MHz Fast RC Oscillator or the Crystal Oscillator can be determined.

191
24.6
Divider and PLLA Block

The PLLA embeds an input divider to increase the accuracy of the resulting clock signals. However, the user must respect the PLLA minimum input frequency when programming the divider. Figure 24-5 shows the block diagram of the divider and PLLA block.
Figure 24-5. Divider and PLLA Block Diagram
DIVA MULA OUTA
MAINCK
Divider
PLLA
PLLACK
PLLACOUNT SLCK
PLLA Counter
LOCKA
24.6.1 Divider and Phase Lock Loop Programming

The divider can be set between 1 and 255 in steps of 1. When a divider field (DIV) is set to 0, the output of the corresponding divider and the PLL output is a continuous signal at level 0. On reset, each DIV field is set to 0, thus the corresponding PLL input clock is set to 0. The PLLA allows multiplication of the dividers outputs. The PLLA clock signal has a frequency that depends on the respective source signal frequency and on the parameters DIVA and MULA. The factor applied to the source signal frequency is (MULA + 1)/DIVA. When MULA is written to 0, the PLLA is disabled and its power consumption is saved. Re-enabling the PLLA can be performed by writing a value higher than 0 in the MUL field. Whenever the PLLA is re-enabled or one of its parameters is changed, the LOCKA bit in PMC_SR is automatically cleared. The values written in the PLLACOUNT field in CKGR_PLLAR are loaded in the PLLA counter. The PLLA counter then decrements at the speed of the Slow Clock until it reaches 0. At this time, the LOCK bit is set in PMC_SR and can trigger an interrupt to the processor. The user has to load the number of Slow Clock cycles required to cover the PLLA transient time into the PLLACOUNT field. The PLLA clock must be divided by 2 by writing the PLLADIV2 bit in the PMC_MCKR register.

192
24.7
UTMI Phase Lock Loop Programming

The source clock of the UTMI PLL is the Main OSC output. When the 12 MHz Fast RC Oscillator is selected as the source of the MAINCK, the 12 MHz frequency must also be selected because the UTMI PLL multiplier contains a built-in multiplier of x 40 to obtain the USB High Speed 480 MHz. A 12 MHz crystal is needed to use the USB.
Figure 24-6. UTMI PLL Block Diagram
UPLLEN
MAINCK
UTMI PLL
UPLLCK
UPLLCOUNT SLCK
UTMI PLL Counter
LOCKU
Whenever the UTMI PLL is enabled by writing UPLLEN in CKGR_UCKR, the LOCKU bit in PMC_SR is automatically cleared. The values written in the PLLCOUNT field in CKGR_UCKR are loaded in the UTMI PLL counter. The UTMI PLL counter then decrements at the speed of the Slow Clock divided by 8 until it reaches 0. At this time, the LOCKU bit is set in PMC_SR and can trigger an interrupt to the processor. The user has to load the number of Slow Clock cycles required to cover the UTMI PLL transient time into the PLLCOUNT field.

193
25.
25.1
Power Management Controller (PMC)

Overview
The Power Management Controller (PMC) optimizes power consumption by controlling all system and user peripheral clocks. The PMC enables/disables the clock inputs to many of the peripherals and the Core. The Power Management Controller provides the following clocks:
z z z z z
MCK, the Master Clock, programmable from a few hundred Hz to the maximum operating frequency of the device. It is available to the modules running permanently. Processor Clock (PCK), must be switched off when entering the processor in Sleep Mode. The USB Device HS Clock (UDPCK) The Software Modem Clock (SMDCK) Peripheral Clocks, typically MCK, provided to the embedded peripherals (USART, SSC, SPI, TWI, TC, HSMCI, etc.) and independently controllable. In order to reduce the number of clock names in a product, the Peripheral Clocks are named MCK in the product datasheet. Programmable Clock Outputs can be selected from the clocks provided by the clock generator and driven on the PCKx pins.
25.2
Master Clock Controller

The Master Clock Controller provides selection and division of the Master Clock (MCK). MCK is the clock provided to all the peripherals and the memory controller. The Master Clock is selected from one of the clocks provided by the Clock Generator. Selecting the Slow Clock provides a Slow Clock signal to the whole device. Selecting the Main Clock saves power consumption of the PLLs. The Master Clock Controller is made up of a clock selector and a prescaler. It also contains a Master Clock divider which allows the processor clock to be faster than the Master Clock. The Master Clock selection is made by writing the CSS field (Clock Source Selection) in PMC_MCKR (Master Clock Register). The prescaler supports the division by a power of 2 of the selected clock between 1 and 64, and the division by 6. The PRES field in PMC_MCKR programs the prescaler. Each time PMC_MCKR is written to define a new Master Clock, the MCKRDY bit is cleared in PMC_SR. It reads 0 until the Master Clock is established. Then, the MCKRDY bit is set and can trigger an interrupt to the processor. This feature is useful when switching from a high-speed clock to a lower one to inform the software when the change is actually done.
Figure 25-1. Master Clock Controller
PMC_MCKR CSS SLCK MAINCK PLLACK UPLLCK To the Processor Clock Controller (PCK) Master Clock Prescaler MCK PMC_MCKR PRES

194
25.3
Block Diagram
Figure 25-2. General Clock Block Diagram

PLLACK
USBS USBDIV+1 /4 UHP48M UHP12M USB OHCI USB EHCI
UPLLCK
Processor Clock Controller
PCK int
Divider X /1 /1.5 /2 MAINCK SLCK

Prescaler /1,/2,/3,/4,...,/64
/2
DDRCK 2x MCK MCK
/1 /2
/3
/4 Peripherals Clock Controller ON/OFF
Master Clock Controller
Divider
Periph_clk[..]
SLCK MAINCK UPLLCK
ON/OFF Prescaler /1,/2,/4,...,/64 pck[..]
Programmable Clock Controller

195
25.4
Processor Clock Controller

The PMC features a Processor Clock Controller (PCK) that implements the Processor Idle Mode. The Processor Clock can be disabled by writing the System Clock Disable Register (PMC_SCDR). The status of this clock (at least for debug purpose) can be read in the System Clock Status Register (PMC_SCSR). The Processor Clock PCK is enabled after a reset and is automatically re-enabled by any enabled interrupt. The Processor Idle Mode is achieved by disabling the Processor Clock, which is automatically re-enabled by any enabled fast or normal interrupt, or by the reset of the product. When the Processor Clock is disabled, the current instruction is finished before the clock is stopped, but this does not prevent data transfers from other masters of the system bus.
25.5
USB Device and Host Clocks

The USB Device and Host High Speed ports clocks are controlled by the UDPHS and UHPHS bits in PMC_PCER. To save power on this peripheral when they are is not used, the user can set these bits in PMC_PCDR. The UDPHS and UHPHS bits PMC_PCSR gives the activity of these clocks. The PMC also provides the clocks UHP48M and UHP12M to the USB Host OHCI. The USB Host OHCI clocks are controlled by the UHP bit in PMC_SCER. To save power on this peripheral when it is not used, the user can set the UHP bit in PMC_SCDR. The UHP bit in PMC_SCSR gives the activity of this clock. The USB host OHCI requires both the 12/48 MHz signal and the Master Clock. USBDIV field in PMC_USB register is to be programmed to 9 (division by 10) for normal operations. To save more power consumption user can stop UTMI PLL, in this case USB high-speed operations are not possible. Nevertheless, as the USB OHCI Input clock can be selected with USBS bit (PLLA or UTMI PLL) in PMC_USB register, OHCI full-speed operation remain possible. The user must program the USB OHCI Input Clock and the USBDIV divider in PMC_USB register to generate a 48 MHz and a 12 MHz signal with an accuracy of 0.25%. The USB clock input is to be defined according to Main Oscillator via the FREQ field. It is defined in the UTMI Clock Trimming Register located in the SFR section (see the Special Function Registers (SFR) section of this datasheet for details). This input clock can be 12, 16, 24, or 48 MHz.
25.6
DDR2/LPDDR/LPDDR2 Clock
The Power Management Controller controls the clocks of the DDR memory. The DDR clock can be enabled and disabled with the DDRCK bit respectively in the PMC_SCER and PMC_SDER registers. At reset, the DDR clock is disabled to save power consumption. In the case MDIV = 00, (PCK = MCK) SysClk DDR and DDRCK clocks are not available. If the Input clock is PLLACK/2, the DDR Controller can drive DDR2, LPDDR and LPDDR2 at up to 166 MHz with MDIV = 11. To save PLLA power consumption, the user can choose UPLLCK an Input clock for the system. In this case the DDR Controller can drive LD-DDR at up to 120 MHz.
25.7
Software Modem Clock

The Power Management Controller controls the clocks of the Software Modem. SMDCK is a division of UPLL or PLLA.

196
25.8
Peripheral Clock Controller

The Power Management Controller controls the clocks of each embedded peripheral by the way of the Peripheral Clock Controller. The user can individually enable and disable the clock on the peripherals and select a division factor from MCK. This is done with the help of the Peripheral Control Register (PMC_PCR). In order to save power consumption, the division factor can be 1, 2, 4 or 8. PMC_PCR register is a register that features a command and acts like a mailbox. To write the division factor on a particular peripheral, user needs to write a WRITE command, the peripheral ID and the chosen division factor. To read the current division factor on a particular peripheral, user just needs to write the READ command and the peripheral ID. Code Example to select divider 8 for peripheral 2 and enable its clock: write_register(PMC_PCR,0x01030102) Code Example to read the divider of peripheral 4: write_register(PMC_PCR,0x00000004) When a peripheral clock is disabled, the clock is immediately stopped. The peripheral clocks are automatically disabled after a reset. In order to stop a peripheral, it is recommended that the system software wait until the peripheral has executed its last programmed operation before disabling the clock. This is to avoid data corruption or erroneous behavior of the system. The bit number within the Peripheral Control registers is the Peripheral Identifier defined at the product level. Generally, the bit number corresponds to the interrupt source number assigned to the peripheral.
25.9
Programmable Clock Output Controller

The PMC controls 2 signals to be outputs on external pins PCKx. Each signal can be independently programmed via the PMC_PCKx registers. PCKx can be independently selected between the Slow clock, the Master Clock, the PLLACK2, the UTMI PLL output and the main clock by writing the CSS field in PMC_PCKx. Each output signal can also be divided by a power of 2 between 1 and 64 by writing the PRES (Prescaler) field in PMC_PCKx. Each output signal can be enabled and disabled by writing 1 in the corresponding bit, PCKx of PMC_SCER and PMC_SCDR, respectively. Status of the active programmable output clocks are given in the PCKx bits of PMC_SCSR (System Clock Status Register). Moreover, like the PCK, a status bit in PMC_SR indicates that the Programmable Clock is actually what has been programmed in the Programmable Clock registers. As the Programmable Clock Controller does not manage with glitch prevention when switching clocks, it is strongly recommended to disable the Programmable Clock before any configuration change and to re-enable it after the change is actually performed.

197
25.10 Main Crystal Clock Failure Detector

The clock failure detector monitors the 3 to 20 MHz Crystal or Ceramic Resonator-based oscillator to identify an eventual defect of this oscillator (for example, if the crystal is unconnected). The clock failure detector can be enabled or disabled by means of the CFDEN bit in the PMC Clock Generator Main Oscillator Register (CKGR_MOR). After reset, the detector is disabled. However, if the 3 to 20 MHz Crystal or Ceramic Resonator-based Oscillator is disabled, the clock failure detector is disabled too. A failure is detected by means of a counter incrementing on the 3 to 20 MHz Crystal oscillator or Ceramic Resonatorbased oscillator clock edge and timing logic clocked on the slow clock RC oscillator controlling the counter. The counter is cleared when the slow clock RC oscillator signal is low and enabled when the slow clock RC oscillator is high. Thus the failure detection time is 1 slow clock RC oscillator clock period. If, during the high level period of the slow clock RC oscillator, less than 8 fast crystal oscillator clock periods have been counted, then a failure is declared. The slow RC oscillator must be enabled. The clock failure detection must be enabled only when system clock MCK selects the fast RC oscillator.Then the status register must be read 2 slow clock cycles after enabling. The clock failure detection must be disabled when the main crystal is disabled. If a failure of the 3 to 20 MHz Crystal or Ceramic Resonator-based oscillator clock is detected, the CFDEV flag is set in the PMC Status Register (PMC_SR), and generates an interrupt if it is not masked. The interrupt remains active until a read operation in the PMC_SR register. The user can know the status of the clock failure detector at any time by reading the CFDS bit in the PMC_SR register. If the 3 to 20 MHz Crystal or Ceramic Resonator-based oscillator clock is selected as the source clock of MAINCK (MOSCSEL = 1), and if the Master Clock Source is PLLACK or UPLLCK (CSS = 2 or 3), a clock failure detection automatically forces MAINCK to be the source clock for the master clock (MCK).Then, regardless of the PMC configuration, a clock failure detection automatically forces the 4/8/12 MHz Fast RC oscillator to be the source clock for MAINCK. If the Fast RC oscillator is disabled when a clock failure detection occurs, it is automatically re-enabled by the clock failure detection mechanism. It takes 2 slow clock RC oscillator cycles to detect and switch from the 3 to 20 MHz Crystal, or Ceramic Resonator-based oscillator, to the 4/8/12 MHz Fast RC Oscillator if the Master Clock source is Main Clock, or 3 slow clock RC oscillator cycles if the Master Clock source is PLLACK or UPLLCK. A clock failure detection activates a fault output that is connected to the Pulse Width Modulator (PWM) Controller. With this connection, the PWM controller is able to force its outputs and to protect the driven device, if a clock failure is detected. This fault output remains active until the defect is detected and until it is cleared by the bit FOCLR in the PMC Fault Output Clear Register (PMC_FOCR). The user can know the status of the fault output at any time by reading the FOS bit in the PMC_SR register.

198
25.11 Programming Sequence

1. Enabling the 12 MHz Main Oscillator: The main oscillator is enabled by setting the MOSCEN field in the CKGR_MOR register. In some cases it may be advantageous to define a start-up time. This can be achieved by writing a value in the OSCOUNT field in the CKGR_MOR register. Once this register has been correctly configured, the user must wait for MOSCS field in the PMC_SR register to be set. This can be done either by polling the status register or by waiting the interrupt line to be raised if the associated interrupt to MOSCS has been enabled in the PMC_IER register. 2. Setting PLLA and divider: All parameters needed to configure PLLA and the divider are located in the CKGR_PLLAR register. The DIVA field is used to control divider itself. A value between 0 and 255 can be programmed. Divider output is divider input divided by DIVA parameter. By default DIVA parameter is set to 0 which means that divider is turned off. The OUTA field is used to select the PLLA output frequency range. The MULA field is the PLLA multiplier factor. This parameter can be programmed between 0 and 127. If MULA is set to 0, PLLA will be turned off, otherwise the PLLA output frequency is PLLA input frequency multiplied by (MULA + 1). The PLLACOUNT field specifies the number of slow clock cycles before LOCKA bit is set in the PMC_SR register after CKGR_PLLAR register has been written. Once the PMC_PLLAR register has been written, the user must wait for the LOCKA bit to be set in the PMC_SR register. This can be done either by polling the status register or by waiting the interrupt line to be raised if the associated interrupt to LOCKA has been enabled in the PMC_IER register. All parameters in CKGR_PLLAR can be programmed in a single write operation. If at some stage one of the following parameters, MULA, DIVA is modified, LOCKA bit will go low to indicate that PLLA is not ready yet. When PLLA is locked, LOCKA will be set again. The user is constrained to wait for LOCKA bit to be set before using the PLLA output clock. Code Example: write_register(CKGR_PLLAR,0x00040805) If PLLA and divider are enabled, the PLLA input clock is the main clock. PLLA output clock is PLLA input clock multiplied by 5. Once CKGR_PLLAR has been written, LOCKA bit will be set after eight slow clock cycles. 3. Setting Bias and High-speed PLL (UPLL) for UTMI The UTMI PLL is enabled by setting the UPLLEN field in the CKGR_UCKR register. The UTMI Bias must is enabled by setting the BIASEN field in the CKGR_UCKR register in the same time. In some cases it may be advantageous to define a start-up time. This can be achieved by writing a value in the PLLCOUNT field in the CKGR_UCKR register. Once this register has been correctly configured, the user must wait for LOCKU field in the PMC_SR register to be set. This can be done either by polling the status register or by waiting the interrupt line to be raised if the associated interrupt to LOCKU has been enabled in the PMC_IER register. 4. Selecting Master Clock and Processor Clock The Master Clock and the Processor Clock are configurable via the PMC_MCKR register. The CSS field is used to select the clock source of the Master Clock and Processor Clock dividers. By default, the selected clock source is slow clock. The PRES field is used to control the Master/Processor Clock prescaler. The user can choose between different values (1, 2, 4, 8, 16, 32, 64). Prescaler output is the selected clock source divided by PRES parameter. By default, PRES parameter is set to 1 which means that the input clock of the Master Clock and Processor Clock dividers is equal to slow clock.

199
The MDIV field is used to control the Master Clock divider. It is possible to choose between different values (0, 1, 2, 3). The Master Clock output is Master/Processor Clock Prescaler output divided by 1, 2, 4 or 3, depending on the value programmed in MDIV. The PMC PLLA Clock input must be divided by 2, thanks to the PLLADIV2 field. By default, MDIV and PLLLADIV2 are set to 0, which indicates that Processor Clock is equal to the Master Clock. Once the PMC_MCKR register has been written, the user must wait for the MCKRDY bit to be set in the PMC_SR register. This can be done either by polling the status register or by waiting for the interrupt line to be raised if the associated interrupt to MCKRDY has been enabled in the PMC_IER register. The PMC_MCKR register must not be programmed in a single write operation. The preferred programming sequence for the PMC_MCKR register is as follows:
z
If a new value for CSS field corresponds to PLLA Clock,

z z z z
Program the PRES field in the PMC_MCKR register. Wait for the MCKRDY bit to be set in the PMC_SR register. Program the CSS field in the PMC_MCKR register. Wait for the MCKRDY bit to be set in the PMC_SR register. Program the CSS field in the PMC_MCKR register. Wait for the MCKRDY bit to be set in the PMC_SR register. Program the PRES field in the PMC_MCKR register. Wait for the MCKRDY bit to be set in the PMC_SR register.
If a new value for CSS field corresponds to Main Clock or Slow Clock,
z z z z
If at some stage one of the following parameters, CSS or PRES, is modified, the MCKRDY bit will go low to indicate that the Master Clock and the Processor Clock are not ready yet. The user must wait for MCKRDY bit to be set again before using the Master and Processor Clocks. Note: IF PLLA clock was selected as the Master Clock and the user decides to modify it by writing in CKGR_PLLAR, the MCKRDY flag will go low while PLLA is unlocked. Once PLLA is locked again, LOCK goes high and MCKRDY is set. While PLLA is unlocked, the Master Clock selection is automatically changed to Main Clock. For further information, see Section 25.12.2. Clock Switching Waveforms on page 203. Code Example: write_register(PMC_MCKR,0x00000001) wait (MCKRDY=1) write_register(PMC_MCKR,0x00000011) wait (MCKRDY=1) The Master Clock is main clock divided by 16. The Processor Clock is the Master Clock. 5. Selecting Programmable Clocks Programmable clocks are controlled via registers; PMC_SCER, PMC_SCDR and PMC_SCSR. Programmable clocks can be enabled and/or disabled via the PMC_SCER and PMC_SCDR registers. Depending on the system used, 3 programmable clocks can be enabled or disabled. The PMC_SCSR provides a clear indication as to which Programmable clock is enabled. By default all Programmable clocks are disabled. PMC_PCKx registers are used to configure programmable clocks. The CSS and CSSMCK fields are used to select the programmable clock divider source. Five clock options are available: main clock, slow clock, master clock, PLLACK, UPLLCK. By default, the clock source selected is slow clock. The PRES field is used to control the programmable clock prescaler. It is possible to choose between different values (1, 2, 4, 8, 16, 32, 64). Programmable clock output is prescaler input divided by PRES parameter. By default, the PRES parameter is set to 1 which means that master clock is equal to slow clock.

200
Once the PMC_PCKx register has been programmed, The corresponding programmable clock must be enabled and the user is constrained to wait for the PCKRDYx bit to be set in the PMC_SR register. This can be done either by polling the status register or by waiting the interrupt line to be raised if the associated interrupt to PCKRDYx has been enabled in the PMC_IER register. All parameters in PMC_PCKx can be programmed in a single write operation. If the CSS and PRES parameters are to be modified, the corresponding programmable clock must be disabled first. The parameters can then be modified. Once this has been done, the user must re-enable the programmable clock and wait for the PCKRDYx bit to be set. Code Example: write_register(PMC_PCK0,0x00000015) Programmable clock 0 is main clock divided by 32. 6. Enabling Peripheral Clocks Once all of the previous steps have been completed, the peripheral clocks can be enabled and/or disabled via registers PMC_PCER and PMC_PCDR. Depending on the system used, 19 peripheral clocks can be enabled or disabled. The PMC_PCSR provides a clear view as to which peripheral clock is enabled. Note: Each enabled peripheral clock corresponds to Master Clock. Code Examples: write_register(PMC_PCER,0x00000110) Peripheral clocks 4 and 8 are enabled. write_register(PMC_PCDR,0x00000010) Peripheral clock 4 is disabled.

201
25.12 Clock Switching Details

25.12.1 Master Clock Switching Timings
Table 25-1 and Table 25-2 give the worst case timings required for the Master Clock to switch from one selected clock to another one. This is in the event that the prescaler is de-activated. When the prescaler is activated, an additional time of 64 clock cycles of the new selected clock has to be added.
Table 25-1. Clock Switching Timings (Worst Case) From To Main Clock SLCK 0.5 x Main Clock + 4.5 x SLCK 0.5 x Main Clock + 4 x SLCK + PLLCOUNT x SLCK + 2.5 x PLLx Clock 4 x SLCK + 2.5 x Main Clock 2.5 x PLL Clock + 5 x SLCK + PLLCOUNT x SLCK 3 x PLL Clock + 4 x SLCK + 1 x Main Clock 3 x PLL Clock + 5 x SLCK 2.5 x PLL Clock + 4 x SLCK + PLLCOUNT x SLCK Main Clock SLCK PLL Clock
PLL Clock
Notes: 1. 2.
PLL designates either the PLLA or the UPLL Clock. PLLCOUNT designates either PLLACOUNT or UPLLCOUNT.
Table 25-2. Clock Switching Timings between Two PLLs (Worst Case) From To PLLA Clock 2.5 x PLLA Clock + 4 x SLCK + PLLACOUNT x SLCK 3 x UPLL Clock + 4 x SLCK + 1.5 x UPLL Clock 3 x PLLA Clock + 4 x SLCK + 1.5 x PLLA Clock 2.5 x UPLL Clock + 4 x SLCK + UPLLCOUNT x SLCK PLLA Clock UPLL Clock
UPLL Clock

202
25.12.2 Clock Switching Waveforms

Figure 25-3. Switch Master Clock from Slow Clock to PLL Clock
Slow Clock
PLL Clock
LOCK
MCKRDY
Master Clock
Write PMC_MCKR
Figure 25-4. Switch Master Clock from Main Clock to Slow Clock
Slow Clock
Main Clock
MCKRDY
Master Clock
Write PMC_MCKR

203
Figure 25-5. Change PLLA Programming
Slow Clock
PLLA Clock
LOCKA
MCKRDY
Master Clock Slow Clock Write CKGR_PLLAR
Figure 25-6. Programmable Clock Output Programming
PLL Clock
PCKRDY
PCKx Output
Write PMC_PCKx
PLL Clock is selected
Write PMC_SCER
PCKx is enabled
Write PMC_SCDR
PCKx is disabled

204
25.13 Power Management Controller (PMC) User Interface

Table 25-3. Register Mapping Offset 0x0000 0x0004 0x0008 0x000C 0x0010 0x0014 0x0018 0x001C 0x0020 0x0024 0x0028 0x002C 0x0030 0x0034 0x0038 0x003C 0x0040 0x0044 0x0048 0x004C - 0x005C 0x0060 0x0064 0x0068 0x006C 0x0070 - 0x0074 0x0078 0x007C 0x0080 0x0084-0x00E0 0x00E4 0x00E8 0x00EC-0x00FC 0x0100 0x0104 Register System Clock Enable Register System Clock Disable Register System Clock Status Register Reserved Peripheral Clock Enable Register 0 Peripheral Clock Disable Register 0 Peripheral Clock Status Register 0 UTMI Clock Register Main Oscillator Register Main Clock Frequency Register PLLA Register Reserved Master Clock Register Reserved USB Clock Register Soft Modem Clock Register Programmable Clock 0 Register Programmable Clock 1 Register Programmable Clock 2 Register Reserved Interrupt Enable Register Interrupt Disable Register Status Register Interrupt Mask Register Reserved Fault Output Clear Register Reserved PLL Charge Pump Current Register Reserved Write Protect Mode Register Write Protect Status Register Reserved Peripheral Clock Enable Register 1 Peripheral Clock Disable Register 1 Name PMC_SCER PMC_SCDR PMC_SCSR PMC_PCER0 PMC_PCDR0 PMC_PCSR0 CKGR_UCKR CKGR_MOR CKGR_MCFR CKGR_PLLAR PMC_MCKR PMC_USB PMC_SMD PMC_PCK0 PMC_PCK1 PMC_PCK2 PMC_IER PMC_IDR PMC_SR PMC_IMR PMC_FOCR PMC_PLLICPR PMC_WPMR PMC_WPSR PMC_PCER1 PMC_PCDR1 Access Write-only Write-only Read-only Write-only Write-only Read-only Read-write Read-write Read-only Read-write Read-write Read/Write Read-write Read-write Read-write Read-write Write-only Write-only Read-only Read-only Write-only Write-only Read-write Read-only Write-only Write-only Reset N.A. N.A. 0x0000_0005 0x0000_0000 0x1020_0000 0x0100_0001 0x0000_0000 0x0000_3F00 0x0000_0001 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 N.A. N.A. 0x0001_0008 0x0000_0000 0x0100_0100 0x0000_0000 0x0000_0000

205
Table 25-3. Register Mapping Offset 0x0108 0x010C 0x0110 Register Peripheral Clock Status Register 1 Peripheral Control Register Oscillator Calibration Register Name PMC_PCSR1 PMC_PCR PMC_OCR Access Read-only Read-write Read-write Reset 0x0000_0000 0x0000_0000 0x0040_4040

206
25.13.1 PMC System Clock Enable Register

31 23 15 7 UDP
PMC_SCER 0xFFFFFC00 Write-only

30 22 14 6 UHP 29 21 13 5 28 20 12 4 SMDCK 27 19 11 3 LCDCK 26 18 10 PCK2 2 DDRCK 25 17 9 PCK1 1 24 16 8 PCK0 0
DDRCK: DDR Clock Enable

0 = No effect. 1 = Enables the DDR clock.
LCDCK: LCD2x Clock Enable

0 = No effect. 1 = Enables the LCD2x clock.
SMDCK: SMD Clock Enable

0 = No effect. 1 = Enables the soft modem clock.
UHP: USB Host OHCI Clocks Enable

0 = No effect. 1 = Enables the UHP48M and UHP12M OHCI clocks.
UDP: USB Device Clock Enable

0 = No effect. 1 = Enables the USB Device clock.
PCKx: Programmable Clock x Output Enable

0 = No effect. 1 = Enables the corresponding Programmable Clock output.

207
25.13.2 PMC System Clock Disable Register

31 23 15 7 UDP
PMC_SCDR 0xFFFFFC04 Write-only

30 22 14 6 UHP 29 21 13 5 28 20 12 4 SMDCK 27 19 11 3 LCDCK 26 18 10 PCK2 2 DDRCK 25 17 9 PCK1 1 24 16 8 PCK0 0 PCK
PCK: Processor Clock Disable

0 = No effect. 1 = Disables the Processor clock. This is used to enter the processor in Idle Mode.
DDRCK: DDR Clock Disable

0 = No effect. 1 = Disables the DDR clock.
LCDCK: LCD2x Clock Disable

0 = No effect. 1 = Disables the LCD2x clock.
SMDCK: SMD Clock Disable

0 = No effect. 1 = Disables the soft modem clock.
UHP: USB Host OHCI Clock Disable

0 = No effect. 1 = Disables the UHP48M and UHP12M OHCI clocks.
UDP: USB Device Clock Enable

0 = No effect. 1 = Disables the USB Device clock.
PCKx: Programmable Clock x Output Disable

0 = No effect. 1 = Disables the corresponding Programmable Clock output.

208
25.13.3 PMC System Clock Status Register

31 23 15 7 UDP
PMC_SCSR 0xFFFFFC08 Read-only

30 22 14 6 UHP 29 21 13 5 28 20 12 4 SMDCK 27 19 11 3 LCDCK 26 18 10 PCK2 2 DDRCK 25 17 9 PCK1 1 24 16 8 PCK0 0 PCK
PCK: Processor Clock Status

0 = The Processor clock is disabled. 1 = The Processor clock is enabled.
DDRCK: DDR Clock Status

0 = The DDR clock is disabled. 1 = The DDR clock is enabled.
LCDCK: LCD2x Clock Status

0 = The LCD2x clock is disabled. 1 = The LCD2x clock is enabled.
SMDCK: SMD Clock Status

0 = The soft modem clock is disabled. 1 = The soft modem clock is enabled.
UHP: USB Host Port Clock Status

0 = The UHP48M and UHP12M OHCI clocks are disabled. 1 = The UHP48M and UHP12M OHCI clocks are enabled.
UDP: USB Device Port Clock Status

0 = The USB Device clock is disabled. 1 = The USB Device clock is enabled.
PCKx: Programmable Clock x Output Status

0 = The corresponding Programmable Clock output is disabled. 1 = The corresponding Programmable Clock output is enabled.

209
25.13.4 PMC Peripheral Clock Enable Register 0

31 PID31 23 PID23 15 PID15 7 PID7
PMC_PCER0 0xFFFFFC10 Write-only

30 PID30 22 PID22 14 PID14 6 PID6 29 PID29 21 PID21 13 PID13 5 PID5 28 PID28 20 PID20 12 PID12 4 PID4 27 PID27 19 PID19 11 PID11 3 PID3 26 PID26 18 PID18 10 PID10 2 PID2 25 PID25 17 PID17 9 PID9 1 24 PID24 16 PID16 8 PID8 0
This register can only be written if the WPEN bit is cleared in PMC Write Protect Mode Register on page 230.
PIDx: Peripheral Clock x Enable

0 = No effect. 1 = Enables the corresponding peripheral clock. Note: Note: PID2 to PID31 refer to identifiers as defined in the section Peripheral Identifiers in the product datasheet. Other peripherals can be enabled in PMC_PCER1. Programming the control bits of the Peripheral ID that are not implemented has no effect on the behavior of the PMC.

210
25.13.5 PMC Peripheral Clock Disable Register 0

31 PID31 23 PID23 15 PID15 7 PID7
PMC_PCDR0 0xFFFFFC14 Write-only

30 PID30 22 PID22 14 PID14 6 PID6 29 PID29 21 PID21 13 PID13 5 PID5 28 PID28 20 PID20 12 PID12 4 PID4 27 PID27 19 PID19 11 PID11 3 PID3 26 PID26 18 PID18 10 PID10 2 PID2 25 PID25 17 PID17 9 PID9 1 24 PID24 16 PID16 8 PID8 0 -
PIDx: Peripheral Clock x Disable

0 = No effect. 1 = Disables the corresponding peripheral clock. Note: PID2 to PID31 refer to identifiers as defined in the section Peripheral Identifiers in the product datasheet. Other peripherals can be disabled in PMC_PCDR1.

211
25.13.6 PMC Peripheral Clock Status Register 0

31 PID31 23 PID23 15 PID15 7 PID7
PMC_PCSR0 0xFFFFFC18 Read-only

30 PID30 22 PID22 14 PID14 6 PID6 29 PID29 21 PID21 13 PID13 5 PID5 28 PID28 20 PID20 12 PID12 4 PID4 27 PID27 19 PID19 11 PID11 3 PID3 26 PID26 18 PID18 10 PID10 2 PID2 25 PID25 17 PID17 9 PID9 1 24 PID24 16 PID16 8 PID8 0
PIDx: Peripheral Clock x Status

0 = The corresponding peripheral clock is disabled. 1 = The corresponding peripheral clock is enabled. Note: PID2 to PID31 refer to identifiers as defined in the section Peripheral Identifiers in the product datasheet. Other peripherals status can be read in PMC_PCSR1.

212
25.13.7 PMC UTMI Clock Configuration Register

31
CKGR_UCKR 0xFFFFFC1C Read-write

30 BIASCOUNT 29 28 27 19 11 3 26 18 10 2 25 17 9 1 24 BIASEN 16 UPLLEN 8 0
23
22 UPLLCOUNT
21
20
15 7
14 6
13 5
12 4
UPLLEN: UTMI PLL Enable

0 = The UTMI PLL is disabled. 1 = The UTMI PLL is enabled. When UPLLEN is set, the LOCKU flag is set once the UTMI PLL startup time is achieved.
UPLLCOUNT: UTMI PLL Start-up Time

Specifies the number of Slow Clock cycles multiplied by 8 for the UTMI PLL start-up time.
BIASEN: UTMI BIAS Enable

0 = The UTMI BIAS is disabled. 1 = The UTMI BIAS is enabled.
BIASCOUNT: UTMI BIAS Start-up Time

Specifies the number of Slow Clock cycles for the UTMI BIAS start-up time.

213
25.13.8 PMC Clock Generator Main Oscillator Register

31 23
CKGR_MOR 0xFFFFFC20 Read-write

30 22 29 21 28 20 KEY 27 19 26 18 25 CFDEN 17 24 MOSCSEL 16
15
14
13
12 MOSCXTST
11
10
5 0
3 MOSCRCEN
1 MOSCXTBY
0 MOSCXTEN
Warning: Bit 4,5,6 must always be set to 0 when programming the CKGR_MOR register.
KEY: Password
0x37 (PASSWD): Should be written at value 0x37. Writing any other value in this field aborts the write operation.
MOSCXTEN: Main Crystal Oscillator Enable

A crystal must be connected between XIN and XOUT. 0 = The Main Crystal Oscillator is disabled. 1 = The Main Crystal Oscillator is enabled. MOSCXTBY must be set to 0. When MOSCXTEN is set, the MOSCXTS flag is set once the Main Crystal Oscillator startup time is achieved.
MOSCXTBY: Main Crystal Oscillator Bypass

0 = No effect. 1 = The Main Crystal Oscillator is bypassed. MOSCXTEN must be set to 0. An external clock must be connected on XIN. When MOSCXTBY is set, the MOSCXTS flag in PMC_SR is automatically set. Clearing MOSCXTEN and MOSCXTBY bits allows resetting the MOSCXTS flag.
MOSCRCEN: Main On-Chip RC Oscillator Enable

0 = The Main On-Chip RC Oscillator is disabled. 1 = The Main On-Chip RC Oscillator is enabled. When MOSCRCEN is set, the MOSCRCS flag is set once the Main On-Chip RC Oscillator startup time is achieved.
MOSCXTST: Main Crystal Oscillator Start-up Time

Specifies the number of Slow Clock cycles multiplied by 8 for the Main Crystal Oscillator start-up time.
MOSCSEL: Main Oscillator Selection

0 = The Main On-Chip RC Oscillator is selected. 1 = The Main Crystal Oscillator is selected.

214
CFDEN: Clock Failure Detector Enable

0 = The Clock Failure Detector is disabled. 1 = The Clock Failure Detector is enabled. The clock failure detection must be disabled when the main crystal is disabled. The slow RC oscillator must be enabled. The clock failure detection must be enabled only when system clock MCK selects the fast RC oscillator.

215
25.13.9 PMC Clock Generator Main Clock Frequency Register

31 23 15
CKGR_MCFR 0xFFFFFC24 Read-only

30 22 14 29 21 13 28 20 RCMEAS 12 MAINF 7 6 5 4 MAINF 3 2 1 0 27 19 11 26 18 10 25 17 9 24 16 MAINFRDY 8
MAINF: Main Clock Frequency

Gives the number of Main Clock cycles within 16 Slow Clock periods.
MAINFRDY: Main Clock Ready

0 = MAINF value is not valid or the Main Oscillator is disabled. 1 = The Main Oscillator has been enabled previously and MAINF value is available.
RCMEAS: RC Oscillator Frequency Measure (write-only)

0 = No effect. 1 = Restarts a measure of the main RC frequency, MAINF will carry the new frequency as soon as a low to high transition occurs on MAINFRDY flag. The measure is performed on the main frequency (i.e. not limited to RC oscillator only) but if the main clock frequency source is the fast crystal oscillator, the restart of the measure is unneeded because of the well known stability of crystal oscillators.

216
25.13.10 PMC Clock Generator PLLA Register

31 23
CKGR_PLLAR 0xFFFFFC28 Read/Write

30 22 29 ONE 21 MULA 28 20 27 19 26 18 25 17 OUTA 12 11 PLLACOUNT 6 5 4 DIVA 3 2 1 0 10 9 8 24 MULA 16
15 OUTA 7
14
13
Possible limitations on PLL input frequencies and multiplier factors should be checked before using the PMC. Warning: Bit 29 must always be set to 1 when programming the CKGR_PLLAR register.
DIVA: Divider A
Value 0 1 2 - 255 Name 0 BYPASS Description Divider output is 0 Divider is bypassed Divider output is the selected clock divided by DIVA.
PLLACOUNT: PLLA Counter

Specifies the number of slow clock cycles before the LOCKA bit is set in PMC_SR after CKGR_PLLAR is written.
OUTA: PLLA Clock Frequency Range

To be programmed to 0.
MULA: PLLA Multiplier

0 = The PLLA is deactivated. 1 up to 127 = The PLLA Clock frequency is the PLLA input frequency multiplied by MULA+ 1.
ONE: Must Be Set to 1

Bit 29 must always be set to 1 when programming the CKGR_PLLAR register.

217
25.13.11 PMC Master Clock Register

31 23 15 7
PMC_MCKR 0xFFFFFC30 Read/Write

30 22 14 6 29 21 13 5 PRES 28 20 12 PLLADIV2 4 27 19 11 3 26 18 10 2 25 17 9 MDIV 1 CSS 0 24 16 8
CSS: Master/Processor Clock Source Selection

Value 0 1 2 3 Name SLOW_CLK MAIN_CLK PLLA_CLK UPLL_CLK Description Slow Clock is selected Main Clock is selected PLLACK2 is selected UPLL Clock is selected
PRES: Master/Processor Clock Prescaler

Value 0 1 2 3 4 5 6 7 Name CLOCK CLOCK_DIV2 CLOCK_DIV4 CLOCK_DIV8 CLOCK_DIV16 CLOCK_DIV32 CLOCK_DIV64 Reserved Description Selected clock Selected clock divided by 2 Selected clock divided by 4 Selected clock divided by 8 Selected clock divided by 16 Selected clock divided by 32 Selected clock divided by 64 Reserved

218
MDIV: Master Clock Division

Value 0 Name EQ_PCK Description Master Clock is Prescaler Output Clock divided by 1. Warning: SysClk DDR and DDRCK are not available. Master Clock is Prescaler Output Clock divided by 2. SysClk DDR is equal to 2 x MCK. DDRCK is equal to MCK. Master Clock is Prescaler Output Clock divided by 4. SysClk DDR is equal to 2 x MCK. DDRCK is equal to MCK. Master Clock is Prescaler Output Clock divided by 3. SysClk DDR is equal to 2 x MCK. DDRCK is equal to MCK.
PCK_DIV2
PCK_DIV4
PCK_DIV3
PLLADIV2: PLLA Divisor by 2

Bit PLLADIV2 must always be set to 1 when MDIV is set to 3.

219
25.13.12 PMC USB Clock Register

31 23 15 7
PMC_USB 0xFFFFFC38 Read/Write

30 22 14 6 29 21 13 5 28 20 12 4 27 19 11 26 18 10 USBDIV 3 2 1 0 USBS 25 17 9 24 16 8
USBS: USB OHCI Input clock selection

0 = USB Clock Input is PLLA 1 = USB Clock Input is UPLL
USBDIV: Divider for USB OHCI Clock.

USB Clock is Input clock divided by USBDIV+1

220
25.13.13 PMC SMD Clock Register

31 23 15 7
PMC_SMD 0xFFFFFC3C Read/Write

30 22 14 6 29 21 13 5 28 20 12 27 19 11 26 18 10 SMDDIV 2 25 17 9 24 16 8
0 SMDS
SMDS: SMD input clock selection

0 = SMD Clock Input is PLLA 1 = SMD Clock Input is UPLL
SMDDIV: Divider for SMD Clock.

SMD Clock is Input clock divided by SMD +1

221
25.13.14 PMC Programmable Clock Register

31 23 15 7
PMC_PCKx 0xFFFFFC40 Read-write

30 22 14 6 29 21 13 5 PRES 28 20 12 4 27 19 11 3 26 18 10 2 25 17 9 1 CSS 24 16 8 0
CSS: Master Clock Source Selection

Value 0 1 2 3 4 name SLOW_CLK MAIN_CLK PLLA_CLK UPLL_CLK MCK_CLK Description Slow Clock is selected Main Clock is selected PLLACK2 is selected UPLL Clock is selected Master Clock is selected
PRES: Programmable Clock Prescaler

Value 0 1 2 3 4 5 6 7 name CLOCK CLOCK_DIV2 CLOCK_DIV4 CLOCK_DIV8 CLOCK_DIV16 CLOCK_DIV32 CLOCK_DIV64 Reserved Description Selected clock Selected clock divided by 2 Selected clock divided by 4 Selected clock divided by 8 Selected clock divided by 16 Selected clock divided by 32 Selected clock divided by 64 Reserved

222
25.13.15 PMC Interrupt Enable Register

31 23 15 7
PMC_IER 0xFFFFFC60 Write-only

30 22 14 6 LOCKU 29 21 13 5 28 20 12 4 27 19 11 3 MCKRDY 26 18 CFDEV 10 PCKRDY2 2 25 17 MOSCRCS 9 PCKRDY1 1 LOCKA 24 16 MOSCSELS 8 PCKRDY0 0 MOSCXTS
MOSCXTS: Main Crystal Oscillator Status Interrupt Enable LOCKA: PLLA Lock Interrupt Enable MCKRDY: Master Clock Ready Interrupt Enable LOCKU: UTMI PLL Lock Interrupt Enable PCKRDYx: Programmable Clock Ready x Interrupt Enable MOSCSELS: Main Oscillator Selection Status Interrupt Enable MOSCRCS: Main On-Chip RC Status Interrupt Enable CFDEV: Clock Failure Detector Event Interrupt Enable

223
25.13.16 PMC Interrupt Disable Register

31 23 15 7
PMC_IDR 0xFFFFFC64 Write-only

30 22 14 6 LOCKU 29 21 13 5 28 20 12 4 27 19 11 3 MCKRDY 26 18 CFDEV 10 PCKRDY2 2 25 17 MOSCRCS 9 PCKRDY1 1 LOCKA 24 16 MOSCSELS 8 PCKRDY0 0 MOSCXTS
MOSCXTS: Main Crystal Oscillator Status Interrupt Disable LOCKA: PLLA Lock Interrupt Disable MCKRDY: Master Clock Ready Interrupt Disable LOCKU: UTMI PLL Lock Interrupt Enable PCKRDYx: Programmable Clock Ready x Interrupt Disable MOSCSELS: Main Oscillator Selection Status Interrupt Disable MOSCRCS: Main On-Chip RC Status Interrupt Disable CFDEV: Clock Failure Detector Event Interrupt Disable

224
25.13.17 PMC Status Register

31 23 15 7 OSCSELS
PMC_SR 0xFFFFFC68 Read-only

30 22 14 6 LOCKU 29 21 13 5 28 20 FOS 12 4 27 19 CFDS 11 3 MCKRDY 26 18 CFDEV 10 PCKRDY2 2 25 17 MOSCRCS 9 PCKRDY1 1 LOCKA 24 16 MOSCSELS 8 PCKRDY0 0 MOSCXTS
MOSCXTS: Main XTAL Oscillator Status

0 = Main XTAL oscillator is not stabilized. 1 = Main XTAL oscillator is stabilized.
LOCKA: PLLA Lock Status

0 = PLLA is not locked 1 = PLLA is locked.
MCKRDY: Master Clock Status

0 = Master Clock is not ready. 1 = Master Clock is ready.
LOCKU: UPLL Clock Status

0 = UPLL Clock is not ready. 1 = UPLL Clock is ready.
OSCSELS: Slow Clock Oscillator Selection

0 = Internal slow clock RC oscillator is selected. 1 = External slow clock 32 kHz oscillator is selected.
PCKRDYx: Programmable Clock Ready Status

0 = Programmable Clock x is not ready. 1 = Programmable Clock x is ready.
MOSCSELS: Main Oscillator Selection Status

0 = Selection is in progress. 1 = Selection is done.
MOSCRCS: Main On-Chip RC Oscillator Status

0 = Main on-chip RC oscillator is not stabilized. 1 = Main on-chip RC oscillator is stabilized.

225
CFDEV: Clock Failure Detector Event

0 = No clock failure detection of the main on-chip RC oscillator clock has occurred since the last read of PMC_SR. 1 = At least one clock failure detection of the main on-chip RC oscillator clock has occurred since the last read of PMC_SR.
CFDS: Clock Failure Detector Status

0 = A clock failure of the main on-chip RC oscillator clock is not detected. 1 = A clock failure of the main on-chip RC oscillator clock is detected.
FOS: Clock Failure Detector Fault Output Status

0 = The fault output of the clock failure detector is inactive. 1 = The fault output of the clock failure detector is active.

226
25.13.18 PMC Interrupt Mask Register

31 23 15 7
PMC_IMR 0xFFFFFC6C Read-only

30 22 14 6 29 21 13 5 28 20 12 4 27 19 11 3 MCKRDY 26 18 CFDEV 10 PCKRDY2 2 25 17 MOSCRCS 9 PCKRDY1 1 LOCKA 24 16 MOSCSELS 8 PCKRDY0 0 MOSCXTS
MOSCXTS: Main Crystal Oscillator Status Interrupt Mask LOCKA: PLLA Lock Interrupt Mask MCKRDY: Master Clock Ready Interrupt Mask PCKRDYx: Programmable Clock Ready x Interrupt Mask MOSCSELS: Main Oscillator Selection Status Interrupt Mask MOSCRCS: Main On-Chip RC Status Interrupt Mask CFDEV: Clock Failure Detector Event Interrupt Mask

227
25.13.19 PMC Fault Output Clear Register

31 23 15 7
PMC_FOCR 0xFFFFFC78 Write-only

30 22 14 6 29 21 13 5 28 20 12 4 27 19 11 3 26 18 10 2 25 17 9 1 24 16 8 0 FOCLR
FOCLR: Fault Output Clear

Clears the clock failure detector fault output.

228
25.13.20 PLL Charge Pump Current Register

31 23 15 7
PMC_PLLICPR 0xFFFFFC80 Write-only

30 22 14 6 29 21 13 5 28 20 12 4 27 19 11 3 26 18 10 25 IVCO_PLLU 17 ICP_PLLU 9 IPLL_PLLA 1 ICP_PLLA 8 16 24
ICP_PLLA: Charge Pump Current PLLA

To optimize clock performance, this field must be programmed as specified in PLL A Characteristics in the Electrical Characteristics section of the product datasheet.
IPLL_PLLA: Engineering Configuration PLLA

Should be written to 0.
ICP_PLLU: Charge Pump Current PLL UTMI

IVCO_PLLU: Voltage Control Output Current PLL UTMI


229
25.13.21 PMC Write Protect Mode Register

31
PMC_WPMR 0xFFFFFCE4 Read-write See Table 25-3

30 29 28 WPKEY 27 26 25 24
23
22
21
20 WPKEY
19
18
17
16
15
14
13
12 WPKEY
11
10
0 WPEN

0 = Disables the Write Protect if WPKEY corresponds to 0x504D43 (PMC in ASCII). 1 = Enables the Write Protect if WPKEY corresponds to 0x504D43 (PMC in ASCII). Protects the registers:
PMC System Clock Enable Register on page 207 PMC System Clock Disable Register on page 208 PMC Clock Generator Main Clock Frequency Register on page 216 PMC Clock Generator PLLA Register on page 217 PMC Clock Generator PLLA Register on page 217 PMC Master Clock Register on page 218 PMC USB Clock Register on page 220 PMC Programmable Clock Register on page 222 PLL Charge Pump Current Register on page 229 PMC Peripheral Clock Enable Register 0 on page 210 PMC Peripheral Clock Disable Register 1 on page 233 PMC Oscillator Calibration Register on page 236 WPKEY: Write Protect KEY
Should be written at value 0x504D43 (PMC in ASCII). Writing any other value in this field aborts the write operation of the WPEN bit. Always reads as 0.

230
25.13.22 PMC Write Protect Status Register

31 23
PMC_WPSR 0xFFFFFCE8 Read-only See Table 25-3

30 22 29 21 28 20 WPVSRC 27 19 26 18 25 17 24 16
15
14
13
12 WPVSRC
11
10
0 WPVS

0 = No Write Protect Violation has occurred since the last read of the PMC_WPSR register. 1 = A Write Protect Violation has occurred since the last read of the PMC_WPSR register. If this violation is an unauthorized attempt to write a protected register, the associated violation is reported into field WPVSRC.

When WPVS is active, this field indicates the write-protected register (through address offset or code) in which a write access has been attempted. Reading PMC_WPSR automatically clears all fields.

231
25.13.23 PMC Peripheral Clock Enable Register 1

31 PID63 23 PID55 15 PID47 7 PID39
PMC_PCER1 0xFFFFFD00 Write-only

PIDx: Peripheral Clock x Enable

0 = No effect. 1 = Enables the corresponding peripheral clock. Notes: 1. PID32 to PID63 refer to identifiers as defined in the section Peripheral Identifiers in the product datasheet. 2. Programming the control bits of the Peripheral ID that are not implemented has no effect on the behavior of the PMC.

232
25.13.24 PMC Peripheral Clock Disable Register 1

31 PID63 23 PID55 15 PID47 7 PID39
PMC_PCDR1 0xFFFFFD04 Write-only

PIDx: Peripheral Clock x Disable

0 = No effect. 1 = Disables the corresponding peripheral clock. Note: PID32 to PID63 refer to identifiers as defined in the section Peripheral Identifiers in the product datasheet.

233
25.13.25 PMC Peripheral Clock Status Register 1

31 PID63 23 PID55 15 PID47 7 PID39
PMC_PCSR1 0xFFFFFD08 Read-only

PIDx: Peripheral Clock x Status

0 = The corresponding peripheral clock is disabled. 1 = The corresponding peripheral clock is enabled. Note: PID32 to PID63 refer to identifiers as defined in the section Peripheral Identifiers in the product datasheet.

234
25.13.26 PMC Peripheral Control Register

31 23 15 7
PMC_PCR 0xFFFFFD0C Write-only

30 22 14 6 29 21 13 5 28 EN 20 12 CMD 4 27 19 11 3 PID 26 18 10 2 25 17 DIV 9 1 8 0 24 16
PID: Peripheral ID
Peripheral ID selection from PID2 to PID31 PID2 to PID31 refer to identifiers as defined in the section Peripheral Identifiers in the product datasheet.
CMD: Command
0: Read mode 1: Write mode
DIV: Divisor value

Value 0 1 2 3 name PERIPH_DIV_MCK PERIPH_DIV2_MCK PERIPH_DIV4_MCK PERIPH_DIV8_MCK Description Peripheral clock is MCK Peripheral clock is MCK/2 Peripheral clock is MCK/4 Peripheral clock is MCK/8
EN: Enable
0: Selected Peripheral clock is disabled 1: Selected Peripheral clock is enabled

235
25.13.27 PMC Oscillator Calibration Register

31 23 15 7 SEL
PMC_OCR 0xFFFFFD10 Read-Write

30 22 14 6 29 21 13 5 28 20 12 4 27 19 11 3 CAL 26 18 10 2 25 17 9 1 24 16 8 0
CAL: 12 MHz RC Oscillator Calibration bits

Calibration bits applied to the RC Oscillator when SEL is set.
SEL: Selection of RC Oscillator Calibration bits

0 = Factory determined value. 1 = Value written by user in CAL field of this register.

236
26.
26.1
Parallel Input/Output (PIO3) Controller

Description
The Parallel Input/Output Controller (PIO) manages up to 32 fully programmable input/output lines. Each I/O line may be dedicated as a general-purpose I/O or be assigned to a function of an embedded peripheral. This assures effective optimization of the pins of a product. Each I/O line is associated with a bit number in all of the 32-bit registers of the 32-bit wide User Interface. Each I/O line of the PIO Controller features:
z z z z z z z
An input change interrupt enabling level change detection on any I/O line. Additional Interrupt modes enabling rising edge, falling edge, low level or high level detection on any I/O line. A glitch filter providing rejection of glitches lower than one-half of PIO clock cycle. A debouncing filter providing rejection of unwanted pulses from key or push button operations. Multi-drive capability similar to an open drain I/O line. Control of the pull-up and pull-down of the I/O line. Input visibility and output control.
The PIO Controller also features a synchronous output providing up to 32 bits of data output in a single write operation.
26.2
z z z z
Up to 32 Programmable I/O Lines Fully Programmable through Set/Clear Registers Multiplexing of Four Peripheral Functions per I/O Line For each I/O Line (Whether Assigned to a Peripheral or Used as General Purpose I/O)
z z z z z z z z
Input Change Interrupt Programmable Glitch Filter Programmable Debouncing Filter Multi-drive Option Enables Driving in Open Drain Programmable Pull Up on Each I/O Line Pin Data Status Register, Supplies Visibility of the Level on the Pin at Any Time Additional Interrupt Modes on a Programmable Event: Rising Edge, Falling Edge, Low Level or High Level Lock of the Configuration by the Connected Peripheral
z z z z
Synchronous Output, Provides Set and Clear of Several I/O lines in a Single Write Write Protect Registers Programmable Schmitt Trigger Inputs Programmable I/O Drive

237
26.3
Block Diagram
PIO Controller
Interrupt Controller PIO Interrupt
PMC
PIO Clock
Data, Enable
Embedded Peripheral
Up to 32 peripheral IOs
PIN 0 Data, Enable PIN 1 Up to 32 pins Embedded Peripheral Up to 32 peripheral IOs PIN 31
APB
Figure 26-2. Application Block Diagram

On-Chip Peripheral Drivers Keyboard Driver Control & Command Driver On-Chip Peripherals
PIO Controller
Keyboard Driver General Purpose I/Os External Devices

238
26.4
26.4.1 Pin Multiplexing

Each pin is configurable, according to product definition as either a general-purpose I/O line only, or as an I/O line multiplexed with one or two peripheral I/Os. As the multiplexing is hardware defined and thus product-dependent, the hardware designer and programmer must carefully determine the configuration of the PIO controllers required by their application. When an I/O line is general-purpose only, i.e. not multiplexed with any peripheral I/O, programming of the PIO Controller regarding the assignment to a peripheral has no effect and only the PIO Controller can control how the pin is driven by the product.
26.4.2 External Interrupt Lines

The interrupt signals FIQ and IRQ0 to IRQn are most generally multiplexed through the PIO Controllers. However, it is not necessary to assign the I/O line to the interrupt function as the PIO Controller has no effect on inputs and the interrupt lines (FIQ or IRQs) are used only as inputs.

The Power Management Controller controls the PIO Controller clock in order to save power. Writing any of the registers of the user interface does not require the PIO Controller clock to be enabled. This means that the configuration of the I/O lines does not require the PIO Controller clock to be enabled. However, when the clock is disabled, not all of the features of the PIO Controller are available, including glitch filtering. Note that the Input Change Interrupt, Interrupt Modes on a programmable event and the read of the pin level require the clock to be validated. After a hardware reset, the PIO clock is disabled by default. The user must configure the Power Management Controller before any access to the input line information.
26.4.4 Interrupt Generation

For interrupt handling, the PIO Controllers are considered as user peripherals. This means that the PIO Controller interrupt lines are connected among the interrupt sources. Refer to the PIO Controller peripheral identifier in the product description to identify the interrupt sources dedicated to the PIO Controllers. Using the PIO Controller requires the Interrupt Controller to be programmed first. The PIO Controller interrupt can be generated only if the PIO Controller clock is enabled.

239
26.5
The PIO Controller features up to 32 fully-programmable I/O lines. Most of the control logic associated to each I/O is represented in Figure 26-3. In this description each signal shown represents but one of up to 32 possible indexes.
Figure 26-3. I/O Line Control Logic

PIO_OER[0] PIO_OSR[0] PIO_ODR[0]
1
PIO_PUER[0] PIO_PUSR[0] PIO_PUDR[0]
Peripheral A Output Enable Peripheral B Output Enable Peripheral C Output Enable Peripheral D Output Enable PIO_ABCDSR1[0] PIO_ABCDSR2[0] Peripheral A Output Peripheral B Output Peripheral C Output Peripheral D Output
00 01 10 11 PIO_PER[0] PIO_PSR[0] PIO_PDR[0] 00 01 10 11
0 0
PIO_MDER[0] PIO_MDSR[0]
0
PIO_MDDR[0] 0
PIO_SODR[0] PIO_ODSR[0] PIO_CODR[0]

1
Pad 1
Peripheral A Input Peripheral B Input Peripheral C Input Peripheral D Input
PIO_PDSR[0] PIO_ISR[0] 0
D Q DFF D DFF Q
PIO Clock Slow Clock Clock Divider
Programmable Glitch or Debouncing Filter
EVENT DETECTOR
(Up to 32 possible inputs) PIO Interrupt
1 Resynchronization Stage PIO_IER[0] PIO_IMR[0] PIO_IDR[0] PIO_ISR[31] PIO_IER[31] PIO_IMR[31] PIO_IDR[31]
PIO_SCDR
PIO_IFER[0] PIO_IFSR[0] PIO_IFSCER[0] PIO_IFSCSR[0] PIO_IFSCDR[0] PIO_IFDR[0]

240
26.5.1 Pull-up and Pull-down Resistor Control

Each I/O line is designed with an embedded pull-up resistor and an embedded pull-down resistor. The pull-up resistor can be enabled or disabled by writing respectively PIO_PUER (Pull-up Enable Register) and PIO_PUDR (Pull-up Disable Resistor). Writing in these registers results in setting or clearing the corresponding bit in PIO_PUSR (Pull-up Status Register). Reading a 1 in PIO_PUSR means the pull-up is disabled and reading a 0 means the pull-up is enabled. The pull-down resistor can be enabled or disabled by writing respectively PIO_PPDER (Pull-down Enable Register) and PIO_PPDDR (Pull-down Disable Resistor). Writing in these registers results in setting or clearing the corresponding bit in PIO_PPDSR (Pull-down Status Register). Reading a 1 in PIO_PPDSR means the pull-up is disabled and reading a 0 means the pull-down is enabled. Enabling the pull-down resistor while the pull-up resistor is still enabled is not possible. In this case, the write of PIO_PPDER for the concerned I/O line is discarded. Likewise, enabling the pull-up resistor while the pull-down resistor is still enabled is not possible. In this case, the write of PIO_PUER for the concerned I/O line is discarded. Control of the pull-up resistor is possible regardless of the configuration of the I/O line. After reset, all of the pull-ups are enabled, i.e. PIO_PUSR resets at the value 0x0, and all the pull-downs are disabled, i.e. PIO_PPDSR resets at the value 0xFFFFFFFF.
26.5.2 I/O Line or Peripheral Function Selection

When a pin is multiplexed with one or two peripheral functions, the selection is controlled with the registers PIO_PER (PIO Enable Register) and PIO_PDR (PIO Disable Register). The register PIO_PSR (PIO Status Register) is the result of the set and clear registers and indicates whether the pin is controlled by the corresponding peripheral or by the PIO Controller. A value of 0 indicates that the pin is controlled by the corresponding on-chip peripheral selected in the PIO_ABCDSR1 and PIO_ABCDSR2 (ABCD Select Registers). A value of 1 indicates the pin is controlled by the PIO controller. If a pin is used as a general purpose I/O line (not multiplexed with an on-chip peripheral), PIO_PER and PIO_PDR have no effect and PIO_PSR returns 1 for the corresponding bit. After reset, most generally, the I/O lines are controlled by the PIO controller, i.e. PIO_PSR resets at 1. However, in some events, it is important that PIO lines are controlled by the peripheral (as in the case of memory chip select lines that must be driven inactive after reset or for address lines that must be driven low for booting out of an external memory). Thus, the reset value of PIO_PSR is defined at the product level, depending on the multiplexing of the device.
26.5.3 Peripheral A or B or C or D Selection

The PIO Controller provides multiplexing of up to four peripheral functions on a single pin. The selection is performed by writing PIO_ABCDSR1 and PIO_ABCDSR2 (ABCD Select Registers). For each pin: z The corresponding bit at level 0 in PIO_ABCDSR1 and the corresponding bit at level 0 in PIO_ABCDSR2 means peripheral A is selected. z The corresponding bit at level 1 in PIO_ABCDSR1 and the corresponding bit at level 0 in PIO_ABCDSR2 means peripheral B is selected. z The corresponding bit at level 0 in PIO_ABCDSR1 and the corresponding bit at level 1 in PIO_ABCDSR2 means peripheral C is selected. z The corresponding bit at level 1 in PIO_ABCDSR1 and the corresponding bit at level 1 in PIO_ABCDSR2 means peripheral D is selected. Note that multiplexing of peripheral lines A, B, C and D only affects the output line. The peripheral input lines are always connected to the pin input. Writing in PIO_ABCDSR1 and PIO_ABCDSR2 manages the multiplexing regardless of the configuration of the pin. However, assignment of a pin to a peripheral function requires a write in the peripheral selection registers (PIO_ABCDSR1 and PIO_ABCDSR2) in addition to a write in PIO_PDR. After reset, PIO_ABCDSR1 and PIO_ABCDSR2 are 0, thus indicating that all the PIO lines are configured on peripheral A. However, peripheral A generally does not drive the pin as the PIO Controller resets in I/O line mode.

241
26.5.4 Output Control

When the I/0 line is assigned to a peripheral function, i.e. the corresponding bit in PIO_PSR is at 0, the drive of the I/O line is controlled by the peripheral. Peripheral A or B or C or D depending on the value in PIO_ABCDSR1 and PIO_ABCDSR2 (ABCD Select Registers) determines whether the pin is driven or not. When the I/O line is controlled by the PIO controller, the pin can be configured to be driven. This is done by writing PIO_OER (Output Enable Register) and PIO_ODR (Output Disable Register). The results of these write operations are detected in PIO_OSR (Output Status Register). When a bit in this register is at 0, the corresponding I/O line is used as an input only. When the bit is at 1, the corresponding I/O line is driven by the PIO controller. The level driven on an I/O line can be determined by writing in PIO_SODR (Set Output Data Register) and PIO_CODR (Clear Output Data Register). These write operations respectively set and clear PIO_ODSR (Output Data Status Register), which represents the data driven on the I/O lines. Writing in PIO_OER and PIO_ODR manages PIO_OSR whether the pin is configured to be controlled by the PIO controller or assigned to a peripheral function. This enables configuration of the I/O line prior to setting it to be managed by the PIO Controller. Similarly, writing in PIO_SODR and PIO_CODR effects PIO_ODSR. This is important as it defines the first level driven on the I/O line.
26.5.5 Synchronous Data Output

Clearing one (or more) PIO line(s) and setting another one (or more) PIO line(s) synchronously cannot be done by using PIO_SODR and PIO_CODR registers. It requires two successive write operations into two different registers. To overcome this, the PIO Controller offers a direct control of PIO outputs by single write access to PIO_ODSR (Output Data Status Register).Only bits unmasked by PIO_OWSR (Output Write Status Register) are written. The mask bits in PIO_OWSR are set by writing to PIO_OWER (Output Write Enable Register) and cleared by writing to PIO_OWDR (Output Write Disable Register). After reset, the synchronous data output is disabled on all the I/O lines as PIO_OWSR resets at 0x0.
26.5.6 Multi Drive Control (Open Drain)

Each I/O can be independently programmed in Open Drain by using the Multi Drive feature. This feature permits several drivers to be connected on the I/O line which is driven low only by each device. An external pull-up resistor (or enabling of the internal one) is generally required to guarantee a high level on the line. The Multi Drive feature is controlled by PIO_MDER (Multi-driver Enable Register) and PIO_MDDR (Multi-driver Disable Register). The Multi Drive can be selected whether the I/O line is controlled by the PIO controller or assigned to a peripheral function. PIO_MDSR (Multi-driver Status Register) indicates the pins that are configured to support external drivers. After reset, the Multi Drive feature is disabled on all pins, i.e. PIO_MDSR resets at value 0x0.

242
26.5.7 Output Line Timings

Figure 26-4 shows how the outputs are driven either by writing PIO_SODR or PIO_CODR, or by directly writing PIO_ODSR. This last case is valid only if the corresponding bit in PIO_OWSR is set. Figure 26-4 also shows when the feedback in PIO_PDSR is available.
Figure 26-4. Output Line Timings
MCK
Write PIO_SODR Write PIO_ODSR at 1 Write PIO_CODR Write PIO_ODSR at 0
APB Access
APB Access
PIO_ODSR 2 cycles PIO_PDSR 2 cycles
26.5.8 Inputs
The level on each I/O line can be read through PIO_PDSR (Pin Data Status Register). This register indicates the level of the I/O lines regardless of their configuration, whether uniquely as an input or driven by the PIO controller or driven by a peripheral. Reading the I/O line levels requires the clock of the PIO controller to be enabled, otherwise PIO_PDSR reads the levels present on the I/O line at the time the clock was disabled.
26.5.9 Input Glitch and Debouncing Filters

Optional input glitch and debouncing filters are independently programmable on each I/O line. The glitch filter can filter a glitch with a duration of less than 1/2 Master Clock (MCK) and the debouncing filter can filter a pulse of less than 1/2 Period of a Programmable Divided Slow Clock. The selection between glitch filtering or debounce filtering is done by writing in the registers PIO_IFSCDR (PIO Input Filter Slow Clock Disable Register) and PIO_IFSCER (PIO Input Filter Slow Clock Enable Register). Writing PIO_IFSCDR and PIO_IFSCER respectively, sets and clears bits in PIO_IFSCSR. The current selection status can be checked by reading the register PIO_IFSCSR (Input Filter Slow Clock Status Register).
z z
If PIO_IFSCSR[i] = 0: The glitch filter can filter a glitch with a duration of less than 1/2 Period of Master Clock. If PIO_IFSCSR[i] = 1: The debouncing filter can filter a pulse with a duration of less than 1/2 Period of the Programmable Divided Slow Clock.
For the debouncing filter, the Period of the Divided Slow Clock is performed by writing in the DIV field of the PIO_SCDR (Slow Clock Divider Register) Tdiv_slclk = ((DIV+1)*2).Tslow_clock When the glitch or debouncing filter is enabled, a glitch or pulse with a duration of less than 1/2 Selected Clock Cycle (Selected Clock represents MCK or Divided Slow Clock depending on PIO_IFSCDR and PIO_IFSCER programming) is automatically rejected, while a pulse with a duration of 1 Selected Clock (MCK or Divided Slow Clock) cycle or more is accepted. For pulse durations between 1/2 Selected Clock cycle and 1 Selected Clock cycle the pulse may or may not be taken into account, depending on the precise timing of its occurrence. Thus for a pulse to be visible it must exceed 1

243
Selected Clock cycle, whereas for a glitch to be reliably filtered out, its duration must not exceed 1/2 Selected Clock cycle. The filters also introduce some latencies, this is illustrated in Figure 26-5 and Figure 26-6. The glitch filters are controlled by the register set: PIO_IFER (Input Filter Enable Register), PIO_IFDR (Input Filter Disable Register) and PIO_IFSR (Input Filter Status Register). Writing PIO_IFER and PIO_IFDR respectively sets and clears bits in PIO_IFSR. This last register enables the glitch filter on the I/O lines. When the glitch and/or debouncing filter is enabled, it does not modify the behavior of the inputs on the peripherals. It acts only on the value read in PIO_PDSR and on the input change interrupt detection. The glitch and debouncing filters require that the PIO Controller clock is enabled.
Figure 26-5. Input Glitch Filter Timing
PIO_IFCSR = 0
MCK up to 1.5 cycles Pin Level 1 cycle PIO_PDSR if PIO_IFSR = 0 2 cycles PIO_PDSR if PIO_IFSR = 1 up to 2.5 cycles up to 2 cycles 1 cycle 1 cycle 1 cycle 1 cycle
Figure 26-6. Input Debouncing Filter Timing
PIO_IFCSR = 1
Divided Slow Clock
Pin Level up to 2 cycles Tmck PIO_PDSR if PIO_IFSR = 0 up to 2 cycles Tmck
1 cycle Tdiv_slclk PIO_PDSR if PIO_IFSR = 1 up to 1.5 cycles Tdiv_slclk
1 cycle Tdiv_slclk
up to 1.5 cycles Tdiv_slclk up to 2 cycles Tmck up to 2 cycles Tmck

244
26.5.10 Input Edge/Level Interrupt

The PIO Controller can be programmed to generate an interrupt when it detects an edge or a level on an I/O line. The Input Edge/Level Interrupt is controlled by writing PIO_IER (Interrupt Enable Register) and PIO_IDR (Interrupt Disable Register), which respectively enable and disable the input change interrupt by setting and clearing the corresponding bit in PIO_IMR (Interrupt Mask Register). As Input change detection is possible only by comparing two successive samplings of the input of the I/O line, the PIO Controller clock must be enabled. The Input Change Interrupt is available, regardless of the configuration of the I/O line, i.e. configured as an input only, controlled by the PIO Controller or assigned to a peripheral function. By default, the interrupt can be generated at any time an edge is detected on the input. Some additional Interrupt modes can be enabled/disabled by writing in the PIO_AIMER (Additional Interrupt Modes Enable Register) and PIO_AIMDR (Additional Interrupt Modes Disable Register). The current state of this selection can be read through the PIO_AIMMR (Additional Interrupt Modes Mask Register) These Additional Modes are:
z z z z
Rising Edge Detection Falling Edge Detection Low Level Detection High Level Detection The type of event detection (Edge or Level) must be selected by writing in the set of registers; PIO_ESR (Edge Select Register) and PIO_LSR (Level Select Register) which enable respectively, the Edge and Level Detection. The current status of this selection is accessible through the PIO_ELSR (Edge/Level Status Register). The Polarity of the event detection (Rising/Falling Edge or High/Low Level) must be selected by writing in the set of registers; PIO_FELLSR (Falling Edge /Low Level Select Register) and PIO_REHLSR (Rising Edge/High Level Select Register) which allow to select Falling or Rising Edge (if Edge is selected in the PIO_ELSR) Edge or High or Low Level Detection (if Level is selected in the PIO_ELSR). The current status of this selection is accessible through the PIO_FRLHSR (Fall/Rise - Low/High Status Register).
In order to select an Additional Interrupt Mode:

z
When an input Edge or Level is detected on an I/O line, the corresponding bit in PIO_ISR (Interrupt Status Register) is set. If the corresponding bit in PIO_IMR is set, the PIO Controller interrupt line is asserted.The interrupt signals of the thirty-two channels are ORed-wired together to generate a single interrupt signal to the interrupt controller. When the software reads PIO_ISR, all the interrupts are automatically cleared. This signifies that all the interrupts that are pending when PIO_ISR is read must be handled. When an Interrupt is enabled on a Level, the interrupt is generated as long as the interrupt source is not cleared, even if some read accesses in PIO_ISR are performed.

245
Figure 26-7. Event Detector on Input Lines (Figure represents line 0)

Event Detector
Rising Edge Detector Falling Edge Detector PIO_REHLSR[0] PIO_FRLHSR[0] PIO_FELLSR[0] Resynchronized input on line 0 High Level Detector Low Level Detector
1 1 0 1 1
0 0
Event detection on line 0
PIO_LSR[0] PIO_ELSR[0] PIO_ESR[0] PIO_AIMER[0] PIO_AIMMR[0] PIO_AIMDR[0] Edge Detector
26.5.10.1 Example If generating an interrupt is required on the following:

z z z z z z z z z
Rising edge on PIO line 0 Falling edge on PIO line 1 Rising edge on PIO line 2 Low Level on PIO line 3 High Level on PIO line 4 High Level on PIO line 5 Falling edge on PIO line 6 Rising edge on PIO line 7 Any edge on the other lines
The configuration required is described below. 26.5.10.2 Interrupt Mode Configuration All the interrupt sources are enabled by writing 32hFFFF_FFFF in PIO_IER. Then the Additional Interrupt Mode is enabled for line 0 to 7 by writing 32h0000_00FF in PIO_AIMER. 26.5.10.3 Edge or Level Detection Configuration Lines 3, 4 and 5 are configured in Level detection by writing 32h0000_0038 in PIO_LSR. The other lines are configured in Edge detection by default, if they have not been previously configured. Otherwise, lines 0, 1, 2, 6 and 7 must be configured in Edge detection by writing 32h0000_00C7 in PIO_ESR. 26.5.10.4 Falling/Rising Edge or Low/High Level Detection Configuration. Lines 0, 2, 4, 5 and 7 are configured in Rising Edge or High Level detection by writing 32h0000_00B5 in PIO_REHLSR. The other lines are configured in Falling Edge or Low Level detection by default, if they have not been previously configured. Otherwise, lines 1, 3 and 6 must be configured in Falling Edge/Low Level detection by writing 32h0000_004A in PIO_FELLSR.

246
Figure 26-8. Input Change Interrupt Timings if there are no Additional Interrupt Modes
MCK
Pin Level
PIO_ISR
Read PIO_ISR
APB Access
APB Access

247
26.5.11 I/O Lines Lock

When an I/O line is controlled by a peripheral (particularly the Pulse Width Modulation Controller PWM), it can become locked by the action of this peripheral via an input of the PIO controller. When an I/O line is locked, the write of the corresponding bit in the registers PIO_PER, PIO_PDR, PIO_MDER, PIO_MDDR, PIO_PUDR, PIO_PUER, PIO_ABCDSR1 and PIO_ABCDSR2 is discarded in order to lock its configuration. The user can know at anytime which I/O line is locked by reading the PIO Lock Status register PIO_LOCKSR. Once an I/O line is locked, the only way to unlock it is to apply a hardware reset to the PIO Controller.
26.5.12 Programmable I/O Drive

It is possible to configure the I/O drive for pads PA0 to PA31. For any details, refer to the product electrical characteristics.
26.5.13 Programmable Schmitt Trigger

It is possible to configure each input for the Schmitt Trigger. By default the Schmitt trigger is active. Disabling the Schmitt Trigger is requested when using the QTouch Library.
26.5.14 Write Protection Registers

To prevent any single software error that may corrupt PIO behavior, certain address spaces can be write-protected by setting the WPEN bit in the PIO Write Protect Mode Register (PIO_WPMR). If a write access to the protected registers is detected, then the WPVS flag in the PIO Write Protect Status Register (PIO_WPSR) is set and the field WPVSRC indicates in which register the write access has been attempted. The WPVS flag is reset by writing the PIO Write Protect Mode Register (PIO_WPMR) with the appropriate access key, WPKEY. The protected registers are:
z z z z z z z z z z z z z z z z
PIO Enable Register on page 252 PIO Disable Register on page 252 PIO Output Enable Register on page 253 PIO Output Disable Register on page 254 PIO Input Filter Enable Register on page 255 PIO Input Filter Disable Register on page 255 PIO Multi-driver Enable Register on page 260 PIO Multi-driver Disable Register on page 261 PIO Pull Up Disable Register on page 262 PIO Pull Up Enable Register on page 262 PIO Peripheral ABCD Select Register 1 on page 264 PIO Peripheral ABCD Select Register 2 on page 265 PIO Output Write Enable Register on page 270 PIO Output Write Disable Register on page 270 PIO Pad Pull Down Disable Register on page 268 PIO Pad Pull Down Status Register on page 269

248
26.6
I/O Lines Programming Example

The programing example as shown in Table 26-1 below is used to obtain the following configuration.
z z z z z z z z
4-bit output port on I/O lines 0 to 3, (should be written in a single write operation), open-drain, with pull-up resistor Four output signals on I/O lines 4 to 7 (to drive LEDs for example), driven high and low, no pull-up resistor, no pulldown resistor Four input signals on I/O lines 8 to 11 (to read push-button states for example), with pull-up resistors, glitch filters and input change interrupts Four input signals on I/O line 12 to 15 to read an external device status (polled, thus no input change interrupt), no pull-up resistor, no glitch filter I/O lines 16 to 19 assigned to peripheral A functions with pull-up resistor I/O lines 20 to 23 assigned to peripheral B functions with pull-down resistor I/O line 24 to 27 assigned to peripheral C with Input Change Interrupt, no pull-up resistor and no pull-down resistor I/O line 28 to 31 assigned to peripheral D, no pull-up resistor and no pull-down resistor
Table 26-1. Programming Example Register PIO_PER PIO_PDR PIO_OER PIO_ODR PIO_IFER PIO_IFDR PIO_SODR PIO_CODR PIO_IER PIO_IDR PIO_MDER PIO_MDDR PIO_PUDR PIO_PUER PIO_PPDDR PIO_PPDER PIO_ABCDSR1 PIO_ABCDSR2 PIO_OWER PIO_OWDR Value to be Written 0x0000_FFFF 0xFFFF_0000 0x0000_00FF 0xFFFF_FF00 0x0000_0F00 0xFFFF_F0FF 0x0000_0000 0x0FFF_FFFF 0x0F00_0F00 0xF0FF_F0FF 0x0000_000F 0xFFFF_FFF0 0xFFF0_00F0 0x000F_FF0F 0xFF0F_FFFF 0x00F0_0000 0xF0F0_0000 0xFF00_0000 0x0000_000F 0x0FFF_ FFF0

249
26.7
Parallel Input/Output Controller (PIO) User Interface

Each I/O line controlled by the PIO Controller is associated with a bit in each of the PIO Controller User Interface registers. Each register is 32 bits wide. If a parallel I/O line is not defined, writing to the corresponding bits has no effect. Undefined bits read zero. If the I/O line is not multiplexed with any peripheral, the I/O line is controlled by the PIO Controller and PIO_PSR returns 1 systematically.
Table 26-2. Register Mapping Offset 0x0000 0x0004 0x0008 0x000C 0x0010 0x0014 0x0018 0x001C 0x0020 0x0024 0x0028 0x002C 0x0030 0x0034 0x0038 0x003C 0x0040 0x0044 0x0048 0x004C 0x0050 0x0054 0x0058 0x005C 0x0060 0x0064 0x0068 0x006C 0x0070 0x0074 0x0078 to 0x007C 0x0080 0x0084 0x0088 Register PIO Enable Register PIO Disable Register PIO Status Register Reserved Output Enable Register Output Disable Register Output Status Register Reserved Glitch Input Filter Enable Register Glitch Input Filter Disable Register Glitch Input Filter Status Register Reserved Set Output Data Register Clear Output Data Register Output Data Status Register Pin Data Status Register Interrupt Enable Register Interrupt Disable Register Interrupt Mask Register Interrupt Status Register
(4)
Name PIO_PER PIO_PDR PIO_PSR PIO_OER PIO_ODR PIO_OSR PIO_IFER PIO_IFDR PIO_IFSR PIO_SODR PIO_CODR PIO_ODSR PIO_PDSR PIO_IER PIO_IDR PIO_IMR PIO_ISR PIO_MDER PIO_MDDR PIO_MDSR PIO_PUDR PIO_PUER PIO_PUSR PIO_ABCDSR1 PIO_ABCDSR2 PIO_IFSCDR PIO_IFSCER PIO_IFSCSR
Access Write-only Write-only Read-only Write-only Write-only Read-only Write-only Write-only Read-only Write-only Write-only Read-only or(2) Read-write Read-only Write-only Write-only Read-only Read-only Write-only Write-only Read-only Write-only Write-only Read-only Read-write Read-write Write-only Write-only Read-only
Reset
(1)
0x0000 0000 0x0000 0000
(3)
0x00000000 0x00000000 0x00000000

(1)
Multi-driver Enable Register Multi-driver Disable Register Multi-driver Status Register Reserved Pull-up Disable Register Pull-up Enable Register Pad Pull-up Status Register Reserved Peripheral Select Register 1 Peripheral Select Register 2 Reserved Input Filter Slow Clock Disable Register Input Filter Slow Clock Enable Register Input Filter Slow Clock Status Register
0x00000000 0x00000000 0x00000000

250
Table 26-2. Register Mapping (Continued) Offset 0x008C 0x0090 0x0094 0x0098 0x009C 0x00A0 0x00A4 0x00A8 0x00AC 0x00B0 0x00B4 0x00B8 0x00BC 0x00C0 0x00C4 0x00C8 0x00CC 0x00D0 0x00D4 0x00D8 0x00DC 0x00E0 0x00E4 0x00E8 0x00EC to 0x00F8 0x0100 0x01040x010C 0x0110 0x0114 0x0118 0x011C 0x0120 to 0x014C Register Slow Clock Divider Debouncing Register Pad Pull-down Disable Register Pad Pull-down Enable Register Pad Pull-down Status Register Reserved Output Write Enable Output Write Disable Output Write Status Register Reserved Additional Interrupt Modes Enable Register Additional Interrupt Modes Disables Register Additional Interrupt Modes Mask Register Reserved Edge Select Register Level Select Register Edge/Level Status Register Reserved Falling Edge/Low Level Select Register Rising Edge/ High Level Select Register Fall/Rise - Low/High Status Register Reserved Lock Status Write Protect Mode Register Write Protect Status Register Reserved Schmitt Trigger Register Reserved Reserved Reserved I/O Drive Register 1 I/O Drive Register 2 Reserved PIO_LOCKSR PIO_WPMR PIO_WPSR PIO_SCHMITT PIO_DRIVER1 PIO_DRIVER2 Read-only Read-write Read-only Read-write Read-write Read-write 0x00000000 0x0 0x0 0x00000000 0x00000000 0x00000000 Name PIO_SCDR PIO_PPDDR PIO_PPDER PIO_PPDSR PIO_OWER PIO_OWDR PIO_OWSR PIO_AIMER PIO_AIMDR PIO_AIMMR PIO_ESR PIO_LSR PIO_ELSR PIO_FELLSR PIO_REHLSR PIO_FRLHSR Access Read-write Write-only Write-only Read-only Write-only Write-only Read-only Write-only Write-only Read-only Write-only Write-only Read-only Write-only Write-only Read-only Reset 0x00000000
(1)
0x00000000 0x00000000 0x00000000 0x00000000
Notes: 1. Reset value depends on the product implementation. 2. PIO_ODSR is Read-only or Read/Write depending on PIO_OWSR I/O lines. 3. Reset value of PIO_PDSR depends on the level of the I/O lines. Reading the I/O line levels requires the clock of the PIO Controller to be enabled, otherwise PIO_PDSR reads the levels present on the I/O line at the time the clock was disabled. 4. PIO_ISR is reset at 0x0. However, the first read of the register may read a different value as input changes may have occurred. Note: If an offset is not listed in the table it must be considered as reserved.

251
26.7.1 PIO Enable Register

31
PIO_PER 0xFFFFF200 (PIOA), 0xFFFFF400 (PIOB), 0xFFFFF600 (PIOC), 0xFFFFF800 (PIOD), 0xFFFFFA00 (PIOE) Write-only
30 29 28 27 26 25 24
P31
23
P30
22
P29
21
P28
20
P27
19
P26
18
P25
17
P24
16
P23
15
P22
14
P21
13
P20
12
P19
11
P18
10
P17
9
P16
8
P15
7
P14
6
P13
5
P12
4
P11
3
P10
2
P9
1
P8
0
P7
P6
P5
P4
P3
P2
P1
P0
This register can only be written if the WPEN bit is cleared in PIO Write Protect Mode Register .
P0-P31: PIO Enable

0: No effect. 1: Enables the PIO to control the corresponding pin (disables peripheral control of the pin).
26.7.2 PIO Disable Register

31
PIO_PDR 0xFFFFF204 (PIOA), 0xFFFFF404 (PIOB), 0xFFFFF604 (PIOC), 0xFFFFF804 (PIOD), 0xFFFFFA04 (PIOE) Write-only
30 29 28 27 26 25 24
P31
23
P30
22
P29
21
P28
20
P27
19
P26
18
P25
17
P24
16
P23
15
P22
14
P21
13
P20
12
P19
11
P18
10
P17
9
P16
8
P15
7
P14
6
P13
5
P12
4
P11
3
P10
2
P9
1
P8
0
P7
P6
P5
P4
P3
P2
P1
P0
P0-P31: PIO Disable

0: No effect. 1: Disables the PIO from controlling the corresponding pin (enables peripheral control of the pin).

252
26.7.3 PIO Status Register

31
PIO_PSR 0xFFFFF208 (PIOA), 0xFFFFF408 (PIOB), 0xFFFFF608 (PIOC), 0xFFFFF808 (PIOD), 0xFFFFFA08 (PIOE) Read-only
30 29 28 27 26 25 24
P31
23
P30
22
P29
21
P28
20
P27
19
P26
18
P25
17
P24
16
P23
15
P22
14
P21
13
P20
12
P19
11
P18
10
P17
9
P16
8
P15
7
P14
6
P13
5
P12
4
P11
3
P10
2
P9
1
P8
0
P7
P6
P5
P4
P3
P2
P1
P0
P0-P31: PIO Status

0: PIO is inactive on the corresponding I/O line (peripheral is active). 1: PIO is active on the corresponding I/O line (peripheral is inactive).
26.7.4 PIO Output Enable Register

31
PIO_OER 0xFFFFF210 (PIOA), 0xFFFFF410 (PIOB), 0xFFFFF610 (PIOC), 0xFFFFF810 (PIOD), 0xFFFFFA10 (PIOE) Write-only
30 29 28 27 26 25 24
P31
23
P30
22
P29
21
P28
20
P27
19
P26
18
P25
17
P24
16
P23
15
P22
14
P21
13
P20
12
P19
11
P18
10
P17
9
P16
8
P15
7
P14
6
P13
5
P12
4
P11
3
P10
2
P9
1
P8
0
P7
P6
P5
P4
P3
P2
P1
P0
P0-P31: Output Enable

0: No effect. 1: Enables the output on the I/O line.

253
26.7.5 PIO Output Disable Register

31
PIO_ODR 0xFFFFF214 (PIOA), 0xFFFFF414 (PIOB), 0xFFFFF614 (PIOC), 0xFFFFF814 (PIOD), 0xFFFFFA14 (PIOE) Write-only
30 29 28 27 26 25 24
P31
23
P30
22
P29
21
P28
20
P27
19
P26
18
P25
17
P24
16
P23
15
P22
14
P21
13
P20
12
P19
11
P18
10
P17
9
P16
8
P15
7
P14
6
P13
5
P12
4
P11
3
P10
2
P9
1
P8
0
P7
P6
P5
P4
P3
P2
P1
P0
P0-P31: Output Disable

0: No effect. 1: Disables the output on the I/O line.
26.7.6 PIO Output Status Register

31
PIO_OSR 0xFFFFF218 (PIOA), 0xFFFFF418 (PIOB), 0xFFFFF618 (PIOC), 0xFFFFF818 (PIOD), 0xFFFFFA18 (PIOE) Read-only
30 29 28 27 26 25 24
P31
23
P30
22
P29
21
P28
20
P27
19
P26
18
P25
17
P24
16
P23
15
P22
14
P21
13
P20
12
P19
11
P18
10
P17
9
P16
8
P15
7
P14
6
P13
5
P12
4
P11
3
P10
2
P9
1
P8
0
P7
P6
P5
P4
P3
P2
P1
P0
P0-P31: Output Status

0: The I/O line is a pure input. 1: The I/O line is enabled in output.

254
26.7.7 PIO Input Filter Enable Register

31
PIO_IFER 0xFFFFF220 (PIOA), 0xFFFFF420 (PIOB), 0xFFFFF620 (PIOC), 0xFFFFF820 (PIOD), 0xFFFFFA20 (PIOE) Write-only
30 29 28 27 26 25 24
P31
23
P30
22
P29
21
P28
20
P27
19
P26
18
P25
17
P24
16
P23
15
P22
14
P21
13
P20
12
P19
11
P18
10
P17
9
P16
8
P15
7
P14
6
P13
5
P12
4
P11
3
P10
2
P9
1
P8
0
P7
P6
P5
P4
P3
P2
P1
P0
P0-P31: Input Filter Enable

0: No effect. 1: Enables the input glitch filter on the I/O line.
26.7.8 PIO Input Filter Disable Register

31
PIO_IFDR 0xFFFFF224 (PIOA), 0xFFFFF424 (PIOB), 0xFFFFF624 (PIOC), 0xFFFFF824 (PIOD), 0xFFFFFA24 (PIOE) Write-only
30 29 28 27 26 25 24
P31
23
P30
22
P29
21
P28
20
P27
19
P26
18
P25
17
P24
16
P23
15
P22
14
P21
13
P20
12
P19
11
P18
10
P17
9
P16
8
P15
7
P14
6
P13
5
P12
4
P11
3
P10
2
P9
1
P8
0
P7
P6
P5
P4
P3
P2
P1
P0
P0-P31: Input Filter Disable

0: No effect. 1: Disables the input glitch filter on the I/O line.

255
26.7.9 PIO Input Filter Status Register

31
PIO_IFSR 0xFFFFF228 (PIOA), 0xFFFFF428 (PIOB), 0xFFFFF628 (PIOC), 0xFFFFF828 (PIOD), 0xFFFFFA28 (PIOE) Read-only
30 29 28 27 26 25 24
P31
23
P30
22
P29
21
P28
20
P27
19
P26
18
P25
17
P24
16
P23
15
P22
14
P21
13
P20
12
P19
11
P18
10
P17
9
P16
8
P15
7
P14
6
P13
5
P12
4
P11
3
P10
2
P9
1
P8
0
P7
P6
P5
P4
P3
P2
P1
P0
P0-P31: Input Filer Status

0: The input glitch filter is disabled on the I/O line. 1: The input glitch filter is enabled on the I/O line.
26.7.10 PIO Set Output Data Register

31
PIO_SODR 0xFFFFF230 (PIOA), 0xFFFFF430 (PIOB), 0xFFFFF630 (PIOC), 0xFFFFF830 (PIOD), 0xFFFFFA30 (PIOE) Write-only
30 29 28 27 26 25 24
P31
23
P30
22
P29
21
P28
20
P27
19
P26
18
P25
17
P24
16
P23
15
P22
14
P21
13
P20
12
P19
11
P18
10
P17
9
P16
8
P15
7
P14
6
P13
5
P12
4
P11
3
P10
2
P9
1
P8
0
P7
P6
P5
P4
P3
P2
P1
P0
P0-P31: Set Output Data

0: No effect. 1: Sets the data to be driven on the I/O line.

256
26.7.11 PIO Clear Output Data Register

31
PIO_CODR 0xFFFFF234 (PIOA), 0xFFFFF434 (PIOB), 0xFFFFF634 (PIOC), 0xFFFFF834 (PIOD), 0xFFFFFA34 (PIOE) Write-only
30 29 28 27 26 25 24
P31
23
P30
22
P29
21
P28
20
P27
19
P26
18
P25
17
P24
16
P23
15
P22
14
P21
13
P20
12
P19
11
P18
10
P17
9
P16
8
P15
7
P14
6
P13
5
P12
4
P11
3
P10
2
P9
1
P8
0
P7
P6
P5
P4
P3
P2
P1
P0
P0-P31: Clear Output Data

0: No effect. 1: Clears the data to be driven on the I/O line.
26.7.12 PIO Output Data Status Register

31
PIO_ODSR 0xFFFFF238 (PIOA), 0xFFFFF438 (PIOB), 0xFFFFF638 (PIOC), 0xFFFFF838 (PIOD), 0xFFFFFA38 (PIOE) Read-only or Read-write
30 29 28 27 26 25 24
P31
23
P30
22
P29
21
P28
20
P27
19
P26
18
P25
17
P24
16
P23
15
P22
14
P21
13
P20
12
P19
11
P18
10
P17
9
P16
8
P15
7
P14
6
P13
5
P12
4
P11
3
P10
2
P9
1
P8
0
P7
P6
P5
P4
P3
P2
P1
P0
P0-P31: Output Data Status

0: The data to be driven on the I/O line is 0. 1: The data to be driven on the I/O line is 1.

257
26.7.13 PIO Pin Data Status Register

31
PIO_PDSR 0xFFFFF23C (PIOA), 0xFFFFF43C (PIOB), 0xFFFFF63C (PIOC), 0xFFFFF83C (PIOD), 0xFFFFFA3C (PIOE) Read-only
30 29 28 27 26 25 24
P31
23
P30
22
P29
21
P28
20
P27
19
P26
18
P25
17
P24
16
P23
15
P22
14
P21
13
P20
12
P19
11
P18
10
P17
9
P16
8
P15
7
P14
6
P13
5
P12
4
P11
3
P10
2
P9
1
P8
0
P7
P6
P5
P4
P3
P2
P1
P0
P0-P31: Output Data Status

0: The I/O line is at level 0. 1: The I/O line is at level 1.
26.7.14 PIO Interrupt Enable Register

31
PIO_IER 0xFFFFF240 (PIOA), 0xFFFFF440 (PIOB), 0xFFFFF640 (PIOC), 0xFFFFF840 (PIOD), 0xFFFFFA40 (PIOE) Write-only
30 29 28 27 26 25 24
P31
23
P30
22
P29
21
P28
20
P27
19
P26
18
P25
17
P24
16
P23
15
P22
14
P21
13
P20
12
P19
11
P18
10
P17
9
P16
8
P15
7
P14
6
P13
5
P12
4
P11
3
P10
2
P9
1
P8
0
P7
P6
P5
P4
P3
P2
P1
P0
P0-P31: Input Change Interrupt Enable

0: No effect. 1: Enables the Input Change Interrupt on the I/O line.

258
26.7.15 PIO Interrupt Disable Register

31
PIO_IDR 0xFFFFF244 (PIOA), 0xFFFFF444 (PIOB), 0xFFFFF644 (PIOC), 0xFFFFF844 (PIOD), 0xFFFFFA44 (PIOE) Write-only
30 29 28 27 26 25 24
P31
23
P30
22
P29
21
P28
20
P27
19
P26
18
P25
17
P24
16
P23
15
P22
14
P21
13
P20
12
P19
11
P18
10
P17
9
P16
8
P15
7
P14
6
P13
5
P12
4
P11
3
P10
2
P9
1
P8
0
P7
P6
P5
P4
P3
P2
P1
P0
P0-P31: Input Change Interrupt Disable

0: No effect. 1: Disables the Input Change Interrupt on the I/O line.
26.7.16 PIO Interrupt Mask Register

31
PIO_IMR 0xFFFFF248 (PIOA), 0xFFFFF448 (PIOB), 0xFFFFF648 (PIOC), 0xFFFFF848 (PIOD), 0xFFFFFA48 (PIOE) Read-only
30 29 28 27 26 25 24
P31
23
P30
22
P29
21
P28
20
P27
19
P26
18
P25
17
P24
16
P23
15
P22
14
P21
13
P20
12
P19
11
P18
10
P17
9
P16
8
P15
7
P14
6
P13
5
P12
4
P11
3
P10
2
P9
1
P8
0
P7
P6
P5
P4
P3
P2
P1
P0
P0-P31: Input Change Interrupt Mask

0: Input Change Interrupt is disabled on the I/O line. 1: Input Change Interrupt is enabled on the I/O line.

259
26.7.17 PIO Interrupt Status Register

31
PIO_ISR 0xFFFFF24C (PIOA), 0xFFFFF44C (PIOB), 0xFFFFF64C (PIOC), 0xFFFFF84C (PIOD), 0xFFFFFA4C (PIOE) Read-only
30 29 28 27 26 25 24
P31
23
P30
22
P29
21
P28
20
P27
19
P26
18
P25
17
P24
16
P23
15
P22
14
P21
13
P20
12
P19
11
P18
10
P17
9
P16
8
P15
7
P14
6
P13
5
P12
4
P11
3
P10
2
P9
1
P8
0
P7
P6
P5
P4
P3
P2
P1
P0
P0-P31: Input Change Interrupt Status

0: No Input Change has been detected on the I/O line since PIO_ISR was last read or since reset. 1: At least one Input Change has been detected on the I/O line since PIO_ISR was last read or since reset.
26.7.18 PIO Multi-driver Enable Register

31
PIO_MDER 0xFFFFF250 (PIOA), 0xFFFFF450 (PIOB), 0xFFFFF650 (PIOC), 0xFFFFF850 (PIOD), 0xFFFFFA50 (PIOE) Write-only
30 29 28 27 26 25 24
P31
23
P30
22
P29
21
P28
20
P27
19
P26
18
P25
17
P24
16
P23
15
P22
14
P21
13
P20
12
P19
11
P18
10
P17
9
P16
8
P15
7
P14
6
P13
5
P12
4
P11
3
P10
2
P9
1
P8
0
P7
P6
P5
P4
P3
P2
P1
P0
P0-P31: Multi Drive Enable.

0: No effect. 1: Enables Multi Drive on the I/O line.

260
26.7.19 PIO Multi-driver Disable Register

31
PIO_MDDR 0xFFFFF254 (PIOA), 0xFFFFF454 (PIOB), 0xFFFFF654 (PIOC), 0xFFFFF854 (PIOD), 0xFFFFFA54 (PIOE) Write-only
30 29 28 27 26 25 24
P31
23
P30
22
P29
21
P28
20
P27
19
P26
18
P25
17
P24
16
P23
15
P22
14
P21
13
P20
12
P19
11
P18
10
P17
9
P16
8
P15
7
P14
6
P13
5
P12
4
P11
3
P10
2
P9
1
P8
0
P7
P6
P5
P4
P3
P2
P1
P0
P0-P31: Multi Drive Disable.

0: No effect. 1: Disables Multi Drive on the I/O line.
26.7.20 PIO Multi-driver Status Register

31
PIO_MDSR 0xFFFFF258 (PIOA), 0xFFFFF458 (PIOB), 0xFFFFF658 (PIOC), 0xFFFFF858 (PIOD), 0xFFFFFA58 (PIOE) Read-only
30 29 28 27 26 25 24
P31
23
P30
22
P29
21
P28
20
P27
19
P26
18
P25
17
P24
16
P23
15
P22
14
P21
13
P20
12
P19
11
P18
10
P17
9
P16
8
P15
7
P14
6
P13
5
P12
4
P11
3
P10
2
P9
1
P8
0
P7
P6
P5
P4
P3
P2
P1
P0
P0-P31: Multi Drive Status.

0: The Multi Drive is disabled on the I/O line. The pin is driven at high and low level. 1: The Multi Drive is enabled on the I/O line. The pin is driven at low level only.

261
26.7.21 PIO Pull Up Disable Register

31
PIO_PUDR 0xFFFFF260 (PIOA), 0xFFFFF460 (PIOB), 0xFFFFF660 (PIOC), 0xFFFFF860 (PIOD), 0xFFFFFA60 (PIOE) Write-only
30 29 28 27 26 25 24
P31
23
P30
22
P29
21
P28
20
P27
19
P26
18
P25
17
P24
16
P23
15
P22
14
P21
13
P20
12
P19
11
P18
10
P17
9
P16
8
P15
7
P14
6
P13
5
P12
4
P11
3
P10
2
P9
1
P8
0
P7
P6
P5
P4
P3
P2
P1
P0
P0-P31: Pull Up Disable.

0: No effect. 1: Disables the pull up resistor on the I/O line.
26.7.22 PIO Pull Up Enable Register

31
PIO_PUER 0xFFFFF264 (PIOA), 0xFFFFF464 (PIOB), 0xFFFFF664 (PIOC), 0xFFFFF864 (PIOD), 0xFFFFFA64 (PIOE) Write-only
30 29 28 27 26 25 24
P31
23
P30
22
P29
21
P28
20
P27
19
P26
18
P25
17
P24
16
P23
15
P22
14
P21
13
P20
12
P19
11
P18
10
P17
9
P16
8
P15
7
P14
6
P13
5
P12
4
P11
3
P10
2
P9
1
P8
0
P7
P6
P5
P4
P3
P2
P1
P0
P0-P31: Pull Up Enable.

0: No effect. 1: Enables the pull up resistor on the I/O line.

262
26.7.23 PIO Pull Up Status Register

31
PIO_PUSR 0xFFFFF268 (PIOA), 0xFFFFF468 (PIOB), 0xFFFFF668 (PIOC), 0xFFFFF868 (PIOD), 0xFFFFFA68 (PIOE) Read-only
30 29 28 27 26 25 24
P31
23
P30
22
P29
21
P28
20
P27
19
P26
18
P25
17
P24
16
P23
15
P22
14
P21
13
P20
12
P19
11
P18
10
P17
9
P16
8
P15
7
P14
6
P13
5
P12
4
P11
3
P10
2
P9
1
P8
0
P7
P6
P5
P4
P3
P2
P1
P0
P0-P31: Pull Up Status.

0: Pull Up resistor is enabled on the I/O line. 1: Pull Up resistor is disabled on the I/O line.

263
26.7.24 PIO Peripheral ABCD Select Register 1

Name: Access:
31
PIO_ABCDSR1 Read-write
30 29 28 27 26 25 24
P31
23
P30
22
P29
21
P28
20
P27
19
P26
18
P25
17
P24
16
P23
15
P22
14
P21
13
P20
12
P19
11
P18
10
P17
9
P16
8
P15
7
P14
6
P13
5
P12
4
P11
3
P10
2
P9
1
P8
0
P7
P6
P5
P4
P3
P2
P1
P0
P0-P31: Peripheral Select. If the same bit is set to 0 in PIO_ABCDSR2:

0: Assigns the I/O line to the Peripheral A function. 1: Assigns the I/O line to the Peripheral B function.
If the same bit is set to 1 in PIO_ABCDSR2:

0: Assigns the I/O line to the Peripheral C function. 1: Assigns the I/O line to the Peripheral D function.

264
26.7.25 PIO Peripheral ABCD Select Register 2

Name: Access:
31
PIO_ABCDSR2 Read-write
30 29 28 27 26 25 24
P31
23
P30
22
P29
21
P28
20
P27
19
P26
18
P25
17
P24
16
P23
15
P22
14
P21
13
P20
12
P19
11
P18
10
P17
9
P16
8
P15
7
P14
6
P13
5
P12
4
P11
3
P10
2
P9
1
P8
0
P7
P6
P5
P4
P3
P2
P1
P0
P0-P31: Peripheral Select. If the same bit is set to 0 in PIO_ABCDSR1:

0: Assigns the I/O line to the Peripheral A function. 1: Assigns the I/O line to the Peripheral C function.
If the same bit is set to 1 in PIO_ABCDSR1:

0: Assigns the I/O line to the Peripheral B function. 1: Assigns the I/O line to the Peripheral D function.

265
26.7.26 PIO Input Filter Slow Clock Disable Register

31
PIO_IFSCDR 0xFFFFF280 (PIOA), 0xFFFFF480 (PIOB), 0xFFFFF680 (PIOC), 0xFFFFF880 (PIOD), 0xFFFFFA80 (PIOE) Write-only
30 29 28 27 26 25 24
P31
23
P30
22
P29
21
P28
20
P27
19
P26
18
P25
17
P24
16
P23
15
P22
14
P21
13
P20
12
P19
11
P18
10
P17
9
P16
8
P15
7
P14
6
P13
5
P12
4
P11
3
P10
2
P9
1
P8
0
P7
P6
P5
P4
P3
P2
P1
P0
P0-P31: PIO Clock Glitch Filtering Select.

0: No Effect. 1: The Glitch Filter is able to filter glitches with a duration < Tmck/2.
26.7.27 PIO Input Filter Slow Clock Enable Register

31
PIO_IFSCER 0xFFFFF284 (PIOA), 0xFFFFF484 (PIOB), 0xFFFFF684 (PIOC), 0xFFFFF884 (PIOD), 0xFFFFFA84 (PIOE) Write-only
30 29 28 27 26 25 24
P31
23
P30
22
P29
21
P28
20
P27
19
P26
18
P25
17
P24
16
P23
15
P22
14
P21
13
P20
12
P19
11
P18
10
P17
9
P16
8
P15
7
P14
6
P13
5
P12
4
P11
3
P10
2
P9
1
P8
0
P7
P6
P5
P4
P3
P2
P1
P0
P0-P31: Debouncing Filtering Select.

0: No Effect. 1: The Debouncing Filter is able to filter pulses with a duration < Tdiv_slclk/2.

266
26.7.28 PIO Input Filter Slow Clock Status Register

31
PIO_IFSCSR 0xFFFFF288 (PIOA), 0xFFFFF488 (PIOB), 0xFFFFF688 (PIOC), 0xFFFFF888 (PIOD), 0xFFFFFA88 (PIOE) Read-only
30 29 28 27 26 25 24
P31
23
P30
22
P29
21
P28
20
P27
19
P26
18
P25
17
P24
16
P23
15
P22
14
P21
13
P20
12
P19
11
P18
10
P17
9
P16
8
P15
7
P14
6
P13
5
P12
4
P11
3
P10
2
P9
1
P8
0
P7
P6
P5
P4
P3
P2
P1
P0
P0-P31: Glitch or Debouncing Filter Selection Status

0: The Glitch Filter is able to filter glitches with a duration < Tmck2. 1: The Debouncing Filter is able to filter pulses with a duration < Tdiv_slclk/2.
26.7.29 PIO Slow Clock Divider Debouncing Register

31
PIO_SCDR 0xFFFFF28C (PIOA), 0xFFFFF48C (PIOB), 0xFFFFF68C (PIOC), 0xFFFFF88C (PIOD), 0xFFFFFA8C (PIOE) Read-write
30 29 28 27 26 25 24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
6 5 4 3
DIV
2 1 0
DIV
DIVx: Slow Clock Divider Selection for Debouncing

Tdiv_slclk = 2*(DIV+1)*Tslow_clock.

267
26.7.30 PIO Pad Pull Down Disable Register

31
PIO_PPDDR 0xFFFFF290 (PIOA), 0xFFFFF490 (PIOB), 0xFFFFF690 (PIOC), 0xFFFFF890 (PIOD), 0xFFFFFA90 (PIOE) Write-only
30 29 28 27 26 25 24
P31
23
P30
22
P29
21
P28
20
P27
19
P26
18
P25
17
P24
16
P23
15
P22
14
P21
13
P20
12
P19
11
P18
10
P17
9
P16
8
P15
7
P14
6
P13
5
P12
4
P11
3
P10
2
P9
1
P8
0
P7
P6
P5
P4
P3
P2
P1
P0
P0-P31: Pull Down Disable.

0: No effect. 1: Disables the pull down resistor on the I/O line.
26.7.31 PIO Pad Pull Down Enable Register

31
PIO_PPDER 0xFFFFF294 (PIOA), 0xFFFFF494 (PIOB), 0xFFFFF694 (PIOC), 0xFFFFF894 (PIOD), 0xFFFFFA94 (PIOE) Write-only
30 29 28 27 26 25 24
P31
23
P30
22
P29
21
P28
20
P27
19
P26
18
P25
17
P24
16
P23
15
P22
14
P21
13
P20
12
P19
11
P18
10
P17
9
P16
8
P15
7
P14
6
P13
5
P12
4
P11
3
P10
2
P9
1
P8
0
P7
P6
P5
P4
P3
P2
P1
P0
P0-P31: Pull Down Enable.

0: No effect. 1: Enables the pull down resistor on the I/O line.

268
26.7.32 PIO Pad Pull Down Status Register

31
PIO_PPDSR 0xFFFFF298 (PIOA), 0xFFFFF498 (PIOB), 0xFFFFF698 (PIOC), 0xFFFFF898 (PIOD), 0xFFFFFA98 (PIOE) Read-only
30 29 28 27 26 25 24
P31
23
P30
22
P29
21
P28
20
P27
19
P26
18
P25
17
P24
16
P23
15
P22
14
P21
13
P20
12
P19
11
P18
10
P17
9
P16
8
P15
7
P14
6
P13
5
P12
4
P11
3
P10
2
P9
1
P8
0
P7
P6
P5
P4
P3
P2
P1
P0
P0-P31: Pull Down Status.

0: Pull Down resistor is enabled on the I/O line. 1: Pull Down resistor is disabled on the I/O line.

269
26.7.33 PIO Output Write Enable Register

31
PIO_OWER 0xFFFFF2A0 (PIOA), 0xFFFFF4A0 (PIOB), 0xFFFFF6A0 (PIOC), 0xFFFFF8A0 (PIOD), 0xFFFFFAA0 (PIOE) Write-only
30 29 28 27 26 25 24
P31
23
P30
22
P29
21
P28
20
P27
19
P26
18
P25
17
P24
16
P23
15
P22
14
P21
13
P20
12
P19
11
P18
10
P17
9
P16
8
P15
7
P14
6
P13
5
P12
4
P11
3
P10
2
P9
1
P8
0
P7
P6
P5
P4
P3
P2
P1
P0
P0-P31: Output Write Enable.

0: No effect. 1: Enables writing PIO_ODSR for the I/O line.
26.7.34 PIO Output Write Disable Register

31
PIO_OWDR 0xFFFFF2A4 (PIOA), 0xFFFFF4A4 (PIOB), 0xFFFFF6A4 (PIOC), 0xFFFFF8A4 (PIOD), 0xFFFFFAA4 (PIOE) Write-only
30 29 28 27 26 25 24
P31
23
P30
22
P29
21
P28
20
P27
19
P26
18
P25
17
P24
16
P23
15
P22
14
P21
13
P20
12
P19
11
P18
10
P17
9
P16
8
P15
7
P14
6
P13
5
P12
4
P11
3
P10
2
P9
1
P8
0
P7
P6
P5
P4
P3
P2
P1
P0
P0-P31: Output Write Disable.

0: No effect. 1: Disables writing PIO_ODSR for the I/O line.

270
26.7.35 PIO Output Write Status Register

31
PIO_OWSR 0xFFFFF2A8 (PIOA), 0xFFFFF4A8 (PIOB), 0xFFFFF6A8 (PIOC), 0xFFFFF8A8 (PIOD), 0xFFFFFAA8 (PIOE) Read-only
30 29 28 27 26 25 24
P31
23
P30
22
P29
21
P28
20
P27
19
P26
18
P25
17
P24
16
P23
15
P22
14
P21
13
P20
12
P19
11
P18
10
P17
9
P16
8
P15
7
P14
6
P13
5
P12
4
P11
3
P10
2
P9
1
P8
0
P7
P6
P5
P4
P3
P2
P1
P0
P0-P31: Output Write Status.

0: Writing PIO_ODSR does not affect the I/O line. 1: Writing PIO_ODSR affects the I/O line.
26.7.36 PIO Additional Interrupt Modes Enable Register

31
PIO_AIMER 0xFFFFF2B0 (PIOA), 0xFFFFF4B0 (PIOB), 0xFFFFF6B0 (PIOC), 0xFFFFF8B0 (PIOD), 0xFFFFFAB0 (PIOE) Write-only
30 29 28 27 26 25 24
P31
23
P30
22
P29
21
P28
20
P27
19
P26
18
P25
17
P24
16
P23
15
P22
14
P21
13
P20
12
P19
11
P18
10
P17
9
P16
8
P15
7
P14
6
P13
5
P12
4
P11
3
P10
2
P9
1
P8
0
P7
P6
P5
P4
P3
P2
P1
P0
P0-P31: Additional Interrupt Modes Enable.

0: No effect. 1: The interrupt source is the event described in PIO_ELSR and PIO_FRLHSR.

271
26.7.37 PIO Additional Interrupt Modes Disable Register

31
PIO_AIMDR 0xFFFFF2B4 (PIOA), 0xFFFFF4B4 (PIOB), 0xFFFFF6B4 (PIOC), 0xFFFFF8B4 (PIOD), 0xFFFFFAB4 (PIOE) Write-only
30 29 28 27 26 25 24
P31
23
P30
22
P29
21
P28
20
P27
19
P26
18
P25
17
P24
16
P23
15
P22
14
P21
13
P20
12
P19
11
P18
10
P17
9
P16
8
P15
7
P14
6
P13
5
P12
4
P11
3
P10
2
P9
1
P8
0
P7
P6
P5
P4
P3
P2
P1
P0
P0-P31: Additional Interrupt Modes Disable.

0: No effect. 1: The interrupt mode is set to the default interrupt mode (Both Edge detection).
26.7.38 PIO Additional Interrupt Modes Mask Register

31
PIO_AIMMR 0xFFFFF2B8 (PIOA), 0xFFFFF4B8 (PIOB), 0xFFFFF6B8 (PIOC), 0xFFFFF8B8 (PIOD), 0xFFFFFAB8 (PIOE) Read-only
30 29 28 27 26 25 24
P31
23
P30
22
P29
21
P28
20
P27
19
P26
18
P25
17
P24
16
P23
15
P22
14
P21
13
P20
12
P19
11
P18
10
P17
9
P16
8
P15
7
P14
6
P13
5
P12
4
P11
3
P10
2
P9
1
P8
0
P7
P6
P5
P4
P3
P2
P1
P0
P0-P31: Peripheral CD Status.

0: The interrupt source is a Both Edge detection event 1: The interrupt source is described by the registers PIO_ELSR and PIO_FRLHSR

272
26.7.39 PIO Edge Select Register

31
PIO_ESR 0xFFFFF2C0 (PIOA), 0xFFFFF4C0 (PIOB), 0xFFFFF6C0 (PIOC), 0xFFFFF8C0 (PIOD), 0xFFFFFAC0 (PIOE) Write-only
30 29 28 27 26 25 24
P31
23
P30
22
P29
21
P28
20
P27
19
P26
18
P25
17
P24
16
P23
15
P22
14
P21
13
P20
12
P19
11
P18
10
P17
9
P16
8
P15
7
P14
6
P13
5
P12
4
P11
3
P10
2
P9
1
P8
0
P7
P6
P5
P4
P3
P2
P1
P0
P0-P31: Edge Interrupt Selection.

0: No effect. 1: The interrupt source is an Edge detection event.
26.7.40 PIO Level Select Register

31
PIO_LSR 0xFFFFF2C4 (PIOA), 0xFFFFF4C4 (PIOB), 0xFFFFF6C4 (PIOC), 0xFFFFF8C4 (PIOD), 0xFFFFFAC4 (PIOE) Write-only
30 29 28 27 26 25 24
P31
23
P30
22
P29
21
P28
20
P27
19
P26
18
P25
17
P24
16
P23
15
P22
14
P21
13
P20
12
P19
11
P18
10
P17
9
P16
8
P15
7
P14
6
P13
5
P12
4
P11
3
P10
2
P9
1
P8
0
P7
P6
P5
P4
P3
P2
P1
P0
P0-P31: Level Interrupt Selection.

0: No effect. 1: The interrupt source is a Level detection event.

273
26.7.41 PIO Edge/Level Status Register

31
PIO_ELSR 0xFFFFF2C8 (PIOA), 0xFFFFF4C8 (PIOB), 0xFFFFF6C8 (PIOC), 0xFFFFF8C8 (PIOD), 0xFFFFFAC8 (PIOE) Read-only
30 29 28 27 26 25 24
P31
23
P30
22
P29
21
P28
20
P27
19
P26
18
P25
17
P24
16
P23
15
P22
14
P21
13
P20
12
P19
11
P18
10
P17
9
P16
8
P15
7
P14
6
P13
5
P12
4
P11
3
P10
2
P9
1
P8
0
P7
P6
P5
P4
P3
P2
P1
P0
P0-P31: Edge/Level Interrupt source selection.

0: The interrupt source is an Edge detection event. 1: The interrupt source is a Level detection event.
26.7.42 PIO Falling Edge/Low Level Select Register

31
PIO_FELLSR 0xFFFFF2D0 (PIOA), 0xFFFFF4D0 (PIOB), 0xFFFFF6D0 (PIOC), 0xFFFFF8D0 (PIOD), 0xFFFFFAD0 (PIOE) Write-only
30 29 28 27 26 25 24
P31
23
P30
22
P29
21
P28
20
P27
19
P26
18
P25
17
P24
16
P23
15
P22
14
P21
13
P20
12
P19
11
P18
10
P17
9
P16
8
P15
7
P14
6
P13
5
P12
4
P11
3
P10
2
P9
1
P8
0
P7
P6
P5
P4
P3
P2
P1
P0
P0-P31: Falling Edge/Low Level Interrupt Selection.

0: No effect. 1: The interrupt source is set to a Falling Edge detection or Low Level detection event, depending on PIO_ELSR.

274
26.7.43 PIO Rising Edge/High Level Select Register

31
PIO_REHLSR 0xFFFFF2D4 (PIOA), 0xFFFFF4D4 (PIOB), 0xFFFFF6D4 (PIOC), 0xFFFFF8D4 (PIOD), 0xFFFFFAD4 (PIOE) Write-only
30 29 28 27 26 25 24
P31
23
P30
22
P29
21
P28
20
P27
19
P26
18
P25
17
P24
16
P23
15
P22
14
P21
13
P20
12
P19
11
P18
10
P17
9
P16
8
P15
7
P14
6
P13
5
P12
4
P11
3
P10
2
P9
1
P8
0
P7
P6
P5
P4
P3
P2
P1
P0
P0-P31: Rising Edge /High Level Interrupt Selection.

0: No effect. 1: The interrupt source is set to a Rising Edge detection or High Level detection event, depending on PIO_ELSR.

275
26.7.44 PIO Fall/Rise - Low/High Status Register

31
PIO_FRLHSR 0xFFFFF2D8 (PIOA), 0xFFFFF4D8 (PIOB), 0xFFFFF6D8 (PIOC), 0xFFFFF8D8 (PIOD), 0xFFFFFAD8 (PIOE) Read-only
30 29 28 27 26 25 24
P31
23
P30
22
P29
21
P28
20
P27
19
P26
18
P25
17
P24
16
P23
15
P22
14
P21
13
P20
12
P19
11
P18
10
P17
9
P16
8
P15
7
P14
6
P13
5
P12
4
P11
3
P10
2
P9
1
P8
0
P7
P6
P5
P4
P3
P2
P1
P0
P0-P31: Edge /Level Interrupt Source Selection.

0: The interrupt source is a Falling Edge detection (if PIO_ELSR = 0) or Low Level detection event (if PIO_ELSR = 1). 1: The interrupt source is a Rising Edge detection (if PIO_ELSR = 0) or High Level detection event (if PIO_ELSR = 1).

276
26.7.45 PIO Lock Status Register

31
PIO_LOCKSR 0xFFFFF2E0 (PIOA), 0xFFFFF4E0 (PIOB), 0xFFFFF6E0 (PIOC), 0xFFFFF8E0 (PIOD), 0xFFFFFAE0 (PIOE) Read-only
30 29 28 27 26 25 24
P31
23
P30
22
P29
21
P28
20
P27
19
P26
18
P25
17
P24
16
P23
15
P22
14
P21
13
P20
12
P19
11
P18
10
P17
9
P16
8
P15
7
P14
6
P13
5
P12
4
P11
3
P10
2
P9
1
P8
0
P7
P6
P5
P4
P3
P2
P1
P0
P0-P31: Lock Status.

0: The I/O line is not locked. 1: The I/O line is locked.

277
26.7.46 PIO Write Protect Mode Register

31
PIO_WPMR 0xFFFFF2E4 (PIOA), 0xFFFFF4E4 (PIOB), 0xFFFFF6E4 (PIOC), 0xFFFFF8E4 (PIOD), 0xFFFFFAE4 (PIOE) Read-write See Table 26-2
For more information on Write Protection Registers, refer to Section 26.7 Parallel Input/Output Controller (PIO) User Interface.

0: Disables the Write Protect if WPKEY corresponds to 0x50494F (PIO in ASCII). 1: Enables the Write Protect if WPKEY corresponds to 0x50494F (PIO in ASCII). Protects the registers:
PIO Enable Register on page 252 PIO Disable Register on page 252 PIO Output Enable Register on page 253 PIO Output Disable Register on page 254 PIO Input Filter Enable Register on page 255 PIO Input Filter Disable Register on page 255 PIO Multi-driver Enable Register on page 260 PIO Multi-driver Disable Register on page 261 PIO Pull Up Disable Register on page 262 PIO Pull Up Enable Register on page 262 PIO Peripheral ABCD Select Register 1 on page 264 PIO Peripheral ABCD Select Register 2 on page 265 PIO Output Write Enable Register on page 270 PIO Output Write Disable Register on page 270 PIO Pad Pull Down Disable Register on page 268 PIO Pad Pull Down Status Register on page 269 WPKEY: Write Protect KEY
Should be written at value 0x50494F (PIO in ASCII). Writing any other value in this field aborts the write operation of the WPEN bit. Always reads as 0.

278
26.7.47 PIO Write Protect Status Register

31
PIO_WPSR 0xFFFFF2E8 (PIOA), 0xFFFFF4E8 (PIOB), 0xFFFFF6E8 (PIOC), 0xFFFFF8E8 (PIOD), 0xFFFFFAE8 (PIOE) Read-only See Table 26-2
30 29 28 27 26 25 24
23
22
21
20 WPVSRC
19
18
17
16
15
14
13
12 WPVSRC
11
10
0 WPVS

0: No Write Protect Violation has occurred since the last read of the PIO_WPSR register. 1: A Write Protect Violation has occurred since the last read of the PIO_WPSR register. If this violation is an unauthorized attempt to write a protected register, the associated violation is reported into field WPVSRC.

When WPVS is active, this field indicates the write-protected register (through address offset or code) in which a write access has been attempted. Note: Reading PIO_WPSR automatically clears all fields.

279
26.7.48 PIO Schmitt Trigger Register

31 SCHMITT31 23 SCHMITT23 15 SCHMITT15 7 SCHMITT7
PIO_SCHMITT 0xFFFFF300 (PIOA), 0xFFFFF500 (PIOB), 0xFFFFF700 (PIOC), 0xFFFFF900 (PIOD), 0xFFFFFB00 (PIOE) Read-write See Table 26-2
30 SCHMITT30 22 SCHMITT22 14 SCHMITT14 6 SCHMITT6 29 SCHMITT29 21 SCHMITT21 13 SCHMITT13 5 SCHMITT5 28 SCHMITT28 20 SCHMITT20 12 SCHMITT12 4 SCHMITT4 27 SCHMITT27 19 SCHMITT19 11 SCHMITT11 3 SCHMITT3 26 SCHMITT26 18 SCHMITT18 10 SCHMITT10 2 SCHMITT2 25 SCHMITT25 17 SCHMITT17 9 SCHMITT9 1 SCHMITT1 24 SCHMITT24 16 SCHMITT16 8 SCHMITT8 0 SCHMITT0
SCHMITTx [x=0..31]:
0: Schmitt Trigger is enabled. 1: Schmitt Trigger is disabled.

280
26.7.49 PIO I/O Drive Register 1

31 LINE15 23 LINE11 15 LINE7 7 LINE3 6 5 LINE2 14 13 LINE6 4 3 LINE1 22 21 LINE10 12 11 LINE5 2 1 LINE0
PIO_DRIVER1 0xFFFFF318 (PIOA), 0xFFFFF518 (PIOB), 0xFFFFF718 (PIOC), 0xFFFFF918 (PIOD), 0xFFFFFB18 (PIOE) Read-write 0x0
30 29 LINE14 20 19 LINE9 10 9 LINE4 0 28 27 LINE13 18 17 LINE8 8 26 25 LINE12 16 24
LINEx [x=0..15]: Drive of PIO Line x

Value 0 1 2 3 Name LO_DRIVE LO_DRIVE ME_DRIVE HI_DRIVE Description Low drive Low drive Medium drive High drive

281
26.7.50 PIO I/O Drive Register 2

31 LINE31 23 LINE27 15 LINE23 7 LINE19 6 5 LINE18 14 13 LINE22 4 3 LINE17 22 21 LINE26 12 11 LINE21 2 1 LINE16
PIO_DRIVER2 0xFFFFF31C (PIOA), 0xFFFFF51C (PIOB), 0xFFFFF71C (PIOC), 0xFFFFF91C (PIOD), 0xFFFFFB1C (PIOE) Read-write 0x0
30 29 LINE30 20 19 LINE25 10 9 LINE20 0 28 27 LINE29 18 17 LINE24 8 26 25 LINE28 16 24
LINEx [x=16..31]: Drive of PIO line x

Value 0 1 2 3 Name LO_DRIVE LO_DRIVE ME_DRIVE HI_DRIVE Description Low drive Low drive Medium drive High drive

282
27.
External Memories
The product features:
z z
Multiport DDR Controller (MPDDRC) External Bus Interface (EBI) that embeds a NAND Flash controller and a Static Memory Controller (HSMC)
Figure 27-1. External Memory Controllers

MPDDRC Port 3 Port 2 Port 1 Port 0 LPDDR DDR2 LPDDR2-S4 Device
EBI NAND Flash Controller Bus Matrix Static Memory Controller Static Memory Device NAND Flash Device
z z
MPDDRC is a standalone multi-port DDRSDR controller. It supports only DDR2, LPDDR, and LPDDR2-S4 devices. Its user interface is located at 0xFFFFEA00. HSMC supports Static Memories and MLC/SLC NAND Flashes. It embeds Multi-Bit ECC. Its user interface is located at 0xFFFFC000.

283
27.1
MPDDRC Multi-port DDRSDR Controller
27.1.1 Description
The DDR2 Controller is dedicated to 8-port DDR2/LPDDR/LPDDR2 support. Data transfers are performed through a 32bit data bus on one chip select. The Controller operates with 1.8V Power Supply for DDR2 and LP-DDR, 1.2V Power Supply for LP-DDR2.
27.1.2 Embedded Characteristics

27.1.2.1 DDR2/LPDDR/LPDDR2 Controller Four AHB Interfaces, Management of All Accesses Maximizes Memory Bandwidth and Minimizes Transaction Latency.
z
Supports AHB Transfers:

z
DWord, Word, Half Word, Byte Access.
z z
Supports Low-Power DDR2-SDRAM-S4, DDR2-SDRAM, Low-Power DDR1-SDRAM Numerous Configurations Supported

z z z z
2K, 4K, 8K, 16K Row Address Memory Parts DDR2 with Four or Eight Internal Banks (DDR2_SDRAM/Low-Power DDR2-SDRAM) DDR2/LPDDR with 32-bit Data Path One Chip Select for DDR2/LPDDR Device (512 Mbytes Address Space) Multibank Ping-pong Access (Up to 4 or 8 Banks Opened at Same Time = Reduces Average Latency of Transactions) Timing Parameters Specified by Software Automatic Refresh Operation, Refresh Rate is Programmable Automatic Update of DS, TCR and PASR Parameters (Low-power DDR-SDRAM Devices) Self-refresh, Power-down, Active Power-down and Deep Power-down Modes Supported
Programming Facilities
z z z z
Energy-saving Capabilities
z
z z z z z z
Power-up Initialization by Software CAS Latency of 2, 3, 4, 5, 6 supported Reset function supported (DDR2) Auto Precharge Command Not Used On Die Termination not supported OCD mode not supported

284
27.1.3 MPDDR Controller Block Diagram

Figure 27-2. Organization of the MPDDRC
MPDDRC
DDR_A0-DDR_A13 DDR_D0-DDR_D31 Bus Matrix DDR_CS DDR_CKE DDR_RAS, DDR_CAS AHB DDR2 LPDDR LPDDR2-S4 Controller DDR_CLK,#DDR_CLK DDR_DQS[3:0] DDR_DQSN[3:0] DDR_DQM[3:0] DDR_WE DDR_BA[2:0] Address Decoders DDR_CALP DDR_CALN
DDR_VREF User Interface
APB

285
27.1.4 I/O Lines Description

Table 27-1. DDR2 I/O Lines Description Name Function DDR2/LPDDR Controller VDDIODDR DDR_VREF DDR_CALP DDR_CALN DDR_D0 - DDR_D31 DDR_A0 - DDR_A13 DDR_DQM0 - DDR_DQM3 DDR_DQS0 - DDR_DQS3 DDR_DQSN0 DDR_DQSN3 DDR_CS DDR_CLK - DDR_CLK# DDR_CKE DDR_RAS DDR_CAS DDR_WE DDR_BA0 - DDR_BA2 Power Supply of memory interface Reference Voltage for DDR2 operations, typically 0.9V Pad positive calibration reference for LP-DDR2 Pad negative calibration reference for LP-DDR2 Data Bus Address Bus Data Mask Data Strobe Negative Data Strobe Chip Select DDR2 Differential Clock Clock enable Row signal Column signal Write enable Bank Select Input Input Input Input I/O Output Output Output Output Output Output Output Output Output Output Output High Low Low Low Low Type Active Level
In LPDDR2 mode, DQS and DQSN are connected to the LPDDR2 memory.

286
27.1.5 Product Dependencies

The pins used for interfacing the DDR2 memory are not multiplexed with the PIO lines.
Table 27-2. DDR2 I/O Lines Usage vs Operating Modes Signal Name DDR_VREF DDR_CALP DDR_CALN DDR_CK, DDR_CKN DDR_CKE DDR_CS DDR_BA[2..0] DDR_WE DDR_RAS - DDR_CAS DDR_A[13..0] DDR_D[31..0] DQS[3..0], DQSN[3..0] DQM[3..0] DDR2 Mode VDDIODDR/2 GND via 200 resistor VDDIODDR via 200 resistor CLK and CLKN CLKE CS BA[2..0] WE RAS, CAS A[13:0] D[31:0] DQS[3:0] DQSN connected to DDR_VREF DQM[3..0] LPDDR2 Mode VDDIODDR/2 GND via 240 resistor VDDIODDR via 240 resistor CLK and CLKN CLKE CS BA[2..0] CA2 CA0, CA1 CAx, with x>2 D[31:0] DQS[3:0] DQSN[3:0] DQM[3..0] LPDDR VDDIODDR/2 GND via 200 resistor VDDIODDR via 200 resistor CLK and CLKN CLKE CS BA[2..0] WE RAS, CAS A[13:0] D[31:0] DQS[3:0] DQSN connected to DDR_VREF DQM[3..0]

287
27.1.6 Implementation Example

The following hardware configuration is given for illustration only. The user should refer to the memory manufacturer web site to check current device availability. 27.1.6.1 2x16-bit DDR2
Hardware Configuration
Figure 27-3. 2x16-bit DDR2 Hardware Configuration
Software Configuration
The following configuration has to be performed:
z
Initialize the DDR2 Controller depending on the DDR2 device and system bus frequency.
The DDR2 initialization sequence is described in the sub-section DDR2 Device Initialization of the DDRSDRC section.

288
27.1.6.2 2x16-bit LPDDR2
Figure 27-4. 2x16-bit LPDDR2 Hardware Configuration
CAx LPDDR2 signals are to be connected as indicated in Table 27-3:
Table 27-3.
CAx LP-DDR2 Signal Connection LP-DDR2 Signal
DDR Controller Signal
RAS CAS WE DDR_A0 DDR_A1 DDR_A2 DDR_A3 DDR_A4 DDR_A5 DDR_A6

Higher addresses
CA0 CA1 CA2 CA3 CA4 CA5 CA6 CA7 CA8 CA9
Higher CAs
z
Initialize the DDR2 Controller depending on the LPDDR2 device and system bus frequency.
The DDR2 initialization sequence is described in the sub-section LPDDR2 Device Initialization of the DDRSDRC section.

289
27.2
External Bus Interface (EBI)
27.2.1 Description
The External Bus Interface is designed to ensure the successful data transfer between several external devices and the ARM processor-based device. The External Bus Interface of the device consists of a Static Memory Controller (HSMC). This HSMC is capable of handling several types of external memory and peripheral devices, such as SRAM, PSRAM, PROM, EPROM, EEPROM, LCD Module, NOR Flash and NAND Flash. The HSMC generates the signals that control the access to external memory devices or peripheral devices. It has 4 Chip Selects and a 26-bit address bus. The 16-bit data bus can be configured to interface with 8- or 16-bit external devices. Separate read and write control signals allow for direct memory and peripheral interfacing. Read and write signal waveforms are fully parametrizable. The HSMC can manage wait requests from external devices to extend the current access. The HSMC is provided with an automatic slow clock mode. In slow clock mode, it switches from user-programmed waveforms to slow-rate specific waveforms on read and write signals. The HSMC embeds a NAND Flash Controller (NFC). The NFC can handle automatic transfers, sending the commands and address cycles to the NAND Flash and transferring the contents of the page (for read and write) to the NFC SRAM. It minimizes the CPU overhead. The HSMC includes programmable hardware error correcting code with one-bit error correction capability and supports two-bit error detection. In order to improve the overall system performance, the DATA phase of the transfer can be DMAassisted. The External Data Bus can be scrambled/unscrambled by means of user keys. The full description is available in the HSMC section. 27.2.1.1 External Bus Interface (EBI)
z
Integrates Two External Memory Controllers:

z z
Static Memory Controller SLC/MLC Nand Flash Controller
z z z z z
Additional logic for NAND Flash Optional 16-bit External Data Bus Up to 26-bit Address Bus (up to 64 MBytes linear per chip select) Up to 4 chip selects, Configurable Assignment NAND Flash chip select is programmable:
27.2.1.2 Static Memory Controller (HSMC)

z z z z z z z z z z z z z
64-MByte Address Space per Chip Select 8- or 16-bit Data Bus Word, Halfword, Byte Transfers Byte Write or Byte Select Lines Programmable Setup, Pulse and Hold Time for Read Signals per Chip Select Programmable Setup, Pulse and Hold Time for Write Signals per Chip Select Programmable Data Float Time per Chip Select External Data Bus Scrambling/Unscrambling Function External Wait Request Automatic Switch to Slow Clock Mode NAND Flash Controller Supporting NAND Flash with Multiplexed Data/Address Buses Supports SLC and MLC NAND Flash Technology Programmable Timing on a per Chip Select Basis

290
27.2.1.3 NAND Flash Controller (NFC)

z z z z z z
Programmable Flash Data Width 8 bits or 16 bits. Supports NAND Flash and SmartMedia Devices with 8- or 16-bit Data Path. Supports 1-bit Correction for a Page of 512, 1024, 2048 and 4096 Bytes with 8- or 16-bit Data Path. Supports 1-bit Correction per 512 Bytes of Data for a Page Size of 512, 2048 and 4096 Bytes with 8-bit Data Path. Supports 1-bit Correction per 256 Bytes of Data for a Page Size of 512, 2048 and 4096 Bytes with 8-bit Data Path. Multibit Error Correcting Code (ECC)
ECC Algorithm based on binary shortened Bose, Chaudhuri and Hocquenghem (BCH) codes. Programmable Error Correcting Capability: 2, 4, 8, 12 and 24 bits of errors per block. Programmable block size: 512 Bytes or 1024 Bytes. Programmable number of block per page: 1, 2, 4 or 8 blocks of data per page. Programmable spare area size. Supports spare area ECC protection. Supports 8 kBytes page size using 1024 Bytes/block and 4 kBytes page size using 512 Bytes/block. Multibit Error detection is interrupt driven. Provides hardware acceleration for determining roots of polynomials defined over a finite field Programmable finite Field GF(2^13) or GF(2^14) Finds roots of error-locator polynomial. Programmable finite Field GF(2^13) or GF(2^14) Finds roots of error-locator polynomial.

291
27.2.2 Implementation Examples

The following hardware configurations are given for illustration only. The user should refer to the memory manufacturer web site to check current device availability. 27.2.2.1 8-bit NAND Flash
z z z z z z
Select the NAND Flash Chip Select by setting the field CSID in NFCADDR_CMD register. Configure the NFC and HSMC according to the used NAND Flash. Enable the NFC with NFCEN bit in HSMC_CTRL register in HSMC User interface. Reserve A21/A22 for ALE/CLE functions. Address and Command Latches are controlled respectively by setting to 1 the address bits A21 and A22 during accesses. Configure a PIO line as an input to manage the Ready/Busy signal. Configure Static Memory Controller CS3 Setup, Pulse, Cycle and Mode accordingly to NAND Flash timings, the data bus width and the system bus frequency.

292
27.2.2.2 16-bit NAND Flash
The software configuration is the same as for an 8-bit NAND Flash except for the data bus width programmed in the mode register of the Static Memory Controller.

293
27.2.2.3 NOR Flash on NCS0
D[0..15] A[1..22] U1 A1 A2 A3 A4 A5 A6 A7 A8 A9 A10 A11 A12 A13 A14 A15 A16 A17 A18 A19 A20 A21 A22
25 24 23 22 21 20 19 18 8 7 6 5 4 3 2 1 48 17 16 15 10 9 12 11 14 13 26 28
A0 A1 A2 A3 A4 A5 A6 A7 A8 A9 A10 A11 A12 A13 A14 A15 A16 A17 A18 A19 A20 A21 RESET WE WP VPP CE OE
DQ0 DQ1 DQ2 DQ3 DQ4 DQ5 DQ6 DQ7 DQ8 DQ9 DQ10 DQ11 DQ12 DQ13 DQ14 DQ15
AT49BV6416
29 31 33 35 38 40 42 44 30 32 34 36 39 41 43 45
D0 D1 D2 D3 D4 D5 D6 D7 D8 D9 D10 D11 D12 D13 D14 D15
3V3
VCCQ VCC VSS VSS
47 37 46 27
NRST NWE 3V3 NCS0 NRD
C2 100NF
C1 100NF
TSOP48 PACKAGE
The default configuration for the Static Memory Controller, byte select mode, 16-bit data bus, Read/Write controlled by Chip Select, allows boot on 16-bit non-volatile memory at slow clock. For another configuration, configure the Static Memory Controller CS0 Setup, Pulse, Cycle and Mode depending on Flash timings and system bus frequency.

294
28.
28.1
Multi-port DDR-SDRAM Controller (MPDDRC)

Description
The Multi-port DDR-SDRAM Controller (MPDDRC) is a multiport memory controller. It comprises four slave AHB interfaces. All simultaneous accesses (four independent AHB ports) are interleaved to maximize memory bandwidth and minimize transaction latency due to DDR-SDRAM protocol. The MPDDRC extends the memory capabilities of a chip by providing the interface to an external 32-bit DDR-SDRAM device. The page size supports ranges from 2048 to 16384 and the number of columns from 256 to 4096. It supports dword (64-bits), word (32-bit), half-word (16-bit), byte (8-bit) accesses. The MPDDRC supports a read or write burst length of 8 locations which frees the command and address bus to anticipate the next command, thus reducing latency imposed by the DDR-SDRAM protocol and improving the DDRSDRAM bandwidth. Moreover, it keeps track of the active row in each bank, thus maximizing DDR-SDRAM performance, e.g., the application may be placed in one bank and data in the other banks. To optimize performance, it is advisable to avoid accessing different rows in the same bank. The MPDDRC supports a CAS latency of 2, 3, 4, 5 or 6 and optimizes the read access depending on the frequency. The features of self refresh, power-down and deep power-down modes minimize the consumption of the DDR-SDRAM device. The MPDDRC user interface is compliant with ARM Advanced Peripheral Bus (APB rev2).

295
28.2
z z z z
Four Advanced High Performance Bus (AHB) Interfaces, Management of All Accesses Maximizes Memory Bandwidth and Minimizes Transaction Latency Bus Transfer: Dword, Word, Half Word, Byte Access Supports Low-power DDR2-SDRAM-S4 (LPDDR2), DDR2-SDRAM, Low-power DDR1-SDRAM (LPDDR1) Numerous Configurations Supported
z z z z
2K, 4K, 8K, 16K Row Address Memory Parts DDR-SDRAM with Four or Eight Internal Banks (DDR2-SDRAM/ Low-power DDR2-SDRAM-S4) DDR-SDRAM with 32-bit Data Path for System Oriented Dword Access One Chip Select for SDRAM Device (512-Mbyte Address Space) Multibank Ping-pong Access (Up to 4 or 8 Banks Opened at the Same Time = Reduced Average Latency of Transactions) Timing Parameters Specified by Software Automatic Refresh Operation, Refresh Rate is Programmable Automatic Update of DS, TCR and PASR Parameters (Low-power DDR-SDRAM Devices) Self-refresh, Power-down, Active Power-down and Deep Power-down Modes Supported
Programming Facilities
z z z z
Energy-saving Capabilities
z
z z z z z z z z
DDR-SDRAM Power-up Initialization by Software CAS Latency of 2, 3, 4, 5, 6 Supported Reset Function Supported (DDR2-SDRAM) ODT (On-die Termination) Not Supported Auto-refresh per bank Supported (Low-power DDR2-SDRAM-S4) Automatic Adjust Refresh Rate (Low-power DDR2-SDRAM-S4) Auto-precharge Command Not Used OCD (Off-chip Driver) Mode Not Supported

296
28.3
MPDDRC Module Diagram
Figure 28-1. MPDDRC Module Diagram
AHB MPDDR Controller
DDR Controller AHB Slave Interface 0 Input Stage Power Management
clk/nclk AHB Slave Interface 1 Input Stage Output Stage AHB Slave Interface 2 Input Stage Arbiter Memory Controller Finite State Machine SDRAM Signal Management ras, cas, we, cke Addr, DQM DDR-Devices DQS Data odt AHB Slave Interface 3 Input Stage Asynchronous Timing Refresh Management
Interconnect Matrix
APB
Interface APB
MPDDRC is partitioned in two blocks (see Figure 28-1):

z z
An Interconnect-Matrix that manages concurrent accesses on the AHB bus between four AHB masters and integrates an arbiter. A controller that translates AHB requests (Read/Write) in the DDR-SDRAM protocol.

297
28.4
Product Dependencies, Initialization Sequence
28.4.1 Low-power DDR1-SDRAM Initialization

The initialization sequence is generated by software. The Low-power DDR1-SDRAM devices are initialized by the following sequence: 1. 2. Program the memory device type into the Memory Device Register (see Section 28.7.15 on page 339). Program the features of the Low-power DDR1-SDRAM device into the Configuration Register: asynchronous timing (TRC, MPDDRC, etc.), number of columns, rows, CAS latency. See Section 28.7.3 on page 321, Section 28.7.4 on page 324 and Section 28.7.5 on page 326. Program temperature compensated self refresh (TCR), Partial array self refresh (PASR) and Drive strength (DS) into the Low-power Register. See Section 28.7.7 on page 328. A NOP command is issued to the Low-power DDR1-SDRAM. Program the NOP command into the Mode Register, the application must set MODE to 1 in the Mode Register (see Section 28.7.1 on page 319). Perform a write access to any Low-power DDR1-SDRAM address to acknowledge this command. Now, clocks which drive Lowpower DDR1-SDRAM device are enabled. A NOP command is issued to the Low-power DDR1-SDRAM. Program the NOP command into the Mode Register, the application must set MODE to 1 in the Mode Register (see Section 28.7.1 on page 319). Perform a write access to any Low-power DDR1-SDRAM address to acknowledge this command. Now, a calibration request is done to the I/O Pad. An all banks precharge command is issued to the Low-power DDR1-SDRAM. Program all banks precharge command into the Mode Register, the application must set MODE to 2 in the Mode Register (see Section 28.7.1 on page 319). Perform a write access to any Low-power DDR1-SDRAM address to acknowledge this command Two auto-refresh (CBR) cycles are provided. Program the auto refresh command (CBR) into the Mode Register, the application must set MODE to 4 in the Mode Register (see Section 28.7.1 on page 319). Perform a write access to any Low-power DDR1-SDRAM location twice to acknowledge these commands. An Extended Mode Register set (EMRS) cycle is issued to program the Low-power DDR1-SDRAM parameters (TCSR, PASR, DS). The application must set MODE to 5 in the Mode Register (see Section 28.7.1 on page 319) and perform a write access to the SDRAM to acknowledge this command. The write address must be chosen so that BA[1] is set to 1 and BA[0] is set to 0. For example, with a 32-bit 1Gbit SDRAM (14 rows, 10 columns, 4 banks), the SDRAM write access should be done at the address (BASE_ADDRESS_DDR + 0x08000000). This address is for example purposes only. The real address is dependent on implementation in the product.
3.
A minimum pause of 200 s is provided to precede any signal toggle. 4.
5.
6.
7.
8.
9.
10. A Mode Register set (MRS) cycle is issued to program the parameters of the DDR1-SDRAM devices, in particular CAS latency, burst length. The application must set MODE to 3 in the Mode Register (see Section 28.7.1 on page 319) and perform a write access to the Low-power DDR1-SDRAM to acknowledge this command. The write address must be chosen so that BA[1:0] bits are set to 0. For example, the SDRAM write access should be done at the address (BASE_ADDRESS_DDR). 11. The application must go into Normal Mode, setting MODE to 0 in the Mode Register (see Section 28.7.1 on page 319) and performing a write access at any location in the Low-power DDR1-SDRAM to acknowledge this command. 12. Perform a write access to any Low-power DDR1-SDRAM address. 13. Write the refresh rate into the COUNT field in the MPDDRC Refresh Timer register (see page 320). (Refresh rate = delay between refresh cycles). The Low-power DDR1-SDRAM device requires a refresh every 15.625 s or 7.81 s. With a 100 MHz frequency, the refresh timer count register must be set with (15.625 * 100 MHz) = 1562 i.e. 0x061A or (7.81 *100 MHz) = 781 i.e. 0x030d. 14. After initialization, the Low-power DDR1-SDRAM device is fully functional.

298
28.4.2 DDR2-SDRAM Initialization

The initialization sequence is generated by software. The DDR2-SDRAM devices are initialized by the following sequence: 1. 2. Program the memory device type into the Memory Device Register (see Section 28.7.15 on page 339). Program the features of DDR2-SDRAM device into the Timing Register (asynchronous timing (TRC TRAS, etc.)), and into the Configuration Register (number of columns, rows, banks, CAS latency and output driver impedance control) (see Section 28.7.3 on page 321, Section 28.7.4 on page 324 and Section 28.7.5 on page 326). A NOP command is issued to the DDR2-SDRAM. Program the NOP command into the Mode Register, the application must set MODE to 1 in the Mode Register (see Section 28.7.1 on page 319). Perform a write access to any DDR2-SDRAM address to acknowledge this command. Now, clocks which drive the DDR2-SDRAM device are enabled. A NOP command is issued to the DDR2-SDRAM. Program the NOP command into the Mode Register, the application must set MODE to 1 in the Mode Register (see Section 28.7.1 on page 319). Perform a write access to any DDR2-SDRAM address to acknowledge this command. Now, CKE is driven high. An all banks precharge command is issued to the DDR2-SDRAM. Program all banks precharge command into the Mode Register, the application must set MODE to 2 in the Mode Register (see Section 28.7.1 on page 319). Perform a write access to any DDR2-SDRAM address to acknowledge this command. An Extended Mode Register set (EMRS2) cycle is issued to choose between commercial or high temperature operations. The application must set MODE to 5 in the Mode Register (see Section 28.7.1 on page 319) and perform a write access to the DDR2-SDRAM to acknowledge this command. The write address must be chosen so that BA[1] is set to 1 and BA[0] are set to 0. For example, with a 32-bit 1Gbit SDRAM (14 rows, 10 columns, 8 banks), the SDRAM write access should be done at the address (BASE_ADDRESS_DDR + 0x08000000) This address is for example purposes only. The real address is dependent on implementation in the product. An Extended Mode Register set (EMRS3) cycle is issued to set the Extended Mode Register to 0. The application must set MODE to 5 in the Mode Register (see Section 28.7.1 on page 319) and perform a write access to the DDR2-SDRAM to acknowledge this command. The write address must be chosen so that BA[1] is set to 1 and BA[0] is set to 1. For example, with a 32-bit 1Gbit SDRAM (14 rows, 10 columns, 8 banks), the SDRAM write access should be done at the address (BASE_ADDRESS_DDR + 0x0C000000). An Extended Mode Register set (EMRS1) cycle is issued to enable DLL and to program D.I.C. (Output Driver Impedance Control). The application must set MODE to 5 in the Mode Register (see Section 28.7.1 on page 319) and perform a write access to the DDR2-SDRAM to acknowledge this command. The write address must be chosen so that BA[1] is set to 0 and BA[0] is set to 1.For example, with a 32-bit 1Gbit SDRAM (14 rows, 10 columns, 8 banks), the SDRAM write access should be done at the address (BASE_ADDRESS_DDR + 0x04000000).
3.
A minimum pause of 200 s is provided to precede any signal toggle. 4.
5.
6.
7. 8.
9.
10. An additional 200 cycles of clock are required for locking DLL 11. Program DLL field into the Configuration Register (see Section 28.7.3 on page 321) to high (Enable DLL reset). 12. A Mode Register set (MRS) cycle is issued to reset DLL. The application must set MODE to 3 in the Mode Register (see Section 28.7.1 on page 319) and perform a write access to the DDR2-SDRAM to acknowledge this command. The write address must be chosen so that BA[1:0] bits are set to 0. For example, the SDRAM write access should be done at the address (BASE_ADDRESS_DDR). 13. An all banks precharge command is issued to the DDR2-SDRAM. Program all banks precharge command into the Mode Register, the application must set MODE to 2 in the Mode Register (see Section 28.7.1 on page 319). Perform a write access to any DDR2-SDRAM address to acknowledge this command 14. Two auto-refresh (CBR) cycles are provided. Program the auto refresh command (CBR) into the Mode Register, the application must set MODE to 4 in the Mode Register (see Section 28.7.1 on page 319). Performs a write access to any DDR2-SDRAM location twice to acknowledge these commands. 15. Program DLL field into the Configuration Register (see Section 28.7.3 on page 321) to low (Disable DLL reset). 16. A Mode Register set (MRS) cycle is issued to program the parameters of the DDR2-SDRAM devices, in particular CAS latency, burst length and to disable DLL reset. The application must set MODE to 3 in the Mode Register (see

299
Section 28.7.1 on page 319) and perform a write access to the DDR2-SDRAM to acknowledge this command. The write address must be chosen so that BA[1:0] are set to 0. 17. Program OCD field into the Configuration Register (see Section 28.7.3 on page 321) to high (OCD calibration default). 18. An Extended Mode Register set (EMRS1) cycle is issued to OCD default value. The application must set MODE to 5 in the Mode Register (see Section 28.7.1 on page 319) and perform a write access to the DDR2-SDRAM to acknowledge this command. The write address must be chosen so that BA[1] is set to 0 and BA[0] is set to 1. For example, with a 32-bit 1Gbit SDRAM (14 rows, 10 columns, 8 banks), the SDRAM write access should be done at the address (BASE_ADDRESS_DDR + 0x04000000). 19. Program OCD field in the Configuration Register (see Section 28.7.3 on page 321) to low (OCD calibration mode exit). 20. An Extended Mode Register set (EMRS1) cycle is issued to enable OCD exit. The application must set MODE to 5 in the Mode Register (see Section 28.7.1 on page 319) and perform a write access to the DDR2-SDRAM to acknowledge this command. The write address must be chosen so that BA[1] is set to 0 and BA[0] is set to 1. For example, with a 32-bit 1Gbit SDRAM (14 rows, 10 columns, 8 banks), the SDRAM write access should be done at the address (BASE_ADDRESS_DDR + 0x04000000). 21. A Normal mode command is provided. Program the normal mode into Mode Register (see Section 28.7.1 on page 319). Perform a write access to any DDR2-SDRAM address to acknowledge this command. 22. Perform a write access to any DDR2-SDRAM address. 23. Write the refresh rate into the count field in the Refresh Timer register (see page 320). (Refresh rate = delay between refresh cycles). The DDR2-SDRAM device requires a refresh every 15.625 s or 7.81 s. With a 133 MHz frequency, the refresh timer count register must be set with (15.625 *133 MHz) = 2079 i.e. 0x081f or (7.81 *133 MHz) = 1039 i.e. 0x040f After initialization, the DDR2-SDRAM devices are fully functional.
28.4.3 Low-power DDR2-SDRAM Initialization

The initialization sequence is generated by software. The Low-power DDR2-SDRAM devices are initialized by the following sequence: 1. 2. Program the memory device type into the Memory Device Register (see Section 28.7.15 on page 339). Program the features of Low-power DDR2-SDRAM device into the Timing Register (asynchronous timing, TRC, TRAS, etc.) and into the Configuration Register (number of columns, rows, banks, CAS latency and output drive strength) (see Section 28.7.3 on page 321, Section 28.7.4 on page 324 and Section 28.7.5 on page 326). A NOP command is issued to the Low-power DDR2-SDRAM. Program the NOP command into the Mode Register, the application must set the MODE (MPDDRC Command Mode) field to 1 in the Mode Register (see Section 28.7.1 on page 319). Perform a write access to any Low-power DDR2-SDRAM address to acknowledge this command. Now, clocks which drive Low-power DDR2-SDRAM devices are enabled. A NOP command is issued to the Low-power DDR2-SDRAM. Program the NOP command into the Mode Register, the application must set MODE to 1 in the Mode Register (see Section 28.7.1 on page 319). Perform a write access to any Low-power DDR2-SDRAM address to acknowledge this command. Now, CKE is driven high. A reset command is issued to the Low-power DDR2-SDRAM. Program LPDDR2_CMD in the MODE (MPDDRC Command Mode) and MRS (Mode Register Select LPDDR2) field of the Mode Register, the application must set MODE to 7 and MRS to 63. (see Section 28.7.1 on page 319). Perform a write access to any Low-power DDR2SDRAM address to acknowledge this command. Now, the reset command is issued. A Mode Register Read command is issued to the Low-power DDR2-SDRAM. Program LPDDR2_CMD in the MODE and MRS field of the Mode Register, the application must set MODE to 7 and must set MRS field to 0.
3.
A minimum pause of 100 ns must be observed to precede any signal toggle. 4.
A minimum pause of 200 s must be satisfied before Reset Command. 5.
A minimum pause of 1 s must be satisfied before any commands. 6.

300
(See Section 28.7.1 on page 319). Perform a write access to any Low-power DDR2-SDRAM address to acknowledge this command. Now, the Mode Register Read command is issued. A minimum pause of 10 s must be satisfied before any commands. 7. A calibration command is issued to the Low-power DDR2-SDRAM. Program the type of calibration into the Configuration Register, ZQ field, RESET value (see Section 28.7.3 MPDDRC Configuration Register on page 321). In the Mode Register, program the MODE field to LPDDR2_CMD value, and the MRS field; the application must set MODE to 7 and MRS to 10 (see Section 28.7.1 MPDDRC Mode Register on page 319). Perform a write access to any Low-power DDR2-SDRAM address to acknowledge this command. Now, the ZQ Calibration command is issued. Program the type of calibration into the Configuration Register, ZQ field, SHORT value (see Section 28.7.3 MPDDRC Configuration Register on page 321). A Mode Register Write command is issued to the Low-power DDR2-SDRAM. Program LPDDR2_CMD in the MODE and MRS field in the Mode Register, the application must set MODE to 7 and must set MRS field to 1. (see Section 28.7.1 on page 319). The Mode Register Write command cycle is issued to program the parameters of the Low-power DDR2-SDRAM devices, in particular burst length. Perform a write access to any Low-power DDR2SDRAM address to acknowledge this command. Now, the Mode Register Write command is issued. A Mode Register Write Command is issued to the Low-power DDR2-SDRAM. Program LPDDR2_CMD in the MODE and MRS field in the Mode Register, the application must set MODE to 7 and must set MRS field to 2. (see Section 28.7.1 on page 319). The Mode Register Write command cycle is issued to program the parameters of the Low-power DDR2-SDRAM devices, in particular CAS latency. Perform a write access to any Low-power DDR2SDRAM address to acknowledge this command. Now, the Mode Register Write command is issued.
8.
9.
10. A Mode Register Write Command is issued to the Low-power DDR2-SDRAM. Program LPDDR2_CMD in the MODE and MRS field of the Mode Register, the application must set MODE to 7 and must set MRS field to 3. (see Section 28.7.1 on page 319). The Mode Register Write command cycle is issued to program the parameters of the Low-power DDR2-SDRAM devices, in particular Drive Strength and Slew Rate. Perform a write access to any Low-power DDR2-SDRAM address to acknowledge this command. Now, the Mode Register Write command is issued. 11. A Mode Register Write Command is issued to the Low-power DDR2-SDRAM. Program LPDDR2_CMD in the MODE and MRS field of the Mode Register, the application must set MODE to 7 and must set MRS field to 16. (see Section 28.7.1 on page 319). Mode Register Write command cycle is issued to program the parameters of the Low-power DDR2-SDRAM devices, in particular Partial Array Self Refresh (PASR). Perform a write access to any Low-power DDR2-SDRAM address to acknowledge this command. Now, the Mode Register Write command is issued. 12. Write the refresh rate into the COUNT field in the Refresh Timer register (see page 320). (Refresh rate = delay between refresh cycles). The Low-power DDR2-SDRAM device requires a refresh every 7.81 s. With a 133 MHz frequency, the refresh timer count register must be set with (7.81 *133 MHz) = 1039 i.e. 0x040f. After initialization, the Low-power DDR2-SDRAM devices are fully functional.

301
28.5
28.5.1 DDR-SDRAM Controller Write Cycle

The MPDDRC allows burst access or single access in normal mode (mode = 000). Whatever the access type, the MPDDRC keeps track of the active row in each bank, thus maximizing performance. The DDR-SDRAM device is programmed with a burst length equal to 8. This determines the length of a sequential data input by the write command that is set to 8. The latency from write command to data input is fixed to 1 in the case of DDR1-SDRAM devices and is fixed to 2/3/4/5 in the case of DDR2-SDRAM in function of programmed latency. To initiate a single access, the MPDDRC checks if the page access is already open. If row/bank addresses match with the previous row/bank addresses, the controller generates a write command. If the bank addresses are not identical or if bank addresses are identical but the row addresses are not identical, the controller generates a precharge command, activates the new row and initiates a write command. To comply with DDR-SDRAM timing parameters, additional clock cycles are inserted between precharge/active (tRP) commands and active/write (tRCD) command. As the burst length is fixed to 8, in the case of single access, it has to stop the burst, otherwise seven invalid values may be written. In the case of a DDR-SDRAM device, Burst Stop command is not supported for the burst write operation. In order to then interrupt the write operation, DM must be set to 1 to mask invalid data (see Figure 28-2 on page 303 and Figure 28-4 on page 304) and DQS must continue to toggle. To initiate a burst access, the MPDDRC controller uses the transfer type signal provided by the master requesting the access. If the next access is a sequential write access, writing to the DDR-SDRAM device is carried out. If the next access is a write non-sequential access, then an automatic access break is inserted, the MPDDRC generates a precharge command, activates the new row and initiates a write command. To comply with DDR-SDRAM timing parameters, additional clock cycles are inserted between precharge/active (t RP) commands and active/write (tRCD) commands. For a definition of timing parameters, refer to Section 28.7.4 MPDDRC Timing Parameter 0 Register on page 324. Write accesses to the DDR-SDRAM device are burst oriented and the burst length is programmed to 8. It determines the maximum number of column locations that can be accessed for a given write command. When the write command is issued, 8 columns are selected. All accesses for that burst take place within these eight columns, thus the burst wraps within these 8 columns if a boundary is reached. These 8 columns are selected by addr[13:3]. addr[2:0] is used to select the starting location within the block. In the case of incrementing burst (INCR/INCR4/INCR8/INCR16), the addresses can cross the 16-byte boundary of the DDR-SDRAM device. For example, when a transfer (INCR4) starts at address 0x0C, the next access is 0x10, but since the burst length is programmed to 8, the next access is at 0x00. Since the boundary is reached, the burst is wrapped. The MPDDRC takes this feature of the DDR-SDRAM device into account. In the case of a transfer starting at address 0x04/0x08/0x0C or starting at address 0x10/0x14/0x18/0x1C, two write commands are issued to avoid wrapping when the boundary is reached. The last write command is subject to DM input logic level. If DM is registered high, the corresponding data input is ignored and the write access is not done. This avoids additional writing.

302
Figure 28-2. Single Write Access, Row Closed, DDR-SDRAM Devices SDCLK
A[12:0] COMMAND BA[1:0] DQS[1:0] DM[1:0] D[15:0]

NOP 00 PRCHG NOP
Row a ACT NOP
col a WRITE NOP
Da
Db
t RP = 2
t RCD = 2
Figure 28-3. Single Write Access, Row Closed, DDR2-SDRAM Devices
SDCLK A[12:0] COMMAND BA[1:0] DQS[1:0] DM[1:0] D[15:0]

3 0 3 00 NOP PRCHG NOP Row a ACT NOP col a WRITE NOP
Da
Db
t RP = 2
t RCD = 2

303
Figure 28-4. Burst Write Access, Row Closed, DDR-SDRAM Devices

SDCLK
A[12:0]
Row a
col a
COMMAND
NOP
PRCHG
NOP
ACT
NOP
WRITE
NOP
BA[1:0] DQS[1:0] DM[1:0] D [15:0]
3 Da Db Dc Dd
0 De Df Dg Dh
t RP = 2
t RCD = 2
Figure 28-5. Burst Write Access, Row Closed, DDR2-SDRAM Devices

SDCLK
A[12:0]
Row a
col a
COMMAND
NOP
PRCHG
NOP
ACT
NOP
WRITE
NOP
BA[1:0] DQS[1:0] DM[1:0] D [15:0]
3 Da Db Dc Dd
0 De Df Dg Dh
t RP = 2
t RCD = 2
A write command can be followed by a read command. To avoid breaking the current write burst, tWTR/tWRD (bl/2 + 2 = 6 cycles) should be met. See Figure 28-6 on page 305.

304
Figure 28-6. Write Command Followed by a Read Command without Burst Write Interrupt, DDR-SDRAM Devices
3 0 Da Db Dc Dd De Df Dg Dh 3 Da Db 0 col a col a
NOP
WRITE
NOP
READ
BST
NOP
t WRD = BL/2 + 2 = 8/2 + 2 = 6 t WR =1
In the case of a single write access, write operation should be interrupted by a read access but DM must be input 1 cycle prior to the read command to avoid writing invalid data. See Figure 28-7 on page 305.
Figure 28-7. SINGLE Write Access followed by a Read Access, DDR-SDRAM Devices
3 0 3 Da Db NOP 0 PRCHG Row a NOP ACT NOP WRITE col a NOP READ BST NOP
Da
Db
Data masked

305
Figure 28-8. SINGLE Write Access followed by a Read Access, DDR2-SDRAM Devices

3 0 3 Da Db Row a NOP PRCHG NOP 0 ACT NOP WRITE col a NOP READ NOP
Da
Db
Data masked
t WTR
28.5.2 DDR-SDRAM Controller Read Cycle

The MPDDRC allows burst access or single access in normal mode (mode = 000). Whatever the access type, the MPDDRC keeps track of the active row in each bank, thus maximizing performance of the MPDDRC. The DDR-SDRAM devices are programmed with a burst length equal to 8 which determines the length of a sequential data output by the read command that is set to 8. The latency from read command to data output is equal to 2, 3, 4, 5 or 6. This value is programmed during the initialization phase (see Section 28.4 Product Dependencies, Initialization Sequence on page 298). To initiate a single access, the MPDDRC checks if the page access is already open. If row/bank addresses match with the previous row/bank addresses, the controller generates a read command. If the bank addresses are not identical or if bank addresses are identical but the row addresses are not identical, the controller generates a precharge command, activates the new row and initiates a read command. To comply with DDR-SDRAM timing parameters, additional clock cycles are inserted between precharge/active (tRP) commands and active/read (tRCD) command. After a read command, additional wait states are generated to comply with CAS latency. The MPDDRC supports a CAS latency of two to six (2 to 6 clocks delay). As the burst length is fixed to 8, in the case of single access or burst access inferior to 8 data requests, it has to stop the burst otherwise an additional seven or X values could be read. Burst Stop Command (BST) is used to stop output during a burst read. If the DDR2-SDRAM Burst Stop Command is not supported by JEDEC standard, in a single read access, an additional seven unwanted data will be read. To initiate a burst access, the MPDDRC checks the transfer type signal. If the next accesses are sequential read accesses, reading to the SDRAM device is carried out. If the next access is a read non-sequential access, then an automatic page break can be inserted. If the bank addresses are not identical or if bank addresses are identical but the row addresses are not identical, the controller generates a precharge command, activates the new row and initiates a read command. In the case where the page access is already open, a read command is generated. To comply with DDR-SDRAM timing parameters, additional clock cycles are inserted between precharge/active (tRP) commands and active/read (tRCD) commands. The MPDDRC supports a CAS latency of two to six (2 to 6 clocks delay). During this delay, the controller uses internal signals to anticipate the next access and improve the performance of the controller. Depending on the latency, the MPDDRC anticipates 2 to 6 read accesses. In the case of burst of specified length, accesses are not anticipated, but if the burst is broken (border, busy mode, etc.), the next access is treated as an incrementing burst of unspecified length, and in function of the latency, the MPDDRC anticipates 2 to 6 read accesses. For a definition of timing parameters, refer to Section 28.7.3 MPDDRC Configuration Register on page 321. Read accesses to the DDR-SDRAM are burst oriented and the burst length is programmed to 8. It determines the maximum number of column locations that can be accessed for a given read command. When the read command is issued, 8 columns are selected. All accesses for that burst take place within these eight columns, meaning that the burst wraps within these 8 columns if the boundary is reached. These 8 columns are selected by addr[13:3]; addr[2:0] is used to select the starting location within the block.

306
In the case of incrementing burst (INCR/INCR4/INCR8/INCR16), the addresses can cross the 16-byte boundary of the DDR-SDRAM device. For example, when a transfer (INCR4) starts at address 0x0C, the next access is 0x10, but since the burst length is programmed to 8, the next access is 0x00. Since the boundary is reached, the burst wraps. The MPDDRC takes into account this feature of the SDRAM device. In the case of a DDR-SDRAM device, transfers start at address 0x04/0x08/0x0C. Two read commands are issued to avoid wrapping when the boundary is reached. The last read command may generate additional reading (1 read cmd = 4 DDR words). To avoid additional reading, it is possible to use the burst stop command to truncate the read burst and to decrease power consumption. DDR2-SDRAM devices do not support the burst stop command.
Figure 28-9. Single Read Access, Row Closed, Latency = 2, DDR-SDRAM Devices
SDCLK A[12:0] COMMAND BA[1:0] DM[3:0] D[31:0] t RP t RCD

NOP 0 PRCHG NOP Row a ACT NOP col a READ BST NOP
3 DaDb
Latency = 2
Figure 28-10. Single Read Access, Row Closed, Latency = 3, DDR2-SDRAM Devices
SDCLK A[12:0] COMMAND BA[1:0] DQS[1] DQS[0] DM[1:0] D[15:0] t RP t RCD Latency = 3
3 Da Db NOP 0 PRCHG NOP Row a ACT Col a NOP READ

307
Figure 28-11. Burst Read Access, Latency = 2, DDR-SDRAM Devices

SDCLKN SDCLK A[12:0] COMMAND BA[1:0] DQS[1:0] DM[1:0] D[15:0] Latency = 2
3 Da Db Dc Dd De Df Dg Dh NOP 0 Col a READ NOP
Figure 28-12. Burst Read Access, Latency = 3, DDR2-SDRAM Devices

SDCLKN SDCLK A[12:0] COMMAND BA[1:0] DQS[1:0] DM[1:0] D[15:0] Latency = 3
3 Da Db Dc Dd De Df Dg Dh NOP 0 Col a READ NOP
28.5.2.1 Auto Refresh all Banks An All Banks Auto Refresh command performs a refresh operation on all banks. An auto refresh command is used to refresh the MPDDRC. Refresh addresses are generated internally by the DDR-SDRAM device and incremented after each auto-refresh automatically. The MPDDRC generates these auto-refresh commands periodically. A timer is loaded with the value in MPDDRC_RTR (see Section 28.7.2 MPDDRC Refresh Timer Register)that indicates the number of clock cycles between refresh cycles. When the MPDDRC initiates a refresh of a DDR-SDRAM device, internal memory accesses are not delayed. However, if the CPU tries to access the DDR-SDRAM device, the slave indicates that the device is busy. A refresh request does not interrupt a burst transfer in progress. This feature is activated by setting refresh per bank bit [REF_PB] to 0 in MPDDRC_RTR (see Section 28.7.2 MPDDRC Refresh Timer Register). 28.5.2.2 Auto Refresh per Bank Low-power DDR2-SDRAM embeds a new command, the Per Bank Auto Refresh command which performs a refresh operation on the bank which is scheduled by the bank counter in the memory device. The bank sequence of Per Bank Refresh is fixed to be a sequential round-robin: 0-1-2-3-4-5-6-7-0-1-.... The bank count is synchronized between the controller and the SDRAM upon issuing a RESET command or at every exit from self refresh, by resetting bank count to zero. The bank addressing for the per-bank refresh count is the same as established in the single-bank Precharge command. This feature is activated by setting refresh per bank bit [REF_PB] to 1 in MPDDRC_RTR (see Section 28.7.2 MPDDRC Refresh Timer Register on page 320). This feature allows to mask the latency do to the refresh procedure.

308
The target bank is inaccessible during the Per Bank Refresh cycle time (tRFCpb), however other banks within the device are accessible and may be addressed during the Per Bank Refresh cycle. During the REFpb operation, any of the banks other than the one being refreshed can be maintained in active state or accessed by a read or a write command. When the Per Bank Refresh cycle has completed, the affected bank will be in the Idle state. 28.5.2.3 Adjust Auto Refresh Rate Low-power DDR2-SDRAM embeds an internal register, Mode Register 19 (Refresh Mode). The content of this register allows to adjust the interval of auto-refresh operations according to temperature variation. This feature is activated by setting adjust refresh bit [ADJ_REF] to 1 in MPDDRC_RTR (see Section 28.7.2 MPDDRC Refresh Timer Register). When this feature is enabled, a mode register read command (MRR) is performed every 16*tREFI (average time between REFRESH commands). In function of the read value, the auto refresh interval will be modified. In the case of high temperature, the interval is reduced and in the case of low temperature, the interval is increased.

28.5.3.1 Self Refresh Mode This mode is activated by setting low-power command bit [LPCB] to 01 in the MPDDRC_LPR Register. Self refresh mode is used in power-down mode, i.e., when no access to the DDR-SDRAM device is possible. In this case, power consumption is very low. In self refresh mode, the DDR-SDRAM device retains data without external clocking and provides its own internal clocking, thus performing its own auto refresh cycles. All the inputs to the DDRSDRAM device become dont care except CKE, which remains low. As soon as the DDR-SDRAM device is selected, the MPDDRC provides a sequence of commands and exits self refresh mode. The MPDDRC re-enables self refresh mode as soon as the DDR-SDRAM device is not selected. It is possible to define when self refresh mode is to be enabled by setting the LPR register (see Section 28.7.7 MPDDRC Low-power Register on page 328), time-out command bit:
z z z
00 = Self refresh mode is enabled as soon as the DDR-SDRAM device is not selected 01 = Self refresh mode is enabled 64 clock cycles after completion of the last access 10 = Self refresh mode is enabled 128 clock cycles after completion of the last access
This controller also interfaces Low-power DDR-SDRAM. These devices add a feature: a single quarter, one-half quarter or all banks of the DDR-SDRAM array can be enabled in self refresh mode. Disabled banks are not refreshed in self refresh mode. This feature permits to reduce the self refresh current. In the case of Low-power DDR1-SDRAM, the extended mode register controls this feature, it includes Temperature Compensated Self Refresh (TSCR), Partial Array Self refresh (PASR) parameters and drives strength (DS) (see Section 28.7.7 MPDDRC Low-power Register on page 328). In the case of Low-power DDR2-SDRAM, the mode register 16 and 17 control this feature, it includes PASR Bank Mask (BK_MASK), PASR Segment Mask (SEG_MASK) parameters and drives strength (DS) (see Section 28.7.8 MPDDRC Low-power DDR2 Low-power Register on page 331). These parameters are set during the initialization phase. After initialization, as soon as PASR/DS/TCSR fields or BK_MASK/SEG_MASK/DS are modified, the Extended Mode Register or Mode Register 3/16/17 in the memory of the external device is accessed automatically and PASR/DS/TCSR and BK_MASK/SEG_MASK/DS bits are updated before entry into self refresh mode if MPDDRC does not share an external bus with another controller or during a refresh command, and a pending read or write access, if MPDDRC does share an external bus with another controller. This type of update is a function of the UPD_MR bit (see Section 28.7.7 MPDDRC Low-power Register on page 328). The Low-power DDR1-SDRAM must remain in self refresh mode for a minimum of TRFC periods and may remain in self refresh mode for an indefinite period. The DDR2-SDRAM must remain in self refresh mode for a minimum of TCKE periods and may remain in self refresh mode for an indefinite period. The Low-power DDR2-SDRAM must remain in self refresh mode for a minimum of TCKESR periods and may remain in self refresh mode for an indefinite period.

309
Figure 28-13. Self Refresh Mode Entry, Time-out = 0
SDCK A[12:0] COMMAND CKE BA[1:0] DQS[0:1] DM[1:0] D[15:0]

3 Da Db 0 NOP READ BST NOP PRCHG NOP ARFSH NOP
t RP
Figure 28-14. Self Refresh Mode Entry, Time-out = 1 or 2
Enter Self Refresh Mode
SDCLK A[12:0] COMMAND CKE BA[1:0] DQS[1:0] DM[1:0] D[15:0]

3 Da Db 0 NOP READ BST NOP PRCHG NOP ARFSH NOP
64 or 128 Wait states
t RP
Enter Self Refresh Mode
Figure 28-15. Self Refresh Mode Exit

SDCLK A[12:0] COMMAND CKE BA[1:0] DQS[1:0] DM[1:0] D[15:0] Exit Self Refresh Mode Clock must be stable before exiting self refresh mode t XNRD / t XSRD t XSR (DDR device) (Low-power DDR device)
3 0 NOP VALID NOP

310
28.5.3.2 Power-down Mode This mode is activated by setting the low-power command bit [LPCB] to 10. Power-down mode is used when no access to the DDR-SDRAM device is possible. In this mode, power consumption is greater than in self refresh mode. This state is similar to normal mode (No low-power mode/No self refresh mode), but the CKE pin is low and the input and output buffers are deactivated as soon the DDR-SDRAM device is no longer accessible. In contrast to self refresh mode, the DDR-SDRAM device cannot remain in low-power mode longer than the refresh period (64 ms). As no auto-refresh operations are performed in this mode, the MPDDRC carries out the refresh operation. For Low-power DDR1-SDRAM devices, the controller generates a NOP command during a delay of at least TXP. In addition, Low-power DDR-SDRAM and DDR2-SDRAM must remain in power-down mode for a minimum period of TCKE periods. The exit procedure is faster than in self refresh mode. See Figure 28-16 on page 311. The MPDDRC returns to powerdown mode as soon as the DDR-SDRAM device is not selected. It is possible to define when power-down mode is enabled by setting the LPR register time-out command bit.
z z z
00 = Power-down mode is enabled as soon as the DDR-SDRAM device is not selected 01 = Power-down mode is enabled 64 clock cycles after completion of the last access 10 = Power-down mode is enabled 128 clock cycles after completion of the last access
Figure 28-16. Power-down Entry/Exit, Time-out = 0
SDCK A[12:0] COMMAND CKE BA[1:0] DQS[1:0] DM[1:0] D[15:0]

3 Da Db 0 READ BST NOP READ
Enter Power-down Mode
Exit Power-down Mode
28.5.3.3 Deep Power-down Mode The deep power-down mode is a feature of Low-power DDR-SDRAM. When this mode is activated, all internal voltage generators inside the device are stopped and all data is lost. Deep power-down mode is activated by setting the low-power command bit [LPCB] to 11. When this mode is enabled, the MPDDRC leaves normal mode (mode == 000) and the controller is frozen. Before enabling this mode, the user must assume there is no any access in progress. To exit deep power-down mode the low-power command bit (LPCB) must be set to 00, an initialization sequence must be generated by software. See Section 28.4.1 Low-power DDR1-SDRAM Initialization on page 298 or Section 28.4.3 Low-power DDR2-SDRAM Initialization on page 300.

311
Figure 28-17. Deep Power-down Mode Entry
SDCLK A[12:0] COMMAND CKE BA[1:0] DQS[1:0] DM[1:0] D[15:0]

3 Da Db 0 NOP READ BST NOP PRCHG NOP DEEPOWER NOP
t RP
Enter Deep Power-down Mode
28.5.3.4 Change Frequency During Power-down Mode with Low-power DDR2-SDRAM Devices To change frequency, power-down mode must be activated by setting the low-power command bit [LPCB] to 10 and changing frequency bit command [CHG_FR] to 1. Once the Low-power DDR2-SDRAM is in precharge power-down mode, the clock frequency may change. The device input clock frequency is allowed to change only within minimum and maximum operating frequencies as specified by Low-power DDR2-SDRAM providers. Once the input clock frequency is changed, new stable clocks must be provided to the device before exiting from the precharge power-down mode. Depending on the new clock frequency, the user can change the CAS latency in the user interface. (See CAS: CAS Latency on page 321.) It is recommended to check that no access is in progress. Once the controller detects a change of latency during the change frequency procedure, a Load Mode Register command is performed. During a change frequency procedure, the change frequency bit command [CHG_FR] sets to 0 automatically. 28.5.3.5 Reset Mode The reset mode is a feature of DDR2-SDRAM. This mode is activated by setting the low-power command bit [LPCB] to 11 and the clock frozen command bit [CLK_FR] to 1. When this mode is enabled, the MPDDRC leaves normal mode (mode == 000) and the controller is frozen. Before enabling this mode, the user must assume there is no any access in progress. To exit reset mode, the low-power command bit [LPCB] must be set to 00, clock frozen command bit [CLK_FR] set to 0 and an initialization sequence must be generated by software. (See Section 28.4.2 DDR2-SDRAM Initialization).
28.5.4 Multi-port Functionality

The DDR-SDRAM protocol imposes a check of timings prior to performing a read or a write access, thus decreasing system performance. An access to DDR-SDRAM is performed if banks and rows are open (or active). To activate a row in a particular bank, it has to de-activate the last open row and open the new row. Two DDR-SDRAM commands must be performed to open a bank: Precharge and Activate Command with respect to tRP timing. Before performing a read or write command, tRCD timing must be checked. This operation generates a significant bandwidth loss (see Figure 28-18.).

312
Figure 28-18. tRP and tRCD Timings
SDCK A[12:0] COMMAND BA[1:0] DQS[1:0] DM1:0] D[15:0] t RP t RCD Latency = 2

3 NOP 0 PRCHG NOP ACT NOP READ BST NOP
Da
Db
4 cycles before performing a read command
The multi-port controller is designed to mask these timings and thus improve the bandwidth of the system. MPDDRC is a multi-port controller whereby four masters can simultaneously reach the controller. This feature improves the bandwidth of the system because it can detect four requests on the AHB slave inputs and thus anticipate the commands that follow, Precharge and Activate Command in bank X during the current access in bank Y. This allows tRP and tRCD timings to be masked (see Figure 28-19). In the best case, all accesses are done as if the banks and rows were already open. The best condition is met when the four masters work in different banks. In the case of four simultaneous read accesses, when the four or eight banks and associated rows are open, the controller reads with a continuous flow and masks the CAS latency for each access. To allow a continuous flow, the read command must be set at 2 or 6 cycles (CAS latency) before the end of the current access. This requires that the scheme of arbitration changes since the roundrobin arbitration cannot be respected. If the controller anticipates a read access, and thus a master with a high priority arises before the end of the current access, then this master will not be serviced.
Figure 28-19. Anticipate Precharge/Activate Command in Bank 2 during Read Access in Bank 1
SDCK A[12:0] COMMAND BA[1:0] DQS[1:0] DM1:0] D[15:0]
3 Da Db Dc Dd De Df Dg Dh Di Dj Dk Dl NOP 0 READ 1 PRECH 2 NOP ACT READ 1 NOP
t RP Anticipate command, Precharge/Active Bank 2 Read Access in Bank 1

313
The arbitration mechanism reduces latency when conflicts occur, that is when two or more masters try to access the DDR-SDRAM device at the same time. The arbitration type is round-robin arbitration. This algorithm dispatches requests from different masters to the DDRSDRAM device in a round-robin manner. If two or more master requests arise at the same time, the master with the lowest number is serviced first, then the others are serviced in a round-robin manner. To avoid burst breaking and to provide the maximum throughput for the DDR-SDRAM device, arbitration must only take place during the following cycles: 1. 2. 3. Idle cycles: When no master is connected to the DDR-SDRAM device. Single cycles: When a slave is currently doing a single access. End of Burst cycles: When the current cycle is the last cycle of a burst transfer.
z z
For bursts of defined length, predicted end of burst matches the size of the transfer. For bursts of undefined length, predicted end of burst is generated at the end of each four-beat boundary inside the INCR transfer.
4.
Anticipated Access: When an anticipated read access is done while the current access is not complete, the arbitration scheme can be changed if the anticipated access is not the next access serviced by the arbitration scheme.

314
28.6
Software Interface/SDRAM Organization, Address Mapping

The DDR-SDRAM address space is organized into banks, rows and columns. The MPDDRC maps different memory types depending on the values set in the MPDDRC Configuration Register (see Section 28.7.3 MPDDRC Configuration Register on page 321). The following tables illustrate the relation between CPU addresses and columns, rows and banks addresses for 32-bit memory data bus widths. The MPDDRC supports address mapping in linear mode. Sequential mode is a method for address mapping where banks alternate at each last DDR-SDRAM page of the current bank. Interleaved mode is a method for address mapping where banks alternate at each SDRAM end page of the current bank. The MPDDRC makes the DDR-SDRAM device access protocol transparent to the user. The tables that follow illustrate the DDR-SDRAM device memory mapping seen by the user in correlation with the device structure. Various configurations are illustrated.
28.6.1 DDR-SDRAM Address Mapping for 32-bit Memory Data Bus Width
Table 28-1. Sequential Mapping DDR-SDRAM Configuration Mapping: 2K Rows /512/1024/2048 Columns, 4 banks
CPU Address Line 2 8 2 7 2 6 2 5 2 4 2 3 2 2 2 1 2 0 1 9 1 8 1 7 1 6 Row[10:0] Row[10:0] Row[10:0] 1 5 1 4 1 3 1 2 1 1 1 0 9 8 7 6 5 4 3 2 1 0
Bk[1:0] Bk[1:0] Bk[1:0]
Column[8:0] Column[9:0] Column[10:0]
M[1:0] M[1:0] M[1:0]
Table 28-2. Interleaved Mapping DDR-SDRAM Configuration Mapping: 2K Rows /512/1024/2048 Columns, 4 banks
CPU Address Line 2 8 2 7 2 6 2 5 2 4 2 3 2 2 2 1 2 0 1 9 1 8 Row[10:0] Row[10:0] Row[10:0] 1 7 1 6 1 5 1 4 1 3 1 2 1 1 1 0 9 8 7 6 5 4 3 2 1 0
Bk[1:0] Bk[1:0] Bk[1:0]
M[1:0] M[1:0] M[1:0]
Table 28-3. Sequential Mapping DDR-SDRAM Configuration Mapping:4K Rows /512/1024/2048 Columns, 4 banks
CPU Address Line 2 8 2 7 2 6 2 5 2 4 2 3 2 2 2 1 2 0 1 9 1 8 1 7 1 6 1 5 1 4 1 3 1 2 1 1 1 0 9 8 7 6 5 4 3 2 1 0
Bk[1:0] Bk[1:0] Bk[1:0]
Row[11:0] Row[11:0] Row[11:0]
M[1:0] M[1:0] M[1:0]
CPU Address Line 2 8 2 7 2 6 2 5 2 4 2 3 2 2 2 1 2 0 1 9 1 8 1 7 1 6 1 5 1 4 1 3 1 2 1 1 1 0 9 8 7 6 5 4 3 2 1 0
Row[11:0] Row[11:0] Row[11:0]
Bk[1:0] Bk[1:0] Bk[1:0]
M[1:0] M[1:0] M[1:0]

315
Table 28-5. Sequential Mapping DDR-SDRAM Configuration Mapping:8K Rows /512/1024/2048 Columns, 4 banks
CPU Address Line 2 8 2 7 2 6 2 5 2 4 2 3 2 2 2 1 2 0 1 9 1 8 1 7 Row[12:0] Row[12:0] Row[12:0] 1 6 1 5 1 4 1 3 1 2 1 1 1 0 9 8 7 6 5 4 3 2 1 0
Bk[1:0] Bk[1:0] Bk[1:0]
M[1:0] M[1:0] M[1:0]
CPU Address Line 2 8 2 7 2 6 2 5 2 4 2 3 2 2 2 1 2 0 1 9 Row[13:0] Row[13:0] Row[13:0] 1 8 1 7 1 6 1 5 1 4 1 3 1 2 1 1 1 0 9 8 7 6 5 4 3 2 1 0
Bk[1:0] Bk[1:0] Bk[1:0]
M[1:0] M[1:0] M[1:0]
Table 28-7. Sequential Mapping DDR-SDRAM Configuration Mapping:8K Rows /1024 Columns, 8banks
CPU Address Line 2 8 2 7 2 6 Bk[2:0] 2 5 2 4 2 3 2 2 2 1 2 0 1 9 1 8 Row[12:0] 1 7 1 6 1 5 1 4 1 3 1 2 1 1 1 0 9 8 7 6 5 4 3 2 1 0
Column[9:0]
M[1:0]
Table 28-8. Interleaved Mapping DDR-SDRAM Configuration Mapping: 8K Rows /1024 Columns, 8banks
CPU Address Line 2 8 2 7 2 6 2 5 2 4 2 3 2 2 2 1 Row[12:0] 2 0 1 9 1 8 1 7 1 6 1 5 1 4 1 3 Bk[2:0] 1 2 1 1 1 0 9 8 7 6 5 4 3 2 1 0
Column[9:0]
M[1:0]
Table 28-9. Sequential Mapping DDR-SDRAM Configuration Mapping:16K Rows /1024 Columns, 4 banks
CPU Address Line 2 8 2 7 2 6 2 5 2 4 2 3 2 2 2 1 2 0 1 9 1 8 1 7 1 6 1 5 1 4 1 3 1 2 1 1 1 0 9 8 7 6 5 4 3 2 1 0
Bk[1:0]
Row[13:0]
Column[9:0]
M[1:0]
Table 28-10. Interleaved Mapping DDR-SDRAM Configuration Mapping: 16K Rows /1024 Columns, 4 banks
CPU Address Line 2 8 2 7 2 6 2 5 2 4 2 3 2 2 2 1 2 0 1 9 1 8 1 7 1 6 1 5 1 4 1 3 1 2 1 1 1 0 9 8 7 6 5 4 3 2 1 0
Row[13:0]
Bk[1:0]
Column[9:0]
M[1:0]

316
Table 28-11. Sequential Mapping DDR-SDRAM Configuration Mapping:16K Rows /1024 Columns, 8 banks
CPU Address Line 2 8 2 7 Bk[2:0] 2 6 2 5 2 4 2 3 2 2 2 1 2 0 1 9 1 8 1 7 1 6 1 5 1 4 1 3 1 2 1 1 1 0 9 8 7 6 5 4 3 2 1 0
Row[13:0]
Column[9:0]
M[1:0]
Table 28-12. Interleaved Mapping DDR-SDRAM Configuration Mapping: 16K Rows /1024 Columns, 8banks
CPU Address Line 2 8 2 7 2 6 2 5 2 4 2 3 2 2 2 1 2 0 1 9 1 8 1 7 1 6 1 5 1 4 1 3 Bk[2:0] 1 2 1 1 1 0 9 8 7 6 5 4 3 2 1 0
Row[13:0]
Column[9:0]
M[1:0]
Notes: 1. M[1:0] is the byte address inside a 32-bit word. 2. BK[2] = BA2, Bk[1] = BA1, Bk[0] = BA0

317
28.7
AHB Multi-port DDR-SDRAM Controller (MPDDRC) User Interface
The User Interface is connected to the APB bus. The MPDDRC is programmed using the registers listed in Table 28-13
Table 28-13. Register Mapping Offset 0x00 0x04 0x08 0x0C 0x10 0x14 0x18 0x1C 0x20 0x24 0x28 0x2C 0x30 0x34 0x38 0x3C 0x40 0x44x70 0x74 0x78 0x7C 0x80 0x84 0x88 0x8C 0x94xE0 0xE4 0xE8 0x1580x1CC 0x1DC0x1F8 Register MPDDRC Mode Register MPDDRC Refresh Timer Register MPDDRC Configuration Register MPDDRC Timing Parameter 0 Register MPDDRC Timing Parameter 1 Register MPDDRC Timing Parameter 2 Register Reserved MPDDRC Low-power Register MPDDRC Memory Device Register MPDDRC High Speed Register MPDDRC LPDDR2 Low-power Register MPDDRC LPDDR2 Calibration and MR4 Register MPDDRC LPDDR2 Timing Calibration Register MPDDRC IO Calibration MPDDRC OCMS Register MPDDRC OCMS KEY1 Register MPDDRC OCMS KEY2 Register Reserved MPDDRC DLL Master Offset Register MPDDRC DLL Slave Offset Register MPDDRC DLL Status Master Register MPDDRC DLL Status Slave 0 Register MPDDRC DLL Status Slave 1 Register MPDDRC DLL Status Slave 2 Register MPDDRC DLL Status Slave 3 Register Reserved MPDDRC Write Protect Control Register MPDDRC Write Protect Status Register Reserved. Reserved. Name MPDDRC_MR MPDDRC_RTR MPDDRC_CR MPDDRC_TPR0 MPDDRC_TPR1 MPDDRC_TPR2 MPDDRC_LPR MPDDRC_MD MPDDRC_HS MPDDRC_LPDDR2_LPR MPDDRC_LPDDR2_CAL_MR4 MPDDRC_LPDDR2_TIM_CAL MPDDRC_IO_CALIBR MPDDRC_OCMS MPDDRC_OCMS_KEY1 MPDDRC_OCMS_KEY2 MPDDRC_DLL_MO MPDDRC_DLL_SOF MPDDRC_DLL_MS MPDDRC_DLL_SS0 MPDDRC_DLL_SS1 MPDDRC_DLL_SS2 MPDDRC_DLL_SS3 MPDDRC_WPCR MPDDRC_WPSR Access Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Write-only Write-only Read-write Read-write Read-only Read-only Read-only Read-only Read-only Read-write Read-only Reset 0x00000000 0x00000000 0x024 0x20227225 0x3c80808 0x00042062 0x0 0x10 0x00000000 0x00000000 0x00000000 0x040 0x00870002 0x00000000 0x00000000 0x00000000 0x-(1) 0x-(1) 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000
Notes: 1. Values in the DLL Master Offset Register and in the DLL Slave Offset Register vary with the product implementation.

318
28.7.1 MPDDRC Mode Register

31 23 15
MPDDRC_MR 0xFFFFEA00 Read-write See Table 28-13

30 22 14 29 21 13 28 20 12 MRS 7 6 5 4 3 2 1 MODE 0 27 19 11 26 18 10 25 17 9 24 16 8
MODE: MPDDRC Command Mode

This field defines the command issued by the MPDDRC when the SDRAM device is accessed. This register is used to initialize the SDRAM device and to activate deep power-down mode.
Value 000 001 Name NORMAL_CMD NOP_CMD PRCGALL_CM D Description Normal Mode. Any access to the MPDDRC will be decoded normally. To activate this mode, the command must be followed by a write to the DDR-SDRAM. The MPDDRC issues a NOP command when the DDR-SDRAM device is accessed regardless of the cycle. To activate this mode, the command must be followed by a write to the DDR-SDRAM. The MPDDRC issues an All Banks Precharge command when the DDR-SDRAM device is accessed regardless of the cycle. To activate this mode, the command must be followed by a write to the SDRAM. The MPDDRC issues a Load Mode Register command when the DDR-SDRAM device is accessed regardless of the cycle. To activate this mode, the command must be followed by a write to the DDRSDRAM. The MPDDRC issues an Auto-Refresh Command when the DDR-SDRAM device is accessed regardless of the cycle. Previously, an All Banks Precharge command must be issued. To activate this mode, the command must be followed by a write to the DDR-SDRAM. The MPDDRC issues an Extended Load Mode Register command when the SDRAM device is accessed regardless of the cycle. To activate this mode, the command must be followed by a write to the DDR-SDRAM. The write in the DDR-SDRAM must be done in the appropriate bank. Deep power mode: Access to deep power-down mode The MPDDRC issues an LPDDR2 Mode Register command when the Low-power DDR2-SDRAM device is accessed regardless of the cycle. To activate this mode, the Mode Register command must be followed by a write to the Low-power DDR2-SDRAM.
010
011
LMR_CMD
100
RFSH_CMD
101 110 111
EXT_LMR_CM D DEEP_CMD LPDDR2_CMD
MRS: Mode Register Select LPDDR2

Configure this 8-bit field to program all Mode Registers included in a Low-power DDR2-SDRAM Device. This field is unique to Low-power DDR2-SDRAM devices.

319
28.7.2 MPDDRC Refresh Timer Register

31 23
MPDDRC_RTR 0xFFFFEA04 Read-write See Table 28-13

30 22 29 21 MR4_VALUE 15 7 14 6 13 5 12 4 COUNT 3 2 28 20 27 19 11 26 18 10 COUNT 1 0 25 17 REF_PB 9 24 16 ADJ_REF 8
COUNT: MPDDRC Refresh Timer Count

This 12-bit field is loaded into a timer which generates the refresh pulse. Each time the refresh pulse is generated, a refresh sequence is initiated. SDRAM devices require a refresh of all rows every 64 ms. The value to be loaded depends on the MPDDRC clock frequency (MCK: Master Clock) and the number of rows in the device. For example, for an SDRAM with 8192 rows and a 100 MHz Master clock, the value of Refresh Timer Count bit is programmed: ((64 x 10-3)/8192) x100 x106 = 781 or 0x030D. Low-power DDR2-SDRAM devices support Per Bank Refresh operation. In this configuration, average time between refresh command is 0.975 s. The value of the Count bit is programmed in function of this value. For example, the value of a 100 MHz Master clock refresh timer is 98 or 0x0062.
ADJ_REF: Adjust Refresh Rate The reset value is 0. 0: Adjust refresh rate is not enabled. 1: Adjust refresh rate is enabled. This mode is unique to Low-power DDR2-SDRAM devices. REF_PB: Refresh Per Bank The reset value is 0. 0: Refresh all banks during auto-refresh operation. 1: Refresh the scheduled bank by the bank counter in the memory interface. This mode is unique to Low-power DDR2-SDRAM devices. MR4_VALUE: Content of MR4 Register
The reset value is 3. This field (read-only) gives the content of MR4 register. This field is updated when MRR command is generated and Adjust Refresh Rate bit is enabled. Update is done when read value is different from MR4_VALUE. LPDDR2 JEDEC memory standards impose derating LPDDR2 AC timings (tRCD, tRC, tRAS, tRP and tRRD) when the value of MR4 is equal to 6. If the application needs to work in extreme conditions, the derating value must be added to AC timings before the power up sequence. This mode is unique to Low-power DDR2-SDRAM devices.

320
28.7.3 MPDDRC Configuration Register

31 23 UNAL 15 7 DLL 6
MPDDRC_CR 0xFFFFEA08 Read-write See Table 28-13

30 22 DECOD 14 29 21 NDQS 13 OCD 5 CAS 4 3 NR 28 20 NB 12 11 ZQ 2 27 19 26 18 10 25 17 ENRDM 9 DIS_DLL 1 NC 24 16 DQMS 8 DIC_DS 0
NC: Number of Column Bits.

The reset value is 9 column bits.
Value 00 01 10 11 Name COL_9 COL_10 COL_11 COL_12 Description 9 DDR column bits 10 DDR column bits 11 DDR column bits 12 DDR column bits
NR: Number of Row Bits

The reset value is 12 row bits.
Value 00 01 10 11 Name ROW_11 ROW_12 ROW_13 ROW_14 Description 11 row bits 12 row bits 13 row bits 14 row bits
CAS: CAS Latency

The reset value is 2 cycles.
Value 000 001 010 011 100 101 110 111 Name DDR_CAS2 DDR_CAS3 DDR_CAS4 DDR_CAS5 DDR_CAS6 Description DDR CAS Latency Reserved DDR CAS Latency Reserved LPDDR1 CAS Latency 2 DDR2/LPDDR2/LPDDR1 CAS Latency 3 DDR2/LPDDR2 CAS Latency 4 DDR2/LPDDR2 CAS Latency 5 DDR2 CAS Latency 6 DDR CAS Latency Reserved

321
DLL: Reset DLL

Reset value is 0. This field defines the value of Reset DLL. 0 (RESET_DISABLED): Disable DLL reset. 1 (RESET_ENABLED): Enable DLL reset. This value is used during the power-up sequence. This field is found only in DDR2-SDRAM devices.
DIC_DS: Output Driver Impedance Control (Drive Strength)

Reset value is 0. This field defines the output drive strength. 0 = Full strength. 1 = Reduced strength. This value is used during the power-up sequence. This field is found only in DDR2-SDRAM devices.
DIS_DLL: DISABLE DLL

Reset value is 0. 0 = Enable DLL. 1 = Disable DLL. This value is used during the power-up sequence. It is only found in DDR2-SDRAM devices.
ZQ: ZQ Calibration
Reset value is 0.
Value 00 01 10 11 Name INIT LONG SHORT RESET Description Calibration command after initialization Long calibration Short calibration ZQ Reset
This parameter is used to calibrate DRAM On resistance (Ron) values over PVT. This field is found only in Low-power DDR2-SDRAM devices.
OCD: Off-chip Driver

Reset value is 7. This field is only found in DDR2-SDRAM devices. Note:
Value 000 111
OCD is NOT supported by the controller, but these values MUST be programmed during the initialization sequence.
Name EXIT DEFAULT Description OCD calibration mode exit, maintain setting OCD calibration default
DQMS: Mask Data is Shared

Reset value is 0. 0 (NOT_SHARED): DQM is not shared with another controller. 1 (SHARED): DQM is shared with another controller.

322
ENRDM: Enable Read Measure

Reset value is 0. 0 (OFF): DQS/DDR_DATA phase error correction is disabled. 1 (ON): DQS/DDR_DATA phase error correction is enabled.
NB: Number of Banks.

The reset value is four banks.
Value 0 1 Name 4 8 Description 4 banks 8 banks
This field is found only in DDR2-SDRAM and Low-power DDR2-SDRAM devices.
NDQS: Not DQS:

The reset value is 1, Not DQS is disabled. 0 (ENABLED): Not DQS is enabled. 1 (DISABLED): Not DQS is disabled. This field is found only in DDR2-SDRAM devices.
DECOD: Type of Decoding

The reset value is 0, Sequential decoding. 0: Sequential decoding. 1: Interleaved decoding.
UNAL: Support Unaligned Access

The reset value is 0, Unaligned access is not supported. 0 (UNSUPPORTED): Unaligned access is not supported. 1 (SUPPORTED): Unaligned access is supported. This mode is enabled with masters which have an AXI interface.

323
28.7.4 MPDDRC Timing Parameter 0 Register

31
MPDDRC_TPR0 0xFFFFEA0C Read-write See Table 28-13

30 TMRD 23 22 TRRD 15 14 TRC 7 6 TRCD 5 4 3 2 TRAS 13 12 11 10 TWR 1 0 21 20 29 28 27 RDC_WRRD 19 18 TRP 9 8 26 25 TWTR 17 16 24
TRAS: Active to Precharge Delay

Reset Value is 5 cycles. This field defines the delay between an Activate command and a Precharge command in number of cycles. The number of cycles is between 0 and 15.
TRCD: Row to Column Delay

Reset Value is 2 cycles. This field defines the delay between an Activate command and a Read/Write command in number of cycles. The number of cycles is between 0 and 15.
TWR: Write Recovery Delay

Reset value is 2 cycles. This field defines the Write Recovery Time in number of cycles. The number of cycles is between 1 and 15.
TRC: Row Cycle Delay

Reset value is 7 cycles. This field defines the delay between an Activate command and Refresh command in number of cycles. The number of cycles is between 0 and 15
TRP: Row Precharge Delay

Reset Value is 2 cycles. This field defines the delay between a Precharge command and another command in number of cycles. The number of cycles is between 0 and 15.
TRRD: Active BankA to Active BankB

Reset value is 2 cycles. This field defines the delay between an Activate command in BankA and an Activate command in BankB in number of cycles. The number of cycles is between 1 and 15.
TWTR: Internal Write to Read Delay

Reset value is 0. This field defines the internal Write to Read command time in number of cycles. The number of cycles is between 1 and 7.

324
RDC_WRRD: Reduce Write to Read Delay

Reset value is 0. This field reduces the delay between write to read access for Low-power DDR-SDRAM devices with a latency equal to 2. To use this feature, TWTR field must be equal to 0. Important to note is that some devices do not support this feature.
TMRD: Load Mode Register Command to Activate or Refresh Command

Reset Value is 2 cycles. This field defines the delay between a Load mode register command and an Activate or Refresh command in number of cycles. The number of cycles is between 0 and 15. For Low-power DDR2-SDRAM, this field is equivalent to TMRW timing.

325

31 23
MPDDRC_TPR1 0xFFFFEA10 Read-write See Table 28-13

30 22 29 21 28 20 TXSRD 15 14 13 12 TXSNR 7 6 5 4 3 TRFC 2 1 0 11 10 9 8 19 18 27 26 TXP 17 16 25 24
TRFC: Row Cycle Delay Reset Value is 8 cycles. This field defines the delay between a Refresh command or a Refresh and Activate command in number of cycles. The number of cycles is between 0 and 127. In the case of Low-power DDR2-SDRAM, this field is equivalent to TRFCab timing. If the user enables the function Refresh Per Bank (See REF_PB: Refresh Per Bank on page 320.) this field is equivalent to TRFCpb. TXSNR: Exit Self Refresh Delay to Non Read Command Reset Value is 8 cycles. This field defines the delay between CKE set high and a Non Read command in number of cycles. The number of cycles is between 0 and 255. This field is used by DDR-SDRAM devices. In the case of Low-power DDR-SDRAM, this field is equivalent to tXSR timing. TXSRD: Exit Self Refresh Delay to Read Command Reset Value is 200 cycles. This field defines the delay between CKE set high and a Read command in number of cycles. The number of cycles is between 0 and 255 cycles. This field is found only in DDR2-SDRAM devices. TXP: Exit Power-down Delay to First Command Reset Value is 3 cycles.
This field defines the delay between CKE set high and a Valid command in number of cycles. The number of cycles is between 0 and 15 cycles.

326

31 23
MPDDRC_TPR2 0xFFFFEA14 Read-write See Table 28-13

30 22 21 20 19 18 TFAW 15 14 13 TRTP 7 6 TXARDS 5 4 3 2 TXARD 12 11 10 TRPA 1 0 9 8 17 16 29 28 27 26 25 24
TXARD: Exit Active Power Down Delay to Read Command in Mode Fast Exit. The Reset Value is 2 cycles. This field defines the delay between CKE set high and a Read Command in number of cycles. The number of cycles is between 0 and 15.
This field is found only in DDR2-SDRAM devices.
TXARDS: Exit Active Power Down Delay to Read Command in Mode Slow Exit. The Reset Value is 6 cycles. This field defines the delay between CKE set high and a Read Command in number of cycles. The number of cycles is between 0 and 15.
This field is found only in DDR2-SDRAM devices.
TRPA: Row Precharge All Delay

The Reset Value is 0 cycles. This field defines the delay between a Precharge All Banks command and another command in number of cycles. The number of cycles is between 0 and 15. This field is found only in DDR2-SDRAM devices.
TRTP: Read to Precharge

The Reset Value is 2 cycles. This field defines the delay between Read command and a Precharge command in number of cycles. The number of cycles is between 0 and 7.
TFAW: Four Active Windows

The Reset Value is 4 cycles. DDR2 devices with 8-banks (1Gbit or larger) have an additional requirement concerning tFAW timing. This requires that no more than four Activate Commands may be issued in any given tFAW (MIN) period. The number of cycles is between 0 and 15. This field is found only in DDR2-SDRAM and LPDDR2-SDRAM devices.

327
28.7.7 MPDDRC Low-power Register

31 23 15 7
MPDDRC_LPR 0xFFFFEA1C Read-write See Table 28-13

30 22 14 6 5 PASR 13 TIMEOUT 4 29 21 UPD_MR 12 28 20 27 19 11 3 LPDDR2_PWOFF 2 CLK_FR 26 18 10 25 17 9 DS 1 LPCB 0 24 16 APDE 8
LPCB: Low-power Command Bit

Reset value is 00.
Value 00 Name DISABLED Description Low-power Feature is inhibited. No power-down, self refresh and deep-power modes are issued to the DDR-SDRAM device. The MPDDRC issues a Self Refresh command to the DDR-SDRAM device, the clock(s) is/are deactivated and the CKE signal is set low. The DDR-SDRAM device leaves the self refresh mode when accessed and reenters it after the access. The MPDDRC issues a Power-down Command to the DDR-SDRAM device after each access, the CKE signal is set low. The DDR-SDRAM device leaves the power-down mode when accessed and reenters it after the access. The MPDDRC issues a Deep Power-down command to the Low-power DDR-SDRAM device.
01
SELFREFRESH
10 11
POWERDOWN DEEP_PWD
CLK_FR: Clock Frozen Command Bit

Reset value is 0. This field sets the clock low during power-down mode. Some DDR-SDRAM devices do not support freezing the clock during power-down mode. Refer to the device datasheet for details on this. 1 (ENABLED): Clock(s) is/are frozen. 0 (DISABLED): Clock(s) is/are not frozen.
LPDDR2_PWOFF: LPDDR2 Power Off Bit

Reset value is 0. LPDDR2 power off sequence must be controlled to preserve the LPDDR2 device. The power fail is handled at system level(IRQ or FIQ) and the LPDDR2 power off sequence is applied using the LPDDR2_PWOFF bit. LPDDR2_PWOFF bit is used to impose CKE low before a power off sequence. Uncontrolled power off sequence can be applied only up to 400 times in the life of a LPDDR2 device. 1 (ENABLED): A power off sequence is applied to the LPDDR2 device. CKE is forced low. 0 (DISABLED): No power off sequence applied to LPDDR2.

328
PASR: Partial Array Self Refresh

Reset value is 0. This field is unique to Low-power DDR1-SDRAM. It is used to specify whether only one-quarter, one-half or all banks of the DDR-SDRAM array are enabled. Disabled banks are not refreshed in self refresh mode. The values of this field are dependant on Low-power DDR-SDRAM devices.
After the initialization sequence, as soon as PASR field is modified, Extended Mode Register in the external device memory is accessed automatically and PASR bits are updated. In function of the UPD_MR bit, update is done before entering in self refresh mode or during a refresh command and a pending read or write access. DS: Drive Strength
Reset value is 0. This field is unique to Low-power DDR-SDRAM. It selects the driver strength of DDR- SDRAM output. After the initialization sequence, as soon as DS field is modified, Extended Mode Register is accessed automatically and DS bits are updated. In function of UPD_MR bit, update is done before entering in self refresh mode or during a refresh command and a pending read or write access.
TIMEOUT: Enter Low-power Mode

Reset value is 00. This field defines when low-power mode is enabled.
Value 00 01 10 11 Name 0 64 128 Description The SDRAM controller activates the SDRAM low-power mode immediately after the end of the last transfer. The SDRAM controller activates the SDRAM low-power mode 64 clock cycles after the end of the last transfer. The SDRAM controller activates the SDRAM low-power mode 128 clock cycles after the end of the last transfer. Reserved
APDE: Active Power Down Exit Time

Reset value is 1. This mode is unique to DDR2-SDRAM devices. This mode allows to determine the active power-down mode, which determines performance versus power saving. 0 (FAST): Fast Exit. 1 (SLOW): Low Exit. After the initialization sequence and as soon as the APDE field is modified, the Extended Mode Register, located in the memory of the external device, is accessed automatically and APDE bits are updated. In function of the UPD_MR bit, update is done before entering in self refresh mode or during a refresh command and a pending read or write access.

329
UPD_MR: Update Load Mode Register and Extended Mode Register

Reset value is 0. This bit is used to enable or disable automatic update of the Load Mode Register and Extended Mode Register. This update is function of MPDDRC integration in a system. MPDDRC can either share or not share an external bus with another controller.
Value 00 01 10 11 RESERVED Name DISABLED Description Update is disabled. MPDDRC shares external bus. Automatic update is done during a refresh command and a pending read or write access in SDRAM device. MPDDRC does not share external bus. Automatic update is done before entering in self refresh mode. Reserved

330
28.7.8 MPDDRC Low-power DDR2 Low-power Register

31
MPDDRC_LPDDR2_LPR 0xFFFFEA28 Read-write See Table 28-13

30 29 28 27 26 DS 23 22 21 20 SEG_MASK 15 14 13 12 SEG_MASK 7 6 5 4 3 BK_MASK_PASR 2 1 0 11 10 9 8 19 18 17 16 25 24
BK_MASK_PASR: Bank Mask Bit/PASR

Partial Array Self-Refresh (S4 device only) Reset value is 00. After the initialization sequence, as soon as field BK_MASK_PASR is modified, Mode Register 16 is accessed automatically and BK_MASK_PASR bits are updated. In function of UPD_MR bit, update is done before entering in self refresh mode or during a refresh command and a pending read or write access 0: Refresh is enabled (= unmasked). 1: Refresh is disabled (= masked). This mode is unique to Low-power DDR2-SDRAM-S4 devices. In self refresh mode, each bank of LPDDR2 can be independently configured whether a self-refresh operation is taking place. After the initialization sequence, as soon as field BK_MASK_PASR is modified, Extended Mode Register is accessed automatically and BK_MASK_PASR bits are updated. In function of UPD_MR bit, update is done before entering in self refresh mode or during a refresh command and a pending read or write access.
SEG_MASK: Segment Mask Bit

Reset value is 00. After the initialization sequence, as soon as field SEG_MASK is modified, Mode Register 17 is accessed automatically and SEG_MASK bits are updated. In function of UPD_MR bit, update is done before entering in self refresh mode or during a refresh command and a pending read or write access. 0: Segment is refreshed (= unmasked). 1: Segment is not refreshed (= masked). This mode is unique to Low-power DDR2-SDRAM-S4 devices. The number of segment mask bits differ with the density. For 1Gbit density, 8 segments are used. In self-refresh mode, when the mask bit is programmed, the refresh operation is masked in the segment.
DS: Drive strength

Reset value is 0 After the initialization sequence, as soon as field DS is modified, Mode Register 3 is accessed automatically and DS bits are updated. In function of UPD_MR bit, update is done before entering in self refresh mode or during a refresh command and a pending read or write access. This field is unique to Low-power DDR2-SDRAM. It selects the driver strength of DDR2-SDRAM I/O.

331
SR: Slew Rate

Reset value is 0. After the initialization sequence, as soon as field SR is modified, Mode Register 3 is accessed automatically and SR bits are updated. In function of UPD_MR bit, update is done before entering in self refresh mode or during a refresh command and a pending read or write access This field is unique to Low-power DDR2-SDRAM. It selects the slew rate of Low-power DDR2-SDRAM I/O.

332
28.7.9 MPDDRC Low-power DDR2 Calibration and MR4 Register

31
MPDDRC_LPDDR2_CAL_MR4 0xFFFFEA2C Read-write See Table 28-13

30 29 28 MR4_READ 23 22 21 20 MR4_READ 15 14 13 12 COUNT_CAL 7 6 5 4 COUNT_CAL 3 2 1 0 11 10 9 8 19 18 17 16 27 26 25 24
COUNT_CAL: LPDDR2 Calibration Timer Count

This 16-bit field is loaded into a timer which generates the calibration pulse. Each time the calibration pulse is generated, a ZQCS calibration sequence is initiated. The ZQCS Calibration command is used to calibrate DRAM Ron values over PVT. One method for calculating the interval between ZQCS commands gives the temperature (Tdriftrate) and voltage (Vdriftrate) drift rates that the SDRAM is subject to in the application. The interval could be defined by the following formula: ZQCorrection/ ((TSens x Tdriftrate) + (VSens x Vdriftrate)) Where TSens=max(dRONdTM) and VSens=max(dRONdVM) define the SDRAM temperature and voltage sensitivities. For example, if TSens=0.75%/C, VSens=0.2%/mV, Tdriftrate=1C/sec and Vdriftrate=15mV/sec, then the interval between ZQCS commands is calculated as: 1.5/((0.75x1) + (0.2x15)) = 0.4sec In this example LPDDR2-SDRAM devices require a calibration every 0.4 s. The value to be loaded depends on average time between REFRESH commands, tREF. For example, for an LPDDR2-SDRAM with a time between refresh of 7.8 s, the value of the Calibration Timer Count bit is programmed: (0.4/7.8 x 10-6) = 0xC852.
MR4_READ: Mode Register 4 Read Interval

MR4_READ defines the time period between MR4 reads (for LPDDR2-SDRAM). The formula is (MR4_READ+1)*tREF. The value to be loaded depends on the average time between REFRESH commands, tREF. For example, for an LPDDR2-SDRAM with a time between refresh of 7.8 s, if the MR4_READ value is 2, the time period between MR4 reads will be 23.4 s. LPDDR2-SDRAM devices feature a temperature sensor whose status can be read from MR4 register. This sensor can be used to determine an appropriate refresh rate. Temperature sensor data may be read from MR4 register using the Mode Register Read protocol. Adjust Refresh Rate field in Refresh Timer Register must be set to 1 to activate these reads.

333
28.7.10 MPDDRC Low-power DDR2 Timing Calibration Register

31 23 15 7
MPDDRC_LPDDR2_TIM_CAL 0xFFFFEA30 Read-write See Table 28-13

30 22 14 6 29 21 13 5 28 20 12 4 ZQCS 27 19 11 3 26 18 10 2 25 17 9 1 24 16 8 0
ZQCS: ZQ Calibration Short

The Reset Value is 6 Cycles This field defines the delay between ZQ Calibration Command and any Valid commands in number of cycles. The number of cycles is between 0 and 255.

334
28.7.11 MPDDRC I/O Calibration Register

31 23
MPDDRC_IO_CALIBR 0xFFFFEA34 Read-write See Table 28-13

30 22 CALCODEN 15 7 14 6 13 5 12 4 11 3 2 10 29 21 28 20 27 19 26 18 CALCODEP 9 TZQIO 1 RDIV 0 8 25 17 24 16
RDIV: Resistor Divider, Output Driver Impedance

The Reset Value is 2, which corresponds to 40 Ohm. With the LPDDR2-SDRAM device, RDIV field must be equal to DS (Drive strength) field DS: Drive Strength on page 329 RDIV is used with an external precision resistor RZQ to define the Output driver impedance. The value of RZQ is either 240 Ohm (LPDDR2 mode) or200 Ohm (DDR2/LPDDR1 device).
Value 0 1 1 1 2 3 4 5 6 7 Name RESERVED RESERVED RESERVED RZQ_34 RZQ_40_RZQ_33_3 RZQ_48_RZQ_40 RZQ_60_RZQ_50 RESERVED RZQ_80_RZQ_66_7 RZQ_120_RZQ_100 Description RZQ RESERVED RZQ RESERVED RZQ RESERVED LPDDR2 RZQ = 34,3 Ohm, DDR2/LPDDR1: Not applicable LPDDR2:RZQ = 40 Ohm, DDR2/LPDDR1: RZQ = 33,3 Ohm LPDDR2:RZQ =48 Ohm, DDR2/LPDDR1: RZQ =40 Ohm LPDDR2:RZQ =60 Ohm, DDR2/LPDDR1: RZQ =50 Ohm RZQ RESERVED LPDDR2: RZQ = 80 Ohm, DDR2/LPDDR1: RZQ = 66,7 Ohm LPDDR2:RZQ = 120 Ohm, DDR2/LPDDR1: RZQ = 100 Ohm
TZQIO: IO Calibration This field defines the delay between an IO Calibration Command and any Valid commands in number of cycles. The number of cycles is between 0 and 7. The TZQIO configuration code must be correctly set depending on the clock frequency using the following formula: TZQIO = (DDRCLK* 20ns) + 1. CALCODEP: Number of Transistor P This Register is Read Only. The Reset Value is 7. This value gives the number of transistor P to perform the calibration. CALCODEN: Number of Transistor N This Register is Read Only. The Reset Value is 8. This value gives the number of transistor N to perform the calibration. SAMA5D3 Series [DATASHEET]
335
28.7.12 MPDDRC OCMS Register

31 23 15 7
MPDDRC_OCMS 0xFFFFEA38 Read-write See Table 28-13

30 22 14 6 29 21 13 5 28 20 12 4 27 19 11 3 26 18 10 2 25 17 9 1 24 16 8 0 SCR_EN
SCR_EN: Scrambling enable

0: Disable Off-chip Scrambling for SDRAM access. 1: Enable Off-chip Scrambling for SDRAM access.

336
28.7.13 MPDDRC OCMS KEY1 Register

31
MPDDRC_OCMS_KEY1 0xFFFFEA3C Write once See Table 28-13

30 29 28 KEY1 27 26 25 24
23
22
21
20 KEY1
19
18
17
16
15
14
13
12 KEY1
11
10
4 KEY1
KEY1: Off-chip Memory Scrambling (OCMS) Key Part 1

When Off-chip Memory Scrambling is enabled setting by the MPDDRC_OMCS register, the data scrambling depends on KEY1 and KEY2 values.

337
28.7.14 MPDDRC OCMS KEY2 Register

31
MPDDRC_OCMS_KEY2 0xFFFFEA40 Write once See Table 28-13

30 29 28 KEY2 23 22 21 20 KEY2 15 14 13 12 KEY2 7 6 5 4 KEY2 3 2 1 0 11 10 9 8 19 18 17 16 27 26 25 24
KEY2: Off-chip Memory Scrambling (OCMS) Key Part 2

When Off-chip Memory Scrambling is enabled setting by the MPDDRC_OMCS register, the data scrambling depends on KEY1 and KEY2 values.

338
28.7.15 MPDDRC Memory Device Register

31 23 15 7
MPDDRC_MD 0xFFFFEA20 Read-write See Table 28-13

30 22 14 6 29 21 13 5 28 20 12 4 DBW 27 19 11 3 26 18 10 2 25 17 9 1 MD 24 16 8 0
MD: Memory Device

Indicates the type of memory used. Reset value is for DDR-SDRAM device.
Table 28-14. Value 000 001 010 011 100 101 110 111 Name LPDDR_SDRAM DDR2_SDRAM LPDDR2_SDRAM Description Reserved Reserved Reserved Low-power DDR1-SDRAM Reserved Reserved DDR2-SDRAM Low-Power DDR2-SDRAM
DBW: Data Bus Width

Reset value is 16 bits. 0 (DBW_32_BITS): Data bus width is 32 bits. 1 (DBW_16_BITS): Data bus width is 16 bits.(1) Note: 1. Only 32-bit value is used.

339
28.7.16 DDRSDRC High Speed Register

31 23 15 7
MPDDRC_HS 0xFFFFEA24 Read-write See Table 28-13

30 22 14 6 29 21 13 5 28 20 12 4 27 19 11 3 26 18 10 2 DIS_ANTICIP_READ 25 17 9 1 24 16 8 0
DIS_ANTICIP_READ: Disable Anticip Read Access

This field allows DDR read access optimization with the multi-port. As this feature is based on the bank open policy, the software must map different buffers in different DDR banks to take advantage of that feature. 0 = Anticip_read access is enabled (default). 1 = Anticip_read access is disabled.

340
28.7.17 MPDDRC Write Protect Control Register

31
MPDDRC_WPCR 0xFFFFEAE4 Read-write See Table 28-13

WPEN: Write Protection Enable

0 = Disables the Write Protection if WPKEY corresponds to the expected one. 1 = Enables the Write Protection if WPKEY corresponds to the expected one.
WPKEY: Write Protection KEY

Only one value is writable. This value is related to the block name DDR.

341
28.7.18 MPDDRC Write Protect Status Register

31 23
MPDDRC_WPSR 0xFFFFEAE8 Read-only See Table 28-13

WPVS: Write Protection Enable

0 = No Write Protection Violation occurred since the last read of this register (MPDDRC_WPSR). 1 = A Write Protection Violation occurred since the last read of this register (MPDDRC_WPSR). If this violation is an unauthorized attempt to write a control register, the associated violation is reported into field WPVSRC.
WPVSRC: Write Protection Violation Source

When WPVS is active, it indicates the register (through address or code) that should have been written if Write Protection had not been previously enabled.

342
28.7.19 MPDDRC DLL Master Offset Register

31 23 15 7 MPDDRC_DLL_MO
0xFFFFEA74 Read-write
See Table 28-13 30 22 14 6 29 21 13 5 28 20 12 27 19 11 26 18 10 CLK90OFF 2 MOFF 25 17 9 24 16 SELOFF 8
MOFF: DLL Master Delay Line Offset

The value stored by this field is unsigned. When this field is written, the programmable Master delay line offset is written. When this field is read:
If SELOFF=0: the hard-coded Master delay line offset is read. If SELOFF=1: the programmable Master delay line offset is read. CLK90OFF: DLL CLK90 Delay Line Offset
The value stored by this field is signed. When this field is written, the programmable CLK90 delay line offset is written. When this field is read:
If SELOFF=0: the hard-coded CLK90 delay line offset is read. If SELOFF=1: the programmable CLK90 delay line offset is read. SELOFF: DLL Offset Selection
0: The hard-coded Master/Slave x/CLK90 delay line offsets are selected. 1: The programmable Master/Slave x/CLK90 delay line offsets are selected.

343
28.7.20 MPDDRC DLL Slave Offset Register

31 23 15 7 MPDDRC_DLL_SOF
0xFFFFEA78 Read-write
See Table 28-13 30 22 14 6 29 21 13 5 28 27 26 S3OFF 18 S2OFF 10 S1OFF 2 S0OFF 25 24
20
19
17
16
12
11
SxOFF: DLL Slave x Delay Line Offset ([x=0..3])

The value stored by this field is signed. When this field is written, the programmable Slave x delay line offset is written. When this field is read:
If MPDDRC_DLL MOR.SELOFF=0: the hard-coded Slave x delay line offset is read. If MPDDRC_DLL MOR.SELOFF=1: the programmable Slave x delay line offset is read.

344
28.7.21 MPDDRC DLL Master Status Register

31 23 15 MPDDRC_DLL_MS
0xFFFFEA7C Read-only
See Table 28-13 30 22 14 29 21 13 28 20 12 MDVAL 27 19 11 26 18 10 25 17 9 24 16 8
2 MDOVF
1 MDDEC
0 MDINC
MDINC: DLL Master Delay Increment

0: The DLL is not incrementing the Master delay counter. 1: The DLL is incrementing the Master delay counter.
MDDEC: DLL Master Delay Decrement

0: The DLL is not decrementing the Master delay counter. 1: The DLL is decrementing the Master delay counter.
MDOVF: DLL Master Delay Overflow Flag

0: The Master delay counter has not reached its maximum value, or the Master is not locked yet 1: The Master delay counter has reached its maximum value, the Master delay counter increment is stopped and the DLL forces the Master lock. If this flag is set, it means the MPDDRC clock frequency is too low compared to Master delay line number of elements.
MDVAL: DLL Master Delay Value

Value of the Master delay counter.

345
28.7.22 MPDDRC DLL Status Slave x Register

31 23 MPDDRC_DLL_SSx
0xFFFFEA80 Read-only
See Table 28-13 30 22 29 21 28 20 SDCVAL 27 19 26 18 25 17 24 16
15
14
13
12 SDVAL
11
10
2 SDERF
1 SDCUDF
0 SDCOVF
SDCOVF: DLL Slave x Delay Correction Overflow Flag

0 = Due to the correction, the Slave x delay counter has not reached its maximum value, or the Slave x is not locked yet. 1 = Due to the correction, the Slave x delay counter has reached its maximum value, the correction is not optimal because it has not been entirely applied.
SDCUDF: DLL Slave x Delay Correction Underflow Flag

0 = Due to the correction, the Slave x delay counter has not reached its minimum value, or the Slave x is not locked yet. 1 = Due to the correction, the Slave x delay counter has reached its minimum value, the correction is not optimal because it has not been entirely applied.
SDERF: DLL Slave x Delay Correction Error Flag

0 = The DLL has succeeded in computing the Slave x delay correction, or the Slave x is not locked yet. 1 = The DLL has not succeeded in computing the Slave x delay correction.
SDVAL: DLL Slave x Delay Value

Value of the Slave x delay counter.
SDCVAL: DLL Slave x Delay Correction Value

Value of the correction applied to the Slave x delay.

346
29.
29.1
Static Memory Controller (HSMC)

Description
The External Bus Interface is designed to ensure the successful data transfer between several external devices and the ARM processor based device. The External Bus Interface of the SAMA5 consists of a Static Memory Controller (HSMC). This HSMC is capable of handling several types of external memory and peripheral devices, such as SRAM, PSRAM, PROM, EPROM, EEPROM, LCD Module, NOR Flash and NAND Flash. The HSMC generates the signals that control the access to external memory devices or peripheral devices. It has 4 Chip Selects and a 26-bit address bus. The 16-bit data bus can be configured to interface with 8- or 16-bit external devices. Separate read and write control signals allow for direct memory and peripheral interfacing. Read and write signal waveforms are fully parametrizable. The HSMC can manage wait requests from external devices to extend the current access. The HSMC is provided with an automatic slow clock mode. In slow clock mode, it switches from user-programmed waveforms to slow-rate specific waveforms on read and write signals. The HSMC embeds a NAND Flash Controller (NFC). The NFC can handle automatic transfers, sending the commands and address cycles to the NAND Flash and transferring the contents of the page (for read and write) to the NFC SRAM. It minimizes the CPU overhead. The HSMC includes programmable hardware error correcting code with one-bit error correction capability and supports two-bit error detection. In order to improve the overall system performance, the DATA phase of the transfer can be DMAassisted. The External Data Bus can be scrambled/unscrambled by means of user keys.

347
29.2
z z z z z z z z z z z z z z z z z z z z z z z z z z z z z z z z z
64-Mbyte Address Space per Chip Select 8-bit or 16-bit Data Bus Word, Halfword, Byte Transfers Byte Write or Byte Select Lines Programmable Setup, Pulse and Hold Time for Read Signals per Chip Select Programmable Setup, Pulse and Hold Time for Write Signals per Chip Select Programmable Data Float Time per Chip Select External Data Bus Scrambling/Unscrambling Function External Wait Request Automatic Switch to Slow Clock Mode NAND Flash Controller Supporting NAND Flash with Multiplexed Data/Address Buses Supports SLC and MLC NAND Flash Technology Hardware Configurable Number of Chip Selects from 1 to 4 Programmable Timing on a Per-chip Select Basis Programmable Flash Data Width 8 bits or 16 bits Supports NAND Flash and SmartMedia Devices with 8-bit or 16-bit Data Path Supports NAND Flash with Page Sizes of 528, 1056, 2112 and 4224 Bytes, Specified by Software Supports 1-bit Correction for a Page of 512, 1024, 2048 and 4096 Bytes with 8- or 16-bit Data Path Supports 1-bit Correction per 512 Bytes of Data for a Page Size of 512, 2048 and 4096 Bytes with 8-bit Data Path Supports 1-bit Correction per 256 Bytes of Data for a Page Size of 512, 2048 and 4096 Bytes with 8-bit Data Path Multibit Error Correcting Code (ECC) ECC Algorithm based on binary shortened Bose, Chaudhuri and Hocquenghem (BCH) codes Programmable Error Correcting Capability: 2, 4, 8, 12 and 24 bits of errors per block Programmable block size: 512 Bytes or 1024 Bytes Programmable number of block per page: 1, 2, 4 or 8 blocks of data per page Programmable spare area size Supports spare area ECC protection Supports 8 Kbytes page size using 1024 Bytes/block and 4 Kbytes page size using 512 Bytes/block Multibit Error detection is interrupt driven Provides hardware acceleration for determining roots of polynomials defined over a finite field Programmable finite Field GF(2^13) or GF(2^14) Finds roots of error-locator polynomial Programmable number of roots

348
29.3
Block Diagram
NAND Flash Controller (NFC)
AHB arbiter
SMC Scrambler
SMC AHB Interface
SMC Interface
D[15:0] A[0]/NBS0 A[20:1] A21/NANDALE A22/NANDCLE A[25:23] NCS[3:0] NRD NWR0/NWE NWR1/NBS1 NANDOE NANDWE NANDRDY NWAIT
SRAM AHB Interface
SRAM Scrambler
User Interface NFC (8 KBytes) Internal SRAM Control & Status Registers
APB interface

349
29.4
Table 29-1. I/O Line Description Name NCS[3:0] NRD NWR0/NWE A0/NBS0 NWR1/NBS1 A[25:1] D[15:0] NWAIT NANDRDY NANDWE NANDOE NANDALE NANDCLE Description Static Memory Controller Chip Select Lines Read Signal Write 0/Write Enable Signal Address Bit 0/Byte 0 Select Signal Write 1/Byte 1 Select Signal Address Bus Data Bus External Wait Signal NAND Flash Ready/Busy NAND Flash Write Enable NAND Flash Output Enable NAND Flash Address Latch Enable NAND Flash Command Latch Enable Type Output Output Output Output Output Output I/O Input Input Output Output Output Output Low Low Low Active Level Low Low Low Low Low
29.5
Multiplexed Signals
Table 29-2. Static Memory Controller (HSMC) Multiplexed Signals Multiplexed Signals NWR0 A0 A22 A21 NWR1 A1 NWE NBS0 NANDCLE NANDALE NBS1 Related Function Byte-write or Byte-select access, see Figure 29-4 "Memory Connection for an 8-bit Data Bus" and Figure 29-5 "Memory Connection for a 16-bit Data Bus" 8-bit or 16-bit data bus, see Section 29.9.1 Data Bus Width NAND Flash Command Latch Enable NAND Flash Address Latch Enable Byte-write or Byte-select access, see Figure 29-4 and Figure 29-5 8-/16-bit data bus, see Section 29.9.1 Data Bus Width Byte-write or Byte-select access, see Figure 29-4 and Figure 29-5

350
29.6
Application Example
29.6.1 Hardware Interface

Figure 29-2. HSMC Connections to Static Memory Devices
D0-D15
A0/NBS0 NWR0/NWE NWR1/NBS1 A1
D0 - D7
128K x 8 SRAM
D0 - D7 CS A0 - A16 A2 - A18
D8-D15
128K x 8 SRAM
D0-D7 CS A0 - A16 A2 - A18
NRD NCS0 NCS1 NCS2 NCS3 NWR0/NWE
OE WE
NRD NWR1/NBS1
OE WE
A2 - A23
Static Memory Controller

351
29.7
29.7.1 I/O Lines

The pins used for interfacing the Static Memory Controller are multiplexed with the PIO lines. The programmer must first program the PIO controller to assign the Static Memory Controller pins to their peripheral function. If I/O Lines of the HSMC are not used by the application, they can be used for other purposes by the PIO Controller.

The HSMC is clocked through the Power Management Controller (PMC), thus the programmer must first configure the PMC to enable the HSMC clock.
29.7.3 Interrupt
The HSMC has an interrupt line connected to the Nested Vector Interrupt Controller (NVIC). Handling the HSMC interrupt requires programming the NVIC before configuring the HSMC.
Table 29-3. Peripheral IDs Instance HSMC ID 5
29.8
External Memory Mapping

The HSMC provides up to 26 address lines, A[25:0]. This allows each chip select line to address up to 64 MBytes of memory. If the physical memory device connected on one chip select is smaller than 64 MBytes, it wraps around and appears to be repeated within this space. The HSMC correctly handles any valid access to the memory device within the page (see Figure 29-3). A[25:0] is only significant for 8-bit memory; A[25:1] is used for 16-bit memory.
Figure 29-3. Memory Connections for External Devices

NCS[0] - NCS[3] NRD SMC NWE A[25:0] D[15:0] NCS3 NCS2 NCS1 NCS0 Memory Enable Memory Enable Memory Enable
Memory Enable Output Enable Write Enable A[25:0]
8 or 16
D[15:0] or D[7:0]

352
29.9
Connection to External Devices
29.9.1 Data Bus Width

A data bus width of 8 or 16 bits can be selected for each chip select. This option is controlled by the field DBW in HSMC_MODE (Mode Register) for the corresponding chip select. Figure 29-4 shows how to connect a 512K x 8-bit memory on NCS2. Figure 29-5 shows how to connect a 512K x 16-bit memory on NCS2.
29.9.2 Byte Write or Byte Select Access

Each chip select with a 16-bit data bus can operate with one of two different types of write access: Byte write or Byte select access. This is controlled by the BAT field of the HSMC_MODE register for the corresponding chip select.
Figure 29-4. Memory Connection for an 8-bit Data Bus
D[7:0] A[18:2] A1 SMC A0 NWE NRD NCS[2] D[7:0]
A[18:2] A1 A0 Write Enable Output Enable Memory Enable
Figure 29-5. Memory Connection for a 16-bit Data Bus

D[15:0] A[19:2] A1 SMC NBS0 NBS1 NWE NRD NCS[2] D[15:0] A[18:1] A[0] Low Byte Enable High Byte Enable Write Enable Output Enable Memory Enable
29.9.2.1 Byte Write Access Byte write access supports one write signal per Byte of the data bus and a single read signal. Note that the HSMC does not allow boot in Byte Write Access mode. z For 16-bit devices: the HSMC provides NWR0 and NWR1 write signals for respectively, Byte0 (lower Byte) and Byte1 (upper Byte) of a 16-bit bus. One single read signal (NRD) is provided. Byte Write Access is used to connect 2 x 8-bit devices as a 16-bit memory. 29.9.2.2 Byte Select Access In this mode, read/write operations can be enabled/disabled at Byte level. One Byte-select line per Byte of the data bus is provided. One NRD and one NWE signal control read and write. z For 16-bit devices: the HSMC provides NBS0 and NBS1 selection signals for respectively Byte0 (lower Byte) and Byte1 (upper Byte) of a 16-bit bus. Byte Select Access is used to connect one 16-bit device.

353
Figure 29-6. Connection of 2 x 8-bit Devices on a 16-bit Bus: Byte Write Option
D[7:0] D[15:8] A[24:2] A[23:1] A[0] Write Enable Read Enable Memory Enable D[7:0]
SMC
A1 NWR0 NWR1 NRD NCS[3]
D[15:8] A[23:1] A[0] Write Enable Read Enable Memory Enable
29.9.2.3 Signal Multiplexing Depending on the Byte Access Type (BAT), only the write signals or the Byte select signals are used. To save IOs at the external bus interface, control signals at the HSMC interface are multiplexed. Table 29-4 shows signal multiplexing depending on the data bus width and the Byte Access Type. For 16-bit devices, bit A0 of address is unused. When Byte Select Option is selected, NWR1 is unused. When Byte Write option is selected, NBS0 is unused.
Table 29-4. HSMC Multiplexed Signal Translation Signal Name Device Type Byte Access Type (BAT) NBS0_A0 NWE_NWR0 NBS1_NWR1 A1 1x16-bit Byte Select NBS0 NWE NBS1 A1 NWR0 NWR1 A1 A1 16-bit Bus 2 x 8-bit Byte Write A0 NWE 8-bit Bus 1 x 8-bit

354
29.10 Standard Read and Write Protocols

In the following sections, the Byte Access Type is not considered. Byte select lines (NBS0 to NBS1) always have the same timing as the A address bus. NWE represents either the NWE signal in Byte select access type or one of the Byte write lines (NWR0 to NWR1) in Byte write access type. NWR0 to NWR1 have the same timings and protocol as NWE. In the same way, NCS represents one of the NCS[0..3] chip select lines.
29.10.1 Read Waveforms

The read cycle is shown on Figure 29-7. The read cycle starts with the address setting on the memory address bus, i.e.: {A[25:2], A1, A0} for 8-bit devices {A[25:2], A1} for 16-bit devices
Figure 29-7. Standard Read Cycle
MCK
A[25:2]
NBS0,NBS1, A0, A1
NRD
NCS
D[15:0] NRD_SETUP NRD_PULSE NRD_HOLD
NCS_RD_SETUP
NCS_RD_PULSE NRD_CYCLE
NCS_RD_HOLD
29.10.1.1 NRD Waveform The NRD signal is characterized by a setup timing, a pulse width and a hold timing. 1. 2. 3. NRD_SETUP: the NRD setup time is defined as the setup of address before the NRD falling edge. NRD_PULSE: the NRD pulse length is the time between NRD falling edge and NRD rising edge. NRD_HOLD: the NRD hold time is defined as the hold time of address after the NRD rising edge.
29.10.1.2 NCS Waveform Similarly, the NCS signal can be divided into a setup time, pulse length and hold time: 1. 2. 3. NCS_RD_SETUP: the NCS setup time is defined as the setup time of address before the NCS falling edge. NCS_RD_PULSE: the NCS pulse length is the time between NCS falling edge and NCS rising edge. NCS_RD_HOLD: the NCS hold time is defined as the hold time of address after the NCS rising edge.

355
29.10.1.3 Read Cycle The NRD_CYCLE time is defined as the total duration of the read cycle, i.e., from the time where address is set on the address bus to the point where address may change. The total read cycle time is equal to: NRD_CYCLE = NRD_SETUP + NRD_PULSE + NRD_HOLD = NCS_RD_SETUP + NCS_RD_PULSE + NCS_RD_HOLD All NRD and NCS timings are defined separately for each chip select as an integer number of Master Clock cycles. To ensure that the NRD and NCS timings are coherent, the user must define the total read cycle instead of the hold timing. NRD_CYCLE implicitly defines the NRD hold time and NCS hold time as: NRD_HOLD = NRD_CYCLE - NRD SETUP - NRD PULSE NCS_RD_HOLD = NRD_CYCLE - NCS_RD_SETUP - NCS_RD_PULSE
29.10.2 Read Mode

As NCS and NRD waveforms are defined independently of one another, the HSMC needs to know when the read data is available on the data bus. The HSMC does not compare NCS and NRD timings to know which signal rises first. The READ_MODE parameter in the HSMC_MODE register of the corresponding chip select indicates which signal of NRD and NCS controls the read operation. 29.10.2.1 Read is Controlled by NRD (READ_MODE = 1): Figure 29-8 shows the waveforms of a read operation of a typical asynchronous RAM. The read data is available tPACC after the falling edge of NRD, and turns to Z after the rising edge of NRD. In this case, the READ_MODE must be set to 1 (read is controlled by NRD), to indicate that data is available with the rising edge of NRD. The HSMC samples the read data internally on the rising edge of the Master Clock that generates the rising edge of NRD, whatever the programmed waveform of NCS.
Figure 29-8. READ_MODE = 1: Data is Sampled by HSMC before the Rising Edge of NRD
MCK
A[25:2]
NBS0,NBS1, A0, A1
NRD
NCS tPACC D[15:0]
Data Sampling

356
29.10.2.2 Read is Controlled by NCS (READ_MODE = 0) Figure 29-9 shows the typical read cycle. The read data is valid tPACC after the falling edge of the NCS signal and remains valid until the rising edge of NCS. Data must be sampled when NCS is raised. In that case, the READ_MODE must be set to 0 (read is controlled by NCS): the HSMC internally samples the data on the rising edge of the Master Clock that generates the rising edge of NCS, whatever the programmed waveform of NRD.
Figure 29-9. READ_MODE = 0: Data is Sampled by HSMC before the Rising Edge of NCS
MCK
A[25:2]
NBS0,NBS1, A0, A1
NRD
NCS tPACC D[15:0]
Data Sampling

357
29.10.3 Write Waveforms

The write protocol is similar to the read protocol. It is depicted in Figure 29-10. The write cycle starts with the address setting on the memory address bus. 29.10.3.1 NWE Waveforms The NWE signal is characterized by a setup timing, a pulse width and a hold timing. 1. 2. 3. NWE_SETUP: the NWE setup time is defined as the setup of address and data before the NWE falling edge. NWE_PULSE: The NWE pulse length is the time between NWE falling edge and NWE rising edge. NWE_HOLD: The NWE hold time is defined as the hold time of address and data after the NWE rising edge.
The NWE waveforms apply to all Byte-write lines in Byte Write access mode: NWR0 to NWR3. 29.10.3.2 NCS Waveforms The NCS signal waveforms in write operation are not the same as those applied in read operations, but are separately defined: 1. 2. 3. NCS_WR_SETUP: the NCS setup time is defined as the setup time of address before the NCS falling edge. NCS_WR_PULSE: the NCS pulse length is the time between NCS falling edge and NCS rising edge. NCS_WR_HOLD: the NCS hold time is defined as the hold time of address after the NCS rising edge.
Figure 29-10. Write Cycle

MCK
A[25:2]
NBS0, NBS1, A0, A1
NWE
NCS
NWE_SETUP
NWE_PULSE
NWE_HOLD
NCS_WR_SETUP
NCS_WR_PULSE NWE_CYCLE
NCS_WR_HOLD

358
29.10.3.3 Write Cycle The write cycle time is defined as the total duration of the write cycle, that is, from the time where address is set on the address bus to the point where address may change. The total write cycle time is equal to: NWE_CYCLE = NWE_SETUP + NWE_PULSE + NWE_HOLD = NCS_WR_SETUP + NCS_WR_PULSE + NCS_WR_HOLD All NWE and NCS (write) timings are defined separately for each chip select as an integer number of Master Clock cycles. To ensure that the NWE and NCS timings are coherent, the user must define the total write cycle instead of the hold timing. This implicitly defines the NWE hold time and NCS (write) hold times as: NWE_HOLD = NWE_CYCLE - NWE_SETUP - NWE_PULSE NCS_WR_HOLD = NWE_CYCLE - NCS_WR_SETUP - NCS_WR_PULSE
29.10.4 Write Mode

The WRITE_MODE parameter in the HSMC_MODE register of the corresponding chip select indicates which signal controls the write operation. 29.10.4.1 Write is Controlled by NWE (WRITE_MODE = 1) Figure 29-11 shows the waveforms of a write operation with WRITE_MODE set to 1. The data is put on the bus during the pulse and hold steps of the NWE signal. The internal data buffers are switched to output mode after the NWE_SETUP time, and until the end of the write cycle, regardless of the programmed waveform on NCS.
Figure 29-11. WRITE_MODE = 1. The write operation is controlled by NWE
MCK
A[25:2]
NBS0, NBS1, A0, A1 NWE, NWR0, NWR1
NCS D[15:0]

359
29.10.4.2 Write is Controlled by NCS (WRITE_MODE = 0) Figure 29-12 shows the waveforms of a write operation with WRITE_MODE set to 0. The data is put on the bus during the pulse and hold steps of the NCS signal. The internal data buffers are switched to output mode after the NCS_WR_SETUP time, and until the end of the write cycle, regardless of the programmed waveform on NWE.
Figure 29-12. WRITE_MODE = 0. The write operation is controlled by NCS
MCK
A[25:2]
NBS0, NBS1, A0, A1 NWE, NWR0, NWR1
NCS
D[15:0]
29.10.5 Coding Timing Parameters

All timing parameters are defined for one chip select and are grouped together in one HSMC_REGISTER according to their type.
z z z
The HSMC_SETUP register groups the definition of all setup parameters: NRD_SETUP, NCS_RD_SETUP, NWE_SETUP, NCS_WR_SETUP The HSMC_PULSE register groups the definition of all pulse parameters: NRD_PULSE, NCS_RD_PULSE, NWE_PULSE, NCS_WR_PULSE The HSMC_CYCLE register groups the definition of all cycle parameters: NRD_CYCLE, NWE_CYCLE
Table 29-5 shows how the timing parameters are coded and their permitted range.
Table 29-5. Coding and Range of Timing Parameters Permitted Range Coded Value setup [5:0] Number of Bits 6 Effective Value 128 x setup[5] + setup[4:0] Coded Value 0 setup 31 32 setup 63 0 pulse 63 64 pulse 127 0 cycle 127 cycle [8:0] 9 256 x cycle[8:7] + cycle[6:0] 128 cycle 255 256 cycle 383 384 cycle 511 Effective Value 0..31 128..(128+31) 0..63 256..(256+63) 0..127 256..(256+127) 512..(512+127) 768..(768+127)
pulse [6:0]
256 x pulse[6] + pulse[5:0]

360
29.10.6 Reset Values of Timing Parameters

Table 29-6 gives the default value of timing parameters at reset.
Table 29-6. Reset Values of Timing Parameters Register HSMC_SETUP HSMC_PULSE HSMC_CYCLE WRITE_MODE READ_MODE Reset Value 1 1 Description All setup timings are set to 1 All pulse timings are set to 1 The read and write operations last 3 Master Clock cycles and provide one hold cycle Write is controlled with NWE Read is controlled with NRD
29.10.7 Usage Restriction

The HSMC does not check the validity of the user-programmed parameters. If the sum of SETUP and PULSE parameters is larger than the corresponding CYCLE parameter, this leads to an unpredictable behavior of the HSMC. 29.10.7.1 For Read Operations Null but positive setup and hold of address and NRD and/or NCS cannot be guaranteed at the memory interface because of the propagation delay of these signals through external logic and pads. When positive setup and hold values must be verified, then it is strictly recommended to program non-null values so as to cover possible skews between address, NCS and NRD signals. 29.10.7.2 For Write Operations If a null hold value is programmed on NWE, the HSMC can guarantee a positive hold of address, Byte select lines, and NCS signal after the rising edge of NWE. This is true for WRITE_MODE = 1 only. See Early Read Wait State on page 364. 29.10.7.3 For Read and Write Operations A null value for pulse parameters is forbidden and may lead to an unpredictable behavior. In read and write cycles, the setup and hold time parameters are defined in reference to the address bus. For external devices that require setup and hold time between NCS and NRD signals (read), or between NCS and NWE signals (write), these setup and hold times must be converted into setup and hold times in reference to the address bus.

361
29.11 Scrambling/Unscrambling Function

The external data bus D[15:0] can be scrambled in order to prevent intellectual property data located in off-chip memories from being easily recovered by analyzing data at the package pin level of either the microcontroller or the memory device. The scrambling and unscrambling are performed on-the-fly without additional wait states. The scrambling method depends on two user-configurable key registers, HSMC_KEY1 and HSMC_KEY2. These key registers are only accessible in write mode. The key must be securely stored in a reliable non-volatile memory in order to recover data from the off-chip memory. Any data scrambled with a given key cannot be recovered if the key is lost. The scrambling/unscrambling function can be enabled or disabled by programming the HSMC_OCMS register. One bit is dedicated to enabling/disabling the NAND Flash scrambling, and one bit is dedicated to enabling/disabling the off chip SRAM scrambling. When at least one external SRAM is scrambled, the SMSC field must be set in the HSMC_OCMS register. When multiple chip selects (external SRAM) are handled, it is possible to configure the scrambling function per chip select using the OCMS field in the HSMC_TIMINGS registers. To scramble the NAND Flash contents, the SRSE field must be set in the HSMC_OCMS register. When the NAND Flash memory content is scrambled, the on-chip SRAM page buffer associated for the transfer is also scrambled.

362
29.12 Automatic Wait States

Under certain circumstances, the HSMC automatically inserts idle cycles between accesses to avoid bus contention or operation conflict.
29.12.1 Chip Select Wait States

The HSMC always inserts an idle cycle between 2 transfers on separate chip selects. This idle cycle ensures that there is no bus contention between the de-activation of one device and the activation of the next one. During chip select wait state, all control lines are turned inactive: NBS0 to NBS1, NWR0 to NWR1, NCS[0..3], and NRD lines. They are all set to 1. Figure 29-13 illustrates a chip select wait state between access on Chip Select 0 and Chip Select 2.
Figure 29-13. Chip Select Wait State between a Read Access on NCS0 and a Write Access on NCS2
MCK
A[25:2]
NBS0, NBS1, A0,A1 NRD NWE
NCS0
NCS2 NRD_CYCLE D[15:0] NWE_CYCLE
Read to Write Chip Select Wait State Wait State

363
29.12.2 Early Read Wait State

In some cases, the HSMC inserts a wait state cycle between a write access and a read access to allow time for the write cycle to end before the subsequent read cycle begins. This wait state is not generated in addition to a chip select wait state. The early read cycle thus only occurs between a write and read access to the same memory device (same chip select). An early read wait state is automatically inserted if at least one of the following conditions is valid:
z z
If the write controlling signal has no hold time and the read controlling signal has no setup time (Figure 29-14). In NCS write controlled mode (WRITE_MODE = 0), if there is no hold timing on the NCS signal and the NCS_RD_SETUP parameter is set to 0, regardless of the read mode (Figure 29-15). The write operation must end with an NCS rising edge. Without an Early Read Wait State, the write operation could not complete properly. In NWE controlled mode (WRITE_MODE = 1) and if there is no hold timing (NWE_HOLD = 0), the feedback of the write control signal is used to control address, data, chip select and Byte select lines. If the external write control signal is not inactivated as expected due to load capacitances, an Early Read Wait State is inserted and address, data and control signals are maintained one more cycle. See Figure 29-16.
Figure 29-14. Early Read Wait State: Write with No Hold Followed by Read with No Setup
MCK
A[25:2]
NBS0, NBS1, A0, A1
NWE
NRD no hold no setup D[15:0]
write cycle
Early Read wait state
read cycle

364
Figure 29-15. Early Read Wait State: NCS Controlled Write with No Hold Followed by a Read with No NCS Setup
MCK
A[25:2]
NBS0, NBS1, A0,A1
NCS
NRD no hold D[15:0] no setup
read cycle write cycle Early Read (WRITE_MODE = 0) wait state (READ_MODE = 0 or READ_MODE = 1)
Figure 29-16. Early Read Wait State: NWE-controlled Write with No Hold Followed by a Read with one Set-up Cycle
MCK
A[25:2]
NBS0, NBS1, A0, A1 Internal write controlling signal External write controlling signal (NWE) No hold NRD Read setup = 1
D[15:0]
Write cycle Early Read Read cycle (WRITE_MODE = 1) wait state (READ_MODE = 0 or READ_MODE = 1)

365
29.12.3 Reload User Configuration Wait State

The user may change any of the configuration parameters by writing the HSMC user interface. When detecting that a new user configuration has been written in the user interface, the HSMC inserts a wait state before starting the next access. The so called Reload User Configuration Wait State is used by the HSMC to load the new set of parameters to apply to next accesses. The Reload Configuration Wait State is not applied in addition to the Chip Select Wait State. If accesses before and after re-programming the user interface are made to different devices (Chip Selects), then one single Chip Select Wait State is applied. On the other hand, if accesses before and after writing the user interface are made to the same device, a Reload Configuration Wait State is inserted, even if the change does not concern the current Chip Select. 29.12.3.1 User Procedure To insert a Reload Configuration Wait State, the HSMC detects a write access to any HSMC_MODE register of the user interface. If only the timing registers are modified (HSMC_SETUP, HSMC_PULSE, HSMC_CYCLE registers) in the user interface, the user must validate the modification by writing the HSMC_MODE register, even if no change was made on the mode parameters. 29.12.3.2 Slow Clock Mode Transition A Reload Configuration Wait State is also inserted when the Slow Clock Mode is entered or exited, after the end of the current transfer (see Section 29.15 Slow Clock Mode).
29.12.4 Read to Write Wait State

Due to an internal mechanism, a wait cycle is always inserted between consecutive read and write HSMC accesses. This wait cycle is referred to as a read to write wait state in this document. This wait cycle is applied in addition to chip select and reload user configuration wait states when they are to be inserted. See Figure 29-13 on page 363.

366
29.13 Data Float Wait States

Some memory devices are slow to release the external bus. For such devices, it is necessary to add wait states (data float wait states) after a read access:
z z
Before starting a read access to a different external memory Before starting a write access to the same device or to a different external one.
The Data Float Output Time (tDF) for each external memory device is programmed in the TDF_CYCLES field of the HSMC_MODE register for the corresponding chip select. The value of TDF_CYCLES indicates the number of data float wait cycles (between 0 and 15) before the external device releases the bus, and represents the time allowed for the data output to go to high impedance after the memory is disabled. Data float wait states do not delay internal memory accesses. Hence, a single access to an external memory with long tDF will not slow down the execution of a program from internal memory. The data float wait states management depends on the READ_MODE and the TDF_MODE fields of the HSMC_MODE register for the corresponding chip select.
29.13.1 READ_MODE
Setting READ_MODE to 1 indicates to the HSMC that the NRD signal is responsible for turning off the tri-state buffers of the external memory device. The Data Float Period then begins after the rising edge of the NRD signal and lasts TDF_CYCLES MCK cycles. When the read operation is controlled by the NCS signal (READ_MODE = 0), the TDF field gives the number of MCK cycles during which the data bus remains busy after the rising edge of NCS. Figure 29-17 illustrates the Data Float Period in NRD-controlled mode (READ_MODE =1), assuming a data float period of 2 cycles (TDF_CYCLES = 2). Figure 29-18 shows the read operation when controlled by NCS (READ_MODE = 0) and the TDF_CYCLES parameter equals 3.
Figure 29-17. TDF Period in NRD Controlled Read Access (TDF = 2)
MCK
A[25:2]
NBS0, NBS1, A0, A1
NRD
NCS tpacc D[15:0] TDF = 2 clock cycles
NRD controlled read operation

367
Figure 29-18. TDF Period in NCS Controlled Read Operation (TDF = 3)

MCK
A[25:2]
NBS0, NBS1, A0,A1
NRD
NCS tpacc D[15:0]
TDF = 3 clock cycles NCS controlled read operation
29.13.2 TDF Optimization Enabled (TDF_MODE = 1)

When the TDF_MODE of the HSMC_MODE register is set to 1 (TDF optimization is enabled), the HSMC takes advantage of the setup period of the next access to optimize the number of wait states cycle to insert. Figure 29-19 shows a read access controlled by NRD, followed by a write access controlled by NWE, on Chip Select 0. Chip Select 0 has been programmed with: NRD_HOLD = 4; READ_MODE = 1 (NRD controlled) NWE_SETUP = 3; WRITE_MODE = 1 (NWE controlled) TDF_CYCLES = 6; TDF_MODE = 1 (optimization enabled).

368
Figure 29-19. TDF Optimization: No TDF wait states are inserted if the TDF period is over when the next access begins
MCK
A[25:2]
NRD NRD_HOLD= 4 NWE
NWE_SETUP= 3 NCS0
TDF_CYCLES = 6
D[15:0]
Read access on NCS0 (NRD controlled)
Read to Write Wait State
Write access on NCS0 (NWE controlled)
29.13.3 TDF Optimization Disabled (TDF_MODE = 0)

When optimization is disabled, TDF wait states are inserted at the end of the read transfer, so that the data float period ends when the second access begins. If the hold period of the read1 controlling signal overlaps the data float period, no additional TDF wait states will be inserted. Figure 29-20, Figure 29-21 and Figure 29-22 illustrate the cases:
z z z
Read access followed by a read access on another chip select, Read access followed by a write access on another chip select, Read access followed by a write access on the same chip select, with no TDF optimization.

369
Figure 29-20. TDF Optimization Disabled (TDF Mode = 0). TDF wait states between 2 read accesses on different chip selects
MCK
A[25:2]
NBS0, NBS1, A0, A1 read1 controlling signal (NRD)
read1 hold = 1
read2 setup = 1
read2 controlling signal (NRD) D[15:0]
TDF_CYCLES = 6
5 TDF WAIT STATES read1 cycle TDF_CYCLES = 6 Chip Select Wait State read2 cycle TDF_MODE = 0 (optimization disabled)
Figure 29-21. TDF Mode = 0: TDF wait states between a read and a write access on different chip selects
MCK
A[25:2]
NBS0, NBS1, A0, A1 read1 controlling signal (NRD)
read1 hold = 1
write2 setup = 1
write2 controlling signal (NWE)
TDF_CYCLES = 4
D[15:0]
read1 cycle TDF_CYCLES = 4
2 TDF WAIT STATES

Read to Write Chip Select Wait State Wait State
write2 cycle TDF_MODE = 0 (optimization disabled)

370
Figure 29-22. TDF Mode = 0: TDF wait states between read and write accesses on the same chip select
MCK
A[25:2] NBS0, NBS1, A0, A1 read1 controlling signal (NRD)
read1 hold = 1
write2 setup = 1
write2 controlling signal (NWE)
TDF_CYCLES = 5
D[15:0]
4 TDF WAIT STATES read1 cycle TDF_CYCLES = 5 write2 cycle TDF_MODE = 0 (optimization disabled)
Read to Write Wait State
29.14 External Wait

Any access can be extended by an external device using the NWAIT input signal of the HSMC. The EXNW_MODE field of the HSMC_MODE register on the corresponding chip select must be set to either 10 (frozen mode) or 11 (ready mode). When the EXNW_MODE is set to 00 (disabled), the NWAIT signal is simply ignored on the corresponding chip select. The NWAIT signal delays the read or write operation in regards to the read or write controlling signal, depending on the read and write modes of the corresponding chip select.
29.14.1 Restriction
When one of the EXNW_MODE is enabled, it is mandatory to program at least one hold cycle for the read/write controlling signal. For that reason, the NWAIT signal cannot be used in Slow Clock Mode (Slow Clock Mode on page 377). The NWAIT signal is assumed to be a response of the external device to the read/write request of the HSMC. NWAIT is then examined by the HSMC in the pulse state of the read or write controlling signal. The assertion of the NWAIT signal outside the expected period has no impact on the HSMC behavior.
29.14.2 Frozen Mode

When the external device asserts the NWAIT signal (active low), and after an internal synchronization of this signal, the HSMC state is frozen, i.e., HSMC internal counters are frozen, and all control signals remain unchanged. When the resynchronized NWAIT signal is deasserted, the HSMC completes the access, resuming the access from the point where it was stopped. See Figure 29-23. This mode must be selected when the external device uses the NWAIT signal to delay the access and to freeze the HSMC. The assertion of the NWAIT signal outside the expected period is ignored as illustrated in Figure 29-24.

371
Figure 29-23. Write Access with NWAIT Assertion in Frozen Mode (EXNW_MODE = 10)
MCK
A[25:2]
NBS0, NBS1, A0,A1 FROZEN STATE 4 NWE 6 NCS 5 4 3 2 2 2 2 1 0 3 2 1 1 1 1 0
D[15:0]
NWAIT
Internally synchronized NWAIT signal
Write cycle
EXNW_MODE = 10 (Frozen) WRITE_MODE = 1 (NWE_controlled) NWE_PULSE = 5 NCS_WR_PULSE = 7

372
Figure 29-24. Read Access with NWAIT Assertion in Frozen Mode (EXNW_MODE = 10)
MCK
A[25:2]
NBS0, NBS1, A0,A1 FROZEN STATE NCS 4 3 2 2 2 1 0 2 1 NRD 0 5 5 5 4 3 2 1 0 1 0
NWAIT
Read cycle
EXNW_MODE = 10 (Frozen) READ_MODE = 0 (NCS_controlled) NRD_PULSE = 2, NRD_HOLD = 6 NCS_RD_PULSE =5, NCS_RD_HOLD =3 Assertion is ignored

373
29.14.3 Ready Mode

In Ready mode (EXNW_MODE = 11), the HSMC behaves differently. Normally, the HSMC begins the access by down counting the setup and pulse counters of the read/write controlling signal. In the last cycle of the pulse phase, the resynchronized NWAIT signal is examined. If asserted, the HSMC suspends the access as shown in Figure 29-25 and Figure 29-26. After deassertion, the access is completed: the hold step of the access is performed. This mode must be selected when the external device uses deassertion of the NWAIT signal to indicate its ability to complete the read or write operation. If the NWAIT signal is deasserted before the end of the pulse, or asserted after the end of the pulse of the controlling read/write signal, it has no impact on the access length as shown in Figure 29-26.
Figure 29-25. NWAIT Assertion in Write Access: Ready Mode (EXNW_MODE = 11)
MCK
A[25:2]
NBS0, NBS1, A0,A1 Wait STATE 4 NWE 6 NCS 5 4 3 2 1 1 1 0 3 2 1 0 0 0
D[15:0]
NWAIT
Write cycle
EXNW_MODE = 11 (Ready mode) WRITE_MODE = 1 (NWE_controlled) NWE_PULSE = 5 NCS_WR_PULSE = 7

374
Figure 29-26. NWAIT Assertion in Read Access: Ready Mode (EXNW_MODE = 11)
MCK
A[25:2]
NBS0, NBS1, A0,A1 Wait STATE 6 NCS 5 4 3 2 1 0 0
NRD
NWAIT
Read cycle
EXNW_MODE = 11(Ready mode) READ_MODE = 0 (NCS_controlled) Assertion is ignored NRD_PULSE = 7 NCS_RD_PULSE =7 Assertion is ignored

375
29.14.4 NWAIT Latency and Read/Write Timings

There may be a latency between the assertion of the read/write controlling signal and the assertion of the NWAIT signal by the device. The programmed pulse length of the read/write controlling signal must be at least equal to this latency plus the 2 cycles of resynchronization + 1 cycle. Otherwise, the HSMC may enter the hold state of the access without detecting the NWAIT signal assertion. This is true in frozen mode as well as in ready mode. This is illustrated on Figure 29-27. When EXNW_MODE is enabled (ready or frozen), the user must program a pulse length of the read and write controlling signal of at least: minimal pulse length = NWAIT latency + 2 resynchronization cycles + 1 cycle
Figure 29-27. NWAIT Latency
MCK A[25:2]
NBS0, NBS1, A0,A1 WAIT STATE 4 NRD Minimal pulse length 3 2 1 0 0 0
NWAIT Internally synchronized NWAIT signal

NWAIT latency 2 resynchronization cycles
Read cycle EXNW_MODE = 10 or 11 READ_MODE = 1 (NRD_controlled) NRD_PULSE = 5

376
29.15 Slow Clock Mode

The HSMC is able to automatically apply a set of slow clock mode read/write waveforms when an internal signal driven by the Power Management Controller is asserted because MCK has been turned to a very slow clock rate (typically 32 kHz clock rate). In this mode, the user-programmed waveforms are ignored and the slow clock mode waveforms are applied. This mode is provided so as to avoid reprogramming the User Interface with appropriate waveforms at very slow clock rate. When activated, the slow mode is active on all chip selects.
29.15.1 Slow Clock Mode Waveforms

Figure 29-28 illustrates the read and write operations in slow clock mode. They are valid on all chip selects. Table 29-7 indicates the value of read and write parameters in slow clock mode.
Figure 29-28. Write/Read Cycles in Slow Clock Mode
MCK MCK
A[25:2] NBS0, NBS1, A0,A1
A[25:2]
NBS0, NBS1, A0,A1
NWE
1 1
NRD 1 1 NCS NRD_CYCLE = 2 SLOW CLOCK MODE READ
NCS NWE_CYCLE = 3 SLOW CLOCK MODE WRITE
Table 29-7.
Read and Write Timing Parameters in Slow Clock Mode Duration (cycles) 1 1 0 2 2 Write Parameters NWE_SETUP NWE_PULSE NCS_WR_SETUP NCS_WR_PULSE NWE_CYCLE Duration (cycles) 1 1 0 3 3
Read Parameters NRD_SETUP NRD_PULSE NCS_RD_SETUP NCS_RD_PULSE NRD_CYCLE

377
29.15.2 Switching from (to) Slow Clock Mode to (from) Normal Mode
When switching from slow clock mode to normal mode, the current slow clock mode transfer is completed at high clock rate, with the set of slow clock mode parameters. See Figure 29-29. The external device may not be fast enough to support such timings. Figure 29-30 illustrates the recommended procedure to properly switch from one mode to the other.
Figure 29-29. Clock Rate Transition occurs while the HSMC is performing a Write Operation
Slow Clock Mode internal signal from PMC MCK A[25:2] NBS0, NBS1, A0, A1 NWE
1 NCS
NWE_CYCLE = 3
SLOW CLOCK MODE WRITE SLOW CLOCK MODE WRITE
NWE_CYCLE = 7
NORMAL MODE WRITE
This write cycle finishes with the slow clock mode set of parameters after the clock rate transition
Slow clock mode transition is detected: Reload Configuration Wait State

378
Figure 29-30. Recommended Procedure to Switch from Slow Clock Mode to Normal Mode or from Normal Mode to Slow Clock Mode
Slow Clock Mode internal signal from PMC
MCK
A[25:2]
NBS0, NBS1, A0, A1
NWE 1 NCS 1 1 2 3 2
SLOW CLOCK MODE WRITE
IDLE STATE
Reload Configuration Wait State
NORMAL MODE WRITE

379
29.16 NAND Flash Controller Operations

29.16.1 NFC Overview
The NAND Flash Controller (NFC) handles all the command, address and data sequences of the NAND low level protocol. An SRAM is used as an internal read/write buffer when data is transferred from or to the NAND.
29.16.2 NFC Control Registers

NAND Flash Read and NAND Flash Program operations can be performed through the NFC Command Registers. In order to minimize CPU intervention and latency, commands are posted in a command buffer. This buffer provides zero wait state latency. The detailed description of the command encoding scheme is explained below. The NFC handles an automatic transfer between the external NAND Flash and the chip via the NFC SRAM. It is done via NFC Command Registers. The NFC Command Registers are very efficient to use. When writing to these registers:
z z
The address of the register (NFCADDR_CMD) is the command used The data of the register (NFCDATA_ADDT) is the address to be sent to the NAND Flash
So, in one single access the command is sent and immediately executed by the NFC. Two commands can even be programmed within a single access (CMD1, CMD2) depending on the VCMD2 value. The NFC can send up to 5 Address cycles. Figure 29-31 below shows a typical NAND Flash Page Read Command of a NAND Flash Memory and correspondence with NFC Address Command Register.
Figure 29-31. NFC/NAND Flash Access Example
00h Col. Add1 Col. Add2 Row Add1 Row Add2 Row Add3 Row Address 30h
Column Address
CMD1
ADD cycles (0 to 5)
CMD2
Depends on ACYCLE value
If VCMD2 = 1
For more details refer to NFC Address Command on page 382. Reading the NFC command register (to any address) will give the status of the NFC. Especially useful to know if the NFC is busy, for example.

380
29.16.2.1 Building NFC Address Command Example The base address is made of HOST_ADDR address. Page read operation example: // Build the Address Command (NFCADDR_CMD) AddressCommand = (HOST_ADDR | NFCCMD=1 | // NFC Command Enable NFCWR=0 |// NFC Read Data from NAND Flash DATAEN=1 | // NFC Data phase Enable. CSID=1 | // Chip Select ID = 1 ACYCLE= 5 | // Number of address cycle. VCMD2=1 | // CMD2 is sent after Address Cycles CMD2=0x30 | // CMD2 = 30h CMD1=0x0) // CMD1 = Read Command = 00h // Set the Address for Cycle 0 HSMC_ADDR = Col. Add1 // Write command with the Address Command built above *AddressCommand = (Col. Add2 |// ADDR_CYCLE1 Row Add1 | // ADDR_CYCLE2 Row Add2 |// ADDR_CYCLE3 Row Add3 )// ADDR_CYCLE4

381
29.16.2.2 NFC Address Command Name: Access: Reset:

31 23 CSID 15 14 13 CMD2 7 6 5 CMD1 4 3 2 1
NFCADDR_CMD Read-write 0x00000000

30 22 29 21 28 20 ACYCLE 12 27 19 26 NFCWR 18 VCMD2 10 25 DATAEN 17 CMD2 9 CMD1 0 8 24 CSID 16
11
CMD1: Command Register Value for Cycle 1

If NFCCMD is set, when a read or write access occurs, the NFC sends this command.

If NFCCMD and VCMD2 field are set to one, the NFC sends this command after CMD1.
VCMD2: Valid Cycle 2 Command

When set to true, the CMD2 field is issued after the address cycle.
ACYCLE: Number of Address required for the current command

When ACYCLE field is different from zero, ACYCLE Address cycles are performed after Command Cycle 1. The maximum number of cycles is 5.
CSID: Chip Select Identifier

Chip select used
DATAEN: NFC data phase enable

When set to true, the NFC will automatically read or write data after the command.
NFCWR: NFC Write Enable

0: The NFC reads data from the NAND Flash. 1: The NFC writes data into the NAND Flash.

382
29.16.2.3 NFC Data Address Name: Access: Reset:

31
NFCDATA_ADDT Write 0x00000000

30 29 28 27 ADDR_CYCLE4 20 19 ADDR_CYCLE3 12 11 ADDR_CYCLE2 4 3 ADDR_CYCLE1 26 25 24
23
22
21
18
17
16
15
14
13
10
ADDR_CYCLE1: NAND Flash Array Address Cycle 1

When less than 5 address cycles are used, ADDR_CYCLE1 is the first Byte written to the NAND Flash When 5 address cycles are used, ADDR_CYCLE1 is the second Byte written to NAND Flash

When less than 5 address cycles are used, ADDR_CYCLE2 is the second Byte written to the NAND Flash When 5 address cycles are used, ADDR_CYCLE2 is the third Byte written to the NAND Flash

When less than 5 address cycles are used, ADDR_CYCLE3 is the third Byte written to the NAND Flash When 5 address cycles are used, ADDR_CYCLE3 is the fourth Byte written to the NAND Flash

When less than 5 address cycles are used, ADDR_CYCLE4 is the fourth Byte written to the NAND Flash When 5 address cycles are used, ADDR_CYCLE4 is the fifth Byte written to the NAND Flash Note: If 5 address cycles are used, the first address cycle is ADDR_CYCLE0. Refer to HSMC_ADDR register.

383
29.16.2.4 NFC DATA Status Name: Access: Reset:

31 23 CSID 15 14 13 CMD2 7 6 5 CMD1 4 3 2 1
NFCDATA_Status Read 0x00000000

30 22 29 21 28 20 ACYCLE 12 27 NFCBUSY 19 26 NFCWR 18 VCMD2 10 25 DATAEN 17 CMD2 9 CMD1 0 8 24 CSID 16
11

When a Read or Write Access occurs, the Physical Memory Interface drives the IO bus with CMD1 field during the Command Latch cycle 1.

When VCMD2 field is set to true, the Physical Memory Interface drives the IO bus with CMD2 field during the Command Latch cycle 2.
VCMD2: Valid Cycle 2 Command

When set to true, the CMD2 field is issued after addressing cycle.
ACYCLE: Number of Address required for the current command

When ACYCLE field is different from zero, ACYCLE Address cycles are performed after Command Cycle 1.
CSID: Chip Select Identifier

Chip select used
DATAEN: NFC Data phase enable

When set to true, The NFC data phase is enabled.
NFCWR: NFC Write Enable

0: The NFC is in read mode. 1: The NFC is in write mode.
NFCBUSY: NFC Busy Status Flag

If set to true, it indicates that the NFC is busy.

384
29.16.3 NFC Initialization

Prior to any Command and Data Transfer, the HSMC User Interface must be configured to meet the device timing requirements.
z
Write enable Configuration
Use NWE_SETUP, NWE_PULSE and NWE_CYCLE to define the write enable waveform according to the device datasheet. Use TADL field in the HSMC_TIMINGS register to configure the timing between the last address latch cycle and the first rising edge of WEN for data input.
Figure 29-32. Write Enable Timing Configuration
mck
wen
tWEN_SETUP
tWEN_PULSE tWEN_CYCLES
tWEN_HOLD
Figure 29-33. Write Enable Timing for NAND Flash Device Data Input Mode.
mck ale
wen
t ADL
z
Read Enable Configuration
Use NRD_SETUP, NRD_PULSE and NRD_CYCLE to define the read enable waveform according to the device datasheet. Use TAR field in the HSMC_TIMINGS register to configure the timings between the address latch enable falling edge to read the enable falling edge. Use TCLR field in the HSMC_TIMINGS register to configure the timings between the command latch enable falling edge to read the enable falling edge.

385
Figure 29-34. Read Enable Timing Configuration Working with NAND Flash Device
mck
cen
ale
cle
ren
tCLR tAR
z
tREN_SETUP
tREN_PULSE tREN_CYCLE
tREH
Ready/Busy Signal Timing configuration working with a NAND Flash device
Use TWB field in HSMC_TIMINGS register to configure the maximum elapsed time between the rising edge of the wen signal and the falling edge of the rbn signal. Use TRR field in the HSMC_TIMINGS register to program the number of clock cycles between the rising edge of the rbn signal and the falling edge of the ren signal.
Figure 29-35. Ready/Busy Timing Configuration
mck
rbn
ren
wen
tWB
busy
tRR

386
29.16.3.1 NAND Flash Controller Timing Engine When the NFC Command register is written, the NFC issues a NAND Flash Command and optionally performs a data transfer between the NFC SRAM and the NAND Flash device. The NAND Flash Controller Timing Engine guarantees valid NAND Flash timings, depending on the set of parameters decoded from the address bus. These timings are defined in the HSMC_TIMINGS register. For information on the timing used depending on the command, see Figure 29-36:
Figure 29-36. NAND Flash Controller Timing Engine
Timing Check Engine
NFCEN=1 NFCWR =1 TADL =1 Wait TADL NFCEN=1 NFCWR=0 TWB != 0 Wait TWB NFCEN=0 VCMD2=1 TCLR != 0 Wait TCLR !NFCEN=1 VCMD2=0 ACYCLE!=0 NFCWR=1 TADL != 0 Wait TADL !NFCEN=1 VCMD2=0 ACYCLE!=0 NFCWR=0 TAR != 0 Wait TAR !NFCEN=1 VCMD2=0 ACYCLE!=0 TCLR != 0 Wait TCLR
See the NFC Address Command register description and the Timings Register.
29.16.4 NFC SRAM

29.16.4.1 NFC SRAM Mapping If the NFC is used to read and write Data from and to the NAND Flash, the configuration depends on the page size. See Table 29-8 for detailed mapping. The NFC SRAM size is 8 KBytes. The NFC can handle the NAND Flash with a page size of 8 KBytes or of a lower size (such as 2 KBytes for example). In case of a 2-KByte or lower page size, the NFC SRAM can be split into two banks. The HSMC_BANK field enables to select the bank used. Note that a ping-pong mode (write or read to a bank while the NFC writes or reads to another bank) is accessible with the NFC (using 2 different banks). If the NFC is not used, the NFC SRAM can be used for a general purpose by the application.
Table 29-8. NFC SRAM Bank Mapping Offset 0x00000000-0x000011FF 0x00001200-0x000023FF Use Bank 0 Bank 1 Access Read-write Read-write

387
29.16.4.2 NFC SRAM Access Prioritization Algorithm When the NAND Flash Controller (NFC) is reading from or writing to an NFC SRAM bank, the other bank is available. If an NFC SRAM access occurs when the NFC performs a read or write operation in the same bank, then the access is discarded. The write operation is not performed. The read operation returns undefined data. If this situation is encountered, the AWB status flag located in the NFC status Register is raised and indicates that a shared resource access violation has occurred.
29.16.5 NAND Flash Operations

This section describes the software operations needed to issue commands to the NAND Flash device and to perform data transfers using the NFC. 29.16.5.1 Page Read
Figure 29-37. Page Read Flow Chart
Configure the device writing in the User Interface
Using NFC
Write the NFC Command registers
Enable XFRDONE interrupt (SMC_IER)
Wait for Interrupt
Copy data from NFC SRAM to application memory (via DMA for example)
Check Error Correcting Codes
Note that, instead of using the interrupt, one can poll the NFCBUSY Flag. For more information on the NFC Control Register, see Section 29.16.2.2 NFC Address Command.

388
29.16.5.2 Program Page

Figure 29-38. Program Page Flow Chart
Configure the device writing in the User interface
Write data in the NFC SRAM (CPU or DMA) Enable XFRDONE
Write the Command Register through the AHB interface
Wait for interrupt
Write ECC
Wait for Ready/Busy interrupt
Writing the ECC cannot be done using the NFC; it needs to be done manually. Note that, instead of using the interrupt, one can poll the NFCBUSY Flag. For more information on the NFC Control Register, see Section 29.16.2.2 NFC Address Command.

389
29.17 PMECC Controller Functional Description

The Programmable Multibit Error Correcting Code (PMECC) controller is a programmable binary BCH (Bose, Chaudhuri and Hocquenghem) encoder/decoder. This controller can be used to generate redundancy information for both SLC and MLC NAND devices. It supports redundancy for correction of 2, 4, 8, 12 or 24 errors per sector of data. The sector size is programmable and can be set to 512 Bytes or 1024 Bytes. The PMECC module generates redundancy at encoding time, when a NAND write page operation is performed. The redundancy is appended to the page and written in the spare area. This operation is performed by the processor. It moves the content of the PMECCX registers into the NAND flash memory. The number of registers depends on the selected error correction capability (see Table 29-9 on page 392). This operation shall be executed for each sector. At Decoding time, the PMECC module generates the remainders of the received codeword by the minimal polynomials. When all remainders for a given sector are set to zero, no error occurred. When the remainders are different from zero, the codeword is corrupted and further processing is required. The PMECC module generates an interrupt indicating that an error occurred. The Processor must read the PMECCISR register. This register indicates which sector is corrupted. The processor must execute the following decoding steps to find the error location within a sector: 1. 2. 3. Syndrome computation. Finding the error location polynomial. Finding the roots of the error location polynomial.
All decoding steps involve finite field computation. It means that a library of finite field arithmetic must be available to perform addition, multiplication and inversion. These arithmetic operations can be performed through the use of a memory mapped look-up table, or direct software implementation. The software implementation presented is based on look-up tables. Two tables named gf_log and gf_antilog are used. If alpha is the primitive element of the field, then a power of alpha is in the field. Assuming that beta = alpha ^ index, then beta belongs to the field, and gf_log(beta) = gf_log(alpha ^ index) = index. The gf_antilog table provides exponent inverse of the element; if beta = alpha ^ index, then gf_antilog(index) = beta. The first step consists in the syndrome computation. The PMECC module computes the remainders and the software must substitute the power of the primitive element. The procedure implementation is given in Section 29.18.1 Remainder Substitution Procedure on page 397. The second step is the most software intensive. It is the Berlekamps iterative algorithm for finding the error-location polynomial. The procedure implementation is given in Section 29.18.2 Find the Error Location Polynomial Sigma(x) on page 398. The Last step is finding the root of the error location polynomial. This step can be very software intensive. Indeed there is no straightforward method of finding the roots, except evaluating each element of the field in the error location polynomial. However, a hardware accelerator can be used to find the roots of the polynomial. The PMERRLOC module provides this kind of hardware acceleration.

390
Figure 29-39. Software Hardware Multibit Error Correction Dataflow
NAND Flash PROGRAM PAGE Operation
NAND Flash READ PAGE Operation
Software
Hardware Accelerator
Software
Hardware Accelerator
Configure PMECC : error correction capability sector size/page size NAND write field set to true spare area desired layout
Configure PMECC : error correction capability sector size/page size NAND write field set to false spare area desired layout
Move the NAND Page to external Memory whether using DMA or Processor
PMECC computes redundancy as the data is written into external memory
Move the NAND Page from external Memory whether using DMA or Processor
PMECC computes polynomial remainders as the data is read from external memory
Copy redundancy from PMECC user interface to user-defined spare area using DMA or Processor. If a sector is corrupted use the substitute() function to determine the syndromes.
PMECC modules indicate if at least one error is detected.
When the table of syndromes is completed, use the get_sigma() function to get the error location polynomial.
Find the error positions finding the roots of the error location polynomial And correct the bits.
This step can be hardware-assisted using the PMERRLOC module.

391
29.17.1 MLC/SLC Write Page Operation Using PMECC

When an MLC write page operation is performed, the PMECC controller is configured with the NANDWR field of the PMECCFG register set to one. When the NAND spare area contains file system information and redundancy (PMECCx), the spare area is error protected, then the SPAREEN bit of the PMECCFG register is set to one. When the NAND spare area contains only redundancy information, the SPAREEN bit is set to zero. When the write page operation is terminated, the user writes the redundancy in the NAND spare area. This operation can be done with DMA assistance.
Table 29-9. Relevant Redundancy Registers BCH_ERR Field 0 1 2 Sector Size Set to 512 Bytes PMECC0 PMECC0, PMECC1 PMECC0, PMECC1, PMECC2, PMECC3 PMECC0, PMECC1, PMECC2, PMECC3, PMECC4, PMECC5, PMECC6 PMECC0, PMECC1, PMECC2, PMECC3, PMECC4, PMECC5, PMECC6, PMECC7, PMECC8, PMECC9 Sector Size Set to 1024 Bytes PMECC0 PMECC0, PMECC1 PMECC0, PMECC1, PMECC2, PMECC3 PMECC0, PMECC1, PMECC2, PMECC3, PMECC4, PMECC5, PMECC6 PMECC0, PMECC1, PMECC2, PMECC3, PMECC4, PMECC5, PMECC6, PMECC7, PMECC8, PMECC9, PMECC10
Table 29-10. Number of Relevant ECC Bytes per Sector, Copied from LSByte to MSByte BCH_ERR Field 0 1 2 3 4 Sector Size Set to 512 Bytes 4 Bytes 7 Bytes 13 Bytes 20 Bytes 39 Bytes Sector Size Set to 1024 Bytes 4 Bytes 7 Bytes 14 Bytes 21 Bytes 42 Bytes

392
29.17.1.1 SLC/MLC Write Operation with Spare Enable Bit Set When the SPAREEN field of the PMECCFG register is set to one, the spare area of the page is encoded with the stream of data of the last sector of the page. This mode is entered by writing 1 in the DATA field of the PMECCTRL register. When the encoding process is over, the redundancy shall be written to the spare area in user mode. The USER field of the PMECCTRL register must be set to one.
Figure 29-40. NAND Write Operation with Spare Encoding
pagesize = n * sectorsize
sparesize Sector 3 Spare
Sector 0 512 or 1024 bytes
Sector 1
Sector 2
ecc_area start_addr ECC computation enable signal end_addr
29.17.1.2 SLC/MLC Write Operation with Spare Disable When the SPAREEN field of PMECCFG is set to zero, the spare area is not encoded with the stream of data. This mode is entered by writing 1 to the DATA field of the PMECCTRL register.
Figure 29-41. NAND Write Operation
pagesize = n * sectorsize
Sector 0 512 or 1024 bytes ECC computation enable signal
Sector 1
Sector 2
Sector 3

393
29.17.2 MLC/SLC Read Page Operation Using PMECC
Table 29-11. Relevant Remainder Registers BCH_ERR Field 0 1 2 Sector Size Set to 512 Bytes PMECCREM0 PMECCREM0, PMECCREM1 PMECCREM0, PMECCREM1, PMECCREM2, PMECCREM3, PMECCREM0, PMECCREM1, PMECCREM2, PMECCREM3, PMECCREM4, PMECCREM5, PMECCREM6, PMECCREM7 PMECCREM0, PMECCREM1, PMECCREM2, PMECCREM3, PMECCREM4, PMECCREM5, PMECCREM6, PMECCREM7, PMECCREM8, PMECCREM9, PMECCREM10, PMECCREM11 Sector Size Set to 1024 Bytes PMECCREM0 PMECCREM0, PMECCREM1 PMECCREM0, PMECCREM1, PMECCREM2, PMECCREM3 PMECCREM0, PMECCREM1, PMECCREM2, PMECCREM3, PMECCREM4, PMECCREM5, PMECCREM6, PMECCREM7 PMECCREM0, PMECCREM1, PMECCREM2, PMECCREM3, PMECCREM4, PMECCREM5, PMECCREM6, PMECCREM7, PMECCREM8, PMECCREM9, PMECCREM10, PMECCREM11
29.17.2.1 MLC/SLC Read Operation with Spare Decoding When the spare area is protected, it contains valid data. As the redundancy may be included in the middle of the information stream, the user shall program the start address and the end address of the ECC area. The controller will automatically skip the ECC area. This mode is entered writing 1 in the DATA field of the PMECCTRL register. When the page has been fully retrieved from the NAND, the ECC area shall be read using the user mode, writing 1 to the USER field of the PMECCTRL register.
Figure 29-42. Read Operation with Spare Decoding
Read NAND operation with SPAREEN set to One and AUTO set to Zero pagesize = n * sectorsize sparesize Sector 3 Spare
Sector 1
Sector 2
ecc_area start_addr Remainder computation enable signal end_addr

394
29.17.2.2 MLC/SLC Read Operation If the spare area is not protected with the error correcting code, the redundancy area is retrieved directly. This mode is entered writing 1 in the DATA field of the PMECCTRL register. When AUTO field is set to one, the ECC is retrieved automatically; otherwise, the ECC must be read using the user mode.
Figure 29-43. Read Operation
Read NAND operation with SPAREEN set to Zero and AUTO set to One pagesize = n * sectorsize sparesize Sector 3 Spare
Sector 1
Sector 2
ecc_area start_addr
ECC_BLK0 ECC_BLK1 ECC_BLK2
end_addr
ECC_BLK3
Remainder computation enable signal
29.17.2.3 MLC/SLC User Read ECC Area This mode allows a manual retrieve of the ECC. It is entered writing 1 in the USER field of the PMECCTRL register.
Figure 29-44. Read User Mode
ecc_area_size ECC
ecc_area addr = 0 end_addr
Remainder computation enable signal

395
29.17.2.4 MLC Controller Working with NAND Flash Controller

Table 29-12. MLC Controller Configuration when the Host Controller is Used Transfer Type Program Page main area is protected, spare is not protected, spare is written manually Program Page main area is protected, spare is protected, spare is written by NFC Read Page main area is protected, spare is not protected, spare is not retrieved by NFC Read Page main area is protected, spare is not protected, spare is retrieved by NFC Read Page main area is protected, spare is protected, spare is retrieved by NFC NFC RSPARE 0 NFC WSPARE 0 PMECC SPAREEN 0 PMECC AUTO 0 PMECC User Mode Not used
Not used
Used
Not used
Used

396
29.18 Software Implementation

29.18.1 Remainder Substitution Procedure
The substitute function evaluates the remainder polynomial, with different values of the field primitive element. The addition arithmetic operation is performed with the exclusive OR. The multiplication arithmetic operation is performed through the gf_log and gf_antilog look-up tables. The REM2NP1 and REMN2NP3 fields of the PMECCREMN registers contain only odd remainders. Each bit indicates whether the coefficient of the remainder polynomial is set to zero or not. NB_ERROR_MAX defines the maximum value of the error correcting capability. NB_ERROR defines the error correcting capability selected at encoding/decoding time. NB_FIELD_ELEMENTS defines the number of elements in the field. si[] is a table that holds the current syndrome value. An element of that table belongs to the field. This is also a shared variable for the next step of the decoding operation. oo[] is a table that contains the degree of the remainders. int { int int for { substitute() i; j; (i = 1; i < 2 * NB_ERROR_MAX; i++)
si[i] = 0; } for (i = 1; i < 2*NB_ERROR; i++) { for (j = 0; j < oo[i]; j++) { if (REM2NPX[i][j]) { si[i] = gf_antilog[(i * j)%NB_FIELD_ELEMENTS] ^ si[i]; } } } return 0; }

397
29.18.2 Find the Error Location Polynomial Sigma(x)

The sample code below gives a Berlekamp iterative procedure for finding the value of the error location polynomial. The input of the procedure is the si[] table defined in the remainder substitution procedure. The output of the procedure is the error location polynomial named smu (sigma mu). The polynomial coefficients belong to the field. The smu[NB_ERROR+1][] is a table that contains all these coefficients. NB_ERROR_MAX defines the maximum value of the error correcting capability. NB_ERROR defines the error correcting capability selected at encoding/decoding time. NB_FIELD_ELEMENTS defines the number of elements in the field. int get_sigma() { int i; int j; int k; /* mu */ int mu[NB_ERROR_MAX+2]; /* sigma ro */ int sro[2*NB_ERROR_MAX+1]; /* discrepancy */ int dmu[NB_ERROR_MAX+2]; /* delta order */ int delta[NB_ERROR_MAX+2]; /* index of largest delta */ int ro; int largest; int diff; /* */ /* First Row */ /* */ /* Mu */ mu[0] = -1; /* Actually -1/2 */ /* Sigma(x) set to 1 */ for (i = 0; i < (2*NB_ERROR_MAX+1); i++) smu[0][i] = 0; smu[0][0] = 1; /* discrepancy set to 1 */ dmu[0] = 1; /* polynom order set to 0 */ lmu[0] = 0; /* delta set to -1 */ delta[0] = (mu[0] * 2 - lmu[0]) >> 1; /* */ /* Second Row */ /* */ /* Mu */ mu[1] = 0; /* Sigma(x) set to 1 */ for (i = 0; i < (2*NB_ERROR_MAX+1); i++) smu[1][i] = 0; smu[1][0] = 1; /* discrepancy set to Syndrome 1 */ dmu[1] = si[1]; /* polynom order set to 0 */ lmu[1] = 0;

398
/* delta set to 0 */ delta[1] = (mu[1] * 2 - lmu[1]) >> 1; for (i=1; i <= NB_ERROR; i++) { mu[i+1] = i << 1; /*************************************************/ /* */ /* */ /* Compute Sigma (Mu+1) */ /* And L(mu) */ /* check if discrepancy is set to 0 */ if (dmu[i] == 0) { /* copy polynom */ for (j=0; j<2*NB_ERROR_MAX+1; j++) { smu[i+1][j] = smu[i][j]; } /* copy previous polynom order to the next */ lmu[i+1] = lmu[i]; } else { ro = 0; largest = -1; /* find largest delta with dmu != 0 */ for (j=0; j<i; j++) { if (dmu[j]) { if (delta[j] > largest) { largest = delta[j]; ro = j; } } } /* initialize signal ro */ for (k = 0; k < 2*NB_ERROR_MAX+1; k ++) { sro[k] = 0; } /* compute difference */ diff = (mu[i] - mu[ro]); /* compute X ^ (2(mu-ro)) */ for (k = 0; k < (2*NB_ERROR_MAX+1); k ++) { sro[k+diff] = smu[ro][k]; } /* multiply by dmu * dmu[ro]^-1 */ for (k = 0; k < 2*NB_ERROR_MAX+1; k ++) { /* dmu[ro] is not equal to zero by definition */ /* check that operand are different from 0 */ if (sro[k] && dmu[i])

399
{ /* galois inverse */ sro[k] = gf_antilog[(gf_log[dmu[i]] + (NB_FIELD_ELEMENTSgf_log[dmu[ro]]) + gf_log[sro[k]]) % NB_FIELD_ELEMENTS]; } } /* multiply by dmu * dmu[ro]^-1 */ for (k = 0; k < 2*NB_ERROR_MAX+1; k++) { smu[i+1][k] = smu[i][k] ^ sro[k]; if (smu[i+1][k]) { /* find the order of the polynom */ lmu[i+1] = k << 1; } } } /* */ /* */ /* End Compute Sigma (Mu+1) */ /* And L(mu) */ /*************************************************/ /* In either case compute delta */ delta[i+1] = (mu[i+1] * 2 - lmu[i+1]) >> 1; /* In either case compute the discrepancy */ for (k = 0 ; k <= (lmu[i+1]>>1); k++) { if (k == 0) dmu[i+1] = si[2*(i-1)+3]; /* check if one operand of the multiplier is null, its index is -1 */ else if (smu[i+1][k] && si[2*(i-1)+3-k]) dmu[i+1] = gf_antilog[(gf_log[smu[i+1][k]] + gf_log[si[2*(i-1)+3-k]])%nn] ^ dmu[i+1]; } } return 0; }

400
29.18.3 Find the Error Position

The output of the get_sigma() procedure is a polynomial stored in the smu[NB_ERROR+1][] table. The error positions are the roots of that polynomial. The degree of that polynomial is a very important information, as it gives the number of errors. PMERRLOC module provides hardware accelerator for that step. 29.18.3.1 Error Location The PMECC Error Location controller provides hardware acceleration for determining roots of polynomials over two finite fields: GF(2^13) and GF(2^14). It integrates 32 fully programmable coefficients. These coefficients belong to GF(2^13) or GF(2^14). The coefficient programmed in the PMERRLOC{i} is the coefficient of X ^ i in the polynomial. The search operation is started as soon as a write access is detected in the ELEN register and can be disabled writing to the ELDIS register. The ENINIT field of the ELEN register shall be initialized with the number of galois field elements to test. The set of the roots can be limited to a valid range.
Table 29-13. ENINIT Field Value for a Sector Size of 512 Bytes Error Correcting Capability 2 4 8 12 24 ENINIT Value 4122 4148 4200 4252 4408
Table 29-14. ENINIT Field Value for a Sector Size of 1024 Bytes Error Correcting Capability 2 4 8 12 24 ENINIT Value 8220 8248 8304 8360 8528
When the PMECC engine is searching for roots, the BUSY field of the ELSR register remains asserted. An interrupt is asserted at the end of the computation, and the DONE bit of the ELSIR register is set. The ERR_CNT field of the ELISR register indicates the number of errors. The error position can be read in the PMERRLOCX registers.

401
29.19 Static Memory Controller (HSMC) User Interface

The HSMC is programmed using the fields listed in Table 29-15. For each chip select, a set of 4 registers is used to program the parameters of the external device. In Table 29-15, CS_number denotes the chip select number. 16 Bytes per chip select are required.
Table 29-15. Register Mapping Offset 0x000 0x004 0x008 0x00C 0x010 0x014 0x018 0x01C 0x020 - 0x06C 0x70 0x74 0x78 0x7C 0x80 0x84 0x88 0x8C 0x90 0x94 0x98 0x9C-AC 0x0B0+sec_num*(0x40)+0x00 0x0B0+sec_num*(0x40)+0x04 0x0B0+sec_num*(0x40)+0x08 0x0B0+sec_num*(0x40)+0x0C 0x0B0+sec_num*(0x40)+0x10 0x0B0+sec_num*(0x40)+0x14 0x0B0+sec_num*(0x40)+0x18 0x0B0+sec_num*(0x40)+0x1C 0x0B0+sec_num*(0x40)+0x20 0x0B0+sec_num*(0x40)+0x24 0x0B0+sec_num*(0x40)+0x28 0x2B0+sec_num*(0x40)+0x00 Register HSMC NFC Configuration Register HSMC NFC Control Register HSMC NFC Status Register HSMC NFC Interrupt Enable Register HSMC NFC Interrupt Disable Register HSMC NFC Interrupt Mask Register HSMC NFC Address Cycle Zero Register HSMC Bank Address Register Reserved PMECC Configuration Register PMECC Spare Area Size Register PMECC Start Address Register PMECC End Address Register Reserved PMECC Control Register PMECC Status Register PMECC Interrupt Enable register PMECC Interrupt Disable Register PMECC Interrupt Mask Register PMECC Interrupt Status Register Reserved PMECC Redundancy 0 Register PMECC Redundancy 1 Register PMECC Redundancy 2 Register PMECC Redundancy 3 Register PMECC Redundancy 4 Register PMECC Redundancy 5 Register PMECC Redundancy 6 Register PMECC Redundancy 7 Register PMECC Redundancy 8 Register PMECC Redundancy 9 Register PMECC Redundancy 10 Register PMECC Remainder 0 Register HSMC_PMECCFG HSMC_PMECCSAREA HSMC_PMECCSADDR HSMC_PMECCEADDR HSMC_PMECCTRL HSMC_PMECCSR HSMC_PMECCIER HSMC_PMECCIDR HSMC_PMECCIMR HSMC_PMECCISR HSMC_PMECC0 HSMC_PMECC1 HSMC_PMECC2 HSMC_PMECC3 HSMC_PMECC4 HSMC_PMECC5 HSMC_PMECC6 HSMC_PMECC7 HSMC_PMECC8 HSMC_PMECC9 HSMC_PMECC10 HSMC_REM0 Read-write Read-write Read-write Read-write Write Only Read-only Write Only Write Only Read-only Read-only Read-only Read-only Read-only Read-only Read-only Read-only Read-only Read-only Read-only Read-only Read-only Read-only 0x0 0x0 0x0 0x0 0x0 0x0 0x0 0x0 0x0 0x0 0x0 0x0 0x0 0x0 0x0 0x0 0x0 0x0 0x0 0x0 0x0 Name HSMC_CFG HSMC_CTRL HSMC_SR HSMC_IER HSMC_IDR HSMC_IMR HSMC_ADDR HSMC_BANK Access Read-write Write-only Read-only Write-only Write-only Read-only Read-write Read-write Reset 0x0 0x0 0x0 0x0 0x0 0x0 0x0 0x0

402
Table 29-15. Register Mapping (Continued) Offset 0x2B0+sec_num*(0x40)+0x04 0x2B0+sec_num*(0x40)+0x08 0x2B0+sec_num*(0x40)+0x0C 0x2B0+sec_num*(0x40)+0x10 0x2B0+sec_num*(0x40)+0x14 0x2B0+sec_num*(0x40)+0x18 0x2B0+sec_num*(0x40)+0x1C 0x2B0+sec_num*(0x40)+0x20 0x2B0+sec_num*(0x40)+0x24 0x2B0+sec_num*(0x40)+0x28 0x2B0+sec_num*(0x40)+0x2C 0x4A0-0x4FC 0x500 0x504 0x508 0x50C 0x510 0x514 0x518 0x51C 0x520 0x524-0x52C 0x528 ... 0x588 0x58C ... 0x5E8 0x5EC-0x5FC 0x14*CS_number+0x600 0x14*CS_number+0x604 0x14*CS_number+0x608 0x14*CS_number+0x60C 0x14*CS_number+0x610 Register PMECC Remainder 1 Register PMECC Remainder 2 Register PMECC Remainder 3 Register PMECC Remainder 4 Register PMECC Remainder 5 Register PMECC Remainder 6 Register PMECC Remainder 7 Register PMECC Remainder 8 Register PMECC Remainder 9 Register PMECC Remainder 10 Register PMECC Remainder 11 Register Reserved PMECC Error Location Configuration Register PMECC Error Location Primitive Register PMECC Error Location Enable Register PMECC Error Location Disable Register PMECC Error Location Status Register PMECC Error Location Interrupt Enable register PMECC Error Location Interrupt Disable Register PMECC Error Location Interrupt Mask Register PMECC Error Location Interrupt Status Register Reserved PMECC Error Location SIGMA 0 Register ... PMECC Error Location SIGMA 24 Register PMECC Error Location 0 Register ... PMECC Error Location 23 Register Reserved HSMC Setup Register HSMC Pulse Register HSMC Cycle Register HSMC Timings Register HSMC Mode Register Name HSMC_REM1 HSMC_REM2 HSMC_REM3 HSMC_REM4 HSMC_REM5 HSMC_REM6 HSMC_REM7 HSMC_REM8 HSMC_REM9 HSMC_REM10 HSMC_REM11 HSMC_ELCFG HSMC_ELPRIM HSMC_ELEN HSMC_ELDIS HSMC_ELSR HSMC_ELIER HSMC_ELIDR HSMC_ELIMR HSMC_ELISR HSMC_SIGMA0 ... HSMC_SIGMA24 HSMC_ERRLOC0 ... HSMC_ERRLOC23 HSMC_SETUP HSMC_PULSE HSMC_CYCLE HSMC_TIMINGS HSMC_MODE Access Read-only Read-only Read-only Read-only Read-only Read-only Read-only Read-only Read-only Read-only Read-only Read-write Read-only Write Write Read Write Write Read Read Read-write ... Read-write Read-only ... Read-only Read-write Read-write Read-write Read-write Read-write Reset 0x0 0x0 0x0 0x0 0x0 0x0 0x0 0x0 0x0 0x0 0x0 0x0 0x401A 0x0 0x0 0x0 0x0 0x0 0x0 0x0 0x1 ... 0x0 0x0 ... 0x0

403
Table 29-15. Register Mapping (Continued) Offset 0x6A0 0x6A4 0x6A8 0x6AC-0x6E0 0x6E4 0x6E8 0x6FC Register HSMC OCMS Register HSMC OCMS KEY1 Register HSMC OCMS KEY2 Register Reserved HSMC Write Protection Control Register HSMC Write Protection Status Register Reserved Name HSMC_OCMS HSMC_KEY1 HSMC_KEY2 HSMC_WPCR HSMC_WPSR Access Read-write Write-only Write-only Write-only Read-only Reset 0x0 0x0 0x0 0x0 0x0

404
29.19.1 HSMC NFC Configuration Register

31 23 15 7
HSMC_CFG 0xFFFFC000 Read-write 0x00000000

30 29 28 27 NFCSPARESIZE 19 26 25 24
22
21 DTOMUL 13 RBEDGE 5
20
18 DTOCYC
17
16
14 6
12 EDGECTRL 4
11 3
10 2
9 RSPARE 1 PAGESIZE
8 WSPARE 0
PAGESIZE
This field defines the page size of the NAND Flash device.
Value 0 1 2 3 4 PS512 PS1024 PS2048 PS4096 PS8192 Name Main area 512 Bytes Main area 1024 Bytes Main area 2048 Bytes Main area 4096 Bytes Main area 8192 Bytes Description
WSPARE: Write Spare Area

0: The NFC skips the spare area in write mode. 1: The NFC writes both main area and spare area in write mode.
RSPARE: Read Spare Area

0: The NFC skips the spare area in read mode. 1: The NFC reads both main area and spare area in read mode.
EDGECTRL: Rising/Falling Edge Detection Control

0: Rising edge is detected. 1: Falling edge is detected.
RBEDGE: Ready/Busy Signal Edge Detection

0: When set to zero, RB_EDGE fields indicate the level of the Ready/Busy lines. 1: When set to one, RB_EDGE fields indicate only transition on Ready/Busy lines.
DTOCYC: Data Timeout Cycle Number

405
DTOMUL: Data Timeout Multiplier

These fields determine the maximum number of Master Clock cycles that the HSMC waits until the detection of a rising edge on Ready/Busy signal. Data Timeout Multiplier is defined by DTOMUL as shown in the following table:
Value 0 1 2 3 4 5 6 7 X1 X16 X128 X256 X1024 X4096 X65536 X1048576 Name Description DTOCYC DTOCYC x 16 DTOCYC x 128 DTOCYC x 256 DTOCYC x 1024 DTOCYC x 4096 DTOCYC x 65536 DTOCYC x 1048576
If the data timeout set by DTOCYC and DTOMUL has been exceeded, the Data Timeout Error flag (DTOE) in the NFC Status Register (NFC_SR) raises.
NFCSPARESIZE: NAND Flash Spare Area Size Retrieved by the Host Controller
The spare size is set to (NFCSPARESIZE+1) * 4 Bytes. The spare area is only retrieved when RSPARE or WSPARE is activated.

406
29.19.2 HSMC NFC Control Register

31 23 15 7
HSMC_CTRL 0xFFFFC004 Write-only 0x00000000

30 22 14 6 29 21 13 5 28 20 12 4 27 19 11 3 26 18 10 2 25 17 9 1 NFCDIS 24 16 8 0 NFCEN
NFCEN: NAND Flash Controller Enable

0: No effect. 1: Enable the NAND Flash controller.
NFCDIS: NAND Flash Controller Disable

0: No effect 1: Disable the NAND Flash controller.

407
29.19.3 HSMC NFC Status Register

31 23 NFCASE 15 7
HSMC_SR 0xFFFFC008 Read-only 0x00000000

30 22 AWB 14 29 21 UNDEF 13 NFCSID 5 RB_FALL 28 20 DTOE 12 27 19 11 NFCWR 3 26 18 10 2 25 17 CMDDONE 9 1 24 RB_EDGE0 16 XFRDONE 8 NFCBUSY 0 SMCSTS
4 RB_RISE
SMCSTS: NAND Flash Controller Status (This field cannot be reset)

0: NAND Flash Controller is disabled. 1: NAND Flash Controller is enabled.
RB_RISE: Selected Ready Busy Rising Edge Detected

When set to one, this flag indicates that a rising edge on the Ready/Busy Line has been detected. This flag is reset after a status read operation. The Ready/Busy line selected is the decoding of the set NFCCSID, RBNSEL fields.
RB_FALL: Selected Ready Busy Falling Edge Detected

When set to one, this flag indicates that a falling edge on the Ready/Busy Line has been detected. This flag is reset after a status read operation. The Ready/Busy line is selected through the decoding of the set NFCSID, RBNSEL fields.
NFCBUSY: NFC Busy (This field cannot be reset)

When set to one, this flag indicates that the Controller is activated and accesses the memory device.
NFCWR: NFC Write/Read Operation (this field cannot be reset)

When a command is issued, this field indicates the current Read or Write Operation.
NFCSID: NFC Chip Select ID (This field cannot be reset)

When a command is issued, this field indicates the value of the targeted chip select.
XFRDONE: NFC Data Transfer Terminated

When set to one, this flag indicates that the NFC has terminated the Data Transfer. This flag is reset after a status read operation.
CMDDONE: Command Done

When set to one, this flag indicates that the NFC has terminated the Command. This flag is reset after a status read operation.
DTOE: Data Timeout Error

When set to one, this flag indicates that the Data timeout set be by DTOMUL and DTOCYC has been exceeded. This flag is reset after a status read operation.

408
UNDEF: Undefined Area Error

When set to one, this flag indicates that the processor performed an access in an undefined memory area. This flag is reset after a status read operation.
AWB: Accessing While Busy

If set to one, this flag indicates that an AHB master has performed an access during the busy phase. This flag is reset after a status read operation.
NFCASE: NFC Access Size Error

If set to one, this flag indicates that an illegal access has been detected in the NFC Memory Area. Only Word Access is allowed within the NFC memory area. This flag is reset after a status read operation.
RB_EDGEx: Ready/Busy Line x Edge Detected

If set to one, this flag indicates that an edge has been detected on the Ready/Busy Line x. Depending on the EDGE CTRL field located in the HSMC_CFG register, only rising or falling edge is detected. This flag is reset after a status read operation.

409
29.19.4 HSMC NFC Interrupt Enable Register

31 23 NFCASE 15 7
HSMC_IER 0xFFFFC00C Write-only 0x00000000

30 22 AWB 14 6 29 21 UNDEF 13 5 RB_FALL 28 20 DTOE 12 4 RB_RISE 27 19 11 3 26 18 10 2 25 17 CMDDONE 9 1 24 RB_EDGE0 16 XFRDONE 8 0
RB_RISE: Ready Busy Rising Edge Detection Interrupt Enable

0: No effect. 1: Interrupt source is enabled.
RB_FALL: Ready Busy Falling Edge Detection Interrupt Enable

XFRDONE: Transfer Done Interrupt Enable

CMDDONE: Command Done Interrupt Enable

DTOE: Data Timeout Error Interrupt Enable

UNDEF: Undefined Area Access Interrupt Enable

AWB: Accessing While Busy Interrupt Enable


410
NFCASE: NFC Access Size Error Interrupt Enable

RB_EDGEx: Ready/Busy Line x Interrupt Enable


411
29.19.5 HSMC NFC Interrupt Disable Register

31 23 NFCASE 15 7
HSMC_IDR 0xFFFFC010 Write-only 0x00000000

RB_RISE: Ready Busy Rising Edge Detection Interrupt Disable

0: No effect. 1: Interrupt source is disabled.
RB_FALL: Ready Busy Falling Edge Detection Interrupt Disable

XFRDONE: Transfer Done Interrupt Disable

CMDDONE: Command Done Interrupt Disable

DTOE: Data Timeout Error Interrupt Disable

UNDEF: Undefined Area Access Interrupt Disable

AWB: Accessing While Busy Interrupt Disable


412
NFCASE: NFC Access Size Error Interrupt Disable

RB_EDGEx: Ready/Busy Line x Interrupt Disable


413
29.19.6 HSMC NFC Interrupt Mask Register

31 23 NFCASE 15 7
HSMC_IMR 0xFFFFC014 Read-only 0x00000000

RB_RISE: Ready Busy Rising Edge Detection Interrupt Mask

0: Interrupt source is disabled. 1: Interrupt source is enabled.
RB_FALL: Ready Busy Falling Edge Detection Interrupt Mask

XFRDONE: Transfer Done Interrupt Mask

CMDDONE: Command Done Interrupt Mask

DTOE: Data Timeout Error Interrupt Mask

UNDEF: Undefined Area Access Interrupt Mask5

AWB: Accessing While Busy Interrupt Mask


414
NFCASE: NFC Access Size Error Interrupt Mask

RB_EDGEx: Ready/Busy Line x Interrupt Mask


415
29.19.7 HSMC NFC Address Cycle Zero Register

31 23 15 7
HSMC_ADDR 0xFFFFC018 Read-Write 0x00000000

30 22 14 6 29 21 13 5 28 20 12 27 19 11 26 18 10 2 25 17 9 1 24 16 8 0
4 3 ADDR_CYCLE0

When 5 address cycles are used, ADDR_CYCLE0 is the first Byte written to the NAND Flash (used by the NFC).

416
29.19.8 HSMC NFC Bank Register

31 23 15 7
HSMC_BANK 0xFFFFC01C Read-write 0x00000000

30 22 14 6 29 21 13 5 28 20 12 4 27 19 11 3 26 18 10 2 25 17 9 1 24 16 8 0 BANK
BANK: Bank Identifier

Number of the bank used.

417
29.19.9 PMECC Configuration Register

31 23 15 7
HSMC_PMECCFG 0xFFFFC070 Read-write 0x00000000

30 22 14 6 29 21 13 5 28 20 AUTO 12 NANDWR 4 SECTORSZ 27 19 11 3 26 18 10 2 25 17 9 PAGESIZE 1 BCH_ERR 0 24 16 SPAREEN 8
BCH_ERR: Error Correcting Capability

Value 0 1 2 3 4 Name BCH_ERR2 BCH_ERR4 BCH_ERR8 BCH_ERR12 BCH_ERR24 Description 2 errors 4 errors 8 errors 12 errors 24 errors
SECTORSZ: Sector Size 0: The ECC computation is based on a sector of 512 Bytes.
1: The ECC computation is based on a sector of 1024 Bytes.
PAGESIZE: Number of Sectors in the Page

Value 0 1 2 3 Name PAGESIZE_1SEC PAGESIZE_2SEC PAGESIZE_4SEC PAGESIZE_8SEC Description 1 sector for main area (512 or 1024 Bytes) 2 sectors for main area (1024 or 2048 Bytes) 4 sectors for main area (2048 or 4096 Bytes) 8 sectors for main area (4096 or 8192 Bytes)
NANDWR: NAND Write Access 0: NAND read access

1: NAND write access
SPAREEN: Spare Enable - For NAND write access:

0: The spare area is skipped 1: The spare area is protected with the last sector of data.
- For NAND read access:

0: The spare area is skipped. 1: The spare area contains protected data or only redundancy information.

418
AUTO: Automatic Mode Enable

This bit is only relevant in NAND Read Mode, when spare enable is activated. 0: Indicates that the spare area is not protected. In that case, the ECC computation takes into account the ECC area located in the spare area. (within the start address and the end address). 1: Indicates that the spare area is error protected. In this case, the ECC computation takes into account the whole spare area minus the ECC area in the ECC computation operation.

419
29.19.10 PMECC Spare Area Size Register

31 23 15 7
HSMC_PMECCSAREA 0xFFFFC074 Read-write 0x00000000

30 22 14 6 29 21 13 5 28 20 12 4 SPARESIZE 27 19 11 3 26 18 10 2 25 17 9 1 24 16 8 SPARESIZE 0
SPARESIZE: Spare Area Size

Number of Bytes in the spare area. The spare area size is equal to (SPARESIZE+1) Bytes.

420
29.19.11 PMECC Start Address Register

31 23 15 7
HSMC_PMECCSADDR 0xFFFFC078 Read-write 0x00000000

30 22 14 6 29 21 13 5 28 20 12 4 STARTADDR 27 19 11 3 26 18 10 2 25 17 9 1 24 16 8 STARTADDR 0
STARTADDR: ECC Area Start Address

This register is programmed with the start ECC start address. When STARTADDR is equal to 0, then the first ECC Byte is located at the first Byte of the spare area.

421
29.19.12 PMECC End Address Register

31 23 15 7
HSMC_PMECCEADDR 0xFFFFC07C Read-write 0x00000000

30 22 14 6 29 21 13 5 28 20 12 4 ENDADDR 27 19 11 3 26 18 10 2 25 17 9 1 24 16 8 ENDADDR 0
ENDADDR: ECC Area End Address

This register is programmed with the start ECC end address. When ENDADDR is equal to N, then the first ECC Byte is located at Byte N of the spare area.

422
29.19.13 PMECC Control Register

31 23 15 7
HSMC_PMECCTRL 0xFFFFC084 Write 0x00000000

30 22 14 6 29 21 13 5 DISABLE 28 20 12 4 ENABLE 27 19 11 3 26 18 10 2 USER 25 17 9 1 DATA 24 16 8 0 RST
RST: Reset the PMECC Module

0: No effect. 1: Reset the PMECC controller.
DATA: Start a Data Phase

0: No effect. 1: The PMECC controller enters a Data phase.
USER: Start a User Mode Phase

0: No effect. 1: The PMECC controller enters a User mode phase.
ENABLE: PMECC Enable

0: No effect. 1: Enable the PMECC controller.
DISABLE: PMECC Enable

0: No effect. 1: Disable the PMECC controller.

423
29.19.14 PMECC Status Register

31 23 15 7
HSMC_PMECCSR 0xFFFFC088 Read-only 0x00000000

30 22 14 6 29 21 13 5 28 20 12 4 ENABLE 27 19 11 3 26 18 10 2 25 17 9 1 24 16 8 0 BUSY
BUSY: The kernel of the PMECC is busy

0: PMECC controller finite state machine reached idle state. 1: PMECC controller finite state machine is processing the incoming Byte stream.
ENABLE: PMECC Enable bit

0: PMECC controller is disabled. 1: PMECC controller is enabled.

424
29.19.15 PMECC Interrupt Enable Register

31 23 15 7
HSMC_PMECCIER 0xFFFFC08C Write 0x00000000

30 22 14 6 29 21 13 5 28 20 12 4 27 19 11 3 26 18 10 2 25 17 9 1 24 16 8 0 ERRIE
ERRIE: Error Interrupt Enable

0: No effect. 1: The Multibit Error interrupt is enabled. An interrupt will be raised if at least one error is detected in at least one sector.

425
29.19.16 PMECC Interrupt Disable Register

31 23 15 7
HSMC_PMECCIDR 0xFFFFC090 Write 0x00000000

30 22 14 6 29 21 13 5 28 20 12 4 27 19 11 3 26 18 10 2 25 17 9 1 24 16 8 0 ERRID
ERRID: Error Interrupt Disable

0: No effect. 1: The Multibit Error interrupt is disabled.

426
29.19.17 PMECC Interrupt Mask Register

31 23 15 7
HSMC_PMECCIMR 0xFFFFC094 Read-only 0x00000000

30 22 14 6 29 21 13 5 28 20 12 4 27 19 11 3 26 18 10 2 25 17 9 1 24 16 8 0 ERRIM
ERRIM: Error Interrupt Mask

0: The Multibit Error is disabled. 1: The Multibit Error is enabled.

427
29.19.18 PMECC Interrupt Status Register

31 23 15 7
HSMC_PMECCISR 0xFFFFC098 Read-only 0x00000000

30 22 14 6 29 21 13 5 28 20 12 4 ERRIS 27 19 11 3 26 18 10 2 25 17 9 1 24 16 8 0
ERRIS: Error Interrupt Status Register

When set to one, bit i of the PMECCISR register indicates that sector i is corrupted.

428
29.19.19 PMECC Redundancy x Register

Name: Address: HSMC_PMECCx [x=0..10] [sec_num=0..7] 0xFFFFC0B0 [0][0] .. 0xFFFFC0D8 [10][0]; 0xFFFFC0F0 [0][1] .. 0xFFFFC118 [10][1]; 0xFFFFC130 [0][2] .. 0xFFFFC158 [10][2]; 0xFFFFC170 [0][3] .. 0xFFFFC198 [10][3]; 0xFFFFC1B0 [0][4] .. 0xFFFFC1D8 [10][4]; 0xFFFFC1F0 [0][5] .. 0xFFFFC218 [10][5]; 0xFFFFC230 [0][6] .. 0xFFFFC258 [10][6]; 0xFFFFC270 [0][7] .. 0xFFFFC298 [10][7] Read-only 0x00000000
30 29 28 ECC 23 22 21 20 ECC 15 14 13 12 ECC 7 6 5 4 ECC 3 2 1 0 11 10 9 8 19 18 17 16 27 26 25 24
Access: Reset:
31
ECC: BCH Redundancy

This register contains the remainder of the division of the codeword by the generator polynomial.

429
29.19.20 PMECC Remainder x Register

Name: Address: HSMC_REMx [x=0..11] [sec_num=0..7] 0xFFFFC2B0 [0][0] .. 0xFFFFC2DC [11][0]; 0xFFFFC2F0 [0][1] .. 0xFFFFC31C [11][1]; 0xFFFFC330 [0][2] .. 0xFFFFC35C [11][2]; 0xFFFFC370 [0][3] .. 0xFFFFC39C [11][3]; 0xFFFFC3B0 [0][4] .. 0xFFFFC3DC [11][4]; 0xFFFFC3F0 [0][5] .. 0xFFFFC41C [11][5]; 0xFFFFC430 [0][6] .. 0xFFFFC45C [11][6]; 0xFFFFC470 [0][7] .. 0xFFFFC49C [11][7] Read-only 0x00000000
30 22 29 28 27 REM2NP3 21 20 REM2NP3 15 7 14 6 13 12 11 REM2NP1 5 4 REM2NP1 3 2 1 0 10 9 8 19 18 17 16 26 25 24
Access: Reset:
31 23
REM2NP1: BCH Remainder 2 * N + 1

When sector size is set to 512 Bytes, bit REM2NP1[13] is not used and read as zero. If bit i of the REM2NP1 field is set to one, then the coefficient of the X ^ i is set to one; otherwise, the coefficient is zero.
REM2NP3: BCH Remainder 2 * N + 3

When sector size is set to 512 Bytes, bit REM2NP3[29] is not used and read as zero. If bit i of the REM2NP3 field is set to one, then the coefficient of the X ^ i is set to one; otherwise, the coefficient is zero.

430
29.19.21 PMECC Error Location Configuration Register

31 23 15 7
HSMC_ELCFG 0xFFFFC500 Read-write 0x00000000

30 22 14 6 29 21 13 5 28 20 27 19 26 18 ERRNUM 10 2 25 17 24 16
12 4
11 3
9 1
8 0 SECTORSZ
ERRNUM: Number of Errors SECTORSZ: Sector Size

0: The ECC computation is based on a 512 Bytes sector 1: The ECC computation is based on a 1024 Bytes sector

431
29.19.22 PMECC Error Location Primitive Register

31 23 15
HSMC_ELPRIM 0xFFFFC504 Read-only 0x401A

30 22 14 29 21 13 28 20 12 PRIMITIV 7 6 5 4 PRIMITIV 3 2 1 0 27 19 11 26 18 10 25 17 9 24 16 8
PRIMITIV: Primitive Polynomial

This field indicates the Primitive Polynomial used in the ECC computation.

432
29.19.23 PMECC Error Location Enable Register

31 23 15 7
HSMC_ELEN 0xFFFFC508 Write 0x00000000

30 22 14 6 29 21 13 28 20 12 27 19 11 ENINIT 5 4 ENINIT 3 2 1 0 26 18 10 25 17 9 24 16 8
ENINIT: Error Location Enable

Initial bit number in the codeword.

433
29.19.24 PMECC Error Location Disable Register

31 23 15 7
HSMC_ELDIS 0xFFFFC50C Write 0x00000000

30 22 14 6 29 21 13 5 28 20 12 4 27 19 11 3 26 18 10 2 25 17 9 1 24 16 8 0 DIS
DIS: Disable Error Location Engine

0: No effect. 1: Disable the Error location engine.

434
29.19.25 PMECC Error Location Status Register

31 23 15 7
HSMC_ELSR 0xFFFFC510 Read 0x00000000

30 22 14 6 29 21 13 5 28 20 12 4 27 19 11 3 26 18 10 2 25 17 9 1 24 16 8 0 BUSY
BUSY: Error Location Engine Busy

0: The Error location engine is disabled. 1: The Error location engine is enabled and is finding roots of the polynomial.

435
29.19.26 PMECC Error Location Interrupt Enable Register

31 23 15 7
HSMC_ELIER 0xFFFFC514 Write 0x00000000

30 22 14 6 29 21 13 5 28 20 12 4 27 19 11 3 26 18 10 2 25 17 9 1 24 16 8 0 DONE
DONE: Computation Terminated Interrupt Enable

0: No Effect. 1: Interrupt Enable.

436
29.19.27 PMECC Error Location Interrupt Disable Register

31 23 15 7
HSMC_ELIDR 0xFFFFC518 Write 0x00000000

30 22 14 6 29 21 13 5 28 20 12 4 27 19 11 3 26 18 10 2 25 17 9 1 24 16 8 0 DONE
DONE: Computation Terminated Interrupt Disable

0: No effect. 1: Interrupt disable.

437
29.19.28 PMECC Error Location Interrupt Mask Register

31 23 15 7
HSMC_ELIMR 0xFFFFC51C Read 0x00000000

30 22 14 6 29 21 13 5 28 20 12 4 27 19 11 3 26 18 10 2 25 17 9 1 24 16 8 0 DONE
DONE: Computation Terminated Interrupt Mask

0: The Computation terminated interrupt is disabled. 1: The Computation terminated interrupt is enabled.

438
29.19.29 PMECC Error Location Interrupt Status Register

31 23 15 7
HSMC_ELISR 0xFFFFC520 Read 0x00000000

30 22 14 6 29 21 13 5 28 20 12 27 19 11 26 18 10 ERR_CNT 2 25 17 9 24 16 8
0 DONE
DONE: Computation Terminated Interrupt Status

When set to one, this indicates that the error location engine has completed the root finding algorithm.
ERR_CNT: Error Counter value

This field indicates the number of roots of the polynomial.

439
29.19.30 PMECC Error Location SIGMAx Register

31 23 15 7
HSMC_SIGMAx [x=0..24] 0xFFFFC528 [0] .. 0xFFFFC588 [24] Read-Write 0x1

30 22 14 6 29 21 13 28 20 12 27 19 11 SIGMAx 5 4 SIGMAx 3 2 1 0 26 18 10 25 17 9 24 16 8
SIGMAx: Coefficient of degree x in the SIGMA polynomial.

SIGMAx belongs to the finite field GF(2^13) when the sector size is set to 512 Bytes. SIGMAx belongs to the finite field GF(2^14) when the sector size is set to 1024 Bytes.

440
29.19.31 PMECC Error Location x Register

31 23 15 7
HSMC_ERRLOCx [x=0..23] 0xFFFFC58C Read-only 0x00000000

30 22 14 6 29 21 13 28 20 12 27 19 11 ERRLOCN 5 4 ERRLOCN 3 2 1 0 26 18 10 25 17 9 24 16 8
ERRLOCN: Error Position within the Set {sector area, spare area}
ERRLOCN points to 1 when the first bit of the main area is corrupted. If the sector size is set to 512 Bytes, the ERRLOCN points to 4096 when the last bit of the sector area is corrupted. If the sector size is set to 1024 Bytes, the ERRLOCN points to 8192 when the last bit of the sector area is corrupted. If the sector size is set to 512 Bytes, the ERRLOCN points to 4097 when the first bit of the spare area is corrupted. If the sector size is set to 1024 Bytes, the ERRLOCN points to 8193 when the first bit of the spare area is corrupted.

441
29.19.32 Setup Register

31 23 15 7
HSMC_SETUPx [x=0..3] 0xFFFFC600 [0], 0xFFFFC614 [1], 0xFFFFC628 [2], 0xFFFFC63C [3] Write-only
30 22 14 6 29 28 27 26 NCS_RD_SETUP 19 NRD_SETUP 13 12 11 10 NCS_WR_SETUP 3 NWE_SETUP 2 9 8 18 25 24
21
20
17
16
NWE_SETUP: NWE Setup Length

The NWE signal setup length is defined as: NWE setup length = (128 * NWE_SETUP[5] + NWE_SETUP[4:0]) clock cycles.
NCS_WR_SETUP: NCS Setup Length in Write Access

In write access, the NCS signal setup length is defined as: NCS setup length = (128 * NCS_WR_SETUP[5] + NCS_WR_SETUP[4:0]) clock cycles.
NRD_SETUP: NRD Setup Length

The NRD signal setup length is defined as: NRD setup length = (128 * NRD_SETUP[5] + NRD_SETUP[4:0]) clock cycles.
NCS_RD_SETUP: NCS Setup Length in Read Access

In Read access, the NCS signal setup length is defined as: NCS setup length = (128 * NCS_RD_SETUP[5] + NCS_RD_SETUP[4:0]) clock cycles.

442
29.19.33 Pulse Register

31 23 15 7
HSMC_PULSEx [x=0..3] 0xFFFFC604 [0], 0xFFFFC618 [1], 0xFFFFC62C [2], 0xFFFFC640 [3] Write-only
30 22 14 6 29 28 27 26 NCS_RD_PULSE 19 NRD_PULSE 13 12 11 10 NCS_WR_PULSE 3 NWE_PULSE 2 9 8 18 25 24
21
20
17
16
NWE_PULSE: NWE Pulse Length

The NWE signal pulse length is defined as: NWE pulse length = (256 * NWE_PULSE[6]+NWE_PULSE[5:0]) clock cycles. The NWE pulse must be at least one clock cycle.
NCS_WR_PULSE: NCS Pulse Length in WRITE Access

In Write access, The NCS signal pulse length is defined as: NCS pulse length = (256 * NCS_WR_PULSE[6] + NCS_WR_PULSE[5:0]) clock cycles. the NCS pulse must be at least one clock cycle.
NRD_PULSE: NRD Pulse Length

The NRD signal pulse length is defined as: NRD pulse length = (256 * NRD_PULSE[6] + NRD_PULSE[5:0]) clock cycles. The NRD pulse width must be as least 1 clock cycle.
NCS_RD_PULSE: NCS Pulse Length in READ Access

In READ mode, The NCS signal pulse length is defined as: NCS pulse length = (256 * NCS_RD_PULSE[6] + NCS_RD_PULSE[5:0]) clock cycles.

443
29.19.34 Cycle Register

31 23
HSMC_CYCLEx [x=0..3] 0xFFFFC608 [0], 0xFFFFC61C [1], 0xFFFFC630 [2], 0xFFFFC644 [3] Read-write
30 22 29 21 28 20 NRD_CYCLE 15 7 14 6 13 5 12 4 NWE_CYCLE 11 3 10 2 9 1 8 NWE_CYCLE 0 27 19 26 18 25 17 24 NRD_CYCLE 16
NWE_CYCLE: Total Write Cycle Length

The total write cycle length is the total duration in clock cycles of the write cycle. It is equal to the sum of the setup, pulse and hold steps of the NWE and NCS signals. It is defined as: Write cycle length = (NWE_CYCLE[8:7] * 256) + NWE_CYCLE[6:0] clock cycles.
NRD_CYCLE: Total Read Cycle Length

The total read cycle length is the total duration in clock cycles of the read cycle. It is equal to the sum of the setup, pulse and hold steps of the NRD and NCS signals. It is defined as: Read cycle length = (NRD_CYCLE[8:7] * 256) + NRD_CYCLE[6:0] clock cycles.

444
29.19.35 Timings Register

31 NFSEL 23 15 7
HSMC_TIMINGSx [x=0..3] 0xFFFFC60C [0], 0xFFFFC620 [1], 0xFFFFC634 [2], 0xFFFFC648 [3] Read-write
30 29 RBNSEL 21 13 5 TADL 28 27 26 TWB 20 12 OCMS 4 19 18 TRR 11 10 TAR 3 2 TCLR 1 0 9 8 17 16 25 24
22 14 6
TCLR: CLE to REN Low Delay

Command Latch Enable falling edge to Read Enable falling edge timing. Latch Enable Falling to Read Enable Falling = (TCLR[3] * 64) + TCLR[2:0] clock cycles.
TADL: ALE to Data Start

Last address latch cycle to the first rising edge of WEN for data input. Last address latch to first rising edge of WEN = (TADL[3] * 64) + TADL[2:0] clock cycles.
TAR: ALE to REN Low Delay

Address Latch Enable falling edge to Read Enable falling edge timing. Address Latch Enable to Read Enable = (TAR[3] * 64) + TAR[2:0] clock cycles.
OCMS: Off Chip Memory Scrambling Enable

When set to one, the memory scrambling is activated.
TRR: Ready to REN Low Delay

Ready/Busy signal to Read Enable falling edge timing. Read to REN = (TRR[3] * 64) + TRR[2:0] clock cycles.
TWB: WEN High to REN to Busy

Write Enable rising edge to Ready/Busy falling edge timing. Write Enable to Read/Busy = (TWB[3] * 64) + TWB[2:0] clock cycles.
RBNSEL: Ready/Busy Line Selection

This field indicates the selected Ready/Busy Line from the RBN bundle.
NFSEL: NAND Flash Selection

If this bit is set to one, the chip select is assigned to NAND Flash write enable and read enable lines drive the Error Correcting Code module.

445
29.19.36 Mode Register

31 23 15 7
HSMC_MODEx [x=0..3] 0xFFFFC610 [0], 0xFFFFC624 [1], 0xFFFFC638 [2], 0xFFFFC64C [3] Read-write
30 22 14 6 29 21 13 5 EXNW_MODE 28 20 TDF_MODE 12 DBW 4 27 19 26 25 24 16
18 17 TDF_CYCLES 10 2 9
11 3
8 BAT
1 0 WRITE_MODE READ_MODE
READ_MODE
1 (NRD_CTRL): The Read operation is controlled by the NRD signal. 0 (NCS_CTRL): The Read operation is controlled by the NCS signal.
WRITE_MODE
1 (NWE_CTRL): The Write operation is controlled by the NWE signal. 0 (NCS_CTRL): The Write operation is controller by the NCS signal.
EXNW_MODE: NWAIT Mode

The NWAIT signal is used to extend the current read or write signal. It is only taken into account during the pulse phase Read and Write controlling signal. When the use of NWAIT is enabled, at least one cycle hold duration must be programmed for the read and write controlling signal.
Value 0 1 2 3 FROZEN READY Name DISABLED Description Disabled Reserved Frozen Mode Ready Mode
Disabled: The NWAIT input signal is ignored on the corresponding Chip Select. Frozen Mode: If asserted, the NWAIT signal freezes the current read or write cycle. After deassertion, the read/write cycle is resumed from the point where it was stopped. Ready Mode: The NWAIT signal indicates the availability of the external device at the end of the pulse of the controlling read or write signal, to complete the access. If high, the access normally completes. If low, the access is extended until NWAIT returns high. BAT: Byte Access Type
This field is used only if DBW defines a 16-bit data bus. 1 (BYTE_WRITE): Byte write access type:
Write operation is controlled using NCS, NWR0, NWR1. Read operation is controlled using NCS and NRD.

446
0 (BYTE_SELECT): Byte select access type:
Write operation is controlled using NCS, NWE, NBS0, NBS1. Read operation is controlled using NCS, NRD, NBS0, NBS1. DBW: Data Bus Width
Value 0 1 Name BIT_8 BIT_16 Description 8-bit bus 16-bit bus
TDF_CYCLES: Data Float Time

This field gives the integer number of clock cycles required by the external device to release the data after the rising edge of the read controlling signal. The HSMC always provide one full cycle of bus turnaround after the TDF_CYCLES period. The external bus cannot be used by another chip select during TDF_CYCLES + 1 cycles. From 0 up to 15 TDF_CYCLES can be set.
TDF_MODE: TDF Optimization

1: TDF optimization is enabled.
The number of TDF wait states is optimized using the setup period of the next read/write access.
0: TDF optimization is disabled.
The number of TDF wait states is inserted before the next access begins.

447
29.19.37 OCMS Register

31 23 15 7
HSMC_OCMS 0xFFFFC6A0 Read-write 0x00000000

30 22 14 6 29 21 13 5 28 20 12 4 27 19 11 3 26 18 10 2 25 17 9 1 SRSE 24 16 8 0 SMSE
SMSE: Static Memory Controller Scrambling Enable

0: Disable Off Chip Scrambling for HSMC access. 1: Enable Off Chip Scrambling for HSMC access. (If OCMS field is set to 1 in the relevant HSMC_TIMINGS register.)
SRSE: SRAM Scrambling Enable

0: Disable SRAM Scrambling for SRAM access. 1: Enable SRAM Scrambling for SRAM access. (If OCMS field is set to 1 in the relevant HSMC_TIMINGS register.)

448
29.19.38 HSMC OCMS Key1 Register

31
HSMC_KEY1 0xFFFFC6A4 Write-once 0x00000000

30 29 28 KEY1 27 26 25 24
23
22
21
20 KEY1
19
18
17
16
15
14
13
12 KEY1
11
10
4 KEY1
KEY1: Off Chip Memory Scrambling (OCMS) Key Part 1

When Off Chip Memory Scrambling is enabled by setting the HSMC_OMCS and HSMC_TIMINGS registers in accordance, the data scrambling depends on KEY1 and KEY2 values.

449
29.19.39 HSMC OCMS Key2 Register

31
HSMC_KEY2 0xFFFFC6A8 Write-once 0x00000000

30 29 28 KEY2 27 26 25 24
23
22
21
20 KEY2
19
18
17
16
15
14
13
12 KEY2
11
10
4 KEY2
KEY2: Off Chip Memory Scrambling (OCMS) Key Part 2

When Off Chip Memory Scrambling is enabled by setting the HSMC_OMCS and HSMC_TIMINGS registers in accordance, the data scrambling depends on KEY2 and KEY1 values.

450
29.19.40 HSMC Write Protection Control

31
HSMC_WPCR 0xFFFFC6E4 Write-only

30 29 28 WP_KEY 27 26 25 24
23
22
21
20 WP_KEY
19
18
17
16
15
14
13
12 WP_KEY
11
10
0 WP_EN
WP_EN: Write Protection Enable

0: Disables the Write Protection if WP_KEY corresponds. 1: Enables the Write Protection if WP_KEY corresponds.
WP_KEY: Write Protection KEY password

Should be written at value 0x534D43 (ASCII code for SMC). Writing any other value in this field has no effect.

451
29.19.41 HSMC Write Protection Status

31 23
HSMC_WPSR 0xFFFFC6E8 Read-only

30 22 29 21 28 20 WP_VSRC 27 19 26 18 25 17 24 16
15
14
13
12 WP_VSRC
11
10
7 -
6 -
5 -
4 -
2 WP_VS
WP_VS: Write Protection Violation Status

0: No Write Protect Violation has occurred since the last read of the HSMC_WPSR register. 1: A Write Protect Violation has occurred since the last read of the HSMC_WPSR register. If this violation is an unauthorized attempt to write a protected register, the associated violation is reported into field WP_VSRC.
WP_VSRC: Write Protection Violation Source

WP_VSRC field Indicates the Register offset where the last violation occurred.

452
30.
30.1
DMA Controller (DMAC)

Description
The DMA Controller (DMAC) is an AHB-central DMA controller core that transfers data from a source peripheral to a destination peripheral over one or more AMBA buses. One channel is required for each source/destination pair. In the most basic configuration, the DMAC has one master interface and one channel. The master interface reads the data from a source and writes it to a destination. Two AMBA transfers are required for each DMAC data transfer. This is also known as a dual-access transfer. The DMAC is programmed via the APB interface.
30.2
z z z z z z z z z z z z z z z z
3 AHB-Lite Master Interfaces DMA Module Supports the Following Transfer Schemes: Peripheral-to-Memory, Memory-to-Peripheral, Peripheralto-Peripheral and Memory-to-Memory Source and Destination Operate independently on BYTE (8-bit), HALF-WORD (16-bit) and WORD (32-bit) Supports Hardware and Software Initiated Transfers Supports Multiple Buffer Chaining Operations Supports Incrementing/decrementing/fixed Addressing Mode Independently for Source and Destination Supports Programmable Address Increment/decrement on User-defined Boundary Condition to Enable Picture-inPicture Mode Programmable Arbitration Policy, Modified Round Robin and Fixed Priority are Available Supports Specified Length and Unspecified Length AMBA AHB Burst Access to Maximize Data Bandwidth AMBA APB Interface Used to Program the DMA Controller 8 DMA Channels on DMAC0 8 DMA Channels on DMAC1 16 External Request Lines on DMAC0 22 External Request Lines on DMAC1 Embedded FIFO Channel Locking and Bus Locking Capability

453
30.2.1 DMA Controller 0

The DMA Controller 0 can handle the transfer between peripherals and memory and so receives the triggers from the peripherals listed below. The hardware interface numbers are given in Table 30-1.
Table 30-1. DMA Channels Definition (DMAC0) Instance name HSMCI0 SPI0 SPI0 USART0 USART0 USART1 USART1 TWI0 TWI0 TWI1 TWI1 UART0 UART0 SSC0 SSC0 SMD SMD Channel T/R Receive/transmit Transmit Receive Transmit Receive Transmit Receive Transmit Receive Transmit Receive Transmit Receive Transmit Receive Transmit Receive Interface number 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

454
30.2.2 DMA Controller 1

The DMA controller 1can handle the transfer between peripherals and memory and thus receives triggers from the peripherals listed below. The hardware interface numbers are given in Table 30-2.
Table 30-2. DMA Channels Definition (DMAC1) Instance name HSMCI1 HSMCI2 ADC SSC1 SSC1 UART1 UART1 USART2 USART2 USART3 USART3 TWI2 TWI2 DBGU DBGU SPI1 SPI1 SHA AES AES TDES TDES Channel T/R Receive/transmit Receive/transmit Receive Transmit Receive Transmit Receive Transmit Receive Transmit Receive Transmit Receive Transmit Receive Transmit Receive Transmit Transmit Receive Transmit Receive Interface Number 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21

455
30.3
Block Diagram
Figure 30-1. DMA Controller (DMAC) Block Diagram
DMA Channel n DMA Destination DMA Channel 2 DMA Channel 1 DMA Channel 0
DMA Channel 0 Write data path to destination
Atmel APB rev2 Interface Status Registers Configuration Registers
DMA Atmel APB Interface
DMA Destination Control State Machine Destination Pointer Management
DMA Interrupt Controller
DMA Interrupt
DMA FIFO Controller
DMA FIFO Up to 64 bytes
Trigger Manager
External Triggers
Soft Triggers
DMA REQ/ACK Interface
DMA Hardware Handshaking Interface
DMA Channel 0 Read data path from source
DMA Source Control State Machine Source Pointer Management
DMA Read Datapath Bundles DMA Global Control and Data Mux
DMA Source Requests Pool DMA Global Request Arbiter
Requests & Data Interconnect
DMAC Master Interface 0
AMBA AHB Layer 0
AMBA AHB Layer 1
AMBA AHB Layer 2

456
30.4
30.4.1 Basic Definitions

Source peripheral: Device on an AMBA layer from where the DMAC reads data, which is then stored in the channel FIFO. The source peripheral teams up with a destination peripheral to form a channel. Destination peripheral: Device to which the DMAC writes the stored data from the FIFO (previously read from the source peripheral). Memory: Source or destination that is always ready for a DMAC transfer and does not require a handshaking interface to interact with the DMAC. Programmable Arbitration Policy: Modified Round Robin and Fixed Priority are available by means of the ARB_CFG bit in the Global Configuration Register (DMAC_GCFG). The fixed priority is linked to the channel number. The highest DMAC channel number has the highest priority. Channel: Read/write datapath between a source peripheral on one configured AMBA layer and a destination peripheral on the same or different AMBA layer that occurs through the channel FIFO. If the source peripheral is not memory, then a source handshaking interface is assigned to the channel. If the destination peripheral is not memory, then a destination handshaking interface is assigned to the channel. Source and destination handshaking interfaces can be assigned dynamically by programming the channel registers. Master interface: DMAC is a master on the AHB bus reading data from the source and writing it to the destination over the AHB bus. Slave interface: The APB interface over which the DMAC is programmed. The slave interface in practice could be on the same layer as any of the master interfaces or on a separate layer. Handshaking interface: A set of signal registers that conform to a protocol and handshake between the DMAC and source or destination peripheral to control the transfer of a single or chunk transfer between them. This interface is used to request, acknowledge, and control a DMAC transaction. A channel can receive a request through one of two types of handshaking interface: hardware or software. Hardware handshaking interface: Uses hardware signals to control the transfer of a single or chunk transfer between the DMAC and the source or destination peripheral. Software handshaking interface: Uses software registers to control the transfer of a single or chunk transfer between the DMAC and the source or destination peripheral. No special DMAC handshaking signals are needed on the I/O of the peripheral. This mode is useful for interfacing an existing peripheral to the DMAC without modifying it. Transfer hierarchy: Figure 30-2 on page 458 illustrates the hierarchy between DMAC transfers, buffer transfers, chunk or single, and AMBA transfers (single or burst) for non-memory peripherals. Figure 30-3 on page 458 shows the transfer hierarchy for memory.

457
Figure 30-2. DMAC Transfer Hierarchy for Non-Memory Peripheral

DMAC Transfer
DMA Transfer Level
Buffer
Buffer
Buffer
Buffer Transfer Level
Chunk Transfer
Chunk Transfer
Chunk Transfer
Single Transfer
DMA Transaction Level
AMBA Burst Transfer
AMBA Burst Transfer
AMBA Burst Transfer
AMBA Single Transfer
AMBA Transfer Level
Figure 30-3. DMAC Transfer Hierarchy for Memory

DMAC Transfer
DMA Transfer Level
Buffer
Buffer
Buffer
Buffer Transfer Level
AMBA Burst Transfer
AMBA Burst Transfer
AMBA Burst Transfer
AMBA Transfer Level
Buffer: A buffer of DMAC data. The amount of data (length) is determined by the flow controller. For transfers between the DMAC and memory, a buffer is broken directly into a sequence of AMBA bursts and AMBA single transfers. For transfers between the DMAC and a non-memory peripheral, a buffer is broken into a sequence of DMAC transactions (single and chunks). These are in turn broken into a sequence of AMBA transfers. Transaction: A basic unit of a DMAC transfer as determined by either the hardware or software handshaking interface. A transaction is only relevant for transfers between the DMAC and a source or destination peripheral if the source or destination peripheral is a non-memory device. There are two types of transactions: single transfer and chunk transfer.
z z
Single transfer: The length of a single transaction is always 1 and is converted to a single AMBA access. Chunk transfer: The length of a chunk is programmed into the DMAC. The chunk is then converted into a sequence of AHB access.DMAC executes each AMBA burst transfer by performing incremental bursts that are no longer than 16 beats.
DMAC transfer: Software controls the number of buffers in a DMAC transfer. Once the DMAC transfer has completed, then hardware within the DMAC disables the channel and can generate an interrupt to signal the completion of the DMAC transfer. You can then re-program the channel for a new DMAC transfer. Single-buffer DMAC transfer: Consists of a single buffer.

458
Multi-buffer DMAC transfer: A DMAC transfer may consist of multiple DMAC buffers. Multi-buffer DMAC transfers are supported through buffer chaining (linked list pointers), auto-reloading of channel registers, and contiguous buffers. The source and destination can independently select which method to use.
z
Linked lists (buffer chaining) A descriptor pointer (DSCR) points to the location in system memory where the next linked list item (LLI) exists. The LLI is a set of registers that describe the next buffer (buffer descriptor) and a descriptor pointer register. The DMAC fetches the LLI at the beginning of every buffer when buffer chaining is enabled. Replay The DMAC automatically reloads the channel registers at the end of each buffers to the value when the channel was first enabled. Contiguous buffers Where the address of the next buffer is selected to be a continuation from the end of the previous buffer.
z z
Picture-in-Picture Mode: DMAC contains a Picture-in-Picture mode support. When this mode is enabled, addresses are automatically incremented by a programmable value when the DMAC channel transfer count reaches a user defined boundary. Figure 30-4 on page 459 illustrates a memory mapped image 4:2:2 encoded located at image_base_address in memory. A user defined start address is defined at Picture_start_address. The incremented value is set to memory_hole_size = image_width - picture_width, and the boundary is set to picture_width.
Figure 30-4. Picture-In-Picture Mode Support
DMAC PIP transfers
Channel locking: Software can program a channel to keep the AHB master interface by locking the arbitration for the master bus interface for the duration of a DMAC transfer, buffer, or chunk. Bus locking: Software can program a channel to maintain control of the AMBA bus by asserting hmastlock for the duration of a DMAC transfer, buffer, or transaction (single or chunk). Channel locking is asserted for the duration of bus locking at a minimum.

459
30.4.2 Memory Peripherals

Figure 30-3 on page 458 shows the DMAC transfer hierarchy of the DMAC for a memory peripheral. There is no handshaking interface with the DMAC, and therefore the memory peripheral can never be a flow controller. Once the channel is enabled, the transfer proceeds immediately without waiting for a transaction request. The alternative to not having a transaction-level handshaking interface is to allow the DMAC to attempt AMBA transfers to the peripheral once the channel is enabled. If the peripheral slave cannot accept these AMBA transfers, it inserts wait states onto the bus until it is ready; it is not recommended that more than 16 wait states be inserted onto the bus. By using the handshaking interface, the peripheral can signal to the DMAC that it is ready to transmit/receive data, and then the DMAC can access the peripheral without the peripheral inserting wait states onto the bus.
30.4.3 Handshaking Interface

Handshaking interfaces are used at the transaction level to control the flow of single or chunk transfers. The operation of the handshaking interface is different and depends on whether the peripheral or the DMAC is the flow controller. The peripheral uses the handshaking interface to indicate to the DMAC that it is ready to transfer/accept data over the AMBA bus. A non-memory peripheral can request a DMAC transfer through the DMAC using one of two handshaking interfaces:
z z
Hardware handshaking Software handshaking
Software selects between the hardware or software handshaking interface on a per-channel basis. Software handshaking is accomplished through memory-mapped registers, while hardware handshaking is accomplished using a dedicated handshaking interface. 30.4.3.1 Software Handshaking When the slave peripheral requires the DMAC to perform a DMAC transaction, it communicates this request by sending an interrupt to the CPU or interrupt controller. The interrupt service routine then uses the software registers to initiate and control a DMAC transaction. These software registers are used to implement the software handshaking interface. The SRC_H2SEL/DST_H2SEL bit in the DMAC_CFGx channel configuration register must be set to zero to enable software handshaking. When the peripheral is not the flow controller, then the last transaction register DMAC_LAST is not used, and the values in these registers are ignored.
Chunk Transactions
Writing a 1 to the DMAC_CREQ[2x] register starts a source chunk transaction request, where x is the channel number. Writing a 1 to the DMAC_CREQ[2x+1] register starts a destination chunk transfer request, where x is the channel number. Upon completion of the chunk transaction, the hardware clears the DMAC_CREQ[2x] or DMAC_CREQ[2x+1].
Single Transactions
Writing a 1 to the DMAC_SREQ[2x] register starts a source single transaction request, where x is the channel number. Writing a 1 to the DMAC_SREQ[2x+1] register starts a destination single transfer request, where x is the channel number. Upon completion of the chunk transaction, the hardware clears the DMAC_SREQ[x] or DMAC_SREQ[2x+1]. The software can poll the relevant channel bit in the DMAC_CREQ[2x]/DMAC_CREQ[2x+1] and DMAC_SREQ[x]/DMAC_SREQ[2x+1] registers. When both are 0, then either the requested chunk or single transaction has completed.

460
30.4.4 DMAC Transfer Types

A DMAC transfer may consist of single or multi-buffer transfers. On successive buffers of a multi-buffer transfer, the DMAC_SADDRx/DMAC_DADDRx registers in the DMAC are reprogrammed using either of the following methods:
z z z
Buffer chaining using linked lists Replay mode Contiguous address between buffers
On successive buffers of a multi-buffer transfer, the DMAC_CTRLAx and DMAC_CTRLBx registers in the DMAC are reprogrammed using either of the following methods:
z z
Buffer chaining using linked lists Replay mode
When buffer chaining using linked lists is the multi-buffer method of choice, and on successive buffers, the DMAC_DSCRx register in the DMAC is re-programmed using the following method:
z
Buffer chaining using linked lists
A buffer descriptor (LLI) consists of following registers, DMAC_SADDRx, DMAC_DADDRx, DMAC_DSCRx, DMAC_CTRLAx, DMAC_CTRLBx.These registers, along with the DMAC_CFGx register, are used by the DMAC to set up and describe the buffer transfer. 30.4.4.1 Multi-buffer Transfers
Buffer Chaining Using Linked Lists

In this case, the DMAC re-programs the channel registers prior to the start of each buffer by fetching the buffer descriptor for that buffer from system memory. This is known as an LLI update. DMAC buffer chaining is supported by using a Descriptor Pointer register (DMAC_DSCRx) that stores the address in memory of the next buffer descriptor. Each buffer descriptor contains the corresponding buffer descriptor (DMAC_SADDRx, DMAC_DADDRx, DMAC_DSCRx, DMAC_CTRLAx DMAC_CTRLBx). To set up buffer chaining, a sequence of linked lists must be programmed in memory. The DMAC_SADDRx, DMAC_DADDRx, DMAC_DSCRx, DMAC_CTRLAx and DMAC_CTRLBx registers are fetched from system memory on an LLI update. The updated content of the DMAC_CTRLAx register is written back to memory on buffer completion. Figure 30-5 on page 462 shows how to use chained linked lists in memory to define multi-buffer transfers using buffer chaining. The Linked List multi-buffer transfer is initiated by programming DMAC_DSCRx with DSCRx(0) (LLI(0) base address) different from zero. Other fields and registers are ignored and overwritten when the descriptor is retrieved from memory. The last transfer descriptor must be written to memory with its next descriptor address set to 0.

461
Figure 30-5. Multi Buffer Transfer Using Linked List
System Memory LLI(0) DSCRx(1)= DSCRx(0) + 0x10 CTRLBx= DSCRx(0) + 0xC CTRLAx= DSCRx(0) + 0x8 DADDRx= DSCRx(0) + 0x4 SADDRx= DSCRx(0) + 0x0 DSCRx(1)
LLI(1) DSCRx(2)= DSCRx(1) + 0x10 CTRLBx= DSCRx(1) + 0xC CTRLBx= DSCRx(1) + 0x8 DADDRx= DSCRx(1) + 0x4 SADDRx= DSCRx(1) + 0x0 DSCRx(2) (points to 0 if LLI(1) is the last transfer descriptor
DSCRx(0)
Descriptor Integrity Check

When the Descriptor Integrity Check is enabled, a cyclic redundancy check information is attached to the descriptor. When fetched from the memory, the descriptor is verified through the use of a CRC16-CCIT (0x1021 polynom) by the DMAC channel. If a CRC error is detected, then the DICERR flag is set in the DMAC_EBCISR register. The CRC16 is computed from MSB to LSB. The BTSIZE and DONE fields of the DMAC_CTRLAx register are ignored and set to zero.
Figure 30-6. Linked List with CRC16 Attached
LLI(0) System Memory
LLI(1)
CRCx(1)= DSCRx(0) + 0x14 DSCRx(1)= DSCRx(0) + 0x10 CTRLBx= DSCRx(0) + 0xC CTRLAx= DSCRx(0) + 0x8 DADDRx= DSCRx(0) + 0x4 SADDRx= DSCRx(0) + 0x0 DSCRx(1)
CRCx(2)= DSCRx(1) + 0x14 DSCRx(2)= DSCRx(1) + 0x10 CTRLBx= DSCRx(1) + 0xC CTRLBx= DSCRx(1) + 0x8 DADDRx= DSCRx(1) + 0x4 SADDRx= DSCRx(1) + 0x0 DSCRx(2) (points to 0 if LLI(1) is the last transfer descriptor
DSCRx(0)

462
30.4.4.2 Programming DMAC for Multiple Buffer Transfers

Table 30-3. Multiple Buffers Transfer Management Table
Transfer Type 1) Single Buffer or Last buffer of a multiple buffer transfer 2) Multi Buffer transfer with contiguous DADDR 3) Multi Buffer transfer with contiguous SADDR 4) Multi Buffer transfer with LLI support 5) Multi Buffer transfer with DADDR reloaded 6) Multi Buffer transfer with SADDR reloaded 7) Multi Buffer transfer with BTSIZE reloaded and contiguous DADDR 8) Multi Buffer transfer with BTSIZE reloaded and contiguous SADDR 9) Automatic mode channel is stalling BTSIZE is reloaded 10) Automatic mode BTSIZE, SADDR and DADDR reloaded 11) Automatic mode BTSIZE, SADDR reloaded and DADDR contiguous AUTO 0 SRC_REP DST_REP SRC_DSCR DST_DSCR BTSIZE USR DSCR 0 SADDR USR DADDR USR Other Fields USR
0 0 0 0 0
0 1
0 1
0 1 0 0 1
1 0 0 1 0
LLI LLI LLI LLI LLI
USR USR USR USR USR
LLI CONT LLI LLI REP
CONT LLI LLI REP LLI
LLI LLI LLI LLI LLI
REP
USR
LLI
CONT
LLI
REP
USR
CONT
LLI
LLI
REP
USR
CONT
CONT
REP
REP
USR
REP
REP
REP
REP
USR
REP
CONT
REP
Notes: 1. USR means that the register field is manually programmed by the user. 2. CONT means that address are contiguous. 3. REP means that the register field is updated with its previous value. If the transfer is the first one, then the user must manually program the value. 4. Channel stalled is true if the relevant BTC interrupt is not masked. 5. LLI means that the register field is updated with the content of the linked list item.
Replay Mode of Channel Registers

During automatic replay mode, the channel registers are reloaded with their initial values at the completion of each buffer and the new values used for the new buffer. Depending on the row number in Table 30-3 on page 463, some or all of the DMAC_SADDRx, DMAC_DADDRx, DMAC_CTRLAx and DMAC_CTRLBx channel registers are reloaded from their initial value at the start of a buffer transfer.
Contiguous Address Between Buffers

In this case, the address between successive buffers is selected to be a continuation from the end of the previous buffer. Enabling the source or destination address to be contiguous between buffers is a function of DMAC_CTRLAx.SRC_DSCR, DMAC_CFGx.DST_REP, DMAC_CFGx.SRC_REP and DMAC_CTRLAx.DST_DSCR registers.

463
Suspension of Transfers Between Buffers

At the end of every buffer transfer, an end of buffer interrupt is asserted if:
z
The channel buffer interrupt is unmasked, DMAC_EBCIMR.BTCx = 1, where x is the channel number. The Buffer Transfer Completed Interrupt is generated at the completion of the buffer transfer to the destination. The channel end of the Chained Buffer Transfer Completed Interrupt is unmasked, DMAC_EBCIMR.CBTCx = 1, when n is the channel number.
Note:
z
At the end of a chain of multiple buffers, an end of linked list interrupt is asserted if:
30.4.4.3 Ending Multi-buffer Transfers All multi-buffer transfers must end as shown in Row 1 of Table 30-3 on page 463. At the end of every buffer transfer, the DMAC samples the row number, and if the DMAC is in Row 1 state, then the previous buffer transferred was the last buffer and the DMAC transfer is terminated. For rows 9, 10 and 11 of Table 30-3 on page 463, (DMAC_DSCRx = 0 and DMAC_CTRLBx.AUTO is set), multi-buffer DMAC transfers continue until the automatic mode is disabled by writing a 1 in DMAC_CTRLBx.AUTO bit. This bit should be programmed to zero in the end of buffer interrupt service routine that services the next-to-last buffer transfer. This puts the DMAC into Row 1 state. For rows 2, 3, 4, 5, and 6 (DMAC_CRTLBx.AUTO cleared), the user must set up the last buffer descriptor in memory so that LLI.DMAC_DSCRx is set to 0.
30.4.5 Programming a Channel

Four registers, the DMAC_DSCRx, the DMAC_CTRLAx, the DMAC_CTRLBx and DMAC_CFGx, need to be programmed to set up whether single or multi-buffer transfers take place, and which type of multi-buffer transfer is used. The different transfer types are shown in Table 30-3 on page 463. The BTSIZE, SADDR and DADDR columns indicate where the values of DMAC_SARx, DMAC_DARx, DMAC_CTLx, and DMAC_LLPx are obtained for the next buffer transfer when multi-buffer DMAC transfers are enabled. 30.4.5.1 Programming Examples
Single-buffer Transfer (Row 1)

1. 2. 3. Read the Channel Handler Status Register DMAC_CHSR.ENAx Field to choose a free (disabled) channel. Clear any pending interrupts on the channel from the previous DMAC transfer by reading the interrupt status register, DMAC_EBCISR. Program the following channel registers: 1. 2. 3. 4. 5. Write the starting source address in the DMAC_SADDRx register for channel x. Write the starting destination address in the DMAC_DADDRx register for channel x. Write the next descriptor address in the DMA_DSCRx register for channel x with 0x0. Program DMAC_CTRLAx, DMAC_CTRLBx and DMAC_CFGx according to Row 1 as shown in Table 30-3 on page 463. Program the DMAC_CTRLBx register with both AUTO fields set to 0. Write the control information for the DMAC transfer in the DMAC_CTRLAx and DMAC_CTRLBx registers for channel x. For example, in the register, you can program the following: i. Set up the transfer type (memory or non-memory peripheral for source and destination) and flow control device by programming the FC of the DMAC_CTRLBx register. ii. Set up the transfer characteristics, such as: z Transfer width for the source in the SRC_WIDTH field. z Transfer width for the destination in the DST_WIDTH field. z Source AHB Master interface layer in the SIF field where source resides. z Destination AHB Master Interface layer in the DIF field where destination resides. z Incrementing/decrementing or fixed address for source in SRC_INC field. z Incrementing/decrementing or fixed address for destination in DST_INC field.

464
6.
Write the channel configuration information into the DMAC_CFGx register for channel x. i. Designate the handshaking interface type (hardware or software) for the source and destination peripherals. This is not required for memory. This step requires programming the SRC_H2SEL/DST_H2SEL bits, respectively. Writing a 1 activates the hardware handshaking interface to handle source/destination requests. Writing a 0 activates the software handshaking interface to handle source/destination requests. ii. If the hardware handshaking interface is activated for the source or destination peripheral, assign a handshaking interface to the source and destination peripheral. This requires programming the SRC_PER and DST_PER bits, respectively.
7. 8. 4.
If source Picture-in-Picture mode is enabled (DMAC_CTRLBx.SRC_PIP is enabled), program the DMAC_SPIPx register for channel x. If destination Picture-in-Picture mode is enabled (DMAC_CTRLBx.DST_PIP is enabled), program the DMAC_DPIPx register for channel x.
After the DMAC selected channel has been programmed, enable the channel by writing a 1 to the DMAC_CHER.ENAx bit, where x is the channel number. Make sure that bit 0 of DMAC_EN.ENABLE register is enabled. Source and destination request single and chunk DMAC transactions to transfer the buffer of data (assuming nonmemory peripherals). The DMAC acknowledges at the completion of every transaction (chunk and single) in the buffer and carries out the buffer transfer. Once the transfer completes, the hardware sets the interrupts and disables the channel. At this time, you can either respond to the Buffer Transfer Completed Interrupt or Chained Buffer Transfer Completed Interrupt, or poll for the Channel Handler Status Register (DMAC_CHSR.ENAx) bit until it is cleared by hardware, to detect when the transfer is complete.
5.
6.
Multi-buffer Transfer with Linked List for Source and Linked List for Destination (Row 4)
1. 2. Read the Channel Handler Status register to choose a free (disabled) channel. Set up the chain of Linked List Items (otherwise known as buffer descriptors) in memory. Write the control information in the LLI.DMAC_CTRLAx and LLI.DMAC_CTRLBx registers location of the buffer descriptor for each LLI in memory (see Figure 30-7 on page 467) for channel x. For example, in the register, you can program the following: 1. 2. Set up the transfer type (memory or non-memory peripheral for source and destination) and flow control device by programming the FC of the DMAC_CTRLBx register. Set up the transfer characteristics, such as: i. Transfer width for the source in the SRC_WIDTH field. ii. Transfer width for the destination in the DST_WIDTH field. iii. Source AHB master interface layer in the SIF field where source resides. iv. Destination AHB master interface layer in the DIF field where destination resides. v. Incrementing/decrementing or fixed address for source in SRC_INCR field. vi. Incrementing/decrementing or fixed address for destination DST_INCR field. 3. Write the channel configuration information into the DMAC_CFGx register for channel x. 1. Designate the handshaking interface type (hardware or software) for the source and destination peripherals. This is not required for memory. This step requires programming the SRC_H2SEL/DST_H2SEL bits, respectively. Writing a 1 activates the hardware handshaking interface to handle source/destination requests for the specific channel. Writing a 0 activates the software handshaking interface to handle source/destination requests. If the hardware handshaking interface is activated for the source or destination peripheral, assign the handshaking interface to the source and destination peripheral. This requires programming the SRC_PER and DST_PER bits, respectively.
2.

465
4.
Make sure that the LLI.DMAC_CTRLBx register locations of all LLI entries in memory (except the last) are set as shown in Row 4 of Table 30-3 on page 463. The LLI.DMAC_CTRLBx register of the last Linked List Item must be set as described in Row 1 of Table 30-3. Figure 30-5 on page 462 shows a Linked List example with two list items. Make sure that the LLI.DMAC_DSCRx register locations of all LLI entries in memory (except the last) are non-zero and point to the base address of the next Linked List Item. Make sure that the LLI.DMAC_SADDRx/LLI.DMAC_DADDRx register locations of all LLI entries in memory point to the start source/destination buffer address preceding that LLI fetch. Make sure that the LLI.DMAC_CTRLAx.DONE field of the LLI.DMAC_CTRLAx register locations of all LLI entries in memory are cleared. If source Picture-in-Picture mode is enabled (DMAC_CTRLBx.SRC_PIP is enabled), program the DMAC_SPIPx register for channel x. If destination Picture-in-Picture is enabled (DMAC_CTRLBx.DST_PIP is enabled), program the DMAC_DPIPx register for channel x.
5. 6. 7. 8. 9.
10. Clear any pending interrupts on the channel from the previous DMAC transfer by reading the status register: DMAC_EBCISR. 11. Program the DMAC_CTRLBx, DMAC_CFGx registers according to Row 4 as shown in Table 30-3 on page 463. 12. Program the DMAC_DSCRx register with DMAC_DSCRx(0), the pointer to the first Linked List item. 13. Finally, enable the channel by writing a 1 to the DMAC_CHER.ENAx bit, where x is the channel number. The transfer is performed. 14. The DMAC fetches the first LLI from the location pointed to by DMAC_DSCRx(0). Note: The LLI.DMAC_SADDRx, LLI. DMAC_DADDRx, LLI.DMAC_DSCRx, LLI.DMAC_CTRLAx and LLI.DMAC_CTRLBx registers are fetched. The DMAC automatically reprograms the DMAC_SADDRx, DMAC_DADDRx, DMAC_DSCRx, DMAC_CTRLBx and DMAC_CTRLAx channel registers from the DMAC_DSCRx(0).
15. Source and destination request single and chunk DMAC transactions to transfer the buffer of data (assuming nonmemory peripheral). The DMAC acknowledges at the completion of every transaction (chunk and single) in the buffer and carries out the buffer transfer. 16. Once the buffer of data is transferred, the DMAC_CTRLAx register is written out to system memory at the same location and on the same layer (DMAC_DSCRx.DSCR_IF) where it was originally fetched, that is, the location of the DMAC_CTRLAx register of the linked list item fetched prior to the start of the buffer transfer. Only DMAC_CTRLAx register is written out because only the DMAC_CTRLAx.BTSIZE and DMAC_CTRLAX.DONE bits have been updated by DMAC hardware. Additionally, the DMAC_CTRLAx.DONE bit is asserted when the buffer transfer has completed. Note: Do not poll the DMAC_CTRLAx.DONE bit in the DMAC memory map. Instead, poll the LLI.DMAC_CTRLAx.DONE bit in the LLI for that buffer. If the poll LLI.DMAC_CTRLAx.DONE bit is asserted, then this buffer transfer has completed. This LLI.DMAC_CTRLAx.DONE bit was cleared at the start of the transfer.
17. The DMAC does not wait for the buffer interrupt to be cleared, but continues fetching the next LLI from the memory location pointed to by current DMAC_DSCRx register and automatically reprograms the DMAC_SADDRx, DMAC_DADDRx, DMAC_DSCRx, DMAC_CTRLAx and DMAC_CTRLBx channel registers. The DMAC transfer continues until the DMAC determines that the DMAC_CTRLBx and DMAC_DSCRx registers at the end of a buffer transfer match described in Row 1 of Table 30-3 on page 463. The DMAC then knows that the previous buffer transferred was the last buffer in the DMAC transfer. The DMAC transfer might look like that shown in Figure 30-7 on page 467.

466
Figure 30-7. Multi-buffer with Linked List Address for Source and Destination
Address of Source Layer
Address of Destination Layer
Buffer 2 SADDR(2) DADDR(2)
Buffer 2
Buffer 1 SADDR(1) DADDR(1)
Buffer 1
Buffer 0 SADDR(0) Source Buffers DADDR(0)
Buffer 0
Destination Buffers
If the user needs to execute a DMAC transfer where the source and destination address are contiguous but the amount of data to be transferred is greater than the maximum buffer size DMAC_CTRLAx.BTSIZE, then this can be achieved using the type of multi-buffer transfer as shown in Figure 30-8 on page 468.

467
Figure 30-8. Multi-buffer with Linked Address for Source and Destination Buffers are Contiguous
Buffer 2 DADDR(3) Buffer 2 SADDR(3) Buffer 2 SADDR(2) Buffer 1 SADDR(1) Buffer 0 SADDR(0) Source Buffers Destination Buffers Buffer 0 DADDR(0) Buffer 1 DADDR(1) Buffer 2 DADDR(2)
The DMAC transfer flow is shown in Figure 30-9 on page 469.

468
Figure 30-9. DMAC Transfer Flow for Source and Destination Linked List Address
Channel enabled by software
LLI Fetch
Hardware reprograms SADDRx, DADDRx, CTRLA/Bx, DSCRx
DMAC buffer transfer
Writeback of DMAC_CTRLAx register in system memory Chained Buffer Transfer Completed Interrupt generated here
Is DMAC in Row 1 of DMAC State Machine Table?
no
DMAC Chained Buffer Transfer Completed Interrupt generated here
yes
Channel disabled by hardware

469
Multi-buffer Transfer with Source Address Auto-reloaded and Destination Address Auto-reloaded (Row 10)
1. 2. Read the Channel Handler Status register to choose an available (disabled) channel. Clear any pending interrupts on the channel from the previous DMAC transfer by reading the interrupt status register. Program the following channel registers: 1. 2. 3. 4. Write the starting source address in the DMAC_SADDRx register for channel x. Write the starting destination address in the DMAC_DADDRx register for channel x. Program DMAC_CTRLAx, DMAC_CTRLBx and DMAC_CFGx according to Row 10 as shown in Table 30-3 on page 463. Program the DMAC_DSCRx register with 0. Write the control information for the DMAC transfer in the DMAC_CTRLAx and DMAC_CTRLBx register for channel x. For example, in the register, you can program the following: i. Set up the transfer type (memory or non-memory peripheral for source and destination) and flow control device by programming the FC of the DMAC_CTRLBx register. ii. Set up the transfer characteristics, such as:
z z z z z z
Transfer width for the source in the SRC_WIDTH field. Transfer width for the destination in the DST_WIDTH field. Source AHB master interface layer in the SIF field where source resides. Destination AHB master interface layer in the DIF field where destination resides. Incrementing/decrementing or fixed address for source in SRC_INCR field. Incrementing/decrementing or fixed address for destination in DST_INCR field.
5. 6. 7.
If source Picture-in-Picture mode is enabled (DMAC_CTRLBx.SPIP is enabled), program the DMAC_SPIPx register for channel x. If destination Picture-in-Picture is enabled (DMAC_CTRLBx.DPIP), program the DMAC_DPIPx register for channel x. Write the channel configuration information into the DMAC_CFGx register for channel x. Ensure that the reload bits, DMAC_CFGx.SRC_REP, DMAC_CFGx.DST_REP and DMAC_CTRLBx.AUTO are enabled. i. Designate the handshaking interface type (hardware or software) for the source and destination peripherals. This is not required for memory. This step requires programming the SRC_H2SEL/DST_h2SEL bits, respectively. Writing a 1 activates the hardware handshaking interface to handle source/destination requests for the specific channel. Writing a 0 activates the software handshaking interface to handle source/destination requests. ii. If the hardware handshaking interface is activated for the source or destination peripheral, assign handshaking interface to the source and destination peripheral. This requires programming the SRC_PER and DST_PER bits, respectively.
3. 4.
After the DMAC selected channel has been programmed, enable the channel by writing a 1 to the DMAC_CHER.ENAx bit where the channel number is. Make sure that bit 0 of the DMAC_EN register is enabled. Source and destination request single and chunk DMAC transactions to transfer the buffer of data (assuming nonmemory peripherals). The DMAC acknowledges on completion of each chunk/single transaction and carries out the buffer transfer. When the buffer transfer has completed, the DMAC reloads the DMAC_SADDRx, DMAC_DADDRx and DMAC_CTRLAx registers. The hardware sets the Buffer Transfer Completed Interrupt. The DMAC then samples the row number as shown in Table 30-3 on page 463. If the DMAC is in Row 1, then the DMAC transfer has completed. The hardware sets the Chained Buffer Transfer Completed Interrupt and disables the channel. So you can either respond to the Buffer Transfer Completed Interrupt or Chained Buffer Transfer Completed Interrupt, or poll for the Channel Enable in the Channel Status Register (DMAC_CHSR.ENAx) until it is disabled, to detect when the transfer is complete. If the DMAC is not in Row 1, the next step is performed.
5.

470
6.
The DMAC transfer proceeds as follows: 1. If the Buffer Transfer Completed Interrupt is unmasked (DMAC_EBCIMR.BTCx = 1, where x is the channel number), the hardware sets the Buffer Transfer Completed Interrupt when the buffer transfer has completed. It then stalls until the STALx bit of DMAC_CHSR register is cleared by software, writing 1 to DMAC_CHER.KEEPx bit, where x is the channel number. If the next buffer is to be the last buffer in the DMAC transfer, then the buffer complete ISR (interrupt service routine) should clear the automatic mode bit in the DMAC_CTRLBx.AUTO bit. This puts the DMAC into Row 1 as shown in Table 30-3 on page 463. If the next buffer is not the last buffer in the DMAC transfer, then the reload bits should remain enabled to keep the DMAC in Row 4. If the Buffer Transfer Completed Interrupt is masked (DMAC_EBCIMR.BTCx = 0, where x is the channel number), the hardware does not stall until it detects a write to the Buffer Transfer Completed Interrupt Enable register DMAC_EBCIER register, but starts the next buffer transfer immediately. In this case, the software must clear the automatic mode bit in the DMAC_CTRLB to put the DMAC into ROW 1 of Table 303 on page 463 before the last buffer of the DMAC transfer has completed. The transfer is similar to that shown in Figure 30-10 on page 471. The DMAC transfer flow is shown in Figure 30-11 on page 472.
2.
Figure 30-10. Multi-buffer DMAC Transfer with Source and Destination Address Auto-reloaded
Address of Source Layer Address of Destination Layer
Block0 Block1 Block2
SADDR
DADDR
BlockN Source Buffers Destination Buffers

471
Figure 30-11. DMAC Transfer Flow for Source and Destination Address Auto-reloaded
Buffer Transfer
Replay mode for SADDRx, DADDRx, CTRLAx, CTRLBx Buffer Transfer Completed Interrupt generated here DMAC Chained Buffer Transfer Completed Interrupt generated here
yes
Is DMAC in Row 1 of DMAC State Machine table?
no
EBCIMR[x]=1?
no
yes
Stall until STALLx is cleared by writing to KEEPx field

472
Multi-buffer Transfer with Source Address Auto-reloaded and Linked List Destination Address (Row 6)
1. 2. Read the Channel Handler Status register to choose a free (disabled) channel. Set up the chain of linked list items (otherwise known as buffer descriptors) in memory. Write the control information in the LLI.DMAC_CTRLAx and DMAC_CTRLBx registers location of the buffer descriptor for each LLI in memory for channel x. For example, in the register, you can program the following: 1. 2. Set up the transfer type (memory or non-memory peripheral for source and destination) and flow control peripheral by programming the FC of the DMAC_CTRLBx register. Set up the transfer characteristics, such as: i. Transfer width for the source in the SRC_WIDTH field. ii. Transfer width for the destination in the DST_WIDTH field. iii. Source AHB master interface layer in the SIF field where source resides. iv. Destination AHB master interface layer in the DIF field where destination resides. v. Incrementing/decrementing or fixed address for source in SRC_INCR field. vi. Incrementing/decrementing or fixed address for destination DST_INCR field. 3. Note: 4. Write the starting source address in the DMAC_SADDRx register for channel x. The values in the LLI.DMAC_SADDRx register locations of each of the Linked List Items (LLIs) set up in memory, although fetched during an LLI fetch, are not used. Write the channel configuration information into the DMAC_CFGx register for channel x. 1. Designate the handshaking interface type (hardware or software) for the source and destination peripherals. This is not required for memory. This step requires programming the SRC_H2SEL/DST_H2SEL bits, respectively. Writing a 1 activates the hardware handshaking interface to handle source/destination requests for the specific channel. Writing a 0 activates the software handshaking interface source/destination requests. If the hardware handshaking interface is activated for the source or destination peripheral, assign handshaking interface to the source and destination peripheral. This requires programming the SRC_PER and DST_PER bits, respectively.
2.
5.
Make sure that the LLI.DMAC_CTRLBx register locations of all LLIs in memory (except the last one) are set as shown in Row 6 of Table 30-3 on page 463 while the LLI.DMAC_CTRLBx register of the last Linked List item must be set as described in Row 1 of Table 30-3. Figure 30-5 on page 462 shows a Linked List example with two list items. Make sure that the LLI.DMAC_DSCRx register locations of all LLIs in memory (except the last one) are non-zero and point to the next Linked List Item. Make sure that the LLI.DMAC_DADDRx register locations of all LLIs in memory point to the start destination buffer address proceeding that LLI fetch. Make sure that the LLI.DMAC_CTLx.DONE field of the LLI.DMAC_CTRLA register locations of all LLIs in memory is cleared. If source Picture-in-Picture is enabled (DMAC_CTRLBx.SPIP is enabled), program the DMAC_SPIPx register for channel x.
6. 7. 8. 9.
10. If destination Picture-in-Picture is enabled (DMAC_CTRLBx.DPIP is enabled), program the DMAC_DPIPx register for channel x. 11. Clear any pending interrupts on the channel from the previous DMAC transfer by reading to the DMAC_EBCISR register. 12. Program the DMAC_CTLx and DMAC_CFGx registers according to Row 6 as shown in Table 30-3 on page 463. 13. Program the DMAC_DSCRx register with DMAC_DSCRx(0), the pointer to the first Linked List item. 14. Finally, enable the channel by writing a 1 to the DMAC_CHER.ENAx bit, where x is the channel number. The transfer is performed. Make sure that bit 0 of the DMAC_EN register is enabled.

473
15. The DMAC fetches the first LLI from the location pointed to by DMAC_DSCRx(0). Note: The LLI.DMAC_SADDRx, LLI.DMAC_DADDRx, LLI. DMAC_LLPx LLI.DMAC_CTRLAx and LLI.DMAC_CTRLBx registers are fetched. The LLI.DMAC_SADDRx register, although fetched, is not used.
16. Source and destination request single and chunk DMAC transactions to transfer the buffer of data (assuming nonmemory peripherals). DMAC acknowledges at the completion of every transaction (chunk and single) in the buffer and carries out the buffer transfer. 17. The DMAC_CTRLAx register is written out to the system memory. The DMAC_CTRLAx register is written out to the same location on the same layer (DMAC_DSCRx.DSCR_IF) where it was originally fetched, that is the location of the DMAC_CTRLAx register of the linked list item fetched prior to the start of the buffer transfer. Only DMAC_CTRLAx register is written out, because only the DMAC_CTRLAx.BTSIZE and DMAC_CTRLAx.DONE fields have been updated by hardware within the DMAC. The LLI.DMAC_CTRLAx.DONE bit is asserted to indicate buffer completion. Therefore, the software can poll the LLI.DMAC_CTRLAx.DONE field of the DMAC_CTRLAx register in the LLi to ascertain when a buffer transfer has completed. Note: Do not poll the DMAC_CTRLAx.DONE bit in the DMAC memory map. Instead, poll the LLI.DMAC_CTRLAx.DONE bit in the LLI for that buffer. If the polled LLI.DMAC_CTRLAx.DONE bit is asserted, then this buffer transfer has completed. This LLI.DMAC_CTRLA.DONE bit was cleared at the start of the transfer.
18. The DMAC reloads the DMAC_SADDRx register from the initial value. The hardware sets the Buffer Transfer Completed Interrupt. The DMAC samples the row number as shown in Table 30-3 on page 463. If the DMAC is in Row 1, then the DMAC transfer has completed. The hardware sets the Chained Buffer Transfer Completed Interrupt and disables the channel. You can either respond to the Buffer Transfer Completed Interrupt or Chained Buffer Transfer Completed Interrupt, or poll for the Channel Enable. (DMAC_CHSR.ENAx) bit until it is cleared by hardware, to detect when the transfer is complete. If the DMAC is not in Row 1 as shown in Table 30-3 on page 463, the following step is performed. 19. The DMAC fetches the next LLI from the memory location pointed to by the current DMAC_DSCRx register, and automatically reprograms the DMAC_DADDRx, DMAC_CTRLAx, DMAC_CTRLBx and DMAC_DSCRx channel registers. Note that the DMAC_SADDRx is not re-programmed as the reloaded value is used for the next DMAC buffer transfer. If the next buffer is the last buffer of the DMAC transfer, then the DMAC_CTRLBx and DMAC_DSCRx registers just fetched from the LLI should match Row 1 of Table 30-3 on page 463. The DMAC transfer might look like that shown in Figure 30-12 on page 475.

474
Figure 30-12. Multi-buffer DMAC Transfer with Source Address Auto-reloaded and Linked List Destination Address
Buffer0 DADDR(0)
Buffer1 DADDR(1) SADDR Buffer2 DADDR(2)
BufferN DADDR(N)
Source Buffers
Destination Buffers
The DMAC Transfer flow is shown in Figure 30-13 on page 476.

475
Figure 30-13. DMAC Transfer Flow for Replay Mode at Source and Linked List Destination Address
LLI Fetch
Hardware reprograms DADDRx, CTRLAx, CTRLBx, DSCRx
Writeback of control status information in LLI
Reload SADDRx Buffer Transfer Completed Interrupt generated here yes DMAC Chained Buffer Transfer Completed Interrupt generated here Is DMAC in Row 1 of DMAC State Machine Table?
no

476
Multi-buffer Transfer with Source Address Auto-reloaded and Contiguous Destination Address (Row 11)
1. 2. 3. Read the Channel Handler Status register to choose a free (disabled) channel. Clear any pending interrupts on the channel from the previous DMAC transfer by reading to the Interrupt Status Register. Program the following channel registers: 1. 2. 3. Write the starting source address in the DMAC_SADDRx register for channel x. Write the starting destination address in the DMAC_DADDRx register for channel x. Program DMAC_CTRLAx, DMAC_CTRLBx and DMAC_CFGx according to Row 11 as shown in Table 30-3 on page 463. Program the DMAC_DSCRx register with 0. DMAC_CTRLBx.AUTO field is set to 1 to enable automatic mode support. Write the control information for the DMAC transfer in the DMAC_CTRLBx and DMAC_CTRLAx register for channel x. For example, in this register, you can program the following: i. Set up the transfer type (memory or non-memory peripheral for source and destination) and flow control device by programming the FC of the DMAC_CTRLBx register. ii. Set up the transfer characteristics, such as:
z z z z z z
4.
Transfer width for the source in the SRC_WIDTH field. Transfer width for the destination in the DST_WIDTH field. Source AHB master interface layer in the SIF field where source resides. Destination AHB master interface master layer in the DIF field where destination resides. Incrementing/decrementing or fixed address for source in SRC_INCR field. Incrementing/decrementing or fixed address for destination in DST_INCR field.
5. 6. 7.
If source Picture-in-Picture is enabled (DMAC_CTRLBx.SPIP is enabled), program the DMAC_SPIPx register for channel x. If destination Picture-in-Picture is enabled (DMAC_CTRLBx.DPIP), program the DMAC_DPIPx register for channel x. Write the channel configuration information into the DMAC_CFGx register for channel x. i. Designate the handshaking interface type (hardware or software) for the source and destination peripherals. This is not required for memory. This step requires programming the SRC_H2SEL/DST_H2SEL bits, respectively. Writing a 1 activates the hardware handshaking interface to handle source/destination requests for the specific channel. Writing a 0 activates the software handshaking interface to handle source/destination requests. ii. If the hardware handshaking interface is activated for the source or destination peripheral, assign the handshaking interface to the source and destination peripheral. This requires programming the SRC_PER and DST_PER bits, respectively.
4. 5.
After the DMAC channel has been programmed, enable the channel by writing a 1 to the DMAC_CHER.ENAx bit, where x is the channel number. Make sure that bit 0 of the DMAC_EN.ENABLE register is enabled. Source and destination request single and chunk DMAC transactions to transfer the buffer of data (assuming nonmemory peripherals). The DMAC acknowledges at the completion of every transaction (chunk and single) in the buffer and carries out the buffer transfer. When the buffer transfer has completed, the DMAC reloads the DMAC_SADDRx register. The DMAC_DADDRx register remains unchanged. The hardware sets the Buffer Transfer Completed Interrupt. The DMAC then samples the row number as shown in Table 30-3 on page 463. If the DMAC is in Row 1, then the DMAC transfer has completed. The hardware sets the Chained Buffer Transfer Completed Interrupt and disables the channel. So you can either respond to the Buffer Transfer Completed Interrupt or Chained Buffer Transfer Completed Interrupt, or poll for the enable (ENAx) field in the Channel Status Register (DMAC_CHSR.ENAx bit) until it is cleared by hardware, to detect when the transfer is complete. If the DMAC is not in Row 1, the next step is performed.
6.

477
7.
The DMAC transfer proceeds as follows: 1. If the Buffer Transfer Completed Interrupt is unmasked (DMAC_EBCIMR.BTCx = 1, where x is the channel number), the hardware sets the Buffer Transfer Completed Interrupt when the buffer transfer has completed. It then stalls until STALx bit of DMAC_CHSR is cleared by writing in the KEEPx field of DMAC_CHER register, where x is the channel number. If the next buffer is to be the last buffer in the DMAC transfer, then the buffer complete ISR (interrupt service routine) should clear the automatic mode bit, DMAC_CTRLBx.AUTO. This puts the DMAC into Row 1 as shown in Table 30-3 on page 463. If the next buffer is not the last buffer in the DMAC transfer, then the automatic transfer mode bit should remain enabled to keep the DMAC in Row 11 as shown in Table 30-3 on page 463. If the Buffer Transfer Completed Interrupt is masked (DMAC_EBCIMR.BTCx = 0, where x is the channel number), the hardware does not stall until it detects a write to the Buffer Transfer Completed Interrupt Enable register, but starts the next buffer transfer immediately. In this case, the software must clear the automatic mode bit, DMAC_CTRLBx.AUTO, to put the device into ROW 1 of Table 30-3 on page 463 before the last buffer of the DMAC transfer has completed.
2.
The transfer is similar to that shown in Figure 30-14 on page 478. The DMAC Transfer flow is shown in Figure 30-15 on page 479.
Figure 30-14. Multi-buffer Transfer with Source Address Auto-reloaded and Contiguous Destination Address
Buffer2 DADDR(2) Buffer1 DADDR(1) Buffer0 SADDR DADDR(0)
Source Buffers
Destination Buffers

478
Figure 30-15. DMAC Transfer Replay Mode is Enabled for the Source and Contiguous Destination Address
Buffer Transfer
Replay mode for SADDRx, Contiguous mode for DADDRx CTRLAx, CTRLBx Buffer Transfer Completed Interrupt generated here Buffer Transfer Completed Interrupt generated here Is DMAC in Row 1 of DMAC State Machine Table?
yes
no
no DMA_EBCIMR[x]=1?
yes Stall until STALLx field is cleared by software writing KEEPx field
Multi-buffer DMAC Transfer with Linked List for Source and Contiguous Destination Address (Row 2)
1. 2. Read the Channel Handler Status register to choose a free (disabled) channel. Set up the linked list in memory. Write the control information in the LLI.DMAC_CTRLAx and LLI.DMAC_CTRLBx register location of the buffer descriptor for each LLI in memory for channel x. For example, in the register, you can program the following: 1. 2. Set up the transfer type (memory or non-memory peripheral for source and destination) and flow control device by programming the FC of the DMAC_CTRLBx register. Set up the transfer characteristics, such as: i. Transfer width for the source in the SRC_WIDTH field. ii. Transfer width for the destination in the DST_WIDTH field. iii. Source AHB master interface layer in the SIF field where source resides. iv. Destination AHB master interface layer in the DIF field where destination resides. v. Incrementing/decrementing or fixed address for source in SRC_INCR field. vi. Incrementing/decrementing or fixed address for destination DST_INCR field.

479
3. Note: 4.
Write the starting destination address in the DMAC_DADDRx register for channel x. The values in the LLI.DMAC_DADDRx register location of each Linked List Item (LLI) in memory, although fetched during an LLI fetch, are not used. Write the channel configuration information into the DMAC_CFGx register for channel x. 1. Designate the handshaking interface type (hardware or software) for the source and destination peripherals. This is not required for memory. This step requires programming the SRC_H2SEL/DST_H2SEL bits, respectively. Writing a 1 activates the hardware handshaking interface to handle source/destination requests for the specific channel. Writing a 0 activates the software handshaking interface to handle source/destination requests. If the hardware handshaking interface is activated for the source or destination peripheral, assign the handshaking interface to the source and destination peripherals. This requires programming the SRC_PER and DST_PER bits, respectively.
2.
5.
Make sure that all LLI.DMAC_CTRLBx register locations of the LLI (except the last) are set as shown in Row 2 of Table 30-3 on page 463, while the LLI.DMAC_CTRLBx register of the last Linked List item must be set as described in Row 1 of Table 30-3. Figure 30-5 on page 462 shows a Linked List example with two list items. Make sure that the LLI.DMAC_DSCRx register locations of all LLIs in memory (except the last) are non-zero and point to the next Linked List Item. Make sure that the LLI.DMAC_SADDRx register locations of all LLIs in memory point to the start source buffer address proceeding that LLI fetch. Make sure that the LLI.DMAC_CTRLAx.DONE field of the LLI.DMAC_CTRLAx register locations of all LLIs in memory is cleared. If source Picture-in-Picture is enabled (DMAC_CTRLBx.SPIP is enabled), program the DMAC_SPIPx register for channel x.
6. 7. 8. 9.
10. If destination Picture-in-Picture is enabled (DMAC_CTRLBx.DPIP is enabled), program the DMAC_DPIPx register for channel x. 11. Clear any pending interrupts on the channel from the previous DMAC transfer by reading the interrupt status register. 12. Program the DMAC_CTRLAx, DMAC_CTRLBx and DMAC_CFGx registers according to Row 2 as shown in Table 30-3 on page 463 13. Program the DMAC_DSCRx register with DMAC_DSCRx(0), the pointer to the first Linked List item. 14. Finally, enable the channel by writing a 1 to the DMAC_CHER.ENAx bit. The transfer is performed. Make sure that bit 0 of the DMAC_EN register is enabled. 15. The DMAC fetches the first LLI from the location pointed to by DMAC_DSCRx(0). Note: The LLI.DMAC_SADDRx, LLI.DMAC_DADDRx, LLI.DMAC_DSCRx and LLI.DMAC_CTRLA/Bx registers are fetched. The LLI.DMAC_DADDRx register location of the LLI, although fetched, is not used. The DMAC_DADDRx register in the DMAC remains unchanged.
16. Source and destination requests single and chunk DMAC transactions to transfer the buffer of data (assuming non-memory peripherals). The DMAC acknowledges at the completion of every transaction (chunk and single) in the buffer and carries out the buffer transfer. 17. Once the buffer of data is transferred, the DMAC_CTRLAx register is written out to the system memory at the same location and on the same layer (DMAC_DSCRx.DSCR_IF) where it was originally fetched, that is, the location of the DMAC_CTRLAx register of the linked list item fetched prior to the start of the buffer transfer. Only DMAC_CTRLAx register is written out because only the DMAC_CTRLAx.BTSIZE and DMAC_CTRLAX.DONE fields have been updated by DMAC hardware. Additionally, the DMAC_CTRLAx.DONE bit is asserted when the buffer transfer has completed. Note: Do not poll the DMAC_CTRLAx.DONE bit in the DMAC memory map. Instead, poll the LLI.DMAC_CTRLAx.DONE bit in the LLI for that buffer. If the poll LLI.DMAC_CTRLAx.DONE bit is asserted, then this buffer transfer has completed. This LLI.DMAC_CTRLAx.DONE bit was cleared at the start of the transfer.

480
18. The DMAC does not wait for the buffer interrupt to be cleared, but continues and fetches the next LLI from the memory location pointed to by the current DMAC_DSCRx register, then automatically reprograms the DMAC_SADDRx, DMAC_CTRLAx, DMAC_CTRLBx and DMAC_DSCRx channel registers. The DMAC_DADDRx register is left unchanged. The DMAC transfer continues until the DMAC samples the DMAC_CTRLAx, DMAC_CTRLBx and DMAC_DSCRx registers at the end of a buffer transfer match that described in Row 1 of Table 30-3 on page 463. The DMAC then knows that the previous buffer transferred was the last buffer in the DMAC transfer. The DMAC transfer might look like that shown in Figure 30-16 on page 481. Note that the destination address is decrementing.
Figure 30-16. DMAC Transfer with Linked List Source Address and Contiguous Destination Address
Buffer 2 SADDR(2) Buffer 2 DADDR(2)
Buffer 1 SADDR(1)
Buffer 1 DADDR(1)
Buffer 0 Buffer 0 SADDR(0) DADDR(0)
Source Buffers
The DMAC transfer flow is shown in Figure 30-17 on page 482.
Destination Buffers

481
Figure 30-17. DMAC Transfer Flow for Linked List Source Address and Contiguous Destination Address
LLI Fetch
Hardware reprograms SADDRx, CTRLAx,CTRLBx, DSCRx
Writeback of control information of LLI
Buffer Transfer Completed Interrupt generated here
Is DMAC in Row 1 ?
no
DMAC Chained Buffer Transfer Completed Interrupt generated here
yes

482
30.4.6 Disabling a Channel Prior to Transfer Completion

Under normal operation, the software enables a channel by writing a 1 to the Channel Handler Enable Register, DMAC_CHER.ENAx, and the hardware disables a channel on transfer completion by clearing the DMAC_CHSR.ENAx register bit. The recommended way for software to disable a channel without losing data is to use the SUSPx bit in conjunction with the EMPTx bit in the Channel Handler Status Register. 1. If the software wishes to disable a channel n prior to the DMAC transfer completion, then it can set the DMAC_CHER.SUSPx bit to tell the DMAC to halt all transfers from the source peripheral. Therefore, the channel FIFO receives no new data. The software can now poll the DMAC_CHSR.EMPTx bit until it indicates that the channel n FIFO is empty, where n is the channel number. The DMAC_CHER.ENAx bit can then be cleared by software once the channel n FIFO is empty, where n is the channel number.
2. 3.
When DMAC_CTRLAx.SRC_WIDTH is less than DMAC_CTRLAx.DST_WIDTH and the DMAC_CHSRx.SUSPx bit is high, the DMAC_CHSRx.EMPTx is asserted once the contents of the FIFO does not permit a single word of DMAC_CTRLAx.DST_WIDTH to be formed. However, there may still be data in the channel FIFO but not enough to form a single transfer of DMAC_CTLx.DST_WIDTH width. In this configuration, once the channel is disabled, the remaining data in the channel FIFO are not transferred to the destination peripheral. It is permitted to remove the channel from the suspension state by writing a 1 to the DMAC_CHER.RESx field register. The DMAC transfer completes in the normal manner. n defines the channel number. Note: If a channel is disabled by software, an active single or chunk transaction is not guaranteed to receive an acknowledgement.
30.4.6.1 Abnormal Transfer Termination A DMAC transfer may be terminated abruptly by software by clearing the channel enable bit, DMAC_CHDR.ENAx, where x is the channel number. This does not mean that the channel is disabled immediately after the DMAC_CHSR.ENAx bit is cleared over the APB interface. Consider this as a request to disable the channel. The DMAC_CHSR.ENAx must be polled and then it must be confirmed that the channel is disabled by reading back 0. The software may terminate all channels abruptly by clearing the global enable bit in the DMAC Configuration Register (DMAC_EN.ENABLE bit). Again, this does not mean that all channels are disabled immediately after the DMAC_EN.ENABLE is cleared over the APB slave interface. Consider this as a request to disable all channels. The DMAC_CHSR.ENABLE must be polled and then it must be confirmed that all channels are disabled by reading back 0. Note: If the channel enable bit is cleared while there is data in the channel FIFO, this data is not sent to the destination peripheral and is not present when the channel is re-enabled. For read sensitive source peripherals, such as a source FIFO, this data is therefore lost. When the source is not a read sensitive device (i.e., memory), disabling a channel without waiting for the channel FIFO to empty may be acceptable as the data is available from the source peripheral upon request and is not lost. If a channel is disabled by software, an active single or chunk transaction is not guaranteed to receive an acknowledgement.
Note:

483
30.5
DMAC Software Requirements

z
There must not be any write operation to Channel registers in an active channel after the channel enable is made HIGH. If any channel parameters must be reprogrammed, this can only be done after disabling the DMAC channel. You must program the DMAC_SADDRx and DMAC_DADDRx channel registers with a byte, half-word and word aligned address depending on the source width and destination width. After the software disables a channel by writing into the channel disable register, it must re-enable the channel only after it has polled a 0 in the corresponding channel enable status register. This is because the current AHB Burst must terminate properly. If you program the BTSIZE field in the DMAC_CTRLA as zero, and the DMAC has been defined as the flow controller, then the channel is automatically disabled. When hardware handshaking interface protocol is fully implemented, a peripheral is expected to deassert any sreq or breq signals on receiving the ack signal irrespective of the request the ack was asserted in response to. Multiple Transfers involving the same peripheral must not be programmed and enabled on different channels, unless this peripheral integrates several hardware handshaking interfaces. When a Peripheral has been defined as the flow controller, the targeted DMAC Channel must be enabled before the Peripheral. If you do not ensure this and the First DMAC request is also the last transfer, the DMAC Channel might miss a Last Transfer Flag. When the AUTO Field is set to TRUE, then the BTSIZE Field is automatically reloaded from its previous value. BTSIZE must be initialized to a non zero value if the first transfer is initiated with the AUTO field set to TRUE, even if LLI mode is enabled, because the LLI fetch operation will not update this field.
z z
z z z z
30.6
Write Protection Registers

To prevent any single software error that may corrupt the DMAC behavior, the DMAC address space can be writeprotected by setting the WPEN bit in the DMAC Write Protect Mode Register (DMAC_WPMR). If a write access to anywhere in the DMAC address space is detected, then the WPVS flag in the DMAC Write Protect Status Register (MCI_WPSR) is set, and the WPVSRC field indicates in which register the write access has been attempted. The WPVS flag is reset by writing the DMAC Write Protect Mode Register (DMAC_WPMR) with the appropriate access key, WPKEY. The protected registers are:
z z z z z z z z
DMAC Global Configuration Register on page 486 DMAC Enable Register on page 487 DMAC Channel x [x = 0..7] Source Address Register on page 498 DMAC Channel x [x = 0..7] Destination Address Register on page 499 DMAC Channel x [x = 0..7] Descriptor Address Register on page 500 DMAC Channel x [x = 0..7] Control A Register on page 501 DMAC Channel x [x = 0..7] Control B Register on page 502 DMAC Channel x [x = 0..7] Configuration Register on page 505

484
30.7
Offset 0x000 0x004 0x008 0x00C 0x010 0x014 0x018 0x01C 0x020 0x024 0x028 0x02C 0x030 0x034 0x038
DMA Controller (DMAC) User Interface

Register DMAC Global Configuration Register DMAC Enable Register DMAC Software Single Request Register DMAC Software Chunk Transfer Request Register DMAC Software Last Transfer Flag Register Reserved DMAC Error, Chained Buffer Transfer Completed Interrupt and Buffer Transfer Completed Interrupt Enable register. DMAC Error, Chained Buffer Transfer Completed Interrupt and Buffer Transfer Completed Interrupt Disable register. DMAC Error, Chained Buffer Transfer Completed Interrupt and Buffer transfer completed Mask Register. DMAC Error, Chained Buffer Transfer Completed Interrupt and Buffer transfer completed Status Register. DMAC Channel Handler Enable Register DMAC Channel Handler Disable Register DMAC Channel Handler Status Register Reserved Reserved DMAC Channel Source Address Register DMAC Channel Destination Address Register DMAC Channel Descriptor Address Register DMAC Channel Control A Register DMAC Channel Control B Register DMAC Channel Configuration Register DMAC Channel Source Picture-in-Picture Configuration Register DMAC Channel Destination Picture-in-Picture Configuration Register Reserved Reserved DMAC Write Protect Mode Register DMAC Write Protect Status Register Reserved DMAC_EBCIER DMAC_EBCIDR DMAC_EBCIMR DMAC_EBCISR DMAC_CHER DMAC_CHDR DMAC_CHSR DMAC_SADDR DMAC_DADDR DMAC_DSCR DMAC_CTRLA DMAC_CTRLB DMAC_CFG DMAC_SPIP DMAC_DPIP DMAC_WPMR DMAC_WPSR Write-only Write-only Read-only Read-only Write-only Write-only Read-only Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-only 0x0 0x0 0x00FF0000 0x0 0x0 0x0 0x0 0x0 0x01000000 0x0 0x0 0x0 0x0 Name DMAC_GCFG DMAC_EN DMAC_SREQ DMAC_CREQ DMAC_LAST Access Read-write Read-write Read-write Read-write Read-write Reset 0x10 0x0 0x0 0x0 0x0
Table 30-5. Register Mapping
0x03C+ch_num*(0x28)+(0x0) 0x03C+ch_num*(0x28)+(0x4) 0x03C+ch_num*(0x28)+(0x8) 0x03C+ch_num*(0x28)+(0xC) 0x03C+ch_num*(0x28)+(0x10) 0x03C+ch_num*(0x28)+(0x14) 0x03C+ch_num*(0x28)+(0x18) 0x03C+ch_num*(0x28)+(0x1C) 0x03C+ch_num*(0x28)+(0x20) 0x03C+ch_num*(0x28)+(0x24) 0x1E4 0x1E8 0x01EC- 0x1FC

485
30.7.1 DMAC Global Configuration Register

31 23 15 7
DMAC_GCFG 0xFFFFE600 (0), 0xFFFFE800 (1) Read-write 0x00000010

30 22 14 6 29 21 13 5 28 20 12 4 ARB_CFG 27 19 11 3 26 18 10 2 25 17 9 1 24 16 8 DICEN 0
Note: Bit fields 0, 1, 2, 3, have a default value of 0. This should not be changed. This register can only be written if the WPEN bit is cleared in DMAC Write Protect Mode Register .
ARB_CFG: Arbiter Configuration

0 (FIXED): Fixed priority arbiter. 1 (ROUND_ROBIN): Modified round robin arbiter.
DICEN: Descriptor Integrity Check

0: Descriptor Integrity Check Interface is Disabled. 1: Descriptor Integrity Check Interface is Enabled.

486
30.7.2 DMAC Enable Register

31 23 15 7
DMAC_EN 0xFFFFE604 (0), 0xFFFFE804 (1) Read-write 0x00000000

30 22 14 6 29 21 13 5 28 20 12 4 27 19 11 3 26 18 10 2 25 17 9 1 24 16 8 0 ENABLE
This register can only be written if the WPEN bit is cleared in DMAC Write Protect Mode Register .
ENABLE: General Enable of DMA

0: DMA Controller is disabled. 1: DMA Controller is enabled.

487
30.7.3 DMAC Software Single Request Register

31 23 15 DSREQ7 7 DSREQ3
DMAC_SREQ 0xFFFFE608 (0), 0xFFFFE808 (1) Read-write 0x00000000

30 22 14 SSREQ7 6 SSREQ3 29 21 13 DSREQ6 5 DSREQ2 28 20 12 SSREQ6 4 SSREQ2 27 19 11 DSREQ5 3 DSREQ1 26 18 10 SSREQ5 2 SSREQ1 25 17 9 DSREQ4 1 DSREQ0 24 16 8 SSREQ4 0 SSREQ0
DSREQx: Destination Request

Request a destination single transfer on channel i.
SSREQx: Source Request

Request a source single transfer on channel i.

488
30.7.4 DMAC Software Chunk Transfer Request Register

31 23 15 DCREQ7 7 DCREQ3
DMAC_CREQ 0xFFFFE60C (0), 0xFFFFE80C (1) Read-write 0x00000000

30 22 14 SCREQ7 6 SCREQ3 29 21 13 DCREQ6 5 DCREQ2 28 20 12 SCREQ6 4 SCREQ2 27 19 11 DCREQ5 3 DCREQ1 26 18 10 SCREQ5 2 SCREQ1 25 17 9 DCREQ4 1 DCREQ0 24 16 8 SCREQ4 0 SCREQ0
DCREQx: Destination Chunk Request

Request a destination chunk transfer on channel i.
SCREQx: Source Chunk Request

Request a source chunk transfer on channel i.

489
30.7.5 DMAC Software Last Transfer Flag Register

31 23 15 DLAST7 7 DLAST3
DMAC_LAST 0xFFFFE610 (0), 0xFFFFE810 (1) Read-write 0x00000000

30 22 14 SLAST7 6 SLAST3 29 21 13 DLAST6 5 DLAST2 28 20 12 SLAST6 4 SLAST2 27 19 11 DLAST5 3 DLAST1 26 18 10 SLAST5 2 SLAST1 25 17 9 DLAST4 1 DLAST0 24 16 8 SLAST4 0 SLAST0
DLASTx: Destination Last

Writing one to DLASTx prior to writing one to DSREQx or DCREQx indicates that this destination request is the last transfer of the buffer.
SLASTx: Source Last

Writing one to SLASTx prior to writing one to SSREQx or SCREQx indicates that this source request is the last transfer of the buffer.

490
30.7.6 DMAC Error, Buffer Transfer and Chained Buffer Transfer Interrupt Enable Register
31 DICERR7 23 ERR7 15 CBTC7 7 BTC7
DMAC_EBCIER 0xFFFFE618 (0), 0xFFFFE818 (1) Write-only 0x00000000

30 DICERR6 22 ERR6 14 CBTC6 6 BTC6 29 DICERR5 21 ERR5 13 CBTC5 5 BTC5 28 DICERR4 20 ERR4 12 CBTC4 4 BTC4 27 DICERR3 19 ERR3 11 CBTC3 3 BTC3 26 DICERR2 18 ERR2 10 CBTC2 2 BTC2 25 DICERR1 17 ERR1 9 CBTC1 1 BTC1 24 DICERR0 16 ERR0 8 CBTC0 0 BTC0
BTCx: Buffer Transfer Completed [7:0]

Buffer Transfer Completed Interrupt Enable Register. Set the relevant bit in the BTC field to enable the interrupt for channel i.
CBTCx: Chained Buffer Transfer Completed [7:0]

Chained Buffer Transfer Completed Interrupt Enable Register. Set the relevant bit in the CBTC field to enable the interrupt for channel i.
ERRx: Access Error [7:0]

Access Error Interrupt Enable Register. Set the relevant bit in the ERR field to enable the interrupt for channel i.
DICERRx: Descriptor Integrity Check Error [7:0]

Descriptor Integrity Check Error Interrupt Enable Register. Set the relevant bit in the DICERR field to enable the interrupt for channel i.

491
30.7.7 DMAC Error, Buffer Transfer and Chained Buffer Transfer Interrupt Disable Register
DMAC_EBCIDR 0xFFFFE61C (0), 0xFFFFE81C (1) Write-only 0x00000000


Buffer transfer completed Disable Interrupt Register. When set, a bit of the BTC field disables the interrupt from the relevant DMAC channel.

Chained Buffer transfer completed Disable Register. When set, a bit of the CBTC field disables the interrupt from the relevant DMAC channel.

Access Error Interrupt Disable Register. When set, a bit of the ERR field disables the interrupt from the relevant DMAC channel.

Descriptor Integrity Check Error Interrupt Disable Register, When set, a bit of the DICERR field disables the interrupt from the relevant DMAC channel.

492
30.7.8 DMAC Error, Buffer Transfer and Chained Buffer Transfer Interrupt Mask Register
DMAC_EBCIMR 0xFFFFE620 (0), 0xFFFFE820 (1) Read-only 0x00000000


0: Buffer Transfer Completed Interrupt is disabled for channel i. 1: Buffer Transfer Completed Interrupt is enabled for channel i.

0: Chained Buffer Transfer interrupt is disabled for channel i. 1: Chained Buffer Transfer interrupt is enabled for channel i.

0: Transfer Error Interrupt is disabled for channel i. 1: Transfer Error Interrupt is enabled for channel i.

0: Descriptor Integrity Check Error Interrupt is disabled for channel i. 1: Descriptor Integrity Check Error Interrupt is enabled for channel i.

493
30.7.9 DMAC Error, Buffer Transfer and Chained Buffer Transfer Status Register
DMAC_EBCISR 0xFFFFE624 (0), 0xFFFFE824 (1) Read-only 0x00000000


When BTC[i] is set, Channel i buffer transfer has terminated.

When CBTC[i] is set, Channel i Chained buffer has terminated. LLI Fetch operation is disabled.

When ERR[i] is set, Channel i has detected an AHB Read or Write Error Access.

When DICERR[i] is set, Channel i has detected a Descriptor Integrity Check Error.

494
30.7.10 DMAC Channel Handler Enable Register

31 KEEP7 23 15 SUSP7 7 ENA7
DMAC_CHER 0xFFFFE628 (0), 0xFFFFE828 (1) Write-only 0x00000000

30 KEEP6 22 14 SUSP6 6 ENA6 29 KEEP5 21 13 SUSP5 5 ENA5 28 KEEP4 20 12 SUSP4 4 ENA4 27 KEEP3 19 11 SUSP3 3 ENA3 26 KEEP2 18 10 SUSP2 2 ENA2 25 KEEP1 17 9 SUSP1 1 ENA1 24 KEEP0 16 8 SUSP0 0 ENA0
ENAx: Enable [7:0]

When set, a bit of the ENA field enables the relevant channel.
SUSPx: Suspend [7:0]

When set, a bit of the SUSP field freezes the relevant channel and its current context.
KEEPx: Keep on [7:0]

When set, a bit of the KEEP field resumes the current channel from an automatic stall state.

495
30.7.11 DMAC Channel Handler Disable Register

31 23 15 RES7 7 DIS7
DMAC_CHDR 0xFFFFE62C (0), 0xFFFFE82C (1) Write-only 0x00000000

30 22 14 RES6 6 DIS6 29 21 13 RES5 5 DIS5 28 20 12 RES4 4 DIS4 27 19 11 RES3 3 DIS3 26 18 10 RES2 2 DIS2 25 17 9 RES1 1 DIS1 24 16 8 RES0 0 DIS0
DISx: Disable [7:0]

Write one to this field to disable the relevant DMAC Channel. The content of the FIFO is lost and the current AHB access is terminated. Software must poll DIS[7:0] field in the DMAC_CHSR register to be sure that the channel is disabled.
RESx: Resume [7:0]

Write one to this field to resume the channel transfer restoring its context.

496
30.7.12 DMAC Channel Handler Status Register

31 STAL7 23 EMPT7 15 SUSP7 7 ENA7
DMAC_CHSR 0xFFFFE630 (0), 0xFFFFE830 (1) Read-only 0x00FF0000

30 STAL6 22 EMPT6 14 SUSP6 6 ENA6 29 STAL5 21 EMPT5 13 SUSP5 5 ENA5 28 STAL4 20 EMPT4 12 SUSP4 4 ENA4 27 STAL3 19 EMPT3 11 SUSP3 3 ENA3 26 STAL2 18 EMPT2 10 SUSP2 2 ENA2 25 STAL1 17 EMPT1 9 SUSP1 1 ENA1 24 STAL0 16 EMPT0 8 SUSP0 0 ENA0
ENAx: Enable [7:0]

A one in any position of this field indicates that the relevant channel is enabled.
SUSPx: Suspend [7:0]

A one in any position of this field indicates that the channel transfer is suspended.
EMPTx: Empty [7:0]

A one in any position of this field indicates that the relevant channel is empty.
STALx: Stalled [7:0]

A one in any position of this field indicates that the relevant channel is stalling.

497
30.7.13 DMAC Channel x [x = 0..7] Source Address Register

Name: Address: DMAC_SADDRx [x = 0..7] 0xFFFFE63C (0)[0], 0xFFFFE664 (0)[1], 0xFFFFE68C (0)[2], 0xFFFFE6B4 (0)[3], 0xFFFFE6DC (0)[4], 0xFFFFE704 (0)[5], 0xFFFFE72C (0)[6], 0xFFFFE754 (0)[7], 0xFFFFE83C (1)[0], 0xFFFFE864 (1)[1], 0xFFFFE88C (1)[2], 0xFFFFE8B4 (1)[3], 0xFFFFE8DC (1)[4], 0xFFFFE904 (1)[5], 0xFFFFE92C (1)[6], 0xFFFFE954 (1)[7] Read-write 0x00000000
30 29 28 SADDR 23 22 21 20 SADDR 15 14 13 12 SADDR 7 6 5 4 SADDR 3 2 1 0 11 10 9 8 19 18 17 16 27 26 25 24
Access: Reset:
31
SADDR: Channel x Source Address

This register must be aligned with the source transfer width.

498
30.7.14 DMAC Channel x [x = 0..7] Destination Address Register

Name: Address: DMAC_DADDRx [x = 0..7] 0xFFFFE640 (0)[0], 0xFFFFE668 (0)[1], 0xFFFFE690 (0)[2], 0xFFFFE6B8 (0)[3], 0xFFFFE6E0 (0)[4], 0xFFFFE708 (0)[5], 0xFFFFE730 (0)[6], 0xFFFFE758 (0)[7], 0xFFFFE840 (1)[0], 0xFFFFE868 (1)[1], 0xFFFFE890 (1)[2], 0xFFFFE8B8 (1)[3], 0xFFFFE8E0 (1)[4], 0xFFFFE908 (1)[5], 0xFFFFE930 (1)[6], 0xFFFFE958 (1)[7] Read-write 0x00000000
30 29 28 DADDR 23 22 21 20 DADDR 15 14 13 12 DADDR 7 6 5 4 DADDR 3 2 1 0 11 10 9 8 19 18 17 16 27 26 25 24
Access: Reset:
31
DADDR: Channel x Destination Address

This register must be aligned with the destination transfer width.

499
30.7.15 DMAC Channel x [x = 0..7] Descriptor Address Register

Name: Address: DMAC_DSCRx [x = 0..7] 0xFFFFE644 (0)[0], 0xFFFFE66C (0)[1], 0xFFFFE694 (0)[2], 0xFFFFE6BC (0)[3], 0xFFFFE6E4 (0)[4], 0xFFFFE70C (0)[5], 0xFFFFE734 (0)[6], 0xFFFFE75C (0)[7], 0xFFFFE844 (1)[0], 0xFFFFE86C (1)[1], 0xFFFFE894 (1)[2], 0xFFFFE8BC (1)[3], 0xFFFFE8E4 (1)[4], 0xFFFFE90C (1)[5], 0xFFFFE934 (1)[6], 0xFFFFE95C (1)[7] Read-write 0x00000000
30 29 28 DSCR 23 22 21 20 DSCR 15 14 13 12 DSCR 7 6 5 DSCR 4 3 2 1 DSCR_IF 0 11 10 9 8 19 18 17 16 27 26 25 24
Access: Reset:
31
DSCR_IF: Descriptor Interface Selection

Value 00 01 10 Name AHB_IF0 AHB_IF1 AHB_IF2 Description The buffer transfer descriptor is fetched via AHB-Lite Interface 0 The buffer transfer descriptor is fetched via AHB-Lite Interface 1 The buffer transfer descriptor is fetched via AHB-Lite Interface 2
DSCR: Buffer Transfer Descriptor Address

This address is word aligned.

500
30.7.16 DMAC Channel x [x = 0..7] Control A Register

Name: Address: DMAC_CTRLAx [x = 0..7] 0xFFFFE648 (0)[0], 0xFFFFE670 (0)[1], 0xFFFFE698 (0)[2], 0xFFFFE6C0 (0)[3], 0xFFFFE6E8 (0)[4], 0xFFFFE710 (0)[5], 0xFFFFE738 (0)[6], 0xFFFFE760 (0)[7], 0xFFFFE848 (1)[0], 0xFFFFE870 (1)[1], 0xFFFFE898 (1)[2], 0xFFFFE8C0 (1)[3], 0xFFFFE8E8 (1)[4], 0xFFFFE910 (1)[5], 0xFFFFE938 (1)[6], 0xFFFFE960 (1)[7] Read-write 0x00000000
30 22 29 DST_WIDTH 21 DCSIZE 13 20 28 27 19 11 BTSIZE 7 6 5 4 BTSIZE 3 2 1 0 26 18 25 SRC_WIDTH 17 SCSIZE 9 16 24
Access: Reset:
31 DONE 23 15
14
12
10
This register can only be written if the WPEN bit is cleared in DMAC Write Protect Mode Register on page 509
BTSIZE: Buffer Transfer Size

The transfer size relates to the number of transfers to be performed, that is, for writes it refers to the number of source width transfers to perform when DMAC is flow controller. For Reads, BTSIZE refers to the number of transfers completed on the Source Interface. When this field is set to 0, the DMAC module is automatically disabled when the relevant channel is enabled.
SCSIZE: Source Chunk Transfer Size.

Value 000 001 010 011 Name CHK_1 CHK_4 CHK_8 CHK_16 Description 1 data transferred 4 data transferred 8 data transferred 16 data transferred
DCSIZE: Destination Chunk Transfer Size

Value 000 001 010 011 Name CHK_1 CHK_4 CHK_8 CHK_16 Description 1 data transferred 4 data transferred 8 data transferred 16 data transferred

501
SRC_WIDTH: Transfer Width for the Source

Value 00 01 10 11 Name BYTE HALF_WORD WORD DWORD Description the transfer size is set to 8-bit width the transfer size is set to 16-bit width the transfer size is set to 32-bit width the transfer size is set to 64-bit width
DST_WIDTH: Transfer Width for the Destination

Value 00 01 10 11 Name BYTE HALF_WORD WORD DWORD Description the transfer size is set to 8-bit width the transfer size is set to 16-bit width the transfer size is set to 32-bit width the transfer size is set to 64-bit width
DONE: Current Descriptor Stop Command and Transfer Completed Memory Indicator
0: The transfer is performed. 1: If SOD field of DMAC_CFG register is set to true, then the DMAC is automatically disabled when an LLI updates the content of this register. The DONE field is written back to memory at the end of the current descriptor transfer.
30.7.17 DMAC Channel x [x = 0..7] Control B Register

Name: DMAC_CTRLBx [x = 0..7]
Address: 0xFFFFE64C (0)[0], 0xFFFFE674 (0)[1], 0xFFFFE69C (0)[2], 0xFFFFE6C4 (0)[3], 0xFFFFE6EC (0)[4], 0xFFFFE714 (0)[5], 0xFFFFE73C (0)[6], 0xFFFFE764 (0)[7], 0xFFFFE84C (1)[0], 0xFFFFE874 (1)[1], 0xFFFFE89C (1)[2], 0xFFFFE8C4 (1)[3], 0xFFFFE8EC (1)[4], 0xFFFFE914 (1)[5], 0xFFFFE93C (1)[6], 0xFFFFE964 (1)[7] Access: Reset:
31 AUTO 23 15 7
Read-write 0x00000000
30 IEN 22 FC 14 6 13 29 DST_INCR 21 20 DST_DSCR 12 DST_PIP 4 DIF 28 27 19 11 3 26 18 10 2 25 SRC_INCR 17 9 1 SIF 16 SRC_DSCR 8 SRC_PIP 0 24

502
SIF: Source Interface Selection Field

Value 00 01 10 Name AHB_IF0 AHB_IF1 AHB_IF2 Description The source transfer is done via AHB_Lite Interface 0 The source transfer is done via AHB_Lite Interface 1 The source transfer is done via AHB_Lite Interface 2
DIF: Destination Interface Selection Field

Value 00 01 10 Name AHB_IF0 AHB_IF1 AHB_IF2 Description The destination transfer is done via AHB_Lite Interface 0 The destination transfer is done via AHB_Lite Interface 1 The destination transfer is done via AHB_Lite Interface 2
SRC_PIP: Source Picture-in-Picture Mode

0 (DISABLE): Picture-in-Picture mode is disabled. The source data area is contiguous. 1 (ENABLE): Picture-in-Picture mode is enabled. When the source PIP counter reaches the programmable boundary, the address is automatically incremented by a user defined amount.
DST_PIP: Destination Picture-in-Picture Mode

0 (DISABLE): Picture-in-Picture mode is disabled. The Destination data area is contiguous. 1 (ENABLE): Picture-in-Picture mode is enabled. When the Destination PIP counter reaches the programmable boundary the address is automatically incremented by a user-defined amount.
SRC_DSCR: Source Address Descriptor

0 (FETCH_FROM_MEM): Source address is updated when the descriptor is fetched from the memory. 1 (FETCH_DISABLE): Buffer Descriptor Fetch operation is disabled for the source.
DST_DSCR: Destination Address Descriptor

0 (FETCH_FROM_MEM): Destination address is updated when the descriptor is fetched from the memory. 1 (FETCH_DISABLE): Buffer Descriptor Fetch operation is disabled for the destination.
FC: Flow Control

This field defines which device controls the size of the buffer transfer, also referred to as the Flow Controller.
Value 00 01 10 11 Name MEM2MEM_DMA_FC MEM2PER_DMA_FC PER2MEM_DMA_FC PER2PER_DMA_FC Description Memory-to-Memory Transfer DMAC is flow controller Memory-to-Peripheral Transfer DMAC is flow controller Peripheral-to-Memory Transfer DMAC is flow controller Peripheral-to-Peripheral Transfer DMAC is flow controller

503
SRC_INCR: Incrementing, Decrementing or Fixed Address for the Source

Value 00 01 10 Name INCREMENTING DECREMENTING FIXED Description The source address is incremented The source address is decremented The source address remains unchanged
DST_INCR: Incrementing, Decrementing or Fixed Address for the Destination

Value 00 01 10 Name INCREMENTING DECREMENTING FIXED Description The destination address is incremented The destination address is decremented The destination address remains unchanged
IEN: Interrupt Enable Not

0: When the buffer transfer is completed, the BTCx flag is set in the EBCISR status register. This bit is active low. 1: When the buffer transfer is completed, the BTCx flag is not set. If this bit is cleared, when the buffer transfer is completed, the BTCx flag is set in the EBCISR status register.
AUTO: Automatic Multiple Buffer Transfer

0 (DISABLE): Automatic multiple buffer transfer is disabled. 1 (ENABLE): Automatic multiple buffer transfer is enabled. This bit enables replay mode or contiguous mode when several buffers are transferred.

504
30.7.18 DMAC Channel x [x = 0..7] Configuration Register

Name: Address: DMAC_CFGx [x = 0..7] 0xFFFFE650 (0)[0], 0xFFFFE678 (0)[1], 0xFFFFE6A0 (0)[2], 0xFFFFE6C8 (0)[3], 0xFFFFE6F0 (0)[4], 0xFFFFE718 (0)[5], 0xFFFFE740 (0)[6], 0xFFFFE768 (0)[7], 0xFFFFE850 (1)[0], 0xFFFFE878 (1)[1], 0xFFFFE8A0 (1)[2], 0xFFFFE8C8 (1)[3], 0xFFFFE8F0 (1)[4], 0xFFFFE918 (1)[5], 0xFFFFE940 (1)[6], 0xFFFFE968 (1)[7] Read-write 0x0100000000
30 22 LOCK_IF_L 29 FIFOCFG 21 LOCK_B 13 DST_H2SEL 5 DST_PER 20 LOCK_IF 12 DST_REP 4 28 27 19 26 25 AHB_PROT 17 9 SRC_H2SEL 1 SRC_PER 24
Access: Reset:
31 23
18
16 SOD 8 SRC_REP 0
15 14 DST_PER_MSB 7 6
11 10 SRC_PER_MSB 3 2
This register can only be written if the WPEN bit is cleared in DMAC Write Protect Mode Register on page 509
SRC_PER: Source with Peripheral identifier

Channel x Source Request is associated with peripheral identifier coded SRC_PER handshaking interface.
DST_PER: Destination with Peripheral identifier

Channel x Destination Request is associated with peripheral identifier coded DST_PER handshaking interface.
SRC_REP: Source Reloaded from Previous

0 (CONTIGUOUS_ADDR): When automatic mode is activated, source address is contiguous between two buffers. 1 (RELOAD_ADDR): When automatic mode is activated, the source address and the control register are reloaded from previous transfer.
SRC_H2SEL: Software or Hardware Selection for the Source

0 (SW): Software handshaking interface is used to trigger a transfer request. 1 (HW): Hardware handshaking interface is used to trigger a transfer request.
SRC_PER_MSB: SRC_PER Most Significant Bits

This field indicates the Most Significant bits of the SRC_PER field.
DST_REP: Destination Reloaded from Previous

0 (CONTIGUOUS_ADDR): When automatic mode is activated, destination address is contiguous between two buffers. 1 (RELOAD_ADDR): When automatic mode is activated, the destination and the control register are reloaded from the previous transfer.
DST_H2SEL: Software or Hardware Selection for the Destination

0 (SW): Software handshaking interface is used to trigger a transfer request. 1 (HW): Hardware handshaking interface is used to trigger a transfer request.

505
DST_PER_MSB: DST_PER Most Significant Bits

This field indicates the Most Significant bits of the DST_PER field.
SOD: Stop On Done

0 (DISABLE): STOP ON DONE disabled, the descriptor fetch operation ignores DONE Field of CTRLA register. 1 (ENABLE): STOP ON DONE activated, the DMAC module is automatically disabled if DONE FIELD is set to 1.
LOCK_IF: Interface Lock

0 (DISABLE): Interface Lock capability is disabled 1 (ENABLE): Interface Lock capability is enabled
LOCK_B: Bus Lock

0 (DISABLE): AHB Bus Locking capability is disabled. 1(ENABLE): AHB Bus Locking capability is enabled.
LOCK_IF_L: Master Interface Arbiter Lock

0 (CHUNK): The Master Interface Arbiter is locked by the channel x for a chunk transfer. 1 (BUFFER): The Master Interface Arbiter is locked by the channel x for a buffer transfer.
AHB_PROT: AHB Protection

AHB_PROT field provides additional information about a bus access and is primarily used to implement some level of protection.
HPROT[3] HPROT[2] HPROT[1] 1 AHB_PROT[0] HPROT[0] Description Data access 0: User Access 1: Privileged Access 0: Not Bufferable 1: Bufferable 0: Not cacheable 1: Cacheable
AHB_PROT[1]
AHB_PROT[2]
FIFOCFG: FIFO Configuration

Value 00 01 10 Name ALAP_CFG HALF_CFG ASAP_CFG Description The largest defined length AHB burst is performed on the destination AHB interface. When half FIFO size is available/filled, a source/destination request is serviced. When there is enough space/data available to perform a single AHB access, then the request is serviced.

506
30.7.19 DMAC Channel x [x = 0..7] Source Picture-in-Picture Configuration Register

Name: Address: DMAC_SPIPx [x = 0..7] 0xFFFFE654 (0)[0], 0xFFFFE67C (0)[1], 0xFFFFE6A4 (0)[2], 0xFFFFE6CC (0)[3], 0xFFFFE6F4 (0)[4], 0xFFFFE71C (0)[5], 0xFFFFE744 (0)[6], 0xFFFFE76C (0)[7], 0xFFFFE854 (1)[0], 0xFFFFE87C (1)[1], 0xFFFFE8A4 (1)[2], 0xFFFFE8CC (1)[3], 0xFFFFE8F4 (1)[4], 0xFFFFE91C (1)[5], 0xFFFFE944 (1)[6], 0xFFFFE96C (1)[7] Read-write 0x00000000
30 22 29 21 28 27 26 18 25 24 SPIP_BOUNDARY 17 16
Access: Reset:
31 23
20 19 SPIP_BOUNDARY 12 SPIP_HOLE 11
15
14
13
10
4 SPIP_HOLE
SPIP_HOLE: Source Picture-in-Picture Hole

This field indicates the value to add to the address when the programmable boundary has been reached.
SPIP_BOUNDARY: Source Picture-in-Picture Boundary

This field indicates the number of source transfers to perform before the automatic address increment operation.

507
30.7.20 DMAC Channel x [x = 0..7] Destination Picture-in-Picture Configuration Register

Name: Address: DMAC_DPIPx [x = 0..7] 0xFFFFE658 (0)[0], 0xFFFFE680 (0)[1], 0xFFFFE6A8 (0)[2], 0xFFFFE6D0 (0)[3], 0xFFFFE6F8 (0)[4], 0xFFFFE720 (0)[5], 0xFFFFE748 (0)[6], 0xFFFFE770 (0)[7], 0xFFFFE858 (1)[0], 0xFFFFE880 (1)[1], 0xFFFFE8A8 (1)[2], 0xFFFFE8D0 (1)[3], 0xFFFFE8F8 (1)[4], 0xFFFFE920 (1)[5], 0xFFFFE948 (1)[6], 0xFFFFE970 (1)[7] Read-write 0x00000000
30 22 29 21 28 27 26 18 25 24 DPIP_BOUNDARY 17 16
Access: Reset:
31 23
20 19 DPIP_BOUNDARY 12 DPIP_HOLE 11
15
14
13
10
4 DPIP_HOLE
DPIP_HOLE: Destination Picture-in-Picture Hole

This field indicates the value to add to the address when the programmable boundary has been reached.
DPIP_BOUNDARY: Destination Picture-in-Picture Boundary

This field indicates the number of source transfers to perform before the automatic address increment operation.

508
30.7.21 DMAC Write Protect Mode Register

31
DMAC_WPMR 0xFFFFE7E4 (0), 0xFFFFE9E4 (1) Read-write See Table 30-5


0 = Disables the Write Protect if WPKEY corresponds to 0x444D41 (DMA in ASCII). 1 = Enables the Write Protect if WPKEY corresponds to 0x444D41 (DMA in ASCII). Protects the registers:
DMAC Global Configuration Register on page 486 DMAC Enable Register on page 487 DMAC Channel x [x = 0..7] Source Address Register on page 498 DMAC Channel x [x = 0..7] Destination Address Register on page 499 DMAC Channel x [x = 0..7] Descriptor Address Register on page 500 DMAC Channel x [x = 0..7] Control A Register on page 501 DMAC Channel x [x = 0..7] Control B Register on page 502 DMAC Channel x [x = 0..7] Configuration Register on page 505 WPKEY: Write Protect KEY
Should be written at value 0x444D41 (DMA in ASCII). Writing any other value in this field aborts the write operation of the WPEN bit. Always reads as 0.

509
30.7.22 DMAC Write Protect Status Register

31 23
DMAC_WPSR 0xFFFFE7E8 (0), 0xFFFFE9E8 (1) Read-only See Table 30-5


0 = No Write Protect Violation has occurred since the last read of the DMAC_WPSR register. 1 = A Write Protect Violation has occurred since the last read of the DMAC_WPSR register. If this violation is an unauthorized attempt to write a protected register, the associated violation is reported into field WPVSRC.

When WPVS is active, this field indicates the write-protected register (through address offset or code) in which a write access has been attempted. Note: Reading DMAC_WPSR automatically clears all fields.

510
31.
31.1
LCD Controller (LCDC)

Description
The LCD Controller (LCDC) consists of logic for transferring LCD image data from an external display buffer to an LCD module. The LCD has one display input buffer per overlay that fetches pixels through the dual AHB master interface and a lookup table to allow palletized display configurations. The LCD controller is programmable on a per overlay basis, and supports different LCD resolutions, window sizes, image formats and pixel depths. The LCD is connected to the ARM Advanced High Performance Bus (AHB) as a master for reading pixel data. It also integrates an APB interface to configure its registers.
31.2
z z z z z z z z z z z z z z z z z z z z z z z z z z z
Dual AHB Master Interface Supports Single Scan Active TFT Display Supports 12-bit, 16-bit, 18-bit and 24-bit Output Mode through the Spatial Dithering Unit Asynchronous Output Mode Supported (at synthesis time) 1, 2, 4, 8 bits per pixel (palletized) 12, 16, 18, 19, 24, 25 and 32 bits per pixel (non palletized) Supports One Base Layer (background) Supports Two Overlay Layer Windows Supports One High End Overlay (HEO) Window Supports One Hardware Cursor, Fixed or Free Size Hardware Cursor Fixed Size on the following patterns: 32x32, 64x64 and 128x128 Little Endian Memory Organization Programmable Timing Engine, with Integer Clock Divider Programmable Polarity for Data, Line Synchro and Frame Synchro. Display Size up to 2048x2048 Color Lookup Table with up to 256 entries and Predefined 8-bit Alpha Programmable Negative and Positive Row Striding for all Layers Programmable Negative and Positive Pixel Striding for all Overlay1, Overlay2 and HEO layers High End Overlay supports 4:2:0 Planar Mode and Semiplanar Mode High End Overlay supports 4:2:2 Planar Mode, Semiplanar Mode and Packed High End Overlay includes Chroma Upsampling Unit Horizontal and Vertical Rescaling unit with Edge Interpolation and Independent Non Integer Ratio Hidden Layer Removal supported. Integrates Fully Programmable Color Space Conversion Overlay1, Overlay2 and High End Overlay Integrate Rotation Engine: 90, 180, 270 Blender Function Supports Arbitrary 8-bit Alpha Value and Chroma Keying DMA User interface uses Linked List Structure and Add-to-queue Structure

511
31.3
Block Diagram

32-bit APB Interface Configuration Registers
SYSCTRL Unit
PP Layer
HCC Layer AHB Bus OVR2 Layer 64-bit Dual AHB Master Interface DEAG Unit CLUT
ROT CLUT
LCD_DAT[23:0] LCD_VSYNC
OVR1 Layer
ROT CLUT GAB Unit LTE Unit
LCD_HSYNC LCD_PCLK LCD_DEN LCD_PWM LCD_DISP
HEO Layer
ROT CSC 2DSC CUE CLUT
Base Layer CLUT
HEO : High End Overlay CUE : Chroma Upsampling Engine CSC : Color Space Conversion 2DSC : Two Dimension Scaler DEAG : DMA Engine Address Generation
HCC: Hardware Cursor Channel GAB : Global Alpha Blender LTE: LCD Timing Engine ROT : Hardware Rotation

512
31.4
Name LCD_PWM LCD_HSYNC LCD_VSYNC LCD_DAT[23:0] LCD_DEN LCD_DISP LCD_PCLK
Description Contrast control signal, using Pulse Width Modulation Horizontal Synchronization Pulse Vertical Synchronization Pulse LCD 24-bit data bus Data Enable Display Enable signal Pixel Clock
Type Output Output Output Output Output Output Output
31.5
31.5.1 I/O Lines

The pins used for interfacing the LCD Controller may be multiplexed with PIO lines. The programmer must first program the PIO Controller to assign the pins to their peripheral function. If I/O lines of the LCD Controller are not used by the application, they can be used for other purposes by the PIO Controller.
Table 31-1. I/O Lines Instance LCDC LCDC LCDC LCDC LCDC LCDC LCDC LCDC LCDC LCDC LCDC LCDC LCDC LCDC LCDC LCDC LCDC LCDC LCDC LCDC Signal LCDDAT0 LCDDAT1 LCDDAT2 LCDDAT3 LCDDAT4 LCDDAT5 LCDDAT6 LCDDAT7 LCDDAT8 LCDDAT9 LCDDAT10 LCDDAT11 LCDDAT12 LCDDAT13 LCDDAT14 LCDDAT15 LCDDAT16 LCDDAT16 LCDDAT17 LCDDAT17 I/O Line PA0 PA1 PA2 PA3 PA4 PA5 PA6 PA7 PA8 PA9 PA10 PA11 PA12 PA13 PA14 PA15 PA16 PC14 PA17 PC13 Peripheral A A A A A A A A A A A A A A A A A C A C

513
Table 31-1. I/O Lines LCDC LCDC LCDC LCDC LCDC LCDC LCDC LCDC LCDC LCDC LCDC LCDC LCDC LCDC LCDC LCDC LCDC LCDC LCDDAT18 LCDDAT18 LCDDAT19 LCDDAT19 LCDDAT20 LCDDAT20 LCDDAT21 LCDDAT21 LCDDAT22 LCDDAT22 LCDDAT23 LCDDAT23 LCDDEN LCDDISP LCDHSYNC LCDPCK LCDPWM LCDVSYNC PA18 PC12 PA19 PC11 PA20 PC10 PA21 PC15 PA22 PE27 PA23 PE28 PA29 PA25 PA27 PA28 PA24 PA26 A C A C A C A C A C A C A A A A A A

The LCD Controller is not continuously clocked. The user must first enable the LCD Controller clock in the Power Management Controller before using it (PMC_PCER).

The LCD Controller interrupt line is connected to one of the internal sources of the Advanced Interrupt Controller. Using the LCD Controller interrupt requires prior programming of the AIC.
Table 31-2. Peripheral IDs Instance LCDC ID 36

514
31.6
The LCD module integrates the following digital blocks:
z z z z z z z z z
DMA Engine Address Generation (DEAG). This block performs data prefetch and requests access to the AHB interface. Input Overlay FIFO stores the stream of pixels. Color Lookup Table (CLUT). These 256 RAM-based lookup table entries are selected when the color depth is set to 1, 2, 4 or 8 bpp. Chroma Upsampling Engine (CUE). This block is selected when the input image sampling format is YUV (YCbCr) 4:2:0 and converts it to higher quality 4:4:4 image. Color Space Conversion (CSC) changes the color spare from YUV to RGB. Two Dimension Scaler (2DSC) resizes the image. Global Alpha Blender (GAB) performs programmable 256 level alpha blending. Output FIFO stores the blended pixel prior to display. LCD Timing Engine provides a fully programmable HSYNC-VSYNC interface.
The DMA controller reads the image through the AHB master interface. The LCD controller engine formats the display data, then the GAB performs alpha blending if required, and writes the final pixel into the output FIFO. The programmable timing engine drives a valid pixel onto the LCD_DAT[23:0] display bus.
31.6.1 Timing Engine Configuration

31.6.1.1 Pixel Clock Period Configuration The pixel clock (PCLK) generated by the timing engine is the source clock (SCLK) divided by the field CLKDIV in the LCDC_LCDCFG0 register. The source clock can be selected between the system clock and the 2x system clock with the field CLKSEL located in the LCDC_LCDCFG0 register. The Pixel Clock period formula is given below:
SCLK PCLK = -------------------------------CLKDIV + 2

The Pixel Clock polarity is also programmable. 31.6.1.2 Horizontal and Vertical Synchronization Configuration The following fields are used to configure the timing engine:
z z z z z z z z
HSPW field VSPW field VFPW field VBPW field HFPW field HBPW field PPL field RPF field
The polarity of output signals is also programmable. 31.6.1.3 Timing Engine Power Up Software Operation The following sequence is used to enable the display: 1. 2. 3. 4. 5. Configure LCD timing parameters, signal polarity and clock period. Enable the Pixel Clock by writing one to the CLKEN field of the LCDC_LCDEN register. Poll CLKSTS field of the LCDC_LCDSR register to check that the clock is running. Enable Horizontal and Vertical Synchronization by writing one to the SYNCEN field of the LCDC_LCDEN register. Poll LCDSTS field of the LCDC_LCDSR register to check that the synchronization is up.

515
6. 7.
Enable the display power signal by writing one to the DISPEN field of the LCDC_LCDEN register. Poll DISPSTS field of the LCDC_LCDSR register to check that the power signal is activated.
The GUARDTIME field of the LCDC_LCDCFG5 register is used to configure the number of frames before the assertion of the DISP signal. 31.6.1.4 Timing Engine Power Down Software Operation The following sequence is used to disable the display: 1. 2. 3. 4. 5. Disable the DISP signal by writing DISPDIS field of the LCDC_LCDDIS register. Poll DISPSTS field of the LCDC_LCDSR register to verify that the DISP is no longer activated. Disable the HSYNC and VSYNC signals by writing one to SYNCDIS field of the LCDC_LCDDIS register. Poll LCDSTS field of the LCDC_LCDSR register to check that the synchronization is off. Disable the Pixel clock by writing one in the CLKDIS field of the LCDC_LCDDIS register.
31.6.2 DMA Software Operations

31.6.2.1 DMA Channel Descriptor (DSCR) Alignment and Structure The DMA Channel Descriptor (DSCR) must be aligned on a 64-bit boundary. The DMA Channel Descriptor structure contains three fields:
z z z
DSCR.CHXADDR: Frame Buffer base address register DSCR.CHXCTRL: Transfer Control register. DSCR.CHXNEXT: Next Descriptor Address register.
Table 31-3. DMA Channel Descriptor Structure System Memory DSCR + 0x0 DSCR + 0x4 DSCR + 0x8 CTRL NEXT Structure Field for Channel CHX ADDR
31.6.2.2 Programming a DMA Channel 1. 2. 3. 4. 5. 6. Check the status of the channel by reading the CHXCHSR register. Write the channel descriptor (DSCR) structure in the system memory by writing DSCR.CHXADDR Frame base address, DSCR.CHXCTRL channel control and DSCR.CHXNEXT next descriptor location. If more than one descriptor is expected, the field DFETCH of DSCR.CHXCTRL is set to one to enable the descriptor fetch operation. Write the DSCR.CHXNEXT register with the address location of the descriptor structure and set DFETCH field of the DSCR.CHXCTRL register to one. Enable the relevant channel by writing one to the CHEN field of the CHXCHER register. An interrupt may be raised if unmasked when the descriptor has been loaded.
31.6.2.3 Disabling a DMA channel 1. 2. 3. 4. 5. Clearing the DFETCH bit in the DSCR.CHXCTRL field of the DSCR structure will disable the channel at the end of the frame. Setting the DSCR.CHXNEXT field of the DSCR structure will disable the channel at the end of the frame. Writing one to the CHDIS field of the CHXCHDR register will disable the channel at the end of the frame. Writing one to the CHRST field of the CHXCHDR register will disable the channel immediately. This may occur in the middle of the image. Polling CHSR field in the CHXCHSR register until the channel is successfully disabled.

516
31.6.2.4 DMA Dynamic Linking of a New Transfer Descriptor 1. 2. 3. 4. 5. Write the new descriptor structure in the system memory. Write the address of the new structure in the CHXHEAD register. Add the new structure to the queue of descriptors by writing one to the A2QEN field of the CHXCHER register. The new descriptor will be added to the queue on the next frame. An interrupt will be raised if unmasked, when the head descriptor structure has been loaded by the DMA channel.
31.6.2.5 DMA Interrupt Generation The DMA controller operation sets the following interrupt flags in the interrupt status register CHXISR:
z z z z
DMA field indicates that the DMA transfer is completed. DSCR field indicates that the descriptor structure is loaded in the DMA controller. ADD field indicates that a descriptor has been added to the descriptor queue. DONE field indicates that the channel transfer has terminated and the channel is automatically disabled.
31.6.2.6 DMA Address Alignment Requirements When programming the DSCR.CHXADDR field of the DSCR structure the following requirement must be met.
Table 31-4. DMA Address Alignment when CLUT Mode is Selected CLUT Mode 1 bpp 2 bpp 4 bpp 8 bpp 8 bits 8 bits 8 bits 8 bits DMA Address Alignment
Table 31-5. DMA Address Alignment when RGB Mode is Selected RGB Mode 12 bpp RGB 444 16 bpp ARGB 4444 16 bpp RGBA 4444 16 bpp RGB 565 16 bpp TRGB 1555 18 bpp RGB 666 18 bpp RGB 666 PACKED 19 bpp TRGB 1666 19 bpp TRGB 1666 24 bpp RGB 888 24 bpp RGB 888 PACKED 25 bpp TRGB 1888 32 bpp ARGB 8888 32 bpp RGBA 8888 16 bits 16 bits 16 bits 16 bits 16 bits 32 bits 8 bits 32 bits 8 bits 32 bits 8 bits 32 bits 32 bits 32 bits DMA Address Alignment

517
Table 31-6. DMA Address Alignment when YUV Mode is Selected YUV Mode 32 bpp AYCrCb 16 bpp YCrCb 4:2:2 16 bpp semiplanar YCrCb 4:2:2 CrCb 16 bits Y 8 bits 16 bpp planar YCrCb 4:2:2 Cr 8 bits Cb 8 bits Y 8 bits 12 bpp YCrCb 4:2:0 CrCb 16 bits Y 8 bits 12 bpp YCrCb 4:2:0 Cr 8 bits Cb 8 bits 32 bits 32 bits Y 8 bits DMA Address Alignment
31.6.3 Overlay Software Configuration

31.6.3.1 System Bus Access Attributes These attributes are defined to improve bandwidth of the overlay.
z z z z z
LOCKDIS field: when set to one the AHB lock signal is not asserted when the PSTRIDE value is different from zero (rotation in progress). ROTDIS field: when set to one the Pixel Striding optimization is disabled. DLBO field: when set to one only defined burst lengths are performed when the DMA channel retrieves the data from the memory. BLEN field: defines the maximum burst length of the DMA channel. SIF field: defines the targeted DMA interface.
31.6.3.2 Color Attributes

z z z
CLUTMODE field: selects one color lookup mode RGBMODE field: selects the RGB mode. YUVMODE field: selects the Luminance Chrominance mode.
31.6.3.3 Window Position, Size, Scaling and Striding Attributes

z z z z z
XPOS: YPOS fields define the position of the overlay window. XSIZE: YSIZE fields define the size of the displayed window. XMEMSIZE: YMEMSIZE fields define the size of the image frame buffer. XSTRIDE: PSTRIDE fields define the line and pixel striding. XFACTOR: YFACTOR fields define the scaling ratio.
The position and size attributes are to be programmed to keep the window within the display area.

518
When the color lookup mode is enabled the following restrictions apply on the horizontal and vertical window size:
Table 31-7. Color Lookup Mode and Window Size CLUT MODE 1 bpp 2 bpp 4 bpp 8 bpp x-y Size Requirement multiple of 8 pixels multiple of 4 pixels multiple of 2 pixels free size
Pixel striding is disabled when CLUT mode is enabled. When YUV mode is enabled the following restrictions apply on the window size:
Table 31-8. YUV Mode and Window Size YUV MODE AYUV YUV 4:2:2 packed x-y Requirement, Scaling Turned Off Free size xsize is greater than 2 pixels x-y Requirement, Scaling Turned On x-y size is greater than 5 x-y size is greater than 5 x-y size is greater than 5 x-y size is greater than 5 x-y size is greater than 5 x-y size is greater than 5
YUV 4:2:2 semiplanar xsize is greater than 2 pixels YUV 4:2:2 planar xsize is greater than 2 pixels
YUV 4:2:0 semiplanar xsize is greater that 2 pixels YUV 4:2:0 planar xsize is greater than 2 pixels
In RGB mode, there is no restriction on the line length. 31.6.3.4 Overlay Blender Attributes When two or more video layers are used, alpha blending is performed to define the final image displayed. Each window has its own blending attributes.
z z z z z z z z z z z z
CRKEY Field: enables the chroma keying and match logic. INV Field: performs bit inversion at pixel level. ITER2BL Field: when set the iterated data path is selected. ITER Field. REVALPHA Field: uses the reverse alpha value. GAEN Field: enables the global alpha value in the data path. LAEN Field: enables the local alpha value from the pixel. OVR Field: when set the overlay is selected as an input of the blender. DMA Field: the DMA data path is activated. REP Field: enables the bit replication to fill the 24-bit internal data path. DSTKEY Field: when set, Destination keying is enabled. GA Field: defines the global alpha value.
31.6.3.5 Overlay Attributes Software Operation 1. 2. 3. When required, write the overlay attributes configuration registers. Set UPDATEEN field of the CHXCHER register. Poll UPDATESR field in the CHXCHSR, the update applies when that field is reset.

519
31.6.4 RGB Frame Buffer Memory Bitmap

31.6.4.1 1 bpp Through Color Lookup Table
Table 31-9. 1 bpp Memory Mapping, Little Endian Organization Mem addr
Bit Pixel 1 bpp
0x3
0x2
0x1
8
0x0
7 6 5 4 3 2 1 0
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9
p3 p3 p2 p2 p2 p2 p2 p2 p2 p2 p2 p2 p1 p1 p1 p1 p1 p1 p1 p1 p1 p11 p9 p8 p7 p6 p5 p4 p3 p2 p1 p0 1 0 9 8 7 6 5 4 3 2 1 0 9 8 7 6 5 4 3 2 0

Bit Pixel 2 bpp
0x3
0x2
0x1
8 p4
0x0
7 6 p3 5 4 p2 3 2 p1 1 0 p0
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 p15 p14 p13 p12 p11 p10 p9 p8 p7 p6 p5

Bit Pixel 4 bpp
0x3
0x2
0x1
8
0x0
7 6 5 4 3 2 1 0
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9
p7
p6
p5
p4
p3
p2
p1
p0

Bit Pixel 8 bpp
0x3
0x2
0x1
8
0x0
7 6 5 4 3 p0 2 1 0
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 p3 p2 p1
31.6.4.5 12 bpp Memory Mapping, RGB 4:4:4

Bit Pixel 12 bpp
0x3
0x2
0x1
8
0x0
7 6 5 4 3 2 1 0
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 R1[3:0] G1[3:0] B1[3:0] R0[3:0]
G0[3:0]
B0[3:0]

520
31.6.4.6 16 bpp Memory Mapping with Alpha Channel, ARGB 4:4:4:4

Bit Pixel 16 bpp
0x3
0x2
0x1
8
0x0
7 6 5 4 3 2 1 0
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 A1[3:0] R1[3:0] G1[3:0] B1[3:0] A0[3:0] R0[3:0]
G0[3:0]
B0[3:0]
31.6.4.7 16 bpp Memory Mapping with Alpha Channel, RGBA 4:4:4:4

Bit Pixel 16 bpp
0x3
0x2
0x1
8
0x0
7 6 5 4 3 2 1 0
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 R1[3:0] G13:0] B1[3:0] A1[3:0] R0[3:0] G0[3:0]
B0[3:0]
A0[3:0]
31.6.4.8 16 bpp Memory Mapping with Alpha Channel, RGB 5:6:5

Bit Pixel 16bpp
0x3
0x2
0x1
8
0x0
7 6 5 4 3 2 1 0
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 R1[4:0] G1[5:0] B1[4:0] R0[4:0]
G0[5:0]
B0[4:0]
31.6.4.9 16 bpp Memory Mapping with Transparency Bit, ARGB 1:5:5:5

Bit Pixel 4 bpp
0x3
0x2
0x1
8
0x0
7 6 5 4 3 2 1 0
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 A1 R1[4:0] G1[4:0] B1[4:0] A0 R0[4:0]
G0[4:0]
B0[4:0]
31.6.4.10 18 bpp Unpacked Memory Mapping with Transparency Bit, RGB 6:6:6
Table 31-18. 18 bpp Unpacked Memory Mapping, Little Endian Organization Mem addr
Bit Pixel 18 bpp
0x3
0x2
0x1
8
0x0
7 6 5 4 3 2 1 0
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 R0[5:0]
G0[5:0]
B0[5:0]

521
31.6.4.11 18 bpp Packed Memory Mapping with Transparency Bit, RGB 6:6:6
Table 31-19. 18 bpp Packed Memory Mapping, Little Endian Organization at Address 0x0, 0x1, 0x2, 0x3 Mem addr
Bit Pixel 18 bpp
0x3
0x2
0x1
8
0x0
7 6 5 4 3 2 1 0
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 G1[1:0] B1[5:0] R0[5:0]
G0[5:0]
B0[5:0]
Bit Pixel 18 bpp
0x7
0x6
0x5
8
0x4
7 6 5 4 3 2 1 0
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 R2[3:0] G2[5:0] B2[5:0]
R1[5:2]
G1[5:2]
Table 31-21. 18 bpp Packed Memory Mapping, Little Endian Organization at Address 0x8, 0x9, 0xA, 0xB Mem addr
Bit Pixel 18 bpp
0xB
0xA
0x9
8
0x8
7 6 5 4 3 2 1 0
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 G4[1:0] B4[5:0] R3[5:0] G3[5:0]
B3[3:0]
R2[5:4]
31.6.4.12 19 bpp Unpacked Memory Mapping with Transparency Bit, RGB 1:6:6:6
Table 31-22. 19 bpp Unpacked Memory Mapping, Little Endian Organization Mem addr
Bit Pixel 19 bpp
0x3
0x2
0x1
8
0x0
7 6 5 4 3 2 1 0
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 A0 R0[5:0]
G0[5:0]
B0[5:0]
31.6.4.13 19 bpp Packed Memory Mapping with Transparency Bit, ARGB 1:6:6:6
Bit Pixel 19 bpp
0x3
0x2
0x1
8
0x0
7 6 5 4 3 2 1 0
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 G1[1:0] B1[5:0] A0 R0[5:0]
G0[5:0]
B0[5:0]
Bit Pixel 19 bpp
0x7
0x6
0x5
8
0x4
7 6 5 4 3 2 1 0
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 R2[3:0] G2[5:0] B2[5:0] A1
R1[5:2]
G1[5:2]

522
Table 31-25. 18 bpp Packed Memory Mapping, Little Endian Organization at Address 0x8, 0x9, 0xA, 0xB Mem addr
Bit Pixel 19 bpp
0xB
0xA
0x9
8
0x8
7 6 5 4 3 2 1 0
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 G4[1:0] B4[5:0] A3 R3[5:0] G3[5:0]
B3[3:0]
R2[5:4]
31.6.4.14 24 bpp Unpacked Memory Mapping, RGB 8:8:8

Bit Pixel 24 bpp
0x3
0x2
0x1
8
0x0
7 6 5 4 3 2 1 0
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 R0[7:0] G0[7:0]
B0[7:0]
31.6.4.15 24 bpp Packed Memory Mapping, RGB 8:8:8

Bit Pixel 24 bpp
0x3
0x2
0x1
8
0x0
7 6 5 4 3 2 1 0
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 B1[7:0] R0[7:0] G0[7:0]
B0[7:0]
Bit Pixel 24 bpp
0x7
0x6
0x5
8
0x4
7 6 5 4 3 2 1 0
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 G2[7:0] B2[7:0] R1[7:0]
G1[7:0]
31.6.4.16 25 bpp Memory Mapping, ARGB 1:8:8:8

Bit Pixel 25 bpp
0x3
0x2
0x1
8
0x0
7 6 5 4 3 2 1 0
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 A0 R0[7:0] G0[7:0]
B0[7:0]
31.6.4.17 32 bpp Memory Mapping, ARGB 8:8:8:8

Bit Pixel 32 bpp
0x3
0x2
0x1
8
0x0
7 6 5 4 3 2 1 0
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 A0[7:0] R0[7:0] G0[7:0]
B0[7:0]

523
31.6.4.18 32 bpp Memory Mapping, RGBA 8:8:8:8

Bit Pixel 32 bpp
0x3
0x2
0x1
8
0x0
7 6 5 4 3 2 1 0
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 R0[7:0] G0[7:0] B0[7:0]
A0[7:0]
31.6.5 YUV Frame Buffer Memory Mapping

31.6.5.1 AYCbCr 4:4:4 Interleaved Frame Buffer Memory Mapping
Bit Pixel 16 bpp
0x3
0x2
0x1
8
0x0
7 6 5 4 3 2 1 0
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 A0[7:0] Y0[7:0] Cb0[7:0]
Cr0[7:0]
31.6.5.2 4:2:2 Interleaved Mode Frame Buffer Memory Mapping

Table 31-33. 16 bpp 4:2:2 interleaved Mode 0 Mem addr
Bit Pixel 16 bpp
0x3
0x2
0x1
8
0x0
7 6 5 4 3 2 1 0
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 Cr0[7:0] Y1[7:0] Cb0[7:0]
Y0[7:0]

Bit Pixel 16 bpp
0x3
0x2
0x1
8
0x0
7 6 5 4 3 2 1 0
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 Y1[7:0] Cr0[7:0] Y0[7:0]
Cb0[7:0]

Bit Pixel 16 bpp
0x3
0x2
0x1
8
0x0
7 6 5 4 3 2 1 0
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 Cb0[7:0] Y1[7:0] Cr0[7:0]
Y0[7:0]

Bit Pixel 16 bpp
0x3
0x2
0x1
8
0x0
7 6 5 4 3 2 1 0
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 Y1[7:0] Cb0[7:0] Y0[7:0]
Cr0[7:0]

524
31.6.5.3 4:2:2 Semiplanar Mode Frame Buffer Memory Mapping

Table 31-37. 4:2:2 Semiplanar Luminance Memory Mapping, Little Endian Organization for Byte 0x0, 0x1, 0x2, 0x3 Mem addr
Bit Pixel 16 bpp
0x3
0x2
0x1
8
0x0
7 6 5 4 3 2 1 0
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 Y3[7:0] Y2[7:0] Y1[7:0]
Y0[7:0]
Table 31-38. 4:2:2 Semiplanar Chrominance Memory Mapping, Little Endian Organization for Byte 0x0, 0x1, 0x2, 0x3 Mem addr
Bit Pixel 16 bpp
0x3
0x2
0x1
8
0x0
7 6 5 4 3 2 1 0
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 Cb2[7:0] Cr2[7:0] Cb0[7:0]
Cr0[7:0]
31.6.5.4 4:2:2 Planar Mode Frame Buffer Memory Mapping

Table 31-39. 4:2:2 Planar Mode Luminance Memory Mapping, Little Endian Organization for Byte 0x0, 0x1, 0x2, 0x3 Mem addr
Bit Pixel 16 bpp
0x3
0x2
0x1
8
0x0
7 6 5 4 3 2 1 0
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 Y3[7:0] Y2[7:0] Y1[7:0]
Y0[7:0]
Table 31-40. 4:2:2 Planar Mode Chrominance Memory Mapping, Little Endian Organization for Byte 0x0, 0x1, 0x2, 0x3 Mem addr
Bit Pixel 16 bpp
0x3
0x2
0x1
8
0x0
7 6 5 4 3 2 1 0
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 C3[7:0] C2[7:0] C1[7:0]
C0[7:0]
31.6.5.5 4:2:0 Planar Mode Frame Buffer Memory Mapping In Planar Mode, the three video components Y, Cr and Cb are split into 3 memory areas and stored in a raster-scan order. These three memory planes are contiguous and always aligned on a 32-bit boundary.
Bit Pixel 12 bpp
0x3
0x2
0x1
8
0x0
7 6 5 4 3 2 1 0
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 Y3[7:0] Y2[7:0] Y1[7:0]
Y0[7:0]
Bit Pixel 12 bpp
0x7
0x6
0x5
8
0x4
7 6 5 4 3 2 1 0
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 Y7[7:0] Y6[7:0] Y5[7:0]
Y4[7:0]

525
Bit Pixel 12 bpp
0x3
0x2
0x1
8
0x0
7 6 5 4 3 2 1 0
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 C3[7:0] C2[7:0] C1[7:0]
C0[7:0]
Bit Pixel 12 bpp
0x7
0x6
0x5
8
0x4
7 6 5 4 3 2 1 0
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 C7[7:0] C6:[7:0] C5[7:0]
C4[7:0]
31.6.5.6 4:2:0 Semiplanar Frame Buffer Memory Mapping

Table 31-45. 4:2:0 Semiplanar Mode Luminance Memory Mapping, Little Endian Organization Mem addr
Bit Pixel 12 bpp
0x7
0x6
0x5
8
0x4
7 6 5 4 3 2 1 0
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 Y3[7:0] Y2[7:0] Y1[7:0]
Y0[7:0]
Table 31-46. 4:2:0 Semiplanar Mode Chrominance Memory Mapping, Little Endian Organization Mem addr
Bit Pixel 12 bpp
0x3
0x2
0x1
8
0x0
7 6 5 4 3 2 1 0
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 Cb1[7:0] Cr1[7:0] Cb0[7:0]
Cr0[7:0]

526
31.6.6 Chrominance Upsampling Unit

Both 4:2:2 and 4:2:0 input formats are supported by the LCD module. In 4:2:2, the two chrominance components are sampled at half the sample rate of the luminance. The horizontal chrominance resolution is halved. When this input format is selected, the chrominance upsampling unit uses two chrominances to interpolate the missing component. In 4:2:0, Cr and Cb components are subsampled at a factor of two vertically and horizontally. When this input mode is selected, the chrominance upsampling unit uses two and four chroma components to generate the missing horizontal and vertical components.
Figure 31-2. 4:2:2 Upsampling Algorithm
Vertical and Horizontal upsampling 4:2:2 to 4:4:4 conversion 0 or 180 degree
C[0,0]
C[x/2,0]
C[x,0]
C[0,y/2]
C[x/2,y/2]
C[x,y/2]
C[0,y]
C[x/2,y]
C[x,y]
Y sample Cr Cb calculated at encoding time Cr Cb interpolated from 2 Chroma Component

527
Figure 31-3. 4:2:2 Packed Upsampling Algorithm
C[0,0]
C[x/2,0]
C[x,0]
C[0,y/2]
C[x/2,y/2]
C[x,y/2]
C[0,y]
C[x/2,y]
C[x,y]
Y sample Cr Cb calculated at encoding time Cr Cb from the previous line (interpolated)

528
Figure 31-4. 4:2:2 Semiplanar and Planar Upsampling Algorithm - 90 or 270 Degree Rotation Activated
C[0,0]
C[x/2,0]
C[x,0]
C[0,y/2]
C[x/2,y/2]
C[x,y/2]
C[0,y]
C[x/2,y]
C[x,y]
Y sample Cr Cb calculated at encoding time Cr Cb interpolated

529
Figure 31-5. 4:2:0 Upsampling Algorithm
Vertical and Horizontal upsampling 4:2:0 to 4:4:4 conversion
C[0,0]
C[x/2,0]
C[x,0]
C[0,y/2]
C[x/2,y/2]
C[x,y/2]
C[0,y]
C[x/2,y]
C[x,y]
Y sample Cr Cb calculated at encoding time Cr Cb interpolated from 2 Chroma Component Cr Cb interpolated from 4 Chroma Component
x Cr [ 0, 0 ] + Cr [ 0, x ] -, 0 = --------------------------------------------------Chroma -2 2
y Cr [ 0, 0 ] + C [ 0, y ] - = -----------------------------------------------Chroma 0, -2 2 x y [ 0, 0 ] + Cr [ x, 0 ] + Cr [ y, 0 ] + Cr [ x, y ] -, -- = Cr Chroma -------------------------------------------------------------------------------------------------------------2 2 4 y [ x, 0 ] + Cr [ x, y ] - = Cr Chroma x, --------------------------------------------------2 2 x [ 0, y ] + Cr [ x, y ] -, y = Cr Chroma --------------------------------------------------2 2

530
31.6.6.1 Chrominance Upsampling Algorithm 1. Read line n from chrominance cache and interpolate [x/2,0] chrominance component filling the 1 x 2 kernel with line n. If the chrominance cache is empty, then fetch the first line from external memory and interpolate from the external memory. Duplicate the last chrominance at the end of line. Fetch line n+1 from external memory, write line n + 1 to chrominance cache, read line n from the chrominance cache. interpolate [0,y/2], [x/2,y/2] and [x, y/2] filling the 2x2 kernel with line n and n+1. Duplicate the last chrominance line to generate the last interpolated line. Repeat step 1 and step 2.
2.
3.
31.6.7 Line and Pixel Striding

The LCD module includes a mechanism to increment the memory address by a programmable amount when the end of line has been reached, this offset is referred to as XSTRIDE and is defined on a per overlay basis. It also contains a PSTRIDE field that allows a programmable jump at the pixel level. Pixel stride is the value from one pixel to the next. 31.6.7.1 Line Striding When the end of line has been reached, the DMA address counter points to the next pixel address. The channel DMA address register is added to the XSTRIDE field, and then updated. If XSTRIDE is set to zero, the DMA address register remains unchanged. The XSTRIDE field of the channel configuration register is aligned to the pixel size boundary. The XSTRIDE field is a twos complement number. 31.6.7.2 Pixel Striding The DMA channel engine may optionally fetch non contiguous pixels. The channel DMA address register is added to the PSTRIDE field and then updated. If PSTRIDE is set to zero, the DMA address register remains unchanged and pixels are contiguous. The PSTRIDE field of the channel configuration register is aligned to the pixel size boundary. The PSTRIDE is a twos complement number.
31.6.8 Color Space Conversion Unit

The color space conversion unit converts Luminance Chrominance color space into the Red Green Blue color space. The conversion matrix is defined below and is fully programmable through the LCD user interface
R CSCRY CSCRU CSCRV Y Yoff = G CSCGY CSCGU CSCGV Cb Cboff B CSCBY CSCBU CSCBV Cr Croff
Color space conversion coefficients are defined with the following equation:
8
1 - CSC ij = ----7 2
2 c9 +
cn
n=0
Color space conversion coefficients are defined with one sign bit, 2 integer bits and 7 fractional bits. The range of the CSCij coefficients is defined below with a step of 1/128.
4 CSC ij 3.9921875
Additionally a set scaling factor {Yoff, Cboff, Croff} can be applied.

531
31.6.9 Two Dimension Scaler

The High End Overlay (HEO) data path includes a hardware scaler that allows an image resize in both horizontal and vertical directions. 31.6.9.1 Video Scaler Description The scaling operation is based on a vertical and horizontal resampling algorithm. The sampling rate of the original image is increased when the video is upscaled, and decreased when the video is downscaled. A Vertical resampler is used to perform a vertical interpolation by a factor of vI, and a decimation by a factor of vD. A Horizontal resampler is used to perform a vertical interpolation by a factor of hI, and a decimation by a factor of hD. Both horizontal and vertical low pass filters are designed to minimize the aliasing effect. The frequency response of the low pass filter has the following characteristics:
I H ( ) = when 0 min ( --,---) ID 0 otherwise
Taking into account the linear phase condition and anticipating the filter length M, the desired frequency response is modified.
--- j M 2 H ( ) = Ie
when 0 min ( --,---) ID 0 otherwise
Figure 31-6. Video Resampler Architecture

Input video stream Vertical Resampler Vertical upsampler Low Pass Filter Vertical downsampler
vI
vD
Horizontal Resampler Horizontal upsampler Low Pass Filter Horizontal downsampler
output video stream
hI
hD
The impulse response of the low pass filter defined is:

h(n) = I c - when n = 0 I --- c sin ( c n ) ---- --------------------- otherwise c n
or, for the filter of length M:

c - when n = M I ------ 2 h(n) = sin c n M ---- c 2 - -------------------------------------- otherwise I --- M n ---c 2

532
This ideal filter is non-causal and cannot be realized. The unit sample response h(n) is infinite in duration and must be truncated depending on the expected length M of the filter. This truncation is equivalent to the multiplication of the impulse response by a window function w(n).
Table 31-47. Window Function for a Filter Length M Name of the Window Function Time Domain Sequence w(n) 1 2 nM ------------2 1 -----------------------------------M1 2 n4 n0.42 0.5 cos ------------+ 0.08 cos ------------M1 M1 2 n0.54 0.46 cos ------------M1 2 n0.5 0.5 cos ------------M1
Barlett
Blackman
Hamming
Hanning
The horizontal resampler includes an 8-phase 5-tap filter equivalent to a 40-tap FIR described in Figure 31-8. The vertical resampler includes an 8-phase 3-tap filter equivalent to a 24-tap FIR described in Figure 31-8.
Figure 31-7. Horizontal Resampler Filter Architecture
x(n)
Coefficient storage
coeff4
coeff3
coeff2
coeff1
coeff0
y(m)
Figure 31-8. Vertical Resampler Filter Architecture

x(n)
-1
-1
Coefficient storage
coeff2
coeff1
coeff0
y(m)

533
31.6.9.2 Horizontal Scaler The XMEMSIZE field of the LCDC_HEOCFG4 register indicates the horizontal size minus one of the image in the system memory. The XSIZE field of the LCDC_HEOCFG3 register contains the horizontal size minus one of the window.The SCALEN field of the LCDC_HEOCFG13 register is set to one. The scaling factor is programmed in the XFACTOR field of the LCDC_HEOCFG13 register. Use the following algorithm to find the XFACTOR value.
8 256 XMEMSIZE 256 XPHIDEF XFACTOR 1 st = floor ------------------------------------------------------------------------------------------------------------ XSIZE
XFACTOR 1 st = XFACTOR 1 st + 1 XFACTOR 1 st XSIZE + 256 XPHIDEF XMEMSIZE max = floor -------------------------------------------------------------------------------------------------------------- 2048
XFACTOR = XFACTOR 1 st 1 XFACTOR = XFACTOR 1 st
when ( XMEMSIZE max > XMEMSIZE ) otherwise
31.6.9.3 Vertical Scaler The YMEMSIZE field of the LCDC_HEOCFG4 register indicates the vertical size minus one of the image in the system memory. The YSIZE field of the LCDC_HEOCFG3 register contains the vertical size minus one of the window. The SCALEN field of the LCDC_HEOCFG13 register is set to one. The scaling factor is programmed in the YFACTOR field of the LCDC_HEOCFG13 register.
8 256 YMEMSIZE 256 YPHIDEF ------------------------------------------------------------------------------------------------------------- YFACTOR 1 st = floor YSIZE
YFACTOR 1 st = YFACTOR 1 st + 1 YFACTOR 1 st YSIZE + 256 YPHIDEF YMEMSIZE max = floor ------------------------------------------------------------------------------------------------------------- 2048
YFACTOR = YFACTOR 1 st 1 YFACTOR = YFACTOR 1 st 31.6.10 Hardware Cursor
when ( YMEMSIZE max > YMEMSIZE ) otherwise
The LCD module integrates a hardware cursor database. This layer features only a minimal set of color among 1, 2, 4 and 8 bpp palletized and 16 bpp to 32 bpp true color. The cursor size is limited to 128 x 128 pixels.

534
31.6.11 Color Combine Unit

31.6.11.1 Window Overlay The LCD module provides hardware support for multiple overlay plane that can be used to display windows on top of the image without destroying the image located below. The overlay image can use any color depth. Using the overlay alleviates the need to re-render the occluded portion of the image. When pixels are combined together through the alpha blending unit, a new color is created. This new pixel is called an iterated pixel and is passed to the next blending stage. Then, this pixel may be combined again with another pixel. The VIDPRI field located in the LCDC_HEOCFG12 register configures the video priority algorithm used to display the layers. When VIDPRI field is set to zero, the OVR1 layer is located above the HEO layer. When VIDPRI field is set to one, OVR1 is located below the HEO layer.
Figure 31-9. Overlay Example with Two Different Video Prioritization Algorithms
HEO width OVR1 width Base width Base height
o0(x,y)
HEO
o1(x,y)
Overl ay1 OVR1
HEO height
HCC
OVR1 height
Base Image Video Prioritization Algorithm 1 : HCC > OVR1 > HEO > BASE
Base Image
HEO
HCC Overl ay1 OVR1
Video Prioritization Algorithm 2 : HCC > HEO > OVR1 > BASE

535
31.6.11.2 Base Layer, with Window Overlay Optimization When the base layer is combined with at least one active overlay, the whole base layer frame is retrieved from the memory though it is not visible. A set of registers is used to disable the Base DMA when this condition is met. These registers are listed below:
z z z z z
LCDC_CFG5: DISCXPOS field discard area horizontal position LCDC_CFG5: DISCYPOS field discard area vertical position LCDC_CFG6: DISCXSIZE field discard area horizontal size LCDC_CFG6: DISCYSIZE field discard area vertical size LCDC_CFG4: DISCEN field discard area enable

536
Figure 31-10. Base Layer Discard Area

Base width
Base Image
Base height
discxsize Base width Base height
{discxpos, discypos}
Overlay1
Discarded Area
discysize
Base Image
HEO width Base width Base height
HEO Video
HEO height
Base Image

537
31.6.11.3 Overlay Blending The blending function requires two pixels (one iterated from the previous blending stage and one from the current overlay color) and a set of blending configuration parameters. These parameters define the color operation.
Figure 31-11. Alpha Blender Function
iter[n-1] la ovr
GA OVR From LAEN Shadow REVALPHA Registers ITER ITER2BL CRKEY INV DMA GAEN RGBKEY RGBMASK OVRDEF
blending function
iter[n]
Figure 31-12. Alpha Blender Database

la ovr iter[n-1]
OVR ITER OVRDEF GA "0" GAEN DMA LAEN

0 0 0
"0"
0
"0"
0
REVALPHA ovr RGBKEY RGBMASK CRKEY
Alpha * ovr + (1 - Alpha) * iter[n-1] ovr iter[n-1]

0
MATCH LOGIC
0
Inverted INV
iter[n]

538
31.6.11.4 Global Alpha Blender

Figure 31-13. Global Alpha Blender
base ovr1la ovr1 la ovr ovr2la ovr2 heola heo hcrla hcr
iter[n-1]
blending function iter[n]
iter[n-1]
la
ovr
iter[n-1]
la
ovr
iter[n-1]
la
ovr

539
31.6.11.5 Window Blending

Figure 31-14. 256-level Alpha Blending
Base Image
OVR1 25 %
HEO 75 %
Video Prioritization Algorithm 1 : OVR1 > HEO > BASE

540
31.6.11.6 Color Keying Color keying involves a method of bit-block image transfer (Blit). This entails blitting one image onto another where not all the pixels are copied. Blitting usually involves two bitmaps, a source bitmap and a destination bitmap. A raster operation (ROP) is performed to define whether the iterated color or the overlay color is to be visible or not.
Source Color Keying

If the masked overlay color matches the color key then the iterated color is selected, Source Color Keying is activated using the following configuration.
z z z z z
Select the Overlay to Blit Set DSTKEY field to zero Activate Color Keying setting CRKEY field to 1 Program Color Key writing RKEY, GKEY and BKEY fields Program Color Mask writing RKEY, GKEY and BKEY fields
When the Mask register is set to zero, the comparison is disabled and the raster operation is activated.
Destination Color Keying

If the iterated masked color matches the color key then the overlay color is selected, Destination Color Keying is activated using the following configuration:
z z z z z
Select the Overlay to Blit Set DSTKEY field to one Activate Color Keying setting CRKEY field to 1 Program Color Key writing RKEY, GKEY and BKEY fields Program Color Mask writing RKEY, GKEY and BKEY fields
When the Mask register is set to zero, the comparison is disabled and the raster operation is activated.
31.6.12 LCDC PWM Controller

This block generates the LCD contrast control signal (LCD_PWM) to make possible the control of the display's contrast by software. This is an 8-bit PWM (Pulse Width Modulation) signal that can be converted to an analog voltage with a simple passive filter. The PWM module has a free-running counter whose value is compared against a compare register (PWMCVAL field of the LCDC_LCDCFG6 register). If the value in the counter is less than that in the register, the output brings the value of the polarity (PWMPOL field) bit in the PWM control register: LCDC_LCDCFG6. Otherwise, the opposite value is output. Thus, a periodic waveform with a pulse width proportional to the value in the compare register is generated. Due to the comparison mechanism, the output pulse has a width between zero and 255 PWM counter cycles. Thus by adding a simple passive filter outside the chip, an analog voltage between 0 and (255/256) VDD can be obtained (for the positive polarity case, or between (1/256) VDD and VDD for the negative polarity case). Other voltage values can be obtained by adding active external circuitry. For PWM mode, the counter frequency can be adjusted to four different values using the PWMPS field of the LCDC_LCDCFG6 register. The PWM module can be fed with the slow clock or the system clock, depending on the CLKPWMSEL field of the LCDC_CFG0 register.

541
31.6.13 Post Processing Controller

The output stream of pixels can be either displayed on the screen or written to the memory using the Post Processing Controller (PPC). When the PPC is used, the screen display is disabled, but synchronization signals remain active (if enabled). The stream of pixel can be written in RGB mode or encoded in YCbCr 422 mode. A programmable color space conversion stage is available.
Y CSCYR CSCYG CSCYB R Yoff U = CSCUR CSCUG CSCUB G + Uoff V CSCVR CSCVG CSCUB B Voff
31.6.14 LCD Overall Performance

31.6.14.1 Color Lookup Table (CLUT)
Table 31-48. CLUT Pixel Performance CLUT MODE 1 bpp 2 bpp 3 bpp 4 bpp Pixels/Cycle 64 32 16 8 ROTATION Not supported Not supported Not supported Not supported SCALING Supported Supported Supported Supported
31.6.14.2 RGB Mode Fetch Performance

Table 31-49. RGB Mode Performance Pixels/Cycle Memory Burst Mode 4 4 2 2.666 2 2.666 2 2.666 2 2 1 1 1 Not supported 1 Not Supported 1 Not Supported 1 1 Rotation Peak Random Memory Access (pixels/cycle) Rotation Optimization(1) 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 Normal Mode SCALING Burst Mode or Rotation Optimization Available Supported Supported Supported Supported Supported Supported Supported Supported Supported Supported
RGB Mode 12 bpp 16 bpp 18 bpp 18 bpp RGB PACKED 19 bpp 19 bpp PACKED 24 bpp 24 bpp PACKED 25 bpp 32 bpp
Note:
1. Rotation optimization = AHB lock asserted on consecutive single access.

542
31.6.14.3 YUV Mode Fetch Performance
Table 31-50. Single Stream for 0 Wait State Memory Pixels/Cycle Memory Burst Mode 2 4 1 Not Supported Rotation Peak Random Memory Access (pixels/cycle) Rotation Optimization 0.2 Not Supported Normal Mode SCALING Burst Mode or Rotation Optimization is Available Supported Supported
YUV Mode 32 bpp AYUV 16 bpp 422
Note:
Rotation optimization = AHB lock asserted on consecutive single access
Table 31-51. Multiple Stream for 0 Wait State Memory Comp/cycle Memory Burst Mode 8 Y, 4 UV 8 Y, 8 U, 8 V 8 Y, 4 UV 8 Y, 8 U, 8 V Rotation Peak Random Memory Access (Pixels/cycle) Rotation Optimization 1 Y, 1 UV (2 streams) 1 Y, 1 U, 1 V (3 streams) 1 Y, 1 UV (2 streams) 1 Y, 1 U, 1 V (3 streams) Normal Mode 0.2 Y 0.2 UV (2 streams) SCALING Burst Mode or Rotation Optimization is Available Supported
YUV Mode 16 bpp 422 semiplanar 16 bpp 422 planar 12 bpp 4:2:0 semiplanar 12 bpp 4:2:0 planar
0.2 Y, 0.2 U, 0.2 V (3 streams) Supported 0.2 Y 0.2 UV (2 streams) Supported
0.2 Y, 0.2 U, 0.2 V (3 streams) Supported
Note:
In order to provide more bandwidth, when multiple streams are used to transfer Y, UV, U or V components, two AHB interfaces are recommended or multiple AXI ID are required.
Table 31-52. YUV Planar Overall Performance 1 AHB Interface for 0 Wait State Memory Pix/cycle Memory Burst Mode 4 4 5.32 5.32 Rotation Peak Random Memory Access (Pixels/cycle) Rotation Optimization 0.66 0.5 0.8 0.66 0.132 0.1 0.16 0.132 Normal Mode SCALING Burst Mode or Rotation Optimization is Available Supported Supported Supported Supported
YUV Mode 16 bpp 422 semiplanar 16 bpp 422 planar 12 bpp 4:2:0 semiplanar 12 bpp 4:2:0 planar
In order to provide more bandwidth, when multiple streams are used to transfer Y, UV, U or V components, two AHB interfaces are recommended or multiple AXI ID are required.
31.6.15 Input FIFO

The LCD module includes one input FIFO per overlay. These input FIFOs are used to buffer the AHB burst and serialize the stream of pixels.
31.6.16 Output FIFO

The LCD module includes one output FIFO that stores the blended pixel.

543
31.6.17 Output Timing Generation

31.6.17.1 Active Display Timing Mode
Figure 31-15. Active Display Timing
LCD_PCLK
LCD_VSYNC
LCD_HSYNC
LCD_BIAS_DEN
LCD_DAT[23:0]
HSW
VSW
VBP
HBP
LCD_PCLK
LCD_VSYNC
LCD_HSYNC
LCD_BIAS_DEN
LCD_DAT[23:0]
HSW
HBP
PPL
HFP
HSW
HBP
LCD_PCLK
LCD_VSYNC
LCD_HSYNC
LCD_BIAS_DEN
LCD_DAT[23:0]
PPL
HFP
HSW
VFP

544
Figure 31-16. Vertical Synchronization Timing (part 1)

VSPDLYS = 0 LCD_PCLK VSPDLYE = 0 VSPSU = 0 VSPHO = 0
LCD_VSYNC
LCD_HSYNC
HSW VSPDLYS = 1 VSPDLYE = 0 VSPSU = 0
VSW VSPHO = 0
VBP
HBP
LCD_PCLK
LCD_VSYNC
LCD_HSYNC
VSW VSPHO = 0
VBP
HBP
LCD_PCLK
LCD_VSYNC
LCD_HSYNC
VSW VSPHO = 0
VBP
HBP
LCD_PCLK
LCD_VSYNC
LCD_HSYNC
VSW VSPHO = 0
VBP
HBP
LCD_PCLK
LCD_VSYNC
LCD_HSYNC
HSW
VSW
VBP
HBP

545
Figure 31-17. Vertical Synchronization Timing (part 2)

VSPDLYS = 1 VSPDLYE = 0 VSPSU = 0 VSPHO = 1
LCD_PCLK
LCD_VSYNC
LCD_HSYNC
VSW VSPHO = 1
VBP
HBP
LCD_PCLK
LCD_VSYNC
LCD_HSYNC
HSW
VSW
VBP
HBP

546
Figure 31-18. DISP Signal Timing Diagram

VSPDLYE = 0 VSPHO = 0 DISPPOL = 0 DISPDLY = 0 LCD_PCLK
LCD_VSYNC
LCD_HSYNC lcd display off lcd display on
LCD_DISP
LCD_VSYNC
LCD_HSYNC lcd display off
LCD_DISP
lcd display on
VSPDLYE = 0 VSPHO = 0 DISPPOL = 0 DISPDLY = 1
LCD_PCLK
LCD_VSYNC
LCD_HSYNC lcd display off lcd display on
LCD_DISP
LCD_VSYNC
LCD_HSYNC lcd display on lcd display off
LCD_DISP

547
31.6.18 Output Format

31.6.18.1 Active Mode Output Pin Assignment
Table 31-53. Active Mode Output with 24 bits Bus Interface Configuration Pin ID LCD_DAT[23] LCD_DAT[22] LCD_DAT[21] LCD_DAT[20] LCD_DAT[19] LCD_DAT[18] LCD_DAT[17] LCD_DAT[16] LCD_DAT[15] LCD_DAT[14] LCD_DAT[13] LCD_DAT[12] LCD_DAT[11] LCD_DAT[10] LCD_DAT[9] LCD_DAT[8] LCD_DAT[7] LCD_DAT[6] LCD_DAT[5] LCD_DAT[4] LCD_DAT[3] LCD_DAT[2] LCD_DAT[1] LCD_DAT[0] TFT 24 bits R[7] R[6] R[5] R[4] R[3] R[2] R[1] R[0] G[7] G[6] G[5] G[4] G[3] G[2] G[1] G[0] B[7] B[6] B[5] B[4] B[3] B[2] B[1] B[0] TFT 18 bits R[5] R[4] R[3] R[2] R[1] R[0] G[5] G[4] G[3] G[2] G[1] G[0] B[5] B[4] B[3] B[2] B[1] B[0] TFT 16 bits R[4] R[3] R[2] R[1] R[0] G[5] G[4] G[3] G[2] G[1] G[0] B[4] B[3] B[2] B[1] B[0] TFT 12 bits R[3] R[2] R[1] R[0] G[3] G[2] G[1] G[0] B[3] B[2] B[1] B[0]

548
31.7
LCD Controller (LCDC) User Interface

Offset Register LCD Controller Configuration Register 0 LCD Controller Configuration Register 1 LCD Controller Configuration Register 2 LCD Controller Configuration Register 3 LCD Controller Configuration Register 4 LCD Controller Configuration Register 5 LCD Controller Configuration Register 6 Reserved LCD Controller Enable Register LCD Controller Disable Register LCD Controller Status Register LCD Controller Interrupt Enable Register LCD Controller Interrupt Disable Register LCD Controller Interrupt Mask Register LCD Controller Interrupt Status Register Reserved Base Layer Channel Enable Register Base Layer Channel Disable Register Base Layer Channel Status Register Base Layer Interrupt Enable Register Base Layer Interrupt Disabled Register Base Layer Interrupt Mask Register Base Layer Interrupt status Register Base DMA Head Register Base DMA Address Register Base DMA Control Register Base DMA Next Register Base Configuration register 0 Base Configuration register 1 Base Configuration register 2 Base Configuration register 3 Base Configuration register 4 Base Configuration register 5 Base Configuration register 6 Reserved Overlay 1 Channel Enable Register Overlay 1 Channel Disable Register Overlay 1 Channel Status Register Name LCDC_LCDCFG0 LCDC_LCDCFG1 LCDC_LCDCFG2 LCDC_LCDCFG3 LCDC_LCDCFG4 LCDC_LCDCFG5 LCDC_LCDCFG6 LCDC_LCDEN LCDC_LCDDIS LCDC_LCDSR LCDC_LCDIER LCDC_LCDIDR LCDC_LCDIMR LCDC_LCDISR LCDC_BASECHER LCDC_BASECHDR LCDC_BASECHSR LCDC_BASEIER LCDC_BASEIDR LCDC_BASEIMR LCDC_BASEISR LCDC_BASEHEAD LCDC_BASEADDR LCDC_BASECTRL LCDC_BASENEXT LCDC_BASECFG0 LCDC_BASECFG1 LCDC_BASECFG2 LCDC_BASECFG3 LCDC_BASECFG4 LCDC_BASECFG5 LCDC_BASECFG6 LCDC_OVR1CHER LCDC_OVR1CHDR LCDC_OVR1CHSR Access Read-write Read-write Read-write Read-write Read-write Read-write Read-write Write-only Write-only Read-only Write-only Write-only Read-only Read-only Write-only Write-only Read-only Write-only Write-only Read-only Read-only Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Write-only Write-only Read-only Reset 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000
Table 31-54. Register Mapping
0x00000000 0x00000004 0x00000008 0x0000000C 0x00000010 0x00000014 0x00000018 0x0000001C 0x00000020 0x00000024 0x00000028 0x0000002C 0x00000030 0x00000034 0x00000038 0x0000003C 0x00000040 0x00000044 0x00000048 0x0000004C 0x00000050 0x00000054 0x00000058 0x0000005C 0x00000060 0x00000064 0x00000068 0x0000006C 0x00000070 0x00000074 0x00000078 0x0000007C 0x00000080 0x00000084 0x88-0x13C 0x00000140 0x00000144 0x00000148

549
Table 31-54. Register Mapping (Continued) 0x0000014C 0x00000150 0x00000154 0x00000158 0x0000015C 0x00000160 0x00000164 0x00000168 0x0000016C 0x00000170 0x00000174 0x00000178 0x0000017C 0x00000180 0x00000184 0x00000188 0x0000018C 0x00000190 0x194-0x23C 0x00000240 0x00000244 0x00000248 0x0000024C 0x00000250 0x00000254 0x00000258 0x0000025C 0x00000260 0x00000264 0x00000268 0x0000026C 0x00000270 0x00000274 0x00000278 0x0000027C 0x00000280 0x00000284 0x00000288 0x0000028C 0x00000290 Overlay 1 Interrupt Enable Register Overlay 1 Interrupt Disable Register Overlay 1 Interrupt Mask Register Overlay 1 Interrupt Status Register Overlay 1 DMA Head Register Overlay 1 DMA Address Register Overlay1 DMA Control Register Overlay1 DMA Next Register Overlay 1 Configuration 0 Register Overlay 1 Configuration 1 Register Overlay 1 Configuration 2 Register Overlay 1 Configuration 3 Register Overlay 1 Configuration 4 Register Overlay 1 Configuration 5 Register Overlay 1 Configuration 6 Register Overlay 1 Configuration 7 Register Overlay 1 Configuration 8Register Overlay 1 Configuration 9 Register Reserved Overlay 2 Channel Enable Register Overlay 2 Channel Disable Register Overlay 2 Channel Status Register Overlay 2 Interrupt Enable Register Overlay 2 Interrupt Disable Register Overlay 2 Interrupt Mask Register Overlay 2 Interrupt status Register Overlay 2 DMA Head Register Overlay 2 DMA Address Register Overlay 2 DMA Control Register Overlay 2 DMA Next Register Overlay 2 Configuration 0 Register Overlay 2 Configuration 1 Register Overlay 2 Configuration 2 Register Overlay 2 Configuration 3 Register Overlay 2 Configuration 4 Register Overlay 2 Configuration 5 Register Overlay 2 Configuration 6 Register Overlay 2 Configuration 7 Register Overlay 2 Configuration 8 Register Overlay 2 Configuration 9 Register LCDC_OVR1IER LCDC_OVR1IDR LCDC_OVR1IMR LCDC_OVR1ISR LCDC_OVR1HEAD LCDC_OVR1ADDR LCDC_OVR1CTRL LCDC_OVR1NEXT LCDC_OVR1CFG0 LCDC_OVR1CFG1 LCDC_OVR1CFG2 LCDC_OVR1CFG3 LCDC_OVR1CFG4 LCDC_OVR1CFG5 LCDC_OVR1CFG6 LCDC_OVR1CFG7 LCDC_OVR1CFG8 LCDC_OVR1CFG9 LCDC_OVR2CHER LCDC_OVR2CHDR LCDC_OVR2CHSR LCDC_OVR2IER LCDC_OVR2IDR LCDC_OVR2IMR LCDC_OVR2ISR LCDC_OVR2HEAD LCDC_OVR2ADDR LCDC_OVR2CTRL LCDC_OVR2NEXT LCDC_OVR2CFG0 LCDC_OVR2CFG1 LCDC_OVR2CFG2 LCDC_OVR2CFG3 LCDC_OVR2CFG4 LCDC_OVR2CFG5 LCDC_OVR2CFG6 LCDC_OVR2CFG7 LCDC_OVR2CFG8 LCDC_OVR2CFG9 Write-only Write-only Read-only Read-only Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Write-only Write-only Read-only Write-only Write-only Read-only Read-only Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000

550
Table 31-54. Register Mapping (Continued) 0x294-33C 0x00000340 0x00000344 0x00000348 0x0000034C 0x00000350 0x00000354 0x00000358 0x0000035C 0x00000360 0x00000364 0x00000368 0x0000036C 0x00000370 0x00000374 0x00000378 0x0000037C 0x00000380 0x00000384 0x00000388 0x0000038C 0x00000390 0x00000394 0x00000398 0x0000039C 0x000003A0 0x000003A4 0x000003A8 0x000003AC 0x000003B0 0x000003B4 0x000003B8 0x000003BC 0x000003C0 0x000003C4 0x000003C8 0x000003CC 0x000003D0 0x000003D4 0x000003D8 Reserved High-End Overlay Channel Enable Register High-End Overlay Channel Disable Register High-End Overlay Channel Status Register High-End Overlay Interrupt Enable Register High-End Overlay Interrupt Disable Register High-End Overlay Interrupt Mask Register High-End Overlay Interrupt Status Register High-End Overlay DMA Head Register High-End Overlay DMA Address Register High-End Overlay DMA Control Register High-End Overlay DMA Next Register High-End Overlay U DMA Head Register High-End Overlay U DMA Address Register High-End Overlay U DMA control Register High-End Overlay U DMA Next Register High-End Overlay V DMA Head Register High-End Overlay V DMA Address Register High-End Overlay V DMA control Register High-End Overlay VDMA Next Register High-End Overlay Configuration Register 0 High-End Overlay Configuration Register 1 High-End Overlay Configuration Register 2 High-End Overlay Configuration Register 3 High-End Overlay Configuration Register 4 High-End Overlay Configuration Register 5 High-End Overlay Configuration Register 6 High-End Overlay Configuration Register 7 High-End Overlay Configuration Register 8 High-End Overlay Configuration Register 9 High-End Overlay Configuration Register 10 High-End Overlay Configuration Register 11 High-End Overlay Configuration Register 12 High-End Overlay Configuration Register 13 High-End Overlay Configuration Register 14 High-End Overlay Configuration Register 15 High-End Overlay Configuration Register 16 High-End Overlay Configuration Register 17 High-End Overlay Configuration Register 18 High-End Overlay Configuration Register 19 LCDC_HEOCHER LCDC_HEOCHDR LCDC_HEOCHSR LCDC_HEOIER LCDC_HEOIDR LCDC_HEOIMR LCDC_HEOISR LCDC_HEOHEAD LCDC_HEOADDR LCDC_HEOCTRL LCDC_HEONEXT LCDC_HEOUHEAD LCDC_HEOUADDR LCDC_HEOUCTRL LCDC_HEOUNEXT LCDC_HEOVHEAD LCDC_HEOVADDR LCDC_HEOVCTRL LCDC_HEOVNEXT LCDC_HEOCFG0 LCDC_HEOCFG1 LCDC_HEOCFG2 LCDC_HEOCFG3 LCDC_HEOCFG4 LCDC_HEOCFG5 LCDC_HEOCFG6 LCDC_HEOCFG7 LCDC_HEOCFG8 LCDC_HEOCFG9 LCDC_HEOCFG10 LCDC_HEOCFG11 LCDC_HEOCFG12 LCDC_HEOCFG13 LCDC_HEOCFG14 LCDC_HEOCFG15 LCDC_HEOCFG16 LCDC_HEOCFG17 LCDC_HEOCFG18 LCDC_HEOCFG19 Write-only Write-only Read-only Write-only Write-only Read-only Read-only Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000

551
Table 31-54. Register Mapping (Continued) 0x000003DC 0x000003E0 0x000003E4 0x000003E8 0x000003EC 0x000003F0 0x000003F4 0x000003F8 0x000003FC 0x00000400 0x00000404 0x00000408 0x0000040C 0x00000410 0x00000414 0x00000418 0x0000041C 0x00000420 0x00000424 0x00000428 0x0000042C 0x00000430 0x434-0x43C 0x00000440 0x00000444 0x00000448 0x0000044C 0x00000450 0x00000454 0x00000458 0x0000045C 0x00000460 0x00000464 0x00000468 0x0000046C 0x00000470 0x00000474 0x00000478 0x0000047C 0x00000480 High-End Overlay Configuration Register 20 High-End Overlay Configuration Register 21 High-End Overlay Configuration Register 22 High-End Overlay Configuration Register 23 High-End Overlay Configuration Register 24 High-End Overlay Configuration Register 25 High-End Overlay Configuration Register 26 High-End Overlay Configuration Register 27 High-End Overlay Configuration Register 28 High-End Overlay Configuration Register 29 High-End Overlay Configuration Register 30 High-End Overlay Configuration Register 31 High-End Overlay Configuration Register 32 High-End Overlay Configuration Register 33 High-End Overlay Configuration Register 34 High-End Overlay Configuration Register 35 High-End Overlay Configuration Register 36 High-End Overlay Configuration Register 37 High-End Overlay Configuration Register 38 High-End Overlay Configuration Register 39 High-End Overlay Configuration Register 40 High-End Overlay Configuration Register 41 Reserved Hardware Cursor Channel Enable Register Hardware Cursor Channel disable Register Hardware Cursor Channel Status Register Hardware Cursor Interrupt Enable Register Hardware Cursor Interrupt Disable Register Hardware Cursor Interrupt Mask Register Hardware Cursor Interrupt Status Register Hardware Cursor DMA Head Register Hardware cursor DMA Address Register Hardware Cursor DMA Control Register Hardware Cursor DMA NExt Register Hardware Cursor Configuration 0 Register Hardware Cursor Configuration 1 Register Hardware Cursor Configuration 2 Register Hardware Cursor Configuration 3 Register Hardware Cursor Configuration 4 Register Reserved LCDC_HEOCFG20 LCDC_HEOCFG21 LCDC_HEOCFG22 LCDC_HEOCFG23 LCDC_HEOCFG24 LCDC_HEOCFG25 LCDC_HEOCFG26 LCDC_HEOCFG27 LCDC_HEOCFG28 LCDC_HEOCFG29 LCDC_HEOCFG30 LCDC_HEOCFG31 LCDC_HEOCFG32 LCDC_HEOCFG33 LCDC_HEOCFG34 LCDC_HEOCFG35 LCDC_HEOCFG36 LCDC_HEOCFG37 LCDC_HEOCFG38 LCDC_HEOCFG39 LCDC_HEOCFG40 LCDC_HEOCFG41 LCDC_HCRCHER LCDC_HCRCHDR LCDC_HCRCHSR LCDC_HCRIER LCDC_HCRIDR LCDC_HCRIMR LCDC_HCRISR LCDC_HCRHEAD LCDC_HCRADDR LCDC_HCRCTRL LCDC_HCRNEXT LCDC_HCRCFG0 LCDC_HCRCFG1 LCDC_HCRCFG2 LCDC_HCRCFG3 LCDC_HCRCFG4 Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Write-only Write-only Read-only Write-only Write-only Read-only Read-only Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000

552
Table 31-54. Register Mapping (Continued) 0x00000484 0x00000488 0x0000048C 0x00000490 0x494-0x53C 0x00000540 0x00000544 0x00000548 0x0000054C 0x00000550 0x00000554 0x00000558 0x0000055C 0x00000560 0x00000564 0x00000568 0x0000056C 0x00000570 0x00000574 0x00000578 0x0000057C 0x00000580 0x584-0x5FC 0x600 ... 0x8FC 0xA00 ... 0xDFC 0xE00 ... 0x11FC 0x1200 ... 0x15FC 0x1600 ... 0x19FC Hardware Cursor Configuration 6 Register Hardware Cursor Configuration 7 Register Hardware Cursor Configuration 8 Register Hardware Cursor Configuration 9 Register Reserved Post Processing Channel Enable Register Post Processing Channel Disable Register Post Processing Channel Status Register Post Processing Interrupt Enable Register Post Processing Interrupt Disable Register Post Processing Interrupt Mask Register Post Processing Interrupt Status Register Post Processing Head Register Post Processing Address Register Post Processing Control Register Post Processing Next Register Post Processing Configuration Register 0 Post Processing Configuration Register 1 Post Processing Configuration Register 2 Post Processing Configuration Register 3 Post Processing Configuration Register 4 Post Processing Configuration Register 5 Reserved Base CLUT Register 0 ... Base CLUT Register 255 Overlay 1 CLUT Register 0 ... Overlay 1 CLUT Register 255 Overlay 2 CLUT Register 0 ... Overlay 2 CLUT Register 255 High End Overlay CLUT Register 0 ... High End Overlay CLUT Register 255 Hardware Cursor CLUT Register 0 ... Hardware Cursor CLUT Register 255 LCDC_HCRCFG6 LCDC_HCRCFG7 LCDC_HCRCFG8 LCDC_HCRCFG9 LCDC_PPCHER LCDC_PPCHDR LCDC_PPCHSR LCDC_PPIER LCDC_PPIDR LCDC_PPIMR LCDC_PPISR LCDC_PPHEAD LCDC_PPADDR LCDC_PPCTRL LCDC_PPNEXT LCDC_PPCFG0 LCDC_PPCFG1 LCDC_PPCFG2 LCDC_PPCFG3 LCDC_PPCFG4 LCDC_PPCFG5 LCDC_BASECLUT0 ... LCDC_BASECLUT255 LCDC_OVR1CLUT0 ... LCDC_OVR1CLUT255 LCDC_OVR2CLUT0 ... LCDC_OVR2CLUT255 LCDC_HEOCLUT0 ... LCDC_HEOCLUT255 LCDC_HCRCLUT0 ... LCDC_HCRCLUT255 Read-write Read-write Read-write Read-write Write-only Write-only Read-only Write-only Write-only Read-only Read-only Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write ... Read-write Read-write ... Read-write Read-write ... Read-write Read-write ... Read-write Read-write ... Read-write 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 ... 0x00000000 0x00000000 ... 0x00000000 0x00000000 ... 0x00000000 0x00000000 ... 0x00000000 0x00000000 ... 0x00000000
0x1A00-0x1FE4 Reserved
Note:
The CLUT registers are located in the RAM.

553
31.7.1 LCD Controller Configuration Register 0

31 23 15 7
LCDC_LCDCFG0 0xF0030000 Read-write 0x00000000

30 22 14 6 29 21 13 CGDISPP 5 28 20 27 19 26 18 10 CGDISOVR2 2 CLKSEL 25 17 9 CGDISOVR1 1 24 16 8 CGDISBASE 0 CLKPOL
CLKDIV 12 11 CGDISHCR CGDISHEO 4 3 CLKPWMSEL
CLKPOL: LCD Controller Clock Polarity

0: Data/Control signals are launched on the rising edge of the Pixel Clock. 1: Data/Control signals are launched on the falling edge of the Pixel Clock.
CLKSEL: LCD Controller Clock Source Selection

0: The Asynchronous output stage of the LCD controller is fed by the System Clock. 1: The Asynchronous output state of the LCD controller is fed by the 2x System Clock.
CLKPWMSEL: LCD Controller PWM Clock Source Selection

0: The slow clock is selected and feeds the PWM module. 1: The system clock is selected and feeds the PWM module.
CGDISBASE: Clock Gating Disable Control for the Base Layer

0: Automatic Clock Gating is enabled for the Base Layer. 1: Clock is running continuously.
CGDISOVR1: Clock Gating Disable Control for the Overlay 1 Layer

0: Automatic Clock Gating is enabled for the Overlay 1 Layer. 1: Clock is running continuously.
CGDISOVR2: Clock Gating Disable Control for the Overlay 2 Layer

0: Automatic Clock Gating is enabled for the Overlay 2 Layer. 1: Clock is running continuously.
CGDISHEO: Clock Gating Disable Control for the High End Overlay
0: Automatic Clock Gating is enabled for the High End Overlay Layer. 1: Clock is running continuously.
CGDISHCR: Clock Gating Disable Control for the Hardware Cursor Layer
0: Automatic Clock Gating is enabled for the Hardware Cursor Layer. 1: Clock is running continuously.

554
CGDISPP: Clock Gating Disable Control for the Post Processing Layer
0: Automatic Clock Gating is enabled for the Post Processing Layer. 1: Clock is running continuously.
CLKDIV: LCD Controller Clock Divider

8-bit width clock divider for pixel clock LCD_PCLK. pixel_clock = selected_clock / (CLKDIV+2) where selected_clock is equal to system_clock when CLKSEL field is set to 0 and system_clock2x when CLKSEL is set to 1.

555

31 23 15 7

30 22 14 6 29 21 13 5 28 20 12 4 27 19 VSPW 11 3 HSPW 10 2 9 1 8 0 26 18 25 17 24 16
HSPW: Horizontal Synchronization Pulse Width

Width of the LCD_HSYNC pulse, given in pixel clock cycles. Width is (HSPW+1) LCD_PCLK cycles.
VSPW: Vertical Synchronization Pulse Width

Width of the LCD_VSYNC pulse, given in number of lines. Width is (VSPW+1) lines.

556

31 23 15 7

30 22 14 6 29 21 13 5 28 20 12 4 27 19 VBPW 11 3 VFPW 10 2 9 1 8 0 26 18 25 17 24 16
VFPW: Vertical Front Porch Width

This field indicates the number of lines at the end of the Frame. The blanking interval is equal to (VFPW+1) lines.
VBPW: Vertical Back Porch Width

This field indicates the number of lines at the beginning of the Frame. The blanking interval is equal to VBPW lines.

557

31 23 15 7
LCDC_LCDCFG3 0xF003000C Read-write 0x00000000

30 22 14 6 29 21 13 5 28 20 HBPW 12 4 HFPW 11 3 10 2 9 1 8 HFPW 0 27 19 26 18 25 17 24 HBPW 16
HFPW: Horizontal Front Porch Width

Number of pixel clock cycles inserted at the end of the active line. The interval is equal to (HFPW+1) LCD_PCLK cycles.
HBPW: Horizontal Back Porch Width

Number of pixel clock cycles inserted at the beginning of the line. The interval is equal to (HBPW+1) LCD_PCLK cycles.

558

31 23 15 7

30 22 14 6 29 21 13 5 28 20 RPF 12 4 PPL 11 3 10 2 9 PPL 1 8 0 27 19 26 18 25 RPF 17 24 16
RPF: Number of Active Row Per Frame

Number of active lines in the frame. The frame height is equal to (RPF+1) lines.
PPL: Number of Pixels Per Line

Number of pixel in the frame. The number of active pixels in the frame is equal to (PPL+1) pixels.

559

31 23 15 7 DISPDLY

30 22 14 6 DITHER 29 21 13 VSPHO 5 28 20 12 VSPSU 4 DISPPOL 27 19 11 3 VSPDLYE 26 18 GUARDTIME 10 PP 2 VSPDLYS 25 17 9 MODE 1 VSPOL 0 HSPOL 24 16 8
HSPOL: Horizontal Synchronization Pulse Polarity

0: Active High 1: Active Low
VSPOL: Vertical Synchronization Pulse Polarity

VSPDLYS: Vertical Synchronization Pulse Start

0: The first active edge of the Vertical synchronization pulse is synchronous with the second edge of the horizontal pulse. 1: The first active edge of the Vertical synchronization pulse is synchronous with the first edge of the horizontal pulse.
VSPDLYE: Vertical Synchronization Pulse End

0: The second active edge of the Vertical synchronization pulse is synchronous with the second edge of the horizontal pulse. 1: The second active edge of the Vertical synchronization pulse is synchronous with the first edge of the horizontal pulse.
DISPPOL: Display Signal Polarity

DITHER: LCD Controller Dithering

0: Dithering logical unit is disabled. 1: Dithering logical unit is activated.
DISPDLY: LCD Controller Display Power Signal Synchronization

0: The LCD_DISP signal is asserted synchronously with the second active edge of the horizontal pulse. 1: The LCD_DISP signal is asserted asynchronously with both edges of the horizontal pulse.

560
MODE: LCD Controller Output Mode

Value 0 1 2 3 Name OUTPUT_12BPP OUTPUT_16BPP OUTPUT_18BPP OUTPUT_24BPP Description LCD output mode is set to 12 bits per pixel LCD output mode is set to 16 bits per pixel LCD output mode is set to 18 bits per pixel LCD output mode is set to 24 bits per pixel
PP: Post Processing Enable

0: The Blended pixel is pushed into the output FIFO. 1: The Blended pixel is written back to memory, the post processing stage is enabled.
VSPSU: LCD Controller Vertical synchronization Pulse Setup Configuration

0: The vertical synchronization pulse is asserted synchronously with horizontal pulse edge. 1: The vertical synchronization pulse is asserted one pixel clock cycle before the horizontal pulse.
VSPHO: LCD Controller Vertical synchronization Pulse Hold Configuration

0: The vertical synchronization pulse is asserted synchronously with horizontal pulse edge. 1: The vertical synchronization pulse is held active one pixel clock cycle after the horizontal pulse.
GUARDTIME: LCD DISPLAY Guard Time

Number of frame inserted during start up before LCD_DISP assertion. Number of frame inserted after LCD_DISP reset.

561

31 23 15 7

30 22 14 6 29 21 13 5 28 20 12 27 19 11 26 18 10 2 25 17 9 1 PWMPS 24 16 8 0
PWMCVAL 4 3 PWMPOL
PWMPS: PWM Clock Prescaler

3-bit value. Selects the configuration of the counter prescaler module. The PWMPS field decoding is listed below.
Value 000 001 010 011 100 101 110 DIV_1 DIV_2 DIV_4 DIV_8 DIV_16 DIV_32 DIV_64 Name Description The counter advances at a rate of fCOUNTER = fPWM_SELECTED_CLOCK The counter advances at a rate of fCOUNTER = fPWM_SELECTED_CLOCK/2 The counter advances at a rate of fCOUNTER = fPWM_SELECTED_CLOCK/4 The counter advances at a rate of fCOUNTER = fPWM_SELECTED_CLOCK/8 The counter advances at a rate of fCOUNTER = fPWM_SELECTED_CLOCK/16 The counter advances at a of rate fCOUNTER = fPWM_SELECTED_CLOCK/32 The counter advances at a of rate fCOUNTER = fPWM_SELECTED_CLOCK/64
PWMPOL: LCD Controller PWM Signal Polarity

This bit defines the polarity of the PWM output signal. If set to one, the output pulses are high level (the output will be high whenever the value in the counter is less than the value CVAL) If set to zero, the output pulses are low level.
PWMCVAL: LCD Controller PWM Compare Value

PWM compare value. Used to adjust the analog value obtained after an external filter to control the contrast of the display.

562
31.7.8 LCD Controller Enable Register

31 23 15 7
LCDC_LCDEN 0xF0030020 Write 0x00000000

30 22 14 6 29 21 13 5 28 20 12 4 27 19 11 3 PWMEN 26 18 10 2 DISPEN 25 17 9 1 SYNCEN 24 16 8 0 CLKEN
CLKEN: LCD Controller Pixel Clock Enable

0: Writing this field to zero has no effect. 1: When set to one the pixel clock logical unit is activated.
SYNCEN: LCD Controller Horizontal and Vertical Synchronization Enable

0: Writing this field to zero has no effect. 1: When set to one, both horizontal and vertical synchronization (LCD_VSYNC and LCD_HSYNC) signals are generated.
DISPEN: LCD Controller DISP Signal Enable

0: Writing this field to zero has no effect. 1: When set to one, LCD_DISP signal is generated.
PWMEN: LCD Controller Pulse Width Modulation Enable

0: Writing this field to zero has no effect. 1: When set to one, the pwm is enabled.

563
31.7.9 LCD Controller Disable Register

31 23 15 7
LCDC_LCDDIS 0xF0030024 Write 0x00000000

30 22 14 6 29 21 13 5 28 20 12 4 27 19 11 PWMRST 3 PWMDIS 26 18 10 DISPRST 2 DISPDIS 25 17 9 SYNCRST 1 SYNCDIS 24 16 8 CLKRST 0 CLKDIS
CLKDIS: LCD Controller Pixel Clock Disable

0: No effect. 1: Disables the pixel clock.
SYNCDIS: LCD Controller Horizontal and Vertical Synchronization Disable

0: No effect. 1: Disables the synchronization signals after the end of the frame.
DISPDIS: LCD Controller DISP Signal Disable

0: No effect 1: Disables the DISP signal.
PWMDIS: LCD Controller Pulse Width Modulation Disable

0: No effect 1: Disables the pulse width modulation signal.
CLKRST: LCD Controller Clock Reset

0: No effect. 1: Resets the pixel clock generator module. The pixel clock duty cycle may be violated.
SYNCRST: LCD Controller Horizontal and Vertical Synchronization Reset

0: No effect. 1: Resets the timing engine. Both Horizontal and vertical pulse width are violated.
DISPRST: LCD Controller DISP Signal Reset

0: No effect. 1: Resets the DISP signal.
PWMRST: LCD Controller PWM Reset

0: No effect. 1: Resets the PWM module, the duty cycle may be violated.

564
31.7.10 LCD Controller Status Register

31 23 15 7
LCDC_LCDSR 0xF0030028 Read-only 0x00000000

30 22 14 6 29 21 13 5 28 20 12 4 SIPSTS 27 19 11 3 PWMSTS 26 18 10 2 DISPSTS 25 17 9 1 LCDSTS 24 16 8 0 CLKSTS
CLKSTS: Clock Status

0: Pixel Clock is disabled. 1: Pixel Clock is running.
LCDSTS: LCD Controller Synchronization status

0: Timing Engine is disabled. 1: Timing Engine is running.
DISPSTS: LCD Controller DISP Signal Status

0: DISP is disabled. 1: DISP signal is activated.
PWMSTS: LCD Controller PWM Signal Status

0: PWM is disabled. 1: PWM signal is activated.
SIPSTS: Synchronization In Progress

0: Clock domain synchronization is terminated. 1: A double domain synchronization is in progress, access to the LCDC_LCDEN and LCDC_LCDDIS registers has no effect.

565
31.7.11 LCD Controller Interrupt Enable Register

31 23 15 7
LCDC_LCDIER 0xF003002C Write-only 0x00000000

30 22 14 6 29 21 13 PPIE 5 28 20 12 HCRIE 4 FIFOERRIE 27 19 11 HEOIE 3 26 18 10 OVR2IE 2 DISPIE 25 17 9 OVR1IE 1 DISIE 24 16 8 BASEIE 0 SOFIE
SOFIE: Start of Frame Interrupt Enable Register

0: No effect 1: Enables the interrupt.
DISIE: LCD Disable Interrupt Enable Register

DISPIE: Power UP/Down Sequence Terminated Interrupt Enable Register

FIFOERRIE: Output FIFO Error Interrupt Enable Register

BASEIE: Base Layer Interrupt Enable Register

OVR1IE: Overlay 1 Interrupt Enable Register

OVR2IE: Overlay 2 Interrupt Enable Register

HEOIE: High End Overlay Interrupt Enable Register


566
HCRIE: Hardware Cursor Interrupt Enable Register

PPIE: Post Processing Interrupt Enable Register


567
31.7.12 LCD Controller Interrupt Disable Register

31 23 15 7
LCDC_LCDIDR 0xF0030030 Write-only 0x00000000

30 22 14 6 29 21 13 PPID 5 28 20 12 HCRID 4 FIFOERRID 27 19 11 HEOID 3 26 18 10 OVR2ID 2 DISPID 25 17 9 OVR1ID 1 DISID 24 16 8 BASEID 0 SOFID
SOFID: Start of Frame Interrupt Disable Register

0: No effect 1: Disables the interrupt.
DISID: LCD Disable Interrupt Disable Register

DISPID: Power UP/Down Sequence Terminated Interrupt Disable Register

FIFOERRID: Output FIFO Error Interrupt Disable Register

BASEID: Base Layer Interrupt Disable Register

OVR1ID: Overlay 1 Interrupt Disable Register

OVR2ID: Overlay 2 Interrupt Disable Register

HEOID: High End Overlay Interrupt Disable Register


568
HCRID: Hardware Cursor Interrupt Disable Register

PPID: Post Processing Interrupt Disable Register


569
31.7.13 LCD Controller Interrupt Mask Register

31 23 15 7
LCDC_LCDIMR 0xF0030034 Read-only 0x00000000

30 22 14 6 29 21 13 PPIM 5 28 20 12 HCRIM 4 FIFOERRIM 27 19 11 HEOIM 3 26 18 10 OVR2IM 2 DISPIM 25 17 9 OVR1IM 1 DISIM 24 16 8 BASEIM 0 SOFIM
SOFIM: Start of Frame Interrupt Mask Register

DISIM: LCD Disable Interrupt Mask Register

DISPIM: Power UP/Down Sequence Terminated Interrupt Mask Register

FIFOERRIM: Output FIFO Error Interrupt Mask Register

BASEIM: Base Layer Interrupt Mask Register

OVR1IM: Overlay 1 Interrupt Mask Register

OVR2IM: Overlay 2 Interrupt Mask Register

HEOIM: High End Overlay Interrupt Mask Register


570
HCRIM: Hardware Cursor Interrupt Mask Register

PPIM: Post Processing Interrupt Mask Register


571
31.7.14 LCD Controller Interrupt Status Register

31 23 15 7
LCDC_LCDISR 0xF0030038 Read-only 0x00000000

30 22 14 6 29 21 13 PP 5 28 20 12 HCR 4 FIFOERR 27 19 11 HEO 3 26 18 10 OVR2 2 DISP 25 17 9 OVR1 1 DIS 24 16 8 BASE 0 SOF
SOF: Start of Frame Interrupt Status Register

When set to one this flag indicates that a start of frame event has been detected. This flag is reset after a read operation.
DIS: LCD Disable Interrupt Status Register

When set to one this flag indicates that the horizontal and vertical timing generator has been successfully disabled. This flag is reset after a read operation.
DISP: Power-up/Power-down Sequence Terminated Interrupt Status Register

When set to one this flag indicates whether the power-up sequence or power-down sequence has terminated. This flag is reset after a read operation.
FIFOERR: Output FIFO Error

When set to one this flag indicates that an underflow occurs in the output FIFO. This flag is reset after a read operation.
BASE: Base Layer Raw Interrupt Status Register

When set to one this flag indicates that a Base layer interrupt is pending. This flag is reset as soon as the BASEISR register is read.
OVR1: Overlay 1 Raw Interrupt Status Register

When set to one this flag indicates that an Overlay 1 layer interrupt is pending. This flag is reset as soon as the OVR1ISR register is read.
OVR2: Overlay 2 Raw Interrupt Status Register

When set to one this flag indicates that an Overlay 1 layer interrupt is pending. This flag is reset as soon as the OVR1ISR register is read.
HEO: High End Overlay Raw Interrupt Status Register

When set to one this flag indicates that a Hi End layer interrupt is pending. This flag is reset as soon as the HEOISR register is read.
HCR: Hardware Cursor Raw Interrupt Status Register

When set to one this flag indicates that a Hardware Cursor layer interrupt is pending. This flag is reset as soon as the HCRISR register is read.
PP: Post Processing Raw Interrupt Status Register

When set to one this flag indicates that Post Processing interrupt is pending. This flag is reset as soon as the PPISR register is read.

572
31.7.15 Base Layer Channel Enable Register

31 23 15 7
LCDC_BASECHER 0xF0030040 Write-only 0x00000000

30 22 14 6 29 21 13 5 28 20 12 4 27 19 11 3 26 18 10 2 A2QEN 25 17 9 1 UPDATEEN 24 16 8 0 CHEN
CHEN: Channel Enable Register

0: No effect. 1: Enables the DMA channel.
UPDATEEN: Update Overlay Attributes Enable Register

0: No effect. 1: Updates windows attributes on the next start of frame.
A2QEN: Add Head Pointer Enable Register

Write this field to one to add the head pointer to the descriptor list. This field is reset by hardware as soon as the head register is added to the list.

573
31.7.16 Base Layer Channel Disable Register

31 23 15 7
LCDC_BASECHDR 0xF0030044 Write-only 0x00000000

30 22 14 6 29 21 13 5 28 20 12 4 27 19 11 3 26 18 10 2 25 17 9 1 24 16 8 CHRST 0 CHDIS
CHDIS: Channel Disable Register

When set to one this field disables the layer at the end of the current frame. The frame is completed.
CHRST: Channel Reset Register

When set to one this field resets the layer immediately. The frame is aborted.

574
31.7.17 Base Layer Channel Status Register

31 23 15 7
LCDC_BASECHSR 0xF0030048 Read-only 0x00000000

30 22 14 6 29 21 13 5 28 20 12 4 27 19 11 3 26 18 10 2 A2QSR 25 17 9 1 UPDATESR 24 16 8 0 CHSR
CHSR: Channel Status Register

When set to one this field disables the layer at the end of the current frame.
UPDATESR: Update Overlay Attributes In Progress

When set to one this bit indicates that the overlay attributes will be updated on the next frame.
A2QSR: Add To Queue Pending Register

When set to one this bit indicates that the head pointer is still pending.

575
31.7.18 Base Layer Interrupt Enable Register

31 23 15 7
LCDC_BASEIER 0xF003004C Write-only 0x00000000

30 22 14 6 OVR 29 21 13 5 DONE 28 20 12 4 ADD 27 19 11 3 DSCR 26 18 10 2 DMA 25 17 9 1 24 16 8 0
DMA: End of DMA Transfer Interrupt Enable Register

DSCR: Descriptor Loaded Interrupt Enable Register

ADD: Head Descriptor Loaded Interrupt Enable Register

DONE: End of List Interrupt Enable Register

OVR: Overflow Interrupt Enable Register


576
31.7.19 Base Layer Interrupt Disable Register

31 23 15 7
LCDC_BASEIDR 0xF0030050 Write-only 0x00000000

DMA: End of DMA Transfer Interrupt Disable Register

DSCR: Descriptor Loaded Interrupt Disable Register

ADD: Head Descriptor Loaded Interrupt Disable Register

DONE: End of List Interrupt Disable Register

0: No effect. 1: interrupt source is disabled.
OVR: Overflow Interrupt Disable Register


577
31.7.20 Base Layer Interrupt Mask Register

31 23 15 7
LCDC_BASEIMR 0xF0030054 Read-only 0x00000000

DMA: End of DMA Transfer Interrupt Mask Register

DSCR: Descriptor Loaded Interrupt Mask Register

ADD: Head Descriptor Loaded Interrupt Mask Register

DONE: End of List Interrupt Mask Register

OVR: Overflow Interrupt Mask Register


578
31.7.21 Base Layer Interrupt Status Register

31 23 15 7
LCDC_BASEISR 0xF0030058 Read-only 0x00000000

DMA: End of DMA Transfer

When set to one this flag indicates that an End of Transfer has been detected. This flag is reset after a read operation.
DSCR: DMA Descriptor Loaded

When set to one this flag indicates that a descriptor has been loaded successfully. This flag is reset after a read operation.
ADD: Head Descriptor Loaded

When set to one this flag indicates that the descriptor pointed to by the head register has been loaded successfully. This flag is reset after a read operation.
DONE: End of List Detected

When set to one this flag indicates that an End of List condition has occurred. This flag is reset after a read operation.
OVR: Overflow Detected

When set to one this flag indicates that an overflow occurred. This flag is reset after a read operation.

579
31.7.22 Base Layer Head Register

31 23 15 7
LCDC_BASEHEAD 0xF003005C Read-write 0x00000000

30 22 14 6 29 21 13 5 HEAD 28 HEAD 20 HEAD 12 HEAD 4 3 2 1 0 11 10 9 8 19 18 17 16 27 26 25 24
HEAD: DMA Head Pointer

The Head Pointer points to a new descriptor.
31.7.23 Base Layer Address Register

31 23 15 7
LCDC_BASEADDR 0xF0030060 Read-write 0x00000000

30 22 14 6 29 21 13 5 28 ADDR 20 ADDR 12 ADDR 4 ADDR 3 2 1 0 11 10 9 8 19 18 17 16 27 26 25 24
ADDR: DMA Transfer Start Address

Frame buffer base address.

580
31.7.24 Base Layer Control Register

31 23 15 7
LCDC_BASECTRL 0xF0030064 Read-write 0x00000000

30 22 14 6 29 21 13 5 DONEIEN 28 20 12 4 ADDIEN 27 19 11 3 DSCRIEN 26 18 10 2 DMAIEN 25 17 9 1 LFETCH 24 16 8 0 DFETCH
DFETCH: Transfer Descriptor Fetch Enable

0: Transfer Descriptor fetch is disabled. 1: Transfer Descriptor fetch is enabled.
LFETCH: Lookup Table Fetch Enable

0: Lookup Table DMA fetch is disabled. 1: Lookup Table DMA fetch is enabled.
DMAIEN: End of DMA Transfer Interrupt Enable

0: DMA transfer completed interrupt is enabled. 1: DMA transfer completed interrupt is disabled.
DSCRIEN: Descriptor Loaded Interrupt Enable

0: Transfer descriptor loaded interrupt is enabled. 1: Transfer descriptor loaded interrupt is disabled.
ADDIEN: Add Head Descriptor to Queue Interrupt Enable

0: Transfer descriptor added to queue interrupt is enabled. 1: Transfer descriptor added to queue interrupt is enabled.
DONEIEN: End of List Interrupt Enable

0: End of list interrupt is disabled. 1: End of list interrupt is enabled.

581
31.7.25 Base Layer Next Register

31 23 15 7
LCDC_BASENEXT 0xF0030068 Read-write 0x00000000

30 22 14 6 29 21 13 5 28 NEXT 20 NEXT 12 NEXT 4 NEXT 3 2 1 0 11 10 9 8 19 18 17 16 27 26 25 24
NEXT: DMA Descriptor Next Address

The transfer descriptor address must be aligned on a 64-bit boundary.

582
31.7.26 Base Layer Configuration 0 Register

31 23 15 7
LCDC_BASECFG0 0xF003006C Read-write 0x00000000

30 22 14 6 29 21 13 5 BLEN 28 20 12 4 27 19 11 3 26 18 10 2 25 17 9 1 24 16 8 DLBO 0 SIF
SIF: Source Interface

0: Base Layer data is retrieved through AHB interface 0. 1: Base Layer data is retrieved through AHB interface 1.
BLEN: AHB Burst Length

Value 0 Name AHB_SINGLE Description AHB Access is started as soon as there is enough space in the FIFO to store one data. SINGLE, INCR, INCR4, INCR8 and INCR16 bursts are used. INCR is used for a burst of 2 and 3 beats. AHB Access is started as soon as there is enough space in the FIFO to store a total amount of 4 data. An AHB INCR4 Burst is used. SINGLE, INCR and INCR4 bursts are used. INCR is used for a burst of 2 and 3 beats. AHB Access is started as soon as there is enough space in the FIFO to store a total amount of 8 data. An AHB INCR8 Burst is used. SINGLE, INCR, INCR4 and INCR8 bursts are used. INCR is used for a burst of 2 and 3 beats. AHB Access is started as soon as there is enough space in the FIFO to store a total amount of 16 data. An AHB INCR16 Burst is used. SINGLE, INCR, INCR4, INCR8 and INCR16 bursts are used. INCR is used for a burst of 2 and 3 beats.
AHB_INCR4
AHB_INCR8
AHB_INCR16
DLBO: Defined Length Burst Only For Channel Bus Transaction.

0: Undefined length INCR burst is used for a burst of 2 and 3 beats. 1: Only Defined Length burst is used (SINGLE, INCR4, INCR8 and INCR16).

583

31 23 15 7
LCDC_BASECFG1 0xF0030070 Read-write 0x00000000

30 22 14 6 RGBMODE 5 4 29 21 13 28 20 12 27 19 11 3 26 18 10 2 25 17 9 CLUTMODE 1 0 CLUTEN 24 16 8
CLUTEN: Color Lookup Table Enable

0: RGB mode is selected. 1: Color lookup table is selected.
RGBMODE: RGB Input Mode Selection

Value 0 1 2 3 4 5 6 7 8 9 10 11 12 13 Name 12BPP_RGB_444 16BPP_ARGB_4444 16BPP_RGBA_4444 16BPP_RGB_565 16BPP_TRGB_1555 18BPP_RGB_666 18BPP_RGB_666PACKED 19BPP_TRGB_1666 19BPP_TRGB_PACKED 24BPP_RGB_888 24BPP_RGB_888_PACKED 25BPP_TRGB_1888 32BPP_ARGB_8888 32BPP_RGBA_8888 Description 12 bpp RGB 444 16 bpp ARGB 4444 16 bpp RGBA 4444 16 bpp RGB 565 16 bpp TRGB 1555 18 bpp RGB 666 18 bpp RGB 666 PACKED 19 bpp TRGB 1666 19 bpp TRGB 1666 PACKED 24 bpp RGB 888 24 bpp RGB 888 PACKED 25 bpp TRGB 1888 32 bpp ARGB 8888 32 bpp RGBA 8888
CLUTMODE: Color Lookup Table Input Mode Selection

Value 0 1 2 3 Name CLUT_1BPP CLUT_2BPP CLUT_4BPP CLUT_8BPP Description color lookup table mode set to 1 bit per pixel color lookup table mode set to 2 bits per pixel color lookup table mode set to 4 bits per pixel color lookup table mode set to 8 bits per pixel

584

31 23 15 7

30 22 14 6 29 21 13 5 28 XSTRIDE 20 XSTRIDE 12 XSTRIDE 4 XSTRIDE 3 2 1 0 11 10 9 8 19 18 17 16 27 26 25 24
XSTRIDE: Horizontal Stride

XSTRIDE represents the memory offset, in bytes, between two rows of the image memory.

585

31 23 15 7

30 22 14 6 29 21 13 5 28 20 RDEF 12 GDEF 4 BDEF 3 2 1 0 11 10 9 8 27 19 26 18 25 17 24 16
RDEF: Red Default

Default Red color when the Base DMA channel is disabled.
GDEF: Green Default

Default Green color when the Base DMA channel is disabled.
BDEF: Blue Default

Default Blue color when the Base DMA channel is disabled.

586

31 23 15 7
LCDC_BASECFG4 0xF003007C Read-write 0x00000000

30 22 14 6 29 21 13 5 28 20 12 4 27 19 11 DISCEN 3 26 18 10 2 25 17 9 REP 1 24 16 8 DMA 0
DMA: Use DMA Data Path

0: The default color is used on the Base Layer. 1: The DMA channel retrieves the pixels stream from the memory.
REP: Use Replication logic to expand RGB color to 24 bits

0: When the selected pixel depth is less than 24 bpp the pixel is shifted and least significant bits are set to 0. 1: When the selected pixel depth is less than 24 bpp the pixel is shifted and the least significant bit replicates the msb.
DISCEN: Discard Area Enable

0: The whole frame is retrieved from memory. 1:When set to one the DMA channel discards the area located at screen coordinate {DISCXPOS, DISCYPOS}.

587

31 23 15 7

30 22 14 6 29 21 13 5 28 20 DISCYPOS 12 4 DISCXPOS 11 3 10 2 9 DISCXPOS 1 8 0 27 19 26 18 25 DISCYPOS 17 24 16
DISCXPOS: Discard Area horizontal coordinate

Horizontal Position of the Discard Area.
DISCYPOS: Discard Area Vertical coordinate

Vertical Position of the Discard Area.

588

31 23 15 7

30 22 14 6 29 21 13 5 28 20 DISCYSIZE 12 4 DISCXSIZE 11 3 10 2 9 DISCXSIZE 1 8 0 27 19 26 18 25 DISCYSIZE 17 24 16
DISCXSIZE: Discard Area Horizontal Size

Discard Horizontal size in pixels. The Discard size is set to (DISCXSIZE+1) pixels in horizontal.
DISCYSIZE: Discard Area Vertical Size

Discard Vertical size in pixels. The Discard size is set to (DISCYSIZE+1) pixels in vertical.

589
31.7.33 Overlay 1 Layer Channel Enable Register

31 23 15 7
LCDC_OVR1CHER 0xF0030140 Write-only 0x00000000



0: No effect. 1: Updates window attributes (size, alpha-blending, etc.) on the next start of frame.


590
31.7.34 Overlay 1 Layer Channel Disable Register

31 23 15 7
LCDC_OVR1CHDR 0xF0030144 Write-only 0x00000000

30 22 14 6 29 21 13 5 28 20 12 4 27 19 11 3 26 18 10 2 25 17 9 1 24 16 8 CHRST 0 CHDIS



591
31.7.35 Overlay 1 Layer Channel Status Register

31 23 15 7
LCDC_OVR1CHSR 0xF0030148 Read-only 0x00000000



A2QSR: Add to Queue Pending Register


592
31.7.36 Overlay 1 Layer Interrupt Enable Register

31 23 15 7
LCDC_OVR1IER 0xF003014C Write-only 0x00000000







593
31.7.37 Overlay 1 Layer Interrupt Disable Register

31 23 15 7
LCDC_OVR1IDR 0xF0030150 Write-only 0x00000000







594
31.7.38 Overlay 1 Layer Interrupt Mask Register

31 23 15 7
LCDC_OVR1IMR 0xF0030154 Read-only 0x00000000







595
31.7.39 Overlay 1 Layer Interrupt Status Register

31 23 15 7
LCDC_OVR1ISR 0xF0030158 Read-only 0x00000000




DONE: End of List Detected Register



596
31.7.40 Overlay 1 Layer Head Register

31 23 15 7
LCDC_OVR1HEAD 0xF003015C Read-write 0x00000000



597
31.7.41 Overlay 1 Layer Address Register

31 23 15 7
LCDC_OVR1ADDR 0xF0030160 Read-write 0x00000000

ADDR: DMA Transfer Overlay 1 Address

Overlay 1 frame buffer base address.

598
31.7.42 Overlay 1 Layer Control Register

31 23 15 7
LCDC_OVR1CTRL 0xF0030164 Read-write 0x00000000








599
31.7.43 Overlay 1 Layer Next Register

31 23 15 7
LCDC_OVR1NEXT 0xF0030168 Read-write 0x00000000



600
31.7.44 Overlay 1 Layer Configuration 0 Register

31 23 15 7
LCDC_OVR1CFG0 0xF003016C Read-write 0x00000000

30 22 14 6 29 21 13 LOCKDIS 5 BLEN 28 20 12 ROTDIS 4 27 19 11 3 26 18 10 2 25 17 9 1 24 16 8 DLBO 0 SIF


Value 0 Name AHB_BLEN_SINGLE Description AHB Access is started as soon as there is enough space in the FIFO to store one data. SINGLE, INCR, INCR4, INCR8 and INCR16 bursts are used. INCR is used for a burst of 2 and 3 beats. AHB Access is started as soon as there is enough space in the FIFO to store a total amount of 4 data. An AHB INCR4 Burst is used. SINGLE, INCR and INCR4 bursts are used. INCR is used for a burst of 2 and 3 beats. AHB Access is started as soon as there is enough space in the FIFO to store a total amount of 8 data. An AHB INCR8 Burst is used. SINGLE, INCR, INCR4 and INCR8 bursts are used. INCR is used for a burst of 2 and 3 beats.
AHB_BLEN_INCR4
AHB_BLEN_INCR8
AHB Access is started as soon as there is enough space in the FIFO to store a total amount of 16 data. AHB_BLEN_INCR16 An AHB INCR16 Burst is used. SINGLE, INCR, INCR4, INCR8 and INCR16 bursts are used. INCR is used for a burst of 2 and 3 beats.
DLBO: Defined Length Burst Only for Channel Bus Transaction.

ROTDIS: Hardware Rotation Optimization Disable

0: Rotation optimization is enabled. 1: Rotation optimization is disabled.
LOCKDIS: Hardware Rotation Lock Disable

0: AHB lock signal is asserted when a rotation is performed. 1: AHB lock signal is cleared when a rotation is performed.

601

31 23 15 7
LCDC_OVR1CFG1 0xF0030170 Read-write 0x00000000


0: RGB mode is selected. 1: Color lookup table is selected.

CLUTMODE: Color Lookup table input mode selection


602

31 23 15 7

30 22 14 6 29 21 13 5 28 20 YPOS 12 4 XPOS 11 3 10 2 9 XPOS 1 8 0 27 19 26 18 25 YPOS 17 24 16
XPOS: Horizontal Window Position

Overlay 1 Horizontal window position.
YPOS: Vertical Window Position

Overlay 1 Vertical window position.

603

31 23 15 7

30 22 14 6 29 21 13 5 28 20 YSIZE 12 4 XSIZE 11 3 10 2 9 XSIZE 1 8 0 27 19 26 18 25 YSIZE 17 24 16
XSIZE: Horizontal Window Size

Overlay 1 window width in pixels. The window width is set to (XSIZE+1). The following constraint must be met:
XPOS + XSIZE PPL
YSIZE: Vertical Window Size

Overlay 1 window height in pixels. The window height is set to (YSIZE+1). The following constrain must be met:
YPOS + YSIZE RPF

604

31 23 15 7



605

31 23 15 7

30 22 14 6 29 21 13 5 28 PSTRIDE 20 PSTRIDE 12 PSTRIDE 4 PSTRIDE 3 2 1 0 11 10 9 8 19 18 17 16 27 26 25 24
PSTRIDE: Pixel Stride

PSTRIDE represents the memory offset, in bytes, between two pixels of the image.

606

31 23 15 7

30 22 14 6 29 21 13 5 28 20 RDEF 12 GDEF 4 BDEF 3 2 1 0 11 10 9 8 27 19 26 18 25 17 24 16
RDEF: Red Default

Default Red color when the Overlay 1 DMA channel is disabled.
GDEF: Green Default

Default Green color when the Overlay 1 DMA channel is disabled.
BDEF: Blue Default

Default Blue color when the Overlay 1 DMA channel is disabled.

607

31 23 15 7

30 22 14 6 29 21 13 5 28 20 RKEY 12 GKEY 4 BKEY 3 2 1 0 11 10 9 8 27 19 26 18 25 17 24 16
RKEY: Red Color Component Chroma Key

Reference Red chroma key used to match the Red color of the current overlay.
GKEY: Green Color Component Chroma Key

Reference Green chroma key used to match the Green color of the current overlay.
BKEY: Blue Color Component Chroma Key

Reference Blue chroma key used to match the Blue color of the current overlay.

608

31 23 15 7

30 22 14 6 29 21 13 5 28 20 RMASK 12 GMASK 4 BMASK 3 2 1 0 11 10 9 8 27 19 26 18 25 17 24 16
RMASK: Red Color Component Chroma Key Mask

Red Mask used when the compare function is used. If a bit is set then this bit is compared.
GMASK: Green Color Component Chroma Key Mask

Green Mask used when the compare function is used. If a bit is set then this bit is compared.
BMASK: Blue Color Component Chroma Key Mask

Blue Mask used when the compare function is used. If a bit is set then this bit is compared.

609

31 23 15 7 OVR

30 22 14 6 LAEN 29 21 13 5 GAEN 28 20 GA 12 4 REVALPHA 11 3 ITER 10 DSTKEY 2 ITER2BL 9 REP 1 INV 8 DMA 0 CRKEY 27 19 26 18 25 17 24 16
CRKEY: Blender Chroma Key Enable

0: Chroma key matching is disabled. 1: Chroma key matching is enabled.
INV: Blender Inverted Blender Output Enable

0: Iterated pixel is the blended pixel. 1: Iterated pixel is the inverted pixel.
ITER2BL: Blender Iterated Color Enable

0: Final adder stage operand is set to 0. 1: Final adder stage operand is set to the iterated pixel value.
ITER: Blender Use Iterated Color

0: Pixel difference is set to 0. 1: Pixel difference is set to the iterated pixel value.
REVALPHA: Blender Reverse Alpha

0: Pixel difference is multiplied by alpha. 1: Pixel difference is multiplied by 1 - alpha.
GAEN: Blender Global Alpha Enable

0: Global alpha blending coefficient is disabled. 1: Global alpha blending coefficient is enabled.
LAEN: Blender Local Alpha Enable

0: Local alpha blending coefficient is disabled. 1: Local alpha blending coefficient is enabled.
OVR: Blender Overlay Layer Enable

0: Overlay pixel color is set to the default overlay pixel color. 1: Overlay pixel color is set to the DMA channel pixel color.

610
DMA: Blender DMA Layer Enable

0: The default color is used on the Overlay 1 Layer. 1: The DMA channel retrieves the pixels stream from the memory.

DSTKEY: Destination Chroma Keying

0: Source Chroma keying is enabled. 1: Destination Chroma keying is used.
GA: Blender Global Alpha

Global alpha blender for the current layer.

611
31.7.54 Overlay 2 Layer Channel Enable Register

31 23 15 7
LCDC_OVR2CHER 0xF0030240 Write-only 0x00000000





612
31.7.55 Overlay 2 Layer Channel Disable Register

31 23 15 7
LCDC_OVR2CHDR 0xF0030244 Write-only 0x00000000

30 22 14 6 29 21 13 5 28 20 12 4 27 19 11 3 26 18 10 2 25 17 9 1 24 16 8 CHRST 0 CHDIS



613
31.7.56 Overlay 2 Layer Channel Status Register

31 23 15 7
LCDC_OVR2CHSR 0xF0030248 Read-only 0x00000000



When set to one this bit indicates that the overlay attributes will update on the next Frame.


614
31.7.57 Overlay 2 Layer Interrupt Enable Register

31 23 15 7
LCDC_OVR2IER 0xF003024C Write-only 0x00000000







615
31.7.58 Overlay 2 Layer Interrupt Disable Register

31 23 15 7
LCDC_OVR2IDR 0xF0030250 Write-only 0x00000000







616
31.7.59 Overlay 2 Layer Interrupt Mask Register

31 23 15 7
LCDC_OVR2IMR 0xF0030254 Read-only 0x00000000







617
31.7.60 Overlay 2 Layer Interrupt Status Register

31 23 15 7
LCDC_OVR2ISR 0xF0030258 Read-only 0x00000000




DONE: End Of List Interrupt Disable Register

When set to one this flag indicates that a End of List condition has occurred. This flag is reset after a read operation.


618
31.7.61 Overlay 2 Layer Head Register

31 23 15 7
LCDC_OVR2HEAD 0xF003025C Read-Write 0x00000000



619
31.7.62 Overlay 2 Layer Address Register

31 23 15 7
LCDC_OVR2ADDR 0xF0030260 Read-Write 0x00000000

ADDR: DMA Transfer Overlay 2 Address

Overlay 2 frame buffer base address.

620
31.7.63 Overlay 2 Layer Control Register

31 23 15 7
LCDC_OVR2CTRL 0xF0030264 Read-Write 0x00000000








621
31.7.64 Overlay 2 Layer Next Register

31 23 15 7
LCDC_OVR2NEXT 0xF0030268 Read-Write 0x00000000



622

31 23 15 7
LCDC_OVR2CFG0 0xF003026C Read-Write 0x00000000

30 22 14 6 29 21 13 LOCKDIS 5 BLEN 28 20 12 ROTDIS 4 27 19 11 3 26 18 10 2 25 17 9 1 24 16 8 DLBO 0

AHB_INCR4
AHB_INCR8
AHB_INCR16

0: Undefined length INCR burst is used for 2 and 3 beats burst. 1: Only Defined Length burst is used (SINGLE, INCR4, INCR8 and INCR16).



623

31 23 15 7
LCDC_OVR2CFG1 0xF0030270 Read-Write 0x00000000


0: RGB mode is selected 1: Color lookup table is selected



624

31 23 15 7


Overlay 2 Horizontal window position.

Overlay 2 Vertical window position.

625

31 23 15 7


Overlay 2 window width in pixels. The window width is set to (XSIZE+1). The following constraint must be met:
XPOS + XSIZE PPL

Overlay 2 window height in pixels. The window height is set to (YSIZE+1). The following constrain must be met:
YPOS + YSIZE RPF

626

31 23 15 7



627

31 23 15 7


PSTRIDE represents the memory offset, in bytes, between two pixels of the image memory.

628

31 23 15 7

30 22 14 6 29 21 13 5 28 20 RDEF 12 GDEF 4 BDEF 3 2 1 0 11 10 9 8 27 19 26 18 25 17 24 16
RDEF: Red Default

Default Red color when the Overlay 1 DMA channel is disabled.
GDEF: Green Default

Default Green color when the Overlay 1 DMA channel is disabled.
BDEF: Blue Default

Default Blue color when the Overlay 1 DMA channel is disabled.

629

31 23 15 7

30 22 14 6 29 21 13 5 28 20 RKEY 12 GKEY 4 BKEY 3 2 1 0 11 10 9 8 27 19 26 18 25 17 24 16




630

31 23 15 7





631

31 23 15 7 OVR










632





633
31.7.75 High End Overlay Layer Channel Enable Register

31 23 15 7
LCDC_HEOCHER 0xF0030340 Write-only 0x00000000





634
31.7.76 High End Overlay Layer Channel Disable Register

31 23 15 7
LCDC_HEOCHDR 0xF0030344 Write-only 0x00000000

30 22 14 6 29 21 13 5 28 20 12 4 27 19 11 3 26 18 10 2 25 17 9 1 24 16 8 CHRST 0 CHDIS



635
31.7.77 High End Overlay Layer Channel Status Register

31 23 15 7
LCDC_HEOCHSR 0xF0030348 Read-only 0x00000000





636
31.7.78 High End Overlay Layer Interrupt Enable Register

31 23 15 7
LCDC_HEOIER 0xF003034C Write-only 0x00000000

30 22 VOVR 14 UOVR 6 OVR 29 21 VDONE 13 UDONE 5 DONE 28 20 VADD 12 UADD 4 ADD 27 19 VDSCR 11 UDSCR 3 DSCR 26 18 VDMA 10 UDMA 2 DMA 25 17 9 1 24 16 8 0





UDMA: End of DMA Transfer for U or UV Chrominance Interrupt Enable Register

UDSCR: Descriptor Loaded for U or UV Chrominance Interrupt Enable Register

UADD: Head Descriptor Loaded for U or UV Chrominance Interrupt Enable Register


637
UDONE: End of List for U or UV Chrominance Interrupt Enable Register

UOVR: Overflow for U or UV Chrominance Interrupt Enable Register

VDMA: End of DMA for V Chrominance Transfer Interrupt Enable Register

VDSCR: Descriptor Loaded for V Chrominance Interrupt Enable Register

VADD: Head Descriptor Loaded for V Chrominance Interrupt Enable Register

VDONE: End of List for V Chrominance Interrupt Enable Register

VOVR: Overflow for V Chrominance Interrupt Enable Register


638
31.7.79 High End Overlay Layer Interrupt Disable Register

31 23 15 7
LCDC_HEOIDR 0xF0030350 Write-only 0x00000000






UDMA: End of DMA Transfer for U or UV Chrominance Component Interrupt Disable Register
UDSCR: Descriptor Loaded for U or UV Chrominance Component Interrupt Disable Register

UADD: Head Descriptor Loaded for U or UV Chrominance Component Interrupt Disable Register

639
UDONE: End of List Interrupt for U or UV Chrominance Component Disable Register

UOVR: Overflow Interrupt for U or UV Chrominance Component Disable Register

VDMA: End of DMA Transfer for V Chrominance Component Interrupt Disable Register
VDSCR: Descriptor Loaded for V Chrominance Component Interrupt Disable Register

VADD: Head Descriptor Loaded for V Chrominance Component Interrupt Disable Register
VDONE: End of List for V Chrominance Component Interrupt Disable Register

VOVR: Overflow for V Chrominance Component Interrupt Disable Register


640
31.7.80 High End Overlay Layer Interrupt Mask Register

31 23 15 7
LCDC_HEOIMR 0xF0030354 Read-only 0x00000000






UDMA: End of DMA Transfer for U or UV Chrominance Component Interrupt Mask Register
UDSCR: Descriptor Loaded for U or UV Chrominance Component Interrupt Mask Register

UADD: Head Descriptor Loaded for U or UV Chrominance Component Mask Register


641
UDONE: End of List for U or UV Chrominance Component Mask Register

UOVR: Overflow for U Chrominance Interrupt Mask Register

VDMA: End of DMA Transfer for V Chrominance Component Interrupt Mask Register
VDSCR: Descriptor Loaded for V Chrominance Component Interrupt Mask Register

VADD: Head Descriptor Loaded for V Chrominance Component Mask Register

VDONE: End of List for V Chrominance Component Mask Register

VOVR: Overflow for V Chrominance Interrupt Mask Register


642
31.7.81 High End Overlay Layer Interrupt Status Register

31 23 15 7
LCDC_HEOISR 0xF0030358 Read-only 0x00000000






UDMA: End of DMA Transfer for U component

UDSCR: DMA Descriptor Loaded for U component

UADD: Head Descriptor Loaded for U component

UDONE: End of List Detected for U component

UOVR: Overflow Detected for U component

VDMA: End of DMA Transfer for V component


643
VDSCR: DMA Descriptor Loaded for V component

VADD: Head Descriptor Loaded for V component

VDONE: End of List Detected for V component

VOVR: Overflow Detected for V component


644
31.7.82 High End Overlay Layer Head Register

31 23 15 7
LCDC_HEOHEAD 0xF003035C Read-write 0x00000000



645
31.7.83 High End Overlay Layer Address Register

31 23 15 7
LCDC_HEOADDR 0xF0030360 Read-write 0x00000000

ADDR: DMA Transfer start Address

Frame Buffer Base Address.

646
31.7.84 High End Overlay Layer Control Register

31 23 15 7
LCDC_HEOCTRL 0xF0030364 Read-write 0x00000000








647
31.7.85 High End Overlay Layer Next Register

31 23 15 7
LCDC_HEONEXT 0xF0030368 Read-write 0x00000000



648
31.7.86 High End Overlay Layer U-UV Head Register

31 23 15 7
LCDC_HEOUHEAD 0xF003036C Read-write 0x00000000

30 22 14 6 29 21 13 5 28 UHEAD 20 UHEAD 12 UHEAD 4 UHEAD 3 2 1 0 11 10 9 8 19 18 17 16 27 26 25 24
UHEAD: DMA Head Pointer


649
31.7.87 High End Overlay Layer U-UV Address Register

31 23 15 7
LCDC_HEOUADDR 0xF0030370 Read-write 0x00000000

30 22 14 6 29 21 13 5 28 UADDR 20 UADDR 12 UADDR 4 UADDR 3 2 1 0 11 10 9 8 19 18 17 16 27 26 25 24
UADDR: DMA Transfer Start Address for U or UV Chrominance

U or UV frame buffer address.

650
31.7.88 High End Overlay Layer U-UV Control Register

31 23 15 7
LCDC_HEOUCTRL 0xF0030374 Read-write 0x00000000

30 22 14 6 29 21 13 5 UDONEIEN 28 20 12 4 UADDIEN 27 19 11 3 UDSCRIEN 26 18 10 2 UDMAIEN 25 17 9 1 24 16 8 0 UDFETCH
UDFETCH: Transfer Descriptor Fetch Enable

UDMAIEN: End of DMA Transfer Interrupt Enable

UDSCRIEN: Descriptor Loaded Interrupt Enable

UADDIEN: Add Head Descriptor to Queue Interrupt Enable

UDONEIEN: End of List Interrupt Enable


651
31.7.89 High End Overlay Layer U-UV Next Register

31 23 15 7
LCDC_HEOUNEXT 0xF0030378 Read-write 0x00000000

30 22 14 6 29 21 13 5 28 UNEXT 20 UNEXT 12 UNEXT 4 UNEXT 3 2 1 0 11 10 9 8 19 18 17 16 27 26 25 24
UNEXT: DMA Descriptor Next Address


652
31.7.90 High End Overlay Layer V Head Register

31 23 15 7
LCDC_HEOVHEAD 0xF003037C Read-write 0x00000000

30 22 14 6 29 21 13 5 28 VHEAD 20 VHEAD 12 VHEAD 4 VHEAD 3 2 1 0 11 10 9 8 19 18 17 16 27 26 25 24
VHEAD: DMA Head Pointer


653
31.7.91 High End Overlay Layer V Address Register

31 23 15 7
LCDC_HEOVADDR 0xF0030380 Read-write 0x00000000

30 22 14 6 29 21 13 5 28 VADDR 20 VADDR 12 VADDR 4 VADDR 3 2 1 0 11 10 9 8 19 18 17 16 27 26 25 24
VADDR: DMA Transfer Start Address for V Chrominance

Frame Buffer Base Address.

654
31.7.92 High End Overlay Layer V Control Register

31 23 15 7
LCDC_HEOVCTRL 0xF0030384 Read-write 0x00000000

30 22 14 6 29 21 13 5 VDONEIEN 28 20 12 4 VADDIEN 27 19 11 3 VDSCRIEN 26 18 10 2 VDMAIEN 25 17 9 1 24 16 8 0 VDFETCH
VDFETCH: Transfer Descriptor Fetch Enable

VDMAIEN: End of DMA Transfer Interrupt Enable

VDSCRIEN: Descriptor Loaded Interrupt Enable

VADDIEN: Add Head Descriptor to Queue Interrupt Enable

VDONEIEN: End of List Interrupt Enable


655
31.7.93 High End Overlay Layer V Next Register

31 23 15 7
LCDC_HEOVNEXT 0xF0030388 Read-write 0x00000000

30 22 14 6 29 21 13 5 28 VNEXT 20 VNEXT 12 VNEXT 4 VNEXT 3 2 1 0 11 10 9 8 19 18 17 16 27 26 25 24
VNEXT: DMA Descriptor Next Address


656
31.7.94 High End Overlay Layer Configuration 0 Register

31 23 15 7 BLENUV
LCDC_HEOCFG0 0xF003038C Read-write 0x00000000

30 22 14 6 29 21 13 LOCKDIS 5 BLEN 28 20 12 ROTDIS 4 27 19 11 3 26 18 10 2 25 17 9 1 24 16 8 DLBO 0 SIF


AHB_BLEN_INCR4
AHB_BLEN_INCR8
BLENUV: AHB Burst Length for U-V channel

AHB_INCR4
AHB_INCR8
AHB_INCR16

657




658

31 23 15 7
LCDC_HEOCFG1 0xF0030390 Read-write 0x00000000

30 22 14 YUVMODE 6 RGBMODE 5 4 29 21 13 28 20 DSCALEOPT 12 27 19 11 3 26 18 10 2 25 24 17 16 YUV422SWP YUV422ROT 9 8 CLUTMODE 1 0 YUVEN CLUTEN

0: RGB mode is selected. 1: Color Lookup table is selected.
YUVEN: YUV Color Space Enable

0: Color space is RGB 1: Color Space is YUV
RGBMODE: RGB input mode selection



659
YUVMODE: YUV input mode selection

Value 0 1 2 3 4 5 6 7 8 Name 32BPP_AYCBCR 16BPP_YCBCR_MODE0 16BPP_YCBCR_MODE1 16BPP_YCBCR_MODE2 16BPP_YCBCR_MODE3 16BPP_YCBCR_SEMIPLANAR 16BPP_YCBCR_PLANAR 12BPP_YCBCR_SEMIPLANAR 12BPP_YCBCR_PLANAR Description 32 bpp AYCbCr 444 16 bpp Cr(n)Y(n+1)Cb(n)Y(n) 422 16 bpp Y(n+1)Cr(n)Y(n)Cb(n) 422 16 bpp Cb(n)Y(+1)Cr(n)Y(n) 422 16 bpp Y(n+1)Cb(n)Y(n)Cr(n) 422 16 bpp Semiplanar 422 YCbCr 16 bpp Planar 422 YCbCr 12 bpp Semiplanar 420 YCbCr 12 bpp Planar 420 YCbCr
YUV422ROT: YUV 4:2:2 Rotation

When set to one this bit indicates that the Chroma Upsampling kernel is configured to use the 4:2:2 Rotation Algorithm.
YUV422SWP: YUV 4:2:2 SWAP

When set to one the Y component of the YUV 4:2:2 packed data stream are swapped.
DSCALEOPT: Down Scaling Bandwidth Optimization

0: Scaler Optimization is disabled. 1: Scaler Optimization is enabled, only relevant pixels are retrieved from memory to fill the scaler filter.

660

31 23 15 7


High End Overlay Horizontal window position.

High End Overlay Vertical window position.

661

31 23 15 7


High End Overlay window width in pixels. The window width is set to (XSIZE+1). The following constraint must be met:
XPOS + XSIZE PPL

High End Overlay window height in pixels. The window height is set to (YSIZE+1). The following constrain must be met:
YPOS + YSIZE RPF

662

31 23 15 7

30 22 14 6 29 21 13 5 28 20 YMEMSIZE 12 4 XMEMSIZE 11 3 10 2 9 XMEMSIZE 1 8 0 27 19 26 18 25 YMEMSIZE 17 24 16
XMEMSIZE: Horizontal image Size in Memory

High End Overlay image width in pixels. The image width is set to (XMEMSIZE+1).
YMEMSIZE: Vertical image Size in Memory

High End Overlay image height in pixels. The image height is set to (YMEMSIZE+1).

663

31 23 15 7
LCDC_HEOCFG5 0xF00303A0 Read-write 0x00000000



664

31 23 15 7


PSTRIDE represents the memory offset, in bytes, between two pixels of the image memory.

665

31 23 15 7

30 22 14 6 29 21 13 5 28 UVXSTRIDE 20 UVXSTRIDE 12 UVXSTRIDE 4 UVXSTRIDE 3 2 1 0 11 10 9 8 19 18 17 16 27 26 25 24
UVXSTRIDE: UV Horizontal Stride

UVXSTRIDE represents the memory offset, in bytes, between two rows of the image memory.

666

31 23 15 7
LCDC_HEOCFG8 0xF00303AC Read-write 0x00000000

30 22 14 6 29 21 13 5 28 UVPSTRIDE 20 UVPSTRIDE 12 UVPSTRIDE 4 UVPSTRIDE 3 2 1 0 11 10 9 8 19 18 17 16 27 26 25 24
UVPSTRIDE: UV Pixel Stride

UVPSTRIDE represents the memory offset, in bytes, between two pixels of the image memory.

667

31 23 15 7
LCDC_HEOCFG9 0xF00303B0 Read-write 0x00000000

30 22 14 6 29 21 13 5 28 20 RDEF 12 GDEF 4 BDEF 3 2 1 0 11 10 9 8 27 19 26 18 25 17 24 16
RDEF: Red Default

Default Red color when the High End Overlay DMA channel is disabled.
GDEF: Green Default

Default Green color when the High End Overlay DMA channel is disabled.
BDEF: Blue Default

Default Blue color when the High End Overlay DMA channel is disabled.

668

31 23 15 7

30 22 14 6 29 21 13 5 28 20 RKEY 12 GKEY 4 BKEY 3 2 1 0 11 10 9 8 27 19 26 18 25 17 24 16




669

31 23 15 7





670

31 23 15 7 OVR
LCDC_HEOCFG12 0xF00303BC Read-write 0x00000000

30 22 14 6 LAEN 29 21 13 5 GAEN 28 20 GA 12 VIDPRI 4 REVALPHA 11 3 ITER 10 DSTKEY 2 ITER2BL 9 REP 1 INV 8 DMA 0 CRKEY 27 19 26 18 25 17 24 16









671



VIDPRI: Video Priority

0: OVR1 layer is above HEO layer. 1: OVR1 layer is below HEO layer.


672

31 SCALEN 23 15 7
LCDC_HEOCFG13 0xF00303C0 Read-write 0x00000000

30 22 14 6 29 21 13 5 28 20 YFACTOR 12 4 XFACTOR 11 XFACTOR 3 2 1 0 10 9 8 27 YFACTOR 19 18 17 16 26 25 24
SCALEN: Hardware Scaler Enable

0: Scaler is disabled 1: Scaler is enabled.
YFACTOR: Vertical Scaling Factor

Scaler Vertical Factor.
XFACTOR: Horizontal Scaling Factor

Scaler Horizontal Factor.

673

31 23 15 7

30 CSCYOFF 22 CSCRV 14 6 29 21 13 CSCRU 5 4 CSCRY 3 2 1 28 20 12 27 CSCRV 19 11 18 CSCRU 10 9 CSCRY 0 8 17 16 26 25 24
CSCRY: Color Space Conversion Y coefficient for Red Component 1:2:7 format
Color Space Conversion coefficient format is 1 sign bit, 2 magnitude bits and 7 fractional bits.
CSCRU: Color Space Conversion U coefficient for Red Component 1:2:7 format
CSCRV: Color Space Conversion V coefficient for Red Component 1:2:7 format
CSCYOFF: Color Space Conversion Offset

0: Offset is set to 0 1: Offset is set to 16

674

31 23 15 7

30 CSCUOFF 22 CSCGV 14 6 29 21 13 CSCGU 5 4 CSCGY 3 2 1 28 20 12 27 CSCGV 19 11 18 CSCGU 10 9 CSCGY 0 8 17 16 26 25 24
CSCGY: Color Space Conversion Y coefficient for Green Component 1:2:7 format
CSCGU: Color Space Conversion U coefficient for Green Component 1:2:7 format
CSCGV: Color Space Conversion V coefficient for Green Component 1:2:7 format
CSCUOFF: Color Space Conversion Offset


675

31 23 15 7
LCDC_HEOCFG16 0xF00303CC Read-write 0x00000000

30 CSCVOFF 22 CSCBV 14 6 29 21 13 CSCBU 5 4 CSCBY 3 2 1 28 20 12 27 CSCBV 19 11 18 CSCBU 10 9 CSCBY 0 8 17 16 26 25 24
CSCBY: Color Space Conversion Y coefficient for Blue Component 1:2:7 format
CSCBU: Color Space Conversion U coefficient for Blue Component 1:2:7 format
CSCBV: Color Space Conversion V coefficient for Blue Component 1:2:7 format
CSCVOFF: Color Space Conversion Offset


676

31 23 15 7
LCDC_HEOCFG17 0xF00303D0 Read-write 0x00000000

30 22 14 6 29 21 13 5 28 27 XPHI0COEFF3 20 19 XPHI0COEFF2 12 11 XPHI0COEFF1 4 3 XPHI0COEFF0 26 18 10 2 25 17 9 1 24 16 8 0
XPHI0COEFF0: Horizontal Coefficient for phase 0 tap 0

Coefficient format is 1 sign bit and 7 fractional bits.


Coefficient format is 1 magnitude bit and 7 fractional bits.


677

31 23 15 7

30 22 14 6 29 21 13 5 28 27 20 19 12 11 4 3 XPHI0COEFF4 26 18 10 2 25 17 9 1 24 16 8 0


678

31 23 15 7






679

31 23 15 7
LCDC_HEOCFG20 0xF00303DC Read-write 0x00000000

30 22 14 6 29 21 13 5 28 27 20 19 12 11 4 3 XPHI1COEFF4 26 18 10 2 25 17 9 1 24 16 8 0


680

31 23 15 7
LCDC_HEOCFG21 0xF00303E0 Read-write 0x00000000






681

31 23 15 7

30 22 14 6 29 21 13 5 28 27 20 19 12 11 4 3 XPHI2COEFF4 26 18 10 2 25 17 9 1 24 16 8 0


682

31 23 15 7






683

31 23 15 7
LCDC_HEOCFG24 0xF00303EC Read-write 0x00000000

30 22 14 6 29 21 13 5 28 27 20 19 12 11 4 3 XPHI3COEFF4 26 18 10 2 25 17 9 1 24 16 8 0


684

31 23 15 7
LCDC_HEOCFG25 0xF00303F0 Read-write 0x00000000






685

31 23 15 7

30 22 14 6 29 21 13 5 28 27 20 19 12 11 4 3 XPHI4COEFF4 26 18 10 2 25 17 9 1 24 16 8 0


686

31 23 15 7






687

31 23 15 7
LCDC_HEOCFG28 0xF00303FC Read-write 0x00000000

30 22 14 6 29 21 13 5 28 27 20 19 12 11 4 3 XPHI5COEFF4 26 18 10 2 25 17 9 1 24 16 8 0


688

31 23 15 7






689

31 23 15 7

30 22 14 6 29 21 13 5 28 27 20 19 12 11 4 3 XPHI6COEFF4 26 18 10 2 25 17 9 1 24 16 8 0


690

31 23 15 7






691

31 23 15 7

30 22 14 6 29 21 13 5 28 27 20 19 12 11 4 3 XPHI7COEFF4 26 18 10 2 25 17 9 1 24 16 8 0


692

31 23 15 7

30 22 14 6 29 21 13 5 28 27 20 19 YPHI0COEFF2 12 11 YPHI0COEFF1 4 3 YPHI0COEFF0 26 18 10 2 25 17 9 1 24 16 8 0
YPHI0COEFF0: Vertical Coefficient for phase 0 tap 0




693

31 23 15 7





694

31 23 15 7





695

31 23 15 7





696

31 23 15 7





697

31 23 15 7





698

31 23 15 7





699

31 23 15 7





700

31 23 15 7

30 22 14 6 29 21 13 5 28 20 12 4 27 19 11 3 26 18 10 2 25 17 YPHIDEF 9 1 XPHIDEF 24 16 8 0
XPHIDEF: Horizontal Filter Phase Offset

XPHIDEF defines the index of the first coefficient set used when the horizontal resampling operation is started.
YPHIDEF: Vertical Filter Phase Offset

XPHIDEF defines the index of the first coefficient set used when the vertical resampling operation is started.

701
31.7.136 Hardware Cursor Layer Channel Enable Register

31 23 15 7
LCDC_HCRCHER 0xF0030440 Write-only 0x00000000





702
31.7.137 Hardware Cursor Layer Channel Disable Register

31 23 15 7
LCDC_HCRCHDR 0xF0030444 Write-only 0x00000000

30 22 14 6 29 21 13 5 28 20 12 4 27 19 11 3 26 18 10 2 25 17 9 1 24 16 8 CHRST 0 CHDIS



703
31.7.138 Hardware Cursor Layer Channel Status Register

31 23 15 7
LCDC_HCRCHSR 0xF0030448 Read-only 0x00000000





704
31.7.139 Hardware Cursor Layer Interrupt Enable Register

31 23 15 7
LCDC_HCRIER 0xF003044C Write-only 0x00000000







705
31.7.140 Hardware Cursor Layer Interrupt Disable Register

31 23 15 7
LCDC_HCRIDR 0xF0030450 Write-only 0x00000000







706
31.7.141 Hardware Cursor Layer Interrupt Mask Register

31 23 15 7
LCDC_HCRIMR 0xF0030454 Read-only 0x00000000







707
31.7.142 Hardware Cursor Layer Interrupt Status Register

31 23 15 7
LCDC_HCRISR 0xF0030458 Read-only 0x00000000






When set to one this flag indicates that an Overflow has occurred. This flag is reset after a read operation.

708
31.7.143 Hardware Cursor Layer Head Register

31 23 15 7
LCDC_HCRHEAD 0xF003045C Read-write 0x00000000



709
31.7.144 Hardware Cursor Layer Address Register

31 23 15 7
LCDC_HCRADDR 0xF0030460 Read-write 0x00000000

ADDR: DMA Transfer start address

Frame buffer start address.

710
31.7.145 Hardware Cursor Layer Control Register

31 23 15 7
LCDC_HCRCTRL 0xF0030464 Read-write 0x00000000








711
31.7.146 Hardware Cursor Layer Next Register

31 23 15 7
LCDC_HCRNEXT 0xF0030468 Read-write 0x00000000



712
31.7.147 Hardware Cursor Layer Configuration 0 Register

31 23 15 7
LCDC_HCRCFG0 0xF003046C Read-write 0x00000000

30 22 14 6 29 21 13 5 BLEN 28 20 12 4 27 19 11 3 26 18 10 2 25 17 9 1 24 16 8 DLBO 0 SIF


AHB_BLEN_INCR4
AHB_BLEN_INCR8
DLBO: Defined Length Burst Only for Channel Bus Transaction.


713

31 23 15 7
LCDC_HCRCFG1 0xF0030470 Read-write 0x00000000


0: RGB mode is selected. 1: Color Lookup table is selected.
RGBMODE: RGB input mode selection



714

31 23 15 7


Hardware Cursor Horizontal window position.

Hardware Cursor Vertical window position.

715

31 23 15 7


Hardware cursor width is limited to 128 pixels Hardware Cursor window width in pixels. The window width is set to (XSIZE+1). The following constraint must be met:
XPOS + XSIZE PPL

Hardware cursor height is limited to 128 pixels Hardware Cursor window height in pixels. The window height is set to (YSIZE+1). The following constrain must be met:
YPOS + YSIZE RPF

716

31 23 15 7



717

31 23 15 7

30 22 14 6 29 21 13 5 28 20 RDEF 12 GDEF 4 BDEF 3 2 1 0 11 10 9 8 27 19 26 18 25 17 24 16
RDEF: Red Default

Default Red color when the Hardware Cursor DMA channel is disabled.
GDEF: Green Default

Default Green color when the Hardware Cursor DMA channel is disabled.
BDEF: Blue Default Default Blue color when the Hardware Cursor DMA channel is disabled.

718

31 23 15 7

30 22 14 6 29 21 13 5 28 20 RKEY 12 GKEY 4 BKEY 3 2 1 0 11 10 9 8 27 19 26 18 25 17 24 16




719

31 23 15 7





720

31 23 15 7 OVR










721





722
31.7.156 Post Processing Layer Channel Enable Register

31 23 15 7
LCDC_PPCHER 0xF0030540 Write-only 0x00000000





723
31.7.157 Post Processing Layer Channel Disable Register

31 23 15 7
LCDC_PPCHDR 0xF0030544 Write-only 0x00000000

30 22 14 6 29 21 13 5 28 20 12 4 27 19 11 3 26 18 10 2 25 17 9 1 24 16 8 CHRST 0 CHDIS



724
31.7.158 Post Processing Layer Channel Status Register

31 23 15 7
LCDC_PPCHSR 0xF0030548 Read-only 0x00000000



When set to one this bit indicates that the overlay attributes will update on the next Frame.


725
31.7.159 Post Processing Layer Interrupt Enable Register

31 23 15 7
LCDC_PPIER 0xF003054C Write-only 0x00000000

30 22 14 6 29 21 13 5 DONE 28 20 12 4 ADD 27 19 11 3 DSCR 26 18 10 2 DMA 25 17 9 1 24 16 8 0





726
31.7.160 Post Processing Layer Interrupt Disable Register

31 23 15 7
LCDC_PPIDR 0xF0030550 Write-only 0x00000000

30 22 14 6 29 21 13 5 DONE 28 20 12 4 ADD 27 19 11 3 DSCR 26 18 10 2 DMA 25 17 9 1 24 16 8 0





727
31.7.161 Post Processing Layer Interrupt Mask Register

31 23 15 7
LCDC_PPIMR 0xF0030554 Read-only 0x00000000

30 22 14 6 29 21 13 5 DONE 28 20 12 4 ADD 27 19 11 3 DSCR 26 18 10 2 DMA 25 17 9 1 24 16 8 0





728
31.7.162 Post Processing Layer Interrupt Status Register

31 23 15 7
LCDC_PPISR 0xF0030558 Read-only 0x00000000

30 22 14 6 29 21 13 5 DONE 28 20 12 4 ADD 27 19 11 3 DSCR 26 18 10 2 DMA 25 17 9 1 24 16 8 0



DONE: End Of List Detected

When set to one this flag indicates that a End of List condition has occurred. This flag is reset after a read operation.

729
31.7.163 Post Processing Layer Head Register

31 23 15 7
LCDC_PPHEAD 0xF003055C Read-Write 0x00000000



730
31.7.164 Post Processing Layer Address Register

31 23 15 7
LCDC_PPADDR 0xF0030560 Read-Write 0x00000000

ADDR: DMA Transfer start address

Post Processing Destination frame buffer address.

731
31.7.165 Post Processing Layer Control Register

31 23 15 7
LCDC_PPCTRL 0xF0030564 Read-Write 0x00000000

30 22 14 6 29 21 13 5 DONEIEN 28 20 12 4 ADDIEN 27 19 11 3 DSCRIEN 26 18 10 2 DMAIEN 25 17 9 1 24 16 8 0 DFETCH






732
31.7.166 Post Processing Layer Next Register

31 23 15 7
LCDC_PPNEXT 0xF0030568 Read-Write 0x00000000



733
31.7.167 Post Processing Layer Configuration 0 Register

31 23 15 7
LCDC_PPCFG0 0xF003056C Read-Write 0x00000000

30 22 14 6 29 21 13 5 BLEN 28 20 12 4 27 19 11 3 26 18 10 2 25 17 9 1 24 16 8 DLBO 0 SIF


AHB_BLEN_INCR4
AHB_BLEN_INCR8

0: Undefined length INCR burst is used for 2 and 3 beats burst. 1: Only Defined Length burst is used (SINGLE, INCR4, INCR8 and INCR16).

734

31 23 15 7
LCDC_PPCFG1 0xF0030570 Read-Write 0x00000000

30 22 14 6 29 21 13 5 28 20 12 4 ITUBT601 27 19 11 3 26 18 10 2 25 17 9 1 PPMODE 24 16 8 0
PPMODE: Post Processing Output Format selection

Table 31-55. PPMODE selection Value 0 1 2 3 4 5 6 Name PPMODE_RGB_16BPP PPMODE_RGB_24BPP_PACKED PPMODE_RGB_24BPP_UNPACKED PPMODE_YCBCR_422_MODE0 PPMODE_YCBCR_422_MODE1 PPMODE_YCBCR_422_MODE2 PPMODE_YCBCR_422_MODE3 Description RGB 16 bpp RGB 24 bpp PACKED RGB 24 bpp UNPACKED YCbCr 422 16 bpp (Mode 0) YCbCr 422 16 bpp (Mode 1) YCbCr 422 16 bpp (Mode 2) YCbCr 422 16 bpp (Mode 3)
ITUBT601: Color Space Conversion Luminance

0: Luminance and chrominance range is [0;255] 1: Luminance values are clamped to [16;235] range. Chrominance values are clamped to [16;240] range.

735

31 23 15 7



736

31 23 15 7

30 CSCYOFF 22 CSCYB 14 6 29 21 13 CSCYG 5 4 CSCYR 3 2 1 28 20 12 27 CSCYB 19 11 18 CSCYG 10 9 CSCYR 0 8 17 16 26 25 24
CSCYR: Color Space Conversion R coefficient for Luminance component, signed format, step set to 1/1024
Color Space Conversion coefficient format is 1 sign bit, 9 fractional bits.
CSCYG: Color Space Conversion G coefficient for Luminance component, signed format, step set to 1/512
CSCYB: Color Space Conversion B coefficient for Luminance component, signed format, step set to 1/1024
CSCYOFF: Color Space Conversion Luminance Offset

0: The Yoff parameter value is set to 0. 1: The Yoff parameter value is set to 16.

737

31 23 15 7
LCDC_PPCFG4 0xF003057C Read-Write 0x00000000

30 CSCUOFF 22 CSCUB 14 6 29 21 13 CSCUG 5 4 CSCUR 3 2 1 28 20 12 27 CSCUB 19 11 18 CSCUG 10 9 CSCUR 0 8 17 16 26 25 24
CSCUR: Color Space Conversion R coefficient for Chrominance B component, signed format. (step 1/1024)
CSCUG: Color Space Conversion G coefficient for Chrominance B component, signed format. (step 1/512)
CSCUB: Color Space Conversion B coefficient for Chrominance B component, signed format. (step 1/512)
CSCUOFF: Color Space Conversion Chrominance B Offset

0: The Cboff parameter value is set to 0. 1: The Cboff parameter value is set to 128.

738

31 23 15 7

30 CSCVOFF 22 CSCVB 14 6 29 21 13 CSCVG 5 4 CSCVR 3 2 1 28 20 12 27 CSCVB 19 11 18 CSCVG 10 9 CSCVR 0 8 17 16 26 25 24
CSCVR: Color Space Conversion R coefficient for Chrominance R component, signed format. (step 1/1024)
CSCVG: Color Space Conversion G coefficient for Chrominance R component, signed format. (step 1/512)
CSCVB: Color Space Conversion B coefficient for Chrominance R component, signed format. (step 1/1024)
CSCVOFF: Color Space Conversion Chrominance R Offset

0: The Croff parameter value is set to 0. 1: The Croff parameter value is set to 128.

739
31.7.173 Base CLUT Register x Register

31 23 15 7
LCDC_BASECLUTx [x=0..255] 0xF0030600 Read-write 0x00000000

30 22 14 6 29 21 13 5 28 20 RCLUT 12 GCLUT 4 BCLUT 3 2 1 0 11 10 9 8 27 19 26 18 25 17 24 16
BCLUT: Blue Color entry

This field indicates the 8-bit width Blue color of the color lookup table.
GCLUT: Green Color entry

This field indicates the 8-bit width Green color of the color lookup table.
RCLUT: Red Color entry

This field indicates the 8-bit width Red color of the color lookup table.

740
31.7.174 Overlay 1 CLUT Register x Register

31 23 15 7
LCDC_OVR1CLUTx [x=0..255] 0xF0030A00 Read-write 0x00000000

30 22 14 6 29 21 13 5 28 ACLUT 20 RCLUT 12 GCLUT 4 BCLUT 3 2 1 0 11 10 9 8 19 18 17 16 27 26 25 24



ACLUT: Alpha Color entry

This field indicates the 8-bit width Alpha channel of the color lookup table.

741
31.7.175 Overlay 2 CLUT Register x Register

31 23 15 7
LCDC_OVR2CLUTx [x=0..255] 0xF0030E00 Read-write 0x00000000






742
31.7.176 High End Overlay CLUT Register x Register

31 23 15 7
LCDC_HEOCLUTx [x=0..255] 0xF0031200 Read-write 0x00000000






743
31.7.177 Hardware Cursor CLUT Register x Register

31 23 15 7
LCDC_HCRCLUTx [x=0..255] 0xF0031600 Read-write 0x00000000






744
32.
32.1
USB High Speed Device Port (UDPHS)

Description
The USB High Speed Device Port (UDPHS) is compliant with the Universal Serial Bus (USB), rev 2.0 High Speed device specification. Each endpoint can be configured in one of several USB transfer types. It can be associated with one, two or three banks of a Dual-port RAM used to store the current data payload. If two or three banks are used, one DPR bank is read or written by the processor, while the other is read or written by the USB device peripheral. This feature is mandatory for isochronous endpoints.
32.2
z z z z z z z z
1 Device High Speed 1 UTMI transceiver shared between Host and Device USB v2.0 High Speed Compliant, 480 Mbits Per Second 16 Endpoints up to 1024 bytes Embedded Dual-port RAM for Endpoints Suspend/Resume Logic (Command of UTMI) Up to Three Memory Banks for Endpoints (Not for Control Endpoint) 8 Kbytes of DPRAM

745
32.3
Block Diagram
APB bus
APB Interface ctrl status AHB1 Rd/Wr/Ready DMA AHB0
DHSDP DHSDM
AHB bus
USB2.0 CORE
UTMI
DFSDP DFSDM
DP DM
AHB bus
AHB Switch Local AHB Slave interface
EPT Alloc
32 bits
DPRAM
16/8 bits
System Clock Domain PMC
USB Clock Domain

746
32.4
Typical Connection
Figure 32-2. Board Schematic
PIO (VBUS DETECT) "B" Receptacle 1 = VBUS 2 = D3 = D+ 4 = GND

1 2
DHSDM/DFSDM
15k 22k
(1)
(1)
Shell = Shield CRPB

3 4
CRPB:1F to 10F
DHSDP/DFSDP 6K8 1% VBG 10 pF GNDUTMI
Note:
The values shown on the 22 k and 15 k resistors are only valid with 3V3 supplied PIOs.

747
32.5

The UDPHS is not continuously clocked. For using the UDPHS, the programmer must first enable the UDPHS Clock in the Power Management Controller (PMC_PCER register). Then enable the PLL (PMC_UCKR register). Finally, enable BIAS in PMC_UCKR register. However, if the application does not require UDPHS operations, the UDPHS clock can be stopped when not needed and restarted later.
32.5.2 Interrupt
The UDPHS interrupt line is connected on one of the internal sources of the Interrupt Controller. Using the UDPHS
Table 32-1. Peripheral IDs Instance UDPHS ID 33
interrupt requires the Interrupt Controller to be programmed first.

748
32.6
32.6.1 UTMI transceivers Sharing

The High Speed USB Host Port A is shared with the High Speed USB Device port and connected to the second UTMI transceiver. The selection between Host Port A and USB Device is controlled by the UDPHS enable bit (EN_UDPHS) located in the UDPHS_CTRL control register.
Figure 32-3. USB Selection
Other Transceivers HS Transceiver
EN_UDPHS 0 Others Ports HS USB Host HS EHCI FS OHCI DMA PA HS USB Device DMA 1
32.6.2 USB V2.0 High Speed Device Port Introduction

The USB V2.0 High Speed Device Port provides communication services between host and attached USB devices. Each device is offered with a collection of communication flows (pipes) associated with each endpoint. Software on the host communicates with a USB Device through a set of communication flows.
32.6.3 USB V2.0 High Speed Transfer Types

A communication flow is carried over one of four transfer types defined by the USB device. A device provides several logical communication pipes with the host. To each logical pipe is associated an endpoint. Transfer through a pipe belongs to one of the four transfer types:
z z z z
Control Transfers: Used to configure a device at attach time and can be used for other device-specific purposes, including control of other pipes on the device. Bulk Data Transfers: Generated or consumed in relatively large burst quantities and have wide dynamic latitude in transmission constraints. Interrupt Data Transfers: Used for timely but reliable delivery of data, for example, characters or coordinates with human-perceptible echo or feedback response characteristics. Isochronous Data Transfers: Occupy a prenegotiated amount of USB bandwidth with a prenegotiated delivery latency. (Also called streaming real time transfers.)
As indicated below, transfers are sequential events carried out on the USB bus. Endpoints must be configured according to the transfer type they handle.
Table 32-2. USB Communication Flow Transfer Control Isochronous Interrupt Bulk Direction Bidirectional Unidirectional Unidirectional Unidirectional Bandwidth Not guaranteed Guaranteed Not guaranteed Not guaranteed Endpoint Size 8, 16, 32, 64 8-1024 8-1024 8-512 Error Detection Yes Yes Yes Yes Retrying Automatic No Yes Yes

749
32.6.4 USB Transfer Event Definitions

A transfer is composed of one or several transactions;
Table 32-3. USB Transfer Events CONTROL (bidirectional) Control Transfers (1) Setup transaction Data IN transactions Status OUT transaction Setup transaction Data OUT transactions Status IN transaction Setup transaction Status IN transaction IN (device toward host) Bulk IN Transfer Interrupt IN Transfer Isochronous IN Transfer (2) OUT (host toward device) Bulk OUT Transfer Interrupt OUT Transfer Isochronous OUT Transfer
(2)
Data IN transaction Data IN transaction Data IN transaction Data IN transaction Data IN transaction Data IN transaction Data OUT transaction Data OUT transaction Data OUT transaction Data OUT transaction Data OUT transactionData OUT transaction
Notes: 1. Control transfer must use endpoints with one bank and can be aborted using a stall handshake. 2. Isochronous transfers must use endpoints configured with two or three banks. An endpoint handles all transactions related to the type of transfer for which it has been configured.
Table 32-4. UDPHS Endpoint Description Endpoint # Mnemonic 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 EPT_0 EPT_1 EPT_2 EPT_3 EPT_4 EPT_5 EPT_6 EPT_7 EPT_8 EPT_9 EPT_10 EPT_11 EPT_12 EPT_13 EPT_14 EPT_15 Nb Bank 1 3 3 2 2 2 2 2 2 2 2 2 2 2 2 2 DMA N Y Y Y Y Y Y Y N N N N N N N N High Band Width N Y Y N N N N N N N N N N N N N Max. Endpoint Size 64 1024 1024 1024 1024 1024 1024 1024 1024 1024 1024 1024 1024 1024 1024 1024 Endpoint Type Control Ctrl/Bulk/Iso(32.3)/Interrupt Ctrl/Bulk/Iso(32.3)/Interrupt Ctrl/Bulk/Iso(32.3)/Interrupt Ctrl/Bulk/Iso(32.3)/Interrupt Ctrl/Bulk/Iso(32.3)/Interrupt Ctrl/Bulk/Iso(32.3)/Interrupt Ctrl/Bulk/Iso(32.3)/Interrupt Ctrl/Bulk/Iso(32.3)/Interrupt Ctrl/Bulk/Iso(32.3)/Interrupt Ctrl/Bulk/Iso(32.3)/Interrupt Ctrl/Bulk/Iso(32.3)/Interrupt Ctrl/Bulk/Iso(32.3)/Interrupt Ctrl/Bulk/Iso(32.3)/Interrupt Ctrl/Bulk/Iso(32.3)/Interrupt Ctrl/Bulk/Iso(32.3)/Interrupt
Note:
1. In Isochronous Mode (Iso), it is preferable that High Band Width capability is available. The size of internal DPRAM is 8 KB. Suspend and resume are automatically detected by the UDPHS device, which notifies the processor by raising an interrupt.

750
32.6.5 USB V2.0 High Speed BUS Transactions

Each transfer results in one or more transactions over the USB bus. There are five kinds of transactions flowing across the bus in packets: 1. 2. 3. 4. 5. Setup Transaction Data IN Transaction Data OUT Transaction Status IN Transaction Status OUT Transaction
Figure 32-4. Control Read and Write Sequences
Setup Stage
Data Stage
Status Stage
Control Write
Setup TX
Data OUT TX
Data OUT TX
Status IN TX
Setup Stage
Data Stage
Status Stage
Control Read
Setup TX
Data IN TX
Data IN TX
Status OUT TX
Setup Stage
Status Stage
No Data Control
Setup TX
Status IN TX
A status IN or OUT transaction is identical to a data IN or OUT transaction.
32.6.6 Endpoint Configuration

The endpoint 0 is always a control endpoint, it must be programmed and active in order to be enabled when the End Of Reset interrupt occurs. To configure the endpoints:
z z
Fill the configuration register (UDPHS_EPTCFG) with the endpoint size, direction (IN or OUT), type (CTRL, Bulk, IT, ISO) and the number of banks. Fill the number of transactions (NB_TRANS) for isochronous endpoints. Verify that the EPT_MAPD flag is set. This flag is set if the endpoint size and the number of banks are correct compared to the FIFO maximum capacity and the maximum number of allowed banks. Configure control flags of the endpoint and enable it in UDPHS_EPTCTLENBx according to UDPHS Endpoint Control Disable Register (Isochronous Endpoint) on page 791.
Note: For control endpoints the direction has no effect.

z z
Control endpoints can generate interrupts and use only 1 bank. All endpoints (except endpoint 0) can be configured either as Bulk, Interrupt or Isochronous. See Table 32-4. UDPHS Endpoint Description.

751
The maximum packet size they can accept corresponds to the maximum endpoint size. Note: The endpoint size of 1024 is reserved for isochronous endpoints. The size of the DPRAM is 8 KB. The DPR is shared by all active endpoints. The memory size required by the active endpoints must not exceed the size of the DPRAM. SIZE_DPRAM = SIZE _EPT0 + NB_BANK_EPT1 x SIZE_EPT1 + NB_BANK_EPT2 x SIZE_EPT2 + NB_BANK_EPT3 x SIZE_EPT3 + NB_BANK_EPT4 x SIZE_EPT4 + NB_BANK_EPT5 x SIZE_EPT5 + NB_BANK_EPT6 x SIZE_EPT6 +... (refer to 32.7.8 UDPHS Endpoint Configuration Register) If a user tries to configure endpoints with a size the sum of which is greater than the DPRAM, then the EPT_MAPD is not set. The application has access to the physical block of DPR reserved for the endpoint through a 64 KB logical address space. The physical block of DPR allocated for the endpoint is remapped all along the 64 KB logical address space. The application can write a 64 KB buffer linearly.
Figure 32-5. Logical Address Space for DPR Access
DPR 8 to 64 B Logical address 8 to 64 B 64 KB EP0 8 to1024 B 64 KB 8 to1024 B 8 to1024 B EP1 ... 64 KB EP2 64 KB EP3 ... 8 to1024 B x banks 1 bank
8 to1024 B
y banks
z banks 8 to1024 B

752
Configuration examples of UDPHS_EPTCTLx (UDPHS Endpoint Control Disable Register (Isochronous Endpoint)) for Bulk IN endpoint type follow below.
z
With DMA
z z
AUTO_VALID: Automatically validate the packet and switch to the next bank. EPT_ENABL: Enable endpoint. TXRDY: An interrupt is generated after each transmission. EPT_ENABL: Enable endpoint.
Without DMA:
z z
Configuration examples of Bulk OUT endpoint type follow below.

z
With DMA
z z
AUTO_VALID: Automatically validate the packet and switch to the next bank. EPT_ENABL: Enable endpoint. RXRDY_TXKL: An interrupt is sent after a new packet has been stored in the endpoint FIFO. EPT_ENABL: Enable endpoint.
Without DMA
z z
32.6.7 DPRAM Management

Endpoints can only be allocated in ascending order, from the endpoint 0 to the last endpoint to be allocated. The user shall therefore configure them in the same order. The allocation of an endpoint x starts when the Number of Banks field in the UDPHS Endpoint Configuration Register (UDPHS_EPTCFGx.BK_NUMBER) is different from zero. Then, the hardware allocates a memory area in the DPRAM and inserts it between the x-1 and x+1 endpoints. The x+1 endpoint memory window slides up and its data is lost. Note that the following endpoint memory windows (from x+2) do not slide. Disabling an endpoint, by writing a one to the Endpoint Disable bit in the UDPHS Endpoint Control Disable Register (UDPHS_EPTCTLDISx.EPT_DISABL), does not reset its configuration:
z z z z
The Endpoint Banks (UDPHS_EPTCFGx.BK_NUMBER), The Endpoint Size (UDPHS_EPTCFGx.EPT_SIZE), The Endpoint Direction (UDPHS_EPTCFGx.EPT_DIR), and The Endpoint Type (UDPHS_EPTCFGx.EPT_TYPE).
To free its memory, the user shall write a zero to the UDPHS_EPTCFGx.BK_NUMBER field. The x+1 endpoint memory window then slides down and its data is lost. Note that the following endpoint memory windows (from x+2) do not slide. Figure 32-6 on page 754 illustrates the allocation and reorganization of the DPRAM in a typical example.

753
Figure 32-6. Allocation and Reorganization of the DPRAM

Free Memory Free Memory Free Memory Free Memory
EPT5
EPT5
EPT5
EPT5 EPT4 Conflict
EPT4 EPT3
EPT4 EPT3 (always allocated) EPT2
EPT4 Lost Memory EPT4 EPT3 (larger size)
EPT2
EPT2
EPT2
EPT1
EPT1
EPT1
EPT1
EPT0 Device: Device:
EPT0 Device:
EPT0 Device:
EPT0
UDPHS_EPTCTLENBx.EPT_ENABL = 1 UDPHS_EPTCTLDIS3.EPT_DISABL = 1 UDPHS_EPTCFG3.BK_NUMBER = 0 UDPHS_EPTCFGx.BK_NUMBER <> 0
UDPHS_EPTCTLENB3.EPT_ENABL = 1 UDPHS_EPTCFG3.BK_NUMBER <> 0
Endpoints 0..5 Activated
Endpoint 3 Disabled
Endpoint 3 Memory Freed
Endpoint 3 Activated
1. 2. 3. 4.
The endpoints 0 to 5 are enabled, configured and allocated in ascending order. Each endpoint then owns a memory area in the DPRAM. The endpoint 3 is disabled, but its memory is kept allocated by the controller. In order to free its memory, its UDPHS_EPTCFGx.BK_NUMBER field is written to zero. The endpoint 4 memory window slides down, but the endpoint 5 does not move. If the user chooses to reconfigure the endpoint 3 with a larger size, the controller allocates a memory area after the endpoint 2 memory area and automatically slides up the endpoint 4 memory window. The endpoint 5 does not move and a memory conflict appears as the memory windows of the endpoints 4 and 5 overlap. The data of these endpoints is potentially lost. There is no way the data of the endpoint 0 can be lost (except if it is de-allocated) as the memory allocation and de-allocation may affect only higher endpoints. Deactivating then reactivating the same endpoint with the same configuration only modifies temporarily the controller DPRAM pointer and size for this endpoint. Nothing changes in the DPRAM, higher endpoints seem not to have been moved and their data is preserved as far as nothing has been written or received into them while changing the allocation state of the first endpoint. When the user writes a value different from zero to the UDPHS_EPTCFGx.BK_NUMBER field, the Endpoint Mapped bit (UDPHS_EPTCFGx.EPT_MAPD) is set only if the configured size and number of banks are correct as compared to the endpoint maximal allowed values and to the maximal FIFO size (i.e. the DPRAM size). The UDPHS_EPTCFGx.EPT_MAPD value does not consider memory allocation conflicts.
Notes: 1. 2.
3.

754
32.6.8 Transfer With DMA

USB packets of any length may be transferred when required by the UDPHS Device. These transfers always feature sequential addressing. Packet data AHB bursts may be locked on a DMA buffer basis for drastic overall AHB bus bandwidth performance boost with paged memories. These clock-cycle consuming memory row (or bank) changes will then likely not occur, or occur only once instead of dozens times, during a single big USB packet DMA transfer in case another AHB master addresses the memory. This means up to 128-word single-cycle unbroken AHB bursts for Bulk endpoints and 256-word single-cycle unbroken bursts for isochronous endpoints. This maximum burst length is then controlled by the lowest programmed USB endpoint size (EPT_SIZE field in the UDPHS_EPTCFGx register) and DMA Size (BUFF_LENGTH field in the UDPHS_DMACONTROLx register).
The USB 2.0 device average throughput may be up to nearly 60 MBytes. Its internal slave average access latency decreases as burst length increases due to the 0 wait-state side effect of unchanged endpoints. If at least 0 wait-state word burst capability is also provided by the external DMA AHB bus slaves, each of both DMA AHB busses need less than 50% bandwidth allocation for full USB 2.0 bandwidth usage at 30 MHz, and less than 25% at 60 MHz. The UDPHS DMA Channel Transfer Descriptor is described in UDPHS DMA Channel Transfer Descriptor on page 810. Note: In case of debug, be careful to address the DMA to an SRAM address even if a remap is done.
Figure 32-7. Example of DMA Chained List

Transfer Descriptor UDPHS Registers (Current Transfer Descriptor) UDPHS Next Descriptor Next Descriptor Address DMA Channel Address DMA Channel Control DMA Channel Address Transfer Descriptor Next Descriptor Address DMA Channel Address DMA Channel Control DMA Channel Control DMA Channel Address DMA Channel Control Null Memory Area Data Buff 1 Transfer Descriptor Next Descriptor Address
Data Buff 2 Data Buff 3

755
32.6.9 Transfer Without DMA

Important. If the DMA is not to be used, it is necessary that it be disabled because otherwise it can be enabled by previous versions of software without warning. If this should occur, the DMA can process data before an interrupt without knowledge of the user. The recommended means to disable DMA is as follows: // Reset IP UDPHS AT91C_BASE_UDPHS->UDPHS_CTRL &= ~AT91C_UDPHS_EN_UDPHS; AT91C_BASE_UDPHS->UDPHS_CTRL |= AT91C_UDPHS_EN_UDPHS; // With OR without DMA !!! for( i=1; i<=((AT91C_BASE_UDPHS->UDPHS_IPFEATURES & AT91C_UDPHS_DMA_CHANNEL_NBR)>>4); i++ ) { // RESET endpoint canal DMA: // DMA stop channel command AT91C_BASE_UDPHS->UDPHS_DMA[i].UDPHS_DMACONTROL = 0; // STOP command // Disable endpoint AT91C_BASE_UDPHS->UDPHS_EPT[i].UDPHS_EPTCTLDIS |= 0XFFFFFFFF; // Reset endpoint config AT91C_BASE_UDPHS->UDPHS_EPT[i].UDPHS_EPTCTLCFG = 0; // Reset DMA channel (Buff count and Control field) AT91C_BASE_UDPHS->UDPHS_DMA[i].UDPHS_DMACONTROL = 0x02; // NON STOP command // Reset DMA channel 0 (STOP) AT91C_BASE_UDPHS->UDPHS_DMA[i].UDPHS_DMACONTROL = 0; // STOP command // Clear DMA channel status (read the register for clear it) AT91C_BASE_UDPHS->UDPHS_DMA[i].UDPHS_DMASTATUS = AT91C_BASE_UDPHS->UDPHS_DMA[i].UDPHS_DMASTATUS; }
32.6.10 Handling Transactions with USB V2.0 Device Peripheral

32.6.10.1 Setup Transaction The setup packet is valid in the DPR while RX_SETUP is set. Once RX_SETUP is cleared by the application, the UDPHS accepts the next packets sent over the device endpoint. When a valid setup packet is accepted by the UDPHS:
z z z z
The UDPHS device automatically acknowledges the setup packet (sends an ACK response) Payload data is written in the endpoint Sets the RX_SETUP interrupt The BYTE_COUNT field in the UDPHS_EPTSTAx register is updated
An endpoint interrupt is generated while RX_SETUP in the UDPHS_EPTSTAx register is not cleared. This interrupt is carried out to the microcontroller if interrupts are enabled for this endpoint. Thus, firmware must detect RX_SETUP polling UDPHS_EPTSTAx or catching an interrupt, read the setup packet in the FIFO, then clear the RX_SETUP bit in the UDPHS_EPTCLRSTA register to acknowledge the setup stage. If STALL_SNT was set to 1, then this bit is automatically reset when a setup token is detected by the device. Then, the device still accepts the setup stage. (See Section 32.6.10.15 STALL on page 766).

756
32.6.10.2 NYET NYET is a High Speed only handshake. It is returned by a High Speed endpoint as part of the PING protocol. High Speed devices must support an improved NAK mechanism for Bulk OUT and control endpoints (except setup stage). This mechanism allows the device to tell the host whether it has sufficient endpoint space for the next OUT transfer (see USB 2.0 spec 8.5.1 NAK Limiting via Ping Flow Control). The NYET/ACK response to a High Speed Bulk OUT transfer and the PING response are automatically handled by hardware in the UDPHS_EPTCTLx register (except when the user wants to force a NAK response by using the NYET_DIS bit). If the endpoint responds instead to the OUT/DATA transaction with an NYET handshake, this means that the endpoint accepted the data but does not have room for another data payload. The host controller must return to using a PING token until the endpoint indicates it has space available.
Figure 32-8. NYET Example with Two Endpoint Banks
data 0 ACK
data 1 NYET
PING
ACK
data 0 NYET
PING
NACK
PING
ACK
t=0
t = 125 s
t = 250 s
t = 375 s
t = 500 s
t = 625 s
E: empty E': begin to empty F: full
Bank 1 E Bank 0 F
Bank 1 F Bank 1 F Bank 0 E' Bank 0 E
Bank 1 F Bank 0 E
Bank 1 F Bank 0 F
Bank 1 E' Bank 0 F
Bank 1 E Bank 0 F
32.6.10.3 Data IN 32.6.10.4 Bulk IN or Interrupt IN Data IN packets are sent by the device during the data or the status stage of a control transfer or during an (interrupt/bulk/isochronous) IN transfer. Data buffers are sent packet by packet under the control of the application or under the control of the DMA channel. There are three ways for an application to transfer a buffer in several packets over the USB:
z z z
packet by packet (see 32.6.10.5 below)

64 KB (see 32.6.10.5 below)
DMA (see 32.6.10.6 below)
32.6.10.5 Bulk IN or Interrupt IN: Sending a Packet Under Application Control (Device to Host) The application can write one or several banks. A simple algorithm can be used by the application to send packets regardless of the number of banks associated to the endpoint. Algorithm Description for Each Packet:
z z z
The application waits for TXRDY flag to be cleared in the UDPHS_EPTSTAx register before it can perform a write access to the DPR. The application writes one USB packet of data in the DPR through the 64 KB endpoint logical memory window. The application sets TXRDY flag in the UDPHS_EPTSETSTAx register.
The application is notified that it is possible to write a new packet to the DPR by the TXRDY interrupt. This interrupt can be enabled or masked by setting the TXRDY bit in the UDPHS_EPTCTLENB/UDPHS_EPTCTLDIS register. Algorithm Description to Fill Several Packets: Using the previous algorithm, the application is interrupted for each packet. It is possible to reduce the application overhead by writing linearly several banks at the same time. The AUTO_VALID bit in the UDPHS_EPTCTLx must be set by writing the AUTO_VALID bit in the UDPHS_EPTCTLENBx register.

757
The auto-valid-bank mechanism allows the transfer of data (IN and OUT) without the intervention of the CPU. This means that bank validation (set TXRDY or clear the RXRDY_TXKL bit) is done by hardware.
z z z
The application checks the BUSY_BANK_STA field in the UDPHS_EPTSTAx register. The application must wait that at least one bank is free. The application writes a number of bytes inferior to the number of free DPR banks for the endpoint. Each time the application writes the last byte of a bank, the TXRDY signal is automatically set by the UDPHS. If the last packet is incomplete (i.e., the last byte of the bank has not been written) the application must set the TXRDY bit in the UDPHS_EPTSETSTAx register.
The application is notified that all banks are free, so that it is possible to write another burst of packets by the BUSY_BANK interrupt. This interrupt can be enabled or masked by setting the BUSY_BANK flag in the UDPHS_EPTCTLENB and UDPHS_EPTCTLDIS registers. This algorithm must not be used for isochronous transfer. In this case, the ping-pong mechanism does not operate. A Zero Length Packet can be sent by setting just the TXRDY flag in the UDPHS_EPTSETSTAx register. 32.6.10.6 Bulk IN or Interrupt IN: Sending a Buffer Using DMA (Device to Host) The UDPHS integrates a DMA host controller. This DMA controller can be used to transfer a buffer from the memory to the DPR or from the DPR to the processor memory under the UDPHS control. The DMA can be used for all transfer types except control transfer. Example DMA configuration: 1. 2. 3. Program UDPHS_DMAADDRESS x with the address of the buffer that should be transferred. Enable the interrupt of the DMA in UDPHS_IEN Program UDPHS_ DMACONTROLx:
z z
Size of buffer to send: size of the buffer to be sent to the host. END_B_EN: The endpoint can validate the packet (according to the values programmed in the AUTO_VALID and SHRT_PCKT fields of UDPHS_EPTCTLx.) (See UDPHS Endpoint Control Disable Register (Isochronous Endpoint) on page 791 and Figure 32-13. Autovalid with DMA) END_BUFFIT: generate an interrupt when the BUFF_COUNT in UDPHS_DMASTATUSx reaches 0. CHANN_ENB: Run and stop at end of buffer
z z
The auto-valid-bank mechanism allows the transfer of data (IN & OUT) without the intervention of the CPU. This means that bank validation (set TXRDY or clear the RXRDY_TXKL bit) is done by hardware. A transfer descriptor can be used. Instead of programming the register directly, a descriptor should be programmed and the address of this descriptor is then given to UDPHS_DMANXTDSC to be processed after setting the LDNXT_DSC field (Load Next Descriptor Now) in UDPHS_DMACONTROLx register. The structure that defines this transfer descriptor must be aligned. Each buffer to be transferred must be described by a DMA Transfer descriptor (see UDPHS DMA Channel Transfer Descriptor on page 810). Transfer descriptors are chained. Before executing transfer of the buffer, the UDPHS may fetch a new transfer descriptor from the memory address pointed by the UDPHS_DMANXTDSCx register. Once the transfer is complete, the transfer status is updated in the UDPHS_DMASTATUSx register. To chain a new transfer descriptor with the current DMA transfer, the DMA channel must be stopped. To do so, INTDIS_DMA and TXRDY may be set in the UDPHS_EPTCTLENBx register. It is also possible for the application to wait for the c omple tion of all trans fers. In t his ca se the LDNXT_DSC f ield in the las t trans fer desc ripto r UDPHS_DMACONTROLx register must be set to 0 and CHANN_ENB set to 1. Then the application can chain a new transfer descriptor. The INTDIS_DMA can be used to stop the current DMA transfer if an enabled interrupt is triggered. This can be used to stop DMA transfers in case of errors. The application can be notified at the end of any buffer transfer (ENB_BUFFIT bit in the UDPHS_DMACONTROLx register).

758
Figure 32-9. Data IN Transfer for Endpoint with One Bank

Prevous Data IN TX Microcontroller Loads Data in FIFO Data is Sent on USB Bus
USB Bus Packets
Token IN
Data IN 1
ACK
Token IN
NAK
Token IN
Data IN 2
ACK
TXRDY Flag (UDPHS_EPTSTAx) Set by firmware
Cleared by hardware
Set by the firmware Interrupt Pending
Cleared by hardware Interrupt Pending Payload in FIFO
TX_COMPLT Flag (UDPHS_EPTSTAx)
Set by hardware DPR access by firmware
Cleared by firmware DPR access by hardware Data IN 2
Cleared by firmware
FIFO Content
Data IN 1
Load in progress
Figure 32-10. Data IN Transfer for Endpoint with Two Banks

Microcontroller Load Data IN Bank 0 USB Bus Packets Microcontroller Load Data IN Bank 1 UDPHS Device Send Bank 0 Microcontroller Load Data IN Bank 0 UDPHS Device Send Bank 1
Token IN
Data IN
ACK
Token IN
Data IN
ACK
Set by Firmware, Cleared by Hardware Data Payload Written switch to next bank in FIFO Bank 0 Virtual TXRDY bank 0 (UDPHS_EPTSTAx) Virtual TXRDY bank 1 (UDPHS_EPTSTAx)
Cleared by Hardware Data Payload Fully Transmitted
Set by Firmware, Data Payload Written in FIFO Bank 1 Interrupt Pending Set by Hardware Interrupt Cleared by Firmware Set by Hardware
TX_COMPLT Flag (UDPHS_EPTSTAx) FIFO (DPR) Bank 0
Written by Microcontroller
Read by USB Device
FIFO (DPR) Bank1
Read by UDPHS Device

759
Figure 32-11. Data IN Followed By Status OUT Transfer at the End of a Control Transfer
Device Sends the Last Data Payload to Host USB Bus Packets Device Sends a Status OUT to Host ACK ACK
Token IN
Data IN
ACK
Token OUT
Data OUT (ZLP)
Token OUT
Data OUT (ZLP)
Interrupt Pending
RXRDY (UDPHS_EPTSTAx) Set by Hardware Cleared by Firmware TX_COMPLT (UDPHS_EPTSTAx) Set by Hardware Cleared by Firmware
Note:
A NAK handshake is always generated at the first status stage token.
Figure 32-12. Data OUT Followed by Status IN Transfer

Host Sends the Last Data Payload to the Device Device Sends a Status IN to the Host
USB Bus Packets
Token OUT
Data OUT
ACK
Token IN
Data IN
ACK
Interrupt Pending RXRDY (UDPHS_EPTSTAx) Set by Hardware TXRDY (UDPHS_EPTSTAx) Set by Firmware Clear by Hardware
Cleared by Firmware
Note:
Before proceeding to the status stage, the software should determine that there is no risk of extra data from the host (data stage). If not certain (non-predictable data stage length), then the software should wait for a NAK-IN interrupt before proceeding to the status stage. This precaution should be taken to avoid collision in the FIFO.

760
Figure 32-13. Autovalid with DMA

Bank (system) Bank 0 Bank 1 Bank 0 Bank 1
Write
write bank 0
write bank 1 bank 1 is full
write bank 0 bank 0 is full
bank 0 is full
Bank 1
Bank 0 IN data 0 IN data 1 IN data 0
Bank (usb)
Bank 0
Bank 1
Bank 0
Bank 1
Virtual TXRDY Bank 0
Virtual TXRDY Bank 1
TXRDY (Virtual 0 & Virtual 1)
Note:
In the illustration above Autovalid validates a bank as full, although this might not be the case, in order to continue processing data and to send to DMA.
32.6.10.7 Isochronous IN Isochronous-IN is used to transmit a stream of data whose timing is implied by the delivery rate. Isochronous transfer provides periodic, continuous communication between host and device. It guarantees bandwidth and low latencies appropriate for telephony, audio, video, etc. If the endpoint is not available (TXRDY_TRER = 0), then the device does not answer to the host. An ERR_FL_ISO interrupt is generated in the UDPHS_EPTSTAx register and once enabled, then sent to the CPU. The STALL_SNT command bit is not used for an ISO-IN endpoint. 32.6.10.8 High Bandwidth Isochronous Endpoint Handling: IN Example For high bandwidth isochronous endpoints, the DMA can be programmed with the number of transactions (BUFF_LENGTH field in UDPHS_DMACONTROLx) and the system should provide the required number of packets per microframe, otherwise, the host will notice a sequencing problem. A response should be made to the first token IN recognized inside a microframe under the following conditions:
z
If at least one bank has been validated, the correct DATAx corresponding to the programmed Number Of Transactions per Microframe (NB_TRANS) should be answered. In case of a subsequent missed or corrupted token IN inside the microframe, the USB 2.0 Core available data bank(s) that should normally have been transmitted during that microframe shall be flushed at its end. If this flush occurs, an error condition is flagged (ERR_FLUSH is set in UDPHS_EPTSTAx).

761
z z
If no bank is validated yet, the default DATA0 ZLP is answered and underflow is flagged (ERR_FL_ISO is set in UDPHS_EPTSTAx). Then, no data bank is flushed at microframe end. If no data bank has been validated at the time when a response should be made for the second transaction of NB_TRANS = 3 transactions microframe, a DATA1 ZLP is answered and underflow is flagged (ERR_FL_ISO is set in UDPHS_EPTSTAx). If and only if remaining untransmitted banks for that microframe are available at its end, they are flushed and an error condition is flagged (ERR_FLUSH is set in UDPHS_EPTSTAx). If no data bank has been validated at the time when a response should be made for the last programmed transaction of a microframe, a DATA0 ZLP is answered and underflow is flagged (ERR_FL_ISO is set in UDPHS_EPTSTAx). If and only if the remaining untransmitted data bank for that microframe is available at its end, it is flushed and an error condition is flagged (ERR_FLUSH is set in UDPHS_EPTSTAx). If at the end of a microframe no valid token IN has been recognized, no data bank is flushed and no error condition is reported.
At the end of a microframe in which at least one data bank has been transmitted, if less than NB_TRANS banks have been validated for that microframe, an error condition is flagged (ERR_TRANS is set in UDPHS_EPTSTAx). Cases of Error (in UDPHS_EPTSTAx)
z z
ERR_FL_ISO: There was no data to transmit inside a microframe, so a ZLP is answered by default. ERR_FLUSH: At least one packet has been sent inside the microframe, but the number of token IN received is lesser than the number of transactions actually validated (TXRDY_TRER) and likewise with the NB_TRANS programmed. ERR_TRANS: At least one packet has been sent inside the microframe, but the number of token IN received is lesser than the number of programmed NB_TRANS transactions and the packets not requested were not validated. ERR_FL_ISO + ERR_FLUSH: At least one packet has been sent inside the microframe, but the data has not been validated in time to answer one of the following token IN. ERR_FL_ISO + ERR_TRANS: At least one packet has been sent inside the microframe, but the data has not been validated in time to answer one of the following token IN and the data can be discarded at the microframe end. ERR_FLUSH + ERR_TRANS: The first token IN has been answered and it was the only one received, a second bank has been validated but not the third, whereas NB_TRANS was waiting for three transactions. ERR_FL_ISO + ERR_FLUSH + ERR_TRANS: The first token IN has been treated, the data for the second Token IN was not available in time, but the second bank has been validated before the end of the microframe. The third bank has not been validated, but three transactions have been set in NB_TRANS.
z z z z
32.6.10.9 Data OUT 32.6.10.10 Bulk OUT or Interrupt OUT Like data IN, data OUT packets are sent by the host during the data or the status stage of control transfer or during an interrupt/bulk/isochronous OUT transfer. Data buffers are sent packet by packet under the control of the application or under the control of the DMA channel. 32.6.10.11 Bulk OUT or Interrupt OUT: Receiving a Packet Under Application Control (Host to Device) Algorithm Description for Each Packet:
z z z z
The application enables an interrupt on RXRDY_TXKL. When an interrupt on RXRDY_TXKL is received, the application knows that UDPHS_EPTSTAx register BYTE_COUNT bytes have been received. The application reads the BYTE_COUNT bytes from the endpoint. The application clears RXRDY_TXKL. If the application does not know the size of the transfer, it may not be a good option to use AUTO_VALID. Because if a zero-length-packet is received, the RXRDY_TXKL is automatically cleared by the AUTO_VALID hardware and if the endpoint interrupt is triggered, the software will not find its originating flag when reading the UDPHS_EPTSTAx register.
Note:

762
Algorithm to Fill Several Packets:

z z
The application enables the interrupts of BUSY_BANK and AUTO_VALID. When a BUSY_BANK interrupt is received, the application knows that all banks available for the endpoint have been filled. Thus, the application can read all banks available.
If the application doesnt know the size of the receive buffer, instead of using the BUSY_BANK interrupt, the application must use RXRDY_TXKL. 32.6.10.12 Bulk OUT or Interrupt OUT: Sending a Buffer Using DMA (Host To Device) To use the DMA setting, the AUTO_VALID field is mandatory. See 32.6.10.6 Bulk IN or Interrupt IN: Sending a Buffer Using DMA (Device to Host) for more information. DMA Configuration Example: 1. 2. 3. First program UDPHS_DMAADDRESSx with the address of the buffer that should be transferred. Enable the interrupt of the DMA in UDPHS_IEN Program the DMA Channelx Control Register:
z z z z z
Size of buffer to be sent. END_B_EN: Can be used for OUT packet truncation (discarding of unbuffered packet data) at the end of DMA buffer. END_BUFFIT: Generate an interrupt when BUFF_COUNT in the UDPHS_DMASTATUSx register reaches 0. END_TR_EN: End of transfer enable, the UDPHS device can put an end to the current DMA transfer, in case of a short packet. END_TR_IT: End of transfer interrupt enable, an interrupt is sent after the last USB packet has been transferred by the DMA, if the USB transfer ended with a short packet. (Beneficial when the receive size is unknown.) CHANN_ENB: Run and stop at end of buffer.
For OUT transfer, the bank will be automatically cleared by hardware when the application has read all the bytes in the bank (the bank is empty). Notes: 1. 2. When a zero-length-packet is received, RXRDY_TXKL bit in UDPHS_EPTSTAx is cleared automatically by AUTO_VALID, and the application knows of the end of buffer by the presence of the END_TR_IT. If the host sends a zero-length packet, and the endpoint is free, then the device sends an ACK. No data is written in the endpoint, the RXRDY_TXKL interrupt is generated, and the BYTE_COUNT field in UDPHS_EPTSTAx is null.

763
Figure 32-14. Data OUT Transfer for Endpoint with One Bank
Host Sends Data Payload Microcontroller Transfers Data Host Sends the Next Data Payload Host Resends the Next Data Payload
USB Bus Packets
Token OUT
Data OUT 1
ACK
Token OUT
Data OUT 2
NAK
Token OUT
Data OUT 2
ACK
RXRDY (UDPHS_EPTSTAx)
Interrupt Pending Set by Hardware Cleared by Firmware, Data Payload Written in FIFO Data OUT 2 Written by UDPHS Device
FIFO (DPR) Content
Data OUT 1 Written by UDPHS Device
Data OUT 1 Microcontroller Read
Figure 32-15. Data OUT Transfer for an Endpoint with Two Banks
Host sends first data payload Microcontroller reads Data 1 in bank 0, Host sends second data payload Microcontroller reads Data 2 in bank 1, Host sends third data payload
USB Bus Packets
Token OUT
Data OUT 1
ACK
Token OUT
Data OUT 2
ACK
Token OUT
Data OUT 3
Virtual RXRDY Bank 0
Interrupt pending Set by Hardware, Data payload written in FIFO endpoint bank 0 Set by Hardware Data Payload written in FIFO endpoint bank 1
Cleared by Firmware
Cleared by Firmware Interrupt pending
Virtual RXRDY Bank 1 RXRDY = (virtual bank 0 | virtual bank 1) (UDPHS_EPTSTAx) FIFO (DPR) Bank 0 Data OUT 1 Write by UDPHS Device FIFO (DPR) Bank 1
Data OUT 1 Read by Microcontroller
Data OUT 3 Write in progress
Data OUT 2 Write by Hardware
Data OUT 2 Read by Microcontroller
32.6.10.13 High Bandwidth Isochronous Endpoint OUT

Figure 32-16. Bank Management, Example of Three Transactions per Microframe
USB bus Transactions
MDATA0
MDATA1
DATA2
MDATA0
MDATA1
DATA2
t=0 RXRDY Microcontroller FIFO (DPR) Access
t = 52.5 s (40% of 125 s)
t = 125 s RXRDY
USB line
Read Bank 1
Read Bank 2
Read Bank 3
Read Bank 1

764
USB 2.0 supports individual High Speed isochronous endpoints that require data rates up to 192 Mb/s (24 MB/s): 3x1024 data bytes per microframe. To support such a rate, two or three banks may be used to buffer the three consecutive data packets. The microcontroller (or the DMA) should be able to empty the banks very rapidly (at least 24 MB/s on average). NB_TRANS field in UDPHS_EPTCFGx register = Number Of Transactions per Microframe. If NB_TRANS > 1 then it is High Bandwidth. Example:
z
If NB_TRANS = 3, the sequence should be either

z z z
MData0 MData0/Data1 MData0/Data1/Data2 MData0 MData0/Data1 Data0
If NB_TRANS = 2, the sequence should be either

z z
If NB_TRANS = 1, the sequence should be

z
32.6.10.14 Isochronous Endpoint Handling: OUT Example The user can ascertain the bank status (free or busy), and the toggle sequencing of the data packet for each bank with the UDPHS_EPTSTAx register in the three bit fields as follows:
z z z
TOGGLESQ_STA: PID of the data stored in the current bank CURBK: Number of the bank currently being accessed by the microcontroller. BUSY_BANK_STA: Number of busy bank
This is particularly useful in case of a missing data packet. If the inter-packet delay between the OUT token and the Data is greater than the USB standard, then the ISO-OUT transaction is ignored. (Payload data is not written, no interrupt is generated to the CPU.) If there is a data CRC (Cyclic Redundancy Check) error, the payload is, none the less, written in the endpoint. The ERR_CRC_NTR flag is set in UDPHS_EPTSTAx register. If the endpoint is already full, the packet is not written in the DPRAM. The ERR_FL_ISO flag is set in UDPHS_EPTSTAx. If the payload data is greater than the maximum size of the endpoint, then the ERR_OVFLW flag is set. It is the task of the CPU to manage this error. The data packet is written in the endpoint (except the extra data). If the host sends a Zero Length Packet, and the endpoint is free, no data is written in the endpoint, the RXRDY_TXKL flag is set, and the BYTE_COUNT field in UDPHS_EPTSTAx register is null. The FRCESTALL command bit is unused for an isochonous endpoint. Otherwise, payload data is written in the endpoint, the RXRDY_TXKL interrupt is generated and the BYTE_COUNT in UDPHS_EPTSTAx register is updated.

765
32.6.10.15 STALL STALL is returned by a function in response to an IN token or after the data phase of an OUT or in response to a PING transaction. STALL indicates that a function is unable to transmit or receive data, or that a control pipe request is not supported.
z
OUT
To stall an endpoint, set the FRCESTALL bit in UDPHS_EPTSETSTAx register and after the STALL_SNT flag has been set, set the TOGGLE_SEG bit in the UDPHS_EPTCLRSTAx register.
z
IN
Set the FRCESTALL bit in UDPHS_EPTSETSTAx register.

Figure 32-17. Stall Handshake Data OUT Transfer
USB Bus Packets Token OUT Data OUT Stall PID
FRCESTALL Set by Firmware Cleared by Firmware Interrupt Pending
STALL_SNT Set by Hardware Cleared by Firmware
Figure 32-18. Stall Handshake Data IN Transfer

USB Bus Packets
Token IN
Stall PID
FRCESTALL Set by Firmware Cleared by Firmware Interrupt Pending STALL_SNT Set by Hardware Cleared by Firmware

766
32.6.11 Speed Identification

The high speed reset is managed by the hardware. At the connection, the host makes a reset which could be a classic reset (full speed) or a high speed reset. At the end of the reset process (full or high), the ENDRESET interrupt is generated. Then the CPU should read the SPEED bit in UDPHS_INTSTAx to ascertain the speed mode of the device.
32.6.12 USB V2.0 High Speed Global Interrupt

Interrupts are defined in Section 32.7.3 UDPHS Interrupt Enable Register (UDPHS_IEN) and in Section 32.7.4 UDPHS Interrupt Status Register (UDPHS_INTSTA).
32.6.13 Endpoint Interrupts

Interrupts are enabled in UDPHS_IEN (see Section 32.7.3 UDPHS Interrupt Enable Register) and individually masked in UDPHS_EPTCTLENBx (see Section 32.7.9 UDPHS Endpoint Control Enable Register (Control, Bulk, Interrupt Endpoints)).
Table 32-5. Endpoint Interrupt Source Masks SHRT_PCKT BUSY_BANK NAK_OUT NAK_IN/ERR_FLUSH STALL_SNT/ERR_CRC_NTR RX_SETUP/ERR_FL_ISO TXRDY_TRER TX_COMPLT RXRDY_TXKL ERR_OVFLW MDATA_RX DATAX_RX Short Packet Interrupt Busy Bank Interrupt NAKOUT Interrupt NAKIN/Error Flush Interrupt Stall Sent/CRC error/Number of Transaction Error Interrupt Received SETUP/Error Flow Interrupt TX Packet Read/Transaction Error Interrupt Transmitted IN Data Complete Interrupt Received OUT Data Interrupt Overflow Error Interrupt MDATA Interrupt DATAx Interrupt

767
Figure 32-19. UDPHS Interrupt Control Interface

(UDPHS_IEN) DET_SUSPD MICRO_SOF USB Global IT Sources INT_SOF ENDRESET WAKE_UP ENDOFRSM UPSTR_RES Global IT mask Global IT sources
(UDPHS_EPTCTLENBx) SHRT_PCKT BUSY_BANK NAK_OUT NAK_IN/ERR_FLUSH STALL_SNT/ER_CRC_NTR EPT0 IT Sources RX_SETUP/ERR_FL_ISO TXRDY_TRER TX_COMPLT RXRDY_TXKL ERR_OVFLW MDATA_RX DATAX_RX EP mask EP sources (UDPHS_IEN) EPT_x EP sources EP mask (UDPHS_IEN) EPT_0 husb2dev interrupt
EPT1-6 IT Sources
(UDPHS_EPTCTLx) INTDIS_DMA disable DMA channelx request
(UDPHS_DMACONTROLx) EN_BUFFIT
mask (UDPHS_IEN) DMA_x mask
DMA CH x
END_TR_IT mask DESC_LD_IT

768
32.6.14 Power Modes

32.6.14.1 Controlling Device States A USB device has several possible states. Refer to Chapter 9 (USB Device Framework) of the Universal Serial Bus Specification, Rev 2.0.
Figure 32-20. UDPHS Device State Diagram
Attached
Hub Reset Hub or Configured Deconfigured Bus Inactive Powered Bus Activity Power Interruption Reset Bus Inactive Default Reset Address Assigned Bus Inactive Address Bus Activity Device Deconfigured Device Configured Bus Inactive Suspended Bus Activity Suspended Bus Activity Suspended Suspended
Configured
Movement from one state to another depends on the USB bus state or on standard requests sent through control transactions via the default endpoint (endpoint 0). After a period of bus inactivity, the USB device enters Suspend Mode. Accepting Suspend/Resume requests from the USB host is mandatory. Constraints in Suspend Mode are very strict for bus-powered applications; devices may not consume more than 500 A on the USB bus. While in Suspend Mode, the host may wake up a device by sending a resume signal (bus activity) or a USB device may send a wake-up request to the host, e.g., waking up a PC by moving a USB mouse. The wake-up feature is not mandatory for all devices and must be negotiated with the host.

769
32.6.14.2 Not Powered State Self powered devices can detect 5V VBUS using a PIO. When the device is not connected to a host, device power consumption can be reduced by the DETACH bit in UDPHS_CTRL. Disabling the transceiver is automatically done. HSDM, HSDP, FSDP and FSDP lines are tied to GND pull-downs integrated in the hub downstream ports. 32.6.14.3 Entering Attached State When no device is connected, the USB FSDP and FSDM signals are tied to GND by 15 K pull-downs integrated in the hub downstream ports. When a device is attached to an hub downstream port, the device connects a 1.5 K pull-up on FSDP. The USB bus line goes into IDLE state, FSDP is pulled-up by the device 1.5 K resistor to 3.3V and FSDM is pulled-down by the 15 K resistor to GND of the host. After pull-up connection, the device enters the powered state. The transceiver remains disabled until bus activity is detected. In case of low power consumption need, the device can be stopped. When the device detects the VBUS, the software must enable the USB transceiver by enabling the EN_UDPHS bit in UDPHS_CTRL register. The software can detach the pull-up by setting DETACH bit in UDPHS_CTRL register. 32.6.14.4 From Powered State to Default State (Reset) After its connection to a USB host, the USB device waits for an end-of-bus reset. The unmasked flag ENDRESET is set in the UDPHS_IEN register and an interrupt is triggered. Once the ENDRESET interrupt has been triggered, the device enters Default State. In this state, the UDPHS software must:
z
Enable the default endpoint, setting the EPT_ENABL flag in the UDPHS_EPTCTLENB[0] register and, optionally, enabling the interrupt for endpoint 0 by writing 1 in EPT_0 of the UDPHS_IEN register. The enumeration then begins by a control transfer. Configure the Interrupt Mask Register which has been reset by the USB reset detection Enable the transceiver.
z z
In this state, the EN_UDPHS bit in UDPHS_CTRL register must be enabled. 32.6.14.5 From Default State to Address State (Address Assigned) After a Set Address standard device request, the USB host peripheral enters the address state. Warning: before the device enters address state, it must achieve the Status IN transaction of the control transfer, i.e., the UDPHS device sets its new address once the TX_COMPLT flag in the UDPHS_EPTCTL[0] register has been received and cleared. To move to address state, the driver software sets the DEV_ADDR field and the FADDR_EN flag in the UDPHS_CTRL register. 32.6.14.6 From Address State to Configured State (Device Configured) Once a valid Set Configuration standard request has been received and acknowledged, the device enables endpoints corresponding to the current configuration. This is done by setting the BK_NUMBER, EPT_TYPE, EPT_DIR and EPT_SIZE fields in the UDPHS_EPTCFGx registers and enabling them by setting the EPT_ENABL flag in the UDPHS_EPTCTLENBx registers, and, optionally, enabling corresponding interrupts in the UDPHS_IEN register. 32.6.14.7 Entering Suspend State (Bus Activity) When a Suspend (no bus activity on the USB bus) is detected, the DET_SUSPD signal in the UDPHS_STA register is set. This triggers an interrupt if the corresponding bit is set in the UDPHS_IEN register. This flag is cleared by writing to the UDPHS_CLRINT register. Then the device enters Suspend Mode. In this state bus powered devices must drain less than 500 A from the 5V VBUS. As an example, the microcontroller switches to slow clock, disables the PLL and main oscillator, and goes into Idle Mode. It may also switch off other devices on the board. The UDPHS device peripheral clocks can be switched off. Resume event is asynchronously detected.

770
32.6.14.8 Receiving a Host Resume In Suspend mode, a resume event on the USB bus line is detected asynchronously, transceiver and clocks disabled (however the pull-up should not be removed). Once the resume is detected on the bus, the signal WAKE_UP in the UDPHS_INTSTA is set. It may generate an interrupt if the corresponding bit in the UDPHS_IEN register is set. This interrupt may be used to wake-up the core, enable PLL and main oscillators and configure clocks. 32.6.14.9 Sending an External Resume In Suspend State it is possible to wake-up the host by sending an external resume. The device waits at least 5 ms after being entered in Suspend State before sending an external resume. The device must force a K state from 1 to 15 ms to resume the host.
32.6.15 Test Mode A device must support the TEST_MODE feature when in the Default, Address or Configured High Speed device states. TEST_MODE can be: z Test_J z Test_K z Test_Packet z Test_SEO_NAK
(See Section 32.7.7 UDPHS Test Register on page 781 for definitions of each test mode.) const char test_packet_buffer[] = { 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, 0xAA,0xAA,0xAA,0xAA,0xAA,0xAA,0xAA,0xAA, 0xEE,0xEE,0xEE,0xEE,0xEE,0xEE,0xEE,0xEE, 0xFE,0xFF,0xFF,0xFF,0xFF,0xFF,0xFF,0xFF,0xFF,0xFF,0xFF,0xFF, JJJJJJJKKKKKKK * 8 0x7F,0xBF,0xDF,0xEF,0xF7,0xFB,0xFD, 0xFC,0x7E,0xBF,0xDF,0xEF,0xF7,0xFB,0xFD,0x7E 10}, JK };
// JKJKJKJK * 9 // JJKKJJKK * 8 // JJKKJJKK * 8 // // JJJJJJJK * 8 // {JKKKKKKK *

771
32.7
Offset
0x00 0x04
USB High Speed Device Port (UDPHS) User Interface

Register
UDPHS Control Register UDPHS Frame Number Register Reserved UDPHS Interrupt Enable Register UDPHS Interrupt Status Register UDPHS Clear Interrupt Register UDPHS Endpoints Reset Register Reserved UDPHS Test Register Reserved UDPHS Endpoint Configuration Register UDPHS Endpoint Control Enable Register UDPHS Endpoint Control Disable Register UDPHS Endpoint Control Register Reserved (for endpoint) UDPHS Endpoint Set Status Register UDPHS Endpoint Clear Status Register UDPHS Endpoint Status Register UDPHS Endpoint1 to 15 (2) Registers UDPHS DMA Next Descriptor Address Register UDPHS DMA Channel Address Register UDPHS DMA Channel Control Register UDPHS DMA Channel Status Register DMA Channel1 to 6
(3)
Table 32-6. Register Mapping Name

UDPHS_CTRL UDPHS_FNUM UDPHS_IEN UDPHS_INTSTA UDPHS_CLRINT UDPHS_EPTRST UDPHS_TST UDPHS_EPTCFG UDPHS_EPTCTLENB UDPHS_EPTCTLDIS UDPHS_EPTCTL UDPHS_EPTSETSTA UDPHS_EPTCLRSTA UDPHS_EPTSTA
Access
Read-write Read-only Read-write Read-only Write-only Write-only Read-write Read-write Write-only Write-only Read-only Write-only Write-only Read-only
Reset
0x0000_0200 0x0000_0000 0x0000_0010 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000(1) 0x0000_0040
0x08 - 0x0C 0x10 0x14 0x18 0x1C 0x20 - 0xCC 0xE0 0xE4 - 0xE8 0x100 + endpoint * 0x20 + 0x00 0x100 + endpoint * 0x20 + 0x04 0x100 + endpoint * 0x20 + 0x08 0x100 + endpoint * 0x20 + 0x0C 0x100 + endpoint * 0x20 + 0x10 0x100 + endpoint * 0x20 + 0x14 0x100 + endpoint * 0x20 + 0x18 0x100 + endpoint * 0x20 + 0x1C 0x120 - 0x2FC 0x300 + channel * 0x10 + 0x00 0x300 + channel * 0x10 + 0x04 0x300 + channel * 0x10 + 0x08 0x300 + channel * 0x10 + 0x0C 0x310 - 0x36C
UDPHS_DMANXTDSC UDPHS_DMAADDRESS UDPHS_DMACONTROL UDPHS_DMASTATUS
Read-write Read-write Read-write Read-write
0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000
Registers
Notes: 1. The reset value for UDPHS_EPTCTL0 is 0x0000_0001. 2. The addresses for the UDPHS Endpoint registers shown here are for UDPHS Endpoint0. The structure of this group of registers is repeated successively for each endpoint according to the consecution of endpoint registers located between 0x120 and 0x2FC. 3. The DMA channel index refers to the corresponding EP number. When no DMA channel is assigned to one EP, the associated registers are reserved. This is the case for EP0, so DMA Channel 0 registers are reserved.

772
32.7.1 UDPHS Control Register

31 23 15 7 FADDR_EN
UDPHS_CTRL 0xF8030000 Read-write

30 22 14 6 29 21 13 5 28 20 12 4 27 19 11 PULLD_DIS 3 DEV_ADDR 26 18 10 REWAKEUP 2 25 17 9 DETACH 1 24 16 8 EN_UDPHS 0
DEV_ADDR: UDPHS Address

This field contains the default address (0) after power-up or UDPHS bus reset (read), or it is written with the value set by a SET_ADDRESS request received by the device firmware (write).
FADDR_EN: Function Address Enable

0 = Device is not in address state (read), or only the default function address is used (write). 1 = Device is in address state (read), or this bit is set by the device firmware after a successful status phase of a SET_ADDRESS transaction (write). When set, the only address accepted by the UDPHS controller is the one stored in the UDPHS Address field. It will not be cleared afterwards by the device firmware. It is cleared by hardware on hardware reset, or when UDPHS bus reset is received.
EN_UDPHS: UDPHS Enable

0 = UDPHS is disabled (read), or this bit disables and resets the UDPHS controller (write). Switch the host to UTMI. 1 = UDPHS is enabled (read), or this bit enables the UDPHS controller (write). Switch the host to UTMI.
DETACH: Detach Command

0 = UDPHS is attached (read), or this bit pulls up the DP line (attach command) (write). 1 = UDPHS is detached, UTMI transceiver is suspended (read), or this bit simulates a detach on the UDPHS line and forces the UTMI transceiver into suspend state (Suspend M = 0) (write). See PULLD_DIS description below.
REWAKEUP: Send Remote Wake Up

0 = Remote Wake Up is disabled (read), or this bit has no effect (write). 1 = Remote Wake Up is enabled (read), or this bit forces an external interrupt on the UDPHS controller for Remote Wake UP purposes. An Upstream Resume is sent only after the UDPHS bus has been in SUSPEND state for at least 5 ms. This bit is automatically cleared by hardware at the end of the Upstream Resume.

773
PULLD_DIS: Pull-Down Disable

When set, there is no pull-down on DP & DM. (DM Pull-Down = DP Pull-Down = 0). Note: If the DETACH bit is also set, device DP & DM are left in high impedance state. (See DETACH description above.)
DETACH 0 0 1 1 PULLD_DIS 0 1 0 1 DP Pull up Pull up Pull down High impedance state DM Pull down High impedance state Pull down High impedance state Condition Not recommended VBUS present No VBUS VBUS present & software disconnect

774
32.7.2 UDPHS Frame Number Register

31 FNUM_ERR 23 15 7
UDPHS_FNUM 0xF8030004 Read-only

30 22 14 6 29 21 13 28 20 12 27 19 26 18 25 17 9 24 16 8
11 10 FRAME_NUMBER 3 2
5 FRAME_NUMBER
1 MICRO_FRAME_NUM
MICRO_FRAME_NUM: Microframe Number

Number of the received microframe (0 to 7) in one frame.This field is reset at the beginning of each new frame (1 ms). One microframe is received each 125 microseconds (1 ms/8).
FRAME_NUMBER: Frame Number as defined in the Packet Field Formats

This field is provided in the last received SOF packet (see INT_SOF in the UDPHS Interrupt Status Register).
FNUM_ERR: Frame Number CRC Error

This bit is set by hardware when a corrupted Frame Number in Start of Frame packet (or Micro SOF) is received. This bit and the INT_SOF (or MICRO_SOF) interrupt are updated at the same time.

775
32.7.3 UDPHS Interrupt Enable Register

31 DMA_7 23 EPT_15 15 EPT_7 7 UPSTR_RES
UDPHS_IEN 0xF8030010 Read-write

30 DMA_6 22 EPT_14 14 EPT_6 6 ENDOFRSM 29 DMA_5 21 EPT_13 13 EPT_5 5 WAKE_UP 28 DMA_4 20 EPT_12 12 EPT_4 4 ENDRESET 27 DMA_3 19 EPT_11 11 EPT_3 3 INT_SOF 26 DMA_2 18 EPT_10 10 EPT_2 2 MICRO_SOF 25 DMA_1 17 EPT_9 9 EPT_1 1 DET_SUSPD 24 16 EPT_8 8 EPT_0 0
DET_SUSPD: Suspend Interrupt Enable 0 = Disable Suspend Interrupt. 1 = Enable Suspend Interrupt. MICRO_SOF: Micro-SOF Interrupt Enable 0 = Disable Micro-SOF Interrupt. 1 = Enable Micro-SOF Interrupt. INT_SOF: SOF Interrupt Enable 0 = Disable SOF Interrupt. 1 = Enable SOF Interrupt. ENDRESET: End Of Reset Interrupt Enable 0 = Disable End Of Reset Interrupt. 1 = Enable End Of Reset Interrupt. Automatically enabled after USB reset. WAKE_UP: Wake Up CPU Interrupt Enable 0 = Disable Wake Up CPU Interrupt. 1 = Enable Wake Up CPU Interrupt. ENDOFRSM: End Of Resume Interrupt Enable 0 = Disable Resume Interrupt. 1 = Enable Resume Interrupt. UPSTR_RES: Upstream Resume Interrupt Enable 0 = Disable Upstream Resume Interrupt. 1 = Enable Upstream Resume Interrupt. EPT_x: Endpoint x Interrupt Enable 0 = Disable the interrupts for this endpoint. 1 = Enable the interrupts for this endpoint. DMA_x: DMA Channel x Interrupt Enable 0 = Disable the interrupts for this channel. 1 = Enable the interrupts for this channel. SAMA5D3 Series [DATASHEET]
776
32.7.4 UDPHS Interrupt Status Register

31 DMA_7 23 EPT_15 15 EPT_7 7 UPSTR_RES
UDPHS_INTSTA 0xF8030014 Read-only

30 DMA_6 22 EPT_14 14 EPT_6 6 ENDOFRSM 29 DMA_5 21 EPT_13 13 EPT_5 5 WAKE_UP 28 DMA_4 20 EPT_12 12 EPT_4 4 ENDRESET 27 DMA_3 19 EPT_11 11 EPT_3 3 INT_SOF 26 DMA_2 18 EPT_10 10 EPT_2 2 MICRO_SOF 25 DMA_1 17 EPT_9 9 EPT_1 1 DET_SUSPD 24 16 EPT_8 8 EPT_0 0 SPEED
SPEED: Speed Status

0 = Reset by hardware when the hardware is in Full Speed mode. 1 = Set by hardware when the hardware is in High Speed mode
DET_SUSPD: Suspend Interrupt

0 = Cleared by setting the DET_SUSPD bit in UDPHS_CLRINT register 1 = Set by hardware when a UDPHS Suspend (Idle bus for three frame periods, a J state for 3 ms) is detected. This triggers a UDPHS interrupt when the DET_SUSPD bit is set in UDPHS_IEN register.
MICRO_SOF: Micro Start Of Frame Interrupt

0 = Cleared by setting the MICRO_SOF bit in UDPHS_CLRINT register. 1 = Set by hardware when an UDPHS micro start of frame PID (SOF) has been detected (every 125 us) or synthesized by the macro. This triggers a UDPHS interrupt when the MICRO_SOF bit is set in UDPHS_IEN. In case of detected SOF, the MICRO_FRAME_NUM field in UDPHS_FNUM register is incremented and the FRAME_NUMBER field doesnt change. Note: The Micro Start Of Frame Interrupt (MICRO_SOF), and the Start Of Frame Interrupt (INT_SOF) are not generated at the same time.
INT_SOF: Start Of Frame Interrupt

0 = Cleared by setting the INT_SOF bit in UDPHS_CLRINT. 1 = Set by hardware when an UDPHS Start Of Frame PID (SOF) has been detected (every 1 ms) or synthesized by the macro. This triggers a UDPHS interrupt when the INT_SOF bit is set in UDPHS_IEN register. In case of detected SOF, in High Speed mode, the MICRO_FRAME_NUMBER field is cleared in UDPHS_FNUM register and the FRAME_NUMBER field is updated.
ENDRESET: End Of Reset Interrupt

0 = Cleared by setting the ENDRESET bit in UDPHS_CLRINT. 1 = Set by hardware when an End Of Reset has been detected by the UDPHS controller. This triggers a UDPHS interrupt when the ENDRESET bit is set in UDPHS_IEN.

777
WAKE_UP: Wake Up CPU Interrupt

0 = Cleared by setting the WAKE_UP bit in UDPHS_CLRINT. 1 = Set by hardware when the UDPHS controller is in SUSPEND state and is re-activated by a filtered non-idle signal from the UDPHS line (not by an upstream resume). This triggers a UDPHS interrupt when the WAKE_UP bit is set in UDPHS_IEN register. When receiving this interrupt, the user has to enable the device controller clock prior to operation. Note: This interrupt is generated even if the device controller clock is disabled.
ENDOFRSM: End Of Resume Interrupt

0 = Cleared by setting the ENDOFRSM bit in UDPHS_CLRINT. 1 = Set by hardware when the UDPHS controller detects a good end of resume signal initiated by the host. This triggers a UDPHS interrupt when the ENDOFRSM bit is set in UDPHS_IEN.
UPSTR_RES: Upstream Resume Interrupt

0 = Cleared by setting the UPSTR_RES bit in UDPHS_CLRINT. 1 = Set by hardware when the UDPHS controller is sending a resume signal called upstream resume. This triggers a UDPHS interrupt when the UPSTR_RES bit is set in UDPHS_IEN.
EPT_x: Endpoint x Interrupt

0 = Reset when the UDPHS_EPTSTAx interrupt source is cleared. 1 = Set by hardware when an interrupt is triggered by the UDPHS_EPTSTAx register and this endpoint interrupt is enabled by the EPT_x bit in UDPHS_IEN.
DMA_x: DMA Channel x Interrupt

0 = Reset when the UDPHS_DMASTATUSx interrupt source is cleared. 1 = Set by hardware when an interrupt is triggered by the DMA Channelx and this endpoint interrupt is enabled by the DMA_x bit in UDPHS_IEN.

778
32.7.5 UDPHS Clear Interrupt Register

31 23 15 7 UPSTR_RES
UDPHS_CLRINT 0xF8030018 Write only

30 22 14 6 ENDOFRSM 29 21 13 5 WAKE_UP 28 20 12 4 ENDRESET 27 19 11 3 INT_SOF 26 18 10 2 MICRO_SOF 25 17 9 1 DET_SUSPD 24 16 8 0
DET_SUSPD: Suspend Interrupt Clear

0 = No effect. 1 = Clear the DET_SUSPD bit in UDPHS_INTSTA.
MICRO_SOF: Micro Start Of Frame Interrupt Clear

0 = No effect. 1 = Clear the MICRO_SOF bit in UDPHS_INTSTA.
INT_SOF: Start Of Frame Interrupt Clear

0 = No effect. 1 = Clear the INT_SOF bit in UDPHS_INTSTA.
ENDRESET: End Of Reset Interrupt Clear

0 = No effect. 1 = Clear the ENDRESET bit in UDPHS_INTSTA.
WAKE_UP: Wake Up CPU Interrupt Clear

0 = No effect. 1 = Clear the WAKE_UP bit in UDPHS_INTSTA.
ENDOFRSM: End Of Resume Interrupt Clear

0 = No effect. 1 = Clear the ENDOFRSM bit in UDPHS_INTSTA.
UPSTR_RES: Upstream Resume Interrupt Clear

0 = No effect. 1 = Clear the UPSTR_RES bit in UDPHS_INTSTA.

779
32.7.6 UDPHS Endpoints Reset Register

31 23 15 EPT_15 7 EPT_7
UDPHS_EPTRST 0xF803001C Write only

30 22 14 EPT_14 6 EPT_6 29 21 13 EPT_13 5 EPT_5 28 20 12 EPT_12 4 EPT_4 27 19 11 EPT_11 3 EPT_3 26 18 10 EPT_10 2 EPT_2 25 17 9 EPT_9 1 EPT_1 24 16 8 EPT_8 0 EPT_0
EPT_x: Endpoint x Reset

0 = No effect. 1 = Reset the Endpointx state. Setting this bit clears the Endpoint status UDPHS_EPTSTAx register, except for the TOGGLESQ_STA field.

780
32.7.7 UDPHS Test Register

31 23 15 7
UDPHS_TST 0xF80300E0 Read-write

30 22 14 6 29 21 13 5 OPMODE2 28 20 12 4 TST_PKT 27 19 11 3 TST_K 26 18 10 2 TST_J 25 17 9 1 SPEED_CFG 24 16 8 0
SPEED_CFG: Speed Configuration

Speed Configuration:
Value 0 1 2 3 HIGH_SPEED FULL_SPEED Name NORMAL Description Normal Mode: The macro is in Full Speed mode, ready to make a High Speed identification, if the host supports it and then to automatically switch to High Speed mode Reserved Force High Speed: Set this value to force the hardware to work in High Speed mode. Only for debug or test purpose. Force Full Speed: Set this value to force the hardware to work only in Full Speed mode. In this configuration, the macro will not respond to a High Speed reset handshake.
TST_J: Test J Mode

0 = No effect. 1 = Set to send the J state on the UDPHS line. This enables the testing of the high output drive level on the D+ line.
TST_K: Test K Mode

0 = No effect. 1 = Set to send the K state on the UDPHS line. This enables the testing of the high output drive level on the D- line.
TST_PKT: Test Packet Mode

0 = No effect. 1 = Set to repetitively transmit the packet stored in the current bank. This enables the testing of rise and fall times, eye patterns, jitter, and any other dynamic waveform specifications.
OPMODE2: OpMode2
0 = No effect. 1 = Set to force the OpMode signal (UTMI interface) to 10, to disable the bit-stuffing and the NRZI encoding. Note: For the Test mode, Test_SE0_NAK (see Universal Serial Bus Specification, Revision 2.0: 7.1.20, Test Mode Support). Force the device in High Speed mode, and configure a bulk-type endpoint. Do not fill this endpoint for sending NAK to the host.

781
Upon command, a ports transceiver must enter the High Speed receive mode and remain in that mode until the exit action is taken. This enables the testing of output impedance, low level output voltage and loading characteristics. In addition, while in this mode, upstream facing ports (and only upstream facing ports) must respond to any IN token packet with a NAK handshake (only if the packet CRC is determined to be correct) within the normal allowed device response time. This enables testing of the device squelch level circuitry and, additionally, provides a general purpose stimulus/response test for basic functional testing.

782
32.7.8 UDPHS Endpoint Configuration Register

Name: Address: UDPHS_EPTCFGx [x=0..15] 0xF8030100 [0], 0xF8030120 [1], 0xF8030140 [2], 0xF8030160 [3], 0xF8030180 [4], 0xF80301A0 [5], 0xF80301C0 [6], 0xF80301E0 [7], 0xF8030200 [8], 0xF8030220 [9], 0xF8030240 [10], 0xF8030260 [11], 0xF8030280 [12], 0xF80302A0 [13], 0xF80302C0 [14], 0xF80302E0 [15] Read-write
30 22 14 6 BK_NUMBER 29 21 13 5 EPT_TYPE 28 20 12 4 27 19 11 3 EPT_DIR 26 18 10 2 25 17 9 NB_TRANS 1 EPT_SIZE 0 24 16 8
Access:
31 EPT_MAPD 23 15 7
EPT_SIZE: Endpoint Size

Set this field according to the endpoint size in bytes (see Section 32.6.6 Endpoint Configuration). Endpoint Size (1)
Value 0 1 2 3 4 5 6 7 Name 8 16 32 64 128 256 512 1024 Description 8 bytes 16 bytes 32 bytes 64 bytes 128 bytes 256 bytes 512 bytes 1024 bytes
Note:
1. 1024 bytes is only for isochronous endpoint.
EPT_DIR: Endpoint Direction

0 = Clear this bit to configure OUT direction for Bulk, Interrupt and Isochronous endpoints. 1 = Set this bit to configure IN direction for Bulk, Interrupt and Isochronous endpoints. For Control endpoints this bit has no effect and should be left at zero.
EPT_TYPE: Endpoint Type

Set this field according to the endpoint type (see Section 32.6.6 Endpoint Configuration). (Endpoint 0 should always be configured as control)

783
Endpoint Type
Value 0 1 2 3 Name CTRL8 ISO BULK INT Description Control endpoint Isochronous endpoint Bulk endpoint Interrupt endpoint
BK_NUMBER: Number of Banks

Set this field according to the endpoints number of banks (see Section 32.6.6 Endpoint Configuration). Number of Banks
Value 0 1 2 3 Name 0 1 2 3 Description Zero bank, the endpoint is not mapped in memory One bank (bank 0) Double bank (Ping-Pong: bank0/bank1) Triple bank (bank0/bank1/bank2)
NB_TRANS: Number Of Transaction per Microframe

The Number of transactions per microframe is set by software. Note: Meaningful for high bandwidth isochronous endpoint only.
EPT_MAPD: Endpoint Mapped

0 = The user should reprogram the register with correct values. 1 = Set by hardware when the endpoint size (EPT_SIZE) and the number of banks (BK_NUMBER) are correct regarding:
The fifo max capacity (FIFO_MAX_SIZE in UDPHS_IPFEATURES register) The number of endpoints/banks already allocated The number of allowed banks for this endpoint

784
32.7.9 UDPHS Endpoint Control Enable Register (Control, Bulk, Interrupt Endpoints)
Name: Access:
31 SHRT_PCKT 23 15 NAK_OUT 7
UDPHS_EPTCTLENBx [x=0..15] Write-only

30 22 14 NAK_IN 6 29 21 13 STALL_SNT 5 28 20 12 RX_SETUP 4 NYET_DIS 27 19 11 TXRDY 3 INTDIS_DMA 26 18 BUSY_BANK 10 TX_COMPLT 2 25 17 9 RXRDY_TXKL 1 AUTO_VALID 24 16 8 ERR_OVFLW 0 EPT_ENABL
This register view is relevant only if EPT_TYPE=0x0, 0x2 or 0x3 in UDPHS Endpoint Configuration Register on page 783 For additional Information, see UDPHS Endpoint Control Register (Control, Bulk, Interrupt Endpoints) on page 793.
EPT_ENABL: Endpoint Enable

0 = No effect. 1 = Enable endpoint according to the device configuration.
AUTO_VALID: Packet Auto-Valid Enable

0 = No effect. 1 = Enable this bit to automatically validate the current packet and switch to the next bank for both IN and OUT transfers.
INTDIS_DMA: Interrupts Disable DMA

0 = No effect. 1 = If set, when an enabled endpoint-originated interrupt is triggered, the DMA request is disabled.
NYET_DIS: NYET Disable (Only for High Speed Bulk OUT endpoints)
0 = No effect. 1 = Forces an ACK response to the next High Speed Bulk OUT transfer instead of a NYET response.
ERR_OVFLW: Overflow Error Interrupt Enable

0 = No effect. 1 = Enable Overflow Error Interrupt.
RXRDY_TXKL: Received OUT Data Interrupt Enable

0 = No effect. 1 = Enable Received OUT Data Interrupt.
TX_COMPLT: Transmitted IN Data Complete Interrupt Enable

0 = No effect. 1 = Enable Transmitted IN Data Complete Interrupt.

785
TXRDY: TX Packet Ready Interrupt Enable

0 = No effect. 1 = Enable TX Packet Ready/Transaction Error Interrupt.
RX_SETUP: Received SETUP

0 = No effect. 1 = Enable RX_SETUP Interrupt.
STALL_SNT: Stall Sent Interrupt Enable

0 = No effect. 1 = Enable Stall Sent Interrupt.
NAK_IN: NAKIN Interrupt Enable

0 = No effect. 1 = Enable NAKIN Interrupt.
NAK_OUT: NAKOUT Interrupt Enable

0 = No effect. 1 = Enable NAKOUT Interrupt.
BUSY_BANK: Busy Bank Interrupt Enable

0 = No effect. 1 = Enable Busy Bank Interrupt.
SHRT_PCKT: Short Packet Send/Short Packet Interrupt Enable

For OUT endpoints: 0 = No effect. 1 = Enable Short Packet Interrupt. For IN endpoints: Guarantees short packet at end of DMA Transfer if the UDPHS_DMACONTROLx register END_B_EN and UDPHS_EPTCTLx register AUTOVALID bits are also set.

786
32.7.10 UDPHS Endpoint Control Enable Register (Isochronous Endpoints)

Name: Address: UDPHS_EPTCTLENBx [x=0..15] (ISOENDPT) 0xF8030104 [0], 0xF8030124 [1], 0xF8030144 [2], 0xF8030164 [3], 0xF8030184 [4], 0xF80301A4 [5], 0xF80301C4 [6], 0xF80301E4 [7], 0xF8030204 [8], 0xF8030224 [9], 0xF8030244 [10], 0xF8030264 [11], 0xF8030284 [12], 0xF80302A4 [13], 0xF80302C4 [14], 0xF80302E4 [15] Write-only
30 22 14 ERR_FLUSH 29 21 13 ERR_CRC_NT R 5 28 20 12 ERR_FL_ISO 27 19 11 TXRDY_TRER 26 18 BUSY_BANK 10 TX_COMPLT 25 17 9 RXRDY_TXKL 24 16 8 ERR_OVFLW
Access:
31 SHRT_PCKT 23 15
7 MDATA_RX
6 DATAX_RX
3 INTDIS_DMA
1 AUTO_VALID
0 EPT_ENABL
This register view is relevant only if EPT_TYPE=0x1 in UDPHS Endpoint Configuration Register on page 783 For additional Information, see UDPHS Endpoint Control Register (Isochronous Endpoint) on page 796.

0 = No effect. 1 = Enable endpoint according to the device configuration.
AUTO_VALID: Packet Auto-Valid Enable

0 = No effect. 1 = Enable this bit to automatically validate the current packet and switch to the next bank for both IN and OUT transfers.

0 = No effect. 1 = If set, when an enabled endpoint-originated interrupt is triggered, the DMA request is disabled.
DATAX_RX: DATAx Interrupt Enable (Only for high bandwidth Isochronous OUT endpoints)
0 = No effect. 1 = Enable DATAx Interrupt.
MDATA_RX: MDATA Interrupt Enable (Only for high bandwidth Isochronous OUT endpoints)
0 = No effect. 1 = Enable MDATA Interrupt.
ERR_OVFLW: Overflow Error Interrupt Enable

0 = No effect. 1 = Enable Overflow Error Interrupt.

787
RXRDY_TXKL: Received OUT Data Interrupt Enable

0 = No effect. 1 = Enable Received OUT Data Interrupt.
TX_COMPLT: Transmitted IN Data Complete Interrupt Enable

0 = No effect. 1 = Enable Transmitted IN Data Complete Interrupt.
TXRDY_TRER: TX Packet Ready/Transaction Error Interrupt Enable

0 = No effect. 1 = Enable TX Packet Ready/Transaction Error Interrupt.
ERR_FL_ISO: Error Flow Interrupt Enable

0 = No effect. 1 = Enable Error Flow ISO Interrupt.
ERR_CRC_NTR: ISO CRC Error/Number of Transaction Error Interrupt Enable

0 = No effect. 1 = Enable Error CRC ISO/Error Number of Transaction Interrupt.
ERR_FLUSH: Bank Flush Error Interrupt Enable

0 = No effect. 1 = Enable Bank Flush Error Interrupt.
BUSY_BANK: Busy Bank Interrupt Enable

0 = No effect. 1 = Enable Busy Bank Interrupt.
SHRT_PCKT: Short Packet Send/Short Packet Interrupt Enable

For OUT endpoints: 0 = No effect. 1 = Enable Short Packet Interrupt. For IN endpoints: Guarantees short packet at end of DMA Transfer if the UDPHS_DMACONTROLx register END_B_EN and UDPHS_EPTCTLx register AUTOVALID bits are also set.

788
32.7.11 UDPHS Endpoint Control Disable Register (Control, Bulk, Interrupt Endpoints)
Name: Address: UDPHS_EPTCTLDISx [x=0..15] 0xF8030108 [0], 0xF8030128 [1], 0xF8030148 [2], 0xF8030168 [3], 0xF8030188 [4], 0xF80301A8 [5], 0xF80301C8 [6], 0xF80301E8 [7], 0xF8030208 [8], 0xF8030228 [9], 0xF8030248 [10], 0xF8030268 [11], 0xF8030288 [12], 0xF80302A8 [13], 0xF80302C8 [14], 0xF80302E8 [15] Write-only
30 22 14 NAK_IN 6 29 21 13 STALL_SNT 5 28 20 12 RX_SETUP 4 NYET_DIS 27 19 11 TXRDY 3 INTDIS_DMA 26 18 BUSY_BANK 10 TX_COMPLT 2 25 17 9 RXRDY_TXKL 1 AUTO_VALID 24 16 8 ERR_OVFLW 0 EPT_DISABL
Access:
This register view is relevant only if EPT_TYPE=0x0, 0x2 or 0x3 in UDPHS Endpoint Configuration Register on page 783 For additional Information, see UDPHS Endpoint Control Register (Control, Bulk, Interrupt Endpoints) on page 793.
EPT_DISABL: Endpoint Disable

0 = No effect. 1 = Disable endpoint.
AUTO_VALID: Packet Auto-Valid Disable

0 = No effect. 1 = Disable this bit to not automatically validate the current packet.

0 = No effect. 1 = Disable the Interrupts Disable DMA.
NYET_DIS: NYET Enable (Only for High Speed Bulk OUT endpoints)
0 = No effect. 1 = Let the hardware handle the handshake response for the High Speed Bulk OUT transfer.
ERR_OVFLW: Overflow Error Interrupt Disable

0 = No effect. 1 = Disable Overflow Error Interrupt.
RXRDY_TXKL: Received OUT Data Interrupt Disable

0 = No effect. 1 = Disable Received OUT Data Interrupt.
TX_COMPLT: Transmitted IN Data Complete Interrupt Disable

0 = No effect. 1 = Disable Transmitted IN Data Complete Interrupt.

789
TXRDY: TX Packet Ready Interrupt Disable

0 = No effect. 1 = Disable TX Packet Ready/Transaction Error Interrupt.
RX_SETUP: Received SETUP Interrupt Disable

0 = No effect. 1 = Disable RX_SETUP Interrupt.
STALL_SNT: Stall Sent Interrupt Disable

0 = No effect. 1 = Disable Stall Sent Interrupt.
NAK_IN: NAKIN Interrupt Disable

0 = No effect. 1 = Disable NAKIN Interrupt.
NAK_OUT: NAKOUT Interrupt Disable

0 = No effect. 1 = Disable NAKOUT Interrupt.
BUSY_BANK: Busy Bank Interrupt Disable

0 = No effect. 1 = Disable Busy Bank Interrupt.
SHRT_PCKT: Short Packet Interrupt Disable

For OUT endpoints: 0 = No effect. 1 = Disable Short Packet Interrupt. For IN endpoints: Never automatically add a zero length packet at end of DMA transfer.

790
32.7.12 UDPHS Endpoint Control Disable Register (Isochronous Endpoint)

Name: Address: UDPHS_EPTCTLDISx [x=0..15] (ISOENDPT) 0xF8030108 [0], 0xF8030128 [1], 0xF8030148 [2], 0xF8030168 [3], 0xF8030188 [4], 0xF80301A8 [5], 0xF80301C8 [6], 0xF80301E8 [7], 0xF8030208 [8], 0xF8030228 [9], 0xF8030248 [10], 0xF8030268 [11], 0xF8030288 [12], 0xF80302A8 [13], 0xF80302C8 [14], 0xF80302E8 [15] Write-only
Access:
31 SHRT_PCKT 23 15
7 MDATA_RX
6 DATAX_RX
3 INTDIS_DMA
1 AUTO_VALID
0 EPT_DISABL
This register view is relevant only if EPT_TYPE=0x1 in UDPHS Endpoint Configuration Register on page 783 For additional Information, see UDPHS Endpoint Control Register (Isochronous Endpoint) on page 796.
EPT_DISABL: Endpoint Disable

0 = No effect. 1 = Disable endpoint.
AUTO_VALID: Packet Auto-Valid Disable

0 = No effect. 1 = Disable this bit to not automatically validate the current packet.

0 = No effect. 1 = Disable the Interrupts Disable DMA.
DATAX_RX: DATAx Interrupt Disable (Only for High Bandwidth Isochronous OUT endpoints)
0 = No effect. 1 = Disable DATAx Interrupt.
MDATA_RX: MDATA Interrupt Disable (Only for High Bandwidth Isochronous OUT endpoints)
0 = No effect. 1 = Disable MDATA Interrupt.
ERR_OVFLW: Overflow Error Interrupt Disable

0 = No effect. 1 = Disable Overflow Error Interrupt.

791
RXRDY_TXKL: Received OUT Data Interrupt Disable

0 = No effect. 1 = Disable Received OUT Data Interrupt.
TX_COMPLT: Transmitted IN Data Complete Interrupt Disable

0 = No effect. 1 = Disable Transmitted IN Data Complete Interrupt.
TXRDY_TRER: TX Packet Ready/Transaction Error Interrupt Disable

0 = No effect. 1 = Disable TX Packet Ready/Transaction Error Interrupt.
ERR_FL_ISO: Error Flow Interrupt Disable

0 = No effect. 1 = Disable Error Flow ISO Interrupt.
ERR_CRC_NTR: ISO CRC Error/Number of Transaction Error Interrupt Disable

0 = No effect. 1 = Disable Error CRC ISO/Error Number of Transaction Interrupt.
ERR_FLUSH: bank flush error Interrupt Disable

0 = No effect. 1 = Disable Bank Flush Error Interrupt.
BUSY_BANK: Busy Bank Interrupt Disable

0 = No effect. 1 = Disable Busy Bank Interrupt.
SHRT_PCKT: Short Packet Interrupt Disable

For OUT endpoints: 0 = No effect. 1 = Disable Short Packet Interrupt. For IN endpoints: Never automatically add a zero length packet at end of DMA transfer.

792
32.7.13 UDPHS Endpoint Control Register (Control, Bulk, Interrupt Endpoints)

Name: Access:
UDPHS_EPTCTLx [x=0..15] Read-only

30 22 14 NAK_IN 6 29 21 13 STALL_SNT 5 28 20 12 RX_SETUP 4 NYET_DIS 27 19 11 TXRDY 3 INTDIS_DMA 26 18 BUSY_BANK 10 TX_COMPLT 2 25 17 9 RXRDY_TXKL 1 AUTO_VALID 24 16 8 ERR_OVFLW 0 EPT_ENABL
This register view is relevant only if EPT_TYPE=0x0, 0x2 or 0x3 in UDPHS Endpoint Configuration Register on page 783

0 = If cleared, the endpoint is disabled according to the device configuration. Endpoint 0 should always be enabled after a hardware or UDPHS bus reset and participate in the device configuration. 1 = If set, the endpoint is enabled according to the device configuration.
AUTO_VALID: Packet Auto-Valid Enabled (Not for CONTROL Endpoints)

Set this bit to automatically validate the current packet and switch to the next bank for both IN and OUT endpoints.
For IN Transfer: If this bit is set, then the UDPHS_EPTSTAx register TXRDY bit is set automatically when the current bank is full and at the end of DMA buffer if the UDPHS_DMACONTROLx register END_B_EN bit is set. The user may still set the UDPHS_EPTSTAx register TXRDY bit if the current bank is not full, unless the user wants to send a Zero Length Packet by software. For OUT Transfer: If this bit is set, then the UDPHS_EPTSTAx register RXRDY_TXKL bit is automatically reset for the current bank when the last packet byte has been read from the bank FIFO or at the end of DMA buffer if the UDPHS_DMACONTROLx register END_B_EN bit is set. For example, to truncate a padded data packet when the actual data transfer size is reached. The user may still clear the UDPHS_EPTSTAx register RXRDY_TXKL bit, for example, after completing a DMA buffer by software if UDPHS_DMACONTROLx register END_B_EN bit was disabled or in order to cancel the read of the remaining data bank(s). INTDIS_DMA: Interrupt Disables DMA
If set, when an enabled endpoint-originated interrupt is triggered, the DMA request is disabled regardless of the UDPHS_IEN register EPT_x bit for this endpoint. Then, the firmware will have to clear or disable the interrupt source or clear this bit if transfer completion is needed. If the exception raised is associated with the new system bank packet, then the previous DMA packet transfer is normally completed, but the new DMA packet transfer is not started (not requested). If the exception raised is not associated to a new system bank packet (NAK_IN, NAK_OUT...), then the request cancellation may happen at any time and may immediately stop the current DMA transfer. This may be used, for example, to identify or prevent an erroneous packet to be transferred into a buffer or to complete a DMA buffer by software after reception of a short packet.

793
NYET_DIS: NYET Disable (Only for High Speed Bulk OUT endpoints)
0 = If cleared, this bit lets the hardware handle the handshake response for the High Speed Bulk OUT transfer. 1 = If set, this bit forces an ACK response to the next High Speed Bulk OUT transfer instead of a NYET response. Note: According to the Universal Serial Bus Specification, Rev 2.0 (8.5.1.1 NAK Responses to OUT/DATA During PING Protocol), a NAK response to an HS Bulk OUT transfer is expected to be an unusual occurrence.
ERR_OVFLW: Overflow Error Interrupt Enabled

0 = Overflow Error Interrupt is masked. 1 = Overflow Error Interrupt is enabled.
RXRDY_TXKL: Received OUT Data Interrupt Enabled

0 = Received OUT Data Interrupt is masked. 1 = Received OUT Data Interrupt is enabled.
TX_COMPLT: Transmitted IN Data Complete Interrupt Enabled

0 = Transmitted IN Data Complete Interrupt is masked. 1 = Transmitted IN Data Complete Interrupt is enabled.
TXRDY: TX Packet Ready Interrupt Enabled

0 = TX Packet Ready Interrupt is masked. 1 = TX Packet Ready Interrupt is enabled.
Caution: Interrupt source is active as long as the corresponding UDPHS_EPTSTAx register TXRDY flag remains low. If there are no more banks available for transmitting after the software has set UDPHS_EPTSTAx/TXRDY for the last transmit packet, then the interrupt source remains inactive until the first bank becomes free again to transmit at UDPHS_EPTSTAx/TXRDY hardware clear. RX_SETUP: Received SETUP Interrupt Enabled
0 = Received SETUP is masked. 1 = Received SETUP is enabled.
STALL_SNT: Stall Sent Interrupt Enabled

0 = Stall Sent Interrupt is masked. 1 = Stall Sent Interrupt is enabled.
NAK_IN: NAKIN Interrupt Enabled

0 = NAKIN Interrupt is masked. 1 = NAKIN Interrupt is enabled.
NAK_OUT: NAKOUT Interrupt Enabled

0 = NAKOUT Interrupt is masked. 1 = NAKOUT Interrupt is enabled.
BUSY_BANK: Busy Bank Interrupt Enabled

0 = BUSY_BANK Interrupt is masked. 1 = BUSY_BANK Interrupt is enabled.
For OUT endpoints: an interrupt is sent when all banks are busy. For IN endpoints: an interrupt is sent when all banks are free. SAMA5D3 Series [DATASHEET]
794
SHRT_PCKT: Short Packet Interrupt Enabled For OUT endpoints: send an Interrupt when a Short Packet has been received. 0 = Short Packet Interrupt is masked. 1 = Short Packet Interrupt is enabled. For IN endpoints: a Short Packet transmission is guaranteed upon end of the DMA Transfer, thus signaling a BULK or INTERRUPT end of transfer, but only if the UDPHS_DMACONTROLx register END_B_EN and UDPHS_EPTCTLx register AUTO_VALID bits are also set.

795
32.7.14 UDPHS Endpoint Control Register (Isochronous Endpoint)

Name: Address: UDPHS_EPTCTLx [x=0..15] (ISOENDPT) 0xF803010C [0], 0xF803012C [1], 0xF803014C [2], 0xF803016C [3], 0xF803018C [4], 0xF80301AC [5], 0xF80301CC [6], 0xF80301EC [7], 0xF803020C [8], 0xF803022C [9], 0xF803024C [10], 0xF803026C [11], 0xF803028C [12], 0xF80302AC [13], 0xF80302CC [14], 0xF80302EC [15] Read-only
Access:
31 SHRT_PCKT 23 15
7 MDATA_RX
6 DATAX_RX
3 INTDIS_DMA
1 AUTO_VALID
0 EPT_ENABL
This register view is relevant only if EPT_TYPE=0x1 in UDPHS Endpoint Configuration Register on page 783

0 = If cleared, the endpoint is disabled according to the device configuration. Endpoint 0 should always be enabled after a hardware or UDPHS bus reset and participate in the device configuration. 1 = If set, the endpoint is enabled according to the device configuration.
AUTO_VALID: Packet Auto-Valid Enabled

Set this bit to automatically validate the current packet and switch to the next bank for both IN and OUT endpoints.
For IN Transfer: If this bit is set, then the UDPHS_EPTSTAx register TXRDY_TRER bit is set automatically when the current bank is full and at the end of DMA buffer if the UDPHS_DMACONTROLx register END_B_EN bit is set. The user may still set the UDPHS_EPTSTAx register TXRDY_TRER bit if the current bank is not full, unless the user wants to send a Zero Length Packet by software. For OUT Transfer: If this bit is set, then the UDPHS_EPTSTAx register RXRDY_TXKL bit is automatically reset for the current bank when the last packet byte has been read from the bank FIFO or at the end of DMA buffer if the UDPHS_DMACONTROLx register END_B_EN bit is set. For example, to truncate a padded data packet when the actual data transfer size is reached. The user may still clear the UDPHS_EPTSTAx register RXRDY_TXKL bit, for example, after completing a DMA buffer by software if UDPHS_DMACONTROLx register END_B_EN bit was disabled or in order to cancel the read of the remaining data bank(s). INTDIS_DMA: Interrupt Disables DMA
If set, when an enabled endpoint-originated interrupt is triggered, the DMA request is disabled regardless of the UDPHS_IEN register EPT_x bit for this endpoint. Then, the firmware will have to clear or disable the interrupt source or clear this bit if transfer completion is needed.

796
If the exception raised is associated with the new system bank packet, then the previous DMA packet transfer is normally completed, but the new DMA packet transfer is not started (not requested). If the exception raised is not associated to a new system bank packet (ex: ERR_FL_ISO), then the request cancellation may happen at any time and may immediately stop the current DMA transfer. This may be used, for example, to identify or prevent an erroneous packet to be transferred into a buffer or to complete a DMA buffer by software after reception of a short packet, or to perform buffer truncation on ERR_FL_ISO interrupt for adaptive rate.
DATAX_RX: DATAx Interrupt Enabled (Only for High Bandwidth Isochronous OUT endpoints)
0 = No effect. 1 = Send an interrupt when a DATA2, DATA1 or DATA0 packet has been received meaning the whole microframe data payload has been received.
MDATA_RX: MDATA Interrupt Enabled (Only for High Bandwidth Isochronous OUT endpoints)
0 = No effect. 1 = Send an interrupt when an MDATA packet has been received and so at least one packet of the microframe data payload has been received.
ERR_OVFLW: Overflow Error Interrupt Enabled

0 = Overflow Error Interrupt is masked. 1 = Overflow Error Interrupt is enabled.
RXRDY_TXKL: Received OUT Data Interrupt Enabled

0 = Received OUT Data Interrupt is masked. 1 = Received OUT Data Interrupt is enabled.
TX_COMPLT: Transmitted IN Data Complete Interrupt Enabled

0 = Transmitted IN Data Complete Interrupt is masked. 1 = Transmitted IN Data Complete Interrupt is enabled.
TXRDY_TRER: TX Packet Ready/Transaction Error Interrupt Enabled

0 = TX Packet Ready/Transaction Error Interrupt is masked. 1 = TX Packet Ready/Transaction Error Interrupt is enabled.
Caution: Interrupt source is active as long as the corresponding UDPHS_EPTSTAx register TXRDY_TRER flag remains low. If there are no more banks available for transmitting after the software has set UDPHS_EPTSTAx/TXRDY_TRER for the last transmit packet, then the interrupt source remains inactive until the first bank becomes free again to transmit at UDPHS_EPTSTAx/TXRDY_TRER hardware clear. ERR_FL_ISO: Error Flow Interrupt Enabled
0 = Error Flow Interrupt is masked. 1 = Error Flow Interrupt is enabled.
ERR_CRC_NTR: ISO CRC Error/Number of Transaction Error Interrupt Enabled

0 = ISO CRC error/number of Transaction Error Interrupt is masked. 1 = ISO CRC error/number of Transaction Error Interrupt is enabled.
ERR_FLUSH: Bank Flush Error Interrupt Enabled

0 = Bank Flush Error Interrupt is masked. 1 = Bank Flush Error Interrupt is enabled.

797
BUSY_BANK: Busy Bank Interrupt Enabled

0 = BUSY_BANK Interrupt is masked. 1 = BUSY_BANK Interrupt is enabled.
For OUT endpoints: An interrupt is sent when all banks are busy. For IN endpoints: An interrupt is sent when all banks are free. SHRT_PCKT: Short Packet Interrupt Enabled For OUT endpoints: send an Interrupt when a Short Packet has been received. 0 = Short Packet Interrupt is masked. 1 = Short Packet Interrupt is enabled. For IN endpoints: a Short Packet transmission is guaranteed upon end of the DMA Transfer, thus signaling an end of isochronous (micro-)frame data, but only if the UDPHS_DMACONTROLx register END_B_EN and UDPHS_EPTCTLx register AUTO_VALID bits are also set.

798
32.7.15 UDPHS Endpoint Set Status Register (Control, Bulk, Interrupt Endpoints)
Name: Access:
31 23 15 7
UDPHS_EPTSETSTAx [x=0..15] Write-only

30 22 14 6 29 21 13 5 FRCESTALL 28 20 12 4 27 19 11 TXRDY 3 26 18 10 2 25 17 9 RXRDY_TXKL 1 24 16 8 0
This register view is relevant only if EPT_TYPE=0x0, 0x2 or 0x3 in UDPHS Endpoint Configuration Register on page 783 For additional Information, see UDPHS Endpoint Status Register (Control, Bulk, Interrupt Endpoints) on page 803
FRCESTALL: Stall Handshake Request Set

0 = No effect. 1 = Set this bit to request a STALL answer to the host for the next handshake Refer to chapters 8.4.5 (Handshake Packets) and 9.4.5 (Get Status) of the Universal Serial Bus Specification, Rev 2.0 for more information on the STALL handshake.
RXRDY_TXKL: KILL Bank Set (for IN Endpoint)

0 = No effect. 1 = Kill the last written bank.
TXRDY: TX Packet Ready Set

0 = No effect. 1 = Set this bit after a packet has been written into the endpoint FIFO for IN data transfers
This flag is used to generate a Data IN transaction (device to host). Device firmware checks that it can write a data payload in the FIFO, checking that TXRDY is cleared. Transfer to the FIFO is done by writing in the Buffer Address register. Once the data payload has been transferred to the FIFO, the firmware notifies the UDPHS device setting TXRDY to one. UDPHS bus transactions can start. TXCOMP is set once the data payload has been received by the host. Data should be written into the endpoint FIFO only after this bit has been cleared. Set this bit without writing data to the endpoint FIFO to send a Zero Length Packet.

799
32.7.16 UDPHS Endpoint Set Status Register (Isochronous Endpoint)

Name: Address: UDPHS_EPTSETSTAx [x=0..15] (ISOENDPT) 0xF8030114 [0], 0xF8030134 [1], 0xF8030154 [2], 0xF8030174 [3], 0xF8030194 [4], 0xF80301B4 [5], 0xF80301D4 [6], 0xF80301F4 [7], 0xF8030214 [8], 0xF8030234 [9], 0xF8030254 [10], 0xF8030274 [11], 0xF8030294 [12], 0xF80302B4 [13], 0xF80302D4 [14], 0xF80302F4 [15] Write-only
30 22 14 6 29 21 13 5 28 20 12 4 27 19 11 TXRDY_TRER 3 26 18 10 2 25 17 9 RXRDY_TXKL 1 24 16 8 0
Access:
31 23 15 7
This register view is relevant only if EPT_TYPE=0x1 in UDPHS Endpoint Configuration Register on page 783 For additional Information, see UDPHS Endpoint Status Register (Isochronous Endpoint) on page 806.
RXRDY_TXKL: KILL Bank Set (for IN Endpoint)

0 = No effect. 1 = Kill the last written bank.
TXRDY_TRER: TX Packet Ready Set

0 = No effect. 1 = Set this bit after a packet has been written into the endpoint FIFO for IN data transfers
This flag is used to generate a Data IN transaction (device to host). Device firmware checks that it can write a data payload in the FIFO, checking that TXRDY_TRER is cleared. Transfer to the FIFO is done by writing in the Buffer Address register. Once the data payload has been transferred to the FIFO, the firmware notifies the UDPHS device setting TXRDY_TRER to one. UDPHS bus transactions can start. TXCOMP is set once the data payload has been sent. Data should be written into the endpoint FIFO only after this bit has been cleared. Set this bit without writing data to the endpoint FIFO to send a Zero Length Packet.

800
32.7.17 UDPHS Endpoint Clear Status Register (Control, Bulk, Interrupt Endpoints)
Name: Access:
31 23 15 NAK_OUT 7
UDPHS_EPTCLRSTAx [x=0..15] Write-only

30 22 14 NAK_IN 6 TOGGLESQ 29 21 13 STALL_SNT 5 FRCESTALL 28 20 12 RX_SETUP 4 27 19 11 3 26 18 10 TX_COMPLT 2 25 17 9 RXRDY_TXKL 1 24 16 8 0
This register view is relevant only if EPT_TYPE=0x0, 0x2 or 0x3 in UDPHS Endpoint Configuration Register on page 783 For additional Information, see UDPHS Endpoint Status Register (Control, Bulk, Interrupt Endpoints) on page 803.
FRCESTALL: Stall Handshake Request Clear 0 = No effect. 1 = Clear the STALL request. The next packets from host will not be STALLed. TOGGLESQ: Data Toggle Clear 0 = No effect. 1 = Clear the PID data of the current bank For OUT endpoints, the next received packet should be a DATA0. For IN endpoints, the next packet will be sent with a DATA0 PID. RXRDY_TXKL: Received OUT Data Clear 0 = No effect. 1 = Clear the RXRDY_TXKL flag of UDPHS_EPTSTAx. TX_COMPLT: Transmitted IN Data Complete Clear 0 = No effect. 1 = Clear the TX_COMPLT flag of UDPHS_EPTSTAx. RX_SETUP: Received SETUP Clear 0 = No effect. 1 = Clear the RX_SETUP flags of UDPHS_EPTSTAx. STALL_SNT: Stall Sent Clear 0 = No effect. 1 = Clear the STALL_SNT flags of UDPHS_EPTSTAx. NAK_IN: NAKIN Clear 0 = No effect. 1 = Clear the NAK_IN flags of UDPHS_EPTSTAx. NAK_OUT: NAKOUT Clear 0 = No effect. 1 = Clear the NAK_OUT flag of UDPHS_EPTSTAx.

801
32.7.18 UDPHS Endpoint Clear Status Register (Isochronous Endpoint)

Name: Address: UDPHS_EPTCLRSTAx [x=0..15] (ISOENDPT) 0xF8030118 [0], 0xF8030138 [1], 0xF8030158 [2], 0xF8030178 [3], 0xF8030198 [4], 0xF80301B8 [5], 0xF80301D8 [6], 0xF80301F8 [7], 0xF8030218 [8], 0xF8030238 [9], 0xF8030258 [10], 0xF8030278 [11], 0xF8030298 [12], 0xF80302B8 [13], 0xF80302D8 [14], 0xF80302F8 [15] Write-only
30 22 14 ERR_FLUSH 29 21 13 ERR_CRC_NT R 5 28 20 12 ERR_FL_ISO 27 19 11 26 18 10 TX_COMPLT 25 17 9 RXRDY_TXKL 24 16 8
Access:
31 23 15
6 TOGGLESQ
This register view is relevant only if EPT_TYPE=0x1 in UDPHS Endpoint Configuration Register on page 783 For additional Information, see UDPHS Endpoint Status Register (Isochronous Endpoint) on page 806.
TOGGLESQ: Data Toggle Clear

0 = No effect. 1 = Clear the PID data of the current bank For OUT endpoints, the next received packet should be a DATA0. For IN endpoints, the next packet will be sent with a DATA0 PID.
RXRDY_TXKL: Received OUT Data Clear

0 = No effect. 1 = Clear the RXRDY_TXKL flag of UDPHS_EPTSTAx.
TX_COMPLT: Transmitted IN Data Complete Clear

0 = No effect. 1 = Clear the TX_COMPLT flag of UDPHS_EPTSTAx.
ERR_FL_ISO: Error Flow Clear

0 = No effect. 1 = Clear the ERR_FL_ISO flags of UDPHS_EPTSTAx.
ERR_CRC_NTR: Number of Transaction Error Clear

0 = No effect. 1 = Clear the ERR_CRC_NTR flags of UDPHS_EPTSTAx.
ERR_FLUSH: Bank Flush Error Clear

0 = No effect. 1 = Clear the ERR_FLUSH flags of UDPHS_EPTSTAx.

802
32.7.19 UDPHS Endpoint Status Register (Control, Bulk, Interrupt Endpoints)

Name: Access:
31 SHRT_PCKT 23
UDPHS_EPTSTAx [x=0..15] Read-only

30 29 28 27 BYTE_COUNT 26 25 24
22 21 BYTE_COUNT 14 NAK_IN 13 STALL_SNT 5 FRCESTALL
20
19 18 BUSY_BANK_STA 11 TXRDY 3 10 TX_COMPLT 2
17 16 CURBK_CTLDIR 9 RXRDY_TXKL 1 8 ERR_OVFLW 0
15 NAK_OUT
12 RX_SETUP 4
7 6 TOGGLESQ_STA
This register view is relevant only if EPT_TYPE=0x0, 0x2 or 0x3 in UDPHS Endpoint Configuration Register on page 783
FRCESTALL: Stall Handshake Request

0 = No effect. 1 = If set a STALL answer will be done to the host for the next handshake. This bit is reset by hardware upon received SETUP.
TOGGLESQ_STA: Toggle Sequencing

Toggle Sequencing:
IN endpoint: It indicates the PID Data Toggle that will be used for the next packet sent. This is not relative to the current bank. CONTROL and OUT endpoint: These bits are set by hardware to indicate the PID data of the current bank:
Value 0 1 2 3 Name DATA0 DATA1 DATA2 MDATA Description DATA0 DATA1 Reserved for High Bandwidth Isochronous Endpoint Reserved for High Bandwidth Isochronous Endpoint
Notes: 1. In OUT transfer, the Toggle information is meaningful only when the current bank is busy (Received OUT Data = 1). 2. These bits are updated for OUT transfer: - A new data has been written into the current bank. - The user has just cleared the Received OUT Data bit to switch to the next bank. 3. This field is reset to DATA1 by the UDPHS_EPTCLRSTAx register TOGGLESQ bit, and by UDPHS_EPTCTLDISx (disable endpoint).
ERR_OVFLW: Overflow Error

This bit is set by hardware when a new too-long packet is received. Example: If the user programs an endpoint 64 bytes wide and the host sends 128 bytes in an OUT transfer, then the Overflow Error bit is set. This bit is updated at the same time as the BYTE_COUNT field. This bit is reset by UDPHS_EPTRST register EPT_x (reset endpoint) and by UDPHS_EPTCTLDISx (disable endpoint).

803
RXRDY_TXKL: Received OUT Data/KILL Bank Received OUT Data (for OUT endpoint or Control endpoint):
This bit is set by hardware after a new packet has been stored in the endpoint FIFO. This bit is cleared by the device firmware after reading the OUT data from the endpoint. For multi-bank endpoints, this bit may remain active even when cleared by the device firmware, this if an other packet has been received meanwhile. Hardware assertion of this bit may generate an interrupt if enabled by the UDPHS_EPTCTLx register RXRDY_TXKL bit. This bit is reset by UDPHS_EPTRST register EPT_x (reset endpoint) and by UDPHS_EPTCTLDISx (disable endpoint).
KILL Bank (for IN endpoint): The bank is really cleared or the bank is sent, BUSY_BANK_STA is decremented. The bank is not cleared but sent on the IN transfer, TX_COMPLT The bank is not cleared because it was empty. The user should wait that this bit is cleared before trying to clear another packet.
Note: Kill a packet may be refused if at the same time, an IN token is coming and the current packet is sent on the UDPHS line. In this case, the TX_COMPLT bit is set. Take notice however, that if at least two banks are ready to be sent, there is no problem to kill a packet even if an IN token is coming. In fact, in that case, the current bank is sent (IN transfer) and the last bank is killed.
TX_COMPLT: Transmitted IN Data Complete

This bit is set by hardware after an IN packet has been accepted (ACKed) by the host. This bit is reset by UDPHS_EPTRST register EPT_x (reset endpoint), and by UDPHS_EPTCTLDISx (disable endpoint).
TXRDY: TX Packet Ready

This bit is cleared by hardware after the host has acknowledged the packet. For Multi-bank endpoints, this bit may remain clear even after software is set if another bank is available to transmit. Hardware clear of this bit may generate an interrupt if enabled by the UDPHS_EPTCTLx register TXRDY bit. This bit is reset by UDPHS_EPTRST register EPT_x (reset endpoint), and by UDPHS_EPTCTLDISx (disable endpoint).
RX_SETUP: Received SETUP (for Control endpoint only)

This bit is set by hardware when a valid SETUP packet has been received from the host. It is cleared by the device firmware after reading the SETUP data from the endpoint FIFO. This bit is reset by UDPHS_EPTRST register EPT_x (reset endpoint), and by UDPHS_EPTCTLDISx (disable endpoint).
STALL_SNT: Stall Sent (for Control, Bulk and Interrupt endpoints)

This bit is set by hardware after a STALL handshake has been sent as requested by the UDPHS_EPTSTAx register FRCESTALL bit. This bit is reset by UDPHS_EPTRST register EPT_x (reset endpoint) and by UDPHS_EPTCTLDISx (disable endpoint).
NAK_IN: NAK IN
This bit is set by hardware when a NAK handshake has been sent in response to an IN request from the Host. This bit is cleared by software.

804
NAK_OUT: NAK OUT

This bit is set by hardware when a NAK handshake has been sent in response to an OUT or PING request from the Host. This bit is reset by UDPHS_EPTRST register EPT_x (reset endpoint) and by EPT_CTL_DISx (disable endpoint).
CURBK_CTLDIR: Current Bank/Control Direction Current Bank (not relevant for Control endpoint):
These bits are set by hardware to indicate the number of the current bank.
Value 0 1 2 Name BANK0 BANK1 BANK2 Description Bank 0 (or single bank) Bank 1 Bank 2
Note:
The current bank is updated each time the user: - Sets the TX Packet Ready bit to prepare the next IN transfer and to switch to the next bank. - Clears the received OUT data bit to access the next bank.
This bit is reset by UDPHS_EPTRST register EPT_x (reset endpoint) and by UDPHS_EPTCTLDISx (disable endpoint).
Control Direction (for Control endpoint only):

0 = A Control Write is requested by the Host. 1 = A Control Read is requested by the Host. Notes: 1. This bit corresponds with the 7th bit of the bmRequestType (Byte 0 of the Setup Data). 2. This bit is updated after receiving new setup data.
BUSY_BANK_STA: Busy Bank Number

These bits are set by hardware to indicate the number of busy banks.
IN endpoint: It indicates the number of busy banks filled by the user, ready for IN transfer. OUT endpoint: It indicates the number of busy banks filled by OUT transaction from the Host.
Value 0 1 2 Name 1BUSYBANK 2BUSYBANKS 3BUSYBANKS Description 1 busy bank 2 busy banks 3 busy banks
BYTE_COUNT: UDPHS Byte Count

Byte count of a received data packet. This field is incremented after each write into the endpoint (to prepare an IN transfer). This field is decremented after each reading into the endpoint (OUT transfer). This field is also updated at RXRDY_TXKL flag clear with the next bank. This field is also updated at TXRDY flag set with the next bank. This field is reset by EPT_x of UDPHS_EPTRST register.
SHRT_PCKT: Short Packet

An OUT Short Packet is detected when the receive byte count is less than the configured UDPHS_EPTCFGx register EPT_Size. This bit is updated at the same time as the BYTE_COUNT field. This bit is reset by UDPHS_EPTRST register EPT_x (reset endpoint) and by UDPHS_EPTCTLDISx (disable endpoint).

805
32.7.20 UDPHS Endpoint Status Register (Isochronous Endpoint)

Name: Address: UDPHS_EPTSTAx [x=0..15] (ISOENDPT) 0xF803011C [0], 0xF803013C [1], 0xF803015C [2], 0xF803017C [3], 0xF803019C [4], 0xF80301BC [5], 0xF80301DC [6], 0xF80301FC [7], 0xF803021C [8], 0xF803023C [9], 0xF803025C [10], 0xF803027C [11], 0xF803029C [12], 0xF80302BC [13], 0xF80302DC [14], 0xF80302FC [15] Read-only
30 29 28 27 BYTE_COUNT 26 25 24
Access:
31 SHRT_PCKT 23
22 21 BYTE_COUNT 14 ERR_FLUSH 13 ERR_CRC_NT R 5
20
19 18 BUSY_BANK_STA 11 TXRDY_TRER 10 TX_COMPLT
17 CURBK 9 RXRDY_TXKL
16
15
12 ERR_FL_ISO
8 ERR_OVFLW
7 6 TOGGLESQ_STA
This register view is relevant only if EPT_TYPE=0x1 in UDPHS Endpoint Configuration Register on page 783
TOGGLESQ_STA: Toggle Sequencing

Toggle Sequencing:
IN endpoint: It indicates the PID Data Toggle that will be used for the next packet sent. This is not relative to the current bank. OUT endpoint: These bits are set by hardware to indicate the PID data of the current bank:
Value 0 1 2 3 Name DATA0 DATA1 DATA2 MDATA Description DATA0 DATA1 Data2 (only for High Bandwidth Isochronous Endpoint) MData (only for High Bandwidth Isochronous Endpoint)
Notes: 1. In OUT transfer, the Toggle information is meaningful only when the current bank is busy (Received OUT Data = 1). 2. These bits are updated for OUT transfer: - A new data has been written into the current bank. - The user has just cleared the Received OUT Data bit to switch to the next bank. 3. For High Bandwidth Isochronous Out endpoint, it is recommended to check the UDPHS_EPTSTAx/TXRDY_TRER bit to know if the toggle sequencing is correct or not. 4. This field is reset to DATA1 by the UDPHS_EPTCLRSTAx register TOGGLESQ bit, and by UDPHS_EPTCTLDISx (disable endpoint).
ERR_OVFLW: Overflow Error

This bit is set by hardware when a new too-long packet is received. Example: If the user programs an endpoint 64 bytes wide and the host sends 128 bytes in an OUT transfer, then the Overflow Error bit is set. This bit is updated at the same time as the BYTE_COUNT field. This bit is reset by UDPHS_EPTRST register EPT_x (reset endpoint) and by UDPHS_EPTCTLDISx (disable endpoint).

806
RXRDY_TXKL: Received OUT Data/KILL Bank Received OUT Data (for OUT endpoint or Control endpoint):
This bit is set by hardware after a new packet has been stored in the endpoint FIFO. This bit is cleared by the device firmware after reading the OUT data from the endpoint. For multi-bank endpoints, this bit may remain active even when cleared by the device firmware, this if an other packet has been received meanwhile. Hardware assertion of this bit may generate an interrupt if enabled by the UDPHS_EPTCTLx register RXRDY_TXKL bit. This bit is reset by UDPHS_EPTRST register EPT_x (reset endpoint) and by UDPHS_EPTCTLDISx (disable endpoint).
KILL Bank (for IN endpoint): The bank is really cleared or the bank is sent, BUSY_BANK_STA is decremented. The bank is not cleared but sent on the IN transfer, TX_COMPLT The bank is not cleared because it was empty. The user should wait that this bit is cleared before trying to clear another packet.
Note: Kill a packet may be refused if at the same time, an IN token is coming and the current packet is sent on the UDPHS line. In this case, the TX_COMPLT bit is set. Take notice however, that if at least two banks are ready to be sent, there is no problem to kill a packet even if an IN token is coming. In fact, in that case, the current bank is sent (IN transfer) and the last bank is killed.
TX_COMPLT: Transmitted IN Data Complete

This bit is set by hardware after an IN packet has been sent. This bit is reset by UDPHS_EPTRST register EPT_x (reset endpoint), and by UDPHS_EPTCTLDISx (disable endpoint).
TXRDY_TRER: TX Packet Ready/Transaction Error TX Packet Ready:

This bit is cleared by hardware, as soon as the packet has been sent. For Multi-bank endpoints, this bit may remain clear even after software is set if another bank is available to transmit. Hardware clear of this bit may generate an interrupt if enabled by the UDPHS_EPTCTLx register TXRDY_TRER bit. This bit is reset by UDPHS_EPTRST register EPT_x (reset endpoint), and by UDPHS_EPTCTLDISx (disable endpoint).
Transaction Error (for high bandwidth isochronous OUT endpoints) (Read-Only):

This bit is set by hardware when a transaction error occurs inside one microframe. If one toggle sequencing problem occurs among the n-transactions (n = 1, 2 or 3) inside a microframe, then this bit is still set as long as the current bank contains one bad n-transaction. (see CURBK: Current Bank on page 808) As soon as the current bank is relative to a new good n-transactions, then this bit is reset. Notes: 1. A transaction error occurs when the toggle sequencing does not respect the Universal Serial Bus Specification, Rev 2.0 (5.9.2 High Bandwidth Isochronous endpoints) (Bad PID, missing data....) 2. When a transaction error occurs, the user may empty all the bad transactions by clearing the Received OUT Data flag (RXRDY_TXKL). If this bit is reset, then the user should consider that a new n-transaction is coming. This bit is reset by UDPHS_EPTRST register EPT_x (reset endpoint), and by UDPHS_EPTCTLDISx (disable endpoint).
ERR_FL_ISO: Error Flow

This bit is set by hardware when a transaction error occurs.
Isochronous IN transaction is missed, the micro has no time to fill the endpoint (underflow). Isochronous OUT data is dropped because the bank is busy (overflow).

807
ERR_CRC_NTR: CRC ISO Error/Number of Transaction Error CRC ISO Error (for Isochronous OUT endpoints) (Read-only):
This bit is set by hardware if the last received data is corrupted (CRC error on data). This bit is updated by hardware when new data is received (Received OUT Data bit).
Number of Transaction Error (for High Bandwidth Isochronous IN endpoints):

This bit is set at the end of a microframe in which at least one data bank has been transmitted, if less than the number of transactions per micro-frame banks (UDPHS_EPTCFGx register NB_TRANS) have been validated for transmission inside this microframe. This bit is reset by UDPHS_EPTRST register EPT_x (reset endpoint) and by UDPHS_EPTCTLDISx (disable endpoint).
ERR_FLUSH: Bank Flush Error (for High Bandwidth Isochronous IN endpoints)

This bit is set when flushing unsent banks at the end of a microframe. This bit is reset by UDPHS_EPTRST register EPT_x (reset endpoint) and by EPT_CTL_DISx (disable endpoint).
CURBK: Current Bank Current Bank:

These bits are set by hardware to indicate the number of the current bank.
Value 0 1 2 Name BANK0 BANK1 BANK2 Description Bank 0 (or single bank) Bank 1 Bank 2
Note:
The current bank is updated each time the user: - Sets the TX Packet Ready bit to prepare the next IN transfer and to switch to the next bank. - Clears the received OUT data bit to access the next bank.
BUSY_BANK_STA: Busy Bank Number

These bits are set by hardware to indicate the number of busy banks.
IN endpoint: It indicates the number of busy banks filled by the user, ready for IN transfer. OUT endpoint: It indicates the number of busy banks filled by OUT transaction from the Host.
Value 0 1 2 Name 1BUSYBANK 2BUSYBANKS 3BUSYBANKS Description 1 busy bank 2 busy banks 3 busy banks
BYTE_COUNT: UDPHS Byte Count

Byte count of a received data packet. This field is incremented after each write into the endpoint (to prepare an IN transfer). This field is decremented after each reading into the endpoint (OUT transfer). This field is also updated at RXRDY_TXKL flag clear with the next bank. This field is also updated at TXRDY_TRER flag set with the next bank. This field is reset by EPT_x of UDPHS_EPTRST register.

808
SHRT_PCKT: Short Packet

An OUT Short Packet is detected when the receive byte count is less than the configured UDPHS_EPTCFGx register EPT_Size. This bit is updated at the same time as the BYTE_COUNT field. This bit is reset by UDPHS_EPTRST register EPT_x (reset endpoint) and by UDPHS_EPTCTLDISx (disable endpoint).

809
32.7.21 UDPHS DMA Channel Transfer Descriptor

The DMA channel transfer descriptor is loaded from the memory. Be careful with the alignment of this buffer. The structure of the DMA channel transfer descriptor is defined by three parameters as described below: Offset 0: The address must be aligned: 0xXXXX0 Next Descriptor Address Register: UDPHS_DMANXTDSCx Offset 4: The address must be aligned: 0xXXXX4 DMA Channelx Address Register: UDPHS_DMAADDRESSx Offset 8: The address must be aligned: 0xXXXX8 DMA Channelx Control Register: UDPHS_DMACONTROLx To use the DMA channel transfer descriptor, fill the structures with the correct value (as described in the following pages). Then write directly in UDPHS_DMANXTDSCx the address of the descriptor to be used first. Then write 1 in the LDNXT_DSC bit of UDPHS_DMACONTROLx (load next channel transfer descriptor). The descriptor is automatically loaded upon Endpointx request for packet transfer.

810
32.7.22 UDPHS DMA Next Descriptor Address Register

31
UDPHS_DMANXTDSCx [x = 0..6] 0xF8030300 [0], 0xF8030310 [1], 0xF8030320 [2], 0xF8030330 [3], 0xF8030340 [4], 0xF8030350 [5], 0xF8030360 [6] Read-write
30 29 28 27 NXT_DSC_ADD 20 19 NXT_DSC_ADD 12 11 NXT_DSC_ADD 4 3 NXT_DSC_ADD 26 25 24
23
22
21
18
17
16
15
14
13
10
Note:
Channel 0 is not used.
NXT_DSC_ADD: Next Descriptor Address

This field points to the next channel descriptor to be processed. This channel descriptor must be aligned, so bits 0 to 3 of the address must be equal to zero.

811
32.7.23 UDPHS DMA Channel Address Register

31
UDPHS_DMAADDRESSx [x = 0..6] 0xF8030304 [0], 0xF8030314 [1], 0xF8030324 [2], 0xF8030334 [3], 0xF8030344 [4], 0xF8030354 [5], 0xF8030364 [6] Read-write
30 29 28 BUFF_ADD 27 26 25 24
23
22
21
20 BUFF_ADD
19
18
17
16
15
14
13
12 BUFF_ADD
11
10
4 BUFF_ADD
Note:
BUFF_ADD: Buffer Address

This field determines the AHB bus starting address of a DMA channel transfer. Channel start and end addresses may be aligned on any byte boundary. The firmware may write this field only when the UDPHS_DMASTATUS register CHANN_ENB bit is clear. This field is updated at the end of the address phase of the current access to the AHB bus. It is incrementing of the access byte width. The access width is 4 bytes (or less) at packet start or end, if the start or end address is not aligned on a word boundary. The packet start address is either the channel start address or the next channel address to be accessed in the channel buffer. The packet end address is either the channel end address or the latest channel address accessed in the channel buffer. The channel start address is written by software or loaded from the descriptor, whereas the channel end address is either determined by the end of buffer or the UDPHS device, USB end of transfer if the UDPHS_DMACONTROLx register END_TR_EN bit is set.

812
32.7.24 UDPHS DMA Channel Control Register

31
UDPHS_DMACONTROLx [x = 0..6] 0xF8030308 [0], 0xF8030318 [1], 0xF8030328 [2], 0xF8030338 [3], 0xF8030348 [4], 0xF8030358 [5], 0xF8030368 [6] Read-write
30 29 28 27 BUFF_LENGTH 20 19 BUFF_LENGTH 12 4 END_TR_IT 11 3 END_B_EN 26 25 24
23
22
21
18
17
16
15 7 BURST_LCK
14 6 DESC_LD_IT
13 5 END_BUFFIT
10 2 END_TR_EN
9 1 LDNXT_DSC
8 0 CHANN_ENB
Note:
CHANN_ENB: (Channel Enable Command)

0 = DMA channel is disabled at and no transfer will occur upon request. This bit is also cleared by hardware when the channel source bus is disabled at end of buffer. If the UDPHS_DMACONTROL register LDNXT_DSC bit has been cleared by descriptor loading, the firmware will have to set the corresponding CHANN_ENB bit to start the described transfer, if needed. If the UDPHS_DMACONTROL register LDNXT_DSC bit is cleared, the channel is frozen and the channel registers may then be read and/or written reliably as soon as both UDPHS_DMASTATUS register CHANN_ENB and CHANN_ACT flags read as 0. If a channel request is currently serviced when this bit is cleared, the DMA FIFO buffer is drained until it is empty, then the UDPHS_DMASTATUS register CHANN_ENB bit is cleared. If the LDNXT_DSC bit is set at or after this bit clearing, then the currently loaded descriptor is skipped (no data transfer occurs) and the next descriptor is immediately loaded. 1 = UDPHS_DMASTATUS register CHANN_ENB bit will be set, thus enabling DMA channel data transfer. Then any pending request will start the transfer. This may be used to start or resume any requested transfer.
LDNXT_DSC: Load Next Channel Transfer Descriptor Enable (Command)

0 = No channel register is loaded after the end of the channel transfer. 1 = The channel controller loads the next descriptor after the end of the current transfer, i.e. when the UDPHS_DMASTATUS/CHANN_ENB bit is reset. If the UDPHS_DMA CONTROL/CHANN_ENB bit is cleared, the next descriptor is immediately loaded upon transfer request. DMA Channel Control Command Summary
LDNXT_DSC 0 0 1 1 CHANN_ENB 0 1 0 1 Description Stop now Run and stop at end of buffer Load next descriptor now Run and link at end of buffer

813
END_TR_EN: End of Transfer Enable (Control)

Used for OUT transfers only. 0 = USB end of transfer is ignored. 1 = UDPHS device can put an end to the current buffer transfer. When set, a BULK or INTERRUPT short packet or the last packet of an ISOCHRONOUS (micro) frame (DATAX) will close the current buffer and the UDPHS_DMASTATUSx register END_TR_ST flag will be raised. This is intended for UDPHS non-prenegotiated end of transfer (BULK or INTERRUPT) or ISOCHRONOUS microframe data buffer closure.
END_B_EN: End of Buffer Enable (Control)

0 = DMA Buffer End has no impact on USB packet transfer. 1 = Endpoint can validate the packet (according to the values programmed in the UDPHS_EPTCTLx register AUTO_VALID and SHRT_PCKT fields) at DMA Buffer End, i.e. when the UDPHS_DMASTATUS register BUFF_COUNT reaches 0. This is mainly for short packet IN validation initiated by the DMA reaching end of buffer, but could be used for OUT packet truncation (discarding of unwanted packet data) at the end of DMA buffer.
END_TR_IT: End of Transfer Interrupt Enable

0 = UDPHS device initiated buffer transfer completion will not trigger any interrupt at UDPHS_STATUSx/END_TR_ST rising. 1 = An interrupt is sent after the buffer transfer is complete, if the UDPHS device has ended the buffer transfer. Use when the receive size is unknown.
END_BUFFIT: End of Buffer Interrupt Enable

0 = UDPHS_DMA_STATUSx/END_BF_ST rising will not trigger any interrupt. 1 = An interrupt is generated when the UDPHS_DMASTATUSx register BUFF_COUNT reaches zero.
DESC_LD_IT: Descriptor Loaded Interrupt Enable

0 = UDPHS_DMASTATUSx/DESC_LDST rising will not trigger any interrupt. 1 = An interrupt is generated when a descriptor has been loaded from the bus.
BURST_LCK: Burst Lock Enable

0 = The DMA never locks bus access. 1 = USB packets AHB data bursts are locked for maximum optimization of the bus bandwidth usage and maximization of fly-by AHB burst duration.
BUFF_LENGTH: Buffer Byte Length (Write-only)

This field determines the number of bytes to be transferred until end of buffer. The maximum channel transfer size (64 KBytes) is reached when this field is 0 (default value). If the transfer size is unknown, this field should be set to 0, but the transfer end may occur earlier under UDPHS device control. When this field is written, The UDPHS_DMASTATUSx register BUFF_COUNT field is updated with the write value. Notes: 1. Bits [31:2] are only writable when issuing a channel Control Command other than Stop Now. 2. For reliability it is highly recommended to wait for both UDPHS_DMASTATUSx register CHAN_ACT and CHAN_ENB flags are at 0, thus ensuring the channel has been stopped before issuing a command other than Stop Now.

814
32.7.25 UDPHS DMA Channel Status Register

31
UDPHS_DMASTATUSx [x = 0..6] 0xF803030C [0], 0xF803031C [1], 0xF803032C [2], 0xF803033C [3], 0xF803034C [4], 0xF803035C [5], 0xF803036C [6] Read-write
30 29 28 27 BUFF_COUNT 20 19 BUFF_COUNT 12 4 END_TR_ST 11 3 26 25 24
23
22
21
18
17
16
15 7
14 6 DESC_LDST
13 5 END_BF_ST
10 2
9 1 CHANN_ACT
8 0 CHANN_ENB
Note:
CHANN_ENB: Channel Enable Status

0 = If cleared, the DMA channel no longer transfers data, and may load the next descriptor if the UDPHS_DMACONTROLx register LDNXT_DSC bit is set. When any transfer is ended either due to an elapsed byte count or a UDPHS device initiated transfer end, this bit is automatically reset. 1 = If set, the DMA channel is currently enabled and transfers data upon request. This bit is normally set or cleared by writing into the UDPHS_DMACONTROLx register CHANN_ENB bit field either by software or descriptor loading. If a channel request is currently serviced when the UDPHS_DMACONTROLx register CHANN_ENB bit is cleared, the DMA FIFO buffer is drained until it is empty, then this status bit is cleared.
CHANN_ACT: Channel Active Status

0 = The DMA channel is no longer trying to source the packet data. When a packet transfer is ended this bit is automatically reset. 1 = The DMA channel is currently trying to source packet data, i.e. selected as the highest-priority requesting channel. When a packet transfer cannot be completed due to an END_BF_ST, this flag stays set during the next channel descriptor load (if any) and potentially until UDPHS packet transfer completion, if allowed by the new descriptor.
END_TR_ST: End of Channel Transfer Status

0 = Cleared automatically when read by software. 1 = Set by hardware when the last packet transfer is complete, if the UDPHS device has ended the transfer. Valid until the CHANN_ENB flag is cleared at the end of the next buffer transfer.
END_BF_ST: End of Channel Buffer Status

0 = Cleared automatically when read by software. 1 = Set by hardware when the BUFF_COUNT downcount reach zero. Valid until the CHANN_ENB flag is cleared at the end of the next buffer transfer.

815
DESC_LDST: Descriptor Loaded Status

0 = Cleared automatically when read by software. 1 = Set by hardware when a descriptor has been loaded from the system bus. Valid until the CHANN_ENB flag is cleared at the end of the next buffer transfer.
BUFF_COUNT: Buffer Byte Count

This field determines the current number of bytes still to be transferred for this buffer. This field is decremented from the AHB source bus access byte width at the end of this bus address phase. The access byte width is 4 by default, or less, at DMA start or end, if the start or end address is not aligned on a word boundary. At the end of buffer, the DMA accesses the UDPHS device only for the number of bytes needed to complete it. This field value is reliable (stable) only if the channel has been stopped or frozen (UDPHS_EPTCTLx register NT_DIS_DMA bit is used to disable the channel request) and the channel is no longer active CHANN_ACT flag is 0. Note: For OUT endpoints, if the receive buffer byte length (BUFF_LENGTH) has been defaulted to zero because the USB transfer length is unknown, the actual buffer byte length received will be 0x10000-BUFF_COUNT.

816
33.
33.1
USB High Speed Host Port (UHPHS)

Description
The USB High Speed Host Port (UHPHS) interfaces the USB with the host application. It handles Open HCI protocol (Open Host Controller Interface) as well as Enhanced HCI protocol (Enhanced Host Controller Interface).
33.2
z
Compliant with Enhanced HCI Rev 1.0 Specification

z z
Compliant with USB V2.0 High-speed Supports High-speed 480 Mbps Compliant with USB V2.0 Full-speed and Low-speed Specification Supports both Low-speed 1.5 Mbps and Full-speed 12 Mbps USB devices
Compliant with OpenHCI Rev 1.0 Specification

z z
z z z z
Root Hub Integrated with 3 Downstream USB HS Ports Embedded USB Transceivers Supports Power Management 3 Hosts (A, B and C) High Speed (EHCI), Port A shared with UDPHS

817
33.3
Block Diagram
HCI Slave Block AHB Slave OHCI Registers Control
List Processor Block ED & TD Registers Root Hub and Host SIE
Root Hub Registers
PORT S/M 1
Embedded USB v2.0 Transceiver USB High-speed Transceiver HFSDPC HFSDMC HHSDPC HHSDMC HFSDPB HFSDMB HHSDPB HHSDMB HFSDPA HFSDMA HHSDPA HHSDMA
Master AHB
HCI Master Block
Data
PORT S/M 0 FIFO 64 x 8
USB High-speed Transceiver
SOF generator HCI Slave Block Slave AHB EHCI Registers Control Packet Buffer FIFO
USB High-speed Transceiver
List Processor Master AHB HCI Master Block Data
Access to the USB host operational registers is achieved through the AHB bus slave interface. The Open HCI host controller and Enhanced HCI host controller initialize master DMA transfers through the AHB bus master interface as follows:
z z z z
Fetches endpoint descriptors and transfer descriptors Access to endpoint data from system memory Access to the HC communication area Write status and retire transfer descriptor
Memory access errors (abort, misalignment) lead to an Unrecoverable Error indicated by the corresponding flag in the host controller operational registers. The USB root hub is integrated in the USB host. Several USB downstream ports are available. The number of downstream ports can be determined by the software driver reading the root hubs operational registers. Device connection is automatically detected by the USB host port logic. USB physical transceivers are integrated in the product and driven by the root hubs ports. Over current protection on ports can be activated by the USB host controller. Atmels standard product does not dedicate pads to external over current protection.

818
33.4
Typical Connection
Figure 33-2. Board Schematic to Interface UHP High-speed Host Controller
PIO (VBUS ENABLE +5V "A" Receptacle 1 = VBUS 2 = D3 = D+ 4 = GND

3 4
HHSDM/HFSDM
Shell = Shield
1 2
HHSDP/HFSDP 6K8 1% VBG 10 pF GNDUTMI

819
33.5
33.5.1 I/O Lines

HFSDPs, HFSDMs, HHSDPs and HHSDMs are not controlled by any PIO controllers. The embedded USB High Speed physical transceivers are controlled by the USB host controller. One transceiver is shared with the USB High Speed Device (port A). The selection between Host Port A and USB Device is controlled by the UDPHS enable bit (EN_UDPHS) located in the UDPHS_CTRL control register. In the case the port A is driven by the USB High Speed Device, the output signals are DFSDP, DFSDM, DHSDP and DHSDM. The transceiver is automatically selected for Device operation once the USB High Speed Device is enabled. In the case the port A is driven by the USB High Speed Host, the output signals are HFSDPA, HFSDMA, HHSDPA and HHSDMA.

The system embeds 3 transceivers. The USB Host High Speed requires a 480 MHz clock for the embedded High-speed transceivers. This clock is provided by the UTMI PLL, it is UPLLCK. In case power consumption is saved by stopping the UTMI PLL, high-speed operations are not possible. Nevertheless, OHCI Full-speed operations remain possible by selecting PLLACK as the input clock of OHCI. The High-speed transceiver returns a 30 MHz clock to the USB Host controller. The USB Host controller requires 48 MHz and 12 MHz clocks for OHCI full-speed operations. These clocks must be generated by a PLL with a correct accuracy of 0.25% thanks to USBDIV field. Thus the USB Host peripheral receives three clocks from the Power Management Controller (PMC): the Peripheral Clock (MCK domain), the UHP48M and the UHP12M (built-in UHP48M divided by four) used by the OHCI to interface with the bus USB signals (Recovered 12 MHz domain) in Full-speed operations. For High-speed operations, the user has to perform the following:
z z z z z z z z
Enable UHP peripheral clock, bit (1 << AT91C_ID_UHPHS) in PMC_PCER register. Write CKGR_PLLCOUNT field in PMC_UCKR register. Enable UPLL, bit AT91C_CKGR_UPLLEN in PMC_UCKR register. Wait until UTMI_PLL is locked. LOCKU bit in PMC_SR register Enable BIAS, bit AT91C_CKGR_BIASEN in PMC_UCKR register. Select UPLLCK as Input clock of OHCI part, USBS bit in PMC_USB register. Program the OHCI clocks (UHP48M and UHP12M) with USBDIV field in PMC_USB register. USBDIV must be 9 (division by 10) if UPLLCK is selected. Enable OHCI clocks, UHP bit in PMC_SCER register. Enable UHP peripheral clock, bit (1 << AT91C_ID_UHPHS) in PMC_PCER register. Select PLLACK as Input clock of OHCI part, USBS bit in PMC_USB register. Program the OHCI clocks (UHP48M and UHP12M) with USBDIV field in PMC_USB register. USBDIV value is to calculated regarding the PLLACK value and USB Full-speed accuracy. Enable the OHCI clocks, UHP bit in PMC_SCER register.
For OHCI Full-speed operations only, the user has to perform the following:
z z z z

820
Figure 33-3. UHP Clock Trees

UPLL (480 MHz)
AHB
EHCI Master Interface EHCI User Interface
30 MHz
UTMI transceiver
USB 2.0 EHCI Host Controller
Port Router
30 MHz 30 MHz
UTMI transceiver UTMI transceiver
MCK OHCI Master Interface OHCI User Interface USB 1.1 OHCI Host Controller
Root Hub and Host SIE
UHP12M
OHCI clocks
33.5.3 Interrupt
The USB host interface has an interrupt line connected to the Advanced Interrupt Controller (AIC). Handling USB host interrupts requires programming the AIC before configuring the UHP HS.
UHP48M

821
33.6
33.6.1 UTMI transceivers Sharing

The High Speed USB Host Port A is shared with the High Speed USB Device port and connected to the second UTMI transceiver. The selection between Host Port A and USB Device is controlled by the UDPHS enable bit (EN_UDPHS) located in the UDPHS_CTRL control register.
Figure 33-4. USB Selection
Other Transceivers HS Transceiver
EN_UDPHS 0 Other Ports HS USB Host HS EHCI FS OHCI DMA PA HS USB Device DMA 1
33.6.2 EHCI
The USB Host Port controller is fully compliant with the Enhanced HCI specification. The USB Host Port User Interface (registers description) can be found in the Enhanced HCI Rev 1.0 Specification available on http://www.intel.com/technology/usb/ehcispec.htm. The standard EHCI USB stack driver can be easily ported to Atmels architecture in the same way all existing class drivers run, without hardware specialization.
33.6.3 OHCI
The USB Host Port integrates a root hub and transceivers on downstream ports. It provides several Full-speed halfduplex serial communication ports at a baud rate of 12 Mbit/s. Up to 127 USB devices (printer, camera, mouse, keyboard, disk, etc.) and the USB hub can be connected to the USB host in the USB tiered star topology. The USB Host Port controller is fully compliant with the Open HCI specification. The USB Host Port User Interface (registers description) can be found in the Open HCI Rev 1.0 Specification available on http://h18000.www1.hp.com/productinfo/development/openhci.html. The standard OHCI USB stack driver can be easily ported to Atmels architecture, in the same way all existing class drivers run without hardware specialization. This means that all standard class devices are automatically detected and available to the users application. As an example, integrating an HID (Human Interface Device) class driver provides a plug & play feature for all USB keyboards and mouses.

822
34.
34.1
Gigabit Ethernet MAC (GMAC)

Description
The Gigabit Ethernet MAC (GMAC) module implements a 10/100/1000 Mbps Ethernet MAC compatible with the IEEE 802.3 standard. The GMAC can operate in either half or full duplex mode at all supported speeds. The Network Configuration Register is used to select the speed, duplex mode and interface type (MII, GMII or RGMII). The GMAC comprises two constituent components:
z z
GEM_MAC controlling transmit, receive, address checking and loopback GEM_TSU calculates the IEEE 1588 timer values
34.2
z z z z z z z z z z z z z z z z z z z z z z z z z
Compatible with IEEE Standard 802.3 10, 100 and 1000 Mbps operation Full and half duplex operation at all supported speeds of operation Statistics Counter Registers for RMON/MIB MII/GMII/RGMII interface to the physical layer Integrated physical coding Direct memory access (DMA) interface to external memory Support for up to eight priority queues in DMA Programmable burst length and endianism for DMA Interrupt generation to signal receive and transmit completion, or errors Automatic pad and cyclic redundancy check (CRC) generation on transmitted frames Frame extension and frame bursting at 1000 Mbps in half duplex mode Automatic discard of frames received with errors Receive and transmit IP, TCP and UDP checksum offload. Both IPv4 and IPv6 packet types supported Address checking logic for four specific 48-bit addresses, four type IDs, promiscuous mode, hash matching of unicast and multicast destination addresses and Wake-on-LAN Management Data Input/Output (MDIO) interface for physical layer management Support for jumbo frames up to 10240 bytes Full duplex flow control with recognition of incoming pause frames and hardware generation of transmitted pause frames Half duplex flow control by forcing collisions on incoming frames Support for 802.1Q VLAN tagging with recognition of incoming VLAN and priority tagged frames Support for 802.1Qbb priority-based flow control Programmable Inter Packet Gap (IPG) Stretch Recognition of IEEE 1588 PTP frames IEEE 1588 time stamp unit (TSU) Support for 802.1AS timing and synchronization

823
34.3
Block Diagram
Status & Statistic Registers APB Register Interface MDIO Control Registers
MAC Transmitter AHB AHB DMA Interface FIFO Interface Media Interface MAC Receiver
Frame Filtering

824
34.4
Signal Interface
The GMAC includes the following signal interfaces
z z z z
MII, GMII and RGMII to an external PHY MDIO interface for external PHY management Slave APB interface for accessing GMAC registers Master AHB interface for memory access
Table 34-1. GMAC connections in the different modes Signal Name GTXCK G125CK G125CKO GTXEN GTX[7..0] GTXER GRXCK GRXDV GRX[7..0] GRXER GCRS GCOL GMDC GMDIO Function Transmit Clock or Reference Clock 125 MHz input Clock 125 MHz output Clock Transmit Enable Transmit Data Transmit Coding Error Receive Clock Receive Data Valid Receive Data Receive Error Carrier Sense and Data Valid Collision Detect Management Data Clock Management Data Input/Output MII TXCK Not Used Not Used TXEN TXD[3:0] TXER RXCK RXDV RXD[3:0] RXER CRS COL MDC MDIO GMII Not Used 125 MHz Ref Clk TXCK TXEN TXD[7:0] TXER RXCK RXDV RXD[7:0] RXER CRS COL MDC MDIO RGMII TXCK 125 MHz Ref Clk Not Used TXCTL TXD[3:0] Not Used RXCK Not Used RXD[3:0] RXCTL Not Used Not Used MDC MDIO

825
34.5
34.5.1 Media Access Controller

The Media Access Controller (MAC) transmit block takes data from FIFO, adds preamble and, if necessary, pad and frame check sequence (FCS). Both half duplex and full duplex Ethernet modes of operation are supported. When operating in half duplex mode, the MAC transmit block generates data according to the carrier sense multiple access with collision detect (CSMA/CD) protocol. The start of transmission is deferred if carrier sense (CRS) is active. If collision (COL) becomes active during transmission, a jam sequence is asserted and the transmission is retried after a random back off. The CRS and COL signals have no effect in full duplex mode. When operating in gigabit mode half duplex, both carrier extension and frame bursting are performed in accordance with the IEEE 802.3 standard. The MAC receive block checks for valid preamble, FCS, alignment and length, and presents received frames to the MAC address checking block and FIFO. Software can configure the GMAC to receive jumbo frames up to 10240 bytes. It can optionally strip CRC from the received frame prior to transfer to FIFO. The address checker recognizes four specific 48-bit addresses, can recognize four different type ID values, and contains a 64-bit Hash Register for matching multicast and unicast addresses as required. It can recognize the broadcast address of all ones and copy all frames. The MAC can also reject all frames that are not VLAN tagged and recognise Wake on LAN events. The MAC receive block supports offloading of IP, TCP and UDP checksum calculations (both IPv4 and IPv6 packet types supported), and can automatically discard bad checksum frames.
34.5.2 1588 Time Stamp Unit

The 1588 time stamp unit (TSU) is a timer implemented as a 62-bit register. The 32 upper bits count seconds and the 30 lower bits count nanoseconds. The 30 lower bits roll over when they have counted to one second. The timer increments by a programmable number of nanoseconds with each MCK period and can be adjusted (incremented or decremented) through APB register accesses.
34.5.3 AHB Direct Memory Access Interface

GMAC is supplied with an AHB DMA interface. When the GMAC is configured to use the DMA, it is attached to the MAC module's FIFO to provide a scatter gather type capability for packet data storage in the embedded processor system or System on Chip. The GMAC DMA controller performs six types of operation on the AHB bus. When the GMAC DMA is configured in internal FIFO mode, in order of priority these are as follows:
z z z z
Receive buffer manager write/read Transmit buffer manager write/read Receive data DMA write Transmit data DMA read
The DMA is configured in packet buffer mode, where dual-port memories are used to buffer multiple frames. The packet buffer DMA features follow. 34.5.3.1 Packet Buffer DMA
z z z z z z
Easier to guarantee maximum line rate due to the ability to store multiple frames in the packet buffer Full store and forward, or partial store and forward programmable options (partial store will cater for shorter latency requirements) Support for Transmit TCP/IP checksum offload Support for priority queueing When a collision on the line occurs during transmission, the packet will be automatically replayed directly from the packet buffer memory rather than having to re-fetch through the AHB (full store and forward ONLY) Received errored packets are automatically dropped before any of the packet is presented to the AHB (full store and forward ONLY), thus reducing AHB activity

826
z z
Supports manual RX packet flush capabilities Optional RX packet flush when there is lack of AHB resource
34.5.3.2 Partial Store and Forward Using Packet Buffer DMA When the DMA is configured to use SRAM-based packet buffers, it can be programmed into a low latency mode, known as partial store and forward. This mode allows the integrator to attach small SRAMs to the GMAC DMA, smaller than the maximum sized packets to be transmitted/received. This allows for a reduced latency as the full packet is not buffered before forwarding. Note that this option is only available when the device is configured for full duplex operation. This feature is enabled via the TX and RX partial store and forward programmable registers. When the transmit partial store and forward mode is activated, the transmitter will only begin to forward the packet to the MAC when there is enough packet data stored in the packet buffer. Likewise, when the receive partial store and forward mode is activated, the receiver will only begin to forward the packet to the AHB when enough packet data is stored in the packet buffer. The amount of packet data required to activate the forwarding process is programmable via watermark registers which are located at the same address as the partial store and forward enable bits. Note that the minimum operational value for the TX partial store and forward watermark is 20. There is no operational limit for the RX partial store and forward watermark. Enabling partial store and forward is a useful means to reduce latency and reduce the size of the attached SRAMs, but there are performance implications. The GMAC DMA uses separate transmit and receive lists of buffer descriptors, with each descriptor describing a buffer area in memory. This allows Ethernet packets to be broken up and scattered around the AHB memory space. 34.5.3.3 Receive AHB Buffers Received frames, optionally including FCS, are written to receive AHB buffers stored in memory. The receive buffer depth is programmable in the range of 64 bytes to 16 Kbytes through the DMA Configuration Register, with the default being 128 bytes. The start location for each receive AHB buffer is stored in memory in a list of receive buffer descriptors at an address location pointed to by the receive buffer queue pointer. The base address for the receive buffer queue pointer is configured in software using the Receive Buffer Queue Base Address Register. Each list entry consists of two words. The first is the address of the receive AHB buffer and the second the receive status. If the length of a receive frame exceeds the AHB buffer length, the status word for the used buffer is written with zeroes except for the start of frame bit, which is always set for the first buffer in a frame. Bit zero of the address field is written to 1 to show the buffer has been used. The receive buffer manager then reads the location of the next receive AHB buffer and fills that with the next part of the received frame data. AHB buffers are filled until the frame is complete and the final buffer descriptor status word contains the complete frame status. Refer to Table 34-2, Receive Buffer Descriptor Entry for details of the receive buffer descriptor list. Each receive AHB buffer start location is a word address. The start of the first AHB buffer in a frame can be offset by up to three bytes, depending on the value written to bits 14 and 15 of the Network Configuration Register. For 64-bit datapaths, the start of the frame can be offset by up to a further four bytes if bit 2 of the AHB buffer start location in the buffer descriptor is set. If the start location of the AHB buffer is offset, the available length of the first AHB buffer is reduced by the corresponding number of bytes.
Table 34-2. Receive Buffer Descriptor Entry Bit Function Word 0 31:2 1 0 Address of beginning of buffer Wrap - marks last descriptor in receive buffer descriptor list. Ownership - needs to be zero for the GMAC to write data to the receive buffer. The GMAC sets this to one once it has successfully written a frame to memory. Software has to clear this bit before the buffer can be used again. Word 1 31 Global all ones broadcast address detected

827
Table 34-2. Receive Buffer Descriptor Entry (Continued) Bit 30 29 28 27 Function Multicast hash match Unicast hash match Specific Address Register match found, bit 25 and bit 26 indicate which Specific Address Register causes the match. Specific Address Register match. Encoded as follows: 00 - Specific Address Register 1 match 26:25 01 - Specific Address Register 2 match 10 - Specific Address Register 3 match 11 - Specific Address Register 4 match If more than one specific address is matched only one is indicated with priority 4 down to 1. This bit has a different meaning depending on whether RX checksum offloading is enabled. With RX checksum offloading disabled: (bit 24 clear in Network Configuration Register) Type ID register match found, bit 22 and bit 23 indicate which type ID register causes the match. 24 With RX checksum offloading enabled: (bit 24 set in Network Configuration Register) 0 - the frame was not SNAP encoded and/or had a VLAN tag with the Canonical Format Indicator (CFI) bit set. 1 - the frame was SNAP encoded and had either no VLAN tag or a VLAN tag with the CFI bit not set. This bit has a different meaning depending on whether RX checksum offloading is enabled. With RX checksum offloading disabled: (bit 24 clear in Network Configuration) Type ID register match. Encoded as follows: 00 - Type ID register 1 match 01 - Type ID register 2 match 10 -Type ID register 3 match 11 - Type ID register 4 match 23:22 If more than one Type ID is matched only one is indicated with priority 4 down to 1. With RX checksum offloading enabled: (bit 24 set in Network Configuration Register) 00 - Neither the IP header checksum nor the TCP/UDP checksum was checked. 01 - The IP header checksum was checked and was correct. Neither the TCP nor UDP checksum was checked. 10 - Both the IP header and TCP checksum were checked and were correct. 11 - Both the IP header and UDP checksum were checked and were correct. 21 VLAN tag detected - type ID of 0x8100. For packets incorporating the stacked VLAN processing feature, this bit will be set if the second VLAN tag has a type ID of 0x8100 Priority tag detected - type ID of 0x8100 and null VLAN identifier. For packets incorporating the stacked VLAN processing feature, this bit will be set if the second VLAN tag has a type ID of 0x8100 and a null VLAN identifier. VLAN priority - only valid if bit 21 is set.
20 19:17

828
Table 34-2. Receive Buffer Descriptor Entry (Continued) Bit 16 15 14 Function Canonical format indicator (CFI) bit (only valid if bit 21 is set). End of frame - when set the buffer contains the end of a frame. If end of frame is not set, then the only valid status bit is start of frame (bit 14). Start of frame - when set the buffer contains the start of a frame. If both bits 15 and 14 are set, the buffer contains a whole frame. This bit has a different meaning depending on whether jumbo frames and ignore FCS modes are enabled. If neither mode is enabled this bit will be zero. With jumbo frame mode enabled: (bit 3 set in Network Configuration Register) Additional bit for length of frame (bit[13]), that is concatenated with bits[12:0] 13 With ignore FCS mode enabled and jumbo frames disabled: (bit 26 set in Network Configuration Register and bit 3 clear in Network Configuration Register) This indicates per frame FCS status as follows: 0 - Frame had good FCS 1 - Frame had bad FCS, but was copied to memory as ignore FCS enabled. These bits represent the length of the received frame which may or may not include FCS depending on whether FCS discard mode is enabled. With FCS discard mode disabled: (bit 17 clear in Network Configuration Register) 12:0 Least significant 12 bits for length of frame including FCS. If jumbo frames are enabled, these 12 bits are concatenated with bit[13] of the descriptor above. With FCS discard mode enabled: (bit 17 set in Network Configuration Register) Least significant 12 bits for length of frame excluding FCS. If jumbo frames are enabled, these 12 bits are concatenated with bit[13] of the descriptor above.
To receive frames, the AHB buffer descriptors must be initialized by writing an appropriate address to bits 31:2 in the first word of each list entry. Bit 0 must be written with zero. Bit 1 is the wrap bit and indicates the last entry in the buffer descriptor list. The start location of the receive buffer descriptor list must be written with the receive buffer queue base address before reception is enabled (receive enable in the Network Control Register). Once reception is enabled, any writes to the Receive Buffer Queue Base Address Register are ignored. When read, it will return the current pointer position in the descriptor list, though this is only valid and stable when receive is disabled. If the filter block indicates that a frame should be copied to memory, the receive data DMA operation starts writing data into the receive buffer. If an error occurs, the buffer is recovered. An internal counter within the GMAC represents the receive buffer queue pointer and it is not visible through the CPU interface. The receive buffer queue pointer increments by two words after each buffer has been used. It re-initializes to the receive buffer queue base address if any descriptor has its wrap bit set. As receive AHB buffers are used, the receive AHB buffer manager sets bit zero of the first word of the descriptor to logic one indicating the AHB buffer has been used. Software should search through the used bits in the AHB buffer descriptors to find out how many frames have been received, checking the start of frame and end of frame bits. When the DMA is configured in the packet buffer partial store and forward mode, received frames are written out to the AHB buffers as soon as enough frame data exists in the packet buffer. For both cases, this may mean several full AHB buffers are used before some error conditions can be detected. If a receive error is detected the receive buffer currently being written will be recovered. Previous buffers will not be recovered. As an example, when receiving frames with cyclic redundancy check (CRC) errors or excessive length, it is possible that a frame fragment might be stored in a sequence of

829
AHB receive buffers. Software can detect this by looking for start of frame bit set in a buffer following a buffer with no end of frame bit set. For a properly working 10/100/1000 Ethernet system there should be no excessive length frames or frames greater than 128 bytes with CRC errors. Collision fragments will be less than 128 bytes long, therefore it will be a rare occurrence to find a frame fragment in a receive AHB buffer, when using the default value of 128 bytes for the receive buffers size. When in packet buffer full store and forward mode, only good received frames are written out of the DMA, so no fragments will exist in the AHB buffers due to MAC receiver errors. There is still the possibility of fragments due to DMA errors, for example used bit read on the second buffer of a multi-buffer frame. If bit zero of the receive buffer descriptor is already set when the receive buffer manager reads the location of the receive AHB buffer, then the buffer has been already used and cannot be used again until software has processed the frame and cleared bit zero. In this case, the buffer not available bit in the receive status register is set and an interrupt triggered. The receive resource error statistics register is also incremented. When the DMA is configured in the packet buffer full store and forward mode, the user can optionally select whether received frames should be automatically discarded when no AHB buffer resource is available. This feature is selected via bit 24 of the DMA Configuration Register (by default, the received frames are not automatically discarded). If this feature is off, then received packets will remain to be stored in the SRAM-based packet buffer until AHB buffer resource next becomes available. This may lead to an eventual packet buffer overflow if packets continue to be received when bit zero (used bit) of the receive buffer descriptor remains set. Note that after a used bit has been read, the receive buffer manager will reread the location of the receive buffer descriptor every time a new packet is received. When the DMA is not configured in the packet buffer full store and forward mode and a used bit is read, the frame currently being received will be automatically discarded. When the DMA is configured in the packet buffer full store and forward mode, a receive overrun condition occurs when the receive SRAM-based packet buffer is full, or because HRESP was not OK. In all other modes, a receive overrun condition occurs when either the AHB bus was not granted quickly enough, or because HRESP was not OK, or because a new frame has been detected by the receive block, but the status update or write back for the previous frame has not yet finished. For a receive overrun condition, the receive overrun interrupt is asserted and the buffer currently being written is recovered. The next frame that is received whose address is recognized reuses the buffer. When the DMA is configured for packet buffer mode, a write to bit 18 of the Network Control Register will force a packet from the external SRAM-based receive packet buffer to be flushed. This feature is only acted upon when the RX DMA is not currently writing packet data out to AHB, i.e., it is in an IDLE state. If the RX DMA is active, a write to this bit is ignored. 34.5.3.4 Transmit AHB Buffers Frames to transmit are stored in one or more transmit AHB buffers. Transmit frames can be between 1 and 16384 bytes long, so it is possible to transmit frames longer than the maximum length specified in the IEEE 802.3 standard. It should be noted that zero length AHB buffers are allowed and that the maximum number of buffers permitted for each transmit frame is 128. The start location for each transmit AHB buffer is stored in memory in a list of transmit buffer descriptors at a location pointed to by the transmit buffer queue pointer. The base address for this queue pointer is set in software using the Transmit Buffer Queue Base Address Register. Each list entry consists of two words. The first is the byte address of the transmit buffer and the second containing the transmit control and status. For the packet buffer DMA, the start location for each AHB buffer is a byte address, the bottom bits of the address being used to offset the start of the data from the data-word boundary (i.e. bits 2,1 and 0 are used to offset the address for 64-bit datapaths). For bus widths of 64 or 128bits however, the address of the buffer must be aligned to the correct 64-bit or 128-bit boundary, plus an offset of less than 4 bytes. (Note this alignment restriction in FIFO-based DMA mode only should be sufficient for applications as the main purpose is to allow alignment of the encapsulated IP packet - Given the 14 bytes of MAC encapsulation, an offset of 2 will always align the IP header to a 128-bit boundary.) Frames can be transmitted with or without automatic CRC generation. If CRC is automatically generated, pad will also be automatically generated to take frames to a minimum length of 64 bytes. When CRC is not automatically generated (as

830
defined in word 1 of the transmit buffer descriptor or through the control bus of FIFO), the frame is assumed to be at least 64 bytes long and pad is not generated. An entry in the transmit buffer descriptor list is described in Table 34-3, Transmit Buffer Descriptor Entry. To transmit frames, the buffer descriptors must be initialized by writing an appropriate byte address to bits [31:0] in the first word of each descriptor list entry. The second word of the transmit buffer descriptor is initialized with control information that indicates the length of the frame, whether or not the MAC is to append CRC and whether the buffer is the last buffer in the frame. After transmission the status bits are written back to the second word of the first buffer along with the used bit. Bit 31 is the used bit which must be zero when the control word is read if transmission is to take place. It is written to one once the frame has been transmitted. Bits[29:20] indicate various transmit error conditions. Bit 30 is the wrap bit which can be set for any buffer within a frame. If no wrap bit is encountered the queue pointer continues to increment. The Transmit Buffer Queue Base Address Register can only be updated while transmission is disabled or halted; otherwise any attempted write will be ignored. When transmission is halted the transmit buffer queue pointer will maintain its value. Therefore when transmission is restarted the next descriptor read from the queue will be from immediately after the last successfully transmitted frame. while transmit is disabled (bit 3 of the network control set low), the transmit buffer queue pointer resets to point to the address indicated by the Transmit Buffer Queue Base Address Register. Note that disabling receive does not have the same effect on the receive buffer queue pointer. Once the transmit queue is initialized, transmit is activated by writing to the transmit start bit (bit 9) of the Network Control Register. Transmit is halted when a buffer descriptor with its used bit set is read, a transmit error occurs, or by writing to the transmit halt bit of the Network Control Register. Transmission is suspended if a pause frame is received while the pause enable bit is set in the Network Configuration Register. Rewriting the start bit while transmission is active is allowed. This is implemented with TXGO variable which is readable in the Transmit Status Register at bit location 3. The TXGO variable is reset when:
z z z z
Transmit is disabled. A buffer descriptor with its ownership bit set is read. Bit 10, THALT, of the Network Control Register is written. There is a transmit error such as too many retries, late collision (gigabit mode only) or a transmit under run.
To set TXGO, write TSTART to the bit 9 of the Network Control Register. Transmit halt does not take effect until any ongoing transmit finishes. If the DMA is configured for internal FIFO mode, or for packet buffer partial store and forward mode and a collision occurs during transmission of a multi-buffer frame, transmission will automatically restart from the first buffer of the frame. For packet buffer mode, the entire contents of the frame are read into the transmit packet buffer memory, so the retry attempt will be replayed directly from the packet buffer memory rather than having to re-fetch through the AHB. If a used bit is read mid way through transmission of a multi buffer frame this is treated as a transmit error. Transmission stops, GTXER is asserted and the FCS will be bad. If transmission stops due to a transmit error or a used bit being read, transmission will restart from the first buffer descriptor of the frame being transmitted when the transmit start bit is rewritten.
Table 34-3. Transmit Buffer Descriptor Entry Bit Function Word 0 31:0 Byte address of buffer Word 1 31 Used - must be zero for the GMAC to read data to the transmit buffer. The GMAC sets this to one for the first buffer of a frame once it has been successfully transmitted. Software must clear this bit before the buffer can be used again. Wrap - marks last descriptor in transmit buffer descriptor list. This can be set for any buffer within the frame.
30

831
Table 34-3. Transmit Buffer Descriptor Entry (Continued) Bit 29 Function Retry limit exceeded, transmit error detected Transmit under run - occurs when the start of packet data has been written into the FIFO and either HRESP is not OK, or the transmit data could not be fetched in time, or when buffers are exhausted. This is not set when the DMA is configured for packet buffer mode. Transmit frame corruption due to AHB error - set if an error occurs while midway through reading transmit frame from the AHB, including HRESP errors and buffers exhausted mid frame (if the buffers run out during transmission of a frame then transmission stops, FCS shall be bad and GTXER asserted). Also set in DMA packet buffer mode if single frame is too large for configured packet buffer memory size. 26 25:23 Late collision, transmit error detected. Late collisions only force this status bit to be set in gigabit mode. Reserved Transmit IP/TCP/UDP checksum generation offload errors: 000 No Error. 001 The Packet was identified as a VLAN type, but the header was not fully complete, or had an error in it. 010 The Packet was identified as a SNAP type, but the header was not fully complete, or had an error in it. 22:20 011 The Packet was not of an IP type, or the IP packet was invalidly short, or the IP was not of type IPv4/IPv6. 100 The Packet was not identified as VLAN, SNAP or IP. 101 110 111 19:17 Non supported packet fragmentation occurred. For IPv4 packets, the IP checksum was generated and inserted. Packet type detected was not TCP or UDP. TCP/UDP checksum was therefore not generated. For IPv4 packets, the IP checksum was generated and inserted. A premature end of packet was detected and the TCP/UDP checksum could not be generated.
28
27
Reserved No CRC to be appended by MAC. When set, this implies that the data in the buffers already contains a valid CRC, hence no CRC or padding is to be appended to the current frame by the MAC.
16
This control bit must be set for the first buffer in a frame and will be ignored for the subsequent buffers of a frame. Note that this bit must be clear when using the transmit IP/TCP/UDP checksum generation offload, otherwise checksum generation and substitution will not occur.
15 14 13:0
Last buffer, when set this bit will indicate the last buffer in the current frame has been reached. Reserved Length of buffer

832
34.5.3.5 DMA Bursting on the AHB The DMA will always use SINGLE, or INCR type AHB accesses for buffer management operations. When performing data transfers, the AHB burst length used can be programmed using bits [4:0] of the DMA Configuration Register so that either SINGLE, INCR or fixed length incrementing bursts (INCR4, INCR8 or INCR16) are used where possible. When there is enough space and enough data to be transferred, the programmed fixed length bursts will be used. If there is not enough data or space available, for example when at the beginning or the end of a buffer, SINGLE type accesses are used. Also SINGLE type accesses are used at 1024 byte boundaries, so that the 1 Kbytes boundaries are not burst over as per AHB requirements. The DMA will not terminate a fixed length burst early, unless an error condition occurs on the AHB or if receive or transmit are disabled in the Network Control Register. 34.5.3.6 DMA Packet Buffer The DMA can be configured to use packet buffers for both transmit and receive paths. This mode allows multiple packets to be buffered in both transmit and receive directions. This allows the DMA to withstand far greater access latencies on the AHB and make more efficient use of the AHB bandwidth. As described earlier, (see Section 34.5.3.2 Partial Store and Forward Using Packet Buffer DMA) when the DMA is configured to use packet buffers, it can be programmed into a low latency mode, known as partial store and forward. Further details of this mode have been given (see Section 34.5.3.2) and are not repeated here. When it is programmed in full store and forward mode, full packets are buffered which provides the possibility to:
z z z
Discard packets with error on the receive path before they are partially written out of the DMA, thus saving AHB bus bandwidth and driver processing overhead, Retry collided transmit frames from the buffer, thus saving AHB bus bandwidth, Implement transmit IP/TCP/UDP checksum generation offload.
With the packet buffers included, the structure of the GMAC data paths is shown in Figure 34-2 below.

833
Figure 34-2. Data Paths with Packet Buffers Included
TX GMII MAC Transmitter
TX Packet Buffer
TX Packet Buffer DPSRAM
TX DMA APB Register Interface Status and Statistic Registers RX DMA AHB DMA AHB
MDIO Control Interface
RX Packet Buffer
RX Packet Buffer DPSRAM
RX GMII MAC Receive
Frame Filtering
Ethernet MAC
34.5.3.7 Transmit Packet Buffer The transmitter packet buffer will continue attempting to fetch frame data from the AHB system memory until the packet buffer itself is full, at which point it will attempt to maintain its full level. To accommodate the status and statistics associated with each frame, 3 words per packet (or 2 if the GMAC is configured in 64-bit datapath mode) are reserved at the end of the packet data. If the packet is bad and requires to be dropped, the status and statistics are the only information held on that packet. Storing the status in the DPRAM is required in order to decouple the DMA interface of the buffer from the MAC interface, to update the MAC status/statistics and to generate interrupts in the order in which the packets that they represent were fetched from the AHB memory. If any errors occur on the AHB while reading the transmit frame, the fetching of packet data from AHB memory is halted. The MAC transmitter will continue to fetch packet data, thereby emptying the packet buffer and allowing any good nonerrored frames to be transmitted successfully. Once these have been fully transmitted, the status/statistics for the errored frame will be updated and software will be informed via an interrupt that an AHB error occurred. This way, the error is reported in the correct packet order.

834
The transmit packet buffer will only attempt to read more frame data from the AHB when space is available in the packet buffer memory. If space is not available it must wait until the a packet fetched by the MAC completes transmission and is subsequently removed from the packet buffer memory. Note that if full store and forward mode is active and if a single frame is fetched that is too large for the packet buffer memory, the frame is flushed and the DMA halted with an error status. This is because a complete frame must be written into the packet buffer before transmission can begin, and therefore the minimum packet buffer memory size should be chosen to satisfy the maximum frame to be transmitted in the application. In full store and forward mode, once the complete transmit frame is written into the packet buffer memory, a trigger is sent across to the MAC transmitter, which will then begin reading the frame from the packet buffer memory. Since the whole frame is present and stable in the packet buffer memory an underflow of the transmitter is not possible. The frame is kept in the packet buffer until notification is received from the MAC that the frame data has either been successfully transmitted or can no longer be re-transmitted (too many retries in half duplex mode). When this notification is received the frame is flushed from memory to make room for a new frame to be fetched from AHB system memory. In partial store and forward mode, a trigger is sent across to the MAC transmitter as soon as sufficient packet data is available, which will then begin fetching the frame from the packet buffer memory. If, after this point, the MAC transmitter is able to fetch data from the packet buffer faster than the AHB DMA can fill it, an underflow of the transmitter is possible. In this case, the transmission is terminated early, and the packet buffer is completely flushed. Transmission can only be restarted by writing to the transmit START bit. In half duplex mode, the frame is kept in the packet buffer until notification is received from the MAC that the frame data has either been successfully transmitted or can no longer be re-transmitted (too many retries in half duplex mode). When this notification is received the frame is flushed from memory to make room for a new frame to be fetched from AHB system memory. In full duplex mode, the frame is removed from the packet buffer on the fly. Other than underflow, the only MAC related errors that can occur are due to collisions during half duplex transmissions. When a collision occurs the frame still exists in the packet buffer memory so can be retried directly from there. Only once the MAC transmitter has failed to transmit after sixteen attempts is the frame finally flushed from the packet buffer. 34.5.3.8 Receive Packet Buffer The receive packet buffer stores frames from the MAC receiver along with their status and statistics. Frames with errors are flushed from the packet buffer memory, while good frames are pushed onto the DMA AHB interface. The receiver packet buffer monitors FIFO from the MAC receiver and translates the FIFO pushes into packet buffer writes. At the end of the received frame the status and statistics are buffered so that the information can be used when the frame is read out. When programmed in full store and forward mode, if the frame has an error the frame data is immediately flushed from the packet buffer memory allowing subsequent frames to utilise the freed up space. The status and statistics for bad frames are still used to update the GMAC registers. To accommodate the status and statistics associated with each frame, 3 words per packet (or 2 if configured in 64 bit datapath mode) are reserved at the end of the packet data. If the packet is bad and requires to be dropped, the status and statistics are the only information held on that packet. The receiver packet buffer will also detect a full condition so that an overflow condition can be detected. If this occurs, subsequent packets are dropped and an RX overflow interrupt is raised. For full store and forward, the DMA only begins packet fetches once the status and statistics for a frame are available. If the frame has a bad status due to a frame error, the status and statistics are passed onto the GMAC registers. If the frame has a good status, the information is used to read the frame from the packet buffer memory and burst onto the AHB using the DMA buffer management protocol. Once the last frame data has been transferred to the FIFO, the status and statistics are updated to the GMAC registers. If partial store and forward mode is active, the DMA will begin fetching the packet data before the status is available. As soon as the status becomes available, the DMA will fetch this information as soon as possible before continuing to fetch the remainder of the frame. Once the last frame data has been transferred to the FIFO, the status and statistics are updated to the GMAC registers.

835
34.5.3.9 Priority Queueing in the DMA The DMA by default uses a single transmit and receive queue. This means the list of transmit/receive buffer descriptors point to data buffers associated with a single transmit/receive data stream. When the DMA is configured to use packet buffer memories, the GMAC can optionally select up to 8 priority queues. Each queue has an independent list of buffer descriptors pointing to separate data streams. In the transmit direction, higher priority queues are always serviced before lower priority queues. This strict priority scheme requires the user to ensure that high priority traffic is constrained so that lower priority traffic will have required bandwidth. The GMAC DMA will determine the next queue to service by initiating a sequence of buffer descriptor reads interrogating the ownership bits of each. The buffer descriptor corresponding to the highest priority queue is read first. If the ownership bit of this descriptor is set, then the DMA will progress to reading the 2nd highest priority queue's descriptor. If that ownership bit read of this lower priority queue is set, then the DMA will read the 3rd highest priority queue's descriptor, and so on. If all the descriptors return an ownership bit set, then a resource error has occurred, an interrupt is generated and transmission is automatically halted. Transmission can only be restarted by setting the START bit in the Network Control Register. The GMAC DMA will need to identify the highest available queue to transmit from when the START bit in the Network Control Register is written to and the TX is in a halted state, or when the last word of any packet has been fetched from external AHB memory. The GMAC transmit DMA maximizes the effectiveness of priority queuing by ensuring that high priority traffic be transmitted as early as possible after being fetched from AHB. High priority traffic fetched from AHB will be pushed to the MAC layer before any lower priority traffic that may pre-exist in the transmit SRAM-based packet buffer. This is achieved by separating the transmit SRAM-based packet buffer into regions, one region per queue. The size of each region determines the amount of SRAM space allocated per queue. For each queue, there is an associated Transmit Buffer Queue Base Address Register. For the lowest priority queue (or the only queue when only 1 queue is selected), the Transmit Buffer Queue Base Address is located at address 0x1C. For all other queues, the Transmit Buffer Queue Base Address Registers are located at sequential addresses starting at address 0x440. In the receive direction each packet is written to AHB data buffers in the order that it is received. For each queue, there is an independent set of receive AHB buffers for each queue. There is therefore a separate Receive Buffer Queue Base Address Register for each queue. For the lowest priority queue (or the only queue when only one queue is selected), the Receive Buffer Queue Base Address is located at address 0x18. For all other queues, the Receive Buffer Queue Base Address Registers are located at sequential addresses starting at address 0x480. Every received packet will pass through a programmable screening algorithm which will allocate a particular queue to that frame. The user interface to the screeners is through two types of programmable registers.
z
Screener Type 1 Registers hold values to match against particular IP and UDP fields of the received frames. The fields used to match against are DS (Differentiated Services field of IPv4 frames), TC (Traffic class field of IPv6 frames) and/or the UDP destination port. Screener Type 2 Registers hold values to match against the VLAN priority field of the received frames.
Each screener type has an enable bit, a match pattern and a queue number. If a received frame matches on an enabled screener register, then the frame will be tagged with the queue value in the associated screener register, and forwarded onto the DMA and subsequently into the external memory associated with that queue. An APB address location is required for each screener register, and 16 of each type have been reserved in the default address map. If two screeners are matched then the one which resides at the lowest register address will take priority so care must be taken on the selection of the screener location. When the priority queuing feature is enabled, the number of interrupt outputs from the GMAC core is increased to match the number of supported queues. The number of Interrupt Status Registers is increased by the same number. Only DMA related events are reported using the individual interrupt outputs, as the GMAC can relate these events to specific queues. All other events generated within the GMAC are reported in the interrupt associated with the lowest priority queue. For the lowest priority queue (or the only queue when only 1 queue is selected), the Interrupt Status Register is located at address 0x24. For all other queues, the Interrupt Status Register is located at sequential addresses starting at address 0x400.

836
34.5.4 MAC Transmit Block

The MAC transmitter can operate in either half duplex or full duplex mode and transmits frames in accordance with the Ethernet IEEE 802.3 standard. In half duplex mode, the CSMA/CD protocol of the IEEE 802.3 specification is followed. A small input buffer receives data through FIFO which, depending on the DMA bus width control bits in the Network Configuration Register, will extract data in 32-bit or 64-bit form. All subsequent processing prior to the final output is performed in bytes. Transmit data can be output using the GMII/MII interface. Frame assembly starts by adding preamble and the start frame delimiter. Data is taken from FIFO a word at a time. When the GMAC is configured for gigabit operation, the data output to the PHY uses all 8 bits of the txd[7:0] output. In 10/100 mode, transmit data to the PHY is nibble wide and least significant nibble first using txd[3:0] with txd[7:4] tied to logic 0. If necessary, padding is added to take the frame length to 60 bytes. CRC is calculated using an order 32-bit polynomial. This is inverted and appended to the end of the frame taking the frame length to a minimum of 64 bytes. If the no CRC bit is set in the second word of the last buffer descriptor of a transmit frame, neither pad nor CRC are appended. The no CRC bit can also be set through FIFO. In full duplex mode (at all data rates), frames are transmitted immediately. Back to back frames are transmitted at least 96 bit times apart to guarantee the interframe gap. In half duplex mode, the transmitter checks carrier sense. If asserted, the transmitter waits for the signal to become inactive, and then starts transmission after the interframe gap of 96 bit times. If the collision signal is asserted during transmission, the transmitter will transmit a jam sequence of 32 bits taken from the data register and then retry transmission after the back off time has elapsed. If the collision occurs during either the preamble or Start Frame Delimiter (SFD), then these fields will be completed prior to generation of the jam sequence. The back off time is based on an XOR of the 10 least significant bits of the data coming from FIFO and a 10 bit pseudo random number generator. The number of bits used depends on the number of collisions seen. After the first collision 1 bit is used, then the second 2 bits and so on up to the maximum of 10 bits. All 10 bits are used above ten collisions. An error will be indicated and no further attempts will be made if 16 consecutive attempts cause collision. This operation is compliant with the description in Clause 4.2.3.2.5 of the IEEE 802.3 standard which refers to the truncated binary exponential back off algorithm. In 10/100 mode, both collisions and late collisions are treated identically, and back off and retry will be performed up to 16 times. When operating in gigabit mode, late collisions are treated as an exception and transmission is aborted, without retry. This condition is reported in the transmit buffer descriptor word 1 (late collision, bit 26) and also in the Transmit Status Register (late collision, bit 7). An interrupt can also be generated (if enabled) when this exception occurs, and bit 5 in the Interrupt Status Register will be set. When operating in gigabit mode (half duplex) both carrier extension and frame bursting are performed in accordance with the IEEE 802.3 standard. For frames less than 512 bytes carrier extension is used to ensure the minimum slot time is not violated. Frame bursting is used by the transmitter in gigabit mode (half duplex) when more than one frame is queued for transmission. The first frame of a burst must be carrier extended (if necessary) to ensure the minimum slot time of 512 bytes is achieved, after which all subsequent frames within the burst must only satisfy the minimum frame length of 64 bytes or greater. Each interframe gap within the burst is filled by the transmitter with carrier extensions, thus ensuring control of the medium is not given up. Several frames may be transmitted up to the burst limit of 65,536 bytes. The transmitter relinquishes control of the medium when there are no more frames queued for transmission or the burst limit is exceeded. In gigabit mode any collisions occurring after the minimum slot time for the first frame within a burst are treated as a late collision. The burst is terminated upon this event. In all modes of operation, if the transmit DMA under runs, a bad CRC is automatically appended using the same mechanism as jam insertion and the GTXER signal is asserted. For a properly configured system this should never happen and also it is impossible if configured to use the DMA with packet buffers, as the complete frame is buffered in local packet buffer memory.

837
By setting when bit 28 is set in the Network Configuration Register, the Inter Packet Gap (IPG) may be stretched beyond 96 bits depending on the length of the previously transmitted frame and the value written to the IPG Stretch Register (GMAC_IPGS). The least significant 8 bits of the IPG Stretch Register multiply the previous frame length (including preamble). The next significant 8 bits (+1 so as not to get a divide by zero) divide the frame length to generate the IPG. IPG stretch only works in full duplex mode and when bit 28 is set in the Network Configuration Register. The IPG Stretch Register cannot be used to shrink the IPG below 96 bits. If the back pressure bit is set in the Network Control Register, or if the HDFC configuration bit is set in the GMAC_UR register (10M or 100M half duplex mode), the transmit block transmits 64 bits of data, which can consist of 16 nibbles of 1011 or in bit rate mode 64 1s, whenever it sees an incoming frame to force a collision. This provides a way of implementing flow control in half duplex mode. Note that this feature is not available in gigabit half duplex mode.
34.5.5 MAC Receive Block

All processing within the MAC receive block is implemented using a 16-bit data path. The MAC receive block checks for valid preamble, FCS, alignment and length, presents received frames to FIFO and stores the frame destination address for use by the address checking block. If, during the frame reception, the frame is found to be too long, a bad frame indication is sent to FIFO. The receiver logic ceases to send data to memory as soon as this condition occurs. At end of frame reception the receive block indicates to the DMA block whether the frame is good or bad. The DMA block will recover the current receive buffer if the frame was bad. Ethernet frames are normally stored in DMA memory or to FIFO complete with the FCS. Setting the FCS remove bit in the network configuration (bit 17) causes frames to be stored without their corresponding FCS. The reported frame length field is reduced by four bytes to reflect this operation. The receive block signals to the register block to increment the alignment, CRC (FCS), short frame, long frame, jabber or receive symbol errors when any of these exception conditions occur. If bit 26 is set in the network configuration, CRC errors will be ignored and CRC errored frames will not be discarded, though the Frame Check Sequence Errors statistic register will still be incremented. Additionally, if configured to use the DMA and not enabled for jumbo frames mode, then bit[13] of the receiver descriptor word 1 will be updated to indicate the FCS validity for the particular frame. This is useful for applications such as EtherCAT whereby individual frames with FCS errors must be identified. Received frames can be checked for length field error by setting the length field error frame discard bit of the Network Configuration Register (bit-16). When this bit is set, the receiver compares a frame's measured length with the length field (bytes 13 and 14) extracted from the frame. The frame is discarded if the measured length is shorter. This checking procedure is for received frames between 64 bytes and 1518 bytes in length. Each discarded frame is counted in the 10 bit length field error statistics register. Frames where the length field is greater than or equal to 0x0600 hex will not be checked. When operating in gigabit mode (half duplex), the receiver will discard frames which do not meet the minimal slot time of 512 bytes. If a burst is detected, the first frame is checked to ensure it meets the slot time, but all subsequent frames of the burst are checked to ensure they meet the minimum frame size of 64 bytes. In gigabit mode (half duplex), carrier extension errors are detected by the receiver during the minimum slot time, and the frame discarded. An error of this nature causes the receive symbol errors statistic register to be incremented. Carrier extension errors occurring during the inter packet gap period are ignored and have no effect on the statistics.
34.5.6 Checksum Offload for IP, TCP and UDP

The GMAC can be programmed to perform IP, TCP and UDP checksum offloading in both receive and transmit directions, which is enabled by setting bit 24 in the Network Configuration Register for receive and bit 11 in the DMA Configuration Register for transmit. IPv4 packets contain a 16-bit checksum field, which is the 16-bit 1's complement of the 1's complement sum of all 16-bit words in the header. TCP and UDP packets contain a 16-bit checksum field, which is the 16-bit 1's complement of the 1's complement sum of all 16-bit words in the header, the data and a conceptual IP pseudo header.

838
To calculate these checksums in software requires each byte of the packet to be processed. For TCP and UDP this can use a large amount of processing power. Offloading the checksum calculation to hardware can result in significant performance improvements. For IP, TCP or UDP checksum offload to be useful, the operating system containing the protocol stack must be aware that this offload is available so that it can make use of the fact that the hardware can either generate or verify the checksum. 34.5.6.1 Receiver Checksum Offload When receive checksum offloading is enabled in the GMAC, the IPv4 header checksum is checked as per RFC 791, where the packet meets the following criteria:
z z z z
If present, the VLAN header must be four octets long and the CFI bit must not be set. Encapsulation must be RFC 894 Ethernet Type Encoding or RFC 1042 SNAP Encoding. IPv4 packet. IP header is of a valid length.
The GMAC also checks the TCP checksum as per RFC 793, or the UDP checksum as per RFC 768, if the following criteria are met:
z z z z
IPv4 or IPv6 packet Good IP header checksum (if IPv4). No IP fragmentation. TCP or UDP packet.
When an IP, TCP or UDP frame is received, the receive buffer descriptor gives an indication if the GMAC was able to verify the checksums. There is also an indication if the frame had SNAP encapsulation. These indication bits will replace the type ID match indication bits when the receive checksum offload is enabled. For details of these indication bits refer to Table 34-2, Receive Buffer Descriptor Entry. If any of the checksums are verified as incorrect by the GMAC, the packet is discarded and the appropriate statistics counter incremented. 34.5.6.2 Transmitter Checksum Offload The transmitter checksum offload is only available if the GMAC is configured to use the DMA in packet buffer mode and full store and forward mode is enabled. This is because the complete frame to be transmitted must be read into the packet buffer memory before the checksum can be calculated and written back into the headers at the beginning of the frame. Transmitter checksum offload is enabled by setting bit [11] in the DMA Configuration Register. When enabled, it will monitor the frame as it is written into the transmitter packet buffer memory to automatically detect the protocol of the frame. Protocol support is identical to the receiver checksum offload. For transmit checksum generation and substitution to occur, the protocol of the frame must be recognized and the frame must be provided without the FCS field, by making sure that bit [16] of the transmit descriptor word 1 is clear. If the frame data already had the FCS field, this would be corrupted by the substitution of the new checksum fields. If these conditions are met, the transmit checksum offload engine will calculate the IP, TCP and UDP checksums as appropriate. Once the full packet is completely written into packet buffer memory, the checksums will be valid and the relevant DPRAM locations will be updated for the new checksum fields as per standard IP/TCP and UDP packet structures. If the transmitter checksum engine is prevented from generating the relevant checksums, bits [22:20] of the transmitter DMA writeback status will be updated to identify the reason for the error. Note that the frame will still be transmitted but without the checksum substitution, as typically the reason that the substitution did not occur was that the protocol was not recognized.

839
34.5.7 MAC Filtering Block

The filter block determines which frames should be written to FIFO and on to the optional DMA. Whether a frame is passed depends on what is enabled in the Network Configuration Register, the state of the external matching pins, the contents of the specific address, type and Hash Registers and the frame's destination address and type field. If bit 25 of the Network Configuration Register is not set, a frame will not be copied to memory if the GMAC is transmitting in half duplex mode at the time a destination address is received. Ethernet frames are transmitted a byte at a time, least significant bit first. The first six bytes (48 bits) of an Ethernet frame make up the destination address. The first bit of the destination address, which is the LSB of the first byte of the frame, is the group or individual bit. This is one for multicast addresses and zero for unicast. The all ones address is the broadcast address and a special case of multicast. The GMAC supports recognition of four specific addresses. Each specific address requires two registers, Specific Address Register Bottom and Specific Address Register Top. Specific Address Register Bottom stores the first four bytes of the destination address and Specific Address Register Top contains the last two bytes. The addresses stored can be specific, group, local or universal. The destination address of received frames is compared against the data stored in the Specific Address Registers once they have been activated. The addresses are deactivated at reset or when their corresponding Specific Address Register Bottom is written. They are activated when Specific Address Register Top is written. If a receive frame address matches an active address, the frame is written to FIFO and on to DMA memory if used. Frames may be filtered using the type ID field for matching. Four type ID registers exist in the register address space and each can be enabled for matching by writing a one to the MSB (bit 31) of the respective register. When a frame is received, the matching is implemented as an OR function of the various types of match. The contents of each type ID register (when enabled) are compared against the length/type ID of the frame being received (e.g. bytes 13 and 14 in non-VLAN and non-SNAP encapsulated frames) and copied to memory if a match is found. The encoded type ID match bits (Word 0, Bit 22 and Bit 23) in the receive buffer descriptor status are set indicating which type ID register generated the match, if the receive checksum offload is disabled. The reset state of the type ID registers is zero, hence each is initially disabled. The following example illustrates the use of the address and type ID match registers for a MAC address of 21:43:65:87:A9:CB:
Preamble SFD DA (Octet 0 - LSB) DA (Octet 1) DA (Octet 2) DA (Octet 3) DA (Octet 4) DA (Octet 5 - MSB) SA (LSB) SA SA SA SA SA (MSB) Type ID (MSB) Type ID (LSB) 55 D5 21 43 65 87 A9 CB 00(1) 00(1) 00(1) 00(1) 00(1) 00(1) 43 21
Note:
1.
Contains the address of the transmitting device

840
The sequence above shows the beginning of an Ethernet frame. Byte order of transmission is from top to bottom as shown. For a successful match to specific address 1, the following address matching registers must be set up: Specific Address 1 Bottom [31:0] Register (GMAC_SAB1)(Address 0x088) 0x87654321 Specific Address 1 Top [47:32] Register (GMAC_SAT1) (Address 0x08C) 0x0000CBA9
And for a successful match to the type ID, the following type ID match 1 register must be set up: Type ID Match 1 Register (GMAC_TIDM1) (Address 0x0A8) 0x80004321
34.5.8 Broadcast Address

Frames with the broadcast address of 0xFFFFFFFFFFFF are stored to memory only if the 'no broadcast' bit in the Network Configuration Register is set to zero.
34.5.9 Hash Addressing

The hash address register is 64 bits long and takes up two locations in the memory map. The least significant bits are stored in Hash Register Bottom and the most significant bits in Hash Register Top. The unicast hash enable and the multicast hash enable bits in the Network Configuration Register enable the reception of hash matched frames. The destination address is reduced to a 6-bit index into the 64-bit Hash Register using the following hash function: The hash function is an XOR of every sixth bit of the destination address. hash_index[05] = da[05] ^ da[11] ^ da[17] ^ da[23] ^ da[29] ^ da[35] ^ da[41] ^ da[47] hash_index[04] = da[04] ^ da[10] ^ da[16] ^ da[22] ^ da[28] ^ da[34] ^ da[40] ^ da[46] hash_index[03] = da[03] ^ da[09] ^ da[15] ^ da[21] ^ da[27] ^ da[33] ^ da[39] ^ da[45] hash_index[02] = da[02] ^ da[08] ^ da[14] ^ da[20] ^ da[26] ^ da[32] ^ da[38] ^ da[44] hash_index[01] = da[01] ^ da[07] ^ da[13] ^ da[19] ^ da[25] ^ da[31] ^ da[37] ^ da[43] hash_index[00] = da[00] ^ da[06] ^ da[12] ^ da[18] ^ da[24] ^ da[30] ^ da[36] ^ da[42]
da[0] represents the least significant bit of the first byte received, that is, the multicast/unicast indicator, and da[47]
represents the most significant bit of the last byte received. If the hash index points to a bit that is set in the Hash Register then the frame will be matched according to whether the frame is multicast or unicast. A multicast match will be signalled if the multicast hash enable bit is set, da[0] is logic 1 and the hash index points to a bit set in the Hash Register. A unicast match will be signalled if the unicast hash enable bit is set, da[0] is logic 0 and the hash index points to a bit set in the Hash Register. To receive all multicast frames, the Hash Register should be set with all ones and the multicast hash enable bit should be set in the Network Configuration Register.
34.5.10 Copy all Frames (Promiscuous Mode)

If the copy all frames bit is set in the Network Configuration Register then all frames (except those that are too long, too short, have FCS errors or have GRXER asserted during reception) will be copied to memory. Frames with FCS errors will be copied if bit 26 is set in the Network Configuration Register.
34.5.11 Disable Copy of Pause Frames

Pause frames can be prevented from being written to memory by setting the disable copying of pause frames control bit 23 in the Network Configuration Register. When set, pause frames are not copied to memory regardless of the copy all frames bit, whether a hash match is found, a type ID match is identified or if a destination address match is found.

841
34.5.12 VLAN Support

An Ethernet encoded 802.1Q VLAN tag looks like this:
Table 34-4. 802.1Q VLAN Tag TPID (Tag Protocol Identifier) 16 bits 0x8100 TCI (Tag Control Information) 16 bits First 3 bits priority, then CFI bit, last 12 bits VID
The VLAN tag is inserted at the 13th byte of the frame adding an extra four bytes to the frame. To support these extra four bytes, the GMAC can accept frame lengths up to 1536 bytes by setting bit 8 in the Network Configuration Register. If the VID (VLAN identifier) is null (0x000) this indicates a priority-tagged frame. The following bits in the receive buffer descriptor status word give information about VLAN tagged frames:z z z z
Bit 21 set if receive frame is VLAN tagged (i.e. type ID of 0x8100). Bit 20 set if receive frame is priority tagged (i.e. type ID of 0x8100 and null VID). (If bit 20 is set, bit 21 will be set also.) Bit 19, 18 and 17 set to priority if bit 21 is set. Bit 16 set to CFI if bit 21 is set.
The GMAC can be configured to reject all frames except VLAN tagged frames by setting the discard non-VLAN frames bit in the Network Configuration Register.
34.5.13 Wake on LAN Support

The receive block supports Wake on LAN by detecting the following events on incoming receive frames:
z z z z
Magic packet Address Resolution Protocol (ARP) request to the device IP address Specific address 1 filter match Multicast hash filter match
These events can be individually enabled through bits [19:16] of the Wake on LAN Register. Also, for Wake on LAN detection to occur, receive enable must be set in the Network Control Register, however a receive buffer does not have to be available. In case of an ARP request, specific address 1 or multicast filter events will occur even if the frame is errored. For magic packet events, the frame must be correctly formed and error free. A magic packet event is detected if all of the following are true:
z z z z z
Magic packet events are enabled through bit 16 of the Wake on LAN Register The frame's destination address matches specific address 1 The frame is correctly formed with no errors The frame contains at least 6 bytes of 0xFF for synchronization There are 16 repetitions of the contents of Specific Address 1 Register immediately following the synchronization ARP request events are enabled through bit 17 of the Wake on LAN Register Broadcasts are allowed by bit 5 in the Network Configuration Register The frame has a broadcast destination address (bytes 1 to 6) The frame has a type ID field of 0x0806 (bytes 13 and 14) The frame has an ARP operation field of 0x0001 (bytes 21 and 22) The least significant 16 bits of the frame's ARP target protocol address (bytes 41 and 42) match the value programmed in bits[15:0] of the Wake on LAN Register
An ARP request event is detected if all of the following are true:

z z z z z z
The decoding of the ARP fields adjusts automatically if a VLAN tag is detected within the frame. The reserved value of 0x0000 for the Wake on LAN target address value will not cause an ARP request event, even if matched by the frame.

842
A specific address 1 filter match event will occur if all of the following are true:
z z
Specific address 1 events are enabled through bit 18 of the Wake on LAN Register The frame's destination address matches the value programmed in the Specific Address 1 Registers Multicast hash events are enabled through bit 19 of the Wake on LAN Register Multicast hash filtering is enabled through bit 6 of the Network Configuration Register The frame destination address matches against the multicast hash filter The frame destination address is not a broadcast
A multicast filter match event will occur if all of the following are true:
z z z z
34.5.14 IEEE 1588 Support

IEEE 1588 is a standard for precision time synchronization in local area networks. It works with the exchange of special Precision Time Protocol (PTP) frames. The PTP messages can be transported over IEEE 802.3/Ethernet, over Internet Protocol Version 4 or over Internet Protocol Version 6 as described in the annex of IEEE P1588.D2.1. GMAC output pins indicate the message time-stamp point (asserted on the start packet delimiter and de-asserted at end of frame) for all frames and the passage of PTP event frames (asserted when a PTP event frame is detected and deasserted at end of frame). Synchronization between master and slave clocks is a two stage process. First, the offset between the master and slave clocks is corrected by the master sending a sync frame to the slave with a follow up frame containing the exact time the sync frame was sent. Hardware assist modules at the master and slave side detect exactly when the sync frame was sent by the master and received by the slave. The slave then corrects its clock to match the master clock. Second, the transmission delay between the master and slave is corrected. The slave sends a delay request frame to the master which sends a delay response frame in reply. Hardware assist modules at the master and slave side detect exactly when the delay request frame was sent by the slave and received by the master. The slave will now have enough information to adjust its clock to account for delay. For example, if the slave was assuming zero delay, the actual delay will be half the difference between the transmit and receive time of the delay request frame (assuming equal transmit and receive times) because the slave clock will be lagging the master clock by the delay time already. The time-stamp is taken when the message time-stamp point passes the clock time-stamp point. For Ethernet the message time-stamp point is the SFD and the clock time-stamp point is the MII interface. (The IEEE 1588 specification refers to sync and delay_req messages as event messages as these require time-stamping. Follow up, delay response and management messages do not require time-stamping and are referred to as general messages.) 1588 version 2 defines two additional PTP event messages. These are the peer delay request (Pdelay_Req) and peer delay response (Pdelay_Resp) messages. These messages are used to calculate the delay on a link. Nodes at both ends of a link send both types of frames (regardless of whether they contain a master or slave clock). The Pdelay_Resp message contains the time at which a Pdelay_Req was received and is itself an event message. The time at which a Pdelay_Resp message is received is returned in a Pdelay_Resp_Follow_Up message. 1588 version 2 introduces transparent clocks of which there are two kinds, peer-to-peer (P2P) and end-to-end (E2E). Transparent clocks measure the transit time of event messages through a bridge and amend a correction field within the message to allow for the transit time. P2P transparent clocks additionally correct for the delay in the receive path of the link using the information gathered from the peer delay frames. With P2P transparent clocks delay_req messages are not used to measure link delay. This simplifies the protocol and makes larger systems more stable. The GMAC recognizes four different encapsulations for PTP event messages: 1. 2. 3. 4. 1588 version 1 (UDP/IPv4 multicast) 1588 version 2 (UDP/IPv4 multicast) 1588 version 2 (UDP/IPv6 multicast) 1588 version 2 (Ethernet multicast)

843
Example of a sync frame in the 1588 version 1 format:

Preamble/SFD DA (Octets 0 - 5) SA (Octets 6 - 11) Type (Octets 12-13) IP stuff (Octets 14-22) UDP (Octet 23) IP stuff (Octets 24-29) IP DA (Octets 30-32) IP DA (Octet 33) source IP port (Octets 34-35) dest IP port (Octets 36-37) other stuff (Octets 38-42) version PTP (Octet 43) other stuff (Octets 44-73) control (Octet 74) other stuff (Octets 75-168) 00 01 013F E00001 81 or 82 or 83 or 84 11 0800 55555555555555D5
Example of a delay request frame in the 1588 version 1 format:

Preamble/SFD DA (Octets 0 - 5) SA (Octets 6 - 11) Type (Octets 12-13) IP stuff (Octets 14-22) UDP (Octet 23) IP stuff (Octets 24-29) IP DA (Octets 30-32) IP DA (Octet 33) source IP port (Octets 34-35) dest IP port (Octets 36-37) other stuff (Octets 38-42) version PTP (Octet 43) other stuff (Octets 44-73) control (Octet 74) other stuff (Octets 75-168) 01 01 013F E00001 81 or 82 or 83 or 84 11 0800 55555555555555D5

844
For 1588 version 1 messages, sync and delay request frames are indicated by the GMAC if the frame type field indicates TCP/IP, UDP protocol is indicated, the destination IP address is 224.0.1.129/130/131 or 132, the destination UDP port is 319 and the control field is correct. The control field is 0x00 for sync frames and 0x01 for delay request frames. For 1588 version 2 messages, the type of frame is determined by looking at the message type field in the first byte of the PTP frame. Whether a frame is version 1 or version 2 can be determined by looking at the version PTP field in the second byte of both version 1 and version 2 PTP frames. In version 2 messages sync frames have a message type value of 0x0, delay_req have 0x1, pdelay_req have 0x2 and pdelay_resp have 0x3. Example of a sync frame in the 1588 version 2 (UDP/IPv4) format
Preamble/SFD DA (Octets 0 - 5) SA (Octets 6 - 11) Type (Octets 12-13) IP stuff (Octets 14-22) UDP (Octet 23) IP stuff (Octets 24-29) IP DA (Octets 30-33) source IP port (Octets 34-35) dest IP port (Octets 36-37) other stuff (Octets 38-41) message type (Octet 42) version PTP (Octet 43) 00 02 013F E0000181 11 0800 55555555555555D5
Example of a pdelay_req frame in the 1588 version 2 (UDP/IPv4) format

Preamble/SFD DA (Octets 0 - 5) SA (Octets 6 - 11) Type (Octets 12-13) IP stuff (Octets 14-22) UDP (Octet 23) IP stuff (Octets 24-29) IP DA (Octets 30-33) source IP port (Octets 34-35) dest IP port (Octets 36-37) other stuff (Octets 38-41) message type (Octet 42) version PTP (Octet 43) 02 02 013F E000006B 11 0800 55555555555555D5

845
Example of a synch frame in the 1588 version 2 (UDP/IPv6) format:

Preamble/SFD DA (Octets 0 - 5) SA (Octets 6 - 11) Type (Octets 12-13) IP stuff (Octets 14-19) UDP (Octet 20) IP stuff (Octets 21-37) IP DA (Octets 38-53) source IP port (Octets 54-55) dest IP port (Octets 56-57) other stuff (Octets 58-61) message type (Octet 62) other stuff (Octets 63-93) version PTP (Octet 94) 02 00 013F FF0X00000000018 11 86dd 55555555555555D5
Example of a pdelay_resp frame in the 1588 version 2 (UDP/IPv6) format:

Preamble/SFD DA (Octets 0 - 5) SA (Octets 6 - 11) Type (Octets 12-13) IP stuff (Octets 14-19) UDP (Octet 20) IP stuff (Octets 21-37) IP DA (Octets 38-53) source IP port (Octets 54-55) dest IP port (Octets 56-57) other stuff (Octets 58-61) message type (Octet 62) other stuff (Octets 63-93) version PTP (Octet 94) 02 03 013F FF0200000000006B 11 86dd 55555555555555D5
Example of a sync frame in the 1588 version 2 (Ethernet multicast) format. For the multicast address 011B19000000 sync and delay request frames are recognized depending on the message type field, 00 for sync and 01 for delay request::
Preamble/SFD DA (Octets 0 - 5) SA (Octets 6 - 11) Type (Octets 12-13) message type (Octet 14) version PTP (Octet 15) 88F7 00 02 55555555555555D5 011B19000000

846
Example of a pdelay_req frame in the 1588 version 2 (Ethernet multicast) format, these need a special multicast address so they can get through ports blocked by the spanning tree protocol. For the multicast address 0180C200000E sync, pdelay request and pdelay response frames are recognized depending on the message type field, 00 for sync, 02 for pdelay request and 03 for pdelay response.
Preamble/SFD DA (Octets 0 - 5) SA (Octets 6 - 11) Type (Octets 12-13) message type (Octet 14) version PTP (Octet 15) 88F7 00 02 55555555555555D5 0180C200000E
34.5.15 Time Stamp Unit

The TSU consists of a timer and registers to capture the time at which PTP event frames cross the message timestamp point. An interrupt is issued when a capture register is updated. The timer is implemented as a 62-bit register with the upper 32 bits counting seconds and the lower 30 bits counting nanoseconds. The lower 30 bits roll over when they have counted to one second. An interrupt is generated when the seconds increment. The timer value can be read, written and adjusted through the APB interface. The timer is clocked by MCK. The amount by which the timer increments each clock cycle is controlled by the timer increment register. Bits [7:0] are the default increment value in nanoseconds. If the rest of the register is written with zero the timer increments by the value in [7:0] each clock cycle. Bits [15:8] of the increment register are the alternative increment value in nanoseconds and bits [23:16] are the number of increments after which the alternative increment value is used. If [23:16] are zero then the alternative increment value will never be used. Taking the example of 10.2 MHz, there are 102 cycles every ten microseconds or 51 every five microseconds. So a timer with a 10.2 MHz clock source is constructed by incrementing by 98 ns for fifty cycles and then incrementing by 100 ns (98*50 + 100 = 5000). This is programmed by setting the 1588 Timer Increment Register to 0x00326462. For a 49.8 MHz clock source it would be 20 ns for 248 cycles followed by an increment of 40 ns (20*248 + 40 = 5000) programmed as 0x00F82814. Having eight bits for the number of increments field allows frequencies up to 50 MHz to be supported with 200 kHz resolution Without the alternative increment field the period of the clock would be limited to an integer number of nanoseconds, resulting in supported clock frequencies of 8, 10, 20, 25, 40, 50, 100, 125, 200 and 250 MHz. GMAC is configured to operate as PTP slave. The timer register increments as normal but the timer value is copied to the sync strobe register. There are six additional 62-bit registers that capture the time at which PTP event frames are transmitted and received. An interrupt is issued when these registers are updated.
34.5.16 MAC 802.3 Pause Frame Support

Note: See Clause 31, and Annex 31A and 31B of the IEEE standard 802.3 for a full description of MAC 802.3 pause operation.

847
The start of a MAC 802.3 pause frame is as follows:

Table 34-5. Start of an 802.3 Pause Frame Destination Address 0x0180C2000001 Source Address 6 bytes Type (Mac Control Frame) 0x8808 Pause Opcode 0x0001 Pause Time 2 bytes
The GMAC supports both hardware controlled pause of the transmitter, upon reception of a pause frame, and hardware generated pause frame transmission. 34.5.16.1 802.3 Pause Frame Reception Bit 13 of the Network Configuration Register is the pause enable control for reception. If this bit is set, transmission will pause if a non zero pause quantum frame is received. If a valid pause frame is received then the Pause Time Register is updated with the new frame's pause time, regardless of whether a previous pause frame is active or not. An interrupt (either bit 12 or bit 13 of the Interrupt Status Register) is triggered when a pause frame is received, but only if the interrupt has been enabled (bit 12 and bit 13 of the Interrupt Mask Register). Pause frames received with non zero quantum are indicated through the interrupt bit 12 of the Interrupt Status Register. Pause frames received with zero quantum are indicated on bit 13 of the Interrupt Status Register. Once the Pause Time Register is loaded and the frame currently being transmitted has been sent, no new frames are transmitted until the pause time reaches zero. The loading of a new pause time, and hence the pausing of transmission, only occurs when the GMAC is configured for full duplex operation. If the GMAC is configured for half duplex there will be no transmission pause, but the pause frame received interrupt will still be triggered. A valid pause frame is defined as having a destination address that matches either the address stored in Specific Address Register 1 or if it matches the reserved address of 0x0180C2000001. It must also have the MAC control frame type ID of 0x8808 and have the pause opcode of 0x0001. Pause frames that have frame check sequence (FCS) or other errors will be treated as invalid and will be discarded. 802.3 Pause frames that are received after Priority-based Flow Control (PFC) has been negotiated will also be discarded. Valid pause frames received will increment the pause frames received statistic register. The pause time register decrements every 512 bit times once transmission has stopped. For test purposes, the retry test bit can be set (bit 12 in the Network Configuration Register) which causes the Pause Time Register to decrement every GTXCK cycle once transmission has stopped. The interrupt (bit 13 in the Interrupt Status Register) is asserted whenever the Pause Time Register decrements to zero (assuming it has been enabled by bit 13 in the Interrupt Mask Register). This interrupt is also set when a zero quantum pause frame is received. 34.5.16.2 802.3 Pause Frame Transmission Automatic transmission of pause frames is supported through the transmit pause frame bits of the Network Control Register. If either bit 11 or bit 12 of the Network Control Register is written with logic 1, an 802.3 pause frame will be transmitted, providing full duplex is selected in the Network Configuration Register and the transmit block is enabled in the Network Control Register. Pause frame transmission will happen immediately if transmit is inactive or if transmit is active between the current frame and the next frame due to be transmitted. Transmitted pause frames comprise the following:
z z z z z z z
A destination address of 01-80-C2-00-00-01 A source address taken from Specific Address Register 1 A type ID of 88-08 (MAC control frame) A pause opcode of 00-01 A pause quantum register Fill of 00 to take the frame to minimum frame length Valid FCS

848
The pause quantum used in the generated frame will depend on the trigger source for the frame as follows:
z z
If bit 11 is written with a one, the pause quantum will be taken from the Transmit Pause Quantum Register. The Transmit Pause Quantum Register resets to a value of 0xFFFF giving maximum pause quantum as default. If bit 12 is written with a one, the pause quantum will be zero.
After transmission, a pause frame transmitted interrupt will be generated (bit 14 of the Interrupt Status Register) and the only statistics register that will be incremented will be the Pause Frames Transmitted Register. Pause frames can also be transmitted by the MAC using normal frame transmission methods.
34.5.17 MAC PFC Priority-based Pause Frame Support

The start of a Priority-based Flow Control (PFC) pause frame is as follows:
Table 34-6. Start of a PFC Pause Frame Source Address 6 bytes Type (Mac Control Frame) 0x8808 Pause Opcode 0x1001 Priority Enable Vector 2 bytes Pause Time 8*2 bytes
Destination Address 0x0180C2000001
The GMAC supports PFC priority-based pause transmission and reception. Before PFC pause frames can be received, bit 16 of the Network Control Register must be set. Note: Refer to the 802.1Qbb standard for a full description of priority-based pause operation.
34.5.17.1 PFC Pause Frame Reception The ability to receive and decode priority-based pause frames is enabled by setting bit 16 of the Network Control Register. When this bit is set, the GMAC will match either classic 802.3 pause frames or PFC priority-based pause frames. Once a priority-based pause frame has been received and matched, then from that moment on the GMAC will only match on priority-based pause frames (this is a 802.1Qbb requirement, known as PFC negotiation). Once prioritybased pause has been negotiated, any received 802.3x format pause frames will not be acted upon. If a valid priority-based pause frame is received then the GMAC will decode the frame and determine which, if any, of the 8 priorities require to be paused. Up to 8 Pause Time Registers are then updated with the 8 pause times extracted from the frame regardless of whether a previous pause operation is active or not. An interrupt (either bit 12 or bit 13 of the Interrupt Status Register) is triggered when a pause frame is received, but only if the interrupt has been enabled (bit 12 and bit 13 of the Interrupt Mask Register). Pause frames received with non zero quantum are indicated through the interrupt bit 12 of the Interrupt Status Register. Pause frames received with zero quantum are indicated on bit 13 of the Interrupt Status Register. The loading of a new pause time only occurs when the GMAC is configured for full duplex operation. If the GMAC is configured for half duplex, the pause time counters will not be loaded, but the pause frame received interrupt will still be triggered. A valid pause frame is defined as having a destination address that matches either the address stored in Specific Address Register 1 or if it matches the reserved address of 0x0180C2000001. It must also have the MAC control frame type ID of 0x8808 and have the pause opcode of 0x0101. Pause frames that have frame check sequence (FCS) or other errors will be treated as invalid and will be discarded. Valid pause frames received will increment the Pause Frames Received Statistic Register. The Pause Time Registers decrement every 512 bit times immediately following the PFC frame reception. For test purposes, the retry test bit can be set (bit 12 in the Network Configuration Register) which causes the Pause Time Register to decrement every GRXCK cycle once transmission has stopped. The interrupt (bit 13 in the Interrupt Status Register) is asserted whenever the Pause Time Register decrements to zero (assuming it has been enabled by bit 13 in the Interrupt Mask Register). This interrupt is also set when a zero quantum pause frame is received.

849
34.5.17.2 PFC Pause Frame Transmission Automatic transmission of pause frames is supported through the transmit priority-based pause frame bit of the Network Control Register. If bit 17 of the Network Control Register is written with logic 1, a PFC pause frame will be transmitted providing full duplex is selected in the Network Configuration Register and the transmit block is enabled in the Network Control Register. When bit 17 of the Network Control Register is set, the fields of the priority-based pause frame will be built using the values stored in the Transmit PFC Pause Register. Pause frame transmission will happen immediately if transmit is inactive or if transmit is active between the current frame and the next frame due to be transmitted. Transmitted pause frames comprise the following:
z z z z z z z z
A destination address of 01-80-C2-00-00-01 A source address taken from Specific Address Register 1 A type ID of 88-08 (MAC control frame) A pause opcode of 01-01 A priority enable vector taken from Transmit PFC Pause Register 8 pause quantum registers Fill of 00 to take the frame to minimum frame length Valid FCS If bit 17 of the Network Control Register is written with a one, then the priority enable vector of the priority-based pause frame will be set equal to the value stored in the Transmit PFC Pause Register [7:0]. For each entry equal to zero in the Transmit PFC Pause Register[15:8], the pause quantum field of the pause frame associated with that entry will be taken from the transmit pause quantum register. For each entry equal to one in the Transmit PFC Pause Register[15:8], the pause quantum associated with that entry will be zero. The Transmit Pause Quantum Register resets to a value of 0xFFFF giving maximum pause quantum as default.
The pause quantum registers used in the generated frame will depend on the trigger source for the frame as follows:
z
After transmission, a pause frame transmitted interrupt will be generated (bit 14 of the Interrupt Status Register) and the only statistics register that will be incremented will be the Pause Frames Transmitted Register. PFC Pause frames can also be transmitted by the MAC using normal frame transmission methods.
34.5.18 PHY Interface

Different PHY interfaces are supported by the Gigabit Ethernet MAC:
z z z
GMII MII RGMII
The GMII and RGMII should only be used for 1000 Mbps operation. The MII interface is provided for 10/100 operation and uses txd[3:0] and rxd[3:0].
34.5.19 10/100/1000 Operation

The gigabit select bit in the Network Configuration Register selects between 10/100 Mbps Ethernet operation and 1000 Mbps mode. The 10/100 Mbps speed bit in the Network Configuration Register is used to select between 10 Mbps and 100 Mbps.
34.5.20 Jumbo Frames

The jumbo frames enable bit in the Network Configuration Register allows the GMAC, in its default configuration, to receive jumbo frames up to 10240 bytes in size. This operation does not form part of the IEEE 802.3 specification and is normally disabled. When jumbo frames are enabled, frames received with a frame size greater than 10240 bytes are discarded.

850
34.6
Programming Interface
34.6.1 Initialization
34.6.1.1 Configuration Initialization of the GMAC configuration (e.g. loop back mode, frequency ratios) must be done while the transmit and receive circuits are disabled. See the description of the Network Control Register and Network Configuration Register earlier in this document. To change loop back mode, the following sequence of operations must be followed: 1. 2. 3. Note: Write to Network Control Register to disable transmit and receive circuits. Write to Network Control Register to change loop back mode. Write to Network Control Register to re-enable transmit or receive circuits. These writes to Network Control Register cannot be combined in any way.
34.6.1.2 Receive Buffer List Receive data is written to areas of data (i.e. buffers) in system memory. These buffers are listed in another data structure that also resides in main memory. This data structure (receive buffer queue) is a sequence of descriptor entries as defined in Table 34-2, Receive Buffer Descriptor Entry The Receive Buffer Queue Pointer Register points to this data structure.
Figure 34-3. Receive Buffer List
Receive Buffer 0 Receive Buffer Queue Pointer (MAC Register) Receive Buffer 1
Receive Buffer N Receive Buffer Descriptor List (In memory) (In memory)
To create the list of buffers: 1. 2. 3. 4. 5. Allocate a number (N) of buffers of X bytes in system memory, where X is the DMA buffer length programmed in the DMA Configuration Register. Allocate an area 8N bytes for the receive buffer descriptor list in system memory and create N entries in this list. Mark all entries in this list as owned by GMAC, i.e., bit 0 of word 0 set to 0. Mark the last descriptor in the queue with the wrap bit (bit 1 in word 0 set to 1). Write address of receive buffer descriptor list and control information to GMAC register receive buffer queue pointer The receive circuits can then be enabled by writing to the address recognition registers and the Network Control Register.

851
34.6.1.3 Transmit Buffer List Transmit data is read from areas of data (the buffers) in system memory. These buffers are listed in another data structure that also resides in main memory. This data structure (Transmit Buffer Queue) is a sequence of descriptor entries as defined in Table 34-3, Transmit Buffer Descriptor Entry. The Transmit Buffer Queue Pointer Register points to this data structure. To create this list of buffers: 1. 2. 3. 4. 5. Allocate a number (N) of buffers of between 1 and 2047 bytes of data to be transmitted in system memory. Up to 128 buffers per frame are allowed. Allocate an area 8N bytes for the transmit buffer descriptor list in system memory and create N entries in this list. Mark all entries in this list as owned by GMAC, i.e., bit 31 of word 1 set to 0. Mark the last descriptor in the queue with the wrap bit (bit 30 in word 1 set to 1). Write address of transmit buffer descriptor list and control information to GMAC register transmit buffer queue pointer. The transmit circuits can then be enabled by writing to the Network Control Register.
34.6.1.4 Address Matching The GMAC register pair hash address and the four Specific Address Register-pairs must be written with the required values. Each register pair comprises of a bottom register and top register with the bottom register being written first. The address matching is disabled for a particular register pair after the bottom register has been written and re-enabled when the top register is written. Each register pair may be written at any time, regardless of whether the receive circuits are enabled or disabled. As an example, to set Specific Address Register 1 to recognize destination address 21:43:65:87:A9:CB, the following values are written to Specific Address Register 1 bottom and Specific Address Register 1 top:
z z
Specific Address Register 1 bottom bits 31:0 (0x98): 8765_4321 hex. Specific Address Register 1 top bits 31:0 (0x9C): 0000_cba9 hex.
34.6.1.5 PHY Maintenance The PHY Maintenance Register is implemented as a shift register. Writing to the register starts a shift operation which is signalled as complete when bit two is set in the Network Status Register (about 2000 MCK cycles later when bits [18:16] are set to 010 in the Network Configuration Register). An interrupt is generated as this bit is set. During this time, the MSB of the register is output on the MDIO pin and the LSB updated from the MDIO pin with each Management Data Clock (MDC) cycle. This causes the transmission of a PHY management frame on MDIO. See section 22.2.4.5 of the IEEE 802.3 standard. Reading during the shift operation will return the current contents of the shift register. At the end of the management operation the bits will have shifted back to their original locations. For a read operation the data bits are updated with data read from the PHY. It is important to write the correct values to the register to ensure a valid PHY management frame is produced. The Management Data Clock (MDC) should not toggle faster than 2.5 MHz (minimum period of 400 ns), as defined by the IEEE 802.3 standard. MDC is generated by dividing down MCK. Three bits in the Network Configuration Register determine by how much MCK should be divided to produce MDC. 34.6.1.6 Interrupts There are 18 interrupt conditions that are detected within the GMAC. These are ORed to make a single interrupt (or multiple interrupts if priority queuing is enabled). Depending on the overall system design this may be passed through a further level of interrupt collection (interrupt controller). On receipt of the interrupt signal, the CPU enters the interrupt handler. Refer to the device interrupt controller documentation to identify that it is the GMAC that is generating the interrupt. To ascertain which interrupt, read the Interrupt Status Register. Note that in the default configuration this register will clear itself after being read, though this may be configured to be write-one-to-clear if desired.

852
At reset all interrupts are disabled. To enable an interrupt, write to Interrupt Enable Register with the pertinent interrupt bit set to 1. To disable an interrupt, write to Interrupt Disable Register with the pertinent interrupt bit set to 1. To check whether an interrupt is enabled or disabled, read Interrupt Mask Register. If the bit is set to 1, the interrupt is disabled. 34.6.1.7 Transmitting Frames To set up a frame for transmission: 1. 2. 3. 4. 5. 6. 7. Enable transmit in the Network Control Register. Allocate an area of system memory for transmit data. This does not have to be contiguous, varying byte lengths can be used if they conclude on byte borders. Set-up the transmit buffer list by writing buffer addresses to word zero of the transmit buffer descriptor entries and control and length to word one. Write data for transmission into the buffers pointed to by the descriptors. Write the address of the first buffer descriptor to transmit buffer descriptor queue pointer. Enable appropriate interrupts. Write to the transmit start bit (TSTART) in the Network Control Register.
34.6.1.8 Receiving Frames When a frame is received and the receive circuits are enabled, the GMAC checks the address and, in the following cases, the frame is written to system memory:
z z z z z
If it matches one of the four Specific Address Registers. If it matches one of the four type ID registers. If it matches the hash address function. If it is a broadcast address (0xFFFFFFFFFFFF) and broadcasts are allowed. If the GMAC is configured to copy all frames.
The register receive buffer queue pointer points to the next entry in the receive buffer descriptor list and the GMAC uses this as the address in system memory to write the frame to. Once the frame has been completely and successfully received and written to system memory, the GMAC then updates the receive buffer descriptor entry (see Table 34-2, Receive Buffer Descriptor Entry) with the reason for the address match and marks the area as being owned by software. Once this is complete, a receive complete interrupt is set. Software is then responsible for copying the data to the application area and releasing the buffer (by writing the ownership bit back to 0). If the GMAC is unable to write the data at a rate to match the incoming frame, then a receive overrun interrupt is set. If there is no receive buffer available, i.e., the next buffer is still owned by software, a receive buffer not available interrupt is set. If the frame is not successfully received, a Statistics Register is incremented and the frame is discarded without informing software.

853
34.6.2 Statistics Registers

Statistics Registers are described in the User Interface beginning with Section 34.7.39 on page 899 and ending with Section 34.7.83 on page 927. The Statistics Register Block begins at 0x100 and runs to 0x1B0, and is comprised of the registers listed below.
Octets Transmitted [31:0] Register Octets Transmitted [47:32] Register Frames Transmitted Register Broadcast Frames Transmitted Register Multicast Frames Transmitted Register Pause Frames Transmitted Register 64 Byte Frames Transmitted Register 65 to 127 Byte Frames Transmitted Register 128 to 255 Byte Frames Transmitted Register 256 to 511 Byte Frames Transmitted Register 512 to 1023 Byte Frames Transmitted Register 1024 to 1518 Byte Frames Transmitted Register Greater Than 1518 Byte Frames Transmitted Register Transmit Under Runs Register Single Collision Frames Register Multiple Collision Frames Register Excessive Collisions Register Late Collisions Register Deferred Transmission Frames Register Carrier Sense Errors Register Octets Received [31:0] Received Octets Received [47:32] Received Frames Received Register Broadcast Frames Received Register Multicast Frames Received Register Pause Frames Received Register 64 Byte Frames Received Register 65 to 127 Byte Frames Received Register 128 to 255 Byte Frames Received Register 256 to 511Byte Frames Received Register 512 to 1023 Byte Frames Received Register 1024 to 1518 Byte Frames Received Register 1519 to Maximum Byte Frames Received Register Undersize Frames Received Register Oversize Frames Received Register Jabbers Received Register Frame Check Sequence Errors Register Length Field Frame Errors Register Receive Symbol Errors Register Alignment Errors Register Receive Resource Errors Register Receive Overrun Register IP Header Checksum Errors Register TCP Checksum Errors Register UDP Checksum Errors Register
These registers reset to zero on a read and stick at all ones when they count to their maximum value. They should be read frequently enough to prevent loss of data. In order to reduce overall design area, the Statistics Registers may be optionally removed in the configuration file if they are deemed unnecessary for a particular design. The Receive Statistics Registers are only incremented when the receive enable bit is set in the Network Control Register. Once a Statistics Register has been read, it is automatically cleared. When reading the Octets Transmitted and Octets Received Registers, bits 31:0 should be read prior to bits 47:32 to ensure reliable operation.

854
34.7
Gigabit Ethernet MAC (GMAC) User Interface
Table 34-7. Register Mapping Offset 0x000 0x004 0x008 0x00C 0x010 0x014 0x018 0x01C 0x020 0x024 0x028 0x02C 0x030 0x034 0x038 0x03C 0x040 0x044 0x048-0x07C 0x080 0x084 0x088 0x08C 0x090 0x094 0x098 0x09C 0x0A0 0x0A4 0x0A8 0x0AC 0x0B0 0x0B4 0x0B8 0x0BC Register Network Control Register Network Configuration Register Network Status Register User Register DMA Configuration Register Transmit Status Register Receive Buffer Queue Base Address Transmit Buffer Queue Base Address Receive Status Register Interrupt Status Register Interrupt Enable Register Interrupt Disable Register Interrupt Mask Register PHY Maintenance Register Received Pause Quantum Register Transmit Pause Quantum Register TX Partial Store and Forward Register RX Partial Store and Forward Register Reserved Hash Register Bottom [31:0] Hash Register Top [63:32] Specific Address 1 Bottom [31:0] Register Specific Address 1 Top [47:32] Register Specific Address 2 Bottom [31:0] Register Specific Address 2 Top [47:32] Register Specific Address 3 Bottom [31:0] Register Specific Address 3 Top [47:32] Register Specific Address 4 Bottom [31:0] Register Specific Address 4 Top [47:32] Register Type ID Match 1 Register Type ID Match 2 Register Type ID Match 3 Register Type ID Match 4 Register Wake on LAN Register IPG Stretch Register Name GMAC_NCR GMAC_NCFGR GMAC_NSR GMAC_UR GMAC_DCFGR GMAC_TSR GMAC_RBQB GMAC_TBQB GMAC_RSR GMAC_ISR GMAC_IER GMAC_IDR GMAC_IMR GMAC_MAN GMAC_RPQ GMAC_TPQ GMAC_TPSF GMAC_RPSF GMAC_HRB GMAC_HRT GMAC_SAB1 GMAC_SAT1 GMAC_SAB2 GMAC_SAT2 GMAC_SAB3 GMAC_SAT3 GMAC_SAB4 GMAC_SAT4 GMAC_TIDM1 GMAC_TIDM2 GMAC_TIDM3 GMAC_TIDM4 GMAC_WOL GMAC_IPGS Access Read-write Read-write Read-only Read-write Read-write Read-write Read-write Read-write Read-write Read-only Write-only Write-only Read-only Read-write Read-only Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Reset 0x0000_0000 0x0008_0000 0b01x0 0x0000_0000 0x0002_0004 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x07FF_FFFF 0x0000_0000 0x0000_0000 0x0000_FFFF 0x0000_0FFF 0x0000_0FFF 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000

855
Table 34-7. Register Mapping (Continued) Offset 0x0C0 0x0C4 0x0C8 0x0CC 0x0FC 0x100 0x104 0x108 0x10C 0x110 0x114 0x118 0x11C 0x120 0x124 0x128 0x12C 0x130 0x134 0x138 0x13C 0x140 0x144 0x148 0x14C 0x150 0x154 0x158 0x15C 0x160 0x164 0x168 0x16C 0x170 0x174 0x178 0x17C Register Stacked VLAN Register Transmit PFC Pause Register Specific Address 1 Mask Bottom [31:0] Register Specific Address 1 Mask Top [47:32] Register Reserved Octets Transmitted [31:0] Register Octets Transmitted [47:32] Register Frames Transmitted Register Broadcast Frames Transmitted Register Multicast Frames Transmitted Register Pause Frames Transmitted Register 64 Byte Frames Transmitted Register 65 to 127 Byte Frames Transmitted Register 128 to 255 Byte Frames Transmitted Register 256 to 511 Byte Frames Transmitted Register 512 to 1023 Byte Frames Transmitted Register 1024 to 1518 Byte Frames Transmitted Register Greater Than 1518 Byte Frames Transmitted Register Transmit Under Runs Register Single Collision Frames Register Multiple Collision Frames Register Excessive Collisions Register Late Collisions Register Deferred Transmission Frames Register Carrier Sense Errors Register Octets Received [31:0] Received Octets Received [47:32] Received Frames Received Register Broadcast Frames Received Register Multicast Frames Received Register Pause Frames Received Register 64 Byte Frames Received Register 65 to 127 Byte Frames Received Register 128 to 255 Byte Frames Received Register 256 to 511Byte Frames Received Register 512 to 1023 Byte Frames Received Register 1024 to 1518 Byte Frames Received Register Name GMAC_SVLAN GMAC_TPFCP GMAC_SAMB1 GMAC_SAMT1 GMAC_OTLO GMAC_OTHI GMAC_FT GMAC_BCFT GMAC_MFT GMAC_PFT GMAC_BFT64 GMAC_TBFT127 GMAC_TBFT255 GMAC_TBFT511 GMAC_TBFT1023 GMAC_TBFT1518 GMAC_GTBFT1518 GMAC_TUR GMAC_SCF GMAC_MCF GMAC_EC GMAC_LC GMAC_DTF GMAC_CSE GMAC_ORLO GMAC_ORHI GMAC_FR GMAC_BCFR GMAC_MFR GMAC_PFR GMAC_BFR64 GMAC_TBFR127 GMAC_TBFR255 GMAC_TBFR511 GMAC_TBFR1023 GMAC_TBFR1518 Access Read-write Read-write Read-write Read-write Read-only Read-only Read-only Read-only Read-only Read-only Read-only Read-only Read-only Read-only Read-only Read-only Read-only Read-only Read-only Read-only Read-only Read-only Read-only Read-only Read-only Read-only Read-only Read-only Read-only Read-only Read-only Read-only Read-only Read-only Read-only Read-only Reset 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000

856
Table 34-7. Register Mapping (Continued) Offset 0x180 0x184 0x188 0x18C 0x190 0x194 0x198 0x19C 0x1A0 0x1A4 0x1A8 0x1AC 0x1B0 0x1C8 0x1CC 0x1D0 0x1D4 0x1D8 0x1DC 0x1E0 0x1E4 0x1E8 0x1EC 0x1F0 0x1F4 0x1F8 0x1FC 0x200-0x23C 0x280-0x298 0x400 ... 0x418 0x440 ... 0x458 0x480 Register 1519 to Maximum Byte Frames Received Register Undersize Frames Received Register Oversize Frames Received Register Jabbers Received Register Frame Check Sequence Errors Register Length Field Frame Errors Register Receive Symbol Errors Register Alignment Errors Register Receive Resource Errors Register Receive Overrun Register IP Header Checksum Errors Register TCP Checksum Errors Register UDP Checksum Errors Register 1588 Timer Sync Strobe Seconds Register 1588 Timer Sync Strobe Nanoseconds Register 1588 Timer Seconds Register 1588 Timer Nanoseconds Register 1588 Timer Adjust Register 1588 Timer Increment Register PTP Event Frame Transmitted Seconds PTP Event Frame Transmitted Nanoseconds PTP Event Frame Received Seconds PTP Event Frame Received Nanoseconds PTP Peer Event Frame Transmitted Seconds PTP Peer Event Frame Transmitted Nanoseconds PTP Peer Event Frame Received Seconds PTP Peer Event Frame Received Nanoseconds Reserved Reserved Interrupt Status Register Priority Queue 0 ... Interrupt Status Register Priority Queue 6 Transmit Buffer Queue Base Address Priority Queue 0 ... Transmit Buffer Queue Base Address Priority Queue 6 Receive Buffer Queue Base Address Priority Queue 0 Name GMAC_TMXBFR GMAC_UFR GMAC_OFR GMAC_JR GMAC_FCSE GMAC_LFFE GMAC_RSE GMAC_AE GMAC_RRE GMAC_ROE GMAC_IHCE GMAC_TCE GMAC_UCE GMAC_TSSS GMAC_TSSN GMAC_TS GMAC_TN GMAC_TA GMAC_TI GMAC_EFTS GMAC_EFTN GMAC_EFRS GMAC_EFRN GMAC_PEFTS GMAC_PEFTN GMAC_PEFRS GMAC_PEFRN GMAC_ISRPQ0 ... GMAC_ISRPQ6 GMAC_TBQBAPQ0 ... GMAC_TBQBAPQ6 GMAC_RBQBAPQ0 Access Read-only Read-only Read-only Read-only Read-only Read-only Read-only Read-only Read-only Read-only Read-only Read-only Read-only Read-write Read-write Read-write Read-write Write-only Read-write Read-only Read-only Read-only Read-only Read-only Read-only Read-only Read-only Read-only ... Read-only Read-write ... Read-write Read-write 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 ... 0x0000_0000 0x0000_0000 ... 0x0000_0000 0x0000_0000 Reset 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000

857
Table 34-7. Register Mapping (Continued) Offset ... 0x498 0x4a0 ... 0x4b8 0x500 ... 0x53c 0x540 ... 0x57c 0x600 ... 0x618 0x620 ... 0x638 0x640 ... 0x658 Register ... Receive Buffer Queue Base Address Priority Queue 6 Receive Buffer Size Register Priority Queue 0 ... Receive Buffer Size Register Priority Queue 6 Screening Type1 Register Priority Queue 0 ... Screening Type1 Register Priority Queue 15 Screening Type2 Register Priority Queue 0 ... Screening Type2 Register Priority Queue 15 Interrupt Enable Register Priority Queue 0 ... Interrupt Enable Register Priority Queue 6 Interrupt Disable Register Priority Queue 0 ... Interrupt Disable Register Priority Queue 6 Interrupt Mask Register Priority Queue 0 ... Interrupt Mask Register Priority Queue 6 Name ... GMAC_RBQBAPQ6 GMAC_RBSRPQ0 ... GMAC_RBSRPQ6 GMAC_ST1RPQ0 ... GMAC_ST1RPQ15 GMAC_ST2RPQ0 ... GMAC_ST2RPQ15 GMAC_IERPQ0 ... GMAC_IERPQ6 GMAC_IDRPQ0 ... GMAC_IDRPQ6 GMAC_IMRPQ0 ... GMAC_IMRPQ6 Access ... Read-write Read-write ... Read-write Read-write ... Read-write Read-write ... Read-write Write-only ... Write-only Write-only ... Write-only Read-write ... Read-write Reset ... 0x0000_0000 0x0000_0002 ... 0x0000_0002 0x0000_0000 ... 0x0000_0000 0x0000_0000 ... 0x0000_0000 0x0000_0000 ... 0x0000_0000 0x0000_0000 ... 0x0000_0000 0x0000_0000 ... 0x0000_0000

858
34.7.1 Network Control Register

31
GMAC_NCR 0xF0028000 Read-write

30 29 28 27 26 25 24
23
22
21
20
19
18 FNP 10 THALT 2 RXEN
17 TXPBPF 9 TSTART 1 LBL
16 ENPBPR 8 BP 0
15 SRTSM 7 WESTAT
14 RDS 6 INCSTAT
13 5 CLRSTAT
12 TXZQPF 4 MPE
11 TXPF 3 TXEN
LBL: Loop Back Local

Connects GTX to GRX, GTXEN to GRXDV and forces full duplex mode. GRXCK and GTXCK may malfunction as the GMAC is switched into and out of internal loop back. It is important that receive and transmit circuits have already been disabled when making the switch into and out of internal loop back.
RXEN: Receive Enable

When set, RXEN enables the GMAC to receive data. When reset frame reception stops immediately and the receive pipeline will be cleared. The Receive Queue Pointer Register is unaffected.
TXEN: Transmit Enable

When set, TXEN enables the GMAC transmitter to send data. When reset transmission will stop immediately, the transmit pipeline and control registers will be cleared and the Transmit Queue Pointer Register will reset to point to the start of the transmit descriptor list.
MPE: Management Port Enable

Set to one to enable the management port. When zero, forces MDIO to high impedance state and MDC low.
CLRSTAT: Clear Statistics Registers

This bit is write only. Writing a one clears the statistics registers.
INCSTAT: Increment Statistics Registers

This bit is write only. Writing a one increments all the Statistics Registers by one for test purposes.
WESTAT: Write Enable for Statistics Registers

Setting this bit to one makes the Statistics Registers writable for functional test purposes.
BP: Back pressure

If set in 10M or 100M half duplex mode, forces collisions on all received frames. Ignored in gigabit half duplex mode.
TSTART: Start Transmission

Writing one to this bit starts transmission.

859
THALT: Transmit Halt

Writing one to this bit halts transmission as soon as any ongoing frame transmission ends.
TXPF: Transmit Pause Frame

Writing one to this bit causes a pause frame to be transmitted.
TXZQPF: Transmit Zero Quantum Pause Frame

Writing one to this bit causes a pause frame with zero quantum to be transmitted.
RDS: Read Snapshot

Writing a one means that the snapshot value of the statistics register will be read back, otherwise the raw statistic register will be read. The default GMAC configuration does not support this function.
SRTSM: Store Receive Time Stamp to Memory

0 = normal operation. 1 = causes the CRC of every received frame to be replaced with the value of the nanoseconds field of the 1588 timer that was captured as the receive frame passed the message time stamp point.
ENPBPR: Enable PFC Priority-based Pause Reception

Enables PFC Priority Based Pause Reception capabilities. Setting this bit enables PFC negotiation and recognition of prioritybased pause frames.
TXPBPF: Transmit PFC Priority-based Pause Frame

Takes the values stored in the Transmit PFC Pause Register.
FNP: Flush Next Packet

Flush the next packet from the external RX DPRAM. Writing one to this bit will only have an effect if the DMA is not currently writing a packet already stored in the DPRAM to memory.

860
34.7.2 Network Configuration Register

31 23 DCPF 15 RXBUFO 7 UNIHEN 6 MTI HEN
GMAC_NCFGR 0xF0028004 Read-write

30 IRXER 22 DBW 14 13 PEN 5 NBC 12 RTY 4 CAF 29 RXBP 21 28 IPGSEN 20 27 19 CLK 11 3 JFRAME 26 IRXFCS 18 25 EFRHD 17 RFCS 9 1 FD 24 RXCOEN 16 LFERD 8 MAXFS 0 SPD
10 GBE 2 DNVLAN
SPD: Speed
Set to logic one to indicate 100 Mbps operation, logic zero for 10 Mbps.
FD: Full Duplex

If set to logic one, the transmit block ignores the state of collision and carrier sense and allows receive while transmitting.
DNVLAN: Discard Non-VLAN FRAMES

When set only VLAN tagged frames will be passed to the address matching logic.
JFRAME: Jumbo Frame Size

Set to one to enable jumbo frames up to 10240 bytes to be accepted. The default length is 10240 bytes.
CAF: Copy All Frames

When set to logic one, all valid frames will be accepted.
NBC: No Broadcast
When set to logic one, frames addressed to the broadcast address of all ones will not be accepted.
MTIHEN: Multicast Hash Enable

When set, multicast frames will be accepted when the 6-bit hash function of the destination address points to a bit that is set in the Hash Register.
UNIHEN: Unicast Hash Enable

When set, unicast frames will be accepted when the 6-bit hash function of the destination address points to a bit that is set in the Hash Register.
MAXFS: 1536 Maximum Frame Size

Setting this bit means the GMAC will accept frames up to 1536 bytes in length. Normally the GMAC would reject any frame above 1518 bytes.

861
GBE: Gigabit Mode Enable

Setting this bit configures the GMAC for 1000 Mbps operation. 0 = 10/100 operation using MII interface. 1 = Gigabit operation using GMII interface.
RTY: Retry Test

Must be set to zero for normal operation. If set to one the backoff between collisions will always be one slot time. Setting this bit to one helps test the too many retries condition. Also used in the pause frame tests to reduce the pause counter's decrement time from 512 bit times, to every GRXCK cycle.
PEN: Pause Enable

when set, transmission will pause if a non zero 802.3 classic pause frame is received and PFC has not been negotiated.
RXBUFO: Receive Buffer Offset

Indicates the number of bytes by which the received data is offset from the start of the receive buffer
LFERD: Length Field Error Frame Discard

Setting this bit causes frames with a measured length shorter than the extracted length field (as indicated by bytes 13 and 14 in a non-VLAN tagged frame) to be discarded. This only applies to frames with a length field less than 0x0600.
RFCS: Remove FCS

Setting this bit will cause received frames to be written to memory without their frame check sequence (last 4 bytes). The frame length indicated will be reduced by four bytes in this mode.
CLK: MDC CLock Division

Set according to MCK speed. These three bits determine the number MCK will be divided by to generate Management Data Clock (MDC). For conformance with the 802.3 specification, MDC must not exceed 2.5 MHz (MDC is only active during MDIO read and write operations).
Value 0 1 2 3 4 5 Name MCK_8 MCK_16 MCK_32 MCK_48 MCK_64 MCK_96 Description MCK divided by 8 (MCK up to 20 MHz) MCK divided by 16 (MCK up to 40 MHz) MCK divided by 32 (MCK up to 80 MHz) MCK divided by 48 (MCK up to 120 MHz) MCK divided by 64 (MCK up to 160 MHz) MCK divided by 96 (MCK up to 240 MHz)
DBW: Data Bus Width

The default value for this register is 64 bits.
Value 0 1 Name DBW32 DBW64 Description 32-bit data bus width 64-bit data bus width

862
DCPF: Disable Copy of Pause Frames

Set to one to prevent valid pause frames being copied to memory. When set, pause frames are not copied to memory regardless of the state of the copy all frames bit, whether a hash match is found or whether a type ID match is identified. If a destination address match is found, the pause frame will be copied to memory. Note that valid pause frames received will still increment pause statistics and pause the transmission of frames as required.
RXCOEN: Receive Checksum Offload Enable

When set, the receive checksum engine is enabled. Frames with bad IP, TCP or UDP checksums are discarded.
EFRHD: Enable Frames Received in Half Duplex

Enable frames to be received in half-duplex mode while transmitting.
IRXFCS: Ignore RX FCS

When set, frames with FCS/CRC errors will not be rejected. FCS error statistics will still be collected for frames with bad FCS and FCS status will be recorded in frame's DMA descriptor. For normal operation this bit must be set to zero.
IPGSEN: IP Stretch Enable

When set, the transmit IPG can be increased above 96 bit times depending on the previous frame length using the IPG Stretch Register.
RXBP: Receive Bad Preamble

When set, frames with non-standard preamble are not rejected.
IRXER: Ignore IPG GRXER

When set, GRXER has no effect on the GMAC's operation when GRXDV is low. Set this bit when using the RGMII wrapper in half-duplex mode.

863
34.7.3 Network Status Register

31
GMAC_NSR 0xF0028008 Read-only

30 29 28 27 26 25 24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
2 IDLE
1 MDIO
MDIO: MDIO Input Status

Returns status of the GMDIO pin.
IDLE: PHY Management Logic Idle

The PHY management logic is idle (i.e. has completed).

864
34.7.4 User Register

31 23 15 7
GMAC_UR 0xF002800C Read-write

30 22 14 6 29 21 13 5 28 20 12 4 27 19 11 3 26 18 10 2 25 17 9 1 24 16 8 0 RGMII
RGMII: Reduced GMII Mode

0: GMII mode is selected (default). 1: RGMII mode is selected.

865
34.7.5 DMA Configuration Register

31
GMAC_DCFGR 0xF0028010 Read-write

30 29 28 20 DRBS 27 26 25 24 DDRP 16
23
22
21
19
18
17
15 7 ESPA
14 6 ESMA
13 5
12 4
11 TXCOEN 3
10 TXPBMS 2 FBLDO
9 RXBMS 1
FBLDO: Fixed Burst Length for DMA Data Operations:

Selects the burst length to attempt to use on the AHB when transferring frame data. Not used for DMA management operations and only used where space and data size allow. Otherwise SINGLE type AHB transfers are used. Upper bits become non-writable if the configured DMA TX and RX FIFO sizes are smaller than required to support the selected burst size. One-hot priority encoding enforced automatically on register writes as follows, where 'x' represents don't care:
Value 0 1 2 4 8 16 Name SINGLE INCR4 INCR8 INCR16 Description Reserved 00001: Always use SINGLE AHB bursts Reserved 001xx: Attempt to use INCR4 AHB bursts (Default) 01xxx: Attempt to use INCR8 AHB bursts 1xxxx: Attempt to use INCR16 AHB bursts
ESMA: Endian Swap Mode Enable for Management Descriptor Accesses

When set, selects swapped endianism for AHB transfers. When clear, selects little endian mode.
ESPA: Endian Swap Mode Enable for Packet Data Accesses

When set, selects swapped endianism for AHB transfers. When clear, selects little endian mode.
RXBMS: Receiver Packet Buffer Memory Size Select

The default receive packet buffer size is 4 Kbytes. The table below shows how to configure this memory to FULL, HALF, QUARTER or EIGHTH of the default size.
Value 0 1 2 3 Name EIGHTH QUARTER HALF FULL Description
1/2 Kbyte Memory Size 1Kbyte Memory Size 2 Kbytes Memory Size 4 Kbytes Memory Size

866
TXPBMS: Transmitter Packet Buffer Memory Size Select

Having this bit at zero halves the amount of memory used for the transmit packet buffer. This reduces the amount of memory used by the GMAC. It is important to set this bit to one if the full configured physical memory is available. The value in brackets below represents the size that would result for the default maximum configured memory size of 4 Kbytes. 0 = Do not use top address bit (2 Kbytes). 1 = Use full configured addressable space (4 Kbytes).
TXCOEN: Transmitter Checksum Generation Offload Enable

Transmitter IP, TCP and UDP checksum generation offload enable. When set, the transmitter checksum generation engine is enabled to calculate and substitute checksums for transmit frames. When clear, frame data is unaffected.
DRBS: DMA Receive Buffer Size

DMA receive buffer size in AHB system memory. The value defined by these bits determines the size of buffer to use in main AHB system memory when writing received data. The value is defined in multiples of 64 bytes, thus a value of 0x01 corresponds to buffers of 64 bytes, 0x02 corresponds to 128 bytes etc.
For example: 0x02: 128 bytes 0x18: 1536 bytes (1*max length frame/buffer) 0xA0: 10240 bytes (1*10K jumbo frame/buffer) Note that this value should never be written as zero. DDRP: DMA Discard Receive Packets
When set, the GMAC DMA will automatically discard receive packets from the receiver packet buffer memory when no AHB resource is available. When low, the received packets will remain to be stored in the SRAM based packet buffer until AHB buffer resource next becomes available. A write to this bit is ignored if the DMA is not configured in the packet buffer full store and forward mode.

867
34.7.6 Transmit Status Register

31
GMAC_TSR 0xF0028014 Read-write

30 29 28 27 26 25 24
23
22
21
20
19
18
17
16
15
14
13
12 4 TFC
11
10
8 HRESP 0 UBR
7 LCO
6 UND
5 TXCOMP
3 TXGO
2 RLE
1 COL
UBR: Used Bit Read

Set when a transmit buffer descriptor is read with its used bit set. Cleared by writing a one to this bit.
COL: Collision Occurred

Set by the assertion of collision. Cleared by writing a one to this bit. When operating in 10/100 mode, this status indicates either a collision or a late collision. In gigabit mode, this status is not set for a late collision.
RLE: Retry Limit Exceeded

Cleared by writing a one to this bit.
TXGO: Transmit Go
Transmit go, if high transmit is active. When using FIFO, this bit represents bit 3 of the Network Control Register. When using the DMA interface this bit represents the TXGO variable as specified in the transmit buffer description.
TFC: Transmit Frame Corruption due to AHB error

Transmit frame corruption due to AHB error. Set if an error occurs while midway through reading transmit frame from the AHB, including HRESP errors and buffers exhausted mid frame (if the buffers run out during transmission of a frame then transmission stops, FCS shall be bad and GTXER asserted). Also set in DMA packet buffer mode if single frame is too large for configured packet buffer memory size. Cleared by writing a one to this bit.
TXCOMP: Transmit Complete

Set when a frame has been transmitted. Cleared by writing a one to this bit.
UND: Transmit Under Run

This bit is set if the transmitter was forced to terminate a frame that it had already began transmitting due to further data being unavailable. This bit is set if a transmitter status write back has not completed when another status write back is attempted. When using the DMA interface configured for internal FIFO mode, this bit is also set when the transmit DMA has written the SOP data into the FIFO and either the AHB bus was not granted in time for further data, or because an AHB not OK response was returned, or because a used bit was read. When using the DMA interface configured for packet buffer mode, this bit will never be set.

868
LCO: Late Collision Occurred

Only set if the condition occurs in gigabit mode, as retry is not attempted. Cleared by writing a one to this bit.
HRESP: HRESP Not OK

Set when the DMA block sees HRESP not OK. Cleared by writing a one to this bit.

869
34.7.7 Receive Buffer Queue Base Address Register

31
GMAC_RBQB 0xF0028018 Read-write

30 29 28 ADDR 27 26 25 24
23
22
21
20 ADDR
19
18
17
16
15
14
13
12 ADDR
11
10
5 ADDR
This register holds the start address of the receive buffer queue (receive buffers descriptor list). The receive buffer queue base address must be initialized before receive is enabled through bit 2 of the Network Control Register. Once reception is enabled, any write to the Receive Buffer Queue Base Address Register is ignored. Reading this register returns the location of the descriptor currently being accessed. This value increments as buffers are used. Software should not use this register for determining where to remove received frames from the queue as it constantly changes as new frames are received. Software should instead work its way through the buffer descriptor queue checking the used bits. In terms of AMBA AHB operation, the descriptors are read from memory using a single 32-bit AHB access. When the datapath is configured at 64 bits, the descriptors should be aligned at 64-bit boundaries and each pair of 32-bit descriptors is written to by using a single AHB access. The descriptors should be aligned at 32-bit boundaries and the descriptors are written to using two individual non sequential accesses for 32-bit datapaths.
ADDR: Receive buffer queue base address

Written with the address of the start of the receive queue.

870
34.7.8 Transmit Buffer Queue Base Address Register

31
GMAC_TBQB 0xF002801C Read-write

30 29 28 ADDR 27 26 25 24
23
22
21
20 ADDR
19
18
17
16
15
14
13
12 ADDR
11
10
5 ADDR
This register holds the start address of the transmit buffer queue (transmit buffers descriptor list). The Transmit Buffer Queue Base Address Register must be initialized before transmit is started through bit 9 of the Network Control Register. Once transmission has started, any write to the Transmit Buffer Queue Base Address Register is illegal and therefore ignored. Note that due to clock boundary synchronization, it takes a maximum of four MCK cycles from the writing of the transmit start bit before the transmitter is active. Writing to the Transmit Buffer Queue Base Address Register during this time may produce unpredictable results. Reading this register returns the location of the descriptor currently being accessed. Since the DMA handles two frames at once, this may not necessarily be pointing to the current frame being transmitted. In terms of AMBA AHB operation, the descriptors are written to memory using a single 32-bit AHB access. When the datapath is configured at 64 bits, the descriptors should be aligned at 64-bit boundaries and each pair of 32-bit descriptors is read from memory using a single AHB access. The descriptors should be aligned at 32-bit boundaries and the descriptors are read from memory using two individual non sequential accesses for 32-bit datapaths.
ADDR: Transmit Buffer Queue Base Address

Written with the address of the start of the transmit queue.

871
34.7.9 Receive Status Register

31
GMAC_RSR 0xF0028020 Read-write

30 29 28 27 26 25 24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
3 HNO
2 RXOVR
1 REC
0 BNA
This register, when read provides details of the status of a receive. Once read, individual bits may be cleared by writing 1 to them. It is not possible to set a bit to 1 by writing to the register.
BNA: Buffer Not Available

An attempt was made to get a new buffer and the pointer indicated that it was owned by the processor. The DMA will reread the pointer each time an end of frame is received until a valid pointer is found. This bit is set following each descriptor read attempt that fails, even if consecutive pointers are unsuccessful and software has in the mean time cleared the status flag. Cleared by writing a one to this bit.
REC: Frame Received

One or more frames have been received and placed in memory. Cleared by writing a one to this bit.
RXOVR: Receive Overrun

This bit is set if RX FIFO is not able to store the receive frame due to a FIFO overflow, or if the receive status was not taken at the end of the frame. This bit is also set in DMA packet buffer mode if the packet buffer overflows. For DMA operation, the buffer will be recovered if an overrun occurs. This bit is cleared when set to 1.
HNO: HRESP Not OK

Set when the DMA block sees HRESP not OK. This bit is cleared when set to 1.

872
34.7.10 Interrupt Status Register

31 23 PDRSFR 15 7 TCOMP
GMAC_ISR 0xF0028024 Read-only

30 22 PDRQFR 14 PFTR 6 TFC 29 21 SFT 13 PTZ 5 RLEX 28 WOL 20 DRQFT 12 PFNZ 4 TUR 27 19 SFR 11 HRESP 3 TXUBR 26 SRI 18 DRQFR 10 ROVR 2 RXUBR 25 PDRSFT 17 9 1 RCOMP 24 PDRQFT 16 8 0 MFS
This register indicates the source of the interrupt. In order the bits of this register are read to 1, the corresponding interrupt source must be enabled in the mask register. If any bit is set in this register, the GMAC interrupt signal will be asserted in the system.
MFS: Management Frame Sent

The PHY Maintenance Register has completed its operation. Cleared on read.
RCOMP: Receive Complete

A frame has been stored in memory. Cleared on read.
RXUBR: RX Used Bit Read

Set when a receive buffer descriptor is read with its used bit set. Cleared on read.
TXUBR: TX Used Bit Read

Set when a transmit buffer descriptor is read with its used bit set. Cleared on read.
TUR: Transmit Under Run

This interrupt is set if the transmitter was forced to terminate a frame that it has already began transmitting due to further data being unavailable. This interrupt is set if a transmitter status write back has not completed when another status write back is attempted. This interrupt is also set when the transmit DMA has written the SOP data into the FIFO and either the AHB bus was not granted in time for further data, or because an AHB not OK response was returned, or because the used bit was read.
RLEX: Retry Limit Exceeded or Late Collision

Transmit error. Late collision will only cause this status bit to be set in gigabit mode, as a retry is not attempted. Cleared on read.

Transmit frame corruption due to AHB error. Set if an error occurs while midway through reading transmit frame from the AHB, including HRESP errors and buffers exhausted mid frame.
TCOMP: Transmit Complete

Set when a frame has been transmitted. Cleared on read.

873
ROVR: Receive Overrun

Set when the receive overrun status bit is set. Cleared on read.
HRESP: HRESP Not OK

Set when the DMA block sees HRESP not OK. Cleared on read.
PFNZ: Pause Frame with Non-zero Pause Quantum Received

Indicates a valid pause has been received that has a non-zero pause quantum field. Cleared on read.
PTZ: Pause Time Zero

Set when either the Pause Time Register at address 0x38 decrements to zero, or when a valid pause frame is received with a zero pause quantum field. Cleared on read.
PFTR: Pause Frame Transmitted

Indicates a pause frame has been successfully transmitted after being initiated from the Network Control Register. Cleared on read.
DRQFR: PTP Delay Request Frame Received

Indicates a PTP delay_req frame has been received. Cleared on read.
SFR: PTP Sync Frame Received

Indicates a PTP sync frame has been received. Cleared on read.
DRQFT: PTP Delay Request Frame Transmitted

Indicates a PTP delay_req frame has been transmitted. Cleared on read.
SFT: PTP Sync Frame Transmitted

Indicates a PTP sync frame has been transmitted. Cleared on read.
PDRQFR: PDelay Request Frame Received

Indicates a PTP pdelay_req frame has been received. Cleared on read.
PDRSFR: PDelay Response Frame Received

Indicates a PTP pdelay_resp frame has been received. Cleared on read.
PDRQFT: PDelay Request Frame Transmitted

Indicates a PTP pdelay_req frame has been transmitted. Cleared on read.
PDRSFT: PDelay Response Frame Transmitted

Indicates a PTP pdelay_resp frame has been transmitted. Cleared on read.
SRI: TSU Seconds Register Increment

Indicates the register has incremented. Cleared on read.
WOL: Wake On LAN

WOL interrupt. Indicates a WOL event has been received.

874
34.7.11 Interrupt Enable Register

31 23 PDRSFR 15 EXINT 7 TCOMP
GMAC_IER 0xF0028028 Write-only

At reset all interrupts are disabled. Writing a one to the relevant bit location enables the required interrupt. This register is write only and when read will return zero.

Enable management done interrupt.

Enable receive complete interrupt.

Enable receive used bit read interrupt.

Enable transmit used bit read interrupt.

Enable transmit buffer under run interrupt.

Enable retry limit exceeded or late collision interrupt.

Enable transmit frame corruption due to AHB error interrupt.

Enable transmit complete interrupt.

Enable receive overrun interrupt.
HRESP: HRESP Not OK

Enable HRESP not OK interrupt.

875

Enable pause frame with non-zero pause quantum interrupt.

Enable pause time zero interrupt.

Enable pause frame transmitted interrupt.
EXINT: External Interrupt

Enable external interrupt.

Enable PTP delay_req frame received.

Enable PTP sync frame received.

Enable PTP delay_req frame transmitted.

Enable PTP sync frame transmitted.

Enable PTP pdelay_req frame received

Enable PTP pdelay_resp frame received.

Enable PTP pdelay_req frame transmitted.

Enable PTP pdelay_resp frame transmitted.

Enable TSU Seconds Register increment.
WOL: Wake On LAN

Enable WOL event received interrupt.

876
34.7.12 Interrupt Disable Register

GMAC_IDR 0xF002802C Write-only

Writing a 1 to the relevant bit location disables that particular interrupt. This register is write only and when read will return zero.

Disable management done interrupt.

Disable receive complete interrupt.

Disable receive used bit read interrupt.

Disable transmit used bit read interrupt.

Disable transmit buffer under run interrupt.

Disable retry limit exceeded or late collision interrupt.

Disable transmit frame corruption due to AHB error interrupt.

Disable transmit complete interrupt.

Disable receive overrun interrupt.
HRESP: HRESP Not OK

Disable HRESP not OK interrupt.

877

Disable pause frame with non-zero pause quantum interrupt.

Disable pause time zero interrupt.

Disable pause frame transmitted interrupt.

Disable external interrupt.

Disable PTP delay_req frame received.

Disable PTP sync frame received.

Disable PTP delay_req frame transmitted.

Disable PTP sync frame transmitted.

Disable PTP pdelay_req frame received

Disable PTP pdelay_resp frame received.

Disable PTP pdelay_req frame transmitted.

Disable PTP pdelay_resp frame transmitted.

Disable TSU Seconds Register increment.
WOL: Wake On LAN

Disable WOL event received interrupt.

878
34.7.13 Interrupt Mask Register

GMAC_IMR 0xF0028030 Read-only

30 22 PDRQFR 14 PFTR 6 TFC 29 21 SFT 13 PTZ 5 RLEX 28 20 DRQFT 12 PFNZ 4 TUR 27 19 SFR 11 HRESP 3 TXUBR 26 18 DRQFR 10 ROVR 2 RXUBR 25 PDRSFT 17 9 1 RCOMP 24 PDRQFT 16 8 0 MFS
The Interrupt Mask Register is a read-only register indicating which interrupts are masked. All bits are set at reset and can be reset individually by writing to the Interrupt Enable Register or set individually by writing to the Interrupt Disable Register. Having separate address locations for enable and disable saves the need for performing a read modify write when updating the Interrupt Mask Register. For test purposes there is a write-only function to this register that allows the bits in the Interrupt Status Register to be set or cleared, regardless of the state of the mask register.

A read of this register returns the value of the management done interrupt mask. 0 = Interrupt is enabled. 1 = Interrupt is disabled. A write to this register directly affects the state of the corresponding bit in the Interrupt Status Register, causing an interrupt to be generated if a 1 is written.

A read of this register returns the value of the receive complete interrupt mask. 0 = Interrupt is enabled. 1 = Interrupt is disabled. A write to this register directly affects the state of the corresponding bit in the Interrupt Status Register, causing an interrupt to be generated if a 1 is written.

A read of this register returns the value of the receive used bit read interrupt mask. 0 = Interrupt is enabled. 1 = Interrupt is disabled. A write to this register directly affects the state of the corresponding bit in the Interrupt Status Register, causing an interrupt to be generated if a 1 is written.

A read of this register returns the value of the transmit used bit read interrupt mask. 0 = Interrupt is enabled. 1 = Interrupt is disabled.

879
A write to this register directly affects the state of the corresponding bit in the Interrupt Status Register, causing an interrupt to be generated if a 1 is written.

A read of this register returns the value of the transmit buffer under run interrupt mask. 0 = Interrupt is enabled. 1 = Interrupt is disabled. A write to this register directly affects the state of the corresponding bit in the Interrupt Status Register, causing an interrupt to be generated if a 1 is written.

A read of this register returns the value of the retry limit exceeded or late collision (gigabit mode only) interrupt mask. 0 = Interrupt is enabled. 1 = Interrupt is disabled. A write to this register directly affects the state of the corresponding bit in the Interrupt Status Register, causing an interrupt to be generated if a 1 is written.

A read of this register returns the value of the transmit frame corruption due to AHB error interrupt mask. 0 = Interrupt is enabled. 1 = Interrupt is disabled. A write to this register directly affects the state of the corresponding bit in the Interrupt Status Register, causing an interrupt to be generated if a 1 is written.

A read of this register returns the value of the transmit complete interrupt mask. 0 = Interrupt is enabled. 1 = Interrupt is disabled. A write to this register directly affects the state of the corresponding bit in the Interrupt Status Register, causing an interrupt to be generated if a 1 is written.

A read of this register returns the value of the receive overrun interrupt mask. 0 = Interrupt is enabled. 1 = Interrupt is disabled. A write to this register directly affects the state of the corresponding bit in the Interrupt Status Register, causing an interrupt to be generated if a 1 is written.
HRESP: HRESP Not OK

A read of this register returns the value of the HRESP not OK interrupt mask. 0 = Interrupt is enabled. 1 = Interrupt is disabled. A write to this register directly affects the state of the corresponding bit in the Interrupt Status Register, causing an interrupt to be generated if a 1 is written.

880

A read of this register returns the value of the pause frame with non-zero pause quantum interrupt mask. 0 = Interrupt is enabled. 1 = Interrupt is disabled. A write to this register directly affects the state of the corresponding bit in the Interrupt Status Register, causing an interrupt to be generated if a 1 is written.

A read of this register returns the value of the pause time zero interrupt mask. 0 = Interrupt is enabled. 1 = Interrupt is disabled. A write to this register directly affects the state of the corresponding bit in the Interrupt Status Register, causing an interrupt to be generated if a 1 is written.

A read of this register returns the value of the pause frame transmitted interrupt mask. 0 = Interrupt is enabled. 1 = Interrupt is disabled. A write to this register directly affects the state of the corresponding bit in the Interrupt Status Register, causing an interrupt to be generated if a 1 is written.

A read of this register returns the value of the external interrupt mask. 0 = Interrupt is enabled. 1 = Interrupt is disabled. A write to this register directly affects the state of the corresponding bit in the Interrupt Status Register, causing an interrupt to be generated if a 1 is written.

A read of this register returns the value of the PTP delay_req frame received mask. 0 = Interrupt is enabled. 1 = Interrupt is disabled. A write to this register directly affects the state of the corresponding bit in the Interrupt Status Register, causing an interrupt to be generated if a 1 is written.

A read of this register returns the value of the PTP sync frame received mask. 0 = Interrupt is enabled. 1 = Interrupt is disabled. A write to this register directly affects the state of the corresponding bit in the Interrupt Status Register, causing an interrupt to be generated if a 1 is written.

881

A read of this register returns the value of the PTP delay_req frame transmitted mask. 0 = Interrupt is enabled. 1 = Interrupt is disabled. A write to this register directly affects the state of the corresponding bit in the Interrupt Status Register, causing an interrupt to be generated if a 1 is written.

A read of this register returns the value of the PTP sync frame transmitted mask. 0 = Interrupt is enabled. 1 = Interrupt is disabled. A write to this register directly affects the state of the corresponding bit in the Interrupt Status Register, causing an interrupt to be generated if a 1 is written.

A read of this register returns the value of the PTP pdelay_req frame received mask. 0 = Interrupt is enabled. 1 = Interrupt is disabled. A write to this register directly affects the state of the corresponding bit in the Interrupt Status Register, causing an interrupt to be generated if a 1 is written.

A read of this register returns the value of the PTP pdelay_resp frame received mask. 0 = Interrupt is enabled. 1 = Interrupt is disabled. A write to this register directly affects the state of the corresponding bit in the Interrupt Status Register, causing an interrupt to be generated if a 1 is written.

A read of this register returns the value of the PTP pdelay_req frame transmitted mask. 0 = Interrupt is enabled. 1 = Interrupt is disabled. A write to this register directly affects the state of the corresponding bit in the Interrupt Status Register, causing an interrupt to be generated if a 1 is written.

A read of this register returns the value of the PTP pdelay_resp frame transmitted mask. 0 = Interrupt is enabled. 1 = Interrupt is disabled. A write to this register directly affects the state of the corresponding bit in the Interrupt Status Register, causing an interrupt to be generated if a 1 is written.

882
34.7.14 PHY Maintenance Register

31 WZO 23 PHYA 15
GMAC_MAN 0xF0028034 Read-write

30 CLTTO 22 29 OP 21 20 REGA 12 DATA 7 6 5 4 DATA 3 2 1 0 19 18 28 27 26 PHYA 17 WTN 11 10 9 8 16 25 24
14
13
The PHY Maintenance Register is implemented as a shift register. Writing to the register starts a shift operation which is signalled as complete when bit 2 is set in the Network Status Register. It takes about 2000 MCK cycles to complete, when MDC is set for MCK divide by 32 in the Network Configuration Register. An interrupt is generated upon completion. During this time, the MSB of the register is output on the MDIO pin and the LSB updated from the MDIO pin with each MDC cycle. This causes transmission of a PHY management frame on MDIO. See Section 22.2.4.5 of the IEEE 802.3 standard. Reading during the shift operation will return the current contents of the shift register. At the end of management operation, the bits will have shifted back to their original locations. For a read operation, the data bits will be updated with data read from the PHY. It is important to write the correct values to the register to ensure a valid PHY management frame is produced. The MDIO interface can read IEEE 802.3 clause 45 PHYs as well as clause 22 PHYs. To read clause 45 PHYs, bit 30 should be written with a 0 rather than a 1. For a description of MDC generation, see Section 34.7.2 Network Configuration Register.
DATA: PHY Data

For a write operation this field is written with the data to be written to the PHY. After a read operation this field contains the data read from the PHY.
WTN: Write Ten

Must be written to 10.
REGA: Register Address

Specifies the register in the PHY to access.
PHYA: PHY Address OP: Operation

10 is read. 01 is write.
CLTTO: Clause 22 Operation

Must be written to 1 for Clause 22 operation. To read clause 45 PHYs, bit 30 should be written with a 0 rather than a 1.
WZO: Write ZERO

Must be written with 0.

883
34.7.15 Receive Pause Quantum Register

31
GMAC_RPQ 0xF0028038 Read-only

30 29 28 27 26 25 24
23
22
21
20
19
18
17
16
15
14
13
12 RPQ
11
10
4 RPQ
RPQ: Received Pause Quantum

Stores the current value of the Receive Pause Quantum Register which is decremented every 512 bit times.
34.7.16 Transmit Pause Quantum Register

31
GMAC_TPQ 0xF002803C Read-write

30 29 28 27 26 25 24
23
22
21
20
19
18
17
16
15
14
13
12 TPQ
11
10
4 TPQ
TPQ: Transmit Pause Quantum

Written with the pause quantum value for pause frame transmission.

884
34.7.17 TX Partial Store and Forward Register

31 ENTXP 23 15 7
GMAC_TPSF 0xF0028040 Read-write

30 22 14 6 29 21 13 5 28 20 12 4 TPB1ADR 27 19 11 26 18 10 TPB1ADR 3 2 1 0 25 17 9 24 16 8
Partial store and forward is only applicable when using the AHB DMA configured in SRAM based packet buffer mode. It is also not available when the priority queueing feature is enabled.
TPB1ADR: Transmit Partial Store and Forward Address

Watermark value. Reset = 1.
ENTXP: Enable TX Partial Store and Forward Operation

885
34.7.18 RX Partial Store and Forward Register

31 ENRXP 23 15 7
GMAC_RPSF 0xF0028044 Read-write

30 22 14 6 29 21 13 5 28 20 12 4 RPB1ADR 27 19 11 26 18 10 RPB1ADR 3 2 1 0 25 17 9 24 16 8
Partial store and forward is only applicable when using the AHB DMA configured in SRAM based packet buffer mode. It is also not available when the priority queueing feature is enabled.
RPB1ADR: Receive Partial Store and Forward Address

Watermark value. Reset = 1.
ENRXP: Enable RX Partial Store and Forward Operation

886
34.7.19 Hash Register Bottom [31:0]

31
GMAC_HRB 0xF0028080 Read-only

30 29 28 ADDR 27 26 25 24
23
22
21
20 ADDR
19
18
17
16
15
14
13
12 ADDR
11
10
4 ADDR
The unicast hash enable (UNIHEN) and the multicast hash enable (MITIHEN) bits in the Network Configuration Register (Section 34.7.2 Network Configuration Register) enable the reception of hash matched frames. See Section 34.5.9 Hash Addressing.
ADDR: Hash Address

The first 32 bits of the Hash Address Register.
34.7.20 Hash Register Top [63:32]

31
GMAC_HRT 0xF0028084 Read-only

30 29 28 ADDR 27 26 25 24
23
22
21
20 ADDR
19
18
17
16
15
14
13
12 ADDR
11
10
4 ADDR
The unicast hash enable (UNIHEN) and the multicast hash enable (MITIHEN) bits in the Network Configuration Register enable the reception of hash matched frames. See Section 34.5.9 Hash Addressing.
ADDR: Hash Address

Bits 63 to 32 of the Hash Address Register.

887
34.7.21 Specific Address 1 Bottom [31:0] Register

31
GMAC_SAB1 0xF0028088 Read-write

30 29 28 ADDR 27 26 25 24
23
22
21
20 ADDR
19
18
17
16
15
14
13
12 ADDR
11
10
4 ADDR
The addresses stored in the Specific Address Registers are deactivated at reset or when their corresponding Specific Address Register Bottom is written. They are activated when Specific Address Register Top is written.
ADDR: Specific Address 1

Least significant 32 bits of the destination address, that is bits 31:0. Bit zero indicates whether the address is multicast or unicast and corresponds to the least significant bit of the first byte received.
34.7.22 Specific Address 1 Top [47:32] Register

31 23 15
GMAC_SAT1 0xF002808C Read-write

30 22 14 29 21 13 28 20 12 ADDR 7 6 5 4 ADDR 3 2 1 0 27 19 11 26 18 10 25 17 9 24 16 8

The most significant bits of the destination address, that is bits 47:32.

888

31

30 29 28 ADDR 27 26 25 24
23
22
21
20 ADDR
19
18
17
16
15
14
13
12 ADDR
11
10
4 ADDR


31 23 15
GMAC_SAT2 0xF0028094 Read-write

30 22 14 29 21 13 28 20 12 ADDR 7 6 5 4 ADDR 3 2 1 0 27 19 11 26 18 10 25 17 9 24 16 8


889

31

30 29 28 ADDR 27 26 25 24
23
22
21
20 ADDR
19
18
17
16
15
14
13
12 ADDR
11
10
4 ADDR


31 23 15
GMAC_SAT3 0xF002809C Read-write

30 22 14 29 21 13 28 20 12 ADDR 7 6 5 4 ADDR 3 2 1 0 27 19 11 26 18 10 25 17 9 24 16 8


890
34.7.27 Specific Address 4 Bottom Register[31:0]

31
GMAC_SAB4 0xF00280A0 Read-write

30 29 28 ADDR 27 26 25 24
23
22
21
20 ADDR
19
18
17
16
15
14
13
12 ADDR
11
10
4 ADDR

34.7.28 Specific Address 4 Top Register[47:32]

31 23 15
GMAC_SAT4 0xF00280A4 Read-write

30 22 14 29 21 13 28 20 12 ADDR 7 6 5 4 ADDR 3 2 1 0 27 19 11 26 18 10 25 17 9 24 16 8


891
34.7.29 Type ID Match 1 Register

Name: Access:
31 23 15
GMAC_TIDM1 Read-write
30 22 14 29 21 13 28 20 12 TID 7 6 5 4 TID 3 2 1 0 27 19 11 26 18 10 25 17 9 24 16 8
TID: Type ID Match 1

For use in comparisons with received frames type ID/length frames.

Name: Access:
31 23 15
30 22 14 29 21 13 28 20 12 TID 7 6 5 4 TID 3 2 1 0 27 19 11 26 18 10 25 17 9 24 16 8


892

Name: Access:
31 23 15
30 22 14 29 21 13 28 20 12 TID 7 6 5 4 TID 3 2 1 0 27 19 11 26 18 10 25 17 9 24 16 8


Name: Access:
31 23 15
30 22 14 29 21 13 28 20 12 TID 7 6 5 4 TID 3 2 1 0 27 19 11 26 18 10 25 17 9 24 16 8


893
34.7.33 Wake on LAN Register

31
GMAC_WOL 0xF00280B8 Read-write

30 29 28 27 26 25 24
23
22
21
20
19 MTI 11 IP
18 SA1 10
17 ARP 9
16 MAG 8
15
14
13
12
4 IP
IP: ARP Request IP Address

Wake on LAN ARP request IP address. Written to define the least significant 16 bits of the target IP address that is matched to generate a Wake on LAN event. A value of zero will not generate an event, even if this is matched by the received frame.
MAG: Magic Packet Event Enable

Wake on LAN magic packet event enable.
ARP: ARP Request Event Enable

Wake on LAN ARP request event enable.
SA1: Specific Address Register 1 Event Enable

Wake on LAN Specific Address Register 1 event enable.
MTI: Multicast Hash Event Enable

Wake on LAN multicast hash event enable.

894
34.7.34 IPG Stretch Register

31 23 15
GMAC_IPGS 0xF00280BC Read-write

30 22 14 29 21 13 28 20 12 FL 7 6 5 4 FL 3 2 1 0 27 19 11 26 18 10 25 17 9 24 16 8
FL: Frame Length

Bits 7:0 are multiplied with the previously transmitted frame length (including preamble). Bits 15:8 +1 divide the frame length. If the resulting number is greater than 96 and bit 28 is set in the Network Configuration Register then the resulting number is used for the transmit inter-packet-gap. 1 is added to bits 15:8 to prevent a divide by zero. See Section 34.5.4 MAC Transmit Block.

895
34.7.35 Stacked VLAN Register

31 ESVLAN 23 15
GMAC_SVLAN 0xF00280C0 Read-write

30 22 14 29 21 13 28 20 12 VLAN_TYPE 7 6 5 4 VLAN_TYPE 3 2 1 0 27 19 11 26 18 10 25 17 9 24 16 8
VLAN_TYPE: User Defined VLAN_TYPE Field

User defined VLAN_TYPE field. When Stacked VLAN is enabled, the first VLAN tag in a received frame will only be accepted if the VLAN type field is equal to this user defined VLAN_TYPE, OR equal to the standard VLAN type (0x8100). Note that the second VLAN tag of a Stacked VLAN packet will only be matched correctly if its VLAN_TYPE field equals 0x8100.
ESVLAN: Enable Stacked VLAN Processing Mode

896
34.7.36 Transmit PFC Pause Register

31 23 15
GMAC_TPFCP 0xF00280C4 Read-write

30 22 14 29 21 13 28 20 12 PQ 7 6 5 4 PEV 3 2 1 0 27 19 11 26 18 10 25 17 9 24 16 8
PEV: Priority Enable Vector

If bit 17 of the Network Control Register is written with a one then the priority enable vector of the PFC priority based pause frame will be set equal to the value stored in this register [7:0].
PQ: Pause Quantum

If bit 17 of the Network Control Register is written with a one then for each entry equal to zero in the Transmit PFC Pause Register[15:8], the PFC pause frame's pause quantum field associated with that entry will be taken from the Transmit Pause Quantum Register. For each entry equal to one in the Transmit PFC Pause Register [15:8], the pause quantum associated with that entry will be zero.

897
34.7.37 Specific Address 1 Mask Bottom [31:0] Register

31
GMAC_SAMB1 0xF00280C8 Read-write

30 29 28 ADDR 27 26 25 24
23
22
21
20 ADDR
19
18
17
16
15
14
13
12 ADDR
11
10
4 ADDR
ADDR: Specific Address 1 Mask

Setting a bit to one masks the corresponding bit in the Specific Address 1 Register.
34.7.38 Specific Address Mask 1 Top [47:32] Register

31 23 15
GMAC_SAMT1 0xF00280CC Read-write

30 22 14 29 21 13 28 20 12 ADDR 7 6 5 4 ADDR 3 2 1 0 27 19 11 26 18 10 25 17 9 24 16 8
ADDR: Specific Address 1 Mask

Setting a bit to one masks the corresponding bit in the Specific Address 1 Register.

898
34.7.39 Octets Transmitted [31:0] Register

31
GMAC_OTLO 0xF0028100 Read-only

30 29 28 TXO 27 26 25 24
23
22
21
20 TXO
19
18
17
16
15
14
13
12 TXO
11
10
4 TXO
When reading the Octets Transmitted and Octets Received Registers, bits 31:0 should be read prior to bits 47:32 to ensure reliable operation.
TXO: Transmitted Octets

Transmitted octets in frame without errors [31:0]. The number of octets transmitted in valid frames of any type. This counter is 48bits, and is read through two registers. This count does not include octets from automatically generated pause frames.
34.7.40 Octets Transmitted [47:32] Register

31
GMAC_OTHI 0xF0028104 Read-only

30 29 28 27 26 25 24
23
22
21
20
19
18
17
16
15
14
13
12 TXO
11
10
4 TXO
TXO: Transmitted Octets

Transmitted octets in frame without errors [47:32]. The number of octets transmitted in valid frames of any type. This counter is 48-bits, and is read through two registers. This count does not include octets from automatically generated pause frames.

899
34.7.41 Frames Transmitted Register

31
GMAC_FT 0xF0028108 Read-only

30 29 28 FTX 27 26 25 24
23
22
21
20 FTX
19
18
17
16
15
14
13
12 FTX
11
10
4 FTX
FTX: Frames Transmitted without Error

Frames transmitted without error. This register counts the number of frames successfully transmitted, i.e., no under run and not too many retries. Excludes pause frames.
34.7.42 Broadcast Frames Transmitted Register

31
GMAC_BCFT 0xF002810C Read-only

30 29 28 BFTX 27 26 25 24
23
22
21
20 BFTX
19
18
17
16
15
14
13
12 BFTX
11
10
4 BFTX
BFTX: Broadcast Frames Transmitted without Error

Broadcast frames transmitted without error. This register counts the number of broadcast frames successfully transmitted without error, i.e., no under run and not too many retries. Excludes pause frames.

900
34.7.43 Multicast Frames Transmitted Register

31
GMAC_MFT 0xF0028110 Read-only

30 29 28 MFTX 27 26 25 24
23
22
21
20 MFTX
19
18
17
16
15
14
13
12 MFTX
11
10
4 MFTX
MFTX: Multicast Frames Transmitted without Error

This register counts the number of multicast frames successfully transmitted without error, i.e., no under run and not too many retries. Excludes pause frames.
34.7.44 Pause Frames Transmitted Register

31 23 15
GMAC_PFT 0xF0028114 Read-only

30 22 14 29 21 13 28 20 12 PFTX 7 6 5 4 PFTX 3 2 1 0 27 19 11 26 18 10 25 17 9 24 16 8
PFTX: Pause Frames Transmitted Register

This register counts the number of pause frames transmitted. Only pause frames triggered by the register interface or through the external pause pins are counted as pause frames. Pause frames received through FIFO are counted in the frames transmitted counter.

901
34.7.45 64 Byte Frames Transmitted Register

31
GMAC_BFT64 0xF0028118 Read-only

30 29 28 NFTX 27 26 25 24
23
22
21
20 NFTX
19
18
17
16
15
14
13
12 NFTX
11
10
4 NFTX
NFTX: 64 Byte Frames Transmitted without Error

This register counts the number of 64 byte frames successfully transmitted without error, i.e., no under run and not too many retries. Excludes pause frames.
34.7.46 65 to 127 Byte Frames Transmitted Register

31
GMAC_TBFT127 0xF002811C Read-only

30 29 28 NFTX 27 26 25 24
23
22
21
20 NFTX
19
18
17
16
15
14
13
12 NFTX
11
10
4 NFTX
NFTX: 65 to 127 Byte Frames Transmitted without Error

This register counts the number of 65 to 127 byte frames successfully transmitted without error, i.e., no under run and not too many retries. Excludes pause frames.

902

31
GMAC_TBFT255 0xF0028120 Read-only

30 29 28 NFTX 27 26 25 24
23
22
21
20 NFTX
19
18
17
16
15
14
13
12 NFTX
11
10
4 NFTX

This register counts the number of 128 to 255 byte frames successfully transmitted without error, i.e., no under run and not too many retries.

31

30 29 28 NFTX 27 26 25 24
23
22
21
20 NFTX
19
18
17
16
15
14
13
12 NFTX
11
10
4 NFTX


903

31

30 29 28 NFTX 27 26 25 24
23
22
21
20 NFTX
19
18
17
16
15
14
13
12 NFTX
11
10
4 NFTX


31
GMAC_TBFT1518 0xF002812C Read-only

30 29 28 NFTX 27 26 25 24
23
22
21
20 NFTX
19
18
17
16
15
14
13
12 NFTX
11
10
4 NFTX


904
34.7.51 Greater Than 1518 Byte Frames Transmitted Register

31
GMAC_GTBFT1518 0xF0028130 Read-only

30 29 28 NFTX 27 26 25 24
23
22
21
20 NFTX
19
18
17
16
15
14
13
12 NFTX
11
10
4 NFTX
NFTX: Greater than 1518 Byte Frames Transmitted without Error

This register counts the number of 1518 or above byte frames successfully transmitted without error i.e., no under run and not too many retries.
34.7.52 Transmit Under Runs Register

31 23 15 7
GMAC_TUR 0xF0028134 Read-only

30 22 14 6 29 21 13 5 28 20 12 4 TXUNR 27 19 11 3 26 18 10 2 25 17 9 TXUNR 1 0 24 16 8
TXUNR: Transmit Under Runs

This register counts the number of frames not transmitted due to a transmit under run. If this register is incremented then no other statistics register is incremented.

905
34.7.53 Single Collision Frames Register

31 23 15
GMAC_SCF 0xF0028138 Read-only

30 22 14 29 21 13 28 20 12 SCOL 7 6 5 4 SCOL 3 2 1 0 27 19 11 26 18 10 25 17 SCOL 9 8 24 16
SCOL: Single Collision

This register counts the number of frames experiencing a single collision before being successfully transmitted i.e., no under run.
34.7.54 Multiple Collision Frames Register

31 23 15
GMAC_MCF 0xF002813C Read-only

30 22 14 29 21 13 28 20 12 MCOL 7 6 5 4 MCOL 3 2 1 0 27 19 11 26 18 10 25 17 MCOL 9 8 24 16
MCOL: Multiple Collision

This register counts the number of frames experiencing between two and fifteen collisions prior to being successfully transmitted, i.e., no under run and not too many retries.

906
34.7.55 Excessive Collisions Register

31 23 15 7
GMAC_EC 0xF0028140 Read-only

30 22 14 6 29 21 13 5 28 20 12 4 XCOL 27 19 11 3 26 18 10 2 25 17 9 XCOL 1 0 24 16 8
XCOL: Excessive Collisions

This register counts the number of frames that failed to be transmitted because they experienced 16 collisions.
34.7.56 Late Collisions Register

31 23 15 7
GMAC_LC 0xF0028144 Read-only

30 22 14 6 29 21 13 5 28 20 12 4 LCOL 27 19 11 3 26 18 10 2 25 17 9 LCOL 1 0 24 16 8
LCOL: Late Collisions

This register counts the number of late collisions occurring after the slot time (512 bits) has expired. In 10/100 mode, late collisions are counted twice i.e., both as a collision and a late collision. In gigabit mode, a late collision causes the transmission to be aborted, thus the single and multi collision registers are not updated.

907
34.7.57 Deferred Transmission Frames Register

31 23 15
GMAC_DTF 0xF0028148 Read-only

30 22 14 29 21 13 28 20 12 DEFT 7 6 5 4 DEFT 3 2 1 0 27 19 11 26 18 10 25 17 DEFT 9 8 24 16
DEFT: Deferred Transmission

This register counts the number of frames experiencing deferral due to carrier sense being active on their first attempt at transmission. Frames involved in any collision are not counted nor are frames that experienced a transmit under run.
34.7.58 Carrier Sense Errors Register

31 23 15 7
GMAC_CSE 0xF002814C Read-only

30 22 14 6 29 21 13 5 28 20 12 4 CSR 27 19 11 3 26 18 10 2 25 17 9 CSR 1 0 24 16 8
CSR: Carrier Sense Error

This register counts the number of frames transmitted where carrier sense was not seen during transmission or where carrier sense was deasserted after being asserted in a transmit frame without collision (no under run). Only incremented in half duplex mode. The only effect of a carrier sense error is to increment this register. The behavior of the other Statistics Registers is unaffected by the detection of a carrier sense error.

908
34.7.59 Octets Received [31:0] Register

31
GMAC_ORLO 0xF0028150 Read-only

30 29 28 RXO 27 26 25 24
23
22
21
20 RXO
19
18
17
16
15
14
13
12 RXO
11
10
4 RXO
When reading the Octets Transmitted and Octets Received Registers, bits [31:0] should be read prior to bits [47:32] to ensure reliable operation.
RXO: Received Octets

Received octets in frame without errors [31:0]. The number of octets received in valid frames of any type. This counter is 48-bits and is read through two registers. This count does not include octets from pause frames, and is only incremented if the frame is successfully filtered and copied to memory.
34.7.60 Octets Received [47:32] Register

31
GMAC_ORHI 0xF0028154 Read-only

30 29 28 27 26 25 24
23
22
21
20
19
18
17
16
15
14
13
12 RXO
11
10
4 RXO
RXO: Received Octets

Received octets in frame without errors [47:32]. The number of octets received in valid frames of any type. This counter is 48-bits and is read through two registers. This count does not include octets from pause frames, and is only incremented if the frame is successfully filtered and copied to memory.

909
34.7.61 Frames Received Register

31
GMAC_FR 0xF0028158 Read-only

30 29 28 FRX 27 26 25 24
23
22
21
20 FRX
19
18
17
16
15
14
13
12 FRX
11
10
4 FRX
FRX: Frames Received without Error

Frames received without error. This register counts the number of frames successfully received. Excludes pause frames, and is only incremented if the frame is successfully filtered and copied to memory.
34.7.62 Broadcast Frames Received Register

31
GMAC_BCFR 0xF002815C Read-only

30 29 28 BFRX 27 26 25 24
23
22
21
20 BFRX
19
18
17
16
15
14
13
12 BFRX
11
10
4 BFRX
BFRX: Broadcast Frames Received without Error

Broadcast frames received without error. This register counts the number of broadcast frames successfully received. Excludes pause frames, and is only incremented if the frame is successfully filtered and copied to memory.

910
34.7.63 Multicast Frames Received Register

31
GMAC_MFR 0xF0028160 Read-only

30 29 28 MFRX 27 26 25 24
23
22
21
20 MFRX
19
18
17
16
15
14
13
12 MFRX
11
10
4 MFRX
MFRX: Multicast Frames Received without Error

This register counts the number of multicast frames successfully received without error. Excludes pause frames, and is only incremented if the frame is successfully filtered and copied to memory.
34.7.64 Pause Frames Received Register

31 23 15
GMAC_PFR 0xF0028164 Read-only

30 22 14 29 21 13 28 20 12 PFRX 7 6 5 4 PFRX 3 2 1 0 27 19 11 26 18 10 25 17 9 24 16 8
PFRX: Pause Frames Received Register

This register counts the number of pause frames received without error.

911
34.7.65 64 Byte Frames Received Register

31
GMAC_BFR64 0xF0028168 Read-only

30 29 28 NFRX 27 26 25 24
23
22
21
20 NFRX
19
18
17
16
15
14
13
12 NFRX
11
10
4 NFRX
NFRX: 64 Byte Frames Received without Error

This register counts the number of 64 byte frames successfully received without error. Excludes pause frames, and is only incremented if the frame is successfully filtered and copied to memory.
34.7.66 65 to 127 Byte Frames Received Register

31
GMAC_TBFR127 0xF002816C Read-only

30 29 28 NFRX 27 26 25 24
23
22
21
20 NFRX
19
18
17
16
15
14
13
12 NFRX
11
10
4 NFRX
NFRX: 65 to 127 Byte Frames Received without Error

This register counts the number of 65 to 127 byte frames successfully received without error. Excludes pause frames, and is only incremented if the frame is successfully filtered and copied to memory.

912

31
GMAC_TBFR255 0xF0028170 Read-only

30 29 28 NFRX 27 26 25 24
23
22
21
20 NFRX
19
18
17
16
15
14
13
12 NFRX
11
10
4 NFRX


31

30 29 28 NFRX 27 26 25 24
23
22
21
20 NFRX
19
18
17
16
15
14
13
12 NFRX
11
10
4 NFRX


913

31

30 29 28 NFRX 27 26 25 24
23
22
21
20 NFRX
19
18
17
16
15
14
13
12 NFRX
11
10
4 NFRX


31
GMAC_TBFR1518 0xF002817C Read-only

30 29 28 NFRX 27 26 25 24
23
22
21
20 NFRX
19
18
17
16
15
14
13
12 NFRX
11
10
4 NFRX

This register counts the number of 1024 to 1518 byte frames successfully received without error, i.e., no under run and not too many retries.

914
34.7.71 1519 to Maximum Byte Frames Received Register

31
GMAC_TMXBFR 0xF0028180 Read-only

30 29 28 NFRX 27 26 25 24
23
22
21
20 NFRX
19
18
17
16
15
14
13
12 NFRX
11
10
4 NFRX
NFRX: 1519 to Maximum Byte Frames Received without Error

This register counts the number of 1519 byte or above frames successfully received without error. Maximum frame size is determined by the Network Configuration Register bit 8 (1536 maximum frame size) or bit 3 (jumbo frame size). Excludes pause frames, and is only incremented if the frame is successfully filtered and copied to memory. See: Section 34.7.2 Network Configuration Register.

915
34.7.72 Undersized Frames Received Register

31 23 15 7
GMAC_UFR 0xF0028184 Read-only

30 22 14 6 29 21 13 5 28 20 12 4 UFRX 27 19 11 3 26 18 10 2 25 17 9 UFRX 1 0 24 16 8
UFRX: Undersize Frames Received

This register counts the number of frames received less than 64 bytes in length (10/100 mode or gigabit mode, full duplex) that do not have either a CRC error or an alignment error. In gigabit mode, half duplex, this register counts either frames not conforming to the minimum slot time of 512 bytes or frames not conforming to the minimum frame size once bursting is active.

916
34.7.73 Oversized Frames Received Register

31 23 15 7
GMAC_OFR 0xF0028188 Read-only

30 22 14 6 29 21 13 5 28 20 12 4 OFRX 27 19 11 3 26 18 10 2 25 17 9 OFRX 1 0 24 16 8
OFRX: Oversized Frames Received

This register counts the number of frames received exceeding 1518 bytes (1536 bytes if bit 8 is set in the Network Configuration Register, 10240 bytes if bit 3 is set in the Network Configuration Register) in length but do not have either a CRC error, an alignment error nor a receive symbol error. See Section 34.7.2 Network Configuration Register.

917
34.7.74 Jabbers Received Register

31 23 15 7
GMAC_JR 0xF002818C Read-only

30 22 14 6 29 21 13 5 28 20 12 4 JRX 27 19 11 3 26 18 10 2 25 17 9 JRX 1 0 24 16 8
JRX: Jabbers Received

The register counts the number of frames received exceeding 1518 bytes in length (1536 if bit 8 is set in Network Configuration Register, 10240 bytes if bit 3 is set in the Network Configuration Register) and have either a CRC error, an alignment error or a receive symbol error. See: Section 34.7.2 Network Configuration Register.

918
34.7.75 Frame Check Sequence Errors Register

31 23 15 7
GMAC_FCSE 0xF0028190 Read-only

30 22 14 6 29 21 13 5 28 20 12 4 FCKR 27 19 11 3 26 18 10 2 25 17 9 FCKR 1 0 24 16 8
FCKR: Frame Check Sequence Errors

The register counts frames that are an integral number of bytes, have bad CRC and are between 64 and 1518 bytes in length (1536 if bit 8 is set in Network Configuration Register, 10240 bytes if bit 3 is set in the Network Configuration Register). This register is also incremented if a symbol error is detected and the frame is of valid length and has an integral number of bytes. This register is incremented for a frame with bad FCS, regardless of whether it is copied to memory due to ignore FCS mode being enabled in bit 26 of the Network Configuration Register. See: Section 34.7.2 Network Configuration Register.

919
34.7.76 Length Field Frame Errors Register

31 23 15 7
GMAC_LFFE 0xF0028194 Read-only

30 22 14 6 29 21 13 5 28 20 12 4 LFER 27 19 11 3 26 18 10 2 25 17 9 LFER 1 0 24 16 8
LFER: Length Field Frame Errors

This register counts the number of frames received that have a measured length shorter than that extracted from the length field (bytes 13 and 14). This condition is only counted if the value of the length field is less than 0x0600, the frame is not of excessive length and checking is enabled through bit 16 of the Network Configuration Register. See: Section 34.7.2 Network Configuration Register.

920
34.7.77 Receive Symbol Errors Register

31 23 15 7
GMAC_RSE 0xF0028198 Read-only

30 22 14 6 29 21 13 5 28 20 12 4 RXSE 27 19 11 3 26 18 10 2 25 17 9 RXSE 1 0 24 16 8
RXSE: Receive Symbol Errors

This register counts the number of frames that had GRXER asserted during reception. For 10/100 mode symbol errors are counted regardless of frame length checks. For gigabit mode the frame must satisfy slot time requirements in order to count a symbol error. Additionally, in gigabit half duplex mode, carrier extension errors are also recorded. Receive symbol errors will also be counted as an FCS or alignment error if the frame is between 64 and 1518 bytes (1536 bytes if bit 8 is set in the Network Configuration Register, 10240 bytes if bit 3 is set in the Network Configuration Register). If the frame is larger it will be recorded as a jabber error. See Section 34.7.2 Network Configuration Register.

921
34.7.78 Alignment Errors Register

31 23 15 7
GMAC_AE 0xF002819C Read-only

30 22 14 6 29 21 13 5 28 20 12 4 AER 27 19 11 3 26 18 10 2 25 17 9 AER 1 0 24 16 8
AER: Alignment Errors

This register counts the frames that are not an integral number of bytes long and have bad CRC when their length is truncated to an integral number of bytes and are between 64 and 1518 bytes in length (1536 if bit 8 is set in Network Configuration Register, 10240 bytes if bit 3 is set in the Network Configuration Register). This register is also incremented if a symbol error is detected and the frame is of valid length and does not have an integral number of bytes. See: Section 34.7.2 Network Configuration Register.

922
34.7.79 Receive Resource Errors Register

31 23 15
GMAC_RRE 0xF00281A0 Read-only

30 22 14 29 21 13 28 20 12 RXRER 7 6 5 4 RXRER 3 2 1 0 27 19 11 26 18 10 25 17 RXRER 9 8 24 16
RXRER: Receive Resource Errors

This register counts frames that are not an integral number of bytes long and have bad CRC when their length is truncated to an integral number of bytes and are between 64 and 1518 bytes in length (1536 if bit 8 is set in Network Configuration Register, 10240 bytes if bit 3 is set in the Network Configuration Register). This register is also incremented if a symbol error is detected and the frame is of valid length and does not have an integral number of bytes. See: Section 34.7.2 Network Configuration Register.

923
34.7.80 Receive Overruns Register

31 23 15 7
GMAC_ROE 0xF00281A4 Read-only

30 22 14 6 29 21 13 5 28 20 12 4 RXOVR 27 19 11 3 26 18 10 2 25 17 9 RXOVR 1 0 24 16 8
RXOVR: Receive Overruns

This register counts the number of frames that are address recognized but were not copied to memory due to a receive overrun.

924
34.7.81 IP Header Checksum Errors Register

31 23 15 7
GMAC_IHCE 0xF00281A8 Read-only

30 22 14 6 29 21 13 5 28 20 12 4 HCKER 27 19 11 3 26 18 10 2 25 17 9 1 24 16 8 0
HCKER: IP Header Checksum Errors

This register counts the number of frames discarded due to an incorrect IP header checksum, but are between 64 and 1518 bytes (1536 bytes if bit 8 is set in the Network Configuration Register or 10240 bytes if bit 3 is set in the Network Configuration Register) and do not have a CRC error, an alignment error, nor a symbol error.

925
34.7.82 TCP Checksum Errors Register

31 23 15 7
GMAC_TCE 0xF00281AC Read-only

30 22 14 6 29 21 13 5 28 20 12 4 TCKER 27 19 11 3 26 18 10 2 25 17 9 1 24 16 8 0
TCKER: TCP Checksum Errors

This register counts the number of frames discarded due to an incorrect TCP checksum, but are between 64 and 1518 bytes (1536 bytes if bit 8 is set in the Network Configuration Register or 10240 bytes if bit 3 is set in the Network Configuration Register) and do not have a CRC error, an alignment error, nor a symbol error.

926
34.7.83 UDP Checksum Errors Register

31 23 15 7
GMAC_UCE 0xF00281B0 Read-only

30 22 14 6 29 21 13 5 28 20 12 4 UCKER 27 19 11 3 26 18 10 2 25 17 9 1 24 16 8 0
UCKER: UDP Checksum Errors

This register counts the number of frames discarded due to an incorrect UDP checksum, but are between 64 and 1518 bytes (1536 bytes if bit 8 is set in the Network Configuration Register or 10240 bytes if bit 3 is set in the Network Configuration Register) and do not have a CRC error, an alignment error, nor a symbol error.

927
34.7.84 1588 Timer Sync Strobe Seconds Register

31
GMAC_TSSS 0xF00281C8 Read-write

30 29 28 VTS 27 26 25 24
23
22
21
20 VTS
19
18
17
16
15
14
13
12 VTS
11
10
4 VTS
VTS: Value of Timer Seconds Register Capture

The value of the Timer Seconds Register is captured.
34.7.85 1588 Timer Sync Strobe Nanoseconds Register

31 23
GMAC_TSSN 0xF00281CC Read-write

30 22 29 28 27 VTN 21 20 VTN 19 18 17 16 26 25 24
15
14
13
12 VTN
11
10
4 VTN
VTN: Value Timer Nanoseconds Register Capture

The value of the Timer Nanoseconds Register is captured.

928
34.7.86 1588 Timer Seconds Register

31
GMAC_TS 0xF00281D0 Read-write

30 29 28 TCS 27 26 25 24
23
22
21
20 TCS
19
18
17
16
15
14
13
12 TCS
11
10
4 TCS
TCS: Timer Count in Seconds

This register is writable. It increments by one when the 1588 nanoseconds counter counts to one second. It may also be incremented when the Timer Adjust Register is written.
34.7.87 1588 Timer Nanoseconds Register

31 23
GMAC_TN 0xF00281D4 Read-write

30 22 29 28 27 TNS 21 20 TNS 19 18 17 16 26 25 24
15
14
13
12 TNS
11
10
4 TNS
TNS: Timer Count in Nanoseconds

This register is writable. It can also be adjusted by writes to the 1588 Timer Adjust Register. It increments by the value of the 1588 Timer Increment Register each clock cycle.

929
34.7.88 1588 Timer Adjust Register

31 ADJ 23
GMAC_TA 0xF00281D8 Write-only

30 22 29 28 27 ITDT 21 20 ITDT 19 18 17 16 26 25 24
15
14
13
12 ITDT
11
10
4 ITDT
ITDT: Increment/Decrement
The number of nanoseconds to increment or decrement the 1588 Timer Nanoseconds Register. If necessary, the 1588 Seconds Register will be incremented or decremented.
ADJ: Adjust 1588 Timer

Write as one to subtract from the 1588 timer. Write as zero to add to it.

930
34.7.89 1588 Timer Increment Register

31 23
GMAC_TI 0xF00281DC Read-write

30 22 29 21 28 20 NIT 27 19 26 18 25 17 24 16
15
14
13
12 AC NS
11
10
4 CNS
CNS: Count Nanoseconds

A count of nanoseconds by which the 1588 Timer Nanoseconds Register will be incremented each clock cycle.
ACNS: Alternative Count Nanoseconds

Alternative count of nanoseconds by which the 1588 Timer Nanoseconds Register will be incremented each clock cycle.
NIT: Number of Increments

The number of increments after which the alternative increment is used.

931
34.7.90 PTP Event Frame Transmitted Seconds Register

31
GMAC_EFTS 0xF00281E0 Read-only

30 29 28 RUD 27 26 25 24
23
22
21
20 RUD
19
18
17
16
15
14
13
12 RUD
11
10
4 RUD
RUD: Register Update

The register is updated with the value that the 1588 Timer Seconds Register holds when the SFD of a PTP transmit peer event crosses the MII interface. An interrupt is issued when the register is updated.
34.7.91 PTP Event Frame Transmitted Nanoseconds Register

31 23
GMAC_EFTN 0xF00281E4 Read-only

30 22 29 28 27 RUD 21 20 RUD 19 18 17 16 26 25 24
15
14
13
12 RUD
11
10
4 RUD

The register is updated with the value that the 1588 Timer Nanoseconds Register holds when the SFD of a PTP transmit peer event crosses the MII interface. An interrupt is issued when the register is updated.

932
34.7.92 PTP Event Frame Received Seconds Register

31
GMAC_EFRS 0xF00281E8 Read-only

30 29 28 RUD 27 26 25 24
23
22
21
20 RUD
19
18
17
16
15
14
13
12 RUD
11
10
4 RUD

The register is updated with the value that the 1588 Timer Seconds Register holds when the SFD of a PTP receive primary event crosses the MII interface. An interrupt is issued when the register is updated.
34.7.93 PTP Event Frame Received Nanoseconds Register

31 23
GMAC_EFRN 0xF00281EC Read-only

30 22 29 28 27 RUD 21 20 RUD 19 18 17 16 26 25 24
15
14
13
12 RUD
11
10
4 RUD

The register is updated with the value that the 1588 Timer Nanoseconds Register holds when the SFD of a PTP receive primary event crosses the MII interface. An interrupt is issued when the register is updated.

933
34.7.94 PTP Peer Event Frame Transmitted Seconds Register

31
GMAC_PEFTS 0xF00281F0 Read-only

30 29 28 RUD 27 26 25 24
23
22
21
20 RUD
19
18
17
16
15
14
13
12 RUD
11
10
4 RUD

The register is updated with the value that the 1588 Timer Seconds Register holds when the SFD of a PTP transmit peer event crosses the MII interface. An interrupt is issued when the register is updated.
34.7.95 PTP Peer Event Frame Transmitted Nanoseconds Register

31 23
GMAC_PEFTN 0xF00281F4 Read-only

30 22 29 28 27 RUD 21 20 RUD 19 18 17 16 26 25 24
15
14
13
12 RUD
11
10
4 RUD

The register is updated with the value that the 1588 Timer Nanoseconds Register holds when the SFD of a PTP transmit peer event crosses the MII interface. An interrupt is issued when the register is updated.

934
34.7.96 PTP Peer Event Frame Received Seconds Register

31
GMAC_PEFRS 0xF00281F8 Read-only

30 29 28 RUD 27 26 25 24
23
22
21
20 RUD
19
18
17
16
15
14
13
12 RUD
11
10
4 RUD

The register is updated with the value that the 1588 Timer Seconds Register holds when the SFD of a PTP receive primary event crosses the MII interface. An interrupt is issued when the register is updated.

935
34.7.97 PTP Peer Event Frame Received Nanoseconds Register

31 23
GMAC_PEFRN 0xF00281FC Read-only

30 22 29 28 27 RUD 21 20 RUD 19 18 17 16 26 25 24
15
14
13
12 RUD
11
10
4 RUD

The register is updated with the value that the 1588 Timer Nanoseconds Register holds when the SFD of a PTP receive primary event crosses the MII interface. An interrupt is issued when the register is updated.

936
34.7.98 Interrupt Status Register Priority Queue

31
GMAC_ISRPQx[x=0..7] 0xF0028400 Read-only

30 29 28 27 26 25 24
23
22
21
20
19
18
17
16
15
14
13
12
11 HRESP 3
10 ROVR 2 RXUBR
9 1 RCOMP
8 0
7 TCOMP
6 TFC
5 RLEX
Seven status registers exist for each priority queue.
RCOMP: Receive Complete RXUBR: RX Used Bit Read RLEX: Retry Limit Exceeded or Late Collision TFC: Transmit Frame Corruption due to AHB error
Transmit frame corruption due to AHB error - set if an error occurs whilst midway through reading transmit frame from the AHB, including HRESP errors and buffers exhausted mid frame
TCOMP: Transmit Complete ROVR: Receive Overrun HRESP: HRESP Not OK

937
34.7.99 Transmit Buffer Queue Base Address Priority Queue

31
GMAC_TBQBAPQx[x=0..7] 0xF0028440 Read-write

30 29 28 27 26 25 24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
5 TXBQBA
These 7 registers hold the start address of the transmit buffer queues (transmit buffers descriptor lists) for the additional queues used when priority queues are employed.
TXBQBA: Transmit Buffer Queue Base Address

Written with the address of the start of the transmit queue.

938
34.7.100 Receive Buffer Queue Base Address Priority Queue

31
GMAC_RBQBAPQx[x=0..7] 0xF0028480 Read-write

30 29 28 27 26 25 24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
5 RXBQBA
These 7 registers hold the start address of the receive buffer queues (receive buffers descriptor lists) for the additional queues used when priority queues are employed.
RXBQBA: Receive Buffer Queue Base Address

Written with the address of the start of the receive queue.

939
34.7.101 Receive Buffer Size Register Priority Queue

31 23 15
GMAC_RBSRPQx[x=0..7] 0xF00284A0 Read-write

30 22 14 29 21 13 28 20 12 RBS 7 6 5 4 RBS 3 2 1 0 27 19 11 26 18 10 25 17 9 24 16 8
There are seven receive buffer size registers, one per priority que.
RBS: Receive Buffer Size

DMA receive buffer size in AHB system memory. The value defined by these bits determines the size of buffer to use in main AHB system memory when writing received data. The value is defined in multiples of 64 bytes such that a value of 0x01 corresponds to buffers of 64 bytes, 0x02 corresponds to 128 bytes etc. For example:
0x02: 128 bytes 0x18: 1536 bytes (1*max length frame/buffer) 0xA0: 10240 bytes (1*10K jumbo frame/buffer)
Note that this value should never be written as zero.

940
34.7.102 Screening Type1 Register Priority Queue

31 23
GMAC_ST1RPQx[x=0..15] 0xF0028500 Read-write

30 22 29 UDPE 21 28 DSTCE 20 UDPM 27 26 UDPM 19 18 17 16 25 24
15
14 UDPM
13
12
11
10 DSTCM
6 DSTCM
2 QNB
Screening type 1 registers are used to allocate up to 8 priority queues to received frames based on certain IP or UDP fields of incoming frames.
1. When DS/TC match enable is set (bit 28), the DS (Differentiated Services) field of the received IPv4 header or TC field (traffic class) of IPv6 headers are matched against bits 11:4. 2. When UDP port match enable is set (bit 29), the UDP Destination Port of the received UDP frame is matched against bits 27:12. Both UDP and DS/TC matching can be enabled simultaneously or individually.
If a match is successful, then the queue value programmed in bits 3:0 is allocated to the frame. The required number of Type 1 screening registers is configured in the gem defines file. Up to 16 type 1 screening registers have been allocated APB address space.
QNB: Que Number (0->7)

If a match is successful, then the queue value programmed in bits 3:0 is allocated to the frame.
DSTCM: Differentiated Services or Traffic Class Match

When DS/TC match enable is set (bit 28), the DS (differentiated services) field of the received IPv4 header or TC field (traffic class) of IPv6 headers are matched against bits 11:4
UDPM: UDP Port Match

When UDP port match enable is set (bit 29), the UDP Destination Port of the received UDP frame is matched against bits 27:12.
DSTCE: Differentiated Services or Traffic Class Match Enable

When DS/TC match enable is set (bit 28), the DS (differentiated services) field of the received IPv4 header or TC field (traffic class) of IPv6 headers are matched against bits 11:4
UDPE: UDP Port Match Enable

When UDP port match enable is set (bit 29), the UDP Destination Port of the received UDP frame is matched against bits 27:12.

941
34.7.103 Screening Type2 Register Priority Queue

31 23 15 7
GMAC_ST2RPQx[x=0..15] 0xF0028540 Read-write

30 22 14 6 VLANP 29 21 13 5 28 20 12 4 27 19 11 3 26 18 10 2 QNB 25 17 9 1 24 16 8 VLANE 0
Screening type 2 registers are used to allocate up to 8 priority queues to received frames based on the VLAN priority field of received ethernet frames. When VLAN match enable is set (bit 8), the VLAN priority field of the received frame is matched against bits [7:4] of this register. If a match is successful, then the queue value programmed in bits [3:0] is allocated to the frame. The required number of Type 2 screening registers is configured in the gem defines file. Up to 16 type 2 screening registers have been allocated APB address space.
QNB: Que Number (0->7)

If a match is successful, then the queue value programmed in bits [3:0] is allocated to the frame.
VLANP: VLAN Priority

When VLAN match enable is set (bit 8), the VLAN priority field of the received frame is matched against bits [7:4] of this register.
VLANE: VLAN Enable

0: VLAN match is disabled. 1: VLAN match is enabled.

942
34.7.104 Interrupt Enable Register Priority Queue

31
GMAC_IERPQx[x=0..6] 0xF0028600 Write-only

30 29 28 27 26 25 24
23
22
21
20
19
18
17
16
15
14
13
12
11 HRESP 3
10 ROVR 2 RXUBR
9 1 RCOMP
8 0
7 TCOMP
6 TFC
5 RLEX
For each priority queue there are seven interrupt enable registers.



Enable retry limit exceeded or late collision interrupt.

Enable transmit frame corruption due to AHB error interrupt.

Enable transmit complete.

HRESP: HRESP Not OK

Enable HRESP not OK interrupt.

943
34.7.105 Interrupt Disable Register Priority Queue

31
GMAC_IDRPQx[x=0..6] 0xF0028620 Write-only

30 29 28 27 26 25 24
23
22
21
20
19
18
17
16
15
14
13
12
11 HRESP 3
10 ROVR 2 RXUBR
9 1 RCOMP
8 0
7 TCOMP
6 TFC
5 RLEX
For each priority queue there are seven interrupt disable registers.



Disable retry limit exceeded or late collision interrupt.

Disable transmit frame corruption due to AHB error interrupt.

Disable transmit complete.

HRESP: HRESP Not OK

Disable HRESP not OK interrupt.

944
34.7.106 Interrupt Mask Register Priority Queue

31
GMAC_IMRPQx[x=0..6] 0xF0028640 Read-write

30 29 28 27 26 25 24
23
22
21
20
19
18
17
16
15
14
13
12
11 HRESP 3
10 ROVR 2 RXUBR
9 1 RCOMP
8 0
7 TCOMP
6 AHB
5 RLEX

A read of this register returns the value of the receive complete interrupt mask. 0 = Interrupt is enabled. 1 = Interrupt is disabled. A write to this register directly affects the state of the corresponding bit in the Interrupt Status Register, causing an interrupt to be generated if a 1 is written.

A read of this register returns the value of the RX used bit interrupt mask. 0 = Interrupt is enabled. 1 = Interrupt is disabled. A write to this register directly affects the state of the corresponding bit in the Interrupt Status Register, causing an interrupt to be generated if a 1 is written.

A read of this register returns the value of the retry limit exceeded or late collision interrupt mask. 0 = Interrupt is enabled. 1 = Interrupt is disabled. A write to this register directly affects the state of the corresponding bit in the Interrupt Status Register, causing an interrupt to be generated if a 1 is written.
AHB: AHB Error

A read of this register returns the value of the AHB error interrupt mask. 0 = Interrupt is enabled. 1 = Interrupt is disabled. A write to this register directly affects the state of the corresponding bit in the Interrupt Status Register, causing an interrupt to be generated if a 1 is written.

945

A read of this register returns the value of the transmit complete interrupt mask. 0 = Interrupt is enabled. 1 = Interrupt is disabled. A write to this register directly affects the state of the corresponding bit in the Interrupt Status Register, causing an interrupt to be generated if a 1 is written.

A read of this register returns the value of the receive overrun interrupt mask. 0 = Interrupt is enabled. 1 = Interrupt is disabled. A write to this register directly affects the state of the corresponding bit in the Interrupt Status Register, causing an interrupt to be generated if a 1 is written.
HRESP: HRESP Not OK

A read of this register returns the value of the HRESP not OK interrupt mask. 0 = Interrupt is enabled. 1 = Interrupt is disabled. A write to this register directly affects the state of the corresponding bit in the Interrupt Status Register, causing an interrupt to be generated if a 1 is written.

946
35.
35.1
Ethernet MAC 10/100 (EMAC)

Description
The EMAC module implements a 10/100 Ethernet MAC compatible with the IEEE 802.3 standard using an address checker, statistics and control registers, receive and transmit blocks, and a DMA interface. The address checker recognizes four specific 48-bit addresses and contains a 64-bit hash register for matching multicast and unicast addresses. It can recognize the broadcast address of all ones, copy all frames, and act on an external address match signal. The statistics register block contains registers for counting various types of event associated with transmit and receive operations. These registers, along with the status words stored in the receive buffer list, enable software to generate network management statistics compatible with IEEE 802.3.
35.2
z z z z z z z z z z z z z z z z z z z z
Supports RMII Interface to the physical layer Compatible with IEEE Standard 802.3 10 and 100 Mbit/s Operation Full-duplex and Half-duplex Operation Statistics Counter Registers Interrupt Generation to Signal Receive and Transmit Completion DMA Master on Receive and Transmit Channels Transmit and Receive FIFOs Automatic Pad and CRC Generation on Transmitted Frames Automatic Discard of Frames Received with Errors Address Checking Logic Supports Up to Four Specific 48-bit Addresses Supports Promiscuous Mode Where All Valid Received Frames are Copied to Memory Hash Matching of Unicast and Multicast Destination Addresses Physical Layer Management through MDIO Interface Half-duplex Flow Control by Forcing Collisions on Incoming Frames Full-duplex Flow Control with Recognition of Incoming Pause Frames Support for 802.1Q VLAN Tagging with Recognition of Incoming VLAN and Priority Tagged Frames Multiple Buffers per Receive and Transmit Frame Wake-on-LAN Support Jumbo Frames Up to 10240 bytes Supported

947
35.3
Block Diagram
Figure 35-1. EMAC Block Diagram
Address Checker
APB Slave
Register Interface
Statistics Registers
MDIO
Control Registers
DMA Interface
RX FIFO TX FIFO
Ethernet Receive
AHB Master MII/RMII
Ethernet Transmit

948
35.4
The MACB has several clock domains:
z z z
System bus clock (AHB and APB): DMA and register blocks Transmit clock: transmit block Receive clock: receive and address checker block
The system bus clock must run at least as fast as the receive clock and transmit clock (25 MHz at 100 Mbps, and 2.5 MHZ at 10 Mbps). Figure 35-1 illustrates the different blocks of the EMAC module. The control registers drive the MDIO interface, setup up DMA activity, start frame transmission and select modes of operation such as full- or half-duplex. The receive block checks for valid preamble, FCS, alignment and length, and presents received frames to the address checking block and DMA interface. The transmit block takes data from the DMA interface, adds preamble and, if necessary, pad and FCS, and transmits data according to the CSMA/CD (carrier sense multiple access with collision detect) protocol. The start of transmission is deferred if CRS (carrier sense) is active. If COL (collision) becomes active during transmission, a jam sequence is asserted and the transmission is retried after a random back off. CRS and COL have no effect in full duplex mode. The DMA block connects to external memory through its AHB bus interface. It contains receive and transmit FIFOs for buffering frame data. It loads the transmit FIFO and empties the receive FIFO using AHB bus master operations. Receive data is not sent to memory until the address checking logic has determined that the frame should be copied. Receive or transmit frames are stored in one or more buffers. Receive buffers have a fixed length of 128 bytes. Transmit buffers range in length between 0 and 2047 bytes, and up to 128 buffers are permitted per frame. The DMA block manages the transmit and receive framebuffer queues. These queues can hold multiple frames.
35.4.1 Clock
Synchronization module in the EMAC requires that the bus clock (MCK) runs at the speed of the macb_tx/rx_clk at least, which is 25 MHz at 100 Mbps, and 2.5 MHz at 10 Mbps.
35.4.2 Memory Interface

Frame data is transferred to and from the EMAC through the DMA interface. All transfers are 32-bit words and may be single accesses or bursts of 2, 3 or 4 words. Burst accesses do not cross sixteen-byte boundaries. Bursts of 4 words are the default data transfer; single accesses or bursts of less than four words may be used to transfer data at the beginning or the end of a buffer. The DMA controller performs six types of operation on the bus. In order of priority, these are: 1. 2. 3. 4. 5. 6. 35.4.2.1 FIFO The FIFO depths are 128 bytes for receive and 128 bytes for transmit and are a function of the system clock speed, memory latency and network speed. Data is typically transferred into and out of the FIFOs in bursts of four words. For receive, a bus request is asserted when the FIFO contains four words and has space for 28 more. For transmit, a bus request is generated when there is space for four words, or when there is space for 27 words if the next transfer is to be only one or two words. Receive buffer manager write Receive buffer manager read Transmit data DMA read Receive data DMA write Transmit buffer manager read Transmit buffer manager write

949
Thus the bus latency must be less than the time it takes to load the FIFO and transmit or receive three words (112 bytes) of data. At 100 Mbit/s, it takes 8960 ns to transmit or receive 112 bytes of data. In addition, six master clock cycles should be allowed for data to be loaded from the bus and to propagate through the FIFOs. For a 133 MHz master clock this takes 45 ns, making the bus latency requirement 8915 ns. 35.4.2.2 Receive Buffers Received frames, including CRC/FCS optionally, are written to receive buffers stored in memory. Each receive buffer is 128 bytes long. The start location for each receive buffer is stored in memory in a list of receive buffer descriptors at a location pointed to by the receive buffer queue pointer register. The receive buffer start location is a word address. For the first buffer of a frame, the start location can be offset by up to three bytes depending on the value written to bits 14 and 15 of the network configuration register. If the start location of the buffer is offset the available length of the first buffer of a frame is reduced by the corresponding number of bytes. Each list entry consists of two words, the first being the address of the receive buffer and the second being the receive status. If the length of a receive frame exceeds the buffer length, the status word for the used buffer is written with zeroes except for the start of frame bit and the offset bits, if appropriate. Bit zero of the address field is written to one to show the buffer has been used. The receive buffer manager then reads the location of the next receive buffer and fills that with receive frame data. The final buffer descriptor status word contains the complete frame status. Refer to Table 35-1 for details of the receive buffer descriptor list.
Table 35-1. Receive Buffer Descriptor Entry Bit Function Word 0 31:2 1 0 Address of beginning of buffer Wrap - marks last descriptor in receive buffer descriptor list. Ownership - needs to be zero for the EMAC to write data to the receive buffer. The EMAC sets this to one once it has successfully written a frame to memory. Software has to clear this bit before the buffer can be used again. Word 1 31 30 29 28 27 26 25 24 23 22 21 20 19:17 16 15 Global all ones broadcast address detected Multicast hash match Unicast hash match External address match Reserved for future use Specific address register 1 match Specific address register 2 match Specific address register 3 match Specific address register 4 match Type ID match VLAN tag detected (i.e., type id of 0x8100) Priority tag detected (i.e., type id of 0x8100 and null VLAN identifier) VLAN priority (only valid if bit 21 is set) Concatenation format indicator (CFI) bit (only valid if bit 21 is set) End of frame - when set the buffer contains the end of a frame. If end of frame is not set, then the only other valid status are bits 12, 13 and 14.

950
Table 35-1. Receive Buffer Descriptor Entry (Continued) Bit 14 Function Start of frame - when set the buffer contains the start of a frame. If both bits 15 and 14 are set, then the buffer contains a whole frame. Receive buffer offset - indicates the number of bytes by which the data in the first buffer is offset from the word address. Updated with the current values of the network configuration register. If jumbo frame mode is enabled through bit 3 of the network configuration register, then bits 13:12 of the receive buffer descriptor entry are used to indicate bits 13:12 of the frame length. Length of frame including FCS (if selected). Bits 13:12 are also used if jumbo frame mode is selected.
13:12
11:0
To receive frames, the buffer descriptors must be initialized by writing an appropriate address to bits 31 to 2 in the first word of each list entry. Bit zero must be written with zero. Bit one is the wrap bit and indicates the last entry in the list. The start location of the receive buffer descriptor list must be written to the receive buffer queue pointer register before setting the receive enable bit in the network control register to enable receive. As soon as the receive block starts writing received frame data to the receive FIFO, the receive buffer manager reads the first receive buffer location pointed to by the receive buffer queue pointer register. If the filter block then indicates that the frame should be copied to memory, the receive data DMA operation starts writing data into the receive buffer. If an error occurs, the buffer is recovered. If the current buffer pointer has its wrap bit set or is the 1024th descriptor, the next receive buffer location is read from the beginning of the receive descriptor list. Otherwise, the next receive buffer location is read from the next word in memory. There is an 11-bit counter to count out the 2048 word locations of a maximum length, receive buffer descriptor list. This is added with the value originally written to the receive buffer queue pointer register to produce a pointer into the list. A read of the receive buffer queue pointer register returns the pointer value, which is the queue entry currently being accessed. The counter is reset after receive status is written to a descriptor that has its wrap bit set or rolls over to zero after 1024 descriptors have been accessed. The value written to the receive buffer pointer register may be any word-aligned address, provided that there are at least 2048 word locations available between the pointer and the top of the memory. Section 3.6 of the AMBA 2.0 specification states that bursts should not cross 1K boundaries. As receive buffer manager writes are bursts of two words, to ensure that this does not occur, it is best to write the pointer register with the least three significant bits set to zero. As receive buffers are used, the receive buffer manager sets bit zero of the first word of the descriptor to indicate used. If a receive error is detected the receive buffer currently being written is recovered. Previous buffers are not recovered. Software should search through the used bits in the buffer descriptors to find out how many frames have been received. It should be checking the start-of-frame and end-of-frame bits, and not rely on the value returned by the receive buffer queue pointer register which changes continuously as more buffers are used. For CRC errored frames, excessive length frames or length field mismatched frames, all of which are counted in the statistics registers, it is possible that a frame fragment might be stored in a sequence of receive buffers. Software can detect this by looking for start of frame bit set in a buffer following a buffer with no end of frame bit set. For a properly working Ethernet system, there should be no excessively long frames or frames greater than 128 bytes with CRC/FCS errors. Collision fragments are less than 128 bytes long. Therefore, it is a rare occurrence to find a frame fragment in a receive buffer. If bit zero is set when the receive buffer manager reads the location of the receive buffer, then the buffer has already been used and cannot be used again until software has processed the frame and cleared bit zero. In this case, the DMA block sets the buffer not available bit in the receive status register and triggers an interrupt. If bit zero is set when the receive buffer manager reads the location of the receive buffer and a frame is being received, the frame is discarded and the receive resource error statistics register is incremented. A receive overrun condition occurs when bus was not granted in time or because HRESP was not OK (bus error). In a receive overrun condition, the receive overrun interrupt is asserted and the buffer currently being written is recovered. The next frame received with an address that is recognized reuses the buffer. If bit 17 of the network configuration register is set, the FCS of received frames shall not be copied to memory. The frame length indicated in the receive status field shall be reduced by four bytes in this case.

951
35.4.2.3 Transmit Buffer Frames to be transmitted are stored in one or more transmit buffers. Transmit buffers can be between 0 and 2047 bytes long, so it is possible to transmit frames longer than the maximum length specified in IEEE Standard 802.3. Zero length buffers are allowed. The maximum number of buffers permitted for each transmit frame is 128. The start location for each transmit buffer is stored in memory in a list of transmit buffer descriptors at a location pointed to by the transmit buffer queue pointer register. Each list entry consists of two words, the first being the byte address of the transmit buffer and the second containing the transmit control and status. Frames can be transmitted with or without automatic CRC generation. If CRC is automatically generated, pad is also automatically generated to take frames to a minimum length of 64 bytes. Table 35-2 on page 952 defines an entry in the transmit buffer descriptor list. To transmit frames, the buffer descriptors must be initialized by writing an appropriate byte address to bits 31 to 0 in the first word of each list entry. The second transmit buffer descriptor is initialized with control information that indicates the length of the buffer, whether or not it is to be transmitted with CRC and whether the buffer is the last buffer in the frame. After transmission, the control bits are written back to the second word of the first buffer along with the used bit and other status information. Bit 31 is the used bit which must be zero when the control word is read if transmission is to happen. It is written to one when a frame has been transmitted. Bits 27, 28 and 29 indicate various transmit error conditions. Bit 30 is the wrap bit which can be set for any buffer within a frame. If no wrap bit is encountered after 1024 descriptors, the queue pointer rolls over to the start in a similar fashion to the receive queue. The transmit buffer queue pointer register must not be written while transmit is active. If a new value is written to the transmit buffer queue pointer register, the queue pointer resets itself to point to the beginning of the new queue. If transmit is disabled by writing to bit 3 of the network control, the transmit buffer queue pointer register resets to point to the beginning of the transmit queue. Note that disabling receive does not have the same effect on the receive queue pointer. Once the transmit queue is initialized, transmit is activated by writing to bit 9, the Transmit Start bit of the network control register. Transmit is halted when a buffer descriptor with its used bit set is read, or if a transmit error occurs, or by writing to the transmit halt bit of the network control register. (Transmission is suspended if a pause frame is received while the pause enable bit is set in the network configuration register.) Rewriting the start bit while transmission is active is allowed. Transmission control is implemented with a Tx_go variable which is readable in the transmit status register at bit location 3. The Tx_go variable is reset when:
z z z z z
Transmit is disabled A buffer descriptor with its ownership bit set is read A new value is written to the transmit buffer queue pointer register Bit 10, tx_halt, of the network control register is written There is a transmit error such as too many retries or a transmit underrun.
To set tx_go, write to bit 9, tx_start, of the network control register. Transmit halt does not take effect until any ongoing transmit finishes. If a collision occurs during transmission of a multi-buffer frame, transmission automatically restarts from the first buffer of the frame. If a used bit is read midway through transmission of a multi-buffer frame, this is treated as a transmit error. Transmission stops, tx_er is asserted and the FCS is bad. If transmission stops due to a transmit error, the transmit queue pointer resets to point to the beginning of the transmit queue. Software needs to re-initialize the transmit queue after a transmit error. If transmission stops due to a used bit being read at the start of the frame, the transmission queue pointer is not reset and transmit starts from the same transmit buffer descriptor when the transmit start bit is written.
Table 35-2. Transmit Buffer Descriptor Entry Bit Function Word 0 31:0 Byte Address of buffer Word 1

952
Table 35-2. Transmit Buffer Descriptor Entry Bit Function Used. Needs to be zero for the EMAC to read data from the transmit buffer. The EMAC sets this to one for the first buffer of a frame once it has been successfully transmitted. 31 Software has to clear this bit before the buffer can be used again.
Note:
30 29 28 27 26:17 16 15 14:11 10:0
This bit is only set for the first buffer in a frame unlike receive where all buffers have the Used bit set once used.
Wrap. Marks last descriptor in transmit buffer descriptor list. Retry limit exceeded, transmit error detected Transmit underrun, occurs either when hresp is not OK (bus error) or the transmit data could not be fetched in time or when buffers are exhausted in mid frame. Buffers exhausted in mid frame Reserved No CRC. When set, no CRC is appended to the current frame. This bit only needs to be set for the last buffer of a frame. Last buffer. When set, this bit indicates the last buffer in the current frame has been reached. Reserved Length of buffer
35.4.3 Transmit Block

This block transmits frames in accordance with the Ethernet IEEE 802.3 CSMA/CD protocol. Frame assembly starts by adding preamble and the start frame delimiter. Data is taken from the transmit FIFO a word at a time. Data is transmitted least significant nibble first. If necessary, padding is added to increase the frame length to 60 bytes. CRC is calculated as a 32-bit polynomial. This is inverted and appended to the end of the frame, taking the frame length to a minimum of 64 bytes. If the No CRC bit is set in the second word of the last buffer descriptor of a transmit frame, neither pad nor CRC are appended. In full-duplex mode, frames are transmitted immediately. Back-to-back frames are transmitted at least 96 bit times apart to guarantee the interframe gap. In half-duplex mode, the transmitter checks carrier sense. If asserted, it waits for it to de-assert and then starts transmission after the interframe gap of 96 bit times. If the collision signal is asserted during transmission, the transmitter transmits a jam sequence of 32 bits taken from the data register and then retry transmission after the back off time has elapsed. The back-off time is based on an XOR of the 10 least significant bits of the data coming from the transmit FIFO and a 10bit pseudo random number generator. The number of bits used depends on the number of collisions seen. After the first collision, 1 bit is used, after the second 2, and so on up to 10. Above 10, all 10 bits are used. An error is indicated and no further attempts are made if 16 attempts cause collisions. If transmit DMA underruns, bad CRC is automatically appended using the same mechanism as jam insertion and the tx_er signal is asserted. For a properly configured system, this should never happen. If the back pressure bit is set in the network control register in half duplex mode, the transmit block transmits 64 bits of data, which can consist of 16 nibbles of 1011 or in bit-rate mode 64 1s, whenever it sees an incoming frame to force a collision. This provides a way of implementing flow control in half-duplex mode.

953
35.4.4 Pause Frame Support

The start of an 802.3 pause frame is as follows:
Table 35-3. Start of an 802.3 Pause Frame Destination Address 0x0180C2000001 Source Address 6 bytes Type (Mac Control Frame) 0x8808 Pause Opcode 0x0001 Pause Time 2 bytes
The network configuration register contains a receive pause enable bit (13). If a valid pause frame is received, the pause time register is updated with the frames pause time, regardless of its current contents and regardless of the state of the configuration register bit 13. An interrupt (12) is triggered when a pause frame is received, assuming it is enabled in the interrupt mask register. If bit 13 is set in the network configuration register and the value of the pause time register is nonzero, no new frame is transmitted until the pause time register has decremented to zero. The loading of a new pause time, and hence the pausing of transmission, only occurs when the EMAC is configured for full-duplex operation. If the EMAC is configured for half-duplex, there is no transmission pause, but the pause frame received interrupt is still triggered. A valid pause frame is defined as having a destination address that matches either the address stored in specific address register 1 or matches 0x0180C2000001 and has the MAC control frame type ID of 0x8808 and the pause opcode of 0x0001. Pause frames that have FCS or other errors are treated as invalid and are discarded. Valid pause frames received increment the Pause Frame Received statistic register. The pause time register decrements every 512 bit times (i.e., 128 rx_clks in nibble mode) once transmission has stopped. For test purposes, the register decrements every rx_clk cycle once transmission has stopped if bit 12 (retry test) is set in the network configuration register. If the pause enable bit (13) is not set in the network configuration register, then the decrementing occurs regardless of whether transmission has stopped or not. An interrupt (13) is asserted whenever the pause time register decrements to zero (assuming it is enabled in the interrupt mask register).
35.4.5 Receive Block

The receive block checks for valid preamble, FCS, alignment and length, presents received frames to the DMA block and stores the frames destination address for use by the address checking block. If, during frame reception, the frame is found to be too long or rx_er is asserted, a bad frame indication is sent to the DMA block. The DMA block then ceases sending data to memory. At the end of frame reception, the receive block indicates to the DMA block whether the frame is good or bad. The DMA block recovers the current receive buffer if the frame was bad. The receive block signals the register block to increment the alignment error, the CRC (FCS) error, the short frame, long frame, jabber error, the receive symbol error statistics and the length field mismatch statistics. The enable bit for jumbo frames in the network configuration register allows the EMAC to receive jumbo frames of up to 10240 bytes in size. This operation does not form part of the IEEE802.3 specification and is disabled by default. When jumbo frames are enabled, frames received with a frame size greater than 10240 bytes are discarded.
35.4.6 Address Checking Block

The address checking (or filter) block indicates to the DMA block which receive frames should be copied to memory. Whether a frame is copied depends on what is enabled in the network configuration register, the state of the external match pin, the contents of the specific address and hash registers and the frames destination address. In this implementation of the EMAC, the frames source address is not checked. Provided that bit 18 of the Network Configuration register is not set, a frame is not copied to memory if the EMAC is transmitting in half duplex mode at the time a destination address is received. If bit 18 of the Network Configuration register is set, frames can be received while transmitting in half-duplex mode.

954
Ethernet frames are transmitted a byte at a time, least significant bit first. The first six bytes (48 bits) of an Ethernet frame make up the destination address. The first bit of the destination address, the LSB of the first byte of the frame, is the group/individual bit: this is One for multicast addresses and Zero for unicast. The All Ones address is the broadcast address, and a special case of multicast. The EMAC supports recognition of four specific addresses. Each specific address requires two registers, specific address register bottom and specific address register top. Specific address register bottom stores the first four bytes of the destination address and specific address register top contains the last two bytes. The addresses stored can be specific, group, local or universal. The destination address of received frames is compared against the data stored in the specific address registers once they have been activated. The addresses are deactivated at reset or when their corresponding specific address register bottom is written. They are activated when specific address register top is written. If a receive frame address matches an active address, the frame is copied to memory. The following example illustrates the use of the address match registers for a MAC address of 21:43:65:87:A9:CB. Preamble 55 SFD D5 DA (Octet0 - LSB) 21 DA(Octet 1) 43 DA(Octet 2) 65 DA(Octet 3) 87 DA(Octet 4) A9 DA (Octet5 - MSB) CB SA (LSB) 00 SA 00 SA 00 SA 00 SA 00 SA (MSB) 43 SA (LSB) 21 The sequence above shows the beginning of an Ethernet frame. Byte order of transmission is from top to bottom as shown. For a successful match to specific address 1, the following address matching registers must be set up:
z z
Base address + 0x98 0x87654321 (Bottom) Base address + 0x9C 0x0000CBA9 (Top) Base address + 0xB8 0x00004321
And for a successful match to the Type ID register, the following should be set up:
z
35.4.7 Broadcast Address

The broadcast address of 0xFFFFFFFFFFFF is recognized if the no broadcast bit in the network configuration register is zero.
35.4.8 Hash Addressing

The hash address register is 64 bits long and takes up two locations in the memory map. The least significant bits are stored in hash register bottom and the most significant bits in hash register top. The unicast hash enable and the multicast hash enable bits in the network configuration register enable the reception of hash matched frames. The destination address is reduced to a 6-bit index into the 64-bit hash register using the following hash function. The hash function is an exclusive or of every sixth bit of the destination address.

955
hash_index[5] = da[5] ^ da[11] ^ da[17] ^ da[23] ^ da[29] ^ da[35] ^ da[41] ^ da[47] hash_index[4] = da[4] ^ da[10] ^ da[16] ^ da[22] ^ da[28] ^ da[34] ^ da[40] ^ da[46] hash_index[3] = da[3] ^ da[09] ^ da[15] ^ da[21] ^ da[27] ^ da[33] ^ da[39] ^ da[45] hash_index[2] = da[2] ^ da[08] ^ da[14] ^ da[20] ^ da[26] ^ da[32] ^ da[38] ^ da[44] hash_index[1] = da[1] ^ da[07] ^ da[13] ^ da[19] ^ da[25] ^ da[31] ^ da[37] ^ da[43] hash_index[0] = da[0] ^ da[06] ^ da[12] ^ da[18] ^ da[24] ^ da[30] ^ da[36] ^ da[42]
da[0] represents the least significant bit of the first byte received, that is, the multicast/unicast indicator, and da[47] represents the most significant bit of the last byte received. If the hash index points to a bit that is set in the hash register, then the frame is matched according to whether the frame is multicast or unicast. A multicast match is signalled if the multicast hash enable bit is set. da[0] is 1 and the hash index points to a bit set in the hash register. A unicast match is signalled if the unicast hash enable bit is set. da[0] is 0 and the hash index points to a bit set in the hash register. To receive all multicast frames, the hash register should be set with all ones and the multicast hash enable bit should be set in the network configuration register.
35.4.9 Copy All Frames (or Promiscuous Mode)

If the copy all frames bit is set in the network configuration register, then all non-errored frames are copied to memory. For example, frames that are too long, too short, or have FCS errors or rx_er asserted during reception are discarded and all others are received. Frames with FCS errors are copied to memory if bit 19 in the network configuration register is set.
35.4.10 Type ID Checking

The contents of the type_id register are compared against the length/type ID of received frames (i.e., bytes 13 and 14). Bit 22 in the receive buffer descriptor status is set if there is a match. The reset state of this register is zero which is unlikely to match the length/type ID of any valid Ethernet frame. Note: A type ID match does not affect whether a frame is copied to memory.
35.4.11 VLAN Support

An Ethernet encoded 802.1Q VLAN tag looks like this:
Table 35-4. 802.1Q VLAN Tag TPID (Tag Protocol Identifier) 16 bits 0x8100 TCI (Tag Control Information) 16 bits First 3 bits priority, then CFI bit, last 12 bits VID
The VLAN tag is inserted at the 13th byte of the frame, adding an extra four bytes to the frame. If the VID (VLAN identifier) is null (0x000), this indicates a priority-tagged frame. The MAC can support frame lengths up to 1536 bytes, 18 bytes more than the original Ethernet maximum frame length of 1518 bytes. This is achieved by setting bit 8 in the network configuration register. The following bits in the receive buffer descriptor status word give information about VLAN tagged frames:
z z z z
Bit 21 set if receive frame is VLAN tagged (i.e. type id of 0x8100) Bit 20 set if receive frame is priority tagged (i.e. type id of 0x8100 and null VID). (If bit 20 is set bit 21 is set also.) Bit 19, 18 and 17 set to priority if bit 21 is set Bit 16 set to CFI if bit 21 is set

956
35.4.12 Wake-on-LAN Support

The receive block supports Wake-on-LAN by detecting the following events on incoming receive frames:
z z z z
Magic packet ARP request to the device IP address Specific address 1 filter match Multicast hash filter match
If one of these events occurs Wake-on-LAN detection is indicated by asserting the wol output pin for 64 rx_clk cycles. These events can be individually enabled through bits[19:16] of the Wake-on-LAN register. Also, for Wake-onLAN detection to occur, receive enable must be set in the network control register, however a receive buffer does not have to be available. wol assertion due to ARP request, specific address 1 or multicast filter events occurs even if the frame is errored. For magic packet events, the frame must be correctly formed and error free. A magic packet event is detected if all of the following are true:
z z z z z
Magic packet events are enabled through bit 16 of the Wake-on-LAN register The frames destination address matches specific address 1 The frame is correctly formed with no errors The frame contains at least 6 bytes of 0xFF for synchronization There are 16 repetitions of the contents of specific address 1 register immediately following the synchronization ARP request events are enabled through bit 17 of the Wake-on-LAN register Broadcasts are allowed by bit 5 in the network configuration register The frame has a broadcast destination address (bytes 1 to 6) The frame has a type ID field of 0x0806 (bytes 13 and 14) The frame has an ARP operation field of 0x0001 (bytes 21 and 22) The least significant 16 bits of the frames ARP target protocol address (bytes 41 and 42) match the value programmed in bits[15:0] of the Wake-on-LAN register
An ARP request event is detected if all of the following are true:

z z z z z z
The decoding of the ARP fields adjusts automatically if a VLAN tag is detected within the frame. The reserved value of 0x0000 for the Wake-on-LAN target address value does not cause an ARP request event, even if matched by the frame. A specific address 1 filter match event occurs if all of the following are true:
z z
Specific address 1 events are enabled through bit 18 of the Wake-on-LAN register The frames destination address matches the value programmed in the specific address 1 registers Multicast hash events are enabled through bit 19 of the Wake-on-LAN register Multicast hash filtering is enabled through bit 6 of the network configuration register The frames destination address matches against the multicast hash filter The frames destination address is not a broadcast
A multicast filter match event occurs if all of the following are true:
z z z z
35.4.13 PHY Maintenance

The register EMAC_MAN enables the EMAC to communicate with a PHY by means of the MDIO interface. It is used during auto-negotiation to ensure that the EMAC and the PHY are configured for the same speed and duplex configuration. The PHY maintenance register is implemented as a shift register. Writing to the register starts a shift operation which is signalled as complete when bit two is set in the network status register (about 2000 MCK cycles later when bit ten is set to zero, and bit eleven is set to one in the network configuration register). An interrupt is generated as this bit is set. During this time, the MSB of the register is output on the MDIO pin and the LSB updated from the MDIO pin with each MDC cycle. This causes transmission of a PHY management frame on MDIO.

957
Reading during the shift operation returns the current contents of the shift register. At the end of management operation, the bits have shifted back to their original locations. For a read operation, the data bits are updated with data read from the PHY. It is important to write the correct values to the register to ensure a valid PHY management frame is produced. The MDIO interface can read IEEE 802.3 clause 45 PHYs as well as clause 22 PHYs. To read clause 45 PHYs, bits[31:28] should be written as 0x0011. For a description of MDC generation, see the network configuration register in the Network Control Register on page 964.
35.4.14 Physical Interface

Depending on products, the Ethernet MAC is capable of interfacing to RMII or MII Interface. The RMII bit in the EMAC_USRIO register controls the interface that is selected. When this bit is set, the RMII interface is selected, else the MII interface is selected. The MII and RMII interfaces are capable of both 10 Mb/s and 100 Mb/s data rates as described in the IEEE 802.3u standard. The signals used by the RMII interface are described in Table 35-5.
Table 35-5. Pin Configuration Pin Name ETXCK_EREFCK ECRS ECOL ERXDV ERX0 - ERX3 ERXER ERXCK ETXEN ETX0-ETX3 ETXER ETXEN: Transmit Enable ETX0 - ETX1: 2-bit Transmit Data ECRSDV: Carrier Sense/Data Valid ERX0 - ERX1: 2-bit Receive Data ERXER: Receive Error RMII EREFCK: Reference Clock
The intent of the RMII is to provide a reduced pin count alternative to the IEEE 802.3u MII. It uses 2 bits for transmit (ETX0 and ETX1) and two bits for receive (ERX0 and ERX1). There is a Transmit Enable (ETXEN), a Receive Error (ERXER), a Carrier Sense (ECRS_DV), and a 50 MHz Reference Clock (ETXCK_EREFCK) for 100Mb/s data rate. 35.4.14.1 RMII Transmit and Receive Operation The RMII maps the signals in a more pin-efficient manner. The transmit and receive bits are converted from a 4-bit parallel format to a 2-bit parallel scheme that is clocked at twice the rate. The carrier sense and data valid signals are combined into the ECRSDV signal. This signal contains information on carrier sense, FIFO status, and validity of the data. Transmit error bit (ETXER) and collision detect (ECOL) are not used in RMII mode.

958
35.5
Programming Interface
35.5.1 Initialization
35.5.1.1 Configuration Initialization of the EMAC configuration (e.g., loop-back mode, frequency ratios) must be done while the transmit and receive circuits are disabled. See the description of the network control register and network configuration register earlier in this document. To change loop-back mode, the following sequence of operations must be followed: 1. 2. 3. Note: Write to network control register to disable transmit and receive circuits. Write to network control register to change loop-back mode. Write to network control register to re-enable transmit or receive circuits. These writes to network control register cannot be combined in any way.
35.5.1.2 Receive Buffer List Receive data is written to areas of data (i.e., buffers) in system memory. These buffers are listed in another data structure that also resides in main memory. This data structure (receive buffer queue) is a sequence of descriptor entries as defined in Receive Buffer Descriptor Entry on page 950. It points to this data structure.
Figure 35-2. Receive Buffer List
Receive Buffer 0 Receive Buffer Queue Pointer (MAC Register) Receive Buffer 1
Receive Buffer N Receive Buffer Descriptor List (In memory) (In memory)
To create the list of buffers: 1. 2. 3. 4. 5. Allocate a number (n) of buffers of 128 bytes in system memory. Allocate an area 2n words for the receive buffer descriptor entry in system memory and create n entries in this list. Mark all entries in this list as owned by EMAC, i.e., bit 0 of word 0 set to 0. If less than 1024 buffers are defined, the last descriptor must be marked with the wrap bit (bit 1 in word 0 set to 1). Write address of receive buffer descriptor entry to EMAC register receive_buffer queue pointer. The receive circuits can then be enabled by writing to the address recognition registers and then to the network control register.

959
35.5.1.3 Transmit Buffer List Transmit data is read from areas of data (the buffers) in system memory These buffers are listed in another data structure that also resides in main memory. This data structure (Transmit Buffer Queue) is a sequence of descriptor entries (as defined in Table 35-2 on page 952) that points to this data structure. To create this list of buffers: 1. 2. 3. 4. 5. Allocate a number (n) of buffers of between 1 and 2047 bytes of data to be transmitted in system memory. Up to 128 buffers per frame are allowed. Allocate an area 2n words for the transmit buffer descriptor entry in system memory and create N entries in this list. Mark all entries in this list as owned by EMAC, i.e. bit 31 of word 1 set to 0. If fewer than 1024 buffers are defined, the last descriptor must be marked with the wrap bit bit 30 in word 1 set to 1. Write address of transmit buffer descriptor entry to EMAC register transmit_buffer queue pointer. The transmit circuits can then be enabled by writing to the network control register.
35.5.1.4 Address Matching The EMAC register-pair hash address and the four specific address register-pairs must be written with the required values. Each register-pair comprises a bottom register and top register, with the bottom register being written first. The address matching is disabled for a particular register-pair after the bottom-register has been written and re-enabled when the top register is written. See Address Checking Block on page 954. for details of address matching. Each register-pair may be written at any time, regardless of whether the receive circuits are enabled or disabled. 35.5.1.5 Interrupts There are 15 interrupt conditions that are detected within the EMAC. These are ORed to make a single interrupt. Depending on the overall system design, this may be passed through a further level of interrupt collection (interrupt controller). On receipt of the interrupt signal, the CPU enters the interrupt handler (Refer to the Interrupt Controller). To ascertain which interrupt has been generated, read the interrupt status register. Note that this register clears itself when read. At reset, all interrupts are disabled. To enable an interrupt, write to interrupt enable register with the pertinent interrupt bit set to 1. To disable an interrupt, write to interrupt disable register with the pertinent interrupt bit set to 1. To check whether an interrupt is enabled or disabled, read interrupt mask register: if the bit is set to 1, the interrupt is disabled. 35.5.1.6 Transmitting Frames To set up a frame for transmission: 1. 2. 3. 4. 5. 6. 7. 8. Enable transmit in the network control register. Allocate an area of system memory for transmit data. This does not have to be contiguous, varying byte lengths can be used as long as they conclude on byte borders. Set-up the transmit buffer list. Set the network control register to enable transmission and enable interrupts. Write data for transmission into these buffers. Write the address to transmit buffer descriptor queue pointer. Write control and length to word one of the transmit buffer descriptor entry. Write to the transmit start bit in the network control register.
35.5.1.7 Receiving Frames When a frame is received and the receive circuits are enabled, the EMAC checks the address and, in the following cases, the frame is written to system memory:
z z z z
If it matches one of the four specific address registers. If it matches the hash address function. If it is a broadcast address (0xFFFFFFFFFFFF) and broadcasts are allowed. If the EMAC is configured to copy all frames.

960
The register receive buffer queue pointer points to the next entry (see Table 35-1 on page 950) and the EMAC uses this as the address in system memory to write the frame to. Once the frame has been completely and successfully received and written to system memory, the EMAC then updates the receive buffer descriptor entry with the reason for the address match and marks the area as being owned by software. Once this is complete an interrupt receive complete is set. Software is then responsible for handling the data in the buffer and then releasing the buffer by writing the ownership bit back to 0. If the EMAC is unable to write the data at a rate to match the incoming frame, then an interrupt receive overrun is set. If there is no receive buffer available, i.e., the next buffer is still owned by software, the interrupt receive buffer not available is set. If the frame is not successfully received, a statistic register is incremented and the frame is discarded without informing software.

961
35.6
Ethernet MAC 10/100 (EMAC) User Interface
Table 35-6. Register Mapping Offset 0x00 0x04 0x08 0x0C 0x10 0x14 0x18 0x1C 0x20 0x24 0x28 0x2C 0x30 0x34 0x38 0x3C 0x40 0x44 0x48 0x4C 0x50 0x54 0x58 0x5C 0x60 0x64 0x68 0x6C 0x70 0x74 0x78 0x7C 0x80 0x84 0x88 Register Network Control Register Network Configuration Register Network Status Register Reserved Reserved Transmit Status Register Receive Buffer Queue Pointer Register Transmit Buffer Queue Pointer Register Receive Status Register Interrupt Status Register Interrupt Enable Register Interrupt Disable Register Interrupt Mask Register Phy Maintenance Register Pause Time Register Pause Frames Received Register Frames Transmitted Ok Register Single Collision Frames Register Multiple Collision Frames Register Frames Received Ok Register Frame Check Sequence Errors Register Alignment Errors Register Deferred Transmission Frames Register Late Collisions Register Excessive Collisions Register Transmit Underrun Errors Register Carrier Sense Errors Register Receive Resource Errors Register Receive Overrun Errors Register Receive Symbol Errors Register Excessive Length Errors Register Receive Jabbers Register Undersize Frames Register SQE Test Errors Register Received Length Field Mismatch Register EMAC_TSR EMAC_RBQP EMAC_TBQP EMAC_RSR EMAC_ISR EMAC_IER EMAC_IDR EMAC_IMR EMAC_MAN EMAC_PTR EMAC_PFR EMAC_FTO EMAC_SCF EMAC_MCF EMAC_FRO EMAC_FCSE EMAC_ALE EMAC_DTF EMAC_LCOL EMAC_ECOL EMAC_TUND EMAC_CSE EMAC_RRE EMAC_ROV EMAC_RSE EMAC_ELE EMAC_RJA EMAC_USF EMAC_STE EMAC_RLE Read-write Read-write Read-write Read-write Read-write Write-only Write-only Read-only Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_7FFF 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 Name EMAC_NCR EMAC_NCFGR EMAC_NSR Access Read-write Read-write Read-only Reset 0 0x800 -

962
Table 35-6. Register Mapping (Continued) Offset 0x90 0x94 0x98 0x9C 0xA0 0xA4 0xA8 0xAC 0xB0 0xB4 0xB8 0xC0 0xC4 0xC8 - 0xFC Register Hash Register Bottom [31:0] Register Hash Register Top [63:32] Register Specific Address 1 Bottom Register Specific Address 1 Top Register Specific Address 2 Bottom Register Specific Address 2 Top Register Specific Address 3 Bottom Register Specific Address 3 Top Register Specific Address 4 Bottom Register Specific Address 4 Top Register Type ID Checking Register User Input/Output Register Wake on LAN Register Reserved Name EMAC_HRB EMAC_HRT EMAC_SA1B EMAC_SA1T EMAC_SA2B EMAC_SA2T EMAC_SA3B EMAC_SA3T EMAC_SA4B EMAC_SA4T EMAC_TID EMAC_USRIO EMAC_WOL Access Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Reset 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000 0x0000_0000

963
35.6.1 Network Control Register

31 23 15 7 WESTAT
EMAC_NCR 0xF802C000 Read-write

30 22 14 6 INCSTAT 29 21 13 5 CLRSTAT 28 20 12 4 MPE 27 19 11 3 TE 26 18 10 THALT 2 RE 25 17 9 TSTART 1 LLB 24 16 8 BP 0 LB
LB: LoopBack Asserts the loopback signal to the PHY. LLB: Loopback local
Connects txd to rxd, tx_en to rx_dv, forces full duplex and drives rx_clk and tx_clk with MCK divided by 4. rx_clk and tx_clk may glitch as the EMAC is switched into and out of internal loop back. It is important that receive and transmit circuits have already been disabled when making the switch into and out of internal loop back.
RE: Receive enable When set, enables the EMAC to receive data. When reset, frame reception stops immediately and the receive FIFO is cleared. The receive queue pointer register is unaffected. TE: Transmit enable When set, enables the Ethernet transmitter to send data. When reset transmission, stops immediately, the transmit FIFO and control registers are cleared and the transmit queue pointer register resets to point to the start of the transmit descriptor list. MPE: Management port enable Set to one to enable the management port. When zero, forces MDIO to high impedance state and MDC low. CLRSTAT: Clear statistics registers This bit is write only. Writing a one clears the statistics registers. INCSTAT: Increment statistics registers This bit is write only. Writing a one increments all the statistics registers by one for test purposes. WESTAT: Write enable for statistics registers Setting this bit to one makes the statistics registers writable for functional test purposes. BP: Back pressure If set in half duplex mode, forces collisions on all received frames. TSTART: Start transmission Writing one to this bit starts transmission. THALT: Transmit halt Writing one to this bit halts transmission as soon as any ongoing frame transmission ends.

964
35.6.2 Network Configuration Register

31 23 15 RBOF 7 UNI 6 MTI
EMAC_NCFGR 0xF802C004 Read-write

30 22 14 29 21 13 PAE 5 NBC 28 20 12 RTY 4 CAF 27 19 IRXFCS 11 CLK 3 JFRAME 2 26 18 EFRHD 10 25 17 DRFCS 9 1 FD 24 16 RLCE 8 BIG 0 SPD
SPD: Speed
Set to 1 to indicate 100 Mbit/s operation, 0 for 10 Mbit/s. The value of this pin is reflected on the speed pin.
FD: Full Duplex

If set to 1, the transmit block ignores the state of collision and carrier sense and allows receive while transmitting. Also controls the half_duplex pin.
CAF: Copy All Frames

When set to 1, all valid frames are received.
JFRAME: Jumbo Frames

Set to one to enable jumbo frames of up to 10240 bytes to be accepted.
NBC: No Broadcast
When set to 1, frames addressed to the broadcast address of all ones are not received.
MTI: Multicast Hash Enable

When set, multicast frames are received when the 6-bit hash function of the destination address points to a bit that is set in the hash register.
UNI: Unicast Hash Enable

When set, unicast frames are received when the 6-bit hash function of the destination address points to a bit that is set in the hash register.
BIG: Receive 1536 bytes frames

Setting this bit means the EMAC receives frames up to 1536 bytes in length. Normally, the EMAC would reject any frame above 1518 bytes.
CLK: MDC clock divider

Set according to system clock speed. This determines by what number system clock is divided to generate MDC. For conformance with 802.3, MDC must not exceed 2.5 MHz (MDC is only active during MDIO read and write operations).

965
.
Value 0 1 2 3 Name MCK_8 MCK_16 MCK_32 MCK_64 Description MCK divided by 8 (MCK up to 20 MHz). MCK divided by 16 (MCK up to 40 MHz). MCK divided by 32 (MCK up to 80 MHz). MCK divided by 64 (MCK up to 160 MHz).
RTY: Retry test

Must be set to zero for normal operation. If set to one, the back off between collisions is always one slot time. Setting this bit to one helps testing the too many retries condition. Also used in the pause frame tests to reduce the pause counters decrement time from 512 bit times, to every rx_clk cycle.
PAE: Pause Enable

When set, transmission pauses when a valid pause frame is received.
RBOF: Receive Buffer Offset

Indicates the number of bytes by which the received data is offset from the start of the first receive buffer.
Value 0 1 2 3 Name OFFSET_0 OFFSET_1 OFFSET_2 OFFSET_3 Description No offset from start of receive buffer. One-byte offset from start of receive buffer. Two-byte offset from start of receive buffer. Three-byte offset from start of receive buffer.
RLCE: Receive Length field Checking Enable

When set, frames with measured lengths shorter than their length fields are discarded. Frames containing a type ID in bytes 13 and 14 length/type ID = 0600 are not counted as length errors.
DRFCS: Discard Receive FCS

When set, the FCS field of received frames is not copied to memory.
EFRHD:
Enable Frames to be received in half-duplex mode while transmitting.
IRXFCS: Ignore RX FCS

When set, frames with FCS/CRC errors are not rejected and no FCS error statistics are counted. For normal operation, this bit must be set to 0.

966
35.6.3 Network Status Register

31 23 15 7
EMAC_NSR 0xF802C008 Read-only

30 22 14 6 29 21 13 5 28 20 12 4 27 19 11 3 26 18 10 2 IDLE 25 17 9 1 MDIO 24 16 8 0
MDIO
Returns status of the mdio_in pin. Use the PHY maintenance register for reading managed frames rather than this bit.
IDLE
0 = The PHY logic is running. 1 = The PHY management logic is idle (i.e., has completed).

967
35.6.4 Transmit Status Register

31 23 15 7
EMAC_TSR 0xF802C014 Read-write

30 22 14 6 UND 29 21 13 5 COMP 28 20 12 4 BEX 27 19 11 3 TGO 26 18 10 2 RLES 25 17 9 1 COL 24 16 8 0 UBR
This register, when read, provides details of the status of a transmit. Once read, individual bits may be cleared by writing 1 to them. It is not possible to set a bit to 1 by writing to the register.
UBR: Used Bit Read

Set when a transmit buffer descriptor is read with its used bit set. Cleared by writing a one to this bit.
COL: Collision Occurred

Set by the assertion of collision. Cleared by writing a one to this bit.
RLES: Retry Limit exceeded

Cleared by writing a one to this bit.
TGO: Transmit Go
If high transmit is active.
BEX: Buffers exhausted mid frame

If the buffers run out during transmission of a frame, then transmission stops, FCS shall be bad and tx_er asserted. Cleared by writing a one to this bit.
COMP: Transmit Complete

Set when a frame has been transmitted. Cleared by writing a one to this bit.
UND: Transmit Underrun

Set when transmit DMA was not able to read data from memory, either because the bus was not granted in time, because a not OK hresp(bus error) was returned or because a used bit was read midway through frame transmission. If this occurs, the transmitter forces bad CRC. Cleared by writing a one to this bit.

968
35.6.5 Receive Buffer Queue Pointer Register

31
EMAC_RBQP 0xF802C018 Read-write

30 29 28 ADDR 27 26 25 24
23
22
21
20 ADDR
19
18
17
16
15
14
13
12 ADDR
11
10
5 ADDR
This register points to the entry in the receive buffer queue (descriptor list) currently being used. It is written with the start location of the receive buffer descriptor list. The lower order bits increment as buffers are used up and wrap to their original values after either 1024 buffers or when the wrap bit of the entry is set. Reading this register returns the location of the descriptor currently being accessed. This value increments as buffers are used. Software should not use this register for determining where to remove received frames from the queue as it constantly changes as new frames are received. Software should instead work its way through the buffer descriptor queue checking the used bits. Receive buffer writes also comprise bursts of two words and, as with transmit buffer reads, it is recommended that bit 2 is always written with zero to prevent a burst crossing a 1K boundary, in violation of section 3.6 of the AMBA specification.
ADDR: Receive buffer queue pointer address

Written with the address of the start of the receive queue, reads as a pointer to the current buffer being used.

969
35.6.6 Transmit Buffer Queue Pointer Register

31
EMAC_TBQP 0xF802C01C Read-write

30 29 28 ADDR 27 26 25 24
23
22
21
20 ADDR
19
18
17
16
15
14
13
12 ADDR
11
10
5 ADDR
This register points to the entry in the transmit buffer queue (descriptor list) currently being used. It is written with the start location of the transmit buffer descriptor list. The lower order bits increment as buffers are used up and wrap to their original values after either 1024 buffers or when the wrap bit of the entry is set. This register can only be written when bit 3 in the transmit status register is low. As transmit buffer reads consist of bursts of two words, it is recommended that bit 2 is always written with zero to prevent a burst crossing a 1K boundary, in violation of section 3.6 of the AMBA specification.
ADDR: Transmit buffer queue pointer address

Written with the address of the start of the transmit queue, reads as a pointer to the first buffer of the frame being transmitted or about to be transmitted.

970
35.6.7 Receive Status Register

31 23 15 7
EMAC_RSR 0xF802C020 Read-write

30 22 14 6 29 21 13 5 28 20 12 4 27 19 11 3 26 18 10 2 OVR 25 17 9 1 REC 24 16 8 0 BNA
This register, when read, provides details of the status of a receive. Once read, individual bits may be cleared by writing 1 to them. It is not possible to set a bit to 1 by writing to the register.
BNA: Buffer Not Available

An attempt was made to get a new buffer and the pointer indicated that it was owned by the processor. The DMA rereads the pointer each time a new frame starts until a valid pointer is found. This bit is set at each attempt that fails even if it has not had a successful pointer read since it has been cleared. Cleared by writing a one to this bit.
REC: Frame Received

One or more frames have been received and placed in memory. Cleared by writing a one to this bit.
OVR: Receive Overrun

The DMA block was unable to store the receive frame to memory, either because the bus was not granted in time or because a not OK hresp(bus error) was returned. The buffer is recovered if this happens. Cleared by writing a one to this bit.

971
35.6.8 Interrupt Status Register

31 23 15 7 TCOMP
EMAC_ISR 0xF802C024 Read-write

30 22 14 WOL 6 TXERR 29 21 13 PTZ 5 RLEX 28 20 12 PFRE 4 TUND 27 19 11 HRESP 3 TXUBR 26 18 10 ROVR 2 RXUBR 25 17 9 1 RCOMP 24 16 8 0 MFD
MFD: Management Frame Done

The PHY maintenance register has completed its operation. Cleared on read.

A frame has been stored in memory. Cleared on read.
RXUBR: Receive Used Bit Read

Set when a receive buffer descriptor is read with its used bit set. Cleared on read.
TXUBR: Transmit Used Bit Read

Set when a transmit buffer descriptor is read with its used bit set. Cleared on read.
TUND: Ethernet Transmit Buffer Underrun

The transmit DMA did not fetch frame data in time for it to be transmitted or hresp returned not OK. Also set if a used bit is read mid-frame or when a new transmit queue pointer is written. Cleared on read.
RLEX: Retry Limit Exceeded

Cleared on read.
TXERR: Transmit Error

Transmit buffers exhausted in mid-frame - transmit error. Cleared on read.

Set when a frame has been transmitted. Cleared on read.

Set when the receive overrun status bit gets set. Cleared on read.
HRESP: Hresp not OK

Set when the DMA block sees a bus error. Cleared on read.

972
PFRE: Pause Frame Received

Indicates a valid pause has been received. Cleared on a read.

Set when the pause time register, 0x38 decrements to zero. Cleared on a read.
WOL: Wake On LAN

Set when a WOL event has been triggered (This flag can be set even if the EMAC is not clocked). Cleared on a read.

973
35.6.9 Interrupt Enable Register

31 23 15 7 TCOMP
EMAC_IER 0xF802C028 Write-only

30 22 14 WOL 6 TXERR 29 21 13 PTZ 5 RLE 28 20 12 PFR 4 TUND 27 19 11 HRESP 3 TXUBR 26 18 10 ROVR 2 RXUBR 25 17 9 1 RCOMP 24 16 8 0 MFD
MFD: Management Frame sent

Enable management done interrupt.



Enable transmit used bit read interrupt.

Enable transmit underrun interrupt.

Enable retry limit exceeded interrupt.
TXERR
Enable transmit buffers exhausted in mid-frame interrupt.

Enable transmit complete interrupt.

HRESP: Hresp not OK

Enable Hresp not OK interrupt.
PFR: Pause Frame Received

Enable pause frame received interrupt.

974

Enable pause time zero interrupt.
WOL: Wake On LAN

Enable Wake On LAN interrupt.

975
35.6.10 Interrupt Disable Register

31 23 15 7 TCOMP
EMAC_IDR 0xF802C02C Write-only


Disable management done interrupt.



Disable transmit used bit read interrupt.

Disable transmit underrun interrupt.

Disable retry limit exceeded interrupt.
TXERR
Disable transmit buffers exhausted in mid-frame interrupt.

Disable transmit complete interrupt.

HRESP: Hresp not OK

Disable Hresp not OK interrupt.

Disable pause frame received interrupt.

976

Disable pause time zero interrupt.
WOL: Wake On LAN

Disable Wake On LAN interrupt.

977
35.6.11 Interrupt Mask Register

31 23 15 7 TCOMP
EMAC_IMR 0xF802C030 Read-only


Management done interrupt masked.

Receive complete interrupt masked.

Receive used bit read interrupt masked.

Transmit used bit read interrupt masked.

Transmit underrun interrupt masked.

Retry limit exceeded interrupt masked.
TXERR
Transmit buffers exhausted in mid-frame interrupt masked.

Transmit complete interrupt masked.

Receive overrun interrupt masked.
HRESP: Hresp not OK

Hresp not OK interrupt masked.

Pause frame received interrupt masked.

978

Pause time zero interrupt masked.
WOL: Wake On LAN

Wake On LAN interrupt masked.

979
35.6.12 PHY Maintenance Register

31 SOF 23 PHYA 15 22 21
EMAC_MAN 0xF802C034 Read-write

30 29 RW 20 REGA 12 DATA 7 6 5 4 DATA 3 2 1 0 19 18 28 27 26 PHYA 17 CODE 11 10 9 8 16 25 24
14
13
DATA
For a write operation this is written with the data to be written to the PHY. After a read operation this contains the data read from the PHY.
CODE:
Must be written to 10. Reads as written.
REGA: Register Address

Specifies the register in the PHY to access.
PHYA: PHY Address RW: Read-write

10 is read; 01 is write. Any other value is an invalid PHY management frame
SOF: Start of frame

Must be written 01 for a valid frame.

980
35.6.13 Pause Time Register

31 23 15
EMAC_PTR 0xF802C038 Read-write

30 22 14 29 21 13 28 20 12 PTIME 7 6 5 4 PTIME 3 2 1 0 27 19 11 26 18 10 25 17 9 24 16 8
PTIME: Pause Time

Stores the current value of the pause time register which is decremented every 512 bit times.

981
35.6.14 Hash Register Bottom

31
EMAC_HRB 0xF802C090 Read-write

30 29 28 ADDR 27 26 25 24
23
22
21
20 ADDR
19
18
17
16
15
14
13
12 ADDR
11
10
4 ADDR
ADDR:
Bits 31:0 of the hash address register. See Hash Addressing on page 955.
35.6.15 Hash Register Top

31
EMAC_HRT 0xF802C094 Read-write

30 29 28 ADDR 27 26 25 24
23
22
21
20 ADDR
19
18
17
16
15
14
13
12 ADDR
11
10
4 ADDR
ADDR:
Bits 63:32 of the hash address register. See Hash Addressing on page 955.

982
35.6.16 Specific Address 1 Bottom Register

31
EMAC_SA1B 0xF802C098 Read-write

30 29 28 ADDR 27 26 25 24
23
22
21
20 ADDR
19
18
17
16
15
14
13
12 ADDR
11
10
4 ADDR
ADDR
Least significant bits of the destination address. Bit zero indicates whether the address is multicast or unicast and corresponds to the least significant bit of the first byte received.
35.6.17 Specific Address 1 Top Register

31 23 15
EMAC_SA1T 0xF802C09C Read-write

30 22 14 29 21 13 28 20 12 ADDR 7 6 5 4 ADDR 3 2 1 0 27 19 11 26 18 10 25 17 9 24 16 8
ADDR
The most significant bits of the destination address, that is bits 47 to 32.

983

31
EMAC_SA2B 0xF802C0A0 Read-write

30 29 28 ADDR 27 26 25 24
23
22
21
20 ADDR
19
18
17
16
15
14
13
12 ADDR
11
10
4 ADDR
ADDR

31 23 15
EMAC_SA2T 0xF802C0A4 Read-write

30 22 14 29 21 13 28 20 12 ADDR 7 6 5 4 ADDR 3 2 1 0 27 19 11 26 18 10 25 17 9 24 16 8
ADDR

984

31
EMAC_SA3B 0xF802C0A8 Read-write

30 29 28 ADDR 27 26 25 24
23
22
21
20 ADDR
19
18
17
16
15
14
13
12 ADDR
11
10
4 ADDR
ADDR

31 23 15
EMAC_SA3T 0xF802C0AC Read-write

30 22 14 29 21 13 28 20 12 ADDR 7 6 5 4 ADDR 3 2 1 0 27 19 11 26 18 10 25 17 9 24 16 8
ADDR

985

31
EMAC_SA4B 0xF802C0B0 Read-write

30 29 28 ADDR 27 26 25 24
23
22
21
20 ADDR
19
18
17
16
15
14
13
12 ADDR
11
10
4 ADDR
ADDR

31 23 15
EMAC_SA4T 0xF802C0B4 Read-write

30 22 14 29 21 13 28 20 12 ADDR 7 6 5 4 ADDR 3 2 1 0 27 19 11 26 18 10 25 17 9 24 16 8
ADDR

986
35.6.24 Type ID Checking Register

31 23 15
EMAC_TID 0xF802C0B8 Read-write

30 22 14 29 21 13 28 20 12 TID 7 6 5 4 TID 3 2 1 0 27 19 11 26 18 10 25 17 9 24 16 8
TID: Type ID checking

For use in comparisons with received frames TypeID/Length field.

987
35.6.25 User Input/Output Register

31 23 15 7
EMAC_USRIO 0xF802C0C0 Read-write

30 22 14 6 29 21 13 5 28 20 12 4 27 19 11 3 26 18 10 2 25 17 9 1 CLKEN 24 16 8 0 RMII
RMII: Reduce MII

When set, this bit enables the RMII operation mode.
CLKEN: Clock Enable

When set, this bit enables the transceiver input clock. Setting this bit to 0 reduces power consumption when the treasurer is not used.

988
35.6.26 Wake-on-LAN Register

Name: Access:
31 23 15
EMAC_WOL Read-write
30 22 14 29 21 13 28 20 12 IP 7 6 5 4 IP 3 2 1 0 27 19 MTI 11 26 18 SA1 10 25 17 ARP 9 24 16 MAG 8
IP: ARP request IP address

Written to define the least significant 16 bits of the target IP address that is matched to generate a Wake-on-LAN event. A value of zero does not generate an event, even if this is matched by the received frame.
MAG: Magic packet event enable

When set, magic packet events causes the wol output to be asserted.
ARP: ARP request event enable

When set, ARP request events causes the wol output to be asserted.
SA1: Specific address register 1 event enable

When set, specific address 1 events causes the wol output to be asserted.
MTI: Multicast hash event enable

When set, multicast hash events causes the wol output to be asserted.

989
35.6.27 EMAC Statistic Registers

These registers reset to zero on a read and stick at all ones when they count to their maximum value. They should be read frequently enough to prevent loss of data. The receive statistics registers are only incremented when the receive enable bit is set in the network control register. To write to these registers, bit 7 must be set in the network control register. The statistics register block contains the following registers. 35.6.27.1 Pause Frames Received Register Name: Address: Access:
31 23 15
EMAC_PFR 0xF802C03C Read-write

30 22 14 29 21 13 28 20 12 FROK 7 6 5 4 FROK 3 2 1 0 27 19 11 26 18 10 25 17 9 24 16 8
FROK: Pause Frames received OK

A 16-bit register counting the number of good pause frames received. A good frame has a length of 64 to 1518 (1536 if bit 8 set in network configuration register) and has no FCS, alignment or receive symbol errors.

990
35.6.27.2 Frames Transmitted OK Register Name: Address: Access:

31 23
EMAC_FTO 0xF802C040 Read-write

30 22 29 21 28 20 FTOK 27 19 26 18 25 17 24 16
15
14
13
12 FTOK
11
10
4 FTOK
FTOK: Frames Transmitted OK

A 24-bit register counting the number of frames successfully transmitted, i.e., no underrun and not too many retries.

991
35.6.27.3 Single Collision Frames Register Name: Address: Access:

31 23 15
EMAC_SCF 0xF802C044 Read-write

30 22 14 29 21 13 28 20 12 SCF 7 6 5 4 SCF 3 2 1 0 27 19 11 26 18 10 25 17 9 24 16 8
SCF: Single Collision Frames

A 16-bit register counting the number of frames experiencing a single collision before being successfully transmitted, i.e., no underrun.
35.6.27.4 Multicollision Frames Register Name: Address: Access:

31 23 15
EMAC_MCF 0xF802C048 Read-write

30 22 14 29 21 13 28 20 12 MCF 7 6 5 4 MCF 3 2 1 0 27 19 11 26 18 10 25 17 9 24 16 8
MCF: Multicollision Frames

A 16-bit register counting the number of frames experiencing between two and fifteen collisions prior to being successfully transmitted, i.e., no underrun and not too many retries.

992
35.6.27.5 Frames Received OK Register Name: Address: Access:

31 23
EMAC_FRO 0xF802C04C Read-write

30 22 29 21 28 20 FROK 27 19 26 18 25 17 24 16
15
14
13
12 FROK
11
10
4 FROK
FROK: Frames Received OK

A 24-bit register counting the number of good frames received, i.e., address recognized and successfully copied to memory. A good frame is of length 64 to 1518 bytes (1536 if bit 8 set in network configuration register) and has no FCS, alignment or receive symbol errors.
35.6.27.6 Frames Check Sequence Errors Register Name: Address: Access:

31 23 15 7
EMAC_FCSE 0xF802C050 Read-write

30 22 14 6 29 21 13 5 28 20 12 4 FCSE 27 19 11 3 26 18 10 2 25 17 9 1 24 16 8 0
FCSE: Frame Check Sequence Errors

An 8-bit register counting frames that are an integral number of bytes, have bad CRC and are between 64 and 1518 bytes in length (1536 if bit 8 set in network configuration register). This register is also incremented if a symbol error is detected and the frame is of valid length and has an integral number of bytes.

993
35.6.27.7 Alignment Errors Register Name: Address: Access:

31 23 15 7
EMAC_ALE 0xF802C054 Read-write

30 22 14 6 29 21 13 5 28 20 12 4 ALE 27 19 11 3 26 18 10 2 25 17 9 1 24 16 8 0
ALE: Alignment Errors

An 8-bit register counting frames that are not an integral number of bytes long and have bad CRC when their length is truncated to an integral number of bytes and are between 64 and 1518 bytes in length (1536 if bit 8 set in network configuration register). This register is also incremented if a symbol error is detected and the frame is of valid length and does not have an integral number of bytes.
35.6.27.8 Deferred Transmission Frames Register Name: Address: Access:

31 23 15
EMAC_DTF 0xF802C058 Read-write

30 22 14 29 21 13 28 20 12 DTF 7 6 5 4 DTF 3 2 1 0 27 19 11 26 18 10 25 17 9 24 16 8
DTF: Deferred Transmission Frames

A 16-bit register counting the number of frames experiencing deferral due to carrier sense being active on their first attempt at transmission. Frames involved in any collision are not counted nor are frames that experienced a transmit underrun.

994
35.6.27.9 Late Collisions Register Name: Address: Access:

31 23 15 7
EMAC_LCOL 0xF802C05C Read-write

30 22 14 6 29 21 13 5 28 20 12 4 LCOL 27 19 11 3 26 18 10 2 25 17 9 1 24 16 8 0
LCOL: Late Collisions

An 8-bit register counting the number of frames that experience a collision after the slot time (512 bits) has expired. A late collision is counted twice; i.e., both as a collision and a late collision.
35.6.27.10 Excessive Collisions Register Name: Address: Access:

31 23 15 7
EMAC_ECOL 0xF802C060 Read-write

30 22 14 6 29 21 13 5 28 20 12 4 EXCOL 27 19 11 3 26 18 10 2 25 17 9 1 24 16 8 0
EXCOL: Excessive Collisions

An 8-bit register counting the number of frames that failed to be transmitted because they experienced 16 collisions.

995
35.6.27.11 Transmit Underrun Errors Register Name: Address: Access:

31 23 15 7
EMAC_TUND 0xF802C064 Read-write

30 22 14 6 29 21 13 5 28 20 12 4 TUND 27 19 11 3 26 18 10 2 25 17 9 1 24 16 8 0
TUND: Transmit Underruns

An 8-bit register counting the number of frames not transmitted due to a transmit DMA underrun. If this register is incremented, then no other statistics register is incremented.
35.6.27.12 Carrier Sense Errors Register Name: Address: Access:

31 23 15 7
EMAC_CSE 0xF802C068 Read-write

30 22 14 6 29 21 13 5 28 20 12 4 CSE 27 19 11 3 26 18 10 2 25 17 9 1 24 16 8 0
CSE: Carrier Sense Errors

An 8-bit register counting the number of frames transmitted where carrier sense was not seen during transmission or where carrier sense was deasserted after being asserted in a transmit frame without collision (no underrun). Only incremented in halfduplex mode. The only effect of a carrier sense error is to increment this register. The behavior of the other statistics registers is unaffected by the detection of a carrier sense error.

996
35.6.27.13 Receive Resource Errors Register Name: Address: Access:

31 23 15
EMAC_RRE 0xF802C06C Read-write

30 22 14 29 21 13 28 20 12 RRE 7 6 5 4 RRE 3 2 1 0 27 19 11 26 18 10 25 17 9 24 16 8
RRE: Receive Resource Errors

A 16-bit register counting the number of frames that were address matched but could not be copied to memory because no receive buffer was available.
35.6.27.14 Receive Overrun Errors Register Name: Address: Access:

31 23 15 7
EMAC_ROV 0xF802C070 Read-write

30 22 14 6 29 21 13 5 28 20 12 4 ROVR 27 19 11 3 26 18 10 2 25 17 9 1 24 16 8 0

An 8-bit register counting the number of frames that are address recognized but were not copied to memory due to a receive DMA overrun.

997
35.6.27.15 Receive Symbol Errors Register Name: Address: Access:

31 23 15 7
EMAC_RSE 0xF802C074 Read-write

30 22 14 6 29 21 13 5 28 20 12 4 RSE 27 19 11 3 26 18 10 2 25 17 9 1 24 16 8 0
RSE: Receive Symbol Errors

An 8-bit register counting the number of frames that had rx_er asserted during reception. Receive symbol errors are also counted as an FCS or alignment error if the frame is between 64 and 1518 bytes in length (1536 if bit 8 is set in the network configuration register). If the frame is larger, it is recorded as a jabber error.
35.6.27.16 Excessive Length Errors Register Name: Address: Access:

31 23 15 7
EMAC_ELE 0xF802C078 Read-write

30 22 14 6 29 21 13 5 28 20 12 4 EXL 27 19 11 3 26 18 10 2 25 17 9 1 24 16 8 0
EXL: Excessive Length Errors

An 8-bit register counting the number of frames received exceeding 1518 bytes (1536 if bit 8 set in network configuration register) in length but do not have either a CRC error, an alignment error nor a receive symbol error.

998
35.6.27.17 Receive Jabbers Register Name: Address: Access:

31 23 15 7
EMAC_RJA 0xF802C07C Read-write

30 22 14 6 29 21 13 5 28 20 12 4 RJB 27 19 11 3 26 18 10 2 25 17 9 1 24 16 8 0
RJB: Receive Jabbers

An 8-bit register counting the number of frames received exceeding 1518 bytes (1536 if bit 8 set in network configuration register) in length and have either a CRC error, an alignment error or a receive symbol error.
35.6.27.18 Undersize Frames Register Name: Address: Access:

31 23 15 7
EMAC_USF 0xF802C080 Read-write

30 22 14 6 29 21 13 5 28 20 12 4 USF 27 19 11 3 26 18 10 2 25 17 9 1 24 16 8 0
USF: Undersize frames

An 8-bit register counting the number of frames received less than 64 bytes in length but do not have either a CRC error, an alignment error or a receive symbol error.

999
35.6.27.19 SQE Test Errors Register Name: Address: Access:

31 23 15 7
EMAC_STE 0xF802C084 Read-write

30 22 14 6 29 21 13 5 28 20 12 4 SQER 27 19 11 3 26 18 10 2 25 17 9 1 24 16 8 0
SQER: SQE test errors

An 8-bit register counting the number of frames where col was not asserted within 96 bit times (an interframe gap) of tx_en being deasserted in half duplex mode.
35.6.27.20 Received Length Field Mismatch Register Name: Address: Access:

31 23 15 7
EMAC_RLE 0xF802C088 Read-write

30 22 14 6 29 21 13 5 28 20 12 4 RLFM 27 19 11 3 26 18 10 2 25 17 9 1 24 16 8 0
RLFM: Receive Length Field Mismatch

An 8-bit register counting the number of frames received that have a measured length shorter than that extracted from its length field. Checking is enabled through bit 16 of the network configuration register. Frames containing a type ID in bytes 13 and 14 (i.e., length/type ID = 0x0600) are not counted as length field errors, neither are excessive length frames.

1000
36.
36.1
Image Sensor Interface (ISI)

Description
The Image Sensor Interface (ISI) connects a CMOS-type image sensor to the processor and provides image capture in various formats. It does data conversion, if necessary, before the storage in memory through DMA. The ISI supports color CMOS image sensor and grayscale image sensors with a reduced set of functionalities. In grayscale mode, the data stream is stored in memory without any processing and so is not compatible with the LCD controller. Internal FIFOs on the preview and codec paths are used to store the incoming data. The RGB output on the preview path is compatible with the LCD controller. This module outputs the data in RGB format (LCD compatible) and has scaling capabilities to make it compliant to the LCD display resolution (See Table 36-3 on page 1005). Several input formats such as preprocessed RGB or YCbCr are supported through the data bus interface. It supports two modes of synchronization: 1. 2. The hardware with ISI_VSYNC and ISI_HSYNC signals The International Telecommunication Union Recommendation ITU-R BT.656-4 Start-of-Active-Video (SAV) and End-of-Active-Video (EAV) synchronization sequence.
Using EAV/SAV for synchronization reduces the pin count (ISI_VSYNC, ISI_HSYNC not used). The polarity of the synchronization pulse is programmable to comply with the sensor signals.
Table 36-1. I/O Description Signal ISI_VSYNC ISI_HSYNC ISI_DATA[11..0] ISI_MCK ISI_PCK Dir IN IN IN OUT IN Description Vertical Synchronization Horizontal Synchronization Sensor Pixel Data Master Clock Provided to the Image Sensor Pixel Clock Provided by the Image Sensor
Figure 36-1. ISI Connection Example

Image Sensor Image Sensor Interface
data[11..0] CLK PCLK VSYNC HSYNC
ISI_DATA[11..0] ISI_MCK ISI_PCK ISI_VSYNC ISI_HSYNC

1001
36.2
Compatible with an Embedded 32-bit Microcontroller ITU-R BT. 601/656 8-bit Mode External Interface Support Supports up to 12-bit Grayscale CMOS Sensors Support for ITU-R BT.656-4 SAV and EAV Synchronization Vertical and Horizontal Resolutions up to 2048 x 2048 Preview Path up to 640*480 32 Bytes FIFO on Codec Path 32 Bytes FIFO on Preview Path Support for Packed Data Formatting for YCbCr 4:2:2 Formats Preview Scaler to Generate Smaller Size image Programmable Frame Capture Rate VGA, QVGA, CIF, QCIF Formats Supported for LCD Preview Custom Formats with Horizontal and Vertical Preview Size as Multiples of 16 Also Supported for LCD Preview

1002
36.3
Block Diagram
Figure 36-2. Image Sensor Interface Block Diagram
Hsync/Len Vsync/Fen
Timing Signals Interface
Camera Interrupt Controller From Rx buffers
Config Registers Camera Interrupt Request Line
APB Interface
CCIR-656 Embedded Timing Decoder(SAV/EAV) CMOS sensor Pixel input up to 12 bit YCbCr 4:2:2 8:8:8 RGB 5:6:5
APB Clock Domain AHB Clock Domain

Camera AHB Master Interface Scatter Mode Support
Frame Rate
Preview path
Clipping + Color Conversion YCC to RGB
Pixel Clock Domain

2-D Image Scaler Pixel Formatter
Pixel Sampling Module
Core Video Arbiter
CMOS sensor pixel clock input
Clipping + Color Conversion RGB to YCC
Packed Formatter
Rx Direct Capture FIFO
codec_on
Codec path

1003
AHB bus
Rx Direct Display FIFO
APB bus
36.4
The Image Sensor Interface (ISI) supports direct connection to the ITU-R BT. 601/656 8-bit mode compliant sensors and up to 12-bit grayscale sensors. It receives the image data stream from the image sensor on the 12-bit data bus. This module receives up to 12 bits for data, the horizontal and vertical synchronizations and the pixel clock. The reduced pin count alternative for synchronization is supported for sensors that embed SAV (start of active video) and EAV (end of active video) delimiters in the data stream. The Image Sensor Interface interrupt line is connected to the Advanced Interrupt Controller and can trigger an interrupt at the beginning of each frame and at the end of a DMA frame transfer. If the SAV/EAV synchronization is used, an interrupt can be triggered on each delimiter event. For 8-bit color sensors, the data stream received can be in several possible formats: YCbCr 4:2:2, RGB 8:8:8, RGB 5:6:5 and may be processed before the storage in memory. When the preview DMA channel is configured and enabled, the preview path is activated and an RGB frame is moved to memory. The preview path frame rate is configured with the FRATE field of the ISI_CFG1 register. When the codec DMA channel is configured and enabled, the codec path is activated and a YCbCr 4:2:2 frame is captured as soon as the ISI_CDC field of the ISI_CTRL register is set to 1. When the FULL field of the ISI_CFG1 register is set to 1, both preview DMA channel and codec DMA channel can operate simultaneously. When the FULL field of the ISI_CFG1 register is set to 0, a hardware scheduler checks the FRATE field. If its value is zero, a preview frame is skipped and a codec frame is moved to memory instead. If its value is different from zero, at least one free frame slot is available. The scheduler postpones the codec frame to that free available frame slot. The data stream may be sent on both preview path and codec path if the bit ISI_CDC in the ISI_CTRL is one. To optimize the bandwidth, the codec path should be enabled only when a capture is required. In grayscale mode, the input data stream is stored in memory without any processing. The 12-bit data, which represent the grayscale level for the pixel, is stored in memory one or two pixels per word, depending on the GS_MODE bit in the ISI_CFG2 register. The codec datapath is not available when grayscale image is selected. A frame rate counter allows users to capture all frames or 1 out of every 2 to 8 frames.
36.4.1 Data Timing

The two data timings using horizontal and vertical synchronization and EAV/SAV sequence synchronization are shown in Figure 36-3 and Figure 36-4. In the VSYNC/HSYNC synchronization, the valid data is captured with the active edge of the pixel clock (ISI_PCK), after SFD lines of vertical blanking and SLD pixel clock periods delay programmed in the control register. The ITU-RBT.656-4 defines the functional timing for an 8-bit wide interface. There are two timing reference signals, one at the beginning of each video data block SAV (0xFF000080) and one at the end of each video data block EAV(0xFF00009D). Only data sent between EAV and SAV is captured. Horizontal blanking and vertical blanking are ignored. Use of the SAV and EAV synchronization eliminates the ISI_VSYNC and ISI_HSYNC signals from the interface, thereby reducing the pin count. In order to retrieve both frame and line synchronization properly, at least one line of vertical blanking is mandatory.

1004
Figure 36-3. HSYNC and VSYNC Synchronization

Frame
ISI_VSYNC
1 line
ISI_HSYNC
ISI_PCK DATA[7..0]
Y Cb Y Cr Y Cb Y Cr Y Cb Y Cr
Figure 36-4. SAV and EAV Sequence Synchronization

ISII_PCK DATA[7..0]
FF 00 00 SAV 80 Y Cb Y Cr Y Cb Y Cr Active Video Y Y Cr Y Cb FF 00 00 EAV 9D
36.4.2 Data Ordering

The RGB color space format is required for viewing images on a display screen preview, and the YCbCr color space format is required for encoding. All the sensors do not output the YCbCr or RGB components in the same order. The ISI allows the user to program the same component order as the sensor, reducing software treatments to restore the right format.
Table 36-2. Data Ordering in YCbCr Mode Mode Default Mode1 Mode2 Mode3 Byte 0 Cb(i) Cr(i) Y(i) Y(i) Byte 1 Y(i) Y(i) Cb(i) Cr(i) Byte 2 Cr(i) Cb(i) Y(i+1) Y(i+1) Byte 3 Y(i+1) Y(i+1) Cr(i) Cb(i)
Table 36-3. RGB Format in Default Mode, RGB_CFG = 00, No Swap Mode Byte Byte 0 Byte 1 RGB 8:8:8 Byte 2 Byte 3 Byte 0 Byte 1 RGB 5:6:5 Byte 2 Byte 3 R4(i+1) G2(i+1) R3(i+1) G1(i+1) R2(i+1) G0(i+1) R1(i+1) B4(i+1) R0(i+1) B3(i+1) G5(i+1) B2(i+1) G4(i+1) B1(i+1) G3(i+1) B0(i+1) B7(i) R7(i+1) R4(i) G2(i) B6(i) R6(i+1) R3(i) G1(i) B5(i) R5(i+1) R2(i) G0(i) B4(i) R4(i+1) R1(i) B4(i) B3(i) R3(i+1) R0(i) B3(i) B2(i) R2(i+1) G5(i) B2(i) B1(i) R1(i+1) G4(i) B1(i) B0(i) R0(i+1) G3(i) B0(i) D7 R7(i) G7(i) D6 R6(i) G6(i) D5 R5(i) G5(i) D4 R4(i) G4(i) D3 R3(i) G3(i) D2 R2(i) G2(i) D1 R1(i) G1(i) D0 R0(i) G0(i)

1005
Table 36-4. RGB Format, RGB_CFG = 10 (Mode 2), No Swap Mode Byte Byte 0 Byte 1 RGB 5:6:5 Byte 2 Byte 3 G2(i+1) B4(i+1) G1(i+1) B3(i+1) G0(i+1) B2(i+1) R4(i+1) B1(i+1) R3(i+1) B0(i+1) R2(i+1) G5(i+1) R1(i+1) G4(i+1) R0(i+1) G3(i+1) D7 G2(i) B4(i) D6 G1(i) B3(i) D5 G0(i) B2(i) D4 R4(i) B1(i) D3 R3(i) B0(i) D2 R2(i) G5(i) D1 R1(i) G4(i) D0 R0(i) G3(i)
Table 36-5. RGB Format in Default Mode, RGB_CFG = 00, Swap Activated Mode Byte Byte 0 Byte 1 RGB 8:8:8 Byte 2 Byte 3 Byte 0 Byte 1 RGB 5:6:5 Byte 2 Byte 3 G3(i+1) B0(i+1) G4(i+1) B1(i+1) G5(i+1) B2(i+1) R0(i+1) B3(i+1) R1(i+1) B4(i+1) R2(i+1) G0(i+1) R3(i+1) G1(i+1) R4(i+1) G2(i+1) B0(i) R0(i+1) G3(i) B0(i) B1(i) R1(i+1) G4(i) B1(i) B2(i) R2(i+1) G5(i) B2(i) B3(i) R3(i+1) R0(i) B3(i) B4(i) R4(i+1) R1(i) B4(i) B5(i) R5(i+1) R2(i) G0(i) B6(i) R6(i+1) R3(i) G1(i) B7(i) R7(i+1) R4(i) G2(i) D7 R0(i) G0(i) D6 R1(i) G1(i) D5 R2(i) G2(i) D4 R3(i) G3(i) D3 R4(i) G4(i) D2 R5(i) G5(i) D1 R6(i) G6(i) D0 R7(i) G7(i)
The RGB 5:6:5 input format is processed to be displayed as RGB 5:6:5 format, compliant with the 16-bit mode of the LCD controller.
36.4.3 Clocks
The sensor master clock (ISI_MCK) can be generated either by the Advanced Power Management Controller (APMC) through a Programmable Clock output or by an external oscillator connected to the sensor. None of the sensors embed a power management controller, so providing the clock by the APMC is a simple and efficient way to control power consumption of the system. Care must be taken when programming the system clock. The ISI has two clock domains, the sensor master clock and the pixel clock provided by sensor. The two clock domains are not synchronized, but the sensor master clock must be faster than the pixel clock.

1006
36.4.4 Preview Path

36.4.4.1 Scaling, Decimation (Subsampling) This module resizes captured 8-bit color sensor images to fit the LCD display format. The resize module performs only downscaling. The same ratio is applied for both horizontal and vertical resize, then a fractional decimation algorithm is applied. The decimation factor is a multiple of 1/16 and values 0 to 15 are forbidden.
Table 36-6. Decimation Factor Dec value Dec Factor 0->15 X 16 1 17 1.063 18 1.125 19 1.188 ... ... 124 7.750 125 7.813 126 7.875 127 7.938
Table 36-7. Decimation and Scaler Offset Values INPUT OUTPUT VGA 640*480 QVGA 320*240 CIF 352*288 QCIF 176*144 F 352*288 640*480 800*600 1280*1024 1600*1200 2048*1536
NA
16
20
32
40
51
16
32
40
64
80
102
16
26
33
56
66
85
32
53
66
113
133
170
Example: Input 1280*1024 Output = 640*480 Hratio = 1280/640 = 2 Vratio = 1024/480 = 2.1333 The decimation factor is 2 so 32/16.

1007
Figure 36-5. Resize Examples

1280 32/16 decimation 640
1024
480
1280
56/16 decimation 352
1024
288
36.4.4.2 Color Space Conversion This module converts YCrCb or YUV pixels to RGB color space. Clipping is performed to ensure that the samples value do not exceed the allowable range. The conversion matrix is defined below and is fully programmable:
C0 0 C1 Y Y off R G = C 0 C 2 C 3 C b C boff B C0 C4 0 C r C roff
Example of programmable value to convert YCrCb to RGB:

R = 1.164 ( Y 16 ) + 1.596 ( C r 128 ) G = 1.164 ( Y 16 ) 0.813 ( C r 128 ) 0.392 ( C b 128 B = 1.164 ( Y 16 ) + 2.107 ( C b 128 )
An example of programmable value to convert from YUV to RGB:

R = Y + 1.596 V G = Y 0.394 U 0.436 V B = Y + 2.032 U

1008
36.4.4.3 Memory Interface Preview datapath contains a data formatter that converts 8:8:8 pixel to RGB 5:6:5 format compliant with 16-bit format of the LCD controller. In general, when converting from a color channel with more bits to one with fewer bits, formatter module discards the lower-order bits. Example: Converting from RGB 8:8:8 to RGB 5:6:5, it discards the three LSBs from the red and blue channels, and two LSBs from the green channel. When grayscale mode is enabled, two memory formats are supported. One mode supports 2 pixels per word, and the other mode supports 1 pixel per word.
Table 36-8. Grayscale Memory Mapping Configuration for 12-bit Data GS_MODE 0 1 DATA[31:24] P_0[11:4] P_0[11:4] DATA[23:16] P_0[3:0], 0000 P_0[3:0], 0000 DATA[15:8] P_1[11:4] 0 DATA[7:0] P_1[3:0], 0000 0
36.4.4.4 FIFO and DMA Features Both preview and Codec datapaths contain FIFOs. These asynchronous buffers are used to safely transfer formatted pixels from Pixel clock domain to AHB clock domain. A video arbiter is used to manage FIFO thresholds and triggers a relevant DMA request through the AHB master interface. Thus, depending on FIFO state, a specified length burst is asserted. Regarding AHB master interface, it supports Scatter DMA mode through linked list operation. This mode of operation improves flexibility of image buffer location and allows the user to allocate two or more frame buffers. The destination frame buffers are defined by a series of Frame Buffer Descriptors (FBD). Each FBD controls the transfer of one entire frame and then optionally loads a further FBD to switch the DMA operation at another frame buffer address. The FBD is defined by a series of three words. The first one defines the current frame buffer address (named DMA_X_ADDR register), the second defines control information (named DMA_X_CTRL register) and the third defines the next descriptor address (named DMA_X_DSCR). DMA transfer mode with linked list support is available for both codec and preview datapath. The data to be transferred described by an FBD requires several burst accesses. In the example below, the use of 2 ping-pong frame buffers is described.
Example
The first FBD, stored at address 0x00030000, defines the location of the first frame buffer. This address is programmed in the ISI user interface DMA_P_DSCR. To enable Descriptor fetch operation DMA_P_CTRL register must be set to 0x00000001. LLI_0 and LLI_1 are the two descriptors of the Linked list. Destination Address: frame buffer ID0 0x02A000 (LLI_0.DMA_P_ADDR) Transfer 0 Control Information, fetch and writeback: 0x00000003 (LLI_0.DMA_P_CTRL) Next FBD address: 0x00030010 (LLI_0.DMA_P_DSCR) Second FBD, stored at address 0x00030010, defines the location of the second frame buffer. Destination Address: frame buffer ID1 0x0003A000 (LLI_1.DMA_P_ADDR Transfer 1 Control information fetch and writeback: 0x00000003 (LLI_1.DMA_P_CTRL) Next FBD address: 0x00030000, wrapping to first FBD (LLI_1.DMA_P_DSCR) Using this technique, several frame buffers can be configured through the linked list. Figure 36-6 illustrates a typical three frame buffer application. Frame n is mapped to frame buffer 0, frame n+1 is mapped to frame buffer 1, frame n+2 is mapped to Frame buffer 2, further frames wrap. A codec request occurs, and the full-size 4:2:2 encoded frame is stored in a dedicated memory space.

1009
Figure 36-6. Three Frame Buffers Application and Memory Mapping

Codec Request Codec Done
frame n-1
frame n
frame n+1
frame n+2
frame n+3
frame n+4
Memory Space
Frame Buffer 3
Frame Buffer 0
LCD
Frame Buffer 1
ISI config Space
4:2:2 Image Full ROI
36.4.5 Codec Path

36.4.5.1 Color Space Conversion Depending on user selection, this module can be bypassed so that input YCrCb stream is directly connected to the format converter module. If the RGB input stream is selected, this module converts RGB to YCrCb color space with the formulas given below:
Y Cr = Cb C0 C1 C2 C3 C4 C5 C6 C7 C8 Y off R G + Cr off B Cb off
An example of coefficients is given below:

Y = 0.257 R + 0.504 G + 0.098 B + 16 C r = 0.439 R 0.368 G 0.071 B + 128 C b = 0.148 R 0.291 G + 0.439 B + 128
36.4.5.2 Memory Interface Dedicated FIFOs are used to support packed memory mapping. YCrCb pixel components are sent in a single 32-bit word in a contiguous space (packed). Data is stored in the order of natural scan lines. Planar mode is not supported. 36.4.5.3 DMA Features
Like preview datapath, codec datapath DMA mode uses linked list operation.

1010
36.5
Image Sensor Interface (ISI) User Interface
Table 36-9. Register Mapping Offset 0x00 0x04 0x08 0x0C 0x10 0x14 0x18 0x1C 0x20 0x24 0x28 0x2C 0x30 0x34 0x38 0x3C 0x40 0x44 0x48 0x4C 0x50 0x54 0x58 0x5C-0xE0 0xE4 0xE8 0xEC-0xF8 0xFC Register ISI Configuration 1 Register ISI Configuration 2 Register ISI Preview Size Register ISI Preview Decimation Factor Register ISI CSC YCrCb To RGB Set 0 Register ISI CSC YCrCb To RGB Set 1 Register ISI CSC RGB To YCrCb Set 0 Register ISI CSC RGB To YCrCb Set 1 Register ISI CSC RGB To YCrCb Set 2 Register ISI Control Register ISI Status Register ISI Interrupt Enable Register ISI Interrupt Disable Register ISI Interrupt Mask Register DMA Channel Enable Register DMA Channel Disable Register DMA Channel Status Register DMA Preview Base Address Register DMA Preview Control Register DMA Preview Descriptor Address Register DMA Codec Base Address Register DMA Codec Control Register DMA Codec Descriptor Address Register Reserved Write Protection Control Register Write Protection Status Register Reserved Reserved Name ISI_CFG1 ISI_CFG2 ISI_PSIZE ISI_PDECF ISI_Y2R_SET0 ISI_Y2R_SET1 ISI_R2Y_SET0 ISI_R2Y_SET1 ISI_R2Y_SET2 ISI_CR ISI_SR ISI_IER ISI_IDR ISI_IMR ISI_DMA_CHER ISI_DMA_CHDR ISI_DMA_CHSR ISI_DMA_P_ADDR ISI_DMA_P_CTRL ISI_DMA_P_DSCR ISI_DMA_C_ADDR ISI_DMA_C_CTRL ISI_DMA_C_DSCR ISI_WPCR ISI_WPSR Access Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-write Write-only Read-only Write-only Write-only Read-only Write-only Write-only Read-only Read-write Read-write Read-write Read-write Read-write Read-write Read-write Read-only Reset Value 0x00000000 0x00000000 0x00000000 0x00000010 0x6832cc95 0x00007102 0x01324145 0x01245e38 0x01384a4b 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000
Note:
Several parts of the ISI controller use the pixel clock provided by the image sensor (ISI_PCK). Thus the user must first program the image sensor to provide this clock (ISI_PCK) before programming the Image Sensor Controller.

1011
36.5.1 ISI Configuration 1 Register

31
ISI_CFG1 0xF0034000 Read-write 0x00000000

30 29 28 SFD 27 26 25 24
23
22
21
20 SLD
19
18
17
16
15 7 CRC_SYNC
14 THMASK 6 EMB_SYNC
13
12 FULL 4 PIXCLK_POL
11 DISCR 3 VSYNC_POL
10
9 FRATE 1
2 HSYNC_POL
HSYNC_POL: Horizontal Synchronization Polarity

0: HSYNC active high. 1: HSYNC active low.
VSYNC_POL: Vertical Synchronization Polarity

0: VSYNC active high. 1: VSYNC active low.
PIXCLK_POL: Pixel Clock Polarity

0: Data is sampled on rising edge of pixel clock. 1: Data is sampled on falling edge of pixel clock.
EMB_SYNC: Embedded Synchronization

0: Synchronization by HSYNC, VSYNC. 1: Synchronization by embedded synchronization sequence SAV/EAV.
CRC_SYNC: Embedded Synchronization Correction

0: No CRC correction is performed on embedded synchronization. 1: CRC correction is performed. if the correction is not possible, the current frame is discarded and the CRC_ERR is set in the status register.
FRATE: Frame Rate [0..7]

0: All the frames are captured, else one frame every FRATE+1 is captured.
DISCR: Disable Codec Request

0 = Codec datapath DMA interface requires a request to restart. 1 = Codec datapath DMA automatically restarts.
FULL: Full Mode is Allowed

0: The codec frame is transferred to memory when an available frame slot is detected. 1: Both preview and codec DMA channels are operating simultaneously.

1012
THMASK: Threshold Mask

Value 0 1 2 Name BEATS_4 BEATS_8 BEATS_16 Description Only 4 beats AHB burst allowed Only 4 and 8 beats AHB burst allowed 4, 8 and 16 beats AHB burst allowed
SLD: Start of Line Delay

SLD pixel clock periods to wait before the beginning of a line.
SFD: Start of Frame Delay

SFD lines are skipped at the beginning of the frame.

1013
36.5.2 ISI Configuration 2 Register

31 RGB_CFG 23 22 21
ISI_CFG2 0xF0034004 Read-write 0x00000000

30 29 YCC_SWAP 20 IM_HSIZE 28 27 19 26 25 IM_HSIZE 17 24
18
16
15 COL_SPACE 7
14 RGB_SWAP 6
13 GRAYSCALE 5
12 RGB_MODE 4
11 GS_MODE 3
10
9 IM_VSIZE 1
IM_VSIZE
IM_VSIZE: Vertical Size of the Image Sensor [0..2047]:

Vertical size = IM_VSIZE + 1.
GS_MODE:
0: 2 pixels per word. 1: 1 pixel per word.
RGB_MODE: RGB Input Mode:

0: RGB 8:8:8 24 bits. 1: RGB 5:6:5 16 bits.
GRAYSCALE:
0: Grayscale mode is disabled. 1: Input image is assumed to be grayscale coded.
RGB_SWAP:
0: D7 -> R7. 1: D0 -> R7. The RGB_SWAP has no effect when the grayscale mode is enabled.
COL_SPACE: Color Space for the Image Data

0: YCbCr. 1: RGB.
IM_HSIZE: Horizontal Size of the Image Sensor [0..2047]

Horizontal size = IM_HSIZE + 1.

1014
YCC_SWAP: Defines the YCC Image Data

YCC_SWAP 00: Default 01: Mode1 10: Mode2 11: Mode3 Byte 0 Cb(i) Cr(i) Y(i) Y(i) Byte 1 Y(i) Y(i) Cb(i) Cr(i) Byte 2 Cr(i) Cb(i) Y(i+1) Y(i+1) Byte 3 Y(i+1) Y(i+1) Cr(i) Cb(i)
RGB_CFG: Defines RGB Pattern when RGB_MODE is set to 1

RGB_CFG 00: Default 01: Mode1 10: Mode2 11: Mode3 Byte 0 R/G(MSB) B/G(MSB) G(LSB)/R G(LSB)/B Byte 1 G(LSB)/B G(LSB)/R B/G(MSB) R/G(MSB) Byte 2 R/G(MSB) B/G(MSB) G(LSB)/R G(LSB)/B Byte 3 G(LSB)/B G(LSB)/R B/G(MSB) R/G(MSB)
If RGB_MODE is set to RGB 8:8:8, then RGB_CFG = 0 implies RGB color sequence, else it implies BGR color sequence.

1015
36.5.3 ISI Preview Register

31 23
ISI_PSIZE 0xF0034008 Read-write 0x00000000

30 22 29 21 28 20 PREV_HSIZE 27 19 26 18 25 PREV_HSIZE 17 16 24
15 7
14 6
13 5
12 4 PREV_VSIZE
11 3
10 2
9 PREV_VSIZE 1
PREV_VSIZE: Vertical Size for the Preview Path

Vertical Preview size = PREV_VSIZE + 1 (480 max only in RGB mode).
PREV_HSIZE: Horizontal Size for the Preview Path

Horizontal Preview size = PREV_HSIZE + 1 (640 max only in RGB mode).

1016
36.5.4 ISI Preview Decimation Factor Register

31 23 15 7
ISI_PDECF 0xF003400C Read-write 0x00000010

30 22 14 6 29 21 13 5 28 20 12 4 DEC_FACTOR 27 19 11 3 26 18 10 2 25 17 9 1 24 16 8 0
DEC_FACTOR: Decimation Factor

DEC_FACTOR is 8-bit width, range is from 16 to 255. Values from 0 to 16 do not perform any decimation.

1017
36.5.5 ISI Color Space Conversion YCrCb to RGB Set 0 Register

31
ISI_Y2R_SET0 0xF0034010 Read-write 0x6832cc95

30 29 28 C3 27 26 25 24
23
22
21
20 C2
19
18
17
16
15
14
13
12 C1
11
10
4 C0
C0: Color Space Conversion Matrix Coefficient C0

C0 element default step is 1/128, ranges from 0 to 1.9921875.




1018
36.5.6 ISI Color Space Conversion YCrCb to RGB Set 1 Register

31 23 15 7
ISI_Y2R_SET1 0xF0034014 Read-write 0x00007102

30 22 14 Cboff 6 29 21 13 Croff 5 28 20 12 Yoff 4 C4 27 19 11 3 26 18 10 2 25 17 9 1 24 16 8 C4 0

Yoff: Color Space Conversion Luminance Default Offset

0: No offset. 1: Offset = 128.
Croff: Color Space Conversion Red Chrominance Default Offset

Cboff: Color Space Conversion Blue Chrominance Default Offset


1019
36.5.7 ISI Color Space Conversion RGB to YCrCb Set 0 Register

31 23 15 7
ISI_R2Y_SET0 0xF0034018 Read-write 0x01324145

30 22 29 21 28 20 27 19 C2 11 C1 3 C0 26 18 25 17 24 Roff 16
14
13
12
10

C0 element default step is 1/256, from 0 to 0.49609375.


Roff: Color Space Conversion Red Component Offset


1020

31 23 15 7
ISI_R2Y_SET1 0xF003401C Read-write 0x01245e38

30 22 29 21 28 20 27 19 C5 11 C4 3 C3 26 18 25 17 24 Goff 16
14
13
12
10



Goff: Color Space Conversion Green Component Offset


1021

31 23 15 7
ISI_R2Y_SET2 0xF0034020 Read-write 0x01384a4b

30 22 29 21 28 20 27 19 C8 11 C7 3 C6 26 18 25 17 24 Boff 16
14
13
12
10



Boff: Color Space Conversion Blue Component Offset


1022
36.5.10 ISI Control Register

31 23 15 7
ISI_CR 0xF0034024 Write-only 0x00000000

30 22 14 6 29 21 13 5 28 20 12 4 27 19 11 3 26 18 10 2 ISI_SRST 25 17 9 1 ISI_DIS 24 16 8 ISI_CDC 0 ISI_EN
ISI_EN: ISI Module Enable Request

Write one to this field to enable the module. Software must poll ENABLE field in the ISI_STATUS register to verify that the command has successfully completed.
ISI_DIS: ISI Module Disable Request

Write one to this field to disable the module. If both ISI_EN and ISI_DIS are asserted at the same time, the disable request is not taken into account. Software must poll DIS_DONE field in the ISI_STATUS register to verify that the command has successfully completed.
ISI_SRST: ISI Software Reset Request

Write one to this field to request a software reset of the module. Software must poll SRST field in the ISI_STATUS register to verify that the software request command has terminated.
ISI_CDC: ISI Codec Request

Write one to this field to enable the codec datapath and capture a Full resolution Frame. A new request cannot be taken into account while CDC_PND bit is active in the ISI_STATUS register.

1023
36.5.11 ISI Status Register

31 23 15 7
ISI_SR 0xF0034028 Read-only 0x00000000

30 22 14 6 29 21 13 5 28 20 12 4 27 FR_OVR 19 SIP 11 3 26 CRC_ERR 18 10 VSYNC 2 SRST 25 C_OVR 17 CXFR_DONE 9 1 DIS_DONE 24 P_OVR 16 PXFR_DONE 8 CDC_PND 0 ENABLE
ENABLE (this bit is a status bit)

0: Module is disabled. 1: Module is enabled.
DIS_DONE: Module Disable Request has Terminated

1: Disable request has completed. This flag is reset after a read operation.
SRST: Module Software Reset Request has Terminated

1: Software reset request has completed. This flag is reset after a read operation.
CDC_PND: Pending Codec Request (this bit is a status bit)

0: Indicates that no Codec request is pending. 1: Indicates that the request has been taken into account but cannot be serviced within the current frame. The operation is postponed to the next frame.
VSYNC: Vertical Synchronization

1: Indicates that a Vertical synchronization has been detected since the last read of the status register.
PXFR_DONE: Preview DMA Transfer has Terminated.

When set to one, this bit indicates that the DATA transfer on the preview channel has completed. This flag is reset after a read operation.
CXFR_DONE: Codec DMA Transfer has Terminated.

When set to one, this bit indicates that the DATA transfer on the codec channel has completed. This flag is reset after a read operation.
SIP: Synchronization in Progress (this is a status bit)

When the status of the preview or codec DMA channel is modified, a minimum amount of time is required to perform the clock domain synchronization. This bit is set when this operation occurs. No modification of the channel status is allowed when this bit is set, to guarantee data integrity.

1024
P_OVR: Preview Datapath Overflow

0: No overflow 1: An overrun condition has occurred in input FIFO on the preview path. The overrun happens when the FIFO is full and an attempt is made to write a new sample to the FIFO. This flag is reset after a read operation.
C_OVR: Codec Datapath Overflow

0: No overflow 1: An overrun condition has occurred in input FIFO on the codec path. The overrun happens when the FIFO is full and an attempt is made to write a new sample to the FIFO. This flag is reset after a read operation.
CRC_ERR: CRC Synchronization Error

0: No CRC error in the embedded synchronization frame (SAV/EAV) 1: The CRC_SYNC is enabled in the control register and an error has been detected and not corrected. The frame is discarded and the ISI waits for a new one. This flag is reset after a read operation.
FR_OVR: Frame Rate Overrun

0: No frame overrun. 1: Frame overrun, the current frame is being skipped because a vsync signal has been detected while flushing FIFOs. This flag is reset after a read operation.

1025
36.5.12 ISI Interrupt Enable Register

31 23 15 7
ISI_IER 0xF003402C Read-write 0x0

30 22 14 6 29 21 13 5 28 20 12 4 27 FR_OVR 19 11 3 26 CRC_ERR 18 10 VSYNC 2 SRST 25 C_OVR 17 CXFR_DONE 9 1 DIS_DONE 24 P_OVR 16 PXFR_DONE 8 0
DIS_DONE: Disable Done Interrupt Enable SRST: Software Reset Interrupt Enable VSYNC: Vertical Synchronization Interrupt Enable PXFR_DONE: Preview DMA Transfer Done Interrupt Enable CXFR_DONE: Codec DMA Transfer Done Interrupt Enable P_OVR: Preview Datapath Overflow Interrupt Enable C_OVR: Codec Datapath Overflow Interrupt Enable CRC_ERR: Embedded Synchronization CRC Error Interrupt Enable FR_OVR: Frame Rate Overflow Interrupt Enable

1026
36.5.13 ISI Interrupt Disable Register

31 23 15 7
ISI_IDR 0xF0034030 Read-write 0x0

DIS_DONE: Disable Done Interrupt Disable SRST: Software Reset Interrupt Disable VSYNC: Vertical Synchronization Interrupt Disable PXFR_DONE: Preview DMA Transfer Done Interrupt Disable CXFR_DONE: Codec DMA Transfer Done Interrupt Disable P_OVR: Preview Datapath Overflow Interrupt Disable C_OVR: Codec Datapath Overflow Interrupt Disable CRC_ERR: Embedded Synchronization CRC Error Interrupt Disable FR_OVR: Frame Rate Overflow Interrupt Disable

1027
36.5.14 ISI Interrupt Mask Register

31 23 15 7
ISI_IMR 0xF0034034 Read-write 0x0

DIS_DONE: Module Disable Operation Completed

0: The disable completed interrupt is disabled. 1: The disable completed interrupt is enabled.
SRST: Software Reset Completed

0: The software reset completed interrupt is disabled. 1: The software reset completed interrupt is enabled.
VSYNC: Vertical Synchronization

0: The vertical synchronization interrupt is enabled. 1: The vertical synchronization interrupt is disabled.
PXFR_DONE: Preview DMA Transfer Interrupt

0: The Preview DMA transfer completed interrupt is enabled 1: The Preview DMA transfer completed interrupt is disabled
CXFR_DONE: Codec DMA Transfer Interrupt

0: The Codec DMA transfer completed interrupt is enabled 1: The Codec DMA transfer completed interrupt
P_OVR: FIFO Preview Overflow

0: The preview FIFO overflow interrupt is disabled. 1: The preview FIFO overflow interrupt is enabled.
C_OVR: FIFO Codec Overflow

0: The codec FIFO overflow interrupt is disabled. 1: The codec FIFO overflow interrupt is enabled.

1028
CRC_ERR: CRC Synchronization Error

0: The crc error interrupt is disabled. 1: The crc error interrupt is enabled.
FR_OVR: Frame Rate Overrun

0: The frame overrun interrupt is disabled. 1: The frame overrun interrupt is enabled.

1029
36.5.15 DMA Channel Enable Register

31 23 15 7
ISI_DMA_CHER 0xF0034038 Write-only 0x00000000

30 22 14 6 29 21 13 5 28 20 12 4 27 19 11 3 26 18 10 2 25 17 9 1 C_CH_EN 24 16 8 0 P_CH_EN
P_CH_EN: Preview Channel Enable

Write one to this field to enable the preview DMA channel.
C_CH_EN: Codec Channel Enable

Write one to this field to enable the codec DMA channel.

1030
36.5.16 DMA Channel Disable Register

31 23 15 7
ISI_DMA_CHDR 0xF003403C Write-only 0x00000000

30 22 14 6 29 21 13 5 28 20 12 4 27 19 11 3 26 18 10 2 25 17 9 1 C_CH_DIS 24 16 8 0 P_CH_DIS
P_CH_DIS
Write one to this field to disable the channel. Poll P_CH_S in DMA_CHSR to verify that the preview channel status has been successfully modified.
C_CH_DIS
Write one to this field to disabled the channel. Poll C_CH_S in DMA_CHSR to verify that the codec channel status has been successfully modified.

1031
36.5.17 DMA Channel Status Register

31 23 15 7
ISI_DMA_CHSR 0xF0034040 Read-only 0x00000000

30 22 14 6 29 21 13 5 28 20 12 4 27 19 11 3 26 18 10 2 25 17 9 1 C_CH_S 24 16 8 0 P_CH_S
P_CH_S:
0: indicates that the Preview DMA channel is disabled 1: indicates that the Preview DMA channel is enabled.
C_CH_S:
0: indicates that the Codec DMA channel is disabled. 1: indicates that the Codec DMA channel is enabled.

1032
36.5.18 DMA Preview Base Address Register

31
ISI_DMA_P_ADDR 0xF0034044 Read-write 0x00000000

30 29 28 P_ADDR 27 26 25 24
23
22
21
20 P_ADDR
19
18
17
16
15
14
13
12 P_ADDR
11
10
5 P_ADDR
P_ADDR: Preview Image Base Address. (This address is word aligned.)

1033
36.5.19 DMA Preview Control Register

31 23 15 7
ISI_DMA_P_CTRL 0xF0034048 Read-write 0x00000000

30 22 14 6 29 21 13 5 28 20 12 4 27 19 11 3 P_DONE 26 18 10 2 P_IEN 25 17 9 1 P_WB 24 16 8 0 P_FETCH
P_FETCH: Descriptor Fetch Control Field

0: Preview channel fetch operation is disabled. 1: Preview channel fetch operation is enabled.
P_WB: Descriptor Writeback Control Field

0: Preview channel writeback operation is disabled. 1: Preview channel writeback operation is enabled.
P_IEN: Transfer Done Flag Control

0: Preview Transfer done flag generation is enabled. 1: Preview Transfer done flag generation is disabled.
P_DONE: (This field is only updated in the memory.)

0: The transfer related to this descriptor has not been performed. 1: The transfer related to this descriptor has completed. This field is updated in memory at the end of the transfer, when writeback operation is enabled.

1034
36.5.20 DMA Preview Descriptor Address Register

31
ISI_DMA_P_DSCR 0xF003404C Read-write 0x00000000

30 29 28 P_DSCR 27 26 25 24
23
22
21
20 P_DSCR
19
18
17
16
15
14
13
12 P_DSCR
11
10
5 P_DSCR
P_DSCR: Preview Descriptor Base Address (This address is word aligned.)

1035
36.5.21 DMA Codec Base Address Register

31
ISI_DMA_C_ADDR 0xF0034050 Read-write 0x00000000

30 29 28 C_ADDR 27 26 25 24
23
22
21
20 C_ADDR
19
18
17
16
15
14
13
12 C_ADDR
11
10
5 C_ADDR
C_ADDR: Codec Image Base Address (This address is word aligned.)

1036
36.5.22 DMA Codec Control Register

31 23 15 7
ISI_DMA_C_CTRL 0xF0034054 Read-write 0x00000000

30 22 14 6 29 21 13 5 28 20 12 4 27 19 11 3 C_DONE 26 18 10 2 C_IEN 25 17 9 1 C_WB 24 16 8 0 C_FETCH
C_FETCH: Descriptor Fetch Control Field

0: Codec channel fetch operation is disabled. 1: Codec channel fetch operation is enabled.
C_WB: Descriptor Writeback Control Field

0: Codec channel writeback operation is disabled. 1: Codec channel writeback operation is enabled.
C_IEN: Transfer Done flag control

0: Codec Transfer done flag generation is enabled. 1: Codec Transfer done flag generation is disabled.
C_DONE: (This field is only updated in the memory.)

0: The transfer related to this descriptor has not been performed. 1: The transfer related to this descriptor has completed. This field is updated in memory at the end of the transfer, when. writeback operation is enabled.

1037
36.5.23 DMA Codec Descriptor Address Register

31
ISI_DMA_C_DSCR 0xF0034058 Read-write 0x00000000

30 29 28 C_DSCR 27 26 25 24
23
22
21
20 C_DSCR
19
18
17
16
15
14
13
12 C_DSCR
11
10
5 C_DSCR
C_DSCR: Codec Descriptor Base Address (This address is word aligned.)

1038
36.5.24 ISI Write Protection Control

31
ISI_WPCR 0xF00340E4 Read-write

30 29 28 27 WP_KEY (0x49 => I) 20 19 WP_KEY (0x53 => S) 12 11 WP_KEY (0x49 => I) 4 3 26 25 24
23
22
21
18
17
16
15
14
13
10
0 WP_EN

0 = Disables the Write Protection if WP_KEY corresponds. 1 = Enables the Write Protection if WP_KEY corresponds.
WP_KEY: Write Protection KEY Password

Should be written at value 0x495349 (ASCII code for ISI). Writing any other value in this field has no effect.

1039
36.5.25 ISI Write Protection Status

31 23
ISI_WPSR 0xF00340E8 Read-write

30 22 29 21 28 20 WP_VSRC 27 19 26 18 25 17 24 16
15
14
13
12 WP_VSRC
11
10
7 -
6 -
5 -
4 -
2 WP_VS

WP_VS 0 0 0 0 0 0 0 0 Other value 0 0 1 1 0 1 0 1 No Write Protection Violation occurred since the last read of this register (WP_SR). Write Protection detected unauthorized attempt to write a control register had occurred (since the last read). Software reset had been performed while Write Protection was enabled (since the last read). Both Write Protection violation and software reset with Write Protection enabled had occurred since the last read. Reserved
WP_VSRC: Write Protection Violation Source

WP_VSRC 0 0 0 0 0 0 0 0 1 1 0 0 0 0 1 1 1 1 0 0 Other value 0 0 1 1 0 0 1 1 0 0 0 1 0 1 0 1 0 1 0 1 No Write Protection Violation occurred since the last read of this register (WP_SR). Write access in ISI_CFG1 while Write Protection was enabled (since the last read). Write access in ISI_CFG2 while Write Protection was enabled (since the last read). Write access in ISI_PSIZE while Write Protection was enabled (since the last read). Write access in ISI_PDECF while Write Protection was enabled (since the last read). Write access in ISI_Y2R_SET0 while Write Protection was enabled (since the last read). Write access in ISI_Y2R_SET1 while Write Protection was enabled (since the last read). Write access in ISI_R2Y_SET0 while Write Protection was enabled (since the last read). Write access in ISI_R2Y_SET1 while Write Protection was enabled (since the last read). Write access in ISI_R2Y_SET2 while Write Protection was enabled (since the last read). Reserved

1040
37.
37.1
High Speed MultiMedia Card Interface (HSMCI)

Description
The High Speed Multimedia Card Interface (HSMCI) supports the MultiMedia Card (MMC) Specification V4.3, the SD Memory Card Specification V2.0, the SDIO V2.0 specification and CE-ATA V1.1. The HSMCI includes a command register, response registers, data registers, timeout counters and error detection logic that automatically handle the transmission of commands and, when required, the reception of the associated responses and data with a limited processor overhead. The HSMCI supports stream, block and multi block data read and write, and is compatible with the DMA Controller (DMAC), minimizing processor intervention for large buffer transfers. The HSMCI operates at a rate of up to Master Clock divided by 2 and supports the interfacing of 1 slot(s). Each slot may be used to interface with a High Speed MultiMedia Card bus (up to 30 Cards) or with an SD Memory Card. Only one slot can be selected at a time (slots are multiplexed). A bit field in the SD Card Register performs this selection. The SD Memory Card communication is based on a 9-pin interface (clock, command, four data and three power lines) and the High Speed MultiMedia Card on a 7-pin interface (clock, command, one data, three power lines and one reserved for future use). The SD Memory Card interface also supports High Speed MultiMedia Card operations. The main differences between SD and High Speed MultiMedia Cards are the initialization process and the bus topology. HSMCI fully supports CE-ATA Revision 1.1, built on the MMC System Specification v4.0. The module includes dedicated hardware to issue the command completion signal and capture the host command completion signal disable.
37.2
z z z z z z z z z
Compatible with MultiMedia Card Specification Version 4.3 Compatible with SD Memory Card Specification Version 2.0 Compatible with SDIO Specification Version 2.0 Compatible with CE-ATA Specification 1.1 Cards Clock Rate Up to Master Clock Divided by 2 Boot Operation Mode Support High Speed Mode Support Embedded Power Management to Slow Down Clock Rate When Not Used Supports 1 Multiplexed Slot
z
Each Slot for either a High Speed MultiMedia Card Bus (Up to 30 Cards) or an SD Memory Card
z z
Support for Stream, Block and Multi-block Data Read and Write Supports Connection to DMA Controller (DMAC)
z
Minimizes Processor Intervention for Large Buffer Transfers
z z z
Built in FIFO (from 16 to 256 bytes) with Large Memory Aperture Supporting Incremental Access Support for CE-ATA Completion Signal Disable Command Protection Against Unexpected Modification On-the-Fly of the Configuration Registers

1041
37.3
Block Diagram

APB Bridge
DMAC APB MCCK

(1) (1)
MCCDA HSMCI Interface PMC MCK PIO MCDA0
(1)
MCDA1 (1) MCDA2 MCDA3 MCDA4 MCDA5 MCDA6

(1)
(1)
(1)
(1)
(1)
Interrupt Control
MCDA7
(1)
HSMCI Interrupt
Note:
1. When several HSMCI (x HSMCI) are embedded in a product, MCCK refers to HSMCIx_CK, MCCDA to HSMCIx_CDA, MCDAy to HSMCIx_DAy.

1042

APB Bridge
DMAC APB MCCK(1) MCCDA(1) HSMCI Interface PIO MCDA0(1) MCDA1(1) MCDA2(1) MCDA3(1) Interrupt Control
PMC
MCK
HSMCI Interrupt
37.4

Application Layer ex: File System, Audio, Security, etc.
Physical Layer HSMCI Interface
1 2 3 4 5 6 7
1 2 3 4 5 6 78 9
9 1011 1213 8
MMC
SDCard

1043
37.5
Pin Name List
Table 37-1. I/O Lines Description for 8-bit Configuration Pin Name(2) MCCDA/MCCDB MCCK MCDA0 - MCDA7 MCDB0 - MCDB7 MCDC0 - MCDC7 MCDD0 - MCDD7 Pin Description Command/response Clock Data 0..7 of Slot A Data 0..7 of Slot B Data 0..7 of Slot C Data 0..7 of Slot D Type(1) I/O/PP/OD I/O I/O/PP I/O/PP I/O/PP I/O/PP Comments CMD of an MMC or SDCard/SDIO CLK of an MMC or SD Card/SDIO DAT[0..7] of an MMC DAT[0..3] of an SD Card/SDIO DAT[0..7] of an MMC DAT[0..3] of an SD Card/SDIO DAT[0..7] of an MMC DAT[0..3] of an SD Card/SDIO DAT[0..7] of an MMC DAT[0..3] of an SD Card/SDIO
Notes: 1. I: Input, O: Output, PP: Push/Pull, OD: Open Drain. 2. When several HSMCI (x HSMCI) are embedded in a product, MCCK refers to HSMCIx_CK, MCCDA to HSMCIx_CDA, MCCDB to HSMCIx_CDB, MCCDC to HSMCIx_CDC, MCCDD to HSMCIx_CDD, MCDAy to HSMCIx_DAy, MCDBy to HSMCIx_DBy, MCDCy to HSMCIx_DCy, MCDDy to HSMCIx_DDy.
Table 37-2. I/O Lines Description for 4-bit Configuration Pin Name(2) MCCDA/MCCDB MCCK MCDA0 - MCDA3 Pin Description Command/response Clock Data 0..3 of Slot A Type(1) I/O/PP/OD I/O I/O/PP Comments CMD of an MMC or SDCard/SDIO CLK of an MMC or SD Card/SDIO DAT[0..3] of an MMC DAT[0..3] of an SD Card/SDIO
Notes: 1. I: Input, O: Output, PP: Push/Pull, OD: Open Drain. 2. When several HSMCI (x HSMCI) are embedded in a product, MCCK refers to HSMCIx_CK, MCCDA to HSMCIx_CDA, MCDAy to HSMCIx_DAy.

1044
37.6
37.6.1 I/O Lines

The pins used for interfacing the High Speed MultiMedia Cards or SD Cards are multiplexed with PIO lines. The programmer must first program the PIO controllers to assign the peripheral functions to HSMCI pins.
Table 37-3. I/O Lines Instance HSMCI0 HSMCI0 HSMCI0 HSMCI0 HSMCI0 HSMCI0 HSMCI0 HSMCI0 HSMCI0 HSMCI0 HSMCI1 HSMCI1 HSMCI1 HSMCI1 HSMCI1 HSMCI1 HSMCI2 HSMCI2 HSMCI2 HSMCI2 HSMCI2 HSMCI2 Signal MCI0_CDA MCI0_CK MCI0_DA0 MCI0_DA1 MCI0_DA2 MCI0_DA3 MCI0_DA4 MCI0_DA5 MCI0_DA6 MCI0_DA7 MCI1_CDA MCI1_CK MCI1_DA0 MCI1_DA1 MCI1_DA2 MCI1_DA3 MCI2_CDA MCI2_CK MCI2_DA0 MCI2_DA1 MCI2_DA2 MCI2_DA3 I/O Line PD0 PD9 PD1 PD2 PD3 PD4 PD5 PD6 PD7 PD8 PB19 PB24 PB20 PB21 PB22 PB23 PC10 PC15 PC11 PC12 PC13 PC14 Peripheral A A A A A A A A A A A A A A A A A A A A A A

The HSMCI is clocked through the Power Management Controller (PMC), so the programmer must first configure the PMC to enable the HSMCI clock.
37.6.3 Interrupt
The HSMCI interface has an interrupt line connected to the interrupt controller. Handling the HSMCI interrupt requires programming the interrupt controller before configuring the HSMCI.
Table 37-4. Peripheral IDs Instance HSMCI0 HSMCI1 HSMCI2 ID 21 22 23

1045
37.7
Bus Topology
Figure 37-4. High Speed MultiMedia Memory Card Bus Topology
1 2 3 4 5 6 7
9 1011
1213 8
MMC
The High Speed MultiMedia Card communication is based on a 13-pin serial bus interface. It has three communication lines and four supply lines.
Table 37-5. Bus Topology Pin Number 1 2 3 4 5 6 7 8 9 10 11 12 13 Name DAT[3] CMD VSS1 VDD CLK VSS2 DAT[0] DAT[1] DAT[2] DAT[4] DAT[5] DAT[6] DAT[7] Type(1) I/O/PP I/O/PP/OD S S I/O S I/O/PP I/O/PP I/O/PP I/O/PP I/O/PP I/O/PP I/O/PP Description Data Command/response Supply voltage ground Supply voltage Clock Supply voltage ground Data 0 Data 1 Data 2 Data 4 Data 5 Data 6 Data 7 HSMCI Pin Name(2) (Slot z) MCDz3 MCCDz VSS VDD MCCK VSS MCDz0 MCDz1 MCDz2 MCDz4 MCDz5 MCDz6 MCDz7
Notes: 1. 2.
I: Input, O: Output, PP: Push/Pull, OD: Open Drain. When several HSMCI (x HSMCI) are embedded in a product, MCCK refers to HSMCIx_CK, MCCDA to HSMCIx_CDA, MCCDB to HSMCIx_CDB, MCDAy to HSMCIx_DAy, MCDBy to HSMCIx_DBy.

1046
Figure 37-5. MMC Bus Connections (One Slot)
HSMCI
MCDA0
MCCDA
MCCK
1 2 3 4 5 6 7
1 2 3 4 5 6 7
1 2 3 4 5 6 7
9 1011
1213 8
9 1011
1213 8
9 1011
1213 8
MMC1
MMC2
MMC3
Note:
When several HSMCI (x HSMCI) are embedded in a product, MCCK refers to HSMCIx_CK, MCCDA to HSMCIx_CDA MCDAy to HSMCIx_DAy.
Figure 37-6. SD Memory Card Bus Topology

1 2 3 4 5 6 78 9
SD CARD
The SD Memory Card bus includes the signals listed in Table 37-6.
Table 37-6. SD Memory Card Bus Signals Pin Number 1 2 3 4 5 6 7 8 9 Name CD/DAT[3] CMD VSS1 VDD CLK VSS2 DAT[0] DAT[1] DAT[2] Type(1) I/O/PP PP S S I/O S I/O/PP I/O/PP I/O/PP Description Card detect/ Data line Bit 3 Command/response Supply voltage ground Supply voltage Clock Supply voltage ground Data line Bit 0 Data line Bit 1 or Interrupt Data line Bit 2 HSMCI Pin Name(2) (Slot z) MCDz3 MCCDz VSS VDD MCCK VSS MCDz0 MCDz1 MCDz2
Notes: 1. 2.
I: input, O: output, PP: Push Pull, OD: Open Drain. When several HSMCI (x HSMCI) are embedded in a product, MCCK refers to HSMCIx_CK, MCCDA to HSMCIx_CDA, MCCDB to HSMCIx_CDB, MCDAy to HSMCIx_DAy, MCDBy to HSMCIx_DBy.

1047
Figure 37-7. SD Card Bus Connections with One Slot

MCDA0 - MCDA3 MCCK MCCDA 1 2 3 4 5 6 78
SD CARD
Note:
When several HSMCI (x HSMCI) are embedded in a product, MCCK refers to HSMCIx_CK, MCCDA to HSMCIx_CDA MCDAy to HSMCIx_DAy.
When the HSMCI is configured to operate with SD memory cards, the width of the data bus can be selected in the HSMCI_SDCR register. Clearing the SDCBUS bit in this register means that the width is one bit; setting it means that the width is four bits. In the case of High Speed MultiMedia cards, only the data line 0 is used. The other data lines can be used as independent PIOs.

1048
37.8
High Speed MultiMedia Card Operations

After a power-on reset, the cards are initialized by a special message-based High Speed MultiMedia Card bus protocol. Each message is represented by one of the following tokens:
z
Command: A command is a token that starts an operation. A command is sent from the host either to a single card (addressed command) or to all connected cards (broadcast command). A command is transferred serially on the CMD line. Response: A response is a token which is sent from an addressed card or (synchronously) from all connected cards to the host as an answer to a previously received command. A response is transferred serially on the CMD line. Data: Data can be transferred from the card to the host or vice versa. Data is transferred via the data line.
Card addressing is implemented using a session address assigned during the initialization phase by the bus controller to all currently connected cards. Their unique CID number identifies individual cards. The structure of commands, responses and data blocks is described in the High Speed MultiMedia Card System Specification. See also Table 37-7 on page 1050. High Speed MultiMedia Card bus data transfers are composed of these tokens. There are different types of operations. Addressed operations always contain a command and a response token. In addition, some operations have a data token; the others transfer their information directly within the command or response structure. In this case, no data token is present in an operation. The bits on the DAT and the CMD lines are transferred synchronous to the clock HSMCI Clock. Two types of data transfer commands are defined:
z z
Sequential commands: These commands initiate a continuous data stream. They are terminated only when a stop command follows on the CMD line. This mode reduces the command overhead to an absolute minimum. Block-oriented commands: These commands send a data block succeeded by CRC bits.
Both read and write operations allow either single or multiple block transmission. A multiple block transmission is terminated when a stop command follows on the CMD line similarly to the sequential read or when a multiple block transmission has a pre-defined block count (See Data Transfer Operation on page 1052.). The HSMCI provides a set of registers to perform the entire range of High Speed MultiMedia Card operations.
37.8.1 Command - Response Operation

After reset, the HSMCI is disabled and becomes valid after setting the MCIEN bit in the HSMCI_CR Control Register. The PWSEN bit saves power by dividing the HSMCI clock by 2PWSDIV + 1 when the bus is inactive. The two bits, RDPROOF and WRPROOF in the HSMCI Mode Register (HSMCI_MR) allow stopping the HSMCI Clock during read or write access if the internal FIFO is full. This will guarantee data integrity, not bandwidth. All the timings for High Speed MultiMedia Card are defined in the High Speed MultiMedia Card System Specification. The two bus modes (open drain and push/pull) needed to process all the operations are defined in the HSMCI Command Register. The HSMCI_CMDR allows a command to be carried out. For example, to perform an ALL_SEND_CID command:
Host Command CMD S T Content CRC E Z NID Cycles ****** Z S T CID Content Z Z Z

1049
The command ALL_SEND_CID and the fields and values for the HSMCI_CMDR Control Register are described in Table 37-7 and Table 37-8.
Table 37-7. ALL_SEND_CID Command Description CMD Index Type Argument Resp Abbreviation Command Description Asks all cards to send their CID numbers on the CMD line
CMD2
bcr(1)
[31:0] stuff bits
R2
ALL_SEND_CID
Note:
1.
bcr means broadcast command with response.
Table 37-8. Fields and Values for HSMCI_CMDR Command Register Field CMDNB (command number) RSPTYP (response type) SPCMD (special command) OPCMD (open drain command) MAXLAT (max latency for command to response) TRCMD (transfer command) TRDIR (transfer direction) TRTYP (transfer type) IOSPCMD (SDIO special command) Value 2 (CMD2) 2 (R2: 136 bits response) 0 (not a special command) 1 0 (NID cycles ==> 5 cycles) 0 (No transfer) X (available only in transfer command) X (available only in transfer command) 0 (not a special command)
The HSMCI_ARGR contains the argument field of the command. To send a command, the user must perform the following steps:
z z
Fill the argument register (HSMCI_ARGR) with the command argument. Set the command register (HSMCI_CMDR) (see Table 37-8).
The command is sent immediately after writing the command register. While the card maintains a busy indication (at the end of a STOP_TRANSMISSION command CMD12, for example), a new command shall not be sent. The NOTBUSY flag in the status register (HSMCI_SR) is asserted when the card releases the busy indication. If the command requires a response, it can be read in the HSMCI Response Register (HSMCI_RSPR). The response size can be from 48 bits up to 136 bits depending on the command. The HSMCI embeds an error detection to prevent any corrupted data during the transfer. The following flowchart shows how to send a command to the card and read the response if needed. In this example, the status register bits are polled but setting the appropriate bits in the Interrupt Enable Register (HSMCI_IER) allows using an interrupt method.

1050
Figure 37-8. Command/Response Functional Flow Diagram

Set the command argument HSMCI_ARGR = Argument(1)
Set the command HSMCI_CMDR = Command
Read HSMCI_SR
Wait for command ready status flag
0 CMDRDY
Check error bits in the status register (1)
Yes Status error flags? (1) RETURN ERROR
Read response if required
Does the command involve a busy indication?
No
RETURN OK
Read HSMCI_SR
0 NOTBUSY
1 RETURN OK
Note:
1. If the command is SEND_OP_COND, the CRC error flag is always present (refer to R3 response in the High Speed MultiMedia Card specification).

1051
37.8.2 Data Transfer Operation

The High Speed MultiMedia Card allows several read/write operations (single block, multiple blocks, stream, etc.). These kinds of transfer can be selected setting the Transfer Type (TRTYP) field in the HSMCI Command Register (HSMCI_CMDR). These operations can be done using the features of the DMA Controller. In all cases, the block length (BLKLEN field) must be defined either in the Mode Register HSMCI_MR, or in the Block Register HSMCI_BLKR. This field determines the size of the data block. Consequent to MMC Specification 3.1, two types of multiple block read (or write) transactions are defined (the host can use either one at any time):
z
Open-ended/Infinite Multiple block read (or write): The number of blocks for the read (or write) multiple block operation is not defined. The card will continuously transfer (or program) data blocks until a stop transmission command is received.
Multiple block read (or write) with pre-defined block count (since version 3.1 and higher): The card will transfer (or program) the requested number of data blocks and terminate the transaction. The stop command is not required at the end of this type of multiple block read (or write), unless terminated with an error. In order to start a multiple block read (or write) with pre-defined block count, the host must correctly program the HSMCI Block Register (HSMCI_BLKR). Otherwise the card will start an open-ended multiple block read. The BCNT field of the Block Register defines the number of blocks to transfer (from 1 to 65535 blocks). Programming the value 0 in the BCNT field corresponds to an infinite block transfer.

1052
37.8.3 Read Operation

The following flowchart (Figure 37-9) shows how to read a single block with or without use of DMAC facilities. In this example, a polling method is used to wait for the end of read. Similarly, the user can configure the Interrupt Enable Register (HSMCI_IER) to trigger an interrupt at the end of read.

1053
Figure 37-9. Read Functional Flow Diagram

Send SELECT/DESELECT_CARD command(1) to select the card
Send SET_BLOCKLEN command(1)
No Read with DMAC
Yes
Reset the DMAEN bit HSMCI_DMA &= ~DMAEN Set the block length (in bytes) HSMCI_MR l= (BlockLength<<16) (2) Set the block count (if neccessary) HSMCI_BLKR l= (BlockCount<<0)
Set the DMAEN bit HSMCI_DMA |= DMAEN Set the block length (in bytes) HSMCI_BLKR |= (BlockLength << 16)(2)
Send READ_SINGLE_BLOCK command(1)
Configure the DMA channel X DMAC_SADDRx = Data Address DMAC_BTSIZE = BlockLength/4 DMACHEN[X] = TRUE
Number of words to read = BlockLength/4 Send READ_SINGLE_BLOCK command(1)
Yes Number of words to read = 0 ? Read status register HSMCI_SR No Read status register HSMCI_SR Poll the bit XFRDONE = 0? Yes
Poll the bit RXRDY = 0?
Yes No
No Read data = HSMCI_RDR Number of words to read = Number of words to read -1
RETURN
RETURN
Notes: 1. It is assumed that this command has been correctly sent (see Figure 37-8). 2. This field is also accessible in the HSMCI Block Register (HSMCI_BLKR).

1054
37.8.4 Write Operation

In write operation, the HSMCI Mode Register (HSMCI_MR) is used to define the padding value when writing non-multiple block size. If the bit PADV is 0, then 0x00 value is used when padding data, otherwise 0xFF is used. If set, the bit DMAEN in the HSMCI_DMA register enables DMA transfer. The following flowchart (Figure 37-10) shows how to write a single block with or without use of DMA facilities. Polling or interrupt method can be used to wait for the end of write according to the contents of the Interrupt Mask Register (HSMCI_IMR).

1055
Figure 37-10. Write Functional Flow Diagram

No Write using DMAC
Yes
Reset theDMAEN bit HSMCI_DMA &= ~DMAEN Set the block length (in bytes) HSMCI_MR |= (BlockLength) <<16)(2) Set the block count (if necessary) HSMCI_BLKR |= (BlockCount << 0)
Set the DMAEN bit HSMCI_DMA |= DMAEN Set the block length (in bytes) HSMCI_BLKR |= (BlockLength << 16)(2)
Send WRITE_SINGLE_BLOCK command(1)
Send WRITE_SINGLE_BLOCK command(1)
Number of words to write = BlockLength/4
Configure the DMA channel X DMAC_DADDRx = Data Address to write DMAC_BTSIZE = BlockLength/4
DMAC_CHEN[X] = TRUE Yes Number of words to write = 0 ? Read status register HSMCI_SR No Read status register HSMCI_SR Poll the bit XFRDONE = 0? Poll the bit TXRDY = 0? Yes No Yes
No HSMCI_TDR = Data to write Number of words to write = Number of words to write -1 RETURN
RETURN
Note:
1. It is assumed that this command has been correctly sent (see Figure 37-8). 2. This field is also accessible in the HSMCI Block Register (HSMCI_BLKR). The following flowchart (Figure 37-11) shows how to manage read multiple block and write multiple block transfers with the DMA Controller. Polling or interrupt method can be used to wait for the end of write according to the contents of the Interrupt Mask Register (HSMCI_IMR).

1056
Figure 37-11. Read Multiple Block and Write Multiple Block

Set the block length HSMCI_MR |= (BlockLength << 16) Set the DMAEN bit HSMCI_DMA |= DMAEN
Send WRITE_MULTIPLE_BLOCK or READ_MULTIPLE_BLOCK command(1)
Configure the HDMA channel X DMAC_SADDRx and DMAC_DADDRx DMAC_BTSIZE = BlockLength/4
DMAC_CHEN[X] = TRUE
Read status register DMAC_EBCISR and Poll Bit CBTC[X]
New Buffer ?(2)
Yes
No Read status register HSMCI_SR and Poll Bit FIFOEMPTY
Send STOP_TRANSMISSION command(1)
Poll the bit XFRDONE = 1
No
Yes RETURN
Notes: 1. 2.
It is assumed that this command has been correctly sent (see Figure 37-8). Handle errors reported in HSMCI_SR.

1057
37.8.5 WRITE_SINGLE_BLOCK Operation using DMA Controller

1. 2. 3. 4. Wait until the current command execution has successfully terminated. 3. Check that CMDRDY and NOTBUSY fields are asserted in HSMCI_SR Program the block length in the card. This value defines the value block_length. Program the block length in the HSMCI Configuration Register with block_length value. Program the HSMCI_DMA register with the following fields:
z z z
OFFSET field with dma_offset. CHKSIZE is user defined and set according to DMAC_DCSIZE. DMAEN is set to true to enable DMA hardware handshaking in the HSMCI. This bit was previously set to false.
5. 6.
Issue a WRITE_SINGLE_BLOCK command writing HSMCI_ARG then HSMCI_CMDR. Program the DMA Controller. 1. 2. 3. 4. Read the channel register to choose an available (disabled) channel. Clear any pending interrupts on the channel from the previous DMAC transfer by reading the DMAC_EBCISR register. Program the channel registers. The DMAC_SADDRx register for Channel x must be set to the location of the source data. When the first data location is not word aligned, the two LSB bits define the temporary value called dma_offset. The two LSB bits of DMAC_SADDRx must be set to 0. The DMAC_DADDRx register for Channel x must be set with the starting address of the HSMCI_FIFO address. Program the DMAC_CTRLAx register of Channel x with the following fields values:
5. 6.
DST_WIDTH is set to WORD. SRC_WIDTH is set to WORD. DCSIZE must be set according to the value of HSMCI_DMA, CHKSIZE field. BTSIZE is programmed with CEILING((block_length + dma_offset) / 4), where the ceiling function is the function that returns the smallest integer not less than x.
7. Program the DMAC_CTRLBx register for Channel x with the following fields values:
DST_INCR is set to INCR, the block_length value must not be larger than the HSMCI_FIFO aperture. SRC_INCR is set to INCR. FC field is programmed with memory to peripheral flow control mode. both DST_DSCR and SRC_DSCR are set to 1 (descriptor fetch is disabled). DIF and SIF are set with their respective layer ID. If SIF is different from DIF, the DMA controller is able to prefetch data and write HSMCI simultaneously.
8. Program the DMAC_CFGx register for Channel x with the following fields values:
FIFOCFG defines the watermark of the DMAC channel FIFO. DST_H2SEL is set to true to enable hardware handshaking on the destination. DST_PER is programmed with the hardware handshaking ID of the targeted HSMCI Host Controller.
9. 7. Enable Channel x, writing one to DMAC_CHER[x]. The DMAC is ready and waiting for request. Wait for XFRDONE in the HSMCI_SR register.

1058
37.8.6 READ_SINGLE_BLOCK Operation using DMA Controller

37.8.6.1 Block Length is Multiple of 4 1. 2. 3. 4. 5. Wait until the current command execution has successfully completed. 1. Check that CMDRDY and NOTBUSY are asserted in HSMCI_SR. Program the block length in the card. This value defines the value block_length. Program the block length in the HSMCI Configuration Register with block_length value. Set RDPROOF bit in HSMCI_MR to avoid overflow. Program HSMCI_DMA register with the following fields:
z z z z
ROPT field is set to 0. OFFSET field is set to 0. CHKSIZE is user defined. DMAEN is set to true to enable DMAC hardware handshaking in the HSMCI. This bit was previously set to false.
6. 7.
Issue a READ_SINGLE_BLOCK command. Program the DMA controller. 1. 2. 3. 4. 5. 6. Read the channel register to choose an available (disabled) channel. Clear any pending interrupts on the channel from the previous DMA transfer by reading the DMAC_EBCISR register. Program the channel registers. The DMAC_SADDRx register for Channel x must be set with the starting address of the HSMCI_FIFO address. The DMAC_DADDRx register for Channel x must be word aligned. Program the DMAC_CTRLAx register of Channel x with the following fields values:
DST_WIDTH is set to WORD. SRC_WIDTH is set to WORD. SCSIZE must be set according to the value of HSMCI_DMA, CHKSIZE field. BTSIZE is programmed with block_length/4.
DST_INCR is set to INCR. SRC_INCR is set to INCR. FC field is programmed with peripheral to memory flow control mode. both DST_DSCR and SRC_DSCR are set to 1 (descriptor fetch is disabled). DIF and SIF are set with their respective layer ID. If SIF is different from DIF, the DMA controller is able to prefetch data and write HSMCI simultaneously.
FIFOCFG defines the watermark of the DMA channel FIFO. SRC_H2SEL is set to true to enable hardware handshaking on the destination. SRC_PER is programmed with the hardware handshaking ID of the targeted HSMCI Host Controller. Enable Channel x, writing one to DMAC_CHER[x]. The DMAC is ready and waiting for request.
8. Wait for XFRDONE in the HSMCI_SR register.

1059
37.8.6.2 Block Length is Not Multiple of 4 and Padding Not Used (ROPT field in HSMCI_DMA register set to 0) In the previous DMA transfer flow (block length multiple of 4), the DMA controller is configured to use only WORD AHB access. When the block length is no longer a multiple of 4 this is no longer true. The DMA controller is programmed to copy exactly the block length number of bytes using 2 transfer descriptors. 1. 2. Use the previous step until READ_SINGLE_BLOCK then Program the DMA controller to use a two descriptors linked list. 1. 2. 3. 4. 5. 6. Read the channel register to choose an available (disabled) channel. Clear any pending interrupts on the channel from the previous DMA transfer by reading the DMAC_EBCISR register. Program the channel registers in the Memory for the first descriptor. This descriptor will be word oriented. This descriptor is referred to as LLI_W, standing for LLI word oriented transfer. The LLI_W.DMAC_SADDRx field in memory must be set with the starting address of the HSMCI_FIFO address. The LLI_W.DMAC_DADDRx field in the memory must be word aligned. Program LLI_W.DMAC_CTRLAx with the following fields values:
DST_WIDTH is set to WORD. SRC_WIDTH is set to WORD. SCSIZE must be set according to the value of HSMCI_DMA, CHKSIZE field. BTSIZE is programmed with block_length/4. If BTSIZE is zero, this descriptor is skipped later.
7. Program LLI_W.DMAC_CTRLBx with the following fields values:
DST_INCR is set to INCR SRC_INCR is set to INCR FC field is programmed with peripheral to memory flow control mode. SRC_DSCR is set to zero. (descriptor fetch is enabled for the SRC) DST_DSCR is set to one. (descriptor fetch is disabled for the DST) DIF and SIF are set with their respective layer ID. If SIF is different from DIF, DMA controller is able to prefetch data and write HSMCI simultaneously.
8. Program the LLI_W.DMAC_CFGx register for Channel x with the following fields values:
FIFOCFG defines the watermark of the DMA channel FIFO. DST_REP is set to zero meaning that address are contiguous. SRC_H2SEL is set to true to enable hardware handshaking on the destination. SRC_PER is programmed with the hardware handshaking ID of the targeted HSMCI Host Controller.
9. Program LLI_W.DMAC_DSCRx with the address of LLI_B descriptor. And set DSCRx_IF to the AHB Layer ID. This operation actually links the Word oriented descriptor on the second byte oriented descriptor. When block_length[1:0] is equal to 0 (multiple of 4) LLI_W.DMAC_DSCRx points to 0, only LLI_W is relevant.
10. Program the channel registers in the Memory for the second descriptor. This descriptor will be byte oriented. This descriptor is referred to as LLI_B, standing for LLI Byte oriented. 11. The LLI_B.DMAC_SADDRx field in memory must be set with the starting address of the HSMCI_FIFO address. 12. The LLI_B.DMAC_DADDRx is not relevant if previous word aligned descriptor was enabled. If 1, 2 or 3 bytes are transferred that address is user defined and not word aligned. 13. Program LLI_B.DMAC_CTRLAx with the following fields values:
DST_WIDTH is set to BYTE. SRC_WIDTH is set to BYTE.

1060
SCSIZE must be set according to the value of HSMCI_DMA, CHKSIZE field. BTSIZE is programmed with block_length[1:0]. (last 1, 2, or 3 bytes of the buffer).
14. Program LLI_B.DMAC_CTRLBx with the following fields values:
DST_INCR is set to INCR SRC_INCR is set to INCR FC field is programmed with peripheral to memory flow control mode. Both SRC_DSCR and DST_DSCR are set to 1 (descriptor fetch is disabled) or Next descriptor location points to 0. DIF and SIF are set with their respective layer ID. If SIF is different from DIF, DMA Controller is able to prefetch data and write HSMCI simultaneously.
15. Program LLI_B.DMAC_CFGx memory location for Channel x with the following fields values:
FIFOCFG defines the watermark of the DMA channel FIFO. SRC_H2SEL is set to true to enable hardware handshaking on the destination. SRC_PER is programmed with the hardware handshaking ID of the targeted HSMCI Host Controller.
16. Program LLI_B.DMAC_DSCR with 0. 17. Program the DMAC_CTRLBx register for Channel x with 0. its content is updated with the LLI fetch operation. 18. Program DMAC_DSCRx with the address of LLI_W if block_length greater than 4 else with address of LLI_B. 19. Enable Channel x writing one to DMAC_CHER[x]. The DMAC is ready and waiting for request. 3. Wait for XFRDONE in the HSMCI_SR register.
37.8.6.3 Block Length is Not Multiple of 4, with Padding Value (ROPT field in HSMCI_DMA register set to 1) When the ROPT field is set to one, The DMA Controller performs only WORD access on the bus to transfer a nonmultiple of 4 block length. Unlike previous flow, in which the transfer size is rounded to the nearest multiple of 4. 1. 2. Program the HSMCI Interface, see previous flow.
z
ROPT field is set to 1. Read the channel register to choose an available (disabled) channel. Clear any pending interrupts on the channel from the previous DMA transfer by reading the DMAC_EBCISR register. Program the channel registers. The DMAC_SADDRx register for Channel x must be set with the starting address of the HSMCI_FIFO address. The DMAC_DADDRx register for Channel x must be word aligned. Program the DMAC_CTRLAx register of Channel x with the following fields values:
Program the DMA Controller 1. 2. 3. 4. 5. 6.
DST_WIDTH is set to WORD SRC_WIDTH is set to WORD SCSIZE must be set according to the value of HSMCI_DMA.CHKSIZE Field. BTSIZE is programmed with CEILING(block_length/4).
DST_INCR is set to INCR SRC_INCR is set to INCR FC field is programmed with peripheral to memory flow control mode.

1061
Both DST_DSCR and SRC_DSCR are set to 1. (descriptor fetch is disabled) DIF and SIF are set with their respective layer ID. If SIF is different from DIF, the DMA Controller is able to prefetch data and write HSMCI simultaneously.
FIFOCFG defines the watermark of the DMA channel FIFO. SRC_H2SEL is set to true to enable hardware handshaking on the destination. SRC_PER is programmed with the hardware handshaking ID of the targeted HSMCI Host Controller. Enable Channel x writing one to DMAC_CHER[x]. The DMAC is ready and waiting for request.
3. Wait for XFRDONE in the HSMCI_SR register.
37.8.7 WRITE_MULTIPLE_BLOCK
37.8.7.1 One Block per Descriptor 1. 2. 3. 4. Wait until the current command execution has successfully terminated. 1. Check that CMDRDY and NOTBUSY are asserted in HSMCI_SR. Program the block length in the card. This value defines the value block_length. Program the block length in the HSMCI Configuration Register with block_length value. Program the HSMCI_DMA register with the following fields:
z z z
OFFSET field with dma_offset. CHKSIZE is user defined. DMAEN is set to true to enable DMAC hardware handshaking in the HSMCI. This bit was previously set to false.
5. 6.
Issue a WRITE_MULTIPLE_BLOCK command. Program the DMA Controller to use a list of descriptors. Each descriptor transfers one block of data. Block n of data is transferred with descriptor LLI(n). 1. 2. 3. 4. Read the channel register to choose an available (disabled) channel. Clear any pending interrupts on the channel from the previous DMAC transfer by reading the DMAC_EBCISR register. Program a List of descriptors. The LLI(n).DMAC_SADDRx memory location for Channel x must be set to the location of the source data. When the first data location is not word aligned, the two LSB bits define the temporary value called dma_offset. The two LSB bits of LLI(n).DMAC_SADDRx must be set to 0. The LLI(n).DMAC_DADDRx register for Channel x must be set with the starting address of the HSMCI_FIFO address. Program the LLI(n).DMAC_CTRLAx register of Channel x with the following fields values:
5. 6.
DST_WIDTH is set to WORD. SRC_WIDTH is set to WORD. DCSIZE must be set according to the value of HSMCI_DMA, CHKSIZE field. BTSIZE is programmed with CEILING((block_length + dma_offset)/4).
7. Program the LLI(n).DMAC_CTRLBx register for Channel x with the following fields values:
DST_INCR is set to INCR. SRC_INCR is set to INCR. DST_DSCR is set to 0 (fetch operation is enabled for the destination). SRC_DSCR is set to 1 (source address is contiguous). FC field is programmed with memory to peripheral flow control mode.

1062
Both DST_DSCR and SRC_DSCR are set to 1 (descriptor fetch is disabled). DIF and SIF are set with their respective layer ID. If SIF is different from DIF, DMA Controller is able to prefetch data and write HSMCI simultaneously.
8. Program the LLI(n).DMAC_CFGx register for Channel x with the following fields values:
FIFOCFG defines the watermark of the DMA channel FIFO. DST_H2SEL is set to true to enable hardware handshaking on the destination. SRC_REP is set to 0. (contiguous memory access at block boundary) DST_PER is programmed with the hardware handshaking ID of the targeted HSMCI Host Controller.
9. If LLI(n) is the last descriptor, then LLI(n).DSCR points to 0 else LLI(n) points to the start address of LLI(n+1).
10. Program DMAC_CTRLBx for the Channel Register x with 0. Its content is updated with the LLI fetch operation. 11. Program DMAC_DSCRx for the Channel Register x with the address of the first descriptor LLI(0). 12. Enable Channel x writing one to DMAC_CHER[x]. The DMA is ready and waiting for request. 7. 8. 9. Poll CBTC[x] bit in the DMAC_EBCISR Register. If a new list of buffers shall be transferred, repeat step 6. Check and handle HSMCI errors. Poll FIFOEMPTY field in the HSMCI_SR.
10. Send The STOP_TRANSMISSION command writing HSMCI_ARG then HSMCI_CMDR. 11. Wait for XFRDONE in the HSMCI_SR register.
37.8.8 READ_MULTIPLE_BLOCK
37.8.8.1 Block Length is a Multiple of 4 1. 2. 3. 4. 5. Wait until the current command execution has successfully terminated. 1. Check that CMDRDY and NOTBUSY are asserted in HSMCI_SR. Program the block length in the card. This value defines the value block_length. Program the block length in the HSMCI Configuration Register with block_length value. Set RDPROOF bit in HSMCI_MR to avoid overflow. Program the HSMCI_DMA register with the following fields:
z z z z
ROPT field is set to 0. OFFSET field is set to 0. CHKSIZE is user defined. DMAEN is set to true to enable DMAC hardware handshaking in the HSMCI. This bit was previously set to false.
6. 7.
Issue a READ_MULTIPLE_BLOCK command. Program the DMA Controller to use a list of descriptors: 1. 2. 3. 4. 5. 6. Read the channel register to choose an available (disabled) channel. Clear any pending interrupts on the channel from the previous DMA transfer by reading the DMAC_EBCISR register. Program the channel registers in the Memory with the first descriptor. This descriptor will be word oriented. This descriptor is referred to as LLI_W(n), standing for LLI word oriented transfer for block n. The LLI_W(n).DMAC_SADDRx field in memory must be set with the starting address of the HSMCI_FIFO address. The LLI_W(n).DMAC_DADDRx field in the memory must be word aligned. Program LLI_W(n).DMAC_CTRLAx with the following fields values:
DST_WIDTH is set to WORD

1063
SRC_WIDTH is set to WORD SCSIZE must be set according to the value of HSMCI_DMA, CHKSIZE field. BTSIZE is programmed with block_length/4.
7. Program LLI_W(n).DMAC_CTRLBx with the following fields values:
DST_INCR is set to INCR. SRC_INCR is set to INCR. FC field is programmed with peripheral to memory flow control mode. SRC_DSCR is set to 0 (descriptor fetch is enabled for the SRC). DST_DSCR is set to TRUE (descriptor fetch is disabled for the DST). DIF and SIF are set with their respective layer ID. If SIF is different from DIF, the DMA Controller is able to prefetch data and write HSMCI simultaneously.
8. Program the LLI_W(n).DMAC_CFGx register for Channel x with the following fields values:
FIFOCFG defines the watermark of the DMA channel FIFO. DST_REP is set to zero. Addresses are contiguous. SRC_H2SEL is set to true to enable hardware handshaking on the destination. SRC_PER is programmed with the hardware handshaking ID of the targeted HSMCI Host Controller.
9. Program LLI_W(n).DMAC_DSCRx with the address of LLI_W(n+1) descriptor. And set the DSCRx_IF to the AHB Layer ID. This operation actually links descriptors together. If LLI_W(n) is the last descriptor then LLI_W(n).DMAC_DSCRx points to 0.
10. Program the DMAC_CTRLBx register for Channel x with 0. its content is updated with the LLI Fetch operation. 11. Program DMAC_DSCRx register for Channel x with the address of LLI_W(0). 12. Enable Channel x writing one to DMAC_CHER[x]. The DMA is ready and waiting for request. 8. 9. Poll CBTC[x] bit in the DMAC_EBCISR Register. If a new list of buffer shall be transferred repeat step 6. Check and handle HSMCI errors.
10. Poll FIFOEMPTY field in the HSMCI_SR. 11. Send The STOP_TRANSMISSION command writing the HSMCI_ARG then the HSMCI_CMDR. 12. Wait for XFRDONE in the HSMCI_SR register. 37.8.8.2 Block Length is Not Multiple of 4. (ROPT field in HSMCI_DMA register set to 0) Two DMA Transfer descriptors are used to perform the HSMCI block transfer. 1. 2. 3. Use the previous step to configure the HSMCI to perform a READ_MULTIPLE_BLOCK command. Issue a READ_MULTIPLE_BLOCK command. Program the DMA Controller to use a list of descriptors. 1. 2. 3. 4. 5. 6. 7. Read the channel register to choose an available (disabled) channel. Clear any pending interrupts on the channel from the previous DMAC transfer by reading the DMAC_EBCISR register. For every block of data repeat the following procedure: Program the channel registers in the Memory for the first descriptor. This descriptor will be word oriented. This descriptor is referred to as LLI_W(n) standing for LLI word oriented transfer for block n. The LLI_W(n).DMAC_SADDRx field in memory must be set with the starting address of the HSMCI_FIFO address. The LLI_W(n).DMAC_DADDRx field in the memory must be word aligned. Program LLI_W(n).DMAC_CTRLAx with the following fields values:
DST_WIDTH is set to WORD. SAMA5D3 Series [DATASHEET]

1064
SRC_WIDTH is set to WORD. SCSIZE must be set according to the value of HSMCI_DMA, CHKSIZE field. BTSIZE is programmed with block_length/4. If BTSIZE is zero, this descriptor is skipped later.
DST_INCR is set to INCR. SRC_INCR is set to INCR. FC field is programmed with peripheral to memory flow control mode. SRC_DSCR is set to 0 (descriptor fetch is enabled for the SRC). DST_DSCR is set to TRUE (descriptor fetch is disabled for the DST). DIF and SIF are set with their respective layer ID. If SIF is different from DIF, the DMA Controller is able to prefetch data and write HSMCI simultaneously.
FIFOCFG defines the watermark of the DMA channel FIFO. DST_REP is set to zero. Address are contiguous. SRC_H2SEL is set to true to enable hardware handshaking on the destination. SRC_PER is programmed with the hardware handshaking ID of the targeted HSMCI Host Controller.
10. Program LLI_W(n).DMAC_DSCRx with the address of LLI_B(n) descriptor. And set the DSCRx_IF to the AHB Layer ID. This operation actually links the Word oriented descriptor on the second byte oriented descriptor. When block_length[1:0] is equal to 0 (multiple of 4) LLI_W(n).DMAC_DSCRx points to 0, only LLI_W(n) is relevant. 11. Program the channel registers in the Memory for the second descriptor. This descriptor will be byte oriented. This descriptor is referred to as LLI_B(n), standing for LLI Byte oriented. 12. The LLI_B(n).DMAC_SADDRx field in memory must be set with the starting address of the HSMCI_FIFO address. 13. The LLI_B(n).DMAC_DADDRx is not relevant if previous word aligned descriptor was enabled. If 1, 2 or 3 bytes are transferred, that address is user defined and not word aligned. 14. Program LLI_B(n).DMAC_CTRLAx with the following fields values:
DST_WIDTH is set to BYTE. SRC_WIDTH is set to BYTE. SCSIZE must be set according to the value of HSMCI_DMA, CHKSIZE field. BTSIZE is programmed with block_length[1:0]. (last 1, 2, or 3 bytes of the buffer).
15. Program LLI_B(n).DMAC_CTRLBx with the following fields values:
DST_INCR is set to INCR. SRC_INCR is set to INCR. FC field is programmed with peripheral to memory flow control mode. Both SRC_DSCR and DST_DSCR are set to 1 (descriptor fetch is disabled) or Next descriptor location points to 0. DIF and SIF are set with their respective layer ID. If SIF is different from DIF, the DMA Controller is able to prefetch data and write HSMCI simultaneously.
16. Program LLI_B(n).DMAC_CFGx memory location for Channel x with the following fields values:
FIFOCFG defines the watermark of the DMAC channel FIFO. SRC_H2SEL is set to true to enable hardware handshaking on the destination.

1065
SRC_PER is programmed with the hardware handshaking ID of the targeted HSMCI Host Controller
17. Program LLI_B(n).DMAC_DSCR with address of descriptor LLI_W(n+1). If LLI_B(n) is the last descriptor, then program LLI_B(n).DMAC_DSCR with 0. 18. Program the DMAC_CTRLBx register for Channel x with 0, its content is updated with the LLI Fetch operation. 19. Program DMAC_DSCRx with the address of LLI_W(0) if block_length is greater than 4 else with address of LLI_B(0). 20. Enable Channel x writing one to DMAC_CHER[x]. The DMAC is ready and waiting for request. 4. 5. 6. 7. 8. 9. Enable DMADONE interrupt in the HSMCI_IER register. Poll CBTC[x] bit in the DMAC_EBCISR Register. If a new list of buffers shall be transferred, repeat step 7. Check and handle HSMCI errors. Poll FIFOEMPTY field in the HSMCI_SR. Send The STOP_TRANSMISSION command writing HSMCI_ARG then HSMCI_CMDR. Wait for XFRDONE in the HSMCI_SR register.
37.8.8.3 Block Length is Not a Multiple of 4. (ROPT field in HSMCI_DMA register set to 1) One DMA Transfer descriptor is used to perform the HSMCI block transfer, the DMA writes a rounded up value to the nearest multiple of 4. 1. 2. 3. 4. Use the previous step to configure the HSMCI to perform a READ_MULTIPLE_BLOCK. Set the ROPT field to 1 in the HSMCI_DMA register. Issue a READ_MULTIPLE_BLOCK command. Program the DMA controller to use a list of descriptors: 1. 2. 3. 4. 5. 6. Read the channel register to choose an available (disabled) channel. Clear any pending interrupts on the channel from the previous DMAC transfer by reading the DMAC_EBCISR register. Program the channel registers in the Memory with the first descriptor. This descriptor will be word oriented. This descriptor is referred to as LLI_W(n), standing for LLI word oriented transfer for block n. The LLI_W(n).DMAC_SADDRx field in memory must be set with the starting address of the HSMCI_FIFO address. The LLI_W(n).DMAC_DADDRx field in the memory must be word aligned. Program LLI_W(n).DMAC_CTRLAx with the following fields values:
DST_WIDTH is set to WORD. SRC_WIDTH is set to WORD. SCSIZE must be set according to the value of HSMCI_DMA, CHKSIZE field. BTSIZE is programmed with Ceiling(block_length/4).
DST_INCR is set to INCR SRC_INCR is set to INCR FC field is programmed with peripheral to memory flow control mode. SRC_DSCR is set to 0. (descriptor fetch is enabled for the SRC) DST_DSCR is set to TRUE. (descriptor fetch is disabled for the DST) DIF and SIF are set with their respective layer ID. If SIF is different from DIF, the DMA Controller is able to prefetch data and write HSMCI simultaneously.
FIFOCFG defines the watermark of the DMA channel FIFO.

1066
DST_REP is set to zero. Address are contiguous. SRC_H2SEL is set to true to enable hardware handshaking on the destination. SRC_PER is programmed with the hardware handshaking ID of the targeted HSMCI Host Controller.
9. Program LLI_W(n).DMAC_DSCRx with the address of LLI_W(n+1) descriptor. And set the DSCRx_IF to the AHB Layer ID. This operation actually links descriptors together. If LLI_W(n) is the last descriptor then LLI_W(n).DMAC_DSCRx points to 0.
10. Program the DMAC_CTRLBx register for Channel x with 0. its content is updated with the LLI Fetch operation. 11. Program the DMAC_DSCRx register for Channel x with the address of LLI_W(0). 12. Enable Channel x writing one to DMAC_CHER[x]. The DMAC is ready and waiting for request. 5. 6. 7. 8. 9. Poll CBTC[x] bit in the DMAC_EBCISR register. If a new list of buffers shall be transferred repeat step 7. Check and handle HSMCI errors. Poll FIFOEMPTY field in the HSMCI_SR. Send The STOP_TRANSMISSION command writing the HSMCI_ARG then the HSMCI_CMDR. Wait for XFRDONE in the HSMCI_SR register.

1067
37.9
SD/SDIO Card Operation

The High Speed MultiMedia Card Interface allows processing of SD Memory (Secure Digital Memory Card) and SDIO (SD Input Output) Card commands. SD/SDIO cards are based on the MultiMedia Card (MMC) format, but are physically slightly thicker and feature higher data transfer rates, a lock switch on the side to prevent accidental overwriting and security features. The physical form factor, pin assignment and data transfer protocol are forward-compatible with the High Speed MultiMedia Card with some additions. SD slots can actually be used for more than flash memory cards. Devices that support SDIO can use small devices designed for the SD form factor, such as GPS receivers, Wi-Fi or Bluetooth adapters, modems, barcode readers, IrDA adapters, FM radio tuners, RFID readers, digital cameras and more. SD/SDIO is covered by numerous patents and trademarks, and licensing is only available through the Secure Digital Card Association. The SD/SDIO Card communication is based on a 9-pin interface (Clock, Command, 4 x Data and 3 x Power lines). The communication protocol is defined as a part of this specification. The main difference between the SD/SDIO Card and the High Speed MultiMedia Card is the initialization process. The SD/SDIO Card Register (HSMCI_SDCR) allows selection of the Card Slot and the data bus width. The SD/SDIO Card bus allows dynamic configuration of the number of data lines. After power up, by default, the SD/SDIO Card uses only DAT0 for data transfer. After initialization, the host can change the bus width (number of active data lines).
37.9.1 SDIO Data Transfer Type

SDIO cards may transfer data in either a multi-byte (1 to 512 bytes) or an optional block format (1 to 511 blocks), while the SD memory cards are fixed in the block transfer mode. The TRTYP field in the HSMCI Command Register (HSMCI_CMDR) allows to choose between SDIO Byte or SDIO Block transfer. The number of bytes/blocks to transfer is set through the BCNT field in the HSMCI Block Register (HSMCI_BLKR). In SDIO Block mode, the field BLKLEN must be set to the data block size while this field is not used in SDIO Byte mode. An SDIO Card can have multiple I/O or combined I/O and memory (called Combo Card). Within a multi-function SDIO or a Combo card, there are multiple devices (I/O and memory) that share access to the SD bus. In order to allow the sharing of access to the host among multiple devices, SDIO and combo cards can implement the optional concept of suspend/resume (Refer to the SDIO Specification for more details). To send a suspend or a resume command, the host must set the SDIO Special Command field (IOSPCMD) in the HSMCI Command Register.
37.9.2 SDIO Interrupts

Each function within an SDIO or Combo card may implement interrupts (Refer to the SDIO Specification for more details). In order to allow the SDIO card to interrupt the host, an interrupt function is added to a pin on the DAT[1] line to signal the cards interrupt to the host. An SDIO interrupt on each slot can be enabled through the HSMCI Interrupt Enable Register. The SDIO interrupt is sampled regardless of the currently selected slot.

1068
37.10 CE-ATA Operation

CE-ATA maps the streamlined ATA command set onto the MMC interface. The ATA task file is mapped onto MMC register space. CE-ATA utilizes five MMC commands:
z z z z z
GO_IDLE_STATE (CMD0): used for hard reset. STOP_TRANSMISSION (CMD12): causes the ATA command currently executing to be aborted. FAST_IO (CMD39): Used for single register access to the ATA taskfile registers, 8 bit access only. RW_MULTIPLE_REGISTERS (CMD60): used to issue an ATA command or to access the control/status registers. RW_MULTIPLE_BLOCK (CMD61): used to transfer data for an ATA command.
CE-ATA utilizes the same MMC command sequences for initialization as traditional MMC devices.
37.10.1 Executing an ATA Polling Command

1. 2. 3. 4. Issue READ_DMA_EXT with RW_MULTIPLE_REGISTER (CMD60) for 8kB of DATA. Read the ATA status register until DRQ is set. Issue RW_MULTIPLE_BLOCK (CMD61) to transfer DATA. Read the ATA status register until DRQ && BSY are set to 0.
37.10.2 Executing an ATA Interrupt Command

1. 2. 3. Issue READ_DMA_EXT with RW_MULTIPLE_REGISTER (CMD60) for 8kB of DATA with nIEN field set to zero to enable the command completion signal in the device. Issue RW_MULTIPLE_BLOCK (CMD61) to transfer DATA. Wait for Completion Signal Received Interrupt.
37.10.3 Aborting an ATA Command

If the host needs to abort an ATA command prior to the completion signal it must send a special command to avoid potential collision on the command line. The SPCMD field of the HSMCI_CMDR must be set to 3 to issue the CE-ATA completion Signal Disable Command.
37.10.4 CE-ATA Error Recovery

Several methods of ATA command failure may occur, including:
z z z z z
No response to an MMC command, such as RW_MULTIPLE_REGISTER (CMD60). CRC is invalid for an MMC command or response. CRC16 is invalid for an MMC data packet. ATA Status register reflects an error by setting the ERR bit to one. The command completion signal does not arrive within a host specified time out period.
Error conditions are expected to happen infrequently. Thus, a robust error recovery mechanism may be used for each error event. The recommended error recovery procedure after a timeout is:
z z z
Issue the command completion signal disable if nIEN was cleared to zero and the RW_MULTIPLE_BLOCK (CMD61) response has been received. Issue STOP_TRANSMISSION (CMD12) and successfully receive the R1 response. Issue a software reset to the CE-ATA device using FAST_IO (CMD39).
If STOP_TRANMISSION (CMD12) is successful, then the device is again ready for ATA commands. However, if the error recovery procedure does not work as expected or there is another timeout, the next step is to issue GO_IDLE_STATE (CMD0) to the device. GO_IDLE_STATE (CMD0) is a hard reset to the device and completely resets all device states. Note that after issuing GO_IDLE_STATE (CMD0), all device initialization needs to be completed again. If the CE-ATA device completes all MMC commands correctly but fails the ATA command with the ERR bit set in the ATA Status

1069
register, no error recovery action is required. The ATA command itself failed implying that the device could not complete the action requested, however, there was no communication or protocol failure. After the device signals an error by setting the ERR bit to one in the ATA Status register, the host may attempt to retry the command.
37.11 HSMCI Boot Operation Mode

37.11.1 In boot operation mode, the processor can read boot data from the slave (MMC device) by keeping the CMD line low after power-on before issuing CMD1. The data can be read from either the boot area or user area, depending on register setting. Boot Procedure, Processor Mode
1. 2. 3. 4. 5. 6. Configure the HSMCI data bus width programming SDCBUS Field in the HSMCI_SDCR register. The BOOT_BUS_WIDTH field located in the device Extended CSD register must be set accordingly. Set the byte count to 512 bytes and the block count to the desired number of blocks, writing BLKLEN and BCNT fields of the HSMCI_BLKR Register. Issue the Boot Operation Request command by writing to the HSMCI_CMDR register with SPCMD field set to BOOTREQ, TRDIR set to READ and TRCMD set to start data transfer. The BOOT_ACK field located in the HSMCI_CMDR register must be set to one, if the BOOT_ACK field of the MMC device located in the Extended CSD register is set to one. Host processor can copy boot data sequentially as soon as the RXRDY flag is asserted. When Data transfer is completed, host processor shall terminate the boot stream by writing the HSMCI_CMDR register with SPCMD field set to BOOTEND.
37.11.2 Boot Procedure DMA Mode

1. 2. 3. 4. 5. 6. 7. Configure the HSMCI data bus width by programming SDCBUS Field in the HSMCI_SDCR register. The BOOT_BUS_WIDTH field in the device Extended CSD register must be set accordingly. Set the byte count to 512 bytes and the block count to the desired number of blocks by writing BLKLEN and BCNT fields of the HSMCI_BLKR register. Enable DMA transfer in the HSMCI_DMA register. Configure DMA controller, program the total amount of data to be transferred and enable the relevant channel. Issue the Boot Operation Request command by writing to the HSMCI_CMDR register with SPCND set to BOOTREQ, TRDIR set to READ and TRCMD set to start data transfer. DMA controller copies the boot partition to the memory. When DMA transfer is completed, host processor shall terminate the boot stream by writing the HSMCI_CMDR register with SPCMD field set to BOOTEND.

1070
37.12 HSMCI Transfer Done Timings

37.12.1 Definition
The XFRDONE flag in the HSMCI_SR indicates exactly when the read or write sequence is finished.
37.12.2 Read Access

During a read access, the XFRDONE flag behaves as shown in Figure 37-12.
Figure 37-12. XFRDONE During a Read Access
CMD line HSMCI read CMD Card response
CMDRDY flag
The CMDRDY flag is released 8 tbit after the end of the card response.
Data 1st Block Not busy flag Last Block
XFRDONE flag
37.12.3 Write Access

During a write access, the XFRDONE flag behaves as shown in Figure 37-13.
Figure 37-13. XFRDONE During a Write Access
CMD line HSMCI write CMD Card response
CMDRDY flag
The CMDRDY flag is released 8 tbit after the end of the card response.
D0 1st Block Data bus - D0 Last Block
D0 is tied by the card D0 is released
1st Block Not busy flag
Last Block
XFRDONE flag

1071
37.13 Write Protection Registers

To prevent any single software error that may corrupt HSMCI behavior, the entire HSMCI address space from address offset 0x000 to 0x00FC can be write-protected by setting the WPEN bit in the HSMCI Write Protect Mode Register (HSMCI_WPMR). If a write access to anywhere in the HSMCI address space from address offset 0x000 to 0x00FC is detected, then the WPVS flag in the HSMCI Write Protect Status Register (HSMCI_WPSR) is set and the field WPVSRC indicates in which register the write access has been attempted. The WPVS flag is reset by writing the HSMCI Write Protect Mode Register (HSMCI_WPMR) with the appropriate access key, WPKEY. The protected registers are:
z z z z z z
HSMCI Mode Register on page 1075 HSMCI Data Timeout Register on page 1076 HSMCI SDCard/SDIO Register on page 1077 HSMCI Completion Signal Timeout Register on page 1082 HSMCI DMA Configuration Register on page 1094 HSMCI Configuration Register on page 1095

1072
37.14 High Speed MultiMedia Card Interface (HSMCI) User Interface

Table 37-9. Register Mapping Offset 0x00 0x04 0x08 0x0C 0x10 0x14 0x18 0x1C 0x20 0x24 0x28 0x2C 0x30 0x34 0x38 - 0x3C 0x40 0x44 0x48 0x4C 0x50 0x54 0x58-0xE0 0xE4 0xE8 0xEC - 0xFC 0x100-0x1FC 0x200 ... 0x5FC Register Control Register Mode Register Data Timeout Register SD/SDIO Card Register Argument Register Command Register Block Register Completion Signal Timeout Register Response Register Response Register Response Register
(1) (1) (1)
Name HSMCI_CR HSMCI_MR HSMCI_DTOR HSMCI_SDCR HSMCI_ARGR HSMCI_CMDR HSMCI_BLKR HSMCI_CSTOR HSMCI_RSPR HSMCI_RSPR HSMCI_RSPR HSMCI_RSPR HSMCI_RDR HSMCI_TDR HSMCI_SR HSMCI_IER HSMCI_IDR HSMCI_IMR HSMCI_DMA HSMCI_CFG HSMCI_WPMR HSMCI_WPSR HSMCI_FIFO0 ... HSMCI_FIFO255
Access Write Read-write Read-write Read-write Read-write Write Read-write Read-write Read Read Read Read Read Write Read Write Write Read Read-write Read-write Read-write Read-only Read-write ... Read-write
Reset 0x0 0x0 0x0 0x0 0x0 0x0 0x0 0x0 0x0 0x0 0x0 0xC0E5 0x0 0x00 0x00 0x0 ... 0x0
Response Register(1) Receive Data Register Transmit Data Register Reserved Status Register Interrupt Enable Register Interrupt Disable Register Interrupt Mask Register DMA Configuration Register Configuration Register Reserved Write Protection Mode Register Write Protection Status Register Reserved Reserved FIFO Memory Aperture0 ... FIFO Memory Aperture255
Notes: 1. The Response Register can be read by N accesses at the same HSMCI_RSPR or at consecutive addresses (0x20 to 0x2C). N depends on the size of the response.

1073
37.14.1 HSMCI Control Register

31 23 15 7 SWRST
HSMCI_CR 0xF0000000 (0), 0xF8000000 (1), 0xF8004000 (2) Write-only

30 22 14 6 29 21 13 5 28 20 12 4 27 19 11 3 PWSDIS 26 18 10 2 PWSEN 25 17 9 1 MCIDIS 24 16 8 0 MCIEN
MCIEN: Multi-Media Interface Enable

0 = No effect. 1 = Enables the Multi-Media Interface if MCDIS is 0.
MCIDIS: Multi-Media Interface Disable

0 = No effect. 1 = Disables the Multi-Media Interface.
PWSEN: Power Save Mode Enable

0 = No effect. 1 = Enables the Power Saving Mode if PWSDIS is 0. Warning: Before enabling this mode, the user must set a value different from 0 in the PWSDIV field (Mode Register, HSMCI_MR).
PWSDIS: Power Save Mode Disable

0 = No effect. 1 = Disables the Power Saving Mode.
SWRST: Software Reset

0 = No effect. 1 = Resets the HSMCI. A software triggered hardware reset of the HSMCI interface is performed.

1074
37.14.2 HSMCI Mode Register

Name: Address: Access: HSMCI_MR 0xF0000004 (0), 0xF8000004 (1), 0xF8004004 (2) Read-write
31 23 15 7
30 22 14 PADV 6
29 21 13 FBYTE 5
28 20 12 WRPROOF 4 CLKDIV
27 19 11 RDPROOF 3
26 18 10
25 17 9 PWSDIV 1
24 16 CLKODD 8
This register can only be written if the WPEN bit is cleared in HSMCI Write Protect Mode Register on page 1096.
CLKDIV: Clock Divider High Speed MultiMedia Card Interface clock (MCCK or HSMCI_CK) is Master Clock (MCK) divider by ({CLKDIV,CLKODD}+2). PWSDIV: Power Saving Divider High Speed MultiMedia Card Interface clock is divided by 2(PWSDIV) + 1 when entering Power Saving Mode. Warning: This value must be different from 0 before enabling the Power Save Mode in the HSMCI_CR (HSMCI_PWSEN bit). RDPROOF: Read Proof Enable Enabling Read Proof allows to stop the HSMCI Clock during read access if the internal FIFO is full. This will guarantee data integrity, not bandwidth. 0 = Disables Read Proof. 1 = Enables Read Proof. WRPROOF: Write Proof Enable Enabling Write Proof allows to stop the HSMCI Clock during write access if the internal FIFO is full. This will guarantee data integrity, not bandwidth. 0 = Disables Write Proof. 1 = Enables Write Proof. FBYTE: Force Byte Transfer Enabling Force Byte Transfer allow byte transfers, so that transfer of blocks with a size different from modulo 4 can be supported. Warning: BLKLEN value depends on FBYTE. 0 = Disables Force Byte Transfer. 1 = Enables Force Byte Transfer. PADV: Padding Value 0 = 0x00 value is used when padding data in write transfer. 1 = 0xFF value is used when padding data in write transfer. PADV may be only in manual transfer. CLKODD: Clock divider is odd This field is the least significant bit of the clock divider and indicates the clock divider parity. SAMA5D3 Series [DATASHEET]
1075
37.14.3 HSMCI Data Timeout Register

31 23 15 7
HSMCI_DTOR 0xF0000008 (0), 0xF8000008 (1), 0xF8004008 (2) Read-write

30 22 14 6 29 21 13 5 DTOMUL 28 20 12 4 27 19 11 3 26 18 10 2 DTOCYC 25 17 9 1 24 16 8 0
DTOCYC: Data Timeout Cycle Number

These fields determine the maximum number of Master Clock cycles that the HSMCI waits between two data block transfers. It equals (DTOCYC x Multiplier).
DTOMUL: Data Timeout Multiplier

Multiplier is defined by DTOMUL as shown in the following table:
Value 0 1 2 3 4 5 6 7 Name 1 16 128 256 1024 4096 65536 1048576 Description DTOCYC DTOCYC x 16 DTOCYC x 128 DTOCYC x 256 DTOCYC x 1024 DTOCYC x 4096 DTOCYC x 65536 DTOCYC x 1048576
If the data time-out set by DTOCYC and DTOMUL has been exceeded, the Data Time-out Error flag (DTOE) in the HSMCI Status Register (HSMCI_SR) rises.

1076
37.14.4 HSMCI SDCard/SDIO Register

31 23 15 7 SDCBUS
HSMCI_SDCR 0xF000000C (0), 0xF800000C (1), 0xF800400C (2) Read-write

30 22 14 6 29 21 13 5 28 20 12 4 27 19 11 3 26 18 10 2 25 17 9 1 SDCSEL 24 16 8 0
SDCSEL: SDCard/SDIO Slot

Value 0 1 2 3 Name SLOTA SLOTB SLOTC SLOTD Description
Slot A is selected.
Reserved Reserved Reserved
SDCBUS: SDCard/SDIO Bus Width

Value 0 1 2 3 Name 1 4 8 Description
1 bit Reserved
4 bit 8 bit

1077
37.14.5 HSMCI Argument Register

31
HSMCI_ARGR 0xF0000010 (0), 0xF8000010 (1), 0xF8004010 (2) Read-write

30 29 28 ARG 27 26 25 24
23
22
21
20 ARG
19
18
17
16
15
14
13
12 ARG
11
10
4 ARG
ARG: Command Argument

1078
37.14.6 HSMCI Command Register

31 23 15 7 RSPTYP
HSMCI_CMDR 0xF0000014 (0), 0xF8000014 (1), 0xF8004014 (2) Write-only

30 22 14 6 29 21 28 20 TRTYP 12 MAXLAT 4 27 BOOT_ACK 19 26 ATACS 18 TRDIR 10 25 IOSPCMD 17 TRCMD 9 SPCMD 1 8 16 24
13 5
11 OPDCMD 3 CMDNB
This register is write-protected while CMDRDY is 0 in HSMCI_SR. If an Interrupt command is sent, this register is only writable by an interrupt response (field SPCMD). This means that the current command execution cannot be interrupted or modified.
CMDNB: Command Number

This is the command index.
RSPTYP: Response Type

Value 0 1 2 3 Name NORESP 48_BIT 136_BIT R1B Description No response. 48-bit response. 136-bit response. R1b response type
SPCMD: Special Command

Value 0 1 2 Name STD INIT SYNC Description Not a special CMD. Initialization CMD: 74 clock cycles for initialization sequence. Synchronized CMD: Wait for the end of the current data block transfer before sending the pending command. CE-ATA Completion Signal disable Command. The host cancels the ability for the device to return a command completion signal on the command line. Interrupt command: Corresponds to the Interrupt Mode (CMD40). Interrupt response: Corresponds to the Interrupt Mode (CMD40). Boot Operation Request. Start a boot operation mode, the host processor can read boot data from the MMC device directly. End Boot Operation. This command allows the host processor to terminate the boot operation mode.
CE_ATA
4 5 6 7
IT_CMD IT_RESP BOR EBO

1079
OPDCMD: Open Drain Command

0 (PUSHPULL) = Push pull command. 1 (OPENDRAIN) = Open drain command.
MAXLAT: Max Latency for Command to Response

0 (5) = 5-cycle max latency. 1 (64) = 64-cycle max latency.
TRCMD: Transfer Command

Value 0 1 2 3 Name NO_DATA START_DATA STOP_DATA Description No data transfer Start data transfer Stop data transfer Reserved
TRDIR: Transfer Direction

0 (WRITE) = Write. 1 (READ) = Read.
TRTYP: Transfer Type

Value 0 1 2 4 5 Name SINGLE MULTIPLE STREAM BYTE BLOCK Description MMC/SD Card Single Block MMC/SD Card Multiple Block MMC Stream SDIO Byte SDIO Block
IOSPCMD: SDIO Special Command

Value 0 1 2 Name STD SUSPEND RESUME Description Not an SDIO Special Command SDIO Suspend Command SDIO Resume Command
ATACS: ATA with Command Completion Signal

0 (NORMAL) = Normal operation mode. 1 (COMPLETION) = This bit indicates that a completion signal is expected within a programmed amount of time (HSMCI_CSTOR).
BOOT_ACK: Boot Operation Acknowledge.

The master can choose to receive the boot acknowledge from the slave when a Boot Request command is issued. When set to one this field indicates that a Boot acknowledge is expected within a programmable amount of time defined with DTOMUL and DTOCYC fields located in the HSMCI_DTOR register. If the acknowledge pattern is not received then an acknowledge timeout error is raised. If the acknowledge pattern is corrupted then an acknowledge pattern error is set.

1080
37.14.7 HSMCI Block Register

31
HSMCI_BLKR 0xF0000018 (0), 0xF8000018 (1), 0xF8004018 (2) Read-write

30 29 28 BLKLEN 27 26 25 24
23
22
21
20 BLKLEN
19
18
17
16
15
14
13
12 BCNT
11
10
4 BCNT
BCNT: MMC/SDIO Block Count - SDIO Byte Count

This field determines the number of data byte(s) or block(s) to transfer. The transfer data type and the authorized values for BCNT field are determined by the TRTYP field in the HSMCI Command Register (HSMCI_CMDR). When TRTYP=1 (MMC/SDCARD Multiple Block), BCNT can be programmed from 1 to 65535, 0 corresponds to an infinite block transfer. When TRTYP=4 (SDIO Byte), BCNT can be programmed from 1 to 511, 0 corresponds to 512-byte transfer. Values in range 512 to 65536 are forbidden. When TRTYP=5 (SDIO Block), BCNT can be programmed from 1 to 511, 0 corresponds to an infinite block transfer. Values in range 512 to 65536 are forbidden. Warning: In SDIO Byte and Block modes (TRTYP=4 or 5), writing the 7 last bits of BCNT field with a value which differs from 0 is forbidden and may lead to unpredictable results.
BLKLEN: Data Block Length

This field determines the size of the data block. This field is also accessible in the HSMCI Mode Register (HSMCI_MR). Bits 16 and 17 must be set to 0 if FBYTE is disabled. Note: In SDIO Byte mode, BLKLEN field is not used.

1081
37.14.8 HSMCI Completion Signal Timeout Register

31 23 15 7
HSMCI_CSTOR 0xF000001C (0), 0xF800001C (1), 0xF800401C (2) Read-write

30 22 14 6 29 21 13 5 CSTOMUL 28 20 12 4 27 19 11 3 26 18 10 2 CSTOCYC 25 17 9 1 24 16 8 0
CSTOCYC: Completion Signal Timeout Cycle Number

These fields determine the maximum number of Master Clock cycles that the HSMCI waits between two data block transfers. Its value is calculated by (CSTOCYC x Multiplier).
CSTOMUL: Completion Signal Timeout Multiplier

These fields determine the maximum number of Master Clock cycles that the HSMCI waits between two data block transfers. Its value is calculated by (CSTOCYC x Multiplier). These fields determine the maximum number of Master Clock cycles that the HSMCI waits between the end of the data transfer and the assertion of the completion signal. The data transfer comprises data phase and the optional busy phase. If a non-DATA ATA command is issued, the HSMCI starts waiting immediately after the end of the response until the completion signal. Multiplier is defined by CSTOMUL as shown in the following table:
Value 0 1 2 3 4 5 6 7 Name 1 16 128 256 1024 4096 65536 1048576 Description CSTOCYC x 1 CSTOCYC x 16 CSTOCYC x 128 CSTOCYC x 256 CSTOCYC x 1024 CSTOCYC x 4096 CSTOCYC x 65536 CSTOCYC x 1048576
If the data time-out set by CSTOCYC and CSTOMUL has been exceeded, the Completion Signal Time-out Error flag (CSTOE) in the HSMCI Status Register (HSMCI_SR) rises.

1082
37.14.9 HSMCI Response Register

31
HSMCI_RSPR 0xF0000020 (0), 0xF8000020 (1), 0xF8004020 (2) Read-only

30 29 28 RSP 23 22 21 20 RSP 15 14 13 12 RSP 7 6 5 4 RSP 3 2 1 0 11 10 9 8 19 18 17 16 27 26 25 24
RSP: Response Note: 1. The response register can be read by N accesses at the same HSMCI_RSPR or at consecutive addresses (0x20 to 0x2C). N depends on the size of the response.

1083
37.14.10 HSMCI Receive Data Register

31
HSMCI_RDR 0xF0000030 (0), 0xF8000030 (1), 0xF8004030 (2) Read-only

30 29 28 DATA 27 26 25 24
23
22
21
20 DATA
19
18
17
16
15
14
13
12 DATA
11
10
4 DATA
DATA: Data to Read
37.14.11 HSMCI Transmit Data Register

31
HSMCI_TDR 0xF0000034 (0), 0xF8000034 (1), 0xF8004034 (2) Write-only

30 29 28 DATA 27 26 25 24
23
22
21
20 DATA
19
18
17
16
15
14
13
12 DATA
11
10
4 DATA
DATA: Data to Write

1084
37.14.12 HSMCI Status Register

31 UNRE 23 CSTOE 15 7
HSMCI_SR 0xF0000040 (0), 0xF8000040 (1), 0xF8004040 (2) Read-only

30 OVRE 22 DTOE 14 6 29 ACKRCVE 21 DCRCE 13 CSRCV 5 NOTBUSY 28 ACKRCV 20 RTOE 12 SDIOWAIT 4 DTIP 27 XFRDONE 19 RENDE 11 3 BLKE 26 FIFOEMPTY 18 RCRCE 10 2 TXRDY 25 DMADONE 17 RDIRE 9 1 RXRDY 24 BLKOVRE 16 RINDE 8 SDIOIRQA 0 CMDRDY
CMDRDY: Command Ready

0 = A command is in progress. 1 = The last command has been sent. Cleared when writing in the HSMCI_CMDR.
RXRDY: Receiver Ready

0 = Data has not yet been received since the last read of HSMCI_RDR. 1 = Data has been received since the last read of HSMCI_RDR.
TXRDY: Transmit Ready

0= The last data written in HSMCI_TDR has not yet been transferred in the Shift Register. 1= The last data written in HSMCI_TDR has been transferred in the Shift Register.
BLKE: Data Block Ended

This flag must be used only for Write Operations. 0 = A data block transfer is not yet finished. Cleared when reading the HSMCI_SR. 1 = A data block transfer has ended, including the CRC16 Status transmission. the flag is set for each transmitted CRC Status. Refer to the MMC or SD Specification for more details concerning the CRC Status.
DTIP: Data Transfer in Progress

0 = No data transfer in progress. 1 = The current data transfer is still in progress, including CRC16 calculation. Cleared at the end of the CRC16 calculation.
NOTBUSY: HSMCI Not Busy

A block write operation uses a simple busy signalling of the write operation duration on the data (DAT0) line: during a data transfer block, if the card does not have a free data receive buffer, the card indicates this condition by pulling down the data line (DAT0) to LOW. The card stops pulling down the data line as soon as at least one receive buffer for the defined data transfer block length becomes free. Refer to the MMC or SD Specification for more details concerning the busy behavior. For all the read operations, the NOTBUSY flag is cleared at the end of the host command. For the Infinite Read Multiple Blocks, the NOTBUSY flag is set at the end of the STOP_TRANSMISSION host command (CMD12).

1085
For the Single Block Reads, the NOTBUSY flag is set at the end of the data read block. For the Multiple Block Reads with pre-defined block count, the NOTBUSY flag is set at the end of the last received data block. The NOTBUSY flag allows to deal with these different states. 0 = The HSMCI is not ready for new data transfer. Cleared at the end of the card response. 1 = The HSMCI is ready for new data transfer. Set when the busy state on the data line has ended. This corresponds to a free internal data receive buffer of the card.
SDIOIRQA: SDIO Interrupt for Slot A

0 = No interrupt detected on SDIO Slot A. 1 = An SDIO Interrupt on Slot A occurred. Cleared when reading the HSMCI_SR.
SDIOWAIT: SDIO Read Wait Operation Status

0 = Normal Bus operation. 1 = The data bus has entered IO wait state.
CSRCV: CE-ATA Completion Signal Received

0 = No completion signal received since last status read operation. 1 = The device has issued a command completion signal on the command line. Cleared by reading in the HSMCI_SR register.
RINDE: Response Index Error

0 = No error. 1 = A mismatch is detected between the command index sent and the response index received. Cleared when writing in the HSMCI_CMDR.
RDIRE: Response Direction Error

0 = No error. 1 = The direction bit from card to host in the response has not been detected.
RCRCE: Response CRC Error

0 = No error. 1 = A CRC7 error has been detected in the response. Cleared when writing in the HSMCI_CMDR.
RENDE: Response End Bit Error

0 = No error. 1 = The end bit of the response has not been detected. Cleared when writing in the HSMCI_CMDR.
RTOE: Response Time-out Error

0 = No error. 1 = The response time-out set by MAXLAT in the HSMCI_CMDR has been exceeded. Cleared when writing in the HSMCI_CMDR.
DCRCE: Data CRC Error

0 = No error. 1 = A CRC16 error has been detected in the last data block. Cleared by reading in the HSMCI_SR register.

1086
DTOE: Data Time-out Error

0 = No error. 1 = The data time-out set by DTOCYC and DTOMUL in HSMCI_DTOR has been exceeded. Cleared by reading in the HSMCI_SR register.
CSTOE: Completion Signal Time-out Error

0 = No error. 1 = The completion signal time-out set by CSTOCYC and CSTOMUL in HSMCI_CSTOR has been exceeded. Cleared by reading in the HSMCI_SR register. Cleared by reading in the HSMCI_SR register.
BLKOVRE: DMA Block Overrun Error

0 = No error. 1 = A new block of data is received and the DMA controller has not started to move the current pending block, a block overrun is raised. Cleared by reading in the HSMCI_SR register.
DMADONE: DMA Transfer done

0 = DMA buffer transfer has not completed since the last read of the HSMCI_SR register. 1 = DMA buffer transfer has completed.
FIFOEMPTY: FIFO empty flag

0 = FIFO contains at least one byte. 1 = FIFO is empty.
XFRDONE: Transfer Done flag

0 = A transfer is in progress. 1 = Command Register is ready to operate and the data bus is in the idle state.
ACKRCV: Boot Operation Acknowledge Received

0 = No Boot acknowledge received since the last read of the status register. 1 = A Boot acknowledge signal has been received. Cleared by reading the HSMCI_SR register.
ACKRCVE: Boot Operation Acknowledge Error

0 = No error 1 = Corrupted Boot Acknowledge signal received.
OVRE: Overrun
0 = No error. 1 = At least one 8-bit received data has been lost (not read). Cleared when sending a new data transfer command. When FERRCTRL in HSMCI_CFG is set to 1, OVRE becomes reset after read.
UNRE: Underrun
0 = No error. 1 = At least one 8-bit data has been sent without valid information (not written). Cleared when sending a new data transfer command or when setting FERRCTRL in HSMCI_CFG to 1. When FERRCTRL in HSMCI_CFG is set to 1, UNRE becomes reset after read.

1087
37.14.13 HSMCI Interrupt Enable Register

HSMCI_IER 0xF0000044 (0), 0xF8000044 (1), 0xF8004044 (2) Write-only

CMDRDY: Command Ready Interrupt Enable RXRDY: Receiver Ready Interrupt Enable TXRDY: Transmit Ready Interrupt Enable BLKE: Data Block Ended Interrupt Enable DTIP: Data Transfer in Progress Interrupt Enable NOTBUSY: Data Not Busy Interrupt Enable SDIOIRQA: SDIO Interrupt for Slot A Interrupt Enable SDIOWAIT: SDIO Read Wait Operation Status Interrupt Enable CSRCV: Completion Signal Received Interrupt Enable RINDE: Response Index Error Interrupt Enable RDIRE: Response Direction Error Interrupt Enable RCRCE: Response CRC Error Interrupt Enable RENDE: Response End Bit Error Interrupt Enable RTOE: Response Time-out Error Interrupt Enable DCRCE: Data CRC Error Interrupt Enable DTOE: Data Time-out Error Interrupt Enable CSTOE: Completion Signal Timeout Error Interrupt Enable BLKOVRE: DMA Block Overrun Error Interrupt Enable DMADONE: DMA Transfer completed Interrupt Enable

1088
FIFOEMPTY: FIFO empty Interrupt enable XFRDONE: Transfer Done Interrupt enable ACKRCV: Boot Acknowledge Interrupt Enable ACKRCVE: Boot Acknowledge Error Interrupt Enable OVRE: Overrun Interrupt Enable UNRE: Underrun Interrupt Enable
0 = No effect. 1 = Enables the corresponding interrupt.

1089
37.14.14 HSMCI Interrupt Disable Register

HSMCI_IDR 0xF0000048 (0), 0xF8000048 (1), 0xF8004048 (2) Write-only

CMDRDY: Command Ready Interrupt Disable RXRDY: Receiver Ready Interrupt Disable TXRDY: Transmit Ready Interrupt Disable BLKE: Data Block Ended Interrupt Disable DTIP: Data Transfer in Progress Interrupt Disable NOTBUSY: Data Not Busy Interrupt Disable SDIOIRQA: SDIO Interrupt for Slot A Interrupt Disable SDIOWAIT: SDIO Read Wait Operation Status Interrupt Disable CSRCV: Completion Signal received interrupt Disable RINDE: Response Index Error Interrupt Disable RDIRE: Response Direction Error Interrupt Disable RCRCE: Response CRC Error Interrupt Disable RENDE: Response End Bit Error Interrupt Disable RTOE: Response Time-out Error Interrupt Disable DCRCE: Data CRC Error Interrupt Disable DTOE: Data Time-out Error Interrupt Disable CSTOE: Completion Signal Time out Error Interrupt Disable BLKOVRE: DMA Block Overrun Error Interrupt Disable DMADONE: DMA Transfer completed Interrupt Disable

1090
FIFOEMPTY: FIFO empty Interrupt Disable XFRDONE: Transfer Done Interrupt Disable ACKRCV: Boot Acknowledge Interrupt Disable ACKRCVE: Boot Acknowledge Error Interrupt Disable OVRE: Overrun Interrupt Disable UNRE: Underrun Interrupt Disable
0 = No effect. 1 = Disables the corresponding interrupt.

1091
37.14.15 HSMCI Interrupt Mask Register

HSMCI_IMR 0xF000004C (0), 0xF800004C (1), 0xF800404C (2) Read-only

CMDRDY: Command Ready Interrupt Mask RXRDY: Receiver Ready Interrupt Mask TXRDY: Transmit Ready Interrupt Mask BLKE: Data Block Ended Interrupt Mask DTIP: Data Transfer in Progress Interrupt Mask NOTBUSY: Data Not Busy Interrupt Mask SDIOIRQA: SDIO Interrupt for Slot A Interrupt Mask SDIOWAIT: SDIO Read Wait Operation Status Interrupt Mask CSRCV: Completion Signal Received Interrupt Mask RINDE: Response Index Error Interrupt Mask RDIRE: Response Direction Error Interrupt Mask RCRCE: Response CRC Error Interrupt Mask RENDE: Response End Bit Error Interrupt Mask RTOE: Response Time-out Error Interrupt Mask DCRCE: Data CRC Error Interrupt Mask DTOE: Data Time-out Error Interrupt Mask CSTOE: Completion Signal Time-out Error Interrupt Mask BLKOVRE: DMA Block Overrun Error Interrupt Mask DMADONE: DMA Transfer Completed Interrupt Mask

1092
FIFOEMPTY: FIFO Empty Interrupt Mask XFRDONE: Transfer Done Interrupt Mask ACKRCV: Boot Operation Acknowledge Received Interrupt Mask ACKRCVE: Boot Operation Acknowledge Error Interrupt Mask OVRE: Overrun Interrupt Mask UNRE: Underrun Interrupt Mask
0 = The corresponding interrupt is not enabled. 1 = The corresponding interrupt is enabled.

1093
37.14.16 HSMCI DMA Configuration Register

31
HSMCI_DMA 0xF0000050 (0), 0xF8000050 (1), 0xF8004050 (2) Read-write

30 29 21 13 5 CHKSIZE 28 20 12 ROPT 4 27 19 11 3 26 18 10 2 25 17 9 1 OFFSET 24 16 8 DMAEN 0
23 15 7
22 14 6
OFFSET: DMA Write Buffer Offset

This field indicates the number of discarded bytes when the DMA writes the first word of the transfer.
CHKSIZE: DMA Channel Read and Write Chunk Size

The CHKSIZE field indicates the number of data available when the DMA chunk transfer request is asserted.
Value 0 1 2 3 4
Name 1 4 8 16 32
Description 1 data available 4 data available 8 data available 16 data available 32 data available
DMAEN: DMA Hardware Handshaking Enable

0 = DMA interface is disabled. 1 = DMA Interface is enabled. Note: To avoid unpredictable behavior, DMA hardware handshaking must be disabled when CPU transfers are performed.
ROPT: Read Optimization with padding

0: BLKLEN bytes are moved from the Memory Card to the system memory, two DMA descriptors are used when the transfer size is not a multiple of 4. 1: Ceiling(BLKLEN/4) * 4 bytes are moved from the Memory Card to the system memory, only one DMA descriptor is used.

1094
37.14.17 HSMCI Configuration Register

31
HSMCI_CFG 0xF0000054 (0), 0xF8000054 (1), 0xF8004054 (2) Read-write

30 29 21 13 5 28 20 12 LSYNC 4 FERRCTRL 27 19 11 3 26 18 10 2 25 17 9 1 24 16 8 HSMODE 0 FIFOMODE
23 15 7
22 14 6
FIFOMODE: HSMCI Internal FIFO control mode

0 = A write transfer starts when a sufficient amount of data is written into the FIFO. When the block length is greater than or equal to 3/4 of the HSMCI internal FIFO size, then the write transfer starts as soon as half the FIFO is filled. When the block length is greater than or equal to half the internal FIFO size, then the write transfer starts as soon as one quarter of the FIFO is filled. In other cases, the transfer starts as soon as the total amount of data is written in the internal FIFO. 1 = A write transfer starts as soon as one data is written into the FIFO.
FERRCTRL: Flow Error flag reset control mode

0= When an underflow/overflow condition flag is set, a new Write/Read command is needed to reset the flag. 1= When an underflow/overflow condition flag is set, a read status resets the flag.
HSMODE: High Speed Mode

0= Default bus timing mode. 1= If set to one, the host controller outputs command line and data lines on the rising edge of the card clock. The Host driver shall check the high speed support in the card registers.
LSYNC: Synchronize on the last block

0= The pending command is sent at the end of the current data block. 1= The pending command is sent at the end of the block transfer when the transfer length is not infinite. (block count shall be different from zero)

1095
37.14.18 HSMCI Write Protect Mode Register

31
HSMCI_WPMR 0xF00000E4 (0), 0xF80000E4 (1), 0xF80040E4 (2) Read-write

30 29 28 27 WP_KEY (0x4D => M) 20 19 WP_KEY (0x43 => C) 12 11 WP_KEY (0x49 => I) 4 3 26 25 24
23
22
21
18
17
16
15
14
13
10
0 WP_EN

0 = Disables the Write Protection if WP_KEY corresponds to 0x4D4349 (MCI in ASCII). 1 = Enables the Write Protection if WP_KEY corresponds to 0x4D4349 (MCI in ASCII).
WP_KEY: Write Protection Key password

Should be written at value 0x4D4349 (ASCII code for MCI). Writing any other value in this field has no effect. Protects the registers:
HSMCI Mode Register on page 1075 HSMCI Data Timeout Register on page 1076 HSMCI SDCard/SDIO Register on page 1077 HSMCI Completion Signal Timeout Register on page 1082 HSMCI DMA Configuration Register on page 1094 HSMCI Configuration Register on page 1095

1096
37.14.19 HSMCI Write Protect Status Register

31 23
HSMCI_WPSR 0xF00000E8 (0), 0xF80000E8 (1), 0xF80040E8 (2) Read-only

30 22 29 21 28 20 WP_VSRC 27 19 26 18 25 17 24 16
15
14
13
12 WP_VSRC
11
10
2 WP_VS

Value 0 1 2 3 Name NONE WRITE RESET BOTH Description No Write Protection Violation occurred since the last read of this register (WP_SR) Write Protection detected unauthorized attempt to write a control register had occurred (since the last read.) Software reset had been performed while Write Protection was enabled (since the last read). Both Write Protection violation and software reset with Write Protection enabled have occurred since the last read.
WP_VSRC: Write Protection Violation SouRCe

When WPVS is active, this field indicates the write-protected register (through address offset or code) in which a write access has been attempted.

1097
37.14.20 HSMCI FIFOx Memory Aperture

31
HSMCI_FIFOx[x=0..255] 0xF0000200 (0), 0xF8000200 (1), 0xF8004200 (2) Read-write

30 29 28 DATA 27 26 25 24
23
22
21
20 DATA
19
18
17
16
15
14
13
12 DATA
11
10
4 DATA
DATA: Data to Read or Data to Write

1098
38.
38.1
Serial Peripheral Interface (SPI)

Description
The Serial Peripheral Interface (SPI) circuit is a synchronous serial data link that provides communication with external devices in Master or Slave Mode. It also enables communication between processors if an external processor is connected to the system. The Serial Peripheral Interface is essentially a shift register that serially transmits data bits to other SPIs. During a data transfer, one SPI system acts as the master' which controls the data flow, while the other devices act as slaves'' which have data shifted into and out by the master. Different CPUs can take turn being masters (Multiple Master Protocol opposite to Single Master Protocol where one CPU is always the master while all of the others are always slaves) and one master may simultaneously shift data into multiple slaves. However, only one slave may drive its output to write data back to the master at any given time. A slave device is selected when the master asserts its NSS signal. If multiple slave devices exist, the master generates a separate slave select signal for each slave (NPCS). The SPI system consists of two data lines and two control lines:
z z z z
Master Out Slave In (MOSI): This data line supplies the output data from the master shifted into the input(s) of the slave(s). Master In Slave Out (MISO): This data line supplies the output data from a slave to the input of the master. There may be no more than one slave transmitting data during any particular transfer. Serial Clock (SPCK): This control line is driven by the master and regulates the flow of the data bits. The master may transmit data at a variety of baud rates; the SPCK line cycles once for each bit that is transmitted. Slave Select (NSS): This control line allows slaves to be turned on and off by hardware.
38.2
z
Supports Communication with Serial External Devices

z z z z z z
Master Mode can drive SPCK up to peripheral clock (bounded by maximum bus clock divided by 2) Slave Mode operates on SPCK, asynchronously to Core and Bus Clock Four Chip Selects with External Decoder Support Allow Communication with Up to 15 Peripherals Serial Memories, such as DataFlash and 3-wire EEPROMs Serial Peripherals, such as ADCs, DACs, LCD Controllers, CAN Controllers and Sensors External Coprocessors 8-bit to 16-bit Programmable Data Length Per Chip Select Programmable Phase and Polarity Per Chip Select Programmable Transfer Delay Between Consecutive Transfers and Delay before SPI Clock per Chip Select Programmable Delay Between Chip Selects Selectable Mode Fault Detection One channel for the Receiver, One Channel for the Transmitter
Master or Slave Serial Peripheral Bus Interface

z z z z z
Connection to DMA Channel Capabilities Optimizes Data Transfers

z

1099
38.3
Block Diagram

AHB Matrix DMA Ch.
Peripheral Bridge APB SPCK MISO MCK SPI Interface PIO MOSI NPCS0/NSS NPCS1 NPCS2 Interrupt Control NPCS3
PMC
SPI Interrupt

1100
38.4
Figure 38-2. Application Block Diagram: Single Master/Multiple Slave Implementation

SPCK MISO MOSI SPI Master NPCS0 NPCS1 NPCS2 NPCS3 NC SPCK MISO Slave 0 MOSI NSS SPCK MISO Slave 1 MOSI NSS SPCK MISO Slave 2 MOSI NSS
38.5
Signal Description
Signal Description
Pin Name Pin Description Master MISO MOSI SPCK NPCS1-NPCS3 NPCS0/NSS Master In Slave Out Master Out Slave In Serial Clock Peripheral Chip Selects Peripheral Chip Select/Slave Select Input Output Output Output Output Type Slave Output Input Input Unused Input

1101
38.6
38.6.1 I/O Lines

The pins used for interfacing the compliant external devices may be multiplexed with PIO lines. The programmer must first program the PIO controllers to assign the SPI pins to their peripheral functions.
Table 38-1. I/O Lines Instance SPI0 SPI0 SPI0 SPI0 SPI0 SPI0 SPI0 SPI1 SPI1 SPI1 SPI1 SPI1 SPI1 SPI1 Signal SPI0_MISO SPI0_MOSI SPI0_NPCS0 SPI0_NPCS1 SPI0_NPCS2 SPI0_NPCS3 SPI0_SPCK SPI1_MISO SPI1_MOSI SPI1_NPCS0 SPI1_NPCS1 SPI1_NPCS2 SPI1_NPCS3 SPI1_SPCK I/O Line PD10 PD11 PD13 PD14 PD15 PD16 PD12 PC22 PC23 PC25 PC26 PC27 PC28 PC24 Peripheral A A A B B B A A A A A A A A

The SPI may be clocked through the Power Management Controller (PMC), thus the programmer must first configure the PMC to enable the SPI clock.
38.6.3 Interrupt
The SPI interface has an interrupt line connected to the Interrupt Controller. Handling the SPI interrupt requires programming the interrupt controller before configuring the SPI.
Table 38-2. Peripheral IDs Instance SPI0 SPI1 ID 24 25
38.6.4 Direct Memory Access Controller (DMAC)

The SPI interface can be used in conjunction with the DMAC in order to reduce processor overhead. For a full description of the DMAC, refer to the corresponding section in this datasheet.

1102
38.7
38.7.1 Modes of Operation

The SPI operates in Master Mode or in Slave Mode. Operation in Master Mode is programmed by writing at 1 the MSTR bit in the Mode Register. The pins NPCS0 to NPCS3 are all configured as outputs, the SPCK pin is driven, the MISO line is wired on the receiver input and the MOSI line driven as an output by the transmitter. If the MSTR bit is written at 0, the SPI operates in Slave Mode. The MISO line is driven by the transmitter output, the MOSI line is wired on the receiver input, the SPCK pin is driven by the transmitter to synchronize the receiver. The NPCS0 pin becomes an input, and is used as a Slave Select signal (NSS). The pins NPCS1 to NPCS3 are not driven and can be used for other purposes. The data transfers are identically programmable for both modes of operations. The baud rate generator is activated only in Master Mode.
38.7.2 Data Transfer

Four combinations of polarity and phase are available for data transfers. The clock polarity is programmed with the CPOL bit in the Chip Select Register. The clock phase is programmed with the NCPHA bit. These two parameters determine the edges of the clock signal on which data is driven and sampled. Each of the two parameters has two possible states, resulting in four possible combinations that are incompatible with one another. Thus, a master/slave pair must use the same parameter pair values to communicate. If multiple slaves are used and fixed in different configurations, the master must reconfigure itself each time it needs to communicate with a different slave. Table 38-3 shows the four modes and corresponding parameter settings.
Table 38-3. SPI Bus Protocol Mode SPI Mode 0 1 2 3 CPOL 0 0 1 1 NCPHA 1 0 1 0 Shift SPCK Edge Falling Rising Rising Falling Capture SPCK Edge Rising Falling Falling Rising SPCK Inactive Level Low Low High High
Figure 38-3 and Figure 38-4 show examples of data transfers.

1103
Figure 38-3. SPI Transfer Format (NCPHA = 1, 8 bits per transfer)

SPCK cycle (for reference) SPCK (CPOL = 0) 1 2 3 4 5 6 7 8
SPCK (CPOL = 1)
MOSI (from master)
MSB
LSB
MISO (from slave)
MSB
LSB
NSS (to slave)
* Not defined, but normally MSB of previous character received.

Figure 38-4. SPI Transfer Format (NCPHA = 0, 8 bits per transfer)
SPCK cycle (for reference) SPCK (CPOL = 0) 1 2 3 4 5 6 7 8
SPCK (CPOL = 1)
MOSI (from master)
MSB
LSB
MISO (from slave)
MSB
LSB
NSS (to slave)
* Not defined but normally LSB of previous character transmitted.

1104
38.7.3 Master Mode Operations

When configured in Master Mode, the SPI operates on the clock generated by the internal programmable baud rate generator. It fully controls the data transfers to and from the slave(s) connected to the SPI bus. The SPI drives the chip select line to the slave and the serial clock signal (SPCK). The SPI features two holding registers, the Transmit Data Register and the Receive Data Register, and a single Shift Register. The holding registers maintain the data flow at a constant rate. After enabling the SPI, a data transfer begins when the processor writes to the SPI_TDR (Transmit Data Register). The written data is immediately transferred in the Shift Register and transfer on the SPI bus starts. While the data in the Shift Register is shifted on the MOSI line, the MISO line is sampled and shifted in the Shift Register. Receiving data cannot occur without transmitting data. If receiving mode is not needed, for example when communicating with a slave receiver only (such as an LCD), the receive status flags in the status register can be discarded. Before writing the TDR, the PCS field in the SPI_MR register must be set in order to select a slave. After enabling the SPI, a data transfer begins when the processor writes to the SPI_TDR (Transmit Data Register). The written data is immediately transferred in the Shift Register and transfer on the SPI bus starts. While the data in the Shift Register is shifted on the MOSI line, the MISO line is sampled and shifted in the Shift Register. Transmission cannot occur without reception. Before writing the TDR, the PCS field must be set in order to select a slave. If new data is written in SPI_TDR during the transfer, it stays in it until the current transfer is completed. Then, the received data is transferred from the Shift Register to SPI_RDR, the data in SPI_TDR is loaded in the Shift Register and a new transfer starts. The transfer of a data written in SPI_TDR in the Shift Register is indicated by the TDRE bit (Transmit Data Register Empty) in the Status Register (SPI_SR). When new data is written in SPI_TDR, this bit is cleared. The TDRE bit is used to trigger the Transmit DMAchannel. The end of transfer is indicated by the TXEMPTY flag in the SPI_SR register. If a transfer delay (DLYBCT) is greater than 0 for the last transfer, TXEMPTY is set after the completion of said delay. The master clock (MCK) can be switched off at this time. The transfer of received data from the Shift Register in SPI_RDR is indicated by the RDRF bit (Receive Data Register Full) in the Status Register (SPI_SR). When the received data is read, the RDRF bit is cleared. If the SPI_RDR (Receive Data Register) has not been read before new data is received, the Overrun Error bit (OVRES) in SPI_SR is set. As long as this flag is set, data is loaded in SPI_RDR. The user has to read the status register to clear the OVRES bit. Figure 38-5, shows a block diagram of the SPI when operating in Master Mode. Figure 38-6 on page 1107 shows a flow chart describing how transfers are handled.

1105
38.7.3.1 Master Mode Block Diagram

Figure 38-5. Master Mode Block Diagram
SPI_CSR0..3 SCBR Baud Rate Generator
MCK
SPCK
SPI Clock SPI_CSR0..3 BITS NCPHA CPOL MISO LSB
SPI_RDR RD
RDRF OVRES
Shift Register
MSB
MOSI
SPI_TDR TD SPI_CSR0..3 SPI_RDR CSAAT PS SPI_MR PCS 0 SPI_TDR PCS 1 NPCS0 PCSDEC Current Peripheral PCS NPCS3 NPCS2 NPCS1 TDRE
MSTR NPCS0 MODFDIS
MODF

1106
38.7.3.2 Master Mode Flow Diagram

Figure 38-6. Master Mode Flow Diagram
SPI Enable - NPCS defines the current Chip Select - CSAAT, DLYBS, DLYBCT refer to the fields of the Chip Select Register corresponding to the Current Chip Select - When NPCS is 0xF, CSAAT is 0. 1 TDRE ?
0 1 CSAAT ? PS ? Variable peripheral yes 0 Fixed peripheral
0 0 PS ? Variable peripheral NPCS = SPI_MR(PCS) Fixed peripheral
SPI_TDR(PCS) = NPCS ? no NPCS = 0xF
SPI_MR(PCS) = NPCS ? no NPCS = 0xF
NPCS = SPI_TDR(PCS)
Delay DLYBCS
Delay DLYBCS
NPCS = SPI_TDR(PCS)
NPCS = SPI_MR(PCS), SPI_TDR(PCS)
Delay DLYBS
Serializer = SPI_TDR(TD) TDRE = 1
Data Transfer
SPI_RDR(RD) = Serializer RDRF = 1
Delay DLYBCT
0 TDRE ?
1 CSAAT ?
0 NPCS = 0xF
Delay DLYBCS

1107
Figure 38-7 shows Transmit Data Register Empty (TDRE), Receive Data Register (RDRF) and Transmission Register Empty (TXEMPTY) status flags behavior within the SPI_SR (Status Register) during an 8-bit data transfer in fixed mode and no Peripheral Data Controller involved.
Figure 38-7. Status Register Flags Behavior
1 SPCK NPCS0 MOSI (from master) TDRE RDR read Write in SPI_TDR RDRF MISO (from slave) TXEMPTY MSB 6 5 4 3 2 1 LSB 2 3 4 5 6 7 8
MSB
LSB
shift register empty
38.7.3.3 Clock Generation The SPI Baud rate clock is generated by dividing the Master Clock (MCK), by a value between 1 and 255. This allows a maximum operating baud rate at up to Master Clock and a minimum operating baud rate of MCK divided by 255. Programming the SCBR field at 0 is forbidden. Triggering a transfer while SCBR is at 0 can lead to unpredictable results. At reset, SCBR is 0 and the user has to program it at a valid value before performing the first transfer. The divisor can be defined independently for each chip select, as it has to be programmed in the SCBR field of the Chip Select Registers. This allows the SPI to automatically adapt the baud rate for each interfaced peripheral without reprogramming. 38.7.3.4 Transfer Delays Figure 38-8 shows a chip select transfer change and consecutive transfers on the same chip select. Three delays can be programmed to modify the transfer waveforms:
z z z
The delay between chip selects, programmable only once for all the chip selects by writing the DLYBCS field in the Mode Register. Allows insertion of a delay between release of one chip select and before assertion of a new one. The delay before SPCK, independently programmable for each chip select by writing the field DLYBS. Allows the start of SPCK to be delayed after the chip select has been asserted. The delay between consecutive transfers, independently programmable for each chip select by writing the DLYBCT field. Allows insertion of a delay between two transfers occurring on the same chip select
These delays allow the SPI to be adapted to the interfaced peripherals and their speed and bus release time.

1108
Figure 38-8. Programmable Delays
Chip Select 1
Chip Select 2
SPCK DLYBCS DLYBS DLYBCT DLYBCT
38.7.3.5 Peripheral Selection The serial peripherals are selected through the assertion of the NPCS0 to NPCS3 signals. By default, all the NPCS signals are high before and after each transfer.
z
Fixed Peripheral Select: SPI exchanges data with only one peripheral
Fixed Peripheral Select is activated by writing the PS bit to zero in SPI_MR (Mode Register). In this case, the current peripheral is defined by the PCS field in SPI_MR and the PCS field in the SPI_TDR has no effect.
z
Variable Peripheral Select: Data can be exchanged with more than one peripheral without having to reprogram the NPCS field in the SPI_MR register.
Variable Peripheral Select is activated by setting PS bit to one. The PCS field in SPI_TDR is used to select the current peripheral. This means that the peripheral selection can be defined for each new data. The value to write in the SPI_TDR register as the following format. [xxxxxxx(7-bit) + LASTXFER(1-bit)(1)+ xxxx(4-bit) + PCS (4-bit) + DATA (8 to 16-bit)] with PCS equals to the chip select to assert as defined in Section 38.8.4 (SPI Transmit Data Register) and LASTXFER bit at 0 or 1 depending on CSAAT bit. Note: 1. Optional. CSAAT, LASTXFER and CSNAAT bits are discussed in Section 38.7.3.8 Peripheral Deselection with DMAC . If LASTXFER is used, the command must be issued before writing the last character. Instead of LASTXFER, the user can use the SPIDIS command. After the end of the DMA transfer, wait for the TXEMPTY flag, then write SPIDIS into the SPI_CR register (this will not change the configuration register values); the NPCS will be deactivated after the last character transfer. Then, another DMA transfer can be started if the SPIEN was previously written in the SPI_CR register. 38.7.3.6 SPI Direct Access Memory Controller (DMAC) In both fixed and variable mode the Direct Memory Access Controller (DMAC) can be used to reduce processor overhead. The Fixed Peripheral Selection allows buffer transfers with a single peripheral. Using the DMAC is an optimal means, as the size of the data transfer between the memory and the SPI is either 8 bits or 16 bits. However, changing the peripheral selection requires the Mode Register to be reprogrammed. The Variable Peripheral Selection allows buffer transfers with multiple peripherals without reprogramming the Mode Register. Data written in SPI_TDR is 32 bits wide and defines the real data to be transmitted and the peripheral it is destined to. Using the DMAC in this mode requires 32-bit wide buffers, with the data in the LSBs and the PCS and LASTXFER fields in the MSBs, however the SPI still controls the number of bits (8 to16) to be transferred through MISO and MOSI lines with the chip select configuration registers. This is not the optimal means in term of memory size for the buffers, but it provides a very effective means to exchange data with several peripherals without any intervention of the processor.

1109
38.7.3.7 Peripheral Chip Select Decoding The user can program the SPI to operate with up to 15 peripherals by decoding the four Chip Select lines, NPCS0 to NPCS3 with 1 of up to 16 decoder/demultiplexer. This can be enabled by writing the PCSDEC bit at 1 in the Mode Register (SPI_MR). When operating without decoding, the SPI makes sure that in any case only one chip select line is activated, i.e., one NPCS line driven low at a time. If two bits are defined low in a PCS field, only the lowest numbered chip select is driven low. When operating with decoding, the SPI directly outputs the value defined by the PCS field on NPCS lines of either the Mode Register or the Transmit Data Register (depending on PS). As the SPI sets a default value of 0xF on the chip select lines (i.e. all chip select lines at 1) when not processing any transfer, only 15 peripherals can be decoded. The SPI has only four Chip Select Registers, not 15. As a result, when decoding is activated, each chip select defines the characteristics of up to four peripherals. As an example, SPI_CRS0 defines the characteristics of the externally decoded peripherals 0 to 3, corresponding to the PCS values 0x0 to 0x3. Thus, the user has to make sure to connect compatible peripherals on the decoded chip select lines 0 to 3, 4 to 7, 8 to 11 and 12 to 14. Figure 38-9 below shows such an implementation. If the CSAAT bit is used, with or without the DMAC, the Mode Fault detection for NPCS0 line must be disabled. This is not needed for all other chip select lines since Mode Fault Detection is only on NPCS0.
Figure 38-9. Chip Select Decoding Application Block Diagram: Single Master/Multiple Slave Implementation
SPCK MISO MOSI SPCK MISO MOSI Slave 0 SPI Master NSS NPCS0 NPCS1 NPCS2 NPCS3 SPCK MISO MOSI Slave 1 NSS SPCK MISO MOSI Slave 14 NSS
1-of-n Decoder/Demultiplexer
Peripheral Deselection without DMA During a transfer of more than one data on a Chip Select without the DMA, the SPI_TDR is loaded by the processor, the flag TDRE rises as soon as the content of the SPI_TDR is transferred into the internal shift register. When this flag is detected high, the SPI_TDR can be reloaded. If this reload by the processor occurs before the end of the current transfer and if the next transfer is performed on the same chip select as the current transfer, the Chip Select is not de-asserted between the two transfers. But depending on the application software handling the SPI status register flags (by interrupt or polling method) or servicing other interrupts or other tasks, the processor may not reload the SPI_TDR in time to keep the chip select active (low). A null Delay Between Consecutive Transfer (DLYBCT) value in the SPI_CSR register, will give even less time for the processor to reload the SPI_TDR. With some SPI slave peripherals, requiring the chip select line to remain active (low) during a full set of transfers might lead to communication errors.

1110
To facilitate interfacing with such devices, the Chip Select Register [CSR0...CSR3] can be programmed with the CSAAT bit (Chip Select Active After Transfer) at 1. This allows the chip select lines to remain in their current state (low = active) until transfer to another chip select is required. Even if the SPI_TDR is not reloaded the chip select will remain active. To have the chip select line to raise at the end of the transfer the Last transfer Bit (LASTXFER) in the SPI_MR register must be set at 1 before writing the last data to transmit into the SPI_TDR. 38.7.3.8 Peripheral Deselection with DMAC When the Direct Memory Access Controller is used, the chip select line will remain low during the whole transfer since the TDRE flag is managed by the DMAC itself. The reloading of the SPI_TDR by the DMAC is done as soon as TDRE flag is set to one. In this case the use of CSAAT bit might not be needed. However, it may happen that when other DMAC channels connected to other peripherals are in use as well, the SPI DMAC might be delayed by another (DMAC with a higher priority on the bus). Having DMAC buffers in slower memories like flash memory or SDRAM compared to fast internal SRAM, may lengthen the reload time of the SPI_TDR by the DMAC as well. This means that the SPI_TDR might not be reloaded in time to keep the chip select line low. In this case the chip select line may toggle between data transfer and according to some SPI Slave devices, the communication might get lost. The use of the CSAAT bit might be needed. If chip select must be deasserted between each data when communicating with a slave device such as an SPI ADC for example, one can workaround this by using another ghost chip select line. In this case, the variable mode must be used. In this mode, the assertion of the corresponding chip select (CS) depends on the PCS field within SPI_TDR.
z z
Thereafter, an example of such implementation: NPCS0(1): Actual SPI Slave Device NPCS1(1): Ghost SPI Slave Device
Transmit Buffer must be initialized in the following way according to the SPI_TDR register (See Section 38.8.4 SPI Transmit Data Register on page 1120) to send data equal to 0xAB with NPCS0 and to send data equal to 0xCD on NPCS1. [BUFFER_ADDR(2)] = 0x0000 00AB <------ Transfer on NPCS0 [BUFFER_ADDR + 0x4] = 0x0001 000CD <------ Dummy transfer on NPCS1 [BUFFER_ADDR + 0x8] = 0x0000 000AB <------ Transfer on NPCS0 [BUFFER_ADDR + 0xC] = 0x0001 000CD <------ Dummy transfer on NPCS1 Please note that in this case the size of the buffer will be twice the actual size needed. Notes: 1. 2. Can be any other chip select line. Transmit Buffer address.
When the CSAAT bit is set at 0, the NPCS does not rise in all cases between two transfers on the same peripheral. During a transfer on a Chip Select, the flag TDRE rises as soon as the content of the SPI_TDR is transferred into the internal shifter. When this flag is detected the SPI_TDR can be reloaded. If this reload occurs before the end of the current transfer and if the next transfer is performed on the same chip select as the current transfer, the Chip Select is not de-asserted between the two transfers. This might lead to difficulties for interfacing with some serial peripherals requiring the chip select to be de-asserted after each transfer. To facilitate interfacing with such devices, the Chip Select Register can be programmed with the CSNAAT bit (Chip Select Not Active After Transfer) at 1. This allows to de-assert systematically the chip select lines during a time DLYBCS. (The value of the CSNAAT bit is taken into account only if the CSAAT bit is set at 0 for the same Chip Select). Figure 38-10 shows different peripheral deselection cases and the effect of the CSAAT and CSNAAT bits.

1111
Figure 38-10. Peripheral Deselection

CSAAT = 0 and CSNAAT = 0 CSAAT = 1 and CSNAAT= 0 / 1
TDRE
DLYBCT A DLYBCS PCS = A A A
DLYBCT A DLYBCS PCS = A A
NPCS[0..3]
Write SPI_TDR
TDRE
DLYBCT A DLYBCS PCS=A A A
DLYBCT A DLYBCS PCS = A A
NPCS[0..3]
Write SPI_TDR
TDRE NPCS[0..3]
DLYBCT A DLYBCS PCS = B B A
DLYBCT B DLYBCS PCS = B
Write SPI_TDR
CSAAT = 0 and CSNAAT = 0 DLYBCT TDRE
CSAAT = 0 and CSNAAT = 1 DLYBCT
NPCS[0..3]
A DLYBCS
PCS = A Write SPI_TDR
PCS = A

1112
38.7.3.9 Mode Fault Detection A mode fault is detected when the SPI is programmed in Master Mode and a low level is driven by an external master on the NPCS0/NSS signal. In this case, multi-master configuration, NPCS0, MOSI, MISO and SPCK pins must be configured in open drain (through the PIO controller). When a mode fault is detected, the MODF bit in the SPI_SR is set until the SPI_SR is read and the SPI is automatically disabled until re-enabled by writing the SPIEN bit in the SPI_CR (Control Register) at 1. By default, the Mode Fault detection circuitry is enabled. The user can disable Mode Fault detection by setting the MODFDIS bit in the SPI Mode Register (SPI_MR).
38.7.4 SPI Slave Mode

When operating in Slave Mode, the SPI processes data bits on the clock provided on the SPI clock pin (SPCK). The SPI waits for NSS to go active before receiving the serial clock from an external master. When NSS falls, the clock is validated on the serializer, which processes the number of bits defined by the BITS field of the Chip Select Register 0 (SPI_CSR0). These bits are processed following a phase and a polarity defined respectively by the NCPHA and CPOL bits of the SPI_CSR0. Note that BITS, CPOL and NCPHA of the other Chip Select Registers have no effect when the SPI is programmed in Slave Mode. The bits are shifted out on the MISO line and sampled on the MOSI line. (For more information on BITS field, see also, the Register on page 1125.)
(Note:)
below the register table; Section 38.8.9 SPI Chip Select
When all the bits are processed, the received data is transferred in the Receive Data Register and the RDRF bit rises. If the SPI_RDR (Receive Data Register) has not been read before new data is received, the Overrun Error bit (OVRES) in SPI_SR is set. As long as this flag is set, data is loaded in SPI_RDR. The user has to read the status register to clear the OVRES bit. When a transfer starts, the data shifted out is the data present in the Shift Register. If no data has been written in the Transmit Data Register (SPI_TDR), the last data received is transferred. If no data has been received since the last reset, all bits are transmitted low, as the Shift Register resets at 0. When a first data is written in SPI_TDR, it is transferred immediately in the Shift Register and the TDRE bit rises. If new data is written, it remains in SPI_TDR until a transfer occurs, i.e. NSS falls and there is a valid clock on the SPCK pin. When the transfer occurs, the last data written in SPI_TDR is transferred in the Shift Register and the TDRE bit rises. This enables frequent updates of critical variables with single transfers. Then, a new data is loaded in the Shift Register from the Transmit Data Register. In case no character is ready to be transmitted, i.e. no character has been written in SPI_TDR since the last load from SPI_TDR to the Shift Register, the Shift Register is not modified and the last received character is retransmitted. In this case the Underrun Error Status Flag (UNDES) is set in the SPI_SR. Figure 38-11 shows a block diagram of the SPI when operating in Slave Mode.

1113
Figure 38-11. Slave Mode Functional Bloc Diagram

SPCK NSS SPIEN SPIENS SPIDIS SPI_CSR0 BITS NCPHA CPOL MOSI LSB SPI_RDR RD RDRF OVRES SPI Clock
Shift Register
MSB
MISO
SPI_TDR TD TDRE

To prevent any single software error that may corrupt SPI behavior, the registers listed below can be write-protected by setting the WPEN bit in the SPI Write Protection Mode Register (SPI_WPMR). If a write access in a write-protected register is detected, then the WPVS flag in the SPI Write Protection Status Register (SPI_WPSR) is set and the field WPVSRC indicates in which register the write access has been attempted. The WPVS flag is automatically reset after reading the SPI Write Protection Status Register (SPI_WPSR). List of the write-protected registers: Section 38.8.2 SPI Mode Register Section 38.8.9 SPI Chip Select Register

1114
38.8
Serial Peripheral Interface (SPI) User Interface
Table 38-4. Register Mapping Offset 0x00 0x04 0x08 0x0C 0x10 0x14 0x18 0x1C 0x20 - 0x2C 0x30 0x34 0x38 0x3C 0x4C - 0xE0 0xE4 0xE8 0x00E8 - 0x00F8 0x00FC 0x100 - 0x124 Control Register Mode Register Receive Data Register Transmit Data Register Status Register Interrupt Enable Register Interrupt Disable Register Interrupt Mask Register Reserved Chip Select Register 0 Chip Select Register 1 Chip Select Register 2 Chip Select Register 3 Reserved Write Protection Control Register Write Protection Status Register Reserved Reserved Reserved for PDC Registers SPI_CSR0 SPI_CSR1 SPI_CSR2 SPI_CSR3 SPI_WPMR SPI_WPSR Read-write Read-write Read-write Read-write Read-write Read-only 0x0 0x0 0x0 0x0 0x0 0x0 Register Name SPI_CR SPI_MR SPI_RDR SPI_TDR SPI_SR SPI_IER SPI_IDR SPI_IMR Access Write-only Read-write Read-only Write-only Read-only Write-only Write-only Read-only Reset --0x0 0x0 --0x000000F0 ----0x0

1115
38.8.1 SPI Control Register

31
SPI_CR 0xF0004000 (0), 0xF8008000 (1) Write-only

30 29 28 27 26 25 24
23
22
21
20
19
18
17
LASTXFER
16
15
14
13
12
11
10
SWRST
SPIDIS
SPIEN
SPIEN: SPI Enable

0 = No effect. 1 = Enables the SPI to transfer and receive data.
SPIDIS: SPI Disable

0 = No effect. 1 = Disables the SPI. As soon as SPIDIS is set, SPI finishes its transfer. All pins are set in input mode and no data is received or transmitted. If a transfer is in progress, the transfer is finished before the SPI is disabled. If both SPIEN and SPIDIS are equal to one when the control register is written, the SPI is disabled.
SWRST: SPI Software Reset

0 = No effect. 1 = Reset the SPI. A software-triggered hardware reset of the SPI interface is performed. The SPI is in slave mode after software reset.
LASTXFER: Last Transfer

0 = No effect. 1 = The current NPCS will be deasserted after the character written in TD has been transferred. When CSAAT is set, this allows to close the communication with the current serial peripheral by raising the corresponding NPCS line as soon as TD transfer has completed. Refer to Section 38.7.3.5 Peripheral Selection for more details.

1116
38.8.2 SPI Mode Register

31
SPI_MR 0xF0004004 (0), 0xF8008004 (1) Read-write

30 29 28 27 26 25 24
DLYBCS
23 22 21 20 19 18 17 16
15
14
13
12 11 10
PCS
9 8
LLB
WDRBT
MODFDIS
PCSDEC
PS
MSTR
This register can only be written if the WPEN bit is cleared in SPI Write Protection Mode Register.
MSTR: Master/Slave Mode

0 = SPI is in Slave mode. 1 = SPI is in Master mode.
PS: Peripheral Select

0 = Fixed Peripheral Select. 1 = Variable Peripheral Select.
PCSDEC: Chip Select Decode

0 = The chip selects are directly connected to a peripheral device. 1 = The four chip select lines are connected to a 4- to 16-bit decoder. When PCSDEC equals one, up to 15 Chip Select signals can be generated with the four lines using an external 4- to 16-bit decoder. The Chip Select Registers define the characteristics of the 15 chip selects according to the following rules: SPI_CSR0 defines peripheral chip select signals 0 to 3. SPI_CSR1 defines peripheral chip select signals 4 to 7. SPI_CSR2 defines peripheral chip select signals 8 to 11. SPI_CSR3 defines peripheral chip select signals 12 to 14.
MODFDIS: Mode Fault Detection

0 = Mode fault detection is enabled. 1 = Mode fault detection is disabled.
WDRBT: Wait Data Read Before Transfer

0 = No Effect. In master mode, a transfer can be initiated whatever the state of the Receive Data Register is. 1 = In Master Mode, a transfer can start only if the Receive Data Register is empty, i.e. does not contain any unread data. This mode prevents overrun error in reception.
LLB: Local Loopback Enable

0 = Local loopback path disabled. 1 = Local loopback path enabled LLB controls the local loopback on the data serializer for testing in Master Mode only. (MISO is internally connected on MOSI.)

1117
PCS: Peripheral Chip Select

This field is only used if Fixed Peripheral Select is active (PS = 0). If PCSDEC = 0: PCS = xxx0 PCS = xx01 PCS = x011 PCS = 0111 PCS = 1111 (x = dont care) If PCSDEC = 1: NPCS[3:0] output signals = PCS. NPCS[3:0] = 1110 NPCS[3:0] = 1101 NPCS[3:0] = 1011 NPCS[3:0] = 0111 forbidden (no peripheral is selected)
DLYBCS: Delay Between Chip Selects

This field defines the delay from NPCS inactive to the activation of another NPCS. The DLYBCS time guarantees non-overlapping chip selects and solves bus contentions in case of peripherals having long data float times. If DLYBCS is less than or equal to six, six MCK periods will be inserted by default. Otherwise, the following equation determines the delay:
DLYBCS Delay Between Chip Selects = ---------------------MCK

1118
38.8.3 SPI Receive Data Register

31
SPI_RDR 0xF0004008 (0), 0xF8008008 (1) Read-only

30 29 28 27 26 25 24
23
22
21
20
19
18
17
16
15
14
13
12 11 10
PCS
9 8
RD
7 6 5 4 3 2 1 0
RD
RD: Receive Data

Data received by the SPI Interface is stored in this register right-justified. Unused bits read zero.

In Master Mode only, these bits indicate the value on the NPCS pins at the end of a transfer. Otherwise, these bits read zero. Note: When using variable peripheral select mode (PS = 1 in SPI_MR) it is mandatory to also set the WDRBT field to 1 if the SPI_RDR PCS field is to be processed.

1119
38.8.4 SPI Transmit Data Register

31
SPI_TDR 0xF000400C (0), 0xF800800C (1) Write-only

30 29 28 27 26 25 24
23
22
21
20
19
18
17
LASTXFER
16
15
14
13
12 11 10
PCS
9 8
TD
7 6 5 4 3 2 1 0
TD
TD: Transmit Data

Data to be transmitted by the SPI Interface is stored in this register. Information to be transmitted must be written to the transmit data register in a right-justified format.

This field is only used if Variable Peripheral Select is active (PS = 1). If PCSDEC = 0: PCS = xxx0 PCS = xx01 PCS = x011 PCS = 0111 PCS = 1111 (x = dont care) If PCSDEC = 1: NPCS[3:0] output signals = PCS NPCS[3:0] = 1110 NPCS[3:0] = 1101 NPCS[3:0] = 1011 NPCS[3:0] = 0111 forbidden (no peripheral is selected)
LASTXFER: Last Transfer

0 = No effect. 1 = The current NPCS will be deasserted after the character written in TD has been transferred. When CSAAT is set, this allows to close the communication with the current serial peripheral by raising the corresponding NPCS line as soon as TD transfer has completed. This field is only used if Variable Peripheral Select is active (PS = 1).

1120
38.8.5 SPI Status Register

31
SPI_SR 0xF0004010 (0), 0xF8008010 (1) Read-only

30 29 28 27 26 25 24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
SPIENS
8
UNDES
2
TXEMPTY
1
NSSR
0
OVRES
MODF
TDRE
RDRF
RDRF: Receive Data Register Full 0 = No data has been received since the last read of SPI_RDR
1 = Data has been received and the received data has been transferred from the serializer to SPI_RDR since the last read of SPI_RDR.
TDRE: Transmit Data Register Empty 0 = Data has been written to SPI_TDR and not yet transferred to the serializer.
1 = The last data written in the Transmit Data Register has been transferred to the serializer. TDRE equals zero when the SPI is disabled or at reset. The SPI enable command sets this bit to one.
MODF: Mode Fault Error 0 = No Mode Fault has been detected since the last read of SPI_SR.
1 = A Mode Fault occurred since the last read of the SPI_SR.
OVRES: Overrun Error Status 0 = No overrun has been detected since the last read of SPI_SR.
1 = An overrun has occurred since the last read of SPI_SR. An overrun occurs when SPI_RDR is loaded at least twice from the serializer since the last read of the SPI_RDR.
NSSR: NSS Rising 0 = No rising edge detected on NSS pin since last read.
1 = A rising edge occurred on NSS pin since last read.
TXEMPTY: Transmission Registers Empty 0 = As soon as data is written in SPI_TDR.

1 = SPI_TDR and internal shifter are empty. If a transfer delay has been defined, TXEMPTY is set after the completion of such delay.
UNDES: Underrun Error Status (Slave Mode Only) 0 = No underrun has been detected since the last read of SPI_SR.
1 = A transfer begins whereas no data has been loaded in the Transmit Data Register.
SPIENS: SPI Enable Status 0 = SPI is disabled.

1 = SPI is enabled.

1121
38.8.6 SPI Interrupt Enable Register

31
SPI_IER 0xF0004014 (0), 0xF8008014 (1) Write-only

30 29 28 27 26 25 24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
UNDES
2
TXEMPTY
1
NSSR
0
OVRES
MODF
TDRE
RDRF
RDRF: Receive Data Register Full Interrupt Enable TDRE: SPI Transmit Data Register Empty Interrupt Enable MODF: Mode Fault Error Interrupt Enable OVRES: Overrun Error Interrupt Enable NSSR: NSS Rising Interrupt Enable TXEMPTY: Transmission Registers Empty Enable UNDES: Underrun Error Interrupt Enable

1122
38.8.7 SPI Interrupt Disable Register

31
SPI_IDR 0xF0004018 (0), 0xF8008018 (1) Write-only

30 29 28 27 26 25 24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
UNDES
2
TXEMPTY
1
NSSR
0
OVRES
MODF
TDRE
RDRF
RDRF: Receive Data Register Full Interrupt Disable TDRE: SPI Transmit Data Register Empty Interrupt Disable MODF: Mode Fault Error Interrupt Disable OVRES: Overrun Error Interrupt Disable NSSR: NSS Rising Interrupt Disable TXEMPTY: Transmission Registers Empty Disable UNDES: Underrun Error Interrupt Disable

1123
38.8.8 SPI Interrupt Mask Register

31
SPI_IMR 0xF000401C (0), 0xF800801C (1) Read-only

30 29 28 27 26 25 24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
UNDES
2
TXEMPTY
1
NSSR
0
OVRES
MODF
TDRE
RDRF
RDRF: Receive Data Register Full Interrupt Mask TDRE: SPI Transmit Data Register Empty Interrupt Mask MODF: Mode Fault Error Interrupt Mask OVRES: Overrun Error Interrupt Mask NSSR: NSS Rising Interrupt Mask TXEMPTY: Transmission Registers Empty Mask UNDES: Underrun Error Interrupt Mask
0 = The corresponding interrupt is not enabled. 1 = The corresponding interrupt is enabled.

1124
38.8.9 SPI Chip Select Register

31
SPI_CSRx[x=0..3] 0xF0004030 (0), 0xF8008030 (1) Read/Write

30 29 28 27 26 25 24
DLYBCT
23 22 21 20 19 18 17 16
DLYBS
15 14 13 12 11 10 9 8
SCBR
7 6 5 4 3 2 1 0
BITS
CSAAT
CSNAAT
NCPHA
CPOL
This register can only be written if the WPEN bit is cleared in SPI Write Protection Mode Register. Note: SPI_CSRx registers must be written even if the user wants to use the defaults. The BITS field will not be updated with the translated value unless the register is written.
CPOL: Clock Polarity

0 = The inactive state value of SPCK is logic level zero. 1 = The inactive state value of SPCK is logic level one. CPOL is used to determine the inactive state value of the serial clock (SPCK). It is used with NCPHA to produce the required clock/data relationship between master and slave devices.
NCPHA: Clock Phase

0 = Data is changed on the leading edge of SPCK and captured on the following edge of SPCK. 1 = Data is captured on the leading edge of SPCK and changed on the following edge of SPCK. NCPHA determines which edge of SPCK causes data to change and which edge causes data to be captured. NCPHA is used with CPOL to produce the required clock/data relationship between master and slave devices.
CSNAAT: Chip Select Not Active After Transfer (Ignored if CSAAT = 1)

0 = The Peripheral Chip Select does not rise between two transfers if the SPI_TDR is reloaded before the end of the first transfer and if the two transfers occur on the same Chip Select. 1 = The Peripheral Chip Select rises systematically after each transfer performed on the same slave. It remains active after the end of transfer for a minimal duration of:
DLYBCT ----------------------- (if DLYBCT field is different from 0) MCK
DLYBCT + 1 - (if DLYBCT field equals 0) -------------------------------MCK

z
CSAAT: Chip Select Active After Transfer

0 = The Peripheral Chip Select Line rises as soon as the last transfer is achieved. 1 = The Peripheral Chip Select does not rise after the last transfer is achieved. It remains active until a new transfer is requested on a different chip select.

1125
BITS: Bits Per Transfer

(See the (Note:) below the register table; Section 38.8.9 SPI Chip Select Register on page 1125.) The BITS field determines the number of data bits transferred. Reserved values should not be used.
Value 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Name 8_BIT 9_BIT 10_BIT 11_BIT 12_BIT 13_BIT 14_BIT 15_BIT 16_BIT Description 8 bits for transfer 9 bits for transfer 10 bits for transfer 11 bits for transfer 12 bits for transfer 13 bits for transfer 14 bits for transfer 15 bits for transfer 16 bits for transfer Reserved Reserved Reserved Reserved Reserved Reserved Reserved
SCBR: Serial Clock Baud Rate

In Master Mode, the SPI Interface uses a modulus counter to derive the SPCK baud rate from the Master Clock MCK. The Baud rate is selected by writing a value from 1 to 255 in the SCBR field. The following equations determine the SPCK baud rate:
MCK SPCK Baudrate = -------------SCBR

Programming the SCBR field at 0 is forbidden. Triggering a transfer while SCBR is at 0 can lead to unpredictable results. At reset, SCBR is 0 and the user has to program it at a valid value before performing the first transfer. Note: If one of the SCBR fields inSPI_CSRx is set to 1, the other SCBR fields in SPI_CSRx must be set to 1 as well, if they are required to process transfers. If they are not used to transfer data, they can be set at any value.
DLYBS: Delay Before SPCK

This field defines the delay from NPCS valid to the first valid SPCK transition. When DLYBS equals zero, the NPCS valid to SPCK transition is 1/2 the SPCK clock period. Otherwise, the following equations determine the delay:
DLYBS Delay Before SPCK = -----------------MCK DLYBCT: Delay Between Consecutive Transfers
This field defines the delay between two consecutive transfers with the same peripheral without removing the chip select. The delay is always inserted after each transfer and before removing the chip select if needed. When DLYBCT equals zero, no delay between consecutive transfers is inserted and the clock keeps its duty cycle over the character transfers. Otherwise, the following equation determines the delay:
DLYBCT Delay Between Consecutive Transfers = 32 -----------------------------------MCK

1126
38.8.10 SPI Write Protection Mode Register

31
SPI_WPMR 0xF00040E4 (0), 0xF80080E4 (1) Read-write

30 29 28 27 26 25 24
WPKEY
23 22 21 20 19 18 17 16
WPKEY
15 14 13 12 11 10 9 8
WPKEY
7 6 5 4 3 2 1 0
WPEN

0: The Write Protection is Disabled 1: The Write Protection is Enabled
WPKEY: Write Protection Key Password

If a value is written in WPEN, the value is taken into account only if WPKEY is written with SPI (SPI written in ASCII Code, ie 0x535049 in hexadecimal). List of the write-protected registers:
Section 38.8.2 SPI Mode Register Section 38.8.9 SPI Chip Select Register

1127
38.8.11 SPI Write Protection Status Register

31 23 15
SPI_WPSR 0xF00040E8 (0), 0xF80080E8 (1) Read-only

30 22 14 29 21 13 28 20 12 27 19 11 26 18 10 25 17 9 24 16 8
WPVSRC
7 6 5 4 3 2 1 0
WPVS
WPVS: Write Protection Violation Status

0: No Write Protect Violation has occurred since the last read of the SPI_WPSR register. 1: A Write Protect Violation has occurred since the last read of the SPI_WPSR register. If this violation is an unauthorized attempt to write a protected register, the associated violation is reported into field WPVSRC.

This Field indicates the APB Offset of the register concerned by the violation (SPI_MR or SPI_CSRx)

1128
39.
39.1
Periodic Interval Timer (PIT)

Description
The Periodic Interval Timer (PIT) provides the operating systems scheduler interrupt. It is designed to offer maximum accuracy and efficient management, even for systems with long response time.
39.2
z z z
20-bit Programmable Counter plus 12-bit Interval Counter Reset-on-read Feature Both Counters Work on Master Clock/16

1129
39.3
Block Diagram
Figure 39-1. Periodic Interval Timer

PIT_MR
PIV
PIT_MR
PITIEN
set
PIT_SR
PITS
reset
pit_irq
12-bit Adder
read PIT_PIVR
MCK
20-bit Counter
Prescaler
MCK/16
CPIV
PIT_PIVR
PICNT
CPIV
PIT_PIIR
PICNT

1130
39.4
The Periodic Interval Timer aims at providing periodic interrupts for use by operating systems. The PIT provides a programmable overflow counter and a reset-on-read feature. It is built around two counters: a 20-bit CPIV counter and a 12-bit PICNT counter. Both counters work at Master Clock /16. The first 20-bit CPIV counter increments from 0 up to a programmable overflow value set in the field PIV of the Mode Register (PIT_MR). When the counter CPIV reaches this value, it resets to 0 and increments the Periodic Interval Counter, PICNT. The status bit PITS in the Status Register (PIT_SR) rises and triggers an interrupt, provided the interrupt is enabled (PITIEN in PIT_MR). Writing a new PIV value in PIT_MR does not reset/restart the counters. When CPIV and PICNT values are obtained by reading the Periodic Interval Value Register (PIT_PIVR), the overflow counter (PICNT) is reset and the PITS is cleared, thus acknowledging the interrupt. The value of PICNT gives the number of periodic intervals elapsed since the last read of PIT_PIVR. When CPIV and PICNT values are obtained by reading the Periodic Interval Image Register (PIT_PIIR), there is no effect on the counters CPIV and PICNT, nor on the bit PITS. For example, a profiler can read PIT_PIIR without clearing any pending interrupt, whereas a timer interrupt clears the interrupt by reading PIT_PIVR. The PIT may be enabled/disabled using the PITEN bit in the PIT_MR register (disabled on reset). The PITEN bit only becomes effective when the CPIV value is 0. Figure 39-2 illustrates the PIT counting. After the PIT Enable bit is reset (PITEN= 0), the CPIV goes on counting until the PIV value is reached, and is then reset. PIT restarts counting, only if the PITEN is set again. The PIT is stopped when the core enters debug state.
Figure 39-2. Enabling/Disabling PIT with PITEN

APB cycle MCK 15 restarts MCK Prescaler MCK Prescaler 0 PITEN APB cycle
CPIV PICNT PITS (PIT_SR) APB Interface
1 0
PIV - 1
PIV 1
0 0
read PIT_PIVR

1131
39.5
Periodic Interval Timer (PIT) User Interface
Table 39-1. Register Mapping Offset 0x00 0x04 0x08 0x0C Register Mode Register Status Register Periodic Interval Value Register Periodic Interval Image Register Name PIT_MR PIT_SR PIT_PIVR PIT_PIIR Access Read-write Read-only Read-only Read-only Reset 0x000F_FFFF 0x0000_0000 0x0000_0000 0x0000_0000

1132
39.5.1 Periodic Interval Timer Mode Register

31 23 15
PIT_MR 0xFFFFFE30 Read/Write

30 22 14 29 21 13 28 20 12 PIV 7 6 5 4 PIV 3 2 1 0 27 19 26 18 PIV 11 10 9 8 25 PITIEN 17 24 PITEN 16
PIV: Periodic Interval Value

Defines the value compared with the primary 20-bit counter of the Periodic Interval Timer (CPIV). The period is equal to (PIV + 1).
PITEN: Period Interval Timer Enabled

0 = The Periodic Interval Timer is disabled when the PIV value is reached. 1 = The Periodic Interval Timer is enabled.
PITIEN: Periodic Interval Timer Interrupt Enable

0 = The bit PITS in PIT_SR has no effect on interrupt. 1 = The bit PITS in PIT_SR asserts interrupt.

1133
39.5.2 Periodic Interval Timer Status Register

31 23 15 7
PIT_SR 0xFFFFFE34 Read-only

30 22 14 6 29 21 13 5 28 20 12 4 27 19 11 3 26 18 10 2 25 17 9 1 24 16 8 0 PITS
PITS: Periodic Interval Timer Status

0 = The Periodic Interval timer has not reached PIV since the last read of PIT_PIVR. 1 = The Periodic Interval timer has reached PIV since the last read of PIT_PIVR.

1134
39.5.3 Periodic Interval Timer Value Register

31
PIT_PIVR 0xFFFFFE38 Read-only

30 29 28 PICNT 27 26 25 24
23
22 PICNT
21
20
19
18 CPIV
17
16
15
14
13
12 CPIV
11
10
4 CPIV
Reading this register clears PITS in PIT_SR.
CPIV: Current Periodic Interval Value

Returns the current value of the periodic interval timer.
PICNT: Periodic Interval Counter

Returns the number of occurrences of periodic intervals since the last read of PIT_PIVR.

1135
39.5.4 Periodic Interval Timer Image Register

31
PIT_PIIR 0xFFFFFE3C Read-only

30 29 28 PICNT 27 26 25 24
23
22 PICNT
21
20
19
18 CPIV
17
16
15
14
13
12 CPIV
11
10
4 CPIV
CPIV: Current Periodic Interval Value

Returns the current value of the periodic interval timer.
PICNT: Periodic Interval Counter

Returns the number of occurrences of periodic intervals since the last read of PIT_PIVR.

1136
40.
40.1
Two-wire Interface (TWI)

Description
The Atmel Two-wire Interface (TWI) interconnects components on a unique two-wire bus, made up of one clock line and one data line with speeds of up to 400 Kbits per second, based on a byte-oriented transfer format. It can be used with any Atmel Two-wire Interface bus Serial EEPROM and IC compatible device such as Real Time Clock (RTC), Dot Matrix/Graphic LCD Controllers and Temperature Sensor, to name but a few. The TWI is programmable as a master or a slave with sequential or single-byte access. Multiple master capability is supported. Arbitration of the bus is performed internally and puts the TWI in slave mode automatically if the bus arbitration is lost. A configurable baud rate generator permits the output data rate to be adapted to a wide range of core clock frequencies. Below, Table 40-1 lists the compatibility level of the Atmel Two-wire Interface in Master Mode and a full I2C compatible device.
Table 40-1. Atmel TWI compatibility with I2C Standard I2C Standard Standard Mode Speed (100 KHz) Fast Mode Speed (400 KHz) 7 or 10 bits Slave Addressing START BYTE(1) Repeated Start (Sr) Condition ACK and NACK Management Slope control and input filtering (Fast mode) Clock stretching Multi Master Capability Atmel TWI Supported Supported Supported Not Supported Supported Supported Not Supported Supported Supported
Note:
1.
START + b000000001 + Ack + Sr

1137
40.2
z z z z z z z z z
3 TWIs Compatible with Atmel Two-wire Interface Serial Memory and IC Compatible Devices(1) One, Two or Three Bytes for Slave Address Sequential Read-write Operations Master, Multi-master and Slave Mode Operation Bit Rate: Up to 400 Kbits General Call Supported in Slave mode SMBUS Quick Command Supported in Master Mode Connection to DMA Controller (DMA) Channel Capabilities Optimizes Data Transfers
1. See Table 40-1 for details on compatibility with IC Standard.
Note:
40.3
List of Abbreviations
Table 40-2. Abbreviations Abbreviation TWI A NA P S Sr SADR ADR R W Description Two-wire Interface Acknowledge Non Acknowledge Stop Start Repeated Start Slave Address Any address except SADR Read Write

1138
40.4
Block Diagram
APB Bridge
TWCK PIO Two-wire Interface TWD
PMC
MCK
TWI Interrupt
Interrupt Controller
40.5

VDD Rp TWD TWCK Rp
Host with TWI Interface
Atmel TWI Serial EEPROM Slave 1
IC RTC Slave 2
IC LCD Controller Slave 3
IC Temp. Sensor Slave 4
Rp: Pull up value as given by the IC Standard
40.5.1 I/O Lines Description

Table 40-3. I/O Lines Description Pin Name TWD TWCK Pin Description Two-wire Serial Data Two-wire Serial Clock Type Input/Output Input/Output

1139
40.6
40.6.1 I/O Lines

Both TWD and TWCK are bidirectional lines, connected to a positive supply voltage via a current source or pull-up resistor (see Figure 40-2 on page 1139). When the bus is free, both lines are high. The output stages of devices connected to the bus must have an open-drain or open-collector to perform the wired-AND function. TWD and TWCK pins may be multiplexed with PIO lines. To enable the TWI, the programmer must perform the following step:
z
Program the PIO controller to dedicate TWD and TWCK as peripheral lines.
The user must not program TWD and TWCK as open-drain. It is already done by the hardware.
Table 40-4. I/O Lines Instance TWI0 TWI0 TWI1 TWI1 TWI2 TWI2 Signal TWCK0 TWD0 TWCK1 TWD1 TWCK2 TWD2 I/O Line PA31 PA30 PC27 PC26 PA19 PA18 Peripheral A A B B B B

z
Enable the peripheral clock.
The TWI interface may be clocked through the Power Management Controller (PMC), thus the programmer must first configure the PMC to enable the TWI clock.
40.6.3 Interrupt
The TWI interface has an interrupt line connected to the Interrupt Controller. In order to handle interrupts, the Interrupt Controller must be programmed before configuring the TWI.
Table 40-5. Peripheral IDs Instance TWI0 TWI1 TWI2 ID 18 19 20

1140
40.7
40.7.1 Transfer Format

The data put on the TWD line must be 8 bits long. Data is transferred MSB first; each byte must be followed by an acknowledgement. The number of bytes per transfer is unlimited (see Figure 40-4). Each transfer begins with a START condition and terminates with a STOP condition (see Figure 40-3).
z z
A high-to-low transition on the TWD line while TWCK is high defines the START condition. A low-to-high transition on the TWD line while TWCK is high defines a STOP condition.
Figure 40-3. START and STOP Conditions
TWD
TWCK Start Stop
Figure 40-4. Transfer Format
TWD
TWCK
Start
Address
R/W
Ack
Data
Ack
Data
Ack
Stop
40.7.2 Modes of Operation

The TWI has different modes of operations:
z z z z z
Master transmitter mode Master receiver mode Multi-master transmitter mode Multi-master receiver mode Slave transmitter mode
z Slave receiver mode These modes are described in the following chapters.

1141
40.8
Master Mode
40.8.1 Definition
The Master is the device that starts a transfer, generates a clock and stops it.
40.8.2 Application Block Diagram

Figure 40-5. Master Mode Typical Application Block Diagram
VDD Rp TWD TWCK Rp
Atmel TWI Serial EEPROM Slave 1
IC RTC Slave 2
IC LCD Controller Slave 3
IC Temp. Sensor Slave 4
Rp: Pull up value as given by the IC Standard
40.8.3 Programming Master Mode

The following registers have to be programmed before entering Master mode: 1. 2. 3. 4. DADR (+ IADRSZ + IADR if a 10 bit device is addressed): The device address is used to access slave devices in read or write mode. CKDIV + CHDIV + CLDIV: Clock Waveform. SVDIS: Disable the slave mode. MSEN: Enable the master mode.
40.8.4 Master Transmitter Mode

After the master initiates a Start condition when writing into the Transmit Holding Register, TWI_THR, it sends a 7-bit slave address, configured in the Master Mode register (DADR in TWI_MMR), to notify the slave device. The bit following the slave address indicates the transfer direction, 0 in this case (MREAD = 0 in TWI_MMR). The TWI transfers require the slave to acknowledge each received byte. During the acknowledge clock pulse (9th pulse), the master releases the data line (HIGH), enabling the slave to pull it down in order to generate the acknowledge. The master polls the data line during this clock pulse and sets the Not Acknowledge bit (NACK) in the status register if the slave does not acknowledge the byte. As with the other status bits, an interrupt can be generated if enabled in the interrupt enable register (TWI_IER). If the slave acknowledges the byte, the data written in the TWI_THR, is then shifted in the internal shifter and transferred. When an acknowledge is detected, the TXRDY bit is set until a new write in the TWI_THR. While no new data is written in the TWI_THR, the Serial Clock Line is tied low. When new data is written in the TWI_THR, the SCL is released and the data is sent. To generate a STOP event, the STOP command must be performed by writing in the STOP field of TWI_CR. After a Master Write transfer, the Serial Clock line is stretched (tied low) while no new data is written in the TWI_THR or until a STOP command is performed. See Figure 40-6, Figure 40-7, and Figure 40-8.

1142
Figure 40-6. Master Write with One Data Byte

STOP Command sent (write in TWI_CR)
TWD
DADR
DATA
TXCOMP
TXRDY Write THR (DATA)
Figure 40-7. Master Write with Multiple Data Bytes

STOP command performed (by writing in the TWI_CR)
TWD
DADR
DATA n
DATA n+1
DATA n+2
TWCK
TXCOMP
TXRDY Write THR (Data n) Write THR (Data n+1) Write THR (Data n+2) Last data sent
Figure 40-8. Master Write with One Byte Internal Address and Multiple Data Bytes
STOP command performed (by writing in the TWI_CR)
TWD
DADR
IADR
DATA n
DATA n+1
DATA n+2
TWCK
TXCOMP
TXRDY Write THR (Data n) Write THR (Data n+1) Write THR (Data n+2) Last data sent

1143
40.8.5 Master Receiver Mode

The read sequence begins by setting the START bit. After the start condition has been sent, the master sends a 7-bit slave address to notify the slave device. The bit following the slave address indicates the transfer direction, 1 in this case (MREAD = 1 in TWI_MMR). During the acknowledge clock pulse (9th pulse), the master releases the data line (HIGH), enabling the slave to pull it down in order to generate the acknowledge. The master polls the data line during this clock pulse and sets the NACK bit in the status register if the slave does not acknowledge the byte. If an acknowledge is received, the master is then ready to receive data from the slave. After data has been received, the master sends an acknowledge condition to notify the slave that the data has been received except for the last data, after the stop condition. See Figure 40-9. When the RXRDY bit is set in the status register, a character has been received in the receive-holding register (TWI_RHR). The RXRDY bit is reset when reading the TWI_RHR. When a single data byte read is performed, with or without internal address (IADR), the START and STOP bits must be set at the same time. See Figure 40-9. When a multiple data byte read is performed, with or without internal address (IADR), the STOP bit must be set after the next-to-last data received. See Figure 40-10. For Internal Address usage see Section 40.8.6. If the receive holding register (TWI_RHR) is full (RXRDY high) and the master is receiving data, the Serial Clock Line will be tied low before receiving the last bit of the data and until the TWI_RHR register is read. Once the TWI_RHR register is read, the master will stop stretching the Serial Clock Line and end the data reception. See Figure 40-11. Warning: When receiving multiple bytes in master read mode, if the next-to-last access is not read (the RXRDY flag remains high), the last access will not be completed until TWI_RHR is read. The last access stops on the next-to-last bit (clock stretching). When the TWI_RHR register is read there is only half a bit period to send the stop bit command, else another read access might occur (spurious access). A possible workaround is to raise the STOP BIT command before reading the TWI_RHR on the next-to-last access (within IT handler).
Figure 40-9. Master Read with One Data Byte
TWD S DADR R A DATA N P
TXCOMP Write START & STOP Bit RXRDY Read RHR
Figure 40-10. Master Read with Multiple Data Bytes

TWD S DADR R A DATA n A DATA (n+1) A DATA (n+m)-1 A DATA (n+m)
N
TXCOMP Write START Bit RXRDY Read RHR DATA n Read RHR DATA (n+1) Read RHR DATA (n+m)-1 Read RHR DATA (n+m)
Write STOP Bit after next-to-last data read

1144
Figure 40-11. Master Read Clock Stretching with Multiple Data Bytes
STOP command performed (by writing in the TWI_CR) Clock Streching
TWD
DADR
DATA n
DATA n+1
DATA n+2
TWCK
TXCOMP
RXRDY
Read RHR (Data n)
Read RHR (Data n+1)
Read RHR (Data n+2)
40.8.6 Internal Address

The TWI interface can perform various transfer formats: Transfers with 7-bit slave address devices and 10-bit slave address devices. 40.8.6.1 7-bit Slave Addressing When Addressing 7-bit slave devices, the internal address bytes are used to perform random address (read or write) accesses to reach one or more data bytes, within a memory page location in a serial memory, for example. When performing read operations with an internal address, the TWI performs a write operation to set the internal address into the slave device, and then switch to Master Receiver mode. Note that the second start condition (after sending the IADR) is sometimes called repeated start (Sr) in I2C fully-compatible devices. See Figure 40-13. See Figure 40-12 and Figure 40-14 for Master Write operation with internal address. The three internal address bytes are configurable through the Master Mode register (TWI_MMR). If the slave device supports only a 7-bit address, i.e. no internal address, IADRSZ must be set to 0. In the figures below the following abbreviations are used:
z z z z z z z z z
S Sr P W R A N DADR IADR
Start Repeated Start Stop Write Read Acknowledge Not Acknowledge Device Address Internal Address

1145
Figure 40-12. Master Write with One, Two or Three Bytes Internal Address and One Data Byte
Three bytes internal address TWD
S DADR W A IADR(23:16) A IADR(15:8) A IADR(7:0) A DATA A P
Two bytes internal address TWD

S DADR W A IADR(15:8) A IADR(7:0) A DATA A P
One byte internal address TWD

S DADR W A IADR(7:0) A DATA A P
Figure 40-13. Master Read with One, Two or Three Bytes Internal Address and One Data Byte
Three bytes internal address TWD S DADR W A IADR(23:16) A IADR(15:8) A IADR(7:0) A Sr DADR R A
DATA Two bytes internal address TWD S DADR W A IADR(15:8) A IADR(7:0) A Sr DADR R A DATA
One byte internal address TWD S DADR W A IADR(7:0) A Sr DADR R A DATA N P
40.8.6.2 10-bit Slave Addressing For a slave address higher than 7 bits, the user must configure the address size (IADRSZ) and set the other slave address bits in the internal address register (TWI_IADR). The two remaining Internal address bytes, IADR[15:8] and IADR[23:16] can be used the same as in 7-bit Slave Addressing. Example: Address a 10-bit device (10-bit device address is b1 b2 b3 b4 b5 b6 b7 b8 b9 b10) 1. 2. 3. Program IADRSZ = 1, Program DADR with 1 1 1 1 0 b1 b2 (b1 is the MSB of the 10-bit address, b2, etc.) Program TWI_IADR with b3 b4 b5 b6 b7 b8 b9 b10 (b10 is the LSB of the 10-bit address)
Figure 40-14 below shows a byte write to an Atmel AT24LC512 EEPROM. This demonstrates the use of internal addresses to access the device.
Figure 40-14. Internal Address Usage
S T A R T W R I T E S T O P
Device Address 0 M S B
FIRST WORD ADDRESS
SECOND WORD ADDRESS
DATA
LR A S / C BW K
M S B
A C K
LA SC BK
A C K
40.8.7 Using the DMA Controller

The use of the DMA significantly reduces the CPU load. To assure correct implementation, respect the programming sequence that follows.

1146
40.8.7.1 Data Transmit with the DMA 1. 2. 3. 4. 5. 6. 7. 8. 9. Initialize the DMA (channels, memory pointers, size -1, etc.); Configure the master mode (DADR, CKDIV, etc.) or slave mode. Enable the DMA. Wait for the DMA BTC flag. Disable the DMA. Wait for the TXRDY flag in TWI_SR register Set the STOP command in TWI_CR. Write the last character in TWI_THR (Optional) Wait for the TXCOMP flag in TWI_SR register before disabling the peripheral clock if required
40.8.7.2 Data Receive with the DMA The DMA transfer size must be defined with the buffer size minus 2. The two remaining characters must be managed without DMA to ensure that the exact number of bytes are received whatever the system bus latency conditions encountered during the end of buffer transfer period. In slave mode, the number of characters to receive must be known in order to configure the DMA. 1. 2. 3. 4. 5. 6. 7. 8. 9. Initialize the DMA (channels, memory pointers, size -2, etc.); Configure the master mode (DADR, CKDIV, etc.) or slave mode. Enable the DMA. (Master Only) Write the START bit in the TWI_CR register to start the transfer Wait for the DMA BTC flag. Disable the DMA. Wait for the RXRDY flag in the TWI_SR register Set the STOP command in TWI_CR Read the penultimate character in TWI_RHR
10. Wait for the RXRDY flag in the TWI_SR register 11. Read the last character in TWI_RHR 12. (Optional) Wait for the TXCOMP flag in TWI_SR register before disabling the peripheral clock if required
40.8.8 SMBUS Quick Command (Master Mode Only)

The TWI interface can perform a Quick Command: 1. 2. 3. Configure the master mode (DADR, CKDIV, etc.). Write the MREAD bit in the TWI_MMR register at the value of the one-bit command to be sent. Start the transfer by setting the QUICK bit in the TWI_CR.
Figure 40-15. SMBUS Quick Command

TWD S DADR R/W A P
TXCOMP
TXRDY Write QUICK command in TWI_CR

1147
40.8.9 Read-write Flowcharts

The following flowcharts shown in Figure 40-17 on page 1149, Figure 40-18 on page 1150, Figure 40-19 on page 1151, Figure 40-20 on page 1152 and Figure 40-21 on page 1153 give examples for read and write operations. A polling or interrupt method can be used to check the status bits. The interrupt method requires that the interrupt enable register (TWI_IER) be configured first.
Figure 40-16. TWI Write Operation with Single Data Byte without Internal Address
BEGIN
Set TWI clock (CLDIV, CHDIV, CKDIV) in TWI_CWGR (Needed only once)
Set the Control register: - Master enable TWI_CR = MSEN + SVDIS
Set the Master Mode register: - Device slave address (DADR) - Transfer direction bit Write ==> bit MREAD = 0
Load Transmit register TWI_THR = Data to send
Write STOP Command TWI_CR = STOP
Read Status register
No TXRDY = 1? Yes Read Status register
No TXCOMP = 1? Yes Transfer finished

1148
Figure 40-17. TWI Write Operation with Single Data Byte and Internal Address
BEGIN
Set the Master Mode register: - Device slave address (DADR) - Internal address size (IADRSZ) - Transfer direction bit Write ==> bit MREAD = 0
Set the internal address TWI_IADR = address
Load transmit register TWI_THR = Data to send
Write STOP command TWI_CR = STOP
No TXRDY = 1? Yes Read Status register
TXCOMP = 1? No Yes Transfer finished

1149
Figure 40-18. TWI Write Operation with Multiple Data Bytes with or without Internal Address
BEGIN
Set the Master Mode register: - Device slave address - Internal address size (if IADR used) - Transfer direction bit Write ==> bit MREAD = 0
No Internal address size = 0? Set the internal address TWI_IADR = address
Yes
Load Transmit register TWI_THR = Data to send
TWI_THR = data to send TXRDY = 1? Yes Data to send? Yes
No
Write STOP Command TWI_CR = STOP
Read Status register Yes No TXCOMP = 1?
END

1150
Figure 40-19. TWI Read Operation with Single Data Byte without Internal Address
BEGIN
Set the Master Mode register: - Device slave address - Transfer direction bit Read ==> bit MREAD = 1
Start the transfer TWI_CR = START | STOP
Read status register
RXRDY = 1? Yes Read Receive Holding Register
No
No TXCOMP = 1? Yes END

1151
Figure 40-20. TWI Read Operation with Single Data Byte and Internal Address
BEGIN
Set the Master Mode register: - Device slave address - Internal address size (IADRSZ) - Transfer direction bit Read ==> bit MREAD = 1
Set the internal address TWI_IADR = address
Start the transfer TWI_CR = START | STOP
No RXRDY = 1? Yes Read Receive Holding register
No TXCOMP = 1? Yes END

1152
Figure 40-21. TWI Read Operation with Multiple Data Bytes with or without Internal Address
BEGIN
Set the Master Mode register: - Device slave address - Internal address size (if IADR used) - Transfer direction bit Read ==> bit MREAD = 1
Internal address size = 0? Set the internal address TWI_IADR = address
Yes Start the transfer TWI_CR = START
RXRDY = 1? Yes Read Receive Holding register (TWI_RHR)
No
No
Last data to read but one? Yes Stop the transfer TWI_CR = STOP
No RXRDY = 1? Yes Read Receive Holding register (TWI_RHR)
Read status register
TXCOMP = 1? Yes END
No

1153
40.9
Multi-master Mode
40.9.1 Definition
More than one master may handle the bus at the same time without data corruption by using arbitration. Arbitration starts as soon as two or more masters place information on the bus at the same time, and stops (arbitration is lost) for the master that intends to send a logical one while the other master sends a logical zero. As soon as arbitration is lost by a master, it stops sending data and listens to the bus in order to detect a stop. When the stop is detected, the master who has lost arbitration may put its data on the bus by respecting arbitration. Arbitration is illustrated in Figure 40-23 on page 1155.
40.9.2 Different Multi-master Modes

Two multi-master modes may be distinguished: 1. 2. Note: TWI is considered as a Master only and will never be addressed. TWI may be either a Master or a Slave and may be addressed. In both Multi-master modes arbitration is supported.
40.9.2.1 TWI as Master Only In this mode, TWI is considered as a Master only (MSEN is always at one) and must be driven like a Master with the ARBLST (ARBitration Lost) flag in addition. If arbitration is lost (ARBLST = 1), the programmer must reinitiate the data transfer. If the user starts a transfer (ex.: DADR + START + W + Write in THR) and if the bus is busy, the TWI automatically waits for a STOP condition on the bus to initiate the transfer (see Figure 40-22 on page 1155). Note: The state of the bus (busy or free) is not indicated in the user interface.
40.9.2.2 TWI as Master or Slave The automatic reversal from Master to Slave is not supported in case of a lost arbitration. Then, in the case where TWI may be either a Master or a Slave, the programmer must manage the pseudo Multi-master mode described in the steps below. 1. 2. 3. 4. 5. 6. 7. Note: Program TWI in Slave mode (SADR + MSDIS + SVEN) and perform Slave Access (if TWI is addressed). If TWI has to be set in Master mode, wait until TXCOMP flag is at 1. Program Master mode (DADR + SVDIS + MSEN) and start the transfer (ex: START + Write in THR). As soon as the Master mode is enabled, TWI scans the bus in order to detect if it is busy or free. When the bus is considered as free, TWI initiates the transfer. As soon as the transfer is initiated and until a STOP condition is sent, the arbitration becomes relevant and the user must monitor the ARBLST flag. If the arbitration is lost (ARBLST is set to 1), the user must program the TWI in Slave mode in the case where the Master that won the arbitration wanted to access the TWI. If TWI has to be set in Slave mode, wait until TXCOMP flag is at 1 and then program the Slave mode. In the case where the arbitration is lost and TWI is addressed, TWI will not acknowledge even if it is programmed in Slave mode as soon as ARBLST is set to 1. Then, the Master must repeat SADR.

1154
Figure 40-22. Programmer Sends Data While the Bus is Busy
TWCK STOP sent by the master TWD DATA sent by a master Bus is busy Bus is free TWI DATA transfer Transfer is kept START sent by the TWI DATA sent by the TWI
A transfer is programmed (DADR + W + START + Write THR)
Bus is considered as free Transfer is initiated
Figure 40-23. Arbitration Cases
TWCK TWD
TWCK Data from a Master Data from TWI TWD S S S 1 1 1 0 0 0 1 1 1

Arbitration is lost TWI stops sending data
S S
1 1 1
0 0 0
Arbitration is lost The master stops sending data
0 1 0 1
1 1
Data from the TWI
0 0
1 1
Data from the master
ARBLST
Bus is busy Bus is free
TWI DATA transfer

A transfer is programmed (DADR + W + START + Write THR) Transfer is stopped
Transfer is kept
Transfer is programmed again (DADR + W + START + Write THR)
Bus is considered as free Transfer is initiated
The flowchart shown in Figure 40-24 on page 1156 gives an example of read and write operations in Multi-master mode.

1155
Figure 40-24. Multi-master Flowchart

START
Programm the SLAVE mode: SADR + MSDIS + SVEN
Read Status Register
SVACC = 1 ? No No EOSACC = 1 ? Yes No TXCOMP = 1 ? Yes No
Yes
GACC = 1 ?
No No No
SVREAD = 1 ? Yes
TXRDY= 1 ? Yes Write in TWI_THR
RXRDY= 1 ? Yes Read TWI_RHR
No
Need to perform a master access ?
GENERAL CALL TREATMENT Yes Decoding of the programming sequence Prog seq OK ? Change SADR Program the Master mode DADR + SVDIS + MSEN + CLK + R / W No
Yes ARBLST = 1 ?
No
Yes Yes RXRDY= 0 ? No Read TWI_RHR Yes Data to read? No
MREAD = 1 ?
No TXRDY= 0 ? No Data to send ? No Yes Write in TWI_THR Yes
Stop Transfer TWI_CR = STOP
Read Status Register Yes No
TXCOMP = 0 ?

1156
40.10 Slave Mode

40.10.1 Definition
The Slave Mode is defined as a mode where the device receives the clock and the address from another device called the master. In this mode, the device never initiates and never completes the transmission (START, REPEATED_START and STOP conditions are always provided by the master).
40.10.2 Application Block Diagram

Figure 40-25. Slave Mode Typical Application Block Diagram
VDD R TWD TWCK R
Master
Host with TWI Interface Slave 1
Host with TWI Interface Slave 2
LCD Controller Slave 3
40.10.3 Programming Slave Mode

The following fields must be programmed before entering Slave mode: 1. 2. 3. SADR (TWI_SMR): The slave device address is used in order to be accessed by master devices in read or write mode. MSDIS (TWI_CR): Disable the master mode. SVEN (TWI_CR): Enable the slave mode.
As the device receives the clock, values written in TWI_CWGR are not taken into account.
40.10.4 Receiving Data

After a Start or Repeated Start condition is detected and if the address sent by the Master matches with the Slave address programmed in the SADR (Slave ADdress) field, SVACC (Slave ACCess) flag is set and SVREAD (Slave READ) indicates the direction of the transfer. SVACC remains high until a STOP condition or a repeated START is detected. When such a condition is detected, EOSACC (End Of Slave ACCess) flag is set. 40.10.4.1 Read Sequence In the case of a Read sequence (SVREAD is high), TWI transfers data written in the TWI_THR (TWI Transmit Holding Register) until a STOP condition or a REPEATED_START + an address different from SADR is detected. Note that at the end of the read sequence TXCOMP (Transmission Complete) flag is set and SVACC reset. As soon as data is written in the TWI_THR, TXRDY (Transmit Holding Register Ready) flag is reset, and it is set when the shift register is empty and the sent data acknowledged or not. If the data is not acknowledged, the NACK flag is set. Note that a STOP or a repeated START always follows a NACK. See Figure 40-26 on page 1158.

1157
40.10.4.2 Write Sequence In the case of a Write sequence (SVREAD is low), the RXRDY (Receive Holding Register Ready) flag is set as soon as a character has been received in the TWI_RHR (TWI Receive Holding Register). RXRDY is reset when reading the TWI_RHR. TWI continues receiving data until a STOP condition or a REPEATED_START + an address different from SADR is detected. Note that at the end of the write sequence TXCOMP flag is set and SVACC reset. See Figure 40-27 on page 1159. 40.10.4.3 Clock Synchronization Sequence In the case where TWI_THR or TWI_RHR is not written/read in time, TWI performs a clock synchronization. Clock stretching information is given by the SCLWS (Clock Wait state) bit. See Figure 40-29 on page 1160 and Figure 40-30 on page 1161. 40.10.4.4 General Call In the case where a GENERAL CALL is performed, GACC (General Call ACCess) flag is set. After GACC is set, it is up to the programmer to interpret the meaning of the GENERAL CALL and to decode the new address programming sequence. See Figure 40-28 on page 1159.
40.10.5 Data Transfer

40.10.5.1 Read Operation The read mode is defined as a data requirement from the master. After a START or a REPEATED START condition is detected, the decoding of the address starts. If the slave address (SADR) is decoded, SVACC is set and SVREAD indicates the direction of the transfer. Until a STOP or REPEATED START condition is detected, TWI continues sending data loaded in the TWI_THR register. If a STOP condition or a REPEATED START + an address different from SADR is detected, SVACC is reset. Figure 40-26 on page 1158 describes the write operation.
Figure 40-26. Read Access Ordered by a MASTER
SADR does not match, TWI answers with a NACK SADR matches, TWI answers with an ACK
ACK/NACK from the Master A DATA NA S/Sr
TWD TXRDY NACK SVACC SVREAD EOSVACC
ADR
NA
DATA
NA
P/S/Sr
SADR R
DATA
Write THR
Read RHR
SVREAD has to be taken into account only while SVACC is active
Notes: 1. When SVACC is low, the state of SVREAD becomes irrelevant. 2. TXRDY is reset when data has been transmitted from TWI_THR to the shift register and set when this data has been acknowledged or non acknowledged.

1158
40.10.5.2 Write Operation The write mode is defined as a data transmission from the master. After a START or a REPEATED START, the decoding of the address starts. If the slave address is decoded, SVACC is set and SVREAD indicates the direction of the transfer (SVREAD is low in this case). Until a STOP or REPEATED START condition is detected, TWI stores the received data in the TWI_RHR register. If a STOP condition or a REPEATED START + an address different from SADR is detected, SVACC is reset. Figure 40-27 on page 1159 describes the Write operation.
Figure 40-27. Write Access Ordered by a Master
SADR does not match, TWI answers with a NACK SADR matches, TWI answers with an ACK Read RHR
TWD RXRDY SVACC SVREAD EOSVACC
ADR
NA
DATA
NA
P/S/Sr
SADR W
DATA
DATA
NA
S/Sr
SVREAD has to be taken into account only while SVACC is active
Notes: 1. When SVACC is low, the state of SVREAD becomes irrelevant. 2. RXRDY is set when data has been transmitted from the shift register to the TWI_RHR and reset when this data is read. 40.10.5.3 General Call The general call is performed in order to change the address of the slave. If a GENERAL CALL is detected, GACC is set. After the detection of General Call, it is up to the programmer to decode the commands which come afterwards. In case of a WRITE command, the programmer has to decode the programming sequence and program a new SADR if the programming sequence matches. Figure 40-28 on page 1159 describes the General Call access.
Figure 40-28. Master Performs a General Call
0000000 + W RESET command = 00000110X WRITE command = 00000100X
TXD
GENERAL CALL
Reset or write DADD
DATA1
DATA2
New SADR
New SADR Programming sequence GCACC
Reset after read

SVACC
Note:
This method allows the user to create an own programming sequence by choosing the programming bytes and the number of them. The programming sequence has to be provided to the master.

1159
40.10.5.4 Clock Synchronization In both read and write modes, it may happen that TWI_THR/TWI_RHR buffer is not filled /emptied before the emission/reception of a new character. In this case, to avoid sending/receiving undesired data, a clock stretching mechanism is implemented.
Clock Synchronization in Read Mode

The clock is tied low if the shift register is empty and if a STOP or REPEATED START condition was not detected. It is tied low until the shift register is loaded. Figure 40-29 on page 1160 describes the clock synchronization in Read mode.
Figure 40-29. Clock Synchronization in Read Mode
TWI_THR
DATA0
DATA1
DATA2
SADR
DATA0
DATA1
XXXXXXX 2
DATA2
NA
TWCK
Write THR CLOCK is tied low by the TWI as long as THR is empty
SCLWS TXRDY SVACC SVREAD TXCOMP

As soon as a START is detected
TWI_THR is transmitted to the shift register 1 2 The data is memorized in TWI_THR until a new value is written
Ack or Nack from the master
The clock is stretched after the ACK, the state of TWD is undefined during clock stretching
Notes: 1. TXRDY is reset when data has been written in the TWI_THR to the shift register and set when this data has been acknowledged or non acknowledged. 2. At the end of the read sequence, TXCOMP is set after a STOP or after a REPEATED_START + an address different from SADR. 3. SCLWS is automatically set when the clock synchronization mechanism is started.

1160
Clock Synchronization in Write Mode

The clock is tied low if the shift register and the TWI_RHR is full. If a STOP or REPEATED_START condition was not detected, it is tied low until TWI_RHR is read. Figure 40-30 on page 1161 describes the clock synchronization in Read mode.
Figure 40-30. Clock Synchronization in Write Mode
TWCK CLOCK is tied low by the TWI as long as RHR is full TWD S SADR W A DATA0 A DATA1 A DATA2
NA
ADR
TWI_RHR SCLWS
DATA0 is not read in the RHR
DATA1
DATA2
SCL is stretched on the last bit of DATA1
RXRDY Rd DATA0 SVACC SVREAD TXCOMP

Rd DATA1
Rd DATA2
Notes: 1. At the end of the read sequence, TXCOMP is set after a STOP or after a REPEATED_START + an address different from SADR. 2. SCLWS is automatically set when the clock synchronization mechanism is started and automatically reset when the mechanism is finished.

1161
40.10.5.5 Reversal after a Repeated Start
Reversal of Read to Write

The master initiates the communication by a read command and finishes it by a write command. Figure 40-31 on page 1162 describes the repeated start + reversal from Read to Write mode.
Figure 40-31. Repeated Start + Reversal from Read to Write Mode
TWI_THR DATA0 DATA1
TWD
SADR
DATA0
DATA1
NA
Sr
SADR
DATA2
DATA3
A DATA3
TWI_RHR SVACC SVREAD TXRDY RXRDY EOSACC TXCOMP

DATA2
Cleared after read
Note:
1.
TXCOMP is only set at the end of the transmission because after the repeated start, SADR is detected again.
Reversal of Write to Read

The master initiates the communication by a write command and finishes it by a read command. Figure 40-32 on page 1162 describes the repeated start + reversal from Write to Read mode.
Figure 40-32. Repeated Start + Reversal from Write to Read Mode
DATA2 DATA3
TWI_THR
TWD TWI_RHR SVACC SVREAD TXRDY RXRDY EOSACC TXCOMP
SADR
DATA0
DATA1
Sr
SADR
DATA2
DATA3
NA
DATA0
DATA1
Read TWI_RHR
Cleared after read
Notes: 1. In this case, if TWI_THR has not been written at the end of the read command, the clock is automatically stretched before the ACK. 2. TXCOMP is only set at the end of the transmission because after the repeated start, SADR is detected again.

1162
40.10.6 Read Write Flowcharts

The flowchart shown in Figure 40-33 on page 1163 gives an example of read and write operations in Slave mode. A polling or interrupt method can be used to check the status bits. The interrupt method requires that the interrupt enable register (TWI_IER) be configured first.
Figure 40-33. Read Write Flowchart in Slave Mode
Set the SLAVE mode: SADR + MSDIS + SVEN
SVACC = 1 ?
No No
GACC = 1 ?
No
SVREAD = 0 ?
No
EOSACC = 1 ?
TXRDY= 1 ?
No
No
Write in TWI_THR TXCOMP = 1 ? RXRDY= 0 ? END Read TWI_RHR

No
GENERAL CALL TREATMENT
Decoding of the programming sequence
Prog seq OK ?
No
Change SADR

1163
40.11 Write Protection System

In order to bring security to the TWI, a write protection system has been implemented. The write protection mode prevents the write of TWI Clock Waveform Generator Register and TWI Slave Mode Register. When this mode is enabled and one of the protected registers is written, an error is generated in the TWI Write Protection Status Register and the register write request is canceled. When a write protection error occurs the WPROTERR flag is set and the address of the corresponding canceled register write is available in the WPROTADRR field of the TWI_WPROT_STATUS register. Due to the nature of the write protection feature, enabling and disabling the write protection mode requires the use of a security code. Thus when enabling or disabling the write protection mode the SECURITY_CODE field of the TWI Write Protection Mode Register must be filled with the TWI ASCII code (0x545749) otherwise the register write will be canceled.

1164
40.12 Two-wire Interface (TWI) User Interface

Table 40-6. Register Mapping Offset 0x00 0x04 0x08 0x0C 0x10 0x14 - 0x1C 0x20 0x24 0x28 0x2C 0x30 0x34 0xE4 0xE8 0xEC - 0xFC
(1)
Register Control Register Master Mode Register Slave Mode Register Internal Address Register Clock Waveform Generator Register Reserved Status Register Interrupt Enable Register Interrupt Disable Register Interrupt Mask Register Receive Holding Register Transmit Holding Register Protection Mode Register Protection Status Register Reserved
Name TWI_CR TWI_MMR TWI_SMR TWI_IADR TWI_CWGR TWI_SR TWI_IER TWI_IDR TWI_IMR TWI_RHR TWI_THR TWI_WPROT_MODE TWI_WPROT_STATUS
Access Write-only Read-write Read-write Read-write Read-write Read-only Write-only Write-only Read-only Read-only Write-only Read-write Read-only
Reset N/A 0x00000000 0x00000000 0x00000000 0x00000000 0x0000F009 N/A N/A 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000
Note:
1. All unlisted offset values are considered as reserved.

1165
40.12.1 TWI Control Register

31 23 15 7 SWRST
TWI_CR 0xF0014000 (0), 0xF0018000 (1), 0xF801C000 (2) Write-only 0x00000000

30 22 14 6 QUICK 29 21 13 5 SVDIS 28 20 12 4 SVEN 27 19 11 3 MSDIS 26 18 10 2 MSEN 25 17 9 1 STOP 24 16 8 0 START
START: Send a START Condition

0 = No effect. 1 = A frame beginning with a START bit is transmitted according to the features defined in the mode register. This action is necessary when the TWI peripheral wants to read data from a slave. When configured in Master Mode with a write operation, a frame is sent as soon as the user writes a character in the Transmit Holding Register (TWI_THR).
STOP: Send a STOP Condition

0 = No effect. 1 = STOP Condition is sent just after completing the current byte transmission in master read mode.
In single data byte master read, the START and STOP must both be set. In multiple data bytes master read, the STOP must be set after the last data received but one. In master read mode, if a NACK bit is received, the STOP is automatically performed. In master data write operation, a STOP condition will be sent after the transmission of the current data is finished. MSEN: TWI Master Mode Enabled
0 = No effect. 1 = If MSDIS = 0, the master mode is enabled. Note: Switching from Slave to Master mode is only permitted when TXCOMP = 1.
MSDIS: TWI Master Mode Disabled

0 = No effect. 1 = The master mode is disabled, all pending data is transmitted. The shifter and holding characters (if it contains data) are transmitted in case of write operation. In read operation, the character being transferred must be completely received before disabling.
SVEN: TWI Slave Mode Enabled

0 = No effect. 1 = If SVDIS = 0, the slave mode is enabled. Note: Switching from Master to Slave mode is only permitted when TXCOMP = 1.

1166
SVDIS: TWI Slave Mode Disabled

0 = No effect. 1 = The slave mode is disabled. The shifter and holding characters (if it contains data) are transmitted in case of read operation. In write operation, the character being transferred must be completely received before disabling.
QUICK: SMBUS Quick Command

0 = No effect. 1 = If Master mode is enabled, a SMBUS Quick Command is sent.

0 = No effect. 1 = Equivalent to a system reset.

1167
40.12.2 TWI Master Mode Register

31 23 15 7
TWI_MMR 0xF0014004 (0), 0xF0018004 (1), 0xF801C004 (2) Read-write 0x00000000

30 22 29 21 28 20 27 19 DADR 11 3 26 18 25 17 24 16
14 6
13 5
12 MREAD 4
10 2
9 IADRSZ 1
IADRSZ: Internal Device Address Size

Value 0 1 2 3 Name NONE 1_BYTE 2_BYTE 3_BYTE Description No internal device address One-byte internal device address Two-byte internal device address Three-byte internal device address
MREAD: Master Read Direction

0 = Master write direction. 1 = Master read direction.
DADR: Device Address

The device address is used to access slave devices in read or write mode. Those bits are only used in Master mode.

1168
40.12.3 TWI Slave Mode Register

31 23 15 7
TWI_SMR 0xF0014008 (0), 0xF0018008 (1), 0xF801C008 (2) Read-write 0x00000000

30 22 29 21 28 20 27 19 SADR 11 3 26 18 25 17 24 16
14 6
13 5
12 4
10 2
This register can only be written if the WPEN bit is cleared in the TWI Write Protection Mode Register.
SADR: Slave Address

The slave device address is used in Slave mode in order to be accessed by master devices in read or write mode. SADR must be programmed before enabling the Slave mode or after a general call. Writes at other times have no effect.

1169
40.12.4 TWI Internal Address Register

31 23
TWI_IADR 0xF001400C (0), 0xF001800C (1), 0xF801C00C (2) Read-write 0x00000000

30 22 29 21 28 20 IADR 15 14 13 12 IADR 7 6 5 4 IADR 3 2 1 0 11 10 9 8 27 19 26 18 25 17 24 16
IADR: Internal Address

0, 1, 2 or 3 bytes depending on IADRSZ.

1170
40.12.5 TWI Clock Waveform Generator Register

31 23
TWI_CWGR 0xF0014010 (0), 0xF0018010 (1), 0xF801C010 (2) Read-write 0x00000000

30 22 29 21 28 20 27 19 26 18 25 17 CKDIV 9 24 16
15
14
13
12 CHDIV
11
10
4 CLDIV
This register can only be written if the WPEN bit is cleared in the TWI Write Protection Mode Register. TWI_CWGR is only used in Master mode.
CLDIV: Clock Low Divider

The SCL low period is defined as follows:
T low = ( ( CLDIV 2
CKDIV
) + 4 ) T MCK
CHDIV: Clock High Divider

The SCL high period is defined as follows:
T high = ( ( CHDIV 2
CKDIV
) + 4 ) T MCK
CKDIV: Clock Divider

The CKDIV is used to increase both SCL high and low periods.

1171
40.12.6 TWI Status Register

31 23 15 7
TWI_SR 0xF0014020 (0), 0xF0018020 (1), 0xF801C020 (2) Read-only 0x0000F009

30 22 14 6 OVRE 29 21 13 5 GACC 28 20 12 4 SVACC 27 19 11 EOSACC 3 SVREAD 26 18 10 SCLWS 2 TXRDY 25 17 9 ARBLST 1 RXRDY 24 16 8 NACK 0 TXCOMP
TXCOMP: Transmission Completed (automatically set / reset)

TXCOMP used in Master mode: 0 = During the length of the current frame. 1 = When both holding and shifter registers are empty and STOP condition has been sent. TXCOMP behavior in Master mode can be seen in Figure 40-8 on page 1143 and in Figure 40-10 on page 1144. TXCOMP used in Slave mode: 0 = As soon as a Start is detected. 1 = After a Stop or a Repeated Start + an address different from SADR is detected. TXCOMP behavior in Slave mode can be seen in Figure 40-29 on page 1160, Figure 40-30 on page 1161, Figure 40-31 on page 1162 and Figure 40-32 on page 1162.
RXRDY: Receive Holding Register Ready (automatically set / reset)

0 = No character has been received since the last TWI_RHR read operation. 1 = A byte has been received in the TWI_RHR since the last read. RXRDY behavior in Master mode can be seen in Figure 40-10 on page 1144. RXRDY behavior in Slave mode can be seen in Figure 40-27 on page 1159, Figure 40-30 on page 1161, Figure 40-31 on page 1162 and Figure 40-32 on page 1162.
TXRDY: Transmit Holding Register Ready (automatically set / reset)

TXRDY used in Master mode: 0 = The transmit holding register has not been transferred into shift register. Set to 0 when writing into TWI_THR register. 1 = As soon as a data byte is transferred from TWI_THR to internal shifter or if a NACK error is detected, TXRDY is set at the same time as TXCOMP and NACK. TXRDY is also set when MSEN is set (enable TWI). TXRDY behavior in Master mode can be seen in Figure 40.8.4 on page 1142. TXRDY used in Slave mode: 0 = As soon as data is written in the TWI_THR, until this data has been transmitted and acknowledged (ACK or NACK). 1 = It indicates that the TWI_THR is empty and that data has been transmitted and acknowledged.

1172
If TXRDY is high and if a NACK has been detected, the transmission will be stopped. Thus when TRDY = NACK = 1, the programmer must not fill TWI_THR to avoid losing it. TXRDY behavior in Slave mode can be seen in Figure 40-26 on page 1158, Figure 40-29 on page 1160, Figure 40-31 on page 1162 and Figure 40-32 on page 1162.
SVREAD: Slave Read (automatically set / reset)

This bit is only used in Slave mode. When SVACC is low (no Slave access has been detected) SVREAD is irrelevant. 0 = Indicates that a write access is performed by a Master. 1 = Indicates that a read access is performed by a Master. SVREAD behavior can be seen in Figure 40-26 on page 1158, Figure 40-27 on page 1159, Figure 40-31 on page 1162 and Figure 40-32 on page 1162.
SVACC: Slave Access (automatically set / reset)

This bit is only used in Slave mode. 0 = TWI is not addressed. SVACC is automatically cleared after a NACK or a STOP condition is detected. 1 = Indicates that the address decoding sequence has matched (A Master has sent SADR). SVACC remains high until a NACK or a STOP condition is detected. SVACC behavior can be seen in Figure 40-26 on page 1158, Figure 40-27 on page 1159, Figure 40-31 on page 1162 and Figure 40-32 on page 1162.
GACC: General Call Access (clear on read)

This bit is only used in Slave mode. 0 = No General Call has been detected. 1 = A General Call has been detected. After the detection of General Call, if need be, the programmer may acknowledge this access and decode the following bytes and respond according to the value of the bytes. GACC behavior can be seen in Figure 40-28 on page 1159.
OVRE: Overrun Error (clear on read)

This bit is only used in Master mode. 0 = TWI_RHR has not been loaded while RXRDY was set 1 = TWI_RHR has been loaded while RXRDY was set. Reset by read in TWI_SR when TXCOMP is set.
NACK: Not Acknowledged (clear on read)

NACK used in Master mode: 0 = Each data byte has been correctly received by the far-end side TWI slave component. 1 = A data byte has not been acknowledged by the slave component. Set at the same time as TXCOMP. NACK used in Slave Read mode: 0 = Each data byte has been correctly received by the Master. 1 = In read mode, a data byte has not been acknowledged by the Master. When NACK is set the programmer must not fill TWI_THR even if TXRDY is set, because it means that the Master will stop the data transfer or re initiate it. Note that in Slave Write mode all data are acknowledged by the TWI.
ARBLST: Arbitration Lost (clear on read)

This bit is only used in Master mode. 0: Arbitration won. 1: Arbitration lost. Another master of the TWI bus has won the multi-master arbitration. TXCOMP is set at the same time.

1173
SCLWS: Clock Wait State (automatically set / reset)

This bit is only used in Slave mode. 0 = The clock is not stretched. 1 = The clock is stretched. TWI_THR / TWI_RHR buffer is not filled / emptied before the emission / reception of a new character. SCLWS behavior can be seen in Figure 40-29 on page 1160 and Figure 40-30 on page 1161.
EOSACC: End Of Slave Access (clear on read)

This bit is only used in Slave mode. 0 = A slave access is being performing. 1 = The Slave Access is finished. End Of Slave Access is automatically set as soon as SVACC is reset. EOSACC behavior can be seen in Figure 40-31 on page 1162 and Figure 40-32 on page 1162

1174
40.12.7 TWI Interrupt Enable Register

31 23 15 7
TWI_IER 0xF0014024 (0), 0xF0018024 (1), 0xF801C024 (2) Write-only 0x00000000

30 22 14 6 OVRE 29 21 13 5 GACC 28 20 12 4 SVACC 27 19 11 EOSACC 3 26 18 10 SCL_WS 2 TXRDY 25 17 9 ARBLST 1 RXRDY 24 16 8 NACK 0 TXCOMP
TXCOMP: Transmission Completed Interrupt Enable RXRDY: Receive Holding Register Ready Interrupt Enable TXRDY: Transmit Holding Register Ready Interrupt Enable SVACC: Slave Access Interrupt Enable GACC: General Call Access Interrupt Enable OVRE: Overrun Error Interrupt Enable NACK: Not Acknowledge Interrupt Enable ARBLST: Arbitration Lost Interrupt Enable SCL_WS: Clock Wait State Interrupt Enable EOSACC: End Of Slave Access Interrupt Enable

1175
40.12.8 TWI Interrupt Disable Register

31 23 15 7
TWI_IDR 0xF0014028 (0), 0xF0018028 (1), 0xF801C028 (2) Write-only 0x00000000

TXCOMP: Transmission Completed Interrupt Disable RXRDY: Receive Holding Register Ready Interrupt Disable TXRDY: Transmit Holding Register Ready Interrupt Disable SVACC: Slave Access Interrupt Disable GACC: General Call Access Interrupt Disable OVRE: Overrun Error Interrupt Disable NACK: Not Acknowledge Interrupt Disable ARBLST: Arbitration Lost Interrupt Disable SCL_WS: Clock Wait State Interrupt Disable EOSACC: End Of Slave Access Interrupt Disable

1176
40.12.9 TWI Interrupt Mask Register

31 23 15 7
TWI_IMR 0xF001402C (0), 0xF001802C (1), 0xF801C02C (2) Read-only 0x00000000

TXCOMP: Transmission Completed Interrupt Mask RXRDY: Receive Holding Register Ready Interrupt Mask TXRDY: Transmit Holding Register Ready Interrupt Mask SVACC: Slave Access Interrupt Mask GACC: General Call Access Interrupt Mask OVRE: Overrun Error Interrupt Mask NACK: Not Acknowledge Interrupt Mask ARBLST: Arbitration Lost Interrupt Mask SCL_WS: Clock Wait State Interrupt Mask EOSACC: End Of Slave Access Interrupt Mask
0 = The corresponding interrupt is disabled. 1 = The corresponding interrupt is enabled.

1177
40.12.10 TWI Receive Holding Register

31 23 15 7
TWI_RHR 0xF0014030 (0), 0xF0018030 (1), 0xF801C030 (2) Read-only 0x00000000

30 22 14 6 29 21 13 5 28 20 12 4 RXDATA 27 19 11 3 26 18 10 2 25 17 9 1 24 16 8 0
RXDATA: Master or Slave Receive Holding Data
40.12.11 TWI Transmit Holding Register

31 23 15 7
TWI_THR 0xF0014034 (0), 0xF0018034 (1), 0xF801C034 (2) Read-write 0x00000000

30 22 14 6 29 21 13 5 28 20 12 4 TXDATA 27 19 11 3 26 18 10 2 25 17 9 1 24 16 8 0
TXDATA: Master or Slave Transmit Holding Data

1178
40.12.12 TWI Write Protection Mode Register

31
TWI_WPROT_MODE 0xF00140E4 (0), 0xF00180E4 (1), 0xF801C0E4 (2) Read-write

30 29 28 27 SECURITY_CODE 20 19 SECURITY_CODE 12 11 SECURITY_CODE 4 3 26 25 24
23
22
21
18
17
16
15
14
13
10
0 WPROT
SECURITY_CODE: Write protection mode security code

This security code is needed to set/reset the WPROT bit value (see Section 40.11 Write Protection System for details). Must be filled with 0x545749 (ASCII code for TWI).
WPROT: Write protection bit

Enables/Disables write protection mode. The write protected registers are:
TWI Clock Waveform Generator Register TWI Slave Mode Register

1179
40.12.13 TWI Write Protection Status Register

31
TWI_WPROT_STATUS 0xF00140E8 (0), 0xF00180E8 (1), 0xF801C0E8 (2) Read-only

30 29 28 27 WPROTADDR 20 19 WPROTADDR 12 11 WPROTADDR 4 3 26 25 24
23
22
21
18
17
16
15
14
13
10
0 WPROTERR
WPROTADDR: Write Protection Error Address

Indicates the address of the register write request which generated the error.
WPROTERR: Write Protection Error

Indicates a write protection error.

1180
41.
41.1
Synchronous Serial Controller (SSC)

Description
The Synchronous Serial Controller (SSC) provides a synchronous communication link with external devices. It supports many serial synchronous communication protocols generally used in audio and telecom applications such as I2S, Short Frame Sync, Long Frame Sync, etc. The SSC contains an independent receiver and transmitter and a common clock divider. The receiver and the transmitter each interface with three signals: the TD/RD signal for data, the TK/RK signal for the clock and the TF/RF signal for the Frame Sync. The transfers can be programmed to start automatically or on different events detected on the Frame Sync signal. The SSC high-level of programmability and its use of DMA permit a continuous high bit rate data transfer without processor intervention. Featuring connection to the DMA, the SSC permits interfacing with low processor overhead to the following:
z z z
CODECs in master or slave mode DAC through dedicated serial interface, particularly I2S Magnetic card reader
41.2
z z z z z z
Provides Serial Synchronous Communication Links Used in Audio and Telecom Applications Contains an Independent Receiver and Transmitter and a Common Clock Divider Interfaced with the DMA Controller (DMAC) to Reduce Processor Overhead Offers a Configurable Frame Sync and Data Length Receiver and Transmitter Can be Programmed to Start Automatically or on Detection of Different Events on the Frame Sync Signal Receiver and Transmitter Include a Data Signal, a Clock Signal and a Frame Synchronization Signal

1181
41.3
Block Diagram

System Bus
APB Bridge
DMA Peripheral Bus TF TK TD SSC Interface PIO RF RK Interrupt Control RD
PMC
MCK
SSC Interrupt

1182
41.4

Power Management SSC Time Slot Management Frame Management Interrupt Management Test Management
OS or RTOS Driver
Serial AUDIO
Codec
Line Interface
41.5
Pin Name List
Table 41-1. I/O Lines Description Pin Name RF RK RD TF TK TD Pin Description Receiver Frame Synchro Receiver Clock Receiver Data Transmitter Frame Synchro Transmitter Clock Transmitter Data Type Input/Output Input/Output Input Input/Output Input/Output Output

1183
41.6
41.6.1 I/O Lines

The pins used for interfacing the compliant external devices may be multiplexed with PIO lines. Before using the SSC receiver, the PIO controller must be configured to dedicate the SSC receiver I/O lines to the SSC peripheral mode. Before using the SSC transmitter, the PIO controller must be configured to dedicate the SSC transmitter I/O lines to the SSC peripheral mode.
Table 41-2. I/O Lines Instance SSC0 SSC0 SSC0 SSC0 SSC0 SSC0 SSC1 SSC1 SSC1 SSC1 SSC1 SSC1 Signal RD0 RF0 RK0 TD0 TF0 TK0 RD1 RF1 RK1 TD1 TF1 TK1 I/O Line PC21 PC20 PC19 PC18 PC17 PC16 PB11 PB10 PB7 PB6 PB3 PB2 Peripheral A A A A A A B B B B B B

The SSC is not continuously clocked. The SSC interface may be clocked through the Power Management Controller (PMC), therefore the programmer must first configure the PMC to enable the SSC clock.
41.6.3 Interrupt
The SSC interface has an interrupt line connected to the interrupt controller. Handling interrupts requires programming the interrupt controller before configuring the SSC. All SSC interrupts can be enabled/disabled configuring the SSC Interrupt mask register. Each pending and unmasked
Table 41-3. Peripheral IDs Instance SSC0 SSC1 ID 38 39
SSC interrupt will assert the SSC interrupt line. The SSC interrupt service routine can get the interrupt origin by reading the SSC interrupt status register.

1184
41.7
This chapter contains the functional description of the following: SSC Functional Block, Clock Management, Data format, Start, Transmitter, Receiver and Frame Sync. The receiver and transmitter operate separately. However, they can work synchronously by programming the receiver to use the transmit clock and/or to start a data transfer when transmission starts. Alternatively, this can be done by programming the transmitter to use the receive clock and/or to start a data transfer when reception starts. The transmitter and the receiver can be programmed to operate with the clock signals provided on either the TK or RK pins. This allows the SSC to support many slave-mode data transfers. The maximum clock speed allowed on the TK and RK pins is the master clock divided by 2.
Figure 41-3. SSC Functional Block Diagram

Transmitter
Clock Output Controller TX clock
TK
MCK
Clock Divider
TK Input RX clock TXEN RX Start Start Selector TF Transmit Clock Controller
Frame Sync Controller
TF
TX Start Transmit Shift Register Transmit Sync Holding Register
Data Controller
TD
APB User Interface
Transmit Holding Register
Receiver
Clock Output Controller
RK
RK Input TX Clock RXEN TX Start Start RF Selector RC0R
Receive Clock RX Clock Controller RX Start Receive Shift Register Receive Sync Holding Register
Frame Sync Controller
RF
Data Controller
RD
Interrupt Control
Receive Holding Register
To Interrupt Controller

1185
41.7.1 Clock Management

The transmitter clock can be generated by:
z z z
An external clock received on the TK I/O pad The receiver clock The internal clock divider An external clock received on the RK I/O pad The transmitter clock The internal clock divider
The receiver clock can be generated by:

z z z
Furthermore, the transmitter block can generate an external clock on the TK I/O pad, and the receiver block can generate an external clock on the RK I/O pad. This allows the SSC to support many Master and Slave Mode data transfers. 41.7.1.1 Clock Divider
Figure 41-4. Divided Clock Block Diagram
Clock Divider SSC_CMR MCK /2
12-bit Counter
Divided Clock
The Master Clock divider is determined by the 12-bit field DIV counter and comparator (so its maximal value is 4095) in the Clock Mode Register SSC_CMR, allowing a Master Clock division by up to 8190. The Divided Clock is provided to both the Receiver and Transmitter. When this field is programmed to 0, the Clock Divider is not used and remains inactive. When DIV is set to a value equal to or greater than 1, the Divided Clock has a frequency of Master Clock divided by 2 times DIV. Each level of the Divided Clock has a duration of the Master Clock multiplied by DIV. This ensures a 50% duty cycle for the Divided Clock regardless of whether the DIV value is even or odd.
Figure 41-5. Divided Clock Generation
Master Clock
Divided Clock DIV = 1 Divided Clock Frequency = MCK/2
Master Clock
Divided Clock DIV = 3 Divided Clock Frequency = MCK/6

1186
41.7.1.2 Transmitter Clock Management The transmitter clock is generated from the receiver clock or the divider clock or an external clock scanned on the TK I/O pad. The transmitter clock is selected by the CKS field in SSC_TCMR (Transmit Clock Mode Register). Transmit Clock can be inverted independently by the CKI bits in SSC_TCMR. The transmitter can also drive the TK I/O pad continuously or be limited to the actual data transfer. The clock output is configured by the SSC_TCMR register. The Transmit Clock Inversion (CKI) bits have no effect on the clock outputs. Programming the TCMR register to select TK pin (CKS field) and at the same time Continuous Transmit Clock (CKO field) might lead to unpredictable results.
Figure 41-6. Transmitter Clock Management
TK (pin)
MUX Receiver Clock
Tri_state Controller
Clock Output
Divider Clock CKO Data Transfer
CKS
INV MUX
Tri-state Controller
Transmitter Clock
CKI
CKG
41.7.1.3 Receiver Clock Management The receiver clock is generated from the transmitter clock or the divider clock or an external clock scanned on the RK I/O pad. The Receive Clock is selected by the CKS field in SSC_RCMR (Receive Clock Mode Register). Receive Clocks can be inverted independently by the CKI bits in SSC_RCMR. The receiver can also drive the RK I/O pad continuously or be limited to the actual data transfer. The clock output is configured by the SSC_RCMR register. The Receive Clock Inversion (CKI) bits have no effect on the clock outputs. Programming the RCMR register to select RK pin (CKS field) and at the same time Continuous Receive Clock (CKO field) can lead to unpredictable results.

1187
Figure 41-7. Receiver Clock Management

RK (pin) Tri-state Controller
MUX Transmitter Clock
Clock Output
Divider Clock CKO Data Transfer
CKS
INV MUX
Tri-state Controller
Receiver Clock
CKI
CKG
41.7.1.4 Serial Clock Ratio Considerations The Transmitter and the Receiver can be programmed to operate with the clock signals provided on either the TK or RK pins. This allows the SSC to support many slave-mode data transfers. In this case, the maximum clock speed allowed on the RK pin is:
z z
Master Clock divided by 2 if Receiver Frame Synchro is input Master Clock divided by 3 if Receiver Frame Synchro is output Master Clock divided by 6 if Transmit Frame Synchro is input Master Clock divided by 2 if Transmit Frame Synchro is output
In addition, the maximum clock speed allowed on the TK pin is:

z z
41.7.2 Transmitter Operations

A transmitted frame is triggered by a start event and can be followed by synchronization data before data transmission. The start event is configured by setting the Transmit Clock Mode Register (SSC_TCMR). See Start on page 1190. The frame synchronization is configured setting the Transmit Frame Mode Register (SSC_TFMR). See Frame Sync on page 1192. To transmit data, the transmitter uses a shift register clocked by the transmitter clock signal and the start mode selected in the SSC_TCMR. Data is written by the application to the SSC_THR register then transferred to the shift register according to the data format selected. When both the SSC_THR and the transmit shift register are empty, the status flag TXEMPTY is set in SSC_SR. When the Transmit Holding register is transferred in the Transmit shift register, the status flag TXRDY is set in SSC_SR and additional data can be loaded in the holding register.

1188
Figure 41-8. Transmitter Block Diagram

SSC_CRTXEN SSC_SRTXEN SSC_CRTXDIS SSC_TCMR.STTDLY SSC_TFMR.FSDEN SSC_RCMR.START SSC_TCMR.START SSC_TFMR.DATNB SSC_TFMR.DATDEF SSC_TFMR.MSBF RXEN TXEN TX Start TX Start Start RX Start Start RF Selector Selector RF RC0R Transmit Shift Register SSC_TFMR.FSDEN SSC_TCMR.STTDLY != 0 TXEN
TX Controller TD
Transmitter Clock
SSC_TFMR.DATLEN
SSC_THR
SSC_TSHR
SSC_TFMR.FSLEN
TX Controller counter reached STTDLY
41.7.3 Receiver Operations

A received frame is triggered by a start event and can be followed by synchronization data before data transmission. The start event is configured setting the Receive Clock Mode Register (SSC_RCMR). See Start on page 1190. The frame synchronization is configured setting the Receive Frame Mode Register (SSC_RFMR). See Frame Sync on page 1192. The receiver uses a shift register clocked by the receiver clock signal and the start mode selected in the SSC_RCMR. The data is transferred from the shift register depending on the data format selected. When the receiver shift register is full, the SSC transfers this data in the holding register, the status flag RXRDY is set in SSC_SR and the data can be read in the receiver holding register. If another transfer occurs before read of the RHR register, the status flag OVERUN is set in SSC_SR and the receiver shift register is transferred in the RHR register.

1189
Figure 41-9. Receiver Block Diagram

SSC_CR.RXEN SSC_SR.RXEN SSC_CR.RXDIS SSC_TCMR.START SSC_RCMR.START TXEN RX Start RF Start Selector RXEN RF RC0R Start Selector SSC_RFMR.MSBF SSC_RFMR.DATNB RX Start RX Controller RD Receive Shift Register
SSC_RCMR.STTDLY != 0 load SSC_RSHR load SSC_RHR
Receiver Clock
SSC_RFMR.FSLEN
SSC_RFMR.DATLEN RX Controller counter reached STTDLY
41.7.4 Start
The transmitter and receiver can both be programmed to start their operations when an event occurs, respectively in the Transmit Start Selection (START) field of SSC_TCMR and in the Receive Start Selection (START) field of SSC_RCMR. Under the following conditions the start event is independently programmable:
z z z z z
Continuous. In this case, the transmission starts as soon as a word is written in SSC_THR and the reception starts as soon as the Receiver is enabled. Synchronously with the transmitter/receiver On detection of a falling/rising edge on TF/RF On detection of a low level/high level on TF/RF On detection of a level change or an edge on TF/RF
A start can be programmed in the same manner on either side of the Transmit/Receive Clock Register (RCMR/TCMR). Thus, the start could be on TF (Transmit) or RF (Receive). Moreover, the Receiver can start when data is detected in the bit stream with the Compare Functions. Detection on TF/RF input/output is done by the field FSOS of the Transmit/Receive Frame Mode Register (TFMR/RFMR).

1190
Figure 41-10. Transmit Start Mode

TK TF (Input)
Start = Low Level on TF
TD (Output) TD (Output)
BO
B1 STTDLY
Start = Falling Edge on TF
BO
B1 STTDLY X BO B1 STTDLY
Start = High Level on TF
TD (Output) TD (Output) X
Start = Rising Edge on TF
BO
B1 STTDLY
TD Start = Level Change on TF (Output) TD (Output)
BO
B1
BO
B1 STTDLY
Start = Any Edge on TF
BO
B1
BO
B1 STTDLY
Figure 41-11. Receive Pulse/Edge Start Modes

RK RF (Input)
Start = Low Level on RF
RD (Input) RD (Input)
BO
B1 STTDLY
Start = Falling Edge on RF
BO
B1 STTDLY X BO B1 STTDLY
Start = High Level on RF
RD (Input) RD (Input) RD (Input) RD (Input) X
Start = Rising Edge on RF
BO
B1 STTDLY
Start = Level Change on RF
BO
B1
BO
B1 STTDLY
Start = Any Edge on RF
BO
B1
BO
B1 STTDLY

1191
41.7.5 Frame Sync

The Transmitter and Receiver Frame Sync pins, TF and RF, can be programmed to generate different kinds of frame synchronization signals. The Frame Sync Output Selection (FSOS) field in the Receive Frame Mode Register (SSC_RFMR) and in the Transmit Frame Mode Register (SSC_TFMR) are used to select the required waveform.
z z
Programmable low or high levels during data transfer are supported. Programmable high levels before the start of data transfers or toggling are also supported.
If a pulse waveform is selected, the Frame Sync Length (FSLEN) field in SSC_RFMR and SSC_TFMR programs the length of the pulse, from 1 bit time up to 256 bit time. The periodicity of the Receive and Transmit Frame Sync pulse output can be programmed through the Period Divider Selection (PERIOD) field in SSC_RCMR and SSC_TCMR. 41.7.5.1 Frame Sync Data Frame Sync Data transmits or receives a specific tag during the Frame Sync signal. During the Frame Sync signal, the Receiver can sample the RD line and store the data in the Receive Sync Holding Register and the transmitter can transfer Transmit Sync Holding Register in the Shifter Register. The data length to be sampled/shifted out during the Frame Sync signal is programmed by the FSLEN field in SSC_RFMR/SSC_TFMR and has a maximum value of 16. Concerning the Receive Frame Sync Data operation, if the Frame Sync Length is equal to or lower than the delay between the start event and the actual data reception, the data sampling operation is performed in the Receive Sync Holding Register through the Receive Shift Register. The Transmit Frame Sync Operation is performed by the transmitter only if the bit Frame Sync Data Enable (FSDEN) in SSC_TFMR is set. If the Frame Sync length is equal to or lower than the delay between the start event and the actual data transmission, the normal transmission has priority and the data contained in the Transmit Sync Holding Register is transferred in the Transmit Register, then shifted out. 41.7.5.2 Frame Sync Edge Detection The Frame Sync Edge detection is programmed by the FSEDGE field in SSC_RFMR/SSC_TFMR. This sets the corresponding flags RXSYN/TXSYN in the SSC Status Register (SSC_SR) on frame synchro edge detection (signals RF/TF).
41.7.6 Receive Compare Modes

Figure 41-12. Receive Compare Modes
RK
RD (Input)
CMP0
CMP1
CMP2
CMP3 Start
Ignored
B0
B1
B2
FSLEN Up to 16 Bits (4 in This Example)
STDLY
DATLEN
41.7.6.1 Compare Functions Length of the comparison patterns (Compare 0, Compare 1) and thus the number of bits they are compared to is defined by FSLEN, but with a maximum value of 16 bits. Comparison is always done by comparing the last bits received with the comparison pattern. Compare 0 can be one start event of the Receiver. In this case, the receiver compares at each new sample the last bits received at the Compare 0 pattern contained in the Compare 0 Register (SSC_RC0R). When this start event is selected, the user can program the Receiver to start a new data transfer either by writing a new Compare 0, or by receiving continuously until Compare 1 occurs. This selection is done with the bit (STOP) in SSC_RCMR.

1192
41.7.7 Data Format

The data framing format of both the transmitter and the receiver are programmable through the Transmitter Frame Mode Register (SSC_TFMR) and the Receiver Frame Mode Register (SSC_RFMR). In either case, the user can independently select:
z z z z z z
The event that starts the data transfer (START) The delay in number of bit periods between the start event and the first data bit (STTDLY) The length of the data (DATLEN) The number of data to be transferred for each start event (DATNB). The length of synchronization transferred for each start event (FSLEN) The bit sense: most or lowest significant bit first (MSBF)
Additionally, the transmitter can be used to transfer synchronization and select the level driven on the TD pin while not in data transfer operation. This is done respectively by the Frame Sync Data Enable (FSDEN) and by the Data Default Value (DATDEF) bits in SSC_TFMR.
Table 41-4. Data Frame Registers Transmitter SSC_TFMR SSC_TFMR SSC_TFMR SSC_TFMR SSC_TFMR SSC_TFMR SSC_TCMR SSC_TCMR SSC_RCMR SSC_RCMR Receiver SSC_RFMR SSC_RFMR SSC_RFMR SSC_RFMR Field DATLEN DATNB MSBF FSLEN DATDEF FSDEN PERIOD STTDLY Up to 512 Up to 255 Up to 16 0 or 1 Length Up to 32 Up to 16 Comment Size of word Number of words transmitted in frame Most significant bit first Size of Synchro data register Data default value ended Enable send SSC_TSHR Frame size Size of transmit start delay
Figure 41-13. Transmit and Receive Frame Format in Edge/Pulse Start Modes
Start PERIOD TF/RF
(1)
Start
FSLEN TD (If FSDEN = 1) Sync Data Default Data From SSC_THR Data From SSC_THR Data To SSC_RHR DATLEN Data From SSC_THR Data From SSC_THR Data To SSC_RHR DATLEN Default FromDATDEF Default From DATDEF Ignored Sync Data Sync Data
From SSC_TSHR FromDATDEF Default From DATDEF Sync Data To SSC_RSHR STTDLY Ignored
TD (If FSDEN = 0) RD
DATNB
Note:
1. Example of input on falling edge of TF/RF.

1193
Figure 41-14. Transmit Frame Format in Continuous Mode

Start
TD
Data From SSC_THR DATLEN
Data From SSC_THR DATLEN
Default
Start: 1. TXEMPTY set to 1 2. Write into the SSC_THR
Note:
1.
STTDLY is set to 0. In this example, SSC_THR is loaded twice. FSDEN value has no effect on the transmission. SyncData cannot be output in continuous mode.
Figure 41-15. Receive Frame Format in Continuous Mode

Start = Enable Receiver
RD
Data To SSC_RHR DATLEN
Data To SSC_RHR DATLEN
Note:
1.
STTDLY is set to 0.
41.7.8 Loop Mode

The receiver can be programmed to receive transmissions from the transmitter. This is done by setting the Loop Mode (LOOP) bit in SSC_RFMR. In this case, RD is connected to TD, RF is connected to TF and RK is connected to TK.
41.7.9 Interrupt
Most bits in SSC_SR have a corresponding bit in interrupt management registers. The SSC can be programmed to generate an interrupt when it detects an event. The interrupt is controlled by writing SSC_IER (Interrupt Enable Register) and SSC_IDR (Interrupt Disable Register) These registers enable and disable, respectively, the corresponding interrupt by setting and clearing the corresponding bit in SSC_IMR (Interrupt Mask Register), which controls the generation of interrupts by asserting the SSC interrupt line connected to the interrupt controller.

1194
Figure 41-16. Interrupt Block Diagram

SSC_IMR SSC_IER Set SSC_IDR Clear
Transmitter TXRDY TXEMPTY TXSYNC Interrupt Control
SSC Interrupt
Receiver RXRDY OVRUN RXSYNC

1195
41.8
SSC Application Examples

The SSC can support several serial communication modes used in audio or high speed serial links. Some standard applications are shown in the following figures. All serial link applications supported by the SSC are not listed here.
Figure 41-17. Audio Application Block Diagram

Clock SCK TK Word Select WS TF Data SD SSC TD RD RF RK I2S RECEIVER
Clock SCK Word Select WS
Data SD
MSB Left Channel
LSB
MSB Right Channel
Figure 41-18. Codec Application Block Diagram

Serial Data Clock (SCLK) TK Frame sync (FSYNC) TF Serial Data Out SSC TD Serial Data In RD RF RK CODEC
Serial Data Clock (SCLK) Frame sync (FSYNC) First Time Slot Dstart Serial Data Out Dend
Serial Data In

1196
Figure 41-19. Time Slot Application Block Diagram

SCLK TK FSYNC TF Data Out TD SSC RD RF RK Data in CODEC First Time Slot
CODEC Second Time Slot
Serial Data Clock (SCLK) Frame sync (FSYNC) Serial Data Out First Time Slot Dstart Second Time Slot Dend
Serial Data in

1197

To prevent any single software error that may corrupt SSC behavior, certain address spaces can be write-protected by setting the WPEN bit in the SSC Write Protect Mode Register (SSC_WPMR). If a write access to the protected registers is detected, then the WPVS flag in the SSC Write Protect Status Register (US_WPSR) is set and the field WPVSRC indicates in which register the write access has been attempted. The WPVS flag is reset by writing the SSC Write Protect Mode Register (SSC_WPMR) with the appropriate access key, WPKEY. The protected registers are:
z z z z z z z
SSC Clock Mode Register on page 1201 SSC Receive Clock Mode Register on page 1202 SSC Receive Frame Mode Register on page 1204 SSC Transmit Clock Mode Register on page 1206 SSC Transmit Frame Mode Register on page 1208 SSC Receive Compare 0 Register on page 1214 SSC Receive Compare 1 Register on page 1215

1198
41.9
Synchronous Serial Controller (SSC) User Interface
Table 41-5. Register Mapping Offset 0x0 0x4 0x8 0xC 0x10 0x14 0x18 0x1C 0x20 0x24 0x28 0x2C 0x30 0x34 0x38 0x3C 0x40 0x44 0x48 0x4C 0xE4 0xE8 0x50-0xFC 0x100-0x124 Register Control Register Clock Mode Register Reserved Reserved Receive Clock Mode Register Receive Frame Mode Register Transmit Clock Mode Register Transmit Frame Mode Register Receive Holding Register Transmit Holding Register Reserved Reserved Receive Sync. Holding Register Transmit Sync. Holding Register Receive Compare 0 Register Receive Compare 1 Register Status Register Interrupt Enable Register Interrupt Disable Register Interrupt Mask Register Write Protect Mode Register Write Protect Status Register Reserved Reserved Name SSC_CR SSC_CMR SSC_RCMR SSC_RFMR SSC_TCMR SSC_TFMR SSC_RHR SSC_THR SSC_RSHR SSC_TSHR SSC_RC0R SSC_RC1R SSC_SR SSC_IER SSC_IDR SSC_IMR SSC_WPMR SSC_WPSR Access Write-only Read-write Read-write Read-write Read-write Read-write Read-only Write-only Read-only Read-write Read-write Read-write Read-only Write-only Write-only Read-only Read-write Read-only Reset 0x0 0x0 0x0 0x0 0x0 0x0 0x0 0x0 0x0 0x0 0x000000CC 0x0 0x0 0x0

1199
41.9.1 SSC Control Register

31 23 15 SWRST 7
SSC_CR: 0xF0008000 (0), 0xF800C000 (1) Write-only

30 22 14 6 29 21 13 5 28 20 12 4 27 19 11 3 26 18 10 2 25 17 9 TXDIS 1 RXDIS 24 16 8 TXEN 0 RXEN

0 = No effect. 1 = Enables Receive if RXDIS is not set.
RXDIS: Receive Disable

0 = No effect. 1 = Disables Receive. If a character is currently being received, disables at end of current character reception.

0 = No effect. 1 = Enables Transmit if TXDIS is not set.
TXDIS: Transmit Disable

0 = No effect. 1 = Disables Transmit. If a character is currently being transmitted, disables at end of current character transmission.

0 = No effect. 1 = Performs a software reset. Has priority on any other bit in SSC_CR.

1200
41.9.2 SSC Clock Mode Register

31 23 15 7
SSC_CMR 0xF0008004 (0), 0xF800C004 (1) Read-write

30 22 14 6 29 21 13 5 28 20 12 4 DIV 27 19 11 26 18 10 DIV 3 2 1 0 25 17 9 24 16 8
This register can only be written if the WPEN bit is cleared in SSC Write Protect Mode Register .
DIV: Clock Divider

0 = The Clock Divider is not active. Any Other Value: The Divided Clock equals the Master Clock divided by 2 times DIV. The maximum bit rate is MCK/2. The minimum bit rate is MCK/2 x 4095 = MCK/8190.

1201
41.9.3 SSC Receive Clock Mode Register

31
SSC_RCMR 0xF0008010 (0), 0xF800C010 (1) Read-write

30 29 28 PERIOD 27 26 25 24
23
22
21
20 STTDLY
19
18
17
16
15 7 CKG
14 6
13 5 CKI
12 STOP 4
11
10 START
3 CKO
1 CKS
CKS: Receive Clock Selection

Value 0 1 2 Name MCK TK RK Description Divided Clock TK Clock signal RK pin
CKO: Receive Clock Output Mode Selection

Value 0 1 2 Name NONE CONTINUOUS TRANSFER Description None, RK pin is an input Continuous Receive Clock, RK pin is an output Receive Clock only during data transfers, RK pin is an output
CKI: Receive Clock Inversion

0 = The data inputs (Data and Frame Sync signals) are sampled on Receive Clock falling edge. The Frame Sync signal output is shifted out on Receive Clock rising edge. 1 = The data inputs (Data and Frame Sync signals) are sampled on Receive Clock rising edge. The Frame Sync signal output is shifted out on Receive Clock falling edge. CKI affects only the Receive Clock and not the output clock signal.

1202
CKG: Receive Clock Gating Selection

Value 0 1 2 Name CONTINUOUS EN_RF_LOW EN_RF_HIGH Description None Receive Clock enabled only if RF Pin is Low Receive Clock enabled only if RF Pin is High
START: Receive Start Selection

Value 0 1 2 3 4 5 6 7 8 Name CONTINUOUS TRANSMIT RF_LOW RF_HIGH RF_FALLING RF_RISING RF_LEVEL RF_EDGE CMP_0 Description Continuous, as soon as the receiver is enabled, and immediately after the end of transfer of the previous data. Transmit start Detection of a low level on RF signal Detection of a high level on RF signal Detection of a falling edge on RF signal Detection of a rising edge on RF signal Detection of any level change on RF signal Detection of any edge on RF signal Compare 0
STOP: Receive Stop Selection

0 = After completion of a data transfer when starting with a Compare 0, the receiver stops the data transfer and waits for a new compare 0. 1 = After starting a receive with a Compare 0, the receiver operates in a continuous mode until a Compare 1 is detected.
STTDLY: Receive Start Delay

If STTDLY is not 0, a delay of STTDLY clock cycles is inserted between the start event and the actual start of reception. When the Receiver is programmed to start synchronously with the Transmitter, the delay is also applied. Note: It is very important that STTDLY be set carefully. If STTDLY must be set, it should be done in relation to TAG (Receive Sync Data) reception.
PERIOD: Receive Period Divider Selection

This field selects the divider to apply to the selected Receive Clock in order to generate a new Frame Sync Signal. If 0, no PERIOD signal is generated. If not 0, a PERIOD signal is generated each 2 x (PERIOD+1) Receive Clock.

1203
41.9.4 SSC Receive Frame Mode Register

31
SSC_RFMR 0xF0008014 (0), 0xF800C014 (1) Read-write

30 FSLEN_EXT 29 28 27 19 26 18 FSLEN 12 4 11 10 DATNB 3 2 DATLEN 1 0 9 8 25 17 24 FSEDGE 16
23 15 7 MSBF
22
21 FSOS 13 5 LOOP
20
14 6
DATLEN: Data Length

0 = Forbidden value (1-bit data length not supported). Any other value: The bit stream contains DATLEN + 1 data bits.
LOOP: Loop Mode

0 = Normal operating mode. 1 = RD is driven by TD, RF is driven by TF and TK drives RK.
MSBF: Most Significant Bit First

0 = The lowest significant bit of the data register is sampled first in the bit stream. 1 = The most significant bit of the data register is sampled first in the bit stream.
DATNB: Data Number per Frame

This field defines the number of data words to be received after each transfer start, which is equal to (DATNB + 1).
FSLEN: Receive Frame Sync Length

This field defines the number of bits sampled and stored in the Receive Sync Data Register. When this mode is selected by the START field in the Receive Clock Mode Register, it also determines the length of the sampled data to be compared to the Compare 0 or Compare 1 register. This field is used with FSLEN_EXT to determine the pulse length of the Receive Frame Sync signal. Pulse length is equal to FSLEN + (FSLEN_EXT * 16) + 1 Receive Clock periods.

1204
FSOS: Receive Frame Sync Output Selection

Value 0 1 2 3 4 5 Name NONE NEGATIVE POSITIVE LOW HIGH TOGGLING Description None, RF pin is an input Negative Pulse, RF pin is an output Positive Pulse, RF pin is an output Driven Low during data transfer, RF pin is an output Driven High during data transfer, RF pin is an output Toggling at each start of data transfer, RF pin is an output
FSEDGE: Frame Sync Edge Detection

Determines which edge on Frame Sync will generate the interrupt RXSYN in the SSC Status Register.
Value 0 1 Name POSITIVE NEGATIVE Description Positive Edge Detection Negative Edge Detection
FSLEN_EXT: FSLEN Field Extension

Extends FSLEN field. For details, refer to FSLEN bit description on page 1204.

1205
41.9.5 SSC Transmit Clock Mode Register

31
SSC_TCMR 0xF0008018 (0), 0xF800C018 (1) Read-write

30 29 28 PERIOD 27 26 25 24
23
22
21
20 STTDLY
19
18
17
16
15 7 CKG
14 6
13 5 CKI
12 4
11
10 START
3 CKO
1 CKS
CKS: Transmit Clock Selection

Value 0 1 2 Name MCK RK TK Description Divided Clock RK Clock signal TK pin
CKO: Transmit Clock Output Mode Selection

Value 0 1 2 Name NONE CONTINUOUS TRANSFER Description None, TK pin is an input Continuous Transmit Clock, TK pin is an output Transmit Clock only during data transfers, TK pin is an output
CKI: Transmit Clock Inversion

0 = The data outputs (Data and Frame Sync signals) are shifted out on Transmit Clock falling edge. The Frame sync signal input is sampled on Transmit clock rising edge. 1 = The data outputs (Data and Frame Sync signals) are shifted out on Transmit Clock rising edge. The Frame sync signal input is sampled on Transmit clock falling edge. CKI affects only the Transmit Clock and not the output clock signal.
CKG: Transmit Clock Gating Selection

Value 0 1 2 Name CONTINUOUS EN_TF_LOW EN_TF_HIGH Description None Transmit Clock enabled only if TF pin is Low Transmit Clock enabled only if TF pin is High

1206
START: Transmit Start Selection

Value 0 1 2 3 4 5 6 7 Name CONTINUOUS RECEIVE TF_LOW TF_HIGH TF_FALLING TF_RISING TF_LEVEL TF_EDGE Description Continuous, as soon as a word is written in the SSC_THR Register (if Transmit is enabled), and immediately after the end of transfer of the previous data. Receive start Detection of a low level on TF signal Detection of a high level on TF signal Detection of a falling edge on TF signal Detection of a rising edge on TF signal Detection of any level change on TF signal Detection of any edge on TF signal
STTDLY: Transmit Start Delay

If STTDLY is not 0, a delay of STTDLY clock cycles is inserted between the start event and the actual start of transmission of data. When the Transmitter is programmed to start synchronously with the Receiver, the delay is also applied. Note: STTDLY must be set carefully. If STTDLY is too short in respect to TAG (Transmit Sync Data) emission, data is emitted instead of the end of TAG.
PERIOD: Transmit Period Divider Selection

This field selects the divider to apply to the selected Transmit Clock to generate a new Frame Sync Signal. If 0, no period signal is generated. If not 0, a period signal is generated at each 2 x (PERIOD+1) Transmit Clock.

1207
41.9.6 SSC Transmit Frame Mode Register

31
SSC_TFMR 0xF000801C (0), 0xF800C01C (1) Read-write

30 FSLEN_EXT 29 28 27 19 26 18 FSLEN 12 4 11 10 DATNB 3 2 DATLEN 1 0 9 8 25 17 24 FSEDGE 16
23 FSDEN 15 7 MSBF
22
21 FSOS 13 5 DATDEF
20
14 6
DATLEN: Data Length

0 = Forbidden value (1-bit data length not supported). Any other value: The bit stream contains DATLEN + 1 data bits.
DATDEF: Data Default Value

This bit defines the level driven on the TD pin while out of transmission. Note that if the pin is defined as multi-drive by the PIO Controller, the pin is enabled only if the SCC TD output is 1.
MSBF: Most Significant Bit First

0 = The lowest significant bit of the data register is shifted out first in the bit stream. 1 = The most significant bit of the data register is shifted out first in the bit stream.
DATNB: Data Number per frame

This field defines the number of data words to be transferred after each transfer start, which is equal to (DATNB +1).
FSLEN: Transmit Frame Sync Length

This field defines the length of the Transmit Frame Sync signal and the number of bits shifted out from the Transmit Sync Data Register if FSDEN is 1. This field is used with FSLEN_EXT to determine the pulse length of the Transmit Frame Sync signal. Pulse length is equal to FSLEN + (FSLEN_EXT * 16) + 1 Transmit Clock period.

1208
FSOS: Transmit Frame Sync Output Selection

Value 0 1 2 3 4 5 Name NONE NEGATIVE POSITIVE LOW HIGH TOGGLING Description None, TF pin is an input Negative Pulse, TF pin is an output Positive Pulse, TF pin is an output TF pin Driven Low during data transfer TF pin Driven High during data transfer TF pin Toggles at each start of data transfer
FSDEN: Frame Sync Data Enable

0 = The TD line is driven with the default value during the Transmit Frame Sync signal. 1 = SSC_TSHR value is shifted out during the transmission of the Transmit Frame Sync signal.
FSEDGE: Frame Sync Edge Detection

Determines which edge on frame sync will generate the interrupt TXSYN (Status Register).
Value 0 1 Name POSITIVE NEGATIVE Description Positive Edge Detection Negative Edge Detection
FSLEN_EXT: FSLEN Field Extension

Extends FSLEN field. For details, refer to FSLEN bit description on page 1208.

1209
41.9.7 SSC Receive Holding Register

31
SSC_RHR 0xF0008020 (0), 0xF800C020 (1) Read-only

30 29 28 RDAT 27 26 25 24
23
22
21
20 RDAT
19
18
17
16
15
14
13
12 RDAT
11
10
4 RDAT
RDAT: Receive Data

Right aligned regardless of the number of data bits defined by DATLEN in SSC_RFMR.

1210
41.9.8 SSC Transmit Holding Register

31
SSC_THR 0xF0008024 (0), 0xF800C024 (1) Write-only

30 29 28 TDAT 27 26 25 24
23
22
21
20 TDAT
19
18
17
16
15
14
13
12 TDAT
11
10
4 TDAT
TDAT: Transmit Data

Right aligned regardless of the number of data bits defined by DATLEN in SSC_TFMR.

1211
41.9.9 SSC Receive Synchronization Holding Register

31 23 15
SSC_RSHR 0xF0008030 (0), 0xF800C030 (1) Read-only

30 22 14 29 21 13 28 20 12 RSDAT 7 6 5 4 RSDAT 3 2 1 0 27 19 11 26 18 10 25 17 9 24 16 8
RSDAT: Receive Synchronization Data

1212
41.9.10 SSC Transmit Synchronization Holding Register

31 23 15
SSC_TSHR 0xF0008034 (0), 0xF800C034 (1) Read-write

30 22 14 29 21 13 28 20 12 TSDAT 7 6 5 4 TSDAT 3 2 1 0 27 19 11 26 18 10 25 17 9 24 16 8
TSDAT: Transmit Synchronization Data

1213
41.9.11 SSC Receive Compare 0 Register

31 23 15
SSC_RC0R 0xF0008038 (0), 0xF800C038 (1) Read-write

30 22 14 29 21 13 28 20 12 CP0 7 6 5 4 CP0 3 2 1 0 27 19 11 26 18 10 25 17 9 24 16 8
CP0: Receive Compare Data 0

1214
41.9.12 SSC Receive Compare 1 Register

31 23 15
SSC_RC1R 0xF000803C (0), 0xF800C03C (1) Read-write

30 22 14 29 21 13 28 20 12 CP1 7 6 5 4 CP1 3 2 1 0 27 19 11 26 18 10 25 17 9 24 16 8
CP1: Receive Compare Data 1

1215
41.9.13 SSC Status Register

31 23 15 7
SSC_SR 0xF0008040 (0), 0xF800C040 (1) Read-only

30 22 14 6 29 21 13 5 OVRUN 28 20 12 4 RXRDY 27 19 11 RXSYN 3 26 18 10 TXSYN 2 25 17 RXEN 9 CP1 1 TXEMPTY 24 16 TXEN 8 CP0 0 TXRDY
TXRDY: Transmit Ready

0 = Data has been loaded in SSC_THR and is waiting to be loaded in the Transmit Shift Register (TSR). 1 = SSC_THR is empty.
TXEMPTY: Transmit Empty

0 = Data remains in SSC_THR or is currently transmitted from TSR. 1 = Last data written in SSC_THR has been loaded in TSR and last data loaded in TSR has been transmitted.
RXRDY: Receive Ready

0 = SSC_RHR is empty. 1 = Data has been received and loaded in SSC_RHR.
OVRUN: Receive Overrun

0 = No data has been loaded in SSC_RHR while previous data has not been read since the last read of the Status Register. 1 = Data has been loaded in SSC_RHR while previous data has not yet been read since the last read of the Status Register.
CP0: Compare 0
0 = A compare 0 has not occurred since the last read of the Status Register. 1 = A compare 0 has occurred since the last read of the Status Register.
CP1: Compare 1
0 = A compare 1 has not occurred since the last read of the Status Register. 1 = A compare 1 has occurred since the last read of the Status Register.
TXSYN: Transmit Sync

0 = A Tx Sync has not occurred since the last read of the Status Register. 1 = A Tx Sync has occurred since the last read of the Status Register.
RXSYN: Receive Sync

0 = An Rx Sync has not occurred since the last read of the Status Register. 1 = An Rx Sync has occurred since the last read of the Status Register.

1216

0 = Transmit is disabled. 1 = Transmit is enabled.

0 = Receive is disabled. 1 = Receive is enabled.

1217
41.9.14 SSC Interrupt Enable Register

31 23 15 7
SSC_IER 0xF0008044 (0), 0xF800C044 (1) Write-only

30 22 14 6 29 21 13 5 OVRUN 28 20 12 4 RXRDY 27 19 11 RXSYN 3 26 18 10 TXSYN 2 25 17 9 CP1 1 TXEMPTY 24 16 8 CP0 0 TXRDY
TXRDY: Transmit Ready Interrupt Enable

0 = 0 = No effect. 1 = Enables the Transmit Ready Interrupt.
TXEMPTY: Transmit Empty Interrupt Enable

0 = No effect. 1 = Enables the Transmit Empty Interrupt.
RXRDY: Receive Ready Interrupt Enable

0 = No effect. 1 = Enables the Receive Ready Interrupt.
OVRUN: Receive Overrun Interrupt Enable

0 = No effect. 1 = Enables the Receive Overrun Interrupt.
CP0: Compare 0 Interrupt Enable

0 = No effect. 1 = Enables the Compare 0 Interrupt.
CP1: Compare 1 Interrupt Enable

0 = No effect. 1 = Enables the Compare 1 Interrupt.
TXSYN: Tx Sync Interrupt Enable

0 = No effect. 1 = Enables the Tx Sync Interrupt.
RXSYN: Rx Sync Interrupt Enable

0 = No effect. 1 = Enables the Rx Sync Interrupt.

1218
41.9.15 SSC Interrupt Disable Register

31 23 15 7
SSC_IDR 0xF0008048 (0), 0xF800C048 (1) Write-only

TXRDY: Transmit Ready Interrupt Disable

0 = No effect. 1 = Disables the Transmit Ready Interrupt.
TXEMPTY: Transmit Empty Interrupt Disable

0 = No effect. 1 = Disables the Transmit Empty Interrupt.
RXRDY: Receive Ready Interrupt Disable

0 = No effect. 1 = Disables the Receive Ready Interrupt.
OVRUN: Receive Overrun Interrupt Disable

0 = No effect. 1 = Disables the Receive Overrun Interrupt.
CP0: Compare 0 Interrupt Disable

0 = No effect. 1 = Disables the Compare 0 Interrupt.
CP1: Compare 1 Interrupt Disable

0 = No effect. 1 = Disables the Compare 1 Interrupt.
TXSYN: Tx Sync Interrupt Enable

0 = No effect. 1 = Disables the Tx Sync Interrupt.
RXSYN: Rx Sync Interrupt Enable

0 = No effect. 1 = Disables the Rx Sync Interrupt.

1219
41.9.16 SSC Interrupt Mask Register

31 23 15 7
SSC_IMR 0xF000804C (0), 0xF800C04C (1) Read-only

TXRDY: Transmit Ready Interrupt Mask

0 = The Transmit Ready Interrupt is disabled. 1 = The Transmit Ready Interrupt is enabled.
TXEMPTY: Transmit Empty Interrupt Mask

0 = The Transmit Empty Interrupt is disabled. 1 = The Transmit Empty Interrupt is enabled.
RXRDY: Receive Ready Interrupt Mask

0 = The Receive Ready Interrupt is disabled. 1 = The Receive Ready Interrupt is enabled.
OVRUN: Receive Overrun Interrupt Mask

0 = The Receive Overrun Interrupt is disabled. 1 = The Receive Overrun Interrupt is enabled.
CP0: Compare 0 Interrupt Mask

0 = The Compare 0 Interrupt is disabled. 1 = The Compare 0 Interrupt is enabled.
CP1: Compare 1 Interrupt Mask

0 = The Compare 1 Interrupt is disabled. 1 = The Compare 1 Interrupt is enabled.
TXSYN: Tx Sync Interrupt Mask

0 = The Tx Sync Interrupt is disabled. 1 = The Tx Sync Interrupt is enabled.
RXSYN: Rx Sync Interrupt Mask

0 = The Rx Sync Interrupt is disabled. 1 = The Rx Sync Interrupt is enabled.

1220
41.9.17 SSC Write Protect Mode Register

31
SSC_WPMR 0xF00080E4 (0), 0xF800C0E4 (1) Read-write See Table 41-5


0 = Disables the Write Protect if WPKEY corresponds to 0x535343 (SSC in ASCII). 1 = Enables the Write Protect if WPKEY corresponds to 0x535343 (SSC in ASCII). Protects the registers:
SSC Clock Mode Register on page 1201 SSC Receive Clock Mode Register on page 1202 SSC Receive Frame Mode Register on page 1204 SSC Transmit Clock Mode Register on page 1206 SSC Transmit Frame Mode Register on page 1208 SSC Receive Compare 0 Register on page 1214 SSC Receive Compare 1 Register on page 1215 WPKEY: Write Protect KEY
Should be written at value 0x535343 (SSC in ASCII). Writing any other value in this field aborts the write operation of the WPEN bit. Always reads as 0.

1221
41.9.18 SSC Write Protect Status Register

31 23
SSC_WPSR 0xF00080E8 (0), 0xF800C0E8 (1) Read-only See Table 41-5


0 = No Write Protect Violation has occurred since the last read of the SSC_WPSR register. 1 = A Write Protect Violation has occurred since the last read of the SSC_WPSR register. If this violation is an unauthorized attempt to write a protected register, the associated violation is reported into field WPVSRC.

When WPVS is active, this field indicates the write-protected register (through address offset or code) in which a write access has been attempted. Note: Reading SSC_WPSR automatically clears all fields.

1222
42.
42.1
Debug Unit (DBGU)

Description
The Debug Unit provides a single entry point from the processor for access to all the debug capabilities of Atmels ARMbased systems. The Debug Unit features a two-pin UART that can be used for several debug and trace purposes and offers an ideal medium for in-situ programming solutions and debug monitor communications. The Debug Unit two-pin UART can be used stand-alone for general purpose serial communication. Moreover, the association with DMA controller channels permits packet handling for these tasks with processor time reduced to a minimum. The Debug Unit also makes the Debug Communication Channel (DCC) signals provided by the In-circuit Emulator of the ARM processor visible to the software. These signals indicate the status of the DCC read and write registers and generate an interrupt to the ARM processor, making possible the handling of the DCC under interrupt control. Chip Identifier registers permit recognition of the device and its revision. These registers inform as to the sizes and types of the on-chip memories, as well as the set of embedded peripherals. Finally, the Debug Unit features a Force NTRST capability that enables the software to decide whether to prevent access to the system via the In-circuit Emulator. This permits protection of the code, stored in ROM.
42.2
z z
System Peripheral to Facilitate Debug of Atmel ARM-based Systems Composed of Four Functions
z z z z
Two-pin UART Debug Communication Channel (DCC) Support Chip ID Registers ICE Access Prevention Implemented Features are USART Compatible Independent Receiver and Transmitter with a Common Programmable Baud Rate Generator Even, Odd, Mark or Space Parity Generation Parity, Framing and Overrun Error Detection Automatic Echo, Local Loopback and Remote Loopback Channel Modes Interrupt Generation Support for Two DMA Channels with Connection to Receiver and Transmitter Offers Visibility of COMMRX and COMMTX Signals from the ARM Processor Interrupt Generation Identification of the Device Revision, Sizes of the Embedded Memories, Set of Peripherals Enables Software to Prevent System Access Through the ARM Processors ICE Prevention is Made by Asserting the NTRST Line of the ARM Processors ICE
Two-pin UART
z z z z z z z
Debug Communication Channel Support

z z
Chip ID Registers
z
ICE Access Prevention

z z

1223
42.3
Block Diagram
Figure 42-1. Debug Unit Functional Block Diagram

Peripheral Bridge (Peripheral) DMA Controller
APB
Debug Unit
DTXD
Transmit Power Management Controller

MCK
Baud Rate Generator Receive
Parallel Input/ Output

DRXD
COMMRX ARM Processor

nTRST
COMMTX
DCC Handler
Chip ID
ICE Access Handler Power-on Reset

force_ntrst
Interrupt Control
dbgu_irq
Table 42-1. Debug Unit Pin Description Pin Name DRXD DTXD Description Debug Receive Data Debug Transmit Data Type Input Output
Figure 42-2. Debug Unit Application Example

Boot Program Debug Monitor Trace Manager
Debug Unit
RS232 Drivers Programming Tool Debug Console Trace Console

1224
42.4
42.4.1 I/O Lines

Depending on product integration, the Debug Unit pins may be multiplexed with PIO lines. In this case, the programmer must first configure the corresponding PIO Controller to enable I/O lines operations of the Debug Unit.
Table 42-2. I/O Lines Instance DBGU DBGU Signal DRXD DTXD I/O Line PB30 PB31 Peripheral A A

Depending on product integration, the Debug Unit clock may be controllable through the Power Management Controller. In this case, the programmer must first configure the PMC to enable the Debug Unit clock. Usually, the peripheral identifier used for this purpose is 1.
42.4.3 Interrupt Source

Depending on product integration, the Debug Unit interrupt line is connected to one of the interrupt sources of the Advanced Interrupt Controller. Interrupt handling requires programming of the AIC before configuring the Debug Unit. Usually, the Debug Unit interrupt line connects to the interrupt source 1 of the AIC, which may be shared with the realtime clock, the system timer interrupt lines and other system peripheral interrupts, as shown in Figure 42-1. This sharing requires the programmer to determine the source of the interrupt when the source 1 is triggered.
42.5
UART Operations
The Debug Unit operates as a UART, (asynchronous mode only) and supports only 8-bit character handling (with parity). It has no clock pin. The Debug Unit's UART is made up of a receiver and a transmitter that operate independently, and a common baud rate generator. Receiver timeout and transmitter time guard are not implemented. However, all the implemented features are compatible with those of a standard USART.
42.5.1 Baud Rate Generator

The baud rate generator provides the bit period clock named baud rate clock to both the receiver and the transmitter. The baud rate clock is the master clock divided by 16 times the value (CD) written in DBGU_BRGR (Baud Rate Generator Register). If DBGU_BRGR is set to 0, the baud rate clock is disabled and the Debug Unit's UART remains inactive. The maximum allowable baud rate is Master Clock divided by 16. The minimum allowable baud rate is Master Clock divided by (16 x 65536).
MCK Baud Rate = --------------------16 CD

1225
Figure 42-3. Baud Rate Generator

CD CD MCK 16-bit Counter
OUT
>1 1 0 0 Receiver Sampling Clock Divide by 16 Baud Rate Clock
42.5.2 Receiver
42.5.2.1 Receiver Reset, Enable and Disable After device reset, the Debug Unit receiver is disabled and must be enabled before being used. The receiver can be enabled by writing the control register DBGU_CR with the bit RXEN at 1. At this command, the receiver starts looking for a start bit. The programmer can disable the receiver by writing DBGU_CR with the bit RXDIS at 1. If the receiver is waiting for a start bit, it is immediately stopped. However, if the receiver has already detected a start bit and is receiving the data, it waits for the stop bit before actually stopping its operation. The programmer can also put the receiver in its reset state by writing DBGU_CR with the bit RSTRX at 1. In doing so, the receiver immediately stops its current operations and is disabled, whatever its current state. If RSTRX is applied when data is being processed, this data is lost. 42.5.2.2 Start Detection and Data Sampling The Debug Unit only supports asynchronous operations, and this affects only its receiver. The Debug Unit receiver detects the start of a received character by sampling the DRXD signal until it detects a valid start bit. A low level (space) on DRXD is interpreted as a valid start bit if it is detected for more than 7 cycles of the sampling clock, which is 16 times the baud rate. Hence, a space that is longer than 7/16 of the bit period is detected as a valid start bit. A space which is 7/16 of a bit period or shorter is ignored and the receiver continues to wait for a valid start bit. When a valid start bit has been detected, the receiver samples the DRXD at the theoretical midpoint of each bit. It is assumed that each bit lasts 16 cycles of the sampling clock (1-bit period) so the bit sampling point is eight cycles (0.5-bit period) after the start of the bit. The first sampling point is therefore 24 cycles (1.5-bit periods) after the falling edge of the start bit was detected. Each subsequent bit is sampled 16 cycles (1-bit period) after the previous one.
Figure 42-4. Start Bit Detection
Sampling Clock
DRXD
True Start Detection Baud Rate Clock
D0

1226
Figure 42-5. Character Reception

Example: 8-bit, parity enabled 1 stop
0.5 bit period 1 bit period
DRXD
Sampling
D0 D1 True Start Detection
D2
D3
D4
D5
D6
D7 Parity Bit
Stop Bit
42.5.2.3 Receiver Ready When a complete character is received, it is transferred to the DBGU_RHR and the RXRDY status bit in DBGU_SR (Status Register) is set. The bit RXRDY is automatically cleared when the receive holding register DBGU_RHR is read.
Figure 42-6. Receiver Ready
DRXD S D0 D1 D2 D3 D4 D5 D6 D7 P S D0 D1 D2 D3 D4 D5 D6 D7 P
RXRDY
Read DBGU_RHR
42.5.2.4 Receiver Overrun If DBGU_RHR has not been read by the software (or the Peripheral Data Controller or DMA Controller) since the last transfer, the RXRDY bit is still set and a new character is received, the OVRE status bit in DBGU_SR is set. OVRE is cleared when the software writes the control register DBGU_CR with the bit RSTSTA (Reset Status) at 1.
Figure 42-7. Receiver Overrun
DRXD S D0 D1 D2 D3 D4 D5 D6 D7 P stop S D0 D1 D2 D3 D4 D5 D6 D7 P stop
RXRDY OVRE
RSTSTA
42.5.2.5 Parity Error Each time a character is received, the receiver calculates the parity of the received data bits, in accordance with the field PAR in DBGU_MR. It then compares the result with the received parity bit. If different, the parity error bit PARE in DBGU_SR is set at the same time the RXRDY is set. The parity bit is cleared when the control register DBGU_CR is written with the bit RSTSTA (Reset Status) at 1. If a new character is received before the reset status command is written, the PARE bit remains at 1.
Figure 42-8. Parity Error
DRXD S D0 D1 D2 D3 D4 D5 D6 D7 P stop
RXRDY PARE
Wrong Parity Bit
RSTSTA

1227
42.5.2.6 Receiver Framing Error When a start bit is detected, it generates a character reception when all the data bits have been sampled. The stop bit is also sampled and when it is detected at 0, the FRAME (Framing Error) bit in DBGU_SR is set at the same time the RXRDY bit is set. The bit FRAME remains high until the control register DBGU_CR is written with the bit RSTSTA at 1.
Figure 42-9. Receiver Framing Error
DRXD S D0 D1 D2 D3 D4 D5 D6 D7 P stop
RXRDY FRAME
Stop Bit Detected at 0
RSTSTA
42.5.3 Transmitter
42.5.3.1 Transmitter Reset, Enable and Disable After device reset, the Debug Unit transmitter is disabled and it must be enabled before being used. The transmitter is enabled by writing the control register DBGU_CR with the bit TXEN at 1. From this command, the transmitter waits for a character to be written in the Transmit Holding Register DBGU_THR before actually starting the transmission. The programmer can disable the transmitter by writing DBGU_CR with the bit TXDIS at 1. If the transmitter is not operating, it is immediately stopped. However, if a character is being processed into the Shift Register and/or a character has been written in the Transmit Holding Register, the characters are completed before the transmitter is actually stopped. The programmer can also put the transmitter in its reset state by writing the DBGU_CR with the bit RSTTX at 1. This immediately stops the transmitter, whether or not it is processing characters. 42.5.3.2 Transmit Format The Debug Unit transmitter drives the pin DTXD at the baud rate clock speed. The line is driven depending on the format defined in the Mode Register and the data stored in the Shift Register. One start bit at level 0, then the 8 data bits, from the lowest to the highest bit, one optional parity bit and one stop bit at 1 are consecutively shifted out as shown on the following figure. The field PARE in the mode register DBGU_MR defines whether or not a parity bit is shifted out. When a parity bit is enabled, it can be selected between an odd parity, an even parity, or a fixed space or mark bit.
Figure 42-10. Character Transmission
Example: Parity enabled Baud Rate Clock DTXD
Start Bit
D0
D1
D2
D3
D4
D5
D6
D7
Parity Bit
Stop Bit

1228
42.5.3.3 Transmitter Control When the transmitter is enabled, the bit TXRDY (Transmitter Ready) is set in the status register DBGU_SR. The transmission starts when the programmer writes in the Transmit Holding Register DBGU_THR, and after the written character is transferred from DBGU_THR to the Shift Register. The bit TXRDY remains high until a second character is written in DBGU_THR. As soon as the first character is completed, the last character written in DBGU_THR is transferred into the shift register and TXRDY rises again, showing that the holding register is empty. When both the Shift Register and the DBGU_THR are empty, i.e., all the characters written in DBGU_THR have been processed, the bit TXEMPTY rises after the last stop bit has been completed.
Figure 42-11. Transmitter Control
DBGU_THR
Data 0 Data 1
Shift Register
Data 0
Data 1
DTXD
Data 0
stop
Data 1
stop
TXRDY TXEMPTY
Write Data 0 in DBGU_THR
Write Data 1 in DBGU_THR
42.5.4 DMA Support

Both the receiver and the transmitter of the Debug Units UART are connected to a DMA Controller (DMAC) channel. The DMA Controller channels are programmed via registers that are mapped within the DMAC user interface.
42.5.5 Test Modes

The Debug Unit supports three tests modes. These modes of operation are programmed by using the field CHMODE (Channel Mode) in the mode register DBGU_MR. The Automatic Echo mode allows bit-by-bit retransmission. When a bit is received on the DRXD line, it is sent to the DTXD line. The transmitter operates normally, but has no effect on the DTXD line. The Local Loopback mode allows the transmitted characters to be received. DTXD and DRXD pins are not used and the output of the transmitter is internally connected to the input of the receiver. The DRXD pin level has no effect and the DTXD line is held high, as in idle state. The Remote Loopback mode directly connects the DRXD pin to the DTXD line. The transmitter and the receiver are disabled and have no effect. This mode allows a bit-by-bit retransmission.

1229
Figure 42-12. Test Modes

Automatic Echo Receiver RXD
Transmitter
Disabled
TXD
Local Loopback Receiver

Disabled
RXD
VDD Transmitter
Disabled
TXD
Remote Loopback Receiver
VDD Disabled RXD
Transmitter
Disabled
TXD
42.5.6 Debug Communication Channel Support

The Debug Unit handles the signals COMMRX and COMMTX that come from the Debug Communication Channel of the ARM Processor and are driven by the In-circuit Emulator. The Debug Communication Channel contains two registers that are accessible through the ICE Breaker on the JTAG side and through the coprocessor 0 on the ARM Processor side. As a reminder, the following instructions are used to read and write the Debug Communication Channel: MRC p14, 0, Rd, c1, c0, 0 Returns the debug communication data read register into Rd MCR p14, 0, Rd, c1, c0, 0
Writes the value in Rd to the debug communication data write register. The bits COMMRX and COMMTX, which indicate, respectively, that the read register has been written by the debugger but not yet read by the processor, and that the write register has been written by the processor and not yet read by the debugger, are wired on the two highest bits of the status register DBGU_SR. These bits can generate an interrupt. This feature permits handling under interrupt a debug link between a debug monitor running on the target system and a debugger.

1230
42.5.7 Chip Identifier

The Debug Unit features two chip identifier registers, DBGU_CIDR (Chip ID Register) and DBGU_EXID (Extension ID). Both registers contain a hard-wired value that is read-only. The first register contains the following fields:
z z z z z z
EXT - shows the use of the extension identifier register NVPTYP and NVPSIZ - identifies the type of embedded non-volatile memory and its size ARCH - identifies the set of embedded peripherals SRAMSIZ - indicates the size of the embedded SRAM EPROC - indicates the embedded ARM processor VERSION - gives the revision of the silicon
The second register is device-dependent and reads 0 if the bit EXT is 0.
42.5.8 ICE Access Prevention

The Debug Unit allows blockage of access to the system through the ARM processor's ICE interface. This feature is implemented via the register Force NTRST (DBGU_FNR), that allows assertion of the NTRST signal of the ICE Interface. Writing the bit FNTRST (Force NTRST) to 1 in this register prevents any activity on the TAP controller. On standard devices, the bit FNTRST resets to 0 and thus does not prevent ICE access. This feature is especially useful on custom ROM devices for customers who do not want their on-chip code to be visible.

1231
42.6
Debug Unit (DBGU) User Interface
Table 42-3. Register Mapping Offset 0x0000 0x0004 0x0008 0x000C 0x0010 0x0014 0x0018 0x001C 0x0020 0x0024 - 0x003C 0x0040 0x0044 0x0048 0x004C - 0x00FC Register Control Register Mode Register Interrupt Enable Register Interrupt Disable Register Interrupt Mask Register Status Register Receive Holding Register Transmit Holding Register Baud Rate Generator Register Reserved Chip ID Register Chip ID Extension Register Force NTRST Register Reserved Name DBGU_CR DBGU_MR DBGU_IER DBGU_IDR DBGU_IMR DBGU_SR DBGU_RHR DBGU_THR DBGU_BRGR DBGU_CIDR DBGU_EXID DBGU_FNR Access Write-only Read-write Write-only Write-only Read-only Read-only Read-only Write-only Read-write Read-only Read-only Read-write Reset 0x0 0x0 0x0 0x0 0x0

1232
42.6.1 Debug Unit Control Register

31
DBGU_CR 0xFFFFEE00 Write-only

30 29 28 27 26 25 24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
8 RSTSTA 0
7 TXDIS
6 TXEN
5 RXDIS
4 RXEN
3 RSTTX
2 RSTRX
RSTRX: Reset Receiver

0 = No effect. 1 = The receiver logic is reset and disabled. If a character is being received, the reception is aborted.
RSTTX: Reset Transmitter

0 = No effect. 1 = The transmitter logic is reset and disabled. If a character is being transmitted, the transmission is aborted.
RXEN: Receiver Enable

0 = No effect. 1 = The receiver is enabled if RXDIS is 0.
RXDIS: Receiver Disable

0 = No effect. 1 = The receiver is disabled. If a character is being processed and RSTRX is not set, the character is completed before the receiver is stopped.
TXEN: Transmitter Enable

0 = No effect. 1 = The transmitter is enabled if TXDIS is 0.
TXDIS: Transmitter Disable

0 = No effect. 1 = The transmitter is disabled. If a character is being processed and a character has been written the DBGU_THR and RSTTX is not set, both characters are completed before the transmitter is stopped.
RSTSTA: Reset Status Bits

0 = No effect. 1 = Resets the status bits PARE, FRAME and OVRE in the DBGU_SR.

1233
42.6.2 Debug Unit Mode Register

31
DBGU_MR 0xFFFFEE04 Read-write

30 29 28 27 26 25 24
23
22
21
20
19
18
17
16
15 CHMODE 7
14
13
12
11
10 PAR
6 5
4 3
1 0
PAR: Parity Type

Value 0b000 0b001 0b010 0b011 0b1xx Name EVEN ODD SPACE MARK NONE Description Even Parity Odd Parity Space: Parity forced to 0 Mark: Parity forced to 1 No Parity
CHMODE: Channel Mode

Value 0b00 0b01 0b10 0b11 Name NORM AUTO LOCLOOP REMLOOP Description Normal Mode Automatic Echo Local Loopback Remote Loopback

1234
42.6.3 Debug Unit Interrupt Enable Register

31 COMMRX 23
DBGU_IER 0xFFFFEE08 Write-only

30 COMMTX 22 29 28 27 26 25 24
21
20
19
18
17
16
15
14
13
12
11
10
9 TXEMPTY 1 TXRDY
7 PARE
6 FRAME
5 OVRE
0 RXRDY
RXRDY: Enable RXRDY Interrupt TXRDY: Enable TXRDY Interrupt OVRE: Enable Overrun Error Interrupt FRAME: Enable Framing Error Interrupt PARE: Enable Parity Error Interrupt TXEMPTY: Enable TXEMPTY Interrupt COMMTX: Enable COMMTX (from ARM) Interrupt COMMRX: Enable COMMRX (from ARM) Interrupt

1235
42.6.4 Debug Unit Interrupt Disable Register

31 COMMRX 23
DBGU_IDR 0xFFFFEE0C Write-only

30 COMMTX 22 29 28 27 26 25 24
21
20
19
18
17
16
15
14
13
12
11
10
9 TXEMPTY 1 TXRDY
7 PARE
6 FRAME
5 OVRE
0 RXRDY
RXRDY: Disable RXRDY Interrupt TXRDY: Disable TXRDY Interrupt OVRE: Disable Overrun Error Interrupt FRAME: Disable Framing Error Interrupt PARE: Disable Parity Error Interrupt TXEMPTY: Disable TXEMPTY Interrupt COMMTX: Disable COMMTX (from ARM) Interrupt COMMRX: Disable COMMRX (from ARM) Interrupt

1236
42.6.5 Debug Unit Interrupt Mask Register

31 COMMRX 23
DBGU_IMR 0xFFFFEE10 Read-only

30 COMMTX 22 29 28 27 26 25 24
21
20
19
18
17
16
15
14
13
12
11
10
9 TXEMPTY 1 TXRDY
7 PARE
6 FRAME
5 OVRE
0 RXRDY
RXRDY: Mask RXRDY Interrupt TXRDY: Disable TXRDY Interrupt OVRE: Mask Overrun Error Interrupt FRAME: Mask Framing Error Interrupt PARE: Mask Parity Error Interrupt TXEMPTY: Mask TXEMPTY Interrupt COMMTX: Mask COMMTX Interrupt COMMRX: Mask COMMRX Interrupt

1237
42.6.6 Debug Unit Status Register

31 COMMRX 23
DBGU_SR 0xFFFFEE14 Read-only

30 COMMTX 22 29 28 27 26 25 24
21
20
19
18
17
16
15
14
13
12
11
10
9 TXEMPTY 1 TXRDY
7 PARE
6 FRAME
5 OVRE
0 RXRDY

0 = No character has been received since the last read of the DBGU_RHR or the receiver is disabled. 1 = At least one complete character has been received, transferred to DBGU_RHR and not yet read.
TXRDY: Transmitter Ready

0 = A character has been written to DBGU_THR and not yet transferred to the Shift Register, or the transmitter is disabled. 1 = There is no character written to DBGU_THR not yet transferred to the Shift Register.
OVRE: Overrun Error

0 = No overrun error has occurred since the last RSTSTA. 1 = At least one overrun error has occurred since the last RSTSTA.
FRAME: Framing Error

0 = No framing error has occurred since the last RSTSTA. 1 = At least one framing error has occurred since the last RSTSTA.
PARE: Parity Error

0 = No parity error has occurred since the last RSTSTA. 1 = At least one parity error has occurred since the last RSTSTA.
TXEMPTY: Transmitter Empty

0 = There are characters in DBGU_THR, or characters being processed by the transmitter, or the transmitter is disabled. 1 = There are no characters in DBGU_THR and there are no characters being processed by the transmitter.
COMMTX: Debug Communication Channel Write Status

0 = COMMTX from the ARM processor is inactive. 1 = COMMTX from the ARM processor is active.
COMMRX: Debug Communication Channel Read Status

0 = COMMRX from the ARM processor is inactive. 1 = COMMRX from the ARM processor is active.

1238
42.6.7 Debug Unit Receiver Holding Register

31
DBGU_RHR 0xFFFFEE18 Read-only

30 29 28 27 26 25 24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
4 RXCHR
RXCHR: Received Character

Last received character if RXRDY is set.

1239
42.6.8 Debug Unit Transmit Holding Register

31
DBGU_THR 0xFFFFEE1C Write-only

30 29 28 27 26 25 24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
4 TXCHR
TXCHR: Character to be Transmitted

Next character to be transmitted after the current character if TXRDY is not set.

1240
42.6.9 Debug Unit Baud Rate Generator Register

31
DBGU_BRGR 0xFFFFEE20 Read-write

30 29 28 27 26 25 24
23
22
21
20
19
18
17
16
15
14
13
12 CD
11
10
4 CD
CD: Clock Divisor

Value 0 1 2 to 65535 Name DISABLED MCK Description DBGU Disabled MCK MCK / (CD x 16)

1241
42.6.10 Debug Unit Chip ID Register

31 EXT 23 22 ARCH 15 14 NVPSIZ2 7 6 EPROC 5 4 3 2 VERSION 13 12 11 10 NVPSIZ 1 0
DBGU_CIDR 0xFFFFEE40 Read-only

30 29 NVPTYP 21 20 19 18 SRAMSIZ 9 8 28 27 26 ARCH 17 16 25 24
VERSION: Version of the Device

Values depend upon the version of the device.
EPROC: Embedded Processor

Value 6 Name CA5 Description Cortex-A5
NVPSIZ: Nonvolatile Program Memory Size

Value 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Name NONE 8K 16K 32K 64K 128K 256K 512K 1024K 2048K Description None 8 Kbytes 16 Kbytes 32 Kbytes Reserved 64 Kbytes Reserved 128 Kbytes Reserved 256 Kbytes 512 Kbytes Reserved 1024 Kbytes Reserved 2048 Kbytes Reserved

1242
NVPSIZ2 Second Nonvolatile Program Memory Size

Value 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 128K 256K 512K 1024K 2048K Name NONE 8K 16K 32K 64K Description None 8 Kbytes 16 Kbytes 32 Kbytes Reserved 64 Kbytes Reserved 128 Kbytes Reserved 256 Kbytes 512 Kbytes Reserved 1024 Kbytes Reserved 2048 Kbytes Reserved
SRAMSIZ: Internal SRAM Size

Value 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Name 1K 2K 6K 112K 4K 80K 160K 8K 16K 32K 64K 128K 256K 96K 512K Description Reserved 1 Kbytes 2 Kbytes 6 Kbytes 112 Kbytes 4 Kbytes 80 Kbytes 160 Kbytes 8 Kbytes 16 Kbytes 32 Kbytes 64 Kbytes 128 Kbytes 256 Kbytes 96 Kbytes 512 Kbytes

1243
ARCH: Architecture Identifier

Value 0xA5 Name ATSAMA5xx Description ATSAMA5xx Series
NVPTYP: Nonvolatile Program Memory Type

Value 0 1 4 2 3 Name ROM ROMLESS SRAM FLASH ROM_FLASH Description ROM ROMless or on-chip Flash SRAM emulating ROM Embedded Flash Memory ROM and Embedded Flash Memory NVPSIZ is ROM size NVPSIZ2 is Flash size
EXT: Extension Flag

0 = Chip ID has a single register definition without extension 1 = An extended Chip ID exists.

1244
42.6.11 Debug Unit Chip ID Extension Register

31
DBGU_EXID 0xFFFFEE44 Read-only

30 29 28 EXID 23 22 21 20 EXID 15 14 13 12 EXID 7 6 5 4 EXID 3 2 1 0 11 10 9 8 19 18 17 16 27 26 25 24
EXID: Chip ID Extension

Reads 0 if the bit EXT in DBGU_CIDR is 0.

1245
42.6.12 Debug Unit Force NTRST Register

31
DBGU_FNR 0xFFFFEE48 Read-write

30 29 28 27 26 25 24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
0 FNTRST
FNTRST: Force NTRST

0 = NTRST of the ARM processors TAP controller is driven by the power_on_reset signal. 1 = NTRST of the ARM processors TAP controller is held low.

1246
43.
43.1
True Random Number Generator (TRNG)

Description
The True Random Number Generator (TRNG) passes the American NIST Special Publication 800-22 and Diehard Random Tests Suites. The TRNG may be used as an entropy source for seeding an NIST approved DRNG (Deterministic RNG) as required by FIPS PUB 140-2 and 140-3.
43.2
z z z z
Passed NIST Special Publication 800-22 Tests Suite Passed Diehard Random Tests Suite May be used as Entropy Source for seeding an NIST approved DRNG (Deterministic RNG) as required by FIPS PUB 140-2 and 140-3 Provides a 32-bit Random Number Every 84 Clock Cycles

1247
43.3
Block Diagram
Figure 43-1. TRNG Block Diagram

Interrupt Controller TRNG Control Logic
MCK PMC
User Interface
Entropy Source
APB

1248
43.4

The TRNG interface may be clocked through the Power Management Controller (PMC), thus the programmer must first configure the PMC to enable the TRNG user interface clock. The user interface clock is independent from any clock that may be used in the logic circuitry used for the source of entropy. The source of entropy can be enabled before enabling the user interface clock.
43.4.2 Interrupt
The TRNG interface has an interrupt line connected to the Interrupt Controller. In order to handle interrupts, the Interrupt Controller must be programmed before configuring the TRNG.
Table 43-1. Peripheral IDs Instance TRNG ID 45
43.5
As soon as the TRNG is enabled (TRNG_CTRL register), the generator provides one 32-bit value every 84 clock cycles. Interrupt trng_int can be enabled through the TRNG_IER register (respectively disabled in TRNG_IDR). This interrupt is set when a new random value is available and is cleared when the status register is read (TRNG_SR register). The flag DATRDY of the status register (TRNG_ISR) is set when the random data is ready to be read out on the 32-bit output data register (TRNG_ODATA). The normal mode of operation checks that the status register flag equals 1 before reading the output data register when a 32-bit random value is required by the software application.
Figure 43-2. TRNG Data Generation Sequence
Clock trng_cr enable
84 clock cycles
84 clock cycles
84 clock cycles
trng_int
Read TRNG_ISR Read TRNG_ODATA
Read TRNG_ISR Read TRNG_ODATA

1249
43.6
True Random Number Generator (TRNG) User Interface
Table 43-2. Register Mapping Offset 0x00 0x10 0x14 0x18 0x1C 0x50 Control Register Interrupt Enable Register Interrupt Disable Register Interrupt Mask Register Interrupt Status Register Output Data Register Register Name TRNG_CR TRNG_IER TRNG_IDR TRNG_IMR TRNG_ISR TRNG_ODATA Access Write-only Write-only Write-only Read-only Read-only Read-only Reset 0x0000_0000 0x0000_0000 0x0000_0000

1250
43.6.1 TRNG Control Register

31
TRNG_CR 0xF8040000 Write-only

30 29 28 27 26 25 24
KEY
23 22 21 20 19 18 17 16
KEY
15 14 13 12 11 10 9 8
KEY
7 6 5 4 3 2 1 0
ENABLE
ENABLE: Enables the TRNG to provide random values

0: Disables the TRNG. 1: Enables the TRNG.
KEY: Security Key

KEY = 0x524e47 (RNG in ASCII) This key is to be written when the ENABLE bit is set or cleared.

1251
43.6.2 TRNG Interrupt Enable Register

31
TRNG_IER 0xF8040010 Write-only

30 29 28 27 26 25 24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
DATRDY
DATRDY: Data Ready Interrupt Enable

0: No effect. 1: Enables the corresponding interrupt.

1252
43.6.3 TRNG Interrupt Disable Register

31
TRNG_IDR 0xF8040014 Write-only

30 29 28 27 26 25 24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
DATRDY
DATRDY: Data Ready Interrupt Disable

0: No effect. 1: Disables the corresponding interrupt.

1253
43.6.4 TRNG Interrupt Mask Register

Name: Address: Reset: Access:
31
TRNG_IMR 0xF8040018
0x0000_0000
Read-only
30 29 28 27 26 25 24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
DATRDY
DATRDY: Data Ready Interrupt Mask

0: The corresponding interrupt is not enabled. 1: The corresponding interrupt is enabled.

1254
43.6.5 TRNG Interrupt Status Register

31
TRNG_ISR 0xF804001C
0x0000_0000
Read-only
30 29 28 27 26 25 24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
7 6 5
DATRDY
DATRDY: Data Ready

0: Output data is not valid or TRNG is disabled. 1: New Random value is completed. DATRDY is cleared when this register is read.

1255
43.6.6 TRNG Output Data Register

31
TRNG_ODATA 0xF8040050
0x0000_0000
Read-only
30 29 28 27 26 25 24
ODATA
23 22 21 20 19 18 17 16
ODATA
15 14 13 12 11 10 9 8
ODATA
7 6 5 4 3 2 1 0
ODATA
ODATA: Output Data

The 32-bit Output Data register contains the 32-bit random data.

1256
44.
44.1
Universal Asynchronous Receiver Transmitter (UART)

Description
The Universal Asynchronous Receiver Transmitter features a two-pin UART that can be used for communication and trace purposes and offers an ideal medium for in-situ programming solutions. Moreover, the association with DMA controller permits packet handling for these tasks with processor time reduced to a minimum.
44.2
z
Two-pin UART
z z z z z z
Independent Receiver and Transmitter with a Common Programmable Baud Rate Generator Even, Odd, Mark or Space Parity Generation Parity, Framing and Overrun Error Detection Automatic Echo, Local Loopback and Remote Loopback Channel Modes Interrupt Generation Support for Two DMA Channels with Connection to Receiver and Transmitter

1257
44.3
Block Diagram
Figure 44-1. UART Functional Block Diagram

Peripheral Bridge DMA Controller
APB
UART
UTXD
Transmit Power Management Controller

MCK
Baud Rate Generator Receive
Parallel Input/ Output

URXD
Interrupt Control
uart_irq
Table 44-1. UART Pin Description Pin Name URXD UTXD Description UART Receive Data UART Transmit Data Type Input Output

1258
44.4
44.4.1 I/O Lines

The UART pins are multiplexed with PIO lines. The programmer must first configure the corresponding PIO Controller to enable I/O line operations of the UART.
Table 44-2. I/O Lines Instance UART0 UART0 UART1 UART1 Signal URXD0 UTXD0 URXD1 UTXD1 I/O Line PC29 PC30 PA30 PA31 Peripheral A A B B

The UART clock is controllable through the Power Management Controller. In this case, the programmer must first configure the PMC to enable the UART clock. Usually, the peripheral identifier used for this purpose is 1.
44.4.3 Interrupt Source

The UART interrupt line is connected to one of the interrupt sources of the Interrupt Controller. Interrupt handling requires programming of the Interrupt Controller before configuring the UART.

1259
44.5
UART Operations
The UART operates in asynchronous mode only and supports only 8-bit character handling (with parity). It has no clock pin. The UART is made up of a receiver and a transmitter that operate independently, and a common baud rate generator. Receiver timeout and transmitter time guard are not implemented. However, all the implemented features are compatible with those of a standard USART.

The baud rate generator provides the bit period clock named baud rate clock to both the receiver and the transmitter.
The baud rate clock is the master clock divided by 16 times the value (CD) written in UART_BRGR (Baud Rate Generator Register). If UART_BRGR is set to 0, the baud rate clock is disabled and the UART remains inactive. The maximum allowable baud rate is Master Clock divided by 16. The minimum allowable baud rate is Master Clock divided by (16 x 65536). MCK Baud Rate = ----------------------16 CD
CD CD MCK 16-bit Counter
OUT
>1 1 0 0 Receiver Sampling Clock Divide by 16 Baud Rate Clock
44.5.2 Receiver
44.5.2.1 Receiver Reset, Enable and Disable After device reset, the UART receiver is disabled and must be enabled before being used. The receiver can be enabled by writing the control register UART_CR with the bit RXEN at 1. At this command, the receiver starts looking for a start bit. The programmer can disable the receiver by writing UART_CR with the bit RXDIS at 1. If the receiver is waiting for a start bit, it is immediately stopped. However, if the receiver has already detected a start bit and is receiving the data, it waits for the stop bit before actually stopping its operation. The programmer can also put the receiver in its reset state by writing UART_CR with the bit RSTRX at 1. In doing so, the receiver immediately stops its current operations and is disabled, whatever its current state. If RSTRX is applied when data is being processed, this data is lost. 44.5.2.2 Start Detection and Data Sampling The UART only supports asynchronous operations, and this affects only its receiver. The UART receiver detects the start of a received character by sampling the URXD signal until it detects a valid start bit. A low level (space) on URXD is interpreted as a valid start bit if it is detected for more than 7 cycles of the sampling clock, which is 16 times the baud rate. Hence, a space that is longer than 7/16 of the bit period is detected as a valid start bit. A space which is 7/16 of a bit period or shorter is ignored and the receiver continues to wait for a valid start bit.

1260
When a valid start bit has been detected, the receiver samples the URXD at the theoretical midpoint of each bit. It is assumed that each bit lasts 16 cycles of the sampling clock (1-bit period) so the bit sampling point is eight cycles (0.5-bit period) after the start of the bit. The first sampling point is therefore 24 cycles (1.5-bit periods) after the falling edge of the start bit was detected. Each subsequent bit is sampled 16 cycles (1-bit period) after the previous one.
Figure 44-3. Start Bit Detection
Sampling Clock
URXD
True Start Detection Baud Rate Clock
D0
Figure 44-4. Character Reception

Example: 8-bit, parity enabled 1 stop
0.5 bit period 1 bit period
URXD
Sampling
D0 D1 True Start Detection
D2
D3
D4
D5
D6
D7 Parity Bit
Stop Bit
44.5.2.3 Receiver Ready When a complete character is received, it is transferred to the UART_RHR and the RXRDY status bit in UART_SR (Status Register) is set. The bit RXRDY is automatically cleared when the receive holding register UART_RHR is read.
Figure 44-5. Receiver Ready
URXD S D0 D1 D2 D3 D4 D5 D6 D7 P S D0 D1 D2 D3 D4 D5 D6 D7 P
RXRDY
Read UART_RHR

1261
44.5.2.4 Receiver Overrun If UART_RHR has not been read by the software (or the Peripheral Data Controller or DMA Controller) since the last transfer, the RXRDY bit is still set and a new character is received, the OVRE status bit in UART_SR is set. OVRE is cleared when the software writes the control register UART_CR with the bit RSTSTA (Reset Status) at 1.
Figure 44-6. Receiver Overrun
URXD S D0 D1 D2 D3 D4 D5 D6 D7 P stop S D0 D1 D2 D3 D4 D5 D6 D7 P stop
RXRDY OVRE
RSTSTA
44.5.2.5 Parity Error Each time a character is received, the receiver calculates the parity of the received data bits, in accordance with the field PAR in UART_MR. It then compares the result with the received parity bit. If different, the parity error bit PARE in UART_SR is set at the same time the RXRDY is set. The parity bit is cleared when the control register UART_CR is written with the bit RSTSTA (Reset Status) at 1. If a new character is received before the reset status command is written, the PARE bit remains at 1.
URXD S D0 D1 D2 D3 D4 D5 D6 D7 P stop
RXRDY PARE
Wrong Parity Bit
RSTSTA
44.5.2.6 Receiver Framing Error When a start bit is detected, it generates a character reception when all the data bits have been sampled. The stop bit is also sampled and when it is detected at 0, the FRAME (Framing Error) bit in UART_SR is set at the same time the RXRDY bit is set. The FRAME bit remains high until the control register UART_CR is written with the bit RSTSTA at 1.
Figure 44-8. Receiver Framing Error
URXD RXRDY FRAME S D0 D1 D2 D3 D4 D5 D6 D7 P stop
Stop Bit Detected at 0
RSTSTA

1262
44.5.3 Transmitter
44.5.3.1 Transmitter Reset, Enable and Disable After device reset, the UART transmitter is disabled and it must be enabled before being used. The transmitter is enabled by writing the control register UART_CR with the bit TXEN at 1. From this command, the transmitter waits for a character to be written in the Transmit Holding Register (UART_THR) before actually starting the transmission. The programmer can disable the transmitter by writing UART_CR with the bit TXDIS at 1. If the transmitter is not operating, it is immediately stopped. However, if a character is being processed into the Shift Register and/or a character has been written in the Transmit Holding Register, the characters are completed before the transmitter is actually stopped. The programmer can also put the transmitter in its reset state by writing the UART_CR with the bit RSTTX at 1. This immediately stops the transmitter, whether or not it is processing characters. 44.5.3.2 Transmit Format The UART transmitter drives the pin UTXD at the baud rate clock speed. The line is driven depending on the format defined in the Mode Register and the data stored in the Shift Register. One start bit at level 0, then the 8 data bits, from the lowest to the highest bit, one optional parity bit and one stop bit at 1 are consecutively shifted out as shown in the following figure. The field PARE in the mode register UART_MR defines whether or not a parity bit is shifted out. When a parity bit is enabled, it can be selected between an odd parity, an even parity, or a fixed space or mark bit.
Figure 44-9. Character Transmission
Example: Parity enabled Baud Rate Clock UTXD
Start Bit
D0
D1
D2
D3
D4
D5
D6
D7
Parity Bit
Stop Bit

1263
44.5.3.3 Transmitter Control When the transmitter is enabled, the bit TXRDY (Transmitter Ready) is set in the status register UART_SR. The transmission starts when the programmer writes in the Transmit Holding Register (UART_THR), and after the written character is transferred from UART_THR to the Shift Register. The TXRDY bit remains high until a second character is written in UART_THR. As soon as the first character is completed, the last character written in UART_THR is transferred into the shift register and TXRDY rises again, showing that the holding register is empty. When both the Shift Register and UART_THR are empty, i.e., all the characters written in UART_THR have been processed, the TXEMPTY bit rises after the last stop bit has been completed.
Figure 44-10. Transmitter Control
UART_THR
Data 0 Data 1
Shift Register
Data 0
Data 1
UTXD
Data 0
stop
Data 1
stop
TXRDY TXEMPTY
Write Data 0 in UART_THR
Write Data 1 in UART_THR

1264
44.5.4 DMA Support

Both the receiver and the transmitter of the UART are connected to a DMA Controller (DMAC) channel. The DMA Controller channels are programmed via registers that are mapped within the DMAC user interface.
44.5.5 Test Modes

The UART supports three test modes. These modes of operation are programmed by using the field CHMODE (Channel Mode) in the mode register (UART_MR). The Automatic Echo mode allows bit-by-bit retransmission. When a bit is received on the URXD line, it is sent to the UTXD line. The transmitter operates normally, but has no effect on the UTXD line. The Local Loopback mode allows the transmitted characters to be received. UTXD and URXD pins are not used and the output of the transmitter is internally connected to the input of the receiver. The URXD pin level has no effect and the UTXD line is held high, as in idle state. The Remote Loopback mode directly connects the URXD pin to the UTXD line. The transmitter and the receiver are disabled and have no effect. This mode allows a bit-by-bit retransmission.
Figure 44-11. Test Modes
Automatic Echo Receiver RXD
Transmitter
Disabled
TXD
Local Loopback Receiver

Disabled
RXD
VDD Transmitter
Disabled
TXD
Remote Loopback Receiver
VDD
Disabled
RXD
Disabled
Transmitter
TXD

1265
44.6
Universal Asynchronous Receiver Transmitter (UART) User Interface
Table 44-3. Register Mapping Offset 0x0000 0x0004 0x0008 0x000C 0x0010 0x0014 0x0018 0x001C 0x0020 0x0024 - 0x003C 0x004C - 0x00FC Register Control Register Mode Register Interrupt Enable Register Interrupt Disable Register Interrupt Mask Register Status Register Receive Holding Register Transmit Holding Register Baud Rate Generator Register Reserved Reserved Name UART_CR UART_MR UART_IER UART_IDR UART_IMR UART_SR UART_RHR UART_THR UART_BRGR Access Write-only Read-write Write-only Write-only Read-only Read-only Read-only Write-only Read-write Reset 0x0 0x0 0x0 0x0

1266
44.6.1 UART Control Register

31
UART_CR 0xF0024000 (0), 0xF8028000 (1) Write-only

30 29 28 27 26 25 24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
8 RSTSTA 0
7 TXDIS
6 TXEN
5 RXDIS
4 RXEN
3 RSTTX
2 RSTRX

0 = No effect. 1 = The receiver logic is reset and disabled. If a character is being received, the reception is aborted.

0 = No effect. 1 = The transmitter logic is reset and disabled. If a character is being transmitted, the transmission is aborted.

0 = No effect. 1 = The receiver is enabled if RXDIS is 0.

0 = No effect. 1 = The receiver is disabled. If a character is being processed and RSTRX is not set, the character is completed before the receiver is stopped.

0 = No effect. 1 = The transmitter is enabled if TXDIS is 0.

0 = No effect. 1 = The transmitter is disabled. If a character is being processed and a character has been written in the UART_THR and RSTTX is not set, both characters are completed before the transmitter is stopped.

0 = No effect. 1 = Resets the status bits PARE, FRAME and OVRE in the UART_SR.

1267
44.6.2 UART Mode Register

31
UART_MR 0xF0024004 (0), 0xF8028004 (1) Read-write

30 29 28 27 26 25 24
23
22
21
20
19
18
17
16
15 CHMODE 7
14
13
12
11
10 PAR
6 5
4 3
1 0
PAR: Parity Type

Value 0 1 2 3 4 Name EVEN ODD SPACE MARK NO Description Even Parity Odd Parity Space: parity forced to 0 Mark: parity forced to 1 No Parity

Value 0 1 2 3 Name NORMAL AUTOMATIC LOCAL_LOOPBACK REMOTE_LOOPBACK Description Normal Mode Automatic Echo Local Loopback Remote Loopback

1268
44.6.3 UART Interrupt Enable Register

31
UART_IER 0xF0024008 (0), 0xF8028008 (1) Write-only

30 29 28 27 26 25 24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9 TXEMPTY 1 TXRDY
7 PARE
6 FRAME
5 OVRE
0 RXRDY
RXRDY: Enable RXRDY Interrupt TXRDY: Enable TXRDY Interrupt OVRE: Enable Overrun Error Interrupt FRAME: Enable Framing Error Interrupt PARE: Enable Parity Error Interrupt TXEMPTY: Enable TXEMPTY Interrupt

1269
44.6.4 UART Interrupt Disable Register

31
UART_IDR 0xF002400C (0), 0xF802800C (1) Write-only

30 29 28 27 26 25 24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9 TXEMPTY 1 TXRDY
7 PARE
6 FRAME
5 OVRE
0 RXRDY
RXRDY: Disable RXRDY Interrupt TXRDY: Disable TXRDY Interrupt OVRE: Disable Overrun Error Interrupt FRAME: Disable Framing Error Interrupt PARE: Disable Parity Error Interrupt TXEMPTY: Disable TXEMPTY Interrupt

1270
44.6.5 UART Interrupt Mask Register

31
UART_IMR 0xF0024010 (0), 0xF8028010 (1) Read-only

30 29 28 27 26 25 24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9 TXEMPTY 1 TXRDY
7 PARE
6 FRAME
5 OVRE
0 RXRDY
RXRDY: Mask RXRDY Interrupt TXRDY: Disable TXRDY Interrupt OVRE: Mask Overrun Error Interrupt FRAME: Mask Framing Error Interrupt PARE: Mask Parity Error Interrupt TXEMPTY: Mask TXEMPTY Interrupt

1271
44.6.6 UART Status Register

31
UART_SR 0xF0024014 (0), 0xF8028014 (1) Read-only

30 29 28 27 26 25 24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9 TXEMPTY 1 TXRDY
7 PARE
6 FRAME
5 OVRE
0 RXRDY

0 = No character has been received since the last read of the UART_RHR or the receiver is disabled. 1 = At least one complete character has been received, transferred to UART_RHR and not yet read.

0 = A character has been written to UART_THR and not yet transferred to the Shift Register, or the transmitter is disabled. 1 = There is no character written to UART_THR not yet transferred to the Shift Register.
OVRE: Overrun Error

0 = No overrun error has occurred since the last RSTSTA. 1 = At least one overrun error has occurred since the last RSTSTA.

0 = No framing error has occurred since the last RSTSTA. 1 = At least one framing error has occurred since the last RSTSTA.
PARE: Parity Error

0 = No parity error has occurred since the last RSTSTA. 1 = At least one parity error has occurred since the last RSTSTA.

0 = There are characters in UART_THR, or characters being processed by the transmitter, or the transmitter is disabled. 1 = There are no characters in UART_THR and there are no characters being processed by the transmitter.

1272
44.6.7 UART Receiver Holding Register

31
UART_RHR 0xF0024018 (0), 0xF8028018 (1) Read-only

30 29 28 27 26 25 24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
4 RXCHR

Last received character if RXRDY is set.
44.6.8 UART Transmit Holding Register

31
UART_THR 0xF002401C (0), 0xF802801C (1) Write-only

30 29 28 27 26 25 24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
4 TXCHR


1273
44.6.9 UART Baud Rate Generator Register

31
UART_BRGR 0xF0024020 (0), 0xF8028020 (1) Read-write

30 29 28 27 26 25 24
23
22
21
20
19
18
17
16
15
14
13
12 CD
11
10
4 CD
CD: Clock Divisor

0 = Baud Rate Clock is disabled 1 to 65,535 = MCK / (CD x 16)

1274
45.
45.1
Universal Synchronous Asynchronous Receiver Transmitter (USART)

Description
The Universal Synchronous Asynchronous Receiver Transceiver (USART) provides one full duplex universal synchronous asynchronous serial link. Data frame format is widely programmable (data length, parity, number of stop bits) to support a maximum of standards. The receiver implements parity error, framing error and overrun error detection. The receiver time-out enables handling variable-length frames and the transmitter timeguard facilitates communications with slow remote devices. Multidrop communications are also supported through address bit handling in reception and transmission. The USART features three test modes: remote loopback, local loopback and automatic echo. The USART supports specific operating modes providing interfaces on RS485and SPI buses, with ISO7816 T = 0 or T = 1 smart card slots and infrared transceivers. The hardware handshaking feature enables an out-of-band flow control by automatic management of the pins RTS and CTS. The USART supports the connection to the DMA Controller, which enables data transfers to the transmitter and from the receiver. The DMAC provides chained buffer management without any intervention of the processor.

1275
45.2
z z
Programmable Baud Rate Generator 5-bit to 9-bit Full-duplex Synchronous or Asynchronous Serial Communications
z z z z z z z z z
1, 1.5 or 2 Stop Bits in Asynchronous Mode or 1 or 2 Stop Bits in Synchronous Mode Parity Generation and Error Detection Framing Error Detection, Overrun Error Detection MSB-first or LSB-first Optional Break Generation and Detection By-8 or by-16 Over-sampling Receiver Frequency Optional Hardware Handshaking RTS-CTS Receiver Time-out and Transmitter Timeguard Optional Multidrop Mode with Address Generation and Detection
z z
RS485 with Driver Control Signal ISO7816, T = 0 or T = 1 Protocols for Interfacing with Smart Cards
z
NACK Handling, Error Counter with Repetition and Iteration Limit Communication at up to 115.2 Kbps Master or Slave Serial Clock Programmable Phase and Polarity SPI Serial Clock (SCK) Frequency up to Internal Clock Frequency MCK/6 Remote Loopback, Local Loopback, Automatic Echo Two DMA Controller Channels (DMAC)
IrDA Modulation and Demodulation

z
SPI Mode
z z z
Test Modes
z
Supports Connection of:

z
Offers Buffer Transfer without Processor Intervention

1276
45.3
Block Diagram
Figure 45-1. USART Block Diagram
(Peripheral) DMA Controller Channel Channel
USART
PIO Controller
RXD Receiver RTS Interrupt Controller USART Interrupt TXD Transmitter CTS
PMC MCK MCK/DIV Baud Rate Generator SCK
DIV
User Interface SLCK
APB
Table 45-1. SPI Operating Mode PIN RXD TXD RTS CTS USART RXD TXD RTS CTS SPI Slave MOSI MISO CS SPI Master MISO MOSI CS

1277
45.4

PPP Serial Driver Field Bus Driver EMV Driver
IrLAP IrDA Driver SPI Driver
USART
RS232 Drivers
RS485 Drivers
Smart Card Slot
IrDA Transceivers
SPI Bus
Serial Port
Differential Bus
45.5
Table 45-2. I/O Line Description Name SCK Description Serial Clock Transmit Serial Data TXD or Master Out Slave In (MOSI) in SPI Master Mode or Master In Slave Out (MISO) in SPI Slave Mode Receive Serial Data RXD or Master In Slave Out (MISO) in SPI Master Mode or Master Out Slave In (MOSI) in SPI Slave Mode CTS Clear to Send or Slave Select (NSS) in SPI Slave Mode Request to Send or Slave Select (NSS) in SPI Master Mode Input Low Input I/O Type I/O Active Level
RTS
Output
Low

1278
45.6
45.6.1 I/O Lines

The pins used for interfacing the USART may be multiplexed with the PIO lines. The programmer must first program the PIO controller to assign the desired USART pins to their peripheral function. If I/O lines of the USART are not used by the application, they can be used for other purposes by the PIO Controller. To prevent the TXD line from falling when the USART is disabled, the use of an internal pull up is mandatory. If the hardware handshaking feature is used, the internal pull up on TXD must also be enabled.
Table 45-3. I/O Lines Instance USART0 USART0 USART0 USART0 USART0 USART1 USART1 USART1 USART1 USART1 USART2 USART2 USART2 USART2 USART2 USART3 USART3 USART3 USART3 USART3 Signal CTS0 RTS0 RXD0 SCK0 TXD0 CTS1 RTS1 RXD1 SCK1 TXD1 CTS2 RTS2 RXD2 SCK2 TXD2 CTS3 RTS3 RXD3 SCK3 TXD3 I/O Line PD15 PD16 PD17 PD14 PD18 PB26 PB27 PB28 PB25 PB29 PE23 PE24 PE25 PE20 PE26 PE16 PE17 PE18 PE15 PE19 Peripheral A A A A A A A A A A B B B B B B B B B B

The USART is not continuously clocked. The programmer must first enable the USART Clock in the Power Management Controller (PMC) before using the USART. However, if the application does not require USART operations, the USART clock can be stopped when not needed and be restarted later. In this case, the USART will resume its operations where it left off. Configuring the USART does not require the USART clock to be enabled.

1279
45.6.3 Interrupt
The USART interrupt line is connected on one of the internal sources of the Interrupt Controller. Using the USART interrupt requires the Interrupt Controller to be programmed first. Note that it is not recommended to use the USART interrupt line in edge sensitive mode.
Table 45-4. Peripheral IDs Instance USART0 USART1 USART2 USART3 ID 12 13 14 15
45.7
The USART is capable of managing several types of serial synchronous or asynchronous communications. It supports the following communication modes:
z
5-bit to 9-bit full-duplex asynchronous serial communication

z z z z z z z
MSB-first or LSB-first 1, 1.5 or 2 stop bits Parity even, odd, marked, space or none By-8 or by-16 over-sampling receiver frequency Optional hardware handshaking Optional break management Optional multidrop serial communication MSB-first or LSB-first 1 or 2 stop bits Parity even, odd, marked, space or none By-8 or by-16 over-sampling frequency Optional hardware handshaking Optional break management Optional multidrop serial communication
High-speed 5- to 9-bit full-duplex synchronous serial communication

z z z z z z z
z z
RS485 with driver control signal ISO7816, T0 or T1 protocols for interfacing with smart cards
z
NACK handling, error counter with repetition and iteration limit, inverted data.
z z
InfraRed IrDA Modulation and Demodulation SPI Mode

z z z
Master or Slave Serial Clock Programmable Phase and Polarity SPI Serial Clock (SCK) Frequency up to Internal Clock Frequency MCK/6 Remote loopback, local loopback, automatic echo
Test modes
z

1280

The Baud Rate Generator provides the bit period clock named the Baud Rate Clock to both the receiver and the transmitter. The Baud Rate Generator clock source can be selected by setting the USCLKS field in the Mode Register (US_MR) between:
z z z
The Master Clock MCK A division of the Master Clock, the divider being product dependent, but generally set to 8 The external clock, available on the SCK pin
The Baud Rate Generator is based upon a 16-bit divider, which is programmed with the CD field of the Baud Rate Generator Register (US_BRGR). If CD is programmed to 0, the Baud Rate Generator does not generate any clock. If CD is programmed to 1, the divider is bypassed and becomes inactive. If the external SCK clock is selected, the duration of the low and high levels of the signal provided on the SCK pin must be longer than a Master Clock (MCK) period. The frequency of the signal provided on SCK must be at least 3 times lower than MCK in USART mode, or 6 times lower in SPI mode.
USCLKS MCK MCK/DIV SCK Reserved CD CD 0 1 2 3 0 16-bit Counter >1 1 0 1 1 SYNC USCLKS = 3 Sampling Clock 0 OVER Sampling Divider 0 Baud Rate Clock FIDI SYNC
SCK
45.7.1.1 Baud Rate in Asynchronous Mode If the USART is programmed to operate in asynchronous mode, the selected clock is first divided by CD, which is field programmed in the Baud Rate Generator Register (US_BRGR). The resulting clock is provided to the receiver as a sampling clock and then divided by 16 or 8, depending on the programming of the OVER bit in US_MR. If OVER is set to 1, the receiver sampling is 8 times higher than the baud rate clock. If OVER is cleared, the sampling is performed at 16 times the baud rate clock. The following formula performs the calculation of the Baud Rate.
SelectedClock Baudrate = -------------------------------------------( 8 ( 2 Over ) CD )
This gives a maximum baud rate of MCK divided by 8, assuming that MCK is the highest possible clock and that OVER is programmed to 1.

1281
Baud Rate Calculation Example

Table 45-5 shows calculations of CD to obtain a baud rate at 38400 bauds for different source clock frequencies. This table also shows the actual resulting baud rate and the error.
Table 45-5. Baud Rate Example (OVER = 0) Source Clock MHz 3 686 400 4 915 200 5 000 000 7 372 800 8 000 000 12 000 000 12 288 000 14 318 180 14 745 600 18 432 000 24 000 000 24 576 000 25 000 000 32 000 000 32 768 000 33 000 000 40 000 000 50 000 000 Expected Baud Rate Bit/s 38 400 38 400 38 400 38 400 38 400 38 400 38 400 38 400 38 400 38 400 38 400 38 400 38 400 38 400 38 400 38 400 38 400 38 400 6.00 8.00 8.14 12.00 13.02 19.53 20.00 23.30 24.00 30.00 39.06 40.00 40.69 52.08 53.33 53.71 65.10 81.38 6 8 8 12 13 20 20 23 24 30 39 40 40 52 53 54 65 81 Calculation Result CD Actual Baud Rate Bit/s 38 400.00 38 400.00 39 062.50 38 400.00 38 461.54 37 500.00 38 400.00 38 908.10 38 400.00 38 400.00 38 461.54 38 400.00 38 109.76 38 461.54 38 641.51 38 194.44 38 461.54 38 580.25 0.00% 0.00% 1.70% 0.00% 0.16% 2.40% 0.00% 1.31% 0.00% 0.00% 0.16% 0.00% 0.76% 0.16% 0.63% 0.54% 0.16% 0.47% Error
The baud rate is calculated with the following formula:
BaudRate = MCK CD 16
The baud rate error is calculated with the following formula. It is not recommended to work with an error higher than 5%.
ExpectedBaudRate -------------------------------------------------- Error = 1 ActualBaudRate
45.7.1.2 Fractional Baud Rate in Asynchronous Mode The Baud Rate generator previously defined is subject to the following limitation: the output frequency changes by only integer multiples of the reference frequency. An approach to this problem is to integrate a fractional N clock generator that has a high resolution. The generator architecture is modified to obtain Baud Rate changes by a fraction of the reference source clock. This fractional part is programmed with the FP field in the Baud Rate Generator Register (US_BRGR). If FP is not 0, the fractional part is activated. The resolution is one eighth of the clock divider. This feature is only available when using USART normal mode. The fractional Baud Rate is calculated using the following formula:
SelectedClock Baudrate = --------------------------------------------------------------- 8 ( 2 Over ) CD + FP ------ 8

1282
The modified architecture is presented below:

Figure 45-4. Fractional Baud Rate Generator
FP
USCLKS MCK MCK/DIV SCK Reserved
CD
Modulus Control FP CD SCK FIDI 0 OVER Sampling Divider 1 1 SYNC USCLKS = 3 Sampling Clock 0 Baud Rate Clock SYNC
0 1 2 3 16-bit Counter Glitch-free Logic
>1 1
45.7.1.3 Baud Rate in Synchronous Mode or SPI Mode If the USART is programmed to operate in synchronous mode, the selected clock is simply divided by the field CD in US_BRGR.
BaudRate = SelectedClock ------------------------------------CD
In synchronous mode, if the external clock is selected (USCLKS = 3), the clock is provided directly by the signal on the USART SCK pin. No division is active. The value written in US_BRGR has no effect. The external clock frequency must be at least 3 times lower than the system clock. In synchronous mode master (USCLKS = 0 or 1, CLK0 set to 1), the receive part limits the SCK maximum frequency to MCK/3 in USART mode, or MCK/6 in SPI mode. When either the external clock SCK or the internal clock divided (MCK/DIV) is selected, the value programmed in CD must be even if the user has to ensure a 50:50 mark/space ratio on the SCK pin. If the internal clock MCK is selected, the Baud Rate Generator ensures a 50:50 duty cycle on the SCK pin, even if the value programmed in CD is odd. 45.7.1.4 Baud Rate in ISO 7816 Mode The ISO7816 specification defines the bit rate with the following formula:
Di - f B = ----Fi
where:
z z z z
B is the bit rate Di is the bit-rate adjustment factor Fi is the clock frequency division factor f is the ISO7816 clock frequency (Hz)

1283
Di is a binary value encoded on a 4-bit field, named DI, as represented in Table 45-6.
Table 45-6. Binary and Decimal Values for Di DI field Di (decimal) 0001 1 0010 2 0011 4 0100 8 0101 16 0110 32 1000 12 1001 20
Fi is a binary value encoded on a 4-bit field, named FI, as represented in Table 45-7.
Table 45-7. Binary and Decimal Values for Fi FI field Fi (decimal) 0000 372 0001 372 0010 558 0011 744 0100 1116 0101 1488 0110 1860 1001 512 1010 768 1011 1024 1100 1536 1101 2048
Table 45-8 shows the resulting Fi/Di Ratio, which is the ratio between the ISO7816 clock and the baud rate clock.
Table 45-8. Possible Values for the Fi/Di Ratio Fi/Di 1 2 4 8 16 32 12 20 372 372 186 93 46.5 23.25 11.62 31 18.6 558 558 279 139.5 69.75 34.87 17.43 46.5 27.9 774 744 372 186 93 46.5 23.25 62 37.2 1116 1116 558 279 139.5 69.75 34.87 93 55.8 1488 1488 744 372 186 93 46.5 124 74.4 1806 1860 930 465 232.5 116.2 58.13 155 93 512 512 256 128 64 32 16 42.66 25.6 768 768 384 192 96 48 24 64 38.4 1024 1024 512 256 128 64 32 85.33 51.2 1536 1536 768 384 192 96 48 128 76.8 2048 2048 1024 512 256 128 64 170.6 102.4
If the USART is configured in ISO7816 Mode, the clock selected by the USCLKS field in the Mode Register (US_MR) is first divided by the value programmed in the field CD in the Baud Rate Generator Register (US_BRGR). The resulting clock can be provided to the SCK pin to feed the smart card clock inputs. This means that the CLKO bit can be set in US_MR. This clock is then divided by the value programmed in the FI_DI_RATIO field in the FI_DI_Ratio register (US_FIDI). This is performed by the Sampling Divider, which performs a division by up to 2047 in ISO7816 Mode. The non-integer values of the Fi/Di Ratio are not supported and the user must program the FI_DI_RATIO field to a value as close as possible to the expected value. The FI_DI_RATIO field resets to the value 0x174 (372 in decimal) and is the most common divider between the ISO7816 clock and the bit rate (Fi = 372, Di = 1). Figure 45-5 shows the relation between the Elementary Time Unit, corresponding to a bit time, and the ISO 7816 clock.
Figure 45-5. Elementary Time Unit (ETU)
FI_DI_RATIO ISO7816 Clock Cycles ISO7816 Clock on SCK ISO7816 I/O Line on TXD
1 ETU

1284
45.7.2 Receiver and Transmitter Control

After reset, the receiver is disabled. The user must enable the receiver by setting the RXEN bit in the Control Register (US_CR). However, the receiver registers can be programmed before the receiver clock is enabled. After reset, the transmitter is disabled. The user must enable it by setting the TXEN bit in the Control Register (US_CR). However, the transmitter registers can be programmed before being enabled. The Receiver and the Transmitter can be enabled together or independently. At any time, the software can perform a reset on the receiver or the transmitter of the USART by setting the corresponding bit, RSTRX and RSTTX respectively, in the Control Register (US_CR). The software resets clear the status flag and reset internal state machines but the user interface configuration registers hold the value configured prior to software reset. Regardless of what the receiver or the transmitter is performing, the communication is immediately stopped. The user can also independently disable the receiver or the transmitter by setting RXDIS and TXDIS respectively in US_CR. If the receiver is disabled during a character reception, the USART waits until the end of reception of the current character, then the reception is stopped. If the transmitter is disabled while it is operating, the USART waits the end of transmission of both the current character and character being stored in the Transmit Holding Register (US_THR). If a timeguard is programmed, it is handled normally.
45.7.3 Synchronous and Asynchronous Modes

45.7.3.1 Transmitter Operations The transmitter performs the same in both synchronous and asynchronous operating modes (SYNC = 0 or SYNC = 1). One start bit, up to 9 data bits, one optional parity bit and up to two stop bits are successively shifted out on the TXD pin at each falling edge of the programmed serial clock. The number of data bits is selected by the CHRL field and the MODE 9 bit in the Mode Register (US_MR). Nine bits are selected by setting the MODE 9 bit regardless of the CHRL field. The parity bit is set according to the PAR field in US_MR. The even, odd, space, marked or none parity bit can be configured. The MSBF field in US_MR configures which data bit is sent first. If written to 1, the most significant bit is sent first. If written to 0, the less significant bit is sent first. The number of stop bits is selected by the NBSTOP field in US_MR. The 1.5 stop bit is supported in asynchronous mode only.
Figure 45-6. Character Transmit
Example: 8-bit, Parity Enabled One Stop Baud Rate Clock TXD
Start Bit
D0
D1
D2
D3
D4
D5
D6
D7
Parity Bit
Stop Bit
The characters are sent by writing in the Transmit Holding Register (US_THR). The transmitter reports two status bits in the Channel Status Register (US_CSR): TXRDY (Transmitter Ready), which indicates that US_THR is empty and TXEMPTY, which indicates that all the characters written in US_THR have been processed. When the current character processing is completed, the last character written in US_THR is transferred into the Shift Register of the transmitter and US_THR becomes empty, thus TXRDY rises.

1285
Both TXRDY and TXEMPTY bits are low when the transmitter is disabled. Writing a character in US_THR while TXRDY is low has no effect and the written character is lost.
Figure 45-7. Transmitter Status
Baud Rate Clock TXD
Start D0 Bit D1 D2 D3 D4 D5 D6 D7 Parity Stop Start D0 Bit Bit Bit D1 D2 D3 D4 D5 D6 D7 Parity Stop Bit Bit
Write US_THR TXRDY
TXEMPTY
45.7.3.2 Manchester Encoder When the Manchester encoder is in use, characters transmitted through the USART are encoded based on biphase Manchester II format. To enable this mode, set the MAN field in the US_MR register to 1. Depending on polarity configuration, a logic level (zero or one), is transmitted as a coded signal one-to-zero or zero-to-one. Thus, a transition always occurs at the midpoint of each bit time. It consumes more bandwidth than the original NRZ signal (2x) but the receiver has more error control since the expected input must show a change at the center of a bit cell. An example of Manchester encoded sequence is: the byte 0xB1 or 10110001 encodes to 10 01 10 10 01 01 01 10, assuming the default polarity of the encoder. Figure 45-8 illustrates this coding scheme.
Figure 45-8. NRZ to Manchester Encoding
NRZ encoded data Manchester encoded data 1 0 1 1 0 0 0 1
Txd
The Manchester encoded character can also be encapsulated by adding both a configurable preamble and a start frame delimiter pattern. Depending on the configuration, the preamble is a training sequence, composed of a pre-defined pattern with a programmable length from 1 to 15 bit times. If the preamble length is set to 0, the preamble waveform is not generated prior to any character. The preamble pattern is chosen among the following sequences: ALL_ONE, ALL_ZERO, ONE_ZERO or ZERO_ONE, writing the field TX_PP in the US_MAN register, the field TX_PL is used to configure the preamble length. Figure 45-9 illustrates and defines the valid patterns. To improve flexibility, the encoding scheme can be configured using the TX_MPOL field in the US_MAN register. If the TX_MPOL field is set to zero (default), a logic zero is encoded with a zero-to-one transition and a logic one is encoded with a one-to-zero transition. If the TX_MPOL field is set to one, a logic one is encoded with a one-to-zero transition and a logic zero is encoded with a zero-to-one transition.

1286
Figure 45-9. Preamble Patterns, Default Polarity Assumed

Manchester encoded data
Txd
SFD
DATA
8 bit width "ALL_ONE" Preamble
Txd
SFD
DATA
8 bit width "ALL_ZERO" Preamble Manchester encoded data
Txd
SFD
DATA
8 bit width "ZERO_ONE" Preamble
Txd
SFD
DATA
8 bit width "ONE_ZERO" Preamble
A start frame delimiter is to be configured using the ONEBIT field in the US_MR register. It consists of a user-defined pattern that indicates the beginning of a valid data. Figure 45-10 illustrates these patterns. If the start frame delimiter, also known as the start bit, is one bit, (ONEBIT to 1), a logic zero is Manchester encoded and indicates that a new character is being sent serially on the line. If the start frame delimiter is a synchronization pattern also referred to as sync (ONEBIT to 0), a sequence of 3 bit times is sent serially on the line to indicate the start of a new character. The sync waveform is in itself an invalid Manchester waveform as the transition occurs at the middle of the second bit time. Two distinct sync patterns are used: the command sync and the data sync. The command sync has a logic one level for one and a half bit times, then a transition to logic zero for the second one and a half bit times. If the MODSYNC field in the US_MR register is set to 1, the next character is a command. If it is set to 0, the next character is a data. When direct memory access is used, the MODSYNC field can be immediately updated with a modified character located in memory. To enable this mode, VAR_SYNC field in US_MR register must be set to 1. In this case, the MODSYNC field in US_MR is bypassed and the sync configuration is held in the TXSYNH in the US_THR register. The USART character format is modified and includes sync information.

1287
Figure 45-10. Start Frame Delimiter

Preamble Length is set to 0 SFD Manchester encoded data Txd DATA One bit start frame delimiter SFD Manchester encoded data Txd DATA
SFD Manchester encoded data Txd
Command Sync start frame delimiter DATA Data Sync start frame delimiter
Drift Compensation
Drift compensation is available only in 16X oversampling mode. An hardware recovery system allows a larger clock drift. To enable the hardware system, the bit in the USART_MAN register must be set. If the RXD edge is one 16X clock cycle from the expected edge, this is considered as normal jitter and no corrective actions is taken. If the RXD event is between 4 and 2 clock cycles before the expected edge, then the current period is shortened by one clock cycle. If the RXD event is between 2 and 3 clock cycles after the expected edge, then the current period is lengthened by one clock cycle. These intervals are considered to be drift and so corrective actions are automatically taken.
Figure 45-11. Bit Resynchronization
Oversampling 16x Clock RXD
Sampling point Expected edge Synchro. Error Synchro. Jump Tolerance Sync Jump Synchro. Error

1288
45.7.3.3 Asynchronous Receiver If the USART is programmed in asynchronous operating mode (SYNC = 0), the receiver oversamples the RXD input line. The oversampling is either 16 or 8 times the Baud Rate clock, depending on the OVER bit in the Mode Register (US_MR). The receiver samples the RXD line. If the line is sampled during one half of a bit time to 0, a start bit is detected and data, parity and stop bits are successively sampled on the bit rate clock. If the oversampling is 16, (OVER to 0), a start is detected at the eighth sample to 0. Then, data bits, parity bit and stop bit are sampled on each 16 sampling clock cycle. If the oversampling is 8 (OVER to 1), a start bit is detected at the fourth sample to 0. Then, data bits, parity bit and stop bit are sampled on each 8 sampling clock cycle. The number of data bits, first bit sent and parity mode are selected by the same fields and bits as the transmitter, i.e. respectively CHRL, MODE9, MSBF and PAR. For the synchronization mechanism only , the number of stop bits has no effect on the receiver as it considers only one stop bit, regardless of the field NBSTOP, so that resynchronization between the receiver and the transmitter can occur. Moreover, as soon as the stop bit is sampled, the receiver starts looking for a new start bit so that resynchronization can also be accomplished when the transmitter is operating with one stop bit. Figure 45-12 and Figure 45-13 illustrate start detection and character reception when USART operates in asynchronous mode.
Figure 45-12. Asynchronous Start Detection
Baud Rate Clock Sampling Clock (x16) RXD Sampling 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 D0 Sampling
Start Detection RXD Sampling
0 1 Start Rejection

1289
Figure 45-13. Asynchronous Character Reception

Example: 8-bit, Parity Enabled
Baud Rate Clock RXD Start Detection

16 16 16 16 16 16 16 16 16 16 samples samples samples samples samples samples samples samples samples samples
D0
D1
D2
D3
D4
D5
D6
D7
Parity Bit
Stop Bit
45.7.3.4 Manchester Decoder When the MAN field in US_MR register is set to 1, the Manchester decoder is enabled. The decoder performs both preamble and start frame delimiter detection. One input line is dedicated to Manchester encoded input data. An optional preamble sequence can be defined, its length is user-defined and totally independent of the emitter side. Use RX_PL in US_MAN register to configure the length of the preamble sequence. If the length is set to 0, no preamble is detected and the function is disabled. In addition, the polarity of the input stream is programmable with RX_MPOL field in US_MAN register. Depending on the desired application the preamble pattern matching is to be defined via the RX_PP field in US_MAN. See Figure 45-9 for available preamble patterns. Unlike preamble, the start frame delimiter is shared between Manchester Encoder and Decoder. So, if ONEBIT field is set to 1, only a zero encoded Manchester can be detected as a valid start frame delimiter. If ONEBIT is set to 0, only a sync pattern is detected as a valid start frame delimiter. Decoder operates by detecting transition on incoming stream. If RXD is sampled during one quarter of a bit time to zero, a start bit is detected. See Figure 45-14. The sample pulse rejection mechanism applies. In order to increase the compatibility the RXIDLV bit in the US_MAN register allows to inform the USART block of the Rx line idle state value (Rx line undriven), it can be either level one (pull-up) or level zero (pull-down). By default this bit is set to one (Rx line is at level 1 if undriven).
Figure 45-14. Asynchronous Start Bit Detection
Sampling Clock (16 x) Manchester encoded data
Txd Start Detection 1 2 3 4
The receiver is activated and starts Preamble and Frame Delimiter detection, sampling the data at one quarter and then three quarters. If a valid preamble pattern or start frame delimiter is detected, the receiver continues decoding with the same synchronization. If the stream does not match a valid pattern or a valid start frame delimiter, the receiver re-synchronizes on the next valid edge.The minimum time threshold to estimate the bit value is three quarters of a bit time. If a valid preamble (if used) followed with a valid start frame delimiter is detected, the incoming stream is decoded into NRZ data and passed to USART for processing. Figure 45-15 illustrates Manchester

1290
pattern mismatch. When incoming data stream is passed to the USART, the receiver is also able to detect Manchester code violation. A code violation is a lack of transition in the middle of a bit cell. In this case, MANE flag in US_CSR register is raised. It is cleared by writing the Control Register (US_CR) with the RSTSTA bit to 1. See Figure 45-16 for an example of Manchester error detection during data phase.
Figure 45-15. Preamble Pattern Mismatch
Preamble Mismatch Manchester coding error Preamble Mismatch invalid pattern
Txd
SFD
DATA
Preamble Length is set to 8
Figure 45-16. Manchester Error Flag

Preamble Length is set to 4 Elementary character bit time SFD Manchester encoded data Txd Entering USART character area
sampling points
Preamble subpacket and Start Frame Delimiter were successfully decoded
Manchester Coding Error detected
When the start frame delimiter is a sync pattern (ONEBIT field to 0), both command and data delimiter are supported. If a valid sync is detected, the received character is written as RXCHR field in the US_RHR register and the RXSYNH is updated. RXCHR is set to 1 when the received character is a command, and it is set to 0 if the received character is a data. This mechanism alleviates and simplifies the direct memory access as the character contains its own sync field in the same register. As the decoder is setup to be used in unipolar mode, the first bit of the frame has to be a zero-to-one transition.

1291
45.7.3.5 Radio Interface: Manchester Encoded USART Application This section describes low data rate RF transmission systems and their integration with a Manchester encoded USART. These systems are based on transmitter and receiver ICs that support ASK and FSK modulation schemes. The goal is to perform full duplex radio transmission of characters using two different frequency carriers. See the configuration in Figure 45-17.
Figure 45-17. Manchester Encoded Characters RF Transmission
Fup frequency Carrier ASK/FSK Upstream Receiver
Upstream Emitter
LNA VCO RF filter Demod
Serial Configuration Interface
control Fdown frequency Carrier bi-dir line ASK/FSK downstream transmitter
Manchester decoder
USART Receiver
Downstream Receiver
Manchester encoder PA RF filter Mod VCO
USART Emitter
control
The USART module is configured as a Manchester encoder/decoder. Looking at the downstream communication channel, Manchester encoded characters are serially sent to the RF emitter. This may also include a user defined preamble and a start frame delimiter. Mostly, preamble is used in the RF receiver to distinguish between a valid data from a transmitter and signals due to noise. The Manchester stream is then modulated. See Figure 45-18 for an example of ASK modulation scheme. When a logic one is sent to the ASK modulator, the power amplifier, referred to as PA, is enabled and transmits an RF signal at downstream frequency. When a logic zero is transmitted, the RF signal is turned off. If the FSK modulator is activated, two different frequencies are used to transmit data. When a logic 1 is sent, the modulator outputs an RF signal at frequency F0 and switches to F1 if the data sent is a 0. See Figure 45-19. From the receiver side, another carrier frequency is used. The RF receiver performs a bit check operation examining demodulated data stream. If a valid pattern is detected, the receiver switches to receiving mode. The demodulated stream is sent to the Manchester decoder. Because of bit checking inside RF IC, the data transferred to the microcontroller is reduced by a user-defined number of bits. The Manchester preamble length is to be defined in accordance with the RF IC configuration.

1292
Figure 45-18. ASK Modulator Output

1 NRZ stream Manchester encoded data default polarity unipolar output ASK Modulator Output Uptstream Frequency F0 0 0 1
Txd
Figure 45-19. FSK Modulator Output

1 NRZ stream Manchester encoded data default polarity unipolar output FSK Modulator Output Uptstream Frequencies [F0, F0+offset] 0 0 1
Txd
45.7.3.6 Synchronous Receiver In synchronous mode (SYNC = 1), the receiver samples the RXD signal on each rising edge of the Baud Rate Clock. If a low level is detected, it is considered as a start. All data bits, the parity bit and the stop bits are sampled and the receiver waits for the next start bit. Synchronous mode operations provide a high speed transfer capability. Configuration fields and bits are the same as in asynchronous mode. Figure 45-20 illustrates a character reception in synchronous mode.
Figure 45-20. Synchronous Mode Character Reception
Example: 8-bit, Parity Enabled 1 Stop Baud Rate Clock
RXD Sampling
Start
D0
D1
D2
D3
D4
D5
D6
D7 Parity Bit
Stop Bit

1293
45.7.3.7 Receiver Operations When a character reception is completed, it is transferred to the Receive Holding Register (US_RHR) and the RXRDY bit in the Status Register (US_CSR) rises. If a character is completed while the RXRDY is set, the OVRE (Overrun Error) bit is set. The last character is transferred into US_RHR and overwrites the previous one. The OVRE bit is cleared by writing the Control Register (US_CR) with the RSTSTA (Reset Status) bit to 1.
Figure 45-21. Receiver Status
Baud Rate Clock RXD
Start D0 Bit D1 D2 D3 D4 D5 D6 D7 Parity Stop Start D0 Bit Bit Bit D1 D2 D3 D4 D5 D6 D7 Parity Stop Bit Bit
RSTSTA = 1
Write US_CR Read US_RHR
RXRDY OVRE
45.7.3.8 Parity The USART supports five parity modes selected by programming the PAR field in the Mode Register (US_MR). The PAR field also enables the Multidrop mode, see Multidrop Mode on page 1295. Even and odd parity bit generation and error detection are supported. If even parity is selected, the parity generator of the transmitter drives the parity bit to 0 if a number of 1s in the character data bit is even, and to 1 if the number of 1s is odd. Accordingly, the receiver parity checker counts the number of received 1s and reports a parity error if the sampled parity bit does not correspond. If odd parity is selected, the parity generator of the transmitter drives the parity bit to 1 if a number of 1s in the character data bit is even, and to 0 if the number of 1s is odd. Accordingly, the receiver parity checker counts the number of received 1s and reports a parity error if the sampled parity bit does not correspond. If the mark parity is used, the parity generator of the transmitter drives the parity bit to 1 for all characters. The receiver parity checker reports an error if the parity bit is sampled to 0. If the space parity is used, the parity generator of the transmitter drives the parity bit to 0 for all characters. The receiver parity checker reports an error if the parity bit is sampled to 1. If parity is disabled, the transmitter does not generate any parity bit and the receiver does not report any parity error. Table 45-9 shows an example of the parity bit for the character 0x41 (character ASCII A) depending on the configuration of the USART. Because there are two bits to 1, 1 bit is added when a parity is odd, or 0 is added when a parity is even.

1294
Table 45-9. Parity Bit Examples Character A A A A A Hexa 0x41 0x41 0x41 0x41 0x41 Binary 0100 0001 0100 0001 0100 0001 0100 0001 0100 0001 Parity Bit 1 0 1 0 None Parity Mode Odd Even Mark Space None
When the receiver detects a parity error, it sets the PARE (Parity Error) bit in the Channel Status Register (US_CSR). The PARE bit can be cleared by writing the Control Register (US_CR) with the RSTSTA bit to 1. Figure 45-22 illustrates the parity bit status setting and clearing.
Baud Rate Clock RXD

Start D0 Bit D1 D2 D3 D4 D5 D6 D7 Bad Stop Parity Bit Bit
RSTSTA = 1
Write US_CR PARE
RXRDY
45.7.3.9 Multidrop Mode If the PAR field in the Mode Register (US_MR) is programmed to the value 0x6 or 0x07, the USART runs in Multidrop Mode. This mode differentiates the data characters and the address characters. Data is transmitted with the parity bit to 0 and addresses are transmitted with the parity bit to 1. If the USART is configured in multidrop mode, the receiver sets the PARE parity error bit when the parity bit is high and the transmitter is able to send a character with the parity bit high when the Control Register is written with the SENDA bit to 1. To handle parity error, the PARE bit is cleared when the Control Register is written with the bit RSTSTA to 1. The transmitter sends an address byte (parity bit set) when SENDA is written to US_CR. In this case, the next byte written to US_THR is transmitted as an address. Any character written in US_THR without having written the command SENDA is transmitted normally with the parity to 0.

1295
45.7.3.10 Transmitter Timeguard The timeguard feature enables the USART interface with slow remote devices. The timeguard function enables the transmitter to insert an idle state on the TXD line between two characters. This idle state actually acts as a long stop bit. The duration of the idle state is programmed in the TG field of the Transmitter Timeguard Register (US_TTGR). When this field is programmed to zero no timeguard is generated. Otherwise, the transmitter holds a high level on TXD after each transmitted byte during the number of bit periods programmed in TG in addition to the number of stop bits. As illustrated in Figure 45-23, the behavior of TXRDY and TXEMPTY status bits is modified by the programming of a timeguard. TXRDY rises only when the start bit of the next character is sent, and thus remains to 0 during the timeguard transmission if a character has been written in US_THR. TXEMPTY remains low until the timeguard transmission is completed as the timeguard is part of the current character being transmitted.
Figure 45-23. Timeguard Operations
TG = 4 Baud Rate Clock TXD
Start D0 Bit D1 D2 D3 D4 D5 D6 D7 Parity Stop Bit Bit Start D0 Bit D1 D2 D3 D4 D5 D6 D7 Parity Stop Bit Bit
TG = 4
Write US_THR TXRDY
TXEMPTY
Table 45-10 indicates the maximum length of a timeguard period that the transmitter can handle in relation to the function of the Baud Rate.
Table 45-10. Maximum Timeguard Length Depending on Baud Rate Baud Rate Bit/sec 1 200 9 600 14400 19200 28800 33400 56000 57600 115200 Bit time s 833 104 69.4 52.1 34.7 29.9 17.9 17.4 8.7 Timeguard ms 212.50 26.56 17.71 13.28 8.85 7.63 4.55 4.43 2.21

1296
45.7.3.11 Receiver Time-out The Receiver Time-out provides support in handling variable-length frames. This feature detects an idle condition on the RXD line. When a time-out is detected, the bit TIMEOUT in the Channel Status Register (US_CSR) rises and can generate an interrupt, thus indicating to the driver an end of frame. The time-out delay period (during which the receiver waits for a new character) is programmed in the TO field of the Receiver Time-out Register (US_RTOR). If the TO field is programmed to 0, the Receiver Time-out is disabled and no time-out is detected. The TIMEOUT bit in US_CSR remains to 0. Otherwise, the receiver loads a 16-bit counter with the value programmed in TO. This counter is decremented at each bit period and reloaded each time a new character is received. If the counter reaches 0, the TIMEOUT bit in the Status Register rises. Then, the user can either:
z
Stop the counter clock until a new character is received. This is performed by writing the Control Register (US_CR) with the STTTO (Start Time-out) bit to 1. In this case, the idle state on RXD before a new character is received will not provide a time-out. This prevents having to handle an interrupt before a character is received and allows waiting for the next idle state on RXD after a frame is received. Obtain an interrupt while no character is received. This is performed by writing US_CR with the RETTO (Reload and Start Time-out) bit to 1. If RETTO is performed, the counter starts counting down immediately from the value TO. This enables generation of a periodic interrupt so that a user time-out can be handled, for example when no key is pressed on a keyboard.
If STTTO is performed, the counter clock is stopped until a first character is received. The idle state on RXD before the start of the frame does not provide a time-out. This prevents having to obtain a periodic interrupt and enables a wait of the end of frame when the idle state on RXD is detected. If RETTO is performed, the counter starts counting down immediately from the value TO. This enables generation of a periodic interrupt so that a user time-out can be handled, for example when no key is pressed on a keyboard. Figure 45-24 shows the block diagram of the Receiver Time-out feature.
Figure 45-24. Receiver Time-out Block Diagram
Baud Rate Clock TO
1 STTTO
Clock
16-bit Time-out Counter Load
16-bit Value = TIMEOUT
Character Received RETTO
Clear

1297
Table 45-11 gives the maximum time-out period for some standard baud rates.
Table 45-11. Maximum Time-out Period Baud Rate bit/sec 600 1 200 2 400 4 800 9 600 14400 19200 28800 33400 56000 57600 200000 Bit Time s 1 667 833 417 208 104 69 52 35 30 18 17 5 Time-out ms 109 225 54 613 27 306 13 653 6 827 4 551 3 413 2 276 1 962 1 170 1 138 328
45.7.3.12 Framing Error The receiver is capable of detecting framing errors. A framing error happens when the stop bit of a received character is detected at level 0. This can occur if the receiver and the transmitter are fully desynchronized. A framing error is reported on the FRAME bit of the Channel Status Register (US_CSR). The FRAME bit is asserted in the middle of the stop bit as soon as the framing error is detected. It is cleared by writing the Control Register (US_CR) with the RSTSTA bit to 1.
Figure 45-25. Framing Error Status
Baud Rate Clock RXD
Start D0 Bit D1 D2 D3 D4 D5 D6 D7 Parity Stop Bit Bit
RSTSTA = 1
Write US_CR FRAME
RXRDY

1298
45.7.3.13 Transmit Break The user can request the transmitter to generate a break condition on the TXD line. A break condition drives the TXD line low during at least one complete character. It appears the same as a 0x00 character sent with the parity and the stop bits to 0. However, the transmitter holds the TXD line at least during one character until the user requests the break condition to be removed. A break is transmitted by writing the Control Register (US_CR) with the STTBRK bit to 1. This can be performed at any time, either while the transmitter is empty (no character in either the Shift Register or in US_THR) or when a character is being transmitted. If a break is requested while a character is being shifted out, the character is first completed before the TXD line is held low. Once STTBRK command is requested further STTBRK commands are ignored until the end of the break is completed. The break condition is removed by writing US_CR with the STPBRK bit to 1. If the STPBRK is requested before the end of the minimum break duration (one character, including start, data, parity and stop bits), the transmitter ensures that the break condition completes. The transmitter considers the break as though it is a character, i.e. the STTBRK and STPBRK commands are taken into account only if the TXRDY bit in US_CSR is to 1 and the start of the break condition clears the TXRDY and TXEMPTY bits as if a character is processed. Writing US_CR with both STTBRK and STPBRK bits to 1 can lead to an unpredictable result. All STPBRK commands requested without a previous STTBRK command are ignored. A byte written into the Transmit Holding Register while a break is pending, but not started, is ignored. After the break condition, the transmitter returns the TXD line to 1 for a minimum of 12 bit times. Thus, the transmitter ensures that the remote receiver detects correctly the end of break and the start of the next character. If the timeguard is programmed with a value higher than 12, the TXD line is held high for the timeguard period. After holding the TXD line for this period, the transmitter resumes normal operations. Figure 45-26 illustrates the effect of both the Start Break (STTBRK) and Stop Break (STPBRK) commands on the TXD line.
Figure 45-26. Break Transmission
Baud Rate Clock TXD
Break Transmission STPBRK = 1
End of Break
STTBRK = 1 Write US_CR TXRDY
TXEMPTY

1299
45.7.3.14 Receive Break The receiver detects a break condition when all data, parity and stop bits are low. This corresponds to detecting a framing error with data to 0x00, but FRAME remains low. When the low stop bit is detected, the receiver asserts the RXBRK bit in US_CSR. This bit may be cleared by writing the Control Register (US_CR) with the bit RSTSTA to 1. An end of receive break is detected by a high level for at least 2/16 of a bit period in asynchronous operating mode or one sample at high level in synchronous operating mode. The end of break detection also asserts the RXBRK bit. 45.7.3.15 Hardware Handshaking The USART features a hardware handshaking out-of-band flow control. The RTS and CTS pins are used to connect with the remote device, as shown in Figure 45-27.
Figure 45-27. Connection with a Remote Device for Hardware Handshaking
USART TXD RXD CTS RTS
Remote Device RXD TXD RTS CTS
Setting the USART to operate with hardware handshaking is performed by writing the USART_MODE field in the Mode Register (US_MR) to the value 0x2. The USART behavior when hardware handshaking is enabled is the same as the behavior in standard synchronous or asynchronous mode, except that the receiver drives the RTS pin as described below and the level on the CTS pin modifies the behavior of the transmitter as described below. Using this mode requires using the DMAC channel for reception. The transmitter can handle hardware handshaking in any case. Figure 45-28 shows how the transmitter operates if hardware handshaking is enabled. The CTS pin disables the transmitter. If a character is being processing, the transmitter is disabled only after the completion of the current character and transmission of the next character happens as soon as the pin CTS falls.
Figure 45-28. Transmitter Behavior when Operating with Hardware Handshaking
CTS
TXD

1300
45.7.4 ISO7816 Mode

The USART features an ISO7816-compatible operating mode. This mode permits interfacing with smart cards and Security Access Modules (SAM) communicating through an ISO7816 link. Both T = 0 and T = 1 protocols defined by the ISO7816 specification are supported. Setting the USART in ISO7816 mode is performed by writing the USART_MODE field in the Mode Register (US_MR) to the value 0x4 for protocol T = 0 and to the value 0x5 for protocol T = 1. 45.7.4.1 ISO7816 Mode Overview The ISO7816 is a half duplex communication on only one bidirectional line. The baud rate is determined by a division of the clock provided to the remote device (see Baud Rate Generator on page 1281). The USART connects to a smart card as shown in Figure 45-29. The TXD line becomes bidirectional and the Baud Rate Generator feeds the ISO7816 clock on the SCK pin. As the TXD pin becomes bidirectional, its output remains driven by the output of the transmitter but only when the transmitter is active while its input is directed to the input of the receiver. The USART is considered as the master of the communication as it generates the clock.
Figure 45-29. Connection of a Smart Card to the USART
USART SCK TXD CLK I/O Smart Card
When operating in ISO7816, either in T = 0 or T = 1 modes, the character format is fixed. The configuration is 8 data bits, even parity and 1 or 2 stop bits, regardless of the values programmed in the CHRL, MODE9, PAR and CHMODE fields. MSBF can be used to transmit LSB or MSB first. Parity Bit (PAR) can be used to transmit in normal or inverse mode. Refer to USART Mode Register on page 1317 and PAR: Parity Type on page 1318. The USART cannot operate concurrently in both receiver and transmitter modes as the communication is unidirectional at a time. It has to be configured according to the required mode by enabling or disabling either the receiver or the transmitter as desired. Enabling both the receiver and the transmitter at the same time in ISO7816 mode may lead to unpredictable results. The ISO7816 specification defines an inverse transmission format. Data bits of the character must be transmitted on the I/O line at their negative value. 45.7.4.2 Protocol T = 0 In T = 0 protocol, a character is made up of one start bit, eight data bits, one parity bit and one guard time, which lasts two bit times. The transmitter shifts out the bits and does not drive the I/O line during the guard time. If no parity error is detected, the I/O line remains to 1 during the guard time and the transmitter can continue with the transmission of the next character, as shown in Figure 45-30. If a parity error is detected by the receiver, it drives the I/O line to 0 during the guard time, as shown in Figure 45-31. This error bit is also named NACK, for Non Acknowledge. In this case, the character lasts 1 bit time more, as the guard time length is the same and is added to the error bit time which lasts 1 bit time.

1301
When the USART is the receiver and it detects an error, it does not load the erroneous character in the Receive Holding Register (US_RHR). It appropriately sets the PARE bit in the Status Register (US_SR) so that the software can handle the error.
Figure 45-30. T = 0 Protocol without Parity Error
Baud Rate Clock RXD Start Bit D0 D1 D2 D3 D4 D5 D6 D7 Parity Guard Guard Next Bit Time 1 Time 2 Start Bit
Figure 45-31. T = 0 Protocol with Parity Error

Baud Rate Clock I/O Start Bit D0 D1 D2 D3 D4 D5 D6 D7 Error Parity Guard Bit Time 1 Guard Start Time 2 Bit D0 D1
Repetition
Receive Error Counter

The USART receiver also records the total number of errors. This can be read in the Number of Error (US_NER) register. The NB_ERRORS field can record up to 255 errors. Reading US_NER automatically clears the NB_ERRORS field.
Receive NACK Inhibit

The USART can also be configured to inhibit an error. This can be achieved by setting the INACK bit in the Mode Register (US_MR). If INACK is to 1, no error signal is driven on the I/O line even if a parity bit is detected. Moreover, if INACK is set, the erroneous received character is stored in the Receive Holding Register, as if no error occurred and the RXRDY bit does rise.
Transmit Character Repetition

When the USART is transmitting a character and gets a NACK, it can automatically repeat the character before moving on to the next one. Repetition is enabled by writing the MAX_ITERATION field in the Mode Register (US_MR) at a value higher than 0. Each character can be transmitted up to eight times; the first transmission plus seven repetitions. If MAX_ITERATION does not equal zero, the USART repeats the character as many times as the value loaded in MAX_ITERATION. When the USART repetition number reaches MAX_ITERATION, the ITERATION bit is set in the Channel Status Register (US_CSR). If the repetition of the character is acknowledged by the receiver, the repetitions are stopped and the iteration counter is cleared. The ITERATION bit in US_CSR can be cleared by writing the Control Register with the RSIT bit to 1. Disable Successive Receive NACK The receiver can limit the number of successive NACKs sent back to the remote transmitter. This is programmed by setting the bit DSNACK in the Mode Register (US_MR). The maximum number of NACK transmitted is programmed in the MAX_ITERATION field. As soon as MAX_ITERATION is reached, the character is considered as correct, an acknowledge is sent on the line and the ITERATION bit in the Channel Status Register is set.

1302
45.7.4.3 Protocol T = 1 When operating in ISO7816 protocol T = 1, the transmission is similar to an asynchronous format with only one stop bit. The parity is generated when transmitting and checked when receiving. Parity error detection sets the PARE bit in the Channel Status Register (US_CSR).
45.7.5 IrDA Mode

The USART features an IrDA mode supplying half-duplex point-to-point wireless communication. It embeds the modulator and demodulator which allows a glueless connection to the infrared transceivers, as shown in Figure 45-32. The modulator and demodulator are compliant with the IrDA specification version 1.1 and support data transfer speeds ranging from 2.4 Kb/s to 115.2 Kb/s. The USART IrDA mode is enabled by setting the USART_MODE field in the Mode Register (US_MR) to the value 0x8. The IrDA Filter Register (US_IF) allows configuring the demodulator filter. The USART transmitter and receiver operate in a normal asynchronous mode and all parameters are accessible. Note that the modulator and the demodulator are activated.
Figure 45-32. Connection to IrDA Transceivers
USART Receiver Demodulator RXD RX TX Transmitter Modulator TXD
IrDA Transceivers
The receiver and the transmitter must be enabled or disabled according to the direction of the transmission to be managed. To receive IrDA signals, the following needs to be done: z Disable TX and Enable RX z Configure the TXD pin as PIO and set it as an output to 0 (to avoid LED emission). Disable the internal pull-up (better for power consumption). z Receive data 45.7.5.1 IrDA Modulation For baud rates up to and including 115.2 Kbits/sec, the RZI modulation scheme is used. 0 is represented by a light pulse of 3/16th of a bit time. Some examples of signal pulse duration are shown in Table 45-12.
Table 45-12. IrDA Pulse Duration Baud Rate 2.4 Kb/s 9.6 Kb/s 19.2 Kb/s 38.4 Kb/s 57.6 Kb/s 115.2 Kb/s Pulse Duration (3/16) 78.13 s 19.53 s 9.77 s 4.88 s 3.26 s 1.63 s

1303
Figure 45-33 shows an example of character transmission.

Figure 45-33. IrDA Modulation
Start Bit Transmitter Output 0 1 0 1 Data Bits 0 0 1 1 0 Stop Bit 1
TXD
Bit Period
3 16 Bit Period
45.7.5.2 IrDA Baud Rate Table 45-13 gives some examples of CD values, baud rate error and pulse duration. Note that the requirement on the maximum acceptable error of 1.87% must be met.
Table 45-13. IrDA Baud Rate Error Peripheral Clock 3 686 400 20 000 000 32 768 000 40 000 000 3 686 400 20 000 000 32 768 000 40 000 000 3 686 400 20 000 000 32 768 000 40 000 000 3 686 400 20 000 000 32 768 000 40 000 000 3 686 400 20 000 000 32 768 000 40 000 000 3 686 400 20 000 000 32 768 000 Baud Rate 115 200 115 200 115 200 115 200 57 600 57 600 57 600 57 600 38 400 38 400 38 400 38 400 19 200 19 200 19 200 19 200 9 600 9 600 9 600 9 600 2 400 2 400 2 400 CD 2 11 18 22 4 22 36 43 6 33 53 65 12 65 107 130 24 130 213 260 96 521 853 Baud Rate Error 0.00% 1.38% 1.25% 1.38% 0.00% 1.38% 1.25% 0.93% 0.00% 1.38% 0.63% 0.16% 0.00% 0.16% 0.31% 0.16% 0.00% 0.16% 0.16% 0.16% 0.00% 0.03% 0.04% Pulse Time 1.63 1.63 1.63 1.63 3.26 3.26 3.26 3.26 4.88 4.88 4.88 4.88 9.77 9.77 9.77 9.77 19.53 19.53 19.53 19.53 78.13 78.13 78.13

1304
45.7.5.3 IrDA Demodulator The demodulator is based on the IrDA Receive filter comprised of an 8-bit down counter which is loaded with the value programmed in US_IF. When a falling edge is detected on the RXD pin, the Filter Counter starts counting down at the Master Clock (MCK) speed. If a rising edge is detected on the RXD pin, the counter stops and is reloaded with US_IF. If no rising edge is detected when the counter reaches 0, the input of the receiver is driven low during one bit time. Figure 45-34 illustrates the operations of the IrDA demodulator.
Figure 45-34. IrDA Demodulator Operations
MCK
RXD
Counter Value
Receiver Input
4 3 Pulse Rejected
Pulse Accepted
The programmed value in the US_IF register must always meet the following criteria: TMCK * (IRDA_FILTER + 3) < 1.41 us As the IrDA mode uses the same logic as the ISO7816, note that the FI_DI_RATIO field in US_FIDI must be set to a value higher than 0 in order to assure IrDA communications operate correctly.
45.7.6 RS485 Mode

The USART features the RS485 mode to enable line driver control. While operating in RS485 mode, the USART behaves as though in asynchronous or synchronous mode and configuration of all the parameters is possible. The difference is that the RTS pin is driven high when the transmitter is operating. The behavior of the RTS pin is controlled by the TXEMPTY bit. A typical connection of the USART to a RS485 bus is shown in Figure 45-35.
Figure 45-35. Typical Connection to a RS485 Bus
USART
RXD
TXD RTS
Differential Bus
The USART is set in RS485 mode by programming the USART_MODE field in the Mode Register (US_MR) to the value 0x1. The RTS pin is at a level inverse to the TXEMPTY bit. Significantly, the RTS pin remains high when a timeguard is programmed so that the line can remain driven after the last character completion. Figure 45-36 gives an example of the RTS waveform during a character transmission when the timeguard is enabled.

1305
Figure 45-36. Example of RTS Drive with Timeguard

TG = 4 Baud Rate Clock TXD
Write US_THR TXRDY
TXEMPTY
RTS
45.7.7 SPI Mode

The Serial Peripheral Interface (SPI) Mode is a synchronous serial data link that provides communication with external devices in Master or Slave Mode. It also enables communication between processors if an external processor is connected to the system. The Serial Peripheral Interface is essentially a shift register that serially transmits data bits to other SPIs. During a data transfer, one SPI system acts as the master which controls the data flow, while the other devices act as slaves'' which have data shifted into and out by the master. Different CPUs can take turns being masters and one master may simultaneously shift data into multiple slaves. (Multiple Master Protocol is the opposite of Single Master Protocol, where one CPU is always the master while all of the others are always slaves.) However, only one slave may drive its output to write data back to the master at any given time. A slave device is selected when its NSS signal is asserted by the master. The USART in SPI Master mode can address only one SPI Slave because it can generate only one NSS signal. The SPI system consists of two data lines and two control lines: z Master Out Slave In (MOSI): This data line supplies the output data from the master shifted into the input of the slave. z Master In Slave Out (MISO): This data line supplies the output data from a slave to the input of the master. z Serial Clock (SCK): This control line is driven by the master and regulates the flow of the data bits. The master may transmit data at a variety of baud rates. The SCK line cycles once for each bit that is transmitted. z Slave Select (NSS): This control line allows the master to select or deselect the slave. 45.7.7.1 Modes of Operation The USART can operate in SPI Master Mode or in SPI Slave Mode. Operation in SPI Master Mode is programmed by writing to 0xE the USART_MODE field in the Mode Register. In this case the SPI lines must be connected as described below: z The MOSI line is driven by the output pin TXD z The MISO line drives the input pin RXD z The SCK line is driven by the output pin SCK z The NSS line is driven by the output pin RTS

1306
Operation in SPI Slave Mode is programmed by writing to 0xF the USART_MODE field in the Mode Register. In this case the SPI lines must be connected as described below: z The MOSI line drives the input pin RXD z The MISO line is driven by the output pin TXD z The SCK line drives the input pin SCK z The NSS line drives the input pin CTS In order to avoid unpredicted behavior, any change of the SPI Mode must be followed by a software reset of the transmitter and of the receiver (except the initial configuration after a hardware reset). 45.7.7.2 Baud Rate In SPI Mode, the baudrate generator operates in the same way as in USART synchronous mode: See Baud Rate in Synchronous Mode or SPI Mode on page 1283. However, there are some restrictions: In SPI Master Mode: z The external clock SCK must not be selected (USCLKS 0x3), and the bit CLKO must be set to 1 in the Mode Register (US_MR), in order to generate correctly the serial clock on the SCK pin. z To obtain correct behavior of the receiver and the transmitter, the value programmed in CD must be superior or equal to 6. z If the internal clock divided (MCK/DIV) is selected, the value programmed in CD must be even to ensure a 50:50 mark/space ratio on the SCK pin, this value can be odd if the internal clock is selected (MCK). In SPI Slave Mode: z The external clock (SCK) selection is forced regardless of the value of the USCLKS field in the Mode Register (US_MR). Likewise, the value written in US_BRGR has no effect, because the clock is provided directly by the signal on the USART SCK pin. z To obtain correct behavior of the receiver and the transmitter, the external clock (SCK) frequency must be at least 6 times lower than the system clock. 45.7.7.3 Data Transfer Up to 9 data bits are successively shifted out on the TXD pin at each rising or falling edge (depending of CPOL and CPHA) of the programmed serial clock. There is no Start bit, no Parity bit and no Stop bit. The number of data bits is selected by the CHRL field and the MODE 9 bit in the Mode Register (US_MR). The 9 bits are selected by setting the MODE 9 bit regardless of the CHRL field. The MSB data bit is always sent first in SPI Mode (Master or Slave). Four combinations of polarity and phase are available for data transfers. The clock polarity is programmed with the CPOL bit in the Mode Register. The clock phase is programmed with the CPHA bit. These two parameters determine the edges of the clock signal upon which data is driven and sampled. Each of the two parameters has two possible states, resulting in four possible combinations that are incompatible with one another. Thus, a master/slave pair must use the same parameter pair values to communicate. If multiple slaves are used and fixed in different configurations, the master must reconfigure itself each time it needs to communicate with a different slave.
Table 45-14. SPI Bus Protocol Mode SPI Bus Protocol Mode 0 1 2 3 CPOL 0 0 1 1 CPHA 1 0 1 0

1307
Figure 45-37. SPI Transfer Format (CPHA=1, 8 bits per transfer)

SCK cycle (for reference) SCK (CPOL = 0) 1 2 3 4 5 6 7 8
SCK (CPOL = 1)
MOSI SPI Master ->TXD SPI Slave -> RXD
MSB
LSB
MISO SPI Master ->RXD SPI Slave -> TXD
MSB
LSB
NSS SPI Master -> RTS SPI Slave -> CTS
Figure 45-38. SPI Transfer Format (CPHA=0, 8 bits per transfer)

SCK cycle (for reference) SCK (CPOL = 0) 1 2 3 4 5 6 7 8
SCK (CPOL = 1)
MOSI SPI Master -> TXD SPI Slave -> RXD
MSB
LSB
MISO SPI Master -> RXD SPI Slave -> TXD
MSB
LSB
NSS SPI Master -> RTS SPI Slave -> CTS

1308
45.7.7.4 Receiver and Transmitter Control See Receiver and Transmitter Control on page 1285. 45.7.7.5 Character Transmission The characters are sent by writing in the Transmit Holding Register (US_THR). An additional condition for transmitting a character can be added when the USART is configured in SPI master mode. In the USART_MR register, the value configured on INACK field can prevent any character transmission (even if US_THR has been written) while the receiver side is not ready (character not read). When WRDBT equals 0, the character is transmitted whatever the receiver status. If WRDBT is set to 1, the transmitter waits for the receiver holding register to be read before transmitting the character (RXRDY flag cleared), thus preventing any overflow (character loss) on the receiver side. The transmitter reports two status bits in the Channel Status Register (US_CSR): TXRDY (Transmitter Ready), which indicates that US_THR is empty and TXEMPTY, which indicates that all the characters written in US_THR have been processed. When the current character processing is completed, the last character written in US_THR is transferred into the Shift Register of the transmitter and US_THR becomes empty, thus TXRDY rises. Both TXRDY and TXEMPTY bits are low when the transmitter is disabled. Writing a character in US_THR while TXRDY is low has no effect and the written character is lost. If the USART is in SPI Slave Mode and if a character must be sent while the Transmit Holding Register (US_THR) is empty, the UNRE (Underrun Error) bit is set. The TXD transmission line stays at high level during all this time. The UNRE bit is cleared by writing the Control Register (US_CR) with the RSTSTA (Reset Status) bit to 1. In SPI Master Mode, the slave select line (NSS) is asserted at low level 1 Tbit (Time bit) before the transmission of the MSB bit and released at high level 1 Tbit after the transmission of the LSB bit. So, the slave select line (NSS) is always released between each character transmission and a minimum delay of 3 Tbits always inserted. However, in order to address slave devices supporting the CSAAT mode (Chip Select Active After Transfer), the slave select line (NSS) can be forced at low level by writing the Control Register (US_CR) with the RTSEN bit to 1. The slave select line (NSS) can be released at high level only by writing the Control Register (US_CR) with the RTSDIS bit to 1 (for example, when all data have been transferred to the slave device). In SPI Slave Mode, the transmitter does not require a falling edge of the slave select line (NSS) to initiate a character transmission but only a low level. However, this low level must be present on the slave select line (NSS) at least 1 Tbit before the first serial clock cycle corresponding to the MSB bit. 45.7.7.6 Character Reception When a character reception is completed, it is transferred to the Receive Holding Register (US_RHR) and the RXRDY bit in the Status Register (US_CSR) rises. If a character is completed while RXRDY is set, the OVRE (Overrun Error) bit is set. The last character is transferred into US_RHR and overwrites the previous one. The OVRE bit is cleared by writing the Control Register (US_CR) with the RSTSTA (Reset Status) bit to 1. To ensure correct behavior of the receiver in SPI Slave Mode, the master device sending the frame must ensure a minimum delay of 1 Tbit between each character transmission. The receiver does not require a falling edge of the slave select line (NSS) to initiate a character reception but only a low level. However, this low level must be present on the slave select line (NSS) at least 1 Tbit before the first serial clock cycle corresponding to the MSB bit. 45.7.7.7 Receiver Timeout Because the receiver baudrate clock is active only during data transfers in SPI Mode, a receiver timeout is impossible in this mode, whatever the Time-out value is (field TO) in the Time-out Register (US_RTOR).

1309
45.7.8 Test Modes

The USART can be programmed to operate in three different test modes. The internal loopback capability allows on-board diagnostics. In the loopback mode the USART interface pins are disconnected or not and reconfigured for loopback internally or externally. 45.7.8.1 Normal Mode Normal mode connects the RXD pin on the receiver input and the transmitter output on the TXD pin.
Figure 45-39. Normal Mode Configuration
RXD Receiver
TXD Transmitter
45.7.8.2 Automatic Echo Mode Automatic echo mode allows bit-by-bit retransmission. When a bit is received on the RXD pin, it is sent to the TXD pin, as shown in Figure 45-40. Programming the transmitter has no effect on the TXD pin. The RXD pin is still connected to the receiver input, thus the receiver remains active.
Figure 45-40. Automatic Echo Mode Configuration
RXD Receiver
TXD Transmitter
45.7.8.3 Local Loopback Mode Local loopback mode connects the output of the transmitter directly to the input of the receiver, as shown in Figure 45-41. The TXD and RXD pins are not used. The RXD pin has no effect on the receiver and the TXD pin is continuously driven high, as in idle state.
Figure 45-41. Local Loopback Mode Configuration
RXD Receiver
Transmitter
TXD

1310
45.7.8.4 Remote Loopback Mode Remote loopback mode directly connects the RXD pin to the TXD pin, as shown in Figure 45-42. The transmitter and the receiver are disabled and have no effect. This mode allows bit-by-bit retransmission.
Figure 45-42. Remote Loopback Mode Configuration
Receiver 1 RXD
TXD Transmitter

To prevent any single software error that may corrupt USART behavior, certain address spaces can be write-protected by setting the WPEN bit in the USART Write Protect Mode Register (US_WPMR). If a write access to the protected registers is detected, then the WPVS flag in the USART Write Protect Status Register (US_WPSR) is set and the field WPVSRC indicates in which register the write access has been attempted. The WPVS flag is automatically reset by reading the USART Write Protect Mode Register (US_WPMR) with the appropriate access key, WPKEY. The protected registers are:
z z z z z z z
USART Mode Register USART Baud Rate Generator Register USART Receiver Time-out Register USART Transmitter Timeguard Register USART FI DI RATIO Register USART IrDA FILTER Register USART Manchester Configuration Register

1311
45.8
Universal Synchronous Asynchronous Receiver Transmitter (USART) User Interface
Table 45-15. Register Mapping Offset 0x0000 0x0004 0x0008 0x000C 0x0010 0x0014 0x0018 0x001C 0x0020 0x0024 0x0028 0x2C - 0x3C 0x0040 0x0044 0x0048 0x004C 0x0050 0xE4 0xE8 0x5C - 0xFC Register Control Register Mode Register Interrupt Enable Register Interrupt Disable Register Interrupt Mask Register Channel Status Register Receiver Holding Register Transmitter Holding Register Baud Rate Generator Register Receiver Time-out Register Transmitter Timeguard Register Reserved FI DI Ratio Register Number of Errors Register Reserved IrDA Filter Register Manchester Encoder Decoder Register Write Protect Mode Register Write Protect Status Register Reserved Name US_CR US_MR US_IER US_IDR US_IMR US_CSR US_RHR US_THR US_BRGR US_RTOR US_TTGR US_FIDI US_NER US_IF US_MAN US_WPMR US_WPSR Access Write-only Read-write Write-only Write-only Read-only Read-only Read-only Write-only Read-write Read-write Read-write Read-write Read-only Read-write Read-write Read-write Read-only Reset 0x0 0x0 0x0 0x0 0x0 0x174 0x0 0xB0011004 0x0 0x0

1312
45.8.1 USART Control Register

31 23 15 RETTO 7 TXDIS
US_CR 0xF001C000 (0), 0xF0020000 (1), 0xF8020000 (2), 0xF8024000 (3) Write-only
30 22 14 RSTNACK 6 TXEN 29 21 13 RSTIT 5 RXDIS 28 20 12 SENDA 4 RXEN 27 19 RTSDIS 11 STTTO 3 RSTTX 26 18 RTSEN 10 STPBRK 2 RSTRX 25 17 9 STTBRK 1 24 16 8 RSTSTA 0
For SPI control, see USART Control Register (SPI_MODE) on page 1315.

0: No effect. 1: Resets the receiver.

0: No effect. 1: Resets the transmitter.

0: No effect. 1: Enables the receiver, if RXDIS is 0.

0: No effect. 1: Disables the receiver.

0: No effect. 1: Enables the transmitter if TXDIS is 0.

0: No effect. 1: Disables the transmitter.

1313

0: No effect. 1: Resets the status bits PARE, FRAME, OVRE, MANERR and RXBRK in US_CSR.
STTBRK: Start Break

0: No effect. 1: Starts transmission of a break after the characters present in US_THR and the Transmit Shift Register have been transmitted. No effect if a break is already being transmitted.
STPBRK: Stop Break

0: No effect. 1: Stops transmission of the break after a minimum of one character length and transmits a high level during 12-bit periods. No effect if no break is being transmitted.
STTTO: Start Time-out

0: No effect. 1: Starts waiting for a character before clocking the time-out counter. Resets the status bit TIMEOUT in US_CSR.
SENDA: Send Address

0: No effect. 1: In Multidrop Mode only, the next character written to the US_THR is sent with the address bit set.
RSTIT: Reset Iterations

0: No effect. 1: Resets ITERATION in US_CSR. No effect if the ISO7816 is not enabled.
RSTNACK: Reset Non Acknowledge

0: No effect 1: Resets NACK in US_CSR.
RETTO: Rearm Time-out

0: No effect 1: Restart Time-out
RTSEN: Request to Send Enable

0: No effect. 1: Drives the pin RTS to 0.
RTSDIS: Request to Send Disable

0: No effect. 1: Drives the pin RTS to 1.

1314
45.8.2 USART Control Register (SPI_MODE)

31 23 15 7 TXDIS
US_CR (SPI_MODE) 0xF001C000 (0), 0xF0020000 (1), 0xF8020000 (2), 0xF8024000 (3) Write-only
30 22 14 6 TXEN 29 21 13 5 RXDIS 28 20 12 4 RXEN 27 19 RCS 11 3 RSTTX 26 18 FCS 10 2 RSTRX 25 17 9 1 24 16 8 RSTSTA 0
This configuration is relevant only if USART_MODE=0xE or 0xF in USART Mode Register on page 1317.

0: No effect. 1: Resets the receiver.

0: No effect. 1: Resets the transmitter.

0: No effect. 1: Enables the receiver, if RXDIS is 0.

0: No effect. 1: Disables the receiver.

0: No effect. 1: Enables the transmitter if TXDIS is 0.

0: No effect. 1: Disables the transmitter.

1315

0: No effect. 1: Resets the status bits OVRE, UNRE in US_CSR.
FCS: Force SPI Chip Select Applicable if USART operates in SPI Master Mode (USART_MODE = 0xE):
FCS = 0: No effect. FCS = 1: Forces the Slave Select Line NSS (RTS pin) to 0, even if USART is no transmitting, in order to address SPI slave devices supporting the CSAAT Mode (Chip Select Active After Transfer).
RCS: Release SPI Chip Select Applicable if USART operates in SPI Master Mode (USART_MODE = 0xE):
RCS = 0: No effect. RCS = 1: Releases the Slave Select Line NSS (RTS pin).

1316
45.8.3 USART Mode Register

31 ONEBIT 23 INVDATA 15 CHMODE 7 CHRL 6 5 USCLKS
US_MR 0xF001C004 (0), 0xF0020004 (1), 0xF8020004 (2), 0xF8024004 (3) Read-write
30 MODSYNC 22 VAR_SYNC 14 29 MAN 21 DSNACK 13 NBSTOP 4 3 28 FILTER 20 INACK 12 27 19 OVER 11 26 25 MAX_ITERATION 17 MODE9 9 24
18 CLKO 10 PAR 2
16 MSBF 8 SYNC 0
1 USART_MODE
This register can only be written if the WPEN bit is cleared in USART Write Protect Mode Register on page 1341. For SPI configuration, see USART Mode Register (SPI_MODE) on page 1320.
USART_MODE: USART Mode of Operation

Value 0x0 0x1 0x2 0x4 0x6 0x8 0xE 0xF Name NORMAL RS485 HW_HANDSHAKING IS07816_T_0 IS07816_T_1 IRDA SPI_MASTER SPI_SLAVE Description Normal mode RS485 Hardware Handshaking IS07816 Protocol: T = 0 IS07816 Protocol: T = 1 IrDA SPI Master SPI Slave
USCLKS: Clock Selection

Value 0 1 3 Name MCK DIV SCK Description Master Clock MCK is selected Internal Clock Divided MCK/DIV (DIV=(DIV=8)) is selected Serial Clock SLK is selected
CHRL: Character Length.

Value 0 1 2 3 Name 5_BIT 6_BIT 7_BIT 8_BIT Description Character length is 5 bits Character length is 6 bits Character length is 7 bits Character length is 8 bits

1317
SYNC: Synchronous Mode Select

0: USART operates in Asynchronous Mode. 1: USART operates in Synchronous Mode.
PAR: Parity Type

Value 0 1 2 3 4 6 Name EVEN ODD SPACE MARK NO MULTIDROP Description Even parity Odd parity Parity forced to 0 (Space) Parity forced to 1 (Mark) No parity Multidrop mode
NBSTOP: Number of Stop Bits

Value 0 1 2 Name 1_BIT 1_5_BIT 2_BIT Description 1 stop bit 1.5 stop bit (SYNC = 0) or reserved (SYNC = 1) 2 stop bits

Value 0 1 2 3 Name NORMAL AUTOMATIC LOCAL_LOOPBACK REMOTE_LOOPBACK Description Normal Mode Automatic Echo. Receiver input is connected to the TXD pin. Local Loopback. Transmitter output is connected to the Receiver Input. Remote Loopback. RXD pin is internally connected to the TXD pin.
MSBF: Bit Order

0: Least Significant Bit is sent/received first. 1: Most Significant Bit is sent/received first.
MODE9: 9-bit Character Length

0: CHRL defines character length. 1: 9-bit character length.
CLKO: Clock Output Select

0: The USART does not drive the SCK pin. 1: The USART drives the SCK pin if USCLKS does not select the external clock SCK.
OVER: Oversampling Mode

0: 16x Oversampling. 1: 8x Oversampling.

1318
INACK: Inhibit Non Acknowledge

0: The NACK is generated. 1: The NACK is not generated.
DSNACK: Disable Successive NACK

0: NACK is sent on the ISO line as soon as a parity error occurs in the received character (unless INACK is set). 1: Successive parity errors are counted up to the value specified in the MAX_ITERATION field. These parity errors generate a NACK on the ISO line. As soon as this value is reached, no additional NACK is sent on the ISO line. The flag ITERATION is asserted.
INVDATA: Inverted Data

0: The data field transmitted on TXD line is the same as the one written in US_THR register or the content read in US_RHR is the same as RXD line. Normal mode of operation. 1: The data field transmitted on TXD line is inverted (voltage polarity only) compared to the value written on US_THR register or the content read in US_RHR is inverted compared to what is received on RXD line (or ISO7816 IO line). Inverted Mode of operation, useful for contactless card application. To be used with configuration bit MSBF.
VAR_SYNC: Variable Synchronization of Command/Data Sync Start Frame Delimiter

0: User defined configuration of command or data sync field depending on MODSYNC value. 1: The sync field is updated when a character is written into US_THR register.
MAX_ITERATION: Maximum Number of Automatic Iteration

0 - 7: Defines the maximum number of iterations in mode ISO7816, protocol T= 0.
FILTER: Infrared Receive Line Filter

0: The USART does not filter the receive line. 1: The USART filters the receive line using a three-sample filter (1/16-bit clock) (2 over 3 majority).
MAN: Manchester Encoder/Decoder Enable

0: Manchester Encoder/Decoder are disabled. 1: Manchester Encoder/Decoder are enabled.
MODSYNC: Manchester Synchronization Mode

0:The Manchester Start bit is a 0 to 1 transition 1: The Manchester Start bit is a 1 to 0 transition.
ONEBIT: Start Frame Delimiter Selector

0: Start Frame delimiter is COMMAND or DATA SYNC. 1: Start Frame delimiter is One Bit.

1319
45.8.4 USART Mode Register (SPI_MODE)

31 23 15 7 CHRL
US_MR (SPI_MODE) 0xF001C004 (0), 0xF0020004 (1), 0xF8020004 (2), 0xF8024004 (3) Read-write
30 22 14 6 29 21 13 5 USCLKS 28 20 WRDBT 12 4 27 19 26 18 10 2 USART_MODE 25 17 24 16 CPOL 8 CPHA 0
11 3
9 1
This configuration is relevant only if USART_MODE=0xE or 0xF in USART Mode Register on page 1317. This register can only be written if the WPEN bit is cleared in USART Write Protect Mode Register on page 1341.
USART_MODE: USART Mode of Operation

Value 0xE 0xF Name SPI_MASTER SPI_SLAVE Description SPI Master SPI Slave
USCLKS: Clock Selection

Value 0 1 3 Name MCK DIV SCK Description Master Clock MCK is selected Internal Clock Divided MCK/DIV (DIV=(DIV=8)) is selected Serial Clock SLK is selected
CHRL: Character Length.

Value 3 Name 8_BIT Description Character length is 8 bits
CPHA: SPI Clock Phase Applicable if USART operates in SPI Mode (USART_MODE = 0xE or 0xF):
CPHA = 0: Data is changed on the leading edge of SPCK and captured on the following edge of SPCK. CPHA = 1: Data is captured on the leading edge of SPCK and changed on the following edge of SPCK. CPHA determines which edge of SPCK causes data to change and which edge causes data to be captured. CPHA is used with CPOL to produce the required clock/data relationship between master and slave devices.

1320

Value 0 1 2 3 Name NORMAL AUTOMATIC LOCAL_LOOPBACK REMOTE_LOOPBACK Description Normal Mode Automatic Echo. Receiver input is connected to the TXD pin. Local Loopback. Transmitter output is connected to the Receiver Input. Remote Loopback. RXD pin is internally connected to the TXD pin.
CPOL: SPI Clock Polarity Applicable if USART operates in SPI Mode (Slave or Master, USART_MODE = 0xE or 0xF):
CPOL = 0: The inactive state value of SPCK is logic level zero. CPOL = 1: The inactive state value of SPCK is logic level one. CPOL is used to determine the inactive state value of the serial clock (SPCK). It is used with CPHA to produce the required clock/data relationship between master and slave devices.
WRDBT: Wait Read Data Before Transfer

0: The character transmission starts as soon as a character is written into US_THR register (assuming TXRDY was set). 1: The character transmission starts when a character is written and only if RXRDY flag is cleared (Receiver Holding Register has been read).

1321
45.8.5 USART Interrupt Enable Register

31 23 15 7 PARE
US_IER 0xF001C008 (0), 0xF0020008 (1), 0xF8020008 (2), 0xF8024008 (3) Write-only
30 22 14 6 FRAME 29 21 13 NACK 5 OVRE 28 20 12 4 27 19 CTSIC 11 3 26 18 10 ITER 2 RXBRK 25 17 9 TXEMPTY 1 TXRDY 24 MANE 16 8 TIMEOUT 0 RXRDY
For SPI specific configuration, see USART Interrupt Enable Register (SPI_MODE) on page 1323.
RXRDY: RXRDY Interrupt Enable TXRDY: TXRDY Interrupt Enable RXBRK: Receiver Break Interrupt Enable OVRE: Overrun Error Interrupt Enable FRAME: Framing Error Interrupt Enable PARE: Parity Error Interrupt Enable TIMEOUT: Time-out Interrupt Enable TXEMPTY: TXEMPTY Interrupt Enable ITER: Max number of Repetitions Reached Interrupt Enable NACK: Non Acknowledge Interrupt Enable CTSIC: Clear to Send Input Change Interrupt Enable MANE: Manchester Error Interrupt Enable
0: No effect 1: Enables the corresponding interrupt.

1322
45.8.6 USART Interrupt Enable Register (SPI_MODE)

31 23 15 7
US_IER (SPI_MODE) 0xF001C008 (0), 0xF0020008 (1), 0xF8020008 (2), 0xF8024008 (3) Write-only
30 22 14 6 29 21 13 5 OVRE 28 20 12 4 27 19 11 3 26 18 10 UNRE 2 25 17 9 TXEMPTY 1 TXRDY 24 16 8 0 RXRDY
RXRDY: RXRDY Interrupt Enable TXRDY: TXRDY Interrupt Enable OVRE: Overrun Error Interrupt Enable TXEMPTY: TXEMPTY Interrupt Enable UNRE: SPI Underrun Error Interrupt Enable
0: No effect 1: Enables the corresponding interrupt.

1323
45.8.7
Name:
USART Interrupt Disable Register

US_IDR 0xF001C00C (0), 0xF002000C (1), 0xF802000C (2), 0xF802400C (3) Write-only
Address: Access:
31 23 15 7 PARE
For SPI specific configuration, see USART Interrupt Disable Register (SPI_MODE) on page 1325.
RXRDY: RXRDY Interrupt Disable TXRDY: TXRDY Interrupt Disable RXBRK: Receiver Break Interrupt Disable OVRE: Overrun Error Interrupt Enable FRAME: Framing Error Interrupt Disable PARE: Parity Error Interrupt Disable TIMEOUT: Time-out Interrupt Disable TXEMPTY: TXEMPTY Interrupt Disable ITER: Max Number of Repetitions Reached Interrupt Disable NACK: Non Acknowledge Interrupt Disable CTSIC: Clear to Send Input Change Interrupt Disable MANE: Manchester Error Interrupt Disable
0: No effect 1: Disables the corresponding interrupt.

1324
45.8.8 USART Interrupt Disable Register (SPI_MODE)

31 23 15 7
US_IDR (SPI_MODE) 0xF001C00C (0), 0xF002000C (1), 0xF802000C (2), 0xF802400C (3) Write-only
RXRDY: RXRDY Interrupt Disable TXRDY: TXRDY Interrupt Disable OVRE: Overrun Error Interrupt Disable TXEMPTY: TXEMPTY Interrupt Disable UNRE: SPI Underrun Error Interrupt Disable
0: No effect 1: Disables the corresponding interrupt.

1325
45.8.9
Name:
USART Interrupt Mask Register

US_IMR 0xF001C010 (0), 0xF0020010 (1), 0xF8020010 (2), 0xF8024010 (3) Read-only
Address: Access:
31 23 15 7 PARE
For SPI specific configuration, see USART Interrupt Mask Register (SPI_MODE) on page 1327.
RXRDY: RXRDY Interrupt Mask TXRDY: TXRDY Interrupt Mask RXBRK: Receiver Break Interrupt Mask OVRE: Overrun Error Interrupt Mask FRAME: Framing Error Interrupt Mask PARE: Parity Error Interrupt Mask TIMEOUT: Time-out Interrupt Mask TXEMPTY: TXEMPTY Interrupt Mask ITER: Max Number of Repetitions Reached Interrupt Mask NACK: Non Acknowledge Interrupt Mask CTSIC: Clear to Send Input Change Interrupt Mask MANE: Manchester Error Interrupt Mask

1326
45.8.10 USART Interrupt Mask Register (SPI_MODE)

31 23 15 7
US_IMR (SPI_MODE) 0xF001C010 (0), 0xF0020010 (1), 0xF8020010 (2), 0xF8024010 (3) Read-only
RXRDY: RXRDY Interrupt Mask TXRDY: TXRDY Interrupt Mask OVRE: Overrun Error Interrupt Mask TXEMPTY: TXEMPTY Interrupt Mask UNRE: SPI Underrun Error Interrupt Mask

1327
45.8.11 USART Channel Status Register

31 23 CTS 15 7 PARE
US_CSR 0xF001C014 (0), 0xF0020014 (1), 0xF8020014 (2), 0xF8024014 (3) Read-only
30 22 14 6 FRAME 29 21 13 NACK 5 OVRE 28 20 12 4 27 19 CTSIC 11 3 26 18 10 ITER 2 RXBRK 25 17 9 TXEMPTY 1 TXRDY 24 MANERR 16 8 TIMEOUT 0 RXRDY
For SPI specific configuration, see USART Channel Status Register (SPI_MODE) on page 1330.

0: No complete character has been received since the last read of US_RHR or the receiver is disabled. If characters were being received when the receiver was disabled, RXRDY changes to 1 when the receiver is enabled. 1: At least one complete character has been received and US_RHR has not yet been read.

0: A character is in the US_THR waiting to be transferred to the Transmit Shift Register, or an STTBRK command has been requested, or the transmitter is disabled. As soon as the transmitter is enabled, TXRDY becomes 1. 1: There is no character in the US_THR.
RXBRK: Break Received/End of Break

0: No Break received or End of Break detected since the last RSTSTA. 1: Break Received or End of Break detected since the last RSTSTA.
OVRE: Overrun Error

0: No overrun error has occurred since the last RSTSTA. 1: At least one overrun error has occurred since the last RSTSTA.

0: No stop bit has been detected low since the last RSTSTA. 1: At least one stop bit has been detected low since the last RSTSTA.
PARE: Parity Error

0: No parity error has been detected since the last RSTSTA. 1: At least one parity error has been detected since the last RSTSTA.

1328
TIMEOUT: Receiver Time-out

0: There has not been a time-out since the last Start Time-out command (STTTO in US_CR) or the Time-out Register is 0. 1: There has been a time-out since the last Start Time-out command (STTTO in US_CR).

0: There are characters in either US_THR or the Transmit Shift Register, or the transmitter is disabled. 1: There are no characters in US_THR, nor in the Transmit Shift Register.
ITER: MaxNumber of Repetitions Reached

0: Maximum number of repetitions has not been reached since the last RSTSTA. 1: Maximum number of repetitions has been reached since the last RSTSTA.
NACK: Non Acknowledge Interrupt

0: Non Acknowledge has not been detected since the last RSTNACK. 1: At least one Non Acknowledge has been detected since the last RSTNACK.
CTSIC: Clear to Send Input Change Flag

0: No input change has been detected on the CTS pin since the last read of US_CSR. 1: At least one input change has been detected on the CTS pin since the last read of US_CSR.
CTS: Image of CTS Input

0: CTS is set to 0. 1: CTS is set to 1.
MANERR: Manchester Error

0: No Manchester error has been detected since the last RSTSTA. 1: At least one Manchester error has been detected since the last RSTSTA.

1329
45.8.12 USART Channel Status Register (SPI_MODE)

31 23 15 7
US_CSR (SPI_MODE) 0xF001C014 (0), 0xF0020014 (1), 0xF8020014 (2), 0xF8024014 (3) Read-only

0: No complete character has been received since the last read of US_RHR or the receiver is disabled. If characters were being received when the receiver was disabled, RXRDY changes to 1 when the receiver is enabled. 1: At least one complete character has been received and US_RHR has not yet been read.

0: A character is in the US_THR waiting to be transferred to the Transmit Shift Register or the transmitter is disabled. As soon as the transmitter is enabled, TXRDY becomes 1. 1: There is no character in the US_THR.
OVRE: Overrun Error

0: No overrun error has occurred since the last RSTSTA. 1: At least one overrun error has occurred since the last RSTSTA.

0: There are characters in either US_THR or the Transmit Shift Register, or the transmitter is disabled. 1: There are no characters in US_THR, nor in the Transmit Shift Register.
UNRE: Underrun Error

0: No SPI underrun error has occurred since the last RSTSTA. 1: At least one SPI underrun error has occurred since the last RSTSTA.

1330
45.8.13 USART Receive Holding Register

31 23 15 RXSYNH 7
US_RHR 0xF001C018 (0), 0xF0020018 (1), 0xF8020018 (2), 0xF8024018 (3) Read-only
30 22 14 6 29 21 13 5 28 20 12 4 RXCHR 27 19 11 3 26 18 10 2 25 17 9 1 24 16 8 RXCHR 0

Last character received if RXRDY is set.
RXSYNH: Received Sync

0: Last Character received is a Data. 1: Last Character received is a Command.

1331
45.8.14 USART Transmit Holding Register

31 23 15 TXSYNH 7
US_THR 0xF001C01C (0), 0xF002001C (1), 0xF802001C (2), 0xF802401C (3) Write-only
30 22 14 6 29 21 13 5 28 20 12 4 TXCHR 27 19 11 3 26 18 10 2 25 17 9 1 24 16 8 TXCHR 0

TXSYNH: Sync Field to be Transmitted

0: The next character sent is encoded as a data. Start Frame Delimiter is DATA SYNC. 1: The next character sent is encoded as a command. Start Frame Delimiter is COMMAND SYNC.

1332
45.8.15 USART Baud Rate Generator Register

31 23 15
US_BRGR 0xF001C020 (0), 0xF0020020 (1), 0xF8020020 (2), 0xF8024020 (3) Read-write
30 22 14 29 21 13 28 20 12 CD 7 6 5 4 CD 3 2 1 0 27 19 11 26 18 25 17 FP 9 24 16
10
This register can only be written if the WPEN bit is cleared in USART Write Protect Mode Register on page 1341.
CD: Clock Divider

USART_MODE ISO7816 SYNC = 1 or USART_MODE = SPI (Master or Slave) OVER = 1 Baud Rate Clock Disabled Baud Rate = Selected Clock/(16*CD) Baud Rate = Selected Clock/(8*CD) Baud Rate = Selected Clock /CD Baud Rate = Selected Clock/(FI_DI_RATIO*CD)
SYNC = 0 CD 0 1 to 65535 OVER = 0
USART_MODE = ISO7816
FP: Fractional Part

0: Fractional divider is disabled. 1 - 7: Baud rate resolution, defined by FP x 1/8.

1333
45.8.16 USART Receiver Time-out Register

31
US_RTOR 0xF001C024 (0), 0xF0020024 (1), 0xF8020024 (2), 0xF8024024 (3) Read-write
30 29 28 27 26 25 24
23 15
22 14
21 13
20 12 TO
19 11
18 10
17 9
16 8
4 TO
TO: Time-out Value

0: The Receiver Time-out is disabled. 1 - 65535: The Receiver Time-out is enabled and the Time-out delay is TO x Bit Period.

1334
45.8.17 USART Transmitter Timeguard Register

31 23 15 7
US_TTGR 0xF001C028 (0), 0xF0020028 (1), 0xF8020028 (2), 0xF8024028 (3) Read-write
30 22 14 6 29 21 13 5 28 20 12 4 TG 27 19 11 3 26 18 10 2 25 17 9 1 24 16 8 0
TG: Timeguard Value

0: The Transmitter Timeguard is disabled. 1 - 255: The Transmitter timeguard is enabled and the timeguard delay is TG x Bit Period.

1335
45.8.18 USART FI DI RATIO Register

31 23 15
US_FIDI 0xF001C040 (0), 0xF0020040 (1), 0xF8020040 (2), 0xF8024040 (3) Read-write 0x174
30 22 14 29 21 13 28 20 12 FI_DI_RATIO 7 6 5 4 FI_DI_RATIO 3 2 1 0 27 19 11 26 18 10 25 17 9 24 16 8
FI_DI_RATIO: FI Over DI Ratio Value

0: If ISO7816 mode is selected, the Baud Rate Generator generates no signal. 1 - 2047: If ISO7816 mode is selected, the Baud Rate is the clock provided on SCK divided by FI_DI_RATIO.

1336
45.8.19 USART Number of Errors Register

31 23 15 7
US_NER 0xF001C044 (0), 0xF0020044 (1), 0xF8020044 (2), 0xF8024044 (3) Read-only
30 22 14 6 29 21 13 5 28 20 12 4 NB_ERRORS 27 19 11 3 26 18 10 2 25 17 9 1 24 16 8 0
This register is relevant only if USART_MODE=0x4 or 0x6 in USART Mode Register on page 1317.
NB_ERRORS: Number of Errors

Total number of errors that occurred during an ISO7816 transfer. This register automatically clears when read.

1337
45.8.20 USART IrDA FILTER Register

31 23 15 7
US_IF 0xF001C04C (0), 0xF002004C (1), 0xF802004C (2), 0xF802404C (3) Read-write
30 22 14 6 29 21 13 5 28 20 12 4 IRDA_FILTER 27 19 11 3 26 18 10 2 25 17 9 1 24 16 8 0
This register is relevant only if USART_MODE=0x8 in USART Mode Register on page 1317. This register can only be written if the WPEN bit is cleared in USART Write Protect Mode Register on page 1341.
IRDA_FILTER: IrDA Filter

The IRDA_FILTER value must be defined to meet the following criteria:
TMCK * (IRDA_FILTER + 3) < 1.41 us

1338
45.8.21 USART Manchester Configuration Register

31 23 15 7
US_MAN 0xF001C050 (0), 0xF0020050 (1), 0xF8020050 (2), 0xF8024050 (3) Read-write
30 DRIFT 22 14 6 29 ONE 21 13 5 28 RX_MPOL 20 12 TX_MPOL 4 27 19 26 18 RX_PL 11 3 10 2 TX_PL 9 TX_PP 1 0 8 25 RX_PP 17 16 24
TX_PL: Transmitter Preamble Length 0: The Transmitter Preamble pattern generation is disabled
1 - 15: The Preamble Length is TX_PL x Bit Period
TX_PP: Transmitter Preamble Pattern The following values assume that TX_MPOL field is not set:
Value 00 01 10 11 Name ALL_ONE ALL_ZERO ZERO_ONE ONE_ZERO Description The preamble is composed of 1s The preamble is composed of 0s The preamble is composed of 01s The preamble is composed of 10s
TX_MPOL: Transmitter Manchester Polarity 0: Logic Zero is coded as a zero-to-one transition, Logic One is coded as a one-to-zero transition.
1: Logic Zero is coded as a one-to-zero transition, Logic One is coded as a zero-to-one transition.
RX_PL: Receiver Preamble Length 0: The receiver preamble pattern detection is disabled
1 - 15: The detected preamble length is RX_PL x Bit Period
RX_PP: Receiver Preamble Pattern detected The following values assume that RX_MPOL field is not set:
Value 00 01 10 11 Name ALL_ONE ALL_ZERO ZERO_ONE ONE_ZERO Description The preamble is composed of 1s The preamble is composed of 0s The preamble is composed of 01s The preamble is composed of 10s

1339
RX_MPOL: Receiver Manchester Polarity 0: Logic Zero is coded as a zero-to-one transition, Logic One is coded as a one-to-zero transition.
1: Logic Zero is coded as a one-to-zero transition, Logic One is coded as a zero-to-one transition.
ONE: Must Be Set to 1 Bit 29 must always be set to 1 when programming the US_MAN register. DRIFT: Drift Compensation 0: The USART can not recover from an important clock drift
1: The USART can recover from clock drift. The 16X clock mode must be enabled.

1340
45.8.22 USART Write Protect Mode Register

31
US_WPMR 0xF001C0E4 (0), 0xF00200E4 (1), 0xF80200E4 (2), 0xF80240E4 (3) Read-write See Table 45-15

0 = Disables the Write Protect if WPKEY corresponds to 0x555341 (USA in ASCII). 1 = Enables the Write Protect if WPKEY corresponds to 0x555341 (USA in ASCII). Protects the registers:
USART Mode Register on page 1317 USART Baud Rate Generator Register on page 1333 USART Receiver Time-out Register on page 1334 USART Transmitter Timeguard Register on page 1335 USART FI DI RATIO Register on page 1336 USART IrDA FILTER Register on page 1338 USART Manchester Configuration Register on page 1339 WPKEY: Write Protect KEY
Should be written at value 0x555341 (USA in ASCII). Writing any other value in this field aborts the write operation of the WPEN bit. Always reads as 0.

1341
45.8.23 USART Write Protect Status Register

31 23
US_WPSR 0xF001C0E8 (0), 0xF00200E8 (1), 0xF80200E8 (2), 0xF80240E8 (3) Read-only See Table 45-15

0 = No Write Protect Violation has occurred since the last read of the US_WPSR register. 1 = A Write Protect Violation has occurred since the last read of the US_WPSR register. If this violation is an unauthorized attempt to write a protected register, the associated violation is reported into field WPVSRC.

When WPVS is active, this field indicates the write-protected register (through address offset or code) in which a write access has been attempted. Note: Reading US_WPSR automatically clears all fields.

1342
46.
46.1
Controller Area Network (CAN)

Description
The CAN controller provides all the features required to implement the serial communication protocol CAN defined by Robert Bosch GmbH, the CAN specification as referred to by ISO/11898A (2.0 Part A and 2.0 Part B) for high speeds and ISO/11519-2 for low speeds. The CAN Controller is able to handle all types of frames (Data, Remote, Error and Overload) and achieves a bitrate of 1 Mbit/sec. CAN controller accesses are made through configuration registers. 8 independent message objects (mailboxes) are implemented. Any mailbox can be programmed as a reception buffer block (even non-consecutive buffers). For the reception of defined messages, one or several message objects can be masked without participating in the buffer feature. An interrupt is generated when the buffer is full. According to the mailbox configuration, the first message received can be locked in the CAN controller registers until the application acknowledges it, or this message can be discarded by new received messages. Any mailbox can be programmed for transmission. Several transmission mailboxes can be enabled in the same time. A priority can be defined for each mailbox independently. An internal 16-bit timer is used to stamp each received and sent message. This timer starts counting as soon as the CAN controller is enabled. This counter can be reset by the application or automatically after a reception in the last mailbox in Time Triggered Mode. The CAN controller offers optimized features to support the Time Triggered Communication (TTC) protocol.
46.2
z z z
Fully Compliant with CAN 2.0 Part A and 2.0 Part B Bit Rates up to 1 Mbps 8 Object Oriented Mailboxes with the Following Properties:
z z z z z z
CAN Specification 2.0 Part A or 2.0 Part B Programmable for Each Message Object Configurable in Receive (with Overwrite or Not) or Transmit Modes Independent 29-bit Identifier and Mask Defined for Each Mailbox 32-bit Access to Data Registers for Each Mailbox Data Object Uses a 16-bit Timestamp on Receive and Transmit Messages Hardware Concatenation of ID Masked Bitfields To Speed Up Family ID Processing
z z z z z z z
16-bit Internal Timer for Timestamping and Network Synchronization Programmable Reception Buffer Length up to 8 Mailbox Objects Priority Management between Transmission Mailboxes Autobaud and Listening Mode Low-power Mode and Programmable Wake-up on Bus Activity or by the Application Data, Remote, Error and Overload Frame Handling Write Protected Registers

1343
46.3
Block Diagram
Figure 46-1. CAN Block Diagram
Controller Area Network

CANRX CAN Protocol Controller
PIO
CANTX
Error Counter
Mailbox Priority Encoder
Control & Status MB0 MB1 MCK PMC
MBx
(x = number of mailboxes - 1)
CAN Interrupt User Interface
Internal Bus
46.4
Figure 46-2.
Layers
CAN-based Profiles CAN-based Application Layer CAN Data Link Layer CAN Physical Layer
Implementation
Software Software CAN Controller Transceiver

1344
46.5
Table 46-1. I/O Lines Description Name CANRX CANTX Description CAN Receive Serial Data CAN Transmit Serial Data Type Input Output
46.6
46.6.1 I/O Lines

The pins used for interfacing the CAN may be multiplexed with the PIO lines. The programmer must first program the PIO controller to assign the desired CAN pins to their peripheral function. If I/O lines of the CAN are not used by the application, they can be used for other purposes by the PIO Controller.
Table 46-2. I/O Lines Instance CAN0 CAN0 CAN1 CAN1 Signal CANRX0 CANTX0 CANRX1 CANTX1 I/O Line PD14 PD15 PB14 PB15 Peripheral C C B B

The programmer must first enable the CAN clock in the Power Management Controller (PMC) before using the CAN. A Low-power Mode is defined for the CAN controller. If the application does not require CAN operations, the CAN clock can be stopped when not needed and be restarted later. Before stopping the clock, the CAN Controller must be in Lowpower Mode to complete the current transfer. After restarting the clock, the application must disable the Low-power Mode of the CAN controller.
46.6.3 Interrupt
The CAN interrupt line is connected on one of the internal sources of the Advanced Interrupt Controller. Using the CAN interrupt requires the AIC to be programmed first. Note that it is not recommended to use the CAN interrupt line in edgesensitive mode.
Table 46-3. Peripheral IDs Instance CAN0 CAN1 ID 40 41

1345
46.7
CAN Controller Features
46.7.1 CAN Protocol Overview

The Controller Area Network (CAN) is a multi-master serial communication protocol that efficiently supports real-time control with a very high level of security with bit rates up to 1 Mbit/s. The CAN protocol supports four different frame types:
z z
Data frames: They carry data from a transmitter node to the receiver nodes. The overall maximum data frame length is 108 bits for a standard frame and 128 bits for an extended frame. Remote frames: A destination node can request data from the source by sending a remote frame with an identifier that matches the identifier of the required data frame. The appropriate data source node then sends a data frame as a response to this node request. Error frames: An error frame is generated by any node that detects a bus error. Overload frames: They provide an extra delay between the preceding and the successive data frames or remote frames.
z z
The Atmel CAN controller provides the CPU with full functionality of the CAN protocol V2.0 Part A and V2.0 Part B. It minimizes the CPU load in communication overhead. The Data Link Layer and part of the physical layer are automatically handled by the CAN controller itself. The CPU reads or writes data or messages via the CAN controller mailboxes. An identifier is assigned to each mailbox. The CAN controller encapsulates or decodes data messages to build or to decode bus data frames. Remote frames, error frames and overload frames are automatically handled by the CAN controller under supervision of the software application.
46.7.2 Mailbox Organization

The CAN module has 8 buffers, also called channels or mailboxes. An identifier that corresponds to the CAN identifier is defined for each active mailbox. Message identifiers can match the standard frame identifier or the extended frame identifier. This identifier is defined for the first time during the CAN initialization, but can be dynamically reconfigured later so that the mailbox can handle a new message family. Several mailboxes can be configured with the same ID. Each mailbox can be configured in receive or in transmit mode independently. The mailbox object type is defined in the MOT field of the CAN_MMRx register. 46.7.2.1 Message Acceptance Procedure If the MIDE field in the CAN_MIDx register is set, the mailbox can handle the extended format identifier; otherwise, the mailbox handles the standard format identifier. Once a new message is received, its ID is masked with the CAN_MAMx value and compared with the CAN_MIDx value. If accepted, the message ID is copied to the CAN_MIDx register.

1346
Figure 46-3. Message Acceptance Procedure

CAN_MIDx CAN_MAMx Message Received
&
&
==
Yes Message Accepted
No
Message Refused
CAN_MFIDx
If a mailbox is dedicated to receiving several messages (a family of messages) with different IDs, the acceptance mask defined in the CAN_MAMx register must mask the variable part of the ID family. Once a message is received, the application must decode the masked bits in the CAN_MIDx. To speed up the decoding, masked bits are grouped in the family ID register (CAN_MFIDx). For example, if the following message IDs are handled by the same mailbox: ID0 101000100100010010000100 0 11 00b ID1 101000100100010010000100 0 11 01b ID2 101000100100010010000100 0 11 10b ID3 101000100100010010000100 0 11 11b ID4 101000100100010010000100 1 11 00b ID5 101000100100010010000100 1 11 01b ID6 101000100100010010000100 1 11 10b ID7 101000100100010010000100 1 11 11b The CAN_MIDx and CAN_MAMx of Mailbox x must be initialized to the corresponding values: CAN_MIDx = 001 101000100100010010000100 x 11 xxb CAN_MAMx = 001 111111111111111111111111 0 11 00b If Mailbox x receives a message with ID6, then CAN_MIDx and CAN_MFIDx are set: CAN_MIDx = 001 101000100100010010000100 1 11 10b CAN_MFIDx = 00000000000000000000000000000110b If the application associates a handler for each message ID, it may define an array of pointers to functions: void (*pHandler[8])(void); When a message is received, the corresponding handler can be invoked using CAN_MFIDx register and there is no need to check masked bits: unsigned int MFID0_register; MFID0_register = Get_CAN_MFID0_Register(); // Get_CAN_MFID0_Register() returns the value of the CAN_MFID0 register pHandler[MFID0_register]();

1347
46.7.2.2 Receive Mailbox When the CAN module receives a message, it looks for the first available mailbox with the lowest number and compares the received message ID with the mailbox ID. If such a mailbox is found, then the message is stored in its data registers. Depending on the configuration, the mailbox is disabled as long as the message has not been acknowledged by the application (Receive only), or, if new messages with the same ID are received, then they overwrite the previous ones (Receive with overwrite). It is also possible to configure a mailbox in Consumer Mode. In this mode, after each transfer request, a remote frame is automatically sent. The first answer received is stored in the corresponding mailbox data registers. Several mailboxes can be chained to receive a buffer. They must be configured with the same ID in Receive Mode, except for the last one, which can be configured in Receive with Overwrite Mode. The last mailbox can be used to detect a buffer overflow.
Table 46-4. Receive Mailbox Object Types Mailbox Object Type Receive Description The first message received is stored in mailbox data registers. Data remain available until the next transfer request. The last message received is stored in mailbox data register. The next message always overwrites the previous one. The application has to check whether a new message has not overwritten the current one while reading the data registers. A remote frame is sent by the mailbox. The answer received is stored in mailbox data register. This extends Receive mailbox features. Data remain available until the next transfer request.
Receive with overwrite
Consumer
46.7.2.3 Transmit Mailbox When transmitting a message, the message length and data are written to the transmit mailbox with the correct identifier. For each transmit mailbox, a priority is assigned. The controller automatically sends the message with the highest priority first (set with the field PRIOR in CAN_MMRx register). It is also possible to configure a mailbox in Producer Mode. In this mode, when a remote frame is received, the mailbox data are sent automatically. By enabling this mode, a producer can be done using only one mailbox instead of two: one to detect the remote frame and one to send the answer.
Table 46-5. Transmit Mailbox Object Types Mailbox Object Type Transmit Description The message stored in the mailbox data registers will try to win the bus arbitration immediately or later according to or not the Time Management Unit configuration (see Section 46.7.3). The application is notified that the message has been sent or aborted. Producer The message prepared in the mailbox data registers will be sent after receiving the next remote frame. This extends transmit mailbox features.

1348
46.7.3 Time Management Unit

The CAN Controller integrates a free-running 16-bit internal timer. The counter is driven by the bit clock of the CAN bus line. It is enabled when the CAN controller is enabled (CANEN set in the CAN_MR register). It is automatically cleared in the following cases:
z z z z
After a reset When the CAN controller is in Low-power Mode is enabled (LPM bit set in the CAN_MR and SLEEP bit set in the CAN_SR) After a reset of the CAN controller (CANEN bit in the CAN_MR register) In Time-triggered Mode, when a message is accepted by the last mailbox (rising edge of the MRDY signal in the CAN_MSRlast_mailbox_number register).
The application can also reset the internal timer by setting TIMRST in the CAN_TCR register. The current value of the internal timer is always accessible by reading the CAN_TIM register. When the timer rolls-over from FFFFh to 0000h, TOVF (Timer Overflow) signal in the CAN_SR register is set. TOVF bit in the CAN_SR register is cleared by reading the CAN_SR register. Depending on the corresponding interrupt mask in the CAN_IMR register, an interrupt is generated while TOVF is set. In a CAN network, some CAN devices may have a larger counter. In this case, the application can also decide to freeze the internal counter when the timer reaches FFFFh and to wait for a restart condition from another device. This feature is enabled by setting TIMFRZ in the CAN_MR register. The CAN_TIM register is frozen to the FFFFh value. A clear condition described above restarts the timer. A timer overflow (TOVF) interrupt is triggered. To monitor the CAN bus activity, the CAN_TIM register is copied to the CAN _TIMESTP register after each start of frame or end of frame and a TSTP interrupt is triggered. If TEOF bit in the CAN_MR register is set, the value is captured at each End Of Frame, else it is captured at each Start Of Frame. Depending on the corresponding mask in the CAN_IMR register, an interrupt is generated while TSTP is set in the CAN_SR. TSTP bit is cleared by reading the CAN_SR register. The time management unit can operate in one of the two following modes:
z z
Timestamping mode: The value of the internal timer is captured at each Start Of Frame or each End Of Frame Time Triggered mode: A mailbox transfer operation is triggered when the internal timer reaches the mailbox trigger.
Timestamping Mode is enabled by clearing TTM field in the CAN_MR register. Time Triggered Mode is enabled by setting TTM field in the CAN_MR register.
46.7.4 CAN 2.0 Standard Features

46.7.4.1 CAN Bit Timing Configuration All controllers on a CAN bus must have the same bit rate and bit length. At different clock frequencies of the individual controllers, the bit rate has to be adjusted by the time segments. The CAN protocol specification partitions the nominal bit time into four different segments:
Figure 46-4. Partition of the CAN Bit Time
NOMINAL BIT TIME
SYNC_SEG
PROP_SEG
PHASE_SEG1
PHASE_SEG2
Sample Point

1349
TIME QUANTUM
The TIME QUANTUM (TQ) is a fixed unit of time derived from the MCK period. The total number of TIME QUANTA in a bit time is programmable from 8 to 25. SYNC SEG: SYNChronization Segment. This part of the bit time is used to synchronize the various nodes on the bus. An edge is expected to lie within this segment. It is 1 TQ long.
z
PROP SEG: PROPagation Segment.
This part of the bit time is used to compensate for the physical delay times within the network. It is twice the sum of the signals propagation time on the bus line, the input comparator delay, and the output driver delay. It is programmable to be 1,2,..., 8 TQ long. This parameter is defined in the PROPAG field of the CAN Baudrate Register.
z
PHASE SEG1, PHASE SEG2: PHASE Segment 1 and 2.
The Phase-Buffer-Segments are used to compensate for edge phase errors. These segments can be lengthened (PHASE SEG1) or shortened (PHASE SEG2) by resynchronization. Phase Segment 1 is programmable to be 1,2,..., 8 TQ long. Phase Segment 2 length has to be at least as long as the Information Processing Time (IPT) and may not be more than the length of Phase Segment 1. These parameters are defined in the PHASE1 and PHASE2 fields of the CAN Baudrate Register.
z
INFORMATION PROCESSING TIME:
The Information Processing Time (IPT) is the time required for the logic to determine the bit level of a sampled bit. The IPT begins at the sample point, is measured in TQ and is fixed at 2 TQ for the Atmel CAN. Since Phase Segment 2 also begins at the sample point and is the last segment in the bit time, PHASE SEG2 shall not be less than the IPT.
z
SAMPLE POINT:
The SAMPLE POINT is the point in time at which the bus level is read and interpreted as the value of that respective bit. Its location is at the end of PHASE_SEG1.
z
SJW: ReSynchronization Jump Width.
The ReSynchronization Jump Width defines the limit to the amount of lengthening or shortening of the Phase Segments. SJW is programmable to be the minimum of PHASE SEG1 and 4 TQ. If the SMP field in the CAN_BR register is set, then the incoming bit stream is sampled three times with a period of half a CAN clock period, centered on sample point. In the CAN controller, the length of a bit on the CAN bus is determined by the parameters (BRP, PROPAG, PHASE1 and PHASE2).
t BIT = t CSC + t PRS + t PHS1 + t PHS2

The time quantum is calculated as follows:
t CSC = ( BRP + 1 ) MCK

Note: The BRP field must be within the range [1, 0x7F], i.e., BRP = 0 is not authorized.
t PRS = t CSC ( PROPAG + 1 ) t PHS1 = t CSC ( PHASE1 + 1 ) t PHS2 = t CSC ( PHASE2 + 1 ) To compensate for phase shifts between clock oscillators of different controllers on the bus, the CAN controller must resynchronize on any relevant signal edge of the current transmission. The resynchronization shortens or lengthens the bit time so that the position of the sample point is shifted with regard to the detected edge. The resynchronization jump width (SJW) defines the maximum of time by which a bit period may be shortened or lengthened by resynchronization. t SJW = t CSC ( SJW + 1 )

1350
Figure 46-5. CAN Bit Timing

MCK
CAN Clock tCSC tPRS tPHS1 tPHS2
NOMINAL BIT TIME

SYNC_ SEG
PROP_SEG
PHASE_SEG1
PHASE_SEG2
Sample Point
Transmission Point
Example of bit timing determination for CAN baudrate of 500 Kbit/s: MCK = 48MHz CAN baudrate= 500kbit/s => bit time= 2us Delay of the bus driver: 50 ns Delay of the receiver: 30ns Delay of the bus line (20m): 110ns The total number of time quanta in a bit time must be comprised between 8 and 25. If we fix the bit time to 16 time quanta: Tcsc = 1 time quanta = bit time / 16 = 125 ns => BRP = (Tcsc x MCK) - 1 = 5 The propagation segment time is equal to twice the sum of the signals propagation time on the bus line, the receiver delay and the output driver delay: Tprs = 2 * (50+30+110) ns = 380 ns = 3 Tcsc => PROPAG = Tprs/Tcsc - 1 = 2 The remaining time for the two phase segments is: Tphs1 + Tphs2 = bit time - Tcsc - Tprs = (16 - 1 - 3)Tcsc Tphs1 + Tphs2 = 12 Tcsc Because this number is even, we choose Tphs2 = Tphs1 (else we would choose Tphs2 = Tphs1 + Tcsc) Tphs1 = Tphs2 = (12/2) Tcsc = 6 Tcsc => PHASE1 = PHASE2 = Tphs1/Tcsc - 1 = 5 The resynchronization jump width must be comprised between 1 Tcsc and the minimum of 4 Tcsc and Tphs1. We choose its maximum value: Tsjw = Min(4 Tcsc,Tphs1) = 4 Tcsc => SJW = Tsjw/Tcsc - 1 = 3 Finally: CAN_BR = 0x00053255

1351
CAN Bus Synchronization

Two types of synchronization are distinguished: hard synchronization at the start of a frame and resynchronization inside a frame. After a hard synchronization, the bit time is restarted with the end of the SYNC_SEG segment, regardless of the phase error. Resynchronization causes a reduction or increase in the bit time so that the position of the sample point is shifted with respect to the detected edge. The effect of resynchronization is the same as that of hard synchronization when the magnitude of the phase error of the edge causing the resynchronization is less than or equal to the programmed value of the resynchronization jump width (tSJW). When the magnitude of the phase error is larger than the resynchronization jump width and
z z
The phase error is positive, then PHASE_SEG1 is lengthened by an amount equal to the resynchronization jump width. The phase error is negative, then PHASE_SEG2 is shortened by an amount equal to the resynchronization jump width.
Figure 46-6. CAN Resynchronization

THE PHASE ERROR IS POSITIVE (the transmitter is slower than the receiver) Received data bit Nominal Sample point Sample point after resynchronization
Nominal bit time (before resynchronization)
SYNC_ SEG
PROP_SEG
PHASE_SEG1 PHASE_SEG2
SYNC_ SEG
Phase error Bit time with resynchronization
Phase error (max Tsjw)
SYNC_ SEG
PROP_SEG
PHASE_SEG1
PHASE_SEG2
SYNC_ SEG
THE PHASE ERROR IS NEGATIVE (the transmitter is faster than the receiver) Received data bit
Sample point after resynchronization
Nominal Sample point
Nominal bit time (before resynchronization)
PHASE_SEG2
SYNC_ SEG
PROP_SEG
SYNC_ SEG
Phase error
Bit time with resynchronization
PHASE_ SYNC_ SEG2 SEG
PROP_SEG
SYNC_ SEG
Phase error (max Tsjw)
Autobaud Mode
The autobaud feature is enabled by setting the ABM field in the CAN_MR register. In this mode, the CAN controller is only listening to the line without acknowledging the received messages. It can not send any message. The errors flags are updated. The bit timing can be adjusted until no error occurs (good configuration found). In this mode, the error counters are frozen. To go back to the standard mode, the ABM bit must be cleared in the CAN_MR register.

1352
46.7.4.2 Error Detection There are five different error types that are not mutually exclusive. Each error concerns only specific fields of the CAN data frame (refer to the Bosch CAN specification for their correspondence):
z
CRC error (CERR bit in the CAN_SR register): With the CRC, the transmitter calculates a checksum for the CRC bit sequence from the Start of Frame bit until the end of the Data Field. This CRC sequence is transmitted in the CRC field of the Data or Remote Frame. Bit-stuffing error (SERR bit in the CAN_SR register): If a node detects a sixth consecutive equal bit level during the bit-stuffing area of a frame, it generates an Error Frame starting with the next bit-time. Bit error (BERR bit in CAN_SR register): A bit error occurs if a transmitter sends a dominant bit but detects a recessive bit on the bus line, or if it sends a recessive bit but detects a dominant bit on the bus line. An error frame is generated and starts with the next bit time. Form Error (FERR bit in the CAN_SR register): If a transmitter detects a dominant bit in one of the fix-formatted segments CRC Delimiter, ACK Delimiter or End of Frame, a form error has occurred and an error frame is generated. Acknowledgment error (AERR bit in the CAN_SR register): The transmitter checks the Acknowledge Slot, which is transmitted by the transmitting node as a recessive bit, contains a dominant bit. If this is the case, at least one other node has received the frame correctly. If not, an Acknowledge Error has occurred and the transmitter will start in the next bit-time an Error Frame transmission.
z z
Fault Confinement
To distinguish between temporary and permanent failures, every CAN controller has two error counters: REC (Receive Error Counter) and TEC (Transmit Error Counter). The two counters are incremented upon detected errors and are decremented upon correct transmissions or receptions, respectively. Depending on the counter values, the state of the node changes: the initial state of the CAN controller is Error Active, meaning that the controller can send Error Active flags. The controller changes to the Error Passive state if there is an accumulation of errors. If the CAN controller fails or if there is an extreme accumulation of errors, there is a state transition to Bus Off.
Figure 46-7. Line Error Mode
Init
TEC < 127 and REC < 127
ERROR ACTIVE TEC > 127 or REC > 127
128 occurences of 11 consecutive recessive bits or CAN controller reset
ERROR PASSIVE
BUS OFF
TEC > 255
An error active unit takes part in bus communication and sends an active error frame when the CAN controller detects an error. An error passive unit cannot send an active error frame. It takes part in bus communication, but when an error is detected, a passive error frame is sent. Also, after a transmission, an error passive unit waits before initiating further transmission. A bus off unit is not allowed to have any influence on the bus.

1353
For fault confinement, two errors counters (TEC and REC) are implemented. These counters are accessible via the CAN_ECR register. The state of the CAN controller is automatically updated according to these counter values. If the CAN controller enters Error Active state, then the ERRA bit is set in the CAN_SR register. The corresponding interrupt is pending while the interrupt is not masked in the CAN_IMR register. If the CAN controller enters Error Passive Mode, then the ERRP bit is set in the CAN_SR register and an interrupt remains pending while the ERRP bit is set in the CAN_IMR register. If the CAN enters Bus Off Mode, then the BOFF bit is set in the CAN_SR register. As for ERRP and ERRA, an interrupt is pending while the BOFF bit is set in the CAN_IMR register. When one of the error counters values exceeds 96, an increased error rate is indicated to the controller through the WARN bit in CAN_SR register, but the node remains error active. The corresponding interrupt is pending while the interrupt is set in the CAN_IMR register. Refer to the Bosch CAN specification v2.0 for details on fault confinement.
Error Interrupt Handler

ERRA, WARN, ERRP and BOFF (CAN_SR) store the key transitions of the CAN bus status as defined in Figure 46-7 on page 1353. The transitions depend on the TEC and REC (CAN_ECR) values as described in Section Fault Confinement on page 1353. These flags are latched to keep from triggering a spurious interrupt in case these bits are used as the source of an interrupt. Thus, these flags may not reflect the current status of the CAN bus. The current CAN bus state can be determined by reading the TEC and REC fields of CAN_ECR. 46.7.4.3 Overload The overload frame is provided to request a delay of the next data or remote frame by the receiver node (Request overload frame) or to signal certain error conditions (Reactive overload frame) related to the intermission field respectively. Reactive overload frames are transmitted after detection of the following error conditions:
z z
Detection of a dominant bit during the first two bits of the intermission field Detection of a dominant bit in the last bit of EOF by a receiver, or detection of a dominant bit by a receiver or a transmitter at the last bit of an error or overload frame delimiter
The CAN controller can generate a request overload frame automatically after each message sent to one of the CAN controller mailboxes. This feature is enabled by setting the OVL bit in the CAN_MR register. Reactive overload frames are automatically handled by the CAN controller even if the OVL bit in the CAN_MR register is not set. An overload flag is generated in the same way as an error flag, but error counters do not increment.
46.7.5 Low-power Mode

In Low-power Mode, the CAN controller cannot send or receive messages. All mailboxes are inactive. In Low-power Mode, the SLEEP signal in the CAN_SR register is set; otherwise, the WAKEUP signal in the CAN_SR register is set. These two fields are exclusive except after a CAN controller reset (WAKEUP and SLEEP are stuck at 0 after a reset). After power-up reset, the Low-power Mode is disabled and the WAKEUP bit is set in the CAN_SR register only after detection of 11 consecutive recessive bits on the bus. 46.7.5.1 Enabling Low-power Mode A software application can enable Low-power Mode by setting the LPM bit in the CAN_MR global register. The CAN controller enters Low-power Mode once all pending transmit messages are sent. When the CAN controller enters Low-power Mode, the SLEEP signal in the CAN_SR register is set. Depending on the corresponding mask in the CAN_IMR register, an interrupt is generated while SLEEP is set. The SLEEP signal in the CAN_SR register is automatically cleared once WAKEUP is set. The WAKEUP signal is automatically cleared once SLEEP is set.

1354
Reception is disabled while the SLEEP signal is set to one in the CAN_SR register. It is important to note that those messages with higher priority than the last message transmitted can be received between the LPM command and entry in Low-power Mode. Once in Low-power Mode, the CAN controller clock can be switched off by programming the chips Power Management Controller (PMC). The CAN controller drains only the static current. Error counters are disabled while the SLEEP signal is set to one. Thus, to enter Low-power Mode, the software application must:
z z
Set LPM field in the CAN_MR register Wait for SLEEP signal rising
Now the CAN Controller clock can be disabled. This is done by programming the Power Management Controller (PMC).
Figure 46-8. Enabling Low-power Mode
Arbitration lost
CAN BUS LPEN= 1 LPM (CAN_MR) SLEEP (CAN_SR) WAKEUP (CAN_SR) MRDY (CAN_MSR1) MRDY (CAN_MSR3) CAN_TIM
Mailbox 1
Mailbox 3
0x0
46.7.5.2 Disabling Low-power Mode The CAN controller can be awake after detecting a CAN bus activity. Bus activity detection is done by an external module that may be embedded in the chip. When it is notified of a CAN bus activity, the software application disables Low-power Mode by programming the CAN controller. To disable Low-power Mode, the software application must:
z z
Enable the CAN Controller clock. This is done by programming the Power Management Controller (PMC). Clear the LPM field in the CAN_MR register
The CAN controller synchronizes itself with the bus activity by checking for eleven consecutive recessive bits. Once synchronized, the WAKEUP signal in the CAN_SR register is set. Depending on the corresponding mask in the CAN_IMR register, an interrupt is generated while WAKEUP is set. The SLEEP signal in the CAN_SR register is automatically cleared once WAKEUP is set. WAKEUP signal is automatically cleared once SLEEP is set. If no message is being sent on the bus, then the CAN controller is able to send a message eleven bit times after disabling Low-power Mode. If there is bus activity when Low-power mode is disabled, the CAN controller is synchronized with the bus activity in the next interframe. The previous message is lost (see Figure 46-9).

1355
Figure 46-9. Disabling Low-power Mode

Bus Activity Detected Message x Interframe synchronization
CAN BUS LPM (CAN_MR) SLEEP (CAN_SR) WAKEUP (CAN_SR)
Message lost
MRDY (CAN_MSRx)

1356
46.8
46.8.1 CAN Controller Initialization

After power-up reset, the CAN controller is disabled. The CAN controller clock must be activated by the Power Management Controller (PMC) and the CAN controller interrupt line must be enabled by the interrupt controller (AIC). The CAN controller must be initialized with the CAN network parameters. The CAN_BR register defines the sampling point in the bit time period. CAN_BR must be set before the CAN controller is enabled by setting the CANEN field in the CAN_MR register. The CAN controller is enabled by setting the CANEN flag in the CAN_MR register. At this stage, the internal CAN controller state machine is reset, error counters are reset to 0, error flags are reset to 0. Once the CAN controller is enabled, bus synchronization is done automatically by scanning eleven recessive bits. The WAKEUP bit in the CAN_SR register is automatically set to 1 when the CAN controller is synchronized (WAKEUP and SLEEP are stuck at 0 after a reset). The CAN controller can start listening to the network in Autobaud Mode. In this case, the error counters are locked and a mailbox may be configured in Receive Mode. By scanning error flags, the CAN_BR register values synchronized with the network. Once no error has been detected, the application disables the Autobaud Mode, clearing the ABM field in the CAN_MR register.
Figure 46-10. Possible Initialization Procedure
Enable CAN Controller Clock (PMC) Enable CAN Controller Interrupt Line (AIC)
Configure a Mailbox in Reception Mode
Change CAN_BR value (ABM == 1 and CANEN == 1)
Errors ? (CAN_SR or CAN_MSRx)
Yes
No ABM = 0 and CANEN = 0
CANEN = 1 (ABM == 0)
End of Initialization

1357
46.8.2 CAN Controller Interrupt Handling

There are two different types of interrupts. One type of interrupt is a message-object related interrupt, the other is a system interrupt that handles errors or system-related interrupt sources. All interrupt sources can be masked by writing the corresponding field in the CAN_IDR register. They can be unmasked by writing to the CAN_IER register. After a power-up reset, all interrupt sources are disabled (masked). The current mask status can be checked by reading the CAN_IMR register. The CAN_SR register gives all interrupt source states. The following events may initiate one of the two interrupts:
z
Message object interrupt

z z
Data registers in the mailbox object are available to the application. In Receive Mode, a new message was received. In Transmit Mode, a message was transmitted successfully. A sent transmission was aborted. Bus off interrupt: The CAN module enters the bus off state. Error passive interrupt: The CAN module enters Error Passive Mode. Error Active Mode: The CAN module is neither in Error Passive Mode nor in Bus Off mode. Warn Limit interrupt: The CAN module is in Error-active Mode, but at least one of its error counter value exceeds 96. Wake-up interrupt: This interrupt is generated after a wake-up and a bus synchronization. Sleep interrupt: This interrupt is generated after a Low-power Mode enable once all pending messages in transmission have been sent. Internal timer counter overflow interrupt: This interrupt is generated when the internal timer rolls over. Timestamp interrupt: This interrupt is generated after the reception or the transmission of a start of frame or an end of frame. The value of the internal counter is copied in the CAN_TIMESTP register.
System interrupts
z z z z z z z z
All interrupts are cleared by clearing the interrupt source except for the internal timer counter overflow interrupt and the timestamp interrupt. These interrupts are cleared by reading the CAN_SR register.
46.8.3 CAN Controller Message Handling

46.8.3.1 Receive Handling Two modes are available to configure a mailbox to receive messages. In Receive Mode, the first message received is stored in the mailbox data register. In Receive with Overwrite Mode, the last message received is stored in the mailbox.
Simple Receive Mailbox

A mailbox is in Receive Mode once the MOT field in the CAN_MMRx register has been configured. Message ID and Message Acceptance Mask must be set before the Receive Mode is enabled. After Receive Mode is enabled, the MRDY flag in the CAN_MSR register is automatically cleared until the first message is received. When the first message has been accepted by the mailbox, the MRDY flag is set. An interrupt is pending for the mailbox while the MRDY flag is set. This interrupt can be masked depending on the mailbox flag in the CAN_IMR global register. Message data are stored in the mailbox data register until the software application notifies that data processing has ended. This is done by asking for a new transfer command, setting the MTCR flag in the CAN_MCRx register. This automatically clears the MRDY signal. The MMI flag in the CAN_MSRx register notifies the software that a message has been lost by the mailbox. This flag is set when messages are received while MRDY is set in the CAN_MSRx register. This flag is cleared by reading the CAN_MSRs register. A receive mailbox prevents from overwriting the first message by new ones while MRDY flag is set in the CAN_MSRx register. See Figure 46-11.

1358
Figure 46-11. Receive Mailbox

Message ID = CAN_MIDx
CAN BUS MRDY (CAN_MSRx) MMI (CAN_MSRx)
Message 1
Message 2 lost
Message 3
(CAN_MDLx CAN_MDHx) MTCR (CAN_MCRx)
Message 1
Message 3
Reading CAN_MSRx Reading CAN_MDHx & CAN_MDLx Writing CAN_MCRx
Note:
In the case of ARM architecture, CAN_MSRx, CAN_MDLx, CAN_MDHx can be read using an optimized ldm assembler instruction.
Receive with Overwrite Mailbox

A mailbox is in Receive with Overwrite Mode once the MOT field in the CAN_MMRx register has been configured. Message ID and Message Acceptance masks must be set before Receive Mode is enabled. After Receive Mode is enabled, the MRDY flag in the CAN_MSR register is automatically cleared until the first message is received. When the first message has been accepted by the mailbox, the MRDY flag is set. An interrupt is pending for the mailbox while the MRDY flag is set. This interrupt is masked depending on the mailbox flag in the CAN_IMR global register. If a new message is received while the MRDY flag is set, this new message is stored in the mailbox data register, overwriting the previous message. The MMI flag in the CAN_MSRx register notifies the software that a message has been dropped by the mailbox. This flag is cleared when reading the CAN_MSRx register. The CAN controller may store a new message in the CAN data registers while the application reads them. To check that CAN_MDHx and CAN_MDLx do not belong to different messages, the application must check the MMI field in the CAN_MSRx register before and after reading CAN_MDHx and CAN_MDLx. If the MMI flag is set again after the data registers have been read, the software application has to re-read CAN_MDHx and CAN_MDLx (see Figure 46-12).

1359
Figure 46-12. Receive with Overwrite Mailbox

Message ID = CAN_MIDx
CAN BUS MRDY (CAN_MSRx) MMI (CAN_MSRx) (CAN_MDLx CAN_MDHx) MTCR (CAN_MCRx)
Message 1
Message 2
Message 3
Message 4
Message 1
Message 2
Message 3
Message 4
Reading CAN_MSRx Reading CAN_MDHx & CAN_MDLx Writing CAN_MCRx
Chaining Mailboxes
Several mailboxes may be used to receive a buffer split into several messages with the same ID. In this case, the mailbox with the lowest number is serviced first. In the receive and receive with overwrite modes, the field PRIOR in the CAN_MMRx register has no effect. If Mailbox 0 and Mailbox 5 accept messages with the same ID, the first message is received by Mailbox 0 and the second message is received by Mailbox 5. Mailbox 0 must be configured in Receive Mode (i.e., the first message received is considered) and Mailbox 5 must be configured in Receive with Overwrite Mode. Mailbox 0 cannot be configured in Receive with Overwrite Mode; otherwise, all messages are accepted by this mailbox and Mailbox 5 is never serviced. If several mailboxes are chained to receive a buffer split into several messages, all mailboxes except the last one (with the highest number) must be configured in Receive Mode. The first message received is handled by the first mailbox, the second one is refused by the first mailbox and accepted by the second mailbox, the last message is accepted by the last mailbox and refused by previous ones (see Figure 46-13).

1360
Figure 46-13. Chaining Three Mailboxes to Receive a Buffer Split into Three Messages
Buffer split in 3 messages
CAN BUS MRDY (CAN_MSRx) MMI (CAN_MSRx) MRDY (CAN_MSRy) MMI (CAN_MSRy) MRDY (CAN_MSRz) MMI (CAN_MSRz)
Message s1
Message s2
Message s3
Reading CAN_MSRx, CAN_MSRy and CAN_MSRz Reading CAN_MDH & CAN_MDL for mailboxes x, y and z Writing MBx MBy MBz in CAN_TCR
If the number of mailboxes is not sufficient (the MMI flag of the last mailbox raises), the user must read each data received on the last mailbox in order to retrieve all the messages of the buffer split (see Figure 46-14).
Figure 46-14. Chaining Three Mailboxes to Receive a Buffer Split into Four Messages
Buffer split in 4 messages
CAN BUS MRDY (CAN_MSRx) MMI (CAN_MSRx) MRDY (CAN_MSRy) MMI (CAN_MSRy) MRDY (CAN_MSRz) MMI (CAN_MSRz)
Message s1
Message s2
Message s3
Message s4
Reading CAN_MSRx, CAN_MSRy and CAN_MSRz Reading CAN_MDH & CAN_MDL for mailboxes x, y and z Writing MBx MBy MBz in CAN_TCR

1361
46.8.3.2 Transmission Handling A mailbox is in Transmit Mode once the MOT field in the CAN_MMRx register has been configured. Message ID and Message Acceptance mask must be set before Receive Mode is enabled. After Transmit Mode is enabled, the MRDY flag in the CAN_MSR register is automatically set until the first command is sent. When the MRDY flag is set, the software application can prepare a message to be sent by writing to the CAN_MDx registers. The message is sent once the software asks for a transfer command setting the MTCR bit and the message data length in the CAN_MCRx register. The MRDY flag remains at zero as long as the message has not been sent or aborted. It is important to note that no access to the mailbox data register is allowed while the MRDY flag is cleared. An interrupt is pending for the mailbox while the MRDY flag is set. This interrupt can be masked depending on the mailbox flag in the CAN_IMR global register. It is also possible to send a remote frame setting the MRTR bit instead of setting the MDLC field. The answer to the remote frame is handled by another reception mailbox. In this case, the device acts as a consumer but with the help of two mailboxes. It is possible to handle the remote frame emission and the answer reception using only one mailbox configured in Consumer Mode. Refer to the section Remote Frame Handling on page 1363. Several messages can try to win the bus arbitration in the same time. The message with the highest priority is sent first. Several transfer request commands can be generated at the same time by setting MBx bits in the CAN_TCR register. The priority is set in the PRIOR field of the CAN_MMRx register. Priority 0 is the highest priority, priority 15 is the lowest priority. Thus it is possible to use a part of the message ID to set the PRIOR field. If two mailboxes have the same priority, the message of the mailbox with the lowest number is sent first. Thus if mailbox 0 and mailbox 5 have the same priority and have a message to send at the same time, then the message of the mailbox 0 is sent first. Setting the MACR bit in the CAN_MCRx register aborts the transmission. Transmission for several mailboxes can be aborted by writing MBx fields in the CAN_MACR register. If the message is being sent when the abort command is set, then the application is notified by the MRDY bit set and not the MABT in the CAN_MSRx register. Otherwise, if the message has not been sent, then the MRDY and the MABT are set in the CAN_MSR register. When the bus arbitration is lost by a mailbox message, the CAN controller tries to win the next bus arbitration with the same message if this one still has the highest priority. Messages to be sent are re-tried automatically until they win the bus arbitration. This feature can be disabled by setting the bit DRPT in the CAN_MR register. In this case if the message was not sent the first time it was transmitted to the CAN transceiver, it is automatically aborted. The MABT flag is set in the CAN_MSRx register until the next transfer command. Figure 46-15 shows three MBx message attempts being made (MRDY of MBx set to 0). The first MBx message is sent, the second is aborted and the last one is trying to be aborted but too late because it has already been transmitted to the CAN transceiver.
Figure 46-15. Transmitting Messages
CAN BUS MRDY (CAN_MSRx) MABT (CAN_MSRx) MTCR (CAN_MCRx) MACR (CAN_MCRx) Reading CAN_MSRx Writing CAN_MDHx & CAN_MDLx Abort MBx message Try to Abort MBx message MBx message MBx message

1362
46.8.3.3 Remote Frame Handling Producer/consumer model is an efficient means of handling broadcasted messages. The push model allows a producer to broadcast messages; the pull model allows a customer to ask for messages.
Figure 46-16. Producer / Consumer Model
Producer
Request
PUSH MODEL
CAN Data Frame
Consumer
Indication(s)
PULL MODEL
Producer
Indications CAN Remote Frame
Consumer
Request(s)
Response
CAN Data Frame Confirmation(s)
In Pull Mode, a consumer transmits a remote frame to the producer. When the producer receives a remote frame, it sends the answer accepted by one or many consumers. Using transmit and receive mailboxes, a consumer must dedicate two mailboxes, one in Transmit Mode to send remote frames, and at least one in Receive Mode to capture the producers answer. The same structure is applicable to a producer: one reception mailbox is required to get the remote frame and one transmit mailbox to answer. Mailboxes can be configured in Producer or Consumer Mode. A lonely mailbox can handle the remote frame and the answer. With 8 mailboxes, the CAN controller can handle 8 independent producers/consumers.
Producer Configuration
A mailbox is in Producer Mode once the MOT field in the CAN_MMRx register has been configured. Message ID and Message Acceptance masks must be set before Receive Mode is enabled. After Producer Mode is enabled, the MRDY flag in the CAN_MSR register is automatically set until the first transfer command. The software application prepares data to be sent by writing to the CAN_MDHx and the CAN_MDLx registers, then by setting the MTCR bit in the CAN_MCRx register. Data is sent after the reception of a remote frame as soon as it wins the bus arbitration. The MRDY flag remains at zero as long as the message has not been sent or aborted. No access to the mailbox data register can be done while MRDY flag is cleared. An interrupt is pending for the mailbox while the MRDY flag is set. This interrupt can be masked according to the mailbox flag in the CAN_IMR global register. If a remote frame is received while no data are ready to be sent (signal MRDY set in the CAN_MSRx register), then the MMI signal is set in the CAN_MSRx register. This bit is cleared by reading the CAN_MSRx register. The MRTR field in the CAN_MSRx register has no meaning. This field is used only when using Receive and Receive with Overwrite modes. After a remote frame has been received, the mailbox functions like a transmit mailbox. The message with the highest priority is sent first. The transmitted message may be aborted by setting the MACR bit in the CAN_MCR register. Please refer to the section Transmission Handling on page 1362.

1363
Figure 46-17. Producer Handling

CAN BUS MRDY (CAN_MSRx) MMI (CAN_MSRx) MTCR (CAN_MCRx) Reading CAN_MSRx Remote Frame Message 1 Remote Frame Remote Frame Message 2
(CAN_MDLx CAN_MDHx)
Message 1
Message 2
Consumer Configuration
A mailbox is in Consumer Mode once the MOT field in the CAN_MMRx register has been configured. Message ID and Message Acceptance masks must be set before Receive Mode is enabled. After Consumer Mode is enabled, the MRDY flag in the CAN_MSR register is automatically cleared until the first transfer request command. The software application sends a remote frame by setting the MTCR bit in the CAN_MCRx register or the MBx bit in the global CAN_TCR register. The application is notified of the answer by the MRDY flag set in the CAN_MSRx register. The application can read the data contents in the CAN_MDHx and CAN_MDLx registers. An interrupt is pending for the mailbox while the MRDY flag is set. This interrupt can be masked according to the mailbox flag in the CAN_IMR global register. The MRTR bit in the CAN_MCRx register has no effect. This field is used only when using Transmit Mode. After a remote frame has been sent, the consumer mailbox functions as a reception mailbox. The first message received is stored in the mailbox data registers. If other messages intended for this mailbox have been sent while the MRDY flag is set in the CAN_MSRx register, they will be lost. The application is notified by reading the MMI field in the CAN_MSRx register. The read operation automatically clears the MMI flag. If several messages are answered by the Producer, the CAN controller may have one mailbox in consumer configuration, zero or several mailboxes in Receive Mode and one mailbox in Receive with Overwrite Mode. In this case, the consumer mailbox must have a lower number than the Receive with Overwrite mailbox. The transfer command can be triggered for all mailboxes at the same time by setting several MBx fields in the CAN_TCR register.
Figure 46-18. Consumer Handling
CAN BUS MRDY (CAN_MSRx) MMI (CAN_MSRx) MTCR (CAN_MCRx) Remote Frame Message x Remote Frame Message y
(CAN_MDLx CAN_MDHx)
Message x
Message y

1364
46.8.4 CAN Controller Timing Modes

Using the free running 16-bit internal timer, the CAN controller can be set in one of the two following timing modes:
z z
Timestamping Mode: The value of the internal timer is captured at each Start Of Frame or each End Of Frame. Time Triggered Mode: The mailbox transfer operation is triggered when the internal timer reaches the mailbox trigger.
Timestamping Mode is enabled by clearing the TTM bit in the CAN_MR register. Time Triggered Mode is enabled by setting the TTM bit in the CAN_MR register. 46.8.4.1 Timestamping Mode Each mailbox has its own timestamp value. Each time a message is sent or received by a mailbox, the 16-bit value MTIMESTAMP of the CAN_TIMESTP register is transferred to the LSB bits of the CAN_MSRx register. The value read in the CAN_MSRx register corresponds to the internal timer value at the Start Of Frame or the End Of Frame of the message handled by the mailbox.
Figure 46-19. Mailbox Timestamp
Start of Frame End of Frame
CAN BUS CAN_TIM
Message 1
Message 2
TEOF (CAN_MR) TIMESTAMP (CAN_TSTP) MTIMESTAMP (CAN_MSRx) MTIMESTAMP (CAN_MSRy) Timestamp 1 Timestamp 2
Timestamp 1 Timestamp 2
46.8.4.2 Time Triggered Mode In Time Triggered Mode, basic cycles can be split into several time windows. A basic cycle starts with a reference message. Each time a window is defined from the reference message, a transmit operation should occur within a predefined time window. A mailbox must not win the arbitration in a previous time window, and it must not be retried if the arbitration is lost in the time window.
Figure 46-20. Time Triggered Principle
Time Cycle Reference Message Reference Message
Time Windows for Messages Global Time
Time Trigger Mode is enabled by setting the TTM field in the CAN_MR register. In Time Triggered Mode, as in Timestamp Mode, the CAN_TIMESTP field captures the values of the internal counter, but the MTIMESTAMP fields in the CAN_MSRx registers are not active and are read at 0.

1365
Synchronization by a Reference Message

In Time Triggered Mode, the internal timer counter is automatically reset when a new message is received in the last mailbox. This reset occurs after the reception of the End Of Frame on the rising edge of the MRDY signal in the CAN_MSRx register. This allows synchronization of the internal timer counter with the reception of a reference message and the start a new time window.
Transmitting within a Time Window

A time mark is defined for each mailbox. It is defined in the 16-bit MTIMEMARK field of the CAN_MMRx register. At each internal timer clock cycle, the value of the CAN_TIM is compared with each mailbox time mark. When the internal timer counter reaches the MTIMEMARK value, an internal timer event for the mailbox is generated for the mailbox. In Time Triggered Mode, transmit operations are delayed until the internal timer event for the mailbox. The application prepares a message to be sent by setting the MTCR in the CAN_MCRx register. The message is not sent until the CAN_TIM value is less than the MTIMEMARK value defined in the CAN_MMRx register. If the transmit operation is failed, i.e., the message loses the bus arbitration and the next transmit attempt is delayed until the next internal time trigger event. This prevents overlapping the next time window, but the message is still pending and is retried in the next time window when CAN_TIM value equals the MTIMEMARK value. It is also possible to prevent a retry by setting the DRPT field in the CAN_MR register.
Freezing the Internal Timer Counter

The internal counter can be frozen by setting TIMFRZ in the CAN_MR register. This prevents an unexpected roll-over when the counter reaches FFFFh. When this occurs, it automatically freezes until a new reset is issued, either due to a message received in the last mailbox or any other reset counter operations. The TOVF bit in the CAN_SR register is set when the counter is frozen. The TOVF bit in the CAN_SR register is cleared by reading the CAN_SR register. Depending on the corresponding interrupt mask in the CAN_IMR register, an interrupt is generated when TOVF is set.

1366
Figure 46-21. Time Triggered Operations

End of Frame Message x Arbitration Lost Message y Arbitration Win
CAN BUS
Reference Message
Message y Internal Counter Reset
CAN_TIM Cleared by software MRDY (CAN_MSRlast_mailbox_number) Timer Event x MRDY (CAN_MSRx) MTIMEMARKy == CAN_TIM MTIMEMARKx == CAN_TIM
Timer Event y MRDY (CAN_MSRy)
Time Window Basic Cycle
End of Frame
Message x Arbitration Win Message x
CAN BUS CAN_TIM
Reference Message
Internal Counter Reset
Cleared by software MRDY (CAN_MSRlast_mailbox_number) Timer Event x MRDY (CAN_MSRx) Time Window Basic Cycle MTIMEMARKx == CAN_TIM

1367

To prevent any single software error that may corrupt CAN behavior, the registers listed below can be write-protected by setting the WPEN bit in the CAN Write Protection Mode Register (CAN_WPMR). If a write access in a write-protected register is detected, then the WPVS flag in the CAN Write Protection Status Register (CAN_WPSR) is set and the field WPVSRC indicates in which register the write access has been attempted. The WPVS flag is automatically reset after reading the CAN Write Protection Status Register (CAN_WPSR). List of the write-protected registers: Section 46.9.1 CAN Mode Register on page 1370 Section 46.9.6 CAN Baudrate Register on page 1380 Section 46.9.14 CAN Message Mode Register on page 1388 Section 46.9.15 CAN Message Acceptance Mask Register on page 1389 Section 46.9.16 CAN Message ID Register on page 1390

1368
46.9
Controller Area Network (CAN) User Interface
Table 46-6. Register Mapping Offset 0x0000 0x0004 0x0008 0x000C 0x0010 0x0014 0x0018 0x001C 0x0020 0x0024 0x0028 0x002C - 0x00E0 0x00E4 0x00E8 0x00EC - 0x01FC 0x0200 + MB * 0x20 + 0x00 0x0200 + MB * 0x20 + 0x04 0x0200 + MB * 0x20 + 0x08 0x0200 + MB * 0x20 + 0x0C 0x0200 + MB * 0x20 + 0x10 0x0200 + MB * 0x20 + 0x14 0x0200 + MB * 0x20 + 0x18 0x0200 + MB * 0x20 + 0x1C Register Mode Register Interrupt Enable Register Interrupt Disable Register Interrupt Mask Register Status Register Baudrate Register Timer Register Timestamp Register Error Counter Register Transfer Command Register Abort Command Register Reserved Write Protect Mode Register Write Protect Status Register Reserved Mailbox Mode Register
(1)
Name CAN_MR CAN_IER CAN_IDR CAN_IMR CAN_SR CAN_BR CAN_TIM CAN_TIMESTP CAN_ECR CAN_TCR CAN_ACR CAN_WPMR CAN_WPSR CAN_MMR CAN_MAM CAN_MID CAN_MFID CAN_MSR CAN_MDL CAN_MDH CAN_MCR
Access Read-write Write-only Write-only Read-only Read-only Read-write Read-only Read-only Read-only Write-only Write-only Read-write Read-only Read-write Read-write Read-write Read-only Read-only Read-write Read-write Write-only
Reset 0x0 0x0 0x0 0x0 0x0 0x0 0x0 0x0 0x0 0x0 0x0 0x0 0x0 0x0 0x0 0x0 -
Mailbox Acceptance Mask Register Mailbox ID Register Mailbox Family ID Register Mailbox Status Register Mailbox Data Low Register Mailbox Data High Register Mailbox Control Register
Note:
1. Mailbox number ranges from 0 to 7.

1369
46.9.1 CAN Mode Register

31 23 15 7 DRPT
CAN_MR 0xF000C000 (0), 0xF8010000 (1) Read-write

30 22 14 6 TIMFRZ 29 21 13 5 TTM 28 20 12 4 TEOF 27 19 11 3 OVL 26 18 10 2 ABM 25 17 9 1 LPM 24 16 8 0 CANEN
This register can only be written if the WPEN bit is cleared in CAN Write Protection Mode Register.
CANEN: CAN Controller Enable 0: The CAN Controller is disabled. 1: The CAN Controller is enabled. LPM: Disable/Enable Low-power Mode 0: Disable low-power Mode 1: Enable low-power Mode CAN controller enters low-power Mode once all pending messages have been transmitted. ABM: Disable/Enable Autobaud/Listen mode 0: Disable Autobaud/listen mode 1: Enable Autobaud/listen mode OVL: Disable/Enable Overload Frame 0: No overload frame is generated. 1: An overload frame is generated after each successful reception for mailboxes configured in Receive with/without overwrite Mode, Producer and Consumer. TEOF: Timestamp Messages at each End of Frame 0: The value of CAN_TIM is captured in the CAN_TIMESTP register at each Start Of Frame. 1: The value of CAN_TIM is captured in the CAN_TIMESTP register at each End Of Frame. TTM: Disable/Enable Time Triggered Mode 0: Time Triggered Mode is disabled 1: Time Triggered Mode is enabled TIMFRZ: Enable Timer Freeze 0: The internal timer continues to be incremented after it reached 0xFFFF. 1: The internal timer stops incrementing after reaching 0xFFFF. It is restarted after a timer reset. See Freezing the Internal Timer Counter on page 1366. DRPT: Disable Repeat 0: When a transmit mailbox loses the bus arbitration, the transfer request remains pending. 1: When a transmit mailbox lose the bus arbitration, the transfer request is automatically aborted. It automatically raises the MABT and MRDT flags in the corresponding CAN_MSRx. SAMA5D3 Series [DATASHEET]
1370
46.9.2 CAN Interrupt Enable Register

31 23 TSTP 15 7 MB7
CAN_IER 0xF000C004 (0), 0xF8010004 (1) Write-only

30 22 TOVF 14 6 MB6 29 21 WAKEUP 13 5 MB5 28 BERR 20 SLEEP 12 4 MB4 27 FERR 19 BOFF 11 3 MB3 26 AERR 18 ERRP 10 2 MB2 25 SERR 17 WARN 9 1 MB1 24 CERR 16 ERRA 8 0 MB0
MBx: Mailbox x Interrupt Enable

0: No effect. 1: Enable Mailbox x interrupt.
ERRA: Error Active Mode Interrupt Enable

0: No effect. 1: Enable ERRA interrupt.
WARN: Warning Limit Interrupt Enable

0: No effect. 1: Enable WARN interrupt.
ERRP: Error Passive Mode Interrupt Enable

0: No effect. 1: Enable ERRP interrupt.
BOFF: Bus Off Mode Interrupt Enable

0: No effect. 1: Enable BOFF interrupt.
SLEEP: Sleep Interrupt Enable

0: No effect. 1: Enable SLEEP interrupt.
WAKEUP: Wake-up Interrupt Enable

0: No effect. 1: Enable SLEEP interrupt.
TOVF: Timer Overflow Interrupt Enable

0: No effect. 1: Enable TOVF interrupt.
TSTP: TimeStamp Interrupt Enable

0: No effect. 1: Enable TSTP interrupt.

1371
CERR: CRC Error Interrupt Enable

0: No effect. 1: Enable CRC Error interrupt.
SERR: Stuffing Error Interrupt Enable

0: No effect. 1: Enable Stuffing Error interrupt.
AERR: Acknowledgment Error Interrupt Enable

0: No effect. 1: Enable Acknowledgment Error interrupt.
FERR: Form Error Interrupt Enable

0: No effect. 1: Enable Form Error interrupt.
BERR: Bit Error Interrupt Enable

0: No effect. 1: Enable Bit Error interrupt.

1372
46.9.3 CAN Interrupt Disable Register

31 23 TSTP 15 7 MB7
CAN_IDR 0xF000C008 (0), 0xF8010008 (1) Write-only

MBx: Mailbox x Interrupt Disable

0: No effect. 1: Disable Mailbox x interrupt.
ERRA: Error Active Mode Interrupt Disable

0: No effect. 1: Disable ERRA interrupt.
WARN: Warning Limit Interrupt Disable

0: No effect. 1: Disable WARN interrupt.
ERRP: Error Passive Mode Interrupt Disable

0: No effect. 1: Disable ERRP interrupt.
BOFF: Bus Off Mode Interrupt Disable

0: No effect. 1: Disable BOFF interrupt.
SLEEP: Sleep Interrupt Disable

0: No effect. 1: Disable SLEEP interrupt.
WAKEUP: Wake-up Interrupt Disable

0: No effect. 1: Disable WAKEUP interrupt.
TOVF: Timer Overflow Interrupt

0: No effect. 1: Disable TOVF interrupt.

1373
TSTP: TimeStamp Interrupt Disable

0: No effect. 1: Disable TSTP interrupt.
CERR: CRC Error Interrupt Disable

0: No effect. 1: Disable CRC Error interrupt.
SERR: Stuffing Error Interrupt Disable

0: No effect. 1: Disable Stuffing Error interrupt.
AERR: Acknowledgment Error Interrupt Disable

0: No effect. 1: Disable Acknowledgment Error interrupt.
FERR: Form Error Interrupt Disable

0: No effect. 1: Disable Form Error interrupt.
BERR: Bit Error Interrupt Disable

0: No effect. 1: Disable Bit Error interrupt.

1374
46.9.4 CAN Interrupt Mask Register

31 23 TSTP 15 7 MB7
CAN_IMR 0xF000C00C (0), 0xF801000C (1) Read-only

MBx: Mailbox x Interrupt Mask

0: Mailbox x interrupt is disabled. 1: Mailbox x interrupt is enabled.
ERRA: Error Active Mode Interrupt Mask

0: ERRA interrupt is disabled. 1: ERRA interrupt is enabled.
WARN: Warning Limit Interrupt Mask

0: Warning Limit interrupt is disabled. 1: Warning Limit interrupt is enabled.
ERRP: Error Passive Mode Interrupt Mask

0: ERRP interrupt is disabled. 1: ERRP interrupt is enabled.
BOFF: Bus Off Mode Interrupt Mask

0: BOFF interrupt is disabled. 1: BOFF interrupt is enabled.
SLEEP: Sleep Interrupt Mask

0: SLEEP interrupt is disabled. 1: SLEEP interrupt is enabled.
WAKEUP: Wake-up Interrupt Mask

0: WAKEUP interrupt is disabled. 1: WAKEUP interrupt is enabled.
TOVF: Timer Overflow Interrupt Mask

0: TOVF interrupt is disabled. 1: TOVF interrupt is enabled.

1375
TSTP: Timestamp Interrupt Mask

0: TSTP interrupt is disabled. 1: TSTP interrupt is enabled.
CERR: CRC Error Interrupt Mask

0: CRC Error interrupt is disabled. 1: CRC Error interrupt is enabled.
SERR: Stuffing Error Interrupt Mask

0: Bit Stuffing Error interrupt is disabled. 1: Bit Stuffing Error interrupt is enabled.
AERR: Acknowledgment Error Interrupt Mask

0: Acknowledgment Error interrupt is disabled. 1: Acknowledgment Error interrupt is enabled.
FERR: Form Error Interrupt Mask

0: Form Error interrupt is disabled. 1: Form Error interrupt is enabled.
BERR: Bit Error Interrupt Mask

0: Bit Error interrupt is disabled. 1: Bit Error interrupt is enabled.

1376
46.9.5 CAN Status Register

31 OVLSY 23 TSTP 15 7 MB7
CAN_SR 0xF000C010 (0), 0xF8010010 (1) Read-only

30 TBSY 22 TOVF 14 6 MB6 29 RBSY 21 WAKEUP 13 5 MB5 28 BERR 20 SLEEP 12 4 MB4 27 FERR 19 BOFF 11 3 MB3 26 AERR 18 ERRP 10 2 MB2 25 SERR 17 WARN 9 1 MB1 24 CERR 16 ERRA 8 0 MB0
MBx: Mailbox x Event

0: No event occurred on Mailbox x. 1: An event occurred on Mailbox x. An event corresponds to MRDY, MABT fields in the CAN_MSRx register.
ERRA: Error Active Mode

0: CAN controller has not reached Error Active Mode since the last read of CAN_SR. 1: CAN controller has reached Error Active Mode since the last read of CAN_SR. This flag is automatically cleared by reading CAN_SR register. This flag is set depending on TEC and REC counter values. It is set when a node is neither in Error Passive Mode nor in Bus Off Mode.
WARN: Warning Limit

0: CAN controller Warning Limit has not been reached since the last read of CAN_SR. 1: CAN controller Warning Limit has been reached since the last read of CAN_SR. This flag is automatically cleared by reading CAN_SR register. This flag is set depending on TEC and REC counter values. It is set when at least one of the counter values has reached a value greater or equal to 96.
ERRP: Error Passive Mode

0: CAN controller has not reached Error Passive Mode since the last read of CAN_SR. 1: CAN controller has reached Error Passive Mode since the last read of CAN_SR. This flag is set depending on TEC and REC counters values. This flag is automatically cleared by reading CAN_SR register. A node is in error passive state when TEC counter is greater or equal to 128 (decimal) or when the REC counter is greater or equal to 128 (decimal).
BOFF: Bus Off Mode

0: CAN controller has not reached Bus Off Mode. 1: CAN controller has reached Bus Off Mode since the last read of CAN_SR. This flag is automatically cleared by reading CAN_SR register. This flag is set depending on TEC counter value. A node is in bus off state when TEC counter is greater or equal to 256 (decimal).

1377
SLEEP: CAN Controller in Low-power Mode

0: CAN controller is not in low-power mode. 1: CAN controller is in low-power mode. This flag is automatically reset when low-power mode is disabled
WAKEUP: CAN Controller is not in Low-power Mode

0: CAN controller is in low-power mode. 1: CAN controller is not in low-power mode. When a WAKEUP event occurs, the CAN controller is synchronized with the bus activity. Messages can be transmitted or received. The CAN controller clock must be available when a WAKEUP event occurs. This flag is automatically reset when the CAN Controller enters low-power mode.
TOVF: Timer Overflow

0: The timer has not rolled-over FFFFh to 0000h. 1: The timer rolls-over FFFFh to 0000h. This flag is automatically cleared by reading CAN_SR register.
TSTP Timestamp
0: No bus activity has been detected. 1: A start of frame or an end of frame has been detected (according to the TEOF field in the CAN_MR register). This flag is automatically cleared by reading the CAN_SR register.
CERR: Mailbox CRC Error

0: No CRC error occurred during a previous transfer. 1: A CRC error occurred during a previous transfer. A CRC error has been detected during last reception. This flag is automatically cleared by reading CAN_SR register.
SERR: Mailbox Stuffing Error

0: No stuffing error occurred during a previous transfer. 1: A stuffing error occurred during a previous transfer. A form error results from the detection of more than five consecutive bit with the same polarity. This flag is automatically cleared by reading CAN_SR register.
AERR: Acknowledgment Error

0: No acknowledgment error occurred during a previous transfer. 1: An acknowledgment error occurred during a previous transfer. An acknowledgment error is detected when no detection of the dominant bit in the acknowledge slot occurs. This flag is automatically cleared by reading CAN_SR register.
FERR: Form Error

0: No form error occurred during a previous transfer 1: A form error occurred during a previous transfer

1378
A form error results from violations on one or more of the fixed form of the following bit fields:
CRC delimiter ACK delimiter End of frame Error delimiter Overload delimiter
This flag is automatically cleared by reading CAN_SR register.
BERR: Bit Error

0: No bit error occurred during a previous transfer. 1: A bit error occurred during a previous transfer. A bit error is set when the bit value monitored on the line is different from the bit value sent. This flag is automatically cleared by reading CAN_SR register.
RBSY: Receiver busy

0: CAN receiver is not receiving a frame. 1: CAN receiver is receiving a frame. Receiver busy. This status bit is set by hardware while CAN receiver is acquiring or monitoring a frame (remote, data, overload or error frame). It is automatically reset when CAN is not receiving.
TBSY: Transmitter busy

0: CAN transmitter is not transmitting a frame. 1: CAN transmitter is transmitting a frame. Transmitter busy. This status bit is set by hardware while CAN transmitter is generating a frame (remote, data, overload or error frame). It is automatically reset when CAN is not transmitting.
OVLSY: Overload busy

0: CAN transmitter is not transmitting an overload frame. 1: CAN transmitter is transmitting a overload frame. It is automatically reset when the bus is not transmitting an overload frame.

1379
46.9.6 CAN Baudrate Register

31 23 15 7
CAN_BR 0xF000C014 (0), 0xF8010014 (1) Read-write

30 22 14 6 29 21 13 SJW 5 PHASE1 4 28 20 12 27 19 BRP 11 3 26 18 10 2 25 17 9 PROPAG 1 PHASE2 24 SMP 16 8 0
This register can only be written if the WPEN bit is cleared in CAN Write Protection Mode Register. Any modification on one of the fields of the CAN_BR register must be done while CAN module is disabled. To compute the different Bit Timings, please refer to the Section 46.7.4.1 CAN Bit Timing Configuration on page 1349.
PHASE2: Phase 2 segment

This phase is used to compensate the edge phase error.
t PHS2 = t CSC ( PHASE2 + 1 ) Warning: PHASE2 value must be different from 0. PHASE1: Phase 1 segment
This phase is used to compensate for edge phase error.
t PHS1 = t CSC ( PHASE1 + 1 ) PROPAG: Programming time segment

This part of the bit time is used to compensate for the physical delay times within the network.
t PRS = t CSC ( PROPAG + 1 ) SJW: Re-synchronization jump width

To compensate for phase shifts between clock oscillators of different controllers on bus. The controller must re-synchronize on any relevant signal edge of the current transmission. The synchronization jump width defines the maximum of clock cycles a bit period may be shortened or lengthened by re-synchronization.
t SJW = t CSC ( SJW + 1 ) BRP: Baudrate Prescaler.

This field allows user to program the period of the CAN system clock to determine the individual bit timing.
t CSC = ( BRP + 1 ) MCK

The BRP field must be within the range [1, 0x7F], i.e., BRP = 0 is not authorized.
SMP: Sampling Mode

0 (ONCE): The incoming bit stream is sampled once at sample point. 1 (THREE): The incoming bit stream is sampled three times with a period of a MCK clock period, centered on sample point. SMP Sampling Mode is automatically disabled if BRP = 0.

1380
46.9.7 CAN Timer Register

31 23 15 7
CAN_TIM 0xF000C018 (0), 0xF8010018 (1) Read-only

30 22 14 6 29 21 13 5 28 20 12 TIMER 4 TIMER 3 2 1 0 27 19 11 26 18 10 25 17 9 24 16 8
TIMER: Timer
This field represents the internal CAN controller 16-bit timer value.

1381
46.9.8 CAN Timestamp Register

31 23 15 7
CAN_TIMESTP 0xF000C01C (0), 0xF801001C (1) Read-only

30 22 14 6 29 21 13 5 28 27 20 19 12 11 MTIMESTAMP 4 3 MTIMESTAMP 26 18 10 2 25 17 9 1 24 16 8 0
MTIMESTAMP: Timestamp
This field carries the value of the internal CAN controller 16-bit timer value at the start or end of frame. If the TEOF bit is cleared in the CAN_MR register, the internal Timer Counter value is captured in the MTIMESTAMP field at each start of frame else the value is captured at each end of frame. When the value is captured, the TSTP flag is set in the CAN_SR register. If the TSTP mask in the CAN_IMR register is set, an interrupt is generated while TSTP flag is set in the CAN_SR register. This flag is cleared by reading the CAN_SR register. Note: The CAN_TIMESTP register is reset when the CAN is disabled then enabled thanks to the CANEN bit in the CAN_MR.

1382
46.9.9 CAN Error Counter Register

31 23 15 7
CAN_ECR 0xF000C020 (0), 0xF8010020 (1) Read-only

30 22 14 6 29 21 13 5 28 20 TEC 12 4 REC 11 3 10 2 9 1 8 0 27 19 26 18 25 17 24 TEC 16
REC: Receive Error Counter

When a receiver detects an error, REC will be increased by one, except when the detected error is a BIT ERROR while sending an ACTIVE ERROR FLAG or an OVERLOAD FLAG. When a receiver detects a dominant bit as the first bit after sending an ERROR FLAG, REC is increased by 8. When a receiver detects a BIT ERROR while sending an ACTIVE ERROR FLAG, REC is increased by 8. Any node tolerates up to 7 consecutive dominant bits after sending an ACTIVE ERROR FLAG, PASSIVE ERROR FLAG or OVERLOAD FLAG. After detecting the 14th consecutive dominant bit (in case of an ACTIVE ERROR FLAG or an OVERLOAD FLAG) or after detecting the 8th consecutive dominant bit following a PASSIVE ERROR FLAG, and after each sequence of additional eight consecutive dominant bits, each receiver increases its REC by 8. After successful reception of a message, REC is decreased by 1 if it was between 1 and 127. If REC was 0, it stays 0, and if it was greater than 127, then it is set to a value between 119 and 127.
TEC: Transmit Error Counter

When a transmitter sends an ERROR FLAG, TEC is increased by 8 except when
the transmitter is error passive and detects an ACKNOWLEDGMENT ERROR because of not detecting a dominant ACK and does not detect a dominant bit while sending its PASSIVE ERROR FLAG. the transmitter sends an ERROR FLAG because a STUFF ERROR occurred during arbitration and should have been recessive and has been sent as recessive but monitored as dominant.
When a transmitter detects a BIT ERROR while sending an ACTIVE ERROR FLAG or an OVERLOAD FLAG, the TEC will be increased by 8. Any node tolerates up to 7 consecutive dominant bits after sending an ACTIVE ERROR FLAG, PASSIVE ERROR FLAG or OVERLOAD FLAG. After detecting the 14th consecutive dominant bit (in case of an ACTIVE ERROR FLAG or an OVERLOAD FLAG) or after detecting the 8th consecutive dominant bit following a PASSIVE ERROR FLAG, and after each sequence of additional eight consecutive dominant bits every transmitter increases its TEC by 8. After a successful transmission the TEC is decreased by 1 unless it was already 0.

1383
46.9.10 CAN Transfer Command Register

31 TIMRST 23 15 7 MB7
CAN_TCR 0xF000C024 (0), 0xF8010024 (1) Write-only

30 22 14 6 MB6 29 21 13 5 MB5 28 20 12 4 MB4 27 19 11 3 MB3 26 18 10 2 MB2 25 17 9 1 MB1 24 16 8 0 MB0
This register initializes several transfer requests at the same time.
MBx: Transfer Request for Mailbox x

Mailbox Object Type Receive Receive with overwrite Transmit Consumer Producer Description It receives the next message. This triggers a new reception. Sends data prepared in the mailbox as soon as possible. Sends a remote frame. Sends data prepared in the mailbox after receiving a remote frame from a consumer.
This flag clears the MRDY and MABT flags in the corresponding CAN_MSRx register. When several mailboxes are requested to be transmitted simultaneously, they are transmitted in turn, starting with the mailbox with the highest priority. If several mailboxes have the same priority, then the mailbox with the lowest number is sent first (i.e., MB0 will be transferred before MB1).
TIMRST: Timer Reset

Resets the internal timer counter. If the internal timer counter is frozen, this command automatically re-enables it. This command is useful in Time Triggered mode.

1384
46.9.11 CAN Abort Command Register

31 23 15 7 MB7
CAN_ACR 0xF000C028 (0), 0xF8010028 (1) Write-only

30 22 14 6 MB6 29 21 13 5 MB5 28 20 12 4 MB4 27 19 11 3 MB3 26 18 10 2 MB2 25 17 9 1 MB1 24 16 8 0 MB0
This register initializes several abort requests at the same time.
MBx: Abort Request for Mailbox x

Mailbox Object Type Receive Receive with overwrite Transmit Consumer Producer Description No action No action Cancels transfer request if the message has not been transmitted to the CAN transceiver. Cancels the current transfer before the remote frame has been sent. Cancels the current transfer. The next remote frame is not serviced.
It is possible to set MACR field (in the CAN_MCRx register) for each mailbox.

1385
46.9.12 CAN Write Protection Mode Register

31
CAN_WPMR 0xF000C0E4 (0), 0xF80100E4 (1) Read-write

30 29 28 27 26 25 24
WPKEY
23 22 21 20 19 18 17 16
WPKEY
15 14 13 12 11 10 9 8
WPKEY
7 6 5 4 3 2 1 0
WPEN

0: The Write Protection is Disabled 1: The Write Protection is Enabled
WPKEY: SPI Write Protection Key Password

If a value is written in WPEN, the value is taken into account only if WPKEY is written with CAN (CAN is written in ASCII Code, i.e. 0x43414E in hexadecimal). Protects the registers: Section 46.9.1 CAN Mode Register on page 1370 Section 46.9.6 CAN Baudrate Register on page 1380 Section 46.9.14 CAN Message Mode Register on page 1388 Section 46.9.15 CAN Message Acceptance Mask Register on page 1389 Section 46.9.16 CAN Message ID Register on page 1390

1386
46.9.13 CAN Write Protection Status Register

31
CAN_WPSR 0xF000C0E8 (0), 0xF80100E8 (1) Read-only

30 29 28 27 26 25 24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
WPVSRC
7 6 5 4 3 2 1 0
WPVS
WPVS: Write Protection Violation Status

0 = No Write Protect Violation has occurred since the last read of the CAN_WPSR register. 1 = A Write Protect Violation has occurred since the last read of the CAN_WPSR register. If this violation is an unauthorized attempt to write a protected register, the associated violation is reported into field WPVSRC.

This Field indicates the offset of the register concerned by the violation

1387
46.9.14 CAN Message Mode Register

Name: Address: CAN_MMRx [x=0..7] 0xF000C200 (0)[0], 0xF000C220 (0)[1], 0xF000C240 (0)[2], 0xF000C260 (0)[3], 0xF000C280 (0)[4], 0xF000C2A0 (0)[5], 0xF000C2C0 (0)[6], 0xF000C2E0 (0)[7], 0xF8010200 (1)[0], 0xF8010220 (1)[1], 0xF8010240 (1)[2], 0xF8010260 (1)[3], 0xF8010280 (1)[4], 0xF80102A0 (1)[5], 0xF80102C0 (1)[6], 0xF80102E0 (1)[7] Access:
31 23 15 7
Read-write
30 22 14 6 29 21 13 5 28 20 12 MTIMEMARK 4 MTIMEMARK 3 2 1 0 27 19 11 26 18 PRIOR 10 9 8 25 MOT 17 24 16
This register can only be written if the WPEN bit is cleared in CAN Write Protection Mode Register.
MTIMEMARK: Mailbox Timemark

This field is active in Time Triggered Mode. Transmit operations are allowed when the internal timer counter reaches the Mailbox Timemark. See Transmitting within a Time Window on page 1366. In Timestamp Mode, MTIMEMARK is set to 0.
PRIOR: Mailbox Priority

This field has no effect in receive and receive with overwrite modes. In these modes, the mailbox with the lowest number is serviced first. When several mailboxes try to transmit a message at the same time, the mailbox with the highest priority is serviced first. If several mailboxes have the same priority, the mailbox with the lowest number is serviced first (i.e., MBx0 is serviced before MBx 15 if they have the same priority).
MOT: Mailbox Object Type

This field allows the user to define the type of the mailbox. All mailboxes are independently configurable. Five different types are possible for each mailbox:
Value 0 1 2 3 4 Name MB_DISABLED MB_RX MB_RX_OVERWRITE MB_TX MB_CONSUMER Description Mailbox is disabled. This prevents receiving or transmitting any messages with this mailbox. Reception Mailbox. Mailbox is configured for reception. If a message is received while the mailbox data register is full, it is discarded. Reception mailbox with overwrite. Mailbox is configured for reception. If a message is received while the mailbox is full, it overwrites the previous message. Transmit mailbox. Mailbox is configured for transmission. Consumer Mailbox. Mailbox is configured in reception but behaves as a Transmit Mailbox, i.e., it sends a remote frame and waits for an answer. Producer Mailbox. Mailbox is configured in transmission but also behaves like a reception mailbox, i.e., it waits to receive a Remote Frame before sending its contents. Reserved
5 6
MB_PRODUCER

1388
46.9.15 CAN Message Acceptance Mask Register

Name: Address: CAN_MAMx [x=0..7] 0xF000C204 (0)[0], 0xF000C224 (0)[1], 0xF000C244 (0)[2], 0xF000C264 (0)[3], 0xF000C284 (0)[4], 0xF000C2A4 (0)[5], 0xF000C2C4 (0)[6], 0xF000C2E4 (0)[7], 0xF8010204 (1)[0], 0xF8010224 (1)[1], 0xF8010244 (1)[2], 0xF8010264 (1)[3], 0xF8010284 (1)[4], 0xF80102A4 (1)[5], 0xF80102C4 (1)[6], 0xF80102E4 (1)[7] Access:
31 23 15 7
Read-write
30 22 14 6 29 MIDE 21 MIDvA 13 5 12 MIDvB 4 MIDvB 3 2 1 0 11 10 9 28 20 27 19 26 MIDvA 18 25 17 MIDvB 8 24 16
This register can only be written if the WPEN bit is cleared in CAN Write Protection Mode Register. To prevent concurrent access with the internal CAN core, the application must disable the mailbox before writing to CAN_MAMx registers.
MIDvB: Complementary bits for identifier in extended frame mode

Acceptance mask for corresponding field of the message IDvB register of the mailbox.
MIDvA: Identifier for standard frame mode

Acceptance mask for corresponding field of the message IDvA register of the mailbox.
MIDE: Identifier Version

0: Compares IDvA from the received frame with the CAN_MIDx register masked with CAN_MAMx register. 1: Compares IDvA and IDvB from the received frame with the CAN_MIDx register masked with CAN_MAMx register.

1389
46.9.16 CAN Message ID Register

Name: Address: CAN_MIDx [x=0..7] 0xF000C208 (0)[0], 0xF000C228 (0)[1], 0xF000C248 (0)[2], 0xF000C268 (0)[3], 0xF000C288 (0)[4], 0xF000C2A8 (0)[5], 0xF000C2C8 (0)[6], 0xF000C2E8 (0)[7], 0xF8010208 (1)[0], 0xF8010228 (1)[1] 0xF8010248 (1)[2], 0xF8010268 (1)[3], 0xF8010288 (1)[4], 0xF80102A8 (1)[5], 0xF80102C8 (1)[6], 0xF80102E8 (1)[7] Access:
31 23 15 7
Read-write
30 22 14 6 29 MIDE 21 MIDvA 13 5 12 MIDvB 4 MIDvB 3 2 1 0 11 10 9 28 20 27 19 26 MIDvA 18 25 17 MIDvB 8 24 16
This register can only be written if the WPEN bit is cleared in CAN Write Protection Mode Register. To prevent concurrent access with the internal CAN core, the application must disable the mailbox before writing to CAN_MIDx registers.
MIDvB: Complementary bits for identifier in extended frame mode

If MIDE is cleared, MIDvB value is 0.
MIDE: Identifier Version

This bit allows the user to define the version of messages processed by the mailbox. If set, mailbox is dealing with version 2.0 Part B messages; otherwise, mailbox is dealing with version 2.0 Part A messages.
MIDvA: Identifier for standard frame mode

1390
46.9.17 CAN Message Family ID Register

Name: Address: CAN_MFIDx [x=0..7] 0xF000C20C (0)[0], 0xF000C22C (0)[1], 0xF000C24C (0)[2], 0xF000C26C (0)[3], 0xF000C28C (0)[4], 0xF000C2AC (0)[5], 0xF000C2CC (0)[6], 0xF000C2EC (0)[7], 0xF801020C (1)[0], 0xF801022C (1)[1], 0xF801024C (1)[2], 0xF801026C (1)[3], 0xF801028C (1)[4], 0xF80102AC (1)[5], 0xF80102CC (1)[6], 0xF80102EC (1)[7] Access:
31 23 15 7
Read-only
30 22 14 6 29 21 13 5 28 20 MFID 12 MFID 4 MFID 3 2 1 0 11 10 9 8 27 19 26 MFID 18 25 17 24 16
MFID: Family ID
This field contains the concatenation of CAN_MIDx register bits masked by the CAN_MAMx register. This field is useful to speed up message ID decoding. The message acceptance procedure is described below. As an example: CAN_MIDx = 0x305A4321 CAN_MAMx = 0x3FF0F0FF CAN_MFIDx = 0x000000A3

1391
46.9.18 CAN Message Status Register

Name: Address: CAN_MSRx [x=0..7] 0xF000C210 (0)[0], 0xF000C230 (0)[1], 0xF000C250 (0)[2], 0xF000C270 (0)[3], 0xF000C290 (0)[4], 0xF000C2B0 (0)[5], 0xF000C2D0 (0)[6], 0xF000C2F0 (0)[7], 0xF8010210 (1)[0], 0xF8010230 (1)[1], 0xF8010250 (1)[2], 0xF8010270 (1)[3], 0xF8010290 (1)[4], 0xF80102B0 (1)[5], 0xF80102D0 (1)[6], 0xF80102F0 (1)[7] Access:
31 23 MRDY 15 7
Read-only
30 22 MABT 14 6 29 21 13 5 28 27 20 19 MRTR 12 11 MTIMESTAMP 4 3 MTIMESTAMP 26 18 MDLC 10 2 9 1 8 0 25 17 24 MMI 16
These register fields are updated each time a message transfer is received or aborted. MMI is cleared by reading the CAN_MSRx register. MRDY, MABT are cleared by writing MTCR or MACR in the CAN_MCRx register. Warning: MRTR and MDLC state depends partly on the mailbox object type.
MTIMESTAMP: Timer value

This field is updated only when time-triggered operations are disabled (TTM cleared in CAN_MR register). If the TEOF field in the CAN_MR register is cleared, TIMESTAMP is the internal timer value at the start of frame of the last message received or sent by the mailbox. If the TEOF field in the CAN_MR register is set, TIMESTAMP is the internal timer value at the end of frame of the last message received or sent by the mailbox. In Time Triggered Mode, MTIMESTAMP is set to 0.
MDLC: Mailbox Data Length Code

Mailbox Object Type Receive Receive with overwrite Transmit Consumer Producer Description Length of the first mailbox message received Length of the last mailbox message received No action Length of the mailbox message received Length of the mailbox message to be sent after the remote frame reception
MRTR: Mailbox Remote Transmission Request

Mailbox Object Type Receive Receive with overwrite Transmit Consumer Producer Description The first frame received has the RTR bit set. The last frame received has the RTR bit set. Reserved Reserved. After setting the MOT field in the CAN_MMR, MRTR is reset to 1. Reserved. After setting the MOT field in the CAN_MMR, MRTR is reset to 0.

1392
MABT: Mailbox Message Abort

An interrupt is triggered when MABT is set. 0: Previous transfer is not aborted. 1: Previous transfer has been aborted. This flag is cleared by writing to CAN_MCRx register
Mailbox Object Type Receive Receive with overwrite Transmit Consumer Producer Description Reserved Reserved Previous transfer has been aborted The remote frame transfer request has been aborted. The response to the remote frame transfer has been aborted.
MRDY: Mailbox Ready

An interrupt is triggered when MRDY is set. 0: Mailbox data registers can not be read/written by the software application. CAN_MDx are locked by the CAN_MDx. 1: Mailbox data registers can be read/written by the software application. This flag is cleared by writing to CAN_MCRx register.
Mailbox Object Type Receive Description At least one message has been received since the last mailbox transfer order. Data from the first frame received can be read in the CAN_MDxx registers. After setting the MOT field in the CAN_MMR, MRDY is reset to 0. Receive with overwrite At least one frame has been received since the last mailbox transfer order. Data from the last frame received can be read in the CAN_MDxx registers. After setting the MOT field in the CAN_MMR, MRDY is reset to 0. Transmit Mailbox data have been transmitted. After setting the MOT field in the CAN_MMR, MRDY is reset to 1. At least one message has been received since the last mailbox transfer order. Data from the first message received can be read in the CAN_MDxx registers. After setting the MOT field in the CAN_MMR, MRDY is reset to 0. Producer A remote frame has been received, mailbox data have been transmitted. After setting the MOT field in the CAN_MMR, MRDY is reset to 1.
Consumer
MMI: Mailbox Message Ignored

0: No message has been ignored during the previous transfer 1: At least one message has been ignored during the previous transfer

1393
Cleared by reading the CAN_MSRx register.

Mailbox Object Type Receive Description Set when at least two messages intended for the mailbox have been sent. The first one is available in the mailbox data register. Others have been ignored. A mailbox with a lower priority may have accepted the message. Set when at least two messages intended for the mailbox have been sent. The last one is available in the mailbox data register. Previous ones have been lost. Reserved A remote frame has been sent by the mailbox but several messages have been received. The first one is available in the mailbox data register. Others have been ignored. Another mailbox with a lower priority may have accepted the message. A remote frame has been received, but no data are available to be sent.
Receive with overwrite Transmit Consumer Producer

1394
46.9.19 CAN Message Data Low Register

Name: Address: CAN_MDLx [x=0..7] 0xF000C214 (0)[0], 0xF000C234 (0)[1], 0xF000C254 (0)[2], 0xF000C274 (0)[3], 0xF000C294 (0)[4], 0xF000C2B4 (0)[5], 0xF000C2D4 (0)[6], 0xF000C2F4 (0)[7], 0xF8010214 (1)[0], 0xF8010234 (1)[1], 0xF8010254 (1)[2], 0xF8010274 (1)[3], 0xF8010294 (1)[4], 0xF80102B4 (1)[5], 0xF80102D4 (1)[6], 0xF80102F4 (1)[7] Access:
31
Read-write
30 29 28 MDL 27 26 25 24
23
22
21
20 MDL
19
18
17
16
15
14
13
12 MDL
11
10
4 MDL
MDL: Message Data Low Value

When MRDY field is set in the CAN_MSRx register, the lower 32 bits of a received message can be read or written by the software application. Otherwise, the MDL value is locked by the CAN controller to send/receive a new message. In Receive with overwrite, the CAN controller may modify MDL value while the software application reads MDH and MDL registers. To check that MDH and MDL do not belong to different messages, the application has to check the MMI field in the CAN_MSRx register. In this mode, the software application must re-read CAN_MDH and CAN_MDL, while the MMI bit in the CAN_MSRx register is set. Bytes are received/sent on the bus in the following order:
1. CAN_MDL[7:0] 2. CAN_MDL[15:8] 3. CAN_MDL[23:16] 4. CAN_MDL[31:24] 5. CAN_MDH[7:0] 6. CAN_MDH[15:8] 7. CAN_MDH[23:16] 8. CAN_MDH[31:24]

1395
46.9.20 CAN Message Data High Register

Name: Address: CAN_MDHx [x=0..7] 0xF000C218 (0)[0], 0xF000C238 (0)[1], 0xF000C258 (0)[2], 0xF000C278 (0)[3], 0xF000C298 (0)[4], 0xF000C2B8 (0)[5], 0xF000C2D8 (0)[6], 0xF000C2F8 (0)[7], 0xF8010218 (1)[0], 0xF8010238 (1)[1], 0xF8010258 (1)[2], 0xF8010278 (1)[3], 0xF8010298 (1)[4], 0xF80102B8 (1)[5], 0xF80102D8 (1)[6], 0xF80102F8 (1)[7] Access:
31
Read-write
30 29 28 MDH 27 26 25 24
23
22
21
20 MDH
19
18
17
16
15
14
13
12 MDH
11
10
4 MDH
MDH: Message Data High Value

When MRDY field is set in the CAN_MSRx register, the upper 32 bits of a received message are read or written by the software application. Otherwise, the MDH value is locked by the CAN controller to send/receive a new message. In Receive with overwrite, the CAN controller may modify MDH value while the software application reads MDH and MDL registers. To check that MDH and MDL do not belong to different messages, the application has to check the MMI field in the CAN_MSRx register. In this mode, the software application must re-read CAN_MDH and CAN_MDL, while the MMI bit in the CAN_MSRx register is set. Bytes are received/sent on the bus in the following order:
1. CAN_MDL[7:0] 2. CAN_MDL[15:8] 3. CAN_MDL[23:16] 4. CAN_MDL[31:24] 5. CAN_MDH[7:0] 6. CAN_MDH[15:8] 7. CAN_MDH[23:16] 8. CAN_MDH[31:24]

1396
46.9.21 CAN Message Control Register

Name: Address: CAN_MCRx [x=0..7] 0xF000C21C (0)[0], 0xF000C23C (0)[1], 0xF000C25C (0)[2], 0xF000C27C (0)[3], 0xF000C29C (0)[4], 0xF000C2BC (0)[5], 0xF000C2DC (0)[6], 0xF000C2FC (0)[7], 0xF801021C (1)[0], 0xF801023C (1)[1], 0xF801025C (1)[2], 0xF801027C (1)[3], 0xF801029C (1)[4], 0xF80102BC (1)[5], 0xF80102DC (1)[6], 0xF80102FC (1)[7] Access:
31 23 MTCR 15 7
Write-only
30 22 MACR 14 6 29 21 13 5 28 20 MRTR 12 4 27 19 26 18 MDLC 11 3 10 2 9 1 8 0 25 17 24 16
MDLC: Mailbox Data Length Code

Mailbox Object Type Receive Receive with overwrite Transmit Consumer Producer Description No action. No action. Length of the mailbox message. No action. Length of the mailbox message to be sent after the remote frame reception.
MRTR: Mailbox Remote Transmission Request

Mailbox Object Type Receive Receive with overwrite Transmit Consumer Producer Description No action No action Set the RTR bit in the sent frame No action, the RTR bit in the sent frame is set automatically No action
Consumer situations can be handled automatically by setting the mailbox object type in Consumer. This requires only one mailbox. It can also be handled using two mailboxes, one in reception, the other in transmission. The MRTR and the MTCR bits must be set in the same time.

1397
MACR: Abort Request for Mailbox x

Mailbox Object Type Receive Receive with overwrite Transmit Consumer Producer Description No action No action Cancels transfer request if the message has not been transmitted to the CAN transceiver. Cancels the current transfer before the remote frame has been sent. Cancels the current transfer. The next remote frame will not be serviced.
It is possible to set MACR field for several mailboxes in the same time, setting several bits to the CAN_ACR register.
MTCR: Mailbox Transfer Command

Mailbox Object Type Receive Receive with overwrite Transmit Consumer Producer Description Allows the reception of the next message. Triggers a new reception. Sends data prepared in the mailbox as soon as possible. Sends a remote transmission frame. Sends data prepared in the mailbox after receiving a remote frame from a Consumer.
This flag clears the MRDY and MABT flags in the CAN_MSRx register. When several mailboxes are requested to be transmitted simultaneously, they are transmitted in turn. The mailbox with the highest priority is serviced first. If several mailboxes have the same priority, the mailbox with the lowest number is serviced first (i.e., MBx0 will be serviced before MBx 15 if they have the same priority). It is possible to set MTCR for several mailboxes at the same time by writing to the CAN_TCR register.

1398
47.
47.1
Software Modem Device (SMD)

Description
The Software Modem Device (SMD) is a block for communication via a modem's Digital Isolation Barrier (DIB) with a complementary Line Side Device (HLSD). SMD and HLSD are two parts of the "Transformer only" solution. The transformer is the only component connecting SMD and HLSD. The transformer is used for power, clock and data transfers. Power and clock are supplied by the SMD and consumed by the HLSD. The data flow is bidirectional. The data transfer is based on pulse width modulation for transmission from the SMD to the HLSD, and for receiving from the HLSD. There are two channels embedded into the protocol of the DIB link:
z z
Data channel, Control channel.
Each channel is bidirectional. The data channel is used to transfer digitized signal samples at a constant rate of 16 bits at 16 kHz. The control channel is used to communicate with control registers of the HLSD at a maximum rate of 8 bits at 16 kHz. The SMD performs all protocol-related data conversion for transmission and received data interpretation in both data and control channels of the link. The SMD incorporates both RX and TX FIFOs, available through the DMAC interface. Each FIFO is able to hold eight 32bit words (equivalent to 16 modem data samples).
47.2
z
z z z z z
Modulations and protocols z V.90 z V.34 z V.32bis, V.32, V.22bis, V.22, V.23, V.21 z V.23 reverse, V.23 half-duplex z Bell 212A/Bell 103 z V.29 FastPOS z V.22bis fast connect z V.80 Synchronous Access Mode Data compression and error correction z V.44 data compression (V.92 model) z V.42bis and MNP 5 data compression z V.42 LAPM and MNP 2-4 error correction z EIA/TIA 578 Class 1 and T.31 Class 1.0 Call Waiting (CW) detection and Type II Caller ID decoding during data mode Type I Caller ID (CID) decoding Sixty-three embedded and upgradable country profiles Embedded AT commands SmartDAA z Extension pick-up detection z Digital line protection z Line reversal detection z Line-in-use detection z Remote hang-up detection z Worldwide compliance

1399
47.3
Block Diagram
Figure 47-1. Software Modem Device Block Diagram
SMD Controller SMD Core
Byte Parallel Interface CPU Interrupt Control Channel Logic Control/Status Registers Ring Detection and Pulse Dialing Machines (masters)
AHB
AHB Wrapper
FIFO Interface 8x32 (2)
DMA Parallel Interface
DMA Channel Logic
FIFO 2x16 FIFO 2x16
DIB Interface Circuitry
DIB Pads
X X

1400
48.
48.1
Timer Counter (TC)

Description
The Timer Counter (TC) includes 6 identical 32-bit Timer Counter channels. Each channel can be independently programmed to perform a wide range of functions including frequency measurement, event counting, interval measurement, pulse generation, delay timing and pulse width modulation. Each channel has three external clock inputs, five internal clock inputs and two multi-purpose input/output signals which can be configured by the user. Each channel drives an internal interrupt signal which can be programmed to generate processor interrupts. The Timer Counter (TC) embeds a quadrature decoder logic connected in front of the timers and driven by TIOA0, TIOB0 and TIOA1 inputs. When enabled, the quadrature decoder performs the input lines filtering, decoding of quadrature signals and connects to the timers/counters in order to read the position and speed of the motor through the user interface. The Timer Counter block has two global registers which act upon all TC channels. The Block Control Register allows the channels to be started simultaneously with the same instruction. The Block Mode Register defines the external clock inputs for each channel, allowing them to be chained. Table 48-1 gives the assignment of the device Timer Counter clock inputs common to Timer Counter 0 to 2.
Table 48-1. Timer Counter Clock Assignment Name TIMER_CLOCK1 TIMER_CLOCK2 TIMER_CLOCK3 TIMER_CLOCK4 TIMER_CLOCK5
(1)
Definition div2 div8 div32 div128 slow_clock
Note:
1.
When Slow Clock is selected for Master Clock (CSS = 0 in PMC Master Clock Register), TIMER_CLOCK5 input is equivalent to Master Clock.

1401
48.2
z z
Provides six 32-bit Timer Counter channels Wide range of functions including:
z z z z z z z z z
Frequency measurement Event counting Interval measurement Pulse generation Delay timing Pulse Width Modulation Up/down capabilities Quadrature decoder logic 2-bit gray up/down count for stepper motor Three external clock inputs Five Internal clock inputs Two multi-purpose input/output signals acting as trigger event
Each channel is user-configurable and contains:

z z z
z z z z
Internal interrupt signal Two global registers that act on all TC channels Read of the Capture registers by the DMAC Configuration registers can be write protected

1402
48.3
Block Diagram
Figure 48-1. Timer Counter Block Diagram
TIMER_CLOCK1
Parallel I/O Controller TCLK0 TIOA1
TIMER_CLOCK2 TIMER_CLOCK3
TCLK0 TCLK1 TCLK2 TIOA0 TIOB0
TIOA2 TCLK1
XC0 XC1 XC2 TC0XC0S
Timer/Counter Channel 0
TIOA
TIOA0
TIOB
TIMER_CLOCK4 TIMER_CLOCK5
TCLK2
TIOB0
SYNC
INT0
TCLK0 TCLK1 TIOA0 TIOA2 TCLK2 XC0 XC1 XC2 TC1XC1S

SYNC
TIOA
TIOA1
TIOB
TIOB1 INT1
TIOA1 TIOB1
TCLK0 TCLK1 TCLK2 TIOA0 TIOA1
XC0 XC1 XC2 TC2XC2S
TIOA
TIOA2
TIOB
TIOB2
SYNC
TIOA2 TIOB2
INT2
Timer Counter Interrupt Controller
Table 48-2. Signal Name Description Block/Channel Signal Name XC0, XC1, XC2 TIOA Channel Signal TIOB INT SYNC Description External Clock Inputs Capture Mode: Timer Counter Input Waveform Mode: Timer Counter Output Capture Mode: Timer Counter Input Waveform Mode: Timer Counter Input/Output Interrupt Signal Output (internal signal) Synchronization Input Signal (from configuration register)

1403
48.4
Pin Name List
Table 48-3. TC Pin List Pin Name TCLK0-TCLK2 TIOA0-TIOA2 TIOB0-TIOB2 Description External Clock Input I/O Line A I/O Line B Type Input I/O I/O
48.5
48.5.1 I/O Lines

The pins used for interfacing the compliant external devices may be multiplexed with PIO lines. The programmer must first program the PIO controllers to assign the TC pins to their peripheral functions.
Table 48-4. I/O Lines Instance TC0 TC0 TC0 TC0 TC0 TC0 TC0 TC0 TC0 TC1 TC1 TC1 TC1 TC1 TC1 TC1 TC1 TC1 Signal TCLK0 TCLK1 TCLK2 TIOA0 TIOA1 TIOA2 TIOB0 TIOB1 TIOB2 TCLK3 TCLK4 TCLK5 TIOA3 TIOA4 TIOA5 TIOB3 TIOB4 TIOB5 I/O Line PD7 PC14 PE29 PD5 PC12 PE27 PD6 PC13 PE28 PC2 PC5 PC8 PC0 PC3 PC6 PC1 PC4 PC7 Peripheral B B B B B B B B B B B B B B B B B B

The TC is clocked through the Power Management Controller (PMC), thus the programmer must first configure the PMC to enable the Timer Counter clock.
48.5.3 Interrupt
The TC has an interrupt line connected to the Interrupt Controller (IC). Handling the TC interrupt requires programming the IC before configuring the TC.

1404
48.6
48.6.1 TC Description
The 6 channels of the Timer Counter are independent and identical in operation except when quadrature decoder is enabled. The registers for channel programming are listed in Table 48-5 on page 1424.
48.6.2 32-bit Counter

Each channel is organized around a 32-bit counter. The value of the counter is incremented at each positive edge of the selected clock. When the counter has reached the value 0xFFFF and passes to 0x0000, an overflow occurs and the COVFS bit in TC_SR (Status Register) is set. The current value of the counter is accessible in real time by reading the Counter Value Register, TC_CV. The counter can be reset by a trigger. In this case, the counter value passes to 0x0000 on the next valid edge of the selected clock.
48.6.3 Clock Selection

At block level, input clock signals of each channel can either be connected to the external inputs TCLK0, TCLK1 or TCLK2, or be connected to the internal I/O signals TIOA0, TIOA1 or TIOA2 for chaining by programming the TC_BMR (Block Mode). See Figure 48-2 Clock Chaining Selection. Each channel can independently select an internal or external clock source for its counter:
z z
Internal clock signals: TIMER_CLOCK1, TIMER_CLOCK2, TIMER_CLOCK3, TIMER_CLOCK4, TIMER_CLOCK5 External clock signals: XC0, XC1 or XC2
This selection is made by the TCCLKS bits in the TC Channel Mode Register. The selected clock can be inverted with the CLKI bit in TC_CMR. This allows counting on the opposite edges of the clock. The burst function allows the clock to be validated when an external signal is high. The BURST parameter in the Mode Register defines this signal (none, XC0, XC1, XC2). See Figure 48-3 Clock Selection Note: In all cases, if an external clock is used, the duration of each of its levels must be longer than the master clock period. The external clock frequency must be at least 2.5 times lower than the master clock

1405
Figure 48-2. Clock Chaining Selection

TC0XC0S Timer/Counter Channel 0 TIOA1 TIOA2 XC0 XC1 = TCLK1 XC2 = TCLK2 TIOB0 TIOA0
TCLK0
SYNC
TC1XC1S Timer/Counter Channel 1 TCLK1 TIOA0 TIOA2 XC0 = TCLK0 XC1 XC2 = TCLK2 TIOB1 TIOA1
SYNC
TC2XC2S
Timer/Counter Channel 2 XC0 = TCLK0 TIOA2
TCLK2 TIOA0 TIOA1
XC1 = TCLK1 XC2 TIOB2
SYNC
Figure 48-3. Clock Selection

TCCLKS TIMER_CLOCK1 TIMER_CLOCK2 TIMER_CLOCK3 TIMER_CLOCK4 TIMER_CLOCK5 XC0 XC1 XC2 MCK BURST CLKI Synchronous Edge Detection
Selected Clock

1406
48.6.4 Clock Control

The clock of each counter can be controlled in two different ways: it can be enabled/disabled and started/stopped. See Figure 48-4. The clock can be enabled or disabled by the user with the CLKEN and the CLKDIS commands in the Control Register. In Capture Mode it can be disabled by an RB load event if LDBDIS is set to 1 in TC_CMR. In Waveform Mode, it can be disabled by an RC Compare event if CPCDIS is set to 1 in TC_CMR. When disabled, the start or the stop actions have no effect: only a CLKEN command in the Control Register can re-enable the clock. When the clock is enabled, the CLKSTA bit is set in the Status Register. The clock can also be started or stopped: a trigger (software, synchro, external or compare) always starts the clock. The clock can be stopped by an RB load event in Capture Mode (LDBSTOP = 1 in TC_CMR) or a RC compare event in Waveform Mode (CPCSTOP = 1 in TC_CMR). The start and the stop commands have effect only if the clock is enabled.
Figure 48-4. Clock Control
Selected Clock Trigger
CLKSTA
CLKEN
CLKDIS
Q Q S R
S R
Counter Clock
Stop Event
Disable Event
48.6.5 TC Operating Modes

Each channel can independently operate in two different modes:
z z
Capture Mode provides measurement on signals. Waveform Mode provides wave generation.
The TC Operating Mode is programmed with the WAVE bit in the TC Channel Mode Register. In Capture Mode, TIOA and TIOB are configured as inputs. In Waveform Mode, TIOA is always configured to be an output and TIOB is an output if it is not selected to be the external trigger.
48.6.6 Trigger
A trigger resets the counter and starts the counter clock. Three types of triggers are common to both modes, and a fourth external trigger is available to each mode. Regardless of the trigger used, it will be taken into account at the following active edge of the selected clock. This means that the counter value can be read differently from zero just after a trigger, especially when a low frequency signal is selected as the clock.

1407
The following triggers are common to both modes:

z z
Software Trigger: Each channel has a software trigger, available by setting SWTRG in TC_CCR. SYNC: Each channel has a synchronization signal SYNC. When asserted, this signal has the same effect as a software trigger. The SYNC signals of all channels are asserted simultaneously by writing TC_BCR (Block Control) with SYNC set. Compare RC Trigger: RC is implemented in each channel and can provide a trigger when the counter value matches the RC value if CPCTRG is set in TC_CMR.
The channel can also be configured to have an external trigger. In Capture Mode, the external trigger signal can be selected between TIOA and TIOB. In Waveform Mode, an external event can be programmed on one of the following signals: TIOB, XC0, XC1 or XC2. This external event can then be programmed to perform a trigger by setting ENETRG in TC_CMR. If an external trigger is used, the duration of the pulses must be longer than the master clock period in order to be detected.
48.6.7 Capture Operating Mode

This mode is entered by clearing the WAVE parameter in TC_CMR (Channel Mode Register). Capture Mode allows the TC channel to perform measurements such as pulse timing, frequency, period, duty cycle and phase on TIOA and TIOB signals which are considered as inputs. Figure 48-6 shows the configuration of the TC channel when programmed in Capture Mode.
48.6.8 Capture Registers A and B

Registers A and B (RA and RB) are used as capture registers. This means that they can be loaded with the counter value when a programmable event occurs on the signal TIOA. The LDRA parameter in TC_CMR defines the TIOA selected edge for the loading of register A, and the LDRB parameter defines the TIOA selected edge for the loading of Register B. RA is loaded only if it has not been loaded since the last trigger or if RB has been loaded since the last loading of RA. RB is loaded only if RA has been loaded since the last trigger or the last loading of RB. Loading RA or RB before the read of the last value loaded sets the Overrun Error Flag (LOVRS) in TC_SR (Status Register). In this case, the old value is overwritten. When DMA is used, the RAB register address must be configured as source address of the transfer. The RAB register provides the next unread value from Register A and Register B. It may be read by the DMA after a request has been triggered upon loading Register A or Register B.
48.6.9 Transfer with DMAC

The DMAC can only perform access from timer to system memory. The figure Example of transfer with DMAC on page 1409 illustrates how the RA and RB registers can be loaded in the system memory without CPU intervention.

1408
Figure 48-5. Example of transfer with DMAC
TIOA RA RB Internal PDC trigger Transfer to System Memory RA T1 RB T2 RA T3 RB T4
T1,T2,T3,T4 = System Bus load dependent (Tmin = 8 MCK)
ETRGEDG=3, LDRA=3, LDRB=0, ABETRG=0 TIOB TIOA RA
Internal PDC trigger Transfer to System Memory RA T1 T1,T2,T3,T4 = System Bus load dependent (Tmin = 8 MCK)
48.6.10 Trigger Conditions
In addition to the SYNC signal, the software trigger and the RC compare trigger, an external trigger can be defined. The ABETRG bit in the TC_CMR register selects TIOA or TIOB input signal as an external trigger. The ETRGEDG parameter defines the edge (rising, falling or both) detected to generate an external trigger. If ETRGEDG = 0 (none), the external trigger is disabled.
RA T2
RA T3
RA T4

1409
TCCLKS
TIMER_CLOCK1
TIMER_CLOCK2
Synchronous Edge Detection
CLKI
CLKSTA
CLKEN
CLKDIS
TIMER_CLOCK3
Figure 48-6. Capture Mode
TIMER_CLOCK4
Q Q R S R
TIMER_CLOCK5
XC0
XC1 LDBSTOP
XC2
MCK
BURST
LDBDIS
Register C
1 Counter CLK OVF RESET Trig ABETRG ETRGEDG Edge Detector LDRA LDRB CPCTRG
Capture Register A
Capture Register B
Compare RC =
SWTRG
SYNC
MTIOB
TIOB
CPCS
LOVRS
LDRAS
LDRBS
ETRGS
COVFS
TC1_SR
MTIOA If RA is not loaded or RB is Loaded
Edge Detector If RA is Loaded
Edge Detector
TC1_IMR
TIOA

Timer/Counter Channel
INT
1410
48.6.11 Waveform Operating Mode

Waveform operating mode is entered by setting the WAVE parameter in TC_CMR (Channel Mode Register). In Waveform Operating Mode the TC channel generates 1 or 2 PWM signals with the same frequency and independently programmable duty cycles, or generates different types of one-shot or repetitive pulses. In this mode, TIOA is configured as an output and TIOB is defined as an output if it is not used as an external event (EEVT parameter in TC_CMR). Figure 48-7 shows the configuration of the TC channel when programmed in Waveform Operating Mode.
48.6.12 Waveform Selection

Depending on the WAVSEL parameter in TC_CMR (Channel Mode Register), the behavior of TC_CV varies. With any selection, RA, RB and RC can all be used as compare registers. RA Compare is used to control the TIOA output, RB Compare is used to control the TIOB output (if correctly configured) and RC Compare is used to control TIOA and/or TIOB outputs.

1411
TCCLKS CLKSTA ACPC

Synchronous Edge Detection
CLKEN
CLKDIS
TIMER_CLOCK1 CLKI
Figure 48-7. Waveform Mode
TIMER_CLOCK2
TIMER_CLOCK3
TIMER_CLOCK4 CPCDIS
Q Q R
CPCSTOP
MCK
S R
ACPA MTIOA
TIMER_CLOCK5
XC0
XC1
BURST Register A WAVSEL Register B Register C
ASWTRG Compare RA = Compare RB = Compare RC =
1
Counter
CLK RESET OVF
SWTRG
BCPC Trig BCPB WAVSEL EEVT BEEVT EEVTEDG ENETRG Edge Detector CPCS CPAS CPBS ETRGS COVFS TC1_SR MTIOB
SYNC
Output Controller
XC2
AEEVT
TIOA
TIOB TC1_IMR
BSWTRG
Timer/Counter Channel
INT
Output Controller
TIOB
1412
48.6.12.1 WAVSEL = 00 When WAVSEL = 00, the value of TC_CV is incremented from 0 to 0xFFFF. Once 0xFFFF has been reached, the value of TC_CV is reset. Incrementation of TC_CV starts again and the cycle continues. See Figure 48-8. An external event trigger or a software trigger can reset the value of TC_CV. It is important to note that the trigger may occur at any time. See Figure 48-9. RC Compare cannot be programmed to generate a trigger in this configuration. At the same time, RC Compare can stop the counter clock (CPCSTOP = 1 in TC_CMR) and/or disable the counter clock (CPCDIS = 1 in TC_CMR).
Figure 48-8. WAVSEL= 00 without trigger
Counter Value 0xFFFF Counter cleared by compare match with 0xFFFF
RC RB
RA
Waveform Examples TIOB
Time
TIOA
Figure 48-9. WAVSEL= 00 with trigger

Counter Value 0xFFFF Counter cleared by trigger Counter cleared by compare match with 0xFFFF
RC RB
RA
Time
TIOA

1413
48.6.12.2 WAVSEL = 10 When WAVSEL = 10, the value of TC_CV is incremented from 0 to the value of RC, then automatically reset on a RC Compare. Once the value of TC_CV has been reset, it is then incremented and so on. See Figure 48-10. It is important to note that TC_CV can be reset at any time by an external event or a software trigger if both are programmed correctly. See Figure 48-11. In addition, RC Compare can stop the counter clock (CPCSTOP = 1 in TC_CMR) and/or disable the counter clock (CPCDIS = 1 in TC_CMR).
Figure 48-10. WAVSEL = 10 Without Trigger
Counter Value 0xFFFF Counter cleared by compare match with RC RC RB
RA
Time
TIOA
Figure 48-11. WAVSEL = 10 With Trigger

Counter Value 0xFFFF Counter cleared by compare match with RC RC RB Counter cleared by trigger
RA
Time
TIOA

1414
48.6.12.3 WAVSEL = 01 When WAVSEL = 01, the value of TC_CV is incremented from 0 to 0xFFFF. Once 0xFFFF is reached, the value of TC_CV is decremented to 0, then re-incremented to 0xFFFF and so on. See Figure 48-12. A trigger such as an external event or a software trigger can modify TC_CV at any time. If a trigger occurs while TC_CV is incrementing, TC_CV then decrements. If a trigger is received while TC_CV is decrementing, TC_CV then increments. See Figure 48-13. RC Compare cannot be programmed to generate a trigger in this configuration. At the same time, RC Compare can stop the counter clock (CPCSTOP = 1) and/or disable the counter clock (CPCDIS = 1).
Counter Value 0xFFFF Counter decremented by compare match with 0xFFFF
RC RB
RA
Time
TIOA

Counter Value 0xFFFF Counter decremented by trigger RC RB Counter decremented by compare match with 0xFFFF
Counter incremented by trigger
RA
Time
TIOA

1415
48.6.12.4 WAVSEL = 11 When WAVSEL = 11, the value of TC_CV is incremented from 0 to RC. Once RC is reached, the value of TC_CV is decremented to 0, then re-incremented to RC and so on. See Figure 48-14. A trigger such as an external event or a software trigger can modify TC_CV at any time. If a trigger occurs while TC_CV is incrementing, TC_CV then decrements. If a trigger is received while TC_CV is decrementing, TC_CV then increments. See Figure 48-15. RC Compare can stop the counter clock (CPCSTOP = 1) and/or disable the counter clock (CPCDIS = 1).
Counter Value 0xFFFF Counter decremented by compare match with RC RC RB
RA
Time
TIOA

Counter Value 0xFFFF Counter decremented by compare match with RC RC RB Counter decremented by trigger Counter incremented by trigger
RA
Time
TIOA

1416
48.6.13 External Event/Trigger Conditions

An external event can be programmed to be detected on one of the clock sources (XC0, XC1, XC2) or TIOB. The external event selected can then be used as a trigger. The EEVT parameter in TC_CMR selects the external trigger. The EEVTEDG parameter defines the trigger edge for each of the possible external triggers (rising, falling or both). If EEVTEDG is cleared (none), no external event is defined. If TIOB is defined as an external event signal (EEVT = 0), TIOB is no longer used as an output and the compare register B is not used to generate waveforms and subsequently no IRQs. In this case the TC channel can only generate a waveform on TIOA. When an external event is defined, it can be used as a trigger by setting bit ENETRG in TC_CMR. As in Capture Mode, the SYNC signal and the software trigger are also available as triggers. RC Compare can also be used as a trigger depending on the parameter WAVSEL.
48.6.14 Output Controller

The output controller defines the output level changes on TIOA and TIOB following an event. TIOB control is used only if TIOB is defined as output (not as an external event). The following events control TIOA and TIOB: software trigger, external event and RC compare. RA compare controls TIOA and RB compare controls TIOB. Each of these events can be programmed to set, clear or toggle the output as defined in the corresponding parameter in TC_CMR.
48.6.15 Quadrature Decoder Logic

48.6.15.1 Description The quadrature decoder logic is driven by TIOA0, TIOB0, TIOB1 input pins and drives the timer/counter of channel 0 and 1. Channel 2 can be used as a time base in case of speed measurement requirements (refer to Figure 48.7 Timer Counter (TC) User Interface). When writing 0 in the QDEN field of the TC_BMR register, the quadrature decoder logic is totally transparent. TIOA0 and TIOB0 are to be driven by the 2 dedicated quadrature signals from a rotary sensor mounted on the shaft of the off-chip motor. A third signal from the rotary sensor can be processed through pin TIOB1 and is typically dedicated to be driven by an index signal if it is provided by the sensor. This signal is not required to decode the quadrature signals PHA, PHB. TCCLKS field of TC_CMR channels must be configured to select XC0 input (i.e. 0x101). TC0XC0S field has no effect as soon as quadrature decoder is enabled. Either speed or position/revolution can be measured. Position channel 0 accumulates the edges of PHA, PHB input signals giving a high accuracy on motor position whereas channel 1 accumulates the index pulses of the sensor, therefore the number of rotations. Concatenation of both values provides a high level of precision on motion system position. In speed mode, position cannot be measured but revolution can be measured. Inputs from the rotary sensor can be filtered prior to down-stream processing. Accommodation of input polarity, phase definition and other factors are configurable. Interruptions can be generated on different events. A compare function (using TC_RC register) is available on channel 0 (speed/position) or channel 1 (rotation) and can generate an interrupt by means of the CPCS flag in the TC_SR registers.

1417
Figure 48-16. Predefined Connection of the Quadrature Decoder with Timer Counters
Reset pulse Quadrature Decoder (Filter + Edge Detect + QD) PHEdges TIOB0 PHA QDEN SPEEDEN 1 1 TIOA TIOA0 QDEN 1 TIOB 1 XC0 XC0 Speed/Position Timer/Counter Channel 0
TIOA0
TIOB0
PHB Index
TIOB1
IDX
1 TIOB1
TIOB 1 XC0
XC0
Rotation Direction Timer/Counter Channel 2
Speed Time Base

1418
48.6.15.2 Input Pre-processing Input pre-processing consists of capabilities to take into account rotary sensor factors such as polarities and phase definition followed by configurable digital filtering. Each input can be negated and swapping PHA, PHB is also configurable. By means of the MAXFILT field in TC_BMR, it is possible to configure a minimum duration for which the pulse is stated as valid. When the filter is active, pulses with a duration lower than MAXFILT+1 * tMCK ns are not passed to downstream logic. Filters can be disabled using the FILTER field in the TC_BMR register.
Figure 48-17. Input Stage
Input Pre-Processing
SWAP 1 TIOA0 INVA 1 TIOB0 INVB 1 1 TIOB1 IDXPHB INVIDX IDX PHB PHA
MAXFILT
FILTER
Filter
PHedge
Direction and Edge Detection Filter 1 DIR
Filter
IDX
Input filtering can efficiently remove spurious pulses that might be generated by the presence of particulate contamination on the optical or magnetic disk of the rotary sensor. Spurious pulses can also occur in environments with high levels of electro-magnetic interference. Or, simply if vibration occurs even when rotation is fully stopped and the shaft of the motor is in such a position that the beginning of one of the reflective or magnetic bars on the rotary sensor disk is aligned with the light or magnetic (Hall) receiver cell of the rotary sensor. Any vibration can make the PHA, PHB signals toggle for a short duration.

1419
Figure 48-18. Filtering Examples
MCK
MAXFILT=2
particulate contamination PHA,B
Filter Out
Optical/Magnetic disk strips PHA
PHB motor shaft stopped in such a position that rotary sensor cell is aligned with an edge of the disk rotation stop PHA PHB Edge area due to system vibration PHB stop mechanical shock on system
Resulting PHA, PHB electrical waveforms PHA
PHB
vibration PHA, PHB electrical waveforms after filtering PHA PHB

1420
48.6.15.3 Direction Status and Change Detection After filtering, the quadrature signals are analyzed to extract the rotation direction and edges of the 2 quadrature signals detected in order to be counted by timer/counter logic downstream. The direction status can be directly read at anytime on TC_QISR register. The polarity of the direction flag status depends on the configuration written in TC_BMR register. INVA, INVB, INVIDX, SWAP modify the polarity of DIR flag. Any change in rotation direction is reported on TC_QISR register and can generate an interrupt. The direction change condition is reported as soon as 2 consecutive edges on a phase signal have sampled the same value on the other phase signal and there is an edge on the other signal. The 2 consecutive edges of 1 phase signal sampling the same value on other phase signal is not sufficient to declare a direction change, for the reason that particulate contamination may mask one or more reflective bar on the optical or magnetic disk of the sensor. (Refer to Figure 48-19 Rotation Change Detection for waveforms.)
Figure 48-19. Rotation Change Detection
Direction Change under normal conditions
PHA
change condition Report Time
PHB
DIR
DIRCHG
No direction change due to particulate contamination masking a reflective bar missing pulse PHA same phase PHB DIR spurious change condition (if detected in a simple way) DIRCHG
The direction change detection is disabled when QDTRANS is set to 1 in TC_BMR. In this case the DIR flag report must not be used. A quadrature error is also reported by the quadrature decoder logic. Rather than reporting an error only when 2 edges occur at the same time on PHA and PHB, which is unlikely to occur in real life, there is a report if the time difference between 2 edges on PHA, PHB is lower than a predefined value. This predefined value is configurable and corresponds

1421
to (MAXFILT+1) * tMCK ns. After being filtered there is no reason to have 2 edges closer than (MAXFILT+1) * tMCK ns under normal mode of operation. In the instance an anomaly occurs, a quadrature error is reported on QERR flag on TC_QISR register.
Figure 48-20. Quadrature Error Detection
MCK
MAXFILT = 2
Abnormally formatted optical disk strips (theoretical view) PHA
PHB strip edge inaccurary due to disk etching/printing process PHA
PHB resulting PHA, PHB electrical waveforms PHA Even with an abnorrmaly formatted disk, there is no occurence of PHA, PHB switching at the same time. PHB duration < MAXFILT QERR
MAXFILT must be tuned according to several factors such as the system clock frequency (MCK), type of rotary sensor and rotation speed to be achieved. 48.6.15.4 Position and Rotation Measurement When POSEN is set in TC_BMR register, position is processed on channel 0 (by means of the PHA,PHB edge detections) and motor revolutions are accumulated in channel 1 timer/counter and can be read through TC_CV0 and/or TC_CV1 register if the IDX signal is provided on TIOB1 input. Channel 0 and 1 must be configured in capture mode (WAVE = 0 in TC_CMR0). In parallel, the number of edges are accumulated on timer/counter channel 0 and can be read on the TC_CV0 register. Therefore, the accurate position can be read on both TC_CV registers and concatenated to form a 32-bit word. The timer/counter channel 0 is cleared for each increment of IDX count value. Depending on the quadrature signals, the direction is decoded and allows to count up or down in timer/counter channels 0 and 1. The direction status is reported on TC_QISR register.

1422
48.6.15.5 Speed Measurement When SPEEDEN is set in TC_BMR register, the speed measure is enabled on channel 0. A time base must be defined on channel 2 by writing the TC_RC2 period register. Channel 2 must be configured in waveform mode (WAVE bit field set) in TC_CMR2 register. WAVSEL bit field must be defined with 0x10 to clear the counter by comparison and matching with TC_RC value. ACPC field must be defined at 0x11 to toggle TIOA output. This time base is automatically fed back to TIOA of channel 0 when QDEN and SPEEDEN are set. Channel 0 must be configured in capture mode (WAVE = 0 in TC_CMR0). ABETRG bit field of TC_CMR0 must be configured at 1 to get TIOA as a trigger for this channel. EDGTRG can be set to 0x01, to clear the counter on a rising edge of the TIOA signal and LDRA field must be set accordingly to 0x01, to load TC_RA0 at the same time as the counter is cleared (LDRB must be set to 0x01). As a consequence, at the end of each time base period the differentiation required for the speed calculation is performed. The process must be started by configuring the TC_CR register with CLKEN and SWTRG. The speed can be read on TC_RA0 register in TC_CMR0. Channel 1 can still be used to count the number of revolutions of the motor.
48.6.16 2-bit Gray Up/Down Counter for Stepper Motor

Each channel can be independently configured to generate a 2-bit gray count waveform on corresponding TIOA,TIOB outputs by means of GCEN bit in TC_SMMRx registers. Up or Down count can be defined by writing bit DOWN in TC_SMMRx registers. It is mandatory to configure the channel in WAVE mode in TC_CMR register. The period of the counters can be programmed on TC_RCx registers.
Figure 48-21. 2-bit Gray Up/Down Counter.
WAVEx = GCENx =1 TIOAx TC_RCx TIOBx
DOWNx
48.6.17 Write Protection System
In order to bring security to the Timer Counter, a write protection system has been implemented. The write protection mode prevent the write of TC_BMR, TC_CMRx, TC_SMMRx, TC_RAx, TC_RBx, TC_RCx registers. When this mode is enabled and one of the protected registers write, the register write request canceled. Due to the nature of the write protection feature, enabling and disabling the write protection mode requires the use of a security code. Thus when enabling or disabling the write protection mode the WPKEY field of the TC_WPMR register must be filled with the TIM ASCII code (corresponding to 0x54494D) otherwise the register write will be canceled.

1423
48.7
Timer Counter (TC) User Interface

Offset(1)
Table 48-5. Register Mapping Register Channel Control Register Channel Mode Register Stepper Motor Mode Register Register AB Counter Value Register A Register B Register C Status Register Interrupt Enable Register Interrupt Disable Register Interrupt Mask Register Block Control Register Block Mode Register QDEC Interrupt Enable Register QDEC Interrupt Disable Register QDEC Interrupt Mask Register QDEC Interrupt Status Register Reserved Write Protect Mode Register Reserved TC_WPMR Read-write 0 Name TC_CCR TC_CMR TC_SMMR TC_RAB TC_CV TC_RA TC_RB TC_RC TC_SR TC_IER TC_IDR TC_IMR TC_BCR TC_BMR TC_QIER TC_QIDR TC_QIMR TC_QISR Access Write-only Read-write Read-write Read-only Read-only Read-write Read-write
(2) (2)
Reset 0 0 0 0 0 0 0 0 0 0 0 0
0x00 + channel * 0x40 + 0x00 0x00 + channel * 0x40 + 0x04 0x00 + channel * 0x40 + 0x08 0x00 + channel * 0x40 + 0x0C 0x00 + channel * 0x40 + 0x10 0x00 + channel * 0x40 + 0x14 0x00 + channel * 0x40 + 0x18 0x00 + channel * 0x40 + 0x1C 0x00 + channel * 0x40 + 0x20 0x00 + channel * 0x40 + 0x24 0x00 + channel * 0x40 + 0x28 0x00 + channel * 0x40 + 0x2C 0xC0 0xC4 0xC8 0xCC 0xD0 0xD4 0xD8 0xE4 0xE8 - 0xFC
Read-write Read-only Write-only Write-only Read-only Write-only Read-write Write-only Write-only Read-only Read-only
Notes: 1. Channel index ranges from 0 to 2. 2. Read-only if WAVE = 0

1424
48.7.1 TC Channel Control Register

31 23 15 7
TC_CCRx [x=0..2] 0xF0010000 (0)[0], 0xF0010040 (0)[1], 0xF0010080 (0)[2], 0xF8014000 (1)[0], 0xF8014040 (1)[1], 0xF8014080 (1)[2] Write-only
30 22 14 6 29 21 13 5 28 20 12 4 27 19 11 3 26 18 10 2 SWTRG 25 17 9 1 CLKDIS 24 16 8 0 CLKEN
CLKEN: Counter Clock Enable Command

0 = no effect. 1 = enables the clock if CLKDIS is not 1.
CLKDIS: Counter Clock Disable Command

0 = no effect. 1 = disables the clock.
SWTRG: Software Trigger Command

0 = no effect. 1 = a software trigger is performed: the counter is reset and the clock is started.

1425
48.7.2 TC Channel Mode Register: Capture Mode

31 23 15 WAVE 7 LDBDIS
TC_CMRx [x=0..2] (WAVE = 0) 0xF0010004 (0)[0], 0xF0010044 (0)[1], 0xF0010084 (0)[2], 0xF8014004 (1)[0], 0xF8014044 (1)[1], 0xF8014084 (1)[2] Read-write
30 22 14 CPCTRG 6 LDBSTOP 29 21 13 5 BURST 28 20 12 4 11 3 CLKI 27 19 LDRB 10 ABETRG 2 1 TCCLKS 9 ETRGEDG 0 26 18 25 17 LDRA 8 24 16
This register can only be written if the WPEN bit is cleared in TC Write Protect Mode Register on page 1450
TCCLKS: Clock Selection

Value 0 1 2 3 4 5 6 7 Name TIMER_CLOCK1 TIMER_CLOCK2 TIMER_CLOCK3 TIMER_CLOCK4 TIMER_CLOCK5 XC0 XC1 XC2 Description Clock selected: TCLK1 Clock selected: TCLK2 Clock selected: TCLK3 Clock selected: TCLK4 Clock selected: TCLK5 Clock selected: XC0 Clock selected: XC1 Clock selected: XC2
CLKI: Clock Invert

0 = Counter is incremented on rising edge of the clock. 1 = Counter is incremented on falling edge of the clock.
BURST: Burst Signal Selection

Value 0 1 2 3 Name NONE XC0 XC1 XC2 Description The clock is not gated by an external signal. XC0 is ANDed with the selected clock. XC1 is ANDed with the selected clock. XC2 is ANDed with the selected clock.
LDBSTOP: Counter Clock Stopped with RB Loading

0 = Counter clock is not stopped when RB loading occurs. 1 = Counter clock is stopped when RB loading occurs.

1426
LDBDIS: Counter Clock Disable with RB Loading

0 = Counter clock is not disabled when RB loading occurs. 1 = Counter clock is disabled when RB loading occurs.
ETRGEDG: External Trigger Edge Selection

Value 0 1 2 3 Name NONE RISING FALLING EDGE Description The clock is not gated by an external signal. Rising edge Falling edge Each edge
ABETRG: TIOA or TIOB External Trigger Selection

0 = TIOB is used as an external trigger. 1 = TIOA is used as an external trigger.
CPCTRG: RC Compare Trigger Enable

0 = RC Compare has no effect on the counter and its clock. 1 = RC Compare resets the counter and starts the counter clock.
WAVE: Waveform Mode

0 = Capture Mode is enabled. 1 = Capture Mode is disabled (Waveform Mode is enabled).
LDRA: RA Loading Edge Selection

Value 0 1 2 3 Name NONE RISING FALLING EDGE Description None Rising edge of TIOA Falling edge of TIOA Each edge of TIOA
LDRB: RB Loading Edge Selection

Value 0 1 2 3 Name NONE RISING FALLING EDGE Description None Rising edge of TIOA Falling edge of TIOA Each edge of TIOA

1427
48.7.3 TC Channel Mode Register: Waveform Mode

Name: Access:
31 BSWTRG 23 ASWTRG 15 WAVE 7 CPCDIS 6 CPCSTOP 14 WAVSEL 5 BURST 13 22 21 AEEVT 12 ENETRG 4 3 CLKI 11 EEVT 2 1 TCCLKS
TC_CMRx [x=0..2] (WAVE = 1) Read-write

30 29 BEEVT 20 19 ACPC 10 9 EEVTEDG 0 28 27 BCPC 18 17 ACPA 8 26 25 BCPB 16 24
TCCLKS: Clock Selection

Value 0 1 2 3 4 5 6 7 Name TIMER_CLOCK1 TIMER_CLOCK2 TIMER_CLOCK3 TIMER_CLOCK4 TIMER_CLOCK5 XC0 XC1 XC2 Description Clock selected: TCLK1 Clock selected: TCLK2 Clock selected: TCLK3 Clock selected: TCLK4 Clock selected: TCLK5 Clock selected: XC0 Clock selected: XC1 Clock selected: XC2
CLKI: Clock Invert

0 = Counter is incremented on rising edge of the clock. 1 = Counter is incremented on falling edge of the clock.
BURST: Burst Signal Selection

Value 0 1 2 3 Name NONE XC0 XC1 XC2 Description The clock is not gated by an external signal. XC0 is ANDed with the selected clock. XC1 is ANDed with the selected clock. XC2 is ANDed with the selected clock.
CPCSTOP: Counter Clock Stopped with RC Compare

0 = Counter clock is not stopped when counter reaches RC. 1 = Counter clock is stopped when counter reaches RC.
CPCDIS: Counter Clock Disable with RC Compare

0 = Counter clock is not disabled when counter reaches RC. 1 = Counter clock is disabled when counter reaches RC.

1428
EEVTEDG: External Event Edge Selection

Value 0 1 2 3 Name NONE RISING FALLING EDGE Description None Rising edge Falling edge Each edge
EEVT: External Event Selection

Signal selected as external event.
Value 0 1 2 3 Name TIOB XC0 XC1 XC2 Description TIOB(1) XC0 XC1 XC2 TIOB Direction input output output output
Note:
1. If TIOB is chosen as the external event signal, it is configured as an input and no longer generates waveforms and subsequently no IRQs.
ENETRG: External Event Trigger Enable

0 = The external event has no effect on the counter and its clock. In this case, the selected external event only controls the TIOA output. 1 = The external event resets the counter and starts the counter clock.
WAVSEL: Waveform Selection

Value 0 1 2 3 Name UP UPDOWN UP_RC UPDOWN_RC Description UP mode without automatic trigger on RC Compare UPDOWN mode without automatic trigger on RC Compare UP mode with automatic trigger on RC Compare UPDOWN mode with automatic trigger on RC Compare
WAVE: Waveform Mode

0 = Waveform Mode is disabled (Capture Mode is enabled). 1 = Waveform Mode is enabled.
ACPA: RA Compare Effect on TIOA

Value 0 1 2 3 Name NONE SET CLEAR TOGGLE Description None Set Clear Toggle

1429
ACPC: RC Compare Effect on TIOA

AEEVT: External Event Effect on TIOA

ASWTRG: Software Trigger Effect on TIOA

BCPB: RB Compare Effect on TIOB

BCPC: RC Compare Effect on TIOB


1430
BEEVT: External Event Effect on TIOB

BSWTRG: Software Trigger Effect on TIOB


1431
48.7.4 TC Stepper Motor Mode Register

31 23 15 7
TC_SMMRx [x=0..2] 0xF0010008 (0)[0], 0xF0010048 (0)[1], 0xF0010088 (0)[2], 0xF8014008 (1)[0], 0xF8014048 (1)[1], 0xF8014088 (1)[2] Read-write
30 22 14 6 29 21 13 5 28 20 12 4 27 19 11 3 26 18 10 2 9 1 DOWN 25 17 8 0 GCEN 24 16
GCEN: Gray Count Enable

0 = TIOAx [x=0..2] and TIOBx [x=0..2] are driven by internal counter of channel x. 1 = TIOAx [x=0..2] and TIOBx [x=0..2] are driven by a 2-bit gray counter.
DOWN: DOWN Count

0 = Up counter. 1 = Down counter.

1432
48.7.5 TC Register AB
31
TC_RABx [x=0..2] 0xF001000C (0)[0], 0xF001004C (0)[1], 0xF001008C (0)[2], 0xF801400C (1)[0], 0xF801404C (1)[1], 0xF801408C (1)[2] Read-only
30 29 28 RAB 23 22 21 20 RAB 15 14 13 12 RAB 7 6 5 4 RAB 3 2 1 0 11 10 9 8 19 18 17 16 27 26 25 24
RAB: Register A or Register B

RAB contains the next unread capture Register A or Register B value in real time. It is usually read by the DMA after a request due to a valid load edge on TIOA. When DMA is used, the RAB register address must be configured as source address of the transfer.

1433
48.7.6 TC Counter Value Register

31
TC_CVx [x=0..2] 0xF0010010 (0)[0], 0xF0010050 (0)[1], 0xF0010090 (0)[2], 0xF8014010 (1)[0], 0xF8014050 (1)[1], 0xF8014090 (1)[2] Read-only
30 29 28 CV 23 22 21 20 CV 15 14 13 12 CV 7 6 5 4 CV 3 2 1 0 11 10 9 8 19 18 17 16 27 26 25 24
CV: Counter Value

CV contains the counter value in real time.

1434
48.7.7 TC Register A
31
TC_RAx [x=0..2] 0xF0010014 (0)[0], 0xF0010054 (0)[1], 0xF0010094 (0)[2], 0xF8014014 (1)[0], 0xF8014054 (1)[1], 0xF8014094 (1)[2] Read-only if WAVE = 0, Read-write if WAVE = 1
30 29 28 RA 23 22 21 20 RA 15 14 13 12 RA 7 6 5 4 RA 3 2 1 0 11 10 9 8 19 18 17 16 27 26 25 24
RA: Register A
RA contains the Register A value in real time.

1435
48.7.8 TC Register B
31
TC_RBx [x=0..2] 0xF0010018 (0)[0], 0xF0010058 (0)[1], 0xF0010098 (0)[2], 0xF8014018 (1)[0], 0xF8014058 (1)[1], 0xF8014098 (1)[2] Read-only if WAVE = 0, Read-write if WAVE = 1
30 29 28 RB 23 22 21 20 RB 15 14 13 12 RB 7 6 5 4 RB 3 2 1 0 11 10 9 8 19 18 17 16 27 26 25 24
RB: Register B
RB contains the Register B value in real time.

1436
48.7.9 TC Register C
31
TC_RCx [x=0..2] 0xF001001C (0)[0], 0xF001005C (0)[1], 0xF001009C (0)[2], 0xF801401C (1)[0], 0xF801405C (1)[1], 0xF801409C (1)[2] Read-write
30 29 28 RC 23 22 21 20 RC 15 14 13 12 RC 7 6 5 4 RC 3 2 1 0 11 10 9 8 19 18 17 16 27 26 25 24
RC: Register C
RC contains the Register C value in real time.

1437
48.7.10 TC Status Register

31 23 15 7 ETRGS
TC_SRx [x=0..2] 0xF0010020 (0)[0], 0xF0010060 (0)[1], 0xF00100A0 (0)[2], 0xF8014020 (1)[0], 0xF8014060 (1)[1], 0xF80140A0 (1)[2] Read-only
30 22 14 6 LDRBS 29 21 13 5 LDRAS 28 20 12 4 CPCS 27 19 11 3 CPBS 26 18 MTIOB 10 2 CPAS 25 17 MTIOA 9 1 LOVRS 24 16 CLKSTA 8 0 COVFS
COVFS: Counter Overflow Status

0 = No counter overflow has occurred since the last read of the Status Register. 1 = A counter overflow has occurred since the last read of the Status Register.
LOVRS: Load Overrun Status

0 = Load overrun has not occurred since the last read of the Status Register or WAVE = 1. 1 = RA or RB have been loaded at least twice without any read of the corresponding register since the last read of the Status Register, if WAVE = 0.
CPAS: RA Compare Status

0 = RA Compare has not occurred since the last read of the Status Register or WAVE = 0. 1 = RA Compare has occurred since the last read of the Status Register, if WAVE = 1.
CPBS: RB Compare Status

0 = RB Compare has not occurred since the last read of the Status Register or WAVE = 0. 1 = RB Compare has occurred since the last read of the Status Register, if WAVE = 1.
CPCS: RC Compare Status

0 = RC Compare has not occurred since the last read of the Status Register. 1 = RC Compare has occurred since the last read of the Status Register.
LDRAS: RA Loading Status

0 = RA Load has not occurred since the last read of the Status Register or WAVE = 1. 1 = RA Load has occurred since the last read of the Status Register, if WAVE = 0.
LDRBS: RB Loading Status

0 = RB Load has not occurred since the last read of the Status Register or WAVE = 1. 1 = RB Load has occurred since the last read of the Status Register, if WAVE = 0.
ETRGS: External Trigger Status

0 = External trigger has not occurred since the last read of the Status Register. 1 = External trigger has occurred since the last read of the Status Register.

1438
CLKSTA: Clock Enabling Status

0 = Clock is disabled. 1 = Clock is enabled.
MTIOA: TIOA Mirror

0 = TIOA is low. If WAVE = 0, this means that TIOA pin is low. If WAVE = 1, this means that TIOA is driven low. 1 = TIOA is high. If WAVE = 0, this means that TIOA pin is high. If WAVE = 1, this means that TIOA is driven high.
MTIOB: TIOB Mirror

0 = TIOB is low. If WAVE = 0, this means that TIOB pin is low. If WAVE = 1, this means that TIOB is driven low. 1 = TIOB is high. If WAVE = 0, this means that TIOB pin is high. If WAVE = 1, this means that TIOB is driven high.

1439
48.7.11 TC Interrupt Enable Register

31 23 15 7 ETRGS
TC_IERx [x=0..2] 0xF0010024 (0)[0], 0xF0010064 (0)[1], 0xF00100A4 (0)[2], 0xF8014024 (1)[0], 0xF8014064 (1)[1], 0xF80140A4 (1)[2] Write-only
30 22 14 6 LDRBS 29 21 13 5 LDRAS 28 20 12 4 CPCS 27 19 11 3 CPBS 26 18 10 2 CPAS 25 17 9 1 LOVRS 24 16 8 0 COVFS
COVFS: Counter Overflow

0 = No effect. 1 = Enables the Counter Overflow Interrupt.
LOVRS: Load Overrun

0 = No effect. 1 = Enables the Load Overrun Interrupt.
CPAS: RA Compare
0 = No effect. 1 = Enables the RA Compare Interrupt.
CPBS: RB Compare
0 = No effect. 1 = Enables the RB Compare Interrupt.
CPCS: RC Compare
0 = No effect. 1 = Enables the RC Compare Interrupt.
LDRAS: RA Loading
0 = No effect. 1 = Enables the RA Load Interrupt.
LDRBS: RB Loading
0 = No effect. 1 = Enables the RB Load Interrupt.
ETRGS: External Trigger

0 = No effect. 1 = Enables the External Trigger Interrupt.

1440
48.7.12 TC Interrupt Disable Register

31 23 15 7 ETRGS
TC_IDRx [x=0..2] 0xF0010028 (0)[0], 0xF0010068 (0)[1], 0xF00100A8 (0)[2], 0xF8014028 (1)[0], 0xF8014068 (1)[1], 0xF80140A8 (1)[2] Write-only

0 = No effect. 1 = Disables the Counter Overflow Interrupt.
LOVRS: Load Overrun

0 = No effect. 1 = Disables the Load Overrun Interrupt (if WAVE = 0).
CPAS: RA Compare
0 = No effect. 1 = Disables the RA Compare Interrupt (if WAVE = 1).
CPBS: RB Compare
0 = No effect. 1 = Disables the RB Compare Interrupt (if WAVE = 1).
CPCS: RC Compare
0 = No effect. 1 = Disables the RC Compare Interrupt.
LDRAS: RA Loading
0 = No effect. 1 = Disables the RA Load Interrupt (if WAVE = 0).
LDRBS: RB Loading
0 = No effect. 1 = Disables the RB Load Interrupt (if WAVE = 0).

0 = No effect. 1 = Disables the External Trigger Interrupt.

1441
48.7.13 TC Interrupt Mask Register

Name: Address: TC_IMRx [x=0..2] 0xF001002C (0)[0], 0xF001006C (0)[1], 0xF00100AC (0)[2], 0xF801402C (1)[0], 0xF801406C (1)[1], 0xF80140AC (1)[2] Access:
31 23 15 7 ETRGS
Read-only

0 = The Counter Overflow Interrupt is disabled. 1 = The Counter Overflow Interrupt is enabled.
LOVRS: Load Overrun

0 = The Load Overrun Interrupt is disabled. 1 = The Load Overrun Interrupt is enabled.
CPAS: RA Compare
0 = The RA Compare Interrupt is disabled. 1 = The RA Compare Interrupt is enabled.
CPBS: RB Compare
0 = The RB Compare Interrupt is disabled. 1 = The RB Compare Interrupt is enabled.
CPCS: RC Compare
0 = The RC Compare Interrupt is disabled. 1 = The RC Compare Interrupt is enabled.
LDRAS: RA Loading
0 = The Load RA Interrupt is disabled. 1 = The Load RA Interrupt is enabled.
LDRBS: RB Loading
0 = The Load RB Interrupt is disabled. 1 = The Load RB Interrupt is enabled.

0 = The External Trigger Interrupt is disabled. 1 = The External Trigger Interrupt is enabled.

1442
48.7.14 TC Block Control Register

31 23 15 7
TC_BCR 0xF00100C0 (0), 0xF80140C0 (1) Write-only

30 22 14 6 29 21 13 5 28 20 12 4 27 19 11 3 26 18 10 2 25 17 9 1 24 16 8 0 SYNC
SYNC: Synchro Command

0 = No effect. 1 = Asserts the SYNC signal which generates a software trigger simultaneously for each of the channels.

1443
48.7.15 TC Block Mode Register

31 23
TC_BMR 0xF00100C4 (0), 0xF80140C4 (1) Read-write

30 22 MAXFILT 15 14 INVB 6 13 INVA 5 TC2XC2S 12 EDGPHA 4 29 21 28 20 27 19 FILTER 11 QDTRANS 3 TC1XC1S 26 18 10 SPEEDEN 2 17 IDXPHB 9 POSEN 1 TC0XC0S 25 MAXFILT 16 SWAP 8 QDEN 0 24
INVIDX 7
This register can only be written if the WPEN bit is cleared in TC Write Protect Mode Register on page 1450.
TC0XC0S: External Clock Signal 0 Selection

Value 0 1 2 3 Name TCLK0 TIOA1 TIOA2 Description Signal connected to XC0: TCLK0 Reserved Signal connected to XC0: TIOA1 Signal connected to XC0: TIOA2


QDEN: Quadrature Decoder ENabled

0 = Disabled 1 = Enables the quadrature decoder logic (filter, edge detection and quadrature decoding). Quadrature decoding (direction change) can be disabled using QDTRANS bit. One of the POSEN or SPEEDEN bits must be also enabled.

1444
POSEN: POSition ENabled

0 = Disable position 1 = Enables the position measure on channel 0 and 1.
SPEEDEN: SPEED ENabled

0 = Disabled 1 = Enables the speed measure on channel 0, the time base being provided by channel 2.
QDTRANS: Quadrature Decoding TRANSparent

0 = Full quadrature decoding logic is active (direction change detected). 1 = Quadrature decoding logic is inactive (direction change inactive) but input filtering and edge detection are performed.
EDGPHA: EDGe on PHA count mode

0 = Edges are detected on both PHA and PHB. 1 = Edges are detected on PHA only.
INVA: INVerted phA

0 = PHA (TIOA0) is directly driving quadrature decoder logic. 1 = PHA is inverted before driving quadrature decoder logic.
INVB: INVerted phB

0 = PHB (TIOB0) is directly driving quadrature decoder logic. 1 = PHB is inverted before driving quadrature decoder logic.
SWAP: SWAP PHA and PHB

0 = No swap between PHA and PHB. 1 = Swap PHA and PHB internally, prior to driving quadrature decoder logic.
INVIDX: INVerted InDeX

0 = IDX (TIOA1) is directly driving quadrature logic. 1 = IDX is inverted before driving quadrature logic.
IDXPHB: InDeX pin is PHB pin

0 = IDX pin of the rotary sensor must drive TIOA1. 1 = IDX pin of the rotary sensor must drive TIOB0.
FILTER:
0 = IDX,PHA, PHB input pins are not filtered. 1 = IDX,PHA, PHB input pins are filtered using MAXFILT value.
MAXFILT: MAXimum FILTer

1.. 63: defines the filtering capabilities. Pulses with a period shorter than MAXFILT+1 MCK clock cycles are discarded.

1445
48.7.16 TC QDEC Interrupt Enable Register

31 23 15 7
TC_QIER 0xF00100C8 (0), 0xF80140C8 (1) Write-only

30 22 14 6 29 21 13 5 28 20 12 4 27 19 11 3 26 18 10 2 QERR 25 17 9 1 DIRCHG 24 16 8 0 IDX
IDX: InDeX
0 = No effect. 1 = Enables the interrupt when a rising edge occurs on IDX input.
DIRCHG: DIRection CHanGe

0 = No effect. 1 = Enables the interrupt when a change on rotation direction is detected.
QERR: Quadrature ERRor

0 = No effect. 1 = Enables the interrupt when a quadrature error occurs on PHA,PHB.

1446
48.7.17 TC QDEC Interrupt Disable Register

31 23 15 7
TC_QIDR 0xF00100CC (0), 0xF80140CC (1) Write-only

30 22 14 6 29 21 13 5 28 20 12 4 27 19 11 3 26 18 10 2 QERR 25 17 9 1 DIRCHG 24 16 8 0 IDX
IDX: InDeX
0 = No effect. 1 = Disables the interrupt when a rising edge occurs on IDX input.

0 = No effect. 1 = Disables the interrupt when a change on rotation direction is detected.

0 = No effect. 1 = Disables the interrupt when a quadrature error occurs on PHA, PHB.

1447
48.7.18 TC QDEC Interrupt Mask Register

31 23 15 7
TC_QIMR 0xF00100D0 (0), 0xF80140D0 (1) Read-only

30 22 14 6 29 21 13 5 28 20 12 4 27 19 11 3 26 18 10 2 QERR 25 17 9 1 DIRCHG 24 16 8 0 IDX
IDX: InDeX
0 = The interrupt on IDX input is disabled. 1 = The interrupt on IDX input is enabled.

0 = The interrupt on rotation direction change is disabled. 1 = The interrupt on rotation direction change is enabled.

0 = The interrupt on quadrature error is disabled. 1 = The interrupt on quadrature error is enabled.

1448
48.7.19 TC QDEC Interrupt Status Register

31 23 15 7
TC_QISR 0xF00100D4 (0), 0xF80140D4 (1) Read-only

30 22 14 6 29 21 13 5 28 20 12 4 27 19 11 3 26 18 10 2 QERR 25 17 9 1 DIRCHG 24 16 8 DIR 0 IDX
IDX: InDeX
0 = No Index input change since the last read of TC_QISR. 1 = The IDX input has changed since the last read of TC_QISR.

0 = No change on rotation direction since the last read of TC_QISR. 1 = The rotation direction changed since the last read of TC_QISR.

0 = No quadrature error since the last read of TC_QISR. 1 = A quadrature error occurred since the last read of TC_QISR.
DIR: DIRection
Returns an image of the actual rotation direction.

1449
48.7.20 TC Write Protect Mode Register

31
TC_WPMR 0xF00100E4 (0), 0xF80140E4 (1) Read-write

30 29 28 WPKEY 27 26 25 24
23
22
21
20 WPKEY
19
18
17
16
15
14
13
12 WPKEY
11
10
0 WPEN

0 = Disables the Write Protect if WPKEY corresponds to 0x54494D (TIM in ASCII). 1 = Enables the Write Protect if WPKEY corresponds to 0x54494D (TIM in ASCII). Protects the registers: TC Block Mode Register TC Channel Mode Register: Capture Mode TC Channel Mode Register: Waveform Mode TC Stepper Motor Mode Register TC Register A TC Register B TC Register C
WPKEY: Write Protect KEY

Should be written at value 0x54494D (TIM in ASCII). Writing any other value in this field aborts the write operation of the WPEN bit. Always reads as 0.

1450
49.
49.1
Pulse Width Modulation Controller (PWM)

Description
The PWM macrocell controls four channels independently. Each channel controls two complementary square output waveforms. Characteristics of the output waveforms such as period, duty-cycle, polarity and dead-times (also called dead-bands or non-overlapping times) are configured through the user interface. Each channel selects and uses one of the clocks provided by the clock generator. The clock generator provides several clocks resulting from the division of the PWM master clock (MCK). All PWM macrocell accesses are made through registers mapped on the peripheral bus. All channels integrate a double buffering system in order to prevent an unexpected output waveform while modifying the period, the duty-cycle or the dead-times. Channels can be linked together as synchronous channels to be able to update their duty-cycle or dead-times at the same time. The PWM macrocell provides 8 independent comparison units capable of comparing a programmed value to the counter of the synchronous channels (counter of channel 0). These comparisons are intended to generate software interrupts, to trigger pulses on the 2 independent event lines (in order to synchronize ADC conversions with a lot of flexibility independently of the PWM outputs). The PWM outputs can be overridden synchronously or asynchronously to their channel counter. The PWM block provides a fault protection mechanism with 1 fault inputs, capable to detect a fault condition and to override the PWM outputs asynchronously (outputs forced to 0, 1). For safety usage, some configuration registers are write-protected.

1451
49.2
z z
4 Channels Common Clock Generator Providing Thirteen Different Clocks

z z
A Modulo n Counter Providing Eleven Clocks Two Independent Linear Dividers Working on Modulo n Counter Outputs Independent 16-bit Counter for Each Channel Independent Complementary Outputs with 12-bit Dead-Time Generator (Also Called Dead-Band or NonOverlapping Time) for Each Channel Independent Enable Disable Command for Each Channel Independent Clock Selection for Each Channel Independent Period, Duty-Cycle and Dead-Time for Each Channel Independent Double Buffering of Period, Duty-Cycle and Dead-Times for Each Channel Independent Programmable Selection of The Output Waveform Polarity for Each Channel Independent Programmable Center or Left Aligned Output Waveform for Each Channel Independent Output Override for Each Channel Independent Interrupt for Each Channel, at Each Period for Left-Aligned or Center-Aligned Configuration
Independent Channels
z z z z z z z z z z
z z
Two 2-bit Gray Up/Down Channels for Stepper Motor Control Synchronous Channel Mode
z z
Synchronous Channels Share the Same Counter Mode to Update the Synchronous Channels Registers after a Programmable Number of Periods Programmable delay for Events Lines to delay ADC measurements
2 Independent Events Lines Intended to Synchronize ADC Conversions

z
z z
8 Comparison Units Intended to Generate Interrupts, Pulses on Event Lines 1 Programmable Fault/Break Inputs Providing an Asynchronous Protection of PWM Outputs
z z z
4 User Driven through PIO inputs PMC Driven when Crystal Oscillator Clock Fails ADC Controller Driven through Configurable Comparison Function
Write Protected Registers

1452
49.3
Block Diagram
Figure 49-1. Pulse Width Modulation Controller Block Diagram

PWM Controller Channel x Update
Period Comparator OCx SYNCx Duty-Cycle MUX Clock Selector Counter Channel x
DTOHx Dead-Time Generator DTOLx
OOOHx Output Override OOOLx
PWMHx Fault Protection PWMLx
PWMHx PWMLx
PIO
Channel 0 Update
Period Comparator Duty-Cycle OC0
DTOH0 Dead-Time Generator DTOL0
OOOH0 Output Override OOOL0
PWMH0 Fault Protection PWML0
PWMH0 PWML0
Clock Selector PWMFIx PIO PWMFI0
Counter Channel 0
event line 0 event line 1 Comparison Units PMC MCK CLOCK Generator Events Generator event line x APB Interface Interrupt Generator Interrupt Controller ADC
APB
49.4

Each channel outputs two complementary external I/O lines.
Table 49-1. I/O Line Description Name PWMHx PWMLx PWMFIx Description PWM Waveform Output High for channel x PWM Waveform Output Low for channel x PWM Fault Input x Type Output Output Input

1453
49.5
49.5.1 I/O Lines

The pins used for interfacing the PWM are multiplexed with PIO lines. The programmer must first program the PIO controller to assign the desired PWM pins to their peripheral function. If I/O lines of the PWM are not used by the application, they can be used for other purposes by the PIO controller. All of the PWM outputs may or may not be enabled. If an application requires only four channels, then only four PIO lines will be assigned to PWM outputs.
Table 49-2. I/O Lines Instance PWM PWM PWM PWM PWM PWM PWM PWM PWM PWM PWM PWM PWM PWM PWM PWM PWM PWM PWM PWM PWM PWM Signal PWMFI0 PWMFI1 PWMFI2 PWMFI3 PWMH0 PWMH0 PWMH1 PWMH1 PWMH1 PWMH2 PWMH2 PWMH3 PWMH3 PWML0 PWML0 PWML1 PWML1 PWML1 PWML2 PWML2 PWML3 PWML3 I/O Line PC28 PC31 PC29 PD16 PA20 PB0 PA22 PB4 PB27 PB8 PD5 PB12 PD7 PA21 PB1 PA23 PB5 PE31 PB9 PD6 PB13 PD8 Peripheral B B B C B B B B C B C B C B B B B B B C B C

1454

The PWM is not continuously clocked. The programmer must first enable the PWM clock in the Power Management Controller (PMC) before using the PWM. However, if the application does not require PWM operations, the PWM clock can be stopped when not needed and be restarted later. In this case, the PWM will resume its operations where it left off. In the PWM description, Master Clock (MCK) is the clock of the peripheral bus to which the PWM is connected.

The PWM interrupt line is connected on one of the internal sources of the Interrupt Controller. Using the PWM interrupt requires the Interrupt Controller to be programmed first. Note that it is not recommended to use the PWM interrupt line in edge sensitive mode.
Table 49-3. Peripheral IDs Instance PWM ID 28
49.5.4 Fault Inputs

The PWM has the FAULT inputs connected to the different modules. Please refer to the implementation of these module within the product for detailed information about the fault generation procedure. The PWM receives faults from PIO inputs, PMC, ADC controller.
Table 49-4. Fault Inputs Fault Inputs PC28 PC31 PC29 PD16 Main OSC (PMC) ADC External PWM Fault Input Number PWMFI0 PWMFI1 PWMFI2 PWMFI3 Polarity Level(1) User Defined User Defined User Defined User Defined 1 1 Fault Input ID 0 1 2 3 4 5
Note:
1. FPOL bit in PWMC_FMR.

1455
49.6
The PWM macrocell is primarily composed of a clock generator module and 4 channels.
z z z
Clocked by the master clock (MCK), the clock generator module provides 13 clocks. Each channel can independently choose one of the clock generator outputs. Each channel generates an output waveform with attributes that can be defined independently for each channel through the user interface registers.
49.6.1 PWM Clock Generator

Figure 49-2. Functional View of the Clock Generator Block Diagram
MCK modulo n counter MCK MCK/2 MCK/4 MCK/8 MCK/16 MCK/32 MCK/64 MCK/128 MCK/256 MCK/512 MCK/1024
Divider A
clkA
PREA
DIVA
PWM_MR
Divider B
clkB
PREB
DIVB
PWM_MR
The PWM master clock (MCK) is divided in the clock generator module to provide different clocks available for all channels. Each channel can independently select one of the divided clocks. The clock generator is divided in three blocks:
z z
A modulo n counter which provides 11 clocks: FMCK, FMCK/2, FMCK/4, FMCK/8, FMCK/16, FMCK/32, FMCK/64, FMCK/128, FMCK/256, FMCK/512, FMCK/1024 Two linear dividers (1, 1/2, 1/3, ... 1/255) that provide two separate clocks: clkA and clkB
Each linear divider can independently divide one of the clocks of the modulo n counter. The selection of the clock to be divided is made according to the PREA (PREB) field of the PWM Clock register (PWM_CLK). The resulting clock clkA (clkB) is the clock selected divided by DIVA (DIVB) field value. After a reset of the PWM controller, DIVA (DIVB) and PREA (PREB) are set to 0. This implies that after reset clkA (clkB) are turned off.

1456
At reset, all clocks provided by the modulo n counter are turned off except clock MCK. This situation is also true when the PWM master clock is turned off through the Power Management Controller. CAUTION: Before using the PWM macrocell, the programmer must first enable the PWM clock in the Power Management Controller (PMC).
49.6.2 PWM Channel

49.6.2.1 Channel Block Diagram
Figure 49-3. Functional View of the Channel Block Diagram
Update
Channel x
Period Comparator x Duty-Cycle MUX From Clock Generator Clock Selector Counter Channel x SYNCx OCx DTOHx Dead-Time Generator DTOLx PWMHx OOOHx Output Fault OOOLx PWMLx Override Protection MUX
from APB Peripheral Bus Counter Channel 0 OCy MUX
Channel y (= x+1)
Comparator y PWMHy OOOHy DTOHy Dead-Time Output Fault Generator DTOLy Override OOOLy Protection PWMLy
z = 0 (x = 0, y = 1), z = 1 (x = 2, y = 3), z = 2 (x = 4, y = 5), z = 3 (x = 6, y = 7)
2-bit gray counter z
Each of the 4 channels is composed of six blocks:

z z z
A clock selector which selects one of the clocks provided by the clock generator (described in Section 49.6.1 on page 1456). A counter clocked by the output of the clock selector. This counter is incremented or decremented according to the channel configuration and comparators matches. The size of the counter is 16 bits. A comparator used to compute the OCx output waveform according to the counter value and the configuration. The counter value can be the one of the channel counter or the one of the channel 0 counter according to SYNCx bit in the PWM Sync Channels Mode Register on page 1486 (PWM_SCM). A 2-bit configurable gray counter enables the stepper motor driver. One gray counter drives 2 channels. A dead-time generator providing two complementary outputs (DTOHx/DTOLx) which allows to drive external power control switches safely. An output override block that can force the two complementary outputs to a programmed value (OOOHx/OOOLx). An asynchronous fault protection mechanism that has the highest priority to override the two complementary outputs (PWMHx/PWMLx) in case of fault detection (outputs forced to 0, 1).
z z z z

1457
49.6.2.2 Comparator The comparator continuously compares its counter value with the channel period defined by CPRD in the PWM Channel Period Register on page 1518 (PWM_CPRDx) and the duty-cycle defined by CDTY in the PWM Channel Duty Cycle Register on page 1516 (PWM_CDTYx) to generate an output signal OCx accordingly. The different properties of the waveform of the output OCx are:
z
The clock selection. The channel counter is clocked by one of the clocks provided by the clock generator described in the previous section. This channel parameter is defined in the CPRE field of the PWM Channel Mode Register on page 1514 (PWM_CMRx). This field is reset at 0. The waveform period. This channel parameter is defined in the CPRD field of the PWM_CPRDx register. If the waveform is left aligned, then the output waveform period depends on the counter source clock and can be calculated: By using the PWM master clock (MCK) divided by an X given prescaler value (with X being 1, 2, 4, 8, 16, 32, 64, 128, 256, 512, or 1024), the resulting period formula will be:
( X CPRD ) ------------------------------MCK
By using the PWM master clock (MCK) divided by one of both DIVA or DIVB divider, the formula becomes, respectively:
( CRPD DIVA ) CRPD DIVB ) ------------------------------------------ or ( -----------------------------------------MCK MCK
If the waveform is center aligned then the output waveform period depends on the counter source clock and can be calculated: By using the PWM master clock (MCK) divided by an X given prescaler value (with X being 1, 2, 4, 8, 16, 32, 64, 128, 256, 512, or 1024). The resulting period formula will be:
( 2 X CPRD ) -----------------------------------------MCK
( 2 CPRD DIVA ) 2 CPRD DIVB ) ----------------------------------------------------- or ( ----------------------------------------------------MCK MCK
z
The waveform duty-cycle. This channel parameter is defined in the CDTY field of the PWM_CDTYx register. If the waveform is left aligned then:
( period 1 fchannel_x_clock CDTY ) duty cycle = ----------------------------------------------------------------------------------------------------------period

If the waveform is center aligned, then:
( period 2 ) 1 fchannel_x_clock CDTY ) ) duty cycle = ( ----------------------------------------------------------------------------------------------------------------------------( period 2 )

z z
The waveform polarity. At the beginning of the period, the signal can be at high or low level. This property is defined in the CPOL field of the PWM_CMRx register. By default the signal starts by a low level. The waveform alignment. The output waveform can be left or center aligned. Center aligned waveforms can be used to generate non overlapped waveforms. This property is defined in the CALG field of the PWM_CMRx register. The default mode is left aligned.

1458
Figure 49-4. Non Overlapped Center Aligned Waveforms

No overlap
OC0
OC1
Period
Note:
1.
See Figure 49-5 on page 1460 for a detailed description of center aligned waveforms.
When center aligned, the channel counter increases up to CPRD and decreases down to 0. This ends the period. When left aligned, the channel counter increases up to CPRD and is reset. This ends the period. Thus, for the same CPRD value, the period for a center aligned channel is twice the period for a left aligned channel. Waveforms are fixed at 0 when:
z z CDTY = CPRD and CPOL = 0 CDTY = 0 and CPOL = 1
Waveforms are fixed at 1 (once the channel is enabled) when:

z z CDTY = 0 and CPOL = 0 CDTY = CPRD and CPOL = 1
The waveform polarity must be set before enabling the channel. This immediately affects the channel output level. Modifying CPOL in PWM Channel Mode Register on page 1514 while the channel is enabled can lead to an unexpected behavior of the device being driven by PWM. Besides generating output signals OCx, the comparator generates interrupts in function of the counter value. When the output waveform is left aligned, the interrupt occurs at the end of the counter period. When the output waveform is center aligned, the bit CES of the PWM_CMRx register defines when the channel counter interrupt occurs. If CES is set to 0, the interrupt occurs at the end of the counter period. If CES is set to 1, the interrupt occurs at the end of the counter period and at half of the counter period. Figure 49-5 Waveform Properties illustrates the counter interrupts in function of the configuration.

1459
Figure 49-5. Waveform Properties

Channel x slected clock CHIDx(PWM_SR)
CHIDx(PWM_ENA) CHIDx(PWM_DIS) Center Aligned CALG(PWM_CMRx) = 1
PWM_CCNTx CPRD(PWM_CPRDx) CDTY(PWM_CDTYx)
Period Output Waveform OCx CPOL(PWM_CMRx) = 0
Output Waveform OCx CPOL(PWM_CMRx) = 1 Counter Event CHIDx(PWM_ISR) CES(PWM_CMRx) = 0 Counter Event CHIDx(PWM_ISR) CES(PWM_CMRx) = 1 Left Aligned CALG(PWM_CMRx) = 0
PWM_CCNTx CPRD(PWM_CPRDx) CDTY(PWM_CDTYx)
Period Output Waveform OCx CPOL(PWM_CMRx) = 0
Output Waveform OCx CPOL(PWM_CMRx) = 1
Counter Event CHIDx(PWM_ISR)

1460
49.6.2.3 2-bit Gray Up/Down Counter for Stepper Motor It is possible to configure a couple of channels to provide a 2-bit gray count waveform on 2 outputs. Dead-Time Generator and other downstream logic can be configured on these channels. Up or down count mode can be configured on-the-fly by means of PWM_SMMR configuration registers. When GCEN0 is set to 1, channels 0 and 1 outputs are driven with gray counter.
Figure 49-6. 2-bit Gray Up/Down Counter
GCEN0 = 1 PWMH0
PWML0
PWMH1
PWML1
DOWNx

1461
49.6.2.4 Dead-time Generator The dead-time generator uses the comparator output OCx to provide the two complementary outputs DTOHx and DTOLx, which allows the PWM macrocell to drive external power control switches safely. When the dead-time generator is enabled by setting the bit DTE to 1 or 0 in the PWM Channel Mode Register (PWM_CMRx), dead-times (also called dead-bands or non-overlapping times) are inserted between the edges of the two complementary outputs DTOHx and DTOLx. Note that enabling or disabling the dead-time generator is allowed only if the channel is disabled. The dead-time is adjustable by the PWM Channel Dead Time Register (PWM_DTx). Both outputs of the dead-time generator can be adjusted separately by DTH and DTL. The dead-time values can be updated synchronously to the PWM period by using the PWM Channel Dead Time Update Register (PWM_DTUPDx). The dead-time is based on a specific counter which uses the same selected clock that feeds the channel counter of the comparator. Depending on the edge and the configuration of the dead-time, DTOHx and DTOLx are delayed until the counter has reached the value defined by DTH or DTL. An inverted configuration bit (DTHI and DTLI bit in the PWM_CMRx register) is provided for each output to invert the dead-time outputs. The following figure shows the waveform of the dead-time generator.
Figure 49-7. Complementary Output Waveforms
Output waveform OCx CPOLx = 0 Output waveform DTOHx DTHIx = 0 Output waveform DTOLx DTLIx = 0 Output waveform DTOHx DTHIx = 1 Output waveform DTOLx DTLIx = 1 DTHx DTLx
Output waveform OCx CPOLx = 1 Output waveform DTOHx DTHIx = 0 Output waveform DTOLx DTLIx = 0 Output waveform DTOHx DTHIx = 1 Output waveform DTOLx DTLIx = 1 DTHx DTLx

1462
49.6.2.5 Output Override The two complementary outputs DTOHx and DTOLx of the dead-time generator can be forced to a value defined by the software.
Figure 49-8. Override Output Selection
DTOHx 0 OOOHx OOVHx 1
OSHx DTOLx 0 OOOLx OOVLx 1
OSLx
The OSHx and OSLx fields in the PWM Output Selection Register (PWM_OS) allow the DTOHx and DTOLx outputs of the dead-time generator to be overridden by the value defined in the OOVHx and OOVLx fields in thePWM Output Override Value Register (PWM_OOV). The PWM Output Selection Set Register and PWM Output Selection Set Update Register set registers (PWM_OSS and PWM_OSSUPD) enable the override of the outputs of a channel regardless of other channels. In the same way, the PWM Output Selection Clear Register and PWM Output Selection Clear Update Register clear registers (PWM_OSC and PWM_OSCUPD) disable the override of the outputs of a channel regardless of other channels. By using the PWM_OSSUPD and PWM_OSCUPD buffer registers, the output selection of PWM outputs is done synchronously to the channel counter, at the beginning of the next PWM period. By using the PWM_OSS and PWM_OSC registers, the output selection of PWM outputs is done asynchronously to the channel counter, as soon as the register is written. The value of the current output selection can be read in PWM_OS. While overriding PWM outputs, the channel counters continue to run, only the PWM outputs are forced to user defined values.

1463
49.6.2.6 Fault Protection 1 input provides fault protection which can force any of the PWM output pair to a programmable value. This mechanism has priority over output overriding.
Figure 49-9. Fault Protection
0 fault input 0 Glitch Filter 1 FIV0
0 FMOD0
SET CLR OUT
Fault 0 Status FS0 FPEx[0] FPE0[0] 0 1
from fault 0
FFIL0
FPOL0
Write FCLR0 at 1
FMOD0
0 fault input 1 Glitch Filter 1
FIV1
0 FMOD1
SET CLR OUT
Fault 1 Status FS1 FPEx[1] FPE0[1]
SYNCx from fault 1 0 1 from fault y SYNCx
From Output Override OOHx FPVHx
0 PWMHx 1 Fault protection on PWM channel x
FFIL1
FPOL1
Write FCLR1 at 1
FMOD1
fault input y FPVLx OOLx From Output Override 1 PWMLx 0
The polarity level of the fault inputs is configured by the FPOL field in the PWM Fault Mode Register (PWM_FMR). For fault inputs coming from internal peripherals such as ADC, Timer Counter, to name but a few, the polarity level must be FPOL = 1. For fault inputs coming from external GPIO pins the polarity level depends on the user's implementation. The configuration of the Fault Activation Mode (FMOD bit in PWMC_FMR) depends on the peripheral generating the fault. If the corresponding peripheral does not have Fault Clear management, then the FMOD configuration to use must be FMOD = 1, to avoid spurious fault detection. Check the corresponding peripheral documentation for details on handling fault generation. The fault inputs can be glitch filtered or not in function of the FFIL field in the PWM_FMR register. When the filter is activated, glitches on fault inputs with a width inferior to the PWM master clock (MCK) period are rejected. A fault becomes active as soon as its corresponding fault input has a transition to the programmed polarity level. If the corresponding bit FMOD is set to 0 in the PWM_FMR register, the fault remains active as long as the fault input is at this polarity level. If the corresponding FMOD bit is set to 1, the fault remains active until the fault input is not at this polarity level anymore and until it is cleared by writing the corresponding bit FCLR in the PWM Fault Clear Register (PWM_FSCR). By reading the PWM Fault Status Register (PWM_FSR), the user can read the current level of the fault inputs by means of the field FIV, and can know which fault is currently active thanks to the FS field. Each fault can be taken into account or not by the fault protection mechanism in each channel. To be taken into account in the channel x, the fault y must be enabled by the bit FPEx[y] in the PWM Fault Protection Enable Registers (PWM_FPE1). However the synchronous channels (see Section 49.6.2.7 Synchronous Channels) do not use their own fault enable bits, but those of the channel 0 (bits FPE0[y]). The fault protection on a channel is triggered when this channel is enabled and when any one of the faults that are enabled for this channel is active. It can be triggered even if the PWM master clock (MCK) is not running but only by a fault input that is not glitch filtered. When the fault protection is triggered on a channel, the fault protection mechanism resets the counter of this channel and forces the channel outputs to the values defined by the fields FPVHx and FPVLx in the PWM Fault Protection Value Register (PWM_FPV). The output forcing is made asynchronously to the channel counter. CAUTION:
z z
To prevent an unexpected activation of the status flag FSy in the PWM_FSR register, the FMODy bit can be set to 1 only if the FPOLy bit has been previously configured to its final value. To prevent an unexpected activation of the Fault Protection on the channel x, the bit FPEx[y] can be set to 1 only if the FPOLy bit has been previously configured to its final value.

1464
If a comparison unit is enabled (see Section 49.6.3 PWM Comparison Units) and if a fault is triggered in the channel 0, in this case the comparison cannot match. As soon as the fault protection is triggered on a channel, an interrupt (different from the interrupt generated at the end of the PWM period) can be generated but only if it is enabled and not masked. The interrupt is reset by reading the interrupt status register, even if the fault which has caused the trigger of the fault protection is kept active. 49.6.2.7 Synchronous Channels Some channels can be linked together as synchronous channels. They have the same source clock, the same period, the same alignment and are started together. In this way, their counters are synchronized together. The synchronous channels are defined by the SYNCx bits in the PWM Sync Channels Mode Register (PWM_SCM). Only one group of synchronous channels is allowed. When a channel is defined as a synchronous channel, the channel 0 is automatically defined as a synchronous channel too, because the channel 0 counter configuration is used by all the synchronous channels. If a channel x is defined as a synchronous channel, it uses the following configuration fields of the channel 0 instead of its own:
z z z
CPRE0 field in PWM_CMR0 register instead of CPREx field in PWM_CMRx register (same source clock) CPRD0 field in PWM_CMR0 register instead of CPRDx field in PWM_CMRx register (same period) CALG0 field in PWM_CMR0 register instead of CALGx field in PWM_CMRx register (same alignment)
Thus writing these fields of a synchronous channel has no effect on the output waveform of this channel (except channel 0 of course). Because counters of synchronous channels must start at the same time, they are all enabled together by enabling the channel 0 (by the CHID0 bit in PWM_ENA register). In the same way, they are all disabled together by disabling channel 0 (by the CHID0 bit in PWM_DIS register). However, a synchronous channel x different from channel 0 can be enabled or disabled independently from others (by the CHIDx bit in PWM_ENA and PWM_DIS registers). Defining a channel as a synchronous channel while it is an asynchronous channel (by writing the bit SYNCx to 1 while it was at 0) is allowed only if the channel is disabled at this time (CHIDx = 0 in PWM_SR register). In the same way, defining a channel as an asynchronous channel while it is a synchronous channel (by writing the SYNCx bit to 0 while it was 1) is allowed only if the channel is disabled at this time. The field UPDM (Update Mode) in the PWM_SCM register allow to select one of the three methods to update the registers of the synchronous channels:
z
Method 1 (UPDM = 0): the period value, the duty-cycle values and the dead-time values must be written by the CPU in their respective update registers (respectively PWM_CPRDUPDx, PWM_CDTYUPDx and PWM_DTUPDx).The update is triggered at the next PWM period as soon as the bit UPDULOCK in the PWM Sync Channels Update Control Register (PWM_SCUC) is set to 1 (see Method 1: Manual write of duty-cycle values and manual trigger of the update on page 1466). Method 2 (UPDM = 1): the period value, the duty-cycle values, the dead-time values and the update period value must be written by the CPU in their respective update registers (respectively PWM_CPRDUPDx, PWM_CDTYUPDx and PWM_DTUPD). The update of the period value and of the dead-time values is triggered at the next PWM period as soon as the bit UPDULOCK in the PWM Sync Channels Update Control Register (PWM_SCUC) is set to 1. The update of the duty-cycle values and the update period value is triggered automatically after an update period defined by the field UPR in the PWM Sync Channels Update Period Register (PWM_SCUP) (see Method 2: Manual write of duty-cycle values and automatic trigger of the update on page 1467).

1465
Table 49-5. Summary of the Update of Registers of Synchronous Channels UPDM=0 Write by the CPU Period Value (PWM_CPRDUPDx) Update is triggered at the next PWM period as soon as the bit UPDULOCK is set to 1 Write by the CPU Dead-Time Values (PWM_DTUPDx) Update is triggered at the next PWM period as soon as the bit UPDULOCK is set to 1 Write by the CPU Duty-Cycle Values (PWM_CDTYUPDx) Update is triggered at the next PWM period as soon as the bit UPDULOCK is set to 1 Not applicable Update Period Value (PWM_SCUPUPD) Not applicable Write by the CPU Update is triggered at the next PWM period as soon as the update period counter has reached the value UPR Write by the CPU Update is triggered at the next PWM period as soon as the update period counter has reached the value UPR UPDM=1
Method 1: Manual write of duty-cycle values and manual trigger of the update
In this mode, the update of the period value, the duty-cycle values and the dead-time values must be done by writing in their respective update registers with the CPU (respectively PWM_CPRDUPDx, PWM_CDTYUPDx and PWM_DTUPDx). To trigger the update, the user must use the bit UPDULOCK of the PWM Sync Channels Update Control Register (PWM_SCUC) which allows to update synchronously (at the same PWM period) the synchronous channels:
z z
If the bit UPDULOCK is set to 1, the update is done at the next PWM period of the synchronous channels. If the UPDULOCK bit is not set to 1, the update is locked and cannot be performed.
After writing the UPDULOCK bit to 1, it is held at this value until the update occurs, then it is read 0. Sequence for Method 1: 1. 2. 3. 4. 5. 6. Select the manual write of duty-cycle values and the manual update by setting the UPDM field to 0 in the PWM_SCM register Define the synchronous channels by the SYNCx bits in the PWM_SCM register. Enable the synchronous channels by writing CHID0 in the PWM_ENA register. If an update of the period value and/or the duty-cycle values and/or the dead-time values is required, write registers that need to be updated (PWM_CPRDUPDx, PWM_CDTYUPDx and PWM_DTUPDx). Set UPDULOCK to 1 in PWM_SCUC. The update of the registers will occur at the beginning of the next PWM period. At this moment the UPDULOCK bit is reset, go to Step 4.) for new values.

1466
Figure 49-10. Method 1 (UPDM = 0)

CCNT0
CDTYUPD
0x20
0x40
0x60
UPDULOCK
CDTY
0x20
0x40
0x60
Method 2: Manual write of duty-cycle values and automatic trigger of the update
In this mode, the update of the period value, the duty-cycle values, the dead-time values and the update period value must be done by writing in their respective update registers with the CPU (respectively PWM_CPRDUPDx, PWM_CDTYUPDx, PWM_DTUPDx and PWM_SCUPUPD). To trigger the update of the period value and the dead-time values, the user must use the bit UPDULOCK of the PWM Sync Channels Update Control Register (PWM_SCUC) which allows to update synchronously (at the same PWM period) the synchronous channels: z If the bit UPDULOCK is set to 1, the update is done at the next PWM period of the synchronous channels. z If the UPDULOCK bit is not set to 1, the update is locked and cannot be performed. After writing the UPDULOCK bit to 1, it is held at this value until the update occurs, then it is read 0. The update of the duty-cycle values and the update period is triggered automatically after an update period. To configure the automatic update, the user must define a value for the Update Period by the UPR field in the PWM Sync Channels Update Period Register (PWM_SCUP). The PWM controller waits UPR+1 period of synchronous channels before updating automatically the duty values and the update period value. The status of the duty-cycle value write is reported in the PWM Interrupt Status Register 2 (PWM_ISR2) by the following flags: z WRDY: this flag is set to 1 when the PWM Controller is ready to receive new duty-cycle values and a new update period value. It is reset to 0 when the PWM_ISR2 register is read. Depending on the interrupt mask in the PWM_IMR2 register, an interrupt can be generated by these flags. Sequence for Method 2: 1. Select the manual write of duty-cycle values and the automatic update by setting the field UPDM to 1 in the PWM_SCM register 2. Define the synchronous channels by the bits SYNCx in the PWM_SCM register. 3. Define the update period by the field UPR in the PWM_SCUP register. 4. Enable the synchronous channels by writing CHID0 in the PWM_ENA register. 5. If an update of the period value and/or of the dead-time values is required, write registers that need to be updated (PWM_CPRDUPDx, PWM_DTUPDx), else go to Step 8. 6. Set UPDULOCK to 1 in PWM_SCUC. 7. The update of these registers will occur at the beginning of the next PWM period. At this moment the bit UPDULOCK is reset, go to Step 5. for new values. 8. If an update of the duty-cycle values and/or the update period is required, check first that write of new update values is possible by polling the flag WRDY (or by waiting for the corresponding interrupt) in the PWM_ISR2 register. 9. Write registers that need to be updated (PWM_CDTYUPDx, PWM_SCUPUPD). 10. The update of these registers will occur at the next PWM period of the synchronous channels when the Update Period is elapsed. Go to Step 8. for new values.

1467
Figure 49-11. Method 2 (UPDM=1)

CCNT0
CDTYUPD
0x20
0x40
0x60
UPRUPD
0x1
0x3
UPR
0x1
0x3
UPRCNT
0x0
0x1
0x0
0x1
0x0
0x1
0x0
0x1
0x2
0x3
0x0
0x1
0x2
CDTY
0x20
0x40
0x60
WRDY
49.6.3 PWM Comparison Units

The PWM provides 8 independent comparison units able to compare a programmed value with the current value of the channel 0 counter (which is the channel counter of all synchronous channels, Section 49.6.2.7 Synchronous Channels). These comparisons are intended to generate pulses on the event lines (used to synchronize ADC, see Section 49.6.4 PWM Event Lines), to generate software interrupts.
Figure 49-12. Comparison Unit Block Diagram
CEN [PWM_CMPM]x fault on channel 0 CV [PWM_CMPVx]
CNT [PWM_CCNT0]
=
1 0 1
Comparison x
CNT [PWM_CCNT0] is decrementing
CVM [PWM_CMPVx] CALG [PWM_CMR0] CPRCNT [PWM_CMPMx]
CTR [PWM_CMPMx]

1468
The comparison x matches when it is enabled by the bit CEN in the PWM Comparison x Mode Register (PWM_CMPMx for the comparison x) and when the counter of the channel 0 reaches the comparison value defined by the field CV in PWM Comparison x Value Register (PWM_CMPVx for the comparison x). If the counter of the channel 0 is center aligned (CALG = 1 in PWM Channel Mode Register ), the bit CVM (in PWM_CMPVx) defines if the comparison is made when the counter is counting up or counting down (in left alignment mode CALG=0, this bit is useless). If a fault is active on the channel 0, the comparison is disabled and cannot match (see Section 49.6.2.6 Fault Protection). The user can define the periodicity of the comparison x by the fields CTR and CPR (in PWM_CMPVx). The comparison is performed periodically once every CPR+1 periods of the counter of the channel 0, when the value of the comparison period counter CPRCNT (in PWM_CMPMx) reaches the value defined by CTR. CPR is the maximum value of the comparison period counter CPRCNT. If CPR=CTR=0, the comparison is performed at each period of the counter of the channel 0. The comparison x configuration can be modified while the channel 0 is enabled by using the PWM Comparison x Mode Update Register (PWM_CMPMUPDx registers for the comparison x). In the same way, the comparison x value can be modified while the channel 0 is enabled by using the PWM Comparison x Value Update Register (PWM_CMPVUPDx registers for the comparison x). The update of the comparison x configuration and the comparison x value is triggered periodically after the comparison x update period. It is defined by the field CUPR in the PWM_CMPMx. The comparison unit has an update period counter independent from the period counter to trigger this update. When the value of the comparison update period counter CUPRCNT (in PWM_CMPMx) reaches the value defined by CUPR, the update is triggered. The comparison x update period CUPR itself can be updated while the channel 0 is enabled by using the PWM_CMPMUPDx register. CAUTION: To be taken into account, the write of the register PWM_CMPVUPDx must be followed by a write of the register PWM_CMPMUPDx.
The comparison match and the comparison update can be source of an interrupt, but only if it is enabled and not masked. These interrupts can be enabled by the PWM Interrupt Enable Register 2 and disabled by the PWM Interrupt Disable Register 2 . The comparison match interrupt and the comparison update interrupt are reset by reading the PWM Interrupt Status Register 2 .

1469
Figure 49-13. Comparison Waveform

CCNT0
CVUPD CVMVUPD CTRUPD CPRUPD CUPRUPD CV CVM CTR CPR CUPR CUPRCNT CPRCNT Comparison Update CMPU Comparison Match CMPM
0x6
0x2
0x6
0x1 0x1 0x3 0x6
0x2 0x3 0x2 0x2 0x6
0x1 0x1 0x3 0x0 0x0 0x1 0x1 0x2 0x0 0x3 0x1
0x2 0x3 0x2 0x0 0x0 0x1 0x1 0x2 0x2 0x0 0x3 0x1 0x0 0x2 0x1 0x0 0x2 0x1 0x3
49.6.4 PWM Event Lines

The PWM provides 2 independent event lines intended to trigger actions in other peripherals (in particular for ADC (Analog-to-Digital Converter)). A pulse (one cycle of the master clock (MCK)) is generated on an event line, when at least one of the selected comparisons is matching. The comparisons can be selected or unselected independently by the CSEL bits in the PWM Event Line x Register (PWM_ELMRx for the Event Line x).
Figure 49-14. Event Line Block Diagram
CMPM0 (PWM_ISR2) CSEL0 (PWM_ELMRx) CMPM1 (PWM_ISR2) CSEL1 (PWM_ELMRx) CMPM2 (PWM_ISR2) CSEL2 (PWM_ELMRx) PULSE GENERATOR Event Line x
CMPM7 (PWM_ISR2) CSEL7 (PWM_ELMRx)

1470
49.6.5 PWM Controller Operations

49.6.5.1 Initialization Before enabling the channels, they must have been configured by the software application: z Unlock User Interface by writing the WPCMD field in the PWM_WPCR Register. z Configuration of the clock generator (DIVA, PREA, DIVB, PREB in the PWM_CLK register if required). z Selection of the clock for each channel (CPRE field in the PWM_CMRx register) z Configuration of the waveform alignment for each channel (CALG field in the PWM_CMRx register) z Selection of the counter event selection (if CALG = 1) for each channel (CES field in the PWM_CMRx register) z Configuration of the output waveform polarity for each channel (CPOL in the PWM_CMRx register) z Configuration of the period for each channel (CPRD in the PWM_CPRDx register). Writing in PWM_CPRDx register is possible while the channel is disabled. After validation of the channel, the user must use PWM_CPRDUPDx register to update PWM_CPRDx as explained below. z Configuration of the duty-cycle for each channel (CDTY in the PWM_CDTYx register). Writing in PWM_CDTYx register is possible while the channel is disabled. After validation of the channel, the user must use PWM_CDTYUPDx register to update PWM_CDTYx as explained below. z Configuration of the dead-time generator for each channel (DTH and DTL in PWM_DTx) if enabled (DTE bit in the PWM_CMRx register). Writing in the PWM_DTx register is possible while the channel is disabled. After validation of the channel, the user must use PWM_DTUPDx register to update PWM_DTx z Selection of the synchronous channels (SYNCx in the PWM_SCM register) z Configuration of the update mode (UPDM in the PWM_SCM register) z Configuration of the update period (UPR in the PWM_SCUP register) if needed. z Configuration of the comparisons (PWM_CMPVx and PWM_CMPMx). z Configuration of the event lines (PWM_ELMRx). z Configuration of the fault inputs polarity (FPOL in PWM_FMR) z Configuration of the fault protection (FMOD and FFIL in PWM_FMR, PWM_FPV and PWM_FPE1) z Enable of the Interrupts (writing CHIDx and FCHIDx in PWM_IER1 register, and writing WRDYE, ENDTXE, TXBUFE, UNRE, CMPMx and CMPUx in PWM_IER2 register) z Enable of the PWM channels (writing CHIDx in the PWM_ENA register) 49.6.5.2 Source Clock Selection Criteria The large number of source clocks can make selection difficult. The relationship between the value in the PWM Channel Period Register (PWM_CPRDx) and the PWM Channel Duty Cycle Register (PWM_CDTYx) can help the user in choosing. The event number written in the Period Register gives the PWM accuracy. The Duty-Cycle quantum cannot be lower than 1/CPRDx value. The higher the value of PWM_CPRDx, the greater the PWM accuracy. For example, if the user sets 15 (in decimal) in PWM_CPRDx, the user is able to set a value from between 1 up to 14 in PWM_CDTYx Register. The resulting duty-cycle quantum cannot be lower than 1/15 of the PWM period. 49.6.5.3 Changing the Duty-Cycle, the Period and the Dead-Times It is possible to modulate the output waveform duty-cycle, period and dead-times. To prevent unexpected output waveform, the user must use the PWM Channel Duty Cycle Update Register , the PWM Channel Period Update Register and the PWM Channel Dead Time Update Register (PWM_CDTYUPDx, PWM_CPRDUPDx and PWM_DTUPDx) to change waveform parameters while the channel is still enabled. z If the channel is an asynchronous channel (SYNCx = 0 in PWM Sync Channels Mode Register (PWM_SCM)), these registers hold the new period, duty-cycle and dead-times values until the end of the current PWM period and update the values for the next period. z If the channel is a synchronous channel and update method 0 is selected (SYNCx = 1 and UPDM = 0 in PWM_SCM register), these registers hold the new period, duty-cycle and dead-times values until the bit UPDULOCK is written at 1 (in PWM Sync Channels Update Control Register (PWM_SCUC)) and the end of the current PWM period, then update the values for the next period.

1471
If the channel is a synchronous channel and update method 1 or 2 is selected (SYNCx=1 and UPDM=1 or 2 in PWM_SCM register): z registers PWM_CPRDUPDx and PWM_DTUPDx hold the new period and dead-times values until the bit UPDULOCK is written at 1 (in PWM_SCUC register) and the end of the current PWM period, then update the values for the next period. z register PWM_CDTYUPDx holds the new duty-cycle value until the end of the update period of synchronous channels (when UPRCNT is equal to UPR in PWM Sync Channels Update Period Register (PWM_SCUP)) and the end of the current PWM period, then updates the value for the next period. If the update registers PWM_CDTYUPDx, PWM_CPRDUPDx and PWM_DTUPDx are written several times between two updates, only the last written value is taken into account.
Note:
Figure 49-15. Synchronized Period, Duty-Cycle and Dead-Times Update

User's Writing User's Writing User's Writing
PWM_DTUPDx Value
PWM_CPRDUPDx Value
PWM_CDTYUPDx Value
PWM_DTx
PWM_CPRDx
PWM_CDTYx
- If Asynchronous Channel -> End of PWM period - If Synchronous Channel -> End of PWM period and UPDULOCK = 1 - If Asynchronous Channel -> End of PWM period - If Synchronous Channel - If UPDM = 0 -> End of PWM period and UPDULOCK = 1 - If UPDM = 1 or 2 -> End of PWM period and end of Update Period

1472
49.6.5.4 Changing the Synchronous Channels Update Period It is possible to change the update period of synchronous channels while they are enabled (see Method 2: Manual write of duty-cycle values and automatic trigger of the update on page 1467). To prevent an unexpected update of the synchronous channels registers, the user must use the PWM Sync Channels Update Period Update Register (PWM_SCUPUPD) to change the update period of synchronous channels while they are still enabled. This register holds the new value until the end of the update period of synchronous channels (when UPRCNT is equal to UPR in PWM Sync Channels Update Period Register (PWM_SCUP)) and the end of the current PWM period, then updates the value for the next period.
Notes: 1. If the update register PWM_SCUPUPD is written several times between two updates, only the last written value is taken into account. 2. Changing the update period does make sense only if there is one or more synchronous channels and if the update method 1 or 2 is selected (UPDM = 1 or 2 in PWM Sync Channels Mode Register ).
Figure 49-16. Synchronized Update of Update Period Value of Synchronous Channels

User's Writing
PWM_SCUPUPD Value
PWM_SCUP
End of PWM period and end of Update Period of Synchronous Channels

1473
49.6.5.5 Changing the Comparison Value and the Comparison Configuration It is possible to change the comparison values and the comparison configurations while the channel 0 is enabled (see Section 49.6.3 PWM Comparison Units). To prevent unexpected comparison match, the user must use the PWM Comparison x Value Update Register and the PWM Comparison x Mode Update Register (PWM_CMPVUPDx and PWM_CMPMUPDx) to change respectively the comparison values and the comparison configurations while the channel 0 is still enabled. These registers hold the new values until the end of the comparison update period (when CUPRCNT is equal to CUPR in PWM Comparison x Mode Register (PWM_CMPMx) and the end of the current PWM period, then update the values for the next period. CAUTION: Note: To be taken into account, the write of the PWM_CMPVUPDx register must be followed by a write of the PWM_CMPMUPDx register.
If the update registers PWM_CMPVUPDx and PWM_CMPMUPDx are written several times between two updates, only the last written value are taken into account.
Figure 49-17. Synchronized Update of Comparison Values and Configurations

User's Writing User's Writing
PWM_CMPVUPDx Value Comparison Value for comparison x
PWM_CMPMUPDx Value Comparison configuration for comparison x
PWM_CMPVx
PWM_CMPMx
End of channel0 PWM period and end of Comparison Update Period and and PWM_CMPMx written End of channel0 PWM period and end of Comparison Update Period
49.6.5.6 Interrupts Depending on the interrupt mask in the PWM_IMR1 and PWM_IMR2 registers, an interrupt can be generated at the end of the corresponding channel period (CHIDx in the PWM_ISR1 register), after a fault event (FCHIDx in the PWM_ISR1 register), after a comparison match (CMPMx in the PWM_ISR2 register), after a comparison update (CMPUx in the PWM_ISR2 register) or according to the transfer mode of the synchronous channels (WRDY, ENDTX, TXBUFE and UNRE in the PWM_ISR2 register). If the interrupt is generated by the flags CHIDx or FCHIDx, the interrupt remains active until a read operation in the PWM_ISR1 register occurs. If the interrupt is generated by the flags WRDY or UNRE or CMPMx or CMPUx, the interrupt remains active until a read operation in the PWM_ISR2 register occurs. A channel interrupt is enabled by setting the corresponding bit in the PWM_IER1 and PWM_IER2 registers. A channel interrupt is disabled by setting the corresponding bit in the PWM_IDR1 and PWM_IDR2 registers.

1474
49.6.5.7 Write Protect Registers To prevent any single software error that may corrupt PWM behavior, the registers listed below can be write-protected by writing the field WPCMD in the PWM Write Protect Control Register on page 1507 (PWM_WPCR). They are divided into 6 groups:
z
Register group 0:
z
PWM Clock Register on page 1479 PWM Disable Register on page 1481 PWM Sync Channels Mode Register on page 1486 PWM Channel Mode Register on page 1514 PWM Stepper Motor Mode Register on page 1506 PWM Channel Period Register on page 1518 PWM Channel Period Update Register on page 1519 PWM Channel Dead Time Register on page 1521 PWM Channel Dead Time Update Register on page 1522 PWM Fault Mode Register on page 1500 PWM Fault Protection Value Register on page 1503
Register group 1:
z
Register group 2:
z z z
Register group 3:
z z
Register group 4:
z z
Register group 5:
z z
There are two types of Write Protect:

z z
Write Protect SW, which can be enabled or disabled. Write Protect HW, which can just be enabled, only a hardware reset of the PWM controller can disable it.
Both types of Write Protect can be applied independently to a particular register group by means of the WPCMD and WPRG fields in PWM_WPCR register. If at least one Write Protect is active, the register group is write-protected. The field WPCMD allows to perform the following actions depending on its value:
z z z
0 = Disabling the Write Protect SW of the register groups of which the bit WPRG is at 1. 1 = Enabling the Write Protect SW of the register groups of which the bit WPRG is at 1. 2 = Enabling the Write Protect HW of the register groups of which the bit WPRG is at 1.
At any time, the user can determine which Write Protect is active in which register group by the fields WPSWS and WPHWS in the PWM Write Protect Status Register on page 1509 (PWM_WPSR). If a write access in a write-protected register is detected, then the WPVS flag in the PWM_WPSR register is set and the field WPVSRC indicates in which register the write access has been attempted, through its address offset without the two LSBs. The WPVS and PWM_WPSR fields are automatically reset after reading the PWM_WPSR register.

1475
49.7
Pulse Width Modulation Controller (PWM) User Interface
Table 49-6. Register Mapping Offset 0x00 0x04 0x08 0x0C 0x10 0x14 0x18 0x1C 0x20 0x24 0x28 0x2C 0x30 0x34 0x38 0x3C 0x40 0x44 0x48 0x4C 0x50 0x54 0x58 0x5C 0x60 0x64 0x68 0x6C 0x70-0x78 0x7C 0x80 0x84-0x9C 0xA0 - 0xAC 0xB0 0xB4-0xBC Register PWM Clock Register PWM Enable Register PWM Disable Register PWM Status Register PWM Interrupt Enable Register 1 PWM Interrupt Disable Register 1 PWM Interrupt Mask Register 1 PWM Interrupt Status Register 1 PWM Sync Channels Mode Register Reserved PWM Sync Channels Update Control Register PWM Sync Channels Update Period Register PWM Sync Channels Update Period Update Register PWM Interrupt Enable Register 2 PWM Interrupt Disable Register 2 PWM Interrupt Mask Register 2 PWM Interrupt Status Register 2 PWM Output Override Value Register PWM Output Selection Register PWM Output Selection Set Register PWM Output Selection Clear Register PWM Output Selection Set Update Register PWM Output Selection Clear Update Register PWM Fault Mode Register PWM Fault Status Register PWM Fault Clear Register PWM Fault Protection Value Register PWM Fault Protection Enable Register Reserved PWM Event Line 0 Mode Register PWM Event Line 1 Mode Register Reserved Reserved PWM Stepper Motor Mode Register Reserved Name PWM_CLK PWM_ENA PWM_DIS PWM_SR PWM_IER1 PWM_IDR1 PWM_IMR1 PWM_ISR1 PWM_SCM PWM_SCUC PWM_SCUP PWM_SCUPUPD PWM_IER2 PWM_IDR2 PWM_IMR2 PWM_ISR2 PWM_OOV PWM_OS PWM_OSS PWM_OSC PWM_OSSUPD PWM_OSCUPD PWM_FMR PWM_FSR PWM_FCR PWM_FPV PWM_FPE PWM_ELMR0 PWM_ELMR1 PWM_SMMR Access Read-write Write-only Write-only Read-only Write-only Write-only Read-only Read-only Read-write Read-write Read-write Write-only Write-only Write-only Read-only Read-only Read-write Read-write Write-only Write-only Write-only Write-only Read-write Read-only Write-only Read-write Read-write Read-write Read-write Read-write Reset 0x0 0x0 0x0 0x0 0x0 0x0 0x0 0x0 0x0 0x0 0x0 0x0 0x0 0x0 0x0 0x0 0x0 0x0 0x0

1476
Table 49-6. Register Mapping (Continued) Offset 0xC0-E0 0xE4 0xE8 0xEC - 0xFC 0x130 0x134 0x138 0x13C 0x140 0x144 0x148 0x14C 0x150 0x154 0x158 0x15C 0x160 0x164 0x168 0x16C 0x170 0x174 0x178 0x17C 0x180 0x184 0x188 0x18C 0x190 0x194 0x198 0x19C 0x1A0 0x1A4 0x1A8 0x1AC 0x1B0 - 0x1FC Register Reserved PWM Write Protect Control Register PWM Write Protect Status Register Reserved PWM Comparison 0 Value Register PWM Comparison 0 Value Update Register PWM Comparison 0 Mode Register PWM Comparison 0 Mode Update Register PWM Comparison 1 Value Register PWM Comparison 1 Value Update Register PWM Comparison 1 Mode Register PWM Comparison 1 Mode Update Register PWM Comparison 2 Value Register PWM Comparison 2 Value Update Register PWM Comparison 2 Mode Register PWM Comparison 2 Mode Update Register PWM Comparison 3 Value Register PWM Comparison 3 Value Update Register PWM Comparison 3 Mode Register PWM Comparison 3 Mode Update Register PWM Comparison 4 Value Register PWM Comparison 4 Value Update Register PWM Comparison 4 Mode Register PWM Comparison 4 Mode Update Register PWM Comparison 5 Value Register PWM Comparison 5 Value Update Register PWM Comparison 5 Mode Register PWM Comparison 5 Mode Update Register PWM Comparison 6 Value Register PWM Comparison 6 Value Update Register PWM Comparison 6 Mode Register PWM Comparison 6 Mode Update Register PWM Comparison 7 Value Register PWM Comparison 7 Value Update Register PWM Comparison 7 Mode Register PWM Comparison 7 Mode Update Register Reserved Name PWM_WPCR PWM_WPSR PWM_CMPV0 PWM_CMPVUPD0 PWM_CMPM0 PWM_CMPMUPD0 PWM_CMPV1 PWM_CMPVUPD1 PWM_CMPM1 PWM_CMPMUPD1 PWM_CMPV2 PWM_CMPVUPD2 PWM_CMPM2 PWM_CMPMUPD2 PWM_CMPV3 PWM_CMPVUPD3 PWM_CMPM3 PWM_CMPMUPD3 PWM_CMPV4 PWM_CMPVUPD4 PWM_CMPM4 PWM_CMPMUPD4 PWM_CMPV5 PWM_CMPVUPD5 PWM_CMPM5 PWM_CMPMUPD5 PWM_CMPV6 PWM_CMPVUPD6 PWM_CMPM6 PWM_CMPMUPD6 PWM_CMPV7 PWM_CMPVUPD7 PWM_CMPM7 PWM_CMPMUPD7 Access Write-only Read-only Read-write Write-only Read-write Write-only Read-write Write-only Read-write Write-only Read-write Write-only Read-write Write-only Read-write Write-only Read-write Write-only Read-write Write-only Read-write Write-only Read-write Write-only Read-write Write-only Read-write Write-only Read-write Write-only Read-write Write-only Read-write Write-only Reset 0x0 0x0 0x0 0x0 0x0 0x0 0x0 0x0 0x0 0x0 0x0 0x0 0x0 0x0 0x0 0x0 0x0

1477
Table 49-6. Register Mapping (Continued) Offset 0x200 + ch_num * 0x20 + 0x00 0x200 + ch_num * 0x20 + 0x04 0x200 + ch_num * 0x20 + 0x08 0x200 + ch_num * 0x20 + 0x0C 0x200 + ch_num * 0x20 + 0x10 0x200 + ch_num * 0x20 + 0x14 0x200 + ch_num * 0x20 + 0x18 0x200 + ch_num * 0x20 + 0x1C Register PWM Channel Mode Register(1) PWM Channel Duty Cycle Register(1) PWM Channel Duty Cycle Update Register(1) PWM Channel Period Register(1) PWM Channel Period Update Register(1) PWM Channel Counter Register(1) PWM Channel Dead Time Register(1) PWM Channel Dead Time Update Register(1) Name PWM_CMR PWM_CDTY PWM_CDTYUPD PWM_CPRD PWM_CPRDUPD PWM_CCNT PWM_DT PWM_DTUPD Access Read-write Read-write Write-only Read-write Write-only Read-only Read-write Write-only Reset 0x0 0x0 0x0 0x0 0x0
Note:
1. Some registers are indexed with ch_num index ranging from 0 to 3.

1478
49.7.1 PWM Clock Register

31 23
PWM_CLK 0xF002C000 Read-write

30 22 29 21 28 20 DIVB 27 26 PREB 19 18 17 16 25 24
15 7
14 6
13 5
12 4 DIVA
11
10 PREA
This register can only be written if the bits WPSWS0 and WPHWS0 are cleared in PWM Write Protect Status Register on page 1509.
DIVA, DIVB: CLKA, CLKB Divide Factor

DIVA, DIVB 0 1 2-255 CLKA, CLKB CLKA, CLKB clock is turned off CLKA, CLKB clock is clock selected by PREA, PREB CLKA, CLKB clock is clock selected by PREA, PREB divided by DIVA, DIVB factor.
PREA, PREB: CLKA, CLKB Source Clock Selection

PREA, PREB 0 0 0 0 0 0 0 0 1 1 1 Other 0 0 0 0 1 1 1 1 0 0 0 0 0 1 1 0 0 1 1 0 0 1 0 1 0 1 0 1 0 1 0 1 0 Divider Input Clock MCK MCK/2 MCK/4 MCK/8 MCK/16 MCK/32 MCK/64 MCK/128 MCK/256 MCK/512 MCK/1024 Reserved

1479
49.7.2 PWM Enable Register

31 23 15 7
PWM_ENA 0xF002C004 Write-only

30 22 14 6 29 21 13 5 28 20 12 4 27 19 11 3 CHID3 26 18 10 2 CHID2 25 17 9 1 CHID1 24 16 8 0 CHID0
CHIDx: Channel ID
0 = No effect. 1 = Enables PWM output for channel x.

1480
49.7.3 PWM Disable Register

31 23 15 7
PWM_DIS 0xF002C008 Write-only

CHIDx: Channel ID
0 = No effect. 1 = Disables PWM output for channel x.

1481
49.7.4 PWM Status Register

31 23 15 7
PWM_SR 0xF002C00C Read-only

CHIDx: Channel ID
0 = PWM output for channel x is disabled. 1 = PWM output for channel x is enabled.

1482
49.7.5 PWM Interrupt Enable Register 1

31 23 15 7
PWM_IER1 0xF002C010 Write-only

30 22 14 6 29 21 13 5 28 20 12 4 27 19 FCHID3 11 3 CHID3 26 18 FCHID2 10 2 CHID2 25 17 FCHID1 9 1 CHID1 24 16 FCHID0 8 0 CHID0
CHIDx: Counter Event on Channel x Interrupt Enable FCHIDx: Fault Protection Trigger on Channel x Interrupt Enable
49.7.6 PWM Interrupt Disable Register 1

31 23 15 7
PWM_IDR1 0xF002C014 Write-only

CHIDx: Counter Event on Channel x Interrupt Disable FCHIDx: Fault Protection Trigger on Channel x Interrupt Disable

1483
49.7.7 PWM Interrupt Mask Register 1

31 23 15 7
PWM_IMR1 0xF002C018 Read-only

CHIDx: Counter Event on Channel x Interrupt Mask FCHIDx: Fault Protection Trigger on Channel x Interrupt Mask

1484
49.7.8 PWM Interrupt Status Register 1

31 23 15 7
PWM_ISR1 0xF002C01C Read-only

CHIDx: Counter Event on Channel x

0 = No new counter event has occurred since the last read of the PWM_ISR1 register. 1 = At least one counter event has occurred since the last read of the PWM_ISR1 register.
FCHIDx: Fault Protection Trigger on Channel x

0 = No new trigger of the fault protection since the last read of the PWM_ISR1 register. 1 = At least one trigger of the fault protection since the last read of the PWM_ISR1 register. Note: Reading PWM_ISR1 automatically clears CHIDx and FCHIDx flags.

1485
49.7.9 PWM Sync Channels Mode Register

31 23 15 7
PWM_SCM 0xF002C020 Read-write

30 22 14 6 29 21 13 5 28 20 12 4 27 19 11 3 SYNC3 26 18 10 2 SYNC2 25 17 UPDM 9 1 SYNC1 8 0 SYNC0 24 16
SYNCx: Synchronous Channel x

0 = Channel x is not a synchronous channel. 1 = Channel x is a synchronous channel.
UPDM: Synchronous Channels Update Mode

Value 0 1 3 Name MODE0 MODE1 Description Manual write of double buffer registers and manual update of synchronous channels(1) Manual write of double buffer registers and automatic update of synchronous channels(2) Reserved
Notes: 1. The update occurs at the beginning of the next PWM period, when the UPDULOCK bit in PWM Sync Channels Update Control Register is set. 2. The update occurs when the Update Period is elapsed.

1486
49.7.10 PWM Sync Channels Update Control Register

31 23 15 7
PWM_SCUC 0xF002C028 Read-write

30 22 14 6 29 21 13 5 28 20 12 4 27 19 11 3 26 18 10 2 25 17 9 1 24 16 8 0 UPDULOCK
UPDULOCK: Synchronous Channels Update Unlock

0 = No effect 1 = If the UPDM field is set to 0 in PWM Sync Channels Mode Register on page 1486, writing the UPDULOCK bit to 1 triggers the update of the period value, the duty-cycle and the dead-time values of synchronous channels at the beginning of the next PWM period. If the field UPDM is set to 1 or 2, writing the UPDULOCK bit to 1 triggers only the update of the period value and of the dead-time values of synchronous channels. This bit is automatically reset when the update is done.

1487
49.7.11 PWM Sync Channels Update Period Register

31 23 15 7
PWM_SCUP 0xF002C02C Read-write

30 22 14 6 UPRCNT 29 21 13 5 28 20 12 4 27 19 11 3 26 18 10 2 UPR 25 17 9 1 24 16 8 0
UPR: Update Period

Defines the time between each update of the synchronous channels if automatic trigger of the update is activated (UPDM = 1 or UPDM = 2 in PWM Sync Channels Mode Register on page 1486). This time is equal to UPR+1 periods of the synchronous channels.
UPRCNT: Update Period Counter

Reports the value of the Update Period Counter.

1488
49.7.12 PWM Sync Channels Update Period Update Register

31 23 15 7
PWM_SCUPUPD 0xF002C030 Write-only

30 22 14 6 29 21 13 5 28 20 12 4 27 19 11 3 26 18 10 2 UPRUPD 25 17 9 1 24 16 8 0
This register acts as a double buffer for the UPR value. This prevents an unexpected automatic trigger of the update of synchronous channels.
UPRUPD: Update Period Update

Defines the wanted time between each update of the synchronous channels if automatic trigger of the update is activated (UPDM = 1 or UPDM = 2 in PWM Sync Channels Mode Register on page 1486). This time is equal to UPR+1 periods of the synchronous channels.

1489
49.7.13 PWM Interrupt Enable Register 2

31 23 CMPU7 15 CMPM7 7
PWM_IER2 0xF002C034 Write-only

30 22 CMPU6 14 CMPM6 6 29 21 CMPU5 13 CMPM5 5 28 20 CMPU4 12 CMPM4 4 27 19 CMPU3 11 CMPM3 3 UNRE 26 18 CMPU2 10 CMPM2 2 25 17 CMPU1 9 CMPM1 1 24 16 CMPU0 8 CMPM0 0 WRDY
WRDY: Write Ready for Synchronous Channels Update Interrupt Enable UNRE: Synchronous Channels Update Underrun Error Interrupt Enable CMPMx: Comparison x Match Interrupt Enable CMPUx: Comparison x Update Interrupt Enable

1490
49.7.14 PWM Interrupt Disable Register 2

31 23 CMPU7 15 CMPM7 7
PWM_IDR2 0xF002C038 Write-only

WRDY: Write Ready for Synchronous Channels Update Interrupt Disable UNRE: Synchronous Channels Update Underrun Error Interrupt Disable CMPMx: Comparison x Match Interrupt Disable CMPUx: Comparison x Update Interrupt Disable

1491
49.7.15 PWM Interrupt Mask Register 2

31 23 CMPU7 15 CMPM7 7
PWM_IMR2 0xF002C03C Read-only

WRDY: Write Ready for Synchronous Channels Update Interrupt Mask UNRE: Synchronous Channels Update Underrun Error Interrupt Mask CMPMx: Comparison x Match Interrupt Mask CMPUx: Comparison x Update Interrupt Mask

1492
49.7.16 PWM Interrupt Status Register 2

31 23 CMPU7 15 CMPM7 7
PWM_ISR2 0xF002C040 Read-only

WRDY: Write Ready for Synchronous Channels Update

0 = New duty-cycle and dead-time values for the synchronous channels cannot be written. 1 = New duty-cycle and dead-time values for the synchronous channels can be written.
UNRE: Synchronous Channels Update Underrun Error

0 = No Synchronous Channels Update Underrun has occurred since the last read of the PWM_ISR2 register. 1 = At least one Synchronous Channels Update Underrun has occurred since the last read of the PWM_ISR2 register.
CMPMx: Comparison x Match

0 = The comparison x has not matched since the last read of the PWM_ISR2 register. 1 = The comparison x has matched at least one time since the last read of the PWM_ISR2 register.
CMPUx: Comparison x Update

0 = The comparison x has not been updated since the last read of the PWM_ISR2 register. 1 = The comparison x has been updated at least one time since the last read of the PWM_ISR2 register. Note: Reading PWM_ISR2 automatically clears flags WRDY, UNRE and CMPSx.

1493
49.7.17 PWM Output Override Value Register

31 23 15 7
PWM_OOV 0xF002C044 Read-write

30 22 14 6 29 21 13 5 28 20 12 4 27 19 OOVL3 11 3 OOVH3 26 18 OOVL2 10 2 OOVH2 25 17 OOVL1 9 1 OOVH1 24 16 OOVL0 8 0 OOVH0
OOVHx: Output Override Value for PWMH output of the channel x

0 = Override value is 0 for PWMH output of channel x. 1 = Override value is 1 for PWMH output of channel x.
OOVLx: Output Override Value for PWML output of the channel x

0 = Override value is 0 for PWML output of channel x. 1 = Override value is 1 for PWML output of channel x.

1494
49.7.18 PWM Output Selection Register

31 23 15 7
PWM_OS 0xF002C048 Read-write

30 22 14 6 29 21 13 5 28 20 12 4 27 19 OSL3 11 3 OSH3 26 18 OSL2 10 2 OSH2 25 17 OSL1 9 1 OSH1 24 16 OSL0 8 0 OSH0
OSHx: Output Selection for PWMH output of the channel x

0 = Dead-time generator output DTOHx selected as PWMH output of channel x. 1 = Output override value OOVHx selected as PWMH output of channel x.
OSLx: Output Selection for PWML output of the channel x

0 = Dead-time generator output DTOLx selected as PWML output of channel x. 1 = Output override value OOVLx selected as PWML output of channel x.

1495
49.7.19 PWM Output Selection Set Register

31 23 15 7
PWM_OSS 0xF002C04C Write-only

30 22 14 6 29 21 13 5 28 20 12 4 27 19 OSSL3 11 3 OSSH3 26 18 OSSL2 10 2 OSSH2 25 17 OSSL1 9 1 OSSH1 24 16 OSSL0 8 0 OSSH0
OSSHx: Output Selection Set for PWMH output of the channel x

0 = No effect. 1 = Output override value OOVHx selected as PWMH output of channel x.
OSSLx: Output Selection Set for PWML output of the channel x

0 = No effect. 1 = Output override value OOVLx selected as PWML output of channel x.

1496
49.7.20 PWM Output Selection Clear Register

31 23 15 7
PWM_OSC 0xF002C050 Write-only

30 22 14 6 29 21 13 5 28 20 12 4 27 19 OSCL3 11 3 OSCH3 26 18 OSCL2 10 2 OSCH2 25 17 OSCL1 9 1 OSCH1 24 16 OSCL0 8 0 OSCH0
OSCHx: Output Selection Clear for PWMH output of the channel x

0 = No effect. 1 = Dead-time generator output DTOHx selected as PWMH output of channel x.
OSCLx: Output Selection Clear for PWML output of the channel x

0 = No effect. 1 = Dead-time generator output DTOLx selected as PWML output of channel x.

1497
49.7.21 PWM Output Selection Set Update Register

31 23 15 7
PWM_OSSUPD 0xF002C054 Write-only

30 22 14 6 29 21 13 5 28 20 12 4 27 19 OSSUPL3 11 3 OSSUPH3 26 18 OSSUPL2 10 2 OSSUPH2 25 17 OSSUPL1 9 1 OSSUPH1 24 16 OSSUPL0 8 0 OSSUPH0
OSSUPHx: Output Selection Set for PWMH output of the channel x

0 = No effect. 1 = Output override value OOVHx selected as PWMH output of channel x at the beginning of the next channel x PWM period.
OSSUPLx: Output Selection Set for PWML output of the channel x

0 = No effect. 1 = Output override value OOVLx selected as PWML output of channel x at the beginning of the next channel x PWM period.

1498
49.7.22 PWM Output Selection Clear Update Register

31 23 15 7
PWM_OSCUPD 0xF002C058 Write-only

30 22 14 6 29 21 13 5 28 20 12 4 27 19 OSCUPL3 11 3 OSCUPH3 26 18 OSCUPL2 10 2 OSCUPH2 25 17 OSCUPL1 9 1 OSCUPH1 24 16 OSCUPL0 8 0 OSCUPH0
OSCUPHx: Output Selection Clear for PWMH output of the channel x

0 = No effect. 1 = Dead-time generator output DTOHx selected as PWMH output of channel x at the beginning of the next channel x PWM period.
OSCUPLx: Output Selection Clear for PWML output of the channel x

0 = No effect. 1 = Dead-time generator output DTOLx selected as PWML output of channel x at the beginning of the next channel x PWM period.

1499
49.7.23 PWM Fault Mode Register

31
PWM_FMR 0xF002C05C Read-write

30 29 28 27 26 25 24
23
22
21
20 FFIL
19
18
17
16
15
14
13
12 FMOD
11
10
4 FPOL
FPOL: Fault Polarity

For each field bit y (fault input number):
0 = The fault y becomes active when the fault input y is at 0. 1 = The fault y becomes active when the fault input y is at 1. FMOD: Fault Activation Mode
0 = The fault y is active until the Fault condition is removed at the peripheral(1) level. 1 = The fault y stays active until the Fault condition is removed at the peripheral(1) level AND until it is cleared in the PWM Fault Clear Register .
Note: 1. The Peripheral generating the fault.
FFIL: Fault Filtering

0 = The fault input y is not filtered. 1 = The fault input y is filtered.
CAUTION: To prevent an unexpected activation of the status flag FSy in the PWM Fault Status Register on page 1501, the bit FMODy can be set to 1 only if the FPOLy bit has been previously configured to its final value.

1500
49.7.24 PWM Fault Status Register

31
PWM_FSR 0xF002C060 Read-only

30 29 28 27 26 25 24
23
22
21
20
19
18
17
16
15
14
13
12 FS
11
10
4 FIV
FIV: Fault Input Value

0 = The current sampled value of the fault input y is 0 (after filtering if enabled). 1 = The current sampled value of the fault input y is 1 (after filtering if enabled). FS: Fault Status
0 = The fault y is not currently active. 1 = The fault y is currently active.

1501
49.7.25 PWM Fault Clear Register

31
PWM_FCR 0xF002C064 Write-only

30 29 28 27 26 25 24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
4 FCLR
FCLR: Fault Clear

0 = No effect. 1 = If bit y of FMOD field is set to 1 and if the fault input y is not at the level defined by the bit y of FPOL field, the fault y is cleared and becomes inactive (FMOD and FPOL fields belong to PWM Fault Mode Register on page 1500), else writing this bit to 1 has no effect.

1502
49.7.26 PWM Fault Protection Value Register

31 23 15 7
PWM_FPV 0xF002C068 Read-write

30 22 14 6 29 21 13 5 28 20 12 4 27 19 FPVL3 11 3 FPVH3 26 18 FPVL2 10 2 FPVH2 25 17 FPVL1 9 1 FPVH1 24 16 FPVL0 8 0 FPVH0
FPVHx: Fault Protection Value for PWMH output on channel x

0 = PWMH output of channel x is forced to 0 when fault occurs. 1 = PWMH output of channel x is forced to 1 when fault occurs.
FPVLx: Fault Protection Value for PWML output on channel x

0 = PWML output of channel x is forced to 0 when fault occurs. 1 = PWML output of channel x is forced to 1 when fault occurs.

1503
49.7.27 PWM Fault Protection Enable Register

31
PWM_FPE 0xF002C06C Read-write

30 29 28 FPE3 27 26 25 24
23
22
21
20 FPE2
19
18
17
16
15
14
13
12 FPE1
11
10
4 FPE0
This register can only be written if the bits WPSWS5 and WPHWS5 are cleared in PWM Write Protect Status Register on page 1509. Only the first 1 bit (number of fault input pins) of the FPE0, FPE1, FPE2 and FPE3 fields is significant.
FPEx: Fault Protection Enable for channel x

For each field bit y (fault input number): 0 = Fault y is not used for the Fault Protection of channel x. 1 = Fault y is used for the Fault Protection of channel x.
CAUTION: To prevent an unexpected activation of the Fault Protection, the bit y of the FPEx field can be set to 1 only if the corresponding FPOL bit has been previously configured to its final value in PWM Fault Mode Register on page 1500.

1504
49.7.28 PWM Event Line x Register

31 23 15 7 CSEL7
PWM_ELMRx 0xF002C07C Read-write

30 22 14 6 CSEL6 29 21 13 5 CSEL5 28 20 12 4 CSEL4 27 19 11 3 CSEL3 26 18 10 2 CSEL2 25 17 9 1 CSEL1 24 16 8 0 CSEL0
CSELy: Comparison y Selection

0 = A pulse is not generated on the event line x when the comparison y matches. 1 = A pulse is generated on the event line x when the comparison y match.

1505
49.7.29 PWM Stepper Motor Mode Register

31 23 15 7
PWM_SMMR 0xF002C0B0 Read-write

30 22 14 6 29 21 13 5 28 20 12 4 27 19 11 3 26 18 10 2 25 17 DOWN1 9 1 GCEN1 24 16 DOWN0 8 0 GCEN0
GCENx: Gray Count ENable

0 = Disables gray count generation on PWML[2*x], PWMH[2*x], PWML[2*x +1], PWMH[2*x +1] 1 = Enables gray count generation on PWML[2*x], PWMH[2*x], PWML[2*x +1], PWMH[2*x +1.
DOWNx: DOWN Count

0 = Up counter 1 = Down counter

1506
49.7.30 PWM Write Protect Control Register

31
PWM_WPCR 0xF002C0E4 Write-only

30 29 28 WPKEY 27 26 25 24
23
22
21
20 WPKEY
19
18
17
16
15
14
13
12 WPKEY
11
10
7 WPRG5
6 WPRG4
5 WPRG3
4 WPRG2
3 WPRG1
2 WPRG0
1 WPCMD
WPCMD: Write Protect Command

This command is performed only if the WPKEY value is correct. 0 = Disables the Write Protect SW of the register groups of which the bit WPRGx is at 1. 1 = Enables the Write Protect SW of the register groups of which the bit WPRGx is at 1. 2 = Enables the Write Protect HW of the register groups of which the bit WPRGx is at 1.
Moreover, to meet security requirements, in this mode of operation, the PIO lines associated with PWM can not be configured through the PIO interface, not even by the PIO controller.
3 = No effect. Note: Only a hardware reset of the PWM controller can disable the Write Protect HW.
WPRGx: Write Protect Register Group x

0 = The WPCMD command has no effect on the register group x. 1 = The WPCMD command is applied to the register group x.
WPKEY: Write Protect Key

Should be written at value 0x50574D (PWM in ASCII). Writing any other value in this field aborts the write operation of the WPCMD field. Always reads as 0. List of register groups:
Register group 0:
PWM Clock Register on page 1479
Register group 1:
PWM Disable Register on page 1481
Register group 2:
PWM Sync Channels Mode Register on page 1486 PWM Channel Mode Register on page 1514 PWM Stepper Motor Mode Register on page 1506
Register group 3:
PWM Channel Period Register on page 1518 PWM Channel Period Update Register on page 1519

1507
Register group 4:
PWM Channel Dead Time Register on page 1521 PWM Channel Dead Time Update Register on page 1522
Register group 5:
PWM Fault Mode Register on page 1500 PWM Fault Protection Value Register on page 1503

1508
49.7.31 PWM Write Protect Status Register

31
PWM_WPSR 0xF002C0E8 Read-only

30 29 28 WPVSRC 27 26 25 24
23
22
21
20 WPVSRC
19
18
17
16
15 7 WPVS
14 6
13 WPHWS5 5 WPSWS5
12 WPHWS4 4 WPSWS4
11 WPHWS3 3 WPSWS3
10 WPHWS2 2 WPSWS2
9 WPHWS1 1 WPSWS1
8 WPHWS0 0 WPSWS0
WPSWSx: Write Protect SW Status

0 = The Write Protect SW x of the register group x is disabled. 1 = The Write Protect SW x of the register group x is enabled.
WPHWSx: Write Protect HW Status

0 = The Write Protect HW x of the register group x is disabled. 1 = The Write Protect HW x of the register group x is enabled.

0 = No Write Protect violation has occurred since the last read of the PWM_WPSR register. 1 = At least one Write Protect violation has occurred since the last read of the PWM_WPSR register. If this violation is an unauthorized attempt to write a protected register, the associated violation is reported into field WPVSRC.

When WPVS is active, this field indicates the write-protected register (through address offset) in which a write access has been attempted. Note: Note: The two LSBs of the address offset of the write-protected register are not reported Reading PWM_WPSR automatically clears WPVS and WPVSRC fields.

1509
49.7.32 PWM Comparison x Value Register

31 23
PWM_CMPVx 0xF002C130 [0], 0xF002C140 [1], 0xF002C150 [2], 0xF002C160 [3], 0xF002C170 [4], 0xF002C180 [5], 0xF002C190 [6], 0xF002C1A0 [7] Read-write
30 22 29 21 28 20 CV 27 19 26 18 25 17 24 CVM 16
15
14
13
12 CV
11
10
4 CV
Only the first 16 bits (channel counter size) of field CV are significant.
CV: Comparison x Value

Defines the comparison x value to be compared with the counter of the channel 0.
CVM: Comparison x Value Mode

0 = The comparison x between the counter of the channel 0 and the comparison x value is performed when this counter is incrementing. 1 = The comparison x between the counter of the channel 0 and the comparison x value is performed when this counter is decrementing. Note: This bit is useless if the counter of the channel 0 is left aligned (CALG = 0 in PWM Channel Mode Register on page 1514)

1510
49.7.33 PWM Comparison x Value Update Register

31 23
PWM_CMPVUPDx 0xF002C134 [0], 0xF002C144 [1], 0xF002C154 [2], 0xF002C164 [3], 0xF002C174 [4], 0xF002C184 [5], 0xF002C194 [6], 0xF002C1A4 [7] Write-only
30 22 29 21 28 20 CVUPD 27 19 26 18 25 17 24 CVMUPD 16
15
14
13
12 CVUPD
11
10
4 CVUPD
This register acts as a double buffer for the CV and CVM values. This prevents an unexpected comparison x match. Only the first 16 bits (channel counter size) of field CVUPD are significant.
CVUPD: Comparison x Value Update

Defines the comparison x value to be compared with the counter of the channel 0.
CVMUPD: Comparison x Value Mode Update

0 = The comparison x between the counter of the channel 0 and the comparison x value is performed when this counter is incrementing. 1 = The comparison x between the counter of the channel 0 and the comparison x value is performed when this counter is decrementing. Note: This bit is useless if the counter of the channel 0 is left aligned (CALG = 0 in PWM Channel Mode Register on page 1514)
CAUTION: to be taken into account, the write of the register PWM_CMPVUPDx must be followed by a write of the register PWM_CMPMUPDx.

1511
49.7.34 PWM Comparison x Mode Register

31 23
PWM_CMPMx 0xF002C138 [0], 0xF002C148 [1], 0xF002C158 [2], 0xF002C168 [3], 0xF002C178 [4], 0xF002C188 [5], 0xF002C198 [6], 0xF002C1A8 [7] Read-write
30 22 CUPRCNT 29 21 28 20 27 19 26 18 CUPR 13 CPRCNT 7 6 CTR 5 4 3 2 12 11 10 CPR 1 0 CEN 9 8 25 17 24 16
15
14
CEN: Comparison x Enable

0 = The comparison x is disabled and can not match. 1 = The comparison x is enabled and can match.
CTR: Comparison x Trigger

The comparison x is performed when the value of the comparison x period counter (CPRCNT) reaches the value defined by CTR.
CPR: Comparison x Period

CPR defines the maximum value of the comparison x period counter (CPRCNT). The comparison x value is performed periodically once every CPR+1 periods of the channel 0 counter.
CPRCNT: Comparison x Period Counter

Reports the value of the comparison x period counter. Note: The field CPRCNT is read-only
CUPR: Comparison x Update Period

Defines the time between each update of the comparison x mode and the comparison x value. This time is equal to CUPR+1 periods of the channel 0 counter.
CUPRCNT: Comparison x Update Period Counter

Reports the value of the comparison x update period counter. Note: The field CUPRCNT is read-only

1512
49.7.35 PWM Comparison x Mode Update Register

31 23 15 7
PWM_CMPMUPDx 0xF002C13C [0], 0xF002C14C [1], 0xF002C15C [2], 0xF002C16C [3], 0xF002C17C [4], 0xF002C18C [5], 0xF002C19C [6], 0xF002C1AC [7] Write-only
30 22 14 6 CTRUPD 29 21 13 5 28 20 12 4 27 19 26 18 CUPRUPD 11 10 CPRUPD 3 2 1 0 CENUPD 9 8 25 17 24 16
This register acts as a double buffer for the CEN, CTR, CPR and CUPR values. This prevents an unexpected comparison x match.
CENUPD: Comparison x Enable Update

0 = The comparison x is disabled and can not match. 1 = The comparison x is enabled and can match.
CTRUPD: Comparison x Trigger Update

The comparison x is performed when the value of the comparison x period counter (CPRCNT) reaches the value defined by CTR.
CPRUPD: Comparison x Period Update

CPR defines the maximum value of the comparison x period counter (CPRCNT). The comparison x value is performed periodically once every CPR+1 periods of the channel 0 counter.
CUPRUPD: Comparison x Update Period Update

Defines the time between each update of the comparison x mode and the comparison x value. This time is equal to CUPR+1 periods of the channel 0 counter.

1513
49.7.36 PWM Channel Mode Register

31 23 15 7
PWM_CMRx [x=0..3] 0xF002C200 [0], 0xF002C220 [1], 0xF002C240 [2], 0xF002C260 [3] Read-write
30 22 14 6 29 21 13 5 28 20 12 4 27 19 11 3 26 18 DTLI 10 CES 2 CPRE 25 17 DTHI 9 CPOL 1 24 16 DTE 8 CALG 0
CPRE: Channel Pre-scaler

Value 0b0000 0b0001 0b0010 0b0011 0b0100 0b0101 0b0110 0b0111 0b1000 0b1001 0b1010 0b1011 0b1100 MCK MCK_DIV_2 MCK_DIV_4 MCK_DIV_8 MCK_DIV_16 MCK_DIV_32 MCK_DIV_64 MCK_DIV_128 MCK_DIV_256 MCK_DIV_512 MCK_DIV_1024 CLKA CLKB Name Master clock Master clock/2 Master clock/4 Master clock/8 Master clock/16 Master clock/32 Master clock/64 Master clock/128 Master clock/256 Master clock/512 Master clock/1024 Clock A Clock B Description
CALG: Channel Alignment

0 = The period is left aligned. 1 = The period is center aligned.
CPOL: Channel Polarity

0 = The OCx output waveform (output from the comparator) starts at a low level. 1 = The OCx output waveform (output from the comparator) starts at a high level.

1514
CES: Counter Event Selection

The bit CES defines when the channel counter event occurs when the period is center aligned (flag CHIDx in the PWM Interrupt Status Register 1 on page 1485).
CALG = 0 (Left Alignment):

0/1 = The channel counter event occurs at the end of the PWM period.
CALG = 1 (Center Alignment):

0 = The channel counter event occurs at the end of the PWM period. 1 = The channel counter event occurs at the end of the PWM period and at half the PWM period.
DTE: Dead-Time Generator Enable

0 = The dead-time generator is disabled. 1 = The dead-time generator is enabled.
DTHI: Dead-Time PWMHx Output Inverted

0 = The dead-time PWMHx output is not inverted. 1 = The dead-time PWMHx output is inverted.
DTLI: Dead-Time PWMLx Output Inverted

0 = The dead-time PWMLx output is not inverted. 1 = The dead-time PWMLx output is inverted.

1515
49.7.37 PWM Channel Duty Cycle Register

31 23
PWM_CDTYx [x=0..3] 0xF002C204 [0], 0xF002C224 [1], 0xF002C244 [2], 0xF002C264 [3] Read-write
30 22 29 21 28 20 CDTY 27 19 26 18 25 17 24 16
15
14
13
12 CDTY
11
10
4 CDTY
Only the first 16 bits (channel counter size) are significant.
CDTY: Channel Duty-Cycle

Defines the waveform duty-cycle. This value must be defined between 0 and CPRD (PWM_CPRx).

1516
49.7.38 PWM Channel Duty Cycle Update Register

31 23
PWM_CDTYUPDx [x=0..3] 0xF002C208 [0], 0xF002C228 [1], 0xF002C248 [2], 0xF002C268 [3] Write-only.
30 22 29 21 28 20 CDTYUPD 27 19 26 18 25 17 24 16
15
14
13
12 CDTYUPD
11
10
4 CDTYUPD
This register acts as a double buffer for the CDTY value. This prevents an unexpected waveform when modifying the waveform duty-cycle. Only the first 16 bits (channel counter size) are significant.
CDTYUPD: Channel Duty-Cycle Update

Defines the waveform duty-cycle. This value must be defined between 0 and CPRD (PWM_CPRx).

1517
49.7.39 PWM Channel Period Register

31 23
PWM_CPRDx [x=0..3] 0xF002C20C [0], 0xF002C22C [1], 0xF002C24C [2], 0xF002C26C [3] Read-write
30 22 29 21 28 20 CPRD 27 19 26 18 25 17 24 16
15
14
13
12 CPRD
11
10
4 CPRD
This register can only be written if the bits WPSWS3 and WPHWS3 are cleared in PWM Write Protect Status Register on page 1509. Only the first 16 bits (channel counter size) are significant.
CPRD: Channel Period

If the waveform is left-aligned, then the output waveform period depends on the channel counter source clock and can be calculated:
By using the PWM master clock (MCK) divided by an X given prescaler value (with X being 1, 2, 4, 8, 16, 32, 64, 128, 256, 512, or 1024). The resulting period formula will be:
( X CPRD ) ------------------------------MCK
( CRPD DIVA ) CRPD DIVB ) ------------------------------------------ or ( -----------------------------------------MCK MCK
If the waveform is center-aligned, then the output waveform period depends on the channel counter source clock and can be calculated:
( 2 X CPRD ) -----------------------------------------MCK
( 2 CPRD DIVA ) 2 CPRD DIVB ) ----------------------------------------------------- or ( ----------------------------------------------------MCK MCK

1518
49.7.40 PWM Channel Period Update Register

31 23
PWM_CPRDUPDx [x=0..3] 0xF002C210 [0], 0xF002C230 [1], 0xF002C250 [2], 0xF002C270 [3] Write-only
30 22 29 21 28 20 CPRDUPD 27 19 26 18 25 17 24 16
15
14
13
12 CPRDUPD
11
10
4 CPRDUPD
This register can only be written if the bits WPSWS3 and WPHWS3 are cleared in PWM Write Protect Status Register on page 1509. This register acts as a double buffer for the CPRD value. This prevents an unexpected waveform when modifying the waveform period. Only the first 16 bits (channel counter size) are significant.
CPRDUPD: Channel Period Update

If the waveform is left-aligned, then the output waveform period depends on the channel counter source clock and can be calculated:
( X CPRDUPD ) --------------------------------------------MCK
( CRPDUPD DIVA ) ( CRPDUPD DIVB ) -------------------------------------------------------or -------------------------------------------------------MCK MCK
If the waveform is center-aligned, then the output waveform period depends on the channel counter source clock and can be calculated:
( 2 X CPRDUPD ) ------------------------------------------------------MCK
( 2 CPRDUPD DIVA ) 2 CPRDUPD DIVB ) ------------------------------------------------------------------ or ( -----------------------------------------------------------------MCK MCK

1519
49.7.41 PWM Channel Counter Register

31 23
PWM_CCNTx [x=0..3] 0xF002C214 [0], 0xF002C234 [1], 0xF002C254 [2], 0xF002C274 [3] Read-only
30 22 29 21 28 20 CNT 27 19 26 18 25 17 24 16
15
14
13
12 CNT
11
10
4 CNT
Only the first 16 bits (channel counter size) are significant.
CNT: Channel Counter Register

Channel counter value. This register is reset when:
The channel is enabled (writing CHIDx in the PWM_ENA register). The channel counter reaches CPRD value defined in the PWM_CPRDx register if the waveform is left aligned.

1520
49.7.42 PWM Channel Dead Time Register

31
PWM_DTx [x=0..3] 0xF002C218 [0], 0xF002C238 [1], 0xF002C258 [2], 0xF002C278 [3] Read-write
30 29 28 DTL 27 26 25 24
23
22
21
20 DTL
19
18
17
16
15
14
13
12 DTH
11
10
4 DTH
This register can only be written if the bits WPSWS4 and WPHWS4 are cleared in PWM Write Protect Status Register on page 1509. Only the first 12 bits (dead-time counter size) of fields DTH and DTL are significant.
DTH: Dead-Time Value for PWMHx Output

Defines the dead-time value for PWMHx output. This value must be defined between 0 and CPRD-CDTY (PWM_CPRx and PWM_CDTYx).
DTL: Dead-Time Value for PWMLx Output

Defines the dead-time value for PWMLx output. This value must be defined between 0 and CDTY (PWM_CDTYx).

1521
49.7.43 PWM Channel Dead Time Update Register

31
PWM_DTUPDx [x=0..3] 0xF002C21C [0], 0xF002C23C [1], 0xF002C25C [2], 0xF002C27C [3] Write-only
30 29 28 DTLUPD 27 26 25 24
23
22
21
20 DTLUPD
19
18
17
16
15
14
13
12 DTHUPD
11
10
4 DTHUPD
This register can only be written if the bits WPSWS4 and WPHWS4 are cleared in PWM Write Protect Status Register on page 1509. This register acts as a double buffer for the DTH and DTL values. This prevents an unexpected waveform when modifying the dead-time values. Only the first 12 bits (dead-time counter size) of fields DTHUPD and DTLUPD are significant.
DTHUPD: Dead-Time Value Update for PWMHx Output

Defines the dead-time value for PWMHx output. This value must be defined between 0 and CPRD-CDTY (PWM_CPRx and PWM_CDTYx). This value is applied only at the beginning of the next channel x PWM period.
DTLUPD: Dead-Time Value Update for PWMLx Output

Defines the dead-time value for PWMLx output. This value must be defined between 0 and CDTY (PWM_CDTYx). This value is applied only at the beginning of the next channel x PWM period.

1522
50.
50.1
Analog-to-Digital Converter (ADC)

Description
The ADC is based on a 12-bit Analog-to-Digital Converter (ADC) managed by an ADC Controller. Refer to the Block Diagram: Figure 50-1. It also integrates a 12-to-1 analog multiplexer, making possible the analog-to-digital conversions of 12 analog lines. The conversions extend from 0V to the voltage carried on pin ADVREF. Conversion results are reported in a common register for all channels, as well as in a channel-dedicated register. Software trigger, external trigger on rising edge of the ADTRG pin or internal triggers from Timer Counter output(s) are configurable. The comparison circuitry allows automatic detection of values below a threshold, higher than a threshold, in a given range or outside the range, thresholds and ranges being fully configurable. The ADC Controller internal fault output is directly connected to PWM Fault input. This input can be asserted by means of comparison circuitry in order to immediately put the PWM outputs in a safe state (pure combinational path). The ADC also integrates a Sleep Mode and a conversion sequencer and connects with a DMA channel. These features reduce both power consumption and processor intervention. This ADC has a selectable single-ended or fully differential input and benefits from a 2-bit programmable gain. A digital error correction circuit based on the multi-bit redundant signed digit (RSD) algorithm is employed in order to reduce INL and DNL errors. Finally, the user can configure ADC timings, such as Startup Time and Tracking Time. This ADC Controller includes a Resistive Touchscreen Controller. It supports 4-wire and 5-wire technologies.

1523
50.2
z z z z z z
12-bit Resolution 1 MHz Conversion Rate Wide Range Power Supply Operation Selectable Single Ended or Differential Input Voltage Programmable Gain For Maximum Full Scale Input Range 0 - VDD Resistive 4-wire and 5-wire Touchscreen Controller
z z z
Position and Pressure Measurement for 4-wire screens Position Measurement for 5-wire screens Average of up to 8 measures for noise filtering
z z z z
Programmable Pen Detection sensitivity Integrated Multiplexer Offering Up to 12 Independent Analog Inputs Individual Enable and Disable of Each Channel Hardware or Software Trigger
z z z z z
External Trigger Pin Timer Counter Outputs (Corresponding TIOA Trigger) Internal Trigger Counter Trigger on Pen Contact Detection PWM Event Line
z z z z
Drive of PWM Fault Input DMA Support Possibility of ADC Timings Configuration Two Sleep Modes and Conversion Sequencer
z z
Automatic Wake-up on Trigger and Back to Sleep Mode after Conversions of all Enabled Channels Possibility of Customized Channel Sequence Power Down Capability
Standby Mode for Fast Wake-up Time Response

z
z z
Automatic Window Comparison of Converted Values Write Protect Registers

1524
50.3
Block Diagram
Figure 50-1. Analog-to-Digital Converter Block Diagram

Timer Counter Channels
PMC MCK
ADC Controller
ADTRG
Trigger Selection
ADC Interrupt
ADC cell
Control Logic
Interrupt Controller
VDDANA ADVREF
SOC
System Bus User Interface DMA
AD0/XP/UL
Touchscreen Analog Inputs
AD1/XM/UR AD2/YP/LL AD3/YM/Sense AD4/LR PIO AD-
Touchscreen Switches
1 2 3 4
VIN+ VIN- OFFSET S/H
Cyclic Pipeline 12-bit Analog-to-Digital Converter
Peripheral Bridge
PGA
APB
Other Analog Inputs
AD-
AD-
CHx
GND
50.4
Signal Description
Description Analog power supply Reference voltage Analog input channels External trigger
Table 50-1. ADC Pin Description Pin Name VDDANA ADVREF AD0 - AD11 ADTRG

1525
50.5

The ADC Controller is not continuously clocked. The programmer must first enable the ADC Controller MCK in the Power Management Controller (PMC) before using the ADC Controller. However, if the application does not require ADC operations, the ADC Controller clock can be stopped when not needed and restarted when necessary. Configuring the ADC Controller does not require the ADC Controller clock to be enabled.

The ADC interrupt line is connected on one of the internal sources of the Interrupt Controller. Using the ADC interrupt requires the interrupt controller to be programmed first.
Table 50-2. Peripheral IDs Instance ADC ID 29
50.5.3 Analog Inputs

The analog input pins can be multiplexed with PIO lines. In this case, the assignment of the ADC input is automatically done as soon as the corresponding channel is enabled by writing the register ADC_CHER. By default, after reset, the PIO line is configured as input with its pull-up enabled and the ADC input is connected to the GND.
50.5.4 I/O Lines

The pin ADTRG may be shared with other peripheral functions through the PIO Controller. In this case, the PIO Controller should be set accordingly to assign the pin ADTRG to the ADC function.
Table 50-3. I/O Lines Instance ADC ADC ADC ADC ADC ADC ADC ADC ADC ADC ADC ADC ADC Signal ADTRG AD0 AD1 AD2 AD3 AD4 AD5 AD6 AD7 AD8 AD9 AD10 AD11 I/O Line PD19 PD20 PD21 PD22 PD23 PD24 PD25 PD26 PD27 PD28 PD29 PD30 PD31 Peripheral A A A A A A A A A A A A A

1526
50.5.5 Timer Triggers

Timer Counters may or may not be used as hardware triggers depending on user requirements. Thus, some or all of the timer counters may be unconnected.
50.5.6 PWM Event Line PWM Event Lines may or may not be used as hardware triggers depending on user requirements. 50.5.7 Fault Output The ADC Controller has the FAULT output connected to the FAULT input of PWM. Please refer to Section 50.7.11 Fault Output and implementation of the PWM in the product. 50.5.8 Conversion Performances
For performance and electrical characteristics of the ADC, see the product DC Characteristics section.

1527
50.6
50.6.1 Analog-to-digital Conversion

The ADC uses the ADC Clock to perform conversions. Converting a single analog value to a 12-bit digital data requires Tracking Clock cycles as defined in the field TRACKTIM of the ADC Mode Register on page 1555. The ADC Clock frequency is selected in the PRESCAL field of the Mode Register (ADC_MR). The tracking phase starts during the conversion of the previous channel. If the tracking time is longer than the conversion time, the tracking phase is extended to the end of the previous conversion. The ADC clock range is between MCK/2, if PRESCAL is 0, and MCK/512, if PRESCAL is set to 255 (0xFF). PRESCAL must be programmed in order to provide an ADC clock frequency according to the parameters given in the product Electrical Characteristics section.
Figure 50-2. Sequence of ADC conversions when Tracking time > Conversion time
ADCClock Trigger event (Hard or Soft)
ADC_ON Commands from controller to analog cell
ADC_Start
ADC_SEL
CH0
CH1
CH2
LCDR
CH0
CH1
DRDY
Start Up Transfer Period Time (and tracking of CH0)
Conversion of CH0 Tracking of CH1
Transfer Period
Conversion of CH1 Tracking of CH2

1528
Figure 50-3. Sequence of ADC conversions when Tracking time < Conversion time
Read the ADC_LCDR
ADCClock Trigger event (Hard or Soft)
ADC_ON Commands from controller to analog cell
ADC_Start
ADC_SEL
CH0
CH1
CH2
CH3
LCDR
CH0
CH1
CH2
DRDY
Start Up Time & Tracking of CH0
Transfer Period
Conversion of CH0 & Tracking of CH1
Transfer Period
Transfer Period
50.6.2 Conversion Reference

The conversion is performed on a full range between 0V and the reference voltage pin ADVREF. Analog inputs between these voltages convert to values based on a linear conversion.
50.6.3 Conversion Resolution 50.6.4 The ADC supports 12-bit resolutions.Conversion Results
When a conversion is completed, the resulting 12-bit digital value is stored in the Channel Data Register (ADC_CDRx) of the current channel and in the ADC Last Converted Data Register (ADC_LCDR). By setting the TAG option in the ADC_EMR, the ADC_LCDR presents the channel number associated to the last converted data in the CHNB field. The channel EOC bit in the Status Register (ADC_SR) is set and the DRDY is set. In the case of a connected DMA channel, DRDY rising triggers a data request. In any case, either EOC and DRDY can trigger an interrupt. Reading one of the ADC_CDR registers clears the corresponding EOC bit. Reading ADC_LCDR clears the DRDY bit.

1529
Figure 50-4. EOCx and DRDY Flag Behavior

Write the ADC_CR with START = 1 Write the ADC_CR with START = 1
Read the ADC_CDRx
Read the ADC_LCDR
CHx (ADC_CHSR) EOCx (ADC_SR)
DRDY (ADC_SR)
If the ADC_CDR is not read before further incoming data is converted, the corresponding Overrun Error (OVREx) flag is set in the Overrun Status Register (ADC_OVER). Likewise, new data converted when DRDY is high sets the GOVRE bit (General Overrun Error) in ADC_SR. The OVREx flag is automatically cleared when ADC_OVER is read, and GOVRE flag is automatically cleared when ADC_SR is read.

1530
Figure 50-5. GOVRE and OVREx Flag Behavior
Trigger event CH0 (ADC_CHSR) CH1 (ADC_CHSR) ADC_LCDR ADC_CDR0 ADC_CDR1 Undefined Data Undefined Data Undefined Data Data A Data A Data B Data C
Data C Data B
EOC0 (ADC_SR)
Conversion A
Conversion C
Read ADC_CDR0
EOC1 (ADC_SR)
Conversion B
Read ADC_CDR1
GOVRE (ADC_SR) DRDY (ADC_SR)
Read ADC_SR
Read ADC_OVER OVRE0 (ADC_OVER) OVRE1 (ADC_OVER)
Warning: If the corresponding channel is disabled during a conversion or if it is disabled and then reenabled during a conversion, its associated data and its corresponding EOC and OVRE flags in ADC_SR are unpredictable.

1531
50.6.5 Conversion Triggers

Conversions of the active analog channels are started with a software or hardware trigger. The software trigger is provided by writing the Control Register (ADC_CR) with the START bit at 1. The hardware trigger can be one of the TIOA outputs of the Timer Counter channels, PWM Event line, or the external trigger input of the ADC (ADTRG). The hardware trigger is selected with the TRGSEL field in the Mode Register (ADC_MR). The selected hardware trigger is enabled if TRGMOD=1,2 or 3 in ADC Trigger Register. The minimum time between 2 consecutive trigger events must be strictly greater than the duration time of the longest conversion sequence according to configuration of registers ADC_MR, ADC_CHSR, ADC_SEQR1, ADC_SEQR2. The hardware trigger can be selected by the TRGMOD field in the ADC Trigger Register between:
z
Any edge, either rising or falling or both, detected on the external trigger pin, TSADTRG. The hardware trigger source is selected with the TRGSEL field in the Mode Register (ADC_MR). The selected hardware trigger is enabled if TRGMOD=1,2 or 3 in ADC Trigger Register. The Pen Detect, depending on how the PENDET bit is set in the ADC Touchscreen Mode Register. A continuous trigger, meaning the ADC Controller restarts the next sequence as soon as it finishes the current one A periodic trigger, which is defined by programming the TRGPER field in the ADC Trigger Register.
z z z
The minimum time between 2 consecutive trigger events must be strictly greater than the duration time of the longest conversion sequence according to configuration of registers ADC_MR, ADC_CHSR, ADC_SEQR1, ADC_SEQR2, ADC_TSMR. If a hardware trigger is selected, the start of a conversion is triggered after a delay starting at each rising edge of the selected signal. Due to asynchronous handling, the delay may vary in a range of 2 MCK clock periods to 1 ADC clock period.
trigger start delay
If one of the TIOA outputs is selected, the corresponding Timer Counter channel must be programmed in Waveform Mode. Only one start command is necessary to initiate a conversion sequence on all the channels. The ADC hardware logic automatically performs the conversions on the active channels, then waits for a new request. The Channel Enable (ADC_CHER) and Channel Disable (ADC_CHDR) Registers permit the analog channels to be enabled or disabled independently. If the ADC is used with a DMA, only the transfers of converted data from enabled channels are performed and the resulting data buffers should be interpreted accordingly.
50.6.6 Sleep Mode and Conversion Sequencer

The ADC Sleep Mode maximizes power saving by automatically deactivating the ADC when it is not being used for conversions. Sleep Mode is selected by setting the SLEEP bit in the Mode Register ADC_MR. The Sleep mode is automatically managed by a conversion sequencer, which can automatically process the conversions of all channels at lowest power consumption. This mode can be used when the minimum period of time between 2 successive trigger events is greater than the startup period of Analog-Digital converter (See the product ADC Characteristics section). When a start conversion request occurs, the ADC is automatically activated. As the analog cell requires a start-up time, the logic waits during this time and starts the conversion on the enabled channels. When all conversions are complete, the ADC is deactivated until the next trigger. Triggers occurring during the sequence are not taken into account. A fast wake-up mode is available in the ADC Mode Register (ADC_MR) as a compromise between power saving strategy and responsiveness. Setting the FWUP bit to 1 enables the fast wake-up mode. In fast wake-up mode the ADC cell is not fully deactivated while no conversion is requested, thereby providing less power saving but faster wake-up.

1532
The conversion sequencer allows automatic processing with minimum processor intervention and optimized power consumption. Conversion sequences can be performed periodically using the internal timer (ADC_TRGR register) or the PWM event line. The periodic acquisition of several samples can be processed automatically without any intervention of the processor thanks to the DMA. The sequence can be customized by programming the Sequence Channel Registers, ADC_SEQR1 and ADC_SEQR2 and setting to 1 the USEQ bit of the Mode Register (ADC_MR). The user can choose a specific order of channels and can program up to 12 conversions by sequence. The user is totally free to create a personal sequence, by writing channel numbers in ADC_SEQR1 and ADC_SEQR2. Not only can channel numbers be written in any sequence, channel numbers can be repeated several times. Only enabled sequence bitfields are converted, consequently to program a 15-conversion sequence, the user can simply put a disable in ADC_CHSR[15], thus disabling the 16THCH field of ADC_SEQR2. If all ADC channels (i.e. 12) are used on an application board, there is no restriction of usage of the user sequence. But as soon as some ADC channels are not enabled for conversion but rather used as pure digital inputs, the respective indexes of these channels cannot be used in the user sequence fields (ADC_SEQR1, ADC_SEQR2 bitfields). For example, if channel 4 is disabled (ADC_CSR[4] = 0), ADC_SEQR1, ADC_SEQR2 register bitfields USCH1 up to USCH12 must not contain the value 4. Thus the length of the user sequence may be limited by this behavior. As an example, if only 4 channels over 12 (CH0 up to CH3) are selected for ADC conversions, the user sequence length cannot exceed 4 channels. Each trigger event may launch up to 4 successive conversions of any combination of channels 0 up to 3 but no more (i.e. in this case the sequence CH0, CH0, CH1, CH1, CH1 is impossible). A sequence that repeats several times the same channel requires more enabled channels than channels actually used for conversion. For example, a sequence like CH0, CH0, CH1, CH1 requires 4 enabled channels (4 free channels on application boards) whereas only CH0, CH1 are really converted. Note: The reference voltage pins always remain connected in normal mode as in sleep mode.
50.6.7 Comparison Window

The ADC Controller features automatic comparison functions. It compares converted values to a low threshold or a high threshold or both, according to the CMPMODE function chosen in the Extended Mode Register (ADC_EMR). The comparison can be done on all channels or only on the channel specified in CMPSEL field of ADC_EMR. To compare all channels the CMP_ALL parameter of ADC_EMR should be set. Moreover a filtering option can be set by writing the number of consecutive comparison errors needed to raise the flag. This number can be written and read in the CMPFILTER field of ADC_EMR. The flag can be read on the COMPE bit of the Interrupt Status Register (ADC_ISR) and can trigger an interrupt. The High Threshold and the Low Threshold can be read/write in the Comparison Window Register (ADC_CWR). If the comparison window is to be used with LOWRES bit in ADC_MR set to 1, the thresholds do not need to be adjusted as adjustment will be done internally. Whether or not the LOWRES bit is set, thresholds must always be configured in consideration of the maximum ADC resolution.
50.6.8 Differential Inputs

The ADC can be used either as a single ended ADC (DIFF bit equal to 0) or as a fully differential ADC (DIFF bit equal to 1) as shown in Figure 50-6. By default, after a reset, the ADC is in single ended mode. If ANACH is set in ADC_MR the ADC can apply a different mode on each channel. Otherwise the parameters of CH0 are applied to all channels. The same inputs are used in single ended or differential mode. In single ended mode, inputs are managed by a 16:1 channels analog multiplexer. In the fully differential mode, inputs are managed by an 8:1 channels analog multiplexer. See Table 50-4 and Table 50-5.

1533
Table 50-4. Input Pins and Channel Number in Single Ended Mode Input Pins AD0 AD1 AD2 AD3 AD4 AD5 AD6 AD7 AD8 AD9 AD10 AD11 AD12 AD13 AD14 AD15 Channel Number CH0 CH1 CH2 CH3 CH4 CH5 CH6 CH7 CH8 CH9 CH10 CH11 CH12 CH13 CH14 CH15
Table 50-5. Input Pins and Channel Number In Differential Mode Input Pins AD0-AD1 AD2-AD3 AD4-AD5 AD6-AD7 AD8-AD9 AD10-AD11 AD12-AD13 AD14-AD15 Channel Number CH0 CH2 CH4 CH6 CH8 CH10 CH12 CH14

1534
50.6.9 Input Gain and Offset

The ADC has a built in Programmable Gain Amplifier (PGA) and Programmable Offset. The Programmable Gain Amplifier can be set to gains of 1/2, 1, 2 and 4. The Programmable Gain Amplifier can be used either for single ended applications or for fully differential applications. If ANACH is set in ADC_MR the ADC can apply different gain and offset on each channel. Otherwise the parameters of CH0 are applied to all channels. The gain is configurable through the GAIN bit of the Channel Gain Register (ADC_CGR) as shown in Table 50-6.
Table 50-6. Gain of the Sample and Hold Unit: GAIN Bits and DIFF Bit. GAIN<0:1> 00 01 10 11 GAIN (DIFF = 0) 1 1 2 4 GAIN (DIFF = 1) 0.5 1 2 2
To allow full range, analog offset of the ADC can be configured by the OFFSET bit of the Channel Offset Register (ADC_COR). The Offset is only available in Single Ended Mode.
Table 50-7. Offset of the Sample and Hold Unit: OFFSET DIFF and Gain (G) OFFSET Bit 0 1 OFFSET (DIFF = 0) 0 0 (G-1)Vrefin/2 OFFSET (DIFF = 1)

1535
Figure 50-6. Analog Full Scale Ranges in Single Ended/Differential Applications Versus Gain and Offset
single ended se0fd1=0

vrefin
VIN+ VIN+
fully differential se0fd1=1
gain=0.5 (00)
same as gain=1
()vrefin
VIN-
0 vrefin
()vrefin
VIN+
VIN+
()vrefin
gain=1 (01)
VIN()vrefin
0 vrefin
offset=1
offset=0
()vrefin (5/8)vrefin
gain=2 (10)
VIN+ VIN+
()vrefin (3/8)vrefin ()vrefin
VIN+ VIN-
0 vrefin
offset=1
VIN+
offset=0
(5/8)vrefin ()vrefin (3/8)vrefin
same as gain=2
VIN+ VIN-
gain=4 (11)
VIN+
()vrefin (1/8)vrefin

1536
50.6.10 ADC Timings

Each ADC has its own minimal Startup Time that is programmed through the field STARTUP in the Mode Register, ADC_MR. A minimal Tracking Time is necessary for the ADC to guarantee the best converted final value between two channel selections. This time has to be programmed through the TRACKTIM bit field in the Mode Register, ADC_MR. When the gain, offset or differential input parameters of the analog cell change between two channels, the analog cell may need a specific settling time before starting the tracking phase. In that case, the controller automatically waits during the settling time defined in the ADC Mode Register. Obviously, if the ANACH option is not set, this time is unused. Warning: No input buffer amplifier to isolate the source is included in the ADC. This must be taken into consideration to program a precise value in the TRACKTIM field. See the product ADC Characteristics section.
50.6.11 Automatic Calibration

The ADC features an automatic calibration (AUTOCALIB) mode for gain errors (calibration). The automatic calibration sequence can be started at any time writing to '1' the AUTOCAL bit of the ADC Control Register. The automatic calibration sequence requires a software reset command (SWRST in the ADC_CR register) prior to write AUTOCAL bit. The end of calibration sequence is given by the EOCAL bit in the interrupt status register (ADC_ISR), and an interrupt is generated if EOCAL interrupt has been enabled (ADC_IER register). The calibration sequence will perform an automatic calibration on all enabled channels. The channels required for conversion do not need to be all enabled during the calibration process if they are programmed with the same gain. Only channels with different gain settings need to be enabled. The gain settings of all enabled channels must be set before starting the AUTOCALIB sequence. If the gain settings (ADC_CGR and ADC_COR registers) for a given channel are changed, the AUTOCALIB sequence must then be started again. The calibration data (on one or more enabled channels) is stored in the internal ADC memory. Then, when a new conversion is started (on one or more enabled channels), the converted value (in ADC_LCDR or ADC_CDRx registers) is a calibrated value. Autocalibration is for settings, not for channels. Therefore, if a specific combination of gain has been already calibrated, and a new channel with the same settings is enabled after the initial calibration, there is no need to restart a calibration. If different enabled channels have different gain settings, the corresponding channels must be enabled before starting the calibration. If a software reset is performed (SWRST bit in ADC_CR) or after power up (or wake-up from Backup mode), the calibration data in the ADC memory is lost. Changing the ADC running mode (in ADC_CR register) does not affect the calibration data. Changing the ADC reference voltage (ADVREF pin) requires a new calibration sequence. For calibration time, gain error after calibration, refer to the 12-bit ADC electrical characteristics section of the product.

1537
50.7
Touchscreen
50.7.1 Touchscreen Mode

The TSMODE parameter of ADC Touchscreen Mode Register is used to enable/disable the Touchscreen functionality, to select the type of screen (4-wire or 5-wire) and, in the case of a 4-wire screen, to activate (or not) the pressure measurement. In 4-wire mode, channel 0, 1, 2 and 3 must not be used for classic ADC conversions. Likewise, in 5-wire mode, channel 0, 1, 2, 3, and 4 must not be used for classic ADC conversions.
50.7.2 4-wire Resistive Touchscreen Principles

A resistive touchscreen is based on two resistive films, each one being fitted with a pair of electrodes, placed at the top and bottom on one film, and on the right and left on the other. In between, there is a layer acting as an insulator, but also enables contact when you press the screen. This is illustrated in Figure 50-7. The TSADC controller has the ability to perform without external components:
z z z
Position Measurement Pressure Measurement Pen Detection
Figure 50-7. Touchscreen Position Measurement

Pen Contact XP
YM
YP
XM
XP
VDD
YP
VDD
YP
XP
Volt XM Vertical Position Detection YM
Volt
GND
GND
Horizontal Position Detection
50.7.3 4-wire Position Measurement Method

As shown in Figure 50-7, to detect the position of a contact, a supply is first applied from top to bottom. Due to the linear resistance of the film, there is a voltage gradient from top to bottom. When a contact is performed on the screen, the voltage propagates at the point the two surfaces come into contact with the second film. If the input impedance on the right and left electrodes sense is high enough, the film does not affect this voltage, despite its resistive nature. For the horizontal direction, the same method is used, but by applying supply from left to right. The range depends on the supply voltage and on the loss in the switches that connect to the top and bottom electrodes. In an ideal world (linear, with no loss through switches), the horizontal position is equal to:

1538
VYM / VDD or VYP / VDD. The implementation with on-chip power switches is shown in Figure 50-8. The voltage measurement at the output of the switch compensates for the switches loss. It is possible to correct for switch loss by performing the operation: [VYP - VXM] / [VXP - VXM]. This requires additional measurements, as shown in Figure 50-8.
Figure 50-8. Touchscreen Switches Implementation
XP VDDANA
0
XM GND
1
YP To the ADC VDDANA
2
YM GND
VDDANA Switch Resistor
VDDANA Switch Resistor
XP
YP
YP
XP
XM Switch Resistor
YM Switch Resistor
GND Horizontal Position Detection
GND Vertical Position Detection

1539
50.7.4 4-wire Pressure Measurement Method

The method to measure the pressure (Rp) applied to the touchscreen is based on the known resistance of the X-Panel resistance (Rxp). Three conversions (Xpos,Z1,Z2) are necessary to determine the value of Rp (Zaxis resistance). Rp = Rxp*(Xpos/1024)*[(Z2/Z1)-1]
Figure 50-9. Pressure Measurement
VDDANA Switch Resistor VDDANA Switch Resistor VDDANA
Open circuit YP XP
Switch Resistor
XP
YP
XP
YP
Rp
Rp
Rp
XM Switch Resistor
YM Open circuit
XM Switch Resistor
YM Open circuit
XM Switch Resistor GND Z2 Measure(Xp)
YM
GND XPos Measure(Yp)
GND Z1 Measure(Xp)

1540
50.7.5 5-wire Resistive Touchscreen Principles

To make a 5-wire touchscreen, a resistive layer with a contact point at each corner and a conductive layer are used. The 5-wire touchscreen differs from the 4-wire type mainly in that the voltage gradient is applied only to one layer, the resistive layer, while the other layer is the sense layer for both measurements. The measurement of the X position is obtained by biasing the upper left corner and lower left corner to VDDANA and the upper right corner and lower right to ground. To measure along the Y axis, bias the upper left corner and upper right corner to VDDANA and bias the lower left corner and lower right corner to ground.
Figure 50-10. 5-Wire principle
UL
Pen Contact
Resistive layer
UR
Sense
LL
Conductive Layer
LR UR
VDDANA for Yp GND for Xp Sense
UL
VDDANA
LL
VDDANA for Xp GND for Yp
LR
GND

1541
50.7.6 5-wire Position Measurement Method

In an application only monitoring clicks, 100 points per second is typically needed. For handwriting or motion detection, the number of measurements to consider is approximately 200 points per second. This must take into account that multiple measurements are included (over sampling, filtering) to compute the correct point. The 5-wire touchscreen panel works by applying a voltage at the corners of the resistive layer and measuring the vertical or horizontal resistive network with the sense input. The ADC converts the voltage measured at the point the panel is touched. A measurement of the Y position of the pointing device is made by:
z z z
Connecting Upper left (UL) and upper right (UR) corners to VDDANA Connecting Lower left (LL) and lower right (LR) corners to ground. The voltage measured is determined by the voltage divider developed at the point of touch (Yposition) and the SENSE input is converted by ADC. Connecting the upper left (UL) and lower left (LL) corners to ground Connecting the upper right and lower right corners to VDDANA. The voltage measured is determined by the voltage divider developed at the point of touch (Xposition) and the SENSE input is converted by ADC.
A measurement of the X position of the pointing device is made by:

z z z
Figure 50-11. Touchscreen Switches Implementation
UL UR
VDDANA
0
GND VDDANA GND
LL
VDDANA
2 3
To the ADC
Sense
LR
GND
UL
VDDANA
UR
VDDANA for Ypos GND for Xpos
Sense
LL
VDDANA for Xpos GND for Ypos
LR
GND

1542
50.7.7 Sequence and Noise Filtering

The ADC Controller can manage ADC conversions and Touchscreen measurement. On each trigger event the sequence of ADC conversions is performed as described in Section 50.6.6 Sleep Mode and Conversion Sequencer. The Touchscreen measure frequency can be specified in number of trigger events by writing the TSFREQ parameter in the ADC Touchscreen Mode Register. An internal counter counts triggers up to TSFREQ, and every time it rolls out, a Touchscreen sequence is appended to the classic ADC conversion sequence (see Figure 50-12). Additionally the user can average multiple Touchscreen measures by writing the TSAV parameter in the ADC Touchscreen Mode Register. This can be 1, 2, 4 or 8 measures performed on consecutive triggers as illustrated in Figure 50-12 bellow. Consequently, the TSFREQ parameter must be greater or equal to the TSAV parameter.
Figure 50-12. Insertion of Touchscreen sequences (TSFREQ = 2; TSAV = 1)
Trigger event
ADC_SEL
C T
C T
C -
C T
C T
C: Classic ADC Conversion Sequence

XRDY
T: Touchscreen Sequence
Read the ADC_XPOSR
Read the ADC_XPOSR
YRDY
Read the ADC_YPOSR
Read the ADC_YPOSR
50.7.8 Measured Values, Registers and Flags

As soon as the controller finishes the Touchscreen sequence, XRDY, YRDY and PRDY are set and can generate an interrupt. These flags can be read in the ADC Interrupt Status Register. They are reset independently by reading in ADC_XPOSR, ADC_YPOSR and ADC_PRESSR. for classic ADC conversions. The ADC Touchscreen X Position Register presents XPOS (VX - VXmin) on its LSB and XSCALE (VXMAX - VXmin) aligned on the 16th bit. The ADC Touchscreen Y Position Register presents YPOS (VY - VYmin) on its LSB and YSCALE (VYMAX - VYmin) aligned on the 16th bit. To improve the quality of the measure, the user must calculate: XPOS/XSCALE and YPOS/YSCALE. VXMAX, VXmin, VYMAX, and VYmin are measured at the first start up of the controller. These values can change during use, so it can be necessary to refresh them. Refresh can be done by writing 1 in the CALIB field of the control register (ADC_CR). The ADC Touchscreen Pressure Register presents Z1 on its LSB and Z2 aligned on the 16th bit. See Section 50.7.4 to know how use them.

1543
50.7.9 Pen Detect Method

When there is no contact, it is not necessary to perform a conversion. However, it is important to detect a contact by keeping the power consumption as low as possible. The implementation polarizes one panel by closing the switch on (XP/UL) and ties the horizontal panel by an embedded resistor connected to YM / Sense. This resistor is enabled by a fifth switch. Since there is no contact, no current is flowing and there is no related power consumption. As soon as a contact occurs, a current is flowing in the Touchscreen and a Schmitt trigger detects the voltage in the resistor. The Touchscreen Interrupt configuration is entered by programming the PENDET bit in the ADC Touchscreen Mode Register. If this bit is written at 1, the controller samples the pen contact state when it is not converting and waiting for a trigger. To complete the circuit, a programmable debouncer is placed at the output of the Schmitt trigger. This debouncer is programmable up to 215 ADC clock periods. The debouncer length can be selected by programming the field PENDBC in ADC Touchscreen Mode Register. Due to the analog switchs structure, the debouncer circuitry is only active when no conversion (Touchscreen or classic ADC channels) is in progress. Thus, if the time between the end of a conversion sequence and the arrival of the next trigger event is lower than the debouncing time configured on PENDBC, the debouncer will not detect any contact.
Figure 50-13. Touchscreen Pen Detect
X+/UL X-/UR
VDDANA
0
GND VDDANA GND
1 2 3 4
PENDBC Debouncer Pen Interrupt To the ADC
Y+/LL
VDDANA
Y-/SENSE
GND
LR
GND
GND
The Touchscreen Pen Detect can be used to generate an ADC interrupt to wake up the system. The Pen Detect generates two types of status, reported in the ADC Interrupt Status Register: z The PEN bit is set as soon as a contact exceeds the debouncing time as defined by PENDBC and remains set until ADC_SR is read. z The NOPEN bit is set as soon as no current flows for a time over the debouncing time as defined by PENDBC and remains set until ADC_SR is read. Both bits are automatically cleared as soon as the Status Register (ADC_SR) is read, and can generate an interrupt by writing the ADC Interrupt Enable Register. Moreover, the rising of either one of them clears the other, they cannot be set at the same time. The PENS bit of the ADC_SR indicates the current status of the pen contact.

1544
50.7.10 Buffer Structure

The DMA read channel is triggered each time a new data is stored in ADC_LCDR register. The same structure of data is repeatedly stored in ADC_LCDR register each time a trigger event occurs. Depending on user mode of operation (ADC_MR, ADC_CHSR, ADC_SEQR1, ADC_SEQR2, ADC_TSMR) the structure differs. Each data read to DMA buffer, carried on a half-word (16-bit), consists of last converted data right aligned and when TAG is set in ADC_EMR register, the 4 most significant bits are carrying the channel number thus allowing an easier post-processing in the DMA buffer or better checking the DMA buffer integrity. As soon as touchscreen conversions are required, the pen detection function may help the post-processing of the buffer. To get more details refer to Section 50.7.10.4 Pen Detection Status. 50.7.10.1 Classical ADC Channels Only When no touchscreen conversion is required (i.e. TSMODE = 0 in ADC_TSMR register), the structure of data within the buffer is defined by the ADC_MR, ADC_CHSR, ADC_SEQR1, ADC_SEQR2 registers. If the user sequence is not used (i.e. USEQ is cleared in ADC_MR register) then only the value of ADC_CHSR register defines the data structure. For each trigger event, enabled channels will be consecutively stored in ADC_LCDR register and automatically read to the buffer. When the user sequence is configured (i.e. USEQ is set in ADC_MR register) not only does ADC_CHSR register modify the data structure of the buffer, but ADC_SEQR1, ADC_SEQR2 registers may modify the data structure of the buffer as well.
Figure 50-14. Buffer Structure when TSMODE = 0
Assuming ADC_CHSR = 0x000_01600 ADC_EMR(TAG) = 1

trig.event1 DMA Buffer Structure
Assuming ADC_CHSR = 0x000_01600 ADC_EMR(TAG) = 0

DMA Transfer trig.event1 Base Address (BA) DMA Buffer Structure BA + 0x02 BA + 0x04 trig.event2 BA + 0x06 BA + 0x08 BA + 0x0A
5 6 8 5 6 8
ADC_CDR5 ADC_CDR6 ADC_CDR8 ADC_CDR5 ADC_CDR6 ADC_CDR8
0 0 0 0 0 0
ADC_CDR5 ADC_CDR6 ADC_CDR8 ADC_CDR5 ADC_CDR6 ADC_CDR8
trig.event2
trig.eventN
trig.eventN
5 6 8
ADC_CDR5 ADC_CDR6 ADC_CDR8
BA + [(N-1) * 6] BA + [(N-1) * 6]+ 0x02 BA + [(N-1) * 6]+ 0x04
0 0 0
ADC_CDR5 ADC_CDR6 ADC_CDR8

1545
50.7.10.2 Touchscreen Channels Only When only touchscreen conversions are required (i.e. TSMODE differs from 0 in ADC_TSMR register and ADC_CHSR equals 0), the structure of data within the buffer is defined by the ADC_TSMR register. When TSMODE = 1 or 3, each trigger event adds 2 half-words in the buffer (assuming TSAV = 0), first half-word being XPOS of ADC_XPOSR register then YPOS of ADC_YPOSR register. If TSAV/TSFREQ differs from 0, the data structure remains unchanged. Not all trigger events add data to the buffer. When TSMODE = 2, each trigger event adds 4 half-words to the buffer (assuming TSAV=0), first half-word being XPOS of ADC_XPOSR register followed by YPOS of ADC_YPOSR register and finally Z1 followed by Z2, both located in ADC_PRESSR register. When TAG is set (ADC_EMR), the CHNB field (4 most significant bit of the ADC_LCDR) register is set to 0 when XPOS is transmitted and set to 1 when YPOS is transmitted, allowing an easier post-processing of the buffer or better checking buffer integrity. In case 4-wire with pressure mode is selected, Z1 value is transmitted to the buffer along with tag set to 2 and Z2 is tagged with value 3. XSCALE and YSCALE (calibration values) are not transmitted to the buffer because they are supposed to be constant and moreover only measured at the very first start up of the controller or upon user request. There is no change in buffer structure whatever the value of PENDET bit configuration in ADC_TSMR register but it is recommended to use the pen detection function for buffer post-processing (refer to Pen Detection Status on page 1550).

1546
Figure 50-15. Buffer Structure when only touchscreen channels are enabled
Assuming ADC_TSMR(TSMOD) = 1 or 3 ADC_TSMR(TSAV) = 0 ADC_CHSR = 0x000_00000 , ADC_EMR(TAG) = 1

trig.event1 DMA Buffer Structure trig.event2
Assuming ADC_TSMR(TSMOD) =1 or 3 ADC_TSMR(TSAV) = 0 ADC_CHSR = 0x000_00000 , ADC_EMR(TAG) = 0

DMA Transfer trig.event1 Base Address (BA) DMA Buffer Structure BA + 0x02 trig.event2 BA + 0x04 BA + 0x06
0 1 0 1
ADC_XPOSR ADC_YPOSR ADC_XPOSR ADC_YPOSR
0 0 0 0
ADC_XPOSR ADC_YPOSR ADC_XPOSR ADC_YPOSR
trig.eventN
0 1
ADC_XPOSR ADC_YPOSR
BA + [(N-1) * 4]
trig.eventN
0 0
ADC_XPOSR ADC_YPOSR
BA + [(N-1) * 4]+ 0x02
Assuming ADC_TSMR(TSMOD) = 2 ADC_TSMR(TSAV) = 0 ADC_CHSR = 0x000_00000 , ADC_EMR(TAG) = 1

trig.event1
Assuming ADC_TSMR(TSMOD) = 2 ADC_TSMR(TSAV) = 0 ADC_CHSR = 0x000_00000 , ADC_EMR(TAG) = 0

trig.event1 DMA Transfer Base Address (BA) BA + 0x02 BA + 0x04 BA + 0x06 trig.event2 BA + 0x08 BA + 0x0A BA + 0x0C BA + 0x0E DMA Buffer Structure
0
DMA Buffer Structure
ADC_XPOSR ADC_YPOSR ADC_PRESSR(Z1) ADC_PRESSR(Z2) ADC_XPOSR ADC_YPOSR ADC_PRESSR(Z1) ADC_PRESSR(Z2)
0 0 0 0 0 0 0 0
ADC_XPOSR ADC_YPOSR ADC_PRESSR(Z1) ADC_PRESSR(Z2) ADC_XPOSR ADC_YPOSR ADC_PRESSR(Z1) ADC_PRESSR(Z2)
1 2 3 0 1 2 3
trig.event2
trig.eventN
0 1 2 3
ADC_XPOSR ADC_YPOSR ADC_PRESSR(Z1) ADC_PRESSR(Z2)
trig.eventN BA + [(N-1) * 8] BA + [(N-1) * 8]+ 0x02 BA + [(N-1) * 8]+ 0x04 BA + [(N-1) * 8]+ 0x06
0 0 0 0
ADC_XPOSR ADC_YPOSR ADC_PRESSR(Z1) ADC_PRESSR(Z2)

1547
50.7.10.3 Interleaved Channels When both classic ADC channels (CH4/CH5 up to CH12 are set in ADC_CHSR) and touchscreen conversions are required (TSMODE differs from 0 in ADC_TSMR register) the structure of the buffer differs according to TSAV and TSFREQ values. If TSFREQ differs from 0, not all events generate touchscreen conversions, therefore buffer structure is based on 2TSFREQ trigger events. Given a TSFREQ value, the location of touchscreen conversion results depends on TSAV value. When TSFREQ = 0, TSAV must equal 0. There is no change in buffer structure whatever the value of PENDET bit configuration in ADC_TSMR register but it is recommended to use the pen detection function for buffer post-processing (refer to Pen Detection Status on page 1550).

1548
Figure 50-16. Buffer Structure when classic ADC and touchscreen channels are interleaved
Assuming ADC_TSMR(TSMOD) = 1 ADC_TSMR(TSAV) = ADC_TSMR(TSFREQ) = 0 ADC_CHSR = 0x000_0100 , ADC_EMR(TAG) =1

trig.event1
Assuming ADC_TSMR(TSMOD) = 1 ADC_TSMR(TSAV) = ADC_TSMR(TSFREQ) = 0 ADC_CHSR = 0x000_0100 , ADC_EMR(TAG) = 0

trig.event1
8
DMA Buffer Structure trig.event2
ADC_CDR8 ADC_XPOSR ADC_YPOSR ADC_CDR8 ADC_XPOSR ADC_YPOSR
DMA Transfer Base Address (BA) BA + 0x02 BA + 0x04
0
DMA Buffer Structure trig.event2
0 1 8 0 1
0 0 0 0 0
BA + 0x06 BA + 0x08 BA + 0x0A
trig.eventN
trig.eventN
8 0 1
ADC_CDR8 ADC_XPOSR ADC_YPOSR
BA + [(N-1) * 6] BA + [(N-1) * 6]+ 0x02 BA + [(N-1) * 6]+ 0x04
0 0 0
ADC_CDR8 ADC_XPOSR ADC_YPOSR
Assuming ADC_TSMR(TSMOD) = 1 ADC_TSMR(TSAV) = 0 ADC_TSMR(TSFREQ) = 1 ADC_CHSR = 0x000_0100 , ADC_EMR(TAG) = 1

trig.event1
Assuming ADC_TSMR(TSMOD) = 1 ADC_TSMR(TSAV) = 1 ADC_TSMR(TSFREQ) = 1 ADC_CHSR = 0x000_0100 , ADC_EMR(TAG) = 1

trig.event1 trig.event2
8
DMA Buffer Structure trig.event2 trig.event3
ADC_CDR8 ADC_XPOSR ADC_YPOSR ADC_CDR8 ADC_CDR8 ADC_XPOSR ADC_YPOSR ADC_CDR8
DMA Transfer Base Address (BA) BA + 0x02 BA + 0x04 BA + 0x06 BA + 0x08 BA + 0x0A BA + 0x0c BA + 0x0e
8 8 0 1
ADC_CDR8 ADC_CDR8 ADC_XPOSR ADC_YPOSR ADC_CDR8 ADC_CDR8 ADC_XPOSR ADC_YPOSR
0 1 8 8 0
DMA Buffer Structure
trig.event3 trig.event4
8 8 0 1
trig.event4
1 8
trig.eventN
trig.eventN
8 0
trig.eventN+1
ADC_CDR8 ADC_XPOSR ADC_YPOSR ADC_CDR8
BA + [(N-1) * 8]
trig.eventN+1
8 8 0 1
ADC_CDR8 ADC_CDR8 ADC_XPOSR ADC_YPOSR
BA + [(N-1) * 8]+ 0x02 BA + [(N-1) * 8]+ 0x04 BA + [(N-1) * 8]+ 0x06
1 8

1549
50.7.10.4 Pen Detection Status If the pen detection measure is enabled (PENDET is set in ADC_TSMR register), the XPOS, YPOS, Z1, Z2 values transmitted to the buffer through ADC_LCDR register are cleared (including the CHNB field), if the PENS flag of ADC_ISR register is 0. When the PENS flag is set, XPOS, YPOS, Z1, Z2 are normally transmitted. Therefore, using pen detection together with tag function eases the post-processing of the buffer, especially to determine which touchscreen converted values correspond to a period of time when the pen was in contact with the screen. When the pen detection is disabled or the tag function is disabled, XPOS, YPOS, Z1, Z2 are normally transmitted without tag and no relationship can be found with pen status, thus post-processing may not be easy.
Figure 50-17. Buffer Structure with and without pen detection enabled
Assuming ADC_TSMR(TSMOD) = 1, PENDET = 1 ADC_TSMR(TSAV) = ADC_TSMR(TSFREQ) = 0 ADC_CHSR = 0x000_0100 , ADC_EMR(TAG) = 1

trig.event1
Assuming ADC_TSMR(TSMOD) = 1, PENDET = 1 ADC_TSMR(TSAV) = ADC_TSMR(TSFREQ) = 0 ADC_CHSR = 0x000_0100 , ADC_EMR(TAG) =0

DMA Transfer Base Address (BA) BA + 0x02 BA + 0x04 trig.event2 BA + 0x08 BA + 0x0A trig.event1
8
DMA buffer Structure trig.event2
0
DMA buffer Structure
0 1 8 0 1
0 0 0 0 0
PENS = 1
trig.eventN
trig.eventN
8 0
trig.eventN+1
ADC_CDR8 0 0 ADC_CDR8 0 0 2 successive tags cleared => PENS = 0
BA + [(N-1) * 6] BA + [(N-1) * 6]+ 0x02 BA + [(N-1) * 6]+ 0x04
PENS = 1
BA + 0x06
0 0 0 0 0 0
ADC_CDR8 ADC_XPOSR* ADC_YPOSR* ADC_CDR8 ADC_XPOSR* ADC_YPOSR*
0 8 0 0
PENS = 0
PENS = 0
ADC_XPOSR*, ADC_YPOSR* can be any value when PENS = 0

1550
50.7.11 Fault Output

The ADC Controller internal fault output is directly connected to PWM fault input. Fault output may be asserted according to the configuration of ADC_EMR (Extended Mode Register) and ADC_CWR (Compare Window Register) and converted values. When the Compare occurs, the ADC fault output generates a pulse of one Master Clock Cycle to the PWM fault input. This fault line can be enabled or disabled within PWM. Should it be activated and asserted by the ADC Controller, the PWM outputs are immediately placed in a safe state (pure combinational path). Note that the ADC fault output connected to the PWM is not the COMPE bit. Thus the Fault Mode (FMOD) within the PWM configuration must be FMOD = 1.

To prevent any single software error that may corrupt ADC behavior, certain address spaces can be write-protected by setting the WPEN bit in the ADC Write Protect Mode Register (ADC_WPMR). If a write access to the protected registers is detected, then the WPVS flag in the ADC Write Protect Status Register (ADC_WPSR) is set and the field WPVSRC indicates in which register the write access has been attempted. The WPVS flag is automatically reset by reading the ADC Write Protect Status Register (ADC_WPSR). The protected registers are:
z z z z z z z z z z z z
ADC Mode Register on page 1555 ADC Channel Sequence 1 Register on page 1557 ADC Channel Sequence 2 Register on page 1558 ADC Channel Enable Register on page 1559 ADC Channel Disable Register on page 1560 ADC Extended Mode Register on page 1569 ADC Compare Window Register on page 1570 ADC Channel Gain Register on page 1571 ADC Channel Offset Register on page 1572 ADC Analog Control Register on page 1574 ADC Touchscreen Mode Register on page 1575 ADC Trigger Register on page 1580

1551
50.8
Analog-to-Digital Converter (ADC) User Interface

Any offset not listed in Table 50-8 must be considered as reserved.
Table 50-8. Register Mapping Offset 0x00 0x04 0x08 0x0C 0x10 0x14 0x18 0x1C 0x20 0x24 0x28 0x2C 0x30 0x34 0x38 0x3C 0x40 0x44 0x48 0x4C 0x50 0x54 ... 0x7C 0x80 - 0x90 0x94 0x98 - 0xAC 0xB0 0xB4 0xB8 0xBC 0xC0 0xC4 - 0xE0 0xE4 Control Register Mode Register Channel Sequence Register 1 Channel Sequence Register 2 Channel Enable Register Channel Disable Register Channel Status Register Reserved Last Converted Data Register Interrupt Enable Register Interrupt Disable Register Interrupt Mask Register Interrupt Status Register Reserved Reserved Overrun Status Register Extended Mode Register Compare Window Register Channel Gain Register Channel Offset Register Channel Data Register 0 Channel Data Register 1 ... Channel Data Register 11 Reserved Analog Control Register Reserved Touchscreen Mode Register Touchscreen X Position Register Touchscreen Y Position Register Touchscreen Pressure Register Trigger Register Reserved Write Protect Mode Register Register Name ADC_CR ADC_MR ADC_SEQR1 ADC_SEQR2 ADC_CHER ADC_CHDR ADC_CHSR ADC_LCDR ADC_IER ADC_IDR ADC_IMR ADC_ISR ADC_OVER ADC_EMR ADC_CWR ADC_CGR ADC_COR ADC_CDR0 ADC_CDR1 ... ADC_CDR11 ADC_ACR ADC_TSMR ADC_XPOSR ADC_YPOSR ADC_PRESSR ADC_TRGR ADC_WPMR Access Write-only Read-write Read-write Read-write Write-only Write-only Read-only Read-only Write-only Write-only Read-only Read-only Read-only Read-write Read-write Read-write Read-write Read-only Read-only ... Read-only Read-write Read-write Read-only Read-only Read-only Read-write Read-write Reset 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 ... 0x00000000 0x00000100 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000

1552
Table 50-8. Register Mapping (Continued) Offset 0xE8 0xEC - 0xF8 0xFC Register Write Protect Status Register Reserved Reserved Name ADC_WPSR Access Read-only Reset 0x00000000

1553
50.8.1 ADC Control Register

31 23 15 7
ADC_CR 0xF8018000 Write-only

30 22 14 6 29 21 13 5 28 20 12 4 27 19 11 3 AUTOCAL 26 18 10 2 TSCALIB 25 17 9 1 START 24 16 8 0 SWRST

0 = No effect. 1 = Resets the ADC simulating a hardware reset.
START: Start Conversion

0 = No effect. 1 = Begins analog-to-digital conversion.
TSCALIB: Touchscreen Calibration

0 = No effect. 1 = Programs screen calibration (VDD/GND measurement) The calibration sequence is performed during the next sequence when command is launched during an already started conversion sequence, or at the start of the second conversion sequence located after the TSCALIB command, if it is launched when no conversion is in progress (sleep mode, waiting a trigger event). TSCALIB measurement sequence does not affect the last data converted register (ADC_LDCR).
AUTOCAL: Automatic Calibration of ADC

0 = No effect. 1 = Launch an automatic calibration of the ADC cell on next sequence.

1554
50.8.2 ADC Mode Register

31 USEQ 23 ANACH 15
ADC_MR 0xF8018004 Read-write

30 22 14 29 TRANSFER 21 SETTLING 13 12 PRESCAL 7 6 FWUP 5 SLEEP 4
28
27
26 TRACKTIM
25
24
20
19
18 STARTUP
17
16
11
10
2 TRGSEL
This register can only be written if the WPEN bit is cleared in ADC Write Protect Mode Register on page 1581.
TRGSEL: Trigger Selection

Value 0 1 2 3 4 5 6 7 Name ADC_TRIG0 ADC_TRIG1 ADC_TRIG2 ADC_TRIG3 ADC_TRIG4 ADC_TRIG5 ADC_TRIG6 Description ADTRG TIOA0 TIOA1 TIOA2 PWM Event Line 0 PWM Event Line 1 Reserved Reserved
Note:
The trigger selection can be performed only if TRGMOD=1,2 or 3 in ADC Trigger Register.
SLEEP: Sleep Mode

Value 0 1 Name NORMAL SLEEP Description Normal Mode: The ADC Core and reference voltage circuitry are kept ON between conversions Sleep Mode: The wake-up time can be modified by programming FWUP bit
FWUP: Fast Wake Up

Value 0 1 Name OFF ON Description If SLEEP is 1 then both ADC Core and reference voltage circuitry are OFF between conversions If SLEEP is 1 then Fast Wake-up Sleep Mode: The Voltage reference is ON between conversions and ADC Core is OFF
PRESCAL: Prescaler Rate Selection ADCClock = MCK / ( (PRESCAL+1) * 2 )

1555
STARTUP: Start Up Time

Value 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Name SUT0 SUT8 SUT16 SUT24 SUT64 SUT80 SUT96 SUT112 SUT512 SUT576 SUT640 SUT704 SUT768 SUT832 SUT896 SUT960 Description 0 periods of ADCClock 8 periods of ADCClock 16 periods of ADCClock 24 periods of ADCClock 64 periods of ADCClock 80 periods of ADCClock 96 periods of ADCClock 112 periods of ADCClock 512 periods of ADCClock 576 periods of ADCClock 640 periods of ADCClock 704 periods of ADCClock 768 periods of ADCClock 832 periods of ADCClock 896 periods of ADCClock 960 periods of ADCClock
SETTLING: Analog Settling Time

Value 0 1 2 3 Name AST3 AST5 AST9 AST17 Description 3 periods of ADCClock 5 periods of ADCClock 9 periods of ADCClock 17 periods of ADCClock
ANACH: Analog Change

Value 0 1 Name NONE ALLOWED Description No analog change on channel switching: DIFF0, GAIN0 and OFF0 are used for all channels Allows different analog settings for each channel. See ADC_CGR and ADC_COR Registers
TRACKTIM: Tracking Time Tracking Time = (TRACKTIM + 1) * ADCClock periods. TRANSFER: Transfer Period This field must be programmed with value 2. USEQ: Use Sequence Enable
Value 0 1 Name NUM_ORDER REG_ORDER Description Normal Mode: The controller converts channels in a simple numeric order depending only on the channel index. User Sequence Mode: The sequence respects what is defined in ADC_SEQR1 and ADC_SEQR2 registers and can be used to convert several times the same channel.

1556
50.8.3 ADC Channel Sequence 1 Register

31
ADC_SEQR1 0xF8018008 Read-write

30 USCH8 29 28 27 26 USCH7 21 USCH6 20 19 18 USCH5 13 USCH4 7 6 USCH2 5 4 3 2 USCH1 12 11 10 USCH3 1 0 9 8 17 16 25 24
23
22
15
14
USCHx: User Sequence Number x

The sequence number x (USCHx) can be programmed by the Channel number CHy where y is the value written in this field. The allowed range is 0 up to 11. So it is only possible to use the sequencer from CH0 to CH11. This register activates only if ADC_MR(USEQ) field is set to 1. Any USCHx field is taken into account only if ADC_CHSR(CHx) register field reads logical 1 else any value written in USCHx does not add the corresponding channel in the conversion sequence. Configuring the same value in different fields leads to multiple samples of the same channel during the conversion sequence. This can be done consecutively, or not, according to user needs.

1557
50.8.4 ADC Channel Sequence 2 Register

31
ADC_SEQR2 0xF801800C Read-write

30 29 28 27 26 21 20 19 18 13 7 6 USCH10 5 4 3 2 USCH9 12 11 10 USCH11 1 0 9 8 17 16 25 24
23
22
15
14
USCHx: User Sequence Number x

The sequence number x (USCHx) can be programmed by the Channel number CHy where y is the value written in this field. The allowed range is 0 up to 11. So it is only possible to use the sequencer from CH0 to CH11. This register activates only if ADC_MR(USEQ) field is set to 1. Any USCHx field is taken into account only if ADC_CHSR(CHx) register field reads logical 1 else any value written in USCHx does not add the corresponding channel in the conversion sequence. Configuring the same value in different fields leads to multiple samples of the same channel during the conversion sequence. This can be done consecutively, or not, according to user needs.

1558
50.8.5 ADC Channel Enable Register

31 23 15 7 CH7
ADC_CHER 0xF8018010 Write-only

30 22 14 6 CH6 29 21 13 5 CH5 28 20 12 4 CH4 27 19 11 CH11 3 CH3 26 18 10 CH10 2 CH2 25 17 9 CH9 1 CH1 24 16 8 CH8 0 CH0
CHx: Channel x Enable

0 = No effect. 1 = Enables the corresponding channel. Note: if USEQ = 1 in ADC_MR register, CHx corresponds to the xth channel of the sequence described in ADC_SEQR1 and ADC_SEQR2.

1559
50.8.6 ADC Channel Disable Register

31 23 15 7 CH7
ADC_CHDR 0xF8018014 Write-only

CHx: Channel x Disable

0 = No effect. 1 = Disables the corresponding channel. Warning: If the corresponding channel is disabled during a conversion or if it is disabled then reenabled during a conversion, its associated data and its corresponding EOC and OVRE flags in ADC_SR are unpredictable.

1560
50.8.7 ADC Channel Status Register

31 23 15 7 CH7
ADC_CHSR 0xF8018018 Read-only

CHx: Channel x Status

0 = The corresponding channel is disabled. 1 = The corresponding channel is enabled.

1561
50.8.8 ADC Last Converted Data Register

31 23 15
ADC_LCDR 0xF8018020 Read-only

30 22 14 CHNB 7 6 5 4 LDATA 3 2 29 21 13 28 20 12 27 19 11 26 18 10 LDATA 1 0 25 17 9 24 16 8
LDATA: Last Data Converted

The analog-to-digital conversion data is placed into this register at the end of a conversion and remains until a new conversion is completed.
CHNB: Channel Number

Indicates the last converted channel when the TAG option is set to 1 in the ADC_EMR register. If the TAG option is not set, CHNB = 0.

1562
50.8.9 ADC Interrupt Enable Register

31 23 EOCAL 15 7 EOC7
ADC_IER 0xF8018024 Write-only

30 NOPEN 22 PRDY 14 6 EOC6 29 PEN 21 YRDY 13 5 EOC5 28 20 XRDY 12 4 EOC4 27 19 11 EOC11 3 EOC3 26 COMPE 18 10 EOC10 2 EOC2 25 GOVRE 17 9 EOC9 1 EOC1 24 DRDY 16 8 EOC8 0 EOC0
EOCx: End of Conversion Interrupt Enable x XRDY: Touchscreen Measure XPOS Ready Interrupt Enable
YRDY: Touchscreen Measure YPOS Ready Interrupt Enable PRDY: Touchscreen Measure Pressure Ready Interrupt Enable EOCAL: End of Calibration Sequence DRDY: Data Ready Interrupt Enable GOVRE: General Overrun Error Interrupt Enable COMPE: Comparison Event Interrupt Enable PEN: Pen Contact Interrupt Enable NOPEN: No Pen Contact Interrupt Enable

1563
50.8.10 ADC Interrupt Disable Register

31 23 EOCAL 15 7 EOC7
ADC_IDR 0xF8018028 Write-only

EOCx: End of Conversion Interrupt Disable x XRDY: Touchscreen Measure XPOS Ready Interrupt Disable YRDY: Touchscreen Measure YPOS Ready Interrupt Disable PRDY: Touchscreen Measure Pressure Ready Interrupt Disable EOCAL: End of Calibration Sequence DRDY: Data Ready Interrupt Disable GOVRE: General Overrun Error Interrupt Disable COMPE: Comparison Event Interrupt Disable PEN: Pen Contact Interrupt Disable NOPEN: No Pen Contact Interrupt Disable

1564
50.8.11 ADC Interrupt Mask Register

31 23 EOCAL 15 7 EOC7
ADC_IMR 0xF801802C Read-only

EOCx: End of Conversion Interrupt Mask x XRDY: Touchscreen Measure XPOS Ready Interrupt Mask YRDY: Touchscreen Measure YPOS Ready Interrupt Mask PRDY: Touchscreen Measure Pressure Ready Interrupt Mask EOCAL: End of Calibration Sequence DRDY: Data Ready Interrupt Mask GOVRE: General Overrun Error Interrupt Mask COMPE: Comparison Event Interrupt Mask PEN: Pen Contact Interrupt Mask NOPEN: No Pen Contact Interrupt Mask

1565
50.8.12 ADC Interrupt Status Register

31 PENS 23 EOCAL 15 7 EOC7
ADC_ISR 0xF8018030 Read-only

EOCx: End of Conversion x

0 = The corresponding analog channel is disabled, or the conversion is not finished. This flag is cleared when reading the corresponding ADC_CDRx registers. 1 = The corresponding analog channel is enabled and conversion is complete.
XRDY: Touchscreen XPOS Measure Ready

0 = No Measure has been performed since the last read of ADC_XPOSR. 1 = At least one Measure has been performed since the last read of ADC_ISR.
YRDY: Touchscreen YPOS Measure Ready

0 = No Measure has been performed since the last read of ADC_YPOSR. 1 = At least one Measure has been performed since the last read of ADC_ISR.
PRDY: Touchscreen Pressure Measure Ready

0 = No Measure has been performed since the last read of ADC_PRESSR. 1 = At least one Measure has been performed since the last read of ADC_ISR.
EOCAL: End of Calibration Sequence

0 = Calibration sequence is on going, or no calibration sequence has been requested. 1 = Calibration sequence is complete.
DRDY: Data Ready

0 = No data has been converted since the last read of ADC_LCDR. 1 = At least one data has been converted and is available in ADC_LCDR.
GOVRE: General Overrun Error

0 = No General Overrun Error occurred since the last read of ADC_ISR. 1 = At least one General Overrun Error has occurred since the last read of ADC_ISR.
COMPE: Comparison Error

0 = No Comparison Error since the last read of ADC_ISR. 1 = At least one Comparison Error (defined in the ADC_EMR and ADC_CWR registers) has occurred since the last read of ADC_ISR.

1566
PEN: Pen contact

0 = No pen contact since the last read of ADC_ISR. 1 = At least one pen contact since the last read of ADC_ISR.
NOPEN: No Pen contact

0 = No loss of pen contact since the last read of ADC_ISR. 1 = At least one loss of pen contact since the last read of ADC_ISR.
PENS: Pen detect Status

0 = The pen does not press the screen. 1 = The pen presses the screen. Note: PENS is not a source of interruption.

1567
50.8.13 ADC Overrun Status Register

31 23 15 7 OVRE7
ADC_OVER 0xF801803C Read-only

30 22 14 6 OVRE6 29 21 13 5 OVRE5 28 20 12 4 OVRE4 27 19 11 OVRE11 3 OVRE3 26 18 10 OVRE10 2 OVRE2 25 17 9 OVRE9 1 OVRE1 24 16 8 OVRE8 0 OVRE0
OVREx: Overrun Error x

0 = No overrun error on the corresponding channel since the last read of ADC_OVER. 1 = There has been an overrun error on the corresponding channel since the last read of ADC_OVER.

1568
50.8.14 ADC Extended Mode Register

31 23 15 7
ADC_EMR 0xF8018040 Read-write

30 22 14 6 CMPSEL 29 21 13 CMPFILTER 5 4 28 20 12 27 19 11 3 26 18 10 2 25 17 9 CMPALL 1 CMPMODE 24 TAG 16 8 0
CMPMODE: Comparison Mode

Value 0 1 2 3 Name LOW HIGH IN OUT Description Generates an event when the converted data is lower than the low threshold of the window. Generates an event when the converted data is higher than the high threshold of the window. Generates an event when the converted data is in the comparison window. Generates an event when the converted data is out of the comparison window.
CMPSEL: Comparison Selected Channel

If CMPALL = 0: CMPSEL indicates which channel has to be compared. If CMPALL = 1: No effect.
CMPALL: Compare All Channels

0 = Only channel indicated in CMPSEL field is compared. 1 = All channels are compared.
CMPFILTER: Compare Event Filtering

Number of consecutive compare events necessary to raise the flag = CMPFILTER+1 When programmed to 0, the flag rises as soon as an event occurs.
TAG: TAG of the ADC_LDCR register

0 = Sets CHNB to zero in ADC_LDCR. 1 = Appends the channel number to the conversion result in ADC_LDCR register.

1569
50.8.15 ADC Compare Window Register

31 23
ADC_CWR 0xF8018044 Read-write

30 22 29 21 28 20 HIGHTHRES 27 26 HIGHTHRES 19 18 17 16 25 24
15 7
14 6
13 5
12 4 LOWTHRES
11
10 LOWTHRES
LOWTHRES: Low Threshold

Low threshold associated to compare settings of the ADC_EMR register.
HIGHTHRES: High Threshold

High threshold associated to compare settings of the ADC_EMR register.

1570
50.8.16 ADC Channel Gain Register

31 23 GAIN11 15 GAIN7 7 GAIN3 6 5 GAIN2 14 13 GAIN6 4 3 GAIN1 22 21 GAIN10 12 11 GAIN5 2 1 GAIN0
ADC_CGR 0xF8018048 Read-write

30 29 20 19 GAIN9 10 9 GAIN4 0 28 27 18 17 GAIN8 8 26 25 16 24
GAINx: Gain for channel x

Gain applied on input of analog-to-digital converter.
GAINx 0 0 1 1 0 1 0 1 Gain applied when DIFFx = 0 1 1 2 4 Gain applied when DIFFx = 1 0.5 1 2 2
The DIFFx mentioned in this table is described in the following register, ADC_COR.

1571
50.8.17 ADC Channel Offset Register

31 23 DIFF7 15 7 OFF7
ADC_COR 0xF801804C Read-write

30 22 DIFF6 14 6 OFF6 29 21 DIFF5 13 5 OFF5 28 20 DIFF4 12 4 OFF4 27 DIFF11 19 DIFF3 11 OFF11 3 OFF3 26 DIFF10 18 DIFF2 10 OFF10 2 OFF2 25 DIFF9 17 DIFF1 9 OFF9 1 OFF1 24 DIFF8 16 DIFF0 8 OFF8 0 OFF0
OFFx: Offset for channel x

0 = No Offset. 1 = Centers the analog signal on Vrefin/2 before the gain scaling. The Offset applied is: (G-1)Vrefin/2 where G is the gain applied (see description of ADC_CGR register).
DIFFx: Differential inputs for channel x

0 = Single Ended Mode. 1 = Fully Differential Mode.

1572
50.8.18 ADC Channel Data Register

31 23 15 7
ADC_CDRx [x=0..11] 0xF8018050 Read-write

30 22 14 6 29 21 13 5 28 20 12 4 DATA 27 19 11 26 18 10 DATA 3 2 1 0 25 17 9 24 16 8
DATA: Converted Data

The analog-to-digital conversion data is placed into this register at the end of a conversion and remains until a new conversion is completed. The Convert Data Register (CDR) is only loaded if the corresponding analog channel is enabled.

1573
50.8.19 ADC Analog Control Register

31 23 15 7
ADC_ACR 0xF8018094 Read-write

30 22 14 6 29 21 13 5 28 20 12 4 27 19 11 3 26 18 10 2 25 17 9 1 PENDETSENS 24 16 8 0
PENDETSENS: Pen Detection Sensitivity

Allows to modify the pen detection input pull-up resistor value. (See the product electrical characteristics for further details).

1574
50.8.20 ADC Touchscreen Mode Register

31
ADC_TSMR 0xF80180B0 Read-write

30 PENDBC 29 28 27 19 26 18 TSSCTIM 11 10 TSFREQ 3 2 1 TSMODE 0 9 8 25 17 24 PENDET 16
23 15 7
22 NOTSDMA 14 6
21 13 5 TSAV
20 12 4
TSMODE: Touchscreen Mode

Value 0 1 2 3 Name NONE 4_WIRE_NO_PM 4_WIRE 5_WIRE Description No Touchscreen 4-wire Touchscreen without pressure measurement 4-wire Touchscreen with pressure measurement 5-wire Touchscreen
When TSMOD equals 01 or 10 (i.e. 4-wire mode), channel 0, 1, 2 and 3 must not be used for classic ADC conversions. When TSMOD equals 11 (i.e. 5-wire mode), channel 0, 1, 2, 3, and 4 must not be used.
TSAV: Touchscreen Average

Value 0 1 2 3 Name NO_FILTER AVG2CONV AVG4CONV AVG8CONV Description No Filtering. Only one ADC conversion per measure Averages 2 ADC conversions Averages 4 ADC conversions Averages 8 ADC conversions
TSFREQ: Touchscreen Frequency

Defines the Touchscreen Frequency compared to the Trigger Frequency. TSFREQ must be greater or equal to TSAV. The Touchscreen Frequency is: Touchscreen Frequency = Trigger Frequency / 2TSFREQ
TSSCTIM: Touchscreen Switches Closure Time

Defines closure time of analog switches necessary to establish the measurement conditions. The Closure Time is: Switch Closure Time = (TSSCTIM * 4) ADCClock periods.

1575
PENDET: Pen Contact Detection Enable

0: Pen contact detection disable. 1: Pen contact detection enable. When PENDET = 1, XPOS, YPOS, Z1, Z2 values of ADC_XPOSR, ADC_YPOSR, ADC_PRESSR registers are automatically cleared when PENS = 0 in ADC_ISR.
NOTSDMA: No TouchScreen DMA

0: XPOS, YPOS, Z1, Z2 are transmitted in ADC_LCDR. 1: XPOS, YPOS, Z1, Z2 are never transmitted in ADC_LCDR, therefore the buffer does not contains touchscreen values.
PENDBC: Pen Detect Debouncing Period

Debouncing period = 2PENDBC ADCClock periods.

1576
50.8.21 ADC Touchscreen X Position Register

31 23
ADC_XPOSR 0xF80180B4 Read-only

30 22 29 21 28 20 XSCALE 27 26 XSCALE 19 18 17 16 25 24
15 7
14 6
13 5
12 4 XPOS
11
10 XPOS
XPOS: X Position
The Position measured is stored here. if XPOS = 0 or XPOS = XSIZE, the pen is on the border. When pen detection is enabled (PENDET set to 1 in ADC_TSMR register), XPOS is tied to 0 while there is no detection of contact on the touchscreen (i.e. when PENS bitfield is cleared in ADC_ISR register).
XSCALE: Scale of XPOS

Indicates the max value that XPOS can reach. This value should be close to 212.

1577
50.8.22 ADC Touchscreen Y Position Register

31 23
ADC_YPOSR 0xF80180B8 Read-only

30 22 29 21 28 20 YSCALE 27 26 YSCALE 19 18 17 16 25 24
15 7
14 6
13 5
12 4 YPOS
11
10 YPOS
YPOS: Y Position
The Position measured is stored here. if YPOS = 0 or YPOS = YSIZE, the pen is on the border. When pen detection is enabled (PENDET set to 1 in ADC_TSMR register), YPOS is tied to 0 while there is no detection of contact on the touchscreen (i.e. when PENS bitfield is cleared in ADC_ISR register).
YSCALE: Scale of YPOS

Indicates the max value that YPOS can reach. This value should be close to 212.

1578
50.8.23 ADC Touchscreen Pressure Register

31 23
ADC_PRESSR 0xF80180BC Read-only

30 22 29 21 28 20 Z2 27 26 Z2 19 18 17 16 25 24
15 7
14 6
13 5
12 4 Z1
11
10 Z1
Z1: Data of Z1 Measurement

Data Z1 necessary to calculate pen pressure. When pen detection is enabled (PENDET set to 1 in ADC_TSMR register), Z1 is tied to 0 while there is no detection of contact on the touchscreen (i.e. when PENS bitfield is cleared in ADC_ISR register).
Z2: Data of Z2 Measurement

Data Z2 necessary to calculate pen pressure. When pen detection is enabled (PENDET set to 1 in ADC_TSMR register), Z2 is tied to 0 while there is no detection of contact on the touchscreen (i.e. when PENS bitfield is cleared in ADC_ISR register). Note: these two values are unavailable if TSMODE is not set to 2 in ADC_TSMR register.

1579
50.8.24 ADC Trigger Register

31
ADC_TRGR 0xF80180C0 Read-write

30 29 28 27 26 25 24
TRGPER
23 22 21 20 19 18 17 16
TRGPER
15 14 13 12 11 10 9 8
TRGMOD
TRGMOD: Trigger Mode

Value 0 1 2 3 4 5 6 7 Name NO_TRIGGER EXT_TRIG_RISE EXT_TRIG_FALL EXT_TRIG_ANY PEN_TRIG PERIOD_TRIG CONTINUOUS Description No trigger, only software trigger can start conversions External Trigger Rising Edge External Trigger Falling Edge External Trigger Any Edge Pen Detect Trigger (shall be selected only if PENDET is set and TSAMOD = Touchscreen only mode) Periodic Trigger (TRGPER shall be initiated appropriately) Continuous Mode Reserved
TRGPER: Trigger Period

Effective only if TRGMOD defines a Periodic Trigger. Defines the periodic trigger period, with the following equation: Trigger Period = (TRGPER+1) /ADCCLK The minimum time between 2 consecutive trigger events must be strictly greater than the duration time of the longest conversion sequence according to configuration of registers ADC_MR, ADC_CHSR, ADC_SEQR1, ADC_SEQR2, ADC_TSMR. When TRGMOD is set to pen detect trigger (i.e. 100) and averaging is used (i.e. bitfield TSAV differs from 0 in ADC_TSMR register) only 1 measure is performed. Thus, XRDY, YRDY, PRDY, DRDY will not rise on pen contact trigger. To achieve measurement, several triggers must be provided either by software or by setting the TRGMOD on continuous trigger (i.e. 110) until flags rise.

1580
50.8.25 ADC Write Protect Mode Register

31
ADC_WPMR 0xF80180E4 Read-write


0 = Disables the Write Protect if WPKEY corresponds to 0x414443 (ADC in ASCII). 1 = Enables the Write Protect if WPKEY corresponds to 0x414443 (ADC in ASCII). Protects the registers: ADC Mode Register on page 1555 ADC Channel Sequence 1 Register on page 1557 ADC Channel Sequence 2 Register on page 1558 ADC Channel Enable Register on page 1559 ADC Channel Disable Register on page 1560 ADC Extended Mode Register on page 1569 ADC Compare Window Register on page 1570 ADC Channel Gain Register on page 1571 ADC Channel Offset Register on page 1572 ADC Analog Control Register on page 1574 ADC Touchscreen Mode Register on page 1575 ADC Trigger Register on page 1580
WPKEY: Write Protect KEY

Should be written at value 0x414443 (ADC in ASCII). Writing any other value in this field aborts the write operation of the WPEN bit. Always reads as 0.

1581
50.8.26 ADC Write Protect Status Register

31 23
ADC_WPSR 0xF80180E8 Read-only


0 = No Write Protect Violation has occurred since the last read of the ADC_WPSR register. 1 = A Write Protect Violation has occurred since the last read of the ADC_WPSR register. If this violation is an unauthorized attempt to write a protected register, the associated violation is reported into field WPVSRC.

When WPVS is active, this field indicates the write-protected register (through address offset or code) in which a write access has been attempted. Reading ADC_WPSR automatically clears all fields.

1582
51.
51.1
Fuse Controller (FUSE)

Description
The Fuse Controller (FUSE) supports software fuse programming through a 32-bit register, only fuses set to level 1 are programmed. The fast (main) RC oscillator must be enabled at startup to access the fuse matrix. It reads the fuse states on startup and stores them into 32-bit registers. The first 8 Fuse Status registers (FUSE_SRx) can be masked and will read as a value of 0 regardless of the fuse state when masked.
51.2
z z z z
Software Fuse Programming User Write Access for Fuse Part of Fuse can be Masked After Read 256 FUSE Bits:
z z z
192 Bits are dedicated to Users 3 Bits are dedicated to Special Functions The Fuse Controller can be hidden thanks to a SFR write-once bit. Please refer to the Special Function Registers (SFR) section of this datasheet for details.
51.2.1 FUSE Bit Mapping

To avoid any misfunctioning, user must ensure to avoid writing DO NOT USE (DNU) FUSE bits.
Table 51-1.
Bits [255:224] [223:192] [191:160] [159:128] [127:96] [95:64] [63:32] [31:0] USER_DATA[159:0] DO NOT USE (DNU) USER_DATA[223:192] DO NOT USE (DNU) B J W
FUSE Bit Mapping
51.2.2 Special Functions

User is allowed to set some special bits with the following function.
Table 51-2. Special Bits Bit Number 162 161 160 Bit Name B J W Function BMS_SAMPLING_DISABLED - BMS sampling is disabled if set JTAG_DISABLED - JTAG is disabled if set FUSE_WRITE_DISABLED - FUSE bit writing is disabled if set

1583
51.3
Block Diagram
Figure 51-1. Fuse Controller Block Diagram
Fuse States
Fuse States
Fuse Cell Fuse Controller Controls
Controls

1584
51.4
51.4.1 Fuse Reading

The fuse states are automatically read on CORE startup and are available for reading in the 8 Fuse Status (FUSE_SRx) registers. The fuse states of bits 31 to 0 will be available at FUSE_SR0, the fuse states of bits 63 to 32 will be available at FUSE_SR1 and so on. FUSE_SRx registers can be updated manually by using the RRQ bit of the Fuse Control register (FUSE_CR). RS and WS bits of the Fuse Index register (FUSE_IR) must be at level one before issuing the read request.
Figure 51-2. Fuse Read
Clock
FUSE_SRx
outdated
up to date
RRQ
WS
RS
51.4.2 Fuse Programming

All the fuses, except Atmel reserved fuses, can be written by software. To program fuses, strictly follow the order of the sequence instructions as provided below: 1. 2. 3. 4. Select the word to write, using the SELW field of the Fuse_Index register (FUSE_IR). Write the word to program in the Fuse_Data register (FUSE_DR). Check that RS and WS bits of the Fuse_Index register are at level one (no read and no write pending). Write the WRQ bit of the Fuse_Control register (FUSE_CR) to begin the fuse programming. The KEY field must be written at the same time with a value 0xFB to make the write request valid. Writing the WRQ bit will clear the WS bit. Check the WS bit of FUSE_SRx, when WS has a value of 1 the fuse write process is over.
5.
Only fuses to be set to level 1 are written.

1585
Figure 51-3. Fuse Write

Clock
WSEL
XX
00
01
DATA
XX
Fuse[31:0]
Fuse[63:32]
WRQ
WS RS
51.4.3 Fuse Masking

It is possible to mask the first 8 FUSE_SRx registers so that they will be read at a value of 0, regardless of the fuse state. To activate fuse masking on the first 8 FUSE_SRx registers, the MSK bit of the Fuse Mode register (FUSE_MR) must be written to level 1. The MSK bit is write-only. Solely a general reset can disable fuse masking.

1586
51.5
Fuse Controller (FUSE) User Interface
Table 51-3. Register Mapping Offset 0x00 0x04 0x08 0x0C 0x10 0x14 ... 0x2C 0x30 - 0xDC 0xE0 - 0xFC Register Fuse Control Register Fuse Mode Register Fuse Index Register Fuse Data Register Fuse Status Register 0 Fuse Status Register 1 ... Fuse Status Register 7 Reserved Reserved Name FUSE_CR FUSE_MR FUSE_IR FUSE_DR FUSE_SR0 FUSE_SR1 ... FUSE_SR7 Access Write-only Write-only Read-write Read-write Read-only Read-only ... Read-only Reset 0x00000000 0x00000000 0x00000000 ... 0x00000000

1587
51.5.1 Fuse Control Register

31 23 15
FUSE_CR 0xFFFFE400 Write-only

30 22 14 29 21 13 28 20 12 KEY 7 6 5 4 3 2 1 RRQ 0 WRQ 27 19 11 26 18 10 25 17 9 24 16 8
WRQ: Write Request

0 = No effect. 1 = Requests the word DATA to be programmed, ignored if KEY field is not filled with 0xFB.
RRQ: Read Request

0 = No effect. 1 = Requests the fuses to be read and FUSE_SRx registers to be updated; ignored if KEY field is not filled with 0xFB.
KEY: Key code

This KEY code is needed to set the WRQ bit. 0xFB (VALID): valid key.

1588
51.5.2 Fuse Mode Register

31 23 15 7
FUSE_MR 0xFFFFE404 Write-only

30 22 14 6 29 21 13 5 28 20 12 4 27 19 11 3 26 18 10 2 25 17 9 1 24 16 8 0 MSK
MSK: Mask Fuse Status Registers

0 = No effect. 1 = Masks the first 8 FUSE_SRx registers.

1589
51.5.3 Fuse Index Register

31 23 15 7
FUSE_IR 0xFFFFE408 Read-write

30 22 14 6 29 21 13 5 28 20 12 4 27 19 11 26 18 10 WSEL 3 2 1 RS 0 WS 25 17 9 24 16 8
WS: Write Status

0 = Write is pending or no write has been requested since general reset. 1 = Write of fuses is done.
RS: Read Status

0 = Read is pending or no read has been requested since general reset. 1 = Read of fuses is done.
WSEL: Word Selection

0-15 = Selects the word to write.

1590
51.5.4 Fuse Data Register

31
FUSE_DR 0xFFFFE40C Read-write

30 29 28 DATA 27 26 25 24
23
22
21
20 DATA
19
18
17
16
15
14
13
12 DATA
11
10
4 DATA
DATA: Data to Program

Data to program. Only bits with a value of 1 will be programmed.

1591
51.5.5
Name:
Fuse Status Register

FUSE_SRx [x=0..7] 0xFFFFE410 Read-only
30 29 28 FUSE 23 22 21 20 FUSE 15 14 13 12 FUSE 7 6 5 4 FUSE 3 2 1 0 11 10 9 8 19 18 17 16 27 26 25 24
Address: Access:
31
FUSE: Fuse Status

Indicates the status of corresponding fuses: 0 = Unprogrammed 1 = Programmed

1592
52.
52.1
Advanced Encryption Standard (AES)

Description
The Advanced Encryption Standard (AES) is compliant with the American FIPS (Federal Information Processing Standard) Publication 197 specification. The AES supports all five confidentiality modes of operation for symmetrical key block cipher algorithms (ECB, CBC, OFB, CFB and CTR), as specified in the NIST Special Publication 800-38A Recommendation. It is compatible with all these modes via Peripheral DMA Controller channels, minimizing processor intervention for large buffer transfers. The 128-bit/192-bit/256-bit key is stored in four/six/eight 32-bit registers (AES_KEYWRx) which are all write-only. The 128-bit input data and initialization vector (for some modes) are each stored in four 32-bit registers (AES_IDATARx and AES_IVRx) which are all write-only. As soon as the initialization vector, the input data and the key are configured, the encryption/decryption process may be started. Then the encrypted/decrypted data are ready to be read out on the four 32-bit output data registers (AES_ODATARx) or through the DMA channels.
52.2
z z z z
Compliant with FIPS Publication 197, Advanced Encryption Standard (AES) 128-bit/192-bit/256-bit Cryptographic Key 12/14/16 Clock Cycles Encryption/Decryption Processing Time with a 128-bit/192-bit/256-bit Cryptographic Key Support of the Five Standard Modes of Operation Specified in the NIST Special Publication 800-38A, Recommendation for Block Cipher Modes of Operation - Methods and Techniques:
z z z z z
Electronic Code Book (ECB) Cipher Block Chaining (CBC) including CBC-MAC Cipher Feedback (CFB) Output Feedback (OFB) Counter (CTR)
z z z
8-bit, 16-bit, 32-bit, 64-bit and 128-bit Data Sizes Possible in CFB Mode Last Output Data Mode Allows Optimized Message Authentication Code (MAC) Generation Connection to DMA Optimizes Data Transfers for all Operating Modes

1593
52.3

The AES may be clocked through the Power Management Controller (PMC), so the programmer must first to configure the PMC to enable the AES clock.
52.3.2 Interrupt
The AES interface has an interrupt line connected to the Interrupt Controller. Handling the AES interrupt requires programming the Interrupt Controller before configuring the AES.
Table 52-1. Peripheral IDs Instance AES ID 43

1594
52.4
The Advanced Encryption Standard (AES) specifies a FIPS-approved cryptographic algorithm that can be used to protect electronic data. The AES algorithm is a symmetric block cipher that can encrypt (encipher) and decrypt (decipher) information. Encryption converts data to an unintelligible form called ciphertext. Decrypting the ciphertext converts the data back into its original form, called plaintext. The CIPHER bit in the AES Mode Register (AES_MR) allows selection between the encryption and the decryption processes. The AES is capable of using cryptographic keys of 128/192/256 bits to encrypt and decrypt data in blocks of 128 bits. This 128-bit/192-bit/256-bit key is defined in the Key Registers (AES_KEYWRx). The input to the encryption processes of the CBC, CFB, and OFB modes includes, in addition to the plaintext, a 128-bit data block called the initialization vector (IV), which must be set in the Initialization Vector Registers (AES_IVRx). The initialization vector is used in an initial step in the encryption of a message and in the corresponding decryption of the message. The Initialization Vector Registers are also used by the CTR mode to set the counter value.
52.4.1 Operation Modes

The AES supports the following modes of operation:
z z z z
ECB: Electronic Code Book CBC: Cipher Block Chaining OFB: Output Feedback CFB: Cipher Feedback
z z z z z
CFB8 (CFB where the length of the data segment is 8 bits) CFB16 (CFB where the length of the data segment is 16 bits) CFB32 (CFB where the length of the data segment is 32 bits) CFB64 (CFB where the length of the data segment is 64 bits) CFB128 (CFB where the length of the data segment is 128 bits)
CTR: Counter
The data pre-processing, post-processing and data chaining for the concerned modes are automatically performed. Refer to the NIST Special Publication 800-38A Recommendation for more complete information. These modes are selected by setting the OPMOD field in the AES Mode Register (AES_MR). In CFB mode, five data sizes are possible (8, 16, 32, 64 or 128 bits), configurable by means of the CFBS field in the mode register. (Section 52.6.2 AES Mode Register on page 1603). In CTR mode, the size of the block counter embedded in the module is 16 bits. Therefore, there is a rollover after processing 1 megabyte of data. If the file to be processed is greater than 1 megabyte, this file must be split into fragments of 1 megabyte. Prior to loading the first fragment into AES_IDATARx registers, the AES_IVRx registers must be cleared. For any fragment, after the transfer is completed and prior to transferring the next fragment, the AES_IVR0 register must be programmed so that the fragment number (0 for the first fragment, 1 for the second one, and so on) is written in the 16 MSB of the AES_IVR0 register.

1595
52.4.2 Start Modes

The SMOD field in the AES Mode Register (AES_MR) allows selection of the encryption (or decryption) start mode. 52.4.2.1 Manual Mode The sequence order is as follows:
z z z
Write the Mode Register (AES_MR) with all required fields, including but not limited to SMOD and OPMOD. Write the 128-bit/192-bit/256-bit key in the Key Registers (AES_KEYWRx). Write the initialization vector (or counter) in the Initialization Vector Registers (AES_IVRx). The Initialization Vector Registers concern all modes except ECB. Set the bit DATRDY (Data Ready) in the AES Interrupt Enable register (AES_IER), depending on whether an interrupt is required or not at the end of processing. Write the data to be encrypted/decrypted in the authorized Input Data Registers (See Table 52-2).
Note:
z z
Table 52-2. Authorized Input Data Registers Operation Mode ECB CBC OFB 128-bit CFB 64-bit CFB 32-bit CFB 16-bit CFB 8-bit CFB CTR Input Data Registers to Write All All All All AES_IDATAR0 and AES_IDATAR1 AES_IDATAR0 AES_IDATAR0 AES_IDATAR0 All
Note: Note:
z z z
In 64-bit CFB mode, writing to AES_IDATAR2 and AES_IDATAR3 registers is not allowed and may lead to errors in processing. In 32-, 16- and 8-bit CFB modes, writing to AES_IDATAR1, AES_IDATAR2 and AES_IDATAR3 registers is not allowed and may lead to errors in processing. Set the START bit in the AES Control register AES_CR to begin the encryption or the decryption process. When processing completes, the DATRDY bit in the AES Interrupt Status Register (AES_ISR) raises. If an interrupt has been enabled by setting the DATRDY bit in AES_IER, the interrupt line of the AES is activated. When the software reads one of the Output Data Registers (AES_ODATARx), the DATRDY bit is automatically cleared.
52.4.2.2 Auto Mode The Auto Mode is similar to the manual one, except that in this mode, as soon as the correct number of Input Data registers is written, processing is automatically started without any action in the Control Register.

1596
52.4.3 DMA Mode

The DMA Controller can be used in association with the AES to perform an encryption/decryption of a buffer without any action by the software during processing. The SMOD field of the AES_MR must be set to 0x2 and the DMA must be configured with non incremental addresses. The start address of any transfer descriptor must be set to AES_IDATAR0 register. The DMA chunk size configuration depends on the AES mode of operation and is listed in Table 52-3 "DMA Data Transfer Type for the Different Operation Modes". When writing data to AES with a first DMA channel, data are first fetched from a memory buffer (source data). It is recommended to configure the size of source data to words even for CFB modes. On the contrary, the destination data size depends on the mode of operation. When reading data from the AES with the second DMA channel, the source data is the data read from AES and data destination is the memory buffer. In this case, source data size depends on the AES mode of operation and is listed in Table 52-3.
Table 52-3. DMA Data Transfer Type for the Different Operation Modes Operation Mode ECB CBC OFB CFB 128-bit CFB 64-bit CFB 32-bit CFB 16-bit CFB 8-bit CTR Chunk Size 4 4 4 4 1 1 1 1 4 Destination/Source Data Transfer Type Word Word Word Word Word Word Half-word Byte Word
52.4.4 Last Output Data Mode

This mode is used to generate cryptographic checksums on data (MAC) by means of cipher block chaining encryption algorithm (CBC-MAC algorithm for example). After each end of encryption/decryption, the output data are available either on the output data registers for Manual and Auto mode or at the address specified in the receive buffer pointer for DMA mode (See Table 52-4 "Last Output Data Mode Behavior versus Start Modes"). The Last Output Data bit (LOD) in the AES Mode Register (AES_MR) allows retrieval of only the last data of several encryption/decryption processes. Therefore, there is no need to define a read buffer in DMA mode. This data are only available on the Output Data Registers (AES_ODATARx).

1597
52.4.5 Manual and Auto Modes

52.4.5.1 If LOD = 0 The DATRDY flag is cleared when at least one of the Output Data Registers is read (See Figure 52-1).
Figure 52-1. Manual and Auto Modes with LOD = 0
Write START bit in AES_CR (Manual mode) or Write AES_IDATARx register(s) (Auto mode)
Read the AES_ODATARx
DATRDY
Encryption or Decryption Process
If the user does not want to read the output data registers between each encryption/decryption, the DATRDY flag will not be cleared. If the DATRDY flag is not cleared, the user cannot know the end of the following encryptions/decryptions. 52.4.5.2 If LOD = 1 The DATRDY flag is cleared when at least one Input Data Register is written, so before the start of a new transfer (See Figure 52-2). No more Output Data Register reads are necessary between consecutive encryptions/decryptions.
Write START bit in AES_CR (Manual mode) or Write AES_IDATARx register(s) (Auto mode)
Write AES_IDATARx register(s)
DATRDY
52.4.6 DMA Mode

52.4.6.1 If LOD = 0 This mode may be used for all AES operating modes except CBC-MAC where LOD=1 mode is recommended. The end of the encryption/decryption is notified by the end of DMA transfer associated to AES_ODATARx registers (see Figure 52-3). 2 channels DMA are required, 1 for writing message blocks to AES_IDATARx registers and the other one to get back the processed from AES_ODATARx registers.

1598
Figure 52-3. DMA transfer with LOD = 0
Enable DMA Channels associated to AES_IDATARx and AES_ODATARx
Multiple Encryption or Decryption Processes BTC /channel 0 BTC /channel 1 Write accesses into AES_IDATARx Read accesses into AES_ODATARx Message fully processed (cipher or decipher) last block can be read
52.4.6.2 If LOD = 1 This mode is recommended to process AES CBC-MAC operating mode. The user must first wait for the DMA flag (BTC = Buffer Transfer Complete) to rise, then for DATRDY to ensure that the encryption/decryption is completed (see Figure 52-4). In this case, no receive buffers are required. The output data are only available on the Output Data Registers (AES_ODATARx).
Enable DMA Channels associated with TDES_IDATARx and TDES_ODATARx registers
Multiple Encryption or Decryption Processes Write accesses into AES_IDATARx BTC / channel 0 DATRDY Message fully transferred Message fully processed (cipher or decipher) MAC result can be read
Table 52-4 summarizes the different cases.

Table 52-4. Last Output Data Mode Behavior versus Start Modes Manual and Auto Modes LOD = 0 DATRDY Flag Clearing Condition(1) Encrypted/Decrypted Data Result Location End of Encryption/Decryption At least one Output Data Register must be read In the Output Data Registers DATRDY LOD = 1 At least one Input Data Register must be written In the Output Data Registers DATRDY LOD = 0 Not used DMA Transfer LOD = 1 Managed by the DMA In the Output Data Registers DMA flag (BTC[n]) then AES DATRDY
Not available 2 DMA flags (BTC[n/m])
Note:
1. Depending on the mode, there are other ways of clearing the DATRDY flag. See: Section 52.6.6 AES Interrupt Status Register.
Warning: In DMA mode, reading to the Output Data registers before the last data transfer may lead to unpredictable results.

1599
52.5
Security Features
52.5.1 Unspecified Register Access Detection

When an unspecified register access occurs, the URAD bit in the Interrupt Status Register (AES_ISR) raises. Its source is then reported in the Unspecified Register Access Type field (URAT). Only the last unspecified register access is available through the URAT field. Several kinds of unspecified register accesses can occur:
z z z z z z
Input Data Register written during the data processing when SMOD=IDATAR0_START Output Data Register read during data processing Mode Register written during data processing Output Data Register read during sub-keys generation Mode Register written during sub-keys generation Write-only register read access
The URAD bit and the URAT field can only be reset by the SWRST bit in the AES_CR control register.

1600
52.6
Advanced Encryption Standard (AES) User Interface
Table 52-5. Register Mapping Offset 0x00 0x04 0x08-0x0C 0x10 0x14 0x18 0x1C 0x20 0x24 0x28 0x2C 0x30 0x34 0x38 0x3C 0x40 0x44 0x48 0x4C 0x50 0x54 0x58 0x5C 0x60 0x64 0x68 0x6C 0x70 - 0xFC Register Control Register Mode Register Reserved Interrupt Enable Register Interrupt Disable Register Interrupt Mask Register Interrupt Status Register Key Word Register 0 Key Word Register 1 Key Word Register 2 Key Word Register 3 Key Word Register 4 Key Word Register 5 Key Word Register 6 Key Word Register 7 Input Data Register 0 Input Data Register 1 Input Data Register 2 Input Data Register 3 Output Data Register 0 Output Data Register 1 Output Data Register 2 Output Data Register 3 Initialization Vector Register 0 Initialization Vector Register 1 Initialization Vector Register 2 Initialization Vector Register 3 Reserved Name AES_CR AES_MR AES_IER AES_IDR AES_IMR AES_ISR AES_KEYWR0 AES_KEYWR1 AES_KEYWR2 AES_KEYWR3 AES_KEYWR4 AES_KEYWR5 AES_KEYWR6 AES_KEYWR7 AES_IDATAR0 AES_IDATAR1 AES_IDATAR2 AES_IDATAR3 AES_ODATAR0 AES_ODATAR1 AES_ODATAR2 AES_ODATAR3 AES_IVR0 AES_IVR1 AES_IVR2 AES_IVR3 Access Write-only Read-write Write-only Write-only Read-only Read-only Write-only Write-only Write-only Write-only Write-only Write-only Write-only Write-only Write-only Write-only Write-only Write-only Read-only Read-only Read-only Read-only Write-only Write-only Write-only Write-only Reset 0x0 0x0 0x0 0x0 0x0 0x0 0x0

1601
52.6.1 AES Control Register

31
AES_CR 0xF8038000 Write-only

30 29 28 27 26 25 24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
SWRST
0
START
START: Start Processing

0: No effect 1: Starts manual encryption/decryption process.

0: No effect. 1: Resets the AES. A software triggered hardware reset of the AES interface is performed.

1602
52.6.2 AES Mode Register

31
AES_MR 0xF8038004 Read-write

30 29 28 27 26 25 24
23
22
21
20
19
18
17
16
CKEY
15 14 13 12
11 10
CFBS
9 8
LOD
7 6
OPMOD
5 4 3
KEYSIZE
2 1
SMOD
0
PROCDLY
CIPHER
CIPHER: Processing Mode

0: Decrypts data. 1: Encrypts data.
PROCDLY: Processing Delay Processing Time = 12 ( PROCDLY + 1 )

The Processing Time represents the number of clock cycles that the AES needs in order to perform one encryption/decryption. Note: The best performance is achieved with PROCDLY equal to 0.
SMOD: Start Mode

Value 0x0 0x1 0x2 Name MANUAL_START AUTO_START IDATAR0_START Description Manual Mode Auto Mode AES_IDATAR0 access only Auto Mode
Values which are not listed in the table must be considered as reserved. If a DMA transfer is used, 0x2 must be configured. Refer to Section 52.4.3 DMA Mode for more details.
KEYSIZE: Key Size

Value 0x0 0x1 0x2 Name AES128 AES192 AES256 Description AES Key Size is 128 bits AES Key Size is 192 bits AES Key Size is 256 bits
Values which are not listed in the table must be considered as reserved.

1603
OPMOD: Operation Mode

Value 0x0 0x1 0x2 0x3 0x4 Name ECB CBC OFB CFB CTR Description ECB: Electronic Code Book mode CBC: Cipher Block Chaining mode OFB: Output Feedback mode CFB: Cipher Feedback mode CTR: Counter mode (16-bit internal counter)
Values which are not listed in the table must be considered as reserved. For CBC-MAC operating mode, please set OPMOD to CBC and LOD to 1.
LOD: Last Output Data Mode

0: No effect. After each end of encryption/decryption, the output data will be available either on the output data registers (Manual and Auto modes). In Manual and Auto modes, the DATRDY flag is cleared when at least one of the Output Data registers is read. 1: The DATRDY flag is cleared when at least one of the Input Data Registers is written. No more Output Data Register reads is necessary between consecutive encryptions/decryptions (see Last Output Data Mode on page 1597). Warning: In DMA mode, reading to the Output Data registers before the last data encryption/decryption process may lead to unpredictable results.
CFBS: Cipher Feedback Data Size

Value 0x0 0x1 0x2 0x3 0x4 Name SIZE_128BIT SIZE_64BIT SIZE_32BIT SIZE_16BIT SIZE_8BIT Description 128-bit 64-bit 32-bit 16-bit 8-bit
Values which are not listed in table must be considered as reserved.
CKEY: Key
This field must be written with the value 0xE for the first AES_MR programming. For further programming of the AES_MR register, any value can be written, including that of 0xE. Note: CKEY field is write-only.

1604
52.6.3 AES Interrupt Enable Register

31
AES_IER 0xF8038010 Write-only

30 29 28 27 26 25 24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
URAD
0
DATRDY
DATRDY: Data Ready Interrupt Enable URAD: Unspecified Register Access Detection Interrupt Enable

1605
52.6.4 AES Interrupt Disable Register

31
AES_IDR 0xF8038014 Write-only

30 29 28 27 26 25 24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
URAD
0
DATRDY
DATRDY: Data Ready Interrupt Disable URAD: Unspecified Register Access Detection Interrupt Disable

1606
52.6.5 AES Interrupt Mask Register

31
AES_IMR 0xF8038018 Read-only

30 29 28 27 26 25 24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
URAD
0
DATRDY
DATRDY: Data Ready Interrupt Mask URAD: Unspecified Register Access Detection Interrupt Mask

1607
52.6.6 AES Interrupt Status Register

31
AES_ISR 0xF803801C Read-only

30 29 28 27 26 25 24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
URAT
7 6 5 4
URAD
0
DATRDY
DATRDY: Data Ready

0: Output data not valid. 1: Encryption or decryption process is completed. DATRDY is cleared when a Manual encryption/decryption occurs (START bit in AES_CR) or when a software triggered hardware reset of the AES interface is performed (SWRST bit in AES_CR). LOD = 0 (AES_MR): In Manual and Auto mode, the DATRDY flag can also be cleared when at least one of the Output Data Registers is read. In DMA mode, DATRDY is set and cleared automatically. LOD = 1 (AES_MR): In Manual and Auto mode, the DATRDY flag can also be cleared when at least one of the Input Data Registers is written. In DMA mode, DATRDY is set and cleared automatically.
URAD: Unspecified Register Access Detection Status

0: No unspecified register access has been detected since the last SWRST. 1: At least one unspecified register access has been detected since the last SWRST. URAD bit is reset only by the SWRST bit in the AES_CR control register.
URAT: Unspecified Register Access:

Value 0x0 0x1 0x2 0x3 0x4 0x5 Name IDR_WR_PROCESSING ODR_RD_PROCESSING MR_WR_PROCESSING ODR_RD_SUBKGEN MR_WR_SUBKGEN WOR_RD_ACCESS Description Input Data Register written during the data processing when SMOD=0x2 mode. Output Data Register read during the data processing. Mode Register written during the data processing. Output Data Register read during the sub-keys generation. Mode Register written during the sub-keys generation. Write-only register read access.
Only the last Unspecified Register Access Type is available through the URAT field. URAT field is reset only by the SWRST bit in the AES_CR control register.

1608
52.6.7 AES Key Word Register x

31
AES_KEYWRx 0xF8038020 Write-only

30 29 28 27 26 25 24
KEYW
23 22 21 20 19 18 17 16
KEYW
15 14 13 12 11 10 9 8
KEYW
7 6 5 4 3 2 1 0
KEYW
KEYW: Key Word

The four/six/eight 32-bit Key Word registers set the 128-bit/192-bit/256-bit cryptographic key used for encryption/decryption. AES_KEYWR0 corresponds to the first word of the key and respectively AES_KEYWR3/AES_KEYWR5/AES_KEYWR7 to the last one. These registers are write-only to prevent the key from being read by another application.

1609
52.6.8 AES Input Data Register x

31
AES_IDATARx 0xF8038040 Write-only

30 29 28 27 26 25 24
IDATA
23 22 21 20 19 18 17 16
IDATA
15 14 13 12 11 10 9 8
IDATA
7 6 5 4 3 2 1 0
IDATA
IDATA: Input Data Word

The four 32-bit Input Data registers set the 128-bit data block used for encryption/decryption. AES_IDATAR0 corresponds to the first word of the data to be encrypted/decrypted, and AES_IDATAR3 to the last one. These registers are write-only to prevent the input data from being read by another application.

1610
52.6.9 AES Output Data Register x

31
AES_ODATARx 0xF8038050 Read-only

30 29 28 27 26 25 24
ODATA
23 22 21 20 19 18 17 16
ODATA
15 14 13 12 11 10 9 8
ODATA
7 6 5 4 3 2 1 0
ODATA
ODATA: Output Data

The four 32-bit Output Data registers contain the 128-bit data block that has been encrypted/decrypted. AES_ODATAR0 corresponds to the first word, AES_ODATAR3 to the last one.

1611
52.6.10 AES Initialization Vector Register x

31
AES_IVRx 0xF8038060 Write-only

30 29 28 27 26 25 24
IV
23 22 21 20 19 18 17 16
IV
15 14 13 12 11 10 9 8
IV
7 6 5 4 3 2 1 0
IV
IV: Initialization Vector

The four 32-bit Initialization Vector registers set the 128-bit Initialization Vector data block that is used by some modes of operation as an additional initial input. AES_IVR0 corresponds to the first word of the Initialization Vector, AES_IVR3 to the last one. These registers are write-only to prevent the Initialization Vector from being read by another application. For CBC, OFB and CFB modes, the Initialization Vector corresponds to the initialization vector. For CTR mode, it corresponds to the counter value. Note: These registers are not used in ECB mode and must not be written.

1612
53.
53.1
Triple Data Encryption Standard (Triple DES)

Description
The Triple Data Encryption Standard (TDES) is compliant with the American FIPS (Federal Information Processing Standard) Publication 46-3 specification. The TDES supports the four different confidentiality modes of operation (ECB, CBC, OFB and CFB), specified in the FIPS (Federal Information Processing Standard) Publication 81 and is compatible with the Peripheral Data Controller channels for all of these modes, minimizing processor intervention for large buffer transfers. The 64-bit long keys and input data (and initialization vector for some modes) are each stored in two 32-bit registers (TDES_KEYxWRx, TDES_IDATARx and TDES_IVRx) which are both write-only. As soon as the initialization vector, the input data and the key are configured, the encryption/decryption process may be started. Then the encrypted/decrypted data is ready to be read out on the two 32-bit output data registers (TDES_ODATARx) or through the DMA channels.
53.2
z z z z z z z
Supports Single Data Encryption Standard (DES) and Triple Data Encryption Algorithm (TDEA or TDES) Compliant with FIPS Publication 46-3, Data Encryption Standard (DES) 64-bit Cryptographic Key for TDES Two-key or Three-key Algorithms for TDES 18-clock Cycles Encryption/Decryption Processing Time for DES 50-clock Cycles Encryption/Decryption Processing Time for TDES Support the Four Standard Modes of Operation specified in the FIPS Publication 81, DES Modes of Operation
z z z z
Electronic Code Book (ECB) Cipher Block Chaining (CBC) Cipher Feedback (CFB) Output Feedback (OFB)
z z z
8-bit, 16-bit, 32-bit and 64-bit Data Sizes Possible in CFB Mode Last Output Data Mode Allowing Optimized Message (Data) Authentication Code (MAC) Generation Connection to DMA Optimizes Data Transfers for all Operating Modes

1613
53.3

The TDES may be clocked through the Power Management Controller (PMC), so the programmer must first configure the PMC to enable the TDES clock.
53.3.2 Interrupt
The TDES interface has an interrupt line connected to the Advanced Interrupt Controller (AIC). Handling the TDES interrupt requires programming the AIC before configuring the TDES. Peripheral IDs
Instance TDES ID 44
53.4
The Data Encryption Standard (DES) and the Triple Data Encryption Algorithm (TDES) specify FIPS-approved cryptographic algorithms that can be used to protect electronic data. The TDES bit in the TDES Mode Register (TDES_MR) is used to select either the single DES or the Triple DES mode. Encryption (enciphering) converts data to an unintelligible form called ciphertext. Decrypting (deciphering) the ciphertext converts the data back into its original form, called plaintext. The CIPHER bit in the TDES Mode Register is used to choose between encryption and decryption. A DES is capable of using cryptographic keys of 64 bits to encrypt and decrypt data in blocks of 64 bits. This 64-bit key is defined in the Key 1 Word Registers (TDES_KEY1WRx). A TDES key consists of three DES keys, which is also referred to as a key bundle. These three 64-bit keys are defined, respectively, in the Key 1, 2 and 3 Word Registers (TDES_KEY1WRx, TDES_KEY2WRx and TDES_KEY3WRx). In Triple DES mode (TDESMOD set to 1), the KEYMOD bit in the TDES Mode Register is used to choose between a twoand a three-key algorithm:
z z z z
In three-key encryption mode, the data is first encrypted with Key 1, then decrypted using Key 2 and then encrypted with Key 3. In three-key decryption mode, the data is decrypted with Key 3, then encrypted with Key 2 and then decrypted using Key 1. In two-key encryption mode, the data is first encrypted with Key 1, then decrypted using Key 2 and then encrypted with Key 1. In two-key decryption mode, the data is decrypted with Key 1, then encrypted with Key 2 and then decrypted using Key 1.
The input to the encryption processes of the CBC, CFB, and OFB modes includes, in addition to the plaintext, a 64-bit data block called the initialization vector (IV), which must be set in the Initialization Vector Registers (TDES_IVRx). The initialization vector is used in an initial step in the encryption of a message and in the corresponding decryption of the message.

1614
53.4.1 Operation Modes

The TDES supports the following modes of operation: z ECB: Electronic Code Book z CBC: Cipher Block Chaining z OFB: Output Feedback z CFB: Cipher Feedback z CFB8 (CFB where the length of the data segment is 8 bits)
z z z
CFB16 (CFB where the length of the data segment is 16 bits) CFB32 (CFB where the length of the data segment is 32 bits) CFB64 (CFB where the length of the data segment is 64 bits)
The data pre-processing, post-processing and data chaining for each mode are automatically performed. Refer to the FIPS Publication 81 for more complete information. These modes are selected by setting the OPMOD field in the TDES Mode Register (TDES_MR). In CFB mode, four data sizes are possible (8, 16, 32 and 64 bits), configurable by means of the CFBS field in the mode register. (See TDES Mode Register on page 1621.). The OFB and CFB modes of operation are only available if 2-key mode is selected (KEYMOD=1 in TDES_MR register).
53.4.2 Start Modes

The SMOD field in the TDES Mode Register (TDES_MR) allows selection of encryption (or decryption) start mode. 53.4.2.1 Manual Mode The sequence order is as follows: z Write the Mode Register (TDES_MR) with all required fields, including but not limited to SMOD and OPMOD. z Write the 64-bit key(s) in the different Key Word Registers (TDES_KEYxWRx), depending on whether one, two or three keys are required. z Write the initialization vector (or counter) in the Initialization Vector Registers (TDES_IVRx). Note:
z z
The Initialization Vector Registers concern all modes except ECB. Set the bit DATRDY (Data Ready) in the TDES Interrupt Enable register (TDES_IER), depending on whether an interrupt is required or not at the end of processing. Write the data to be encrypted/decrypted in the authorized Input Data Registers (See Table 53-1).
Table 53-1. Authorized Input Data Registers Operation Mode ECB CBC OFB CFB 64-bit CFB 32-bit CFB 16-bit CFB 8-bit Input Data Registers to Write All All All All TDES_IDATAR0 TDES_IDATAR0 TDES_IDATAR0
Note:
z z z
In 32-bit, 16-bit and 8-bit CFB mode, writing to TDES_IDATAR1 register is not allowed and may lead to errors in processing. Set the START bit in the TDES Control register TDES_CR to begin the encryption or the decryption process. When the processing completes, the bit DATRDY in the TDES Interrupt Status Register (TDES_ISR) raises. If an interrupt has been enabled by setting the bit DATRDY in TDES_IER, the interrupt line of the TDES is activated. When the software reads one of the Output Data Registers (TDES_ODATARx), the DATRDY bit is automatically cleared.

1615
53.4.2.2 Auto Mode The Auto Mode is similar to Manual Mode, except that, as soon as the correct number of Input Data registers is written, processing is automatically started without any action in the control register. 53.4.2.3 DMA Mode The DMA Controller can be used in association with the TDES to perform an encryption/decryption of a buffer without any action by the software during processing. The SMOD field of TDES_MR must be set to 0x2 and the DMA must be configured with non-incremental addresses. The start address of any transfer descriptor must be set in TDES_IDATAR0 register. The DMA chunk size configuration depends on the TDES mode of operation and is listed in Table 53-2 "DMA Data Transfer Type for the Different Operation Modes". When writing data to TDES with the first DMA channel, data will be fetched from a memory buffer (source data). It is recommended to configure the size of source data to words even for CFB modes. On the contrary, the destination data size depends on the mode of operation. When reading data from the TDES with the second DMA channel, the source data is the data read from TDES and data destination is the memory buffer. In this case, source data size depends on the TDES mode of operation and is listed in Table 53-2.
Table 53-2. DMA Data Transfer Type for the Different Operation Modes Operation Mode ECB CBC OFB CFB 64-bit CFB 32-bit CFB 16-bit CFB 8-bit Chunk Size 1 1 1 1 1 1 1 Destination/Source Data Transfer Type Word Word Word Word Word Half-word Byte
53.4.3 Last Output Data Mode

This mode is used to generate cryptographic checksums on data (MAC) using a CBC-MAC or a CFB encryption algorithm (See FIPS Publication 81 Appendix F). After each end of encryption/decryption, the output data is available either on the output data registers for Manual and Auto mode or at the address specified in the receive buffer pointer for DMA mode (See Table 53-3 "Last Output Data Mode Behavior versus Start Modes"). The Last Output Data bit (LOD) in the TDES Mode Register (TDES_MR) allows retrieval of only the last data of several encryption/decryption processes. This data is only available on the Output Data Registers (TDES_ODATARx). No more Output Data Register reads are necessary between consecutive encryptions/decryptions. Therefore, there is no need to define a read buffer in DMA mode. 53.4.3.1 Manual and Auto Modes If LOD = 0: The DATRDY flag is cleared when at least one of the Output Data Registers is read. See Figure 53-1.

1616
Write START bit in TDES_CR (Manual mode) or Write TDES_IDATARx register(s) (Auto mode)
Read the TDES_ODATARx
DATRDY
If the user does not want to read the output data registers between each encryption/decryption, the DATRDY flag will not be cleared. If the DATRDY flag is not cleared, the user will not be informed of the end of the encryptions/decryptions that follow. If LOD = 1: The DATRDY flag is cleared when at least one Input Data Register is written, before the start of a new transfer. See Figure 53-2. No further Output Data Register reads are necessary between consecutive encryptions/decryptions.
Write START bit in TDES_CR (Manual mode) or Write TDES_IDATARx register(s) (Auto mode)
Write TDES_IDATARx register(
DATRDY
53.4.3.2 DMA Mode
If LOD = 0
This mode may be used for all TDES operating modes except CBC-MAC where LOD = 1 mode is recommended. The end of the encryption/decryption is notified by the end of DMA transfer associated to TDES_ODATARx registers (see Figure 53-3). Two DMA channels are required, one for writing message blocks to TDES_IDATARx registers and the other one to get back the processed from TDES_ODATARx registers.
Enable DMA Channels associated to TDES_IDATARx and TDES_ODATARx
Multiple Encryption or Decryption Processes Write accesses into TDES_IDATARx BTC /channel 0 Read accesses into TDES_ODATARx BTC /channel 1 Message fully processed (cipher or decipher) last block can be read

1617
If LOD = 1
This mode is recommended to process TDES CBC-MAC operating mode. The user must first wait for the DMA flag (BTC = Buffer Transfer Complete) to rise, then for DATRDY to ensure that the encryption/decryption is completed (see Figure 53-4). In this case, no receive buffers are required. The output data is only available on the Output Data Registers (TDES_ODATARx).
Enable DMA Channels associated with TDES_IDATARx and TDES_ODATARx registers
Multiple Encryption or Decryption Processes Write accesses into TDES_IDATARx BTC / channel 0 DATRDY Message fully transferred Message fully processed (cipher or decipher) MAC result can be read
Table 53-3 summarizes the different cases.

Table 53-3. Last Output Data Mode Behavior versus Start Modes Manual and Auto Modes LOD = 0 DATRDY Flag Clearing Condition(1) Encrypted/Decrypted Data Result Location End of Encryption/Decryption At least one Output Data Register must be read In the Output Data Registers DATRDY LOD = 1 At least one Input Data Register must be written In the Output Data Registers DATRDY LOD = 0 Not used DMA Transfer LOD = 1 Managed by the DMA In the Output Data Registers DMA flag (BTC[n]) then TDES DATRDY
Not available 2 DMA flags (BTC[n/m])
Note:
Depending on the mode, there are other ways of clearing the DATRDY flag. See: Section 53.5.6 TDES Interrupt Status Register.
Warning: In DMA mode, reading to the Output Data registers before the last data transfer may lead to unpredictable results.
53.4.4 Security Features

53.4.4.1 Unspecified Register Access Detection When an unspecified register access occurs, the URAD bit in the Interrupt Status Register (TDES_ISR) raises. Its source is then reported in the Unspecified Register Access Type field (URAT). Only the last unspecified register access is available through the URAT field. Several kinds of unspecified register accesses can occur: z Input Data Register written during the data processing in DMA mode z Output Data Register read during the data processing z Mode Register written during the data processing z Write-only register read access The URAD bit and the URAT field can only be reset by the SWRST bit in the TDES_CR control register.

1618
53.5
Triple Data Encryption Standard (TDES) User Interface
Table 53-4. Register Mapping Offset 0x00 0x04 0x08-0x0C 0x10 0x14 0x18 0x1C 0x20 0x24 0x28 0x2C 0x30 0x34 0x38-0x3C 0x40 0x44 0x48-0x4C 0x50 0x54 0x58-0x5C 0x60 0x64 0x68 - 0xFC Register Control Register Mode Register Reserved Interrupt Enable Register Interrupt Disable Register Interrupt Mask Register Interrupt Status Register Key 1 Word Register 0 Key 1 Word Register 1 Key 2 Word Register 0 Key 2 Word Register 1 Key 3 Word Register 0 Key 3 Word Register 1 Reserved Input Data Register 0 Input Data Register 1 Reserved Output Data Register 0 Output Data Register 1 Reserved Initialization Vector Register 0 Initialization Vector Register 1 Reserved Name TDES_CR TDES_MR TDES_IER TDES_IDR TDES_IMR TDES_ISR TDES_KEY1WR0 TDES_KEY1WR1 TDES_KEY2WR0 TDES_KEY2WR1 TDES_KEY3WR0 TDES_KEY3WR1 TDES_IDATAR0 TDES_IDATAR1 TDES_ODATAR0 TDES_ODATAR1 TDES_IVR0 TDES_IVR1 Access Write-only Read-write Write-only Write-only Read-only Read-only Write-only Write-only Write-only Write-only Write-only Write-only Write-only Write-only Read-only Read-only Write-only Write-only Reset 0x2 0x0 0x0000001E 0x0 0x0

1619
53.5.1 TDES Control Register

31
TDES_CR 0xF803C000 Write-only

30 29 28 27 26 25 24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
SWRST
0
START

0: No effect 1: Starts Manual encryption/decryption process.

0: No effect. 1: Resets the TDES. A software triggered hardware reset of the TDES interface is performed.

1620
53.5.2 TDES Mode Register

31
TDES_MR 0xF803C004 Read-write

30 29 28 27 26 25 24
23
22
21
20
19
18
17
16
15
14
13
12
11
10 9
CFBS
8
LOD
7
6 5
OPMOD
4
2 1
SMOD
0
KEYMOD
TDESMOD
CIPHER
CIPHER: Processing Mode

0 (DECRYPT): Decrypts data. 1 (ENCRYPT): Encrypts data.
TDESMOD: ALGORITHM mode

0 (SINGLE_DES): Single DES processing using TDES_KEY1WRx registers. 1 (TRIPLE_DES): Triple DES processing using registers TDES_KEY1WRx, TDES_KEY2WRx and TDES_KEY3WRx if KEYMOD is set.
KEYMOD: Key Mode

0: Three-key algorithm is selected. 1: Two-key algorithm is selected. There is no need to write TDES_KEY3WRx registers.
SMOD: Start Mode

Value 0x0 0x1 0x2 Name MANUAL_START AUTO_START IDATAR0_START Description Manual Mode Auto Mode TDES_IDATAR0 access only Auto Mode
Values which are not listed in the table must be considered as reserved. If a DMA transfer is used, 0x2 must be configured. Refer to Section 53.4.3.2 DMA Mode for more details.
OPMOD: Operation Mode

Value 0x0 0x1 0x2 0x3 Name ECB CBC OFB CFB Description ECB: Electronic Code Book mode CBC: Cipher Block Chaining mode OFB: Output Feedback mode CFB: Cipher Feedback mode
For CBC-MAC operating mode, please set OPMOD to CBC and LOD to 1. The OFB and CFB modes of operation are only available if 2-key mode is selected (KEYMOD=1).

1621
LOD: Last Output Data Mode

0: No effect. After each end of encryption/decryption, the output data is available either on the output data registers (Manual and Auto modes). In Manual and Auto modes, the DATRDY flag is cleared when at least one of the Output Data registers is read. 1: The DATRDY flag is cleared when at least one of the Input Data Registers is written. No more Output Data Register reads are necessary between consecutive encryptions/decryptions (See Last Output Data Mode on page 1616.). Warning: In DMA mode, reading to the Output Data registers before the last data encryption/decryption process may lead to unpredictable result.
CFBS: Cipher Feedback Data Size

Value 0x0 0x1 0x2 0x3 Name SIZE_64BIT SIZE_32BIT SIZE_16BIT SIZE_8BIT Description 64-bit 32-bit 16-bit 8-bit

1622
53.5.3 TDES Interrupt Enable Register

31
TDES_IER 0xF803C010 Write-only

30 29 28 27 26 25 24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
URAD
0
DATRDY

1623
53.5.4 TDES Interrupt Disable Register

31
TDES_IDR 0xF803C014 Write-only

30 29 28 27 26 25 24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
URAD
0
DATRDY

1624
53.5.5 TDES Interrupt Mask Register

31
TDES_IMR 0xF803C018 Read-only

30 29 28 27 26 25 24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
URAD
0
DATRDY

1625
53.5.6 TDES Interrupt Status Register

31
TDES_ISR 0xF803C01C Read-only

30 29 28 27 26 25 24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
URAT
7 6 5 4
URAD
0
DATRDY
DATRDY: Data Ready

0: Output data is not valid. 1: Encryption or decryption process is completed. DATRDY is cleared when a Manual encryption/decryption occurs (START bit in TDES_CR) or when a software triggered hardware reset of the TDES interface is performed (SWRST bit in TDES_CR). LOD = 0 (TDES_MR): In Manual and Auto mode, the DATRDY flag can also be cleared when at least one of the Output Data Registers is read. LOD = 1 (TDES_MR): In Manual and Auto mode, the DATRDY flag can also be cleared when at least one of the Input Data Registers is written.

0: No unspecified register access has been detected since the last SWRST. 1: At least one unspecified register access has been detected since the last SWRST. URAD bit is reset only by the SWRST bit in the TDES_CR control register.
URAT: Unspecified Register Access

Value 0x0 0x1 0x2 0x3 Name IDR_WR_PROCESSING ODR_RD_PROCESSING MR_WR_PROCESSING WOR_RD_ACCESS Description Input Data Register written during the data processing when SMOD=0x2 mode. Output Data Register read during the data processing. Mode Register written during the data processing. Write-only register read access.
Only the last Unspecified Register Access Type is available through the URAT field. URAT field is reset only by the SWRST bit in the TDES_CR control register.

1626
53.5.7 TDES Key 1 Word Register x

31
TDES_KEY1WRx 0xF803C020 Write-only

30 29 28 27 26 25 24
KEY1W
23 22 21 20 19 18 17 16
KEY1W
15 14 13 12 11 10 9 8
KEY1W
7 6 5 4 3 2 1 0
KEY1W
KEY1W: Key 1 Word

The two 32-bit Key 1 Word Registers allow to set the 64-bit cryptographic key used for encryption/decryption. KEY1W0 corresponds to the first word of the key and KEY1W1 to the last one. These registers are write-only to prevent the key from being read by another application.

1627

31

30 29 28 27 26 25 24
KEY2W
23 22 21 20 19 18 17 16
KEY2W
15 14 13 12 11 10 9 8
KEY2W
7 6 5 4 3 2 1 0
KEY2W
KEY2W: Key 2 Word

The two 32-bit Key 2 Word Registers allow to set the 64-bit cryptographic key used for encryption/decryption. KEY2W0 corresponds to the first word of the key and KEY2W1 to the last one. These registers are write-only to prevent the key from being read by another application. Note: . KEY2WRx registers are not used in DES mode.

1628

31

30 29 28 27 26 25 24
KEY3W
23 22 21 20 19 18 17 16
KEY3W
15 14 13 12 11 10 9 8
KEY3W
7 6 5 4 3 2 1 0
KEY3W
KEY3W: Key 3 Word

The two 32-bit Key 3 Word Registers allow to set the 64-bit cryptographic key used for encryption/decryption. KEY3W0 corresponds to the first word of the key and KEY3W1 to the last one. These registers are write-only to prevent the key from being read by another application. Note: KEY3WRx registers are not used in DES mode, TDES with two-key algorithm selected.

1629
53.5.10 TDES Input Data Register x

31
TDES_IDATARx 0xF803C040 Write-only

30 29 28 27 26 25 24
IDATA
23 22 21 20 19 18 17 16
IDATA
15 14 13 12 11 10 9 8
IDATA
7 6 5 4 3 2 1 0
IDATA
IDATA: Input Data

The two 32-bit Input Data registers allow to set the 64-bit data block used for encryption/decryption. IDATA0 corresponds to the first word of the data to be encrypted/decrypted, and IDATA1 to the last one. These registers are write-only to prevent the input data from being read by another application.

1630
53.5.11 TDES Output Data Register x

31
TDES_ODATARx 0xF803C050 Read-only

30 29 28 27 26 25 24
ODATA
23 22 21 20 19 18 17 16
ODATA
15 14 13 12 11 10 9 8
ODATA
7 6 5 4 3 2 1 0
ODATA
ODATA: Output Data

The two 32-bit Output Data registers contain the 64-bit data block which has been encrypted/decrypted. ODATA1 corresponds to the first word, ODATA2 to the last one.

1631
53.5.12 TDES Initialization Vector Register x

31
TDES_IVRx 0xF803C060 Write-only

30 29 28 27 26 25 24
IV
23 22 21 20 19 18 17 16
IV
15 14 13 12 11 10 9 8
IV
7 6 5 4 3 2 1 0
IV
IV: Initialization Vector

The two 32-bit Initialization Vector registers are used to set the 64-bit initialization vector data block, which is used by some modes of operation as an additional initial input. IV1 corresponds to the first word of the Initialization Vector, IV2 to the last one. These registers are write-only to prevent the Initialization Vector from being read by another application. Note: These registers are not used for ECB mode and must not be written.

1632
54.
54.1
Secure Hash Algorithm (SHA)

Description
The Secure Hash Algorithm (SHA) is compliant with the American FIPS (Federal Information Processing Standard) Publication 180-2 specification. The 512/1024-bit block of message is respectively stored in 16/32 x 32-bit registers (SHA_IDATARx/SHA_IODATARx) which are all write-only. As soon as the input data is written, the hash processing may be started. The registers comprising the block of a padded message must be entered consecutively. Then the message digest is ready to be read out on the 5 up to 8/16 x 32-bit output data registers (SHA_IODATARx) or through the DMA channels.
54.2
z z z
Supports Secure Hash Algorithm (SHA1, SHA224, SHA256, SHA384, SHA512) Compliant with FIPS Publication 180-2 Configurable Processing Period:
z z
85 Clock Cycles to get a fast SHA1 runtime, 88 clock cycles for SHA384,SHA512 or 209 Clock Cycles for Maximizing Bandwidth of Other Applications 72 Clock Cycles to get a fast SHA224, SHA256 runtime or 194 Clock Cycles for Maximizing Bandwidth of Other Applications
z z
Connection to DMA Channel Capabilities Optimizes Data Transfers Double Input Buffer Optimizes Runtime
54.3

The SHA may be clocked through the Power Management Controller (PMC), so the programmer must first configure the PMC to enable the SHA clock.
54.3.2 Interrupt
The SHA interface has an interrupt line connected to the Interrupt Controller. Handling the SHA interrupt requires programming the interrupt controller before configuring the SHA.
Table 54-1. Peripheral IDs Instance SHA ID 42

1633
54.4
The Secure Hash Algorithm (SHA) module requires a padded message according to FIPS180-2 specification. The first block of the message must be indicated to the module by a specific command. The SHA module produces a N-bit message digest each time a block is written and processing period ends. N is 160 for SHA1, 224 for SHA224, 256 for SHA256, 384 for SHA384, 512 for SHA512.
54.4.1 SHA Algorithm

The module can process SHA1, SHA224, SHA256, SHA384, SHA512 by means of a configuration field in the SHA_MR register.
54.4.2 Processing Period

The processing period can be configured. The short processing period allows to allocate bandwidth to the SHA module whereas the long processing period allocates more bandwidth on the system bus to other applications (example: DMA channels not associated with SHA). In SHA1 mode, the shortest processing period is 85 clock cycles + 2 clock cycles for start command synchronization. The longest period is 209 clock cycles + 2 clock cycles. In SHA384, SHA512 mode, the shortest processing period is 88 clock cycles + 2 clock cycles for start command synchronization. The longest period is 209 clock cycles + 2 clock cycles. In SHA256 and SHA224 mode, the shortest processing period is 72 clock cycles + 2 clock cycles for start command synchronization. The longest period is 194 clock cycles + 2 clock cycles.
54.4.3 Double Input Buffer

The input data register can be double-buffered to reduce the runtime of large files. This mode allows to write a new message block while the previous message block is being processed. This is only possible when DMA accesses are performed (SMOD=0x2). The DUALBUFF field in SHA_MR register must be set to 1 to get access to double buffer.
54.4.4 Start Modes

The SMOD field in the SHA Mode Register (SHA_MR) is used to select the hash processing start mode. 54.4.4.1 Manual Mode The sequence is as follows:
z z z z z z z
Set the bit DATRDY (Data Ready) in the SHA Interrupt Enable Register (SHA_IER), depending on whether an interrupt is required or not at the end of processing. For the first block of a message, the FIRST command must be set by writing a 1 into the corresponding bit of the Control Register (SHA_CR). For the other blocks, there is nothing to write in this Control Register. Write the block to be processed in the Input Data Registers. Set the START bit in the SHA Control Register SHA_CR to begin the processing. When the processing completes, the bit DATRDY in the SHA Interrupt Status Register (SHA_ISR) raises. If an interrupt has been enabled by setting the bit DATRDY in SHA_IER, the interrupt line of the SHA is activated. Repeat the write procedure for each block, start procedure and wait for the interrupt procedure up to the last block of the entire message. Each time the start procedure is complete, the DATRDY flag is cleared. After the last block is processed (DATRDY flag is set, if an interrupt has been enabled by setting the bit DATRDY in SHA_IER, the interrupt line of the SHA is activated), read the message digest in the Output Data Registers. The DATRDY flag is automatically cleared when reading the SHA_IODATARx registers.

1634
54.4.4.2 Auto Mode Auto Mode is similar to Manual Mode, except that in this mode, as soon as the correct number of Input Data Registers is written, processing is automatically started without any action in the control register. 54.4.4.3 DMA Mode The DMA can be used in association with the SHA to perform the algorithm on a complete message without any action by the software during processing. The SMOD field of the SHA_MR must be set to 0x2. The DMA must be configured with non incremental addresses. The start address of any transfer descriptor must be set to point to the SHA_IDATAR0 register. The DMA chunk size must be set to transfer, for each trigger request, 16 words of 32 bits when processing SHA1/SHA256 algorithms or 32 words of 32 bits when SHA384/SHA512 are being used.
Figure 54-1. Enable DMA Channels
Enable DMA Channels associated with SHA_IDATARx registers
Message Processing (Multiple Block) BTC/ channel 0 DATRDY Message fully transferred Message fully processed SHA result can be read Write accesses into SHA_IDATARx
54.4.4.4 SHA Register Endianism In ARM processor based products, the AHB bus and processors manipulate data in Little Endian form. However, following the protocol of FIPS 180-2 specification, data is collected, processed and stored by the SHA module in a Big Endian form. The data presented to the SHA module (written to SHA_IDATAxR) must be in Little Endian form. The data read from the SHA module (read from SHA_IODATAxR) will be in Little Endian form. The SHA interface automatically converts into Big Endian format words that are presented into Little Endian. Likewise, the SHA interface returns hash results into Little Endian format even if the internal processing is Big Endian. Managing how data is presented to the SHA registers should be managed by software. As a clarification of this process consider the following example. If the first 64 bits of a message (according to FIPS180-2, i.e. Big Endian format) to be processed is 0xcafe_dede_0123_4567 then the SHA_IDATA0R and SHA_IDATA1R registers should be written with the following pattern:
z z
SHA_IDATA0R = 0xefac SHA_IDATA1R = 0xeded
In a Little Endian system, the message starting with pattern 0xcafe_dede_0123_4567 will be stored into memories as follows:
z z z z
0xca stored at initial offset (for example 0x00), then 0xfe stored at initial offset + 1 (i.e. 0x01), 0xde stored at initial offset + 2 (i.e. 0x02), 0xde stored at initial offset +3 (i.e. 0x03).

1635
Lets assume the message is received through a serial to parallel communication channel, the first received character is 0xca and stored at first memory location (initial offset), second octet being 0xfe is stored at initial offset + 1. When reading on a 32-bit Little Endian system bus, the first word read back from system memory is 0x_dede_feca. When the SHA_IODATAxR registers are read, the hash result is organized in Little Endian format, allowing system memory storage in the same format as the message. Taking an example from the FIPS 180-2 specification Appendix B.1, the endianism conversion can be observed. For this example, the 512-bit message is: 0x6162638000000000000000000000000000000000000000000000000000000000000000000000000000000000 0000000000000000000000000000000000000018 and the expected SHA-256 result is: 0xba7816bf_8f01cfea_414140de_5dae2223_b00361a3_96177a9c_b410ff61_f20015ad If the message has not already been stored in the system memory, the first step is to convert the input message to Little Endian before writing to the SHA_IDATAxR registers. This would result in a write of: SHA_IDATA0R = 0x80636261...... SHA_IDATA15R = 0x18000000 The data in the output message digest registers, SHA_IODATAxR, contain SHA_IODATAxR = 0xbf1678ba... SHA_IODATA7R = 0xad1500f2 which is the Little Endian format of 0xba7816bf,..., 0xf20015ad. Reading SHA_IODATA0R to SHA_IODATA1R and storing into a Little Endian memory system forces hash results to be stored in the same format as the message. When the output message is read, the user can convert back to Big Endian for a resulting message value of: 0xba7816bf_8f01cfea_414140de_5dae2223_b00361a3_96177a9c_b410ff61_f20015ad
54.4.5 Security Features

When an unspecified register access occurs, the URAD bit in the Interrupt Status Register (SHA_ISR) raises. Its source is then reported in the Unspecified Register Access Type field (URAT). Only the last unspecified register access is available through the URAT field. Several kinds of unspecified register accesses can occur:
z z z z
Input Data Register written during the data processing in DMA mode Input/Output Data Register read during the data processing Mode Register written during the data processing Write-only register read access
The URAD bit and the URAT field can only be reset by the SWRST bit in the SHA_CR control register.

1636
54.5
Secure Hash Algorithm (SHA) User Interface
Table 54-2. Register Mapping Offset 0x00 0x04 0x08-0x0C 0x10 0x14 0x18 0x1C 0x20-0x3C 0x40 ... 0x7C 0x80 ... 0x9C 0xA0 ... 0xBC 0x94-0xFC Register Control Register Mode Register Reserved Interrupt Enable Register Interrupt Disable Register Interrupt Mask Register Interrupt Status Register Reserved Input Data 0 Register ... Input Data 15 Register Input/Output Data 0 Register ... Input/Output Data 7 Register Input/Output Data 8 Register ... Input/Output Data 15 Register Reserved SHA_IDATAR0 ... SHA_IDATAR15 SHA_IODATAR0 ... SHA_IODATAR7 SHA_IODATAR8 ... SHA_IODATAR15 Write-only ... Write-only Read-write ... Read-write Read-write ... Read-write Name SHA_CR SHA_MR SHA_IER SHA_IDR SHA_IMR SHA_ISR Access Write-only Read-write Write-only Write-only Read-only Read-only Reset 0x0000100 0x0 0x0 ... 0x0 ... 0x0 0x0 ... 0x0

1637
54.5.1 SHA Control Register

31
SHA_CR 0xF8034000 Write-only

30 29 28 27 26 25 24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
SWRST
0
FIRST
START

0: No effect 1: Starts Manual hash algorithm process
FIRST: First Block of a Message

0: No effect 1: Indicates that the next block to process is the first one of a message.

0: No effect. 1: Resets the SHA. A software triggered hardware reset of the SHA interface is performed.

1638
54.5.2 SHA Mode Register

31
SHA_MR 0xF8034004 Read-write

30 29 28 27 26 25 24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
DUALBUFF
8
7 6 5
3 2
ALGO
1 0
PROCDLY
SMOD
SMOD: Start Mode

Value 0x0 0x1 0x2 Name MANUAL_START AUTO_START IDATAR0_START Description Manual Mode Auto Mode SHA_IDATAR0 access only Auto Mode
Values which are not listed in table must be considered as reserved. If a DMA transfer is used, either 0x1 or 0x2 must be configured. Refer to DMA Mode on page 1635 for more details.
PROCDLY: Processing Delay

Value 0x0 0x1 Name SHORTEST LONGEST Description SHA processing runtime is the shortest one SHA processing runtime is the longest one
When SHA1 algorithm is processed, runtime period is either 85 or 209 clock cycles. When SHA256 or SHA224 algorithm is processed, runtime period is either 72 or 194 clock cycles. When SHA384 or SHA512 algorithm is processed, runtime period is either 88 or 209 clock cycles.
ALGO: SHA Algorithm.

Value 0x0 0x1 0x2 0x3 0x4 Name SHA1 SHA256 SHA384 SHA512 SHA224 Description SHA1 algorithm processed SHA256 algorithm processed SHA384 algorithm processed SHA512 algorithm processed SHA224 algorithm processed
Values which are not listed in table must be considered as reserved.

1639
DUALBUFF: Dual Input BUFFer

Value 0x0 0x1 Name INACTIVE ACTIVE Description SHA_IDATARx and SHA_IODATARx cannot be written during processing of previous block. SHA_IDATARx and SHA_IODATARx can be written during processing of previous block when SMOD=0x2. It speeds up the overall runtime of large files.

1640
54.5.3 SHA Interrupt Enable Register

31
SHA_IER 0xF8034010 Write-only

30 29 28 27 26 25 24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
URAD
0
DATRDY

1641
54.5.4 SHA Interrupt Disable Register

31
SHA_IDR 0xF8034014 Write-only

30 29 28 27 26 25 24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
URAD
0
DATRDY

1642
54.5.5 SHA Interrupt Mask Register

31
SHA_IMR 0xF8034018 Read-only

30 29 28 27 26 25 24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
URAD
0
DATRDY

1643
54.5.6 SHA Interrupt Status Register

31
SHA_ISR 0xF803401C Read-only

30 29 28 27 26 25 24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
URAT
7 6 5 4
URAD
0
DATRDY
DATRDY: Data Ready

0: Output data is not valid. 1: 512-bit block process is completed. DATRDY is cleared when a Manual process occurs (START bit in SHA_CR) or when a software triggered hardware reset of the SHA interface is performed (SWRST bit in SHA_CR).

0: No unspecified register access has been detected since the last SWRST. 1: At least one unspecified register access has been detected since the last SWRST. URAD bit is reset only by the SWRST bit in the SHA_CR control register. URAT field indicates the unspecified access type.
URAT: Unspecified Register Access Type

Value 0x0 0x1 0x2 0x3 Description Input Data Register 0 to 15 written during the data processing in DMA mode. (URAD=0x1 and URAT=0x0 can occur only if DUALBUFF is cleared in SHA_MR) Output Data Register read during the data processing. Mode Register written during the data processing. Write-only register read access.
Only the last Unspecified Register Access Type is available through the URAT field. URAT field is reset only by the SWRST bit in the SHA_CR control register.

1644
54.5.7 SHA Input Data x Register

31
SHA_IDATARx [x=0.15] 0xF8034040 Write-only

30 29 28 27 26 25 24
IDATA
23 22 21 20 19 18 17 16
IDATA
15 14 13 12 11 10 9 8
IDATA
7 6 5 4 3 2 1 0
IDATA
IDATA: Input Data

The 32-bit Input Data registers allow to load the data block used for hash processing. These registers are write-only to prevent the input data from being read by another application. SHA_IDATAR0 corresponds to the first word of the block, SHA_IDATAR15 to the last word of the last block in case SHA algorithm is set to SHA1,SHA224,SHA256 or SHA_IODATA15R to the last word of the block if SHA algorithm is SHA384 or SHA512 (please refer to SHA Input/Output Data x Register on page 1646).

1645
54.5.8 SHA Input/Output Data x Register

31
SHA_IODATARx [x=0.15] 0xF8034080 Read-write

30 29 28 27 26 25 24
IODATA
23 22 21 20 19 18 17 16
IODATA
15 14 13 12 11 10 9 8
IODATA
7 6 5 4 3 2 1 0
IODATA
IODATA: Input/Output Data

These registers can be used to read the resulting message digest and to write the second part of the message block when the SHA algorithm is SHA-384 or SHA-512. SHA_IODATA0R to SHA_IODATA15R registers can be written or read but reading these offsets does not return the content of corresponding parts (words) of the message block. Only results from SHA calculation can be read through these registers. When SHA processing is in progress, these registers return 0x0000. SHA_IODATA0R corresponds to the first word of message digest; SHA_IODATA5R to the last one in SHA1 mode, SHA_ODATA6R in SHA224, SHA_IODATA7R in SHA256, SHA_IODATA11R in SHA384 or SHA_IODATA15R in SHA512. When SHA224 is selected, the content of SHA_ODATA7R must not be taken into account. When SHA384 is selected, the content of SHA_IODATA12R to SHA_IODATA15R must not be taken into account.

1646
55.
55.1
Electrical Characteristics
Absolute Maximum Ratings
Table 55-1. Absolute Maximum Ratings* Operating Temperature (Industrial)..............-40 C to + 85 C Junction Temperature..................................................125C Storage Temperature.................................-60C to + 150C Voltage on Input Pins with Respect to Ground.....-0.3V to VDDIO+0.3V(+ 4V max) Maximum Operating Voltage (VDDCORE, VDDPLLA, VDDUTMIC)...........................1.5V (VDDIODDR).................................................................2.0V (VDDIOM, VDDIOPx, VDDUTMII, VDDOSC, VDDANA and VDDBU)..................................................4.0V Total DC Output Current on all I/O lines...................350 mA *NOTICE: Stresses beyond those listed under Absolute Maximum Ratings may cause permanent damage to the device. This is a stress rating only and functional operation of the device at these or other conditions beyond those indicated in the operational sections of this specification is not implied. Exposure to absolute maximum rating conditions for extended periods may affect device reliability.

1647
55.2
DC Characteristics
The following characteristics are applicable to the operating temperature range: TA = -40C to +85C, unless otherwise specified.
Table 55-2. DC Characteristics
Symbol VVDDCORE VVDDCORErip VVDDUTMIC VVDDUTMII VVDDBU VVDDBUrip VVDDPLLA VVDDPLLArip VVDDOSC VVDDOSCrip VVDDIOM VVDDIODDR VVDDIOP0 VVDDIOP1 VVDDANA VIL Parameter DC Supply Core VDDCORE ripple DC Supply UDPHS and UHPHS UTMI+ Core DC Supply UDPHS and UHPHS UTMI+ Interface DC Supply Backup VDDBU ripple DC Supply PLLA VDDPLLA ripple DC Supply Oscillator VDDOSC ripple DC Supply EBI I/Os DC Supply SDRAM I/Os DC Supply Peripheral I/Os DC Supply Peripheral I/Os DC Supply Analog Input Low-level Voltage VVDDIO in 3.3V range VVDDIO in 1.8V range Input High-level Voltage VVDDIO in 3.3V range VVDDIO in 1.8V range All PIO lines VVDDIOx in 3.3V range All PIO lines VVDDIOx in 1.8V range Output Low-level Voltage Output High-level Voltage IO Max IO Max All PIO lines VVDDIOx in 1.8V range All PIO lines NTRST and NRST VVDDIOx in 3.3V range VVDDIO - 0.4 100 160 310 k 45 70 130 DDR2 or LP-DDR usage LP-DDR2 usage 1.65/3.0 1.7 1.14 1.65 1.65 3.0 -0.3 -0.3 2 0.7 x VVDDIO 0.34 V 0.21 0.4 V V 3.3 1.8/3.3 1.8 1.2 1.65 1.08 1.2 1.08 3.0 1.65 1.2 3.3 Conditions Min 1.08 Typ 1.2 Max 1.32 20 1.32 3.6 3.6 30 1.32 10 3.6 30 1.95/3.6 1.95 1.30 3.6 3.6 3.6 0.8 V 0.3 x VVDDIO VVDDIO + 0.3 VVDDIO + 0.3 V Units V mVrms V V V mVrms V mVrms V mVrms V V V V V
VIH
VHYS
Schmitt trigger Hysteresis
VOL VOH
RPULL
Pull-up Resistance Pull-down Resistance

1648
Table 55-2. DC Characteristics (Continued)

TDI, TCK, NTRST, BMS, TMS, TDO, NRST, NANDRDY, PA0-PA27, PA29PA31, PB28-PB31, PC0-PC14, PC16PC23, PC25-PC31, PD0-PD8, PD10PD11, PD13-PD31 VVDDIOPx in 1.8V range IO Output Current TDI, TCK, NTRST, BMS, TMS, TDO, NRST, NANDRDY, PA0-PA27, PA29PA31, PB28-PB31, PC0-PC14, PC16PC23, PC25-PC31, PD0-PD8, PD10PD11, PD13-PD31 VVDDIOPx in 3.3V range NCS3, NRD, NWE, PA28, PB0-PB27, PC15, PC24, PD9, PD12, PE0-PE31, D0D15 TMS, TDO, NRST, NANDRDY, PA0PA27, PA29-PA31, PB28-PB31, PC0PC14, PC16-PC23, PC25-PC31, PD0PD8, PD10-PD11, PD13-PD19 NCS3, NRD, NWE, PA28, PB0-PB27, PC15, PC24, PD9, PD12, PE0-PE31, D0D15 On VVDDCORE = 1.0V, MCK = 0 Hz, excluding POR All inputs driven TMS, TDI, TCK, NRST = 1 On VVDDBU = 3.3V, Logic cells consumption, excluding POR All inputs driven WKUP = 0 TA = 25C TA = 85C 1.2 A 1.6 TA = 25C 0.3 mA TA = 85C 4.8 3 (LO_DRIVE) 6 (ME_DRIVE) 7 (HI_DRIVE) mA 8 (LO_DRIVE) 26 (ME_DRIVE) 38 (HI_DRIVE)
5 (LO_DRIVE)
16 (ME_DRIVE)
22 (HI_DRIVE)
30 13
RSERIAL
Serial Resistor
ISC
Static Current

1649
55.3
Power Consumption
z z z z
Typical power consumption of PLLs, Slow Clock and Main Oscillator Power consumption of power supply in four different modes: Active, Idle, Ultra Low-power and Backup Power consumption by peripheral: calculated as the difference in current measurement after having enabled then disabled the corresponding clock Software used for power consumption measurements: DHRYSTONE
55.3.1 Power Consumption versus Modes

The values in Table 55-3, Table 55-4 and Table 55-6 are estimated values of the power consumption with operating conditions as follows:
z z z z z z z
VDDIOM = 1.8V VDDIOP0 and 1= 3.3V VDDPLLA = 1.2V VDDCORE = 1.2V VDDBU = 3.3V TA = 25 C There is no consumption on the I/Os of the device
Figure 55-1. Measures Schematics

VDDBU AMP1 VDDCORE AMP2
55.3.2 Active Mode

Active Mode is the normal running mode with the core clock running from a PLL. The power management controller can be used to adapt the frequency and to disable the peripheral clocks.

1650
Table 55-3 represents the power consumption estimated on the power supplies.
Table 55-3. Power Consumption in Active Mode Mode Conditions ARM Core clock is 400 MHz. Active MCK is 133 MHz. All peripheral clocks activated. Onto AMP2. ARM Core clock is 498 MHz. MCK is 166 MHz. Active All peripheral clocks activated. VDDCORE = 1.25V. Onto AMP2. ARM Core clock is 536 MHz. MCK is 134 MHz. Active All peripheral clocks activated. VDDCORE = 1.25V. Onto AMP2. 110 mA 106 mA 88 mA Consumption Unit
Table 55-4. Power Consumption by Peripheral in Active Mode Peripheral PIO Controller USART UHPHS UDPHS ADC TWI SPI PWM HSMCI SSC Timer Counter Channels DMA SMD ISI EMAC GEMAC LCDC CAN SHA AES TDES TRNG Consumption 4.5 2.0 22.5 24.5 3.0 1.0 2.5 6.0 13.5 3.0 2.5 27.5 5.0 4.0 10.0 21.5 13.0 4.5 4.0 2.0 1.0 0.5 A/MHz(1) Unit
Note:
1.
Reference frequency is the peripheral frequency. It can be a division (1,2,4,8) of MCK. Please refer to the PMC section for more details.

1651
In order to reduce power consumption, each Peripheral Clock has been timed to an optimum value. The peripheral frequency is programmable with the help of a divider in the PMC_PCR register. Programming a divider giving higher frequency than required will lead to an unpredictable behavior. Table Table 55-5 gives the maximum frequency values for each peripheral.
Table 55-5. Maximum Values for each Peripheral Peripheral name DBGU PIT SMD USART[3..0] UART[1..0] TWI[2..0] HSMCI[2..0] SPI[1..0] TC[1..0] PWM ADC SSC[1..0] CAN[1..0] Description Debug Unit Interrupt Periodic Interval Timer Interrupt SMD Soft Modem USART 0 UART Two-wire Interface High Speed Multimedia Card Interface 0 Serial Peripheral Interface 0 Timer Counter Pulse Width Modulation Controller Touch Screen ADC Controller Synchronous Serial Controller CAN Controller 0 Max Frequency 66 MHz 133 MHz 24 MHz 66 MHz 66 MHz 66 MHz 133 MHz 133 MHz 66 MHz 133 MHz 66 MHz 66 MHz 66 MHz
55.3.3 Low-power Modes

The various low-power modes are described below: 55.3.3.1 Backup Mode The purpose of Backup Mode is to achieve the lowest power consumption possible in a system which is performing periodic wake-ups to perform tasks but not requiring fast startup time. The Zero-power Power-on Reset, RTC, Backup registers and 32 kHz oscillator (RC or Crystal Oscillator selected by software in the Supply Controller) are running. The core supply is off. The system can be awakened from this mode through the WKUP0 pin or an RTC wake-up event. Backup mode is entered with the help of the Shutdown Controller that asserts the SHDN output pin. The SHDN pin is to be connected to the Enable of the VDDCORE regulator. Exit from Backup mode happens if one of the following enable wake-up events occurs:
z z
WKUP0 pin (level transition, configurable debouncing) RTC alarm
The system will restart as for a reset event. 55.3.3.2 Idle Mode The purpose of Idle Mode is to optimize power consumption of the device versus response time. In this mode, only the core clock is stopped. The peripheral clocks, including the DDR Controller clock, can be enabled. The current consumption in this mode is application dependent. This mode is entered via the Wait for Interrupt (WFI) instruction and PCK disabling. The processor can be awakened from an interrupt. The system will resume where it was before entering in WFI mode.

1652
55.3.3.3 Ultra Low-power Mode The purpose of Ultra Low-power Mode is to reduce the power consumption of the device to the minimum without disconnecting VDDCORE power supply. It is a combination of very low frequency operations and Idle Mode. This mode is entered via the following steps: 1. 2. Set the DDR in Self Refresh Mode Reduce the system clock (PCK and MCK) to the minimum with the help of the PMC:
z z z z
PCK and MCK configuration is to be defined regarding the expected power consumption and wake-up time. Please refer to Table 55-6 for details PLLs are disabled. CKGR_PLLAR (eventually CKGR_PLLBR) is set to 0x3F00. CKGR_UCKR is set to 0. Main Oscillator is disabled. MOSCXTEN is set to 0 in CKGR_MOR. Eventually 12 MHz RC Oscillator is disabled. MOSCRCEN is set to 0 in CKGR_MOR.
3/ Enter in Wait for Interrupt (WFI) mode and disable the PCK clock. The processor can be awakened from an interrupt. Once revived, the system must reprogram the system clocks (OSC, PLL, PCK, MCK, DDRCK) to recover the previous state. Data is maintained in the external memory. 55.3.3.4 Low-power Mode Summary Table The modes detailed above are the main low-power modes. Each part can be set to on or off separately and wake-up sources can be individually configured. Table 55-6 below shows a summary of the configurations of the low-power modes.
Table 55-6. Low-power Mode Configuration Summary
32K RC, 12 MHz RC, 32 kHz Osc, RTC, Backup Registers, Core POR VDDCORE Memory (Backup Region) Regulator Peripherals OFF Backup ON OFF (Not powered)
Mode
Mode Entry
PIO State Potential while in Wake-Up Core at Low- power PIO State at Sources Wake Up Mode Wake Up Reset Inputs with pull-ups
Consumption(2) 1.2 A typ(3) 19 mA on VDDCORE
Wake-Up Time(1) Start-up time
Shutdown WKUP0 pin Reset Controller RTC alarm
Idle
ON
ON
Powered WFI (Not clocked)
Any interrupt
Clocked Previous back at state saved full speed
Unchanged @ 133 MHz 5 - 8 mA for each PLL(4)
350 ns @ 133 MHz
Ultra Lowpower
ON
ON
Powered (Not clocked)
DDR in Self Refresh PMC WFI
1780 A @ 12 MHz 3.9 s @ 12 MHz Any interrupt Clocked back at previous one 520 A @ 750 kHz Previous state saved Unchanged 455 A @ 187 kHz 432 A @ 32 kHz 429 A @ 512 Hz 60 s @ 750 kHz 230 s @ 187 kHz 1.4 ms @ 32 kHz 89 ms @ 512 Hz
Notes: 1. When considering wake-up time, the time required to start the PLL is not taken into account. Once started, the device works with the Main Oscillator. The user has to add the PLL start-up time if it is needed in the system. The wake-up time is defined as the time taken for wake up until the first instruction is fetched. 2. The external loads on PIOs are not taken into account in the calculation. 3. Total Current consumption. 4. Depends on MCK frequency.

1653
55.4
Clock Characteristics
55.4.1 Processor Clock Characteristics

Table 55-7. Processor Clock Waveform Parameters Symbol Parameter Conditions VDDCORE[1.08V, 1.32V] T = 85C 1/(tCPPCK) Processor Clock Frequency VDDCORE[1.2V, 1.32V] T = 85C 125(1) 536 Min 125(1) Max 400 MHz Units
Note:
1. For DDR2 usage only, there are no limitations to LP-DDR and LP-DDR2.
55.4.2 Master Clock Characteristics

The master clock is the maximum clock at which the system is able to run. It is given by the smallest value of the internal bus clock and EBI clock.
Table 55-8. Master Clock Waveform Parameters Symbol Parameter Conditions VDDCORE[1.08V, 1.32V] T = 85C 1/(tCPMCK) Master Clock Frequency VDDCORE[1.2V, 1.32V], VDDIODDR[1.75V, 1.9V], DDR2 mode only, T = 85C Min 125(1) Max 134 MHz 125
(1)
Units
166
Note:
1. For DDR2 usage only, there are no limitations to LP-DDR, LP-DDR2.

1654
55.5
Main Oscillator Characteristics
Table 55-9. Main Oscillator Characteristics Symbol Parameter Conditions FREQ = 00, 01(2) Freq Operating Frequency FREQ = 10 FREQ = 11 FREQ = 00, 01 CLOAD(1) CLEXT(1) CPARA Load Capacitance FREQ = 10 FREQ = 11 External Load Capacitance Parasitic Load Capacitance Duty Cycle tST IDDST PON IDD ON Startup Time Standby Current Consumption Standby mode FREQ = 00, 01 Drive Level FREQ = 10 FREQ = 11 Current Dissipation @ 12 MHz 0.5 0.6 40 0.7 0.8 60 2 0.1 150 300 400 1 mA W Min 8 20 44 12.5 4 4 Typ Max 16 24 48 17.5 8 6 pF pF % ms A pF MHz Unit
Note:
1. The external capacitors value can be determined by using the following formula: CLEXT = (2*CLOAD) - CBOARD - CROUTING - CPACKAGE - (CPARA*2), where CLEXT: external capacitor value which must be soldered from XIN to GND and XOUT to GND, CLOAD: crystal targeted load. Please refer to CLOAD parameter in the electrical specification, CBOARD: external calculated (or measured) parasitic value due to board, CROUTING: parasitic capacitance due to internal chip routing, typically 1.5 pF, CPACKAGE: parasitic capacitance due to package and bonding, typically 0.75 pF, CPARA: internal parasitic load due to internal structure. 2. FREQ field defines the input frequency for the UTMI and the Main Oscillator. It is defined in the UTMI Clock Trimming Register (SFR_UTMICKTRIM) located in the SFR section. Take care to select the correct FREQ value, this has a direct influence on the USB frequency.
Figure 55-2. Main Oscillator Schematics
XIN
XOUT GNDPLL 1K CCRYSTAL
CLEXT
CLEXT

1655
55.5.1 Crystal Oscillator Characteristics

The following characteristics are applicable to the operating temperature range: TA = -40C to 85C and worst case of power supply, unless otherwise specified.
Table 55-10. Crystal Characteristics Symbol Parameter Conditions @16 MHz, FREQ = 00, 01 ESR Equivalent Series Resistance @24 MHz, FREQ = 10 @48 MHz, FREQ = 11 FREQ = 00, 01 CM Motional Capacitance
(1)
Min
Typ
Max 80 150 100
Unit
5 1.3 2
9 2 3.2 7 1 1.3 pF fF
FREQ = 10 with ESRmax = 150 FREQ = 11 with ESRmax = 100 FREQ = 00, 01
CS
Shunt Capacitance
FREQ = 10 with ESRmax = 150 FREQ = 11 with ESRmax = 100
Note:
1. FREQ field defines the input frequency for the UTMI and the Main Oscillator. It is defined in UTMI Clock Trimming Register, located in SFR section. Take care to select the correct FREQ value, this has a direct influence on USB frequency.
55.5.2 XIN Clock Characteristics

Table 55-11. XIN Clock Electrical Characteristics Symbol 1/(tCPXIN) tCPXIN tCHXIN tCLXIN CIN RIN VIN Parameter XIN Clock Frequency XIN Clock Period XIN Clock High Half-period XIN Clock Low Half-period XIN Input Capacitance XIN Pulldown Resistor XIN Voltage
(1) (1) (1)
Conditions
Min
Max 50
Units MHz ns
20 0.4 x tCPXIN 0.4 x tCPXIN 0.6 x tCPXIN 0.6 x tCPXIN 25 500 VDDOSC VDDOSC
ns ns pF k V
Note:
1. These characteristics apply only when the Main Oscillator is in bypass mode (i.e. when MOSCEN = 0 and OSCBYPASS = 1) in the CKGR_MOR register. See PMC Clock Generator Main Oscillator Register in the PMC section.

1656
55.6
12 MHz RC Oscillator Characteristics
Table 55-12. 12 MHz RC Oscillator Characteristics Symbol F0 Duty IDD ON tON IDD STDBY Parameter Nominal Frequency Duty Cycle Power Consumption Oscillation Startup time Standby consumption Conditions Min 11.4 45 Typ 12 50 220 10 22 Max 12.6 55 Units MHz % A s A
55.7
32 kHz Oscillator Characteristics
Table 55-13. 32 kHz Oscillator Characteristics Symbol 1/(tCP32KHz) CCRYSTAL32 CLEXT32(2) Parameter Crystal Oscillator Frequency Load Capacitance External Load Capacitance Duty Cycle RS = 50 k(1) tST Startup Time RS = 100 k(1) PON Drive Level CCRYSTAL32 = 6 pF CCRYSTAL32 = 12.5 pF CCRYSTAL32 = 6 pF CCRYSTAL32 = 12.5 pF Crystal @ 32.768 kHz CCRYSTAL32 = 6 pF CCRYSTAL32 = 12.5 pF 40 6 6 19 50 60 400 900 600 1200 0.2 Conditions Min Typ 32 768 12.5 Max Unit kHz pF pF pF % ms ms ms ms W
Notes: 1. RS is the equivalent series resistance. 2. CLEXT32 is determined by taking into account internal, parasitic and package load capacitance.
Figure 55-3. 32 kHz Oscillator Schematics
XIN32
XOUT32
GNDBU
CCRYSTAL32
CLEXT32
CLEXT32

1657
55.7.1 32 kHz Crystal Characteristics

Table 55-14. 32 kHz Crystal Characteristics Symbol ESR CM CS Parameter Equivalent Series Resistor Rs Motional Capacitance Shunt Capacitance Conditions Crystal @ 32768 kHz Crystal @ 32768 kHz Crystal @ 32768 kHz RS = 50 k IDD ON Current dissipation RS = 50 k
(1) (1)
Min
Typ 50
Max 100 3 2
Unit k fF pF A A A A nA
0.6 0.6 0.3 0.45 0.45 0.45 5
CCRYSTAL32 = 6 pF CCRYSTAL32 = 12.5pF CCRYSTAL32 = 6 pF
0.65 0.7 0.75 0.65 10
RS = 100 k
(1)
RS = 100 k(1) CCRYSTAL32 = 12.5 pF IDD STDBY Standby consumption
55.7.2 XIN32 Clock Characteristics

Table 55-15. XIN32 Clock Characteristics Symbol 1/(tCPXIN32) tCPXIN32 tCHXIN32 tCLXIN32 tCLCH32 tCLCL32 CIN32 RIN32 VIN32 VINIL32 VINIH32 Parameter XIN32 Clock Frequency XIN32 Clock Period XIN32 Clock High Half-period XIN32 Clock Low Half-period XIN32 Clock Rise time XIN32 Clock Fall time XIN32 Input Capacitance XIN32 Pulldown Resistor XIN32 Voltage XIN32 Input Low Level Voltage XIN32 Input High Level Voltage
(1) (1) (1) (1) (1)
Conditions
Min
Max 44
Units kHz s s s ns ns
22 11 11 400 400 6 4 VDDBU -0.3 0.7 x VVDDBU VDDBU 0.3 x VVDDBU VVDDBU + 0.3
pF M V V V
Note:
1. These characteristics apply only when the 32768 kHz Oscillator is in bypass mode (i.e. when RCEN = 0, OSC32EN = 0, OSCSEL = 1 and OSC32BYP = 1) in the SCKCR register. See Slow Clock Selection in the PMC section.

1658
55.8
32 kHz RC Oscillator Characteristics

Parameter Crystal Oscillator Frequency Duty Cycle Conditions Min 30.4 45 290 After startup time 1.1 Typ 32 Max 33.6 55 500 2.1 0.2 Unit kHz % s A A
Table 55-16. 32 kHz RC Oscillator Characteristics Symbol 1/(tCPRCz) tST IDD ON IDD STDBY
Startup Time Power Consumption Oscillation Standby consumption
55.9
PLL Characteristics
Parameter Output Frequency Input Frequency Current Consumption Startup Time Active mode @ 800 MHz Standby mode Conditions VDDPLLA[1.08V, 1.32V] VDDPLLA[1.2V, 1.32V] Min 400 400 8 13 1 25 Typ Max 800 1000 50 14 200 100 Unit MHz MHz mA A s
Table 55-17. PLLA Characteristics Symbol FOUT FIN IPLL T
55.9.1 UTMI PLL Characteristics

Table 55-18. Phase Lock Loop Characteristics Symbol FIN FOUT IPLL T Parameter Input Frequency Output Frequency Current Consumption Startup Time Active mode Standby mode Conditions Min Typ 12 480 5 8 1.5 50 Max Unit MHz MHz mA A s
55.10 USB HS Characteristics

55.10.1 Electrical Characteristics
Table 55-19. Electrical Parameters Symbol RPUI RPUA Setting time TBIAS TOSC TSETTLING Bias settling time Oscillator settling time Settling time With Crystal 12 MHz FIN = 12 MHz 0.3 20 2 0.5 s ms ms Parameter Bus Pull-up Resistor on Upstream Port (idle bus) Bus Pull-up Resistor on Upstream Port (upstream port receiving) Conditions in LS or FS Mode Min Typ 1.5 Max Unit k k
in LS or FS Mode
15

1659
55.10.2 Static Power Consumption

Table 55-20. Static Power Consumption Symbol IBIAS Parameter Bias current consumption on VBG HS Transceiver and I/O current consumption LS / FS Transceiver and I/O current consumption Core, PLL, and Oscillator current consumption no connection (1) Conditions Min Typ Max 1 8 3 2 Unit A A A A
IVDDUTMII
IVDDUTMIC
Note:
1. If cable is connected, add 200 A (Typical) due to pull-up/pull-down current consumption.
55.10.3 Dynamic Power Consumption

Table 55-21. Dynamic Power Consumption Symbol IBIAS Parameter Bias current consumption on VBG HS Transceiver current consumption HS Transceiver current consumption IVDDUTMII LS / FS Transceiver current consumption LS / FS Transceiver current consumption LS / FS Transceiver current consumption IVDDUTMIC PLL, Core and Oscillator current consumption HS transmission HS reception FS transmission 0m cable(1) FS transmission 5m cable(1) FS reception(1) Conditions Min Typ 0.7 47 18 4 26 3 5.5 Max 0.8 60 27 6 30 4.5 9 Unit mA mA mA mA mA mA mA
Note:
1. Including 1 mA due to pull-up/pull-down current consumption.

1660
55.11 12-Bit ADC Characteristics

Table 55-22. Analog Power Supply Characteristics Symbol Parameter ADC Analog Supply VVDDANA Max. Voltage Ripple RMS value, 10 kHz to 20 MHz Sleep Mode IVDDANA Current Consumption Fast Wake Up Mode Normal Mode (IBCTL= 00) Normal Mode (IBCTL= 01) 4 2.2 3.3 4.4 20 8 3.4 4.8 6.4 mV A mA mA mA Conditions 12-bit Min 2.4 Typ 3.0 Max 3.6 Units V
Note:
Use IBCTL = 00 for Sampling Frequency below 500 kHz and IBCTL = 01 between 500 kHz and 1 MHz.
Table 55-23. Channel Conversion Time and ADC Clock Symbol fADC tCP_ADC fS Parameter ADC Clock Frequency ADC Clock Period Sampling Frequency From OFF Mode to Normal Mode: - Voltage Reference OFF - Analog Circuitry OFF tSTART-UP ADC Startup time From Standby Mode to Normal Mode: - Voltage Reference ON - Analog Circuitry OFF tTRACKTIM tCONV tSETTLE Track and Hold Time Conversion Time Settling Time Settling time to change offset and gain 200 See Section 55.11.1.1 Track and Hold Time versus Source Output Impedance for more details 4 8 12 20 30 Conditions Min 1 50 Typ Max 20 1000 1 40 s Units MHz ns MHz
160 20
ns TCP_ADC ns
Table 55-24. External Voltage Reference Input Parameter ADVREF Input Voltage Range, 12-bit ADVREF Current ADVREF Input DC impedance 6 8 Conditions 2.4V < VVDDIN < 3.6V Min 2.4 Typ Max VDDANA 600 10 Units V A k

1661
55.11.1 Static performance characteristics

In the tables that follow, the LSB is relative to the analog scale:
z
Single Ended (ex: ADVREF=3.0V),

z z z
Gain = 1, LSB = (3.0V / 4096) = 732 uV Gain = 2, LSB = (1.5V / 4096) = 366 uV Gain = 4, LSB = (750mV / 4096) = 183 uV Gain = 0.5, LSB = (6.0V / 4096) = 1465 uV Gain = 1, LSB = (3.0V / 4096) = 732 uV Gain = 4, LSB = (750mV / 4096) = 366 uV
Differential (ex: ADVREF=3.0V),

z z z
Table 55-25. INL, DNL, 12-bit mode, VDDANA 2.4V to 3.6V supply voltage conditions Parameter Resolution Integral Non-linearity (INL) All mode, all gain Typ=3.0V supply, 27C All mode, all gain Typ=3.0V supply, 27C -4 Conditions Min Typ 12 1 +4 Max Units Bit LSB
Differential Non-linearity (DNL)
-2
0.5
+2
LSB
Table 55-26. Gain Error, 12-bit Mode, VDDANA 2.4V to 3.6V supply voltage conditions Parameter Gain Error, Differential mode, uncalibrated Gain Error, Differential mode, after calibration Gain Error, Single Ended, uncalibrated Conditions Any gain and offset values Gain = 0.5 Gain = 1 Gain = 2 Any gain and offset values Gain = 1 Gain Error, Single Ended, after calibration Gain = 2 Gain = 4 Min -45 -8 -16 -24 -45 -8 -16 -24 Typ Max +45 +8 +16 +24 LSB +45 +8 +16 +24 Units
Table 55-27. Error offset with or without calibration, 12-bit Mode, VDDANA 2.4V to 3.6V supply voltage conditions Parameter Offset Error Conditions Single Ended, Gain = 1 / Differential mode Gain = 0.5 Single Ended, Gain = 2 / Differential mode Gain = 1 Single Ended, Gain = 4 / Differential mode Gain = 2 Min -8 -16 -32 Typ Max +8 +16 +32 LSB Units

1662
Table 55-28. Dynamic Performance Characteristics in Single ended and 12 bits mode (1) Parameter Signal to Noise Ratio - SNR Total Harmonic Distortion - THD Signal to Noise and Distortion - SINAD Effective Number of Bits - ENOB
(1)
Conditions
Min 54
Typ 61 -80
Max
Units dB
-68
dB dB Bits
54 8.7
61 9.8
Note:
1. ADC Clock (FADC) = 20 MHz, Fs=1MHz, Fin = 127 kHz, IBCTL = 01, FFT using 1024 points or more, Frequency band = [1 kHz, 500 kHz] Nyquist conditions fulfilled.
Table 55-29. Dynamic Performance Characteristics in Differential and 12 bits mode(1) Parameter Signal to Noise Ratio - SNR Total Harmonic Distortion - THD Signal to Noise and Distortion - SINAD Effective Number of Bits - ENOB
(1)
Conditions
Min 58
Typ 63 -80
Max
Units dB
-72
dB dB Bits
58 9.3
63 10.2
Note:
1. ADC Clock (FADC)= 20 MHz, Fs=1 MHz, Fin=127 kHz, IBCTL = 01, FFT using 1024 points or more, Frequency band = [1 kHz, 500 kHz] Nyquist conditions fulfilled.

1663
55.11.1.1 Track and Hold Time versus Source Output Impedance The following figure gives a simplified acquisition path.
Figure 55-4. Simplified Acquisition Path
ADC Input Zsource Mux. Sample & Hold 12-bit ADC Core
Ron
Csample
During the tracking phase the ADC needs to track the input signal during the tracking time shown below:
z z
10-bit mode: tTRACK = 0.042 x ZSOURCE + 160 12-bit mode: tTRACK = 0.054 x ZSOURCE + 205
With tTRACK expressed in ns and ZSOURCE expressed in Ohms. Two cases must be considered: 1. The calculated tracking time (tTRACK) is lower than 15 tCP_ADC. Set TRANSFER = 1 and TRACTIM = 0. In this case, the allowed ZSOURCE can be computed versus the ADC frequency with the hypothesis of tTRACK = 15 tCP_ADC: Where tCP_ADC = 1/fADC . See Table 55-30 on page 1664.
Table 55-30. Source Impedance Values fADC = ADC clock (MHz) 20.00 16.00 10.67 8.00 6.40 5.33 4.57 4.00 3.56 3.20 2.91 2.67 2.46 2.29 2.13 2.00 1.00 ZSOURCE (k) for 12 bits 10 14 22 31 40 48 57 66 74 83 92 100 109 118 126 135 274 ZSOURCE (k) for 10 bits 14 19 30 41 52 63 74 85 97 108 119 130 141 152 164 175 353

1664
2.
The calculated tracking time (tTRACK) is higher than 15 tCP_ADC.
Set TRANSFER = 1 and TRACTIM = 0. In this case, a timer will trigger the ADC in order to set the correct sampling rate according to the Track time. The maximum possible sampling frequency will be defined by tTRACK in nanoseconds, computed by the previous formula but with minus 15 tCP_ADC and plus TRANSFER time.
z z
10 bit mode: 1/fS = tTRACK - 15 tCP_ADC + 5 tCP_ADC 12 bit mode: 1/fS = tTRACK - 15 tCP_ADC + 5 tCP_ADC CSAMPLE and RON are taken into account in the formulas
Note:
Table 55-31. Analog Inputs Parameter Input Voltage Range Input Leakage Current Input Capacitance Min 0 Typ Max VADVREF 0.5 8 A pF Units
Note:
Input Voltage range can be up to VDDANA without destruction or over-consumption. If VDDIO < VADVREF max input voltage is VDDIO.
55.11.1.2 ADC Application Information For more information on data converter terminology, please refer to the application note: Data Converter Terminology, Atmel lit 6022. http://www.atmel.com/dyn/resources/prod_documents/doc6022.pdf

1665
55.12 POR Characteristics

General presentation of Power-On-Reset (POR) characteristics is provided in Figure 55-5.
Figure 55-5. General Presentation of POR Behavior
VDD
Vth+ VthVop Vnop
Static
Dynamic
NRST Tres
When a very slow (versus TRES) supply rising slope is applied on POR VDD pin, the reset time becomes negligible and the reset signal is released when VDD raises upper than Vth+. When a very fast (versus TRES) supply rising slope is applied on POR VDD pin, the voltage threshold becomes negligible and the reset signal is released after TRES time. It is the smallest possible reset time.
55.12.1 Core Power Supply POR Characteristics

Table 55-32. Core Power Supply POR Characteristics Symbol Vth+ VthTRES IDD Parameter Threshold Voltage Rising Threshold Voltage Falling Reset Time Current Consumption After TRES Conditions Minimum Slope of +2.3 V/ms Min 0.85 0.8 100 Typ 1.02 0.96 260 250 Max 1.08 1.06 600 500 Units V V s nA
55.12.2 Backup Power Supply POR Characteristics

Table 55-33. Backup Power Supply POR Characteristics Symbol Vth+ VthTRES IDD Parameter Threshold Voltage Rising Threshold Voltage Falling Reset Time Current Consumption After TRES Conditions Minimum Slope of +1.9 V/ms Min 1.50 1.39 80 Typ 1.54 1.46 180 300 Max 1.60 1.53 600 500 Units V V s nA

1666
55.13 HSMC Timings

55.13.1 Timing conditions
HSMC Timings are given in MAX corners. Timings assuming a capacitance load on data, control and address pads are given in Table 55-34.
Table 55-34. Capacitance Load Corner Supply 3.3V 1.8V 50 pF 30 pF MAX 50 pF 30 pF MIN
In tables that follow, tCPMCK is MCK period.
55.13.2 Timing extraction

55.13.2.1 Read Timings
Table 55-35. SMC Read Signals - NRD Controlled (READ_MODE = 1) Symbol Parameter VDDIOM supply 1.8V Min 3.3V 1.8V Max 3.3V Units
NO HOLD SETTINGS (nrd hold = 0) SMC1 SMC2 Data Setup before NRD High Data Hold after NRD High 9.6 0 6.8 0 ns ns
HOLD SETTINGS (nrd hold 0) SMC3 SMC4 Data Setup before NRD High Data Hold after NRD High 9.4 0 6.6 0 ns ns
HOLD or NO HOLD SETTINGS (nrd hold 0, nrd hold =0) SMC5 NBS0/A0, NBS1, NBS2/A1, NBS3, A2 - A25 Valid before NRD High (nrd setup + nrd pulse)* tCPMCK + 2.4 (nrd setup + nrd pulse - ncs rd setup) * tCPMCK + 2 nrd pulse * tCPMCK + 1.8 (nrd setup + nrd pulse)* tCPMCK + 2.1 (nrd setup + nrd pulse - ncs rd setup) * tCPMCK + 1.9 nrd pulse * tCPMCK +1.2 ns
SMC6
NCS low before NRD High
ns
SMC7
NRD Pulse Width
ns
Table 55-36. SMC Read Signals - NCS Controlled (READ_MODE = 0) Symbol Parameter VDDIOM supply 1.8V Min 3.3V 1.8V Max 3.3V Units
NO HOLD SETTINGS (ncs rd hold = 0) SMC8 SMC9 Data Setup before NCS High Data Hold after NCS High 18.5 0 16.2 0 ns ns

1667
Table 55-36. SMC Read Signals - NCS Controlled (READ_MODE = 0) HOLD SETTINGS (ncs rd hold 0) SMC10 SMC11 Data Setup before NCS High Data Hold after NCS High 17 0 14.7 0 ns ns
HOLD or NO HOLD SETTINGS (ncs rd hold 0, ncs rd hold = 0) NBS0/A0, NBS1, NBS2/A1, NBS3, A2 - A25 valid before NCS High (ncs rd setup + ncs rd pulse)* tCPMCK + 19.1 (ncs rd setup + ncs rd pulse nrd setup)* tCPMCK + 16.8 ncs rd pulse length * tCPMCK +2 (ncs rd setup + ncs rd pulse)* tCPMCK + 18.8 (ncs rd setup + ncs rd pulse - nrd setup)* tCPMCK +16.1 ncs rd pulse length * tCPMCK + 1.5
SMC12
ns
SMC13
NRD low before NCS High
ns
SMC14
NCS Pulse Width
ns
55.13.2.2 Write Timings

Table 55-37. SMC Write Signals - NWE Controlled (WRITE_MODE = 1) Symbol Parameter 1.8V Supply Min 3.3V Supply 1.8V Supply Max 3.3V Supply Units
HOLD or NO HOLD SETTINGS (nwe hold 0, nwe hold = 0) SMC15 SMC16 SMC17 Data Out Valid before NWE High NWE Pulse Width NBS0/A0 NBS1, NBS2/A1, NBS3, A2 - A25 valid before NWE low nwe pulse * tCPMCK + 2.4 nwe pulse * tCPMCK + 1.9 nwe setup * tCPMCK + 2.6 (nwe setup ncs rd setup + nwe pulse) * tCPMCK + 2.2 nwe pulse * tCPMCK + 1.6 nwe pulse * tCPMCK + 1.3 nwe pulse * tCPMCK + 2 (nwe setup ncs rd setup + nwe pulse) * tCPMCK + 1.9 ns ns ns
SMC18
NCS low before NWE high
ns
HOLD SETTINGS (nwe hold 0) SMC19 NWE High to Data OUT, NBS0/A0 NBS1, NBS2/A1, NBS3, A2 - A25 change NWE High to NCS Inactive (1) nwe hold * tCPMCK + 2.3 (nwe hold - ncs wr hold)* tCPMCK + 1.9 nwe hold * tCPMCK +1.1 (nwe hold - ncs wr hold)* tCPMCK +1 ns
SMC20
ns
NO HOLD SETTINGS (nwe hold = 0) SMC21 NWE High to Data OUT, NBS0/A0 NBS1, NBS2/A1, NBS3, A2 - A25, NCS change(1) 2 1.6 ns
Notes: 1. hold length = total cycle duration - setup duration - pulse duration. hold length is for ncs wr hold length or NWE hold length.

1668
Table 55-38. SMC Write NCS Controlled (WRITE_MODE = 0) Symbol Parameter 1.8V Supply SMC22 SMC23 SMC24 Data Out Valid before NCS High NCS Pulse Width NBS0/A0 NBS1, NBS2/A1, NBS3, A2 - A25 valid before NCS low ncs wr pulse * tCPMCK + 9.8 ncs wr pulse * tCPMCK + 2 ncs wr setup * tCPMCK + 4.3 (ncs wr setup nwe setup + ncs pulse)* tCPMCK + 2.4 ncs wr hold * tCPMCK + 2.6 (ncs wr hold nwe hold)* tCPMCK + 5.5 Min 3.3V Supply ncs wr pulse * tCPMCK + 9.1 ncs wr pulse * tCPMCK + 1.5 ncs wr setup * tCPMCK + 3.7 (ncs wr setup nwe setup + ncs pulse)* tCPMCK + 2.1 ncs wr hold * tCPMCK + 1.7 (ncs wr hold nwe hold)* tCPMCK + 5.6 1.8V Supply Max 3.3V Supply Units ns ns ns
SMC25
NWE low before NCS high
ns
SMC26
NCS High to Data Out, NBS0/A0, NBS1, NBS2/A1, NBS3, A2 - A25, change NCS High to NWE Inactive
ns
SMC27
ns
Figure 55-6. SMC Timings - NCS Controlled Read and Write
SMC12
SMC12
SMC24
SMC26
A0/A1/NBS[3:0]/A2-A25
SMC13
SMC13
NRD
NCS
SMC14 SMC9
SMC14
SMC23
SMC8
SMC10
SMC11
SMC22
SMC26
D0 - D15
SMC25 SMC27
NWE
NCS Controlled READ with NO HOLD
NCS Controlled READ with HOLD
NCS Controlled WRITE

1669
Figure 55-7. SMC Timings - NRD Controlled Read and NWE Controlled Write
SMC5 SMC17 SMC21 SMC5 SMC17 SMC19
A0/A1/NBS[3:0]/A2-A25
SMC6 SMC18 SMC21 SMC6 SMC18 SMC20
NCS
NRD
SMC7
SMC7
SMC1
SMC2
SMC15
SMC21
SMC3
SMC4
SMC15
SMC19
D0 - D31
NWE
SMC16
SMC16
NRD Controlled READ with NO HOLD
NWE Controlled WRITE with NO HOLD
NRD Controlled READ with HOLD
NWE Controlled WRITE with HOLD
55.13.2.3 FPGA Timings SMC and PCK2 can be used to interface a FPGA. PCK2 is to be programmed to output MCK. READ_MODE and WRITE_MODE are to be configured to 0.
Figure 55-8. FPGA Timings
PCK2 FPGA1 ADD, NCS, NDR, NWE FPGA3 DATA Output FPGA4 FPGA2
FPGA5 DATA Input
FPGA6
Shaded areas are not valid
Table 55-39. FPGA Timings vs PCK2 Symbol Parameter VDDIOM Supply VDDCORE Supply FPGA1 Address, NCS, NRD, NWE Setup before PCK2 falling edge Address, NCS, NRD, NWE Hold after PCK2 falling edge Data Out Setup before PCK2 falling edge [1.65:1.95] [1.08:1.32] 1.78 Min [3.0:3.6V] [1.08:1.32] 2.07 [3.0:3.6V] [1.2:1.32] 1.78 [1.65:1.95] [1.08:1.32] Max [3.0:3.6V] [1.08:1.32] [3.0:3.6V] [1.2:1.32] Units V V ns
FPGA2
0.29
0.35
1.22
ns
FPGA3
1.46
1.77
1.59
ns

1670
Table 55-39. FPGA Timings vs PCK2 FPGA4 FPGA5 FPGA6 Data Out Hold after PCK2 falling edge Data In Setup before PCK2 falling edge Data In Hold after PCK2 falling edge 0 0.27 1.10 14.15 0 12.08 0 10.39 0 ns ns ns
55.14 SPI Timings

55.14.1 Timing Conditions
Timings assuming a capacitance load on MISO, SPCK and MOSI are given in Table 55-40.
Table 55-40. Capacitance Load for MISO, SPCK and MOSI (Product Dependent) Corner Supply 3.3V 1.8V 40 pF 20 pF MAX 40 pF 40 pF MIN
These values may be product dependant and should be confirmed by the specification.
55.14.2 Timing Extraction

Figure 55-9. SPI Master mode 1 and 2
SPCK
SPI0 MISO
SPI1
SPI2 MOSI
Figure 55-10. SPI Master mode 0 and 3
SPCK
SPI3 MISO
SPI4
SPI5 MOSI

1671
Figure 55-11. SPI Slave Mode 0 and 3
NPCS0
SPI12 SPCK
SPI13
SPI6 MISO
SPI7 MOSI
SPI8
Figure 55-12. SPI Slave Mode 1 and 2
NPCS0
SPI13 SPI12
SPCK
SPI9 MISO
SPI10 MOSI
SPI11
Figure 55-13. SPI Slave Mode - NPCS Timings
SPI14 SPI6 SPCK (CPOL = 0) SPI12 SPCK (CPOL = 1) SPI16 MISO SPI9
SPI15
SPI13

1672
Table 55-41. SPI Timings with 3v3 peripheral supply Symbol Parameter Cond Master Mode SPI0 SPI1 SPI2 SPI3 SPI4 SPI5 MISO Setup time before SPCK rises MISO Hold time after SPCK rises SPCK rising to MOSI MISO Setup time before SPCK falls MISO Hold time after SPCK falls SPCK falling to MOSI Slave Mode SPI6 SPI7 SPI8 SPI9 SPI10 SPI11 SPI12 SPI13 SPI14 SPI15 SPI16 SPCK falling to MISO MOSI Setup time before SPCK rises MOSI Hold time after SPCK rises SPCK rising to MISO MOSI Setup time before SPCK falls MOSI Hold time after SPCK falls NPCS0 setup to SPCK rising NPCS0 hold after SPCK falling NPCS0 setup to SPCK falling NPCS0 hold after SPCK rising NPCS0 falling to MISO valid 5.8(1) 2 0.3 5.8(1) 2 0.3 3.5 19.5 1.9 19.8 8.8 9.4(1) 9.4(1) ns ns ns ns ns ns ns ns ns ns ns 8.7 5.9 0.8(1) 8.7 5.8 0.9(1) 3.8(1) 3.7(1) ns ns ns ns ns ns Min Max Units
Notes: 1. For output signals, Min and Max access time must be extracted. The Min access time is the time between the SPCK rising or falling edge and the signal change. The Max access timing is the time between the SPCK rising or falling edge and the signal stabilizes. Figure 55-13 illustrates Min and Max accesses for SPI2. The same applies for SPI5, SPI6, SPI9.
Table 55-42. SPI Timings with 1v8 Peripheral Supply Symbol Parameter Cond Master Mode SPI0 SPI1 SPI2 SPI3 SPI4 SPI5 MISO Setup time before SPCK rises MISO Hold time after SPCK rises SPCK rising to MOSI MISO Setup time before SPCK falls MISO Hold time after SPCK falls SPCK falling to MOSI Slave Mode SPI6 SPI7 SPCK falling to MISO MOSI Setup time before SPCK rises 3.2(1) 1.7 9.4(1) ns ns 10.9 0 0.8
(1)
Min
Max
Units
ns ns 3.5
(1)
ns ns ns
10.4 0 1.2(1) 3.2(1)
ns

1673
Table 55-42. SPI Timings with 1v8 Peripheral Supply Symbol SPI8 SPI9 SPI10 SPI11 SPI12 SPI13 SPI14 SPI15 SPI16 Parameter MOSI Hold time after SPCK rises SPCK rising to MISO MOSI Setup time before SPCK falls MOSI Hold time after SPCK falls NPCS0 setup to SPCK rising NPCS0 hold after SPCK falling NPCS0 setup to SPCK falling NPCS0 hold after SPCK rising NPCS0 falling to MISO valid Cond Min 1 2.9
(1)
Max
Units ns
10.9
(1)
ns ns ns ns ns ns ns
1.7 1 2.1 19 3.2 19.6 10.6
ns
Figure 55-14. Min and Max access time for SPI output signal
SPCK
SPI0 MISO
SPI1
SPI2max MOSI
SPI2min

1674
55.15 MPDDRC Timings

55.15.1 Board Design Constraints
As SAMA5D3 series embeds impedance calibrated pads, there is no capacitive constraints on DDR signals. However, a board must be designed and equipped in order to respect DQS setup times, including propagation time and intrinsic delay in SDRAM device. In all cases, line length to memory device must not exceed 5 cm.
55.15.2 DDR2-SDRAM
Table 55-43. System Clock Waveform Parameters Symbol Parameter Conditions VDDCORE[1.08V,1.32V] T = 85C tDDRCK DDRCK Cycle time VDDCORE[1.2V,1.32V], VDDIODDR[1.75V,1.9V], T = 85C VDDCORE[1.08V,1.32V] T = 85C tSDQS DQS setup time VDDCORE[1.2V,1.32V], VDDIODDR[1.75V,1.9V], T = 85C 0.7 ns 0.6 ns 6.0 8.0 ns Min 7.5 Max 8.0 Units ns
55.15.3 LPDDR1-SDRAM
Table 55-44. System Clock Waveform Parameters Symbol tDDRCK tSDQS Parameter DDRCK Cycle time Conditions VDDCORE[1.08V,1.32V] T = 85C VDDCORE[1.08V,1.32V] T = 85C Min 7.5 Max Units ns
DQS setup time
7.3
ns
55.15.4 LPDDR2-SDRAM
Table 55-45. System Clock Waveform Parameters Symbol tDDRCK tSDQS Parameter DDRCK Cycle time Conditions VDDCORE[1.08V,1.32V] T = 85C VDDCORE[1.08V,1.32V] T = 85C Min 7.5 Max Units ns
DQS setup time
6.2
ns

1675
Figure 55-15. DQS Timings

tDDRCK DDRCK
tSDQS
DQS
55.16 SSC Timings

Timings assuming a capacitance load are given in Table 55-46.

Figure 55-16. SSC Transmitter, TK and TF in Output
TK (CKI =0)
TK (CKI =1)
SSC0 TF/TD

1676
Figure 55-17. SSC Transmitter, TK in Input and TF in Output
TK (CKI =0)
TK (CKI =1)
SSC1 TF/TD
Figure 55-18. SSC Transmitter, TK in Output and TF in Input

TK (CKI=0)
TK (CKI=1)
SSC2 TF
SSC3
SSC4 TD
Figure 55-19. SSC Transmitter, TK and TF in Input

TK (CKI=1)
TK (CKI=0)
SSC5 TF SSC7 TD
SSC6

1677
Figure 55-20. SSC Receiver RK and RF in Input

RK (CKI=0)
RK (CKI=1)
SSC8 RF/RD
SSC9
Figure 55-21. SSC Receiver, RK in Input and RF in Output

RK (CKI=1)
RK (CKI=0)
SSC8 RD SSC10 RF
SSC9
Figure 55-22. SSC Receiver, RK and RF in Output

RK (CKI=1)
RK (CKI=0)
SSC11 RD SSC13 RF
SSC12
Figure 55-23. SSC Receiver, RK in Output and RF in Input

RK (CKI=0)
RK (CKI=1)
SSC11 RF/RD
SSC12

1678
Table 55-47. SSC Timings with 3.3V Peripheral supply Symbol Parameter Cond Transmitter SSC0 SSC1 SSC2 SSC3 SSC4(1) SSC5 SSC6 SSC7(1) TK edge to TF/TD (TK output, TF output) TK edge to TF/TD (TK input, TF output) TF setup time before TK edge (TK output) TF hold time after TK edge (TK output) TK edge to TF/TD (TK output, TF input) TF setup time before TK edge (TK input) TF hold time after TK edge (TK input) TK edge to TF/TD (TK input, TF input) Receiver SSC8 SSC9 SSC10 SSC11 SSC12 SSC13 RF/RD setup time before RK edge (RK input) RF/RD hold time after RK edge (RK input) RK edge to RF (RK input) RF/RD setup time before RK edge (RK output) RF/RD hold time after RK edge (RK output) RK edge to RF (RK output) 0 tCPMCK 2.5
(2)
Min
Max
Units
1.2(2) 2.3 (2) 7.7 7.3 1.2 (+2*tCPMCK) 0 tCPMCK 2.4 (+3*tCPMCK)
(1)(2) (1)(2)
3.8 (2) 10.6 (2)
ns ns ns ns
2.1(+2*tCPMCK)
(1)(2)
ns ns ns
8.8 (+3*tCPMCK)
(1)(2)
ns
ns ns 8.8
(2)
ns ns ns
7.6 - tCPMCK tCPMCK - 2.2 0.7(2) 2.2(2)
ns
Notes: 1. Timings SSC4 and SSC7 depend on the start condition. When STTDLY = 0 (Receive start delay) and START = 4, or 5 or 7 (Receive Start Selection), two Periods of the MCK must be added to timings. 2. For output signals (TF, TD, RF), Min and Max access times are defined. The Min access time is the time between the TK (or RK) edge and the signal change. The Max access timing is the time between the TK edge and the signal stabilization. Figure 55-24 illustrates Min and Max accesses for SSC0. The same applies for SSC1, SSC4, and SSC7, SSC10 and SSC13.

1679
Table 55-48. SSC Timing with 1V8 Peripheral supply Symbol Parameter Cond Transmitter SSC0 SSC1 SSC2 SSC3 SSC4(1) SSC5 SSC6 SSC7(1) TK edge to TF/TD (TK output, TF output) TK edge to TF/TD (TK input, TF output) TF setup time before TK edge (TK output) TF hold time after TK edge (TK output) TK edge to TF/TD (TK output, TF input) TF setup time before TK edge (TK input) TF hold time after TK edge (TK input) TK edge to TF/TD (TK input, TF input) Receiver SSC8 SSC9 SSC10 SSC11 SSC12 SSC13 RF/RD setup time before RK edge (RK input) RF/RD hold time after RK edge (RK input) RK edge to RF (RK input) RF/RD setup time before RK edge (RK output) RF/RD hold time after RK edge (RK output) RK edge to RF (RK output) 0 tCPMCK 2.5
(2)
Min
Max
Units
1.2(2) 2.3 (2) 7.7 7.3 1.2 (+2*tCPMCK) 0 tCPMCK 2.4 (+3*tCPMCK)
(1)(2) (1)(2)
3.8 (2) 10.6 (2)
ns ns ns ns
2.1(+2*tCPMCK)
(1)(2)
ns ns ns
8.8 (+3*tCPMCK)
(1)(2)
ns
ns ns 8.8
(2)
ns ns ns
7.6 - tCPMCK tCPMCK - 2.2 0.7(2) 2.2(2)
ns
Figure 55-24. Min and Max Access Time of Output Signals
TK (CKI =1)
TK (CKI =0) SSC0min SSC0max TF/TD

1680
55.17 ISI Timings

Timings assuming capacitance loads are given in Table 55-49.

Figure 55-25. ISI Timing Diagram
PIXCLK
3
DATA[7:0] VSYNC HSYNC

Table 55-50. ISI Timings with Peripheral Supply 3.3V Symbol ISI1 ISI2 ISI3 Parameter DATA/VSYNC/HSYNC setup time DATA/VSYNC/HSYNC hold time PIXCLK frequency
Valid Data
Valid Data
Valid Data
Min 3.1 1
Max
Units ns ns
80
MHz
Table 55-51. ISI Timings with Peripheral Supply 1.8V Symbol ISI1 ISI2 ISI3 Parameter DATA/VSYNC/HSYNC setup time DATA/VSYNC/HSYNC hold time PIXCLK frequency Min 3.4 1.2 80 Max Units ns ns MHz

1681
55.18 MCI Timings

Figure 55-26. MCI Timings
MCI1 CLK MCI2 CMD_DAT Input MCI4 CMD_DAT Output Shaded areas are not valid MCI5 MCI3
Table 55-52. MCI Timings Symbol Parameter CLoad C = 25 pF MCI1 CLK frequency at Data transfer Mode C= 100 pF C= 250 pF CLK frequency at Identification Mode CLK Low time CLK High time CLK Rise time CLK Fall time CLK Low time CLK High time CLK Rise time CLK Fall time MCI2 MCI3 MCI4 MCI5 Input hold time Input setup time Output change after CLK rising Output valid before CLK rising C= 100 pF C= 100 pF C= 100 pF C= 100 pF C= 250 pF C= 250 pF C= 250 pF C= 250 pF 3 3 5 5 50 50 50 50 10 10 10 10 Min Max 25 20 20 400 Units MHz MHz MHz kHz ns ns ns ns ns ns ns ns ns ns ns ns

1682
55.19 EMAC/GMAC Timings

Timings assuming a capacitance load on data and clock are given in Table 55-53.
Table 55-53. Capacitance Load on Data, Clock Pads Corner Supply 3.3V 1.8V MAX 20 pF 20 pF 0 pF 0 pF MIN
55.19.2 Timing constraints

To satisfy the timings of standard given in Table 55-54 and Table 55-55, the Ethernet controller (EMAC) must be constrained in MAX corners.
Table 55-54. EMAC Signals Relative to EMDC Symbol EMAC1 EMAC2 EMAC3 Parameter Setup for EMDIO from EMDC rising Hold for EMDIO from EMDC rising EMDIO toggling from EMDC rising Min (ns) 10 ns 10 ns 0 ns(1) 300 ns(1) Max (ns)
Notes: 1. For EMAC output signals, Min and Max access time are defined. The Min access time is the time between the EDMC rising edge and the signal change. The Max access timing is the time between the EDMC rising edge and the signal stabilizes. Figure 55-27 illustrates Min and Max accesses for EMAC3.
Figure 55-27. Min and Max access time of EMAC output signals
EMDC EMAC1 EMDIO EMAC4 EMAC5 EMAC3 min EMAC2 EMAC3 max

1683
55.19.2.1 MII Mode

Table 55-55. EMAC MII Specific Signals Symbol EMAC4 EMAC5 EMAC6 EMAC7 EMAC8 EMAC9 EMAC10 EMAC11 EMAC12 EMAC13 EMAC14 EMAC15 EMAC16 Parameter Setup for ECOL from ETXCK rising Hold for ECOL from ETXCK rising Setup for ECRS from ETXCK rising Hold for ECRS from ETXCK rising ETXER toggling from ETXCK rising ETXEN toggling from ETXCK rising ETX toggling from ETXCK rising Setup for ERX from ERXCK Hold for ERX from ERXCK Setup for ERXER from ERXCK Hold for ERXER from ERXCK Setup for ERXDV from ERXCK Hold for ERXDV from ERXCK Min (ns) 10 10 10 10 10(1) 10(1) 10(1) 10 10 10 10 10 10 25(1) 25(1) 25(1) Max (ns)
Notes: 1. See Note (2) of Table 55-54.

1684
Figure 55-28. EMAC MII Mode

EMDC EMAC1 EMDIO EMAC4 ECOL EMAC6 ECRS EMAC7 EMAC5 EMAC2 EMAC3
ETXCK EMAC8 ETXER EMAC9 ETXEN EMAC10 ETX[3:0]
ERXCK EMAC11 ERX[3:0] EMAC13 ERXER EMAC15 ERXDV EMAC16 EMAC14 EMAC12

1685
Table 55-56. RMII Mode Symbol EMAC21 EMAC22 EMAC23 EMAC24 EMAC25 EMAC26 EMAC27 EMAC28 Parameter ETXEN toggling from EREFCK rising ETX toggling from EREFCK rising Setup for ERX from EREFCK rising Hold for ERX from EREFCK rising Setup for ERXER from EREFCK rising Hold for ERXER from EREFCK rising Setup for ECRSDV from EREFCK rising Hold for ECRSDV from EREFCK rising Min (ns) 2
(1)
Max (ns) 16(1) 16(1)
2(1) 4 2 4 2 4 2
Notes: 1. See Note (2) of Table 55-54.
Figure 55-29. EMAC RMII Timings

EREFCK EMAC21 ETXEN EMAC22 ETX[1:0] EMAC23 ERX[1:0] EMAC25 ERXER EMAC27 ECRSDV EMAC28 EMAC26 EMAC24

1686
55.20 UART in SPI Mode Timings

Timings assuming a capacitance load on MISO, SPCK and MOSI are given in Table 55-57.
Table 55-57. Capacitance Load for MISO, SPCK and MOSI (product dependent) Corner Supply 3.3V 1.8V 40 pF 20 pF MAX 40 pF 40 pF MIN

Figure 55-30. UART SPI Master mode
NPCSx SPI3 CPOL=1 SPI0 SPCK=0 CPOL=0 SPI4 MISO MOSI SPI4 MSB SPI1 SPI2 LSB
SPI5
Figure 55-31. UART SPI Slave mode
NPCS0
SPI12 SPCK
SPI13
SPI6 MISO
SPI7 MOSI
SPI8

1687
Figure 55-32. SPI Slave mode - NPCS timings
SPI14 SPI6 SPCK (CPOL = 0) SPI12 SPCK (CPOL = 1) SPI16 MISO SPI9
SPI15
SPI13
Table 55-58. UART SPI Timings 3.3V Peripheral Supply Symbol Parameter Cond Master Mode SPI0 SPI1 SPI2 SPI3 SPI4 SPI5 SPCK Period Input Data Setup Time Input Data Hold Time Chip Select Active to Serial Clock Output Data Setup Time Serial Clock to Chip Select Inactive Slave Mode SPI6 SPI7 SPI8 SPI9 SPI10 SPI11 SPI12 SPI13 SPI14 SPI15 SPI16 SPCK falling to MISO MOSI Setup time before SPCK rises MOSI Hold time after SPCK rises SPCK rising to MISO MOSI Setup time before SPCK falls MOSI Hold time after SPCK falls NPCS0 setup to SPCK rising NPCS0 hold after SPCK falling NPCS0 setup to SPCK falling NPCS0 hold after SPCK rising NPCS0 falling to MISO valid 0.1 5.6
(1)
Min
Max
Units
MCK/6 2.4 0.4 1.1 6.5 0.6
ns ns ns ns ns ns
2.2(1)
8.7(1)
ns ns ns
(1)
ns ns ns ns ns ns ns
2 0.5 1 0.9 0.8 0.4 7.6
ns
Notes: 1. For output signals, Min and Max access time must be extracted. The Min access time is the time between the SPCK rising or falling edge and the signal change. The Max access timing is the time between the SPCK rising or falling edge and the signal stabilizes. Figure 55-13 illustrates Min and Max accesses for SPI2. The same applies for SPI5, SPI6, SPI9.

1688
55.21 Two-wire Interface Characteristics

Table 55-59 describes the requirements for devices connected to the Two-wire Serial Bus. For timing symbols, please refer to Figure 55-33.
Table 55-59. Two-wire Serial Bus Requirements Symbol VIL VIH Vhys(1) VOL(1) tr(1) tof(1) Ii Ci(1) fTWCK(4) Parameter Input Low-voltage Input High-voltage Hysteresis of Schmitt Trigger Inputs Output Low-voltage Rise Time for both TWD and TWCK Output Fall Time from VIHmin to VILmax Input Current each I/O Pin Capacitance for each I/O Pin TWCK Clock Frequency MCK
(5)
Condition
Min -0.5 0.7 VCC 0.05 VCC(2)
Max 0.3 VCC VCC + 0.5 0.4 300 250 10 10 400 1000 ns -------------------Cb 300 ns ----------------Cb 3.45 0.9
Units V V V V ns ns A pF kHz s s s s s s s s s s ns ns s s s s
3 mA sink current 20 + 10 pF < Cb < 400 pF

(3)
0 20 + 0.1Cb(3)(2) 0.1Cb(3)(2) -10
0.1 VCC < Vi < 0.9 VCC > max(16 fTWCK, 250 kHz)(6)
0 V CC 0,4V ---------------------------3mA V CC 0,4V ---------------------------3mA 4.0 0.6 4.7 1.3 4.0 0.6 4.7 0.6 0 0 250 100 4.0 0.6 4.7 1.3
fTWCK 100 kHz Rp Value of Pull-up resistor fTWCK > 100 kHz fTWCK 100 kHz fTWCK > 100 kHz fTWCK 100 kHz fTWCK > 100 kHz fTWCK 100 kHz fTWCK > 100 kHz fTWCK 100 kHz fTWCK > 100 kHz fTWCK 100 kHz fTWCK > 100 kHz fTWCK 100 kHz fTWCK > 100 kHz fTWCK 100 kHz fTWCK > 100 kHz fTWCK 100 kHz fTWCK > 100 kHz
tHD;STA tLOW tHIGH tSU;STA tHD;DAT tSU;DAT tSU;STO tBUF
Hold Time (repeated) START Condition Low Period of the SCL Clock High period of the SCL clock Set-up time for a repeated START condition Data hold time Data setup time Setup time for STOP condition Bus free time between a STOP and START condition
Notes: 1. This parameter is characterized in Chip Number; it is not 100% tested. 2. Required only for fTWCK > 100 kHz. 3. Cb = capacitance of one bus line in pF. Per I2C Standard, Cb Max = 400pF 4. This requirement applies to all Chip Number 2-wire Serial Interface operation. Other devices connected to the 2-wire Serial Bus need only to satisfy the general fTWCK requirement. 5. fCK = CPU clock frequency 6. This requirement applies to all Chip Number 2-wire Serial Interface operation. Other devices connected to the Twowire Serial Bus need only obey the general fSCL requirement.

1689
Figure 55-33. Two-wire Serial Bus Timing

tof tLOW SCL tSU;STA SDA tHD;STA tHD;DAT tSU;DAT tSU;STO tHIGH tLOW tr
tBUF

1690
56.
Mechanical Characteristics
Figure 56-1. 324-ball LFBGA Package Drawing
TITLE
Low Profile Fine Pitch Ball Grid Array
324

1691
Table 56-1. 324-ball LFBGA Package Characteristics Moisture Sensitivity Level 3
Table 56-2. Device and 324-ball LFBGA Package Maximum Weight 400 mg
Table 56-3. Package Reference JEDEC Drawing Reference JESD97 Classification MO-275-KAAE-1 e8
Table 56-4. Package Information Ball Land Nominal Ball Diameter Solder Mask Opening Solder Mask Definition Solder 0.350 mm 0.05 0.30 mm 0.275 mm 0.05 SMD SAC105

1692
56.1
Marking
All devices are marked with the Atmel logo and the ordering code. Additional marking may be in one of the following formats:
YYWW V XXXXXXXXX
where
z z z z
ARM
YY: manufactory year WW: manufactory week V: revision XXXXXXXXX: lot number

1693
57.
SAMA5D3 Ordering Information

Ordering Code ATSAMA5D31A-CU Package BGA324 Package Type Green Temperature Operating Range Industrial -40C to 85C Industrial -40C to 85C Industrial -40C to 85C Industrial -40C to 85C
Table 57-1. SAMA5D3 Ordering Information
ATSAMA5D33A-CU
BGA324
Green
ATSAMA5D34A-CU
BGA324
Green
ATSAMA5D35A-CU
BGA324
Green

1694
58.
SAMA5D3 Series Errata

The present errata provides information on Rev. A parts.
58.1
Standard Boot Strategies
58.1.1 Boot ROM: Boot on MCI1 is Not Working

Boot on MCI1 is not working. Problem Fix/Workaround Use a different boot media.
58.1.2 Boot ROM: NAND Flash Detection using ONFI Parameters does Not Work
During Nandflash initialization, the ONFI parameter detection may not work correctly. This can lead to incorrect configuration of ECC settings and to reading wrong data from the Nandflash memory, thus making it impossible to boot from this memory. Problem Fix/Workaround When programming the bootable program in the Nandflash, always use the header method, with any Nandflash memory, ONFI compliant or not.
58.2
Static Memory Controller (HSMC)
58.2.1 HSMC: ECC Value after ERASE is Not Correct

When ERASE page command is issued, the NAND Flash page resets to "ALL_ONE" pattern. The ECC value of ALL_ONE is a fixed but non-zero pattern leading to error detection after page erase. Problem Fix/Workaround None.
58.3
LCD Controller (LCDC)
58.3.1 LCDC: LCDC PWM Is Not Usable

When slow clock is selected as the source clock to supply PWM with (CLKPWMSEL in LCDC_LCDCFG0), the output waveform generated is corrupted. When the MCK is selected the prescaler (PWMPS in LCDC_LCDCFG6) is not sized to generate the PWM output in a range of 200 Hz - 1 kHz. Problem Fix/Workaround Use standalone PWM output instead of LCDC embedded PWM.
58.4
PWM Controller (PWM)
58.4.1 PWM: Wrong Counter Event.

A synchronous channel (but not channel 0) can generate a wrong counter event, which does not match the alignment configuration of the channel 0. The issue occurs if the alignment configuration of the channel is different from that of channel 0. Problem Fix/Workaround None.

1695
58.5
Gigabit Ethernet MAC (GMAC)
58.5.1 GMAC: TX Packet Buffer DMA Lockup when Reading Used Status
In the following conditions:
z z z z z
Frame N-1 has completed the transmission on GMII and its post-transmit status has been passed from the MAC to DMA. The descriptor write back for frame N-1 has been requested on AHB and is awaiting completion. Frame N has completed the transmission on GMII and its post-transmit status has been passed from the MAC to DMA. A used bit has been read from the buffer descriptor queue after the descriptors associated with frame N. This means a buffer descriptor with bit [31] set in word 1 of the descriptor pair. Status associated with this used bit is passed from MAC to DMA on the exact AHB clock cycle that the writeback request for frame N-1 completes (either; BREADY high in the case of HREADY or high in the data phase in the case of AHB).
When this precise sequence occurs, the AHB write back request for frame N never actually occurs, but the TX DMA state machine waits in the DMA_MANWR state, awaiting the AHB response for the write. This response never happens since the request was never sent. Problem Fix/Workaround 1. Avoid the issue by sending only one frame at a time.
To send only a single frame at a time, each frame queued in the descriptor buffers must be followed by a buffer descriptor with the used bit set. GEM will send the frame and then halt and update status. Once frame status indicates that the frame has completed, then a new frame may be queued and transmit DMA restarted. 2. Avoid issue by changing used bit status to exhausted mid-frame status. For this workaround the buffer descriptors with used bit set, that mark the end of the queues are replaced with two zero length descriptors with particular bits set or clear. The effect is an exhausted mid-frame status instead of the usual used bit read status. This requires changes to SW for adding new frames to queues. For this discussion, assume that the last descriptor associated with the last valid frame queued is descriptor N, which is at address offset X from the queue base pointer. Currently the end of the queue is marked with descriptor N+1 with its used bit set as follows:
z
descriptor N: X+0x00: FC800000 X+0x04: 0000803C - Last frame queued descriptor N+1: X+0x08: xxxxxxxx +0x0C: 8xxxxxxx - Used bit set (x is a dont care value)
For the workaround buffer descriptor N+1 and also N+2 should be replaced with the following: z descriptor N: X+0x00: FC800000
z
X+0x04: 0000803C - Last frame queued descriptor N+1: X+0x08: 00000000 X+0x0C: 00000000 descriptor N+2: X+0x10: 00000000 X+0x14: 80008000 - last buffer and used bit set

1696
58.5.2 GMAC: TX Packet Buffer DMA Writeback Status Overflow

The issue occurs due to an overflow of the pipeline for the post-transmit status that is returned from the read side of the packet buffer memory back to the write side. Problem Fix/Workaround As there are four pipeline stages, having only 3 outstanding frames at any one time will avoid the issue.

1697
Revision History
In the tables that follow, the most recent version of the document appears first. rfo indicates changes requested during document review and approval loop.
Doc. Rev. Comments 11121B Introduction: Section 8.1 Chip Identification, updated Chip ID: 0x8A5C07C1 --> 0x8A5C07C2. Description, added a cross-reference to Table 1-1 SAMA5D3 Devices in the last paragraph. Replaced Cortex references with Cortex in Description and further on in the entire document. Removed AT91SAM from the document title. Standard Boot Strategies: Section 12.3.4.1 NAND Flash Boot: NAND Flash Detection, added the eccBitReq field description in NAND Flash Specific Header Detection. SFR: Added a row for SFR_UTMICKTRIM register (offset value 0x30) in Table 16-1 Register Mapping and the corresponding Section 16.3.4 UTMI Clock Trimming Register. External Memories: Section 27.1.5 Product Dependencies, updated LPDDR2 Mode data in Table 27-2 DDR2 I/O Lines Usage vs rfo Operating Modes (DDR_WE, DDR_RAS - DDR_CAS, and DDR_A[13..0]). Section 27.1.6.2 2x16-bit LPDDR2, added references on CAx LP-DDR2 signals and Table 27-3 CAx LPDDR2 Signal Connection. USART Added a paragraph on IRDA_FILTER programming criteria in Section 45.7.5.3 IrDA Demodulator and in the corresponding bitfield description in Section 45.8.20 USART IrDA FILTER Register. 8508 8683 8796

Change Request Ref.
rfo rfo rfo
Section 45.8.18 USART FI DI RATIO Register, expanded FI_DI_RATIO bitfield to 16 bits in the register table. 8643 FUSE: Section 51.2 Embedded Characteristics, added references on FUSE bits. Added Section 51.2.1 FUSE Bit Mapping and Section 51.2.2 Special Functions. 8785

1698
Doc. Rev. Comments 11121B Electrical Characteristics: Section 55.6 12 MHz RC Oscillator Characteristics, updated the IDDON value in Table 55-12. Section 55.7 32 kHz Oscillator Characteristics, added a row on PON in Table 55-13. Section 55.11 12-Bit ADC Characteristics, updated data in: - Table 55-22 Analog Power Supply Characteristics - Table 55-23 Channel Conversion Time and ADC Clock - Table 55-24 External Voltage Reference Input - Table 55-25 INL, DNL, 12-bit mode, VDDANA 2.4V to 3.6V supply voltage conditions - Table 55-26 Gain Error, 12-bit Mode, VDDANA 2.4V to 3.6V supply voltage conditions - Table 55-27 Error offset with or without calibration, 12-bit Mode, VDDANA 2.4V to 3.6V supply voltage conditions - Table 55-28 Dynamic Performance Characteristics in Single ended and 12 bits mode (1); and - Table 55-29 Dynamic Performance Characteristics in Differential and 12 bits mode(1) Added Section 55.15 MPDDRC Timings.
Change Request Ref.
rfo
rfo
Section 55.16 SSC Timings, updated PIXCLK frequency maximum value and fixed transmitter parameter data rfo for SSC7 in Table 55-47 SSC Timings with 3.3V Peripheral supply and Table 55-48 SSC Timing with 1V8 Peripheral supply. Mechanical Characteristics: Added Nominal Ball Diameter and Solder rows in Table 56-4 Package Information. Errata: Added the introduction paragraph in Section 58. SAMA5D3 Series Errata. Added Section 58.1.2 Boot ROM: NAND Flash Detection using ONFI Parameters does Not Work. rfo 8785 rfo
Doc. Rev. Comments 11121A First Issue
Change Request Ref.

1699
Table of Contents
Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1. Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 2. Block Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 3. Signal Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 4. Package and Pinout . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10

4.1 4.2 4.3 Mechanical Overview of the 324-ball LFBGA Package . . . . . . . . . . . . . . . . . 10 324-ball LFBGA Package Pinout . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 Input/Output Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
5. Power Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
5.1 5.2 5.3 Power Supplies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 Power-up Consideration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 Power-down Consideration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
6. Memories . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
6.1 6.2 Embedded Memories . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 External Memory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
7. Real-time Event Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24

7.1 7.2 Embedded Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 Real-time Event Mapping List . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
8. System Controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
8.1 8.2 Chip Identification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 Backup Section . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
9. Peripherals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
9.1 9.2 9.3 9.4 Peripheral Mapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Peripheral Identifiers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Peripheral Signal Multiplexing on I/O Lines . . . . . . . . . . . . . . . . . . . . . . . . . . Peripheral Clock Type . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28 28 30 30
10. Cortex-A5 ARM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31

10.1 10.2 10.3 10.4 10.5 Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Embedded Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Block Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Programmer Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Memory Management Unit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 31 32 33 39
11. SAMA5D3 Series Debug and Test . . . . . . . . . . . . . . . . . . . . . . . . . 43

11.1 11.2 11.3 11.4 11.5 11.6 Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Embedded Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Block Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Application Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Debug and Test Pin Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Functional Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43 43 44 45 47 48

11.7 11.8
The Boundary JTAG ID Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50 The Cortex-A5 DP Identification Code Register IDCODE . . . . . . . . . . . . . . . 51
12. Standard Boot Strategies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53

12.1 12.2 12.3 12.4 Flow Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Chip Setup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . NVM Boot . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . SAM-BA Monitor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54 54 55 65
13. Boot Sequence Controller (BSC) . . . . . . . . . . . . . . . . . . . . . . . . . . . 69

13.1 13.2 13.3 13.4 Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Embedded Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Product Dependencies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Boot Sequence Controller (BSC) User Interface . . . . . . . . . . . . . . . . . . . . . 69 69 69 70
14. AXI Bus Matrix (AXIMX) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71

14.1 14.2 14.3 14.4 Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Embedded Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . AXI Matrix (AXIMX) User Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71 71 72 75
15. Bus Matrix (MATRIX) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77

15.1 15.2 15.3 15.4 15.5 15.6 15.7 15.8 15.9 15.10 Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Embedded Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Memory Mapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Special Bus Granting Mechanism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . No Default Master . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Last Access Master . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Fixed Default Master . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Arbitration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Write Protect Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Bus Matrix (MATRIX) User Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77 79 80 80 80 80 81 81 83 84
16. Special Function Registers (SFR) . . . . . . . . . . . . . . . . . . . . . . . . . . 93

16.1 16.2 16.3 Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93 Embedded Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93 Special Function Registers (SFR) User Interface . . . . . . . . . . . . . . . . . . . . . 93
17. Advanced Interrupt Controller (AIC) . . . . . . . . . . . . . . . . . . . . . . . . . 99

17.1 17.2 17.3 17.4 17.5 17.6 17.7 17.8 17.9 Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99 Embedded Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99 Block Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100 Application Block Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100 AIC Detailed Block Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100 I/O Line Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101 Product Dependencies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101 Functional Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102 Advanced Interrupt Controller (AIC) User Interface . . . . . . . . . . . . . . . . . . . 112
18. Watchdog Timer (WDT) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133

18.1 18.2 Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133 Embedded Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133

ii
18.3 18.4 18.5
Block Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134 Functional Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135 Watchdog Timer (WDT) User Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137
19. Reset Controller (RSTC) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141

19.1 19.2 19.3 19.4 19.5 Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Embedded Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Block Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Functional Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Reset Controller (RSTC) User Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . 141 141 142 143 151
20. Shutdown Controller (SHDWC) . . . . . . . . . . . . . . . . . . . . . . . . . . . 155

20.1 20.2 20.3 20.4 20.5 20.6 20.7 Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Embedded Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Block Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . I/O Lines Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Product Dependencies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Functional Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Shutdown Controller (SHDWC) User Interface . . . . . . . . . . . . . . . . . . . . . . 155 155 155 156 156 156 157
21. General Purpose Backup Registers (GPBR) . . . . . . . . . . . . . . . . . 161

21.1 21.2 21.3 Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161 Embedded Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161 General Purpose Backup Registers (GPBR) User Interface . . . . . . . . . . . . 162
22. Real-time Clock (RTC) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163

22.1 22.2 22.3 22.4 22.5 22.6 Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Embedded Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Block Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Product Dependencies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Functional Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Real-time Clock (RTC) User Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163 163 164 165 165 168
23. Slow Clock Controller (SCKC) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181

23.1 23.2 23.3 23.4 Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Embedded Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Block Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Slow Clock Configuration (SCKC) User Interface . . . . . . . . . . . . . . . . . . . . 181 181 182 184
24. Clock Generator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 185

24.1 24.2 24.3 24.4 24.5 24.6 24.7 Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Embedded characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Block Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Main Clock Selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Main Clock . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Divider and PLLA Block . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . UTMI Phase Lock Loop Programming . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 185 185 186 187 189 192 193
25. Power Management Controller (PMC) . . . . . . . . . . . . . . . . . . . . . . 194

25.1 25.2 25.3 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 194 Master Clock Controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 194 Block Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 195

iii
25.4 25.5 25.6 25.7 25.8 25.9 25.10 25.11 25.12 25.13
Processor Clock Controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . USB Device and Host Clocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . DDR2/LPDDR/LPDDR2 Clock . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Software Modem Clock . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Peripheral Clock Controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Programmable Clock Output Controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . Main Crystal Clock Failure Detector . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Programming Sequence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Clock Switching Details . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Power Management Controller (PMC) User Interface . . . . . . . . . . . . . . . .
196 196 196 196 197 197 198 199 202 205
26. Parallel Input/Output (PIO3) Controller . . . . . . . . . . . . . . . . . . . . . 237

26.1 26.2 26.3 26.4 26.5 26.6 26.7 Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Embedded Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Block Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Product Dependencies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Functional Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . I/O Lines Programming Example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Parallel Input/Output Controller (PIO) User Interface . . . . . . . . . . . . . . . . . 237 237 238 239 240 249 250
27. External Memories . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 283

27.1 27.2 MPDDRC Multi-port DDRSDR Controller . . . . . . . . . . . . . . . . . . . . . . . . . . 284 External Bus Interface (EBI) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 290
28. Multi-port DDR-SDRAM Controller (MPDDRC) . . . . . . . . . . . . . . . 295

28.1 28.2 28.3 28.4 28.5 28.6 28.7 Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Embedded Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . MPDDRC Module Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Product Dependencies, Initialization Sequence . . . . . . . . . . . . . . . . . . . . . Functional Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Software Interface/SDRAM Organization, Address Mapping . . . . . . . . . . . AHB Multi-port DDR-SDRAM Controller (MPDDRC) User Interface . . . . . . 295 296 297 298 302 315 318
29. Static Memory Controller (HSMC) . . . . . . . . . . . . . . . . . . . . . . . . . 347

29.1 29.2 29.3 29.4 29.5 29.6 29.7 29.8 29.9 29.10 29.11 29.12 29.13 29.14 29.15 29.16 29.17 Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Embedded Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Block Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . I/O Lines Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Multiplexed Signals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Application Example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Product Dependencies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . External Memory Mapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Connection to External Devices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Standard Read and Write Protocols . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Scrambling/Unscrambling Function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Automatic Wait States . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Data Float Wait States . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . External Wait . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Slow Clock Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . NAND Flash Controller Operations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . PMECC Controller Functional Description . . . . . . . . . . . . . . . . . . . . . . . . . . 347 348 349 350 350 351 352 352 353 355 362 363 367 371 377 380 390

iv
29.18 Software Implementation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 397 29.19 Static Memory Controller (HSMC) User Interface . . . . . . . . . . . . . . . . . . . . 402
30. DMA Controller (DMAC) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 453

30.1 30.2 30.3 30.4 30.5 30.6 30.7 Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Embedded Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Block Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Functional Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . DMAC Software Requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Write Protection Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . DMA Controller (DMAC) User Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . 453 453 456 457 484 484 485
31. LCD Controller (LCDC) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 511

31.1 31.2 31.3 31.4 31.5 31.6 31.7 Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Embedded Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Block Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . I/O Lines Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Product Dependencies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Functional Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . LCD Controller (LCDC) User Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . 511 511 512 513 513 515 549
32. USB High Speed Device Port (UDPHS) . . . . . . . . . . . . . . . . . . . . 745

32.1 32.2 32.3 32.4 32.5 32.6 32.7 Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Embedded Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Block Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Typical Connection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Product Dependencies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Functional Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . USB High Speed Device Port (UDPHS) User Interface . . . . . . . . . . . . . . . . 745 745 746 747 748 749 772
33. USB High Speed Host Port (UHPHS) . . . . . . . . . . . . . . . . . . . . . . 817

33.1 33.2 33.3 33.4 33.5 33.6 Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Embedded Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Block Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Typical Connection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Product Dependencies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Functional Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 817 817 818 819 820 822
34. Gigabit Ethernet MAC (GMAC) . . . . . . . . . . . . . . . . . . . . . . . . . . . 823

34.1 34.2 34.3 34.4 34.5 34.6 34.7 Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Embedded Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Block Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Signal Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Functional Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Programming Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Gigabit Ethernet MAC (GMAC) User Interface . . . . . . . . . . . . . . . . . . . . . . 823 823 824 825 826 851 855
35. Ethernet MAC 10/100 (EMAC) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 947

35.1 35.2 35.3 35.4 Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Embedded Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Block Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Functional Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 947 947 948 949

35.5 35.6
Programming Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 959 Ethernet MAC 10/100 (EMAC) User Interface . . . . . . . . . . . . . . . . . . . . . . . 962
36. Image Sensor Interface (ISI) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1001

36.1 36.2 36.3 36.4 36.5 Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Embedded Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Block Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Functional Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Image Sensor Interface (ISI) User Interface . . . . . . . . . . . . . . . . . . . . . . . 1001 1002 1003 1004 1011
37. High Speed MultiMedia Card Interface (HSMCI) . . . . . . . . . . . . . 1041

37.1 37.2 37.3 37.4 37.5 37.6 37.7 37.8 37.9 37.10 37.11 37.12 37.13 37.14 Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Embedded Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Block Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Application Block Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Pin Name List . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Product Dependencies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Bus Topology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . High Speed MultiMedia Card Operations . . . . . . . . . . . . . . . . . . . . . . . . . SD/SDIO Card Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . CE-ATA Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . HSMCI Boot Operation Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . HSMCI Transfer Done Timings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Write Protection Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . High Speed MultiMedia Card Interface (HSMCI) User Interface . . . . . . . . 1041 1041 1042 1043 1044 1045 1046 1049 1068 1069 1070 1071 1072 1073
38. Serial Peripheral Interface (SPI) . . . . . . . . . . . . . . . . . . . . . . . . . 1099

38.1 38.2 38.3 38.4 38.5 38.6 38.7 38.8 Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Embedded Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Block Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Application Block Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Signal Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Product Dependencies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Functional Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Serial Peripheral Interface (SPI) User Interface . . . . . . . . . . . . . . . . . . . . . 1099 1099 1100 1101 1101 1102 1103 1115
39. Periodic Interval Timer (PIT) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1129

39.1 39.2 39.3 39.4 39.5 Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Embedded Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Block Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Functional Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Periodic Interval Timer (PIT) User Interface . . . . . . . . . . . . . . . . . . . . . . . 1129 1129 1130 1131 1132
40. Two-wire Interface (TWI) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1137

40.1 40.2 40.3 40.4 40.5 40.6 40.7 40.8 Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Embedded Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . List of Abbreviations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Block Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Application Block Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Product Dependencies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Functional Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Master Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1137 1138 1138 1139 1139 1140 1141 1142

vi
40.9 40.10 40.11 40.12
Multi-master Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Slave Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Write Protection System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Two-wire Interface (TWI) User Interface . . . . . . . . . . . . . . . . . . . . . . . . . .
1154 1157 1164 1165
41. Synchronous Serial Controller (SSC) . . . . . . . . . . . . . . . . . . . . . 1181

41.1 41.2 41.3 41.4 41.5 41.6 41.7 41.8 41.9 Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Embedded Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Block Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Application Block Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Pin Name List . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Product Dependencies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Functional Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . SSC Application Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Synchronous Serial Controller (SSC) User Interface . . . . . . . . . . . . . . . . . 1181 1181 1182 1183 1183 1184 1185 1196 1199
42. Debug Unit (DBGU) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1223

42.1 42.2 42.3 42.4 42.5 42.6 Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Embedded Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Block Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Product Dependencies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . UART Operations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Debug Unit (DBGU) User Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1223 1223 1224 1225 1225 1232
43. True Random Number Generator (TRNG) . . . . . . . . . . . . . . . . . 1247

43.1 43.2 43.3 43.4 43.5 43.6 Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Embedded Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Block Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Product Dependencies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Functional Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . True Random Number Generator (TRNG) User Interface . . . . . . . . . . . . . 1247 1247 1248 1249 1249 1250
44. Universal Asynchronous Receiver Transmitter (UART) . . . . . . . . 1257

44.1 44.2 44.3 44.4 44.5 44.6 Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Embedded Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Block Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Product Dependencies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . UART Operations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Universal Asynchronous Receiver Transmitter (UART) User Interface . . 1257 1257 1258 1259 1260 1266
45. Universal Synchronous Asynchronous Receiver Transmitter (USART) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1275

45.1 45.2 45.3 45.4 45.5 45.6 45.7 45.8 Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Embedded Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Block Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Application Block Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . I/O Lines Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Product Dependencies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Functional Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Universal Synchronous Asynchronous Receiver Transmitter (USART) User Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1275 1276 1277 1278 1278 1279 1280 1312

vii
46. Controller Area Network (CAN) . . . . . . . . . . . . . . . . . . . . . . . . . . 1343

46.1 46.2 46.3 46.4 46.5 46.6 46.7 46.8 46.9 Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Embedded Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Block Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Application Block Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . I/O Lines Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Product Dependencies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . CAN Controller Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Functional Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Controller Area Network (CAN) User Interface . . . . . . . . . . . . . . . . . . . . . 1343 1343 1344 1344 1345 1345 1346 1357 1369
47. Software Modem Device (SMD) . . . . . . . . . . . . . . . . . . . . . . . . . 1399

47.1 47.2 47.3 Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1399 Embedded Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1399 Block Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1400
48. Timer Counter (TC) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1401

48.1 48.2 48.3 48.4 48.5 48.6 48.7 Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Embedded Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Block Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Pin Name List . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Product Dependencies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Functional Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Timer Counter (TC) User Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1401 1402 1403 1404 1404 1405 1424
49. Pulse Width Modulation Controller (PWM) . . . . . . . . . . . . . . . . . 1451

49.1 49.2 49.3 49.4 49.5 49.6 49.7 Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Embedded Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Block Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . I/O Lines Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Product Dependencies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Functional Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Pulse Width Modulation Controller (PWM) User Interface . . . . . . . . . . . . . 1451 1452 1453 1453 1454 1456 1476
50. Analog-to-Digital Converter (ADC) . . . . . . . . . . . . . . . . . . . . . . . . 1523

50.1 50.2 50.3 50.4 50.5 50.6 50.7 50.8 Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Embedded Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Block Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Signal Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Product Dependencies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Functional Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Touchscreen . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Analog-to-Digital Converter (ADC) User Interface . . . . . . . . . . . . . . . . . . . 1523 1524 1525 1525 1526 1528 1538 1552
51. Fuse Controller (FUSE) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1583

51.1 51.2 51.3 51.4 51.5 Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Embedded Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Block Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Functional Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Fuse Controller (FUSE) User Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . 1583 1583 1584 1585 1587

viii
52. Advanced Encryption Standard (AES) . . . . . . . . . . . . . . . . . . . . . 1593

52.1 52.2 52.3 52.4 52.5 52.6 Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Embedded Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Product Dependencies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Functional Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Security Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Advanced Encryption Standard (AES) User Interface . . . . . . . . . . . . . . . . 1593 1593 1594 1595 1600 1601
53. Triple Data Encryption Standard (Triple DES) . . . . . . . . . . . . . . . 1613

53.1 53.2 53.3 53.4 53.5 Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Embedded Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Product Dependencies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Functional Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Triple Data Encryption Standard (TDES) User Interface . . . . . . . . . . . . . . 1613 1613 1614 1614 1619
54. Secure Hash Algorithm (SHA) . . . . . . . . . . . . . . . . . . . . . . . . . . . 1633

54.1 54.2 54.3 54.4 54.5 Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Embedded Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Product Dependencies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Functional Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Secure Hash Algorithm (SHA) User Interface . . . . . . . . . . . . . . . . . . . . . . 1633 1633 1633 1634 1637
55. Electrical Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1647

55.1 55.2 55.3 55.4 55.5 55.6 55.7 55.8 55.9 55.10 55.11 55.12 55.13 55.14 55.15 55.16 55.17 55.18 55.19 55.20 55.21 Absolute Maximum Ratings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . DC Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Power Consumption . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Clock Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Main Oscillator Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 MHz RC Oscillator Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32 kHz Oscillator Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32 kHz RC Oscillator Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . PLL Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . USB HS Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12-Bit ADC Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . POR Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . HSMC Timings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . SPI Timings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . MPDDRC Timings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . SSC Timings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ISI Timings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . MCI Timings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . EMAC/GMAC Timings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . UART in SPI Mode Timings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Two-wire Interface Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1647 1648 1650 1654 1655 1657 1657 1659 1659 1659 1661 1666 1667 1671 1675 1676 1681 1682 1683 1687 1689
56. Mechanical Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1691

56.1 Marking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1693
57. SAMA5D3 Ordering Information . . . . . . . . . . . . . . . . . . . . . . . . . 1694 58. SAMA5D3 Series Errata . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1695
58.1 58.2 Standard Boot Strategies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1695 Static Memory Controller (HSMC) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1695

ix
58.3 58.4 58.5
LCD Controller (LCDC) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1695 PWM Controller (PWM) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1695 Gigabit Ethernet MAC (GMAC) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1696
Revision History . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1698 Table of Contents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . i

Atmel Corporation 1600 Technology Drive San Jose, CA 95110 USA Tel: (+1) (408) 441-0311 Fax: (+1) (408) 487-2600 www.atmel.com
Atmel Asia Limited Unit 01-5 & 16, 19F BEA Tower, Millennium City 5 418 Kwun Tong Road Kwun Tong, Kowloon HONG KONG Tel: (+852) 2245-6100 Fax: (+852) 2722-1369
Atmel Munich GmbH Business Campus Parkring 4 D-85748 Garching b. Munich GERMANY Tel: (+49) 89-31970-0 Fax: (+49) 89-3194621
Atmel Japan G.K. 16F Shin-Osaki Kangyo Bldg 1-6-4 Osaki, Shinagawa-ku Tokyo 141-0032 JAPAN Tel: (+81) (3) 6417-0300 Fax: (+81) (3) 6417-0370
2013 Atmel Corporation. All rights reserved. / Rev.: 11121BATARM08-Mar-13 Atmel, Atmel logo and combinations thereof, Enabling Unlimited Possibilities, QTouch, SAM-BA and others are registered trademarks or trademarks of Atmel Corporation or its subsidiaries. Windows and others, are registered trademarks or trademarks of Microsoft Corporation in the US and/or other countries. ARM, ARMPowered logo, Cortex, Thumb-2 and others are registered trademarks or trademarks of ARM Ltd. Other terms and product names may be trademarks of others.
Disclaimer: The information in this document is provided in connection with Atmel products. No license, express or implied, by estoppel or otherwise, to any intellectual property right is granted by this document or in connection with the sale of Atmel products. EXCEPT AS SET FORTH IN THE ATMEL TERMS AND CONDITIONS OF SALES LOCATED ON THE ATMEL WEBSITE, ATMEL ASSUMES NO LIABILITY WHATSOEVER AND DISCLAIMS ANY EXPRESS, IMPLIED OR STATUTORY WARRANTY RELATING TO ITS PRODUCTS INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTY OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE, OR NON-INFRINGEMENT. IN NO EVENT SHALL ATMEL BE LIABLE FOR ANY DIRECT, INDIRECT, CONSEQUENTIAL, PUNITIVE, SPECIAL OR INCIDENTAL DAMAGES (INCLUDING, WITHOUT LIMITATION, DAMAGES FOR LOSS AND PROFITS, BUSINESS INTERRUPTION, OR LOSS OF INFORMATION) ARISING OUT OF THE USE OR INABILITY TO USE THIS DOCUMENT, EVEN IF ATMEL HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH DAMAGES. Atmel makes no representations or warranties with respect to the accuracy or completeness of the contents of this document and reserves the right to make changes to specifications and products descriptions at any time without notice. Atmel does not make any commitment to update the information contained herein. Unless specifically provided otherwise, Atmel products are not suitable for, and shall not be used in, automotive applications. Atmel products are not intended, authorized, or warranted for use as components in applications intended to support or sustain life.

Atmel 11121 32 Bit Cortex A5 Microcontroller SAMA5D3 Datasheet

Uploaded by

Copyright:

Available Formats

Atmel 11121 32 Bit Cortex A5 Microcontroller SAMA5D3 Datasheet

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Atmel 11121 32 Bit Cortex A5 Microcontroller SAMA5D3 Datasheet

Uploaded by

Copyright:

Available Formats

ARM-based Embedded MPU

ARM Cortex-A5 Processor with ARM v7-A Thumb2 Instruction Set

CPU Frequency up to 536 MHz

System running up to 166 MHz

Low Power Management

SAMA5D3 Series [DATASHEET]

SAMA5D3 Devices SAMA5D31 SAMA5D33 X X X X X X X X SAMA5D34 X X X X X X X X X SAMA5D35

LCDC GMAC EMAC CAN0, CAN1 HSMCI2 UART0 UART1 TC1

SAMA5D3 Series [DATASHEET]

Figure 2-1. SAMA5D3 Block Diagram

NT RS T TD I TD O TM TC S/SW K/S D WC IO LK JTA GS EL

DDR_VREF DDR_A0-DDR_A13 DDR_D0-DDR_D31 DDR_DQM[3..0] DDR_DQS[3..0]

DRXD DTXD PCK0-PCK2

PA HS USB Device DMA GMAC 10/100/1000 EMAC 10/100 LCD ISI

HS EHCI USB HOST DMA

DDR2 LPDDR2 512 MB

12 MHZ RC Osc WDT RC

SHDC POR POR PIOA PIOC PIOE

NAND Flash Controller MCL/SLC ECC (4 KB SRAM)

Reduced Static Memory Controller

DMA CAN0 CAN1 TWI0 TWI1 TWI2

DMA USART0 USART1 USART2 USART3

DMA SPI0 SPI1

DMA SSC0 SSC1

DMA 4-CH PWM

DMA 12-CH 12-bit ADC TouchScreen SMD

SAMA5D3 Series [DATASHEET]

SAMA5D3 Series [DATASHEET]

Low Low Low Low Low Low Low

High Low Low Low Low

SAMA5D3 Series [DATASHEET]

SAMA5D3 Series [DATASHEET]

SAMA5D3 Series [DATASHEET]

SAMA5D3 Series [DATASHEET]

Package and Pinout

Mechanical Overview of the 324-ball LFBGA Package

Figure 4-1. Orientation of the 324-ball LFBGA Package

SAMA5D3 Series [DATASHEET]

324-ball LFBGA Package Pinout

SAMA5D3 Series [DATASHEET]

SAMA5D3 Pinout for 324-ball LFBGA Package (Continued)

SAMA5D3 Series [DATASHEET]

SAMA5D3 Pinout for 324-ball LFBGA Package (Continued)

SAMA5D3 Series [DATASHEET]

SAMA5D3 Pinout for 324-ball LFBGA Package (Continued)

SAMA5D3 Series [DATASHEET]

SAMA5D3 Pinout for 324-ball LFBGA Package (Continued)

SAMA5D3 Series [DATASHEET]

SAMA5D3 Pinout for 324-ball LFBGA Package (Continued)

SAMA5D3 Series [DATASHEET]

SAMA5D3 Pinout for 324-ball LFBGA Package (Continued)

SAMA5D3 Series [DATASHEET]

SAMA5D3 Pinout for 324-ball LFBGA Package (Continued)

SAMA5D3 Series [DATASHEET]

SAMA5 I/O Type Assignation and Frequency

EBI DDR_IO RST JTAG SYSC VBG USBHS CLOCK GMAC

66 166 3 10 0.25 0.25 480 50 125

SAMA5D3 Series [DATASHEET]

VDDIOM VDDIOP0 VDDIOP1 VDDBU VDDUTMIC VDDUTMII VDDPLLA VDDOSC VDDANA

GNDIOM GNDIOP GNDIOP GNDBU GNDUTMI GNDUTMI GNDPLL GNDOSC GNDANA