rust-lang · bors · Jan 26, 2025 · Jan 20, 2025 · Dec 6, 2024 · Jan 9, 2025
diff --git a/compiler/rustc_borrowck/src/polonius/dump.rs b/compiler/rustc_borrowck/src/polonius/dump.rs
@@ -1,7 +1,9 @@
 use std::io;
 
-use rustc_middle::mir::pretty::{PrettyPrintMirOptions, dump_mir_with_options};
-use rustc_middle::mir::{Body, ClosureRegionRequirements, PassWhere};
+use rustc_middle::mir::pretty::{
+    PassWhere, PrettyPrintMirOptions, create_dump_file, dump_enabled, dump_mir_to_writer,
+};
+use rustc_middle::mir::{Body, ClosureRegionRequirements};
 use rustc_middle::ty::TyCtxt;
 use rustc_session::config::MirIncludeSpans;
 
@@ -10,9 +12,6 @@ use crate::polonius::{LocalizedOutlivesConstraint, LocalizedOutlivesConstraintSe
 use crate::{BorrowckInferCtxt, RegionInferenceContext};
 
 /// `-Zdump-mir=polonius` dumps MIR annotated with NLL and polonius specific information.
-// Note: this currently duplicates most of NLL MIR, with some additions for the localized outlives
-// constraints. This is ok for now as this dump will change in the near future to an HTML file to
-// become more useful.
 pub(crate) fn dump_polonius_mir<'tcx>(
     infcx: &BorrowckInferCtxt<'tcx>,
     body: &Body<'tcx>,
@@ -26,25 +25,113 @@ pub(crate) fn dump_polonius_mir<'tcx>(
         return;
     }
 
+    if !dump_enabled(tcx, "polonius", body.source.def_id()) {
+        return;
+    }
+
     let localized_outlives_constraints = localized_outlives_constraints
         .expect("missing localized constraints with `-Zpolonius=next`");
 
-    // We want the NLL extra comments printed by default in NLL MIR dumps (they were removed in
-    // #112346). Specifying `-Z mir-include-spans` on the CLI still has priority: for example,
-    // they're always disabled in mir-opt tests to make working with blessed dumps easier.
+    let _: io::Result<()> = try {
+        let mut file = create_dump_file(tcx, "html", false, "polonius", &0, body)?;
+        emit_polonius_dump(
+            tcx,
+            body,
+            regioncx,
+            borrow_set,
+            localized_outlives_constraints,
+            closure_region_requirements,
+            &mut file,
+        )?;
+    };
+}
+
+/// The polonius dump consists of:
+/// - the NLL MIR
+/// - the list of polonius localized constraints
+/// - a mermaid graph of the CFG
+fn emit_polonius_dump<'tcx>(
+    tcx: TyCtxt<'tcx>,
+    body: &Body<'tcx>,
+    regioncx: &RegionInferenceContext<'tcx>,
+    borrow_set: &BorrowSet<'tcx>,
+    localized_outlives_constraints: LocalizedOutlivesConstraintSet,
+    closure_region_requirements: &Option<ClosureRegionRequirements<'tcx>>,
+    out: &mut dyn io::Write,
+) -> io::Result<()> {
+    // Prepare the HTML dump file prologue.
+    writeln!(out, "<!DOCTYPE html>")?;
+    writeln!(out, "<html>")?;
+    writeln!(out, "<head><title>Polonius MIR dump</title></head>")?;
+    writeln!(out, "<body>")?;
+
+    // Section 1: the NLL + Polonius MIR.
+    writeln!(out, "<div>")?;
+    writeln!(out, "Raw MIR dump")?;
+    writeln!(out, "<code><pre>")?;
+    emit_html_mir(
+        tcx,
+        body,
+        regioncx,
+        borrow_set,
+        localized_outlives_constraints,
+        closure_region_requirements,
+        out,
+    )?;
+    writeln!(out, "</pre></code>")?;
+    writeln!(out, "</div>")?;
+
+    // Section 2: mermaid visualization of the CFG.
+    writeln!(out, "<div>")?;
+    writeln!(out, "Control-flow graph")?;
+    writeln!(out, "<code><pre class='mermaid'>")?;
+    emit_mermaid_cfg(body, out)?;
+    writeln!(out, "</pre></code>")?;
+    writeln!(out, "</div>")?;
+
+    // Finalize the dump with the HTML epilogue.
+    writeln!(
+        out,
+        "<script src='https://cdn.jsdelivr.net/npm/mermaid/dist/mermaid.min.js'></script>"
+    )?;
+    writeln!(out, "<script>")?;
+    writeln!(out, "mermaid.initialize({{ startOnLoad: false, maxEdges: 100 }});")?;
+    writeln!(out, "mermaid.run({{ querySelector: '.mermaid' }})")?;
+    writeln!(out, "</script>")?;
+    writeln!(out, "</body>")?;
+    writeln!(out, "</html>")?;
+
+    Ok(())
+}
+
+/// Emits the polonius MIR, as escaped HTML.
+fn emit_html_mir<'tcx>(
+    tcx: TyCtxt<'tcx>,
+    body: &Body<'tcx>,
+    regioncx: &RegionInferenceContext<'tcx>,
+    borrow_set: &BorrowSet<'tcx>,
+    localized_outlives_constraints: LocalizedOutlivesConstraintSet,
+    closure_region_requirements: &Option<ClosureRegionRequirements<'tcx>>,
+    out: &mut dyn io::Write,
+) -> io::Result<()> {
+    // Buffer the regular MIR dump to be able to escape it.
+    let mut buffer = Vec::new();
+
+    // We want the NLL extra comments printed by default in NLL MIR dumps. Specifying `-Z
+    // mir-include-spans` on the CLI still has priority.
     let options = PrettyPrintMirOptions {
         include_extra_comments: matches!(
             tcx.sess.opts.unstable_opts.mir_include_spans,
             MirIncludeSpans::On | MirIncludeSpans::Nll
         ),
     };
 
-    dump_mir_with_options(
+    dump_mir_to_writer(
         tcx,
-        false,
         "polonius",
         &0,
         body,
+        &mut buffer,
         |pass_where, out| {
             emit_polonius_mir(
                 tcx,
@@ -57,7 +144,27 @@ pub(crate) fn dump_polonius_mir<'tcx>(
             )
         },
         options,
-    );
+    )?;
+
+    // Escape the handful of characters that need it. We don't need to be particularly efficient:
+    // we're actually writing into a buffered writer already. Note that MIR dumps are valid UTF-8.
+    let buffer = String::from_utf8_lossy(&buffer);
+    for ch in buffer.chars() {
+        let escaped = match ch {
+            '>' => "&gt;",
+            '<' => "&lt;",
+            '&' => "&amp;",
+            '\'' => "&#39;",
+            '"' => "&quot;",
+            _ => {
+                // The common case, no escaping needed.
+                write!(out, "{}", ch)?;
+                continue;
+            }
+        };
+        write!(out, "{}", escaped)?;
+    }
+    Ok(())
 }
 
 /// Produces the actual NLL + Polonius MIR sections to emit during the dumping process.
@@ -102,3 +209,55 @@ fn emit_polonius_mir<'tcx>(
 
     Ok(())
 }
+
+/// Emits a mermaid flowchart of the CFG blocks and edges, similar to the graphviz version.
+fn emit_mermaid_cfg(body: &Body<'_>, out: &mut dyn io::Write) -> io::Result<()> {
+    use rustc_middle::mir::{TerminatorEdges, TerminatorKind};
+
+    // The mermaid chart type: a top-down flowchart.
+    writeln!(out, "flowchart TD")?;
+
+    // Emit the block nodes.
+    for (block_idx, block) in body.basic_blocks.iter_enumerated() {
+        let block_idx = block_idx.as_usize();
+        let cleanup = if block.is_cleanup { " (cleanup)" } else { "" };
+        writeln!(out, "{block_idx}[\"bb{block_idx}{cleanup}\"]")?;
+    }
+
+    // Emit the edges between blocks, from the terminator edges.
+    for (block_idx, block) in body.basic_blocks.iter_enumerated() {
+        let block_idx = block_idx.as_usize();
+        let terminator = block.terminator();
+        match terminator.edges() {
+            TerminatorEdges::None => {}
+            TerminatorEdges::Single(bb) => {
+                writeln!(out, "{block_idx} --> {}", bb.as_usize())?;
+            }
+            TerminatorEdges::Double(bb1, bb2) => {
+                if matches!(terminator.kind, TerminatorKind::FalseEdge { .. }) {
+                    writeln!(out, "{block_idx} --> {}", bb1.as_usize())?;
+                    writeln!(out, "{block_idx} -- imaginary --> {}", bb2.as_usize())?;
+                } else {
+                    writeln!(out, "{block_idx} --> {}", bb1.as_usize())?;
+                    writeln!(out, "{block_idx} -- unwind --> {}", bb2.as_usize())?;
+                }
+            }
+            TerminatorEdges::AssignOnReturn { return_, cleanup, .. } => {
+                for to_idx in return_ {
+                    writeln!(out, "{block_idx} --> {}", to_idx.as_usize())?;
+                }
+
+                if let Some(to_idx) = cleanup {
+                    writeln!(out, "{block_idx} -- unwind --> {}", to_idx.as_usize())?;
+                }
+            }
+            TerminatorEdges::SwitchInt { targets, .. } => {
+                for to_idx in targets.all_targets() {
+                    writeln!(out, "{block_idx} --> {}", to_idx.as_usize())?;
+                }
+            }
+        }
+    }
+
+    Ok(())
+}
diff --git a/compiler/rustc_hir/src/hir.rs b/compiler/rustc_hir/src/hir.rs
@@ -1285,13 +1285,13 @@ impl fmt::Debug for OwnerNodes<'_> {
             .field("node", &self.nodes[ItemLocalId::ZERO])
             .field(
                 "parents",
-                &self
-                    .nodes
-                    .iter_enumerated()
-                    .map(|(id, parented_node)| {
-                        debug_fn(move |f| write!(f, "({id:?}, {:?})", parented_node.parent))
-                    })
-                    .collect::<Vec<_>>(),
+                &fmt::from_fn(|f| {
+                    f.debug_list()
+                        .entries(self.nodes.iter_enumerated().map(|(id, parented_node)| {
+                            fmt::from_fn(move |f| write!(f, "({id:?}, {:?})", parented_node.parent))
+                        }))
+                        .finish()
+                }),
             )
             .field("bodies", &self.bodies)
             .field("opt_hash_including_bodies", &self.opt_hash_including_bodies)
@@ -4638,15 +4638,5 @@ mod size_asserts {
     // tidy-alphabetical-end
 }
 
-fn debug_fn(f: impl Fn(&mut fmt::Formatter<'_>) -> fmt::Result) -> impl fmt::Debug {
-    struct DebugFn<F>(F);
-    impl<F: Fn(&mut fmt::Formatter<'_>) -> fmt::Result> fmt::Debug for DebugFn<F> {
-        fn fmt(&self, fmt: &mut fmt::Formatter<'_>) -> fmt::Result {
-            (self.0)(fmt)
-        }
-    }
-    DebugFn(f)
-}
-
 #[cfg(test)]
 mod tests;
diff --git a/compiler/rustc_hir/src/lib.rs b/compiler/rustc_hir/src/lib.rs
@@ -6,6 +6,7 @@
 #![allow(internal_features)]
 #![feature(associated_type_defaults)]
 #![feature(closure_track_caller)]
+#![feature(debug_closure_helpers)]
 #![feature(exhaustive_patterns)]
 #![feature(let_chains)]
 #![feature(never_type)]

diff --git a/compiler/rustc_hir_analysis/src/collect/resolve_bound_vars.rs b/compiler/rustc_hir_analysis/src/collect/resolve_bound_vars.rs
@@ -184,68 +184,66 @@ enum Scope<'a> {
     },
 }
 
-#[derive(Copy, Clone, Debug)]
-enum BinderScopeType {
-    /// Any non-concatenating binder scopes.
-    Normal,
-    /// Within a syntactic trait ref, there may be multiple poly trait refs that
-    /// are nested (under the `associated_type_bounds` feature). The binders of
-    /// the inner poly trait refs are extended from the outer poly trait refs
-    /// and don't increase the late bound depth. If you had
-    /// `T: for<'a>  Foo<Bar: for<'b> Baz<'a, 'b>>`, then the `for<'b>` scope
-    /// would be `Concatenating`. This also used in trait refs in where clauses
-    /// where we have two binders `for<> T: for<> Foo` (I've intentionally left
-    /// out any lifetimes because they aren't needed to show the two scopes).
-    /// The inner `for<>` has a scope of `Concatenating`.
-    Concatenating,
-}
-
-// A helper struct for debugging scopes without printing parent scopes
-struct TruncatedScopeDebug<'a>(&'a Scope<'a>);
-
-impl<'a> fmt::Debug for TruncatedScopeDebug<'a> {
-    fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
-        match self.0 {
-            Scope::Binder { bound_vars, scope_type, hir_id, where_bound_origin, s: _ } => f
+impl<'a> Scope<'a> {
+    // A helper for debugging scopes without printing parent scopes
+    fn debug_truncated(&'a self) -> impl fmt::Debug + 'a {
+        fmt::from_fn(move |f| match self {
+            Self::Binder { bound_vars, scope_type, hir_id, where_bound_origin, s: _ } => f
                 .debug_struct("Binder")
                 .field("bound_vars", bound_vars)
                 .field("scope_type", scope_type)
                 .field("hir_id", hir_id)
                 .field("where_bound_origin", where_bound_origin)
                 .field("s", &"..")
                 .finish(),
-            Scope::Opaque { captures, def_id, s: _ } => f
+            Self::Opaque { captures, def_id, s: _ } => f
                 .debug_struct("Opaque")
                 .field("def_id", def_id)
                 .field("captures", &captures.borrow())
                 .field("s", &"..")
                 .finish(),
-            Scope::Body { id, s: _ } => {
+            Self::Body { id, s: _ } => {
                 f.debug_struct("Body").field("id", id).field("s", &"..").finish()
             }
-            Scope::ObjectLifetimeDefault { lifetime, s: _ } => f
+            Self::ObjectLifetimeDefault { lifetime, s: _ } => f
                 .debug_struct("ObjectLifetimeDefault")
                 .field("lifetime", lifetime)
                 .field("s", &"..")
                 .finish(),
-            Scope::Supertrait { bound_vars, s: _ } => f
+            Self::Supertrait { bound_vars, s: _ } => f
                 .debug_struct("Supertrait")
                 .field("bound_vars", bound_vars)
                 .field("s", &"..")
                 .finish(),
-            Scope::TraitRefBoundary { s: _ } => f.debug_struct("TraitRefBoundary").finish(),
-            Scope::LateBoundary { s: _, what, deny_late_regions } => f
+            Self::TraitRefBoundary { s: _ } => f.debug_struct("TraitRefBoundary").finish(),
+            Self::LateBoundary { s: _, what, deny_late_regions } => f
                 .debug_struct("LateBoundary")
                 .field("what", what)
                 .field("deny_late_regions", deny_late_regions)
                 .finish(),
-            Scope::Root { opt_parent_item } => {
+            Self::Root { opt_parent_item } => {
                 f.debug_struct("Root").field("opt_parent_item", &opt_parent_item).finish()
             }
-        }
+        })
     }
 }
 
+#[derive(Copy, Clone, Debug)]
+enum BinderScopeType {
+    /// Any non-concatenating binder scopes.
+    Normal,
+    /// Within a syntactic trait ref, there may be multiple poly trait refs that
+    /// are nested (under the `associated_type_bounds` feature). The binders of
+    /// the inner poly trait refs are extended from the outer poly trait refs
+    /// and don't increase the late bound depth. If you had
+    /// `T: for<'a>  Foo<Bar: for<'b> Baz<'a, 'b>>`, then the `for<'b>` scope
+    /// would be `Concatenating`. This also used in trait refs in where clauses
+    /// where we have two binders `for<> T: for<> Foo` (I've intentionally left
+    /// out any lifetimes because they aren't needed to show the two scopes).
+    /// The inner `for<>` has a scope of `Concatenating`.
+    Concatenating,
+}
+
 type ScopeRef<'a> = &'a Scope<'a>;
 
 pub(crate) fn provide(providers: &mut Providers) {
@@ -1144,7 +1142,7 @@ impl<'a, 'tcx> BoundVarContext<'a, 'tcx> {
     {
         let BoundVarContext { tcx, map, .. } = self;
         let mut this = BoundVarContext { tcx: *tcx, map, scope: &wrap_scope };
-        let span = debug_span!("scope", scope = ?TruncatedScopeDebug(this.scope));
+        let span = debug_span!("scope", scope = ?this.scope.debug_truncated());
         {
             let _enter = span.enter();
             f(&mut this);

diff --git a/compiler/rustc_hir_analysis/src/lib.rs b/compiler/rustc_hir_analysis/src/lib.rs
@@ -63,6 +63,7 @@ This API is completely unstable and subject to change.
 #![doc(rust_logo)]
 #![feature(assert_matches)]
 #![feature(coroutines)]
+#![feature(debug_closure_helpers)]
 #![feature(if_let_guard)]
 #![feature(iter_from_coroutine)]
 #![feature(iter_intersperse)]