Expand Miri's BorTag GC to a Provenance GC

Author: saethlin

compiler/rustc_const_eval/src/const_eval/machine.rs

@@ -107,6 +107,14 @@ impl<K: Hash + Eq, V> interpret::AllocMap<K, V> for FxIndexMap<K, V> {
         FxIndexMap::contains_key(self, k)
     }
 
+    #[inline(always)]
+    fn contains_key_ref<Q: ?Sized + Hash + Eq>(&self, k: &Q) -> bool
+    where
+        K: Borrow<Q>,
+    {
+        FxIndexMap::contains_key(self, k)
+    }
+
     #[inline(always)]
     fn insert(&mut self, k: K, v: V) -> Option<V> {
         FxIndexMap::insert(self, k, v)

compiler/rustc_const_eval/src/interpret/machine.rs

@@ -49,6 +49,14 @@ pub trait AllocMap<K: Hash + Eq, V> {
     where
         K: Borrow<Q>;
 
+    /// Callers should prefer [`AllocMap::contains_key`] when it is possible to call because it may
+    /// be more efficient. This function exists for callers that only have a shared reference
+    /// (which might make it slightly less efficient than `contains_key`, e.g. if
+    /// the data is stored inside a `RefCell`).
+    fn contains_key_ref<Q: ?Sized + Hash + Eq>(&self, k: &Q) -> bool
+    where
+        K: Borrow<Q>;
+
     /// Inserts a new entry into the map.
     fn insert(&mut self, k: K, v: V) -> Option<V>;
 

compiler/rustc_const_eval/src/interpret/memory.rs

@@ -692,6 +692,15 @@ impl<'mir, 'tcx: 'mir, M: Machine<'mir, 'tcx>> InterpCx<'mir, 'tcx, M> {
         Ok((&mut alloc.extra, machine))
     }
 
+    /// Check whether an allocation is live. This is faster than calling
+    /// [`InterpCx::get_alloc_info`] if all you need to check is whether the kind is
+    /// [`AllocKind::Dead`] because it doesn't have to look up the type and layout of statics.
+    pub fn is_alloc_live(&self, id: AllocId) -> bool {
+        self.tcx.try_get_global_alloc(id).is_some()
+            || self.memory.alloc_map.contains_key_ref(&id)
+            || self.memory.extra_fn_ptr_map.contains_key(&id)
+    }
+
     /// Obtain the size and alignment of an allocation, even if that allocation has
     /// been deallocated.
     pub fn get_alloc_info(&self, id: AllocId) -> (Size, Align, AllocKind) {

compiler/rustc_middle/src/mir/interpret/mod.rs

@@ -525,13 +525,6 @@ impl<'tcx> TyCtxt<'tcx> {
         self.alloc_map.lock().reserve()
     }
 
-    /// Miri's provenance GC needs to see all live allocations. The interpreter manages most
-    /// allocations but some are managed by [`TyCtxt`] and without this method the interpreter
-    /// doesn't know their [`AllocId`]s are in use.
-    pub fn iter_allocs<F: FnMut(AllocId)>(self, func: F) {
-        self.alloc_map.lock().alloc_map.keys().copied().for_each(func)
-    }
-
     /// Reserves a new ID *if* this allocation has not been dedup-reserved before.
     /// Should only be used for "symbolic" allocations (function pointers, vtables, statics), we
     /// don't want to dedup IDs for "real" memory!

src/tools/miri/src/borrow_tracker/mod.rs

@@ -75,8 +75,8 @@ pub struct FrameState {
     protected_tags: SmallVec<[(AllocId, BorTag); 2]>,
 }
 
-impl VisitTags for FrameState {
-    fn visit_tags(&self, _visit: &mut dyn FnMut(BorTag)) {
+impl VisitProvenance for FrameState {
+    fn visit_provenance(&self, _visit: &mut VisitWith<'_>) {
         // `protected_tags` are already recorded by `GlobalStateInner`.
     }
 }
@@ -110,10 +110,10 @@ pub struct GlobalStateInner {
     unique_is_unique: bool,
 }
 
-impl VisitTags for GlobalStateInner {
-    fn visit_tags(&self, visit: &mut dyn FnMut(BorTag)) {
+impl VisitProvenance for GlobalStateInner {
+    fn visit_provenance(&self, visit: &mut VisitWith<'_>) {
         for &tag in self.protected_tags.keys() {
-            visit(tag);
+            visit(None, Some(tag));
         }
         // The only other candidate is base_ptr_tags, and that does not need visiting since we don't ever
         // GC the bottommost/root tag.
@@ -236,6 +236,10 @@ impl GlobalStateInner {
             tag
         })
     }
+
+    pub fn remove_unreachable_allocs(&mut self, allocs: &LiveAllocs<'_, '_, '_>) {
+        self.base_ptr_tags.retain(|id, _| allocs.is_live(*id));
+    }
 }
 
 /// Which borrow tracking method to use
@@ -503,11 +507,11 @@ impl AllocState {
     }
 }
 
-impl VisitTags for AllocState {
-    fn visit_tags(&self, visit: &mut dyn FnMut(BorTag)) {
+impl VisitProvenance for AllocState {
+    fn visit_provenance(&self, visit: &mut VisitWith<'_>) {
         match self {
-            AllocState::StackedBorrows(sb) => sb.visit_tags(visit),
-            AllocState::TreeBorrows(tb) => tb.visit_tags(visit),
+            AllocState::StackedBorrows(sb) => sb.visit_provenance(visit),
+            AllocState::TreeBorrows(tb) => tb.visit_provenance(visit),
         }
     }
 }

src/tools/miri/src/borrow_tracker/stacked_borrows/mod.rs

@@ -462,10 +462,10 @@ impl Stacks {
     }
 }
 
-impl VisitTags for Stacks {
-    fn visit_tags(&self, visit: &mut dyn FnMut(BorTag)) {
+impl VisitProvenance for Stacks {
+    fn visit_provenance(&self, visit: &mut VisitWith<'_>) {
         for tag in self.exposed_tags.iter().copied() {
-            visit(tag);
+            visit(None, Some(tag));
         }
     }
 }

src/tools/miri/src/borrow_tracker/tree_borrows/tree.rs

@@ -742,11 +742,11 @@ impl Tree {
     }
 }
 
-impl VisitTags for Tree {
-    fn visit_tags(&self, visit: &mut dyn FnMut(BorTag)) {
+impl VisitProvenance for Tree {
+    fn visit_provenance(&self, visit: &mut VisitWith<'_>) {
         // To ensure that the root never gets removed, we visit it
         // (the `root` node of `Tree` is not an `Option<_>`)
-        visit(self.nodes.get(self.root).unwrap().tag)
+        visit(None, Some(self.nodes.get(self.root).unwrap().tag))
     }
 }
 

src/tools/miri/src/concurrency/data_race.rs

@@ -790,9 +790,9 @@ pub struct VClockAlloc {
     alloc_ranges: RefCell<RangeMap<MemoryCellClocks>>,
 }
 
-impl VisitTags for VClockAlloc {
-    fn visit_tags(&self, _visit: &mut dyn FnMut(BorTag)) {
-        // No tags here.
+impl VisitProvenance for VClockAlloc {
+    fn visit_provenance(&self, _visit: &mut VisitWith<'_>) {
+        // No tags or allocIds here.
     }
 }
 
@@ -1404,8 +1404,8 @@ pub struct GlobalState {
     pub track_outdated_loads: bool,
 }
 
-impl VisitTags for GlobalState {
-    fn visit_tags(&self, _visit: &mut dyn FnMut(BorTag)) {
+impl VisitProvenance for GlobalState {
+    fn visit_provenance(&self, _visit: &mut VisitWith<'_>) {
         // We don't have any tags.
     }
 }

src/tools/miri/src/concurrency/init_once.rs

@@ -45,10 +45,10 @@ pub(super) struct InitOnce<'mir, 'tcx> {
     data_race: VClock,
 }
 
-impl<'mir, 'tcx> VisitTags for InitOnce<'mir, 'tcx> {
-    fn visit_tags(&self, visit: &mut dyn FnMut(BorTag)) {
+impl<'mir, 'tcx> VisitProvenance for InitOnce<'mir, 'tcx> {
+    fn visit_provenance(&self, visit: &mut VisitWith<'_>) {
         for waiter in self.waiters.iter() {
-            waiter.callback.visit_tags(visit);
+            waiter.callback.visit_provenance(visit);
         }
     }
 }

src/tools/miri/src/concurrency/sync.rs

@@ -181,10 +181,10 @@ pub(crate) struct SynchronizationState<'mir, 'tcx> {
     pub(super) init_onces: IndexVec<InitOnceId, InitOnce<'mir, 'tcx>>,
 }
 
-impl<'mir, 'tcx> VisitTags for SynchronizationState<'mir, 'tcx> {
-    fn visit_tags(&self, visit: &mut dyn FnMut(BorTag)) {
+impl<'mir, 'tcx> VisitProvenance for SynchronizationState<'mir, 'tcx> {
+    fn visit_provenance(&self, visit: &mut VisitWith<'_>) {
         for init_once in self.init_onces.iter() {
-            init_once.visit_tags(visit);
+            init_once.visit_provenance(visit);
         }
     }
 }