Rollup merge of #72796 - RalfJung:mir-assign-sanity, r=matthewjasper

MIR sanity check: validate types on assignment

This expands the MIR validation added by @jonas-schievink in https://github.com/rust-lang/rust/pull/72093 to also check that on an assignment, the types of both sides match.

Cc @eddyb @oli-obk

src/librustc_mir/interpret/eval_context.rs

@@ -15,7 +15,7 @@
 };
 use rustc_middle::ty::layout::{self, TyAndLayout};
 use rustc_middle::ty::{
-    self, fold::BottomUpFolder, query::TyCtxtAt, subst::SubstsRef, Ty, TyCtxt, TypeFoldable,
+    self, query::TyCtxtAt, subst::SubstsRef, ParamEnv, Ty, TyCtxt, TypeFoldable,
 };
 use rustc_span::{source_map::DUMMY_SP, Span};
 use rustc_target::abi::{Align, HasDataLayout, LayoutOf, Size, TargetDataLayout};
@@ -24,6 +24,7 @@
     Immediate, MPlaceTy, Machine, MemPlace, MemPlaceMeta, Memory, OpTy, Operand, Place, PlaceTy,
     ScalarMaybeUninit, StackPopJump,
 };
+use crate::transform::validate::equal_up_to_regions;
 use crate::util::storage::AlwaysLiveLocals;
 
 pub struct InterpCx<'mir, 'tcx, M: Machine<'mir, 'tcx>> {
@@ -220,42 +221,27 @@ fn layout_of(&self, ty: Ty<'tcx>) -> Self::TyAndLayout {
 /// This test should be symmetric, as it is primarily about layout compatibility.
 pub(super) fn mir_assign_valid_types<'tcx>(
     tcx: TyCtxt<'tcx>,
+    param_env: ParamEnv<'tcx>,
     src: TyAndLayout<'tcx>,
     dest: TyAndLayout<'tcx>,
 ) -> bool {
-    if src.ty == dest.ty {
-        // Equal types, all is good.
-        return true;
+    // Type-changing assignments can happen when subtyping is used. While
+    // all normal lifetimes are erased, higher-ranked types with their
+    // late-bound lifetimes are still around and can lead to type
+    // differences. So we compare ignoring lifetimes.
+    if equal_up_to_regions(tcx, param_env, src.ty, dest.ty) {
+        // Make sure the layout is equal, too -- just to be safe. Miri really
+        // needs layout equality. For performance reason we skip this check when
+        // the types are equal. Equal types *can* have different layouts when
+        // enum downcast is involved (as enum variants carry the type of the
+        // enum), but those should never occur in assignments.
+        if cfg!(debug_assertions) || src.ty != dest.ty {
+            assert_eq!(src.layout, dest.layout);
         }
-    if src.layout != dest.layout {
-        // Layout differs, definitely not equal.
-        // We do this here because Miri would *do the wrong thing* if we allowed layout-changing
-        // assignments.
-        return false;
+        true
+    } else {
+        false
     }
-
-    // Type-changing assignments can happen for (at least) two reasons:
-    // 1. `&mut T` -> `&T` gets optimized from a reborrow to a mere assignment.
-    // 2. Subtyping is used. While all normal lifetimes are erased, higher-ranked types
-    //    with their late-bound lifetimes are still around and can lead to type differences.
-    // Normalize both of them away.
-    let normalize = |ty: Ty<'tcx>| {
-        ty.fold_with(&mut BottomUpFolder {
-            tcx,
-            // Normalize all references to immutable.
-            ty_op: |ty| match ty.kind {
-                ty::Ref(_, pointee, _) => tcx.mk_imm_ref(tcx.lifetimes.re_erased, pointee),
-                _ => ty,
-            },
-            // We just erase all late-bound lifetimes, but this is not fully correct (FIXME):
-            // lifetimes in invariant positions could matter (e.g. through associated types).
-            // We rely on the fact that layout was confirmed to be equal above.
-            lt_op: |_| tcx.lifetimes.re_erased,
-            // Leave consts unchanged.
-            ct_op: |ct| ct,
-        })
-    };
-    normalize(src.ty) == normalize(dest.ty)
 }
 
 /// Use the already known layout if given (but sanity check in debug mode),
@@ -263,6 +249,7 @@ pub(super) fn mir_assign_valid_types<'tcx>(
 #[cfg_attr(not(debug_assertions), inline(always))]
 pub(super) fn from_known_layout<'tcx>(
     tcx: TyCtxtAt<'tcx>,
+    param_env: ParamEnv<'tcx>,
     known_layout: Option<TyAndLayout<'tcx>>,
     compute: impl FnOnce() -> InterpResult<'tcx, TyAndLayout<'tcx>>,
 ) -> InterpResult<'tcx, TyAndLayout<'tcx>> {
@@ -271,7 +258,7 @@ pub(super) fn from_known_layout<'tcx>(
         Some(known_layout) => {
             if cfg!(debug_assertions) {
                 let check_layout = compute()?;
-                if !mir_assign_valid_types(tcx.tcx, check_layout, known_layout) {
+                if !mir_assign_valid_types(tcx.tcx, param_env, check_layout, known_layout) {
                     span_bug!(
                         tcx.span,
                         "expected type differs from actual type.\nexpected: {:?}\nactual: {:?}",
@@ -475,7 +462,7 @@ pub fn layout_of_local(
         // have to support that case (mostly by skipping all caching).
         match frame.locals.get(local).and_then(|state| state.layout.get()) {
             None => {
-                let layout = from_known_layout(self.tcx, layout, || {
+                let layout = from_known_layout(self.tcx, self.param_env, layout, || {
                     let local_ty = frame.body.local_decls[local].ty;
                     let local_ty =
                         self.subst_from_frame_and_normalize_erasing_regions(frame, local_ty);

src/librustc_mir/interpret/operand.rs

@@ -488,6 +488,7 @@ pub fn eval_place_to_op(
         // Sanity-check the type we ended up with.
         debug_assert!(mir_assign_valid_types(
             *self.tcx,
+            self.param_env,
             self.layout_of(self.subst_from_current_frame_and_normalize_erasing_regions(
                 place.ty(&self.frame().body.local_decls, *self.tcx).ty
             ))?,
@@ -570,7 +571,8 @@ pub(super) fn eval_operands(
         // documentation).
         let val_val = M::adjust_global_const(self, val_val)?;
         // Other cases need layout.
-        let layout = from_known_layout(self.tcx, layout, || self.layout_of(val.ty))?;
+        let layout =
+            from_known_layout(self.tcx, self.param_env, layout, || self.layout_of(val.ty))?;
         let op = match val_val {
             ConstValue::ByRef { alloc, offset } => {
                 let id = self.tcx.create_memory_alloc(alloc);

src/librustc_mir/interpret/place.rs

@@ -652,6 +652,7 @@ pub fn eval_place(
         // Sanity-check the type we ended up with.
         debug_assert!(mir_assign_valid_types(
             *self.tcx,
+            self.param_env,
             self.layout_of(self.subst_from_current_frame_and_normalize_erasing_regions(
                 place.ty(&self.frame().body.local_decls, *self.tcx).ty
             ))?,
@@ -855,7 +856,7 @@ fn copy_op_no_validate(
     ) -> InterpResult<'tcx> {
         // We do NOT compare the types for equality, because well-typed code can
         // actually "transmute" `&mut T` to `&T` in an assignment without a cast.
-        if !mir_assign_valid_types(*self.tcx, src.layout, dest.layout) {
+        if !mir_assign_valid_types(*self.tcx, self.param_env, src.layout, dest.layout) {
             span_bug!(
                 self.cur_span(),
                 "type mismatch when copying!\nsrc: {:?},\ndest: {:?}",
@@ -912,7 +913,7 @@ pub fn copy_op_transmute(
         src: OpTy<'tcx, M::PointerTag>,
         dest: PlaceTy<'tcx, M::PointerTag>,
     ) -> InterpResult<'tcx> {
-        if mir_assign_valid_types(*self.tcx, src.layout, dest.layout) {
+        if mir_assign_valid_types(*self.tcx, self.param_env, src.layout, dest.layout) {
             // Fast path: Just use normal `copy_op`
             return self.copy_op(src, dest);
         }

src/librustc_mir/transform/validate.rs

@@ -7,7 +7,11 @@
         BasicBlock, Body, Location, Operand, Rvalue, Statement, StatementKind, Terminator,
         TerminatorKind,
     },
-    ty::{self, ParamEnv, TyCtxt},
+    ty::{
+        self,
+        relate::{Relate, RelateResult, TypeRelation},
+        ParamEnv, Ty, TyCtxt,
+    },
 };
 
 #[derive(Copy, Clone, Debug)]
@@ -28,6 +32,98 @@ fn run_pass(&self, tcx: TyCtxt<'tcx>, source: MirSource<'tcx>, body: &mut Body<'
     }
 }
 
+/// Returns whether the two types are equal up to lifetimes.
+/// All lifetimes, including higher-ranked ones, get ignored for this comparison.
+/// (This is unlike the `erasing_regions` methods, which keep higher-ranked lifetimes for soundness reasons.)
+///
+/// The point of this function is to approximate "equal up to subtyping".  However,
+/// the approximation is incorrect as variance is ignored.
+pub fn equal_up_to_regions(
+    tcx: TyCtxt<'tcx>,
+    param_env: ParamEnv<'tcx>,
+    src: Ty<'tcx>,
+    dest: Ty<'tcx>,
+) -> bool {
+    // Fast path.
+    if src == dest {
+        return true;
+    }
+
+    struct LifetimeIgnoreRelation<'tcx> {
+        tcx: TyCtxt<'tcx>,
+        param_env: ty::ParamEnv<'tcx>,
+    }
+
+    impl TypeRelation<'tcx> for LifetimeIgnoreRelation<'tcx> {
+        fn tcx(&self) -> TyCtxt<'tcx> {
+            self.tcx
+        }
+
+        fn param_env(&self) -> ty::ParamEnv<'tcx> {
+            self.param_env
+        }
+
+        fn tag(&self) -> &'static str {
+            "librustc_mir::transform::validate"
+        }
+
+        fn a_is_expected(&self) -> bool {
+            true
+        }
+
+        fn relate_with_variance<T: Relate<'tcx>>(
+            &mut self,
+            _: ty::Variance,
+            a: &T,
+            b: &T,
+        ) -> RelateResult<'tcx, T> {
+            // Ignore variance, require types to be exactly the same.
+            self.relate(a, b)
+        }
+
+        fn tys(&mut self, a: Ty<'tcx>, b: Ty<'tcx>) -> RelateResult<'tcx, Ty<'tcx>> {
+            if a == b {
+                // Short-circuit.
+                return Ok(a);
+            }
+            ty::relate::super_relate_tys(self, a, b)
+        }
+
+        fn regions(
+            &mut self,
+            a: ty::Region<'tcx>,
+            _b: ty::Region<'tcx>,
+        ) -> RelateResult<'tcx, ty::Region<'tcx>> {
+            // Ignore regions.
+            Ok(a)
+        }
+
+        fn consts(
+            &mut self,
+            a: &'tcx ty::Const<'tcx>,
+            b: &'tcx ty::Const<'tcx>,
+        ) -> RelateResult<'tcx, &'tcx ty::Const<'tcx>> {
+            ty::relate::super_relate_consts(self, a, b)
+        }
+
+        fn binders<T>(
+            &mut self,
+            a: &ty::Binder<T>,
+            b: &ty::Binder<T>,
+        ) -> RelateResult<'tcx, ty::Binder<T>>
+        where
+            T: Relate<'tcx>,
+        {
+            self.relate(a.skip_binder(), b.skip_binder())?;
+            Ok(a.clone())
+        }
+    }
+
+    // Instantiate and run relation.
+    let mut relator: LifetimeIgnoreRelation<'tcx> = LifetimeIgnoreRelation { tcx: tcx, param_env };
+    relator.relate(&src, &dest).is_ok()
+}
+
 struct TypeChecker<'a, 'tcx> {
     when: &'a str,
     source: MirSource<'tcx>,
@@ -81,6 +177,28 @@ fn check_edge(&self, location: Location, bb: BasicBlock, edge_kind: EdgeKind) {
             self.fail(location, format!("encountered jump to invalid basic block {:?}", bb))
         }
     }
+
+    /// Check if src can be assigned into dest.
+    /// This is not precise, it will accept some incorrect assignments.
+    fn mir_assign_valid_types(&self, src: Ty<'tcx>, dest: Ty<'tcx>) -> bool {
+        // Fast path before we normalize.
+        if src == dest {
+            // Equal types, all is good.
+            return true;
+        }
+        // Normalize projections and things like that.
+        // FIXME: We need to reveal_all, as some optimizations change types in ways
+        // that require unfolding opaque types.
+        let param_env = self.param_env.with_reveal_all();
+        let src = self.tcx.normalize_erasing_regions(param_env, src);
+        let dest = self.tcx.normalize_erasing_regions(param_env, dest);
+
+        // Type-changing assignments can happen when subtyping is used. While
+        // all normal lifetimes are erased, higher-ranked types with their
+        // late-bound lifetimes are still around and can lead to type
+        // differences. So we compare ignoring lifetimes.
+        equal_up_to_regions(self.tcx, param_env, src, dest)
+    }
 }
 
 impl<'a, 'tcx> Visitor<'tcx> for TypeChecker<'a, 'tcx> {
@@ -99,9 +217,24 @@ fn visit_operand(&mut self, operand: &Operand<'tcx>, location: Location) {
     }
 
     fn visit_statement(&mut self, statement: &Statement<'tcx>, location: Location) {
+        match &statement.kind {
+            StatementKind::Assign(box (dest, rvalue)) => {
+                // LHS and RHS of the assignment must have the same type.
+                let left_ty = dest.ty(&self.body.local_decls, self.tcx).ty;
+                let right_ty = rvalue.ty(&self.body.local_decls, self.tcx);
+                if !self.mir_assign_valid_types(right_ty, left_ty) {
+                    self.fail(
+                        location,
+                        format!(
+                            "encountered `Assign` statement with incompatible types:\n\
+                            left-hand side has type: {}\n\
+                            right-hand side has type: {}",
+                            left_ty, right_ty,
+                        ),
+                    );
+                }
                 // The sides of an assignment must not alias. Currently this just checks whether the places
                 // are identical.
-        if let StatementKind::Assign(box (dest, rvalue)) = &statement.kind {
                 match rvalue {
                     Rvalue::Use(Operand::Copy(src) | Operand::Move(src)) => {
                         if dest == src {
@@ -114,6 +247,8 @@ fn visit_statement(&mut self, statement: &Statement<'tcx>, location: Location) {
                     _ => {}
                 }
             }
+            _ => {}
+        }
     }
 
     fn visit_terminator(&mut self, terminator: &Terminator<'tcx>, location: Location) {