Auto merge of #120405 - cjgillot:gvn-pointer, r=oli-obk

Fold pointer operations in GVN

This PR proposes 2 combinations of cast operations in MIR GVN:
- a chain of `PtrToPtr` or `MutToConstPointer` casts can be folded together into a single `PtrToPtr` cast;
- we attempt to evaluate more ptr ops when there is no provenance.

In particular, this allows to read from static slices.

This is not yet sufficient to see through slice operations that use `PtrComponents` (because that's a union), but still a step forward.

r? `@ghost`

compiler/rustc_mir_transform/src/dataflow_const_prop.rs

@@ -2,7 +2,9 @@
 //!
 //! Currently, this pass only propagates scalar values.
 
-use rustc_const_eval::interpret::{ImmTy, Immediate, InterpCx, OpTy, PlaceTy, Projectable};
+use rustc_const_eval::interpret::{
+    ImmTy, Immediate, InterpCx, OpTy, PlaceTy, PointerArithmetic, Projectable,
+};
 use rustc_data_structures::fx::FxHashMap;
 use rustc_hir::def::DefKind;
 use rustc_middle::mir::interpret::{AllocId, ConstAllocation, InterpResult, Scalar};
@@ -936,12 +938,50 @@ impl<'mir, 'tcx: 'mir> rustc_const_eval::interpret::Machine<'mir, 'tcx> for Dumm
     }
 
     fn binary_ptr_op(
-        _ecx: &InterpCx<'mir, 'tcx, Self>,
-        _bin_op: BinOp,
-        _left: &rustc_const_eval::interpret::ImmTy<'tcx, Self::Provenance>,
-        _right: &rustc_const_eval::interpret::ImmTy<'tcx, Self::Provenance>,
+        ecx: &InterpCx<'mir, 'tcx, Self>,
+        bin_op: BinOp,
+        left: &rustc_const_eval::interpret::ImmTy<'tcx, Self::Provenance>,
+        right: &rustc_const_eval::interpret::ImmTy<'tcx, Self::Provenance>,
     ) -> interpret::InterpResult<'tcx, (ImmTy<'tcx, Self::Provenance>, bool)> {
-        throw_machine_stop_str!("can't do pointer arithmetic");
+        use rustc_middle::mir::BinOp::*;
+        Ok(match bin_op {
+            Eq | Ne | Lt | Le | Gt | Ge => {
+                // Types can differ, e.g. fn ptrs with different `for`.
+                assert_eq!(left.layout.abi, right.layout.abi);
+                let size = ecx.pointer_size();
+                // Just compare the bits. ScalarPairs are compared lexicographically.
+                // We thus always compare pairs and simply fill scalars up with 0.
+                // If the pointer has provenance, `to_bits` will return `Err` and we bail out.
+                let left = match **left {
+                    Immediate::Scalar(l) => (l.to_bits(size)?, 0),
+                    Immediate::ScalarPair(l1, l2) => (l1.to_bits(size)?, l2.to_bits(size)?),
+                    Immediate::Uninit => panic!("we should never see uninit data here"),
+                };
+                let right = match **right {
+                    Immediate::Scalar(r) => (r.to_bits(size)?, 0),
+                    Immediate::ScalarPair(r1, r2) => (r1.to_bits(size)?, r2.to_bits(size)?),
+                    Immediate::Uninit => panic!("we should never see uninit data here"),
+                };
+                let res = match bin_op {
+                    Eq => left == right,
+                    Ne => left != right,
+                    Lt => left < right,
+                    Le => left <= right,
+                    Gt => left > right,
+                    Ge => left >= right,
+                    _ => bug!(),
+                };
+                (ImmTy::from_bool(res, *ecx.tcx), false)
+            }
+
+            // Some more operations are possible with atomics.
+            // The return value always has the provenance of the *left* operand.
+            Add | Sub | BitOr | BitAnd | BitXor => {
+                throw_machine_stop_str!("pointer arithmetic is not handled")
+            }
+
+            _ => span_bug!(ecx.cur_span(), "Invalid operator on pointers: {:?}", bin_op),
+        })
     }
 
     fn expose_ptr(

compiler/rustc_mir_transform/src/gvn.rs

@@ -93,7 +93,6 @@ use rustc_index::IndexVec;
 use rustc_middle::mir::interpret::GlobalAlloc;
 use rustc_middle::mir::visit::*;
 use rustc_middle::mir::*;
-use rustc_middle::ty::adjustment::PointerCoercion;
 use rustc_middle::ty::layout::LayoutOf;
 use rustc_middle::ty::{self, Ty, TyCtxt, TypeAndMut};
 use rustc_span::def_id::DefId;
@@ -552,6 +551,29 @@ impl<'body, 'tcx> VnState<'body, 'tcx> {
                     }
                     value.offset(Size::ZERO, to, &self.ecx).ok()?
                 }
+                CastKind::PointerCoercion(ty::adjustment::PointerCoercion::Unsize) => {
+                    let src = self.evaluated[value].as_ref()?;
+                    let to = self.ecx.layout_of(to).ok()?;
+                    let dest = self.ecx.allocate(to, MemoryKind::Stack).ok()?;
+                    self.ecx.unsize_into(src, to, &dest.clone().into()).ok()?;
+                    self.ecx
+                        .alloc_mark_immutable(dest.ptr().provenance.unwrap().alloc_id())
+                        .ok()?;
+                    dest.into()
+                }
+                CastKind::FnPtrToPtr
+                | CastKind::PtrToPtr
+                | CastKind::PointerCoercion(
+                    ty::adjustment::PointerCoercion::MutToConstPointer
+                    | ty::adjustment::PointerCoercion::ArrayToPointer
+                    | ty::adjustment::PointerCoercion::UnsafeFnPointer,
+                ) => {
+                    let src = self.evaluated[value].as_ref()?;
+                    let src = self.ecx.read_immediate(src).ok()?;
+                    let to = self.ecx.layout_of(to).ok()?;
+                    let ret = self.ecx.ptr_to_ptr(&src, to).ok()?;
+                    ret.into()
+                }
                 _ => return None,
             },
         };
@@ -778,18 +800,8 @@ impl<'body, 'tcx> VnState<'body, 'tcx> {
 
             // Operations.
             Rvalue::Len(ref mut place) => return self.simplify_len(place, location),
-            Rvalue::Cast(kind, ref mut value, to) => {
-                let from = value.ty(self.local_decls, self.tcx);
-                let value = self.simplify_operand(value, location)?;
-                if let CastKind::PointerCoercion(
-                    PointerCoercion::ReifyFnPointer | PointerCoercion::ClosureFnPointer(_),
-                ) = kind
-                {
-                    // Each reification of a generic fn may get a different pointer.
-                    // Do not try to merge them.
-                    return self.new_opaque();
-                }
-                Value::Cast { kind, value, from, to }
+            Rvalue::Cast(ref mut kind, ref mut value, to) => {
+                return self.simplify_cast(kind, value, to, location);
             }
             Rvalue::BinaryOp(op, box (ref mut lhs, ref mut rhs)) => {
                 let ty = lhs.ty(self.local_decls, self.tcx);
@@ -1035,6 +1047,50 @@ impl<'body, 'tcx> VnState<'body, 'tcx> {
         }
     }
 
+    fn simplify_cast(
+        &mut self,
+        kind: &mut CastKind,
+        operand: &mut Operand<'tcx>,
+        to: Ty<'tcx>,
+        location: Location,
+    ) -> Option<VnIndex> {
+        use rustc_middle::ty::adjustment::PointerCoercion::*;
+        use CastKind::*;
+
+        let mut from = operand.ty(self.local_decls, self.tcx);
+        let mut value = self.simplify_operand(operand, location)?;
+        if from == to {
+            return Some(value);
+        }
+
+        if let CastKind::PointerCoercion(ReifyFnPointer | ClosureFnPointer(_)) = kind {
+            // Each reification of a generic fn may get a different pointer.
+            // Do not try to merge them.
+            return self.new_opaque();
+        }
+
+        if let PtrToPtr | PointerCoercion(MutToConstPointer) = kind
+            && let Value::Cast { kind: inner_kind, value: inner_value, from: inner_from, to: _ } =
+                *self.get(value)
+            && let PtrToPtr | PointerCoercion(MutToConstPointer) = inner_kind
+        {
+            from = inner_from;
+            value = inner_value;
+            *kind = PtrToPtr;
+            if inner_from == to {
+                return Some(inner_value);
+            }
+            if let Some(const_) = self.try_as_constant(value) {
+                *operand = Operand::Constant(Box::new(const_));
+            } else if let Some(local) = self.try_as_local(value, location) {
+                *operand = Operand::Copy(local.into());
+                self.reused_locals.insert(local);
+            }
+        }
+
+        Some(self.insert(Value::Cast { kind: *kind, value, from, to }))
+    }
+
     fn simplify_len(&mut self, place: &mut Place<'tcx>, location: Location) -> Option<VnIndex> {
         // Trivial case: we are fetching a statically known length.
         let place_ty = place.ty(self.local_decls, self.tcx).ty;