Rollup merge of #108442 - scottmcm:mir-transmute, r=oli-obk

Add `CastKind::Transmute` to MIR

~~Nothing actually produces it in this commit, so I don't know how to test it, but it also means it shouldn't be possible for it to break anything.~~

Includes lowering `transmute` calls to it, so it's used.

Zulip Conversation: <https://rust-lang.zulipchat.com/#narrow/stream/189540-t-compiler.2Fwg-mir-opt/topic/Good.20first.20isssue/near/321849610>

Author: matthiaskrgr

compiler/rustc_borrowck/src/type_check/mod.rs

@@ -2242,6 +2242,13 @@ impl<'a, 'tcx> TypeChecker<'a, 'tcx> {
                             }
                         }
                     }
+                    CastKind::Transmute => {
+                        span_mirbug!(
+                            self,
+                            rvalue,
+                            "Unexpected CastKind::Transmute, which is not permitted in Analysis MIR",
+                        );
+                    }
                 }
             }
 

compiler/rustc_codegen_cranelift/src/base.rs

@@ -720,6 +720,10 @@ fn codegen_stmt<'tcx>(
                     let operand = codegen_operand(fx, operand);
                     operand.coerce_dyn_star(fx, lval);
                 }
+                Rvalue::Cast(CastKind::Transmute, ref operand, _to_ty) => {
+                    let operand = codegen_operand(fx, operand);
+                    lval.write_cvalue_transmute(fx, operand);
+                }
                 Rvalue::Discriminant(place) => {
                     let place = codegen_place(fx, place);
                     let value = place.to_cvalue(fx);

compiler/rustc_codegen_ssa/src/mir/block.rs

@@ -16,7 +16,7 @@ use rustc_index::vec::Idx;
 use rustc_middle::mir::{self, AssertKind, SwitchTargets};
 use rustc_middle::ty::layout::{HasTyCtxt, LayoutOf, ValidityRequirement};
 use rustc_middle::ty::print::{with_no_trimmed_paths, with_no_visible_paths};
-use rustc_middle::ty::{self, Instance, Ty, TypeVisitableExt};
+use rustc_middle::ty::{self, Instance, Ty};
 use rustc_session::config::OptLevel;
 use rustc_span::source_map::Span;
 use rustc_span::{sym, Symbol};
@@ -765,6 +765,10 @@ impl<'a, 'tcx, Bx: BuilderMethods<'a, 'tcx>> FunctionCx<'a, 'tcx, Bx> {
             _ => None,
         };
 
+        if let Some(intrinsic @ (sym::copy_nonoverlapping | sym::transmute)) = intrinsic {
+            bug!("Intrinsic {intrinsic} should have been lowered before codegen");
+        }
+
         let extra_args = &args[sig.inputs().skip_binder().len()..];
         let extra_args = bx.tcx().mk_type_list_from_iter(extra_args.iter().map(|op_arg| {
             let op_ty = op_arg.ty(self.mir, bx.tcx());
@@ -776,23 +780,6 @@ impl<'a, 'tcx, Bx: BuilderMethods<'a, 'tcx>> FunctionCx<'a, 'tcx, Bx> {
             None => bx.fn_abi_of_fn_ptr(sig, extra_args),
         };
 
-        if intrinsic == Some(sym::transmute) {
-            return if let Some(target) = target {
-                self.codegen_transmute(bx, &args[0], destination);
-                helper.funclet_br(self, bx, target, mergeable_succ)
-            } else {
-                // If we are trying to transmute to an uninhabited type,
-                // it is likely there is no allotted destination. In fact,
-                // transmuting to an uninhabited type is UB, which means
-                // we can do what we like. Here, we declare that transmuting
-                // into an uninhabited type is impossible, so anything following
-                // it must be unreachable.
-                assert_eq!(fn_abi.ret.layout.abi, abi::Abi::Uninhabited);
-                bx.unreachable();
-                MergingSucc::False
-            };
-        }
-
         if let Some(merging_succ) = self.codegen_panic_intrinsic(
             &helper,
             bx,
@@ -835,7 +822,6 @@ impl<'a, 'tcx, Bx: BuilderMethods<'a, 'tcx>> FunctionCx<'a, 'tcx, Bx> {
 
         match intrinsic {
             None | Some(sym::drop_in_place) => {}
-            Some(sym::copy_nonoverlapping) => unreachable!(),
             Some(intrinsic) => {
                 let dest = match ret_dest {
                     _ if fn_abi.ret.is_indirect() => llargs[0],
@@ -1746,33 +1732,7 @@ impl<'a, 'tcx, Bx: BuilderMethods<'a, 'tcx>> FunctionCx<'a, 'tcx, Bx> {
         }
     }
 
-    fn codegen_transmute(&mut self, bx: &mut Bx, src: &mir::Operand<'tcx>, dst: mir::Place<'tcx>) {
-        if let Some(index) = dst.as_local() {
-            match self.locals[index] {
-                LocalRef::Place(place) => self.codegen_transmute_into(bx, src, place),
-                LocalRef::UnsizedPlace(_) => bug!("transmute must not involve unsized locals"),
-                LocalRef::Operand(None) => {
-                    let dst_layout = bx.layout_of(self.monomorphized_place_ty(dst.as_ref()));
-                    assert!(!dst_layout.ty.has_erasable_regions());
-                    let place = PlaceRef::alloca(bx, dst_layout);
-                    place.storage_live(bx);
-                    self.codegen_transmute_into(bx, src, place);
-                    let op = bx.load_operand(place);
-                    place.storage_dead(bx);
-                    self.locals[index] = LocalRef::Operand(Some(op));
-                    self.debug_introduce_local(bx, index);
-                }
-                LocalRef::Operand(Some(op)) => {
-                    assert!(op.layout.is_zst(), "assigning to initialized SSAtemp");
-                }
-            }
-        } else {
-            let dst = self.codegen_place(bx, dst.as_ref());
-            self.codegen_transmute_into(bx, src, dst);
-        }
-    }
-
-    fn codegen_transmute_into(
+    pub(crate) fn codegen_transmute_into(
         &mut self,
         bx: &mut Bx,
         src: &mir::Operand<'tcx>,

compiler/rustc_codegen_ssa/src/mir/rvalue.rs

@@ -135,6 +135,10 @@ impl<'a, 'tcx, Bx: BuilderMethods<'a, 'tcx>> FunctionCx<'a, 'tcx, Bx> {
                 dest.codegen_set_discr(bx, variant_index);
             }
 
+            mir::Rvalue::Cast(mir::CastKind::Transmute, ref operand, _ty) => {
+                self.codegen_transmute_into(bx, operand, dest);
+            }
+
             _ => {
                 assert!(self.rvalue_creates_operand(rvalue, DUMMY_SP));
                 let temp = self.codegen_rvalue_operand(bx, rvalue);
@@ -344,6 +348,9 @@ impl<'a, 'tcx, Bx: BuilderMethods<'a, 'tcx>> FunctionCx<'a, 'tcx, Bx> {
                         };
                         OperandValue::Immediate(newval)
                     }
+                    mir::CastKind::Transmute => {
+                        bug!("Transmute operand {:?} in `codegen_rvalue_operand`", operand);
+                    }
                 };
                 OperandRef { val, layout: cast }
             }
@@ -684,6 +691,9 @@ impl<'a, 'tcx, Bx: BuilderMethods<'a, 'tcx>> FunctionCx<'a, 'tcx, Bx> {
 impl<'a, 'tcx, Bx: BuilderMethods<'a, 'tcx>> FunctionCx<'a, 'tcx, Bx> {
     pub fn rvalue_creates_operand(&self, rvalue: &mir::Rvalue<'tcx>, span: Span) -> bool {
         match *rvalue {
+            mir::Rvalue::Cast(mir::CastKind::Transmute, ..) =>
+                // sometimes this could, but for aggregates it can't
+                false,
             mir::Rvalue::Ref(..) |
             mir::Rvalue::CopyForDeref(..) |
             mir::Rvalue::AddressOf(..) |

compiler/rustc_const_eval/src/interpret/cast.rs

@@ -133,6 +133,10 @@ impl<'mir, 'tcx: 'mir, M: Machine<'mir, 'tcx>> InterpCx<'mir, 'tcx, M> {
                     bug!()
                 }
             }
+
+            Transmute => {
+                self.copy_op(src, dest, /*allow_transmute*/ true)?;
+            }
         }
         Ok(())
     }

compiler/rustc_const_eval/src/interpret/intrinsics.rs

@@ -124,10 +124,13 @@ impl<'mir, 'tcx: 'mir, M: Machine<'mir, 'tcx>> InterpCx<'mir, 'tcx, M> {
         let substs = instance.substs;
         let intrinsic_name = self.tcx.item_name(instance.def_id());
 
+        if let sym::copy_nonoverlapping | sym::transmute = intrinsic_name {
+            bug!("Intrinsic {intrinsic_name} should have been lowered before interpreting MIR");
+        }
+
         // First handle intrinsics without return place.
         let ret = match ret {
             None => match intrinsic_name {
-                sym::transmute => throw_ub_format!("transmuting to uninhabited type"),
                 sym::abort => M::abort(self, "the program aborted execution".to_owned())?,
                 // Unsupported diverging intrinsic.
                 _ => return Ok(false),
@@ -411,9 +414,6 @@ impl<'mir, 'tcx: 'mir, M: Machine<'mir, 'tcx>> InterpCx<'mir, 'tcx, M> {
                 self.exact_div(&val, &size, dest)?;
             }
 
-            sym::transmute => {
-                self.copy_op(&args[0], dest, /*allow_transmute*/ true)?;
-            }
             sym::assert_inhabited
             | sym::assert_zero_valid
             | sym::assert_mem_uninitialized_valid => {

compiler/rustc_const_eval/src/transform/validate.rs

@@ -619,6 +619,43 @@ impl<'a, 'tcx> Visitor<'tcx> for TypeChecker<'a, 'tcx> {
                             );
                         }
                     }
+                    CastKind::Transmute => {
+                        if let MirPhase::Runtime(..) = self.mir_phase {
+                            // `mem::transmute` currently requires types concrete enough
+                            // to *know* that their sizes are the same, and we repeat
+                            // that check here. This restriction may be lifted in future,
+                            // should some optimizations need to it be more general.
+                            // (It might just end up being UB if they don't match, say.)
+
+                            let src_layout = self.tcx.layout_of(self.param_env.and(op_ty));
+                            let dst_layout = self.tcx.layout_of(self.param_env.and(*target_type));
+                            if let Err(e) = src_layout {
+                                self.fail(
+                                    location,
+                                    format!(
+                                        "Unable to compute layout for source type {op_ty:?}: {e}"
+                                    ),
+                                );
+                            }
+                            if let Err(e) = dst_layout {
+                                self.fail(location, format!("Unable to compute layout for destination type {target_type:?}: {e}"));
+                            }
+
+                            if let (Ok(src_layout), Ok(dst_layout)) = (src_layout, dst_layout) {
+                                if src_layout.layout.size() != dst_layout.layout.size() {
+                                    self.fail(location, format!("Source and destination layouts have different sizes: {src_layout:?} vs {dst_layout:?}"));
+                                }
+                            }
+                        } else {
+                            self.fail(
+                                location,
+                                format!(
+                                    "Transmute is not supported in non-runtime phase {:?}.",
+                                    self.mir_phase
+                                ),
+                            );
+                        }
+                    }
                 }
             }
             Rvalue::Repeat(_, _)

compiler/rustc_middle/src/mir/mod.rs

@@ -1963,7 +1963,8 @@ impl<'tcx> Rvalue<'tcx> {
                 | CastKind::PtrToPtr
                 | CastKind::Pointer(_)
                 | CastKind::PointerFromExposedAddress
-                | CastKind::DynStar,
+                | CastKind::DynStar
+                | CastKind::Transmute,
                 _,
                 _,
             )

compiler/rustc_middle/src/mir/syntax.rs

@@ -83,6 +83,9 @@ pub enum MirPhase {
     ///    access to. This occurs in generator bodies. Such locals do not behave like other locals,
     ///    because they eg may be aliased in surprising ways. Runtime MIR has no such special locals -
     ///    all generator bodies are lowered and so all places that look like locals really are locals.
+    ///  - Intrinsics: In analysis MIR, intrinsics are typically represented as [`TerminatorKind::Call`]s
+    ///    to their `extern "rust-intrinsic"` declarations. In runtime MIR, some intrinsics are lowered
+    ///    to MIR constructs not used earlier, such as `mem::transmute` → [`CastKind::Transmute`].
     ///
     /// Also note that the lint pass which reports eg `200_u8 + 200_u8` as an error is run as a part
     /// of analysis to runtime MIR lowering. To ensure lints are reported reliably, this means that
@@ -1189,6 +1192,10 @@ pub enum CastKind {
     IntToFloat,
     PtrToPtr,
     FnPtrToPtr,
+    /// Reinterpret the bits of the input as a different type.
+    ///
+    /// Allowed only in [`MirPhase::Runtime`]; Earlier it's a [`TerminatorKind::Call`].
+    Transmute,
 }
 
 #[derive(Clone, Debug, PartialEq, Eq, TyEncodable, TyDecodable, Hash, HashStable)]

compiler/rustc_mir_transform/src/check_unsafety.rs

@@ -132,6 +132,12 @@ impl<'tcx> Visitor<'tcx> for UnsafetyChecker<'_, 'tcx> {
                     self.register_violations(violations, used_unsafe_blocks.iter().copied());
                 }
             },
+            Rvalue::Cast(CastKind::Transmute, ..) => {
+                self.require_unsafe(
+                    UnsafetyViolationKind::General,
+                    UnsafetyViolationDetails::CallToUnsafeFunction,
+                );
+            }
             _ => {}
         }
         self.super_rvalue(rvalue, location);