cheri: Modify LLVM wrapper to support CHERI tag options for memcpy and memmove

S Harris · xdoardo · xdoardo · commit b4835c371dc6 · 2025-08-26T14:08:28.000+02:00
Co-authored-by: Edoardo Marangoni &lt;ecmm@anche.no&gt;
diff --git a/.cirrus.yml b/.cirrus.yml
@@ -32,7 +32,12 @@ pr_check_commits_task:
 x_check_task:
   skip: $CIRRUS_PR == ""
   name: (PR) Run ./x check
-  depends_on: (PR) Check commits structure
+  # TODO(xdoardo): Figure out when it makes sense to have this dependency, or
+  #   how that dependency should actually work. Having this dependency would
+  #   pretty much mean making every PR that merges changes on `beta` to `master`
+  #   to have failing CI, which would be the reason why we want to make PRs and
+  #   not direct pushes in the first place.
+  # depends_on: (PR) Check commits structure
   timeout_in: 240m
   gce_instance: &arm_vm
     image_project: ubuntu-os-cloud
@@ -54,6 +59,8 @@ x_check_task:
     - export CCACHE_REMOTE_STORAGE="http://${CIRRUS_HTTP_CACHE_HOST}/${CIRRUS_OS}/"
     - export CCACHE_REMOTE_ONLY=1
     - env
-  test_script: CC="clang" CXX="clang++" ./x check
+  pull_master_script: git fetch origin master
+  tidy_script: CC="clang" CXX="clang++" ./x test tidy
+  check_script: CC="clang" CXX="clang++" ./x check
 
 # -- End PR tasks
diff --git a/cheri/gen_bootstrap.sh b/cheri/gen_bootstrap.sh
@@ -17,7 +17,7 @@ change-id = 140732
 ccache = true
 
 [rust]
-#channel = "nightly"
+channel = "beta"
 #codegen-backends = ["llvm"]
 #debug = true
 #debuginfo-level = 2
diff --git a/compiler/rustc_codegen_gcc/src/builder.rs b/compiler/rustc_codegen_gcc/src/builder.rs
@@ -12,7 +12,7 @@ use rustc_abi::{Align, HasDataLayout, Size, TargetDataLayout, WrappingRange};
 use rustc_apfloat::{Float, Round, Status, ieee};
 use rustc_codegen_ssa::MemFlags;
 use rustc_codegen_ssa::common::{
-    AtomicRmwBinOp, IntPredicate, RealPredicate, SynchronizationScope, TypeKind,
+    AtomicRmwBinOp, IntPredicate, RealPredicate, SynchronizationScope, TypeKind, PreserveCheriTags,
 };
 use rustc_codegen_ssa::mir::operand::{OperandRef, OperandValue};
 use rustc_codegen_ssa::mir::place::PlaceRef;
@@ -1369,6 +1369,7 @@ impl<'a, 'gcc, 'tcx> BuilderMethods<'a, 'tcx> for Builder<'a, 'gcc, 'tcx> {
         _src_align: Align,
         size: RValue<'gcc>,
         flags: MemFlags,
+        _preserve_cheri_tags: PreserveCheriTags,
     ) {
         assert!(!flags.contains(MemFlags::NONTEMPORAL), "non-temporal memcpy not supported");
         let size = self.intcast(size, self.type_size_t(), false);
@@ -1391,6 +1392,7 @@ impl<'a, 'gcc, 'tcx> BuilderMethods<'a, 'tcx> for Builder<'a, 'gcc, 'tcx> {
         _src_align: Align,
         size: RValue<'gcc>,
         flags: MemFlags,
+        _preserve_cheri_flags: PreserveCheriTags,
     ) {
         assert!(!flags.contains(MemFlags::NONTEMPORAL), "non-temporal memmove not supported");
         let size = self.intcast(size, self.type_size_t(), false);
diff --git a/compiler/rustc_codegen_gcc/src/intrinsic/mod.rs b/compiler/rustc_codegen_gcc/src/intrinsic/mod.rs
@@ -12,7 +12,7 @@ use rustc_abi::ExternAbi;
 use rustc_abi::{BackendRepr, HasDataLayout};
 use rustc_codegen_ssa::MemFlags;
 use rustc_codegen_ssa::base::wants_msvc_seh;
-use rustc_codegen_ssa::common::IntPredicate;
+use rustc_codegen_ssa::common::{IntPredicate, PreserveCheriTags};
 use rustc_codegen_ssa::errors::InvalidMonomorphization;
 use rustc_codegen_ssa::mir::operand::{OperandRef, OperandValue};
 use rustc_codegen_ssa::mir::place::{PlaceRef, PlaceValue};
@@ -771,6 +771,7 @@ impl<'gcc, 'tcx> ArgAbiExt<'gcc, 'tcx> for ArgAbi<'tcx, Ty<'tcx>> {
                     scratch_align,
                     bx.const_usize(self.layout.size.bytes()),
                     MemFlags::empty(),
+                    PreserveCheriTags::Unknown,
                 );
 
                 bx.lifetime_end(scratch, scratch_size);
diff --git a/compiler/rustc_codegen_llvm/src/abi.rs b/compiler/rustc_codegen_llvm/src/abi.rs
@@ -7,6 +7,7 @@ use rustc_abi::{
     X86Call,
 };
 use rustc_codegen_ssa::MemFlags;
+use rustc_codegen_ssa::common::PreserveCheriTags;
 use rustc_codegen_ssa::mir::operand::{OperandRef, OperandValue};
 use rustc_codegen_ssa::mir::place::{PlaceRef, PlaceValue};
 use rustc_codegen_ssa::traits::*;
@@ -238,6 +239,8 @@ impl<'ll, 'tcx> ArgAbiExt<'ll, 'tcx> for ArgAbi<'tcx, Ty<'tcx>> {
                     scratch_align,
                     bx.const_usize(copy_bytes),
                     MemFlags::empty(),
+                    // Handling of CHERI capabilities could probably be more efficient.
+                    PreserveCheriTags::Unknown,
                 );
                 bx.lifetime_end(llscratch, scratch_size);
             }
diff --git a/compiler/rustc_codegen_llvm/src/builder.rs b/compiler/rustc_codegen_llvm/src/builder.rs
@@ -35,7 +35,7 @@ use crate::attributes;
 use crate::common::Funclet;
 use crate::context::{CodegenCx, FullCx, GenericCx, SCx};
 use crate::llvm::{
-    self, AtomicOrdering, AtomicRmwBinOp, BasicBlock, False, GEPNoWrapFlags, Metadata, True,
+    self, AtomicOrdering, AtomicRmwBinOp, BasicBlock, False, GEPNoWrapFlags, Metadata, True, PreserveCheriTags
 };
 use crate::type_::Type;
 use crate::type_of::LayoutLlvmExt;
@@ -1106,9 +1106,11 @@ impl<'a, 'll, 'tcx> BuilderMethods<'a, 'tcx> for Builder<'a, 'll, 'tcx> {
         src_align: Align,
         size: &'ll Value,
         flags: MemFlags,
+        preserve_tags: rustc_codegen_ssa::common::PreserveCheriTags,
     ) {
         assert!(!flags.contains(MemFlags::NONTEMPORAL), "non-temporal memcpy not supported");
         let size = self.intcast(size, self.type_isize(), false);
+        let preserve_tags = PreserveCheriTags::from_generic(preserve_tags);
         let is_volatile = flags.contains(MemFlags::VOLATILE);
         unsafe {
             llvm::LLVMRustBuildMemCpy(
@@ -1118,6 +1120,7 @@ impl<'a, 'll, 'tcx> BuilderMethods<'a, 'tcx> for Builder<'a, 'll, 'tcx> {
                 src,
                 src_align.bytes() as c_uint,
                 size,
+                preserve_tags,
                 is_volatile,
             );
         }
@@ -1131,9 +1134,11 @@ impl<'a, 'll, 'tcx> BuilderMethods<'a, 'tcx> for Builder<'a, 'll, 'tcx> {
         src_align: Align,
         size: &'ll Value,
         flags: MemFlags,
+        preserve_tags: rustc_codegen_ssa::common::PreserveCheriTags,
     ) {
         assert!(!flags.contains(MemFlags::NONTEMPORAL), "non-temporal memmove not supported");
         let size = self.intcast(size, self.type_isize(), false);
+        let preserve_tags = PreserveCheriTags::from_generic(preserve_tags);
         let is_volatile = flags.contains(MemFlags::VOLATILE);
         unsafe {
             llvm::LLVMRustBuildMemMove(
@@ -1143,6 +1148,7 @@ impl<'a, 'll, 'tcx> BuilderMethods<'a, 'tcx> for Builder<'a, 'll, 'tcx> {
                 src,
                 src_align.bytes() as c_uint,
                 size,
+                preserve_tags,
                 is_volatile,
             );
         }
diff --git a/compiler/rustc_codegen_llvm/src/llvm/ffi.rs b/compiler/rustc_codegen_llvm/src/llvm/ffi.rs
@@ -741,6 +741,25 @@ pub(crate) enum Opcode {
     CatchSwitch = 65,
 }
 
+/// LLVMPreserveCheriTags
+#[derive(Copy, Clone)]
+#[repr(C)]
+pub(crate) enum PreserveCheriTags {
+    Unknown,
+    Required,
+    Unnecessary,
+}
+
+impl PreserveCheriTags {
+    pub(crate) fn from_generic(value: rustc_codegen_ssa::common::PreserveCheriTags) -> Self {
+        match value {
+            rustc_codegen_ssa::common::PreserveCheriTags::Unknown => PreserveCheriTags::Unknown,
+            rustc_codegen_ssa::common::PreserveCheriTags::Required => PreserveCheriTags::Required,
+            rustc_codegen_ssa::common::PreserveCheriTags::Unnecessary => PreserveCheriTags::Unnecessary,
+        }
+    }
+}
+
 unsafe extern "C" {
     type Opaque;
 }
@@ -1933,6 +1952,7 @@ unsafe extern "C" {
         Src: &'a Value,
         SrcAlign: c_uint,
         Size: &'a Value,
+        PreserveTags: PreserveCheriTags,
         IsVolatile: bool,
     ) -> &'a Value;
     pub(crate) fn LLVMRustBuildMemMove<'a>(
@@ -1942,6 +1962,7 @@ unsafe extern "C" {
         Src: &'a Value,
         SrcAlign: c_uint,
         Size: &'a Value,
+        PreserveTags: PreserveCheriTags,
         IsVolatile: bool,
     ) -> &'a Value;
     pub(crate) fn LLVMRustBuildMemSet<'a>(
diff --git a/compiler/rustc_codegen_llvm/src/va_arg.rs b/compiler/rustc_codegen_llvm/src/va_arg.rs
@@ -1,6 +1,6 @@
 use rustc_abi::{Align, BackendRepr, Endian, HasDataLayout, Primitive, Size, TyAndLayout};
 use rustc_codegen_ssa::MemFlags;
-use rustc_codegen_ssa::common::IntPredicate;
+use rustc_codegen_ssa::common::{IntPredicate, PreserveCheriTags};
 use rustc_codegen_ssa::mir::operand::OperandRef;
 use rustc_codegen_ssa::traits::{
     BaseTypeCodegenMethods, BuilderMethods, ConstCodegenMethods, LayoutTypeCodegenMethods,
@@ -735,6 +735,7 @@ fn copy_to_temporary_if_more_aligned<'ll, 'tcx>(
             src_align,
             bx.const_u32(layout.layout.size().bytes() as u32),
             MemFlags::empty(),
+            PreserveCheriTags::Unknown,
         );
         tmp
     } else {
diff --git a/compiler/rustc_codegen_ssa/src/common.rs b/compiler/rustc_codegen_ssa/src/common.rs
@@ -59,6 +59,13 @@ pub enum AtomicRmwBinOp {
     AtomicUMin,
 }
 
+#[derive(Copy, Clone, Debug)]
+pub enum PreserveCheriTags {
+    Unknown,
+    Required,
+    Unnecessary,
+}
+
 #[derive(Copy, Clone, Debug)]
 pub enum SynchronizationScope {
     SingleThread,
diff --git a/compiler/rustc_codegen_ssa/src/mir/block.rs b/compiler/rustc_codegen_ssa/src/mir/block.rs
@@ -21,7 +21,7 @@ use super::operand::OperandValue::{Immediate, Pair, Ref, ZeroSized};
 use super::place::{PlaceRef, PlaceValue};
 use super::{CachedLlbb, FunctionCx, LocalRef};
 use crate::base::{self, is_call_from_compiler_builtins_to_upstream_monomorphization};
-use crate::common::{self, IntPredicate};
+use crate::common::{self, IntPredicate, PreserveCheriTags};
 use crate::errors::CompilerBuiltinsCannotCall;
 use crate::traits::*;
 use crate::{MemFlags, meth};
@@ -1601,6 +1601,7 @@ impl<'a, 'tcx, Bx: BuilderMethods<'a, 'tcx>> FunctionCx<'a, 'tcx, Bx> {
                     align,
                     bx.const_usize(copy_bytes),
                     MemFlags::empty(),
+                    PreserveCheriTags::Unknown,
                 );
                 // ...and then load it with the ABI type.
                 llval = load_cast(bx, cast, llscratch, scratch_align);
diff --git a/compiler/rustc_codegen_ssa/src/mir/intrinsic.rs b/compiler/rustc_codegen_ssa/src/mir/intrinsic.rs
@@ -8,7 +8,7 @@ use rustc_span::sym;
 use super::FunctionCx;
 use super::operand::OperandRef;
 use super::place::PlaceRef;
-use crate::common::{AtomicRmwBinOp, SynchronizationScope};
+use crate::common::{AtomicRmwBinOp, SynchronizationScope, PreserveCheriTags};
 use crate::errors::InvalidMonomorphization;
 use crate::traits::*;
 use crate::{MemFlags, meth, size_of_val};
@@ -27,10 +27,12 @@ fn copy_intrinsic<'a, 'tcx, Bx: BuilderMethods<'a, 'tcx>>(
     let align = layout.align.abi;
     let size = bx.mul(bx.const_usize(size.bytes()), count);
     let flags = if volatile { MemFlags::VOLATILE } else { MemFlags::empty() };
+    // Handling of CHERI capabilities could probably be more efficient.
+    let preserve_tags = PreserveCheriTags::Unknown;
     if allow_overlap {
-        bx.memmove(dst, align, src, align, size, flags);
+        bx.memmove(dst, align, src, align, size, flags, preserve_tags);
     } else {
-        bx.memcpy(dst, align, src, align, size, flags);
+        bx.memcpy(dst, align, src, align, size, flags, preserve_tags);
     }
 }
 
diff --git a/compiler/rustc_codegen_ssa/src/mir/statement.rs b/compiler/rustc_codegen_ssa/src/mir/statement.rs
@@ -3,6 +3,7 @@ use rustc_middle::span_bug;
 use tracing::instrument;
 
 use super::{FunctionCx, LocalRef};
+use crate::common::PreserveCheriTags;
 use crate::traits::*;
 
 impl<'a, 'tcx, Bx: BuilderMethods<'a, 'tcx>> FunctionCx<'a, 'tcx, Bx> {
@@ -90,7 +91,8 @@ impl<'a, 'tcx, Bx: BuilderMethods<'a, 'tcx>> FunctionCx<'a, 'tcx, Bx> {
                 let align = pointee_layout.align;
                 let dst = dst_val.immediate();
                 let src = src_val.immediate();
-                bx.memcpy(dst, align, src, align, bytes, crate::MemFlags::empty());
+                // Handling of CHERI capabilities could probably be more efficient.
+                bx.memcpy(dst, align, src, align, bytes, crate::MemFlags::empty(), PreserveCheriTags::Unknown);
             }
             mir::StatementKind::FakeRead(..)
             | mir::StatementKind::Retag { .. }
diff --git a/compiler/rustc_codegen_ssa/src/traits/builder.rs b/compiler/rustc_codegen_ssa/src/traits/builder.rs
@@ -19,7 +19,9 @@ use super::misc::MiscCodegenMethods;
 use super::type_::{ArgAbiBuilderMethods, BaseTypeCodegenMethods, LayoutTypeCodegenMethods};
 use super::{CodegenMethods, StaticBuilderMethods};
 use crate::MemFlags;
-use crate::common::{AtomicRmwBinOp, IntPredicate, RealPredicate, SynchronizationScope, TypeKind};
+use crate::common::{
+     PreserveCheriTags, AtomicRmwBinOp, IntPredicate, RealPredicate, SynchronizationScope, TypeKind,
+};
 use crate::mir::operand::{OperandRef, OperandValue};
 use crate::mir::place::{PlaceRef, PlaceValue};
 
@@ -424,6 +426,7 @@ pub trait BuilderMethods<'a, 'tcx>:
         src_align: Align,
         size: Self::Value,
         flags: MemFlags,
+        preserve_tags: PreserveCheriTags,
     );
     fn memmove(
         &mut self,
@@ -433,6 +436,7 @@ pub trait BuilderMethods<'a, 'tcx>:
         src_align: Align,
         size: Self::Value,
         flags: MemFlags,
+        preserve_tags: PreserveCheriTags,
     );
     fn memset(
         &mut self,
@@ -480,7 +484,8 @@ pub trait BuilderMethods<'a, 'tcx>:
             temp.val.store_with_flags(self, dst.with_type(layout), flags);
         } else if !layout.is_zst() {
             let bytes = self.const_usize(layout.size.bytes());
-            self.memcpy(dst.llval, dst.align, src.llval, src.align, bytes, flags);
+            self.memcpy(dst.llval, dst.align, src.llval, src.align, bytes, flags,
+                PreserveCheriTags::Unknown);
         }
     }
 
diff --git a/compiler/rustc_llvm/llvm-wrapper/RustWrapper.cpp b/compiler/rustc_llvm/llvm-wrapper/RustWrapper.cpp
@@ -1564,23 +1564,47 @@ LLVMRustUnpackSMDiagnostic(LLVMSMDiagnosticRef DRef, RustStringRef MessageOut,
   return true;
 }
 
-extern "C" LLVMValueRef LLVMRustBuildMemCpy(LLVMBuilderRef B, LLVMValueRef Dst,
-                                            unsigned DstAlign, LLVMValueRef Src,
-                                            unsigned SrcAlign,
-                                            LLVMValueRef Size,
-                                            bool IsVolatile) {
-  return wrap(unwrap(B)->CreateMemCpy(unwrap(Dst), MaybeAlign(DstAlign),
-                                      unwrap(Src), MaybeAlign(SrcAlign),
-                                      unwrap(Size), IsVolatile));
+enum class LLVMPreserveCheriTags {
+    Unknown,
+    Required,
+    Unnecessary,
+};
+
+static PreserveCheriTags fromRust(LLVMPreserveCheriTags PreserveTags) {
+  switch (PreserveTags) {
+  case LLVMPreserveCheriTags::Unknown:
+    return PreserveCheriTags::Unknown;
+  case LLVMPreserveCheriTags::Required:
+    return PreserveCheriTags::Required;
+  case LLVMPreserveCheriTags::Unnecessary:
+    return PreserveCheriTags::Unnecessary;
+  }
+
+  report_fatal_error("Invalid LLVMPreserveCheriTags value!");
 }
 
-extern "C" LLVMValueRef
-LLVMRustBuildMemMove(LLVMBuilderRef B, LLVMValueRef Dst, unsigned DstAlign,
-                     LLVMValueRef Src, unsigned SrcAlign, LLVMValueRef Size,
-                     bool IsVolatile) {
-  return wrap(unwrap(B)->CreateMemMove(unwrap(Dst), MaybeAlign(DstAlign),
-                                       unwrap(Src), MaybeAlign(SrcAlign),
-                                       unwrap(Size), IsVolatile));
+extern "C" LLVMValueRef LLVMRustBuildMemCpy(LLVMBuilderRef B,
+                                            LLVMValueRef Dst, unsigned DstAlign,
+                                            LLVMValueRef Src, unsigned SrcAlign,
+                                            LLVMValueRef Size, LLVMPreserveCheriTags PreserveTags,
+                                            bool IsVolatile) {
+  return wrap(unwrap(B)->CreateMemCpy(
+      unwrap(Dst), MaybeAlign(DstAlign),
+      unwrap(Src), MaybeAlign(SrcAlign),
+      unwrap(Size), fromRust(PreserveTags),
+      IsVolatile));
+}
+
+extern "C" LLVMValueRef LLVMRustBuildMemMove(LLVMBuilderRef B,
+                                             LLVMValueRef Dst, unsigned DstAlign,
+                                             LLVMValueRef Src, unsigned SrcAlign,
+                                             LLVMValueRef Size, LLVMPreserveCheriTags PreserveTags,
+                                             bool IsVolatile) {
+  return wrap(unwrap(B)->CreateMemMove(
+      unwrap(Dst), MaybeAlign(DstAlign),
+      unwrap(Src), MaybeAlign(SrcAlign),
+      unwrap(Size), fromRust(PreserveTags),
+      IsVolatile));
 }
 
 extern "C" LLVMValueRef LLVMRustBuildMemSet(LLVMBuilderRef B, LLVMValueRef Dst,
