Auto merge of #132549 - Zalathar:rust-string, r=cuviper

Make `RustString` an extern type to avoid `improper_ctypes` warnings

Currently, any FFI function that uses `&RustString` needs to also add `#[ignore(improper_ctypes)]` to silence a warning.

The warning is not _completely_ bogus, because `RustString` contains `Vec<u8>` and therefore does not have a guaranteed layout. But we have no way of telling the lint that this doesn't matter, because the C++ code only uses that pointer opaquely and never relies on its underlying layout.

Ideally there would be some way to silence `improper_ctypes` at the type-definition site. But because there isn't, casting to and from a separate extern type is better than having to annotate every single use site.

compiler/rustc_codegen_llvm/src/llvm/ffi.rs

@@ -1766,11 +1766,9 @@ unsafe extern "C" {
     pub fn LLVMRustGetLastError() -> *const c_char;
 
     /// Prints the timing information collected by `-Ztime-llvm-passes`.
-    #[expect(improper_ctypes)]
     pub(crate) fn LLVMRustPrintPassTimings(OutStr: &RustString);
 
     /// Prints the statistics collected by `-Zprint-codegen-stats`.
-    #[expect(improper_ctypes)]
     pub(crate) fn LLVMRustPrintStatistics(OutStr: &RustString);
 
     /// Prepares inline assembly.
@@ -1791,7 +1789,6 @@ unsafe extern "C" {
         ConstraintsLen: size_t,
     ) -> bool;
 
-    #[allow(improper_ctypes)]
     pub(crate) fn LLVMRustCoverageWriteFilenamesToBuffer(
         Filenames: *const *const c_char,
         FilenamesLen: size_t,
@@ -1800,7 +1797,6 @@ unsafe extern "C" {
         BufferOut: &RustString,
     );
 
-    #[allow(improper_ctypes)]
     pub(crate) fn LLVMRustCoverageWriteFunctionMappingsToBuffer(
         VirtualFileMappingIDs: *const c_uint,
         NumVirtualFileMappingIDs: size_t,
@@ -1824,13 +1820,10 @@ unsafe extern "C" {
     ) -> &Value;
     pub(crate) fn LLVMRustCoverageHashBytes(Bytes: *const c_char, NumBytes: size_t) -> u64;
 
-    #[allow(improper_ctypes)]
     pub(crate) fn LLVMRustCoverageWriteCovmapSectionNameToString(M: &Module, OutStr: &RustString);
 
-    #[allow(improper_ctypes)]
     pub(crate) fn LLVMRustCoverageWriteCovfunSectionNameToString(M: &Module, OutStr: &RustString);
 
-    #[allow(improper_ctypes)]
     pub(crate) fn LLVMRustCoverageWriteCovmapVarNameToString(OutStr: &RustString);
 
     pub(crate) fn LLVMRustCoverageMappingVersion() -> u32;
@@ -2185,14 +2178,11 @@ unsafe extern "C" {
     pub fn LLVMRustDIBuilderCreateOpPlusUconst() -> u64;
     pub fn LLVMRustDIBuilderCreateOpLLVMFragment() -> u64;
 
-    #[allow(improper_ctypes)]
     pub fn LLVMRustWriteTypeToString(Type: &Type, s: &RustString);
-    #[allow(improper_ctypes)]
     pub fn LLVMRustWriteValueToString(value_ref: &Value, s: &RustString);
 
     pub fn LLVMRustHasFeature(T: &TargetMachine, s: *const c_char) -> bool;
 
-    #[allow(improper_ctypes)]
     pub(crate) fn LLVMRustPrintTargetCPUs(TM: &TargetMachine, OutStr: &RustString);
     pub fn LLVMRustGetTargetFeaturesCount(T: &TargetMachine) -> size_t;
     pub fn LLVMRustGetTargetFeature(
@@ -2297,10 +2287,8 @@ unsafe extern "C" {
     pub fn LLVMRustArchiveIteratorFree<'a>(AIR: &'a mut ArchiveIterator<'a>);
     pub fn LLVMRustDestroyArchive(AR: &'static mut Archive);
 
-    #[allow(improper_ctypes)]
     pub fn LLVMRustWriteTwineToString(T: &Twine, s: &RustString);
 
-    #[allow(improper_ctypes)]
     pub fn LLVMRustUnpackOptimizationDiagnostic<'a>(
         DI: &'a DiagnosticInfo,
         pass_name_out: &RustString,
@@ -2318,7 +2306,6 @@ unsafe extern "C" {
         message_out: &mut Option<&'a Twine>,
     );
 
-    #[allow(improper_ctypes)]
     pub fn LLVMRustWriteDiagnosticInfoToString(DI: &DiagnosticInfo, s: &RustString);
     pub fn LLVMRustGetDiagInfoKind(DI: &DiagnosticInfo) -> DiagnosticKind;
 
@@ -2327,7 +2314,6 @@ unsafe extern "C" {
         cookie_out: &mut c_uint,
     ) -> &'a SMDiagnostic;
 
-    #[allow(improper_ctypes)]
     pub fn LLVMRustUnpackSMDiagnostic(
         d: &SMDiagnostic,
         message_out: &RustString,
@@ -2374,7 +2360,6 @@ unsafe extern "C" {
     pub fn LLVMRustModuleBufferLen(p: &ModuleBuffer) -> usize;
     pub fn LLVMRustModuleBufferFree(p: &'static mut ModuleBuffer);
     pub fn LLVMRustModuleCost(M: &Module) -> u64;
-    #[allow(improper_ctypes)]
     pub fn LLVMRustModuleInstructionStats(M: &Module, Str: &RustString);
 
     pub fn LLVMRustThinLTOBufferCreate(
@@ -2427,7 +2412,6 @@ unsafe extern "C" {
         bytecode_len: usize,
     ) -> bool;
     pub fn LLVMRustLinkerFree<'a>(linker: &'a mut Linker<'a>);
-    #[allow(improper_ctypes)]
     pub fn LLVMRustComputeLTOCacheKey(
         key_out: &RustString,
         mod_id: *const c_char,
@@ -2450,7 +2434,6 @@ unsafe extern "C" {
         pgo_available: bool,
     );
 
-    #[allow(improper_ctypes)]
     pub fn LLVMRustGetMangledName(V: &Value, out: &RustString);
 
     pub fn LLVMRustGetElementTypeArgIndex(CallSite: &Value) -> i32;

compiler/rustc_codegen_llvm/src/llvm/mod.rs

@@ -1,6 +1,5 @@
 #![allow(non_snake_case)]
 
-use std::cell::RefCell;
 use std::ffi::{CStr, CString};
 use std::ops::Deref;
 use std::ptr;
@@ -301,15 +300,11 @@ pub fn set_value_name(value: &Value, name: &[u8]) {
 }
 
 pub fn build_string(f: impl FnOnce(&RustString)) -> Result<String, FromUtf8Error> {
-    let sr = RustString { bytes: RefCell::new(Vec::new()) };
-    f(&sr);
-    String::from_utf8(sr.bytes.into_inner())
+    String::from_utf8(RustString::build_byte_buffer(f))
 }
 
 pub fn build_byte_buffer(f: impl FnOnce(&RustString)) -> Vec<u8> {
-    let sr = RustString { bytes: RefCell::new(Vec::new()) };
-    f(&sr);
-    sr.bytes.into_inner()
+    RustString::build_byte_buffer(f)
 }
 
 pub fn twine_to_string(tr: &Twine) -> String {

compiler/rustc_llvm/llvm-wrapper/LLVMWrapper.h

@@ -104,8 +104,9 @@ typedef struct OpaqueRustString *RustStringRef;
 typedef struct LLVMOpaqueTwine *LLVMTwineRef;
 typedef struct LLVMOpaqueSMDiagnostic *LLVMSMDiagnosticRef;
 
-extern "C" void LLVMRustStringWriteImpl(RustStringRef Str, const char *Ptr,
-                                        size_t Size);
+extern "C" void LLVMRustStringWriteImpl(RustStringRef buf,
+                                        const char *slice_ptr,
+                                        size_t slice_len);
 
 class RawRustStringOstream : public llvm::raw_ostream {
   RustStringRef Str;

compiler/rustc_llvm/llvm-wrapper/PassWrapper.cpp

@@ -1624,5 +1624,6 @@ extern "C" void LLVMRustComputeLTOCacheKey(RustStringRef KeyOut,
                            CfiFunctionDefs, CfiFunctionDecls);
 #endif
 
-  LLVMRustStringWriteImpl(KeyOut, Key.c_str(), Key.size());
+  auto OS = RawRustStringOstream(KeyOut);
+  OS << Key.str();
 }

compiler/rustc_llvm/llvm-wrapper/RustWrapper.cpp

@@ -1510,8 +1510,8 @@ LLVMRustUnpackSMDiagnostic(LLVMSMDiagnosticRef DRef, RustStringRef MessageOut,
   const SourceMgr &LSM = *D.getSourceMgr();
   const MemoryBuffer *LBuf =
       LSM.getMemoryBuffer(LSM.FindBufferContainingLoc(D.getLoc()));
-  LLVMRustStringWriteImpl(BufferOut, LBuf->getBufferStart(),
-                          LBuf->getBufferSize());
+  auto BufferOS = RawRustStringOstream(BufferOut);
+  BufferOS << LBuf->getBuffer();
 
   *LocOut = D.getLoc().getPointer() - LBuf->getBufferStart();
 

compiler/rustc_llvm/src/lib.rs

@@ -2,42 +2,75 @@
 #![allow(internal_features)]
 #![doc(html_root_url = "https://doc.rust-lang.org/nightly/nightly-rustc/")]
 #![doc(rust_logo)]
+#![feature(extern_types)]
 #![feature(rustdoc_internals)]
 #![warn(unreachable_pub)]
 // tidy-alphabetical-end
 
-// NOTE: This crate only exists to allow linking on mingw targets.
-
 use std::cell::RefCell;
-use std::slice;
+use std::{ptr, slice};
 
-use libc::{c_char, size_t};
+use libc::size_t;
 
-#[repr(C)]
-pub struct RustString {
-    pub bytes: RefCell<Vec<u8>>,
+unsafe extern "C" {
+    /// Opaque type that allows C++ code to write bytes to a Rust-side buffer,
+    /// in conjunction with `RawRustStringOstream`. Use this as `&RustString`
+    /// (Rust) and `RustStringRef` (C++) in FFI signatures.
+    pub type RustString;
 }
 
 impl RustString {
-    pub fn len(&self) -> usize {
-        self.bytes.borrow().len()
-    }
-
-    pub fn is_empty(&self) -> bool {
-        self.bytes.borrow().is_empty()
+    pub fn build_byte_buffer(closure: impl FnOnce(&Self)) -> Vec<u8> {
+        let buf = RustStringInner::default();
+        closure(buf.as_opaque());
+        buf.into_inner()
     }
 }
 
-/// Appending to a Rust string -- used by RawRustStringOstream.
+/// Underlying implementation of [`RustString`].
+///
+/// Having two separate types makes it possible to use the opaque [`RustString`]
+/// in FFI signatures without `improper_ctypes` warnings. This is a workaround
+/// for the fact that there is no way to opt out of `improper_ctypes` when
+/// _declaring_ a type (as opposed to using that type).
+#[derive(Default)]
+struct RustStringInner {
+    bytes: RefCell<Vec<u8>>,
+}
+
+impl RustStringInner {
+    fn as_opaque(&self) -> &RustString {
+        let ptr: *const RustStringInner = ptr::from_ref(self);
+        // We can't use `ptr::cast` here because extern types are `!Sized`.
+        let ptr = ptr as *const RustString;
+        unsafe { &*ptr }
+    }
+
+    fn from_opaque(opaque: &RustString) -> &Self {
+        // SAFETY: A valid `&RustString` must have been created via `as_opaque`.
+        let ptr: *const RustString = ptr::from_ref(opaque);
+        let ptr: *const RustStringInner = ptr.cast();
+        unsafe { &*ptr }
+    }
+
+    fn into_inner(self) -> Vec<u8> {
+        self.bytes.into_inner()
+    }
+}
+
+/// Appends the contents of a byte slice to a [`RustString`].
+///
+/// This function is implemented in `rustc_llvm` so that the C++ code in this
+/// crate can link to it directly, without an implied link-time dependency on
+/// `rustc_codegen_llvm`.
 #[unsafe(no_mangle)]
 pub unsafe extern "C" fn LLVMRustStringWriteImpl(
-    sr: &RustString,
-    ptr: *const c_char,
-    size: size_t,
+    buf: &RustString,
+    slice_ptr: *const u8, // Same ABI as `*const c_char`
+    slice_len: size_t,
 ) {
-    let slice = unsafe { slice::from_raw_parts(ptr as *const u8, size) };
-
-    sr.bytes.borrow_mut().extend_from_slice(slice);
+    let slice = unsafe { slice::from_raw_parts(slice_ptr, slice_len) };
+    RustStringInner::from_opaque(buf).bytes.borrow_mut().extend_from_slice(slice);
 }
 
 /// Initialize targets enabled by the build script via `cfg(llvm_component = "...")`.