From 580b4891aa23c0625539bf5ee55270f27af09072 Mon Sep 17 00:00:00 2001 From: Josh Stone Date: Thu, 7 Aug 2025 14:29:00 -0700 Subject: Update the minimum external LLVM to 20 --- compiler/rustc_llvm/llvm-wrapper/PassWrapper.cpp | 140 +++++------------------ 1 file changed, 30 insertions(+), 110 deletions(-) (limited to 'compiler/rustc_llvm/llvm-wrapper/PassWrapper.cpp') diff --git a/compiler/rustc_llvm/llvm-wrapper/PassWrapper.cpp b/compiler/rustc_llvm/llvm-wrapper/PassWrapper.cpp index 91d11ba317a..ab5d5c03e81 100644 --- a/compiler/rustc_llvm/llvm-wrapper/PassWrapper.cpp +++ b/compiler/rustc_llvm/llvm-wrapper/PassWrapper.cpp @@ -344,7 +344,6 @@ extern "C" LLVMTargetMachineRef LLVMRustCreateTargetMachine( Options.EmitStackSizeSection = EmitStackSizeSection; if (ArgsCstrBuff != nullptr) { -#if LLVM_VERSION_GE(20, 0) size_t buffer_offset = 0; assert(ArgsCstrBuff[ArgsCstrBuffLen - 1] == '\0'); auto Arg0 = std::string(ArgsCstrBuff); @@ -362,33 +361,6 @@ extern "C" LLVMTargetMachineRef LLVMRustCreateTargetMachine( OS.flush(); Options.MCOptions.Argv0 = Arg0; Options.MCOptions.CommandlineArgs = CommandlineArgs; -#else - size_t buffer_offset = 0; - assert(ArgsCstrBuff[ArgsCstrBuffLen - 1] == '\0'); - - const size_t arg0_len = std::strlen(ArgsCstrBuff); - char *arg0 = new char[arg0_len + 1]; - memcpy(arg0, ArgsCstrBuff, arg0_len); - arg0[arg0_len] = '\0'; - buffer_offset += arg0_len + 1; - - const size_t num_cmd_arg_strings = std::count( - &ArgsCstrBuff[buffer_offset], &ArgsCstrBuff[ArgsCstrBuffLen], '\0'); - - std::string *cmd_arg_strings = new std::string[num_cmd_arg_strings]; - for (size_t i = 0; i < num_cmd_arg_strings; ++i) { - assert(buffer_offset < ArgsCstrBuffLen); - const size_t len = std::strlen(ArgsCstrBuff + buffer_offset); - cmd_arg_strings[i] = std::string(&ArgsCstrBuff[buffer_offset], len); - buffer_offset += len + 1; - } - - assert(buffer_offset == ArgsCstrBuffLen); - - Options.MCOptions.Argv0 = arg0; - Options.MCOptions.CommandLineArgs = - llvm::ArrayRef(cmd_arg_strings, num_cmd_arg_strings); -#endif } #if LLVM_VERSION_GE(21, 0) @@ -402,12 +374,6 @@ extern "C" LLVMTargetMachineRef LLVMRustCreateTargetMachine( } extern "C" void LLVMRustDisposeTargetMachine(LLVMTargetMachineRef TM) { -#if LLVM_VERSION_LT(20, 0) - MCTargetOptions &MCOptions = unwrap(TM)->Options.MCOptions; - delete[] MCOptions.Argv0; - delete[] MCOptions.CommandLineArgs.data(); -#endif - delete unwrap(TM); } @@ -688,14 +654,9 @@ extern "C" LLVMRustResult LLVMRustOptimize( // the PassBuilder does not create a pipeline. std::vector> PipelineStartEPCallbacks; -#if LLVM_VERSION_GE(20, 0) std::vector> OptimizerLastEPCallbacks; -#else - std::vector> - OptimizerLastEPCallbacks; -#endif if (!IsLinkerPluginLTO && SanitizerOptions && SanitizerOptions->SanitizeCFI && !NoPrepopulatePasses) { @@ -747,12 +708,8 @@ extern "C" LLVMRustResult LLVMRustOptimize( SanitizerOptions->SanitizeDataFlowABIList + SanitizerOptions->SanitizeDataFlowABIListLen); OptimizerLastEPCallbacks.push_back( -#if LLVM_VERSION_GE(20, 0) [ABIListFiles](ModulePassManager &MPM, OptimizationLevel Level, ThinOrFullLTOPhase phase) { -#else - [ABIListFiles](ModulePassManager &MPM, OptimizationLevel Level) { -#endif MPM.addPass(DataFlowSanitizerPass(ABIListFiles)); }); } @@ -763,66 +720,48 @@ extern "C" LLVMRustResult LLVMRustOptimize( SanitizerOptions->SanitizeMemoryRecover, /*CompileKernel=*/false, /*EagerChecks=*/true); - OptimizerLastEPCallbacks.push_back( -#if LLVM_VERSION_GE(20, 0) - [Options](ModulePassManager &MPM, OptimizationLevel Level, - ThinOrFullLTOPhase phase) { -#else - [Options](ModulePassManager &MPM, OptimizationLevel Level) { -#endif - MPM.addPass(MemorySanitizerPass(Options)); - }); + OptimizerLastEPCallbacks.push_back([Options](ModulePassManager &MPM, + OptimizationLevel Level, + ThinOrFullLTOPhase phase) { + MPM.addPass(MemorySanitizerPass(Options)); + }); } if (SanitizerOptions->SanitizeThread) { - OptimizerLastEPCallbacks.push_back( -#if LLVM_VERSION_GE(20, 0) - [](ModulePassManager &MPM, OptimizationLevel Level, - ThinOrFullLTOPhase phase) { -#else - [](ModulePassManager &MPM, OptimizationLevel Level) { -#endif - MPM.addPass(ModuleThreadSanitizerPass()); - MPM.addPass( - createModuleToFunctionPassAdaptor(ThreadSanitizerPass())); - }); + OptimizerLastEPCallbacks.push_back([](ModulePassManager &MPM, + OptimizationLevel Level, + ThinOrFullLTOPhase phase) { + MPM.addPass(ModuleThreadSanitizerPass()); + MPM.addPass(createModuleToFunctionPassAdaptor(ThreadSanitizerPass())); + }); } if (SanitizerOptions->SanitizeAddress || SanitizerOptions->SanitizeKernelAddress) { - OptimizerLastEPCallbacks.push_back( -#if LLVM_VERSION_GE(20, 0) - [SanitizerOptions, TM](ModulePassManager &MPM, - OptimizationLevel Level, - ThinOrFullLTOPhase phase) { -#else - [SanitizerOptions, TM](ModulePassManager &MPM, - OptimizationLevel Level) { -#endif - auto CompileKernel = SanitizerOptions->SanitizeKernelAddress; - AddressSanitizerOptions opts = AddressSanitizerOptions{ - CompileKernel, - SanitizerOptions->SanitizeAddressRecover || - SanitizerOptions->SanitizeKernelAddressRecover, - /*UseAfterScope=*/true, - AsanDetectStackUseAfterReturnMode::Runtime, - }; - MPM.addPass(AddressSanitizerPass( - opts, - /*UseGlobalGC*/ true, - // UseOdrIndicator should be false on windows machines - // https://reviews.llvm.org/D137227 - !TM->getTargetTriple().isOSWindows())); - }); + OptimizerLastEPCallbacks.push_back([SanitizerOptions, + TM](ModulePassManager &MPM, + OptimizationLevel Level, + ThinOrFullLTOPhase phase) { + auto CompileKernel = SanitizerOptions->SanitizeKernelAddress; + AddressSanitizerOptions opts = AddressSanitizerOptions{ + CompileKernel, + SanitizerOptions->SanitizeAddressRecover || + SanitizerOptions->SanitizeKernelAddressRecover, + /*UseAfterScope=*/true, + AsanDetectStackUseAfterReturnMode::Runtime, + }; + MPM.addPass( + AddressSanitizerPass(opts, + /*UseGlobalGC*/ true, + // UseOdrIndicator should be false on windows + // machines https://reviews.llvm.org/D137227 + !TM->getTargetTriple().isOSWindows())); + }); } if (SanitizerOptions->SanitizeHWAddress) { OptimizerLastEPCallbacks.push_back( -#if LLVM_VERSION_GE(20, 0) [SanitizerOptions](ModulePassManager &MPM, OptimizationLevel Level, ThinOrFullLTOPhase phase) { -#else - [SanitizerOptions](ModulePassManager &MPM, OptimizationLevel Level) { -#endif HWAddressSanitizerOptions opts( /*CompileKernel=*/false, SanitizerOptions->SanitizeHWAddressRecover, @@ -904,11 +843,7 @@ extern "C" LLVMRustResult LLVMRustOptimize( for (const auto &C : PipelineStartEPCallbacks) C(MPM, OptLevel); for (const auto &C : OptimizerLastEPCallbacks) -#if LLVM_VERSION_GE(20, 0) C(MPM, OptLevel, ThinOrFullLTOPhase::None); -#else - C(MPM, OptLevel); -#endif } if (ExtraPassesLen) { @@ -1185,11 +1120,7 @@ struct LLVMRustThinLTOData { // Not 100% sure what these are, but they impact what's internalized and // what's inlined across modules, I believe. -#if LLVM_VERSION_GE(20, 0) FunctionImporter::ImportListsTy ImportLists; -#else - DenseMap ImportLists; -#endif DenseMap ExportLists; DenseMap ModuleToDefinedGVSummaries; StringMap> ResolvedODR; @@ -1531,13 +1462,8 @@ extern "C" void LLVMRustComputeLTOCacheKey(RustStringRef KeyOut, const auto &ExportList = Data->ExportLists.lookup(ModId); const auto &ResolvedODR = Data->ResolvedODR.lookup(ModId); const auto &DefinedGlobals = Data->ModuleToDefinedGVSummaries.lookup(ModId); -#if LLVM_VERSION_GE(20, 0) DenseSet CfiFunctionDefs; DenseSet CfiFunctionDecls; -#else - std::set CfiFunctionDefs; - std::set CfiFunctionDecls; -#endif // Based on the 'InProcessThinBackend' constructor in LLVM #if LLVM_VERSION_GE(21, 0) @@ -1556,15 +1482,9 @@ extern "C" void LLVMRustComputeLTOCacheKey(RustStringRef KeyOut, GlobalValue::getGUID(GlobalValue::dropLLVMManglingEscape(Name))); #endif -#if LLVM_VERSION_GE(20, 0) Key = llvm::computeLTOCacheKey(conf, Data->Index, ModId, ImportList, ExportList, ResolvedODR, DefinedGlobals, CfiFunctionDefs, CfiFunctionDecls); -#else - llvm::computeLTOCacheKey(Key, conf, Data->Index, ModId, ImportList, - ExportList, ResolvedODR, DefinedGlobals, - CfiFunctionDefs, CfiFunctionDecls); -#endif auto OS = RawRustStringOstream(KeyOut); OS << Key.str(); -- cgit 1.4.1-3-g733a5 From 8b0a25486040b91ede0120ff5a797517e0973895 Mon Sep 17 00:00:00 2001 From: Zalathar Date: Wed, 17 Sep 2025 13:46:46 +1000 Subject: Move target machine command-line quoting from C++ to Rust --- .../src/back/command_line_args.rs | 37 ++++++++++++++++++++++ .../src/back/command_line_args/tests.rs | 25 +++++++++++++++ compiler/rustc_codegen_llvm/src/back/mod.rs | 1 + .../src/back/owned_target_machine.rs | 14 ++++---- compiler/rustc_codegen_llvm/src/back/write.rs | 30 +++++++----------- compiler/rustc_codegen_llvm/src/llvm/ffi.rs | 6 ++-- compiler/rustc_llvm/llvm-wrapper/PassWrapper.cpp | 28 ++++------------ 7 files changed, 92 insertions(+), 49 deletions(-) create mode 100644 compiler/rustc_codegen_llvm/src/back/command_line_args.rs create mode 100644 compiler/rustc_codegen_llvm/src/back/command_line_args/tests.rs (limited to 'compiler/rustc_llvm/llvm-wrapper/PassWrapper.cpp') diff --git a/compiler/rustc_codegen_llvm/src/back/command_line_args.rs b/compiler/rustc_codegen_llvm/src/back/command_line_args.rs new file mode 100644 index 00000000000..b14713969b3 --- /dev/null +++ b/compiler/rustc_codegen_llvm/src/back/command_line_args.rs @@ -0,0 +1,37 @@ +#[cfg(test)] +mod tests; + +/// Joins command-line arguments into a single space-separated string, quoting +/// and escaping individual arguments as necessary. +/// +/// The result is intended to be informational, for embedding in debug metadata, +/// and might not be properly quoted/escaped for actual command-line use. +pub(crate) fn quote_command_line_args(args: &[String]) -> String { + // Start with a decent-sized buffer, since rustc invocations tend to be long. + let mut buf = String::with_capacity(128); + + for arg in args { + if !buf.is_empty() { + buf.push(' '); + } + + print_arg_quoted(&mut buf, arg); + } + + buf +} + +/// Equivalent to LLVM's `sys::printArg` with quoting always enabled +/// (see llvm/lib/Support/Program.cpp). +fn print_arg_quoted(buf: &mut String, arg: &str) { + buf.reserve(arg.len() + 2); + + buf.push('"'); + for ch in arg.chars() { + if matches!(ch, '"' | '\\' | '$') { + buf.push('\\'); + } + buf.push(ch); + } + buf.push('"'); +} diff --git a/compiler/rustc_codegen_llvm/src/back/command_line_args/tests.rs b/compiler/rustc_codegen_llvm/src/back/command_line_args/tests.rs new file mode 100644 index 00000000000..69641fed3bc --- /dev/null +++ b/compiler/rustc_codegen_llvm/src/back/command_line_args/tests.rs @@ -0,0 +1,25 @@ +#[test] +fn quote_command_line_args() { + use super::quote_command_line_args; + + struct Case<'a> { + args: &'a [&'a str], + expected: &'a str, + } + + let cases = &[ + Case { args: &[], expected: "" }, + Case { args: &["--hello", "world"], expected: r#""--hello" "world""# }, + Case { args: &["--hello world"], expected: r#""--hello world""# }, + Case { + args: &["plain", "$dollar", "spa ce", r"back\slash", r#""quote""#, "plain"], + expected: r#""plain" "\$dollar" "spa ce" "back\\slash" "\"quote\"" "plain""#, + }, + ]; + + for &Case { args, expected } in cases { + let args = args.iter().copied().map(str::to_owned).collect::>(); + let actual = quote_command_line_args(&args); + assert_eq!(actual, expected, "args {args:?}"); + } +} diff --git a/compiler/rustc_codegen_llvm/src/back/mod.rs b/compiler/rustc_codegen_llvm/src/back/mod.rs index 6cb89f80ab8..fe3883e8c73 100644 --- a/compiler/rustc_codegen_llvm/src/back/mod.rs +++ b/compiler/rustc_codegen_llvm/src/back/mod.rs @@ -1,4 +1,5 @@ pub(crate) mod archive; +mod command_line_args; pub(crate) mod lto; pub(crate) mod owned_target_machine; mod profiling; diff --git a/compiler/rustc_codegen_llvm/src/back/owned_target_machine.rs b/compiler/rustc_codegen_llvm/src/back/owned_target_machine.rs index 6d8178320fe..d5228f0e0de 100644 --- a/compiler/rustc_codegen_llvm/src/back/owned_target_machine.rs +++ b/compiler/rustc_codegen_llvm/src/back/owned_target_machine.rs @@ -1,4 +1,3 @@ -use std::assert_matches::assert_matches; use std::ffi::CStr; use std::marker::PhantomData; use std::ptr::NonNull; @@ -39,13 +38,10 @@ impl OwnedTargetMachine { output_obj_file: &CStr, debug_info_compression: &CStr, use_emulated_tls: bool, - args_cstr_buff: &[u8], + argv0: &str, + command_line_args: &str, use_wasm_eh: bool, ) -> Result> { - // The argument list is passed as the concatenation of one or more C strings. - // This implies that there must be a last byte, and it must be 0. - assert_matches!(args_cstr_buff, [.., b'\0'], "the last byte must be a NUL terminator"); - // SAFETY: llvm::LLVMRustCreateTargetMachine copies pointed to data let tm_ptr = unsafe { llvm::LLVMRustCreateTargetMachine( @@ -70,8 +66,10 @@ impl OwnedTargetMachine { output_obj_file.as_ptr(), debug_info_compression.as_ptr(), use_emulated_tls, - args_cstr_buff.as_ptr(), - args_cstr_buff.len(), + argv0.as_ptr(), + argv0.len(), + command_line_args.as_ptr(), + command_line_args.len(), use_wasm_eh, ) }; diff --git a/compiler/rustc_codegen_llvm/src/back/write.rs b/compiler/rustc_codegen_llvm/src/back/write.rs index bda81fbd19e..c4881f0aafc 100644 --- a/compiler/rustc_codegen_llvm/src/back/write.rs +++ b/compiler/rustc_codegen_llvm/src/back/write.rs @@ -31,6 +31,7 @@ use rustc_span::{BytePos, InnerSpan, Pos, SpanData, SyntaxContext, sym}; use rustc_target::spec::{CodeModel, FloatAbi, RelocModel, SanitizerSet, SplitDebuginfo, TlsModel}; use tracing::{debug, trace}; +use crate::back::command_line_args::quote_command_line_args; use crate::back::lto::ThinBuffer; use crate::back::owned_target_machine::OwnedTargetMachine; use crate::back::profiling::{ @@ -249,23 +250,15 @@ pub(crate) fn target_machine_factory( let use_emulated_tls = matches!(sess.tls_model(), TlsModel::Emulated); - // copy the exe path, followed by path all into one buffer - // null terminating them so we can use them as null terminated strings - let args_cstr_buff = { - let mut args_cstr_buff: Vec = Vec::new(); - let exe_path = std::env::current_exe().unwrap_or_default(); - let exe_path_str = exe_path.into_os_string().into_string().unwrap_or_default(); - - args_cstr_buff.extend_from_slice(exe_path_str.as_bytes()); - args_cstr_buff.push(0); - - for arg in sess.expanded_args.iter() { - args_cstr_buff.extend_from_slice(arg.as_bytes()); - args_cstr_buff.push(0); - } - - args_cstr_buff - }; + // Command-line information to be included in the target machine. + // This seems to only be used for embedding in PDB debuginfo files. + // FIXME(Zalathar): Maybe skip this for non-PDB targets? + let argv0 = std::env::current_exe() + .unwrap_or_default() + .into_os_string() + .into_string() + .unwrap_or_default(); + let command_line_args = quote_command_line_args(&sess.expanded_args); let debuginfo_compression = sess.opts.debuginfo_compression.to_string(); match sess.opts.debuginfo_compression { @@ -323,7 +316,8 @@ pub(crate) fn target_machine_factory( &output_obj_file, &debuginfo_compression, use_emulated_tls, - &args_cstr_buff, + &argv0, + &command_line_args, use_wasm_eh, ) }) diff --git a/compiler/rustc_codegen_llvm/src/llvm/ffi.rs b/compiler/rustc_codegen_llvm/src/llvm/ffi.rs index 71d8b7d25fe..c1b5649a582 100644 --- a/compiler/rustc_codegen_llvm/src/llvm/ffi.rs +++ b/compiler/rustc_codegen_llvm/src/llvm/ffi.rs @@ -2488,8 +2488,10 @@ unsafe extern "C" { OutputObjFile: *const c_char, DebugInfoCompression: *const c_char, UseEmulatedTls: bool, - ArgsCstrBuff: *const c_uchar, // See "PTR_LEN_STR". - ArgsCstrBuffLen: usize, + Argv0: *const c_uchar, // See "PTR_LEN_STR". + Argv0Len: size_t, + CommandLineArgs: *const c_uchar, // See "PTR_LEN_STR". + CommandLineArgsLen: size_t, UseWasmEH: bool, ) -> *mut TargetMachine; diff --git a/compiler/rustc_llvm/llvm-wrapper/PassWrapper.cpp b/compiler/rustc_llvm/llvm-wrapper/PassWrapper.cpp index ab5d5c03e81..7518b40799b 100644 --- a/compiler/rustc_llvm/llvm-wrapper/PassWrapper.cpp +++ b/compiler/rustc_llvm/llvm-wrapper/PassWrapper.cpp @@ -271,8 +271,9 @@ extern "C" LLVMTargetMachineRef LLVMRustCreateTargetMachine( bool TrapUnreachable, bool Singlethread, bool VerboseAsm, bool EmitStackSizeSection, bool RelaxELFRelocations, bool UseInitArray, const char *SplitDwarfFile, const char *OutputObjFile, - const char *DebugInfoCompression, bool UseEmulatedTls, - const char *ArgsCstrBuff, size_t ArgsCstrBuffLen, bool UseWasmEH) { + const char *DebugInfoCompression, bool UseEmulatedTls, const char *Argv0, + size_t Argv0Len, const char *CommandLineArgs, size_t CommandLineArgsLen, + bool UseWasmEH) { auto OptLevel = fromRust(RustOptLevel); auto RM = fromRust(RustReloc); @@ -343,25 +344,10 @@ extern "C" LLVMTargetMachineRef LLVMRustCreateTargetMachine( Options.EmitStackSizeSection = EmitStackSizeSection; - if (ArgsCstrBuff != nullptr) { - size_t buffer_offset = 0; - assert(ArgsCstrBuff[ArgsCstrBuffLen - 1] == '\0'); - auto Arg0 = std::string(ArgsCstrBuff); - buffer_offset = Arg0.size() + 1; - - std::string CommandlineArgs; - raw_string_ostream OS(CommandlineArgs); - ListSeparator LS(" "); - for (StringRef Arg : split(StringRef(ArgsCstrBuff + buffer_offset, - ArgsCstrBuffLen - buffer_offset), - '\0')) { - OS << LS; - sys::printArg(OS, Arg, /*Quote=*/true); - } - OS.flush(); - Options.MCOptions.Argv0 = Arg0; - Options.MCOptions.CommandlineArgs = CommandlineArgs; - } + if (Argv0 != nullptr) + Options.MCOptions.Argv0 = {Argv0, Argv0Len}; + if (CommandLineArgs != nullptr) + Options.MCOptions.CommandlineArgs = {CommandLineArgs, CommandLineArgsLen}; #if LLVM_VERSION_GE(21, 0) TargetMachine *TM = TheTarget->createTargetMachine(Trip, CPU, Feature, -- cgit 1.4.1-3-g733a5