[clang] AMDGPU: Remove "gws" from the “read-only” target feature list (PR #148141)

@@ -816,12 +816,12 @@ kernel void test_target_features_kernel(global int *i) { // NOCPU: attributes #[[ATTR10]] = { convergent nounwind } //. // GFX900: attributes #[[ATTR0:[0-9]+]] = { "objc_arc_inert" } -// GFX900: attributes #[[ATTR1]] = { convergent norecurse nounwind "den

[clang] AMDGPU: Remove "gws" from the “read-only” target feature list (PR #148141)

@@ -1,6 +0,0 @@ -// RUN: %clang_cc1 -triple amdgcn -target-feature +gws -o /dev/null %s 2>&1 \ -// RUN: | FileCheck --check-prefix=GWS %s - -// GWS: warning: feature flag '+gws' is ignored since the feature is read only [-Winvalid-command-line-argument] arsenm

[libclc] [libclc] Change symlinks to copies on Windows (PR #147759)

https://github.com/arsenm approved this pull request. https://github.com/llvm/llvm-project/pull/147759 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] [AMDGPU][Clang] Support bfloat16 arithmetic. (PR #147541)

https://github.com/arsenm approved this pull request. I thought this was enabled a long time ago? https://github.com/llvm/llvm-project/pull/147541 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/

[clang] [Sema] Remove an unnecessary cast (NFC) (PR #147546)

https://github.com/arsenm approved this pull request. https://github.com/llvm/llvm-project/pull/147546 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] [AMDGPU][Clang] Support bfloat16 arithmetic for GFX950 and GFX1250. (PR #147541)

https://github.com/arsenm edited https://github.com/llvm/llvm-project/pull/147541 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] [AMDGPU][Clang] Support bfloat16 arithmetic for GFX950 and GFX1250. (PR #147541)

@@ -451,6 +451,8 @@ class LLVM_LIBRARY_VISIBILITY AMDGPUTargetInfo final : public TargetInfo { CUMode = false; else if (F == "+image-insts") HasImage = true; + else if (F == "+gfx950-insts" || F == "+gfx1250-insts") +HasFullBFloat16 = true; -

[clang] [llvm] [PowerPC][clang] Fix triple constructor ambiguity causing "unknown" target triple on AIX (PR #147488)

https://github.com/arsenm edited https://github.com/llvm/llvm-project/pull/147488 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] [llvm] [PowerPC][clang] Fix triple constructor ambiguity causing "unknown" target triple on AIX (PR #147488)

https://github.com/arsenm approved this pull request. lgtm with test nits https://github.com/llvm/llvm-project/pull/147488 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] [llvm] [PowerPC][clang] Fix triple constructor ambiguity causing "unknown" target triple on AIX (PR #147488)

@@ -0,0 +1,15 @@ +// Test for the Triple constructor ambiguity fix on AIX +// This test verifies that the default target triple is correctly resolved +// and doesn't fall back to "unknown" due to constructor ambiguity. + +// REQUIRES: system-aix +// RUN: %clang -v %s -c 2>&1 | Fil

[clang] [llvm] [PowerPC][clang] Fix triple constructor ambiguity causing "unknown" target triple on AIX (PR #147488)

@@ -0,0 +1,15 @@ +// Test for the Triple constructor ambiguity fix on AIX +// This test verifies that the default target triple is correctly resolved +// and doesn't fall back to "unknown" due to constructor ambiguity. + +// REQUIRES: system-aix +// RUN: %clang -v %s -c 2>&1 | Fil

[clang] [llvm] [AMDGPU] Add support for `v_tanh_bf16` on gfx1250 (PR #147425)

@@ -13658,6 +13658,7 @@ bool SITargetLowering::isCanonicalized(Register Reg, const MachineFunction &MF, case Intrinsic::amdgcn_frexp_mant: case Intrinsic::amdgcn_fdot2: case Intrinsic::amdgcn_trig_preop: +case Intrinsic::amdgcn_tanh: arsenm wro

[libclc] [libclc] Declare workitem built-ins in clc, move ptx-nvidiacl workitem built-ins into clc (PR #144333)

@@ -0,0 +1,22 @@ +//===--===// +// +// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions. +// See https://llvm.org/LICENSE.txt for license information. +// SPDX-License-Identifier: Apac

[clang] [llvm] [AMDGPU] Add support for `v_tanh_bf16` on gfx1250 (PR #147425)

@@ -700,7 +700,8 @@ GCNTTIImpl::instCombineIntrinsic(InstCombiner &IC, IntrinsicInst &II) const { break; } case Intrinsic::amdgcn_sqrt: - case Intrinsic::amdgcn_rsq: { + case Intrinsic::amdgcn_rsq: + case Intrinsic::amdgcn_tanh: { arsenm wrote: Thi

[clang] [llvm] [AMDGPU] Add support for `v_tanh_bf16` on gfx1250 (PR #147425)

@@ -13658,6 +13658,7 @@ bool SITargetLowering::isCanonicalized(Register Reg, const MachineFunction &MF, case Intrinsic::amdgcn_frexp_mant: case Intrinsic::amdgcn_fdot2: case Intrinsic::amdgcn_trig_preop: +case Intrinsic::amdgcn_tanh: arsenm wro

[libclc] [libclc] Declare workitem built-ins in clc, move ptx-nvidiacl workitem built-ins into clc (PR #144333)

@@ -0,0 +1,22 @@ +//===--===// +// +// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions. +// See https://llvm.org/LICENSE.txt for license information. +// SPDX-License-Identifier: Apac

[libclc] [libclc] Declare workitem built-ins in clc, move ptx-nvidiacl workitem built-ins into clc (PR #144333)

@@ -0,0 +1,22 @@ +//===--===// +// +// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions. +// See https://llvm.org/LICENSE.txt for license information. +// SPDX-License-Identifier: Apac

[clang] [libclc] [llvm] [libclc] Add initial LIT tests (PR #87989)

@@ -0,0 +1,32 @@ +// NOTE: Assertions have been autogenerated by utils/update_cc_test_checks.py UTC_ARGS: --version 5 +//===--===// +// +// Part of the LLVM Project, under the Apache License v2.0 with LLVM Excep

[clang] [llvm] [AMDGPU] Add alignment attr & propagate alignment through make.buffer.rsrc inst (PR #145278)

https://github.com/arsenm commented: This is still doing the target intrinsic instead of handling ptrmask first? https://github.com/llvm/llvm-project/pull/145278 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/ma

[clang] [libclc] [llvm] [libclc] Add initial LIT tests (PR #87989)

@@ -215,6 +215,9 @@ void addOpenMPDeviceRTL(const Driver &D, const llvm::opt::ArgList &DriverArgs, StringRef BitcodeSuffix, const llvm::Triple &Triple, const ToolChain &HostTC); +void addOpenCLBuiltinsLib(const Driver &D, const

[clang] [libclc] [clang] Add the ability to link libclc OpenCL libraries (PR #146503)

@@ -2942,6 +2942,65 @@ void tools::addHIPRuntimeLibArgs(const ToolChain &TC, Compilation &C, } } +void tools::addOpenCLBuiltinsLib(const Driver &D, + const llvm::opt::ArgList &DriverArgs, + llvm::opt::ArgString

[clang] [libclc] [clang] Add the ability to link libclc OpenCL libraries (PR #146503)

@@ -2942,6 +2942,65 @@ void tools::addHIPRuntimeLibArgs(const ToolChain &TC, Compilation &C, } } +void tools::addOpenCLBuiltinsLib(const Driver &D, + const llvm::opt::ArgList &DriverArgs, + llvm::opt::ArgString

[clang] [libclc] [clang] Add the ability to link libclc OpenCL libraries (PR #146503)

@@ -2942,6 +2942,65 @@ void tools::addHIPRuntimeLibArgs(const ToolChain &TC, Compilation &C, } } +void tools::addOpenCLBuiltinsLib(const Driver &D, + const llvm::opt::ArgList &DriverArgs, + llvm::opt::ArgString

[clang] [Sema] Remove an unnecessary cast (NFC) (PR #147203)

https://github.com/arsenm approved this pull request. https://github.com/llvm/llvm-project/pull/147203 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] [Serialization] Remove an unnecessarycast (NFC) (PR #147204)

https://github.com/arsenm approved this pull request. https://github.com/llvm/llvm-project/pull/147204 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] [Clang][OpenCL] Declare cl_amd_media_ops/cl_amd_media_ops2 builtins with -fdeclare-opencl-builtins (PR #143507)

Juan Manuel Martinez =?utf-8?q?Caama=C3=B1o?= , Juan Manuel Martinez =?utf-8?q?Caama=C3=B1o?= , Juan Manuel Martinez =?utf-8?q?Caama=C3=B1o?= , Juan Manuel Martinez =?utf-8?q?Caama=C3=B1o?= , Juan Manuel Martinez =?utf-8?q?Caama=C3=B1o?= Message-ID: In-Reply-To: https://github.com/arsenm approv

[clang] clang: Forward exception_model flag for bitcode inputs (PR #146342)

arsenm wrote: > This broke building C++ modules for mingw targets. Repro: > > `empty.cppm`: > > ```c++ > export module empty; > ``` > > ``` > $ clang -target x86_64-windows-gnu -x c++-module empty.cppm -c -o > empty.cppm.obj -std=gnu++23 > error: invalid value 'seh' in '-exception-model=seh'

[clang] clang: Fix parsing of seh exception model (PR #146643)

https://github.com/arsenm ready_for_review https://github.com/llvm/llvm-project/pull/146643 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] clang: Fix parsing of seh exception model (PR #146643)

arsenm wrote: * **#146643** https://app.graphite.dev/github/pr/llvm/llvm-project/146643?utm_source=stack-comment-icon"; target="_blank">https://static.graphite.dev/graphite-32x32-black.png"; alt="Graphite" width="10px" height="10px"/> 👈 https://app.graphite.dev/github/pr/llvm/llvm-project/14664

[clang] clang: Fix parsing of seh exception model (PR #146643)

https://github.com/arsenm created https://github.com/llvm/llvm-project/pull/146643 Fixes regression reported https://github.com/llvm/llvm-project/pull/146342#issuecomment-3026600152 The test could probably be better. I'm not sure what special is happening with the module compile, but I can't

[clang] [llvm] WebAssembly: Stop changing MCAsmInfo's ExceptionsType based on flags (PR #146343)

@@ -1,4 +1,4 @@ -# RUN: llc -mtriple=wasm32-unknown-unknown -wasm-use-legacy-eh -exception-model=wasm -mattr=+exception-handling -run-pass wasm-cfg-stackify %s -o - | FileCheck %s +# RUN: llc -mtriple=wasm32-unknown-unknown -wasm-use-legacy-eh -wasm-enable-sjlj -exception-model

[clang] [llvm] AMDGPU: Implement tensor load and store instructions for gfx1250 (PR #146636)

@@ -4280,6 +4281,20 @@ bool AMDGPUAsmParser::validateMIMGD16(const MCInst &Inst) { return true; } +bool AMDGPUAsmParser::validateTensorR128(const MCInst &Inst) { + const unsigned Opc = Inst.getOpcode(); + const MCInstrDesc &Desc = MII.get(Opc); + + if ((Desc.TSFlags & SI

[clang] [llvm] AMDGPU: Implement tensor load and store instructions for gfx1250 (PR #146636)

@@ -621,6 +621,32 @@ Value *CodeGenFunction::EmitAMDGPUBuiltinExpr(unsigned BuiltinID, llvm::Function *F = CGM.getIntrinsic(IID, {LoadTy}); return Builder.CreateCall(F, {Addr}); } + case AMDGPU::BI__builtin_amdgcn_tensor_load_to_lds: + case AMDGPU::BI__builtin_amdg

[clang] [llvm] AMDGPU: Implement tensor load and store instructions for gfx1250 (PR #146636)

https://github.com/arsenm edited https://github.com/llvm/llvm-project/pull/146636 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] [llvm] AMDGPU: Implement tensor load and store instructions for gfx1250 (PR #146636)

https://github.com/arsenm commented: Fix clang-format error https://github.com/llvm/llvm-project/pull/146636 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] [llvm] WebAssembly: Move validation of EH flags to TargetMachine construct time (PR #146634)

https://github.com/arsenm ready_for_review https://github.com/llvm/llvm-project/pull/146634 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] [llvm] WebAssembly: Move validation of EH flags to TargetMachine construct time (PR #146634)

arsenm wrote: * **#146634** https://app.graphite.dev/github/pr/llvm/llvm-project/146634?utm_source=stack-comment-icon"; target="_blank">https://static.graphite.dev/graphite-32x32-black.png"; alt="Graphite" width="10px" height="10px"/> 👈 https://app.graphite.dev/github/pr/llvm/llvm-project/14663

[clang] [llvm] WebAssembly: Move validation of EH flags to TargetMachine construct time (PR #146634)

https://github.com/arsenm created https://github.com/llvm/llvm-project/pull/146634 This was enforcing some rules about the combinations of flags at a random point in the IR pass pipeline configuration. This is a module property that should be handled at TargetMachine construction time at the lat

[clang] [llvm] WebAssembly: Stop changing MCAsmInfo's ExceptionsType based on flags (PR #146343)

arsenm wrote: > And it was a little confusing that the PR description argues we should remove > `-wasm-enable-eh/sjlj` options, which disagree, when the code seems to fix > something else. Those flags control behaviors, not exception models, so we > still need them. For example, someone might

[clang] [llvm] WebAssembly: Stop changing MCAsmInfo's ExceptionsType based on flags (PR #146343)

@@ -1,4 +1,4 @@ -# RUN: llc -mtriple=wasm32-unknown-unknown -wasm-use-legacy-eh -exception-model=wasm -mattr=+exception-handling -run-pass wasm-cfg-stackify %s -o - | FileCheck %s +# RUN: llc -mtriple=wasm32-unknown-unknown -wasm-use-legacy-eh -wasm-enable-sjlj -exception-model

[clang-tools-extra] [clangd] Drop const from a return type (NFC) (PR #146623)

https://github.com/arsenm approved this pull request. https://github.com/llvm/llvm-project/pull/146623 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] [Sema] Remove an unnecessary cast (NFC) (PR #146622)

https://github.com/arsenm approved this pull request. https://github.com/llvm/llvm-project/pull/146622 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] [WIP][AMDGPU] Support for type inferring image load/store builtins for AMDGPU (PR #140210)

@@ -635,5 +635,66 @@ TARGET_BUILTIN(__builtin_amdgcn_bitop3_b16, "IUi", "nc", "bitop3-insts") TARGET_BUILTIN(__builtin_amdgcn_cvt_sr_bf16_f32, "V2yV2yfUiIb", "nc", "f32-to-f16bf16-cvt-sr-insts") TARGET_BUILTIN(__builtin_amdgcn_cvt_sr_f16_f32, "V2hV2hfUiIb", "nc", "f32-to-

[clang] [WIP][AMDGPU] Support for type inferring image load/store builtins for AMDGPU (PR #140210)

https://github.com/arsenm edited https://github.com/llvm/llvm-project/pull/140210 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] [WIP][AMDGPU] Support for type inferring image load/store builtins for AMDGPU (PR #140210)

@@ -0,0 +1,132 @@ +// RUN: %clang_cc1 -triple amdgcn-- -target-cpu gfx90a -S -verify=expected -o - %s arsenm wrote: Needs another semi test for the rejected target case https://github.com/llvm/llvm-project/pull/140210 __

[clang] [WIP][AMDGPU] Support for type inferring image load/store builtins for AMDGPU (PR #140210)

@@ -635,5 +635,66 @@ TARGET_BUILTIN(__builtin_amdgcn_bitop3_b16, "IUi", "nc", "bitop3-insts") TARGET_BUILTIN(__builtin_amdgcn_cvt_sr_bf16_f32, "V2yV2yfUiIb", "nc", "f32-to-f16bf16-cvt-sr-insts") TARGET_BUILTIN(__builtin_amdgcn_cvt_sr_f16_f32, "V2hV2hfUiIb", "nc", "f32-to-

[clang] [WIP][AMDGPU] Support for type inferring image load/store builtins for AMDGPU (PR #140210)

https://github.com/arsenm edited https://github.com/llvm/llvm-project/pull/140210 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] [WIP][AMDGPU] Support for type inferring image load/store builtins for AMDGPU (PR #140210)

https://github.com/arsenm commented: Title mentions type inferring but nothing is type inferring about these https://github.com/llvm/llvm-project/pull/140210 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailma

[clang] clang: Forward exception_model flag for bitcode inputs (PR #146342)

@@ -3679,6 +3679,22 @@ static StringRef GetInputKindName(InputKind IK) { llvm_unreachable("unknown input language"); } +static StringRef getExceptionHandlingName(unsigned EHK) { + switch (static_cast(EHK)) { + case LangOptions::ExceptionHandlingKind::None: + default: +

[clang] clang: Forward exception_model flag for bitcode inputs (PR #146342)

@@ -3694,6 +3710,10 @@ void CompilerInvocationBase::GenerateLangArgs(const LangOptions &Opts, GenerateArg(Consumer, OPT_pic_is_pie); for (StringRef Sanitizer : serializeSanitizerKinds(Opts.Sanitize)) GenerateArg(Consumer, OPT_fsanitize_EQ, Sanitizer); +if (

[clang] clang: Forward exception_model flag for bitcode inputs (PR #146342)

@@ -4002,6 +4022,24 @@ bool CompilerInvocation::ParseLangArgs(LangOptions &Opts, ArgList &Args, parseSanitizerKinds("-fsanitize=", Args.getAllArgValues(OPT_fsanitize_EQ), Diags, Opts.Sanitize); +if (const Arg *A = Args.getLastArg(options::OPT_e

[clang] clang: Forward exception_model flag for bitcode inputs (PR #146342)

https://github.com/arsenm ready_for_review https://github.com/llvm/llvm-project/pull/146342 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] clang: Forward exception_model flag for bitcode inputs (PR #146342)

arsenm wrote: * **#146343** https://app.graphite.dev/github/pr/llvm/llvm-project/146343?utm_source=stack-comment-icon"; target="_blank">https://static.graphite.dev/graphite-32x32-black.png"; alt="Graphite" width="10px" height="10px"/> * **#146342** https://app.graphite.dev/github/pr/llvm/llvm-p

[clang] clang: Forward exception_model flag for bitcode inputs (PR #146342)

https://github.com/arsenm created https://github.com/llvm/llvm-project/pull/146342 This will enable removal of a hack from the wasm backend in a future change. This feels unnecessarily clunky. I would assume something was automatically parsing this and propagating it in the C++ case, but I can'

[clang] [llvm] [AMDGPU] Add support for `v_cvt_f16_fp8` on gfx1250 (PR #146302)

https://github.com/arsenm approved this pull request. https://github.com/llvm/llvm-project/pull/146302 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] [llvm] [AMDGPU] Add support for `v_cvt_f16_fp8` on gfx1250 (PR #146302)

arsenm wrote: > > > Co-authored-by: Shilei Tian > > > [i...@tianshilei.me](mailto:i...@tianshilei.me) > > > > > > Co authored by yourself? > > Yes, because if you check the commit, the author is @rampitec. After merging on GitHub, the PR author is listed as the author. You should swap these

[clang] [llvm] [AMDGPU] Add support for `v_cvt_f16_fp8` on gfx1250 (PR #146302)

arsenm wrote: > Co-authored-by: Shilei Tian [i...@tianshilei.me](mailto:i...@tianshilei.me) Co authored by yourself? https://github.com/llvm/llvm-project/pull/146302 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-b

[clang] [llvm] Fix Windows EH IP2State tables (remove +1 bias) (PR #144745)

@@ -2660,6 +2644,122 @@ void X86AsmPrinter::emitCallInstruction(const llvm::MCInst &MCI) { OutStreamer->emitInstruction(MCI, getSubtargetInfo()); } +// Checks whether a NOP is required after a CALL and inserts the NOP, if +// necessary. +void X86AsmPrinter::maybeEmitNopAfte

[clang] [llvm] AMDGPU: support s_monitor_sleep on gfx1250 (PR #146293)

https://github.com/arsenm approved this pull request. https://github.com/llvm/llvm-project/pull/146293 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] [clang] Remove unused includes (NFC) (PR #146254)

https://github.com/arsenm approved this pull request. https://github.com/llvm/llvm-project/pull/146254 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] [llvm] Fix Windows EH IP2State tables (remove +1 bias) (PR #144745)

@@ -2660,6 +2644,122 @@ void X86AsmPrinter::emitCallInstruction(const llvm::MCInst &MCI) { OutStreamer->emitInstruction(MCI, getSubtargetInfo()); } +// Checks whether a NOP is required after a CALL and inserts the NOP, if +// necessary. +void X86AsmPrinter::maybeEmitNopAfte

[clang] [llvm] [AMDGPU][Clang] Allow amdgpu-waves-per-eu attribute to lower target occupancy range (PR #138284)

@@ -2519,6 +2519,29 @@ void Verifier::verifyFunctionAttrs(FunctionType *FT, AttributeList Attrs, CheckFailed("invalid value for 'denormal-fp-math-f32' attribute: " + S, V); } + + if (TT.isAMDGPU()) { +if (auto A = Attrs.getFnAttr("amdgpu-waves-pe

[clang] [llvm] [AMDGPU][Clang] Allow amdgpu-waves-per-eu attribute to lower target occupancy range (PR #138284)

@@ -0,0 +1,40 @@ +; RUN: not llvm-as -disable-output %s 2>&1 | FileCheck %s + +target triple = "amdgcn-amd-amdhsa" + +define void @valid_amdgpu_waves_per_eu_range() "amdgpu-waves-per-eu"="2,4" { + ret void +} arsenm wrote: valid cases are tested in test/Assemble

[clang] [llvm] [AMDGPU][Clang] Allow amdgpu-waves-per-eu attribute to lower target occupancy range (PR #138284)

@@ -0,0 +1,40 @@ +; RUN: not llvm-as -disable-output %s 2>&1 | FileCheck %s + +target triple = "amdgcn-amd-amdhsa" + +define void @valid_amdgpu_waves_per_eu_range() "amdgpu-waves-per-eu"="2,4" { + ret void +} + +define void @valid_amdgpu_waves_per_eu_min_only() "amdgpu-waves-per-

[clang] [llvm] [AMDGPU][Clang] Allow amdgpu-waves-per-eu attribute to lower target occupancy range (PR #138284)

@@ -2519,6 +2519,29 @@ void Verifier::verifyFunctionAttrs(FunctionType *FT, AttributeList Attrs, CheckFailed("invalid value for 'denormal-fp-math-f32' attribute: " + S, V); } + + if (TT.isAMDGPU()) { +if (auto A = Attrs.getFnAttr("amdgpu-waves-pe

[clang] [llvm] [AMDGPU][Clang] Allow amdgpu-waves-per-eu attribute to lower target occupancy range (PR #138284)

@@ -46,7 +46,6 @@ __attribute__((amdgpu_num_sgpr(4294967296))) kernel void kernel_num_sgpr_L() {} __attribute__((amdgpu_num_vgpr(4294967296))) kernel void kernel_num_vgpr_L() {} // expected-error {{integer constant expression evaluates to value 4294967296 that cannot be repres

[clang] [clang] Use a new constructor of ArrayRef (NFC) (PR #146007)

https://github.com/arsenm approved this pull request. https://github.com/llvm/llvm-project/pull/146007 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] [llvm] AMDGPU: Add MC layer support for load transpose instructions for gfx1250 (PR #146024)

https://github.com/arsenm approved this pull request. https://github.com/llvm/llvm-project/pull/146024 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] [ASTMatchers] Migrate away from ArrayRef(std::nullopt) (NFC) (PR #145840)

https://github.com/arsenm approved this pull request. https://github.com/llvm/llvm-project/pull/145840 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang-tools-extra] [clangd] Migrate away from std::nullopt (NFC) (PR #145841)

https://github.com/arsenm approved this pull request. https://github.com/llvm/llvm-project/pull/145841 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] [clang] Use llvm::is_contained instead of llvm::all_of (NFC) (PR #145843)

https://github.com/arsenm approved this pull request. https://github.com/llvm/llvm-project/pull/145843 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] [llvm] Triple: Forward declare Twine and remove include (PR #145685)

https://github.com/arsenm closed https://github.com/llvm/llvm-project/pull/145685 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] [llvm] Triple: Forward declare Twine and remove include (PR #145685)

@@ -349,7 +350,12 @@ class Triple { /// triple fields unknown. Triple() = default; + LLVM_ABI explicit Triple(std::string &&Str); + explicit Triple(StringRef Str) : Triple(Str.str()) {} + explicit Triple(const char *Str) : Triple(std::string(Str)) {} + explicit Triple(

[clang] [Clang] Determine offloading architectures at Toolchain creation (PR #145799)

https://github.com/arsenm approved this pull request. https://github.com/llvm/llvm-project/pull/145799 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] [llvm] Triple: Forward declare Twine and remove include (PR #145685)

https://github.com/arsenm ready_for_review https://github.com/llvm/llvm-project/pull/145685 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] [llvm] Triple: Forward declare Twine and remove include (PR #145685)

arsenm wrote: * **#145685** https://app.graphite.dev/github/pr/llvm/llvm-project/145685?utm_source=stack-comment-icon"; target="_blank">https://static.graphite.dev/graphite-32x32-black.png"; alt="Graphite" width="10px" height="10px"/> 👈 https://app.graphite.dev/github/pr/llvm/llvm-project/14568

[clang] [llvm] Triple: Forward declare Twine and remove include (PR #145685)

https://github.com/arsenm created https://github.com/llvm/llvm-project/pull/145685 It turns out real Twine usage is scarce, and seems to only be used from clang. Add a few overloads for the common cases, and introduce a string&& case as the base case. >From a0b41a898e9b220e1b2754967d88cdd667d64

[clang] [llvm] [AMDGPU][clang][CodeGen][opt] Add late-resolved feature identifying predicates (PR #134016)

@@ -0,0 +1,157 @@ +//===- AMDGPUExpandFeaturePredicates.cpp - Feature Predicate Expander Pass ===// +// +// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions. +// See https://llvm.org/LICENSE.txt for license information. +// SPDX-License-Identifier: Ap

[clang] [llvm] [AMDGPU][clang][CodeGen][opt] Add late-resolved feature identifying predicates (PR #134016)

@@ -0,0 +1,157 @@ +//===- AMDGPUExpandFeaturePredicates.cpp - Feature Predicate Expander Pass ===// +// +// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions. +// See https://llvm.org/LICENSE.txt for license information. +// SPDX-License-Identifier: Ap

[clang] [llvm] [ValueTracking] Return true for AddrSpaceCast in canCreateUndefOrPoison (PR #144686)

@@ -0,0 +1,25 @@ +; NOTE: Assertions have been autogenerated by utils/update_test_checks.py UTC_ARGS: --version 5 +; RUN: opt -S -mtriple=amdgcn-amd-amdhsa -passes=instcombine %s | FileCheck %s + +; Check that `select B, true, C` isn't optimized to `or B, C`, +; because the inval

[clang] [clang] Migrate away from ArrayRef(std::nullopt) (NFC) (PR #144982)

https://github.com/arsenm approved this pull request. https://github.com/llvm/llvm-project/pull/144982 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] [llvm] [AMDGPU] Initial support for gfx1250 target. (PR #144965)

https://github.com/arsenm approved this pull request. https://github.com/llvm/llvm-project/pull/144965 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[libclc] [libclc] Declare workitem built-ins in clc, move ptx-nvidiacl workitem built-ins into clc (PR #144333)

arsenm wrote: > Moving them into clc should resolve the concern, but I don't know where is > destination directory in clc. I don't know how to map `opencl/lib/amdgcn` and > `opencl/lib/r600` to `clc/lib/amdgcn` and `clc/lib/amdgpu`. Do you have > suggestions about the mapping? r600 and amdgcn

[clang] [Clang] Add standalone AMDGPU SPIR-V toolchain (PR #144576)

@@ -417,3 +417,15 @@ void HIPAMDToolChain::checkTargetID( getDriver().Diag(clang::diag::err_drv_bad_target_id) << *PTID.OptionalTargetID; } + +SPIRVAMDToolChain::SPIRVAMDToolChain(const Driver &D, + const llvm::Triple &Triple, +

[clang] [HIP] Emit the CUID value in the module with the new driver (PR #144570)

arsenm wrote: > `llvm.compiler.used` global uses AS(0) which makes SPIR-V unhappy, but with > this global it's AS(4) which makes it happy. Either way, this should be fixed. llvm.used and llvm.compiler.used should universally use addrspace(0) and SPIRV should be fixed to not break on this

[clang] [llvm] [SROA] Vector promote some memsets (PR #133301)

arsenm wrote: > and that's why you want to have single element loads that get inserted into a > vector and then perform sub-vector extracts on it? We care about eliminate stack usage at any cost. Big load are always better, even if not naturally aligned. When we do care about alignment, we onl

[clang] [llvm] [SROA] Vector promote some memsets (PR #133301)

arsenm wrote: > We do support unaligned access IIUC, and it is actually required by HSA ABI. > CC @arsenm Unaligned access has been supported since almost always. There's just a mode control for it, you can disable it in the driver though https://github.com/llvm/llvm-project/pull/133301 __

[clang] [clang] Use StringRef in range-based for loops (NFC) (PR #144242)

https://github.com/arsenm approved this pull request. https://github.com/llvm/llvm-project/pull/144242 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang-tools-extra] [modularize] Use range-based for loops (NFC) (PR #144244)

https://github.com/arsenm approved this pull request. https://github.com/llvm/llvm-project/pull/144244 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] [TableGen] Use range-based for loops (NFC) (PR #144250)

https://github.com/arsenm approved this pull request. https://github.com/llvm/llvm-project/pull/144250 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] [Clang][OpenCL] Declare cl_amd_media_ops/cl_amd_media_ops2 builtins with -fdeclare-opencl-builtins (PR #143507)

Juan Manuel Martinez =?utf-8?q?Caamaño?= , Juan Manuel Martinez =?utf-8?q?Caamaño?= , Juan Manuel Martinez =?utf-8?q?Caamaño?= Message-ID: In-Reply-To: @@ -0,0 +1,98 @@ +// RUN: %clang_cc1 %s -triple amdgcn-unknown-unknown -verify -pedantic -Wconversion -Werror -fsyntax-only -

[clang] [Clang][OpenCL] Declare cl_amd_media_ops/cl_amd_media_ops2 builtins with -fdeclare-opencl-builtins (PR #143507)

Juan Manuel Martinez =?utf-8?q?Caamaño?= , Juan Manuel Martinez =?utf-8?q?Caamaño?= , Juan Manuel Martinez =?utf-8?q?Caamaño?= Message-ID: In-Reply-To: @@ -0,0 +1,98 @@ +// RUN: %clang_cc1 %s -triple amdgcn-unknown-unknown -verify -pedantic -Wconversion -Werror -fsyntax-only -

[clang] [Clang][OpenCL] Declare cl_amd_media_ops/cl_amd_media_ops2 builtins with -fdeclare-opencl-builtins (PR #143507)

Juan Manuel Martinez =?utf-8?q?Caamaño?= , Juan Manuel Martinez =?utf-8?q?Caamaño?= , Juan Manuel Martinez =?utf-8?q?Caamaño?= Message-ID: In-Reply-To: @@ -0,0 +1,98 @@ +// RUN: %clang_cc1 %s -triple amdgcn-unknown-unknown -verify -pedantic -Wconversion -Werror -fsyntax-only -

[clang] [DirectoryWatcher] Use llvm::find (NFC) (PR #143337)

https://github.com/arsenm approved this pull request. https://github.com/llvm/llvm-project/pull/143337 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang-tools-extra] [clang-tools-extra] Remove unused local variables (NFC) (PR #142881)

https://github.com/arsenm approved this pull request. https://github.com/llvm/llvm-project/pull/142881 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] [tools] Remove unused local variables (NFC) (PR #142880)

https://github.com/arsenm approved this pull request. https://github.com/llvm/llvm-project/pull/142880 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] clang: Fix wrong C++ marker in tablegen file (PR #142534)

https://github.com/arsenm closed https://github.com/llvm/llvm-project/pull/142534 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] clang: Fix wrong C++ marker in tablegen file (PR #142534)

https://github.com/arsenm ready_for_review https://github.com/llvm/llvm-project/pull/142534 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] clang: Fix wrong C++ marker in tablegen file (PR #142534)

arsenm wrote: * **#142534** https://app.graphite.dev/github/pr/llvm/llvm-project/142534?utm_source=stack-comment-icon"; target="_blank">https://static.graphite.dev/graphite-32x32-black.png"; alt="Graphite" width="10px" height="10px"/> 👈 https://app.graphite.dev/github/pr/llvm/llvm-project/14253

[clang] clang: Fix wrong C++ marker in tablegen file (PR #142534)

https://github.com/arsenm created https://github.com/llvm/llvm-project/pull/142534 None >From ef0db5d83400a2d43a8706cbfb7374a91d68f72f Mon Sep 17 00:00:00 2001 From: Matt Arsenault Date: Mon, 2 Jun 2025 16:58:10 +0200 Subject: [PATCH] clang: Fix wrong C++ marker in tablegen file --- clang/in

  1   2   3   4   5   6   7   8   9   10   >