Author: Matt Arsenault Date: 2020-04-15T09:17:26-04:00 New Revision: 3a612450508b314b7a6f4db142d0c619031d760e
URL: https://github.com/llvm/llvm-project/commit/3a612450508b314b7a6f4db142d0c619031d760e DIFF: https://github.com/llvm/llvm-project/commit/3a612450508b314b7a6f4db142d0c619031d760e.diff LOG: clang/AMDGPU: Assume denormals are enabled for the default target. Since the default logic was based on having fast denormal/fma features, and the default target has no features, we assumed flushing by default. This fixes incorrectly assuming flushing in builds for "generic" IR libraries. The handling for no specified --cuda-gpu-arch in HIP is kind of broken. Somewhere else forces a default target of gfx803, which does not enable denormal handling by default. We don't see this default switching here, so you'll end up with a different denormal mode depending on whether you explicitly requested gfx803, or used it by default. Added: Modified: clang/lib/Driver/ToolChains/AMDGPU.cpp clang/test/Driver/cl-denorms-are-zero.cl clang/test/Driver/cuda-flush-denormals-to-zero.cu Removed: ################################################################################ diff --git a/clang/lib/Driver/ToolChains/AMDGPU.cpp b/clang/lib/Driver/ToolChains/AMDGPU.cpp index 2a796f28403f..e8c0b824ace1 100644 --- a/clang/lib/Driver/ToolChains/AMDGPU.cpp +++ b/clang/lib/Driver/ToolChains/AMDGPU.cpp @@ -262,6 +262,11 @@ AMDGPUToolChain::TranslateArgs(const DerivedArgList &Args, StringRef BoundArch, bool AMDGPUToolChain::getDefaultDenormsAreZeroForTarget( llvm::AMDGPU::GPUKind Kind) { + + // Assume nothing without a specific target. + if (Kind == llvm::AMDGPU::GK_NONE) + return false; + const unsigned ArchAttr = llvm::AMDGPU::getArchAttrAMDGCN(Kind); // Default to enabling f32 denormals by default on subtargets where fma is diff --git a/clang/test/Driver/cl-denorms-are-zero.cl b/clang/test/Driver/cl-denorms-are-zero.cl index 7774c0d60da8..e3fd095e5831 100644 --- a/clang/test/Driver/cl-denorms-are-zero.cl +++ b/clang/test/Driver/cl-denorms-are-zero.cl @@ -1,20 +1,24 @@ // Slow FMAF and slow f32 denormals -// RUN: %clang -### -target amdgcn--amdhsa -c -mcpu=pitcairn %s 2>&1 | FileCheck -check-prefixes=AMDGCN,AMDGCN-FLUSH %s +// RUN: %clang -### -target amdgcn--amdhsa -nogpulib -c -mcpu=pitcairn %s 2>&1 | FileCheck -check-prefixes=AMDGCN,AMDGCN-FLUSH %s // RUN: %clang -### -cl-denorms-are-zero -o - -target amdgcn--amdhsa -c -mcpu=pitcairn %s 2>&1 | FileCheck -check-prefixes=AMDGCN,AMDGCN-FLUSH %s // Fast FMAF, but slow f32 denormals -// RUN: %clang -### -target amdgcn--amdhsa -c -mcpu=tahiti %s 2>&1 | FileCheck -check-prefixes=AMDGCN,AMDGCN-FLUSH %s +// RUN: %clang -### -target amdgcn--amdhsa -nogpulib -c -mcpu=tahiti %s 2>&1 | FileCheck -check-prefixes=AMDGCN,AMDGCN-FLUSH %s // RUN: %clang -### -cl-denorms-are-zero -o - -target amdgcn--amdhsa -c -mcpu=tahiti %s 2>&1 | FileCheck -check-prefixes=AMDGCN,AMDGCN-FLUSH %s // Fast F32 denormals, but slow FMAF -// RUN: %clang -### -target amdgcn--amdhsa -c -mcpu=fiji %s 2>&1 | FileCheck -check-prefixes=AMDGCN,AMDGCN-FLUSH %s +// RUN: %clang -### -target amdgcn--amdhsa -nogpulib -c -mcpu=fiji %s 2>&1 | FileCheck -check-prefixes=AMDGCN,AMDGCN-FLUSH %s // RUN: %clang -### -cl-denorms-are-zero -o - -target amdgcn--amdhsa -c -mcpu=fiji %s 2>&1 | FileCheck -check-prefixes=AMDGCN,AMDGCN-FLUSH %s // Fast F32 denormals and fast FMAF -// RUN: %clang -### -target amdgcn--amdhsa -c -mcpu=gfx900 %s 2>&1 | FileCheck -check-prefixes=AMDGCN,AMDGCN-DENORM %s -// RUN: %clang -### -cl-denorms-are-zero -o - -target amdgcn--amdhsa -c -mcpu=gfx900 %s 2>&1 | FileCheck -check-prefixes=AMDGCN,AMDGCN-FLUSH %s +// RUN: %clang -### -target amdgcn--amdhsa -nogpulib -c -mcpu=gfx900 %s 2>&1 | FileCheck -check-prefixes=AMDGCN,AMDGCN-DENORM %s +// RUN: %clang -### -cl-denorms-are-zero -o - -target amdgcn--amdhsa -nogpulib -c -mcpu=gfx900 %s 2>&1 | FileCheck -check-prefixes=AMDGCN,AMDGCN-FLUSH %s + +// Default target is artificial, but should assume a conservative default. +// RUN: %clang -### -target amdgcn--amdhsa -nogpulib -c %s 2>&1 | FileCheck -check-prefixes=AMDGCN,AMDGCN-DENORM %s +// RUN: %clang -### -cl-denorms-are-zero -o - -target amdgcn--amdhsa -nogpulib -c %s 2>&1 | FileCheck -check-prefixes=AMDGCN,AMDGCN-FLUSH %s // AMDGCN-FLUSH: "-fdenormal-fp-math-f32=preserve-sign,preserve-sign" // This should be omitted and default to ieee -// AMDGCN-DENORM-NOT: "-fdenormal-fp-math-f32" +// AMDGCN-DENORM-NOT: denormal-fp-math diff --git a/clang/test/Driver/cuda-flush-denormals-to-zero.cu b/clang/test/Driver/cuda-flush-denormals-to-zero.cu index a515b5f8ca07..4a7b88823771 100644 --- a/clang/test/Driver/cuda-flush-denormals-to-zero.cu +++ b/clang/test/Driver/cuda-flush-denormals-to-zero.cu @@ -22,6 +22,8 @@ // RUN: %clang -x hip -no-canonical-prefixes -### -target x86_64-linux-gnu -c -march=haswell --cuda-gpu-arch=gfx803 -nocudainc -nogpulib %s 2>&1 | FileCheck -check-prefix=FTZ %s // RUN: %clang -x hip -no-canonical-prefixes -### -target x86_64-linux-gnu -c -march=haswell --cuda-gpu-arch=gfx900 -nocudainc -nogpulib %s 2>&1 | FileCheck -check-prefix=NOFTZ %s +// Test no subtarget, which should get the denormal setting of the default gfx803 +// RUN: %clang -x hip -no-canonical-prefixes -### -target x86_64-linux-gnu -c -march=haswell -nocudainc -nogpulib %s 2>&1 | FileCheck -check-prefix=FTZ %s // Test multiple offload archs with diff erent defaults. // RUN: %clang -x hip -no-canonical-prefixes -### -target x86_64-linux-gnu -c -march=haswell --cuda-gpu-arch=gfx803 --cuda-gpu-arch=gfx900 -nocudainc -nogpulib %s 2>&1 | FileCheck -check-prefix=MIXED-DEFAULT-MODE %s _______________________________________________ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
