From: Matt Arsenault Date: Mon, 7 Aug 2017 14:58:04 +0000 (+0000) Subject: AMDGPU: Cleanup subtarget features X-Git-Url: https://granicus.if.org/sourcecode?a=commitdiff_plain;h=2e48864110d45366e68565f60adc9bd96df6248c;p=llvm AMDGPU: Cleanup subtarget features Try to avoid mutually exclusive features. Don't use a real default GPU, and use a fake "generic". The goal is to make it easier to see which set of features are incompatible between feature strings. Most of the test changes are due to random scheduling changes from not having a default fullspeed model. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@310258 91177308-0d34-0410-b5e6-96231b3b80d8 --- diff --git a/lib/Target/AMDGPU/AMDGPU.td b/lib/Target/AMDGPU/AMDGPU.td index 6d0bb5ef19c..959a43bfa51 100644 --- a/lib/Target/AMDGPU/AMDGPU.td +++ b/lib/Target/AMDGPU/AMDGPU.td @@ -172,12 +172,6 @@ def FeatureGCN : SubtargetFeature<"gcn", "GCN or newer GPU" >; -def FeatureGCN1Encoding : SubtargetFeature<"gcn1-encoding", - "GCN1Encoding", - "true", - "Encoding format for SI and CI" ->; - def FeatureGCN3Encoding : SubtargetFeature<"gcn3-encoding", "GCN3Encoding", "true", @@ -442,14 +436,14 @@ def FeatureNorthernIslands : SubtargetFeatureGeneration<"NORTHERN_ISLANDS", def FeatureSouthernIslands : SubtargetFeatureGeneration<"SOUTHERN_ISLANDS", [FeatureFP64, FeatureLocalMemorySize32768, - FeatureWavefrontSize64, FeatureGCN, FeatureGCN1Encoding, + FeatureWavefrontSize64, FeatureGCN, FeatureLDSBankCount32, FeatureMovrel] >; def FeatureSeaIslands : SubtargetFeatureGeneration<"SEA_ISLANDS", [FeatureFP64, FeatureLocalMemorySize65536, FeatureWavefrontSize64, FeatureGCN, FeatureFlatAddressSpace, - FeatureGCN1Encoding, FeatureCIInsts, FeatureMovrel] + FeatureCIInsts, FeatureMovrel] >; def FeatureVolcanicIslands : SubtargetFeatureGeneration<"VOLCANIC_ISLANDS", @@ -667,7 +661,7 @@ def TruePredicate : Predicate<"true">; def isSICI : Predicate< "Subtarget->getGeneration() == AMDGPUSubtarget::SOUTHERN_ISLANDS ||" "Subtarget->getGeneration() == AMDGPUSubtarget::SEA_ISLANDS" ->, AssemblerPredicate<"FeatureGCN1Encoding">; +>, AssemblerPredicate<"!FeatureGCN3Encoding">; def isVI : Predicate < "Subtarget->getGeneration() >= AMDGPUSubtarget::VOLCANIC_ISLANDS">, diff --git a/lib/Target/AMDGPU/AMDGPUSubtarget.cpp b/lib/Target/AMDGPU/AMDGPUSubtarget.cpp index b69611817c3..deba76a207c 100644 --- a/lib/Target/AMDGPU/AMDGPUSubtarget.cpp +++ b/lib/Target/AMDGPU/AMDGPUSubtarget.cpp @@ -50,7 +50,7 @@ AMDGPUSubtarget::initializeSubtargetDependencies(const Triple &TT, SmallString<256> FullFS("+promote-alloca,+fp64-fp16-denormals,+dx10-clamp,+load-store-opt,"); if (isAmdHsaOS()) // Turn on FlatForGlobal for HSA. - FullFS += "+flat-for-global,+unaligned-buffer-access,+trap-handler,"; + FullFS += "+flat-address-space,+flat-for-global,+unaligned-buffer-access,+trap-handler,"; FullFS += FS; @@ -75,6 +75,18 @@ AMDGPUSubtarget::initializeSubtargetDependencies(const Triple &TT, if (MaxPrivateElementSize == 0) MaxPrivateElementSize = 4; + if (LDSBankCount == 0) + LDSBankCount = 32; + + if (TT.getArch() == Triple::amdgcn) { + if (LocalMemorySize == 0) + LocalMemorySize = 32768; + + // Do something sensible for unspecified target. + if (!HasMovrel && !HasVGPRIndexMode) + HasMovrel = true; + } + return *this; } @@ -117,7 +129,6 @@ AMDGPUSubtarget::AMDGPUSubtarget(const Triple &TT, StringRef GPU, StringRef FS, FP64(false), IsGCN(false), - GCN1Encoding(false), GCN3Encoding(false), CIInsts(false), GFX9Insts(false), diff --git a/lib/Target/AMDGPU/AMDGPUSubtarget.h b/lib/Target/AMDGPU/AMDGPUSubtarget.h index 346788dd8ed..8aad97caa97 100644 --- a/lib/Target/AMDGPU/AMDGPUSubtarget.h +++ b/lib/Target/AMDGPU/AMDGPUSubtarget.h @@ -139,7 +139,6 @@ protected: // Subtarget statically properties set by tablegen bool FP64; bool IsGCN; - bool GCN1Encoding; bool GCN3Encoding; bool CIInsts; bool GFX9Insts; diff --git a/lib/Target/AMDGPU/AMDGPUTargetMachine.cpp b/lib/Target/AMDGPU/AMDGPUTargetMachine.cpp index 7a8fd019593..3ad61049996 100644 --- a/lib/Target/AMDGPU/AMDGPUTargetMachine.cpp +++ b/lib/Target/AMDGPU/AMDGPUTargetMachine.cpp @@ -256,9 +256,8 @@ static StringRef getGPUOrDefault(const Triple &TT, StringRef GPU) { if (!GPU.empty()) return GPU; - // HSA only supports CI+, so change the default GPU to a CI for HSA. if (TT.getArch() == Triple::amdgcn) - return (TT.getOS() == Triple::AMDHSA) ? "kaveri" : "tahiti"; + return "generic"; return "r600"; } diff --git a/lib/Target/AMDGPU/FLATInstructions.td b/lib/Target/AMDGPU/FLATInstructions.td index df9fcff509e..a47bb52ee6c 100644 --- a/lib/Target/AMDGPU/FLATInstructions.td +++ b/lib/Target/AMDGPU/FLATInstructions.td @@ -25,8 +25,6 @@ class FLAT_Pseudo has_glc = 1; bits<1> glcValue = 0; + let SubtargetPredicate = !if(is_flat_global, HasFlatGlobalInsts, + !if(is_flat_scratch, HasFlatScratchInsts, HasFlatAddressSpace)); + // TODO: M0 if it could possibly access LDS (before gfx9? only)? let Uses = !if(is_flat_global, [EXEC], [EXEC, FLAT_SCR]); @@ -667,7 +668,7 @@ class FlatSignedAtomicPat ; -let Predicates = [isCIVI] in { +let Predicates = [HasFlatAddressSpace] in { def : FlatLoadPat ; def : FlatLoadPat ; @@ -719,13 +720,9 @@ def : FlatAtomicPat ; def : FlatAtomicPat ; def : FlatAtomicPat ; -} // End Predicates = [isCIVI] - -let Predicates = [isVI] in { - def : FlatStorePat ; - def : FlatStorePat ; -} - +def : FlatStorePat ; +def : FlatStorePat ; +} // End Predicates = [HasFlatAddressSpace] let Predicates = [HasFlatGlobalInsts], AddedComplexity = 10 in { diff --git a/lib/Target/AMDGPU/Processors.td b/lib/Target/AMDGPU/Processors.td index d30d1d38258..23bf66463bf 100644 --- a/lib/Target/AMDGPU/Processors.td +++ b/lib/Target/AMDGPU/Processors.td @@ -10,6 +10,12 @@ class Proc Features> : Processor; +// The code produced for "generic" is only useful for tests and cannot +// reasonably be expected to execute on any particular target. +def : ProcessorModel<"generic", NoSchedModel, [ + FeatureGCN +]>; + //===----------------------------------------------------------------------===// // R600 //===----------------------------------------------------------------------===// diff --git a/lib/Target/AMDGPU/Utils/AMDGPUBaseInfo.cpp b/lib/Target/AMDGPU/Utils/AMDGPUBaseInfo.cpp index f6d5853b2ba..03c9f7f4f39 100644 --- a/lib/Target/AMDGPU/Utils/AMDGPUBaseInfo.cpp +++ b/lib/Target/AMDGPU/Utils/AMDGPUBaseInfo.cpp @@ -531,6 +531,10 @@ bool isGFX9(const MCSubtargetInfo &STI) { return STI.getFeatureBits()[AMDGPU::FeatureGFX9]; } +bool isGCN3Encoding(const MCSubtargetInfo &STI) { + return STI.getFeatureBits()[AMDGPU::FeatureGCN3Encoding]; +} + bool isSGPR(unsigned Reg, const MCRegisterInfo* TRI) { const MCRegisterClass SGPRClass = TRI->getRegClass(AMDGPU::SReg_32RegClassID); const unsigned FirstSubReg = TRI->getSubReg(Reg, 1); @@ -773,16 +777,15 @@ bool isUniformMMO(const MachineMemOperand *MMO) { } int64_t getSMRDEncodedOffset(const MCSubtargetInfo &ST, int64_t ByteOffset) { - if (isSI(ST) || isCI(ST)) - return ByteOffset >> 2; - - return ByteOffset; + if (isGCN3Encoding(ST)) + return ByteOffset; + return ByteOffset >> 2; } bool isLegalSMRDImmOffset(const MCSubtargetInfo &ST, int64_t ByteOffset) { int64_t EncodedOffset = getSMRDEncodedOffset(ST, ByteOffset); - return isSI(ST) || isCI(ST) ? isUInt<8>(EncodedOffset) : - isUInt<20>(EncodedOffset); + return isGCN3Encoding(ST) ? + isUInt<20>(EncodedOffset) : isUInt<8>(EncodedOffset); } } // end namespace AMDGPU diff --git a/test/CodeGen/AMDGPU/addrspacecast.ll b/test/CodeGen/AMDGPU/addrspacecast.ll index 93f6d88a02c..901b197b1b8 100644 --- a/test/CodeGen/AMDGPU/addrspacecast.ll +++ b/test/CodeGen/AMDGPU/addrspacecast.ll @@ -1,4 +1,4 @@ -; RUN: llc -march=amdgcn -mtriple=amdgcn-amd-amdhsa -mattr=-promote-alloca -verify-machineinstrs < %s | FileCheck -enable-var-scope -check-prefix=HSA -check-prefix=CI %s +; RUN: llc -march=amdgcn -mtriple=amdgcn-amd-amdhsa -mcpu=kaveri -mattr=-promote-alloca -verify-machineinstrs < %s | FileCheck -enable-var-scope -check-prefix=HSA -check-prefix=CI %s ; RUN: llc -march=amdgcn -mtriple=amdgcn-amd-amdhsa -mcpu=gfx900 -mattr=-promote-alloca -verify-machineinstrs < %s | FileCheck -enable-var-scope -check-prefix=HSA -check-prefix=GFX9 %s ; HSA-LABEL: {{^}}use_group_to_flat_addrspacecast: diff --git a/test/CodeGen/AMDGPU/array-ptr-calc-i32.ll b/test/CodeGen/AMDGPU/array-ptr-calc-i32.ll index daa3442097c..90bc6a274f1 100644 --- a/test/CodeGen/AMDGPU/array-ptr-calc-i32.ll +++ b/test/CodeGen/AMDGPU/array-ptr-calc-i32.ll @@ -1,5 +1,5 @@ -; RUN: llc -verify-machineinstrs -march=amdgcn -mattr=-promote-alloca < %s | FileCheck -check-prefix=SI-ALLOCA -check-prefix=SI %s -; RUN: llc -verify-machineinstrs -march=amdgcn -mattr=+promote-alloca < %s | FileCheck -check-prefix=SI-PROMOTE -check-prefix=SI %s +; RUN: llc -verify-machineinstrs -march=amdgcn -mcpu=tahiti -mattr=-promote-alloca < %s | FileCheck -check-prefix=SI-ALLOCA -check-prefix=SI %s +; RUN: llc -verify-machineinstrs -march=amdgcn -mcpu=tahiti -mattr=+promote-alloca < %s | FileCheck -check-prefix=SI-PROMOTE -check-prefix=SI %s declare i32 @llvm.amdgcn.mbcnt.lo(i32, i32) #1 declare i32 @llvm.amdgcn.mbcnt.hi(i32, i32) #1 diff --git a/test/CodeGen/AMDGPU/bitcast-vector-extract.ll b/test/CodeGen/AMDGPU/bitcast-vector-extract.ll index cf95f74afb8..722099fec33 100644 --- a/test/CodeGen/AMDGPU/bitcast-vector-extract.ll +++ b/test/CodeGen/AMDGPU/bitcast-vector-extract.ll @@ -1,4 +1,4 @@ -; RUN: llc -march=amdgcn -verify-machineinstrs < %s | FileCheck -check-prefix=GCN %s +; RUN: llc -march=amdgcn -mcpu=tahiti -verify-machineinstrs < %s | FileCheck -check-prefix=GCN %s ; RUN: llc -march=amdgcn -mcpu=tonga -mattr=-flat-for-global -verify-machineinstrs < %s | FileCheck -check-prefix=GCN %s ; The bitcast should be pushed through the bitcasts so the vectors can diff --git a/test/CodeGen/AMDGPU/br_cc.f16.ll b/test/CodeGen/AMDGPU/br_cc.f16.ll index b7a0c8738df..283384dc298 100644 --- a/test/CodeGen/AMDGPU/br_cc.f16.ll +++ b/test/CodeGen/AMDGPU/br_cc.f16.ll @@ -1,4 +1,4 @@ -; RUN: llc -march=amdgcn -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=SI %s +; RUN: llc -march=amdgcn -mcpu=tahiti -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=SI %s ; RUN: llc -march=amdgcn -mcpu=fiji -mattr=-flat-for-global -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=VI %s ; GCN-LABEL: {{^}}br_cc_f16: diff --git a/test/CodeGen/AMDGPU/branch-relaxation.ll b/test/CodeGen/AMDGPU/branch-relaxation.ll index 233d5a5822f..661a6bd9958 100644 --- a/test/CodeGen/AMDGPU/branch-relaxation.ll +++ b/test/CodeGen/AMDGPU/branch-relaxation.ll @@ -1,4 +1,4 @@ -; RUN: llc -march=amdgcn -verify-machineinstrs -amdgpu-s-branch-bits=4 < %s | FileCheck -check-prefix=GCN %s +; RUN: llc -march=amdgcn -mcpu=tahiti -verify-machineinstrs -amdgpu-s-branch-bits=4 < %s | FileCheck -check-prefix=GCN %s ; FIXME: We should use llvm-mc for this, but we can't even parse our own output. diff --git a/test/CodeGen/AMDGPU/callee-special-input-sgprs.ll b/test/CodeGen/AMDGPU/callee-special-input-sgprs.ll index 11e1a95802b..1518c0e503e 100644 --- a/test/CodeGen/AMDGPU/callee-special-input-sgprs.ll +++ b/test/CodeGen/AMDGPU/callee-special-input-sgprs.ll @@ -1,4 +1,4 @@ -; RUN: llc -mtriple=amdgcn-amd-amdhsa -verify-machineinstrs < %s | FileCheck -enable-var-scope -check-prefixes=GCN,CIVI %s +; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=kaveri -verify-machineinstrs < %s | FileCheck -enable-var-scope -check-prefixes=GCN,CIVI %s ; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx900 -verify-machineinstrs < %s | FileCheck -enable-var-scope -check-prefixes=GCN,GFX9 %s ; GCN-LABEL: {{^}}use_dispatch_ptr: diff --git a/test/CodeGen/AMDGPU/callee-special-input-vgprs.ll b/test/CodeGen/AMDGPU/callee-special-input-vgprs.ll index 10ccce65ba7..50da6b44625 100644 --- a/test/CodeGen/AMDGPU/callee-special-input-vgprs.ll +++ b/test/CodeGen/AMDGPU/callee-special-input-vgprs.ll @@ -1,4 +1,4 @@ -; RUN: llc -mtriple=amdgcn-amd-amdhsa -verify-machineinstrs < %s | FileCheck -enable-var-scope -check-prefix=GCN %s +; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=kaveri -verify-machineinstrs < %s | FileCheck -enable-var-scope -check-prefix=GCN %s ; GCN-LABEL: {{^}}use_workitem_id_x: ; GCN: s_waitcnt diff --git a/test/CodeGen/AMDGPU/captured-frame-index.ll b/test/CodeGen/AMDGPU/captured-frame-index.ll index 5fe1b272850..99f6ec48be7 100644 --- a/test/CodeGen/AMDGPU/captured-frame-index.ll +++ b/test/CodeGen/AMDGPU/captured-frame-index.ll @@ -1,4 +1,4 @@ -; RUN: llc -mtriple=amdgcn-- -mattr=-promote-alloca -amdgpu-sroa=0 -verify-machineinstrs < %s | FileCheck -check-prefix=GCN %s +; RUN: llc -mtriple=amdgcn-- -mcpu=tahiti -mattr=-promote-alloca -amdgpu-sroa=0 -verify-machineinstrs < %s | FileCheck -check-prefix=GCN %s ; GCN-LABEL: {{^}}store_fi_lifetime: ; GCN: v_mov_b32_e32 [[FI:v[0-9]+]], 4{{$}} diff --git a/test/CodeGen/AMDGPU/collapse-endcf.ll b/test/CodeGen/AMDGPU/collapse-endcf.ll index f4199d41221..ec3bbc814c7 100644 --- a/test/CodeGen/AMDGPU/collapse-endcf.ll +++ b/test/CodeGen/AMDGPU/collapse-endcf.ll @@ -1,4 +1,4 @@ -; RUN: llc -march=amdgcn -verify-machineinstrs < %s | FileCheck -check-prefix=GCN %s +; RUN: llc -march=amdgcn -mcpu=tahiti -verify-machineinstrs < %s | FileCheck -check-prefix=GCN %s ; GCN-LABEL: {{^}}simple_nested_if: ; GCN: s_and_saveexec_b64 [[SAVEEXEC:s\[[0-9:]+\]]] diff --git a/test/CodeGen/AMDGPU/ctpop.ll b/test/CodeGen/AMDGPU/ctpop.ll index 68b39bad2bc..4a6743ee98b 100644 --- a/test/CodeGen/AMDGPU/ctpop.ll +++ b/test/CodeGen/AMDGPU/ctpop.ll @@ -1,4 +1,4 @@ -; RUN: llc -march=amdgcn -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=FUNC -check-prefix=SI %s +; RUN: llc -march=amdgcn -mcpu=tahiti -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=FUNC -check-prefix=SI %s ; RUN: llc -march=amdgcn -mcpu=tonga -mattr=-flat-for-global -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=FUNC -check-prefix=VI %s ; RUN: llc -march=r600 -mcpu=cypress -verify-machineinstrs < %s | FileCheck -check-prefix=EG -check-prefix=FUNC %s diff --git a/test/CodeGen/AMDGPU/ctpop64.ll b/test/CodeGen/AMDGPU/ctpop64.ll index 4850370851f..8d2dac9538f 100644 --- a/test/CodeGen/AMDGPU/ctpop64.ll +++ b/test/CodeGen/AMDGPU/ctpop64.ll @@ -1,4 +1,4 @@ -; RUN: llc -march=amdgcn -verify-machineinstrs < %s | FileCheck -check-prefix=SI -check-prefix=GCN -check-prefix=FUNC %s +; RUN: llc -march=amdgcn -mcpu=tahiti -verify-machineinstrs < %s | FileCheck -check-prefix=SI -check-prefix=GCN -check-prefix=FUNC %s ; RUN: llc -march=amdgcn -mcpu=tonga -mattr=-flat-for-global -verify-machineinstrs < %s | FileCheck -check-prefix=VI -check-prefix=GCN -check-prefix=FUNC %s declare i32 @llvm.r600.read.tidig.x() nounwind readnone diff --git a/test/CodeGen/AMDGPU/cvt_f32_ubyte.ll b/test/CodeGen/AMDGPU/cvt_f32_ubyte.ll index f839129fc3d..0600633aa27 100644 --- a/test/CodeGen/AMDGPU/cvt_f32_ubyte.ll +++ b/test/CodeGen/AMDGPU/cvt_f32_ubyte.ll @@ -1,4 +1,4 @@ -; RUN: llc -march=amdgcn -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=SI %s +; RUN: llc -march=amdgcn -mcpu=tahiti -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=SI %s ; RUN: llc -march=amdgcn -mcpu=tonga -mattr=-flat-for-global -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=VI %s declare i32 @llvm.amdgcn.workitem.id.x() nounwind readnone @@ -12,7 +12,7 @@ declare i32 @llvm.amdgcn.workitem.id.y() nounwind readnone ; GCN: buffer_store_dword [[CONV]], define amdgpu_kernel void @load_i8_to_f32(float addrspace(1)* noalias %out, i8 addrspace(1)* noalias %in) nounwind { %tid = call i32 @llvm.amdgcn.workitem.id.x() - %gep = getelementptr i8, i8 addrspace(1)* %in, i32 %tid + %gep = getelementptr i8, i8 addrspace(1)* %in, i32 %tid %load = load i8, i8 addrspace(1)* %gep, align 1 %cvt = uitofp i8 %load to float store float %cvt, float addrspace(1)* %out, align 4 @@ -26,7 +26,7 @@ define amdgpu_kernel void @load_i8_to_f32(float addrspace(1)* noalias %out, i8 a ; GCN: buffer_store_dwordx2 v{{\[}}[[LORESULT]]:[[HIRESULT]]{{\]}}, define amdgpu_kernel void @load_v2i8_to_v2f32(<2 x float> addrspace(1)* noalias %out, <2 x i8> addrspace(1)* noalias %in) nounwind { %tid = call i32 @llvm.amdgcn.workitem.id.x() - %gep = getelementptr <2 x i8>, <2 x i8> addrspace(1)* %in, i32 %tid + %gep = getelementptr <2 x i8>, <2 x i8> addrspace(1)* %in, i32 %tid %load = load <2 x i8>, <2 x i8> addrspace(1)* %gep, align 2 %cvt = uitofp <2 x i8> %load to <2 x float> store <2 x float> %cvt, <2 x float> addrspace(1)* %out, align 16 @@ -42,7 +42,7 @@ define amdgpu_kernel void @load_v2i8_to_v2f32(<2 x float> addrspace(1)* noalias ; GCN: buffer_store_dwordx2 v{{\[}}[[LORESULT]]:[[HIRESULT]]{{\]}}, define amdgpu_kernel void @load_v3i8_to_v3f32(<3 x float> addrspace(1)* noalias %out, <3 x i8> addrspace(1)* noalias %in) nounwind { %tid = call i32 @llvm.amdgcn.workitem.id.x() - %gep = getelementptr <3 x i8>, <3 x i8> addrspace(1)* %in, i32 %tid + %gep = getelementptr <3 x i8>, <3 x i8> addrspace(1)* %in, i32 %tid %load = load <3 x i8>, <3 x i8> addrspace(1)* %gep, align 4 %cvt = uitofp <3 x i8> %load to <3 x float> store <3 x float> %cvt, <3 x float> addrspace(1)* %out, align 16 @@ -60,7 +60,7 @@ define amdgpu_kernel void @load_v3i8_to_v3f32(<3 x float> addrspace(1)* noalias ; GCN: buffer_store_dwordx4 v{{\[}}[[LORESULT]]:[[HIRESULT]]{{\]}}, define amdgpu_kernel void @load_v4i8_to_v4f32(<4 x float> addrspace(1)* noalias %out, <4 x i8> addrspace(1)* noalias %in) nounwind { %tid = call i32 @llvm.amdgcn.workitem.id.x() - %gep = getelementptr <4 x i8>, <4 x i8> addrspace(1)* %in, i32 %tid + %gep = getelementptr <4 x i8>, <4 x i8> addrspace(1)* %in, i32 %tid %load = load <4 x i8>, <4 x i8> addrspace(1)* %gep, align 4 %cvt = uitofp <4 x i8> %load to <4 x float> store <4 x float> %cvt, <4 x float> addrspace(1)* %out, align 16 @@ -86,7 +86,7 @@ define amdgpu_kernel void @load_v4i8_to_v4f32(<4 x float> addrspace(1)* noalias ; GCN: buffer_store_dwordx4 define amdgpu_kernel void @load_v4i8_to_v4f32_unaligned(<4 x float> addrspace(1)* noalias %out, <4 x i8> addrspace(1)* noalias %in) nounwind { %tid = call i32 @llvm.amdgcn.workitem.id.x() - %gep = getelementptr <4 x i8>, <4 x i8> addrspace(1)* %in, i32 %tid + %gep = getelementptr <4 x i8>, <4 x i8> addrspace(1)* %in, i32 %tid %load = load <4 x i8>, <4 x i8> addrspace(1)* %gep, align 1 %cvt = uitofp <4 x i8> %load to <4 x float> store <4 x float> %cvt, <4 x float> addrspace(1)* %out, align 16 @@ -135,7 +135,7 @@ define amdgpu_kernel void @load_v4i8_to_v4f32_2_uses(<4 x float> addrspace(1)* n ; GCN: s_endpgm define amdgpu_kernel void @load_v7i8_to_v7f32(<7 x float> addrspace(1)* noalias %out, <7 x i8> addrspace(1)* noalias %in) nounwind { %tid = call i32 @llvm.amdgcn.workitem.id.x() - %gep = getelementptr <7 x i8>, <7 x i8> addrspace(1)* %in, i32 %tid + %gep = getelementptr <7 x i8>, <7 x i8> addrspace(1)* %in, i32 %tid %load = load <7 x i8>, <7 x i8> addrspace(1)* %gep, align 1 %cvt = uitofp <7 x i8> %load to <7 x float> store <7 x float> %cvt, <7 x float> addrspace(1)* %out, align 16 @@ -160,7 +160,7 @@ define amdgpu_kernel void @load_v7i8_to_v7f32(<7 x float> addrspace(1)* noalias ; GCN: buffer_store_dwordx4 define amdgpu_kernel void @load_v8i8_to_v8f32(<8 x float> addrspace(1)* noalias %out, <8 x i8> addrspace(1)* noalias %in) nounwind { %tid = call i32 @llvm.amdgcn.workitem.id.x() - %gep = getelementptr <8 x i8>, <8 x i8> addrspace(1)* %in, i32 %tid + %gep = getelementptr <8 x i8>, <8 x i8> addrspace(1)* %in, i32 %tid %load = load <8 x i8>, <8 x i8> addrspace(1)* %gep, align 8 %cvt = uitofp <8 x i8> %load to <8 x float> store <8 x float> %cvt, <8 x float> addrspace(1)* %out, align 16 @@ -174,7 +174,7 @@ define amdgpu_kernel void @load_v8i8_to_v8f32(<8 x float> addrspace(1)* noalias ; GCN: buffer_store_dword [[CONV]], define amdgpu_kernel void @i8_zext_inreg_i32_to_f32(float addrspace(1)* noalias %out, i32 addrspace(1)* noalias %in) nounwind { %tid = call i32 @llvm.amdgcn.workitem.id.x() - %gep = getelementptr i32, i32 addrspace(1)* %in, i32 %tid + %gep = getelementptr i32, i32 addrspace(1)* %in, i32 %tid %load = load i32, i32 addrspace(1)* %gep, align 4 %add = add i32 %load, 2 %inreg = and i32 %add, 255 @@ -186,7 +186,7 @@ define amdgpu_kernel void @i8_zext_inreg_i32_to_f32(float addrspace(1)* noalias ; GCN-LABEL: {{^}}i8_zext_inreg_hi1_to_f32: define amdgpu_kernel void @i8_zext_inreg_hi1_to_f32(float addrspace(1)* noalias %out, i32 addrspace(1)* noalias %in) nounwind { %tid = call i32 @llvm.amdgcn.workitem.id.x() - %gep = getelementptr i32, i32 addrspace(1)* %in, i32 %tid + %gep = getelementptr i32, i32 addrspace(1)* %in, i32 %tid %load = load i32, i32 addrspace(1)* %gep, align 4 %inreg = and i32 %load, 65280 %shr = lshr i32 %inreg, 8 @@ -200,7 +200,7 @@ define amdgpu_kernel void @i8_zext_inreg_hi1_to_f32(float addrspace(1)* noalias ; GCN-LABEL: {{^}}i8_zext_i32_to_f32: define amdgpu_kernel void @i8_zext_i32_to_f32(float addrspace(1)* noalias %out, i8 addrspace(1)* noalias %in) nounwind { %tid = call i32 @llvm.amdgcn.workitem.id.x() - %gep = getelementptr i8, i8 addrspace(1)* %in, i32 %tid + %gep = getelementptr i8, i8 addrspace(1)* %in, i32 %tid %load = load i8, i8 addrspace(1)* %gep, align 1 %ext = zext i8 %load to i32 %cvt = uitofp i32 %ext to float @@ -211,7 +211,7 @@ define amdgpu_kernel void @i8_zext_i32_to_f32(float addrspace(1)* noalias %out, ; GCN-LABEL: {{^}}v4i8_zext_v4i32_to_v4f32: define amdgpu_kernel void @v4i8_zext_v4i32_to_v4f32(<4 x float> addrspace(1)* noalias %out, <4 x i8> addrspace(1)* noalias %in) nounwind { %tid = call i32 @llvm.amdgcn.workitem.id.x() - %gep = getelementptr <4 x i8>, <4 x i8> addrspace(1)* %in, i32 %tid + %gep = getelementptr <4 x i8>, <4 x i8> addrspace(1)* %in, i32 %tid %load = load <4 x i8>, <4 x i8> addrspace(1)* %gep, align 1 %ext = zext <4 x i8> %load to <4 x i32> %cvt = uitofp <4 x i32> %ext to <4 x float> @@ -226,7 +226,7 @@ define amdgpu_kernel void @v4i8_zext_v4i32_to_v4f32(<4 x float> addrspace(1)* no ; GCN: buffer_store_dword [[CONV]] define amdgpu_kernel void @extract_byte0_to_f32(float addrspace(1)* noalias %out, i32 addrspace(1)* noalias %in) nounwind { %tid = call i32 @llvm.amdgcn.workitem.id.x() - %gep = getelementptr i32, i32 addrspace(1)* %in, i32 %tid + %gep = getelementptr i32, i32 addrspace(1)* %in, i32 %tid %val = load i32, i32 addrspace(1)* %gep %and = and i32 %val, 255 %cvt = uitofp i32 %and to float @@ -241,7 +241,7 @@ define amdgpu_kernel void @extract_byte0_to_f32(float addrspace(1)* noalias %out ; GCN: buffer_store_dword [[CONV]] define amdgpu_kernel void @extract_byte1_to_f32(float addrspace(1)* noalias %out, i32 addrspace(1)* noalias %in) nounwind { %tid = call i32 @llvm.amdgcn.workitem.id.x() - %gep = getelementptr i32, i32 addrspace(1)* %in, i32 %tid + %gep = getelementptr i32, i32 addrspace(1)* %in, i32 %tid %val = load i32, i32 addrspace(1)* %gep %srl = lshr i32 %val, 8 %and = and i32 %srl, 255 @@ -257,7 +257,7 @@ define amdgpu_kernel void @extract_byte1_to_f32(float addrspace(1)* noalias %out ; GCN: buffer_store_dword [[CONV]] define amdgpu_kernel void @extract_byte2_to_f32(float addrspace(1)* noalias %out, i32 addrspace(1)* noalias %in) nounwind { %tid = call i32 @llvm.amdgcn.workitem.id.x() - %gep = getelementptr i32, i32 addrspace(1)* %in, i32 %tid + %gep = getelementptr i32, i32 addrspace(1)* %in, i32 %tid %val = load i32, i32 addrspace(1)* %gep %srl = lshr i32 %val, 16 %and = and i32 %srl, 255 @@ -273,7 +273,7 @@ define amdgpu_kernel void @extract_byte2_to_f32(float addrspace(1)* noalias %out ; GCN: buffer_store_dword [[CONV]] define amdgpu_kernel void @extract_byte3_to_f32(float addrspace(1)* noalias %out, i32 addrspace(1)* noalias %in) nounwind { %tid = call i32 @llvm.amdgcn.workitem.id.x() - %gep = getelementptr i32, i32 addrspace(1)* %in, i32 %tid + %gep = getelementptr i32, i32 addrspace(1)* %in, i32 %tid %val = load i32, i32 addrspace(1)* %gep %srl = lshr i32 %val, 24 %and = and i32 %srl, 255 diff --git a/test/CodeGen/AMDGPU/fabs.f16.ll b/test/CodeGen/AMDGPU/fabs.f16.ll index b90290e9724..9da2479853b 100644 --- a/test/CodeGen/AMDGPU/fabs.f16.ll +++ b/test/CodeGen/AMDGPU/fabs.f16.ll @@ -1,4 +1,4 @@ -; RUN: llc -mtriple=amdgcn--amdhsa -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=CI %s +; RUN: llc -mtriple=amdgcn--amdhsa -mcpu=kaveri -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=CI %s ; RUN: llc -mtriple=amdgcn--amdhsa -mcpu=tonga -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=VI %s ; RUN: llc -mtriple=amdgcn--amdhsa -mcpu=gfx901 -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=GFX9 %s diff --git a/test/CodeGen/AMDGPU/fadd.f16.ll b/test/CodeGen/AMDGPU/fadd.f16.ll index 88b3be0e0d3..d32b6201f28 100644 --- a/test/CodeGen/AMDGPU/fadd.f16.ll +++ b/test/CodeGen/AMDGPU/fadd.f16.ll @@ -1,4 +1,4 @@ -; RUN: llc -march=amdgcn -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=SI %s +; RUN: llc -march=amdgcn -mcpu=tahiti -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=SI %s ; RUN: llc -march=amdgcn -mcpu=fiji -mattr=-flat-for-global -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=VI %s ; GCN-LABEL: {{^}}fadd_f16 diff --git a/test/CodeGen/AMDGPU/fcopysign.f16.ll b/test/CodeGen/AMDGPU/fcopysign.f16.ll index 335c6808a94..2f24f5f51ce 100644 --- a/test/CodeGen/AMDGPU/fcopysign.f16.ll +++ b/test/CodeGen/AMDGPU/fcopysign.f16.ll @@ -1,6 +1,6 @@ -; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -verify-machineinstrs < %s | FileCheck -enable-var-scope -check-prefix=GCN -check-prefix=SI %s -; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=tonga -mattr=-flat-for-global -verify-machineinstrs < %s | FileCheck -enable-var-scope -check-prefix=GCN -check-prefix=GFX89 -check-prefix=GFX8 %s -; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=gfx901 -mattr=-flat-for-global -verify-machineinstrs < %s | FileCheck -enable-var-scope -check-prefix=GCN -check-prefix=GFX89 -check-prefix=GFX9 %s +; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=tahiti -verify-machineinstrs < %s | FileCheck -enable-var-scope -check-prefix=GCN -check-prefix=SI %s +; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=tonga -mattr=-flat-for-global -verify-machineinstrs < %s | FileCheck -enable-var-scope -check-prefix=GCN -check-prefix=GFX89 -check-prefix=GFX8 %s +; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=gfx901 -mattr=-flat-for-global -verify-machineinstrs < %s | FileCheck -enable-var-scope -check-prefix=GCN -check-prefix=GFX89 -check-prefix=GFX9 %s declare half @llvm.copysign.f16(half, half) declare float @llvm.copysign.f32(float, float) diff --git a/test/CodeGen/AMDGPU/fence-amdgiz.ll b/test/CodeGen/AMDGPU/fence-amdgiz.ll index 98e9d2485d4..3055f325f3f 100644 --- a/test/CodeGen/AMDGPU/fence-amdgiz.ll +++ b/test/CodeGen/AMDGPU/fence-amdgiz.ll @@ -1,7 +1,6 @@ -; RUN: llc < %s | FileCheck %s +; RUN: llc -mtriple=amdgcn-amd-amdhsa-amdgizcl -mcpu=kaveri < %s | FileCheck %s target datalayout = "e-p:64:64-p1:64:64-p2:64:64-p3:32:32-p4:32:32-p5:32:32-i64:64-v16:16-v24:32-v32:32-v48:64-v96:128-v192:256-v256:256-v512:512-v1024:1024-v2048:2048-n32:64-A5" -target triple = "amdgcn-amd-amdhsa-amdgizcl" ; CHECK-LABEL: atomic_fence ; CHECK: BB#0: diff --git a/test/CodeGen/AMDGPU/fmax3.f64.ll b/test/CodeGen/AMDGPU/fmax3.f64.ll index 8b9104b79e7..a56a5866aad 100644 --- a/test/CodeGen/AMDGPU/fmax3.f64.ll +++ b/test/CodeGen/AMDGPU/fmax3.f64.ll @@ -1,4 +1,4 @@ -; RUN: llc -march=amdgcn -verify-machineinstrs < %s | FileCheck -check-prefix=SI %s +; RUN: llc -march=amdgcn -mcpu=tahiti -verify-machineinstrs < %s | FileCheck -check-prefix=SI %s ; RUN: llc -march=amdgcn -mcpu=tonga -mattr=-flat-for-global -verify-machineinstrs < %s | FileCheck -check-prefix=SI %s declare double @llvm.maxnum.f64(double, double) nounwind readnone diff --git a/test/CodeGen/AMDGPU/fmul.f16.ll b/test/CodeGen/AMDGPU/fmul.f16.ll index 5f120f63d7f..fb8aeb80e15 100644 --- a/test/CodeGen/AMDGPU/fmul.f16.ll +++ b/test/CodeGen/AMDGPU/fmul.f16.ll @@ -1,5 +1,5 @@ -; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=SI %s -; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=fiji -mattr=-flat-for-global -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=VI %s +; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=tahiti -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=SI %s +; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=fiji -mattr=-flat-for-global -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=VI %s ; GCN-LABEL: {{^}}fmul_f16 ; GCN: buffer_load_ushort v[[A_F16:[0-9]+]] diff --git a/test/CodeGen/AMDGPU/fpext.f16.ll b/test/CodeGen/AMDGPU/fpext.f16.ll index ec19fd199b4..ae82973a7ac 100644 --- a/test/CodeGen/AMDGPU/fpext.f16.ll +++ b/test/CodeGen/AMDGPU/fpext.f16.ll @@ -1,6 +1,6 @@ -; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -verify-machineinstrs -enable-unsafe-fp-math < %s | FileCheck -check-prefix=GCN -check-prefix=SI -check-prefix=SIVI %s -; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=fiji -mattr=-flat-for-global -verify-machineinstrs -enable-unsafe-fp-math < %s | FileCheck -check-prefix=GCN -check-prefix=VI -check-prefix=GFX89 %s -; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=gfx901 -mattr=-flat-for-global -verify-machineinstrs -enable-unsafe-fp-math < %s | FileCheck -check-prefix=GCN -check-prefix=GFX9 -check-prefix=GFX89 %s +; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=tahiti -verify-machineinstrs -enable-unsafe-fp-math < %s | FileCheck -check-prefix=GCN -check-prefix=SI -check-prefix=SIVI %s +; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=fiji -mattr=-flat-for-global -verify-machineinstrs -enable-unsafe-fp-math < %s | FileCheck -check-prefix=GCN -check-prefix=VI -check-prefix=GFX89 %s +; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=gfx901 -mattr=-flat-for-global -verify-machineinstrs -enable-unsafe-fp-math < %s | FileCheck -check-prefix=GCN -check-prefix=GFX9 -check-prefix=GFX89 %s ; GCN-LABEL: {{^}}fpext_f16_to_f32 ; GCN: buffer_load_ushort v[[A_F16:[0-9]+]] diff --git a/test/CodeGen/AMDGPU/fptosi.f16.ll b/test/CodeGen/AMDGPU/fptosi.f16.ll index f593030764a..57d8c62ec7c 100644 --- a/test/CodeGen/AMDGPU/fptosi.f16.ll +++ b/test/CodeGen/AMDGPU/fptosi.f16.ll @@ -57,8 +57,8 @@ entry: ; SI: v_cvt_f32_f16_e32 v[[A_F32_0:[0-9]+]], v[[A_V2_F16]] ; SI: v_cvt_f32_f16_e32 v[[A_F32_1:[0-9]+]], v[[A_F16_1]] ; SI: v_cvt_i32_f32_e32 v[[R_I16_0:[0-9]+]], v[[A_F32_0]] -; SI: v_cvt_i32_f32_e32 v[[R_I16_1:[0-9]+]], v[[A_F32_1]] -; SI: v_and_b32_e32 v[[R_I16_LO:[0-9]+]], 0xffff, v[[R_I16_0]] +; SI-DAG: v_cvt_i32_f32_e32 v[[R_I16_1:[0-9]+]], v[[A_F32_1]] +; SI-DAG: v_and_b32_e32 v[[R_I16_LO:[0-9]+]], 0xffff, v[[R_I16_0]] ; SI: v_lshlrev_b32_e32 v[[R_I16_HI:[0-9]+]], 16, v[[R_I16_1]] ; SI: v_or_b32_e32 v[[R_V2_I16:[0-9]+]], v[[R_I16_LO]], v[[R_I16_HI]] diff --git a/test/CodeGen/AMDGPU/fptoui.f16.ll b/test/CodeGen/AMDGPU/fptoui.f16.ll index cebe3304d54..e2c8aacfb2a 100644 --- a/test/CodeGen/AMDGPU/fptoui.f16.ll +++ b/test/CodeGen/AMDGPU/fptoui.f16.ll @@ -1,5 +1,5 @@ -; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -verify-machineinstrs -enable-unsafe-fp-math < %s | FileCheck -check-prefix=GCN -check-prefix=SI %s -; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=fiji -mattr=-flat-for-global -verify-machineinstrs -enable-unsafe-fp-math < %s | FileCheck -check-prefix=GCN -check-prefix=VI %s +; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=tahiti -verify-machineinstrs -enable-unsafe-fp-math < %s | FileCheck -check-prefix=GCN -check-prefix=SI %s +; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=fiji -mattr=-flat-for-global -verify-machineinstrs -enable-unsafe-fp-math < %s | FileCheck -check-prefix=GCN -check-prefix=VI %s ; GCN-LABEL: {{^}}fptoui_f16_to_i16 ; GCN: buffer_load_ushort v[[A_F16:[0-9]+]] diff --git a/test/CodeGen/AMDGPU/frame-index-amdgiz.ll b/test/CodeGen/AMDGPU/frame-index-amdgiz.ll index dd46403b68a..47716a40ffe 100644 --- a/test/CodeGen/AMDGPU/frame-index-amdgiz.ll +++ b/test/CodeGen/AMDGPU/frame-index-amdgiz.ll @@ -1,4 +1,4 @@ -; RUN: llc -verify-machineinstrs < %s | FileCheck %s +; RUN: llc -mtriple=amdgcn---amdgiz -mcpu=kaveri -verify-machineinstrs < %s | FileCheck %s ; ; The original OpenCL kernel: ; kernel void f(global int *a, int i, int j) { @@ -6,10 +6,9 @@ ; x[i] = 7; ; a[0] = x[j]; ; } -; clang -cc1 -triple amdgcn---amdgizcl -emit-llvm -o - +; clang -cc1 -triple amdgcn---amdgizcl -emit-llvm -o - target datalayout = "e-p:64:64-p1:64:64-p2:64:64-p3:32:32-p4:32:32-p5:32:32-i64:64-v16:16-v24:32-v32:32-v48:64-v96:128-v192:256-v256:256-v512:512-v1024:1024-v2048:2048-n32:64-A5" -target triple = "amdgcn---amdgiz" define amdgpu_kernel void @f(i32 addrspace(1)* nocapture %a, i32 %i, i32 %j) local_unnamed_addr #0 { entry: diff --git a/test/CodeGen/AMDGPU/frame-index-elimination.ll b/test/CodeGen/AMDGPU/frame-index-elimination.ll index 61b09822f2b..b98428a49fb 100644 --- a/test/CodeGen/AMDGPU/frame-index-elimination.ll +++ b/test/CodeGen/AMDGPU/frame-index-elimination.ll @@ -1,4 +1,4 @@ -; RUN: llc -mtriple=amdgcn-amd-amdhsa -mattr=-promote-alloca -amdgpu-sroa=0 -verify-machineinstrs < %s | FileCheck -enable-var-scope -check-prefix=GCN %s +; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=kaveri -mattr=-promote-alloca -amdgpu-sroa=0 -verify-machineinstrs < %s | FileCheck -enable-var-scope -check-prefix=GCN %s ; Test that non-entry function frame indices are expanded properly to ; give an index relative to the scratch wave offset register diff --git a/test/CodeGen/AMDGPU/fsub.f16.ll b/test/CodeGen/AMDGPU/fsub.f16.ll index 15a4ce2d88f..5e3668a699f 100644 --- a/test/CodeGen/AMDGPU/fsub.f16.ll +++ b/test/CodeGen/AMDGPU/fsub.f16.ll @@ -1,6 +1,6 @@ -; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=SI -check-prefix=SIVI %s -; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=fiji -mattr=-flat-for-global -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=GFX89 -check-prefix=VI -check-prefix=SIVI %s -; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=gfx901 -mattr=-flat-for-global -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=GFX89 -check-prefix=GFX9 %s +; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=tahiti -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=SI -check-prefix=SIVI %s +; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=fiji -mattr=-flat-for-global -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=GFX89 -check-prefix=VI -check-prefix=SIVI %s +; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=gfx901 -mattr=-flat-for-global -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=GFX89 -check-prefix=GFX9 %s ; GCN-LABEL: {{^}}fsub_f16: ; GCN: buffer_load_ushort v[[A_F16:[0-9]+]] diff --git a/test/CodeGen/AMDGPU/half.ll b/test/CodeGen/AMDGPU/half.ll index 43745d4b3da..0df58519a62 100644 --- a/test/CodeGen/AMDGPU/half.ll +++ b/test/CodeGen/AMDGPU/half.ll @@ -1,5 +1,5 @@ -; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=SI %s -; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=tonga -mattr=-flat-for-global -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=VI %s +; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=tahiti -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=SI %s +; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=tonga -mattr=-flat-for-global -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=VI %s ; half args should be promoted to float for SI and lower. diff --git a/test/CodeGen/AMDGPU/indirect-addressing-si-noopt.ll b/test/CodeGen/AMDGPU/indirect-addressing-si-noopt.ll index 8e207a38c84..b2873402da4 100644 --- a/test/CodeGen/AMDGPU/indirect-addressing-si-noopt.ll +++ b/test/CodeGen/AMDGPU/indirect-addressing-si-noopt.ll @@ -1,4 +1,4 @@ -; RUN: llc -O0 -march=amdgcn -verify-machineinstrs < %s | FileCheck %s +; RUN: llc -O0 -march=amdgcn -mcpu=tahiti -verify-machineinstrs < %s | FileCheck %s ; FIXME: Merge into indirect-addressing-si.ll diff --git a/test/CodeGen/AMDGPU/indirect-addressing-si.ll b/test/CodeGen/AMDGPU/indirect-addressing-si.ll index 799be7fcf37..906a1f113cc 100644 --- a/test/CodeGen/AMDGPU/indirect-addressing-si.ll +++ b/test/CodeGen/AMDGPU/indirect-addressing-si.ll @@ -1,7 +1,7 @@ -; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=MOVREL %s -; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=tonga -mattr=-flat-for-global -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=MOVREL %s -; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=tonga -mattr=-flat-for-global -amdgpu-vgpr-index-mode -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=IDXMODE %s -; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=gfx900 -mattr=-flat-for-global -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=IDXMODE %s +; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=tahiti -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=MOVREL %s +; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=tonga -mattr=-flat-for-global -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=MOVREL %s +; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=tonga -mattr=-flat-for-global -amdgpu-vgpr-index-mode -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=IDXMODE %s +; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=gfx900 -mattr=-flat-for-global -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=IDXMODE %s ; Tests for indirect addressing on SI, which is implemented using dynamic ; indexing of vectors. diff --git a/test/CodeGen/AMDGPU/insert_vector_elt.ll b/test/CodeGen/AMDGPU/insert_vector_elt.ll index 6391b6b5407..463e7ba53dd 100644 --- a/test/CodeGen/AMDGPU/insert_vector_elt.ll +++ b/test/CodeGen/AMDGPU/insert_vector_elt.ll @@ -1,4 +1,4 @@ -; RUN: llc -verify-machineinstrs -march=amdgcn -mattr=+max-private-element-size-16 < %s | FileCheck -check-prefix=GCN -check-prefix=SI -check-prefix=GCN-NO-TONGA %s +; RUN: llc -verify-machineinstrs -march=amdgcn -mcpu=tahiti -mattr=+max-private-element-size-16 < %s | FileCheck -check-prefix=GCN -check-prefix=SI -check-prefix=GCN-NO-TONGA %s ; RUN: llc -verify-machineinstrs -march=amdgcn -mcpu=tonga -mattr=-flat-for-global -mattr=+max-private-element-size-16 < %s | FileCheck -check-prefix=GCN -check-prefix=SI -check-prefix=GCN-TONGA %s ; FIXME: Broken on evergreen diff --git a/test/CodeGen/AMDGPU/llvm.ceil.f16.ll b/test/CodeGen/AMDGPU/llvm.ceil.f16.ll index 071f2a6de4c..0f3cf79e257 100644 --- a/test/CodeGen/AMDGPU/llvm.ceil.f16.ll +++ b/test/CodeGen/AMDGPU/llvm.ceil.f16.ll @@ -1,5 +1,5 @@ -; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=SI %s -; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=fiji -mattr=-flat-for-global -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=VI %s +; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=tahiti -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=SI %s +; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=fiji -mattr=-flat-for-global -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=VI %s declare half @llvm.ceil.f16(half %a) declare <2 x half> @llvm.ceil.v2f16(<2 x half> %a) diff --git a/test/CodeGen/AMDGPU/llvm.cos.f16.ll b/test/CodeGen/AMDGPU/llvm.cos.f16.ll index 8931de63e74..16f167ae9e3 100644 --- a/test/CodeGen/AMDGPU/llvm.cos.f16.ll +++ b/test/CodeGen/AMDGPU/llvm.cos.f16.ll @@ -1,5 +1,5 @@ -; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=SI %s -; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=fiji -mattr=-flat-for-global -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=VI %s +; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=tahiti -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=SI %s +; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=fiji -mattr=-flat-for-global -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=VI %s declare half @llvm.cos.f16(half %a) declare <2 x half> @llvm.cos.v2f16(<2 x half> %a) diff --git a/test/CodeGen/AMDGPU/llvm.exp2.f16.ll b/test/CodeGen/AMDGPU/llvm.exp2.f16.ll index 4e96a761971..5764224ccd6 100644 --- a/test/CodeGen/AMDGPU/llvm.exp2.f16.ll +++ b/test/CodeGen/AMDGPU/llvm.exp2.f16.ll @@ -1,5 +1,5 @@ -; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=SI %s -; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=fiji -mattr=-flat-for-global -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=VI %s +; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=tahiti -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=SI %s +; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=fiji -mattr=-flat-for-global -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=VI %s declare half @llvm.exp2.f16(half %a) declare <2 x half> @llvm.exp2.v2f16(<2 x half> %a) diff --git a/test/CodeGen/AMDGPU/llvm.floor.f16.ll b/test/CodeGen/AMDGPU/llvm.floor.f16.ll index 74d1e694ffb..907fd488fa7 100644 --- a/test/CodeGen/AMDGPU/llvm.floor.f16.ll +++ b/test/CodeGen/AMDGPU/llvm.floor.f16.ll @@ -1,5 +1,5 @@ -; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=SI %s -; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=fiji -mattr=-flat-for-global -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=VI %s +; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=tahiti -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=SI %s +; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=fiji -mattr=-flat-for-global -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=VI %s declare half @llvm.floor.f16(half %a) declare <2 x half> @llvm.floor.v2f16(<2 x half> %a) diff --git a/test/CodeGen/AMDGPU/llvm.fma.f16.ll b/test/CodeGen/AMDGPU/llvm.fma.f16.ll index a379b18ffb8..738f4293414 100644 --- a/test/CodeGen/AMDGPU/llvm.fma.f16.ll +++ b/test/CodeGen/AMDGPU/llvm.fma.f16.ll @@ -1,5 +1,5 @@ -; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=SI %s -; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=fiji -mattr=-flat-for-global -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=VI %s +; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=tahiti -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=SI %s +; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=fiji -mattr=-flat-for-global -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=VI %s declare half @llvm.fma.f16(half %a, half %b, half %c) declare <2 x half> @llvm.fma.v2f16(<2 x half> %a, <2 x half> %b, <2 x half> %c) diff --git a/test/CodeGen/AMDGPU/llvm.fmuladd.f16.ll b/test/CodeGen/AMDGPU/llvm.fmuladd.f16.ll index 2d4fe08d8bd..add22dbfab5 100644 --- a/test/CodeGen/AMDGPU/llvm.fmuladd.f16.ll +++ b/test/CodeGen/AMDGPU/llvm.fmuladd.f16.ll @@ -1,7 +1,7 @@ -; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mattr=-fp64-fp16-denormals -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=SI -check-prefix=SI-FLUSH %s -; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=fiji -mattr=-fp64-fp16-denormals,-flat-for-global -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=VI -check-prefix=VI-FLUSH %s -; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mattr=+fp64-fp16-denormals -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=SI -check-prefix=SI-DENORM %s -; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=fiji -mattr=+fp64-fp16-denormals,-flat-for-global -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=VI -check-prefix=VI-DENORM %s +; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=tahiti -mattr=-fp64-fp16-denormals -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=SI -check-prefix=SI-FLUSH %s +; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=fiji -mattr=-fp64-fp16-denormals,-flat-for-global -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=VI -check-prefix=VI-FLUSH %s +; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=tahiti -mattr=+fp64-fp16-denormals -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=SI -check-prefix=SI-DENORM %s +; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=fiji -mattr=+fp64-fp16-denormals,-flat-for-global -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=VI -check-prefix=VI-DENORM %s declare half @llvm.fmuladd.f16(half %a, half %b, half %c) declare <2 x half> @llvm.fmuladd.v2f16(<2 x half> %a, <2 x half> %b, <2 x half> %c) diff --git a/test/CodeGen/AMDGPU/llvm.log2.f16.ll b/test/CodeGen/AMDGPU/llvm.log2.f16.ll index 277195c5320..7228c4044c7 100644 --- a/test/CodeGen/AMDGPU/llvm.log2.f16.ll +++ b/test/CodeGen/AMDGPU/llvm.log2.f16.ll @@ -1,5 +1,5 @@ -; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=SI %s -; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=fiji -mattr=-flat-for-global -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=VI %s +; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=tahiti -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=SI %s +; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=fiji -mattr=-flat-for-global -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=VI %s declare half @llvm.log2.f16(half %a) declare <2 x half> @llvm.log2.v2f16(<2 x half> %a) diff --git a/test/CodeGen/AMDGPU/llvm.maxnum.f16.ll b/test/CodeGen/AMDGPU/llvm.maxnum.f16.ll index c72716439a7..59a4f611bd3 100644 --- a/test/CodeGen/AMDGPU/llvm.maxnum.f16.ll +++ b/test/CodeGen/AMDGPU/llvm.maxnum.f16.ll @@ -1,5 +1,5 @@ -; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=SI %s -; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=fiji -mattr=-flat-for-global -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=VI %s +; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=tahiti -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=SI %s +; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=fiji -mattr=-flat-for-global -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=VI %s declare half @llvm.maxnum.f16(half %a, half %b) declare <2 x half> @llvm.maxnum.v2f16(<2 x half> %a, <2 x half> %b) diff --git a/test/CodeGen/AMDGPU/llvm.memcpy.ll b/test/CodeGen/AMDGPU/llvm.memcpy.ll index 7b4db55155e..4068c020e70 100644 --- a/test/CodeGen/AMDGPU/llvm.memcpy.ll +++ b/test/CodeGen/AMDGPU/llvm.memcpy.ll @@ -1,4 +1,4 @@ -; RUN: llc -march=amdgcn -verify-machineinstrs < %s | FileCheck -check-prefix=SI -check-prefix=FUNC %s +; RUN: llc -march=amdgcn -mcpu=tahiti -verify-machineinstrs < %s | FileCheck -check-prefix=SI -check-prefix=FUNC %s ; RUN: llc -march=amdgcn -mcpu=tonga -mattr=-flat-for-global -verify-machineinstrs < %s | FileCheck -check-prefix=SI -check-prefix=FUNC %s declare void @llvm.memcpy.p3i8.p3i8.i32(i8 addrspace(3)* nocapture, i8 addrspace(3)* nocapture, i32, i32, i1) nounwind diff --git a/test/CodeGen/AMDGPU/llvm.minnum.f16.ll b/test/CodeGen/AMDGPU/llvm.minnum.f16.ll index 0e93acc27dc..5fcbd1e0d85 100644 --- a/test/CodeGen/AMDGPU/llvm.minnum.f16.ll +++ b/test/CodeGen/AMDGPU/llvm.minnum.f16.ll @@ -1,5 +1,5 @@ -; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=SI %s -; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=fiji -mattr=-flat-for-global -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=VI %s +; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=tahiti -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=SI %s +; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=fiji -mattr=-flat-for-global -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=VI %s declare half @llvm.minnum.f16(half %a, half %b) declare <2 x half> @llvm.minnum.v2f16(<2 x half> %a, <2 x half> %b) diff --git a/test/CodeGen/AMDGPU/llvm.rint.f16.ll b/test/CodeGen/AMDGPU/llvm.rint.f16.ll index 92282083984..0e60f273129 100644 --- a/test/CodeGen/AMDGPU/llvm.rint.f16.ll +++ b/test/CodeGen/AMDGPU/llvm.rint.f16.ll @@ -1,6 +1,6 @@ -; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=SI -check-prefix=SIVI %s -; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=fiji -mattr=-flat-for-global -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=SIVI -check-prefix=VI -check-prefix=GFX89 %s -; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=gfx900 -mattr=-flat-for-global -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=GFX89 -check-prefix=GFX9 %s +; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=tahiti -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=SI -check-prefix=SIVI %s +; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=fiji -mattr=-flat-for-global -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=SIVI -check-prefix=VI -check-prefix=GFX89 %s +; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=gfx900 -mattr=-flat-for-global -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=GFX89 -check-prefix=GFX9 %s declare half @llvm.rint.f16(half %a) declare <2 x half> @llvm.rint.v2f16(<2 x half> %a) diff --git a/test/CodeGen/AMDGPU/llvm.round.ll b/test/CodeGen/AMDGPU/llvm.round.ll index 7e29147571f..48f1efb12fa 100644 --- a/test/CodeGen/AMDGPU/llvm.round.ll +++ b/test/CodeGen/AMDGPU/llvm.round.ll @@ -1,4 +1,4 @@ -; RUN: llc -march=amdgcn < %s | FileCheck -check-prefix=GCN -check-prefix=SI -check-prefix=FUNC %s +; RUN: llc -march=amdgcn -mcpu=tahiti < %s | FileCheck -check-prefix=GCN -check-prefix=SI -check-prefix=FUNC %s ; RUN: llc -march=amdgcn -mcpu=tonga -mattr=-flat-for-global < %s | FileCheck -check-prefix=GCN -check-prefix=GFX89 -check-prefix=FUNC %s ; RUN: llc -march=amdgcn -mcpu=gfx900 -mattr=-flat-for-global < %s | FileCheck -check-prefix=GCN -check-prefix=GFX89 -check-prefix=GFX9 -check-prefix=FUNC %s ; RUN: llc -march=r600 -mcpu=redwood < %s | FileCheck -check-prefix=R600 -check-prefix=FUNC %s diff --git a/test/CodeGen/AMDGPU/llvm.sin.f16.ll b/test/CodeGen/AMDGPU/llvm.sin.f16.ll index 08b9d9d873b..839ff0636f9 100644 --- a/test/CodeGen/AMDGPU/llvm.sin.f16.ll +++ b/test/CodeGen/AMDGPU/llvm.sin.f16.ll @@ -1,5 +1,5 @@ -; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=SI %s -; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=fiji -mattr=-flat-for-global -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=VI %s +; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=tahiti -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=SI %s +; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=fiji -mattr=-flat-for-global -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=VI %s declare half @llvm.sin.f16(half %a) declare <2 x half> @llvm.sin.v2f16(<2 x half> %a) diff --git a/test/CodeGen/AMDGPU/llvm.sqrt.f16.ll b/test/CodeGen/AMDGPU/llvm.sqrt.f16.ll index 0e1358ecca2..48c82512466 100644 --- a/test/CodeGen/AMDGPU/llvm.sqrt.f16.ll +++ b/test/CodeGen/AMDGPU/llvm.sqrt.f16.ll @@ -1,5 +1,5 @@ -; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=SI %s -; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=fiji -mattr=-flat-for-global -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=VI %s +; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=tahiti -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=SI %s +; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=fiji -mattr=-flat-for-global -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=VI %s declare half @llvm.sqrt.f16(half %a) declare <2 x half> @llvm.sqrt.v2f16(<2 x half> %a) diff --git a/test/CodeGen/AMDGPU/llvm.trunc.f16.ll b/test/CodeGen/AMDGPU/llvm.trunc.f16.ll index 37ee4e92c63..a8ff7cbb0de 100644 --- a/test/CodeGen/AMDGPU/llvm.trunc.f16.ll +++ b/test/CodeGen/AMDGPU/llvm.trunc.f16.ll @@ -1,5 +1,5 @@ -; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=SI %s -; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=fiji -mattr=-flat-for-global -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=VI %s +; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=tahiti -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=SI %s +; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=fiji -mattr=-flat-for-global -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=VI %s declare half @llvm.trunc.f16(half %a) declare <2 x half> @llvm.trunc.v2f16(<2 x half> %a) diff --git a/test/CodeGen/AMDGPU/macro-fusion-cluster-vcc-uses.mir b/test/CodeGen/AMDGPU/macro-fusion-cluster-vcc-uses.mir index 768acf35eea..2a431fe7946 100644 --- a/test/CodeGen/AMDGPU/macro-fusion-cluster-vcc-uses.mir +++ b/test/CodeGen/AMDGPU/macro-fusion-cluster-vcc-uses.mir @@ -1,4 +1,4 @@ -# RUN: llc -march=amdgcn -verify-machineinstrs -run-pass machine-scheduler -o - %s | FileCheck -check-prefix=GCN %s +# RUN: llc -march=amdgcn -mcpu=tahiti -verify-machineinstrs -run-pass machine-scheduler -o - %s | FileCheck -check-prefix=GCN %s # GCN-LABEL: name: cluster_add_addc # GCN: S_NOP 0, implicit-def %vcc diff --git a/test/CodeGen/AMDGPU/merge-stores.ll b/test/CodeGen/AMDGPU/merge-stores.ll index 59aa89f69ed..e26276beaf9 100644 --- a/test/CodeGen/AMDGPU/merge-stores.ll +++ b/test/CodeGen/AMDGPU/merge-stores.ll @@ -1,5 +1,5 @@ -; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -verify-machineinstrs -amdgpu-load-store-vectorizer=0 < %s | FileCheck -check-prefix=SI -check-prefix=GCN -check-prefix=GCN-AA %s -; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=bonaire -verify-machineinstrs -amdgpu-load-store-vectorizer=0 < %s | FileCheck -check-prefix=SI -check-prefix=GCN -check-prefix=GCN-AA %s +; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=tahiti -verify-machineinstrs -amdgpu-load-store-vectorizer=0 < %s | FileCheck -check-prefix=SI -check-prefix=GCN -check-prefix=GCN-AA %s +; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=bonaire -verify-machineinstrs -amdgpu-load-store-vectorizer=0 < %s | FileCheck -check-prefix=SI -check-prefix=GCN -check-prefix=GCN-AA %s ; This test is mostly to test DAG store merging, so disable the vectorizer. ; Run with devices with different unaligned load restrictions. diff --git a/test/CodeGen/AMDGPU/promote-alloca-mem-intrinsics.ll b/test/CodeGen/AMDGPU/promote-alloca-mem-intrinsics.ll index 7a4a451ff36..7343dd6bbda 100644 --- a/test/CodeGen/AMDGPU/promote-alloca-mem-intrinsics.ll +++ b/test/CodeGen/AMDGPU/promote-alloca-mem-intrinsics.ll @@ -1,4 +1,4 @@ -; RUN: opt -S -mtriple=amdgcn-unknown-amdhsa -amdgpu-promote-alloca < %s | FileCheck %s +; RUN: opt -S -mtriple=amdgcn-unknown-amdhsa -mcpu=kaveri -amdgpu-promote-alloca < %s | FileCheck %s declare void @llvm.memcpy.p0i8.p1i8.i32(i8* nocapture, i8 addrspace(1)* nocapture, i32, i32, i1) #0 declare void @llvm.memcpy.p1i8.p0i8.i32(i8 addrspace(1)* nocapture, i8* nocapture, i32, i32, i1) #0 diff --git a/test/CodeGen/AMDGPU/promote-alloca-no-opts.ll b/test/CodeGen/AMDGPU/promote-alloca-no-opts.ll index 9f22f207179..653e259e155 100644 --- a/test/CodeGen/AMDGPU/promote-alloca-no-opts.ll +++ b/test/CodeGen/AMDGPU/promote-alloca-no-opts.ll @@ -1,5 +1,5 @@ -; RUN: llc -O0 -march=amdgcn -mtriple=amdgcn-unknown-amdhsa -mattr=+promote-alloca < %s | FileCheck -check-prefix=NOOPTS -check-prefix=ALL %s -; RUN: llc -O1 -march=amdgcn -mtriple=amdgcn-unknown-amdhsa -mattr=+promote-alloca < %s | FileCheck -check-prefix=OPTS -check-prefix=ALL %s +; RUN: llc -O0 -march=amdgcn -mtriple=amdgcn-unknown-amdhsa -mcpu=kaveri -mattr=+promote-alloca < %s | FileCheck -check-prefix=NOOPTS -check-prefix=ALL %s +; RUN: llc -O1 -march=amdgcn -mtriple=amdgcn-unknown-amdhsa -mcpu=kaveri -mattr=+promote-alloca < %s | FileCheck -check-prefix=OPTS -check-prefix=ALL %s ; ALL-LABEL: {{^}}promote_alloca_i32_array_array: ; NOOPTS: workgroup_group_segment_byte_size = 0{{$}} diff --git a/test/CodeGen/AMDGPU/promote-alloca-padding-size-estimate.ll b/test/CodeGen/AMDGPU/promote-alloca-padding-size-estimate.ll index bf3bc493a4b..137b41df55c 100644 --- a/test/CodeGen/AMDGPU/promote-alloca-padding-size-estimate.ll +++ b/test/CodeGen/AMDGPU/promote-alloca-padding-size-estimate.ll @@ -1,4 +1,4 @@ -; RUN: llc -mtriple=amdgcn-amd-amdhsa < %s | FileCheck -check-prefix=GCN %s +; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=kaveri < %s | FileCheck -check-prefix=GCN %s ; This shows that the amount of LDS estimate is sensitive to the order ; of the LDS globals. diff --git a/test/CodeGen/AMDGPU/s_addk_i32.ll b/test/CodeGen/AMDGPU/s_addk_i32.ll index deef24cea37..ac7a33aff52 100644 --- a/test/CodeGen/AMDGPU/s_addk_i32.ll +++ b/test/CodeGen/AMDGPU/s_addk_i32.ll @@ -1,4 +1,4 @@ -; RUN: llc -march=amdgcn -verify-machineinstrs < %s | FileCheck -check-prefix=SI %s +; RUN: llc -march=amdgcn -mcpu=tahiti -verify-machineinstrs < %s | FileCheck -check-prefix=SI %s ; RUN: llc -march=amdgcn -mcpu=tonga -mattr=-flat-for-global -verify-machineinstrs < %s | FileCheck -check-prefix=SI %s ; SI-LABEL: {{^}}s_addk_i32_k0: diff --git a/test/CodeGen/AMDGPU/s_mulk_i32.ll b/test/CodeGen/AMDGPU/s_mulk_i32.ll index f6ed5408ba4..78f5e8afef4 100644 --- a/test/CodeGen/AMDGPU/s_mulk_i32.ll +++ b/test/CodeGen/AMDGPU/s_mulk_i32.ll @@ -1,4 +1,4 @@ -; RUN: llc -march=amdgcn -verify-machineinstrs < %s | FileCheck -check-prefix=SI %s +; RUN: llc -march=amdgcn -mcpu=tahiti -verify-machineinstrs < %s | FileCheck -check-prefix=SI %s ; RUN: llc -march=amdgcn -mcpu=tonga -mattr=-flat-for-global -verify-machineinstrs < %s | FileCheck -check-prefix=SI %s ; SI-LABEL: {{^}}s_mulk_i32_k0: diff --git a/test/CodeGen/AMDGPU/sad.ll b/test/CodeGen/AMDGPU/sad.ll index ee56e9053fd..5d981a225d7 100644 --- a/test/CodeGen/AMDGPU/sad.ll +++ b/test/CodeGen/AMDGPU/sad.ll @@ -1,4 +1,4 @@ -; RUN: llc -march=amdgcn -verify-machineinstrs < %s | FileCheck -check-prefix=GCN %s +; RUN: llc -march=amdgcn -mcpu=tahiti -verify-machineinstrs < %s | FileCheck -check-prefix=GCN %s ; GCN-LABEL: {{^}}v_sad_u32_pat1: ; GCN: v_sad_u32 v{{[0-9]+}}, s{{[0-9]+}}, v{{[0-9]+}}, v{{[0-9]+}} diff --git a/test/CodeGen/AMDGPU/schedule-kernel-arg-loads.ll b/test/CodeGen/AMDGPU/schedule-kernel-arg-loads.ll index 5c47c163dcc..fd34978dc07 100644 --- a/test/CodeGen/AMDGPU/schedule-kernel-arg-loads.ll +++ b/test/CodeGen/AMDGPU/schedule-kernel-arg-loads.ll @@ -1,4 +1,4 @@ -; RUN: llc -march=amdgcn -verify-machineinstrs < %s | FileCheck -check-prefix=FUNC -check-prefix=SI -check-prefix=GCN %s +; RUN: llc -march=amdgcn -mcpu=tahiti -verify-machineinstrs < %s | FileCheck -check-prefix=FUNC -check-prefix=SI -check-prefix=GCN %s ; RUN: llc -march=amdgcn -mcpu=tonga -mattr=-flat-for-global -verify-machineinstrs < %s | FileCheck -check-prefix=FUNC -check-prefix=VI -check-prefix=GCN %s ; FUNC-LABEL: {{^}}cluster_arg_loads: diff --git a/test/CodeGen/AMDGPU/schedule-regpressure-limit2.ll b/test/CodeGen/AMDGPU/schedule-regpressure-limit2.ll index 0d19c1e6a8f..cae713be2ea 100644 --- a/test/CodeGen/AMDGPU/schedule-regpressure-limit2.ll +++ b/test/CodeGen/AMDGPU/schedule-regpressure-limit2.ll @@ -1,5 +1,5 @@ -; RUN: llc -march=amdgcn -misched=gcn-minreg -verify-machineinstrs < %s | FileCheck -check-prefix=SI %s -; RUN: llc -march=amdgcn -misched=gcn-max-occupancy-experimental -verify-machineinstrs < %s | FileCheck -check-prefix=SI %s +; RUN: llc -march=amdgcn -mcpu=tahiti -misched=gcn-minreg -verify-machineinstrs < %s | FileCheck -check-prefix=SI %s +; RUN: llc -march=amdgcn -mcpu=tahiti -misched=gcn-max-occupancy-experimental -verify-machineinstrs < %s | FileCheck -check-prefix=SI %s ; RUN: llc -march=amdgcn -mcpu=fiji -misched=gcn-minreg -verify-machineinstrs < %s | FileCheck -check-prefix=VI %s ; RUN: llc -march=amdgcn -mcpu=fiji -misched=gcn-max-occupancy-experimental -verify-machineinstrs < %s | FileCheck -check-prefix=VI %s diff --git a/test/CodeGen/AMDGPU/select-vectors.ll b/test/CodeGen/AMDGPU/select-vectors.ll index b77ebcf5bf5..4f47bffaf4e 100644 --- a/test/CodeGen/AMDGPU/select-vectors.ll +++ b/test/CodeGen/AMDGPU/select-vectors.ll @@ -1,6 +1,6 @@ -; RUN: llc -amdgpu-scalarize-global-loads=false -verify-machineinstrs -march=amdgcn < %s | FileCheck -check-prefix=GCN -check-prefix=SI %s -; RUN: llc -amdgpu-scalarize-global-loads=false -verify-machineinstrs -march=amdgcn -mcpu=tonga -mattr=-flat-for-global < %s | FileCheck -check-prefix=GCN -check-prefix=VI %s -; RUN: llc -amdgpu-scalarize-global-loads=false -verify-machineinstrs -march=amdgcn -mcpu=gfx900 -mattr=-flat-for-global < %s | FileCheck -check-prefix=GCN -check-prefix=GFX9 %s +; RUN: llc -amdgpu-scalarize-global-loads=false -verify-machineinstrs -march=amdgcn -mcpu=tahiti < %s | FileCheck -check-prefix=GCN -check-prefix=SI %s +; RUN: llc -amdgpu-scalarize-global-loads=false -verify-machineinstrs -march=amdgcn -mcpu=tonga -mattr=-flat-for-global < %s | FileCheck -check-prefix=GCN -check-prefix=VI %s +; RUN: llc -amdgpu-scalarize-global-loads=false -verify-machineinstrs -march=amdgcn -mcpu=gfx900 -mattr=-flat-for-global < %s | FileCheck -check-prefix=GCN -check-prefix=GFX9 %s ; Test expansion of scalar selects on vectors. ; Evergreen not enabled since it seems to be having problems with doubles. diff --git a/test/CodeGen/AMDGPU/select.f16.ll b/test/CodeGen/AMDGPU/select.f16.ll index e79ce3af0cf..ce024edf510 100644 --- a/test/CodeGen/AMDGPU/select.f16.ll +++ b/test/CodeGen/AMDGPU/select.f16.ll @@ -1,5 +1,5 @@ -; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=SI %s -; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=fiji -mattr=-flat-for-global -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=VI %s +; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=tahiti -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=SI %s +; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=fiji -mattr=-flat-for-global -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=VI %s ; GCN-LABEL: {{^}}select_f16: ; GCN: buffer_load_ushort v[[A_F16:[0-9]+]] diff --git a/test/CodeGen/AMDGPU/setcc-fneg-constant.ll b/test/CodeGen/AMDGPU/setcc-fneg-constant.ll index bcaa1aa54c1..22827b2cc2b 100644 --- a/test/CodeGen/AMDGPU/setcc-fneg-constant.ll +++ b/test/CodeGen/AMDGPU/setcc-fneg-constant.ll @@ -1,4 +1,4 @@ -; RUN: llc -march=amdgcn -verify-machineinstrs < %s | FileCheck -check-prefix=SI -check-prefix=GCN -check-prefix=FUNC %s +; RUN: llc -march=amdgcn -mcpu=tahiti -verify-machineinstrs < %s | FileCheck -check-prefix=SI -check-prefix=GCN -check-prefix=FUNC %s ; Test fcmp pred (fneg x), c -> fcmp (swapped pred) x, -c combine. diff --git a/test/CodeGen/AMDGPU/shift-and-i128-ubfe.ll b/test/CodeGen/AMDGPU/shift-and-i128-ubfe.ll index 2f9eed457ab..72767d22efc 100644 --- a/test/CodeGen/AMDGPU/shift-and-i128-ubfe.ll +++ b/test/CodeGen/AMDGPU/shift-and-i128-ubfe.ll @@ -1,4 +1,4 @@ -; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -verify-machineinstrs < %s | FileCheck -check-prefix=GCN %s +; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=tahiti -verify-machineinstrs < %s | FileCheck -check-prefix=GCN %s ; Extract the high bit of the 1st quarter ; GCN-LABEL: {{^}}v_uextract_bit_31_i128: diff --git a/test/CodeGen/AMDGPU/shift-and-i64-ubfe.ll b/test/CodeGen/AMDGPU/shift-and-i64-ubfe.ll index 670287ba793..dae22ecbc66 100644 --- a/test/CodeGen/AMDGPU/shift-and-i64-ubfe.ll +++ b/test/CodeGen/AMDGPU/shift-and-i64-ubfe.ll @@ -1,4 +1,5 @@ -; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -verify-machineinstrs < %s | FileCheck -check-prefix=GCN %s +; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=tahiti -verify-machineinstrs < %s | FileCheck -check-prefix=GCN %s +; FIXME: Fails with -enable-var-scope ; Make sure 64-bit BFE pattern does a 32-bit BFE on the relevant half. diff --git a/test/CodeGen/AMDGPU/shl_add_constant.ll b/test/CodeGen/AMDGPU/shl_add_constant.ll index 9da4bc02801..26cdae3cbb0 100644 --- a/test/CodeGen/AMDGPU/shl_add_constant.ll +++ b/test/CodeGen/AMDGPU/shl_add_constant.ll @@ -1,4 +1,4 @@ -; RUN: llc -march=amdgcn -verify-machineinstrs < %s | FileCheck -check-prefix=SI %s +; RUN: llc -march=amdgcn -mcpu=tahiti -verify-machineinstrs < %s | FileCheck -check-prefix=SI %s declare i32 @llvm.amdgcn.workitem.id.x() #1 @@ -56,7 +56,7 @@ define amdgpu_kernel void @shl_2_add_999_i32(i32 addrspace(1)* %out, i32 addrspa ; FUNC-LABEL: {{^}}test_add_shl_add_constant: ; SI-DAG: s_load_dword [[X:s[0-9]+]], s{{\[[0-9]+:[0-9]+\]}}, 0xb ; SI-DAG: s_load_dword [[Y:s[0-9]+]], s{{\[[0-9]+:[0-9]+\]}}, 0xc -; SI: s_lshl_b32 [[SHL3:s[0-9]+]], [[X]], 3 +; SI-DAG: s_lshl_b32 [[SHL3:s[0-9]+]], [[X]], 3 ; SI: s_add_i32 [[RESULT:s[0-9]+]], [[SHL3]], [[Y]] ; SI: s_addk_i32 [[RESULT]], 0x3d8 ; SI: v_mov_b32_e32 [[VRESULT:v[0-9]+]], [[RESULT]] diff --git a/test/CodeGen/AMDGPU/sitofp.f16.ll b/test/CodeGen/AMDGPU/sitofp.f16.ll index 0bcef99df39..31d34663489 100644 --- a/test/CodeGen/AMDGPU/sitofp.f16.ll +++ b/test/CodeGen/AMDGPU/sitofp.f16.ll @@ -1,5 +1,5 @@ -; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -verify-machineinstrs -enable-unsafe-fp-math < %s | FileCheck -check-prefix=GCN -check-prefix=SI %s -; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=fiji -mattr=-flat-for-global -verify-machineinstrs -enable-unsafe-fp-math < %s | FileCheck -check-prefix=GCN -check-prefix=VI %s +; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=tahiti -verify-machineinstrs -enable-unsafe-fp-math < %s | FileCheck -check-prefix=GCN -check-prefix=SI %s +; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=fiji -mattr=-flat-for-global -verify-machineinstrs -enable-unsafe-fp-math < %s | FileCheck -check-prefix=GCN -check-prefix=VI %s ; GCN-LABEL: {{^}}sitofp_i16_to_f16 ; GCN: buffer_load_{{sshort|ushort}} v[[A_I16:[0-9]+]] diff --git a/test/CodeGen/AMDGPU/trunc.ll b/test/CodeGen/AMDGPU/trunc.ll index da038f4b059..b390958ab4c 100644 --- a/test/CodeGen/AMDGPU/trunc.ll +++ b/test/CodeGen/AMDGPU/trunc.ll @@ -1,6 +1,6 @@ -; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -verify-machineinstrs< %s | FileCheck -check-prefix=GCN -check-prefix=SI %s -; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=fiji -verify-machineinstrs< %s | FileCheck -check-prefix=GCN -check-prefix=VI %s -; RUN: llc -amdgpu-scalarize-global-loads=false -march=r600 -mcpu=cypress < %s | FileCheck -check-prefix=EG %s +; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=tahiti -verify-machineinstrs< %s | FileCheck -check-prefix=GCN -check-prefix=SI %s +; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=fiji -verify-machineinstrs< %s | FileCheck -check-prefix=GCN -check-prefix=VI %s +; RUN: llc -amdgpu-scalarize-global-loads=false -march=r600 -mcpu=cypress < %s | FileCheck -check-prefix=EG %s declare i32 @llvm.r600.read.tidig.x() nounwind readnone diff --git a/test/CodeGen/AMDGPU/udivrem.ll b/test/CodeGen/AMDGPU/udivrem.ll index 9507a49cfc8..b89f3268976 100644 --- a/test/CodeGen/AMDGPU/udivrem.ll +++ b/test/CodeGen/AMDGPU/udivrem.ll @@ -1,4 +1,4 @@ -; RUN: llc -march=amdgcn -verify-machineinstrs < %s | FileCheck --check-prefix=SI --check-prefix=FUNC %s +; RUN: llc -march=amdgcn -mcpu=tahiti -verify-machineinstrs < %s | FileCheck --check-prefix=SI --check-prefix=FUNC %s ; RUN: llc -march=amdgcn -mcpu=tonga -mattr=-flat-for-global -verify-machineinstrs < %s | FileCheck --check-prefix=SI --check-prefix=FUNC %s ; RUN: llc -march=r600 -mcpu=redwood < %s | FileCheck --check-prefix=EG --check-prefix=FUNC %s diff --git a/test/CodeGen/AMDGPU/uitofp.f16.ll b/test/CodeGen/AMDGPU/uitofp.f16.ll index eaa1d073caf..0e7250fadbb 100644 --- a/test/CodeGen/AMDGPU/uitofp.f16.ll +++ b/test/CodeGen/AMDGPU/uitofp.f16.ll @@ -1,5 +1,5 @@ -; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -verify-machineinstrs -enable-unsafe-fp-math < %s | FileCheck -check-prefix=GCN -check-prefix=SI %s -; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=fiji -mattr=-flat-for-global -verify-machineinstrs -enable-unsafe-fp-math < %s | FileCheck -check-prefix=GCN -check-prefix=VI %s +; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=tahiti -verify-machineinstrs -enable-unsafe-fp-math < %s | FileCheck -check-prefix=GCN -check-prefix=SI %s +; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=fiji -mattr=-flat-for-global -verify-machineinstrs -enable-unsafe-fp-math < %s | FileCheck -check-prefix=GCN -check-prefix=VI %s ; GCN-LABEL: {{^}}uitofp_i16_to_f16 ; GCN: buffer_load_ushort v[[A_I16:[0-9]+]] diff --git a/test/CodeGen/AMDGPU/v_mac_f16.ll b/test/CodeGen/AMDGPU/v_mac_f16.ll index 46c9b7ee1a3..fe5e5a18ab5 100644 --- a/test/CodeGen/AMDGPU/v_mac_f16.ll +++ b/test/CodeGen/AMDGPU/v_mac_f16.ll @@ -1,4 +1,4 @@ -; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mattr=-fp64-fp16-denormals -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=SI %s +; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=tahiti -mattr=-fp64-fp16-denormals -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=SI %s ; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=fiji -mattr=-fp64-fp16-denormals,-flat-for-global -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=VI %s ; GCN-LABEL: {{^}}mac_f16: diff --git a/test/CodeGen/AMDGPU/vector-extract-insert.ll b/test/CodeGen/AMDGPU/vector-extract-insert.ll index ab2bfcfd1fb..e3027d09309 100644 --- a/test/CodeGen/AMDGPU/vector-extract-insert.ll +++ b/test/CodeGen/AMDGPU/vector-extract-insert.ll @@ -1,4 +1,4 @@ -; RUN: llc -march=amdgcn -verify-machineinstrs < %s | FileCheck -check-prefix=GCN %s +; RUN: llc -march=amdgcn -mcpu=tahiti -verify-machineinstrs < %s | FileCheck -check-prefix=GCN %s ; Test that when extracting the same unknown vector index from an ; insertelement the dynamic indexing is folded away. @@ -81,4 +81,4 @@ define amdgpu_kernel void @extract_insert_same_dynelt_v4f32(float addrspace(1)* } attributes #0 = { nounwind readnone } -attributes #1 = { nounwind } \ No newline at end of file +attributes #1 = { nounwind }