diff options
| author | Bruno Cardoso Lopes <bruno.cardoso@gmail.com> | 2011-07-22 20:53:20 +0000 |
|---|---|---|
| committer | Bruno Cardoso Lopes <bruno.cardoso@gmail.com> | 2011-07-22 20:53:20 +0000 |
| commit | dad38638e172a4c53caf1d9dac9b92228bd4be7a (patch) | |
| tree | c8e3167061d84cacf2d6484423ead5ae3ce3d2ec | |
| parent | 44a456332f1f41d1e0b2815d93e47a88d501ee6e (diff) | |
| download | external_llvm-dad38638e172a4c53caf1d9dac9b92228bd4be7a.zip external_llvm-dad38638e172a4c53caf1d9dac9b92228bd4be7a.tar.gz external_llvm-dad38638e172a4c53caf1d9dac9b92228bd4be7a.tar.bz2 | |
Fix PR10422 by adding the necessary AVX UCOMISD memory versions to
load folding logic
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@135801 91177308-0d34-0410-b5e6-96231b3b80d8
| -rw-r--r-- | lib/Target/X86/X86InstrInfo.cpp | 10 | ||||
| -rw-r--r-- | test/CodeGen/X86/avx-cmp-fp.ll | 28 |
2 files changed, 36 insertions, 2 deletions
diff --git a/lib/Target/X86/X86InstrInfo.cpp b/lib/Target/X86/X86InstrInfo.cpp index 8dc6822..990ceef 100644 --- a/lib/Target/X86/X86InstrInfo.cpp +++ b/lib/Target/X86/X86InstrInfo.cpp @@ -407,6 +407,8 @@ X86InstrInfo::X86InstrInfo(X86TargetMachine &tm) { X86::Int_CVTTSS2SIrr, X86::Int_CVTTSS2SIrm, 0 }, { X86::Int_UCOMISDrr, X86::Int_UCOMISDrm, 0 }, { X86::Int_UCOMISSrr, X86::Int_UCOMISSrm, 0 }, + { X86::Int_VUCOMISDrr, X86::Int_VUCOMISDrm, 0 }, + { X86::Int_VUCOMISSrr, X86::Int_VUCOMISSrm, 0 }, { X86::MOV16rr, X86::MOV16rm, 0 }, { X86::MOV32rr, X86::MOV32rm, 0 }, { X86::MOV64rr, X86::MOV64rm, 0 }, @@ -467,7 +469,9 @@ X86InstrInfo::X86InstrInfo(X86TargetMachine &tm) { X86::TEST8rr, X86::TEST8rm, 0 }, // FIXME: TEST*rr EAX,EAX ---> CMP [mem], 0 { X86::UCOMISDrr, X86::UCOMISDrm, 0 }, - { X86::UCOMISSrr, X86::UCOMISSrm, 0 } + { X86::UCOMISSrr, X86::UCOMISSrm, 0 }, + { X86::VUCOMISDrr, X86::VUCOMISDrm, 0 }, + { X86::VUCOMISSrr, X86::VUCOMISSrm, 0 } }; for (unsigned i = 0, e = array_lengthof(OpTbl1); i != e; ++i) { @@ -2491,7 +2495,9 @@ MachineInstr* X86InstrInfo::foldMemoryOperandImpl(MachineFunction &MF, case X86::AVX_SET0PSY: case X86::AVX_SET0PDY: case X86::FsFLD0SD: - case X86::FsFLD0SS: { + case X86::FsFLD0SS: + case X86::VFsFLD0SD: + case X86::VFsFLD0SS: { // Folding a V_SET0P? or V_SETALLONES as a load, to ease register pressure. // Create a constant-pool entry and operands to load from it. diff --git a/test/CodeGen/X86/avx-cmp-fp.ll b/test/CodeGen/X86/avx-cmp-fp.ll new file mode 100644 index 0000000..18d381b --- /dev/null +++ b/test/CodeGen/X86/avx-cmp-fp.ll @@ -0,0 +1,28 @@ +; RUN: llc < %s -mtriple=x86_64-apple-darwin -mcpu=corei7-avx -mattr=+avx | FileCheck %s + +declare void @scale() nounwind uwtable + +; CHECK: vucomisd .LCPI +define void @render() nounwind uwtable { +entry: + br i1 undef, label %for.cond5, label %for.end52 + +for.cond5: + %or.cond = and i1 undef, false + br i1 %or.cond, label %for.body33, label %for.cond5 + +for.cond30: + br i1 false, label %for.body33, label %for.cond5 + +for.body33: + %tobool = fcmp une double undef, 0.000000e+00 + br i1 %tobool, label %if.then, label %for.cond30 + +if.then: + call void @scale() + br label %for.cond30 + +for.end52: + ret void +} + |