From df9c9da88402e393343d5e199b0ebaa681b3a1ce Mon Sep 17 00:00:00 2001 From: Arnold Schwaighofer Date: Tue, 3 Dec 2013 16:33:06 +0000 Subject: [PATCH] opt: Mirror vectorization presets of clang clang enables vectorization at optimization levels > 1 and size level < 2. opt should behave similarily. Loop vectorization and SLP vectorization can be disabled with the flags -disable-(loop/slp)-vectorization. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@196294 91177308-0d34-0410-b5e6-96231b3b80d8 --- test/Transforms/LoopVectorize/opt.ll | 28 ++++++++++++++++++++++ test/Transforms/SLPVectorizer/X86/opt.ll | 30 ++++++++++++++++++++++++ tools/opt/opt.cpp | 16 +++++++++++-- 3 files changed, 72 insertions(+), 2 deletions(-) create mode 100644 test/Transforms/LoopVectorize/opt.ll create mode 100644 test/Transforms/SLPVectorizer/X86/opt.ll diff --git a/test/Transforms/LoopVectorize/opt.ll b/test/Transforms/LoopVectorize/opt.ll new file mode 100644 index 00000000000..27030a2ff2a --- /dev/null +++ b/test/Transforms/LoopVectorize/opt.ll @@ -0,0 +1,28 @@ +; RUN: opt -S -O3 -force-vector-width=2 -force-vector-unroll=1 < %s | FileCheck --check-prefix=LOOPVEC %s +; RUN: opt -S -O3 -disable-loop-vectorization -force-vector-width=2 -force-vector-unroll=1 < %s | FileCheck --check-prefix=NOLOOPVEC %s + +target datalayout = "e-p:64:64:64-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:64:64-f32:32:32-f64:64:64-v64:64:64-v128:128:128-a0:0:64-s0:64:64-f80:128:128-n8:16:32:64-S128" +target triple = "x86_64-apple-macosx10.8.0" + +; Make sure we can disable vectorization in opt. + +; LOOPVEC: add <2 x i32> +; NOLOOPVEC-NOT: add <2 x i32> + +define i32 @vect(i32* %a) { +entry: + br label %for.body + +for.body: + %indvars.iv = phi i64 [ 0, %entry ], [ %indvars.iv.next, %for.body ] + %red.05 = phi i32 [ 0, %entry ], [ %add, %for.body ] + %arrayidx = getelementptr inbounds i32* %a, i64 %indvars.iv + %0 = load i32* %arrayidx, align 4 + %add = add nsw i32 %0, %red.05 + %indvars.iv.next = add nuw nsw i64 %indvars.iv, 1 + %exitcond = icmp eq i64 %indvars.iv.next, 255 + br i1 %exitcond, label %for.end, label %for.body + +for.end: + ret i32 %add +} diff --git a/test/Transforms/SLPVectorizer/X86/opt.ll b/test/Transforms/SLPVectorizer/X86/opt.ll new file mode 100644 index 00000000000..14137c11ee4 --- /dev/null +++ b/test/Transforms/SLPVectorizer/X86/opt.ll @@ -0,0 +1,30 @@ +; RUN: opt < %s -O3 -S -mtriple=x86_64-apple-macosx10.8.0 -mcpu=corei7-avx | FileCheck %s --check-prefix=SLP +; RUN: opt < %s -O3 -disable-slp-vectorization -S -mtriple=x86_64-apple-macosx10.8.0 -mcpu=corei7-avx | FileCheck %s --check-prefix=NOSLP + +target datalayout = "e-p:64:64:64-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:64:64-f32:32:32-f64:64:64-v64:64:64-v128:128:128-a0:0:64-s0:64:64-f80:128:128-n8:16:32:64-S128" +target triple = "x86_64-apple-macosx10.8.0" + +; Make sure we can disable slp vectorization in opt. + +; SLP-LABEL: test1 +; SLP: store <2 x double> + +; NOSLP-LABEL: test1 +; NOSLP-NOT: store <2 x double> + + +define void @test1(double* %a, double* %b, double* %c) { +entry: + %i0 = load double* %a, align 8 + %i1 = load double* %b, align 8 + %mul = fmul double %i0, %i1 + %arrayidx3 = getelementptr inbounds double* %a, i64 1 + %i3 = load double* %arrayidx3, align 8 + %arrayidx4 = getelementptr inbounds double* %b, i64 1 + %i4 = load double* %arrayidx4, align 8 + %mul5 = fmul double %i3, %i4 + store double %mul, double* %c, align 8 + %arrayidx5 = getelementptr inbounds double* %c, i64 1 + store double %mul5, double* %arrayidx5, align 8 + ret void +} diff --git a/tools/opt/opt.cpp b/tools/opt/opt.cpp index 8a8da01a99e..dba16f72dab 100644 --- a/tools/opt/opt.cpp +++ b/tools/opt/opt.cpp @@ -139,6 +139,16 @@ static cl::opt DisableLoopUnrolling("disable-loop-unrolling", cl::desc("Disable loop unrolling in all relevant passes"), cl::init(false)); +static cl::opt +DisableLoopVectorization("disable-loop-vectorization", + cl::desc("Disable the loop vectorization pass"), + cl::init(false)); + +static cl::opt +DisableSLPVectorization("disable-slp-vectorization", + cl::desc("Disable the slp vectorization pass"), + cl::init(false)); + static cl::opt DisableSimplifyLibCalls("disable-simplify-libcalls", @@ -461,8 +471,10 @@ static void AddOptimizationPasses(PassManagerBase &MPM,FunctionPassManager &FPM, Builder.DisableUnrollLoops = (DisableLoopUnrolling.getNumOccurrences() > 0) ? DisableLoopUnrolling : OptLevel == 0; - Builder.LoopVectorize = OptLevel > 1 && SizeLevel < 2; - Builder.SLPVectorize = true; + Builder.LoopVectorize = + DisableLoopVectorization ? false : OptLevel > 1 && SizeLevel < 2; + Builder.SLPVectorize = + DisableSLPVectorization ? false : OptLevel > 1 && SizeLevel < 2; Builder.populateFunctionPassManager(FPM); Builder.populateModulePassManager(MPM); -- 2.34.1