From eb95cec176ad4154ac4ed509baf1c9c51ca29be0 Mon Sep 17 00:00:00 2001 From: Arnold Schwaighofer Date: Tue, 7 May 2013 04:37:05 +0000 Subject: [PATCH] LoopVectorize: getConsecutiveVector must respect signed arithmetic We were passing an i32 to ConstantInt::get where an i64 was needed and we must also pass the sign if we pass negatives numbers. The start index passed to getConsecutiveVector must also be signed. Should fix PR15882. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@181286 91177308-0d34-0410-b5e6-96231b3b80d8 --- lib/Transforms/Vectorize/LoopVectorize.cpp | 11 +-- .../LoopVectorize/reverse_induction.ll | 79 +++++++++++++++++++ 2 files changed, 85 insertions(+), 5 deletions(-) create mode 100644 test/Transforms/LoopVectorize/reverse_induction.ll diff --git a/lib/Transforms/Vectorize/LoopVectorize.cpp b/lib/Transforms/Vectorize/LoopVectorize.cpp index 491e9cc2480..ad4a6c7c419 100644 --- a/lib/Transforms/Vectorize/LoopVectorize.cpp +++ b/lib/Transforms/Vectorize/LoopVectorize.cpp @@ -216,7 +216,7 @@ private: /// This function adds 0, 1, 2 ... to each vector element, starting at zero. /// If Negate is set then negative numbers are added e.g. (0, -1, -2, ...). /// The sequence starts at StartIndex. - Value *getConsecutiveVector(Value* Val, unsigned StartIdx, bool Negate); + Value *getConsecutiveVector(Value* Val, int StartIdx, bool Negate); /// When we go over instructions in the basic block we rely on previous /// values within the current basic block or on loop invariant values. @@ -829,7 +829,7 @@ Value *InnerLoopVectorizer::getBroadcastInstrs(Value *V) { return Shuf; } -Value *InnerLoopVectorizer::getConsecutiveVector(Value* Val, unsigned StartIdx, +Value *InnerLoopVectorizer::getConsecutiveVector(Value* Val, int StartIdx, bool Negate) { assert(Val->getType()->isVectorTy() && "Must be a vector"); assert(Val->getType()->getScalarType()->isIntegerTy() && @@ -842,8 +842,8 @@ Value *InnerLoopVectorizer::getConsecutiveVector(Value* Val, unsigned StartIdx, // Create a vector of consecutive numbers from zero to VF. for (int i = 0; i < VLen; ++i) { - int Idx = Negate ? (-i): i; - Indices.push_back(ConstantInt::get(ITy, StartIdx + Idx)); + int64_t Idx = Negate ? (-i) : i; + Indices.push_back(ConstantInt::get(ITy, StartIdx + Idx, Negate)); } // Add the consecutive indices to the vector value. @@ -2072,7 +2072,8 @@ InnerLoopVectorizer::vectorizeBlockInLoop(LoopVectorizationLegality *Legal, // After broadcasting the induction variable we need to make the // vector consecutive by adding ... -3, -2, -1, 0. for (unsigned part = 0; part < UF; ++part) - Entry[part] = getConsecutiveVector(Broadcasted, -VF * part, true); + Entry[part] = getConsecutiveVector(Broadcasted, -(int)VF * part, + true); continue; } diff --git a/test/Transforms/LoopVectorize/reverse_induction.ll b/test/Transforms/LoopVectorize/reverse_induction.ll new file mode 100644 index 00000000000..f43f02bc313 --- /dev/null +++ b/test/Transforms/LoopVectorize/reverse_induction.ll @@ -0,0 +1,79 @@ +; RUN: opt < %s -loop-vectorize -force-vector-unroll=2 -force-vector-width=4 -S | FileCheck %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" + +; Make sure consecutive vector generates correct negative indices. +; PR15882 + +; CHECK: reverse_induction_i64 +; CHECK: add <4 x i64> %[[SPLAT:.*]], +; CHECK: add <4 x i64> %[[SPLAT]], + +define i32 @reverse_induction_i64(i64 %startval, i32 * %ptr) { +entry: + br label %for.body + +for.body: + %add.i7 = phi i64 [ %startval, %entry ], [ %add.i, %for.body ] + %i.06 = phi i32 [ 0, %entry ], [ %inc4, %for.body ] + %redux5 = phi i32 [ 0, %entry ], [ %inc.redux, %for.body ] + %add.i = add i64 %add.i7, -1 + %kind_.i = getelementptr inbounds i32* %ptr, i64 %add.i + %tmp.i1 = load i32* %kind_.i, align 4 + %inc.redux = add i32 %tmp.i1, %redux5 + %inc4 = add i32 %i.06, 1 + %exitcond = icmp ne i32 %inc4, 1024 + br i1 %exitcond, label %for.body, label %loopend + +loopend: + ret i32 %inc.redux +} + +; CHECK: reverse_induction_i128 +; CHECK: add <4 x i128> %[[SPLAT:.*]], +; CHECK: add <4 x i128> %[[SPLAT]], +define i32 @reverse_induction_i128(i128 %startval, i32 * %ptr) { +entry: + br label %for.body + +for.body: + %add.i7 = phi i128 [ %startval, %entry ], [ %add.i, %for.body ] + %i.06 = phi i32 [ 0, %entry ], [ %inc4, %for.body ] + %redux5 = phi i32 [ 0, %entry ], [ %inc.redux, %for.body ] + %add.i = add i128 %add.i7, -1 + %kind_.i = getelementptr inbounds i32* %ptr, i128 %add.i + %tmp.i1 = load i32* %kind_.i, align 4 + %inc.redux = add i32 %tmp.i1, %redux5 + %inc4 = add i32 %i.06, 1 + %exitcond = icmp ne i32 %inc4, 1024 + br i1 %exitcond, label %for.body, label %loopend + +loopend: + ret i32 %inc.redux +} + +; CHECK: reverse_induction_i16 +; CHECK: add <4 x i16> %[[SPLAT:.*]], +; CHECK: add <4 x i16> %[[SPLAT]], + +define i32 @reverse_induction_i16(i16 %startval, i32 * %ptr) { +entry: + br label %for.body + +for.body: + %add.i7 = phi i16 [ %startval, %entry ], [ %add.i, %for.body ] + %i.06 = phi i32 [ 0, %entry ], [ %inc4, %for.body ] + %redux5 = phi i32 [ 0, %entry ], [ %inc.redux, %for.body ] + %add.i = add i16 %add.i7, -1 + %kind_.i = getelementptr inbounds i32* %ptr, i16 %add.i + %tmp.i1 = load i32* %kind_.i, align 4 + %inc.redux = add i32 %tmp.i1, %redux5 + %inc4 = add i32 %i.06, 1 + %exitcond = icmp ne i32 %inc4, 1024 + br i1 %exitcond, label %for.body, label %loopend + +loopend: + ret i32 %inc.redux +} + +