diff --git a/lib/Transforms/Vectorize/LoopVectorize.cpp b/lib/Transforms/Vectorize/LoopVectorize.cpp index 5a79c33bdc8..1b6011bd1a9 100644 --- a/lib/Transforms/Vectorize/LoopVectorize.cpp +++ b/lib/Transforms/Vectorize/LoopVectorize.cpp @@ -740,10 +740,15 @@ SingleBlockLoopVectorizer::vectorizeLoop(LoopVectorizationLegality *Legal) { break; } + // The last index does not have to be the induction. It can be + // consecutive and be a function of the index. For example A[I+1]; + unsigned NumOperands = Gep->getNumOperands(); + Value *LastIndex = getVectorValue(Gep->getOperand(NumOperands -1)); + LastIndex = Builder.CreateExtractElement(LastIndex, Builder.getInt32(0)); + // Create the new GEP with the new induction variable. GetElementPtrInst *Gep2 = cast(Gep->clone()); - unsigned NumOperands = Gep->getNumOperands(); - Gep2->setOperand(NumOperands - 1, Induction); + Gep2->setOperand(NumOperands - 1, LastIndex); Ptr = Builder.Insert(Gep2); Ptr = Builder.CreateBitCast(Ptr, StTy->getPointerTo()); Value *Val = getVectorValue(SI->getValueOperand()); @@ -764,10 +769,15 @@ SingleBlockLoopVectorizer::vectorizeLoop(LoopVectorizationLegality *Legal) { break; } + // The last index does not have to be the induction. It can be + // consecutive and be a function of the index. For example A[I+1]; + unsigned NumOperands = Gep->getNumOperands(); + Value *LastIndex = getVectorValue(Gep->getOperand(NumOperands -1)); + LastIndex = Builder.CreateExtractElement(LastIndex, Builder.getInt32(0)); + // Create the new GEP with the new induction variable. GetElementPtrInst *Gep2 = cast(Gep->clone()); - unsigned NumOperands = Gep->getNumOperands(); - Gep2->setOperand(NumOperands - 1, Induction); + Gep2->setOperand(NumOperands - 1, LastIndex); Ptr = Builder.Insert(Gep2); Ptr = Builder.CreateBitCast(Ptr, RetTy->getPointerTo()); LI = Builder.CreateLoad(Ptr); diff --git a/test/Transforms/LoopVectorize/induction_plus.ll b/test/Transforms/LoopVectorize/induction_plus.ll new file mode 100644 index 00000000000..bd90113e523 --- /dev/null +++ b/test/Transforms/LoopVectorize/induction_plus.ll @@ -0,0 +1,30 @@ +; RUN: opt < %s -loop-vectorize -instcombine -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" +target triple = "x86_64-apple-macosx10.8.0" + +@array = common global [1024 x i32] zeroinitializer, align 16 + +;CHECK: @array_at_plus_one +;CHECK: add <4 x i64> +;CHECK: trunc <4 x i64> +;CHECK: add i64 %index, 12 +;CHECK: ret i32 +define i32 @array_at_plus_one(i32 %n) nounwind uwtable ssp { + %1 = icmp sgt i32 %n, 0 + br i1 %1, label %.lr.ph, label %._crit_edge + +.lr.ph: ; preds = %0, %.lr.ph + %indvars.iv = phi i64 [ %indvars.iv.next, %.lr.ph ], [ 0, %0 ] + %2 = add nsw i64 %indvars.iv, 12 + %3 = getelementptr inbounds [1024 x i32]* @array, i64 0, i64 %2 + %4 = trunc i64 %indvars.iv to i32 + store i32 %4, i32* %3, align 4 + %indvars.iv.next = add i64 %indvars.iv, 1 + %lftr.wideiv = trunc i64 %indvars.iv.next to i32 + %exitcond = icmp eq i32 %lftr.wideiv, %n + br i1 %exitcond, label %._crit_edge, label %.lr.ph + +._crit_edge: ; preds = %.lr.ph, %0 + ret i32 undef +}