Fix: Correct the vector pointer when UF>1

Author: Mel-Chen

llvm/lib/Transforms/Vectorize/VPlanRecipes.cpp

@@ -2125,11 +2125,14 @@ void VPVectorPointerRecipe::execute(VPTransformState &State) {
                                 CurrentPart, Builder);
   Value *Ptr = State.get(getOperand(0), VPLane(0));
 
+  Value *Increment = createStepForVF(Builder, IndexTy, State.VF, CurrentPart);
   // TODO: Support non-unit-reverse strided accesses.
-  int64_t Step = Strided ? -1 * CurrentPart : CurrentPart;
-  Value *Increment = createStepForVF(Builder, IndexTy, State.VF, Step);
+  Value *Index =
+      Strided
+          ? Builder.CreateMul(Increment, ConstantInt::getSigned(IndexTy, -1))
+          : Increment;
   Value *ResultPtr =
-      Builder.CreateGEP(IndexedTy, Ptr, Increment, "", getGEPNoWrapFlags());
+      Builder.CreateGEP(IndexedTy, Ptr, Index, "", getGEPNoWrapFlags());
 
   State.set(this, ResultPtr, /*IsScalar*/ true);
 }

llvm/test/Transforms/LoopVectorize/RISCV/riscv-vector-reverse-output.ll

@@ -213,7 +213,8 @@ define void @vector_reverse_i64(ptr noalias %A, ptr noalias %B, i32 %n) {
 ; RV64-UF2-NEXT:    [[TMP21:%.*]] = getelementptr inbounds i32, ptr [[B]], i64 [[TMP20]]
 ; RV64-UF2-NEXT:    [[TMP23:%.*]] = getelementptr inbounds i32, ptr [[TMP21]], i32 0
 ; RV64-UF2-NEXT:    [[TMP22:%.*]] = call i64 @llvm.vscale.i64()
-; RV64-UF2-NEXT:    [[TMP26:%.*]] = mul i64 [[TMP22]], 4294967292
+; RV64-UF2-NEXT:    [[TMP24:%.*]] = mul i64 [[TMP22]], 4
+; RV64-UF2-NEXT:    [[TMP26:%.*]] = mul i64 [[TMP24]], -1
 ; RV64-UF2-NEXT:    [[TMP28:%.*]] = getelementptr inbounds i32, ptr [[TMP21]], i64 [[TMP26]]
 ; RV64-UF2-NEXT:    [[TMP25:%.*]] = call i32 @llvm.vscale.i32()
 ; RV64-UF2-NEXT:    [[TMP29:%.*]] = mul i32 [[TMP25]], 4
@@ -226,7 +227,8 @@ define void @vector_reverse_i64(ptr noalias %A, ptr noalias %B, i32 %n) {
 ; RV64-UF2-NEXT:    [[TMP32:%.*]] = getelementptr inbounds i32, ptr [[A]], i64 [[TMP20]]
 ; RV64-UF2-NEXT:    [[TMP35:%.*]] = getelementptr inbounds i32, ptr [[TMP32]], i32 0
 ; RV64-UF2-NEXT:    [[TMP33:%.*]] = call i64 @llvm.vscale.i64()
-; RV64-UF2-NEXT:    [[TMP37:%.*]] = mul i64 [[TMP33]], 4294967292
+; RV64-UF2-NEXT:    [[TMP44:%.*]] = mul i64 [[TMP33]], 4
+; RV64-UF2-NEXT:    [[TMP37:%.*]] = mul i64 [[TMP44]], -1
 ; RV64-UF2-NEXT:    [[TMP39:%.*]] = getelementptr inbounds i32, ptr [[TMP32]], i64 [[TMP37]]
 ; RV64-UF2-NEXT:    [[TMP36:%.*]] = call i32 @llvm.vscale.i32()
 ; RV64-UF2-NEXT:    [[TMP40:%.*]] = mul i32 [[TMP36]], 4
@@ -483,7 +485,8 @@ define void @vector_reverse_f32(ptr noalias %A, ptr noalias %B, i32 %n) {
 ; RV64-UF2-NEXT:    [[TMP21:%.*]] = getelementptr inbounds float, ptr [[B]], i64 [[TMP20]]
 ; RV64-UF2-NEXT:    [[TMP23:%.*]] = getelementptr inbounds float, ptr [[TMP21]], i32 0
 ; RV64-UF2-NEXT:    [[TMP22:%.*]] = call i64 @llvm.vscale.i64()
-; RV64-UF2-NEXT:    [[TMP26:%.*]] = mul i64 [[TMP22]], 4294967292
+; RV64-UF2-NEXT:    [[TMP24:%.*]] = mul i64 [[TMP22]], 4
+; RV64-UF2-NEXT:    [[TMP26:%.*]] = mul i64 [[TMP24]], -1
 ; RV64-UF2-NEXT:    [[TMP28:%.*]] = getelementptr inbounds float, ptr [[TMP21]], i64 [[TMP26]]
 ; RV64-UF2-NEXT:    [[TMP25:%.*]] = call i32 @llvm.vscale.i32()
 ; RV64-UF2-NEXT:    [[TMP29:%.*]] = mul i32 [[TMP25]], 4
@@ -496,7 +499,8 @@ define void @vector_reverse_f32(ptr noalias %A, ptr noalias %B, i32 %n) {
 ; RV64-UF2-NEXT:    [[TMP32:%.*]] = getelementptr inbounds float, ptr [[A]], i64 [[TMP20]]
 ; RV64-UF2-NEXT:    [[TMP35:%.*]] = getelementptr inbounds float, ptr [[TMP32]], i32 0
 ; RV64-UF2-NEXT:    [[TMP33:%.*]] = call i64 @llvm.vscale.i64()
-; RV64-UF2-NEXT:    [[TMP37:%.*]] = mul i64 [[TMP33]], 4294967292
+; RV64-UF2-NEXT:    [[TMP44:%.*]] = mul i64 [[TMP33]], 4
+; RV64-UF2-NEXT:    [[TMP37:%.*]] = mul i64 [[TMP44]], -1
 ; RV64-UF2-NEXT:    [[TMP39:%.*]] = getelementptr inbounds float, ptr [[TMP32]], i64 [[TMP37]]
 ; RV64-UF2-NEXT:    [[TMP36:%.*]] = call i32 @llvm.vscale.i32()
 ; RV64-UF2-NEXT:    [[TMP40:%.*]] = mul i32 [[TMP36]], 4