[libcxx-commits] [clang] [libcxx] [flang] [mlir] [clang-tools-extra] [llvm] [LV] Improve AnyOf reduction codegen. (PR #78304)
Florian Hahn via libcxx-commits
libcxx-commits at lists.llvm.org
Mon Feb 5 06:28:18 PST 2024
https://github.com/fhahn updated https://github.com/llvm/llvm-project/pull/78304
>From 9846f970b6b394ccc3af25b92f238377a8ae7807 Mon Sep 17 00:00:00 2001
From: Florian Hahn <flo at fhahn.com>
Date: Sun, 14 Jan 2024 18:06:36 +0000
Subject: [PATCH 1/3] [LV] Improve AnyOf reduction codegen.
Update AnyOf reduction code generation to only keep track of the AnyOf
property in a boolean vector in the loop, only selecting either the new
or start value in the middle block.
This fixes the #62565, as now there aren't multiple uses of the
start/new values.
Fixes https://github.com/llvm/llvm-project/issues/62565
---
.../include/llvm/Transforms/Utils/LoopUtils.h | 9 --
llvm/lib/Transforms/Utils/LoopUtils.cpp | 24 +--
.../Vectorize/LoopVectorizationPlanner.h | 1 +
.../Transforms/Vectorize/LoopVectorize.cpp | 44 +++++-
.../lib/Transforms/Vectorize/VPlanRecipes.cpp | 7 +-
.../LoopVectorize/AArch64/sve-select-cmp.ll | 38 ++---
.../RISCV/select-cmp-reduction.ll | 120 ++++++--------
.../LoopVectorize/select-cmp-predicated.ll | 29 ++--
.../Transforms/LoopVectorize/select-cmp.ll | 146 +++++++++---------
...tion-start-value-may-be-undef-or-poison.ll | 43 +++---
10 files changed, 218 insertions(+), 243 deletions(-)
diff --git a/llvm/include/llvm/Transforms/Utils/LoopUtils.h b/llvm/include/llvm/Transforms/Utils/LoopUtils.h
index 5a1385d01d8e4..3bad7b616d9d7 100644
--- a/llvm/include/llvm/Transforms/Utils/LoopUtils.h
+++ b/llvm/include/llvm/Transforms/Utils/LoopUtils.h
@@ -363,15 +363,6 @@ Intrinsic::ID getMinMaxReductionIntrinsicOp(RecurKind RK);
/// Returns the comparison predicate used when expanding a min/max reduction.
CmpInst::Predicate getMinMaxReductionPredicate(RecurKind RK);
-/// See RecurrenceDescriptor::isAnyOfPattern for a description of the pattern we
-/// are trying to match. In this pattern, we are only ever selecting between two
-/// values: 1) an initial start value \p StartVal of the reduction PHI, and 2) a
-/// loop invariant value. If any of lane value in \p Left, \p Right is not equal
-/// to \p StartVal, select the loop invariant value. This is done by selecting
-/// \p Right iff \p Left is equal to \p StartVal.
-Value *createAnyOfOp(IRBuilderBase &Builder, Value *StartVal, RecurKind RK,
- Value *Left, Value *Right);
-
/// Returns a Min/Max operation corresponding to MinMaxRecurrenceKind.
/// The Builder's fast-math-flags must be set to propagate the expected values.
Value *createMinMaxOp(IRBuilderBase &Builder, RecurKind RK, Value *Left,
diff --git a/llvm/lib/Transforms/Utils/LoopUtils.cpp b/llvm/lib/Transforms/Utils/LoopUtils.cpp
index 59485126b280a..c0582fb7d7e15 100644
--- a/llvm/lib/Transforms/Utils/LoopUtils.cpp
+++ b/llvm/lib/Transforms/Utils/LoopUtils.cpp
@@ -962,15 +962,6 @@ CmpInst::Predicate llvm::getMinMaxReductionPredicate(RecurKind RK) {
}
}
-Value *llvm::createAnyOfOp(IRBuilderBase &Builder, Value *StartVal,
- RecurKind RK, Value *Left, Value *Right) {
- if (auto VTy = dyn_cast<VectorType>(Left->getType()))
- StartVal = Builder.CreateVectorSplat(VTy->getElementCount(), StartVal);
- Value *Cmp =
- Builder.CreateCmp(CmpInst::ICMP_NE, Left, StartVal, "rdx.select.cmp");
- return Builder.CreateSelect(Cmp, Left, Right, "rdx.select");
-}
-
Value *llvm::createMinMaxOp(IRBuilderBase &Builder, RecurKind RK, Value *Left,
Value *Right) {
Type *Ty = Left->getType();
@@ -1079,16 +1070,13 @@ Value *llvm::createAnyOfTargetReduction(IRBuilderBase &Builder, Value *Src,
NewVal = SI->getTrueValue();
}
- // Create a splat vector with the new value and compare this to the vector
- // we want to reduce.
- ElementCount EC = cast<VectorType>(Src->getType())->getElementCount();
- Value *Right = Builder.CreateVectorSplat(EC, InitVal);
- Value *Cmp =
- Builder.CreateCmp(CmpInst::ICMP_NE, Src, Right, "rdx.select.cmp");
-
// If any predicate is true it means that we want to select the new value.
- Cmp = Builder.CreateOrReduce(Cmp);
- return Builder.CreateSelect(Cmp, NewVal, InitVal, "rdx.select");
+ Value *AnyOf =
+ Src->getType()->isVectorTy() ? Builder.CreateOrReduce(Src) : Src;
+ // The compares in the loop may yield poison, which propagates through the
+ // bitwise ORs. Freeze it here before the condition is used.
+ AnyOf = Builder.CreateFreeze(AnyOf);
+ return Builder.CreateSelect(AnyOf, NewVal, InitVal, "rdx.select");
}
Value *llvm::createSimpleTargetReduction(IRBuilderBase &Builder, Value *Src,
diff --git a/llvm/lib/Transforms/Vectorize/LoopVectorizationPlanner.h b/llvm/lib/Transforms/Vectorize/LoopVectorizationPlanner.h
index a7ebf78e54ceb..9d3ef5b96c72f 100644
--- a/llvm/lib/Transforms/Vectorize/LoopVectorizationPlanner.h
+++ b/llvm/lib/Transforms/Vectorize/LoopVectorizationPlanner.h
@@ -68,6 +68,7 @@ class VPBuilder {
public:
VPBuilder() = default;
VPBuilder(VPBasicBlock *InsertBB) { setInsertPoint(InsertBB); }
+ VPBuilder(VPRecipeBase *InsertPt) { setInsertPoint(InsertPt); }
/// Clear the insertion point: created instructions will not be inserted into
/// a block.
diff --git a/llvm/lib/Transforms/Vectorize/LoopVectorize.cpp b/llvm/lib/Transforms/Vectorize/LoopVectorize.cpp
index aa5d1bfa57d53..973bcf4f6c2d7 100644
--- a/llvm/lib/Transforms/Vectorize/LoopVectorize.cpp
+++ b/llvm/lib/Transforms/Vectorize/LoopVectorize.cpp
@@ -7428,7 +7428,8 @@ static void createAndCollectMergePhiForReduction(
auto *PhiR = cast<VPReductionPHIRecipe>(RedResult->getOperand(0));
const RecurrenceDescriptor &RdxDesc = PhiR->getRecurrenceDescriptor();
- TrackingVH<Value> ReductionStartValue = RdxDesc.getRecurrenceStartValue();
+ TrackingVH<Value> ReductionStartValue =
+ State.get(PhiR->getStartValue(), VPIteration(0, 0));
Value *FinalValue =
State.get(RedResult, VPIteration(State.UF - 1, VPLane::getFirstLane()));
auto *ResumePhi =
@@ -7452,7 +7453,7 @@ static void createAndCollectMergePhiForReduction(
BCBlockPhi->addIncoming(ResumePhi->getIncomingValueForBlock(Incoming),
Incoming);
else
- BCBlockPhi->addIncoming(ReductionStartValue, Incoming);
+ BCBlockPhi->addIncoming(RdxDesc.getRecurrenceStartValue(), Incoming);
}
auto *OrigPhi = cast<PHINode>(PhiR->getUnderlyingValue());
@@ -9054,6 +9055,41 @@ void LoopVectorizationPlanner::adjustRecipesForReductions(
continue;
const RecurrenceDescriptor &RdxDesc = PhiR->getRecurrenceDescriptor();
+ // Adjust AnyOf reductions; replace the reduction phi for the selected value
+ // with a boolean reduction phi node to check if the condition is true in
+ // any iteration. The final value is selected by the final
+ // ComputeReductionResult.
+ if (RecurrenceDescriptor::isAnyOfRecurrenceKind(
+ RdxDesc.getRecurrenceKind())) {
+ auto *Select = cast<VPRecipeBase>(*find_if(PhiR->users(), [](VPUser *U) {
+ return isa<VPWidenSelectRecipe>(U) ||
+ (isa<VPReplicateRecipe>(U) &&
+ cast<VPReplicateRecipe>(U)->getUnderlyingInstr()->getOpcode() ==
+ Instruction::Select);
+ }));
+ VPValue *Cmp = Select->getOperand(0);
+ // If the compare is checking the reduction PHI node, adjust it to check
+ // the start value.
+ if (VPRecipeBase *CmpR = Cmp->getDefiningRecipe()) {
+ for (unsigned I = 0; I != CmpR->getNumOperands(); ++I)
+ if (CmpR->getOperand(I) == PhiR)
+ CmpR->setOperand(I, PhiR->getStartValue());
+ }
+ VPBuilder::InsertPointGuard Guard(Builder);
+ Builder.setInsertPoint(Select);
+
+ // If the true value of the select is the reduction phi, the new value is
+ // selected if the negated condition is true in any iteration.
+ if (Select->getOperand(1) == PhiR)
+ Cmp = Builder.createNot(Cmp);
+ VPValue *Or = Builder.createOr(PhiR, Cmp);
+ Select->getVPSingleValue()->replaceAllUsesWith(Or);
+
+ // Convert the reduction phi to operate on bools.
+ PhiR->setOperand(0, Plan->getVPValueOrAddLiveIn(ConstantInt::getFalse(
+ OrigLoop->getHeader()->getContext())));
+ }
+
// If tail is folded by masking, introduce selects between the phi
// and the live-out instruction of each reduction, at the beginning of the
// dedicated latch block.
@@ -9086,7 +9122,9 @@ void LoopVectorizationPlanner::adjustRecipesForReductions(
// then extend the loop exit value to enable InstCombine to evaluate the
// entire expression in the smaller type.
Type *PhiTy = PhiR->getStartValue()->getLiveInIRValue()->getType();
- if (MinVF.isVector() && PhiTy != RdxDesc.getRecurrenceType()) {
+ if (MinVF.isVector() && PhiTy != RdxDesc.getRecurrenceType() &&
+ !RecurrenceDescriptor::isAnyOfRecurrenceKind(
+ RdxDesc.getRecurrenceKind())) {
assert(!PhiR->isInLoop() && "Unexpected truncated inloop reduction!");
Type *RdxTy = RdxDesc.getRecurrenceType();
auto *Trunc =
diff --git a/llvm/lib/Transforms/Vectorize/VPlanRecipes.cpp b/llvm/lib/Transforms/Vectorize/VPlanRecipes.cpp
index c553e2c9e7683..f87461fd548bd 100644
--- a/llvm/lib/Transforms/Vectorize/VPlanRecipes.cpp
+++ b/llvm/lib/Transforms/Vectorize/VPlanRecipes.cpp
@@ -452,8 +452,7 @@ Value *VPInstruction::generateInstruction(VPTransformState &State,
else if (RecurrenceDescriptor::isAnyOfRecurrenceKind(RK)) {
TrackingVH<Value> ReductionStartValue =
RdxDesc.getRecurrenceStartValue();
- ReducedPartRdx = createAnyOfOp(Builder, ReductionStartValue, RK,
- ReducedPartRdx, RdxPart);
+ ReducedPartRdx = Builder.CreateOr(ReducedPartRdx, RdxPart);
} else
ReducedPartRdx = createMinMaxOp(Builder, RK, ReducedPartRdx, RdxPart);
}
@@ -461,7 +460,9 @@ Value *VPInstruction::generateInstruction(VPTransformState &State,
// Create the reduction after the loop. Note that inloop reductions create
// the target reduction in the loop using a Reduction recipe.
- if (State.VF.isVector() && !PhiR->isInLoop()) {
+ if ((State.VF.isVector() ||
+ RecurrenceDescriptor::isAnyOfRecurrenceKind(RK)) &&
+ !PhiR->isInLoop()) {
ReducedPartRdx =
createTargetReduction(Builder, RdxDesc, ReducedPartRdx, OrigPhi);
// If the reduction can be performed in a smaller type, we need to extend
diff --git a/llvm/test/Transforms/LoopVectorize/AArch64/sve-select-cmp.ll b/llvm/test/Transforms/LoopVectorize/AArch64/sve-select-cmp.ll
index 1c26ee8479e57..493ffd2ca569d 100644
--- a/llvm/test/Transforms/LoopVectorize/AArch64/sve-select-cmp.ll
+++ b/llvm/test/Transforms/LoopVectorize/AArch64/sve-select-cmp.ll
@@ -8,13 +8,12 @@ target triple = "aarch64-linux-gnu"
define i32 @select_const_i32_from_icmp(ptr nocapture readonly %v, i64 %n) #0 {
; CHECK-VF4IC1-LABEL: @select_const_i32_from_icmp
; CHECK-VF4IC1: vector.body:
-; CHECK-VF4IC1: [[VEC_PHI:%.*]] = phi <vscale x 4 x i32> [ shufflevector (<vscale x 4 x i32> insertelement (<vscale x 4 x i32> poison, i32 3, i64 0), <vscale x 4 x i32> poison, <vscale x 4 x i32> zeroinitializer), %vector.ph ], [ [[VEC_SEL:%.*]], %vector.body ]
+; CHECK-VF4IC1: [[VEC_PHI:%.*]] = phi <vscale x 4 x i1> [ zeroinitializer, %vector.ph ], [ [[VEC_SEL:%.*]], %vector.body ]
; CHECK-VF4IC1: [[VEC_LOAD:%.*]] = load <vscale x 4 x i32>
; CHECK-VF4IC1-NEXT: [[VEC_ICMP:%.*]] = icmp eq <vscale x 4 x i32> [[VEC_LOAD]], shufflevector (<vscale x 4 x i32> insertelement (<vscale x 4 x i32> poison, i32 3, i64 0), <vscale x 4 x i32> poison, <vscale x 4 x i32> zeroinitializer)
-; CHECK-VF4IC1-NEXT: [[VEC_SEL]] = select <vscale x 4 x i1> [[VEC_ICMP]], <vscale x 4 x i32> [[VEC_PHI]], <vscale x 4 x i32> shufflevector (<vscale x 4 x i32> insertelement (<vscale x 4 x i32> poison, i32 7, i64 0), <vscale x 4 x i32> poison, <vscale x 4 x i32> zeroinitializer)
+; CHECK-VF4IC1-NEXT: [[VEC_SEL]] = or <vscale x 4 x i1> [[VEC_PHI]], [[VEC_ICMP]]
; CHECK-VF4IC1: middle.block:
-; CHECK-VF4IC1-NEXT: [[FIN_ICMP:%.*]] = icmp ne <vscale x 4 x i32> [[VEC_SEL]], shufflevector (<vscale x 4 x i32> insertelement (<vscale x 4 x i32> poison, i32 3, i64 0), <vscale x 4 x i32> poison, <vscale x 4 x i32> zeroinitializer)
-; CHECK-VF4IC1-NEXT: [[OR_RDX:%.*]] = call i1 @llvm.vector.reduce.or.nxv4i1(<vscale x 4 x i1> [[FIN_ICMP]])
+; CHECK-VF4IC1-NEXT: [[OR_RDX:%.*]] = call i1 @llvm.vector.reduce.or.nxv4i1(<vscale x 4 x i1> [[VEC_SEL]])
; CHECK-VF4IC1-NEXT: {{.*}} = select i1 [[OR_RDX]], i32 7, i32 3
; CHECK-VF4IC4-LABEL: @select_const_i32_from_icmp
@@ -62,20 +61,15 @@ exit: ; preds = %for.body
define i32 @select_i32_from_icmp(ptr nocapture readonly %v, i32 %a, i32 %b, i64 %n) #0 {
; CHECK-VF4IC1-LABEL: @select_i32_from_icmp
; CHECK-VF4IC1: vector.ph:
-; CHECK-VF4IC1: [[TMP1:%.*]] = insertelement <vscale x 4 x i32> poison, i32 %a, i64 0
-; CHECK-VF4IC1-NEXT: [[SPLAT_OF_A:%.*]] = shufflevector <vscale x 4 x i32> [[TMP1]], <vscale x 4 x i32> poison, <vscale x 4 x i32> zeroinitializer
-; CHECK-VF4IC1-NEXT: [[TMP2:%.*]] = insertelement <vscale x 4 x i32> poison, i32 %b, i64 0
-; CHECK-VF4IC1-NEXT: [[SPLAT_OF_B:%.*]] = shufflevector <vscale x 4 x i32> [[TMP2]], <vscale x 4 x i32> poison, <vscale x 4 x i32> zeroinitializer
+; CHECK-VF4IC1-NOT: shufflevector <vscale x 4 x i32>
+; CHECK-VF4IC1-NOT: shufflevector <vscale x 4 x i32>
; CHECK-VF4IC1: vector.body:
-; CHECK-VF4IC1: [[VEC_PHI:%.*]] = phi <vscale x 4 x i32> [ [[SPLAT_OF_A]], %vector.ph ], [ [[VEC_SEL:%.*]], %vector.body ]
+; CHECK-VF4IC1: [[VEC_PHI:%.*]] = phi <vscale x 4 x i1> [ zeroinitializer, %vector.ph ], [ [[VEC_SEL:%.*]], %vector.body ]
; CHECK-VF4IC1: [[VEC_LOAD:%.*]] = load <vscale x 4 x i32>
; CHECK-VF4IC1-NEXT: [[VEC_ICMP:%.*]] = icmp eq <vscale x 4 x i32> [[VEC_LOAD]], shufflevector (<vscale x 4 x i32> insertelement (<vscale x 4 x i32> poison, i32 3, i64 0), <vscale x 4 x i32> poison, <vscale x 4 x i32> zeroinitializer)
-; CHECK-VF4IC1-NEXT: [[VEC_SEL]] = select <vscale x 4 x i1> [[VEC_ICMP]], <vscale x 4 x i32> [[VEC_PHI]], <vscale x 4 x i32> [[SPLAT_OF_B]]
+; CHECK-VF4IC1-NEXT: [[VEC_SEL]] = or <vscale x 4 x i1> [[VEC_PHI]], [[VEC_ICMP]]
; CHECK-VF4IC1: middle.block:
-; CHECK-VF4IC1-NEXT: [[FIN_INS:%.*]] = insertelement <vscale x 4 x i32> poison, i32 %a, i64 0
-; CHECK-VF4IC1-NEXT: [[FIN_SPLAT:%.*]] = shufflevector <vscale x 4 x i32> [[FIN_INS]], <vscale x 4 x i32> poison, <vscale x 4 x i32> zeroinitializer
-; CHECK-VF4IC1-NEXT: [[FIN_CMP:%.*]] = icmp ne <vscale x 4 x i32> [[VEC_SEL]], [[FIN_SPLAT]]
-; CHECK-VF4IC1-NEXT: [[OR_RDX:%.*]] = call i1 @llvm.vector.reduce.or.nxv4i1(<vscale x 4 x i1> [[FIN_CMP]])
+; CHECK-VF4IC1-NEXT: [[OR_RDX:%.*]] = call i1 @llvm.vector.reduce.or.nxv4i1(<vscale x 4 x i1> [[VEC_SEL]])
; CHECK-VF4IC1-NEXT: {{.*}} = select i1 [[OR_RDX]], i32 %b, i32 %a
; CHECK-VF4IC4-LABEL: @select_i32_from_icmp
@@ -101,13 +95,12 @@ exit: ; preds = %for.body
define i32 @select_const_i32_from_fcmp(ptr nocapture readonly %v, i64 %n) #0 {
; CHECK-VF4IC1-LABEL: @select_const_i32_from_fcmp
; CHECK-VF4IC1: vector.body:
-; CHECK-VF4IC1: [[VEC_PHI:%.*]] = phi <vscale x 4 x i32> [ shufflevector (<vscale x 4 x i32> insertelement (<vscale x 4 x i32> poison, i32 2, i64 0), <vscale x 4 x i32> poison, <vscale x 4 x i32> zeroinitializer), %vector.ph ], [ [[VEC_SEL:%.*]], %vector.body ]
+; CHECK-VF4IC1: [[VEC_PHI:%.*]] = phi <vscale x 4 x i1> [ zeroinitializer, %vector.ph ], [ [[VEC_SEL:%.*]], %vector.body ]
; CHECK-VF4IC1: [[VEC_LOAD:%.*]] = load <vscale x 4 x float>
; CHECK-VF4IC1-NEXT: [[VEC_ICMP:%.*]] = fcmp fast ueq <vscale x 4 x float> [[VEC_LOAD]], shufflevector (<vscale x 4 x float> insertelement (<vscale x 4 x float> poison, float 3.000000e+00, i64 0), <vscale x 4 x float> poison, <vscale x 4 x i32> zeroinitializer)
-; CHECK-VF4IC1-NEXT: [[VEC_SEL]] = select <vscale x 4 x i1> [[VEC_ICMP]], <vscale x 4 x i32> [[VEC_PHI]], <vscale x 4 x i32> shufflevector (<vscale x 4 x i32> insertelement (<vscale x 4 x i32> poison, i32 1, i64 0), <vscale x 4 x i32> poison, <vscale x 4 x i32> zeroinitializer)
+; CHECK-VF4IC1-NEXT: [[VEC_SEL]] = or <vscale x 4 x i1> [[VEC_PHI]], [[VEC_ICMP]]
; CHECK-VF4IC1: middle.block:
-; CHECK-VF4IC1-NEXT: [[FIN_ICMP:%.*]] = icmp ne <vscale x 4 x i32> [[VEC_SEL]], shufflevector (<vscale x 4 x i32> insertelement (<vscale x 4 x i32> poison, i32 2, i64 0), <vscale x 4 x i32> poison, <vscale x 4 x i32> zeroinitializer)
-; CHECK-VF4IC1-NEXT: [[OR_RDX:%.*]] = call i1 @llvm.vector.reduce.or.nxv4i1(<vscale x 4 x i1> [[FIN_ICMP]])
+; CHECK-VF4IC1-NEXT: [[OR_RDX:%.*]] = call i1 @llvm.vector.reduce.or.nxv4i1(<vscale x 4 x i1> [[VEC_SEL]])
; CHECK-VF4IC1-NEXT: {{.*}} = select i1 [[OR_RDX]], i32 1, i32 2
; CHECK-VF4IC4-LABEL: @select_const_i32_from_fcmp
@@ -156,16 +149,15 @@ exit: ; preds = %for.body
define i32 @pred_select_const_i32_from_icmp(ptr noalias nocapture readonly %src1, ptr noalias nocapture readonly %src2, i64 %n) #0 {
; CHECK-VF4IC1-LABEL: @pred_select_const_i32_from_icmp
; CHECK-VF4IC1: vector.body:
-; CHECK-VF4IC1: [[VEC_PHI:%.*]] = phi <vscale x 4 x i32> [ zeroinitializer, %vector.ph ], [ [[VEC_SEL:%.*]], %vector.body ]
+; CHECK-VF4IC1: [[VEC_PHI:%.*]] = phi <vscale x 4 x i1> [ zeroinitializer, %vector.ph ], [ [[VEC_SEL:%.*]], %vector.body ]
; CHECK-VF4IC1: [[VEC_LOAD:%.*]] = load <vscale x 4 x i32>
; CHECK-VF4IC1: [[MASK:%.*]] = icmp sgt <vscale x 4 x i32> [[VEC_LOAD]], shufflevector (<vscale x 4 x i32> insertelement (<vscale x 4 x i32> poison, i32 35, i64 0), <vscale x 4 x i32> poison, <vscale x 4 x i32> zeroinitializer)
; CHECK-VF4IC1: [[MASKED_LOAD:%.*]] = call <vscale x 4 x i32> @llvm.masked.load.nxv4i32.p0(ptr {{%.*}}, i32 4, <vscale x 4 x i1> [[MASK]], <vscale x 4 x i32> poison)
; CHECK-VF4IC1-NEXT: [[VEC_ICMP:%.*]] = icmp eq <vscale x 4 x i32> [[MASKED_LOAD]], shufflevector (<vscale x 4 x i32> insertelement (<vscale x 4 x i32> poison, i32 2, i64 0), <vscale x 4 x i32> poison, <vscale x 4 x i32> zeroinitializer)
-; CHECK-VF4IC1-NEXT: [[VEC_SEL_TMP:%.*]] = select <vscale x 4 x i1> [[VEC_ICMP]], <vscale x 4 x i32> shufflevector (<vscale x 4 x i32> insertelement (<vscale x 4 x i32> poison, i32 1, i64 0), <vscale x 4 x i32> poison, <vscale x 4 x i32> zeroinitializer), <vscale x 4 x i32> [[VEC_PHI]]
-; CHECK-VF4IC1: [[VEC_SEL:%.*]] = select <vscale x 4 x i1> [[MASK]], <vscale x 4 x i32> [[VEC_SEL_TMP]], <vscale x 4 x i32> [[VEC_PHI]]
+; CHECK-VF4IC1-NEXT: [[VEC_SEL_TMP:%.*]] = or <vscale x 4 x i1> [[VEC_PHI]], [[VEC_ICMP]]
+; CHECK-VF4IC1: [[VEC_SEL:%.*]] = select <vscale x 4 x i1> [[MASK]], <vscale x 4 x i1> [[VEC_SEL_TMP]], <vscale x 4 x i1> [[VEC_PHI]]
; CHECK-VF4IC1: middle.block:
-; CHECK-VF4IC1-NEXT: [[FIN_ICMP:%.*]] = icmp ne <vscale x 4 x i32> [[VEC_SEL]], zeroinitializer
-; CHECK-VF4IC1-NEXT: [[OR_RDX:%.*]] = call i1 @llvm.vector.reduce.or.nxv4i1(<vscale x 4 x i1> [[FIN_ICMP]])
+; CHECK-VF4IC1-NEXT: [[OR_RDX:%.*]] = call i1 @llvm.vector.reduce.or.nxv4i1(<vscale x 4 x i1> [[VEC_SEL]])
; CHECK-VF4IC1-NEXT: {{.*}} = select i1 [[OR_RDX]], i32 1, i32 0
; CHECK-VF4IC4-LABEL: @pred_select_const_i32_from_icmp
diff --git a/llvm/test/Transforms/LoopVectorize/RISCV/select-cmp-reduction.ll b/llvm/test/Transforms/LoopVectorize/RISCV/select-cmp-reduction.ll
index 34a7987bb40ab..6c50915c88e82 100644
--- a/llvm/test/Transforms/LoopVectorize/RISCV/select-cmp-reduction.ll
+++ b/llvm/test/Transforms/LoopVectorize/RISCV/select-cmp-reduction.ll
@@ -12,25 +12,22 @@ define i32 @select_icmp(i32 %x, i32 %y, ptr nocapture readonly %c, i64 %n) #0 {
; CHECK-NEXT: [[N_VEC:%.*]] = sub i64 %n, [[N_MOD_VF]]
; CHECK-NEXT: [[BROADCAST_SPLATINSERT:%.*]] = insertelement <4 x i32> poison, i32 [[X:%.*]], i64 0
; CHECK-NEXT: [[BROADCAST_SPLAT:%.*]] = shufflevector <4 x i32> [[BROADCAST_SPLATINSERT]], <4 x i32> poison, <4 x i32> zeroinitializer
-; CHECK-NEXT: [[BROADCAST_SPLATINSERT1:%.*]] = insertelement <4 x i32> poison, i32 [[Y:%.*]], i64 0
-; CHECK-NEXT: [[BROADCAST_SPLAT2:%.*]] = shufflevector <4 x i32> [[BROADCAST_SPLATINSERT1]], <4 x i32> poison, <4 x i32> zeroinitializer
; CHECK-NEXT: br label [[VECTOR_BODY:%.*]]
; CHECK: vector.body:
; CHECK-NEXT: [[INDEX:%.*]] = phi i64 [ 0, [[VECTOR_PH:%.*]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ]
-; CHECK-NEXT: [[VEC_PHI:%.*]] = phi <4 x i32> [ zeroinitializer, [[VECTOR_PH]] ], [ [[TMP5:%.*]], [[VECTOR_BODY]] ]
+; CHECK-NEXT: [[VEC_PHI:%.*]] = phi <4 x i1> [ zeroinitializer, [[VECTOR_PH]] ], [ [[TMP5:%.*]], [[VECTOR_BODY]] ]
; CHECK-NEXT: [[TMP0:%.*]] = add i64 [[INDEX]], 0
; CHECK-NEXT: [[TMP1:%.*]] = getelementptr inbounds i32, ptr [[C:%.*]], i64 [[TMP0]]
; CHECK-NEXT: [[TMP2:%.*]] = getelementptr inbounds i32, ptr [[TMP1]], i32 0
; CHECK-NEXT: [[WIDE_LOAD:%.*]] = load <4 x i32>, ptr [[TMP2]], align 4
; CHECK-NEXT: [[TMP4:%.*]] = icmp slt <4 x i32> [[WIDE_LOAD]], [[BROADCAST_SPLAT]]
-; CHECK-NEXT: [[TMP5]] = select <4 x i1> [[TMP4]], <4 x i32> [[VEC_PHI]], <4 x i32> [[BROADCAST_SPLAT2]]
+; CHECK-NEXT: [[TMP5]] = or <4 x i1> [[VEC_PHI]], [[TMP4]]
; CHECK-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX]], 4
; CHECK-NEXT: [[TMP6:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]]
; CHECK-NEXT: br i1 [[TMP6]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]]
; CHECK: middle.block:
-; CHECK-NEXT: [[RDX_SELECT_CMP:%.*]] = icmp ne <4 x i32> [[TMP5]], zeroinitializer
-; CHECK-NEXT: [[TMP7:%.*]] = call i1 @llvm.vector.reduce.or.v4i1(<4 x i1> [[RDX_SELECT_CMP]])
-; CHECK-NEXT: [[RDX_SELECT:%.*]] = select i1 [[TMP7]], i32 [[Y]], i32 0
+; CHECK-NEXT: [[TMP7:%.*]] = call i1 @llvm.vector.reduce.or.v4i1(<4 x i1> [[TMP5]])
+; CHECK-NEXT: [[RDX_SELECT:%.*]] = select i1 [[TMP7]], i32 %y, i32 0
;
; SCALABLE-LABEL: @select_icmp
; SCALABLE: vector.ph:
@@ -42,25 +39,22 @@ define i32 @select_icmp(i32 %x, i32 %y, ptr nocapture readonly %c, i64 %n) #0 {
; SCALABLE-NEXT: [[TMP11:%.*]] = mul i64 [[TMP10]], 4
; SCALABLE-NEXT: [[BROADCAST_SPLATINSERT:%.*]] = insertelement <vscale x 4 x i32> poison, i32 [[X:%.*]], i64 0
; SCALABLE-NEXT: [[BROADCAST_SPLAT:%.*]] = shufflevector <vscale x 4 x i32> [[BROADCAST_SPLATINSERT]], <vscale x 4 x i32> poison, <vscale x 4 x i32> zeroinitializer
-; SCALABLE-NEXT: [[BROADCAST_SPLATINSERT1:%.*]] = insertelement <vscale x 4 x i32> poison, i32 [[Y:%.*]], i64 0
-; SCALABLE-NEXT: [[BROADCAST_SPLAT2:%.*]] = shufflevector <vscale x 4 x i32> [[BROADCAST_SPLATINSERT1]], <vscale x 4 x i32> poison, <vscale x 4 x i32> zeroinitializer
; SCALABLE-NEXT: br label [[VECTOR_BODY:%.*]]
; SCALABLE: vector.body:
; SCALABLE-NEXT: [[INDEX:%.*]] = phi i64 [ 0, [[VECTOR_PH:%.*]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ]
-; SCALABLE-NEXT: [[VEC_PHI:%.*]] = phi <vscale x 4 x i32> [ zeroinitializer, [[VECTOR_PH]] ], [ [[TMP9:%.*]], [[VECTOR_BODY]] ]
+; SCALABLE-NEXT: [[VEC_PHI:%.*]] = phi <vscale x 4 x i1> [ zeroinitializer, [[VECTOR_PH]] ], [ [[TMP9:%.*]], [[VECTOR_BODY]] ]
; SCALABLE-NEXT: [[TMP4:%.*]] = add i64 [[INDEX]], 0
; SCALABLE-NEXT: [[TMP5:%.*]] = getelementptr inbounds i32, ptr [[C:%.*]], i64 [[TMP4]]
; SCALABLE-NEXT: [[TMP6:%.*]] = getelementptr inbounds i32, ptr [[TMP5]], i32 0
; SCALABLE-NEXT: [[WIDE_LOAD:%.*]] = load <vscale x 4 x i32>, ptr [[TMP6]], align 4
; SCALABLE-NEXT: [[TMP8:%.*]] = icmp slt <vscale x 4 x i32> [[WIDE_LOAD]], [[BROADCAST_SPLAT]]
-; SCALABLE-NEXT: [[TMP9]] = select <vscale x 4 x i1> [[TMP8]], <vscale x 4 x i32> [[VEC_PHI]], <vscale x 4 x i32> [[BROADCAST_SPLAT2]]
+; SCALABLE-NEXT: [[TMP9]] = or <vscale x 4 x i1> [[VEC_PHI]], [[TMP8]]
; SCALABLE-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX]], [[TMP11]]
; SCALABLE-NEXT: [[TMP12:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]]
; SCALABLE-NEXT: br i1 [[TMP12]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]]
; SCALABLE: middle.block:
-; SCALABLE-NEXT: [[RDX_SELECT_CMP:%.*]] = icmp ne <vscale x 4 x i32> [[TMP9]], zeroinitializer
-; SCALABLE-NEXT: [[TMP13:%.*]] = call i1 @llvm.vector.reduce.or.nxv4i1(<vscale x 4 x i1> [[RDX_SELECT_CMP]])
-; SCALABLE-NEXT: [[RDX_SELECT:%.*]] = select i1 [[TMP13]], i32 [[Y]], i32 0
+; SCALABLE-NEXT: [[TMP13:%.*]] = call i1 @llvm.vector.reduce.or.nxv4i1(<vscale x 4 x i1> [[TMP9]])
+; SCALABLE-NEXT: [[RDX_SELECT:%.*]] = select i1 [[TMP13]], i32 %y, i32 0
;
entry:
br label %for.body
@@ -87,25 +81,22 @@ define i32 @select_fcmp(float %x, i32 %y, ptr nocapture readonly %c, i64 %n) #0
; CHECK-NEXT: [[N_VEC:%.*]] = sub i64 %n, [[N_MOD_VF]]
; CHECK-NEXT: [[BROADCAST_SPLATINSERT:%.*]] = insertelement <4 x float> poison, float [[X:%.*]], i64 0
; CHECK-NEXT: [[BROADCAST_SPLAT:%.*]] = shufflevector <4 x float> [[BROADCAST_SPLATINSERT]], <4 x float> poison, <4 x i32> zeroinitializer
-; CHECK-NEXT: [[BROADCAST_SPLATINSERT1:%.*]] = insertelement <4 x i32> poison, i32 [[Y:%.*]], i64 0
-; CHECK-NEXT: [[BROADCAST_SPLAT2:%.*]] = shufflevector <4 x i32> [[BROADCAST_SPLATINSERT1]], <4 x i32> poison, <4 x i32> zeroinitializer
; CHECK-NEXT: br label [[VECTOR_BODY:%.*]]
; CHECK: vector.body:
; CHECK-NEXT: [[INDEX:%.*]] = phi i64 [ 0, [[VECTOR_PH:%.*]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ]
-; CHECK-NEXT: [[VEC_PHI:%.*]] = phi <4 x i32> [ zeroinitializer, [[VECTOR_PH]] ], [ [[TMP5:%.*]], [[VECTOR_BODY]] ]
+; CHECK-NEXT: [[VEC_PHI:%.*]] = phi <4 x i1> [ zeroinitializer, [[VECTOR_PH]] ], [ [[TMP5:%.*]], [[VECTOR_BODY]] ]
; CHECK-NEXT: [[TMP0:%.*]] = add i64 [[INDEX]], 0
; CHECK-NEXT: [[TMP1:%.*]] = getelementptr inbounds float, ptr [[C:%.*]], i64 [[TMP0]]
; CHECK-NEXT: [[TMP2:%.*]] = getelementptr inbounds float, ptr [[TMP1]], i32 0
; CHECK-NEXT: [[WIDE_LOAD:%.*]] = load <4 x float>, ptr [[TMP2]], align 4
; CHECK-NEXT: [[TMP4:%.*]] = fcmp fast olt <4 x float> [[WIDE_LOAD]], [[BROADCAST_SPLAT]]
-; CHECK-NEXT: [[TMP5]] = select <4 x i1> [[TMP4]], <4 x i32> [[VEC_PHI]], <4 x i32> [[BROADCAST_SPLAT2]]
+; CHECK-NEXT: [[TMP5]] = or <4 x i1> [[VEC_PHI]], [[TMP4]]
; CHECK-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX]], 4
; CHECK-NEXT: [[TMP6:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]]
; CHECK-NEXT: br i1 [[TMP6]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP4:![0-9]+]]
; CHECK: middle.block:
-; CHECK-NEXT: [[RDX_SELECT_CMP:%.*]] = icmp ne <4 x i32> [[TMP5]], zeroinitializer
-; CHECK-NEXT: [[TMP7:%.*]] = call i1 @llvm.vector.reduce.or.v4i1(<4 x i1> [[RDX_SELECT_CMP]])
-; CHECK-NEXT: [[RDX_SELECT:%.*]] = select i1 [[TMP7]], i32 [[Y]], i32 0
+; CHECK-NEXT: [[TMP7:%.*]] = call i1 @llvm.vector.reduce.or.v4i1(<4 x i1> [[TMP5]])
+; CHECK-NEXT: [[RDX_SELECT:%.*]] = select i1 [[TMP7]], i32 %y, i32 0
;
; SCALABLE-LABEL: @select_fcmp
; SCALABLE: vector.ph:
@@ -117,25 +108,22 @@ define i32 @select_fcmp(float %x, i32 %y, ptr nocapture readonly %c, i64 %n) #0
; SCALABLE-NEXT: [[TMP11:%.*]] = mul i64 [[TMP10]], 4
; SCALABLE-NEXT: [[BROADCAST_SPLATINSERT:%.*]] = insertelement <vscale x 4 x float> poison, float [[X:%.*]], i64 0
; SCALABLE-NEXT: [[BROADCAST_SPLAT:%.*]] = shufflevector <vscale x 4 x float> [[BROADCAST_SPLATINSERT]], <vscale x 4 x float> poison, <vscale x 4 x i32> zeroinitializer
-; SCALABLE-NEXT: [[BROADCAST_SPLATINSERT1:%.*]] = insertelement <vscale x 4 x i32> poison, i32 [[Y:%.*]], i64 0
-; SCALABLE-NEXT: [[BROADCAST_SPLAT2:%.*]] = shufflevector <vscale x 4 x i32> [[BROADCAST_SPLATINSERT1]], <vscale x 4 x i32> poison, <vscale x 4 x i32> zeroinitializer
; SCALABLE-NEXT: br label [[VECTOR_BODY:%.*]]
; SCALABLE: vector.body:
; SCALABLE-NEXT: [[INDEX:%.*]] = phi i64 [ 0, [[VECTOR_PH:%.*]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ]
-; SCALABLE-NEXT: [[VEC_PHI:%.*]] = phi <vscale x 4 x i32> [ zeroinitializer, [[VECTOR_PH]] ], [ [[TMP9:%.*]], [[VECTOR_BODY]] ]
+; SCALABLE-NEXT: [[VEC_PHI:%.*]] = phi <vscale x 4 x i1> [ zeroinitializer, [[VECTOR_PH]] ], [ [[TMP9:%.*]], [[VECTOR_BODY]] ]
; SCALABLE-NEXT: [[TMP4:%.*]] = add i64 [[INDEX]], 0
; SCALABLE-NEXT: [[TMP5:%.*]] = getelementptr inbounds float, ptr [[C:%.*]], i64 [[TMP4]]
; SCALABLE-NEXT: [[TMP6:%.*]] = getelementptr inbounds float, ptr [[TMP5]], i32 0
; SCALABLE-NEXT: [[WIDE_LOAD:%.*]] = load <vscale x 4 x float>, ptr [[TMP6]], align 4
; SCALABLE-NEXT: [[TMP8:%.*]] = fcmp fast olt <vscale x 4 x float> [[WIDE_LOAD]], [[BROADCAST_SPLAT]]
-; SCALABLE-NEXT: [[TMP9]] = select <vscale x 4 x i1> [[TMP8]], <vscale x 4 x i32> [[VEC_PHI]], <vscale x 4 x i32> [[BROADCAST_SPLAT2]]
+; SCALABLE-NEXT: [[TMP9]] = or <vscale x 4 x i1> [[VEC_PHI]], [[TMP8]]
; SCALABLE-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX]], [[TMP11]]
; SCALABLE-NEXT: [[TMP12:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]]
; SCALABLE-NEXT: br i1 [[TMP12]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP4:![0-9]+]]
; SCALABLE: middle.block:
-; SCALABLE-NEXT: [[RDX_SELECT_CMP:%.*]] = icmp ne <vscale x 4 x i32> [[TMP9]], zeroinitializer
-; SCALABLE-NEXT: [[TMP13:%.*]] = call i1 @llvm.vector.reduce.or.nxv4i1(<vscale x 4 x i1> [[RDX_SELECT_CMP]])
-; SCALABLE-NEXT: [[RDX_SELECT:%.*]] = select i1 [[TMP13]], i32 [[Y]], i32 0
+; SCALABLE-NEXT: [[TMP13:%.*]] = call i1 @llvm.vector.reduce.or.nxv4i1(<vscale x 4 x i1> [[TMP9]])
+; SCALABLE-NEXT: [[RDX_SELECT:%.*]] = select i1 [[TMP13]], i32 %y, i32 0
;
entry:
br label %for.body
@@ -163,19 +151,18 @@ define i32 @select_const_i32_from_icmp(ptr nocapture readonly %v, i64 %n) #0 {
; CHECK-NEXT: br label [[VECTOR_BODY:%.*]]
; CHECK: vector.body:
; CHECK-NEXT: [[INDEX:%.*]] = phi i64 [ 0, [[VECTOR_PH:%.*]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ]
-; CHECK-NEXT: [[VEC_PHI:%.*]] = phi <4 x i32> [ <i32 3, i32 3, i32 3, i32 3>, [[VECTOR_PH]] ], [ [[TMP5:%.*]], [[VECTOR_BODY]] ]
+; CHECK-NEXT: [[VEC_PHI:%.*]] = phi <4 x i1> [ zeroinitializer, [[VECTOR_PH]] ], [ [[TMP5:%.*]], [[VECTOR_BODY]] ]
; CHECK-NEXT: [[TMP0:%.*]] = add i64 [[INDEX]], 0
; CHECK-NEXT: [[TMP1:%.*]] = getelementptr inbounds i32, ptr [[V:%.*]], i64 [[TMP0]]
; CHECK-NEXT: [[TMP2:%.*]] = getelementptr inbounds i32, ptr [[TMP1]], i32 0
; CHECK-NEXT: [[WIDE_LOAD:%.*]] = load <4 x i32>, ptr [[TMP2]], align 4
; CHECK-NEXT: [[TMP4:%.*]] = icmp eq <4 x i32> [[WIDE_LOAD]], <i32 3, i32 3, i32 3, i32 3>
-; CHECK-NEXT: [[TMP5]] = select <4 x i1> [[TMP4]], <4 x i32> [[VEC_PHI]], <4 x i32> <i32 7, i32 7, i32 7, i32 7>
+; CHECK-NEXT: [[TMP5]] = or <4 x i1> [[VEC_PHI]], [[TMP4]]
; CHECK-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX]], 4
; CHECK-NEXT: [[TMP6:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]]
; CHECK-NEXT: br i1 [[TMP6]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP6:![0-9]+]]
; CHECK: middle.block:
-; CHECK-NEXT: [[RDX_SELECT_CMP:%.*]] = icmp ne <4 x i32> [[TMP5]], <i32 3, i32 3, i32 3, i32 3>
-; CHECK-NEXT: [[TMP7:%.*]] = call i1 @llvm.vector.reduce.or.v4i1(<4 x i1> [[RDX_SELECT_CMP]])
+; CHECK-NEXT: [[TMP7:%.*]] = call i1 @llvm.vector.reduce.or.v4i1(<4 x i1> [[TMP5]])
; CHECK-NEXT: [[RDX_SELECT:%.*]] = select i1 [[TMP7]], i32 7, i32 3
;
; SCALABLE-LABEL: @select_const_i32_from_icmp
@@ -189,19 +176,18 @@ define i32 @select_const_i32_from_icmp(ptr nocapture readonly %v, i64 %n) #0 {
; SCALABLE-NEXT: br label [[VECTOR_BODY:%.*]]
; SCALABLE: vector.body:
; SCALABLE-NEXT: [[INDEX:%.*]] = phi i64 [ 0, [[VECTOR_PH:%.*]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ]
-; SCALABLE-NEXT: [[VEC_PHI:%.*]] = phi <vscale x 4 x i32> [ shufflevector (<vscale x 4 x i32> insertelement (<vscale x 4 x i32> poison, i32 3, i64 0), <vscale x 4 x i32> poison, <vscale x 4 x i32> zeroinitializer), [[VECTOR_PH]] ], [ [[TMP9:%.*]], [[VECTOR_BODY]] ]
+; SCALABLE-NEXT: [[VEC_PHI:%.*]] = phi <vscale x 4 x i1> [ zeroinitializer, [[VECTOR_PH]] ], [ [[TMP9:%.*]], [[VECTOR_BODY]] ]
; SCALABLE-NEXT: [[TMP4:%.*]] = add i64 [[INDEX]], 0
; SCALABLE-NEXT: [[TMP5:%.*]] = getelementptr inbounds i32, ptr [[V:%.*]], i64 [[TMP4]]
; SCALABLE-NEXT: [[TMP6:%.*]] = getelementptr inbounds i32, ptr [[TMP5]], i32 0
; SCALABLE-NEXT: [[WIDE_LOAD:%.*]] = load <vscale x 4 x i32>, ptr [[TMP6]], align 4
; SCALABLE-NEXT: [[TMP8:%.*]] = icmp eq <vscale x 4 x i32> [[WIDE_LOAD]], shufflevector (<vscale x 4 x i32> insertelement (<vscale x 4 x i32> poison, i32 3, i64 0), <vscale x 4 x i32> poison, <vscale x 4 x i32> zeroinitializer)
-; SCALABLE-NEXT: [[TMP9]] = select <vscale x 4 x i1> [[TMP8]], <vscale x 4 x i32> [[VEC_PHI]], <vscale x 4 x i32> shufflevector (<vscale x 4 x i32> insertelement (<vscale x 4 x i32> poison, i32 7, i64 0), <vscale x 4 x i32> poison, <vscale x 4 x i32> zeroinitializer)
+; SCALABLE-NEXT: [[TMP9]] = or <vscale x 4 x i1> [[VEC_PHI]], [[TMP8]]
; SCALABLE-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX]], [[TMP11]]
; SCALABLE-NEXT: [[TMP12:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]]
; SCALABLE-NEXT: br i1 [[TMP12]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP6:![0-9]+]]
; SCALABLE: middle.block:
-; SCALABLE-NEXT: [[RDX_SELECT_CMP:%.*]] = icmp ne <vscale x 4 x i32> [[TMP9]], shufflevector (<vscale x 4 x i32> insertelement (<vscale x 4 x i32> poison, i32 3, i64 0), <vscale x 4 x i32> poison, <vscale x 4 x i32> zeroinitializer)
-; SCALABLE-NEXT: [[TMP13:%.*]] = call i1 @llvm.vector.reduce.or.nxv4i1(<vscale x 4 x i1> [[RDX_SELECT_CMP]])
+; SCALABLE-NEXT: [[TMP13:%.*]] = call i1 @llvm.vector.reduce.or.nxv4i1(<vscale x 4 x i1> [[TMP9]])
; SCALABLE-NEXT: [[RDX_SELECT:%.*]] = select i1 [[TMP13]], i32 7, i32 3
;
entry:
@@ -227,29 +213,22 @@ define i32 @select_i32_from_icmp(ptr nocapture readonly %v, i32 %a, i32 %b, i64
; CHECK: vector.ph:
; CHECK-NEXT: [[N_MOD_VF:%.*]] = urem i64 %n, 4
; CHECK-NEXT: [[N_VEC:%.*]] = sub i64 %n, [[N_MOD_VF]]
-; CHECK-NEXT: [[MINMAX_IDENT_SPLATINSERT:%.*]] = insertelement <4 x i32> poison, i32 [[A:%.*]], i64 0
-; CHECK-NEXT: [[MINMAX_IDENT_SPLAT:%.*]] = shufflevector <4 x i32> [[MINMAX_IDENT_SPLATINSERT]], <4 x i32> poison, <4 x i32> zeroinitializer
-; CHECK-NEXT: [[BROADCAST_SPLATINSERT:%.*]] = insertelement <4 x i32> poison, i32 [[B:%.*]], i64 0
-; CHECK-NEXT: [[BROADCAST_SPLAT:%.*]] = shufflevector <4 x i32> [[BROADCAST_SPLATINSERT]], <4 x i32> poison, <4 x i32> zeroinitializer
; CHECK-NEXT: br label [[VECTOR_BODY:%.*]]
; CHECK: vector.body:
; CHECK-NEXT: [[INDEX:%.*]] = phi i64 [ 0, [[VECTOR_PH:%.*]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ]
-; CHECK-NEXT: [[VEC_PHI:%.*]] = phi <4 x i32> [ [[MINMAX_IDENT_SPLAT]], [[VECTOR_PH]] ], [ [[TMP5:%.*]], [[VECTOR_BODY]] ]
+; CHECK-NEXT: [[VEC_PHI:%.*]] = phi <4 x i1> [ zeroinitializer, [[VECTOR_PH]] ], [ [[TMP5:%.*]], [[VECTOR_BODY]] ]
; CHECK-NEXT: [[TMP0:%.*]] = add i64 [[INDEX]], 0
; CHECK-NEXT: [[TMP1:%.*]] = getelementptr inbounds i32, ptr [[V:%.*]], i64 [[TMP0]]
; CHECK-NEXT: [[TMP2:%.*]] = getelementptr inbounds i32, ptr [[TMP1]], i32 0
; CHECK-NEXT: [[WIDE_LOAD:%.*]] = load <4 x i32>, ptr [[TMP2]], align 4
; CHECK-NEXT: [[TMP4:%.*]] = icmp eq <4 x i32> [[WIDE_LOAD]], <i32 3, i32 3, i32 3, i32 3>
-; CHECK-NEXT: [[TMP5]] = select <4 x i1> [[TMP4]], <4 x i32> [[VEC_PHI]], <4 x i32> [[BROADCAST_SPLAT]]
+; CHECK-NEXT: [[TMP5]] = or <4 x i1> [[VEC_PHI]], [[TMP4]]
; CHECK-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX]], 4
; CHECK-NEXT: [[TMP6:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]]
; CHECK-NEXT: br i1 [[TMP6]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP8:![0-9]+]]
; CHECK: middle.block:
-; CHECK-NEXT: [[DOTSPLATINSERT:%.*]] = insertelement <4 x i32> poison, i32 [[A]], i64 0
-; CHECK-NEXT: [[DOTSPLAT:%.*]] = shufflevector <4 x i32> [[DOTSPLATINSERT]], <4 x i32> poison, <4 x i32> zeroinitializer
-; CHECK-NEXT: [[RDX_SELECT_CMP:%.*]] = icmp ne <4 x i32> [[TMP5]], [[DOTSPLAT]]
-; CHECK-NEXT: [[TMP7:%.*]] = call i1 @llvm.vector.reduce.or.v4i1(<4 x i1> [[RDX_SELECT_CMP]])
-; CHECK-NEXT: [[RDX_SELECT:%.*]] = select i1 [[TMP7]], i32 [[B]], i32 [[A]]
+; CHECK-NEXT: [[TMP7:%.*]] = call i1 @llvm.vector.reduce.or.v4i1(<4 x i1> [[TMP5]])
+; CHECK-NEXT: [[RDX_SELECT:%.*]] = select i1 [[TMP7]], i32 %b, i32 %a
;
; SCALABLE-LABEL: @select_i32_from_icmp
; SCALABLE: vector.ph:
@@ -259,29 +238,22 @@ define i32 @select_i32_from_icmp(ptr nocapture readonly %v, i32 %a, i32 %b, i64
; SCALABLE-NEXT: [[N_VEC:%.*]] = sub i64 %n, [[N_MOD_VF]]
; SCALABLE-NEXT: [[TMP10:%.*]] = call i64 @llvm.vscale.i64()
; SCALABLE-NEXT: [[TMP11:%.*]] = mul i64 [[TMP10]], 4
-; SCALABLE-NEXT: [[MINMAX_IDENT_SPLATINSERT:%.*]] = insertelement <vscale x 4 x i32> poison, i32 [[A:%.*]], i64 0
-; SCALABLE-NEXT: [[MINMAX_IDENT_SPLAT:%.*]] = shufflevector <vscale x 4 x i32> [[MINMAX_IDENT_SPLATINSERT]], <vscale x 4 x i32> poison, <vscale x 4 x i32> zeroinitializer
-; SCALABLE-NEXT: [[BROADCAST_SPLATINSERT:%.*]] = insertelement <vscale x 4 x i32> poison, i32 [[B:%.*]], i64 0
-; SCALABLE-NEXT: [[BROADCAST_SPLAT:%.*]] = shufflevector <vscale x 4 x i32> [[BROADCAST_SPLATINSERT]], <vscale x 4 x i32> poison, <vscale x 4 x i32> zeroinitializer
; SCALABLE-NEXT: br label [[VECTOR_BODY:%.*]]
; SCALABLE: vector.body:
; SCALABLE-NEXT: [[INDEX:%.*]] = phi i64 [ 0, [[VECTOR_PH:%.*]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ]
-; SCALABLE-NEXT: [[VEC_PHI:%.*]] = phi <vscale x 4 x i32> [ [[MINMAX_IDENT_SPLAT]], [[VECTOR_PH]] ], [ [[TMP9:%.*]], [[VECTOR_BODY]] ]
+; SCALABLE-NEXT: [[VEC_PHI:%.*]] = phi <vscale x 4 x i1> [ zeroinitializer, [[VECTOR_PH]] ], [ [[TMP9:%.*]], [[VECTOR_BODY]] ]
; SCALABLE-NEXT: [[TMP4:%.*]] = add i64 [[INDEX]], 0
; SCALABLE-NEXT: [[TMP5:%.*]] = getelementptr inbounds i32, ptr [[V:%.*]], i64 [[TMP4]]
; SCALABLE-NEXT: [[TMP6:%.*]] = getelementptr inbounds i32, ptr [[TMP5]], i32 0
; SCALABLE-NEXT: [[WIDE_LOAD:%.*]] = load <vscale x 4 x i32>, ptr [[TMP6]], align 4
; SCALABLE-NEXT: [[TMP8:%.*]] = icmp eq <vscale x 4 x i32> [[WIDE_LOAD]], shufflevector (<vscale x 4 x i32> insertelement (<vscale x 4 x i32> poison, i32 3, i64 0), <vscale x 4 x i32> poison, <vscale x 4 x i32> zeroinitializer)
-; SCALABLE-NEXT: [[TMP9]] = select <vscale x 4 x i1> [[TMP8]], <vscale x 4 x i32> [[VEC_PHI]], <vscale x 4 x i32> [[BROADCAST_SPLAT]]
+; SCALABLE-NEXT: [[TMP9]] = or <vscale x 4 x i1> [[VEC_PHI]], [[TMP8]]
; SCALABLE-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX]], [[TMP11]]
; SCALABLE-NEXT: [[TMP12:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]]
; SCALABLE-NEXT: br i1 [[TMP12]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP8:![0-9]+]]
; SCALABLE: middle.block:
-; SCALABLE-NEXT: [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 4 x i32> poison, i32 [[A]], i64 0
-; SCALABLE-NEXT: [[DOTSPLAT:%.*]] = shufflevector <vscale x 4 x i32> [[DOTSPLATINSERT]], <vscale x 4 x i32> poison, <vscale x 4 x i32> zeroinitializer
-; SCALABLE-NEXT: [[RDX_SELECT_CMP:%.*]] = icmp ne <vscale x 4 x i32> [[TMP9]], [[DOTSPLAT]]
-; SCALABLE-NEXT: [[TMP13:%.*]] = call i1 @llvm.vector.reduce.or.nxv4i1(<vscale x 4 x i1> [[RDX_SELECT_CMP]])
-; SCALABLE-NEXT: [[RDX_SELECT:%.*]] = select i1 [[TMP13]], i32 [[B]], i32 [[A]]
+; SCALABLE-NEXT: [[TMP13:%.*]] = call i1 @llvm.vector.reduce.or.nxv4i1(<vscale x 4 x i1> [[TMP9]])
+; SCALABLE-NEXT: [[RDX_SELECT:%.*]] = select i1 [[TMP13]], i32 %b, i32 %a
;
entry:
br label %for.body
@@ -309,19 +281,18 @@ define i32 @select_const_i32_from_fcmp(ptr nocapture readonly %v, i64 %n) #0 {
; CHECK-NEXT: br label [[VECTOR_BODY:%.*]]
; CHECK: vector.body:
; CHECK-NEXT: [[INDEX:%.*]] = phi i64 [ 0, [[VECTOR_PH:%.*]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ]
-; CHECK-NEXT: [[VEC_PHI:%.*]] = phi <4 x i32> [ <i32 2, i32 2, i32 2, i32 2>, [[VECTOR_PH]] ], [ [[TMP5:%.*]], [[VECTOR_BODY]] ]
+; CHECK-NEXT: [[VEC_PHI:%.*]] = phi <4 x i1> [ zeroinitializer, [[VECTOR_PH]] ], [ [[TMP5:%.*]], [[VECTOR_BODY]] ]
; CHECK-NEXT: [[TMP0:%.*]] = add i64 [[INDEX]], 0
; CHECK-NEXT: [[TMP1:%.*]] = getelementptr inbounds float, ptr [[V:%.*]], i64 [[TMP0]]
; CHECK-NEXT: [[TMP2:%.*]] = getelementptr inbounds float, ptr [[TMP1]], i32 0
; CHECK-NEXT: [[WIDE_LOAD:%.*]] = load <4 x float>, ptr [[TMP2]], align 4
; CHECK-NEXT: [[TMP4:%.*]] = fcmp fast ueq <4 x float> [[WIDE_LOAD]], <float 3.000000e+00, float 3.000000e+00, float 3.000000e+00, float 3.000000e+00>
-; CHECK-NEXT: [[TMP5]] = select <4 x i1> [[TMP4]], <4 x i32> [[VEC_PHI]], <4 x i32> <i32 1, i32 1, i32 1, i32 1>
+; CHECK-NEXT: [[TMP5]] = or <4 x i1> [[VEC_PHI]], [[TMP4]]
; CHECK-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX]], 4
; CHECK-NEXT: [[TMP6:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]]
; CHECK-NEXT: br i1 [[TMP6]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP10:![0-9]+]]
; CHECK: middle.block:
-; CHECK-NEXT: [[RDX_SELECT_CMP:%.*]] = icmp ne <4 x i32> [[TMP5]], <i32 2, i32 2, i32 2, i32 2>
-; CHECK-NEXT: [[TMP7:%.*]] = call i1 @llvm.vector.reduce.or.v4i1(<4 x i1> [[RDX_SELECT_CMP]])
+; CHECK-NEXT: [[TMP7:%.*]] = call i1 @llvm.vector.reduce.or.v4i1(<4 x i1> [[TMP5]])
; CHECK-NEXT: [[RDX_SELECT:%.*]] = select i1 [[TMP7]], i32 1, i32 2
;
; SCALABLE-LABEL: @select_const_i32_from_fcmp
@@ -335,19 +306,18 @@ define i32 @select_const_i32_from_fcmp(ptr nocapture readonly %v, i64 %n) #0 {
; SCALABLE-NEXT: br label [[VECTOR_BODY:%.*]]
; SCALABLE: vector.body:
; SCALABLE-NEXT: [[INDEX:%.*]] = phi i64 [ 0, [[VECTOR_PH:%.*]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ]
-; SCALABLE-NEXT: [[VEC_PHI:%.*]] = phi <vscale x 4 x i32> [ shufflevector (<vscale x 4 x i32> insertelement (<vscale x 4 x i32> poison, i32 2, i64 0), <vscale x 4 x i32> poison, <vscale x 4 x i32> zeroinitializer), [[VECTOR_PH]] ], [ [[TMP9:%.*]], [[VECTOR_BODY]] ]
+; SCALABLE-NEXT: [[VEC_PHI:%.*]] = phi <vscale x 4 x i1> [ zeroinitializer, [[VECTOR_PH]] ], [ [[TMP9:%.*]], [[VECTOR_BODY]] ]
; SCALABLE-NEXT: [[TMP4:%.*]] = add i64 [[INDEX]], 0
; SCALABLE-NEXT: [[TMP5:%.*]] = getelementptr inbounds float, ptr [[V:%.*]], i64 [[TMP4]]
; SCALABLE-NEXT: [[TMP6:%.*]] = getelementptr inbounds float, ptr [[TMP5]], i32 0
; SCALABLE-NEXT: [[WIDE_LOAD:%.*]] = load <vscale x 4 x float>, ptr [[TMP6]], align 4
; SCALABLE-NEXT: [[TMP8:%.*]] = fcmp fast ueq <vscale x 4 x float> [[WIDE_LOAD]], shufflevector (<vscale x 4 x float> insertelement (<vscale x 4 x float> poison, float 3.000000e+00, i64 0), <vscale x 4 x float> poison, <vscale x 4 x i32> zeroinitializer)
-; SCALABLE-NEXT: [[TMP9]] = select <vscale x 4 x i1> [[TMP8]], <vscale x 4 x i32> [[VEC_PHI]], <vscale x 4 x i32> shufflevector (<vscale x 4 x i32> insertelement (<vscale x 4 x i32> poison, i32 1, i64 0), <vscale x 4 x i32> poison, <vscale x 4 x i32> zeroinitializer)
+; SCALABLE-NEXT: [[TMP9]] = or <vscale x 4 x i1> [[VEC_PHI]], [[TMP8]]
; SCALABLE-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX]], [[TMP11]]
; SCALABLE-NEXT: [[TMP12:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]]
; SCALABLE-NEXT: br i1 [[TMP12]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP10:![0-9]+]]
; SCALABLE: middle.block:
-; SCALABLE-NEXT: [[RDX_SELECT_CMP:%.*]] = icmp ne <vscale x 4 x i32> [[TMP9]], shufflevector (<vscale x 4 x i32> insertelement (<vscale x 4 x i32> poison, i32 2, i64 0), <vscale x 4 x i32> poison, <vscale x 4 x i32> zeroinitializer)
-; SCALABLE-NEXT: [[TMP13:%.*]] = call i1 @llvm.vector.reduce.or.nxv4i1(<vscale x 4 x i1> [[RDX_SELECT_CMP]])
+; SCALABLE-NEXT: [[TMP13:%.*]] = call i1 @llvm.vector.reduce.or.nxv4i1(<vscale x 4 x i1> [[TMP9]])
; SCALABLE-NEXT: [[RDX_SELECT:%.*]] = select i1 [[TMP13]], i32 1, i32 2
;
entry:
@@ -401,7 +371,7 @@ define i32 @pred_select_const_i32_from_icmp(ptr noalias nocapture readonly %src1
; CHECK-NEXT: br label [[VECTOR_BODY:%.*]]
; CHECK: vector.body:
; CHECK-NEXT: [[INDEX:%.*]] = phi i64 [ 0, [[VECTOR_PH:%.*]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ]
-; CHECK-NEXT: [[VEC_PHI:%.*]] = phi <4 x i32> [ zeroinitializer, [[VECTOR_PH]] ], [ [[PREDPHI:%.*]], [[VECTOR_BODY]] ]
+; CHECK-NEXT: [[VEC_PHI:%.*]] = phi <4 x i1> [ zeroinitializer, [[VECTOR_PH]] ], [ [[PREDPHI:%.*]], [[VECTOR_BODY]] ]
; CHECK-NEXT: [[TMP0:%.*]] = add i64 [[INDEX]], 0
; CHECK-NEXT: [[TMP1:%.*]] = getelementptr inbounds i32, ptr [[SRC1:%.*]], i64 [[TMP0]]
; CHECK-NEXT: [[TMP2:%.*]] = getelementptr inbounds i32, ptr [[TMP1]], i32 0
@@ -411,15 +381,14 @@ define i32 @pred_select_const_i32_from_icmp(ptr noalias nocapture readonly %src1
; CHECK-NEXT: [[TMP6:%.*]] = getelementptr i32, ptr [[TMP5]], i32 0
; CHECK-NEXT: [[WIDE_MASKED_LOAD:%.*]] = call <4 x i32> @llvm.masked.load.v4i32.p0(ptr [[TMP6]], i32 4, <4 x i1> [[TMP4]], <4 x i32> poison)
; CHECK-NEXT: [[TMP8:%.*]] = icmp eq <4 x i32> [[WIDE_MASKED_LOAD]], <i32 2, i32 2, i32 2, i32 2>
-; CHECK-NEXT: [[TMP9:%.*]] = select <4 x i1> [[TMP8]], <4 x i32> <i32 1, i32 1, i32 1, i32 1>, <4 x i32> [[VEC_PHI]]
+; CHECK-NEXT: [[TMP9:%.*]] = or <4 x i1> [[VEC_PHI]], [[TMP8]]
; CHECK-NEXT: [[TMP10:%.*]] = xor <4 x i1> [[TMP4]], <i1 true, i1 true, i1 true, i1 true>
-; CHECK-NEXT: [[PREDPHI]] = select <4 x i1> [[TMP4]], <4 x i32> [[TMP9]], <4 x i32> [[VEC_PHI]]
+; CHECK-NEXT: [[PREDPHI]] = select <4 x i1> [[TMP4]], <4 x i1> [[TMP9]], <4 x i1> [[VEC_PHI]]
; CHECK-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX]], 4
; CHECK-NEXT: [[TMP11:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]]
; CHECK-NEXT: br i1 [[TMP11]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP12:![0-9]+]]
; CHECK: middle.block:
-; CHECK-NEXT: [[RDX_SELECT_CMP:%.*]] = icmp ne <4 x i32> [[PREDPHI]], zeroinitializer
-; CHECK-NEXT: [[TMP12:%.*]] = call i1 @llvm.vector.reduce.or.v4i1(<4 x i1> [[RDX_SELECT_CMP]])
+; CHECK-NEXT: [[TMP12:%.*]] = call i1 @llvm.vector.reduce.or.v4i1(<4 x i1> [[PREDPHI]])
; CHECK-NEXT: [[RDX_SELECT:%.*]] = select i1 [[TMP12]], i32 1, i32 0
;
; SCALABLE-LABEL: @pred_select_const_i32_from_icmp
@@ -433,7 +402,7 @@ define i32 @pred_select_const_i32_from_icmp(ptr noalias nocapture readonly %src1
; SCALABLE-NEXT: br label [[VECTOR_BODY:%.*]]
; SCALABLE: vector.body:
; SCALABLE-NEXT: [[INDEX:%.*]] = phi i64 [ 0, [[VECTOR_PH:%.*]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ]
-; SCALABLE-NEXT: [[VEC_PHI:%.*]] = phi <vscale x 4 x i32> [ zeroinitializer, [[VECTOR_PH]] ], [ [[PREDPHI:%.*]], [[VECTOR_BODY]] ]
+; SCALABLE-NEXT: [[VEC_PHI:%.*]] = phi <vscale x 4 x i1> [ zeroinitializer, [[VECTOR_PH]] ], [ [[PREDPHI:%.*]], [[VECTOR_BODY]] ]
; SCALABLE-NEXT: [[TMP4:%.*]] = add i64 [[INDEX]], 0
; SCALABLE-NEXT: [[TMP5:%.*]] = getelementptr inbounds i32, ptr [[SRC1:%.*]], i64 [[TMP4]]
; SCALABLE-NEXT: [[TMP6:%.*]] = getelementptr inbounds i32, ptr [[TMP5]], i32 0
@@ -443,15 +412,14 @@ define i32 @pred_select_const_i32_from_icmp(ptr noalias nocapture readonly %src1
; SCALABLE-NEXT: [[TMP10:%.*]] = getelementptr i32, ptr [[TMP9]], i32 0
; SCALABLE-NEXT: [[WIDE_MASKED_LOAD:%.*]] = call <vscale x 4 x i32> @llvm.masked.load.nxv4i32.p0(ptr [[TMP10]], i32 4, <vscale x 4 x i1> [[TMP8]], <vscale x 4 x i32> poison)
; SCALABLE-NEXT: [[TMP12:%.*]] = icmp eq <vscale x 4 x i32> [[WIDE_MASKED_LOAD]], shufflevector (<vscale x 4 x i32> insertelement (<vscale x 4 x i32> poison, i32 2, i64 0), <vscale x 4 x i32> poison, <vscale x 4 x i32> zeroinitializer)
-; SCALABLE-NEXT: [[TMP13:%.*]] = select <vscale x 4 x i1> [[TMP12]], <vscale x 4 x i32> shufflevector (<vscale x 4 x i32> insertelement (<vscale x 4 x i32> poison, i32 1, i64 0), <vscale x 4 x i32> poison, <vscale x 4 x i32> zeroinitializer), <vscale x 4 x i32> [[VEC_PHI]]
+; SCALABLE-NEXT: [[TMP13:%.*]] = or <vscale x 4 x i1> [[VEC_PHI]], [[TMP12]]
; SCALABLE-NEXT: [[TMP14:%.*]] = xor <vscale x 4 x i1> [[TMP8]], shufflevector (<vscale x 4 x i1> insertelement (<vscale x 4 x i1> poison, i1 true, i64 0), <vscale x 4 x i1> poison, <vscale x 4 x i32> zeroinitializer)
-; SCALABLE-NEXT: [[PREDPHI]] = select <vscale x 4 x i1> [[TMP8]], <vscale x 4 x i32> [[TMP13]], <vscale x 4 x i32> [[VEC_PHI]]
+; SCALABLE-NEXT: [[PREDPHI]] = select <vscale x 4 x i1> [[TMP8]], <vscale x 4 x i1> [[TMP13]], <vscale x 4 x i1> [[VEC_PHI]]
; SCALABLE-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX]], [[TMP16]]
; SCALABLE-NEXT: [[TMP17:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]]
; SCALABLE-NEXT: br i1 [[TMP17]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP12:![0-9]+]]
; SCALABLE: middle.block:
-; SCALABLE-NEXT: [[RDX_SELECT_CMP:%.*]] = icmp ne <vscale x 4 x i32> [[PREDPHI]], zeroinitializer
-; SCALABLE-NEXT: [[TMP18:%.*]] = call i1 @llvm.vector.reduce.or.nxv4i1(<vscale x 4 x i1> [[RDX_SELECT_CMP]])
+; SCALABLE-NEXT: [[TMP18:%.*]] = call i1 @llvm.vector.reduce.or.nxv4i1(<vscale x 4 x i1> [[PREDPHI]])
; SCALABLE-NEXT: [[RDX_SELECT:%.*]] = select i1 [[TMP18]], i32 1, i32 0
;
entry:
diff --git a/llvm/test/Transforms/LoopVectorize/select-cmp-predicated.ll b/llvm/test/Transforms/LoopVectorize/select-cmp-predicated.ll
index e8c75635f2c69..98962fd4e371f 100644
--- a/llvm/test/Transforms/LoopVectorize/select-cmp-predicated.ll
+++ b/llvm/test/Transforms/LoopVectorize/select-cmp-predicated.ll
@@ -4,7 +4,7 @@
define i32 @pred_select_const_i32_from_icmp(ptr noalias nocapture readonly %src1, ptr noalias nocapture readonly %src2, i64 %n) {
; CHECK-VF2IC1-LABEL: @pred_select_const_i32_from_icmp(
; CHECK-VF2IC1: vector.body:
-; CHECK-VF2IC1: [[VEC_PHI:%.*]] = phi <2 x i32> [ zeroinitializer, %vector.ph ], [ [[PREDPHI:%.*]], %pred.load.continue2 ]
+; CHECK-VF2IC1: [[VEC_PHI:%.*]] = phi <2 x i1> [ zeroinitializer, %vector.ph ], [ [[PREDPHI:%.*]], %pred.load.continue2 ]
; CHECK-VF2IC1: [[WIDE_LOAD:%.*]] = load <2 x i32>, ptr {{%.*}}, align 4
; CHECK-VF2IC1-NEXT: [[TMP4:%.*]] = icmp sgt <2 x i32> [[WIDE_LOAD]], <i32 35, i32 35>
; CHECK-VF2IC1-NEXT: [[TMP5:%.*]] = extractelement <2 x i1> [[TMP4]], i32 0
@@ -26,14 +26,14 @@ define i32 @pred_select_const_i32_from_icmp(ptr noalias nocapture readonly %src1
; CHECK-VF2IC1: pred.load.continue2:
; CHECK-VF2IC1-NEXT: [[TMP15:%.*]] = phi <2 x i32> [ [[TMP9]], %pred.load.continue ], [ [[TMP14]], %pred.load.if1 ]
; CHECK-VF2IC1-NEXT: [[TMP16:%.*]] = icmp eq <2 x i32> [[TMP15]], <i32 2, i32 2>
-; CHECK-VF2IC1-NEXT: [[TMP17:%.*]] = select <2 x i1> [[TMP16]], <2 x i32> <i32 1, i32 1>, <2 x i32> [[VEC_PHI]]
+; CHECK-VF2IC1-NEXT: [[TMP17:%.*]] = or <2 x i1> [[VEC_PHI]], [[TMP16]]
; CHECK-VF2IC1-NEXT: [[TMP18:%.*]] = xor <2 x i1> [[TMP4]], <i1 true, i1 true>
-; CHECK-VF2IC1-NEXT: [[PREDPHI]] = select <2 x i1> [[TMP4]], <2 x i32> [[TMP17]], <2 x i32> [[VEC_PHI]]
+; CHECK-VF2IC1-NEXT: [[PREDPHI]] = select <2 x i1> [[TMP4]], <2 x i1> [[TMP17]], <2 x i1> [[VEC_PHI]]
; CHECK-VF2IC1: br i1 {{%.*}}, label %middle.block, label %vector.body
; CHECK-VF2IC1: middle.block:
-; CHECK-VF2IC1-NEXT: [[RDX_SELECT_CMP:%.*]] = icmp ne <2 x i32> [[PREDPHI]], zeroinitializer
-; CHECK-VF2IC1-NEXT: [[TMP20:%.*]] = call i1 @llvm.vector.reduce.or.v2i1(<2 x i1> [[RDX_SELECT_CMP]])
-; CHECK-VF2IC1-NEXT: [[RDX_SELECT:%.*]] = select i1 [[TMP20]], i32 1, i32 0
+; CHECK-VF2IC1-NEXT: [[TMP20:%.*]] = call i1 @llvm.vector.reduce.or.v2i1(<2 x i1> [[PREDPHI]])
+; CHECK-VF2IC1-NEXT: [[FR_TMP20:%.*]] = freeze i1 [[TMP20]]
+; CHECK-VF2IC1-NEXT: [[RDX_SELECT:%.*]] = select i1 [[FR_TMP20]], i32 1, i32 0
; CHECK-VF2IC1: scalar.ph:
; CHECK-VF2IC1: [[BC_RESUME_VAL:%.*]] = phi i64 [ {{%.*}}, %middle.block ], [ 0, %entry ]
; CHECK-VF2IC1-NEXT: [[BC_MERGE_RDX:%.*]] = phi i32 [ 0, %entry ], [ [[RDX_SELECT]], %middle.block ]
@@ -56,8 +56,8 @@ define i32 @pred_select_const_i32_from_icmp(ptr noalias nocapture readonly %src1
;
; CHECK-VF1IC2-LABEL: @pred_select_const_i32_from_icmp(
; CHECK-VF1IC2: vector.body:
-; CHECK-VF1IC2: [[VEC_PHI:%.*]] = phi i32 [ 0, %vector.ph ], [ [[PREDPHI:%.*]], %pred.load.continue3 ]
-; CHECK-VF1IC2-NEXT: [[VEC_PHI2:%.*]] = phi i32 [ 0, %vector.ph ], [ [[PREDPHI5:%.*]], %pred.load.continue3 ]
+; CHECK-VF1IC2: [[VEC_PHI:%.*]] = phi i1 [ false, %vector.ph ], [ [[PREDPHI:%.*]], %pred.load.continue3 ]
+; CHECK-VF1IC2-NEXT: [[VEC_PHI2:%.*]] = phi i1 [ false, %vector.ph ], [ [[PREDPHI5:%.*]], %pred.load.continue3 ]
; CHECK-VF1IC2: [[TMP0:%.*]] = getelementptr inbounds i32, ptr [[SRC1:%.*]], i64 {{%.*}}
; CHECK-VF1IC2-NEXT: [[TMP1:%.*]] = getelementptr inbounds i32, ptr [[SRC1]], i64 {{%.*}}
; CHECK-VF1IC2-NEXT: [[TMP2:%.*]] = load i32, ptr [[TMP0]], align 4
@@ -80,16 +80,17 @@ define i32 @pred_select_const_i32_from_icmp(ptr noalias nocapture readonly %src1
; CHECK-VF1IC2-NEXT: [[TMP11:%.*]] = phi i32 [ poison, %pred.load.continue ], [ [[TMP10]], %pred.load.if2 ]
; CHECK-VF1IC2-NEXT: [[TMP12:%.*]] = icmp eq i32 [[TMP8]], 2
; CHECK-VF1IC2-NEXT: [[TMP13:%.*]] = icmp eq i32 [[TMP11]], 2
-; CHECK-VF1IC2-NEXT: [[TMP14:%.*]] = select i1 [[TMP12]], i32 1, i32 [[VEC_PHI]]
-; CHECK-VF1IC2-NEXT: [[TMP15:%.*]] = select i1 [[TMP13]], i32 1, i32 [[VEC_PHI2]]
+; CHECK-VF1IC2-NEXT: [[TMP14:%.*]] = or i1 [[VEC_PHI]], [[TMP12]]
+; CHECK-VF1IC2-NEXT: [[TMP15:%.*]] = or i1 [[VEC_PHI2]], [[TMP13]]
; CHECK-VF1IC2-NEXT: [[TMP16:%.*]] = xor i1 [[TMP4]], true
; CHECK-VF1IC2-NEXT: [[TMP17:%.*]] = xor i1 [[TMP5]], true
-; CHECK-VF1IC2-NEXT: [[PREDPHI]] = select i1 [[TMP4]], i32 [[TMP14]], i32 [[VEC_PHI]]
-; CHECK-VF1IC2-NEXT: [[PREDPHI5]] = select i1 [[TMP5]], i32 [[TMP15]], i32 [[VEC_PHI2]]
+; CHECK-VF1IC2-NEXT: [[PREDPHI]] = select i1 [[TMP4]], i1 [[TMP14]], i1 [[VEC_PHI]]
+; CHECK-VF1IC2-NEXT: [[PREDPHI5]] = select i1 [[TMP5]], i1 [[TMP15]], i1 [[VEC_PHI2]]
; CHECK-VF1IC2: br i1 {{%.*}}, label %middle.block, label %vector.body
; CHECK-VF1IC2: middle.block:
-; CHECK-VF1IC2-NEXT: [[RDX_SELECT_CMP:%.*]] = icmp ne i32 [[PREDPHI]], 0
-; CHECK-VF1IC2-NEXT: [[RDX_SELECT:%.*]] = select i1 [[RDX_SELECT_CMP]], i32 [[PREDPHI]], i32 [[PREDPHI5]]
+; CHECK-VF1IC2-NEXT: [[OR:%.*]] = or i1 [[PREDPHI]], [[PREDPHI5]]
+; CHECK-VF1IC2-NEXT: [[FR_OR:%.*]] = freeze i1 [[OR]]
+; CHECK-VF1IC2-NEXT: [[RDX_SELECT:%.*]] = select i1 [[FR_OR]], i32 1, i32 0
; CHECK-VF1IC2: br i1 {{%.*}}, label %for.end.loopexit, label %scalar.ph
; CHECK-VF1IC2: scalar.ph:
; CHECK-VF1IC2-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ {{%.*}}, %middle.block ], [ 0, %entry ]
diff --git a/llvm/test/Transforms/LoopVectorize/select-cmp.ll b/llvm/test/Transforms/LoopVectorize/select-cmp.ll
index c9f2aaef6d5c8..a9157ec118b01 100644
--- a/llvm/test/Transforms/LoopVectorize/select-cmp.ll
+++ b/llvm/test/Transforms/LoopVectorize/select-cmp.ll
@@ -5,45 +5,47 @@
define i32 @select_const_i32_from_icmp(ptr nocapture readonly %v, i64 %n) {
; CHECK-LABEL: @select_const_i32_from_icmp
; CHECK-VF4IC1: vector.body:
-; CHECK-VF4IC1: [[VEC_PHI:%.*]] = phi <4 x i32> [ <i32 3, i32 3, i32 3, i32 3>, %vector.ph ], [ [[VEC_SEL:%.*]], %vector.body ]
+; CHECK-VF4IC1: [[VEC_PHI:%.*]] = phi <4 x i1> [ zeroinitializer, %vector.ph ], [ [[VEC_SEL:%.*]], %vector.body ]
; CHECK-VF4IC1: [[VEC_LOAD:%.*]] = load <4 x i32>
; CHECK-VF4IC1-NEXT: [[VEC_ICMP:%.*]] = icmp eq <4 x i32> [[VEC_LOAD]], <i32 3, i32 3, i32 3, i32 3>
-; CHECK-VF4IC1-NEXT: [[VEC_SEL]] = select <4 x i1> [[VEC_ICMP]], <4 x i32> [[VEC_PHI]], <4 x i32> <i32 7, i32 7, i32 7, i32 7>
+; CHECK-VF4IC1-NEXT: [[NOT:%.*]] = xor <4 x i1> [[VEC_ICMP]], <i1 true, i1 true, i1 true, i1 true>
+; CHECK-VF4IC1-NEXT: [[VEC_SEL]] = or <4 x i1> [[VEC_PHI]], [[NOT]]
; CHECK-VF4IC1: middle.block:
-; CHECK-VF4IC1-NEXT: [[FIN_ICMP:%.*]] = icmp ne <4 x i32> [[VEC_SEL]], <i32 3, i32 3, i32 3, i32 3>
-; CHECK-VF4IC1-NEXT: [[OR_RDX:%.*]] = call i1 @llvm.vector.reduce.or.v4i1(<4 x i1> [[FIN_ICMP]])
-; CHECK-VF4IC1-NEXT: {{.*}} = select i1 [[OR_RDX]], i32 7, i32 3
+; CHECK-VF4IC1-NEXT: [[OR_RDX:%.*]] = call i1 @llvm.vector.reduce.or.v4i1(<4 x i1> [[VEC_SEL]])
+; CHECK-VF4IC1-NEXT: [[FR_OR_RDX:%.*]] = freeze i1 [[OR_RDX]]
+; CHECK-VF4IC1-NEXT: {{.*}} = select i1 [[FR_OR_RDX]], i32 7, i32 3
; CHECK-VF4IC4: vector.body:
-; CHECK-VF4IC4: [[VEC_PHI1:%.*]] = phi <4 x i32> [ <i32 3, i32 3, i32 3, i32 3>, %vector.ph ], [ [[VEC_SEL1:%.*]], %vector.body ]
-; CHECK-VF4IC4-NEXT: [[VEC_PHI2:%.*]] = phi <4 x i32> [ <i32 3, i32 3, i32 3, i32 3>, %vector.ph ], [ [[VEC_SEL2:%.*]], %vector.body ]
-; CHECK-VF4IC4-NEXT: [[VEC_PHI3:%.*]] = phi <4 x i32> [ <i32 3, i32 3, i32 3, i32 3>, %vector.ph ], [ [[VEC_SEL3:%.*]], %vector.body ]
-; CHECK-VF4IC4-NEXT: [[VEC_PHI4:%.*]] = phi <4 x i32> [ <i32 3, i32 3, i32 3, i32 3>, %vector.ph ], [ [[VEC_SEL4:%.*]], %vector.body ]
+; CHECK-VF4IC4: [[VEC_PHI1:%.*]] = phi <4 x i1> [ zeroinitializer, %vector.ph ], [ [[VEC_SEL1:%.*]], %vector.body ]
+; CHECK-VF4IC4-NEXT: [[VEC_PHI2:%.*]] = phi <4 x i1> [ zeroinitializer, %vector.ph ], [ [[VEC_SEL2:%.*]], %vector.body ]
+; CHECK-VF4IC4-NEXT: [[VEC_PHI3:%.*]] = phi <4 x i1> [ zeroinitializer, %vector.ph ], [ [[VEC_SEL3:%.*]], %vector.body ]
+; CHECK-VF4IC4-NEXT: [[VEC_PHI4:%.*]] = phi <4 x i1> [ zeroinitializer, %vector.ph ], [ [[VEC_SEL4:%.*]], %vector.body ]
; CHECK-VF4IC4: [[VEC_ICMP1:%.*]] = icmp eq <4 x i32> {{.*}}, <i32 3, i32 3, i32 3, i32 3>
; CHECK-VF4IC4-NEXT: [[VEC_ICMP2:%.*]] = icmp eq <4 x i32> {{.*}}, <i32 3, i32 3, i32 3, i32 3>
; CHECK-VF4IC4-NEXT: [[VEC_ICMP3:%.*]] = icmp eq <4 x i32> {{.*}}, <i32 3, i32 3, i32 3, i32 3>
; CHECK-VF4IC4-NEXT: [[VEC_ICMP4:%.*]] = icmp eq <4 x i32> {{.*}}, <i32 3, i32 3, i32 3, i32 3>
-; CHECK-VF4IC4-NEXT: [[VEC_SEL1:%.*]] = select <4 x i1> [[VEC_ICMP1]], <4 x i32> [[VEC_PHI1]], <4 x i32> <i32 7, i32 7, i32 7, i32 7>
-; CHECK-VF4IC4-NEXT: [[VEC_SEL2:%.*]] = select <4 x i1> [[VEC_ICMP2]], <4 x i32> [[VEC_PHI2]], <4 x i32> <i32 7, i32 7, i32 7, i32 7>
-; CHECK-VF4IC4-NEXT: [[VEC_SEL3:%.*]] = select <4 x i1> [[VEC_ICMP3]], <4 x i32> [[VEC_PHI3]], <4 x i32> <i32 7, i32 7, i32 7, i32 7>
-; CHECK-VF4IC4-NEXT: [[VEC_SEL4:%.*]] = select <4 x i1> [[VEC_ICMP4]], <4 x i32> [[VEC_PHI4]], <4 x i32> <i32 7, i32 7, i32 7, i32 7>
+; CHECK-VF4IC4-NEXT: [[NOT1:%.*]] = xor <4 x i1> [[VEC_ICMP1]], <i1 true, i1 true, i1 true, i1 true>
+; CHECK-VF4IC4-NEXT: [[NOT2:%.*]] = xor <4 x i1> [[VEC_ICMP2]], <i1 true, i1 true, i1 true, i1 true>
+; CHECK-VF4IC4-NEXT: [[NOT3:%.*]] = xor <4 x i1> [[VEC_ICMP3]], <i1 true, i1 true, i1 true, i1 true>
+; CHECK-VF4IC4-NEXT: [[NOT4:%.*]] = xor <4 x i1> [[VEC_ICMP4]], <i1 true, i1 true, i1 true, i1 true>
+; CHECK-VF4IC4-NEXT: [[VEC_SEL1:%.*]] = or <4 x i1> [[VEC_PHI1]], [[NOT1]]
+; CHECK-VF4IC4-NEXT: [[VEC_SEL2:%.*]] = or <4 x i1> [[VEC_PHI2]], [[NOT2]]
+; CHECK-VF4IC4-NEXT: [[VEC_SEL3:%.*]] = or <4 x i1> [[VEC_PHI3]], [[NOT3]]
+; CHECK-VF4IC4-NEXT: [[VEC_SEL4:%.*]] = or <4 x i1> [[VEC_PHI4]], [[NOT4]]
; CHECK-VF4IC4: middle.block:
-; CHECK-VF4IC4-NEXT: [[VEC_ICMP5:%.*]] = icmp ne <4 x i32> [[VEC_SEL1]], <i32 3, i32 3, i32 3, i32 3>
-; CHECK-VF4IC4-NEXT: [[VEC_SEL5:%.*]] = select <4 x i1> [[VEC_ICMP5]], <4 x i32> [[VEC_SEL1]], <4 x i32> [[VEC_SEL2]]
-; CHECK-VF4IC4-NEXT: [[VEC_ICMP6:%.*]] = icmp ne <4 x i32> [[VEC_SEL5]], <i32 3, i32 3, i32 3, i32 3>
-; CHECK-VF4IC4-NEXT: [[VEC_SEL6:%.*]] = select <4 x i1> [[VEC_ICMP6]], <4 x i32> [[VEC_SEL5]], <4 x i32> [[VEC_SEL3]]
-; CHECK-VF4IC4-NEXT: [[VEC_ICMP7:%.*]] = icmp ne <4 x i32> [[VEC_SEL6]], <i32 3, i32 3, i32 3, i32 3>
-; CHECK-VF4IC4-NEXT: [[VEC_SEL_FIN:%.*]] = select <4 x i1> [[VEC_ICMP7]], <4 x i32> [[VEC_SEL6]], <4 x i32> [[VEC_SEL4]]
-; CHECK-VF4IC4-NEXT: [[FIN_ICMP:%.*]] = icmp ne <4 x i32> [[VEC_SEL_FIN]], <i32 3, i32 3, i32 3, i32 3>
-; CHECK-VF4IC4-NEXT: [[OR_RDX:%.*]] = call i1 @llvm.vector.reduce.or.v4i1(<4 x i1> [[FIN_ICMP]])
-; CHECK-VF4IC4-NEXT: {{.*}} = select i1 [[OR_RDX]], i32 7, i32 3
+; CHECK-VF4IC4-NEXT: [[VEC_SEL5:%.*]] = or <4 x i1> [[VEC_SEL1]], [[VEC_SEL2]]
+; CHECK-VF4IC4-NEXT: [[VEC_SEL6:%.*]] = or <4 x i1> [[VEC_SEL5]], [[VEC_SEL3]]
+; CHECK-VF4IC4-NEXT: [[VEC_SEL7:%.*]] = or <4 x i1> [[VEC_SEL6]], [[VEC_SEL4]]
+; CHECK-VF4IC4-NEXT: [[OR_RDX:%.*]] = call i1 @llvm.vector.reduce.or.v4i1(<4 x i1> [[VEC_SEL7]])
+; CHECK-VF4IC4-NEXT: [[FR_OR_RDX:%.*]] = freeze i1 [[OR_RDX]]
+; CHECK-VF4IC4-NEXT: {{.*}} = select i1 [[FR_OR_RDX]], i32 7, i32 3
; CHECK-VF1IC4: vector.body:
-; CHECK-VF1IC4: [[VEC_PHI1:%.*]] = phi i32 [ 3, %vector.ph ], [ [[VEC_SEL1:%.*]], %vector.body ]
-; CHECK-VF1IC4-NEXT: [[VEC_PHI2:%.*]] = phi i32 [ 3, %vector.ph ], [ [[VEC_SEL2:%.*]], %vector.body ]
-; CHECK-VF1IC4-NEXT: [[VEC_PHI3:%.*]] = phi i32 [ 3, %vector.ph ], [ [[VEC_SEL3:%.*]], %vector.body ]
-; CHECK-VF1IC4-NEXT: [[VEC_PHI4:%.*]] = phi i32 [ 3, %vector.ph ], [ [[VEC_SEL4:%.*]], %vector.body ]
+; CHECK-VF1IC4: [[VEC_PHI1:%.*]] = phi i1 [ false, %vector.ph ], [ [[VEC_SEL1:%.*]], %vector.body ]
+; CHECK-VF1IC4-NEXT: [[VEC_PHI2:%.*]] = phi i1 [ false, %vector.ph ], [ [[VEC_SEL2:%.*]], %vector.body ]
+; CHECK-VF1IC4-NEXT: [[VEC_PHI3:%.*]] = phi i1 [ false, %vector.ph ], [ [[VEC_SEL3:%.*]], %vector.body ]
+; CHECK-VF1IC4-NEXT: [[VEC_PHI4:%.*]] = phi i1 [ false, %vector.ph ], [ [[VEC_SEL4:%.*]], %vector.body ]
; CHECK-VF1IC4: [[VEC_LOAD1:%.*]] = load i32
; CHECK-VF1IC4-NEXT: [[VEC_LOAD2:%.*]] = load i32
; CHECK-VF1IC4-NEXT: [[VEC_LOAD3:%.*]] = load i32
@@ -52,17 +54,20 @@ define i32 @select_const_i32_from_icmp(ptr nocapture readonly %v, i64 %n) {
; CHECK-VF1IC4-NEXT: [[VEC_ICMP2:%.*]] = icmp eq i32 [[VEC_LOAD2]], 3
; CHECK-VF1IC4-NEXT: [[VEC_ICMP3:%.*]] = icmp eq i32 [[VEC_LOAD3]], 3
; CHECK-VF1IC4-NEXT: [[VEC_ICMP4:%.*]] = icmp eq i32 [[VEC_LOAD4]], 3
-; CHECK-VF1IC4-NEXT: [[VEC_SEL1]] = select i1 [[VEC_ICMP1]], i32 [[VEC_PHI1]], i32 7
-; CHECK-VF1IC4-NEXT: [[VEC_SEL2]] = select i1 [[VEC_ICMP2]], i32 [[VEC_PHI2]], i32 7
-; CHECK-VF1IC4-NEXT: [[VEC_SEL3]] = select i1 [[VEC_ICMP3]], i32 [[VEC_PHI3]], i32 7
-; CHECK-VF1IC4-NEXT: [[VEC_SEL4]] = select i1 [[VEC_ICMP4]], i32 [[VEC_PHI4]], i32 7
+; CHECK-VF1IC4-NEXT: [[NOT1:%.*]] = xor i1 [[VEC_ICMP1]], true
+; CHECK-VF1IC4-NEXT: [[NOT2:%.*]] = xor i1 [[VEC_ICMP2]], true
+; CHECK-VF1IC4-NEXT: [[NOT3:%.*]] = xor i1 [[VEC_ICMP3]], true
+; CHECK-VF1IC4-NEXT: [[NOT4:%.*]] = xor i1 [[VEC_ICMP4]], true
+; CHECK-VF1IC4-NEXT: [[VEC_SEL1:%.*]] = or i1 [[VEC_PHI1]], [[NOT1]]
+; CHECK-VF1IC4-NEXT: [[VEC_SEL2:%.*]] = or i1 [[VEC_PHI2]], [[NOT2]]
+; CHECK-VF1IC4-NEXT: [[VEC_SEL3:%.*]] = or i1 [[VEC_PHI3]], [[NOT3]]
+; CHECK-VF1IC4-NEXT: [[VEC_SEL4:%.*]] = or i1 [[VEC_PHI4]], [[NOT4]]
; CHECK-VF1IC4: middle.block:
-; CHECK-VF1IC4-NEXT: [[VEC_ICMP4:%.*]] = icmp ne i32 [[VEC_SEL1]], 3
-; CHECK-VF1IC4-NEXT: [[VEC_SEL5:%.*]] = select i1 [[VEC_ICMP4]], i32 [[VEC_SEL1]], i32 [[VEC_SEL2]]
-; CHECK-VF1IC4-NEXT: [[VEC_ICMP5:%.*]] = icmp ne i32 [[VEC_SEL5]], 3
-; CHECK-VF1IC4-NEXT: [[VEC_SEL6:%.*]] = select i1 [[VEC_ICMP5]], i32 [[VEC_SEL5]], i32 [[VEC_SEL3]]
-; CHECK-VF1IC4-NEXT: [[VEC_ICMP6:%.*]] = icmp ne i32 [[VEC_SEL6]], 3
-; CHECK-VF1IC4-NEXT: {{.*}} = select i1 [[VEC_ICMP6]], i32 [[VEC_SEL6]], i32 [[VEC_SEL4]]
+; CHECK-VF1IC4-NEXT: [[VEC_SEL5:%.*]] = or i1 [[VEC_SEL1]], [[VEC_SEL2]]
+; CHECK-VF1IC4-NEXT: [[VEC_SEL6:%.*]] = or i1 [[VEC_SEL5]], [[VEC_SEL3]]
+; CHECK-VF1IC4-NEXT: [[OR_RDX:%.*]] = or i1 [[VEC_SEL6]], [[VEC_SEL4]]
+; CHECK-VF1IC4-NEXT: [[FR_OR_RDX:%.*]] = freeze i1 [[OR_RDX]]
+; CHECK-VF1IC4-NEXT: {{.*}} = select i1 [[FR_OR_RDX]], i32 7, i32 3
entry:
br label %for.body
@@ -86,14 +91,14 @@ exit: ; preds = %for.body
define i32 @select_const_i32_from_icmp2(ptr nocapture readonly %v, i64 %n) {
; CHECK-LABEL: @select_const_i32_from_icmp2
; CHECK-VF4IC1: vector.body:
-; CHECK-VF4IC1: [[VEC_PHI:%.*]] = phi <4 x i32> [ <i32 3, i32 3, i32 3, i32 3>, %vector.ph ], [ [[VEC_SEL:%.*]], %vector.body ]
+; CHECK-VF4IC1: [[VEC_PHI:%.*]] = phi <4 x i1> [ zeroinitializer, %vector.ph ], [ [[VEC_SEL:%.*]], %vector.body ]
; CHECK-VF4IC1: [[VEC_LOAD:%.*]] = load <4 x i32>
; CHECK-VF4IC1-NEXT: [[VEC_ICMP:%.*]] = icmp eq <4 x i32> [[VEC_LOAD]], <i32 3, i32 3, i32 3, i32 3>
-; CHECK-VF4IC1-NEXT: [[VEC_SEL]] = select <4 x i1> [[VEC_ICMP]], <4 x i32> <i32 7, i32 7, i32 7, i32 7>, <4 x i32> [[VEC_PHI]]
+; CHECK-VF4IC1-NEXT: [[VEC_SEL]] = or <4 x i1> [[VEC_PHI]], [[VEC_ICMP]]
; CHECK-VF4IC1: middle.block:
-; CHECK-VF4IC1-NEXT: [[FIN_ICMP:%.*]] = icmp ne <4 x i32> [[VEC_SEL]], <i32 3, i32 3, i32 3, i32 3>
-; CHECK-VF4IC1-NEXT: [[OR_RDX:%.*]] = call i1 @llvm.vector.reduce.or.v4i1(<4 x i1> [[FIN_ICMP]])
-; CHECK-VF4IC1-NEXT: {{.*}} = select i1 [[OR_RDX]], i32 7, i32 3
+; CHECK-VF4IC1-NEXT: [[OR_RDX:%.*]] = call i1 @llvm.vector.reduce.or.v4i1(<4 x i1> [[VEC_SEL]])
+; CHECK-VF4IC1-NEXT: [[FR_OR_RDX:%.*]] = freeze i1 [[OR_RDX]]
+; CHECK-VF4IC1-NEXT: {{.*}} = select i1 [[FR_OR_RDX]], i32 7, i32 3
entry:
br label %for.body
@@ -117,21 +122,18 @@ exit: ; preds = %for.body
define i32 @select_i32_from_icmp(ptr nocapture readonly %v, i32 %a, i32 %b, i64 %n) {
; CHECK-LABEL: @select_i32_from_icmp
; CHECK-VF4IC1: vector.ph:
-; CHECK-VF4IC1: [[TMP1:%.*]] = insertelement <4 x i32> poison, i32 %a, i64 0
-; CHECK-VF4IC1-NEXT: [[SPLAT_OF_A:%.*]] = shufflevector <4 x i32> [[TMP1]], <4 x i32> poison, <4 x i32> zeroinitializer
-; CHECK-VF4IC1-NEXT: [[TMP2:%.*]] = insertelement <4 x i32> poison, i32 %b, i64 0
-; CHECK-VF4IC1-NEXT: [[SPLAT_OF_B:%.*]] = shufflevector <4 x i32> [[TMP2]], <4 x i32> poison, <4 x i32> zeroinitializer
+; CHECK-VF4IC1-NOT: shufflevector <4 x i32>
+; CHECK-VF4IC1-NOT: shufflevector <4 x i32>
; CHECK-VF4IC1: vector.body:
-; CHECK-VF4IC1: [[VEC_PHI:%.*]] = phi <4 x i32> [ [[SPLAT_OF_A]], %vector.ph ], [ [[VEC_SEL:%.*]], %vector.body ]
+; CHECK-VF4IC1: [[VEC_PHI:%.*]] = phi <4 x i1> [ zeroinitializer, %vector.ph ], [ [[VEC_SEL:%.*]], %vector.body ]
; CHECK-VF4IC1: [[VEC_LOAD:%.*]] = load <4 x i32>
; CHECK-VF4IC1-NEXT: [[VEC_ICMP:%.*]] = icmp eq <4 x i32> [[VEC_LOAD]], <i32 3, i32 3, i32 3, i32 3>
-; CHECK-VF4IC1-NEXT: [[VEC_SEL]] = select <4 x i1> [[VEC_ICMP]], <4 x i32> [[VEC_PHI]], <4 x i32> [[SPLAT_OF_B]]
+; CHECK-VF4IC1-NEXT: [[NOT:%.*]] = xor <4 x i1> [[VEC_ICMP]], <i1 true, i1 true, i1 true, i1 true>
+; CHECK-VF4IC1-NEXT: [[VEC_SEL]] = or <4 x i1> [[VEC_PHI]], [[NOT]]
; CHECK-VF4IC1: middle.block:
-; CHECK-VF4IC1-NEXT: [[FIN_INS:%.*]] = insertelement <4 x i32> poison, i32 %a, i64 0
-; CHECK-VF4IC1-NEXT: [[FIN_SPLAT:%.*]] = shufflevector <4 x i32> [[FIN_INS]], <4 x i32> poison, <4 x i32> zeroinitializer
-; CHECK-VF4IC1-NEXT: [[FIN_CMP:%.*]] = icmp ne <4 x i32> [[VEC_SEL]], [[FIN_SPLAT]]
-; CHECK-VF4IC1-NEXT: [[OR_RDX:%.*]] = call i1 @llvm.vector.reduce.or.v4i1(<4 x i1> [[FIN_CMP]])
-; CHECK-VF4IC1-NEXT: {{.*}} = select i1 [[OR_RDX]], i32 %b, i32 %a
+; CHECK-VF4IC1-NEXT: [[OR_RDX:%.*]] = call i1 @llvm.vector.reduce.or.v4i1(<4 x i1> [[VEC_SEL]])
+; CHECK-VF4IC1-NEXT: [[FR_OR_RDX:%.*]] = freeze i1 [[OR_RDX]]
+; CHECK-VF4IC1-NEXT: {{.*}} = select i1 [[FR_OR_RDX]], i32 %b, i32 %a
entry:
br label %for.body
@@ -154,14 +156,15 @@ exit: ; preds = %for.body
define i32 @select_const_i32_from_fcmp_fast(ptr nocapture readonly %v, i64 %n) {
; CHECK-LABEL: @select_const_i32_from_fcmp_fast
; CHECK-VF4IC1: vector.body:
-; CHECK-VF4IC1: [[VEC_PHI:%.*]] = phi <4 x i32> [ <i32 2, i32 2, i32 2, i32 2>, %vector.ph ], [ [[VEC_SEL:%.*]], %vector.body ]
+; CHECK-VF4IC1: [[VEC_PHI:%.*]] = phi <4 x i1> [ zeroinitializer, %vector.ph ], [ [[VEC_SEL:%.*]], %vector.body ]
; CHECK-VF4IC1: [[VEC_LOAD:%.*]] = load <4 x float>
; CHECK-VF4IC1-NEXT: [[VEC_FCMP:%.*]] = fcmp fast ueq <4 x float> [[VEC_LOAD]], <float 3.000000e+00, float 3.000000e+00, float 3.000000e+00, float 3.000000e+00>
-; CHECK-VF4IC1-NEXT: [[VEC_SEL]] = select <4 x i1> [[VEC_FCMP]], <4 x i32> [[VEC_PHI]], <4 x i32> <i32 1, i32 1, i32 1, i32 1>
+; CHECK-VF4IC1-NEXT: [[NOT:%.*]] = xor <4 x i1> [[VEC_FCMP]], <i1 true, i1 true, i1 true, i1 true>
+; CHECK-VF4IC1-NEXT: [[VEC_SEL]] = or <4 x i1> [[VEC_PHI]], [[NOT]]
; CHECK-VF4IC1: middle.block:
-; CHECK-VF4IC1-NEXT: [[FIN_ICMP:%.*]] = icmp ne <4 x i32> [[VEC_SEL]], <i32 2, i32 2, i32 2, i32 2>
-; CHECK-VF4IC1-NEXT: [[OR_RDX:%.*]] = call i1 @llvm.vector.reduce.or.v4i1(<4 x i1> [[FIN_ICMP]])
-; CHECK-VF4IC1-NEXT: {{.*}} = select i1 [[OR_RDX]], i32 1, i32 2
+; CHECK-VF4IC1-NEXT: [[OR_RDX:%.*]] = call i1 @llvm.vector.reduce.or.v4i1(<4 x i1> [[VEC_SEL]])
+; CHECK-VF4IC1-NEXT: [[FR_OR_RDX:%.*]] = freeze i1 [[OR_RDX]]
+; CHECK-VF4IC1-NEXT: {{.*}} = select i1 [[FR_OR_RDX]], i32 1, i32 2
entry:
br label %for.body
@@ -184,14 +187,15 @@ exit: ; preds = %for.body
define i32 @select_const_i32_from_fcmp(ptr nocapture readonly %v, i64 %n) {
; CHECK-LABEL: @select_const_i32_from_fcmp
; CHECK-VF4IC1: vector.body:
-; CHECK-VF4IC1: [[VEC_PHI:%.*]] = phi <4 x i32> [ <i32 2, i32 2, i32 2, i32 2>, %vector.ph ], [ [[VEC_SEL:%.*]], %vector.body ]
+; CHECK-VF4IC1: [[VEC_PHI:%.*]] = phi <4 x i1> [ zeroinitializer, %vector.ph ], [ [[VEC_SEL:%.*]], %vector.body ]
; CHECK-VF4IC1: [[VEC_LOAD:%.*]] = load <4 x float>
; CHECK-VF4IC1-NEXT: [[VEC_FCMP:%.*]] = fcmp ueq <4 x float> [[VEC_LOAD]], <float 3.000000e+00, float 3.000000e+00, float 3.000000e+00, float 3.000000e+00>
-; CHECK-VF4IC1-NEXT: [[VEC_SEL]] = select <4 x i1> [[VEC_FCMP]], <4 x i32> [[VEC_PHI]], <4 x i32> <i32 1, i32 1, i32 1, i32 1>
+; CHECK-VF4IC1-NEXT: [[NOT:%.*]] = xor <4 x i1> [[VEC_FCMP]], <i1 true, i1 true, i1 true, i1 true>
+; CHECK-VF4IC1-NEXT: [[VEC_SEL]] = or <4 x i1> [[VEC_PHI]], [[NOT]]
; CHECK-VF4IC1: middle.block:
-; CHECK-VF4IC1-NEXT: [[FIN_ICMP:%.*]] = icmp ne <4 x i32> [[VEC_SEL]], <i32 2, i32 2, i32 2, i32 2>
-; CHECK-VF4IC1-NEXT: [[OR_RDX:%.*]] = call i1 @llvm.vector.reduce.or.v4i1(<4 x i1> [[FIN_ICMP]])
-; CHECK-VF4IC1-NEXT: {{.*}} = select i1 [[OR_RDX]], i32 1, i32 2
+; CHECK-VF4IC1-NEXT: [[OR_RDX:%.*]] = call i1 @llvm.vector.reduce.or.v4i1(<4 x i1> [[VEC_SEL]])
+; CHECK-VF4IC1-NEXT: [[FR_OR_RDX:%.*]] = freeze i1 [[OR_RDX]]
+; CHECK-VF4IC1-NEXT: {{.*}} = select i1 [[FR_OR_RDX]], i32 1, i32 2
entry:
br label %for.body
@@ -216,18 +220,16 @@ define i32 @select_i32_from_icmp_same_inputs(i32 %a, i32 %b, i64 %n) {
; CHECK-VF4IC1: vector.ph:
; CHECK-VF4IC1: [[TMP1:%.*]] = insertelement <4 x i32> poison, i32 %a, i64 0
; CHECK-VF4IC1-NEXT: [[SPLAT_OF_A:%.*]] = shufflevector <4 x i32> [[TMP1]], <4 x i32> poison, <4 x i32> zeroinitializer
-; CHECK-VF4IC1-NEXT: [[TMP2:%.*]] = insertelement <4 x i32> poison, i32 %b, i64 0
-; CHECK-VF4IC1-NEXT: [[SPLAT_OF_B:%.*]] = shufflevector <4 x i32> [[TMP2]], <4 x i32> poison, <4 x i32> zeroinitializer
+; CHECK-VF4IC1-NOT: [[TMP2:%.*]] = insertelement <4 x i32> poison, i32 %b, i64 0
; CHECK-VF4IC1: vector.body:
-; CHECK-VF4IC1: [[VEC_PHI:%.*]] = phi <4 x i32> [ [[SPLAT_OF_A]], %vector.ph ], [ [[VEC_SEL:%.*]], %vector.body ]
-; CHECK-VF4IC1: [[VEC_ICMP:%.*]] = icmp eq <4 x i32> [[VEC_PHI]], <i32 3, i32 3, i32 3, i32 3>
-; CHECK-VF4IC1-NEXT: [[VEC_SEL]] = select <4 x i1> [[VEC_ICMP]], <4 x i32> [[VEC_PHI]], <4 x i32> [[SPLAT_OF_B]]
+; CHECK-VF4IC1: [[VEC_PHI:%.*]] = phi <4 x i1> [ zeroinitializer, %vector.ph ], [ [[VEC_SEL:%.*]], %vector.body ]
+; CHECK-VF4IC1: [[VEC_ICMP:%.*]] = icmp eq <4 x i32> [[SPLAT_OF_A]], <i32 3, i32 3, i32 3, i32 3>
+; CHECK-VF4IC1-NEXT: [[NOT:%.*]] = xor <4 x i1> [[VEC_ICMP]], <i1 true, i1 true, i1 true, i1 true>
+; CHECK-VF4IC1-NEXT: [[VEC_SEL]] = or <4 x i1> [[VEC_PHI]], [[NOT]]
; CHECK-VF4IC1: middle.block:
-; CHECK-VF4IC1-NEXT: [[FIN_INS:%.*]] = insertelement <4 x i32> poison, i32 %a, i64 0
-; CHECK-VF4IC1-NEXT: [[FIN_SPLAT:%.*]] = shufflevector <4 x i32> [[FIN_INS]], <4 x i32> poison, <4 x i32> zeroinitializer
-; CHECK-VF4IC1-NEXT: [[FIN_CMP:%.*]] = icmp ne <4 x i32> [[VEC_SEL]], [[FIN_SPLAT]]
-; CHECK-VF4IC1-NEXT: [[OR_RDX:%.*]] = call i1 @llvm.vector.reduce.or.v4i1(<4 x i1> [[FIN_CMP]])
-; CHECK-VF4IC1-NEXT: {{.*}} = select i1 [[OR_RDX]], i32 %b, i32 %a
+; CHECK-VF4IC1-NEXT: [[OR_RDX:%.*]] = call i1 @llvm.vector.reduce.or.v4i1(<4 x i1> [[VEC_SEL]])
+; CHECK-VF4IC1-NEXT: [[FR_OR_RDX:%.*]] = freeze i1 [[OR_RDX]]
+; CHECK-VF4IC1-NEXT: {{.*}} = select i1 [[FR_OR_RDX]], i32 %b, i32 %a
entry:
br label %for.body
diff --git a/llvm/test/Transforms/LoopVectorize/select-reduction-start-value-may-be-undef-or-poison.ll b/llvm/test/Transforms/LoopVectorize/select-reduction-start-value-may-be-undef-or-poison.ll
index 16ab45415b5cc..55e61158a79c6 100644
--- a/llvm/test/Transforms/LoopVectorize/select-reduction-start-value-may-be-undef-or-poison.ll
+++ b/llvm/test/Transforms/LoopVectorize/select-reduction-start-value-may-be-undef-or-poison.ll
@@ -8,26 +8,25 @@ define i64 @pr62565_incoming_value_known_undef(i64 %a, ptr %src) {
; CHECK-NEXT: entry:
; CHECK-NEXT: br i1 false, label [[SCALAR_PH:%.*]], label [[VECTOR_PH:%.*]]
; CHECK: vector.ph:
-; CHECK-NEXT: [[BROADCAST_SPLATINSERT:%.*]] = insertelement <2 x i64> poison, i64 [[A]], i64 0
-; CHECK-NEXT: [[BROADCAST_SPLAT:%.*]] = shufflevector <2 x i64> [[BROADCAST_SPLATINSERT]], <2 x i64> poison, <2 x i32> zeroinitializer
; CHECK-NEXT: br label [[VECTOR_BODY:%.*]]
; CHECK: vector.body:
; CHECK-NEXT: [[INDEX:%.*]] = phi i32 [ 0, [[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ]
-; CHECK-NEXT: [[VEC_PHI:%.*]] = phi <2 x i64> [ undef, [[VECTOR_PH]] ], [ [[TMP4:%.*]], [[VECTOR_BODY]] ]
+; CHECK-NEXT: [[VEC_PHI:%.*]] = phi <2 x i1> [ zeroinitializer, [[VECTOR_PH]] ], [ [[TMP4:%.*]], [[VECTOR_BODY]] ]
; CHECK-NEXT: [[OFFSET_IDX:%.*]] = add i32 1, [[INDEX]]
; CHECK-NEXT: [[TMP0:%.*]] = add i32 [[OFFSET_IDX]], 0
; CHECK-NEXT: [[TMP1:%.*]] = getelementptr inbounds i32, ptr [[SRC]], i32 [[TMP0]]
; CHECK-NEXT: [[TMP2:%.*]] = getelementptr inbounds i32, ptr [[TMP1]], i32 0
; CHECK-NEXT: [[WIDE_LOAD:%.*]] = load <2 x i32>, ptr [[TMP2]], align 4
; CHECK-NEXT: [[TMP3:%.*]] = icmp eq <2 x i32> [[WIDE_LOAD]], <i32 1, i32 1>
-; CHECK-NEXT: [[TMP4]] = select <2 x i1> [[TMP3]], <2 x i64> [[VEC_PHI]], <2 x i64> [[BROADCAST_SPLAT]]
+; CHECK-NEXT: [[NOT:%*]] = xor <2 x i1> [[TMP3]], <i1 true, i1 true>
+; CHECK-NEXT: [[TMP4]] = or <2 x i1> [[VEC_PHI]], [[NOT]]
; CHECK-NEXT: [[INDEX_NEXT]] = add nuw i32 [[INDEX]], 2
; CHECK-NEXT: [[TMP5:%.*]] = icmp eq i32 [[INDEX_NEXT]], 32
; CHECK-NEXT: br i1 [[TMP5]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]]
; CHECK: middle.block:
-; CHECK-NEXT: [[RDX_SELECT_CMP:%.*]] = icmp ne <2 x i64> [[TMP4]], undef
-; CHECK-NEXT: [[TMP6:%.*]] = call i1 @llvm.vector.reduce.or.v2i1(<2 x i1> [[RDX_SELECT_CMP]])
-; CHECK-NEXT: [[RDX_SELECT:%.*]] = select i1 [[TMP6]], i64 [[A]], i64 undef
+; CHECK-NEXT: [[TMP6:%.*]] = call i1 @llvm.vector.reduce.or.v2i1(<2 x i1> [[TMP4]])
+; CHECK-NEXT: [[FR_TMP6:%.*]] = freeze i1 [[TMP6]]
+; CHECK-NEXT: [[RDX_SELECT:%.*]] = select i1 [[FR_TMP6]], i64 [[A]], i64 undef
; CHECK-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]]
; CHECK: scalar.ph:
; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i32 [ 33, [[MIDDLE_BLOCK]] ], [ 1, [[ENTRY:%.*]] ]
@@ -72,26 +71,25 @@ define i64 @pr62565_incoming_value_known_poison(i64 %a, ptr %src) {
; CHECK-NEXT: entry:
; CHECK-NEXT: br i1 false, label [[SCALAR_PH:%.*]], label [[VECTOR_PH:%.*]]
; CHECK: vector.ph:
-; CHECK-NEXT: [[BROADCAST_SPLATINSERT:%.*]] = insertelement <2 x i64> poison, i64 [[A]], i64 0
-; CHECK-NEXT: [[BROADCAST_SPLAT:%.*]] = shufflevector <2 x i64> [[BROADCAST_SPLATINSERT]], <2 x i64> poison, <2 x i32> zeroinitializer
; CHECK-NEXT: br label [[VECTOR_BODY:%.*]]
; CHECK: vector.body:
; CHECK-NEXT: [[INDEX:%.*]] = phi i32 [ 0, [[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ]
-; CHECK-NEXT: [[VEC_PHI:%.*]] = phi <2 x i64> [ poison, [[VECTOR_PH]] ], [ [[TMP4:%.*]], [[VECTOR_BODY]] ]
+; CHECK-NEXT: [[VEC_PHI:%.*]] = phi <2 x i1> [ zeroinitializer, [[VECTOR_PH]] ], [ [[TMP4:%.*]], [[VECTOR_BODY]] ]
; CHECK-NEXT: [[OFFSET_IDX:%.*]] = add i32 1, [[INDEX]]
; CHECK-NEXT: [[TMP0:%.*]] = add i32 [[OFFSET_IDX]], 0
; CHECK-NEXT: [[TMP1:%.*]] = getelementptr inbounds i32, ptr [[SRC]], i32 [[TMP0]]
; CHECK-NEXT: [[TMP2:%.*]] = getelementptr inbounds i32, ptr [[TMP1]], i32 0
; CHECK-NEXT: [[WIDE_LOAD:%.*]] = load <2 x i32>, ptr [[TMP2]], align 4
; CHECK-NEXT: [[TMP3:%.*]] = icmp eq <2 x i32> [[WIDE_LOAD]], <i32 1, i32 1>
-; CHECK-NEXT: [[TMP4]] = select <2 x i1> [[TMP3]], <2 x i64> [[VEC_PHI]], <2 x i64> [[BROADCAST_SPLAT]]
+; CHECK-NEXT: [[NOT:%.*]] = xor <2 x i1> [[TMP3]], <i1 true, i1 true>
+; CHECK-NEXT: [[TMP4]] = or <2 x i1> [[VEC_PHI]], [[NOT]]
; CHECK-NEXT: [[INDEX_NEXT]] = add nuw i32 [[INDEX]], 2
; CHECK-NEXT: [[TMP5:%.*]] = icmp eq i32 [[INDEX_NEXT]], 32
; CHECK-NEXT: br i1 [[TMP5]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP4:![0-9]+]]
; CHECK: middle.block:
-; CHECK-NEXT: [[RDX_SELECT_CMP:%.*]] = icmp ne <2 x i64> [[TMP4]], poison
-; CHECK-NEXT: [[TMP6:%.*]] = call i1 @llvm.vector.reduce.or.v2i1(<2 x i1> [[RDX_SELECT_CMP]])
-; CHECK-NEXT: [[RDX_SELECT:%.*]] = select i1 [[TMP6]], i64 [[A]], i64 poison
+; CHECK-NEXT: [[TMP6:%.*]] = call i1 @llvm.vector.reduce.or.v2i1(<2 x i1> [[TMP4]])
+; CHECK-NEXT: [[FR_TMP6:%.*]] = freeze i1 [[TMP6]]
+; CHECK-NEXT: [[RDX_SELECT:%.*]] = select i1 [[FR_TMP6]], i64 [[A]], i64 poison
; CHECK-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]]
; CHECK: scalar.ph:
; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i32 [ 33, [[MIDDLE_BLOCK]] ], [ 1, [[ENTRY:%.*]] ]
@@ -136,30 +134,25 @@ define i64 @pr62565_incoming_value_may_be_poison(i64 %a, ptr %src, i64 %start) {
; CHECK-NEXT: entry:
; CHECK-NEXT: br i1 false, label [[SCALAR_PH:%.*]], label [[VECTOR_PH:%.*]]
; CHECK: vector.ph:
-; CHECK-NEXT: [[MINMAX_IDENT_SPLATINSERT:%.*]] = insertelement <2 x i64> poison, i64 [[START]], i64 0
-; CHECK-NEXT: [[MINMAX_IDENT_SPLAT:%.*]] = shufflevector <2 x i64> [[MINMAX_IDENT_SPLATINSERT]], <2 x i64> poison, <2 x i32> zeroinitializer
-; CHECK-NEXT: [[BROADCAST_SPLATINSERT:%.*]] = insertelement <2 x i64> poison, i64 [[A]], i64 0
-; CHECK-NEXT: [[BROADCAST_SPLAT:%.*]] = shufflevector <2 x i64> [[BROADCAST_SPLATINSERT]], <2 x i64> poison, <2 x i32> zeroinitializer
; CHECK-NEXT: br label [[VECTOR_BODY:%.*]]
; CHECK: vector.body:
; CHECK-NEXT: [[INDEX:%.*]] = phi i32 [ 0, [[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ]
-; CHECK-NEXT: [[VEC_PHI:%.*]] = phi <2 x i64> [ [[MINMAX_IDENT_SPLAT]], [[VECTOR_PH]] ], [ [[TMP4:%.*]], [[VECTOR_BODY]] ]
+; CHECK-NEXT: [[VEC_PHI:%.*]] = phi <2 x i1> [ zeroinitializer, [[VECTOR_PH]] ], [ [[TMP4:%.*]], [[VECTOR_BODY]] ]
; CHECK-NEXT: [[OFFSET_IDX:%.*]] = add i32 1, [[INDEX]]
; CHECK-NEXT: [[TMP0:%.*]] = add i32 [[OFFSET_IDX]], 0
; CHECK-NEXT: [[TMP1:%.*]] = getelementptr inbounds i32, ptr [[SRC]], i32 [[TMP0]]
; CHECK-NEXT: [[TMP2:%.*]] = getelementptr inbounds i32, ptr [[TMP1]], i32 0
; CHECK-NEXT: [[WIDE_LOAD:%.*]] = load <2 x i32>, ptr [[TMP2]], align 4
; CHECK-NEXT: [[TMP3:%.*]] = icmp eq <2 x i32> [[WIDE_LOAD]], <i32 1, i32 1>
-; CHECK-NEXT: [[TMP4]] = select <2 x i1> [[TMP3]], <2 x i64> [[VEC_PHI]], <2 x i64> [[BROADCAST_SPLAT]]
+; CHECK-NEXT: [[NOT:%.*]] = xor <2 x i1> [[TMP3]], <i1 true, i1 true>
+; CHECK-NEXT: [[TMP4]] = or <2 x i1> [[VEC_PHI]], [[NOT]]
; CHECK-NEXT: [[INDEX_NEXT]] = add nuw i32 [[INDEX]], 2
; CHECK-NEXT: [[TMP5:%.*]] = icmp eq i32 [[INDEX_NEXT]], 32
; CHECK-NEXT: br i1 [[TMP5]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP6:![0-9]+]]
; CHECK: middle.block:
-; CHECK-NEXT: [[DOTSPLATINSERT:%.*]] = insertelement <2 x i64> poison, i64 [[START]], i64 0
-; CHECK-NEXT: [[DOTSPLAT:%.*]] = shufflevector <2 x i64> [[DOTSPLATINSERT]], <2 x i64> poison, <2 x i32> zeroinitializer
-; CHECK-NEXT: [[RDX_SELECT_CMP:%.*]] = icmp ne <2 x i64> [[TMP4]], [[DOTSPLAT]]
-; CHECK-NEXT: [[TMP6:%.*]] = call i1 @llvm.vector.reduce.or.v2i1(<2 x i1> [[RDX_SELECT_CMP]])
-; CHECK-NEXT: [[RDX_SELECT:%.*]] = select i1 [[TMP6]], i64 [[A]], i64 [[START]]
+; CHECK-NEXT: [[TMP6:%.*]] = call i1 @llvm.vector.reduce.or.v2i1(<2 x i1> [[TMP4]])
+; CHECK-NEXT: [[FR_TMP6:%.*]] = freeze i1 [[TMP6]]
+; CHECK-NEXT: [[RDX_SELECT:%.*]] = select i1 [[FR_TMP6]], i64 [[A]], i64 [[START]]
; CHECK-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]]
; CHECK: scalar.ph:
; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i32 [ 33, [[MIDDLE_BLOCK]] ], [ 1, [[ENTRY:%.*]] ]
>From 0057eb2c33205179cf277f03ffc440e2bd20fd0d Mon Sep 17 00:00:00 2001
From: Florian Hahn <flo at fhahn.com>
Date: Mon, 29 Jan 2024 11:23:42 +0000
Subject: [PATCH 2/3] !fixup update remainng tests.
---
.../LoopVectorize/AArch64/sve-select-cmp.ll | 56 +++++++++-------
.../RISCV/select-cmp-reduction.ll | 66 ++++++++++++-------
2 files changed, 76 insertions(+), 46 deletions(-)
diff --git a/llvm/test/Transforms/LoopVectorize/AArch64/sve-select-cmp.ll b/llvm/test/Transforms/LoopVectorize/AArch64/sve-select-cmp.ll
index 493ffd2ca569d..b43774446b589 100644
--- a/llvm/test/Transforms/LoopVectorize/AArch64/sve-select-cmp.ll
+++ b/llvm/test/Transforms/LoopVectorize/AArch64/sve-select-cmp.ll
@@ -11,35 +11,38 @@ define i32 @select_const_i32_from_icmp(ptr nocapture readonly %v, i64 %n) #0 {
; CHECK-VF4IC1: [[VEC_PHI:%.*]] = phi <vscale x 4 x i1> [ zeroinitializer, %vector.ph ], [ [[VEC_SEL:%.*]], %vector.body ]
; CHECK-VF4IC1: [[VEC_LOAD:%.*]] = load <vscale x 4 x i32>
; CHECK-VF4IC1-NEXT: [[VEC_ICMP:%.*]] = icmp eq <vscale x 4 x i32> [[VEC_LOAD]], shufflevector (<vscale x 4 x i32> insertelement (<vscale x 4 x i32> poison, i32 3, i64 0), <vscale x 4 x i32> poison, <vscale x 4 x i32> zeroinitializer)
-; CHECK-VF4IC1-NEXT: [[VEC_SEL]] = or <vscale x 4 x i1> [[VEC_PHI]], [[VEC_ICMP]]
+; CHECK-VF4IC1-NEXT: [[NOT:%*]] = xor <vscale x 4 x i1> [[VEC_ICMP]], shufflevector (<vscale x 4 x i1> insertelement (<vscale x 4 x i1> poison, i1 true, i64 0), <vscale x 4 x i1> poison, <vscale x 4 x i32> zeroinitializer)
+; CHECK-VF4IC1-NEXT: [[VEC_SEL]] = or <vscale x 4 x i1> [[VEC_PHI]], [[NOT]]
; CHECK-VF4IC1: middle.block:
; CHECK-VF4IC1-NEXT: [[OR_RDX:%.*]] = call i1 @llvm.vector.reduce.or.nxv4i1(<vscale x 4 x i1> [[VEC_SEL]])
-; CHECK-VF4IC1-NEXT: {{.*}} = select i1 [[OR_RDX]], i32 7, i32 3
+; CHECK-VF4IC1-NEXT: [[FR:%.*]] = freeze i1 [[OR_RDX]]
+; CHECK-VF4IC1-NEXT: {{.*}} = select i1 [[FR]], i32 7, i32 3
; CHECK-VF4IC4-LABEL: @select_const_i32_from_icmp
; CHECK-VF4IC4: vector.body:
-; CHECK-VF4IC4: [[VEC_PHI1:%.*]] = phi <vscale x 4 x i32> [ shufflevector (<vscale x 4 x i32> insertelement (<vscale x 4 x i32> poison, i32 3, i64 0), <vscale x 4 x i32> poison, <vscale x 4 x i32> zeroinitializer), %vector.ph ], [ [[VEC_SEL1:%.*]], %vector.body ]
-; CHECK-VF4IC4: [[VEC_PHI2:%.*]] = phi <vscale x 4 x i32> [ shufflevector (<vscale x 4 x i32> insertelement (<vscale x 4 x i32> poison, i32 3, i64 0), <vscale x 4 x i32> poison, <vscale x 4 x i32> zeroinitializer), %vector.ph ], [ [[VEC_SEL2:%.*]], %vector.body ]
-; CHECK-VF4IC4: [[VEC_PHI3:%.*]] = phi <vscale x 4 x i32> [ shufflevector (<vscale x 4 x i32> insertelement (<vscale x 4 x i32> poison, i32 3, i64 0), <vscale x 4 x i32> poison, <vscale x 4 x i32> zeroinitializer), %vector.ph ], [ [[VEC_SEL3:%.*]], %vector.body ]
-; CHECK-VF4IC4: [[VEC_PHI4:%.*]] = phi <vscale x 4 x i32> [ shufflevector (<vscale x 4 x i32> insertelement (<vscale x 4 x i32> poison, i32 3, i64 0), <vscale x 4 x i32> poison, <vscale x 4 x i32> zeroinitializer), %vector.ph ], [ [[VEC_SEL4:%.*]], %vector.body ]
+; CHECK-VF4IC4: [[VEC_PHI1:%.*]] = phi <vscale x 4 x i1> [ zeroinitializer, %vector.ph ], [ [[VEC_SEL1:%.*]], %vector.body ]
+; CHECK-VF4IC4: [[VEC_PHI2:%.*]] = phi <vscale x 4 x i1> [ zeroinitializer, %vector.ph ], [ [[VEC_SEL2:%.*]], %vector.body ]
+; CHECK-VF4IC4: [[VEC_PHI3:%.*]] = phi <vscale x 4 x i1> [ zeroinitializer, %vector.ph ], [ [[VEC_SEL3:%.*]], %vector.body ]
+; CHECK-VF4IC4: [[VEC_PHI4:%.*]] = phi <vscale x 4 x i1> [ zeroinitializer, %vector.ph ], [ [[VEC_SEL4:%.*]], %vector.body ]
; CHECK-VF4IC4: [[VEC_ICMP1:%.*]] = icmp eq <vscale x 4 x i32> {{.*}}, shufflevector (<vscale x 4 x i32> insertelement (<vscale x 4 x i32> poison, i32 3, i64 0), <vscale x 4 x i32> poison, <vscale x 4 x i32> zeroinitializer)
; CHECK-VF4IC4-NEXT: [[VEC_ICMP2:%.*]] = icmp eq <vscale x 4 x i32> {{.*}}, shufflevector (<vscale x 4 x i32> insertelement (<vscale x 4 x i32> poison, i32 3, i64 0), <vscale x 4 x i32> poison, <vscale x 4 x i32> zeroinitializer)
; CHECK-VF4IC4-NEXT: [[VEC_ICMP3:%.*]] = icmp eq <vscale x 4 x i32> {{.*}}, shufflevector (<vscale x 4 x i32> insertelement (<vscale x 4 x i32> poison, i32 3, i64 0), <vscale x 4 x i32> poison, <vscale x 4 x i32> zeroinitializer)
; CHECK-VF4IC4-NEXT: [[VEC_ICMP4:%.*]] = icmp eq <vscale x 4 x i32> {{.*}}, shufflevector (<vscale x 4 x i32> insertelement (<vscale x 4 x i32> poison, i32 3, i64 0), <vscale x 4 x i32> poison, <vscale x 4 x i32> zeroinitializer)
-; CHECK-VF4IC4-NEXT: [[VEC_SEL1]] = select <vscale x 4 x i1> [[VEC_ICMP1]], <vscale x 4 x i32> [[VEC_PHI1]], <vscale x 4 x i32> shufflevector (<vscale x 4 x i32> insertelement (<vscale x 4 x i32> poison, i32 7, i64 0), <vscale x 4 x i32> poison, <vscale x 4 x i32> zeroinitializer)
-; CHECK-VF4IC4-NEXT: [[VEC_SEL2]] = select <vscale x 4 x i1> [[VEC_ICMP2]], <vscale x 4 x i32> [[VEC_PHI2]], <vscale x 4 x i32> shufflevector (<vscale x 4 x i32> insertelement (<vscale x 4 x i32> poison, i32 7, i64 0), <vscale x 4 x i32> poison, <vscale x 4 x i32> zeroinitializer)
-; CHECK-VF4IC4-NEXT: [[VEC_SEL3]] = select <vscale x 4 x i1> [[VEC_ICMP3]], <vscale x 4 x i32> [[VEC_PHI3]], <vscale x 4 x i32> shufflevector (<vscale x 4 x i32> insertelement (<vscale x 4 x i32> poison, i32 7, i64 0), <vscale x 4 x i32> poison, <vscale x 4 x i32> zeroinitializer)
-; CHECK-VF4IC4-NEXT: [[VEC_SEL4]] = select <vscale x 4 x i1> [[VEC_ICMP4]], <vscale x 4 x i32> [[VEC_PHI4]], <vscale x 4 x i32> shufflevector (<vscale x 4 x i32> insertelement (<vscale x 4 x i32> poison, i32 7, i64 0), <vscale x 4 x i32> poison, <vscale x 4 x i32> zeroinitializer)
+; CHECK-VF4IC4-NEXT: [[NOT1:%.*]] = xor <vscale x 4 x i1> [[VEC_ICMP1]], shufflevector (<vscale x 4 x i1> insertelement (<vscale x 4 x i1> poison, i1 true, i64 0), <vscale x 4 x i1> poison, <vscale x 4 x i32> zeroinitializer)
+; CHECK-VF4IC4-NEXT: [[NOT2:%.*]] = xor <vscale x 4 x i1> [[VEC_ICMP2]], shufflevector (<vscale x 4 x i1> insertelement (<vscale x 4 x i1> poison, i1 true, i64 0), <vscale x 4 x i1> poison, <vscale x 4 x i32> zeroinitializer)
+; CHECK-VF4IC4-NEXT: [[NOT3:%.*]] = xor <vscale x 4 x i1> [[VEC_ICMP3]], shufflevector (<vscale x 4 x i1> insertelement (<vscale x 4 x i1> poison, i1 true, i64 0), <vscale x 4 x i1> poison, <vscale x 4 x i32> zeroinitializer)
+; CHECK-VF4IC4-NEXT: [[NOT4:%.*]] = xor <vscale x 4 x i1> [[VEC_ICMP4]], shufflevector (<vscale x 4 x i1> insertelement (<vscale x 4 x i1> poison, i1 true, i64 0), <vscale x 4 x i1> poison, <vscale x 4 x i32> zeroinitializer)
+; CHECK-VF4IC4-NEXT: [[VEC_SEL1:%.*]] = or <vscale x 4 x i1> [[VEC_PHI1]], [[NOT1]]
+; CHECK-VF4IC4-NEXT: [[VEC_SEL2:%.*]] = or <vscale x 4 x i1> [[VEC_PHI2]], [[NOT2]]
+; CHECK-VF4IC4-NEXT: [[VEC_SEL3:%.*]] = or <vscale x 4 x i1> [[VEC_PHI3]], [[NOT3]]
+; CHECK-VF4IC4-NEXT: [[VEC_SEL4:%.*]] = or <vscale x 4 x i1> [[VEC_PHI4]], [[NOT4]]
; CHECK-VF4IC4: middle.block:
-; CHECK-VF4IC4-NEXT: [[VEC_ICMP5:%.*]] = icmp ne <vscale x 4 x i32> [[VEC_SEL1]], shufflevector (<vscale x 4 x i32> insertelement (<vscale x 4 x i32> poison, i32 3, i64 0), <vscale x 4 x i32> poison, <vscale x 4 x i32> zeroinitializer)
-; CHECK-VF4IC4-NEXT: [[VEC_SEL5:%.*]] = select <vscale x 4 x i1> [[VEC_ICMP5]], <vscale x 4 x i32> [[VEC_SEL1]], <vscale x 4 x i32> [[VEC_SEL2]]
-; CHECK-VF4IC4-NEXT: [[VEC_ICMP6:%.*]] = icmp ne <vscale x 4 x i32> [[VEC_SEL5]], shufflevector (<vscale x 4 x i32> insertelement (<vscale x 4 x i32> poison, i32 3, i64 0), <vscale x 4 x i32> poison, <vscale x 4 x i32> zeroinitializer)
-; CHECK-VF4IC4-NEXT: [[VEC_SEL6:%.*]] = select <vscale x 4 x i1> [[VEC_ICMP6]], <vscale x 4 x i32> [[VEC_SEL5]], <vscale x 4 x i32> [[VEC_SEL3]]
-; CHECK-VF4IC4-NEXT: [[VEC_ICMP7:%.*]] = icmp ne <vscale x 4 x i32> [[VEC_SEL6]], shufflevector (<vscale x 4 x i32> insertelement (<vscale x 4 x i32> poison, i32 3, i64 0), <vscale x 4 x i32> poison, <vscale x 4 x i32> zeroinitializer)
-; CHECK-VF4IC4-NEXT: [[VEC_SEL7:%.*]] = select <vscale x 4 x i1> [[VEC_ICMP7]], <vscale x 4 x i32> [[VEC_SEL6]], <vscale x 4 x i32> [[VEC_SEL4]]
-; CHECK-VF4IC4-NEXT: [[FIN_ICMP:%.*]] = icmp ne <vscale x 4 x i32> [[VEC_SEL7]], shufflevector (<vscale x 4 x i32> insertelement (<vscale x 4 x i32> poison, i32 3, i64 0), <vscale x 4 x i32> poison, <vscale x 4 x i32> zeroinitializer)
-; CHECK-VF4IC4-NEXT: [[OR_RDX:%.*]] = call i1 @llvm.vector.reduce.or.nxv4i1(<vscale x 4 x i1> [[FIN_ICMP]])
-; CHECK-VF4IC4-NEXT: {{.*}} = select i1 [[OR_RDX]], i32 7, i32 3
+; CHECK-VF4IC4-NEXT: [[OR1:%.*]] = or <vscale x 4 x i1> [[VEC_SEL1]], [[VEC_SEL2]]
+; CHECK-VF4IC4-NEXT: [[OR2:%.*]] = or <vscale x 4 x i1> [[OR1]], [[VEC_SEL3]]
+; CHECK-VF4IC4-NEXT: [[OR3:%.*]] = or <vscale x 4 x i1> [[OR2]], [[VEC_SEL4]]
+; CHECK-VF4IC4-NEXT: [[OR_RDX:%.*]] = call i1 @llvm.vector.reduce.or.nxv4i1(<vscale x 4 x i1> [[OR3]])
+; CHECK-VF4IC4-NEXT: [[FR:%.*]] = freeze i1 [[OR_RDX]]
+; CHECK-VF4IC4-NEXT: {{.*}} = select i1 [[FR]], i32 7, i32 3
entry:
br label %for.body
@@ -67,10 +70,12 @@ define i32 @select_i32_from_icmp(ptr nocapture readonly %v, i32 %a, i32 %b, i64
; CHECK-VF4IC1: [[VEC_PHI:%.*]] = phi <vscale x 4 x i1> [ zeroinitializer, %vector.ph ], [ [[VEC_SEL:%.*]], %vector.body ]
; CHECK-VF4IC1: [[VEC_LOAD:%.*]] = load <vscale x 4 x i32>
; CHECK-VF4IC1-NEXT: [[VEC_ICMP:%.*]] = icmp eq <vscale x 4 x i32> [[VEC_LOAD]], shufflevector (<vscale x 4 x i32> insertelement (<vscale x 4 x i32> poison, i32 3, i64 0), <vscale x 4 x i32> poison, <vscale x 4 x i32> zeroinitializer)
-; CHECK-VF4IC1-NEXT: [[VEC_SEL]] = or <vscale x 4 x i1> [[VEC_PHI]], [[VEC_ICMP]]
+; CHECK-VF4IC1-NEXT: [[NOT:%*]] = xor <vscale x 4 x i1> [[VEC_ICMP]], shufflevector (<vscale x 4 x i1> insertelement (<vscale x 4 x i1> poison, i1 true, i64 0), <vscale x 4 x i1> poison, <vscale x 4 x i32> zeroinitializer)
+; CHECK-VF4IC1-NEXT: [[VEC_SEL]] = or <vscale x 4 x i1> [[VEC_PHI]], [[NOT]]
; CHECK-VF4IC1: middle.block:
; CHECK-VF4IC1-NEXT: [[OR_RDX:%.*]] = call i1 @llvm.vector.reduce.or.nxv4i1(<vscale x 4 x i1> [[VEC_SEL]])
-; CHECK-VF4IC1-NEXT: {{.*}} = select i1 [[OR_RDX]], i32 %b, i32 %a
+; CHECK-VF4IC1-NEXT: [[FR:%.*]] = freeze i1 [[OR_RDX]]
+; CHECK-VF4IC1-NEXT: {{.*}} = select i1 [[FR]], i32 %b, i32 %a
; CHECK-VF4IC4-LABEL: @select_i32_from_icmp
; CHECK-VF4IC4: vector.body:
@@ -98,10 +103,12 @@ define i32 @select_const_i32_from_fcmp(ptr nocapture readonly %v, i64 %n) #0 {
; CHECK-VF4IC1: [[VEC_PHI:%.*]] = phi <vscale x 4 x i1> [ zeroinitializer, %vector.ph ], [ [[VEC_SEL:%.*]], %vector.body ]
; CHECK-VF4IC1: [[VEC_LOAD:%.*]] = load <vscale x 4 x float>
; CHECK-VF4IC1-NEXT: [[VEC_ICMP:%.*]] = fcmp fast ueq <vscale x 4 x float> [[VEC_LOAD]], shufflevector (<vscale x 4 x float> insertelement (<vscale x 4 x float> poison, float 3.000000e+00, i64 0), <vscale x 4 x float> poison, <vscale x 4 x i32> zeroinitializer)
-; CHECK-VF4IC1-NEXT: [[VEC_SEL]] = or <vscale x 4 x i1> [[VEC_PHI]], [[VEC_ICMP]]
+; CHECK-VF4IC1-NEXT: [[NOT:%*]] = xor <vscale x 4 x i1> [[VEC_ICMP]], shufflevector (<vscale x 4 x i1> insertelement (<vscale x 4 x i1> poison, i1 true, i64 0), <vscale x 4 x i1> poison, <vscale x 4 x i32> zeroinitializer)
+; CHECK-VF4IC1-NEXT: [[VEC_SEL]] = or <vscale x 4 x i1> [[VEC_PHI]], [[NOT]]
; CHECK-VF4IC1: middle.block:
; CHECK-VF4IC1-NEXT: [[OR_RDX:%.*]] = call i1 @llvm.vector.reduce.or.nxv4i1(<vscale x 4 x i1> [[VEC_SEL]])
-; CHECK-VF4IC1-NEXT: {{.*}} = select i1 [[OR_RDX]], i32 1, i32 2
+; CHECK-VF4IC1-NEXT: [[FR:%.*]] = freeze i1 [[OR_RDX]]
+; CHECK-VF4IC1-NEXT: {{.*}} = select i1 [[FR]], i32 1, i32 2
; CHECK-VF4IC4-LABEL: @select_const_i32_from_fcmp
; CHECK-VF4IC4: vector.body:
@@ -158,7 +165,8 @@ define i32 @pred_select_const_i32_from_icmp(ptr noalias nocapture readonly %src1
; CHECK-VF4IC1: [[VEC_SEL:%.*]] = select <vscale x 4 x i1> [[MASK]], <vscale x 4 x i1> [[VEC_SEL_TMP]], <vscale x 4 x i1> [[VEC_PHI]]
; CHECK-VF4IC1: middle.block:
; CHECK-VF4IC1-NEXT: [[OR_RDX:%.*]] = call i1 @llvm.vector.reduce.or.nxv4i1(<vscale x 4 x i1> [[VEC_SEL]])
-; CHECK-VF4IC1-NEXT: {{.*}} = select i1 [[OR_RDX]], i32 1, i32 0
+; CHECK-VF4IC1-NEXT: [[FR:%.*]] = freeze i1 [[OR_RDX]]
+; CHECK-VF4IC1-NEXT: {{.*}} = select i1 [[FR]], i32 1, i32 0
; CHECK-VF4IC4-LABEL: @pred_select_const_i32_from_icmp
; CHECK-VF4IC4: vector.body:
diff --git a/llvm/test/Transforms/LoopVectorize/RISCV/select-cmp-reduction.ll b/llvm/test/Transforms/LoopVectorize/RISCV/select-cmp-reduction.ll
index 6c50915c88e82..a235da069b325 100644
--- a/llvm/test/Transforms/LoopVectorize/RISCV/select-cmp-reduction.ll
+++ b/llvm/test/Transforms/LoopVectorize/RISCV/select-cmp-reduction.ll
@@ -21,13 +21,15 @@ define i32 @select_icmp(i32 %x, i32 %y, ptr nocapture readonly %c, i64 %n) #0 {
; CHECK-NEXT: [[TMP2:%.*]] = getelementptr inbounds i32, ptr [[TMP1]], i32 0
; CHECK-NEXT: [[WIDE_LOAD:%.*]] = load <4 x i32>, ptr [[TMP2]], align 4
; CHECK-NEXT: [[TMP4:%.*]] = icmp slt <4 x i32> [[WIDE_LOAD]], [[BROADCAST_SPLAT]]
-; CHECK-NEXT: [[TMP5]] = or <4 x i1> [[VEC_PHI]], [[TMP4]]
+; CHECK-NEXT: [[NOT:%.*]] = xor <4 x i1> [[TMP4]], <i1 true, i1 true, i1 true, i1 true>
+; CHECK-NEXT: [[TMP5]] = or <4 x i1> [[VEC_PHI]], [[NOT]]
; CHECK-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX]], 4
; CHECK-NEXT: [[TMP6:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]]
; CHECK-NEXT: br i1 [[TMP6]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]]
; CHECK: middle.block:
; CHECK-NEXT: [[TMP7:%.*]] = call i1 @llvm.vector.reduce.or.v4i1(<4 x i1> [[TMP5]])
-; CHECK-NEXT: [[RDX_SELECT:%.*]] = select i1 [[TMP7]], i32 %y, i32 0
+; CHECK-NEXT: [[FR:%.*]] = freeze i1 [[TMP7]]
+; CHECK-NEXT: [[RDX_SELECT:%.*]] = select i1 [[FR]], i32 %y, i32 0
;
; SCALABLE-LABEL: @select_icmp
; SCALABLE: vector.ph:
@@ -48,13 +50,15 @@ define i32 @select_icmp(i32 %x, i32 %y, ptr nocapture readonly %c, i64 %n) #0 {
; SCALABLE-NEXT: [[TMP6:%.*]] = getelementptr inbounds i32, ptr [[TMP5]], i32 0
; SCALABLE-NEXT: [[WIDE_LOAD:%.*]] = load <vscale x 4 x i32>, ptr [[TMP6]], align 4
; SCALABLE-NEXT: [[TMP8:%.*]] = icmp slt <vscale x 4 x i32> [[WIDE_LOAD]], [[BROADCAST_SPLAT]]
-; SCALABLE-NEXT: [[TMP9]] = or <vscale x 4 x i1> [[VEC_PHI]], [[TMP8]]
+; SCALABLE-NEXT: [[NOT:%.*]] = xor <vscale x 4 x i1> [[TMP8]], shufflevector (<vscale x 4 x i1> insertelement (<vscale x 4 x i1> poison, i1 true, i64 0), <vscale x 4 x i1> poison, <vscale x 4 x i32> zeroinitializer)
+; SCALABLE-NEXT: [[TMP9]] = or <vscale x 4 x i1> [[VEC_PHI]], [[NOT]]
; SCALABLE-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX]], [[TMP11]]
; SCALABLE-NEXT: [[TMP12:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]]
; SCALABLE-NEXT: br i1 [[TMP12]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]]
; SCALABLE: middle.block:
; SCALABLE-NEXT: [[TMP13:%.*]] = call i1 @llvm.vector.reduce.or.nxv4i1(<vscale x 4 x i1> [[TMP9]])
-; SCALABLE-NEXT: [[RDX_SELECT:%.*]] = select i1 [[TMP13]], i32 %y, i32 0
+; SCALABLE-NEXT: [[FR:%.*]] = freeze i1 [[TMP13]]
+; SCALABLE-NEXT: [[RDX_SELECT:%.*]] = select i1 [[FR]], i32 %y, i32 0
;
entry:
br label %for.body
@@ -90,13 +94,15 @@ define i32 @select_fcmp(float %x, i32 %y, ptr nocapture readonly %c, i64 %n) #0
; CHECK-NEXT: [[TMP2:%.*]] = getelementptr inbounds float, ptr [[TMP1]], i32 0
; CHECK-NEXT: [[WIDE_LOAD:%.*]] = load <4 x float>, ptr [[TMP2]], align 4
; CHECK-NEXT: [[TMP4:%.*]] = fcmp fast olt <4 x float> [[WIDE_LOAD]], [[BROADCAST_SPLAT]]
-; CHECK-NEXT: [[TMP5]] = or <4 x i1> [[VEC_PHI]], [[TMP4]]
+; CHECK-NEXT: [[NOT:%.*]] = xor <4 x i1> [[TMP4]], <i1 true, i1 true, i1 true, i1 true>
+; CHECK-NEXT: [[TMP5]] = or <4 x i1> [[VEC_PHI]], [[NOT]]
; CHECK-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX]], 4
; CHECK-NEXT: [[TMP6:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]]
; CHECK-NEXT: br i1 [[TMP6]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP4:![0-9]+]]
; CHECK: middle.block:
; CHECK-NEXT: [[TMP7:%.*]] = call i1 @llvm.vector.reduce.or.v4i1(<4 x i1> [[TMP5]])
-; CHECK-NEXT: [[RDX_SELECT:%.*]] = select i1 [[TMP7]], i32 %y, i32 0
+; CHECK-NEXT: [[FR:%.*]] = freeze i1 [[TMP7]]
+; CHECK-NEXT: [[RDX_SELECT:%.*]] = select i1 [[FR]], i32 %y, i32 0
;
; SCALABLE-LABEL: @select_fcmp
; SCALABLE: vector.ph:
@@ -117,13 +123,15 @@ define i32 @select_fcmp(float %x, i32 %y, ptr nocapture readonly %c, i64 %n) #0
; SCALABLE-NEXT: [[TMP6:%.*]] = getelementptr inbounds float, ptr [[TMP5]], i32 0
; SCALABLE-NEXT: [[WIDE_LOAD:%.*]] = load <vscale x 4 x float>, ptr [[TMP6]], align 4
; SCALABLE-NEXT: [[TMP8:%.*]] = fcmp fast olt <vscale x 4 x float> [[WIDE_LOAD]], [[BROADCAST_SPLAT]]
-; SCALABLE-NEXT: [[TMP9]] = or <vscale x 4 x i1> [[VEC_PHI]], [[TMP8]]
+; SCALABLE-NEXT: [[NOT:%.*]] = xor <vscale x 4 x i1> [[TMP8]], shufflevector (<vscale x 4 x i1> insertelement (<vscale x 4 x i1> poison, i1 true, i64 0), <vscale x 4 x i1> poison, <vscale x 4 x i32> zeroinitializer)
+; SCALABLE-NEXT: [[TMP9]] = or <vscale x 4 x i1> [[VEC_PHI]], [[NOT]]
; SCALABLE-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX]], [[TMP11]]
; SCALABLE-NEXT: [[TMP12:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]]
; SCALABLE-NEXT: br i1 [[TMP12]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP4:![0-9]+]]
; SCALABLE: middle.block:
; SCALABLE-NEXT: [[TMP13:%.*]] = call i1 @llvm.vector.reduce.or.nxv4i1(<vscale x 4 x i1> [[TMP9]])
-; SCALABLE-NEXT: [[RDX_SELECT:%.*]] = select i1 [[TMP13]], i32 %y, i32 0
+; SCALABLE-NEXT: [[FR:%.*]] = freeze i1 [[TMP13]]
+; SCALABLE-NEXT: [[RDX_SELECT:%.*]] = select i1 [[FR]], i32 %y, i32 0
;
entry:
br label %for.body
@@ -157,13 +165,15 @@ define i32 @select_const_i32_from_icmp(ptr nocapture readonly %v, i64 %n) #0 {
; CHECK-NEXT: [[TMP2:%.*]] = getelementptr inbounds i32, ptr [[TMP1]], i32 0
; CHECK-NEXT: [[WIDE_LOAD:%.*]] = load <4 x i32>, ptr [[TMP2]], align 4
; CHECK-NEXT: [[TMP4:%.*]] = icmp eq <4 x i32> [[WIDE_LOAD]], <i32 3, i32 3, i32 3, i32 3>
-; CHECK-NEXT: [[TMP5]] = or <4 x i1> [[VEC_PHI]], [[TMP4]]
+; CHECK-NEXT: [[NOT:%.*]] = xor <4 x i1> [[TMP4]], <i1 true, i1 true, i1 true, i1 true>
+; CHECK-NEXT: [[TMP5]] = or <4 x i1> [[VEC_PHI]], [[NOT]]
; CHECK-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX]], 4
; CHECK-NEXT: [[TMP6:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]]
; CHECK-NEXT: br i1 [[TMP6]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP6:![0-9]+]]
; CHECK: middle.block:
; CHECK-NEXT: [[TMP7:%.*]] = call i1 @llvm.vector.reduce.or.v4i1(<4 x i1> [[TMP5]])
-; CHECK-NEXT: [[RDX_SELECT:%.*]] = select i1 [[TMP7]], i32 7, i32 3
+; CHECK-NEXT: [[FR:%.*]] = freeze i1 [[TMP7]]
+; CHECK-NEXT: [[RDX_SELECT:%.*]] = select i1 [[FR]], i32 7, i32 3
;
; SCALABLE-LABEL: @select_const_i32_from_icmp
; SCALABLE: vector.ph:
@@ -182,13 +192,15 @@ define i32 @select_const_i32_from_icmp(ptr nocapture readonly %v, i64 %n) #0 {
; SCALABLE-NEXT: [[TMP6:%.*]] = getelementptr inbounds i32, ptr [[TMP5]], i32 0
; SCALABLE-NEXT: [[WIDE_LOAD:%.*]] = load <vscale x 4 x i32>, ptr [[TMP6]], align 4
; SCALABLE-NEXT: [[TMP8:%.*]] = icmp eq <vscale x 4 x i32> [[WIDE_LOAD]], shufflevector (<vscale x 4 x i32> insertelement (<vscale x 4 x i32> poison, i32 3, i64 0), <vscale x 4 x i32> poison, <vscale x 4 x i32> zeroinitializer)
-; SCALABLE-NEXT: [[TMP9]] = or <vscale x 4 x i1> [[VEC_PHI]], [[TMP8]]
+; SCALABLE-NEXT: [[NOT:%.*]] = xor <vscale x 4 x i1> [[TMP8]], shufflevector (<vscale x 4 x i1> insertelement (<vscale x 4 x i1> poison, i1 true, i64 0), <vscale x 4 x i1> poison, <vscale x 4 x i32> zeroinitializer)
+; SCALABLE-NEXT: [[TMP9]] = or <vscale x 4 x i1> [[VEC_PHI]], [[NOT]]
; SCALABLE-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX]], [[TMP11]]
; SCALABLE-NEXT: [[TMP12:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]]
; SCALABLE-NEXT: br i1 [[TMP12]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP6:![0-9]+]]
; SCALABLE: middle.block:
; SCALABLE-NEXT: [[TMP13:%.*]] = call i1 @llvm.vector.reduce.or.nxv4i1(<vscale x 4 x i1> [[TMP9]])
-; SCALABLE-NEXT: [[RDX_SELECT:%.*]] = select i1 [[TMP13]], i32 7, i32 3
+; SCALABLE-NEXT: [[FR:%.*]] = freeze i1 [[TMP13]]
+; SCALABLE-NEXT: [[RDX_SELECT:%.*]] = select i1 [[FR]], i32 7, i32 3
;
entry:
br label %for.body
@@ -222,13 +234,15 @@ define i32 @select_i32_from_icmp(ptr nocapture readonly %v, i32 %a, i32 %b, i64
; CHECK-NEXT: [[TMP2:%.*]] = getelementptr inbounds i32, ptr [[TMP1]], i32 0
; CHECK-NEXT: [[WIDE_LOAD:%.*]] = load <4 x i32>, ptr [[TMP2]], align 4
; CHECK-NEXT: [[TMP4:%.*]] = icmp eq <4 x i32> [[WIDE_LOAD]], <i32 3, i32 3, i32 3, i32 3>
-; CHECK-NEXT: [[TMP5]] = or <4 x i1> [[VEC_PHI]], [[TMP4]]
+; CHECK-NEXT: [[NOT:%.*]] = xor <4 x i1> [[TMP4]], <i1 true, i1 true, i1 true, i1 true>
+; CHECK-NEXT: [[TMP5]] = or <4 x i1> [[VEC_PHI]], [[NOT]]
; CHECK-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX]], 4
; CHECK-NEXT: [[TMP6:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]]
; CHECK-NEXT: br i1 [[TMP6]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP8:![0-9]+]]
; CHECK: middle.block:
; CHECK-NEXT: [[TMP7:%.*]] = call i1 @llvm.vector.reduce.or.v4i1(<4 x i1> [[TMP5]])
-; CHECK-NEXT: [[RDX_SELECT:%.*]] = select i1 [[TMP7]], i32 %b, i32 %a
+; CHECK-NEXT: [[FR:%.*]] = freeze i1 [[TMP7]]
+; CHECK-NEXT: [[RDX_SELECT:%.*]] = select i1 [[FR]], i32 %b, i32 %a
;
; SCALABLE-LABEL: @select_i32_from_icmp
; SCALABLE: vector.ph:
@@ -247,13 +261,15 @@ define i32 @select_i32_from_icmp(ptr nocapture readonly %v, i32 %a, i32 %b, i64
; SCALABLE-NEXT: [[TMP6:%.*]] = getelementptr inbounds i32, ptr [[TMP5]], i32 0
; SCALABLE-NEXT: [[WIDE_LOAD:%.*]] = load <vscale x 4 x i32>, ptr [[TMP6]], align 4
; SCALABLE-NEXT: [[TMP8:%.*]] = icmp eq <vscale x 4 x i32> [[WIDE_LOAD]], shufflevector (<vscale x 4 x i32> insertelement (<vscale x 4 x i32> poison, i32 3, i64 0), <vscale x 4 x i32> poison, <vscale x 4 x i32> zeroinitializer)
-; SCALABLE-NEXT: [[TMP9]] = or <vscale x 4 x i1> [[VEC_PHI]], [[TMP8]]
+; SCALABLE-NEXT: [[NOT:%.*]] = xor <vscale x 4 x i1> [[TMP8]], shufflevector (<vscale x 4 x i1> insertelement (<vscale x 4 x i1> poison, i1 true, i64 0), <vscale x 4 x i1> poison, <vscale x 4 x i32> zeroinitializer)
+; SCALABLE-NEXT: [[TMP9]] = or <vscale x 4 x i1> [[VEC_PHI]], [[NOT]]
; SCALABLE-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX]], [[TMP11]]
; SCALABLE-NEXT: [[TMP12:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]]
; SCALABLE-NEXT: br i1 [[TMP12]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP8:![0-9]+]]
; SCALABLE: middle.block:
; SCALABLE-NEXT: [[TMP13:%.*]] = call i1 @llvm.vector.reduce.or.nxv4i1(<vscale x 4 x i1> [[TMP9]])
-; SCALABLE-NEXT: [[RDX_SELECT:%.*]] = select i1 [[TMP13]], i32 %b, i32 %a
+; SCALABLE-NEXT: [[FR:%.*]] = freeze i1 [[TMP13]]
+; SCALABLE-NEXT: [[RDX_SELECT:%.*]] = select i1 [[FR]], i32 %b, i32 %a
;
entry:
br label %for.body
@@ -287,13 +303,15 @@ define i32 @select_const_i32_from_fcmp(ptr nocapture readonly %v, i64 %n) #0 {
; CHECK-NEXT: [[TMP2:%.*]] = getelementptr inbounds float, ptr [[TMP1]], i32 0
; CHECK-NEXT: [[WIDE_LOAD:%.*]] = load <4 x float>, ptr [[TMP2]], align 4
; CHECK-NEXT: [[TMP4:%.*]] = fcmp fast ueq <4 x float> [[WIDE_LOAD]], <float 3.000000e+00, float 3.000000e+00, float 3.000000e+00, float 3.000000e+00>
-; CHECK-NEXT: [[TMP5]] = or <4 x i1> [[VEC_PHI]], [[TMP4]]
+; CHECK-NEXT: [[NOT:%.*]] = xor <4 x i1> [[TMP4]], <i1 true, i1 true, i1 true, i1 true>
+; CHECK-NEXT: [[TMP5]] = or <4 x i1> [[VEC_PHI]], [[NOT]]
; CHECK-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX]], 4
; CHECK-NEXT: [[TMP6:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]]
; CHECK-NEXT: br i1 [[TMP6]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP10:![0-9]+]]
; CHECK: middle.block:
; CHECK-NEXT: [[TMP7:%.*]] = call i1 @llvm.vector.reduce.or.v4i1(<4 x i1> [[TMP5]])
-; CHECK-NEXT: [[RDX_SELECT:%.*]] = select i1 [[TMP7]], i32 1, i32 2
+; CHECK-NEXT: [[FR:%.*]] = freeze i1 [[TMP7]]
+; CHECK-NEXT: [[RDX_SELECT:%.*]] = select i1 [[FR]], i32 1, i32 2
;
; SCALABLE-LABEL: @select_const_i32_from_fcmp
; SCALABLE: vector.ph:
@@ -312,13 +330,15 @@ define i32 @select_const_i32_from_fcmp(ptr nocapture readonly %v, i64 %n) #0 {
; SCALABLE-NEXT: [[TMP6:%.*]] = getelementptr inbounds float, ptr [[TMP5]], i32 0
; SCALABLE-NEXT: [[WIDE_LOAD:%.*]] = load <vscale x 4 x float>, ptr [[TMP6]], align 4
; SCALABLE-NEXT: [[TMP8:%.*]] = fcmp fast ueq <vscale x 4 x float> [[WIDE_LOAD]], shufflevector (<vscale x 4 x float> insertelement (<vscale x 4 x float> poison, float 3.000000e+00, i64 0), <vscale x 4 x float> poison, <vscale x 4 x i32> zeroinitializer)
-; SCALABLE-NEXT: [[TMP9]] = or <vscale x 4 x i1> [[VEC_PHI]], [[TMP8]]
+; SCALABLE-NEXT: [[NOT:%.*]] = xor <vscale x 4 x i1> [[TMP8]], shufflevector (<vscale x 4 x i1> insertelement (<vscale x 4 x i1> poison, i1 true, i64 0), <vscale x 4 x i1> poison, <vscale x 4 x i32> zeroinitializer)
+; SCALABLE-NEXT: [[TMP9]] = or <vscale x 4 x i1> [[VEC_PHI]], [[NOT]]
; SCALABLE-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX]], [[TMP11]]
; SCALABLE-NEXT: [[TMP12:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]]
; SCALABLE-NEXT: br i1 [[TMP12]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP10:![0-9]+]]
; SCALABLE: middle.block:
; SCALABLE-NEXT: [[TMP13:%.*]] = call i1 @llvm.vector.reduce.or.nxv4i1(<vscale x 4 x i1> [[TMP9]])
-; SCALABLE-NEXT: [[RDX_SELECT:%.*]] = select i1 [[TMP13]], i32 1, i32 2
+; SCALABLE-NEXT: [[FR:%.*]] = freeze i1 [[TMP13]]
+; SCALABLE-NEXT: [[RDX_SELECT:%.*]] = select i1 [[FR]], i32 1, i32 2
;
entry:
br label %for.body
@@ -389,7 +409,8 @@ define i32 @pred_select_const_i32_from_icmp(ptr noalias nocapture readonly %src1
; CHECK-NEXT: br i1 [[TMP11]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP12:![0-9]+]]
; CHECK: middle.block:
; CHECK-NEXT: [[TMP12:%.*]] = call i1 @llvm.vector.reduce.or.v4i1(<4 x i1> [[PREDPHI]])
-; CHECK-NEXT: [[RDX_SELECT:%.*]] = select i1 [[TMP12]], i32 1, i32 0
+; CHECK-NEXT: [[FR:%.*]] = freeze i1 [[TMP12]]
+; CHECK-NEXT: [[RDX_SELECT:%.*]] = select i1 [[FR]], i32 1, i32 0
;
; SCALABLE-LABEL: @pred_select_const_i32_from_icmp
; SCALABLE: vector.ph:
@@ -420,7 +441,8 @@ define i32 @pred_select_const_i32_from_icmp(ptr noalias nocapture readonly %src1
; SCALABLE-NEXT: br i1 [[TMP17]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP12:![0-9]+]]
; SCALABLE: middle.block:
; SCALABLE-NEXT: [[TMP18:%.*]] = call i1 @llvm.vector.reduce.or.nxv4i1(<vscale x 4 x i1> [[PREDPHI]])
-; SCALABLE-NEXT: [[RDX_SELECT:%.*]] = select i1 [[TMP18]], i32 1, i32 0
+; SCALABLE-NEXT: [[FR:%.*]] = freeze i1 [[TMP18]]
+; SCALABLE-NEXT: [[RDX_SELECT:%.*]] = select i1 [[FR]], i32 1, i32 0
;
entry:
br label %for.body
>From f9fea8e00193d79abfd6ce24af1a323948bb74ea Mon Sep 17 00:00:00 2001
From: Florian Hahn <flo at fhahn.com>
Date: Mon, 5 Feb 2024 14:26:34 +0000
Subject: [PATCH 3/3] !fixup address latest comments, thanks!
---
llvm/lib/Transforms/Vectorize/LoopVectorize.cpp | 6 ++++--
llvm/lib/Transforms/Vectorize/VPlanRecipes.cpp | 8 +++-----
.../LoopVectorize/AArch64/sve-select-cmp.ll | 6 +++---
.../LoopVectorize/select-cmp-predicated.ll | 2 +-
llvm/test/Transforms/LoopVectorize/select-cmp.ll | 12 ++++++------
5 files changed, 17 insertions(+), 17 deletions(-)
diff --git a/llvm/lib/Transforms/Vectorize/LoopVectorize.cpp b/llvm/lib/Transforms/Vectorize/LoopVectorize.cpp
index 65ccd5849e9da..70975b2ef8630 100644
--- a/llvm/lib/Transforms/Vectorize/LoopVectorize.cpp
+++ b/llvm/lib/Transforms/Vectorize/LoopVectorize.cpp
@@ -7499,8 +7499,6 @@ static void createAndCollectMergePhiForReduction(
auto *PhiR = cast<VPReductionPHIRecipe>(RedResult->getOperand(0));
const RecurrenceDescriptor &RdxDesc = PhiR->getRecurrenceDescriptor();
- TrackingVH<Value> ReductionStartValue =
- State.get(PhiR->getStartValue(), VPIteration(0, 0));
Value *FinalValue =
State.get(RedResult, VPIteration(State.UF - 1, VPLane::getFirstLane()));
auto *ResumePhi =
@@ -8948,6 +8946,10 @@ VPlanPtr LoopVectorizationPlanner::buildVPlan(VFRange &Range) {
// A ComputeReductionResult recipe is added to the middle block, also for
// in-loop reductions which compute their result in-loop, because generating
// the subsequent bc.merge.rdx phi is driven by ComputeReductionResult recipes.
+//
+// Adjust AnyOf reductions; replace the reduction phi for the selected value
+// with a boolean reduction phi node to check if the condition is true in any
+// iteration. The final value is selected by the final ComputeReductionResult.
void LoopVectorizationPlanner::adjustRecipesForReductions(
VPBasicBlock *LatchVPBB, VPlanPtr &Plan, VPRecipeBuilder &RecipeBuilder,
ElementCount MinVF) {
diff --git a/llvm/lib/Transforms/Vectorize/VPlanRecipes.cpp b/llvm/lib/Transforms/Vectorize/VPlanRecipes.cpp
index 03f68e67c4cb2..3925980633f42 100644
--- a/llvm/lib/Transforms/Vectorize/VPlanRecipes.cpp
+++ b/llvm/lib/Transforms/Vectorize/VPlanRecipes.cpp
@@ -438,6 +438,8 @@ Value *VPInstruction::generateInstruction(VPTransformState &State,
// Reduce all of the unrolled parts into a single vector.
Value *ReducedPartRdx = RdxParts[0];
unsigned Op = RecurrenceDescriptor::getOpcode(RK);
+ if (RecurrenceDescriptor::isAnyOfRecurrenceKind(RK))
+ Op = Instruction::Or;
if (PhiR->isOrdered()) {
ReducedPartRdx = RdxParts[State.UF - 1];
@@ -450,11 +452,7 @@ Value *VPInstruction::generateInstruction(VPTransformState &State,
if (Op != Instruction::ICmp && Op != Instruction::FCmp)
ReducedPartRdx = Builder.CreateBinOp(
(Instruction::BinaryOps)Op, RdxPart, ReducedPartRdx, "bin.rdx");
- else if (RecurrenceDescriptor::isAnyOfRecurrenceKind(RK)) {
- TrackingVH<Value> ReductionStartValue =
- RdxDesc.getRecurrenceStartValue();
- ReducedPartRdx = Builder.CreateOr(ReducedPartRdx, RdxPart);
- } else
+ else
ReducedPartRdx = createMinMaxOp(Builder, RK, ReducedPartRdx, RdxPart);
}
}
diff --git a/llvm/test/Transforms/LoopVectorize/AArch64/sve-select-cmp.ll b/llvm/test/Transforms/LoopVectorize/AArch64/sve-select-cmp.ll
index b43774446b589..2470bca1e17b9 100644
--- a/llvm/test/Transforms/LoopVectorize/AArch64/sve-select-cmp.ll
+++ b/llvm/test/Transforms/LoopVectorize/AArch64/sve-select-cmp.ll
@@ -37,9 +37,9 @@ define i32 @select_const_i32_from_icmp(ptr nocapture readonly %v, i64 %n) #0 {
; CHECK-VF4IC4-NEXT: [[VEC_SEL3:%.*]] = or <vscale x 4 x i1> [[VEC_PHI3]], [[NOT3]]
; CHECK-VF4IC4-NEXT: [[VEC_SEL4:%.*]] = or <vscale x 4 x i1> [[VEC_PHI4]], [[NOT4]]
; CHECK-VF4IC4: middle.block:
-; CHECK-VF4IC4-NEXT: [[OR1:%.*]] = or <vscale x 4 x i1> [[VEC_SEL1]], [[VEC_SEL2]]
-; CHECK-VF4IC4-NEXT: [[OR2:%.*]] = or <vscale x 4 x i1> [[OR1]], [[VEC_SEL3]]
-; CHECK-VF4IC4-NEXT: [[OR3:%.*]] = or <vscale x 4 x i1> [[OR2]], [[VEC_SEL4]]
+; CHECK-VF4IC4-NEXT: [[OR1:%.*]] = or <vscale x 4 x i1> [[VEC_SEL2]], [[VEC_SEL1]]
+; CHECK-VF4IC4-NEXT: [[OR2:%.*]] = or <vscale x 4 x i1> [[VEC_SEL3]], [[OR1]]
+; CHECK-VF4IC4-NEXT: [[OR3:%.*]] = or <vscale x 4 x i1> [[VEC_SEL4]], [[OR2]]
; CHECK-VF4IC4-NEXT: [[OR_RDX:%.*]] = call i1 @llvm.vector.reduce.or.nxv4i1(<vscale x 4 x i1> [[OR3]])
; CHECK-VF4IC4-NEXT: [[FR:%.*]] = freeze i1 [[OR_RDX]]
; CHECK-VF4IC4-NEXT: {{.*}} = select i1 [[FR]], i32 7, i32 3
diff --git a/llvm/test/Transforms/LoopVectorize/select-cmp-predicated.ll b/llvm/test/Transforms/LoopVectorize/select-cmp-predicated.ll
index 98962fd4e371f..cedc769b811fc 100644
--- a/llvm/test/Transforms/LoopVectorize/select-cmp-predicated.ll
+++ b/llvm/test/Transforms/LoopVectorize/select-cmp-predicated.ll
@@ -88,7 +88,7 @@ define i32 @pred_select_const_i32_from_icmp(ptr noalias nocapture readonly %src1
; CHECK-VF1IC2-NEXT: [[PREDPHI5]] = select i1 [[TMP5]], i1 [[TMP15]], i1 [[VEC_PHI2]]
; CHECK-VF1IC2: br i1 {{%.*}}, label %middle.block, label %vector.body
; CHECK-VF1IC2: middle.block:
-; CHECK-VF1IC2-NEXT: [[OR:%.*]] = or i1 [[PREDPHI]], [[PREDPHI5]]
+; CHECK-VF1IC2-NEXT: [[OR:%.*]] = or i1 [[PREDPHI5]], [[PREDPHI]]
; CHECK-VF1IC2-NEXT: [[FR_OR:%.*]] = freeze i1 [[OR]]
; CHECK-VF1IC2-NEXT: [[RDX_SELECT:%.*]] = select i1 [[FR_OR]], i32 1, i32 0
; CHECK-VF1IC2: br i1 {{%.*}}, label %for.end.loopexit, label %scalar.ph
diff --git a/llvm/test/Transforms/LoopVectorize/select-cmp.ll b/llvm/test/Transforms/LoopVectorize/select-cmp.ll
index a9157ec118b01..993b56a05207b 100644
--- a/llvm/test/Transforms/LoopVectorize/select-cmp.ll
+++ b/llvm/test/Transforms/LoopVectorize/select-cmp.ll
@@ -33,9 +33,9 @@ define i32 @select_const_i32_from_icmp(ptr nocapture readonly %v, i64 %n) {
; CHECK-VF4IC4-NEXT: [[VEC_SEL3:%.*]] = or <4 x i1> [[VEC_PHI3]], [[NOT3]]
; CHECK-VF4IC4-NEXT: [[VEC_SEL4:%.*]] = or <4 x i1> [[VEC_PHI4]], [[NOT4]]
; CHECK-VF4IC4: middle.block:
-; CHECK-VF4IC4-NEXT: [[VEC_SEL5:%.*]] = or <4 x i1> [[VEC_SEL1]], [[VEC_SEL2]]
-; CHECK-VF4IC4-NEXT: [[VEC_SEL6:%.*]] = or <4 x i1> [[VEC_SEL5]], [[VEC_SEL3]]
-; CHECK-VF4IC4-NEXT: [[VEC_SEL7:%.*]] = or <4 x i1> [[VEC_SEL6]], [[VEC_SEL4]]
+; CHECK-VF4IC4-NEXT: [[VEC_SEL5:%.*]] = or <4 x i1> [[VEC_SEL2]], [[VEC_SEL1]]
+; CHECK-VF4IC4-NEXT: [[VEC_SEL6:%.*]] = or <4 x i1> [[VEC_SEL3]], [[VEC_SEL5]]
+; CHECK-VF4IC4-NEXT: [[VEC_SEL7:%.*]] = or <4 x i1> [[VEC_SEL4]], [[VEC_SEL6]]
; CHECK-VF4IC4-NEXT: [[OR_RDX:%.*]] = call i1 @llvm.vector.reduce.or.v4i1(<4 x i1> [[VEC_SEL7]])
; CHECK-VF4IC4-NEXT: [[FR_OR_RDX:%.*]] = freeze i1 [[OR_RDX]]
; CHECK-VF4IC4-NEXT: {{.*}} = select i1 [[FR_OR_RDX]], i32 7, i32 3
@@ -63,9 +63,9 @@ define i32 @select_const_i32_from_icmp(ptr nocapture readonly %v, i64 %n) {
; CHECK-VF1IC4-NEXT: [[VEC_SEL3:%.*]] = or i1 [[VEC_PHI3]], [[NOT3]]
; CHECK-VF1IC4-NEXT: [[VEC_SEL4:%.*]] = or i1 [[VEC_PHI4]], [[NOT4]]
; CHECK-VF1IC4: middle.block:
-; CHECK-VF1IC4-NEXT: [[VEC_SEL5:%.*]] = or i1 [[VEC_SEL1]], [[VEC_SEL2]]
-; CHECK-VF1IC4-NEXT: [[VEC_SEL6:%.*]] = or i1 [[VEC_SEL5]], [[VEC_SEL3]]
-; CHECK-VF1IC4-NEXT: [[OR_RDX:%.*]] = or i1 [[VEC_SEL6]], [[VEC_SEL4]]
+; CHECK-VF1IC4-NEXT: [[VEC_SEL5:%.*]] = or i1 [[VEC_SEL2]], [[VEC_SEL1]]
+; CHECK-VF1IC4-NEXT: [[VEC_SEL6:%.*]] = or i1 [[VEC_SEL3]], [[VEC_SEL5]]
+; CHECK-VF1IC4-NEXT: [[OR_RDX:%.*]] = or i1 [[VEC_SEL4]], [[VEC_SEL6]]
; CHECK-VF1IC4-NEXT: [[FR_OR_RDX:%.*]] = freeze i1 [[OR_RDX]]
; CHECK-VF1IC4-NEXT: {{.*}} = select i1 [[FR_OR_RDX]], i32 7, i32 3
More information about the libcxx-commits
mailing list