[llvm] 8107b43 - [VPlan] Simplify select c, x, x -> x (#133731)
via llvm-commits
llvm-commits at lists.llvm.org
Wed Apr 2 02:26:52 PDT 2025
Author: Luke Lau
Date: 2025-04-02T10:26:48+01:00
New Revision: 8107b430edb38dc798fe4e710df13d4b2e98acfb
URL: https://github.com/llvm/llvm-project/commit/8107b430edb38dc798fe4e710df13d4b2e98acfb
DIFF: https://github.com/llvm/llvm-project/commit/8107b430edb38dc798fe4e710df13d4b2e98acfb.diff
LOG: [VPlan] Simplify select c, x, x -> x (#133731)
As noted in 1a9358c090d0507be21c5e9b2d97a23ef1de8ab0, some
simplifications can produce a redundant select where the true and false
operands are the same, which this patch removes.
The is_fpclass test was changed so the condition wasn't made dead.
Added:
Modified:
llvm/lib/Transforms/Vectorize/VPlanTransforms.cpp
llvm/test/Transforms/LoopVectorize/X86/cost-model.ll
llvm/test/Transforms/LoopVectorize/is_fpclass.ll
llvm/test/Transforms/LoopVectorize/select-reduction.ll
Removed:
################################################################################
diff --git a/llvm/lib/Transforms/Vectorize/VPlanTransforms.cpp b/llvm/lib/Transforms/Vectorize/VPlanTransforms.cpp
index 9a041c83438dc..0c37db7f9d3a3 100644
--- a/llvm/lib/Transforms/Vectorize/VPlanTransforms.cpp
+++ b/llvm/lib/Transforms/Vectorize/VPlanTransforms.cpp
@@ -1062,6 +1062,9 @@ static void simplifyRecipe(VPRecipeBase &R, VPTypeAnalysis &TypeInfo) {
return;
}
+ if (match(&R, m_Select(m_VPValue(), m_VPValue(X), m_Deferred(X))))
+ return R.getVPSingleValue()->replaceAllUsesWith(X);
+
if (match(&R, m_c_Mul(m_VPValue(A), m_SpecificInt(1))))
return R.getVPSingleValue()->replaceAllUsesWith(A);
diff --git a/llvm/test/Transforms/LoopVectorize/X86/cost-model.ll b/llvm/test/Transforms/LoopVectorize/X86/cost-model.ll
index 1f171414242bc..c39dd6ffe5f01 100644
--- a/llvm/test/Transforms/LoopVectorize/X86/cost-model.ll
+++ b/llvm/test/Transforms/LoopVectorize/X86/cost-model.ll
@@ -1046,16 +1046,11 @@ define i64 @live_in_known_1_via_scev() {
; CHECK: vector.body:
; CHECK-NEXT: [[INDEX:%.*]] = phi i32 [ 0, [[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ]
; CHECK-NEXT: [[VEC_PHI:%.*]] = phi <4 x i64> [ <i64 3, i64 1, i64 1, i64 1>, [[VECTOR_PH]] ], [ [[VEC_PHI]], [[VECTOR_BODY]] ]
-; CHECK-NEXT: [[BROADCAST_SPLATINSERT:%.*]] = insertelement <4 x i32> poison, i32 [[INDEX]], i64 0
-; CHECK-NEXT: [[BROADCAST_SPLAT:%.*]] = shufflevector <4 x i32> [[BROADCAST_SPLATINSERT]], <4 x i32> poison, <4 x i32> zeroinitializer
-; CHECK-NEXT: [[VEC_IV:%.*]] = add <4 x i32> [[BROADCAST_SPLAT]], <i32 0, i32 1, i32 2, i32 3>
-; CHECK-NEXT: [[TMP0:%.*]] = icmp ule <4 x i32> [[VEC_IV]], splat (i32 5)
-; CHECK-NEXT: [[TMP1:%.*]] = select <4 x i1> [[TMP0]], <4 x i64> [[VEC_PHI]], <4 x i64> [[VEC_PHI]]
; CHECK-NEXT: [[INDEX_NEXT]] = add nuw i32 [[INDEX]], 4
; CHECK-NEXT: [[TMP2:%.*]] = icmp eq i32 [[INDEX_NEXT]], 8
; CHECK-NEXT: br i1 [[TMP2]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP24:![0-9]+]]
; CHECK: middle.block:
-; CHECK-NEXT: [[TMP3:%.*]] = call i64 @llvm.vector.reduce.mul.v4i64(<4 x i64> [[TMP1]])
+; CHECK-NEXT: [[TMP3:%.*]] = call i64 @llvm.vector.reduce.mul.v4i64(<4 x i64> [[VEC_PHI]])
; CHECK-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]]
; CHECK: scalar.ph:
; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i32 [ 8, [[MIDDLE_BLOCK]] ], [ 0, [[PH]] ]
diff --git a/llvm/test/Transforms/LoopVectorize/is_fpclass.ll b/llvm/test/Transforms/LoopVectorize/is_fpclass.ll
index c429092a50e6a..9f82c93d3579d 100644
--- a/llvm/test/Transforms/LoopVectorize/is_fpclass.ll
+++ b/llvm/test/Transforms/LoopVectorize/is_fpclass.ll
@@ -9,12 +9,12 @@ define void @d() {
; CHECK-NEXT: br label [[VECTOR_BODY:%.*]]
; CHECK: vector.body:
; CHECK-NEXT: [[INDEX:%.*]] = phi i64 [ 0, [[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ]
-; CHECK-NEXT: [[TMP6:%.*]] = load float, ptr null, align 4
-; CHECK-NEXT: [[BROADCAST_SPLATINSERT:%.*]] = insertelement <2 x float> poison, float [[TMP6]], i64 0
+; CHECK-NEXT: [[TMP0:%.*]] = load float, ptr null, align 4
+; CHECK-NEXT: [[BROADCAST_SPLATINSERT:%.*]] = insertelement <2 x float> poison, float [[TMP0]], i64 0
; CHECK-NEXT: [[BROADCAST_SPLAT:%.*]] = shufflevector <2 x float> [[BROADCAST_SPLATINSERT]], <2 x float> poison, <2 x i32> zeroinitializer
; CHECK-NEXT: [[TMP1:%.*]] = getelementptr float, ptr @d, i64 [[INDEX]]
; CHECK-NEXT: [[TMP2:%.*]] = call <2 x i1> @llvm.is.fpclass.v2f32(<2 x float> [[BROADCAST_SPLAT]], i32 0)
-; CHECK-NEXT: [[TMP3:%.*]] = select <2 x i1> [[TMP2]], <2 x float> zeroinitializer, <2 x float> zeroinitializer
+; CHECK-NEXT: [[TMP3:%.*]] = select <2 x i1> [[TMP2]], <2 x float> zeroinitializer, <2 x float> splat (float 1.000000e+00)
; CHECK-NEXT: [[TMP4:%.*]] = getelementptr float, ptr [[TMP1]], i32 0
; CHECK-NEXT: store <2 x float> [[TMP3]], ptr [[TMP4]], align 4
; CHECK-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX]], 2
@@ -30,7 +30,7 @@ define void @d() {
; CHECK-NEXT: [[I3:%.*]] = load float, ptr null, align 4
; CHECK-NEXT: [[I4:%.*]] = getelementptr float, ptr @d, i64 [[I]]
; CHECK-NEXT: [[I5:%.*]] = tail call i1 @llvm.is.fpclass.f32(float [[I3]], i32 0)
-; CHECK-NEXT: [[I6:%.*]] = select i1 [[I5]], float 0.000000e+00, float 0.000000e+00
+; CHECK-NEXT: [[I6:%.*]] = select i1 [[I5]], float 0.000000e+00, float 1.000000e+00
; CHECK-NEXT: store float [[I6]], ptr [[I4]], align 4
; CHECK-NEXT: [[I7]] = add i64 [[I]], 1
; CHECK-NEXT: [[I8:%.*]] = icmp eq i64 [[I7]], 128
@@ -46,7 +46,7 @@ loop:
%i3 = load float, ptr null, align 4
%i4 = getelementptr float, ptr @d, i64 %i
%i5 = tail call i1 @llvm.is.fpclass.f32(float %i3, i32 0)
- %i6 = select i1 %i5, float 0.0, float 0.0
+ %i6 = select i1 %i5, float 0.0, float 1.0
store float %i6, ptr %i4, align 4
%i7 = add i64 %i, 1
%i8 = icmp eq i64 %i7, 128
diff --git a/llvm/test/Transforms/LoopVectorize/select-reduction.ll b/llvm/test/Transforms/LoopVectorize/select-reduction.ll
index 836115f381382..16cc05da17a63 100644
--- a/llvm/test/Transforms/LoopVectorize/select-reduction.ll
+++ b/llvm/test/Transforms/LoopVectorize/select-reduction.ll
@@ -88,16 +88,11 @@ define i32 @pr66895_tail_fold_reduction_exit_inst_gets_simplified(i32 %n) {
; CHECK: vector.body:
; CHECK-NEXT: [[INDEX:%.*]] = phi i32 [ 0, [[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ]
; CHECK-NEXT: [[VEC_PHI:%.*]] = phi <4 x i32> [ <i32 0, i32 1, i32 1, i32 1>, [[VECTOR_PH]] ], [ [[VEC_PHI]], [[VECTOR_BODY]] ]
-; CHECK-NEXT: [[BROADCAST_SPLATINSERT:%.*]] = insertelement <4 x i32> poison, i32 [[INDEX]], i64 0
-; CHECK-NEXT: [[BROADCAST_SPLAT:%.*]] = shufflevector <4 x i32> [[BROADCAST_SPLATINSERT]], <4 x i32> poison, <4 x i32> zeroinitializer
-; CHECK-NEXT: [[VEC_IV:%.*]] = add <4 x i32> [[BROADCAST_SPLAT]], <i32 0, i32 1, i32 2, i32 3>
-; CHECK-NEXT: [[TMP0:%.*]] = icmp ule <4 x i32> [[VEC_IV]], splat (i32 12)
-; CHECK-NEXT: [[TMP1:%.*]] = select <4 x i1> [[TMP0]], <4 x i32> [[VEC_PHI]], <4 x i32> [[VEC_PHI]]
; CHECK-NEXT: [[INDEX_NEXT]] = add nuw i32 [[INDEX]], 4
; CHECK-NEXT: [[TMP2:%.*]] = icmp eq i32 [[INDEX_NEXT]], 16
; CHECK-NEXT: br i1 [[TMP2]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP4:![0-9]+]]
; CHECK: middle.block:
-; CHECK-NEXT: [[TMP3:%.*]] = call i32 @llvm.vector.reduce.mul.v4i32(<4 x i32> [[TMP1]])
+; CHECK-NEXT: [[TMP3:%.*]] = call i32 @llvm.vector.reduce.mul.v4i32(<4 x i32> [[VEC_PHI]])
; CHECK-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]]
; CHECK: scalar.ph:
; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i32 [ -4, [[MIDDLE_BLOCK]] ], [ 12, [[ENTRY:%.*]] ]
More information about the llvm-commits
mailing list