[llvm] 120d52b - [SLP]Fix PR55653: emit undefs where required, not poison.
Alexey Bataev via llvm-commits
llvm-commits at lists.llvm.org
Thu May 26 08:39:51 PDT 2022
Author: Alexey Bataev
Date: 2022-05-26T08:38:50-07:00
New Revision: 120d52b0ef8b0d41d5ce5898422d2d2f42047d0b
URL: https://github.com/llvm/llvm-project/commit/120d52b0ef8b0d41d5ce5898422d2d2f42047d0b
DIFF: https://github.com/llvm/llvm-project/commit/120d52b0ef8b0d41d5ce5898422d2d2f42047d0b.diff
LOG: [SLP]Fix PR55653: emit undefs where required, not poison.
Need to handle a corner case correctly, if all elements are Undefs/Poisons,
need to emit actual values, not just poisons.
Differential Revision: https://reviews.llvm.org/D126298
Added:
Modified:
llvm/lib/Transforms/Vectorize/SLPVectorizer.cpp
llvm/test/Transforms/SLPVectorizer/AArch64/reorder-fmuladd-crash.ll
llvm/test/Transforms/SLPVectorizer/SystemZ/pr34619.ll
llvm/test/Transforms/SLPVectorizer/X86/horizontal-minmax.ll
llvm/test/Transforms/SLPVectorizer/X86/insert-shuffle.ll
llvm/test/Transforms/SLPVectorizer/X86/jumbled-load-used-in-phi.ll
llvm/test/Transforms/SLPVectorizer/X86/partail.ll
llvm/test/Transforms/SLPVectorizer/X86/shrink_after_reorder.ll
llvm/test/Transforms/SLPVectorizer/X86/shrink_after_reorder2.ll
llvm/test/Transforms/SLPVectorizer/X86/vectorize-reorder-alt-shuffle.ll
llvm/test/Transforms/SLPVectorizer/slp-umax-rdx-matcher-crash.ll
Removed:
################################################################################
diff --git a/llvm/lib/Transforms/Vectorize/SLPVectorizer.cpp b/llvm/lib/Transforms/Vectorize/SLPVectorizer.cpp
index f4e94050fc5f4..791e0193951dd 100644
--- a/llvm/lib/Transforms/Vectorize/SLPVectorizer.cpp
+++ b/llvm/lib/Transforms/Vectorize/SLPVectorizer.cpp
@@ -7543,6 +7543,10 @@ Value *BoUpSLP::createBuildVector(ArrayRef<Value *> VL) {
ReuseShuffleIndicies.append(VF - ReuseShuffleIndicies.size(),
UndefMaskElem);
} else if (UniqueValues.size() >= VF - 1 || UniqueValues.size() <= 1) {
+ if (UniqueValues.empty()) {
+ assert(all_of(VL, UndefValue::classof) && "Expected list of undefs.");
+ NumValues = VF;
+ }
ReuseShuffleIndicies.clear();
UniqueValues.clear();
UniqueValues.append(VL.begin(), std::next(VL.begin(), NumValues));
diff --git a/llvm/test/Transforms/SLPVectorizer/AArch64/reorder-fmuladd-crash.ll b/llvm/test/Transforms/SLPVectorizer/AArch64/reorder-fmuladd-crash.ll
index 70bbe65fa2cd5..18365e015e3c3 100644
--- a/llvm/test/Transforms/SLPVectorizer/AArch64/reorder-fmuladd-crash.ll
+++ b/llvm/test/Transforms/SLPVectorizer/AArch64/reorder-fmuladd-crash.ll
@@ -21,14 +21,14 @@ define i32 @foo() {
; CHECK-NEXT: [[TMP0:%.*]] = bitcast double* [[ARRAYIDX43]] to <4 x double>*
; CHECK-NEXT: [[TMP1:%.*]] = load <4 x double>, <4 x double>* [[TMP0]], align 8
; CHECK-NEXT: [[TMP2:%.*]] = fmul <4 x double> [[TMP1]], <double 0x7FF8000000000000, double 0x7FF8000000000000, double 0x7FF8000000000000, double 0x7FF8000000000000>
-; CHECK-NEXT: [[TMP3:%.*]] = call <4 x double> @llvm.fmuladd.v4f64(<4 x double> poison, <4 x double> zeroinitializer, <4 x double> [[TMP2]])
+; CHECK-NEXT: [[TMP3:%.*]] = call <4 x double> @llvm.fmuladd.v4f64(<4 x double> undef, <4 x double> zeroinitializer, <4 x double> [[TMP2]])
; CHECK-NEXT: br label [[SW_EPILOG:%.*]]
; CHECK: sw.bb195:
; CHECK-NEXT: br label [[SW_EPILOG]]
; CHECK: do.body:
; CHECK-NEXT: unreachable
; CHECK: sw.epilog:
-; CHECK-NEXT: [[TMP4:%.*]] = phi <4 x double> [ poison, [[SW_BB195]] ], [ [[TMP3]], [[SW_BB]] ]
+; CHECK-NEXT: [[TMP4:%.*]] = phi <4 x double> [ undef, [[SW_BB195]] ], [ [[TMP3]], [[SW_BB]] ]
; CHECK-NEXT: ret i32 undef
; CHECK: if.end.1:
; CHECK-NEXT: br label [[FOR_COND15_1:%.*]]
diff --git a/llvm/test/Transforms/SLPVectorizer/SystemZ/pr34619.ll b/llvm/test/Transforms/SLPVectorizer/SystemZ/pr34619.ll
index 89d6681278217..51f7c0c52ac9b 100644
--- a/llvm/test/Transforms/SLPVectorizer/SystemZ/pr34619.ll
+++ b/llvm/test/Transforms/SLPVectorizer/SystemZ/pr34619.ll
@@ -16,7 +16,7 @@ define void @foo() local_unnamed_addr {
; CHECK-NEXT: [[TMP3:%.*]] = insertelement <4 x i32> [[TMP2]], i32 [[ADD277]], i32 1
; CHECK-NEXT: [[TMP4:%.*]] = shufflevector <2 x i32> [[TMP1]], <2 x i32> poison, <4 x i32> <i32 0, i32 1, i32 undef, i32 undef>
; CHECK-NEXT: [[TMP5:%.*]] = shufflevector <4 x i32> [[TMP3]], <4 x i32> [[TMP4]], <4 x i32> <i32 0, i32 1, i32 4, i32 5>
-; CHECK-NEXT: [[TMP6:%.*]] = add nsw <4 x i32> poison, [[TMP5]]
+; CHECK-NEXT: [[TMP6:%.*]] = add nsw <4 x i32> undef, [[TMP5]]
; CHECK-NEXT: [[TMP7:%.*]] = ashr <4 x i32> [[TMP6]], <i32 6, i32 6, i32 6, i32 6>
; CHECK-NEXT: [[TMP8:%.*]] = bitcast i32* [[ARRAYIDX372]] to <4 x i32>*
; CHECK-NEXT: store <4 x i32> [[TMP7]], <4 x i32>* [[TMP8]], align 4
diff --git a/llvm/test/Transforms/SLPVectorizer/X86/horizontal-minmax.ll b/llvm/test/Transforms/SLPVectorizer/X86/horizontal-minmax.ll
index 801b4863e7c19..262359c0eea86 100644
--- a/llvm/test/Transforms/SLPVectorizer/X86/horizontal-minmax.ll
+++ b/llvm/test/Transforms/SLPVectorizer/X86/horizontal-minmax.ll
@@ -1430,8 +1430,8 @@ define void @PR49730() {
; SSE-NEXT: ret void
;
; AVX-LABEL: @PR49730(
-; AVX-NEXT: [[TMP1:%.*]] = call <4 x i32> @llvm.smin.v4i32(<4 x i32> poison, <4 x i32> <i32 2, i32 2, i32 1, i32 1>)
-; AVX-NEXT: [[TMP2:%.*]] = sub nsw <4 x i32> poison, [[TMP1]]
+; AVX-NEXT: [[TMP1:%.*]] = call <4 x i32> @llvm.smin.v4i32(<4 x i32> undef, <4 x i32> <i32 2, i32 2, i32 1, i32 1>)
+; AVX-NEXT: [[TMP2:%.*]] = sub nsw <4 x i32> undef, [[TMP1]]
; AVX-NEXT: [[T12:%.*]] = sub nsw i32 undef, undef
; AVX-NEXT: [[TMP3:%.*]] = call i32 @llvm.vector.reduce.umin.v4i32(<4 x i32> [[TMP2]])
; AVX-NEXT: [[TMP4:%.*]] = call i32 @llvm.umin.i32(i32 [[T12]], i32 undef)
@@ -1440,8 +1440,8 @@ define void @PR49730() {
; AVX-NEXT: ret void
;
; AVX2-LABEL: @PR49730(
-; AVX2-NEXT: [[TMP1:%.*]] = call <4 x i32> @llvm.smin.v4i32(<4 x i32> poison, <4 x i32> <i32 2, i32 2, i32 1, i32 1>)
-; AVX2-NEXT: [[TMP2:%.*]] = sub nsw <4 x i32> poison, [[TMP1]]
+; AVX2-NEXT: [[TMP1:%.*]] = call <4 x i32> @llvm.smin.v4i32(<4 x i32> undef, <4 x i32> <i32 2, i32 2, i32 1, i32 1>)
+; AVX2-NEXT: [[TMP2:%.*]] = sub nsw <4 x i32> undef, [[TMP1]]
; AVX2-NEXT: [[T12:%.*]] = sub nsw i32 undef, undef
; AVX2-NEXT: [[TMP3:%.*]] = call i32 @llvm.vector.reduce.umin.v4i32(<4 x i32> [[TMP2]])
; AVX2-NEXT: [[TMP4:%.*]] = call i32 @llvm.umin.i32(i32 [[T12]], i32 undef)
@@ -1450,8 +1450,8 @@ define void @PR49730() {
; AVX2-NEXT: ret void
;
; THRESH-LABEL: @PR49730(
-; THRESH-NEXT: [[TMP1:%.*]] = call <4 x i32> @llvm.smin.v4i32(<4 x i32> poison, <4 x i32> <i32 2, i32 2, i32 1, i32 1>)
-; THRESH-NEXT: [[TMP2:%.*]] = sub nsw <4 x i32> poison, [[TMP1]]
+; THRESH-NEXT: [[TMP1:%.*]] = call <4 x i32> @llvm.smin.v4i32(<4 x i32> undef, <4 x i32> <i32 2, i32 2, i32 1, i32 1>)
+; THRESH-NEXT: [[TMP2:%.*]] = sub nsw <4 x i32> undef, [[TMP1]]
; THRESH-NEXT: [[T12:%.*]] = sub nsw i32 undef, undef
; THRESH-NEXT: [[TMP3:%.*]] = call i32 @llvm.vector.reduce.umin.v4i32(<4 x i32> [[TMP2]])
; THRESH-NEXT: [[TMP4:%.*]] = call i32 @llvm.umin.i32(i32 [[T12]], i32 undef)
diff --git a/llvm/test/Transforms/SLPVectorizer/X86/insert-shuffle.ll b/llvm/test/Transforms/SLPVectorizer/X86/insert-shuffle.ll
index b64c0e4e35a23..b3879bda22356 100644
--- a/llvm/test/Transforms/SLPVectorizer/X86/insert-shuffle.ll
+++ b/llvm/test/Transforms/SLPVectorizer/X86/insert-shuffle.ll
@@ -16,9 +16,9 @@ define { <2 x float>, <2 x float> } @foo(%struct.sw* %v) {
; CHECK-NEXT: [[TMP5:%.*]] = insertelement <4 x float> [[TMP4]], float [[TMP1]], i32 1
; CHECK-NEXT: [[SHUFFLE1:%.*]] = shufflevector <4 x float> [[TMP5]], <4 x float> poison, <4 x i32> <i32 0, i32 undef, i32 1, i32 undef>
; CHECK-NEXT: [[TMP6:%.*]] = fmul <4 x float> [[SHUFFLE]], [[SHUFFLE1]]
-; CHECK-NEXT: [[TMP7:%.*]] = fadd <4 x float> [[TMP6]], poison
-; CHECK-NEXT: [[TMP8:%.*]] = fadd <4 x float> [[TMP7]], poison
-; CHECK-NEXT: [[TMP9:%.*]] = fadd <4 x float> [[TMP8]], poison
+; CHECK-NEXT: [[TMP7:%.*]] = fadd <4 x float> [[TMP6]], undef
+; CHECK-NEXT: [[TMP8:%.*]] = fadd <4 x float> [[TMP7]], undef
+; CHECK-NEXT: [[TMP9:%.*]] = fadd <4 x float> [[TMP8]], undef
; CHECK-NEXT: [[TMP10:%.*]] = shufflevector <4 x float> [[TMP9]], <4 x float> poison, <2 x i32> <i32 0, i32 1>
; CHECK-NEXT: [[TMP11:%.*]] = shufflevector <4 x float> [[TMP9]], <4 x float> poison, <2 x i32> <i32 2, i32 3>
; CHECK-NEXT: [[INS1:%.*]] = insertvalue { <2 x float>, <2 x float> } undef, <2 x float> [[TMP10]], 0
diff --git a/llvm/test/Transforms/SLPVectorizer/X86/jumbled-load-used-in-phi.ll b/llvm/test/Transforms/SLPVectorizer/X86/jumbled-load-used-in-phi.ll
index 8de5c70e5f058..f387f9c1ff8a5 100644
--- a/llvm/test/Transforms/SLPVectorizer/X86/jumbled-load-used-in-phi.ll
+++ b/llvm/test/Transforms/SLPVectorizer/X86/jumbled-load-used-in-phi.ll
@@ -49,7 +49,7 @@ define void @phiUsingLoads(i32* noalias nocapture readonly %A, i32* noalias noca
; CHECK-NEXT: ret void
; CHECK: for.body:
; CHECK-NEXT: [[INDVARS_IV:%.*]] = phi i64 [ 0, [[ENTRY:%.*]] ], [ [[INDVARS_IV_NEXT:%.*]], [[FOR_INC:%.*]] ]
-; CHECK-NEXT: [[TMP2:%.*]] = phi <4 x i32> [ poison, [[ENTRY]] ], [ [[TMP14]], [[FOR_INC]] ]
+; CHECK-NEXT: [[TMP2:%.*]] = phi <4 x i32> [ undef, [[ENTRY]] ], [ [[TMP14]], [[FOR_INC]] ]
; CHECK-NEXT: br i1 [[CMP1]], label [[IF_THEN:%.*]], label [[IF_ELSE:%.*]]
; CHECK: if.then:
; CHECK-NEXT: [[ARRAYIDX2:%.*]] = getelementptr inbounds i32, i32* [[A]], i64 [[INDVARS_IV]]
diff --git a/llvm/test/Transforms/SLPVectorizer/X86/partail.ll b/llvm/test/Transforms/SLPVectorizer/X86/partail.ll
index ce195e8d75242..c4787a6cb8ebc 100644
--- a/llvm/test/Transforms/SLPVectorizer/X86/partail.ll
+++ b/llvm/test/Transforms/SLPVectorizer/X86/partail.ll
@@ -18,8 +18,8 @@ define void @get_block(i32 %y_pos) local_unnamed_addr #0 {
; CHECK-NEXT: [[SHUFFLE:%.*]] = shufflevector <4 x i32> [[TMP1]], <4 x i32> poison, <4 x i32> <i32 0, i32 1, i32 1, i32 1>
; CHECK-NEXT: [[TMP2:%.*]] = icmp sgt <4 x i32> [[SHUFFLE]], <i32 0, i32 -1, i32 -5, i32 -9>
; CHECK-NEXT: [[TMP3:%.*]] = select <4 x i1> [[TMP2]], <4 x i32> [[TMP0]], <4 x i32> zeroinitializer
-; CHECK-NEXT: [[TMP4:%.*]] = icmp slt <4 x i32> [[TMP3]], poison
-; CHECK-NEXT: [[TMP5:%.*]] = select <4 x i1> [[TMP4]], <4 x i32> [[TMP3]], <4 x i32> poison
+; CHECK-NEXT: [[TMP4:%.*]] = icmp slt <4 x i32> [[TMP3]], undef
+; CHECK-NEXT: [[TMP5:%.*]] = select <4 x i1> [[TMP4]], <4 x i32> [[TMP3]], <4 x i32> undef
; CHECK-NEXT: [[TMP6:%.*]] = sext <4 x i32> [[TMP5]] to <4 x i64>
; CHECK-NEXT: [[TMP7:%.*]] = trunc <4 x i64> [[TMP6]] to <4 x i32>
; CHECK-NEXT: [[TMP8:%.*]] = extractelement <4 x i32> [[TMP7]], i32 0
diff --git a/llvm/test/Transforms/SLPVectorizer/X86/shrink_after_reorder.ll b/llvm/test/Transforms/SLPVectorizer/X86/shrink_after_reorder.ll
index e64969b09d4de..c31947a9d7b1c 100644
--- a/llvm/test/Transforms/SLPVectorizer/X86/shrink_after_reorder.ll
+++ b/llvm/test/Transforms/SLPVectorizer/X86/shrink_after_reorder.ll
@@ -12,9 +12,9 @@ define void @wombat(i32* %ptr, i32* %ptr1) {
; CHECK-NEXT: [[SHRINK_SHUFFLE:%.*]] = shufflevector <4 x i32> [[SHUFFLE]], <4 x i32> poison, <2 x i32> <i32 0, i32 1>
; CHECK-NEXT: [[TMP2:%.*]] = add nsw <2 x i32> [[SHRINK_SHUFFLE]], <i32 -1, i32 -1>
; CHECK-NEXT: [[SHUFFLE1:%.*]] = shufflevector <2 x i32> [[TMP2]], <2 x i32> poison, <4 x i32> <i32 0, i32 1, i32 0, i32 1>
-; CHECK-NEXT: [[TMP3:%.*]] = icmp sgt <4 x i32> [[SHUFFLE]], poison
-; CHECK-NEXT: [[TMP4:%.*]] = select <4 x i1> [[TMP3]], <4 x i32> poison, <4 x i32> [[SHUFFLE1]]
-; CHECK-NEXT: [[TMP5:%.*]] = select <4 x i1> poison, <4 x i32> zeroinitializer, <4 x i32> [[TMP4]]
+; CHECK-NEXT: [[TMP3:%.*]] = icmp sgt <4 x i32> [[SHUFFLE]], undef
+; CHECK-NEXT: [[TMP4:%.*]] = select <4 x i1> [[TMP3]], <4 x i32> undef, <4 x i32> [[SHUFFLE1]]
+; CHECK-NEXT: [[TMP5:%.*]] = select <4 x i1> zeroinitializer, <4 x i32> zeroinitializer, <4 x i32> [[TMP4]]
; CHECK-NEXT: [[TMP6:%.*]] = bitcast i32* [[TMP27]] to <4 x i32>*
; CHECK-NEXT: store <4 x i32> [[TMP5]], <4 x i32>* [[TMP6]], align 8
; CHECK-NEXT: ret void
diff --git a/llvm/test/Transforms/SLPVectorizer/X86/shrink_after_reorder2.ll b/llvm/test/Transforms/SLPVectorizer/X86/shrink_after_reorder2.ll
index caf0a3d835748..533b0dad45a5b 100644
--- a/llvm/test/Transforms/SLPVectorizer/X86/shrink_after_reorder2.ll
+++ b/llvm/test/Transforms/SLPVectorizer/X86/shrink_after_reorder2.ll
@@ -26,7 +26,7 @@ define void @foo(%class.e* %this, %struct.a* %p, i32 %add7) {
; CHECK: sw.epilog:
; CHECK-NEXT: [[TMP6:%.*]] = phi <2 x i32> [ undef, [[ENTRY:%.*]] ], [ [[TMP5]], [[SW_BB]] ]
; CHECK-NEXT: [[SHUFFLE1:%.*]] = shufflevector <2 x i32> [[TMP6]], <2 x i32> poison, <4 x i32> <i32 1, i32 1, i32 0, i32 0>
-; CHECK-NEXT: [[TMP7:%.*]] = sub <4 x i32> poison, [[SHUFFLE]]
+; CHECK-NEXT: [[TMP7:%.*]] = sub <4 x i32> undef, [[SHUFFLE]]
; CHECK-NEXT: [[TMP8:%.*]] = add <4 x i32> [[TMP7]], [[SHUFFLE1]]
; CHECK-NEXT: [[TMP9:%.*]] = bitcast i32* [[B]] to <4 x i32>*
; CHECK-NEXT: store <4 x i32> [[TMP8]], <4 x i32>* [[TMP9]], align 4
diff --git a/llvm/test/Transforms/SLPVectorizer/X86/vectorize-reorder-alt-shuffle.ll b/llvm/test/Transforms/SLPVectorizer/X86/vectorize-reorder-alt-shuffle.ll
index 120eec38c6db3..290c0d215cc56 100644
--- a/llvm/test/Transforms/SLPVectorizer/X86/vectorize-reorder-alt-shuffle.ll
+++ b/llvm/test/Transforms/SLPVectorizer/X86/vectorize-reorder-alt-shuffle.ll
@@ -12,9 +12,9 @@ define void @foo(i8* %c, float* %d) {
; CHECK-NEXT: [[TMP3:%.*]] = shl nuw nsw <4 x i32> [[TMP2]], <i32 2, i32 2, i32 2, i32 3>
; CHECK-NEXT: [[TMP4:%.*]] = and <4 x i32> [[TMP2]], <i32 2, i32 2, i32 2, i32 3>
; CHECK-NEXT: [[TMP5:%.*]] = shufflevector <4 x i32> [[TMP3]], <4 x i32> [[TMP4]], <4 x i32> <i32 1, i32 2, i32 7, i32 0>
-; CHECK-NEXT: [[TMP6:%.*]] = add nsw <4 x i32> poison, [[TMP5]]
+; CHECK-NEXT: [[TMP6:%.*]] = add nsw <4 x i32> undef, [[TMP5]]
; CHECK-NEXT: [[TMP7:%.*]] = sitofp <4 x i32> [[TMP6]] to <4 x float>
-; CHECK-NEXT: [[TMP8:%.*]] = fdiv <4 x float> [[TMP7]], poison
+; CHECK-NEXT: [[TMP8:%.*]] = fdiv <4 x float> [[TMP7]], undef
; CHECK-NEXT: [[TMP9:%.*]] = bitcast float* [[ADD_PTR53]] to <4 x float>*
; CHECK-NEXT: store <4 x float> [[TMP8]], <4 x float>* [[TMP9]], align 4
; CHECK-NEXT: ret void
diff --git a/llvm/test/Transforms/SLPVectorizer/slp-umax-rdx-matcher-crash.ll b/llvm/test/Transforms/SLPVectorizer/slp-umax-rdx-matcher-crash.ll
index b2ae15cf6166f..13c8109b016ed 100644
--- a/llvm/test/Transforms/SLPVectorizer/slp-umax-rdx-matcher-crash.ll
+++ b/llvm/test/Transforms/SLPVectorizer/slp-umax-rdx-matcher-crash.ll
@@ -43,8 +43,8 @@ declare i32 @llvm.umin.i32(i32, i32)
define void @test2() {
; CHECK-LABEL: @test2(
; CHECK-NEXT: entry:
-; CHECK-NEXT: [[TMP0:%.*]] = call <4 x i32> @llvm.smin.v4i32(<4 x i32> poison, <4 x i32> <i32 2, i32 3, i32 1, i32 0>)
-; CHECK-NEXT: [[TMP1:%.*]] = sub nsw <4 x i32> poison, [[TMP0]]
+; CHECK-NEXT: [[TMP0:%.*]] = call <4 x i32> @llvm.smin.v4i32(<4 x i32> undef, <4 x i32> <i32 2, i32 3, i32 1, i32 0>)
+; CHECK-NEXT: [[TMP1:%.*]] = sub nsw <4 x i32> undef, [[TMP0]]
; CHECK-NEXT: [[TMP2:%.*]] = call i32 @llvm.vector.reduce.umin.v4i32(<4 x i32> [[TMP1]])
; CHECK-NEXT: [[TMP3:%.*]] = call i32 @llvm.umin.i32(i32 [[TMP2]], i32 77)
; CHECK-NEXT: [[E:%.*]] = icmp ugt i32 [[TMP3]], 1
More information about the llvm-commits
mailing list