[SLP][REVEC] reorderNodeWithReuses should not be called if all users of a TreeEntry are ShuffleVectorInst. #118260
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
…of a TreeEntry are ShuffleVectorInst.
Your org has enabled the Graphite merge queue for merging into mainAdd the label “FP Bundles” to the PR and Graphite will automatically add it to the merge queue when it’s ready to merge. You must have a Graphite account and log in to Graphite in order to use the merge queue. Sign up using this link. |
Member
|
@llvm/pr-subscribers-vectorizers @llvm/pr-subscribers-llvm-transforms Author: Han-Kuan Chen (HanKuanChen) ChangesFull diff: https://github.com/llvm/llvm-project/pull/118260.diff 2 Files Affected:
diff --git a/llvm/lib/Transforms/Vectorize/SLPVectorizer.cpp b/llvm/lib/Transforms/Vectorize/SLPVectorizer.cpp
index 04755102643364..74f5df11ed6a9e 100644
--- a/llvm/lib/Transforms/Vectorize/SLPVectorizer.cpp
+++ b/llvm/lib/Transforms/Vectorize/SLPVectorizer.cpp
@@ -6073,6 +6073,23 @@ void BoUpSLP::reorderTopToBottom() {
TE->Scalars.size();
}) &&
"All users must be of VF size.");
+ if (SLPReVec) {
+ assert(SLPReVec && "Only supported by REVEC.");
+ // ShuffleVectorInst does not do reorderOperands (and it should not
+ // because ShuffleVectorInst supports only a limited set of
+ // patterns). Only do reorderNodeWithReuses if all of the users are
+ // not ShuffleVectorInst.
+ if (all_of(TE->UserTreeIndices, [&](const EdgeInfo &EI) {
+ return isa<ShuffleVectorInst>(EI.UserTE->getMainOp());
+ }))
+ continue;
+ assert(all_of(TE->UserTreeIndices,
+ [&](const EdgeInfo &EI) {
+ return !isa<ShuffleVectorInst>(
+ EI.UserTE->getMainOp());
+ }) &&
+ "Does not know how to reorder.");
+ }
// Update ordering of the operands with the smaller VF than the given
// one.
reorderNodeWithReuses(*TE, Mask);
diff --git a/llvm/test/Transforms/SLPVectorizer/revec.ll b/llvm/test/Transforms/SLPVectorizer/revec.ll
index b160c0174c0a76..ce13f478d38116 100644
--- a/llvm/test/Transforms/SLPVectorizer/revec.ll
+++ b/llvm/test/Transforms/SLPVectorizer/revec.ll
@@ -447,3 +447,37 @@ for.end.loopexit:
store <4 x i32> %4, ptr %out2, align 4
ret void
}
+
+define void @test14(<8 x i1> %0) {
+; CHECK-LABEL: @test14(
+; CHECK-NEXT: entry:
+; CHECK-NEXT: [[TMP1:%.*]] = call <16 x i1> @llvm.vector.insert.v16i1.v8i1(<16 x i1> poison, <8 x i1> [[TMP0:%.*]], i64 0)
+; CHECK-NEXT: [[TMP2:%.*]] = shufflevector <16 x i1> [[TMP1]], <16 x i1> poison, <16 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7, i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7>
+; CHECK-NEXT: [[TMP3:%.*]] = sext <16 x i1> [[TMP2]] to <16 x i16>
+; CHECK-NEXT: [[TMP4:%.*]] = shufflevector <16 x i16> [[TMP3]], <16 x i16> poison, <32 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7, i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7, i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15, i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15>
+; CHECK-NEXT: [[TMP5:%.*]] = shufflevector <16 x i16> [[TMP3]], <16 x i16> poison, <16 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7, i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15>
+; CHECK-NEXT: [[TMP6:%.*]] = shufflevector <16 x i16> [[TMP3]], <16 x i16> poison, <16 x i32> <i32 0, i32 1, i32 2, i32 3, i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15, i32 4, i32 5, i32 6, i32 7>
+; CHECK-NEXT: br label [[FOR_END_LOOPEXIT:%.*]]
+; CHECK: for.end.loopexit:
+; CHECK-NEXT: [[TMP7:%.*]] = phi <16 x i16> [ [[TMP6]], [[ENTRY:%.*]] ]
+; CHECK-NEXT: [[TMP8:%.*]] = call <4 x i16> @llvm.vector.extract.v4i16.v16i16(<16 x i16> [[TMP7]], i64 12)
+; CHECK-NEXT: [[OR0:%.*]] = or <4 x i16> [[TMP8]], zeroinitializer
+; CHECK-NEXT: ret void
+;
+entry:
+ %sext0 = sext <8 x i1> %0 to <8 x i16>
+ %sext1 = sext <8 x i1> %0 to <8 x i16>
+ %1 = shufflevector <8 x i16> %sext0, <8 x i16> zeroinitializer, <4 x i32> <i32 0, i32 1, i32 2, i32 3>
+ %2 = shufflevector <8 x i16> %sext0, <8 x i16> zeroinitializer, <4 x i32> <i32 4, i32 5, i32 6, i32 7>
+ %3 = shufflevector <8 x i16> %sext1, <8 x i16> zeroinitializer, <4 x i32> <i32 0, i32 1, i32 2, i32 3>
+ %4 = shufflevector <8 x i16> %sext1, <8 x i16> zeroinitializer, <4 x i32> <i32 4, i32 5, i32 6, i32 7>
+ br label %for.end.loopexit
+
+for.end.loopexit:
+ %phi0 = phi <4 x i16> [ %1, %entry ]
+ %phi1 = phi <4 x i16> [ %2, %entry ]
+ %phi2 = phi <4 x i16> [ %3, %entry ]
+ %phi3 = phi <4 x i16> [ %4, %entry ]
+ %or0 = or <4 x i16> %phi1, zeroinitializer
+ ret void
+}
|
Co-authored-by: Alexey Bataev <[email protected]>
|
✅ With the latest revision this PR passed the C/C++ code formatter. |
alexey-bataev
approved these changes
Dec 2, 2024
Collaborator
|
LLVM Buildbot has detected a new failure on builder Full details are available at: https://lab.llvm.org/buildbot/#/builders/131/builds/11538 Here is the relevant piece of the build log for the reference |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
4 participants
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
No description provided.