[LAA] Fix non-NFC parts of 1aded51 #160701
Merged
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
1aded51 ([LAA] Prepare to handle diff type sizes (NFC)) was supposed to be a non-functional patch, but introduced functional changes as known-non-negative and known-non-positive is not equivalent to !known-non-zero. Fix this.
llvmbot
added
the
llvm:analysis
|
@llvm/pr-subscribers-llvm-analysis Author: Ramkumar Ramachandra (artagnon) Changes1aded51 ([LAA] Prepare to handle diff type sizes (NFC)) was supposed to be a non-functional patch, but introduced functional changes as known-non-negative and known-non-positive is not equivalent to !known-non-zero. Fix this. Full diff: https://github.com/llvm/llvm-project/pull/160701.diff 2 Files Affected:
diff --git a/llvm/lib/Analysis/LoopAccessAnalysis.cpp b/llvm/lib/Analysis/LoopAccessAnalysis.cpp
index d6ad855cad9a7..512ae415d1c3b 100644
--- a/llvm/lib/Analysis/LoopAccessAnalysis.cpp
+++ b/llvm/lib/Analysis/LoopAccessAnalysis.cpp
@@ -2122,7 +2122,8 @@ MemoryDepChecker::getDependenceDistanceStrideAndSize(
// dependence.
TypeSize AStoreSz = DL.getTypeStoreSize(ATy);
TypeSize BStoreSz = DL.getTypeStoreSize(BTy);
- if (AStoreSz != BStoreSz && !SE.isKnownNonZero(Dist)) {
+ if (AStoreSz != BStoreSz && SE.isKnownNonPositive(Dist) &&
+ SE.isKnownNonNegative(Dist)) {
LLVM_DEBUG(dbgs() << "LAA: possibly zero dependence distance with "
"different type sizes\n");
return Dependence::Unknown;
diff --git a/llvm/test/Analysis/LoopAccessAnalysis/depend_diff_types.ll b/llvm/test/Analysis/LoopAccessAnalysis/depend_diff_types.ll
index 023a8c056968f..c367b31f6d445 100644
--- a/llvm/test/Analysis/LoopAccessAnalysis/depend_diff_types.ll
+++ b/llvm/test/Analysis/LoopAccessAnalysis/depend_diff_types.ll
@@ -187,6 +187,45 @@ exit:
ret void
}
+; In the following test, dependence distance is possibly zero,
+; but this is not equivalent to the condition known-non-positive
+; and known-non-negative.
+
+define void @possibly_zero_dist_diff_typesz(ptr %p) {
+; CHECK-LABEL: 'possibly_zero_dist_diff_typesz'
+; CHECK-NEXT: loop:
+; CHECK-NEXT: Memory dependences are safe
+; CHECK-NEXT: Dependences:
+; CHECK-NEXT: Forward:
+; CHECK-NEXT: %ld.p = load i32, ptr %gep.p.iv.i32, align 1 ->
+; CHECK-NEXT: store i16 %trunc, ptr %gep.p.iv.i16, align 1
+; CHECK-EMPTY:
+; CHECK-NEXT: Run-time memory checks:
+; CHECK-NEXT: Grouped accesses:
+; CHECK-EMPTY:
+; CHECK-NEXT: Non vectorizable stores to invariant address were not found in loop.
+; CHECK-NEXT: SCEV assumptions:
+; CHECK-EMPTY:
+; CHECK-NEXT: Expressions re-written:
+;
+entry:
+ br label %loop
+
+loop:
+ %iv = phi i16 [ 0, %entry ], [ %iv.next, %loop ]
+ %gep.p.iv.i32 = getelementptr inbounds nuw i32, ptr %p, i16 %iv
+ %ld.p = load i32, ptr %gep.p.iv.i32, align 1
+ %trunc = trunc i32 %ld.p to i16
+ %gep.p.iv.i16 = getelementptr inbounds nuw i16, ptr %p, i16 %iv
+ store i16 %trunc, ptr %gep.p.iv.i16, align 1
+ %iv.next = add nuw nsw i16 %iv, 1
+ %exit.cond = icmp eq i16 %iv.next, 32
+ br i1 %exit.cond, label %exit, label %loop
+
+exit:
+ ret void
+}
+
; In the following test, the sink is loop-invariant.
define void @type_size_equivalence_sink_loopinv(ptr nocapture %vec, i64 %n) {
|
igogo-x86
approved these changes
Sep 25, 2025
There was a problem hiding this comment.
LGTM!
I tried to find a nicer way to check SCEV for zero and found this in ScalarEvolution.cpp:
const SCEV *Zero = SE.getZero(B->getType());
if (!SE.isKnownPredicate(CmpInst::ICMP_EQ, URem, Zero))
But I am not sure we need the strength behind isKnownPredicate and it may become non-NFC again
fhahn
reviewed
Sep 25, 2025
fhahn
approved these changes
Sep 25, 2025
|
Thank you! This does seem to fix the regressions we saw. |
|
Hi again @artagnon After some more testing we see that the following starts crashing with this patch and it still crashes on trunk (d48bda5): |
Thanks again for the subtle bug report with if (!LoopGuards)
LoopGuards.emplace(
ScalarEvolution::LoopGuards::collect(InnermostLoop, SE));
Dist = SE.applyLoopGuards(Dist, *LoopGuards);Will fix in the morning. |
|
We are starting to see the same assertion failure in our builders at Google. |
|
@artagnon -- Is it alright to land a revert to unblock our builders with weekend approaching? |
mahesh-attarde
pushed a commit
to mahesh-attarde/llvm-project
that referenced
this pull request
Oct 3, 2025
1aded51 ([LAA] Prepare to handle diff type sizes (NFC)) was supposed to be a non-functional patch, but introduced functional changes as known-non-negative and known-non-positive is not equivalent to !known-non-zero. Fix this.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
6 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.
1aded51 ([LAA] Prepare to handle diff type sizes (NFC)) was supposed to be a non-functional patch, but introduced functional changes as known-non-negative and known-non-positive is not equivalent to !known-non-zero. Fix this.