[LV] Clarify code in isPredicatedInst (NFC)

[artagnon]

No description provided.

[llvmbot]

@llvm/pr-subscribers-llvm-transforms

@llvm/pr-subscribers-vectorizers

Author: Ramkumar Ramachandra (artagnon)

Changes

---

Full diff: https://github.com/llvm/llvm-project/pull/134251.diff

1 Files Affected:

• (modified) llvm/lib/Transforms/Vectorize/LoopVectorize.cpp (+5-3)

diff --git a/llvm/lib/Transforms/Vectorize/LoopVectorize.cpp b/llvm/lib/Transforms/Vectorize/LoopVectorize.cpp
index 55cc801e91452..727098bd19a56 100644
--- a/llvm/lib/Transforms/Vectorize/LoopVectorize.cpp
+++ b/llvm/lib/Transforms/Vectorize/LoopVectorize.cpp
@@ -3244,11 +3244,9 @@ bool LoopVectorizationCostModel::isScalarWithPredication(
 
 // TODO: Fold into LoopVectorizationLegality::isMaskRequired.
 bool LoopVectorizationCostModel::isPredicatedInst(Instruction *I) const {
-  // If predication is not needed, avoid it.
   // TODO: We can use the loop-preheader as context point here and get
   // context sensitive reasoning for isSafeToSpeculativelyExecute.
-  if (!blockNeedsPredicationForAnyReason(I->getParent()) ||
-      isSafeToSpeculativelyExecute(I) ||
+  if (isSafeToSpeculativelyExecute(I) ||
       (isa<LoadInst, StoreInst, CallInst>(I) && !Legal->isMaskRequired(I)) ||
       isa<BranchInst, SwitchInst, PHINode, AllocaInst>(I))
     return false;
@@ -3258,6 +3256,10 @@ bool LoopVectorizationCostModel::isPredicatedInst(Instruction *I) const {
   if (Legal->blockNeedsPredication(I->getParent()))
     return true;
 
+  // If we're not folding the tail by masking, predication is unnecessary.
+  if (!foldTailByMasking())
+    return false;
+
   // All that remain are instructions with side-effects originally executed in
   // the loop unconditionally, but now execute under a tail-fold mask (only)
   // having at least one active lane (the first). If the side-effects of the

[fhahn]

I am not sure which aspect this clarifies?

[artagnon]

I am not sure which aspect this clarifies?

See following comment:

  // All that remain are instructions with side-effects originally executed in
  // the loop unconditionally, but now execute under a tail-fold mask (only)
  // having at least one active lane (the first). If the side-effects of the
  // instruction are invariant, executing it w/o (the tail-folding) mask is safe
  // - it will cause the same side-effects as when masked.

We make it clear that we execute under a tail-fold mask (only). Also, this is a preparatory step for #134261.

[fhahn]

LGTM, thanks