[llvm] [LV] Strengthen calls to collectInstsToScalarize (NFC) (PR #130642)

[Ramkumar Ramachandra via llvm-commits]

================
@@ -5393,20 +5393,21 @@ bool LoopVectorizationCostModel::useEmulatedMaskMemRefHack(Instruction *I,
 }
 
 void LoopVectorizationCostModel::collectInstsToScalarize(ElementCount VF) {
-  // If we aren't vectorizing the loop, or if we've already collected the
-  // instructions to scalarize, there's nothing to do. Collection may already
-  // have occurred if we have a user-selected VF and are now computing the
-  // expected cost for interleaving.
-  if (VF.isScalar() || VF.isZero() || InstsToScalarize.contains(VF))
+  assert(VF.isVector() && "Expected VF >= 2");
+
+  // If we've already collected the instructions to scalarize or the predicated
+  // BBs after vectorization, there's nothing to do. Collection may already have
+  // occurred if we have a user-selected VF and are now computing the expected
+  // cost for interleaving.
+  if (InstsToScalarize.contains(VF) ||
+      PredicatedBBsAfterVectorization.contains(VF))
     return;
 
   // Initialize a mapping for VF in InstsToScalalarize. If we find that it's
   // not profitable to scalarize any instructions, the presence of VF in the
   // map will indicate that we've analyzed it already.
   ScalarCostsTy &ScalarCostsVF = InstsToScalarize[VF];
 
-  PredicatedBBsAfterVectorization[VF].clear();
----------------
artagnon wrote:

Isn't the ` PredicatedBBsAfterVectorization.contains(VF)` added as part of this patch? So, before the change, we always clear the entries corresponding to VF before inserting new BBs (which is necessary for correctness), and after the change, we check if there are any entries corresponding to VF, and bail out early.

https://github.com/llvm/llvm-project/pull/130642