[PATCH] D138591: [AAch64] Don't treat SVE scalable extends as free widening instructions

[Sander de Smalen via Phabricator via llvm-commits]

sdesmalen added inline comments.


================
Comment at: llvm/lib/Target/AArch64/AArch64TargetTransformInfo.cpp:1570
+  // interleaving to be used with zext/sext.
+  if (!isa<FixedVectorType>(DstTy) || DstTy->getScalarSizeInBits() < 16)
     return false;
----------------
It's probably better to use `!useNeonVector(DstTy)` for this, given that fixed-width SVE has the same issue.

You may want to add a similar test to llvm/test/Analysis/CostModel/AArch64/sve-fixed-length.ll for this as well.


CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D138591/new/

https://reviews.llvm.org/D138591