[PATCH] D86114: [SVE] Lower fixed length vXi8/vXi16 SDIV to scalable

[Cameron McInally via Phabricator via llvm-commits]

cameron.mcinally added a comment.

In D86114#2222692 <https://reviews.llvm.org/D86114#2222692>, @efriedma wrote:

> I don't see any obvious difference in the RUN lines?

It's the VBITS_* differences. The VBITS_LE_* prefixes are better for checking the legalizations (really splitting), when the current VL >= the fixed VL. The VBITS_GE_* prefixes are better for ignoring the legalizations, i.e. the current VL < the fixed VL.

For the i8/i16 SDIV cases, the lowerings are pretty long, so it's nicer to check the cases where legalization was boring.


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D86114/new/

https://reviews.llvm.org/D86114