[PATCH] D129994: [RISCV] Add cost modelling for vector widenning integer reduction.

[Philip Reames via Phabricator via llvm-commits]

reames accepted this revision.
reames added a comment.
This revision is now accepted and ready to land.

LGTM

If I read your last comment correctly, the new test doesn't actually hit the relevant cost model change without follow up work right?  If so, please land it separately - as having cover for inloop reduction seems useful - and then land the code change on its own.  Given the false impression that the code is tested by landing them together would be potentially confusing,

Side note, I don't think we're going to ever prefer inloop reductions over out of loop ones.  Reduction instructions are generally log(VL) complexity or worse.  Unless maybe there's a case I'm missing?


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D129994/new/

https://reviews.llvm.org/D129994