[PATCH] D151703: [AMDGPU][LSV] Restrict forming extra large vectors

[Piotr Sobczak via Phabricator via llvm-commits]

piotr added a comment.

In D151703#4380561 <https://reviews.llvm.org/D151703#4380561>, @arsenm wrote:

> What we should do is teach rematerialization to split scalar loads

Yes, that would work as well. Just to note, currently scalar loads are not rematerializable at all.


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D151703/new/

https://reviews.llvm.org/D151703