[PATCH] D90404: [AMDGPU] Add alignment check for v3 to v4 load type promotion

[Jay Foad via Phabricator via llvm-commits]

foad added a comment.

In D90404#2364292 <https://reviews.llvm.org/D90404#2364292>, @foad wrote:

> As a further enhancement you could call isDereferenceable on the MachineMemOperand to see if the extra 4 bytes that a widened load would access are guaranteed to be dereferenceable.

Here's a quick demo of this, on top of your patch: https://reviews.llvm.org/differential/diff/301868/
It fixes all the regressions in kernel-args.ll and store-local.96.ll.


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D90404/new/

https://reviews.llvm.org/D90404