[PATCH] D141277: [InstCombine] Don't optimize idempotent `atomicrmw <op>, 0` into `load atomic`

[Matt Arsenault via Phabricator via llvm-commits]

arsenm added a comment.

In D141277#4060040 <https://reviews.llvm.org/D141277#4060040>, @arsenm wrote:

> Can you just insert a fence to go with the load?

AtomicExpand also tries the same transform (which for some reason is done in an x86 specific way, so this might be incomplete anyway)


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D141277/new/

https://reviews.llvm.org/D141277