[PATCH] D138050: AMDGPU/GlobalISel: Insert freeze when splitting vector G_SEXT_INREG

[Pierre van Houtryve via Phabricator via llvm-commits]

Pierre-vh accepted this revision.
Pierre-vh added a comment.
This revision is now accepted and ready to land.

In D138050#3930981 <https://reviews.llvm.org/D138050#3930981>, @arsenm wrote:

> In D138050#3930454 <https://reviews.llvm.org/D138050#3930454>, @Pierre-vh wrote:
>
>> What makes this combine specifically require a freeze? Could we have more combine that need it to or is it something with G_SEXT_INREG's semantics that makes it need the G_FREEZE?
>
> It's introducing an expectation for potentially poisonous bits in two distinct uses since the low half is used in two different places. Both uses need to use the same value. There are plenty of places that are probably missing freezes
>
> There have been a number of talks on freeze, e.g. https://www.youtube.com/watch?v=ZMaZH3YYJqY

Interesting, I'll definitely be looking out for cases like that in the future. I'll also watch that talk as soon as possible
Thanks :)


CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D138050/new/

https://reviews.llvm.org/D138050