[PATCH] D136432: [AMDGPU] Combine BFI instructions.

[Thomas Symalla via Phabricator via llvm-commits]

tsymalla added a comment.

In D136432#3873859 <https://reviews.llvm.org/D136432#3873859>, @jsilvanus wrote:

> One inline comment on a comment :).
>
> Slightly related: How do we handle `(X1 & Y) | (X2 & ~Y)` where `Y` is a variable?

Thanks!
As such case cannot be combined into a BFI instruction, I guess it's relatively straight-forwardly translated into something like this:

  v_and_b32 v0, x1, y ; v0 = X1 & Y
  v_not_b32 v1, y ; v1 = ~Y
  v_and_or_b32 v1, v1, x2, v1 ; v1 = (~Y & X2) | (X1 & Y)


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D136432/new/

https://reviews.llvm.org/D136432