[PATCH] D133840: AMDGPU: Add a pass to rewrite certain undef in PHI

[Sameer Sahasrabuddhe via Phabricator via llvm-commits]

sameerds added inline comments.


================
Comment at: llvm/lib/Target/AMDGPU/AMDGPURewriteUndefForPHI.cpp:122
+            // Update DominateBB if necessary.
+            if (DT->dominates(IncomingBB, DominateBB))
+              DominateBB = IncomingBB;
----------------
ruiling wrote:
> sameerds wrote:
> > If this is not true, then does DominateBB dominate IncomingBB?
> No, this is used to handle case like with_uniform_region_inside added in this change. If IncomingBB does not dominate DominateBB, they may be successors of the same block in case their common predecessor has uniform branch.
My confusion is that the "all_of" check later in the code only ensures that DominateBB dominates all the undefined values. But how does one prove that UniqueDefinedIncoming dominates all the uses of the PHI? 


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D133840/new/

https://reviews.llvm.org/D133840