[PATCH] D150390: [AMDGPU] Introduce and use the new PRED_COPY opcode

Mon May 22 10:34:30 PDT 2023

yassingh added a comment.

In D150390#4361124 <https://reviews.llvm.org/D150390#4361124>, @arsenm wrote:

> In D150390#4361021 <https://reviews.llvm.org/D150390#4361021>, @cdevadas wrote:
>
>> In D150390#4360314 <https://reviews.llvm.org/D150390#4360314>, @arsenm wrote:
>>
>>> I don’t understand why you are updating this patch when it’s replaced with the new live range splitting hooks
>>
>> This is target handling for introducing the LiveRange Split opcode (Pred-COPY) for AMDGPU and together with D150388 <https://reviews.llvm.org/D150388>, this patch will avoid my initial Pred-COPY generic opcode implementation patches D143754 <https://reviews.llvm.org/D143754> & D143757 <https://reviews.llvm.org/D143757>.
>
> More specifically I meant SimplifyPredicatedCopies. If they're only inserted in contexts where they are required, then we wouldn't need a pass to remove the unnecessary cases

We could merge WWM code insertion logic with SIFixVGPRCopies while still lowering PRED_COPY back to COPY there(to prevent duplication of COPY lowering logic)? That would avoid running a separate pass. Or are you suggesting another approach altogether to handle PRED_COPY instructions?

Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D150390/new/

https://reviews.llvm.org/D150390