[PATCH] D147408: [AMDGPU] Iterative scan implementation for atomic optimizer.

[Pravin Jagtap via Phabricator via llvm-commits]

pravinjagtap updated this revision to Diff 515605.
pravinjagtap retitled this revision from "[AMDGPU] Enable AMDGPU Atomic Optimizer Pass by default." to "[AMDGPU] Iterative scan implementation for atomic optimizer.".
pravinjagtap edited the summary of this revision.
pravinjagtap added a comment.

Introduced new command line flag `amdgpu-atomic-optimizer-use-dpp` which selects the scan implementation(DPP/Iterative).

Following is the plan for submission:

1. Submit this change with -amdgpu-atomic-optimizer-use-dpp=on and -amdgpu-atomic-optimizations=off
2. Downstream clients can opt in using -amdgpu-atomic-optimizer-use-dpp=on and -amdgpu-atomic-optimizations=on
3. Change default -amdgpu-atomic-optimizations to on and -amdgpu-atomic-optimizer-use-dpp to off so that compute uses iterative approach


CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D147408/new/

https://reviews.llvm.org/D147408

Files:
  llvm/lib/Target/AMDGPU/AMDGPU.h
  llvm/lib/Target/AMDGPU/AMDGPUAtomicOptimizer.cpp
  llvm/lib/Target/AMDGPU/AMDGPUTargetMachine.cpp
  llvm/test/CodeGen/AMDGPU/GlobalISel/atomic_optimizations_mul_one.ll
  llvm/test/CodeGen/AMDGPU/atomic_optimizations_buffer.ll
  llvm/test/CodeGen/AMDGPU/atomic_optimizations_global_pointer.ll
  llvm/test/CodeGen/AMDGPU/atomic_optimizations_local_pointer.ll
  llvm/test/CodeGen/AMDGPU/atomic_optimizations_raw_buffer.ll
  llvm/test/CodeGen/AMDGPU/atomic_optimizations_struct_buffer.ll
  llvm/test/CodeGen/AMDGPU/buffer-intrinsics-mmo-offsets.ll
  llvm/test/CodeGen/AMDGPU/llc-pipeline.ll
  llvm/test/CodeGen/AMDGPU/noclobber-barrier.ll

-------------- next part --------------
A non-text attachment was scrubbed...
Name: D147408.515605.patch
Type: text/x-patch
Size: 506207 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20230421/c6ba8bed/attachment-0001.bin>