[all-commits] [llvm/llvm-project] 3680b5: AMDGPU: Add baseline tests for mbcnt.hi combine

Matt Arsenault via All-commits all-commits at lists.llvm.org
Fri Jun 30 12:04:18 PDT 2023


  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: 3680b57a88233cd87928aa6741c5e04522b3b4a5
      https://github.com/llvm/llvm-project/commit/3680b57a88233cd87928aa6741c5e04522b3b4a5
  Author: Matt Arsenault <Matthew.Arsenault at amd.com>
  Date:   2023-06-30 (Fri, 30 Jun 2023)

  Changed paths:
    A llvm/test/Transforms/InstCombine/AMDGPU/mbcnt.ll

  Log Message:
  -----------
  AMDGPU: Add baseline tests for mbcnt.hi combine


  Commit: 5ccfc4543d1784ed98cfae52364e438de5a993a6
      https://github.com/llvm/llvm-project/commit/5ccfc4543d1784ed98cfae52364e438de5a993a6
  Author: Matt Arsenault <Matthew.Arsenault at amd.com>
  Date:   2023-06-30 (Fri, 30 Jun 2023)

  Changed paths:
    M llvm/lib/Target/AMDGPU/AMDGPUInstCombineIntrinsic.cpp
    M llvm/test/Transforms/InstCombine/AMDGPU/mbcnt.ll

  Log Message:
  -----------
  AMDGPU: Fold away mbcnt.hi in wave32 mode

This will allow libraries to drop some of the special casing based on
wave size.


  Commit: 8f9eee3602e01ba136660183e08d6f01da036bc0
      https://github.com/llvm/llvm-project/commit/8f9eee3602e01ba136660183e08d6f01da036bc0
  Author: Matt Arsenault <Matthew.Arsenault at amd.com>
  Date:   2023-06-30 (Fri, 30 Jun 2023)

  Changed paths:
    M llvm/test/CodeGen/AMDGPU/si-triv-disjoint-mem-access.ll

  Log Message:
  -----------
  AMDGPU: Fix opaque pointer conversion error in test

The * was in the wrong place so this was missed by the script.


  Commit: 917eddfdcb15bddf67a54ede1f1643d5fc83628d
      https://github.com/llvm/llvm-project/commit/917eddfdcb15bddf67a54ede1f1643d5fc83628d
  Author: Matt Arsenault <Matthew.Arsenault at amd.com>
  Date:   2023-06-30 (Fri, 30 Jun 2023)

  Changed paths:
    M clang/test/Headers/amdgcn_openmp_device_math_c.c

  Log Message:
  -----------
  OpenMP: Don't use target regions in library function test

Use pragma omp declare target to reduce the IR noise around the call
we're actually trying to test.


Compare: https://github.com/llvm/llvm-project/compare/03b118c7e456...917eddfdcb15


More information about the All-commits mailing list