[all-commits] [llvm/llvm-project] e935fb: Remat for the first loop

[Jeffrey Byrnes via All-commits]

  Branch: refs/heads/users/jebyrnes/TritonInterleaveAndRematRebase0
  Home:   https://github.com/llvm/llvm-project
  Commit: e935fb93f5c13f723e3f6c0ad608e9277b1481ed
      https://github.com/llvm/llvm-project/commit/e935fb93f5c13f723e3f6c0ad608e9277b1481ed
  Author: Jeffrey Byrnes <Jeffrey.Byrnes at amd.com>
  Date:   2025-05-21 (Wed, 21 May 2025)

  Changed paths:
    M llvm/lib/CodeGen/RegAllocGreedy.cpp
    M llvm/lib/Target/AMDGPU/AMDGPUSubtarget.cpp
    M llvm/lib/Target/AMDGPU/AMDGPUSubtarget.h
    M llvm/lib/Target/AMDGPU/GCNSchedStrategy.cpp
    M llvm/lib/Target/AMDGPU/GCNSchedStrategy.h
    M llvm/lib/Target/AMDGPU/GCNSubtarget.cpp
    M llvm/lib/Target/AMDGPU/GCNSubtarget.h
    M llvm/lib/Target/AMDGPU/Utils/AMDGPUBaseInfo.cpp
    M llvm/lib/Target/AMDGPU/Utils/AMDGPUBaseInfo.h

  Log Message:
  -----------
  Remat for the first loop

Change-Id: Ic390b34e4c921325607d245a2cba4fb14f5b35ee


  Commit: f205dee1a5547715950bcf9eca7ec9617d75ebcb
      https://github.com/llvm/llvm-project/commit/f205dee1a5547715950bcf9eca7ec9617d75ebcb
  Author: Austin Kerbow <Austin.Kerbow at amd.com>
  Date:   2025-05-21 (Wed, 21 May 2025)

  Changed paths:
    M llvm/lib/Target/AMDGPU/AMDGPUIGroupLP.cpp
    M llvm/lib/Target/AMDGPU/AMDGPUTargetMachine.cpp
    M llvm/lib/Target/AMDGPU/GCNSchedStrategy.cpp
    M llvm/lib/Target/AMDGPU/GCNSchedStrategy.h
    M llvm/lib/Target/AMDGPU/SIInstrInfo.cpp
    M llvm/lib/Target/AMDGPU/SISchedule.td

  Log Message:
  -----------
  Experimental FA scheduling.

WIP

Changes
- Custom XDL resource tracking. Assume first 8 cycles of 8-pass mfma stalls the whole pipeline. Prefer MFMA if XDL is free, see `tryXDL` in `tryCand` for heuristic priorities.
- Fix exp latency.
- Force topdown scheduling in marked regions.
- Add `GCNPostSchedStrategy` and `tryCand` overrides for both the default schedulers.
- PostRA prefer MFMA with more successors.

Change-Id: Id1afc349c575484a093e9e179a93b1e93f2f4b4b


  Commit: eecfa53ab7d1bef8aba8073d1b2c45e4efee54fa
      https://github.com/llvm/llvm-project/commit/eecfa53ab7d1bef8aba8073d1b2c45e4efee54fa
  Author: Jeffrey Byrnes <Jeffrey.Byrnes at amd.com>
  Date:   2025-05-21 (Wed, 21 May 2025)

  Changed paths:
    M llvm/lib/Target/AMDGPU/GCNSchedStrategy.cpp

  Log Message:
  -----------
  [AMDGPU] Add option to disable scheduling rematerializaion


  Commit: 566172d1006d51403e8d0bc22b41b43d01f34e27
      https://github.com/llvm/llvm-project/commit/566172d1006d51403e8d0bc22b41b43d01f34e27
  Author: root <root at bg-1w300-g1-2a.mkm.dcgpu>
  Date:   2025-05-21 (Wed, 21 May 2025)

  Changed paths:
    M llvm/lib/Target/AMDGPU/GCNSchedStrategy.cpp

  Log Message:
  -----------
  switching

Change-Id: Ib54872aa396664dbc734201a665b9fd544e0fec6


  Commit: dbdcf7b1f2f84720c472b12a7ac1476fa6d85296
      https://github.com/llvm/llvm-project/commit/dbdcf7b1f2f84720c472b12a7ac1476fa6d85296
  Author: Austin Kerbow <Austin.Kerbow at amd.com>
  Date:   2025-05-21 (Wed, 21 May 2025)

  Changed paths:
    M llvm/lib/Target/AMDGPU/GCNSchedStrategy.cpp

  Log Message:
  -----------
  Experimental FA scheduling v2

WIP

Changes
- Fix crash on SU without instructions.
- Handle memops.


  Commit: 4871a63fde85c4501ad60fc0c55cd431524a5fbc
      https://github.com/llvm/llvm-project/commit/4871a63fde85c4501ad60fc0c55cd431524a5fbc
  Author: root <root at bg-1w300-g1-2a.mkm.dcgpu>
  Date:   2025-05-21 (Wed, 21 May 2025)

  Changed paths:
    M llvm/lib/CodeGen/RegAllocGreedy.cpp

  Log Message:
  -----------
  Fix some insert pt related bugs

Change-Id: If81d902530e427ec9e797f7d9e2d33dd68c3b198


  Commit: dba8becf82c0af66c1790be9f6280447d4083640
      https://github.com/llvm/llvm-project/commit/dba8becf82c0af66c1790be9f6280447d4083640
  Author: Jeffrey Byrnes <Jeffrey.Byrnes at amd.com>
  Date:   2025-05-21 (Wed, 21 May 2025)

  Changed paths:
    M llvm/lib/Target/AMDGPU/SIFoldOperands.cpp
    M llvm/test/CodeGen/AMDGPU/packed-fp32.ll

  Log Message:
  -----------
  Revert "AMDGPU: Move reg_sequence splat handling (#140313)"

This reverts commit 4ddab1252fe6a90111a034cef184549882aaba2b.

Change-Id: I861cf070d77cab094dda697ba4b4ce017407f1e0


  Commit: 0a49c1fa2cc72927ab31c852f05822bd8df4d95b
      https://github.com/llvm/llvm-project/commit/0a49c1fa2cc72927ab31c852f05822bd8df4d95b
  Author: Jeffrey Byrnes <Jeffrey.Byrnes at amd.com>
  Date:   2025-05-22 (Thu, 22 May 2025)

  Changed paths:
    M llvm/include/llvm/CodeGen/RegAllocPriorityAdvisor.h
    M llvm/include/llvm/CodeGen/TargetRegisterInfo.h
    M llvm/lib/CodeGen/RegAllocGreedy.cpp
    M llvm/lib/CodeGen/RegAllocGreedy.h
    M llvm/lib/CodeGen/RegAllocPriorityAdvisor.cpp
    M llvm/lib/Target/AMDGPU/GCNSchedStrategy.cpp
    M llvm/lib/Target/AMDGPU/GCNSchedStrategy.h
    M llvm/lib/Target/AMDGPU/SIRegisterInfo.h

  Log Message:
  -----------
  Auto select good flags

Change-Id: I5aa9f71c403215205fb98d79d866453659a42661


Compare: https://github.com/llvm/llvm-project/compare/e935fb93f5c1%5E...0a49c1fa2cc7

To unsubscribe from these emails, change your notification settings at https://github.com/llvm/llvm-project/settings/notifications