[all-commits] [llvm/llvm-project] 5d0ff9: AMDGPU: Promote array alloca if used by memmove/me...

[Ruiling, Song via All-commits]

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: 5d0ff923c3a7cc6c47b6010bbaf68592124110a5
      https://github.com/llvm/llvm-project/commit/5d0ff923c3a7cc6c47b6010bbaf68592124110a5
  Author: Ruiling Song <ruiling.song at amd.com>
  Date:   2023-01-11 (Wed, 11 Jan 2023)

  Changed paths:
    M llvm/lib/Target/AMDGPU/AMDGPUPromoteAlloca.cpp
    M llvm/test/CodeGen/AMDGPU/promote-alloca-array-aggregate.ll

  Log Message:
  -----------
  AMDGPU: Promote array alloca if used by memmove/memcpy

Reviewed by: arsenm

Differential Revision: https://reviews.llvm.org/D140599


  Commit: cce24b6af0999c658fd3e4931eb9bc58252478b8
      https://github.com/llvm/llvm-project/commit/cce24b6af0999c658fd3e4931eb9bc58252478b8
  Author: Ruiling Song <ruiling.song at amd.com>
  Date:   2023-01-11 (Wed, 11 Jan 2023)

  Changed paths:
    M llvm/lib/Target/AMDGPU/SIDefines.h
    M llvm/lib/Target/AMDGPU/SIInstrInfo.cpp

  Log Message:
  -----------
  AMDGPU: Remove IsSourceOfDivergence check

This bit is not set/reserved in td file. Let's remove it for now,
we can always add it back if we need it.

Reviewed by: foad

Differential Revision: https://reviews.llvm.org/D141223


  Commit: 9119d9bfcef47b245d15fc9d2e5044bc67724bfc
      https://github.com/llvm/llvm-project/commit/9119d9bfcef47b245d15fc9d2e5044bc67724bfc
  Author: Ruiling Song <ruiling.song at amd.com>
  Date:   2023-01-11 (Wed, 11 Jan 2023)

  Changed paths:
    M llvm/lib/Target/AMDGPU/BUFInstructions.td
    M llvm/lib/Target/AMDGPU/DSInstructions.td
    M llvm/lib/Target/AMDGPU/FLATInstructions.td
    M llvm/lib/Target/AMDGPU/SIDefines.h
    M llvm/lib/Target/AMDGPU/SIInsertWaitcnts.cpp
    M llvm/lib/Target/AMDGPU/SIInstrFormats.td
    M llvm/lib/Target/AMDGPU/SIInstrInfo.h
    M llvm/test/CodeGen/AMDGPU/chain-hi-to-lo.ll
    M llvm/test/CodeGen/AMDGPU/load-hi16.ll
    M llvm/test/CodeGen/AMDGPU/vector_shuffle.packed.ll

  Log Message:
  -----------
  AMDGPU/SIInsertWait: Skip dummy tied source

For D16 memory load instructions, the hardware usually only write to half
of the 32bit register, but we define the destination register using
32bit register for the MachineIR instruction. Without the extra tied
source register, LLVM framework will think previous write to the other
half of the register being dead. This is because by using 32bit register
as the destination register, LLVM will think the instruction will always
overwrite the whole 32bit register. By adding the extra tied source,
LLVM will think we are reading the register, so previous write to the
register will not be dead. This dummy tied source is introducing
unnecessary read-after-write dependency. The change here is to bypass the
tied source that can be skipped, thus avoiding an unnecessary s_waitcnt.

Reviewed by: foad

Differential Revision: https://reviews.llvm.org/D140537


Compare: https://github.com/llvm/llvm-project/compare/c0b475bd5ec9...9119d9bfcef4