[PATCH] D138494: [AMDGPU][MC][MI100+] Enable VOP3 variants of dot2c/dot4c/dot8c opcodes

Dmitry Preobrazhensky via Phabricator via llvm-commits llvm-commits at lists.llvm.org
Tue Nov 22 06:25:57 PST 2022


dp created this revision.
dp added reviewers: rampitec, foad.
Herald added subscribers: kosarev, kerbowa, hiraditya, tpr, dstuttard, yaxunl, jvesely, kzhuravl, arsenm.
Herald added a project: All.
dp requested review of this revision.
Herald added subscribers: llvm-commits, wdng.
Herald added a project: LLVM.

Unfortunately AMD documentation does not specify which modifiers should be supported for these instructions.
The patch enables `abs`, `neg`, `clamp` and `omod` modifiers for `v_dot2c_f32_f16` and `clamp` for remaining `dotc` opcodes, but it is a somewhat arbitrary decision.


https://reviews.llvm.org/D138494

Files:
  llvm/lib/Target/AMDGPU/AsmParser/AMDGPUAsmParser.cpp
  llvm/lib/Target/AMDGPU/Disassembler/AMDGPUDisassembler.cpp
  llvm/lib/Target/AMDGPU/Utils/AMDGPUBaseInfo.cpp
  llvm/lib/Target/AMDGPU/Utils/AMDGPUBaseInfo.h
  llvm/lib/Target/AMDGPU/VOP2Instructions.td
  llvm/test/MC/AMDGPU/xdl-insts-err.s
  llvm/test/MC/AMDGPU/xdl-insts-gfx908.s
  llvm/test/MC/Disassembler/AMDGPU/gfx908-xdl-insts.txt

-------------- next part --------------
A non-text attachment was scrubbed...
Name: D138494.477152.patch
Type: text/x-patch
Size: 18154 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20221122/e2b629cd/attachment.bin>


More information about the llvm-commits mailing list