diff options
| author | Christopher Bate <cbate@nvidia.com> | 2022-11-05 09:43:24 -0600 |
|---|---|---|
| committer | Christopher Bate <cbate@nvidia.com> | 2022-11-07 09:43:03 -0700 |
| commit | 708185f03ff480b3481132802b7b63461564f0ab (patch) | |
| tree | 4d4d94334b5ed977416ef1d4f1f8015602cfd7bf /flang/lib/Frontend/CompilerInvocation.cpp | |
| parent | 6ebca0302126c43ec8614d26aa444060e7a6da76 (diff) | |
| download | llvm-708185f03ff480b3481132802b7b63461564f0ab.zip llvm-708185f03ff480b3481132802b7b63461564f0ab.tar.gz llvm-708185f03ff480b3481132802b7b63461564f0ab.tar.bz2 | |
[mlir][NVGPU] Add support for structured sparsity MMA variants
This change adds a new NVGPU operation that targets the PTX `mma.sp.sync`
instruction variants. A lowering to NVVM is provided using inline
assembly.
Reviewed By: ThomasRaoux, manishucsd
Differential Revision: https://reviews.llvm.org/D137202
Diffstat (limited to 'flang/lib/Frontend/CompilerInvocation.cpp')
0 files changed, 0 insertions, 0 deletions