projects / platform / upstream / llvm.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: 75a6289) | patch
[mlir][AMDGPU] Add lds_barrier op
authorKrzysztof Drewniak <Krzysztof.Drewniak@amd.com>
Fri, 8 Jul 2022 19:11:03 +0000 (19:11 +0000)
committerKrzysztof Drewniak <Krzysztof.Drewniak@amd.com>
Thu, 14 Jul 2022 20:45:26 +0000 (20:45 +0000)
commitbc61cc9a2db56eeb5fd299132037757da339aebd
tree0aca14ca35b2085b3f1bafbbe44bf46bd67084c3tree | snapshot
parent75a628925e88c62aa192a0c8ab165b840fc334a6commit | diff
[mlir][AMDGPU] Add lds_barrier op

The lds_barrier op allows workgroups to wait at a barrier for
operations to/from their local data store (LDS) to complete without
incurring the performance penalties of a full memory fence.

Reviewed By: nirvedhmeshram

Differential Revision: https://reviews.llvm.org/D129522
mlir/include/mlir/Dialect/AMDGPU/AMDGPU.td diff | blob | history
mlir/lib/Conversion/AMDGPUToROCDL/AMDGPUToROCDL.cpp diff | blob | history
mlir/test/Conversion/AMDGPUToROCDL/amdgpu-to-rocdl.mlir diff | blob | history
mlir/test/Dialect/AMDGPU/ops.mlir diff | blob | history
Domain: System / Toolchain;