projects / platform / upstream / llvm.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: 7811ad3) | patch
Add gpu.shuffle op.
authorChristian Sigg <csigg@google.com>
Fri, 20 Dec 2019 10:52:21 +0000 (02:52 -0800)
committerA. Unique TensorFlower <gardener@tensorflow.org>
Fri, 20 Dec 2019 10:52:52 +0000 (02:52 -0800)
commit42d46b4efa9d7b596c74c101a777f4464a8f1fd0
tree1329db2e71549b610e2eb121528ea8b3284411e7tree | snapshot
parent7811ad3c2b312fb5eda5ed5f3a1d15b8e6085b24commit | diff
Add gpu.shuffle op.

This will allow us to lower most of gpu.all_reduce (when all_reduce
doesn't exist in the target dialect) within the GPU dialect, and only do
target-specific lowering for the shuffle op.

PiperOrigin-RevId: 286548256
mlir/include/mlir/Dialect/GPU/GPUDialect.h diff | blob | history
mlir/include/mlir/Dialect/GPU/GPUOps.td diff | blob | history
mlir/lib/Conversion/GPUToNVVM/LowerGpuOpsToNVVMOps.cpp diff | blob | history
mlir/lib/Dialect/GPU/IR/GPUDialect.cpp diff | blob | history
mlir/test/Conversion/GPUToNVVM/gpu-to-nvvm.mlir diff | blob | history
mlir/test/Dialect/GPU/invalid.mlir diff | blob | history
mlir/test/Dialect/GPU/ops.mlir diff | blob | history
mlir/test/mlir-cuda-runner/shuffle.mlir [new file with mode: 0644] blob
Domain: System / Toolchain;