projects / platform / upstream / llvm.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: 2fe30a3) | patch
[mlir] Add an option to control the number of loops in affine parallelizer
authorAlex Zinenko <zinenko@google.com>
Mon, 7 Dec 2020 15:18:32 +0000 (16:18 +0100)
committerAlex Zinenko <zinenko@google.com>
Tue, 8 Dec 2020 09:44:37 +0000 (10:44 +0100)
commit80766ecc65096deeb4ff6f03562dcad94c54b862
tree2b1c3c71448d0fef446467b9e22ae25764cd94e7tree | snapshot
parent2fe30a3534dad9f982a3d840b4bfa4870b2ba5bccommit | diff
[mlir] Add an option to control the number of loops in affine parallelizer

Add a pass option to control the number of nested parallel loops produced by
the parallelization passes. This is useful to build end-to-end passes targeting
systems that don't need multiple parallel dimensions (e.g., CPUs typically need
only one).

Reviewed By: wsmoses, chelini

Differential Revision: https://reviews.llvm.org/D92765
mlir/include/mlir/Dialect/Affine/Passes.td diff | blob | history
mlir/lib/Dialect/Affine/Transforms/AffineParallelize.cpp diff | blob | history
mlir/test/Dialect/Affine/parallelize.mlir diff | blob | history
Domain: System / Toolchain;