projects / platform / upstream / llvm.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: 53f738c) | patch
[mlir][sparse] Folding operations that try to insert zero into an all-zero sparse...
authorPeiming Liu <peiming@google.com>
Tue, 23 Aug 2022 21:15:10 +0000 (21:15 +0000)
committerPeiming Liu <peiming@google.com>
Thu, 25 Aug 2022 17:00:04 +0000 (17:00 +0000)
commitec495b53f8b2ddf02626fd3d12d1ec7d36bd2bca
treea434d79d833063234acbf7fe22e55f48923005a7tree | snapshot
parent53f738ce7e6cc22d58c6cc1042c7b49cd81f75aecommit | diff
[mlir][sparse] Folding operations that try to insert zero into an all-zero sparse tensor

The operations to fill zero into newly allocated sparse tensor are redundant, plus it failed
to lowering the test cases provided in the patch as well.

Reviewed By: aartbik

Differential Revision: https://reviews.llvm.org/D132500
mlir/lib/Dialect/SparseTensor/Transforms/SparseTensorRewriting.cpp diff | blob | history
mlir/test/Dialect/SparseTensor/sparse_fill_zero.mlir [new file with mode: 0644] blob
Domain: System / Toolchain;