projects / platform / upstream / llvm.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: 4f5355e) | patch
[mlir][sparse] generalize invariant expression handling in sparse compiler
authorAart Bik <ajcbik@google.com>
Mon, 23 Nov 2020 19:01:39 +0000 (11:01 -0800)
committerAart Bik <ajcbik@google.com>
Tue, 24 Nov 2020 21:41:14 +0000 (13:41 -0800)
commitb228e2bd92a7959b6a48ac208871ad0cb5c09ebf
tree780c8e69b0b7230fbb9738be623d807380c1bbbatree | snapshot
parent4f5355ee73626f8b8fe6bf0dd6d167fea7628a2ccommit | diff
[mlir][sparse] generalize invariant expression handling in sparse compiler

Generalizes invariant handling to anything defined outside the Linalg op
(parameters and SSA computations). Fixes bug that was using parameter number
as tensor number.

Reviewed By: penpornk

Differential Revision: https://reviews.llvm.org/D91985
mlir/lib/Dialect/Linalg/Transforms/Sparsification.cpp diff | blob | history
mlir/test/Dialect/Linalg/sparse_2d.mlir diff | blob | history
Domain: System / Toolchain;