projects / platform / upstream / llvm.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: 0d37efd) | patch
[mlir][memref] Use folded composed affine apply ops in FoldMemRefAliasOps
authorNicolas Vasilache <nicolas.vasilache@gmail.com>
Mon, 20 Mar 2023 21:22:00 +0000 (14:22 -0700)
committerNicolas Vasilache <nicolas.vasilache@gmail.com>
Wed, 22 Mar 2023 05:17:36 +0000 (22:17 -0700)
commit829446cb459fee326fc7f6ebf84b357e1665e4d1
tree64595d8c1da5c142755bd7b43acecbeaa0a002edtree | snapshot
parent0d37efdbc599e61ce2a0418723a66d6b45aea8d7commit | diff
[mlir][memref] Use folded composed affine apply ops in FoldMemRefAliasOps

Creating maximally folded and composd affine.apply operation during
FoldMemRefAliasOps composes better with other transformations without having
to interleave canonicalization passes.

Differential Revision: https://reviews.llvm.org/D146515
mlir/lib/Dialect/MemRef/Transforms/FoldMemRefAliasOps.cpp diff | blob | history
mlir/test/Dialect/MemRef/fold-memref-alias-ops.mlir diff | blob | history
Domain: System / Toolchain;