diff options
| author | MaheshRavishankar <ravishankarm@google.com> | 2021-02-16 13:15:55 -0800 |
|---|---|---|
| committer | MaheshRavishankar <ravishankarm@google.com> | 2021-02-16 13:42:08 -0800 |
| commit | 81264dfbe80df08668a325a61613b64243b99c01 (patch) | |
| tree | c0f09d209bd1f9822a57182bb0e88aa20f22ba31 /llvm/lib/MC/MCObjectFileInfo.cpp | |
| parent | 6ffcb2937c96bd0d7a55b984b5eb8f381b68e322 (diff) | |
| download | llvm-81264dfbe80df08668a325a61613b64243b99c01.zip llvm-81264dfbe80df08668a325a61613b64243b99c01.tar.gz llvm-81264dfbe80df08668a325a61613b64243b99c01.tar.bz2 | |
[mlir][Linalg] Add utility method to reshape ops to express output shape in terms of input shape.
Resolving the dim of outputs of a tensor_reshape op in terms of its
input shape allows the op to be eliminated when its used only in its
dims. The init_tensor -> tensor_reshape canonicalization can be
simplified to use the dims of the output of the tensor_reshape which
gets canonicalized away later making the tensor_reshape dead.
Differential Revision: https://reviews.llvm.org/D96635
Diffstat (limited to 'llvm/lib/MC/MCObjectFileInfo.cpp')
0 files changed, 0 insertions, 0 deletions