MaheshRavishankar 81264dfbe8 [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

2021-02-16 13:42:08 -08:00

..

CMakeLists.txt

[mlir][Linalg] NFC - Extract a standalone LinalgInterfaces

2021-02-04 07:19:38 +00:00

LinalgInterfaces.cpp

[mlir][linalg] Verify indexing map required attributes

2021-02-09 08:48:29 -05:00

LinalgOps.cpp

[mlir][Linalg] Add utility method to reshape ops to express output shape in terms of input shape.

2021-02-16 13:42:08 -08:00

LinalgTypes.cpp

[mlir][sparse] sparse tensor storage implementation

2021-02-10 11:57:24 -08:00