diff options
| author | Smit Hinsu <hinsu@google.com> | 2019-02-05 12:02:53 -0800 |
|---|---|---|
| committer | jpienaar <jpienaar@google.com> | 2019-03-29 16:15:08 -0700 |
| commit | 2927297a1cc85dec2e19df339dac696d271aff59 (patch) | |
| tree | 7fd57a5ed70df0a3f633f96492fd1235e533fade /mlir/lib/Transforms/LoopTiling.cpp | |
| parent | 40d5d09f9d52c581fd4419e8a54f4b952d904bb2 (diff) | |
| download | bcm5719-llvm-2927297a1cc85dec2e19df339dac696d271aff59.tar.gz bcm5719-llvm-2927297a1cc85dec2e19df339dac696d271aff59.zip | |
Add derived type attributes for TensorFlow ops generated by TableGen
Motivation for this change is to remove redundant TF type attributes for
TensorFlow ops. For example, tf$T: "tfdtype$DT_FLOAT". Type attributes can be derived using the MLIR operand or result MLIR types, attribute names and their mapping. This will also allow constant folding of instructions generated within MLIR (and not imported from TensorFlow) without adding type attributes for the instruction.
Derived attributes are populated while exporting MLIR to TF GraphDef using
auto-generated populators. Populators are only available for the ops that are generated by the TableGen.
Also, fixed Operator::getNumArgs method to exclude derived attributes as they are not
part of the arguments.
TESTED with unit test
PiperOrigin-RevId: 232531561
Diffstat (limited to 'mlir/lib/Transforms/LoopTiling.cpp')
0 files changed, 0 insertions, 0 deletions
