mirror of https://github.com/llvm/torch-mlir
a82ba1c422
Added lit tests since these scalar operations don't trace well through the `fx_importer` route. `XOR` and `NE` are equivalent binary operators, so `aten.ne.bool` is lowered to `arith.xori`. |
||
---|---|---|
.. | ||
CMakeLists.txt | ||
TorchToArith.cpp |