In the case of all inputs being constant and equal, cstr_eq will be
replaced with a true_witness.
Details
Details
Diff Detail
Diff Detail
- Repository
- rG LLVM Github Monorepo
Event Timeline
Comment Actions
LGTM, but please look into making these be folds instead of canonicalization patterns. That should also make the code simpler since the folding infra will give you the folded attributes pre-canned without having to do as much getDefiningOp<ConstShapeOp>().
mlir/lib/Dialect/Shape/IR/ShapeCanonicalization.td | ||
---|---|---|
14 | This is technically a fold, right? |
I think we are using true and successful witness interchangeably, should we just pick one?