[Mlir-commits] [mlir] [mlir][affine][gpu] support unroll dynamic value and apply it to gpu.thread_id op (PR #128113)

[Krzysztof Drewniak]

krzysz00 wrote:

Also, for `affine`, I think you want `ValueBoundsOpInterface` and the like, not `IntegerRangeInference`

https://github.com/llvm/llvm-project/pull/128113