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

[Krzysztof Drewniak]

krzysz00 wrote:

(Re the lattice, you need to have `DeadCodeAnalysis` loaded)

But also look into `ValueBoundsConstraintSet` - it works better with `affine`

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