[Mlir-commits] [mlir] [mlir][scf] Tighten description of scf.for (PR #203713)

[Reinhard Stahn]

rainij wrote:

Ready for review: @rengolin , @ftynse , @joker-eph 

Related to my discourse post: https://discourse.llvm.org/t/question-on-the-semantics-of-scf-for-around-integer-overflow-and-why-canonicalize-changes-it/91048

I might also open a similar PR on `affine.for` in the future as suggested in the post, but I decided to keep this PR small. I already wrestled a lot with the description of `scf.for` alone. I hope it is in good shape now.

https://github.com/llvm/llvm-project/pull/203713