[Mlir-commits] [mlir] [mlir][gpu] Add innermost-first policy when mapping loops to GPU IDs (PR #160634)

[Georgios Pinitas]

GeorgeARM wrote:

> The `scf.parallel` semantics already say it shouldn't have any side-effects:
> 
> ```
>     Semantically we require that the iteration space can be iterated in any
>     order, and the loop body can be executed in parallel. If there are data
>     races, the behavior is undefined.
> ```
> 
> If a reordering produced an undesired side-effect, then the op is already in violation of its assumptions.

Sure you are right on the semantics of `scf.parallel` when it comes to legality on interchange but I don't think it should be the default solution here. I will add the `TODO` and revisit this.
Thank you @fabianmcg 🙇 

https://github.com/llvm/llvm-project/pull/160634