[clang] [llvm] [AMDGPU] Introduce asyncmark/wait intrinsics (PR #173259)

[Nicolai Hähnle via cfe-commits]

https://github.com/nhaehnle commented:

Mostly LGTM.

I suspect there's an infinite loop due to how `mergeAsyncMarkers` works when the IR contains a loop that has an `asyncmark` intrinsic but no `wait.asyncmark`. What prevents the AsyncMarkers array from just growing indefinitely?

It would generally be nice to have some additional tests where the asyncmarkers in control flow aren't balanced. BTW, I think it'd be fine to have tests that are a bit more bare-bone, e.g. always using the same pointer arguments for the load.lds. That would result in less "noise" in the  assembly.

https://github.com/llvm/llvm-project/pull/173259