[clang] [llvm] [AMDGPU] Implement AsyncMark Stages (PR #220442)

[Ryan Mitchell via cfe-commits]

RyanRio wrote:

> Supporting an asyncmark instruction that does not explicitly specify the types of asyncmarks it is pointing to (whether it is TENSOR , ASYNC or whatever it is) seems unnecessarily ambiguous, I don't see the benefit from doing so. Instead if we had to specify the types of asyncmarks explicitly (which we don't currently do) it would be easier to associate the asyncwaits to the corresponding asyncmarks.

I'd appreciate an example to ensure I'm understanding your point correctly, I can think of a few different interpretations of this

https://github.com/llvm/llvm-project/pull/220442