[llvm] [AMDGPU] CodeGen for GFX12 64-bit scalar add/sub (PR #75070)

[Jay Foad via llvm-commits]

jayfoad wrote:

> > does it need a change in the moveToVALU as well?
> 
> I think the way it works is that we always select S_ADD_U64_PSEUDO, which is already handled in moveToVALU, and only convert to S_ADD_NC_U64 later. But I will check...

Yes that is how it works. I have added a test for this.

https://github.com/llvm/llvm-project/pull/75070