[llvm] [AMDGPU] Form V_MAD_U64_U32 from mul24 (PR #72393)

[Pierre van Houtryve via llvm-commits]

Pierre-vh wrote:

@jayfoad I did the refactor you suggested, seems to be looking better indeed.

However I still haven't done the SpeedModel part because:
- (Apologies but) I'm still confused. Do we need it on those new patterns (we only form the V_MAD_I64_I32 if on FullSpeed) or in #72983 to emit a i64 mul instead of i24 muls on non-FullSpeed models?
- We don't have patterns like that currently so I'm wondering if there is a good reason behind it? (cc @arsenm)


https://github.com/llvm/llvm-project/pull/72393