[llvm] [AMDGPU] Wave Reduce Intrinsics for i32 type (PR #126469)

[via llvm-commits]

easyonaadit wrote:

> We need clang builtins and unit tests to ensure these really work

The long term plan was to replace the CodeGen currently done in AtomicOptimizer, with these intrinsics.
Are the builtins required just for testing purposes, or do we need to expose them to the user?

https://github.com/llvm/llvm-project/pull/126469