[clang] [llvm] [AMDGPU[ Expand `av` built-ins and intrinsics to b8/b16/b32/b64 (PR #229429)

[Pierre van Houtryve via cfe-commits]

Pierre-vh wrote:

@pabloantoniom There seems to be a bug with i64 atomic stores in the DAG. I switched to using <2 x i32> like other b64 atomics instead

https://github.com/llvm/llvm-project/pull/229429