[llvm-branch-commits] [llvm] [AMDGPU] Extract SIAtomicAddrSpace and fence MMRA parsing from SIMemoryLegalizer (PR #231854)

[Fabian Ritter via llvm-branch-commits]

================
@@ -478,4 +480,67 @@ bool isClobberedInFunction(const LoadInst *Load, MemorySSA *MSSA,
   return false;
 }
 
+SIAtomicAddrSpace toSIAtomicAddrSpace(unsigned AS) {
----------------
ritter-x2a wrote:

I've moved the logic into a table and adjusted it to be compatible with `isExtendedGlobalAddrSpace`, to also address @arsenm's comment.

https://github.com/llvm/llvm-project/pull/231854