[llvm] AMDGPU: Add a subtarget feature for fine-grained remote memory support (PR #96442)

[Matt Arsenault via llvm-commits]

================
@@ -788,6 +788,16 @@ def FeatureFlatAtomicFaddF32Inst
   "Has flat_atomic_add_f32 instruction"
 >;
 
+def FeatureAgentScopeFineGrainedRemoteMemoryAtomics
----------------
arsenm wrote:

I originally went with device and changed it to agent to match the syncscope spelling. This was the HSA name 

https://github.com/llvm/llvm-project/pull/96442