[clang] [WIP][AMDGPU] Support for type inferring image load/store builtins for AMDGPU (PR #140210)

[Matt Arsenault via cfe-commits]

================
@@ -160,6 +163,27 @@ static Value *EmitAMDGCNBallotForExec(CodeGenFunction &CGF, const CallExpr *E,
   return Call;
 }
 
+template <unsigned N>
----------------
arsenm wrote:

This doesn't need to be a template function. You can just check E->getNumArgs for the loop bounds, and use the known constant maximum number of arguments for the small vector size 

https://github.com/llvm/llvm-project/pull/140210