[clang] [llvm] [Clang][AMDGPU] Expose buffer load lds as a clang builtin (PR #132048)

[Juan Manuel Martinez Caamaño via llvm-commits]

================
@@ -162,6 +162,8 @@ BUILTIN(__builtin_amdgcn_raw_buffer_load_b64, "V2UiQbiiIi", "n")
 BUILTIN(__builtin_amdgcn_raw_buffer_load_b96, "V3UiQbiiIi", "n")
 BUILTIN(__builtin_amdgcn_raw_buffer_load_b128, "V4UiQbiiIi", "n")
 
+BUILTIN(__builtin_amdgcn_raw_ptr_buffer_load_lds, "vQbv*3IUiiiIiIi", "t")
----------------
jmmartinez wrote:

I've added "lds-buffer-load-insts" in the first commit, and the test in the last commit. I can post the first one as a separate PR if everything is good with this change.

https://github.com/llvm/llvm-project/pull/132048