[clang] [llvm] [RFC][WIP][AMDGPU] Use `bf16` instead of `i16` for bfloat (PR #80908)

[Matt Arsenault via cfe-commits]

================
@@ -2835,8 +2835,8 @@ def int_amdgcn_fdot2_f32_bf16 :
   DefaultAttrsIntrinsic<
     [llvm_float_ty], // %r
     [
-      llvm_v2i16_ty, // %a
-      llvm_v2i16_ty, // %b
+      llvm_v2bf16_ty, // %a
+      llvm_v2bf16_ty, // %b
----------------
arsenm wrote:

For potential revert pain minimization, I would do each intrinsic in a separate commit 

https://github.com/llvm/llvm-project/pull/80908