[llvm] [Windows][Arm64EC] Enable thunk generation for bfloat16 (PR #206710)

[David Truby via llvm-commits]

================
@@ -323,6 +323,11 @@ ThunkArgInfo AArch64Arm64ECCallLowering::canonicalizeThunkType(
     return direct(T);
   }
 
+  if (T->isBFloatTy()) {
+    Out << "bf16";
----------------
DavidTruby wrote:

This is a good point; I hadn't really considered that we probably need to coordinate with MS here assuming they're planning on implementing bfloat16 at some point...

https://github.com/llvm/llvm-project/pull/206710