[libclc] libclc: Use frexp builtins to implement frexp for amdgpu (PR #185637)

[Matt Arsenault via cfe-commits]

================
@@ -0,0 +1,47 @@
+//===----------------------------------------------------------------------===//
----------------
arsenm wrote:

It's still a template inc file, it's not a header. I think this is the right way to do it, the generic code provides a library of possible implementation choices and the target just includes the one it wants from the available options 

https://github.com/llvm/llvm-project/pull/185637