[clang] [llvm] [AMDGPU] Implement 'llvm.get.fpenv' and 'llvm.set.fpenv' (PR #83906)

[Joseph Huber via cfe-commits]

================
@@ -325,6 +325,9 @@ BUILTIN(__builtin_amdgcn_read_exec_hi, "Ui", "nc")
 
 BUILTIN(__builtin_amdgcn_endpgm, "v", "nr")
 
+BUILTIN(__builtin_amdgcn_get_fpenv, "WUi", "n")
----------------
jhuber6 wrote:

There's no builtin as far as I'm aware. I think there might be some pragmas however. Because the intrinsic takes different kinds I figured it was easiest to just specify my use-case. The alternative would be to make this generic and have a LUT by target triple in `CGBuiltin` that just contains the one entry for AMDGPU.

https://github.com/llvm/llvm-project/pull/83906