[clang] [CIR] Implement device-wide nounwind attribute (PR #227016)

[Mariya Podchishchaeva via cfe-commits]

================
@@ -0,0 +1,68 @@
+#include "Inputs/cuda.h"
+
+// REQUIRES: nvptx-registered-target
+// RUN: %clang_cc1 -triple nvptx64-nvidia-cuda -x cuda -fcuda-is-device \
----------------
Fznamznon wrote:

Can we just check AMDGPU in the same file instead of copying it to CodeGenHIP?

https://github.com/llvm/llvm-project/pull/227016