[clang] [Clang] Correctly enable the f16 type for offloading (PR #98331)

[Joseph Huber via cfe-commits]

================
@@ -0,0 +1,117 @@
+// REQUIRES: nvptx-registered-target
+//
+// RUN: %clang_cc1 -ffp-contract=off -triple nvptx-unknown-unknown -target-cpu \
+// RUN:   sm_86 -target-feature +ptx72 -fcuda-is-device -x cuda -emit-llvm -o - %s \
----------------
jhuber6 wrote:

They're probably required since the NVPTX intrinsics have requirements on sm and ptx versions.

https://github.com/llvm/llvm-project/pull/98331