[Chapel Merge] Fix a compiler error about not using printf format with LLVM-22

Branch: refs/heads/main
Revision: d0e1e8f26116645f0b24ac3a1a6450e8f45394de
Author: dlongnecke-cray
Link: Fix a compiler error about not using printf format with LLVM-22 by dlongnecke-cray · Pull Request #28825 · chapel-lang/chapel · GitHub
Log Message:
Fix a compiler error about not using printf format with LLVM-22 (#28825)

A header in the runtime used by the GPU locale model has to define a
format attribute under the LLVM-22 clang. Do so by copying a macro used
in another similar GPU header.

I tested by doing a make check on a CHPL_GPU=cpu config.

Reviewed by @jabraham17. Thanks!

Compare: Comparing 24ef4eb3eac2f4a2ec1daa4e5c2974a20adb35be...04195d3593faed9deef40db7c7078a6a86772a25 · chapel-lang/chapel · GitHub

Diff:
M runtime/include/gpu/cpu/chpl-gpu-gen-includes.h
https://github.com/chapel-lang/chapel/pull/28825.diff