You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
The idea is that downstream users of libCEED can/should inherit the compilation flags of upstream as they should respond similarly to the same optimization flags.
To do that, we could add a cflags_extra key to ceed.pc to store then, similar to how PETSc does it. And in that we'd just store the OPT variable from the build process.
The text was updated successfully, but these errors were encountered:
I'm fine with providing this key, though it may be tricky to record the right one. For example, when PETSc --download-libceed, it sets CFLAGS directly, not via the OPT convenience (whose sole purpose is to make it easy to give optimization flags for C, C++, and Fortran in one shot (only works if they have compatible vendors). https://gitlab.com/petsc/petsc/-/blob/main/config/BuildSystem/config/packages/libceed.py?ref_type=heads#L34
Good point. I'm fine with it being CFLAGS instead of OPT. The thought behind OPT is that it just trims down on the intake of flags. CFLAGS seems like the better one for robustness sake.
See HONEE thread for background.
The idea is that downstream users of libCEED can/should inherit the compilation flags of upstream as they should respond similarly to the same optimization flags.
To do that, we could add a
cflags_extra
key toceed.pc
to store then, similar to how PETSc does it. And in that we'd just store theOPT
variable from the build process.The text was updated successfully, but these errors were encountered: