Reply to this email directly, view it on GitHub, or unsubscribe.
You are receiving this because you are subscribed to this thread.
That is the default one used if none is provided.
I think, the default value should be a USE_SYSV_SEM as in Makefiles.defs:
https://github.com/kamailio/kamailio/blob/6b0b8cb84b7d0e965d50bdf0dfa5423004879f55/src/Makefile.defs#L2156-L2160
If FreeBSD requires some other we can overwrite the value in the
freebsd.cmaketo have a more reasonable one.
Do you think it's need to explicit set LOCK_METHOD to USE_PTHREAD_MUTEX (example) in the freebsd.cmake and also duplicate it as target_compile_definitions in os-specific, isn't it?
https://github.com/kamailio/kamailio/blob/6b0b8cb84b7d0e965d50bdf0dfa5423004879f55/cmake/os-specific/darwin.cmake#L25-L27
For example for freebsd.cmake:
message(STATUS "USE_FAST_LOCK = ${USE_FAST_LOCK}")
if(NOT ${USE_FAST_LOCK})
target_compile_definitions(common INTERFACE USE_PTHREAD_MUTEX)
endif()
set(LOCK_METHOD USE_PTHREAD_MUTEX)
I think it seems ugly.
Also MacOS doesn't tested with futexlock.h earlier too.
I think freebsd just uses fastlock if executed on a standard architecture
Yes. But I said, the explicit USE_FUTEX as a default isn't correct, in my mind.
So, for you to build you have removed the link against
resolv
Yep, I've built with removed link only.
—
Reply to this email directly, view it on GitHub, or unsubscribe.
You are receiving this because you are subscribed to this thread.