On 02.06.17 15:41, Aurélien Larcher <aurelien.larc...@gmail.com> wrote: > > > > > > > On Fri, Jun 2, 2017 at 3:26 PM, Carsten Grzemba <grze...@contac-dt.de> wrote: > > > > > > > On 28.05.17 13:11, Aurélien Larcher <aurelien.larc...@gmail.com> wrote: > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > BTW: What about publishing these packages with exiv2 enabled? > > > > > > > > > > > > > > > > > Speaking of which, there is a new version of exiv2 so that could be a > > > good opportunity to look at it. > > > > > > > > > > > > > > > > > > > > 0.26 uses pthread PTHREAD_RECURSIVE_MUTEX_INITIALIZER_NP which is not > > contained in our pthread.h :( > > > > > Is it really necessary? > > > I guess in /usr/include/synch.h is a valid replacement:
RECURSIVE_ERRORCHECKMUTEX But this definition raise a gcc error: actions.cpp:75:120: error: braces around scalar initializer for type ‘uint16_t {aka short unsigned int}’ {{0, 0, 0, {USYNC_THREAD|LOCK_RECURSIVE|LOCK_ERRORCHECK}, MUTEX_MAGIC}, {{{0, 0, 0, 0, 0, 0, 0, 0}}}, 0} so these macros in synch.h should not work in general with g++-4.9 ? and this definition without braces compiles instead: {{0, 0, 0, USYNC_THREAD|LOCK_RECURSIVE|LOCK_ERRORCHECK, MUTEX_MAGIC}, {{{0, 0, 0, 0, 0, 0, 0, 0}}}, 0} Any hints?
_______________________________________________ oi-dev mailing list oi-dev@openindiana.org https://openindiana.org/mailman/listinfo/oi-dev