Hi, Stipe Tolj wrote:
> Alexander Malysh wrote: > >> Hi, >> >> Stipe Tolj wrote: >> >> >>>Michael R. Pierotti wrote: >>> >>>>Checking for POSIX threads support ... >>>>checking for working pthreads... no >>>>checking for pthread_spinlock support... no >>>>checking for pthread_rwlock support... no >>> >>>this is obviously the reason for it. configure is unable to find the >>>pthread library/headers on your system. And since the #include >>><pthread.h> inside gwlib/gw-rwlock.h does not complain about none >>>existence, but the pthread_rwlock_t is not defined, there seems to be >>>things messed up on your system. >>> >>>Or the pthread library does not support the required read/write >>>locking. So, don't blame Kannel here ;) >>> >>>@Alex: any investigatory comments from your side? >> >> >> not really... all I can say: upgrade your glibc! it's very very old... >> >> Q: should configure fail if no pthread_rwlock supported detected? > > now, if we use it inside the make build process in an assumptive way > that it would be present, then yes, configure should fail then, IMO. heh, instead of just failing in configure if no rwlock support detected, I have implemented "ugly" emulation code that uses List... This emulation code has overhead but works in all cases where rwlock used in kannel. Please get a latest cvs version a try... > > Stipe > > mailto:stolj_{at}_wapme.de > ------------------------------------------------------------------- > Wapme Systems AG > > Vogelsanger Weg 80 > 40470 D�sseldorf, NRW, Germany > > phone: +49.211.74845.0 > fax: +49.211.74845.299 > > mailto:info_{at}_wapme-systems.de > http://www.wapme-systems.de/ > ------------------------------------------------------------------- > -- Thanks, Alex
