From: Michael Widenius Date: March 30 1999 8:54am Subject: Re: error while loading UDF List-Archive: http://lists.mysql.com/mysql/1179 Message-Id: <14080.36954.509463.790890@monty.pp.sci.fi> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit >>>>> "Curt" == Curt W Zirzow writes: Curt> On 29 Mar 1999 05:34:20 -0800, Curt> Michael Widenius wrote: >> >> Did you upgrade FreeBSD version between 3.22.16 gamma and Curt> 3.22.19b ? Curt> Yes, 3.22.16 gamma is running on FreeBSD 3.0-RELEASE. Curt> 3.22.19b is running on FreeBSD 3.1-RELEASE Curt> Also, the new version we have has been compiled --with-mit- Curt> threads=yes; while the 3.22.16 gamma was not, if this makes any Curt> difference. Yes, that should explain the problem. Curt> A way I found to fix this problem is to compile my .so as so... Curt> c++ -shared -DMYSQL_SERVER udf_relevant.cc ../mit- Curt> pthreads/obj/libpthread.a -o udf_relevant.so Curt> Would this be a reasonable aproach to fixing my problem? I must confess I don't understand why udf_relevant.so must be linked with libpthread.a; The pthread_init() call already exists in the mysqld core and udf_relevant.so should be able to find it there! Do you have any pthread calls in the udf_relevant.cc file? (The only reason I can think off is that udf_relevant.cc uses mit-pthreads and mysqld doesn't, but according to your statement, this shouldn't be the case) Curt> Thanks much... Curt> Curt Zirzow Regards, Monty