Hi Ross, Yes, the patch is undergoing some changes and currently under (hopefully) final review. Regards, Ming On Mon, Oct 3, 2016 at 5:23 PM, Ross Beer <ross.beer@xxxxxxxxxxx> wrote: > Hi PJSIP team, > > > Could this patch be considered for inclusion in the project? > > > This issue is causing segfaults a least once a day. > > > Kind regards, > > > Ross > > > > ________________________________ > From: pjsip <pjsip-bounces@xxxxxxxxxxxxxxx> on behalf of Alexei Gradinari > <alex2grad@xxxxxxxxx> > Sent: 21 September 2016 16:38 > To: pjsip list > Subject: patch: crash on using already destroyed ssl socket > > Hello, > > On heavy loaded system with TLS, > one thread could destroy the ssl socket on SSL_ERROR_SYSCALL > while another thread still uses this socket which > was already freed, so we get segfault. > Attached 2 backtraces. > > To avoid race condition need to lock the socket before destroying it. > > The attached patch adds the socket lock on destroying it > and adds a checking on all openssl calls that the socket wasn't destroyed. > > Regards, > Alexei > > _______________________________________________ > Visit our blog: http://blog.pjsip.org > > pjsip mailing list > pjsip@xxxxxxxxxxxxxxx > http://lists.pjsip.org/mailman/listinfo/pjsip_lists.pjsip.org > _______________________________________________ Visit our blog: http://blog.pjsip.org pjsip mailing list pjsip@xxxxxxxxxxxxxxx http://lists.pjsip.org/mailman/listinfo/pjsip_lists.pjsip.org
