Consequently, whenever a SIGQUIT arrives (e.g. bos restart) it has a chance of catching it. If it does so while holding a lock such as the vol_glock_mutex just because a 'vos something' was going on resulting in e.g. VOffline_r being called, it'll call ShutDown() itself. Doing that it'll need the vol_glock_mutex again when doing VPrintCacheStats() and therefore block on the mutex held by itself.
Attached is a little patch to vol/fssync.c fixing this.
--
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
Rainer Toebbicke http://cern.ch/~rtb [EMAIL PROTECTED] O__
European Laboratory for Particle Physics(CERN) - Geneva, Switzerland > |
Phone: +41 22 767 8985 Fax: +41 22 767 7155 ( )\( )
p_mask-signals-in-FSYNC
Description: application/java-vm
