Gconf deadlock at startup



Just got a very annoying problem with GConf 1.0.7 but I don't know how
we could fix it cleanly..

-run a GConf-enabled application with a NFSv3 (with lockd running) (in
my example, Nautilus and Evolution with gconf support in gtkhtml)
-turn off the computer (well, I found my computer off without any good
reason)
-turn on and try to launch Evolution (or Nautilus) => GConf isn't able
to launch correctly..

Since there is no error popup nor on console (but only in
/var/log/message), (newbie) user doesn't know what is the problem..

After digging in /var/log/message, I found :
Failed to get lock for daemon, exiting: Failed to lock
'/home/mandrakesoft/fcrozat/.gconfd/lock/ior': another process has the
lock (Resource temporarily unavailable)

Lock is still kept (in gconf-internal(open_empty_locked_file)) even if
there is no longer any gconfd deamon running.. After checking
gconf_get_lock_or_current_holder, it seems we try to "ping" daemon
before starting a new one but in this particular case, we don't even try
to ping it since we can't get a lock..
-- 
Frédéric Crozat
MandrakeSoft




[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]