From Bugzilla Helper: User-Agent: Mozilla/4.72 [en] (X11; U; Linux 2.2.18-dsp1 i686) Description of problem: In the first Gnome login session in which I run Nautilus (or "Start here"), it works fine, and I can start it as many times as I want. However, if I then logout and back in again, nautilus no longer comes up. If I run "nautilus" in a terminal window, it gives the error message "Gdk-WARNING **: locale not supported by Xlib, local set to C" (this message appears to be normal), and the command just sits there and no windows come up (I can press Ctrl-C to get back to the command line). After a logout, "/usr/bin/gconfd-1" and sometimes "nautilus" processes are still running by the user in question. It appears to be necessary to reboot the computer in order to run nautilus again, regardless of whether I clean out the /tmp directory and/or kill any leftover processes such as gconfd or nautilus. Version-Release number of selected component (if applicable): How reproducible: Always Steps to Reproduce: 1. (Re)boot the computer. 2. Log into Gnome. 3. Click on the "Start here" icon. Close the window and repeat as many times as desired. Nautilus comes up correctly each time. 4. Log out of Gnome. 5. Log back into Gnome. 6. Click on "Start here". Nautilus doesn't come up at all, but it should have. Additional info:
Some things to try: - be sure nautilus is in ~/.gnome/session - try changing your locale to something definitely valid like en_US to avoid the Xlib warning, that isn't normal (but I think it's a fixed bug, just Xlib not understanding iso885915 locales) - "strace -o output -f nautilus" from a terminal, wait for nautilus to hang, hit control-C, send along the file "output" thanks!
Created attachment 55085 [details] Requested systrace output
Created attachment 55086 [details] nautilus is not anywhere in ~/.gnome/session
"LANG=en_US nautilus &" made the Xlib warning go away. (LANG was set to "en_US.iso885915".)
It still hangs when the warning goes away, right.
What version of gconf? Does "gconftool-1 --shutdown" help any while it's stuck? Does gconf-sanity-check-1 report any issues? The lack of nautilus in the session and the hang are perhaps different issues...
Yes, with or without the Xlib warning, nautilus hangs on the select() call near the end of the strace file, and I pressed Ctrl-C to kill it. (I didn't start it in the background.) "rpm -qf /usr/bin/gconf-config" reports package "GConf-devel-1.0.9-3". "rpm -qf /usr/bin/gconftool-1" reports package "GConf-1.0.9-3". "gconftool-1 --shutdown" gets stuck the same way nautilus does. [dpaschal@paschal6 dpaschal]$ gconf-sanity-check-1 Please contact your system administrator to resolve the following problem: Failed to get a file lock: Failed to lock '/home/dpaschal/.gconfd/lock/ior': probably another process has the lock, or your operating system has NFS file locking misconfigured, or a hard NFS client crash caused a stale lock (Resource temporarily unavailable) - run gconf-sanity-check-1 for possible diagnosis, see http://www.gnome.org/projects/gconf/ for more information [dpaschal@paschal6 dpaschal]$ ls -slap /home/dpaschal/.gconfd/lock/ior 4 -rwx------ 1 dpaschal dpaschal 265 Apr 15 17:42 /home/dpaschal/.gconfd/lock/ior [dpaschal@paschal6 dpaschal]$ /usr/sbin/lsof |grep gconfd gconfd-1 2105 dpaschal cwd DIR 3,2 4096 2 / gconfd-1 2105 dpaschal rtd DIR 3,2 4096 2 / gconfd-1 2105 dpaschal txt REG 3,2 53073 16246 /usr/bin/gconfd-1 gconfd-1 2105 dpaschal mem REG 3,2 90327 241009 /lib/ld-2.2.5.so gconfd-1 2105 dpaschal mem REG 3,2 42514 353550 /usr/lib/GConf/1/libgconfbackend-xml.so gconfd-1 2105 dpaschal mem REG 3,2 265551 96676 /usr/lib/libgconf-1.so.1.0.4 gconfd-1 2105 dpaschal mem REG 3,2 123100 96573 /usr/lib/liboaf.so.0.0.0 gconfd-1 2105 dpaschal mem REG 3,2 29376 96464 /usr/lib/libpopt.so.0.0.0 gconfd-1 2105 dpaschal mem REG 3,2 41594 96504 /usr/lib/libORBitCosNaming.so.0.5.13 gconfd-1 2105 dpaschal mem REG 3,2 288277 96502 /usr/lib/libORBit.so.0.5.13 gconfd-1 2105 dpaschal mem REG 3,2 64549 96500 /usr/lib/libIIOP.so.0.5.13 gconfd-1 2105 dpaschal mem REG 3,2 7043 96506 /usr/lib/libORBitutil.so.0.5.13 gconfd-1 2105 dpaschal mem REG 3,2 89424 241026 /lib/libnsl-2.2.5.so gconfd-1 2105 dpaschal mem REG 3,2 173487 241005 /lib/i686/libm-2.2.5.so gconfd-1 2105 dpaschal mem REG 3,2 11931 96453 /usr/lib/libgmodule-1.2.so.0.0.10 gconfd-1 2105 dpaschal mem REG 3,2 182363 96451 /usr/lib/libglib-1.2.so.0.0.10 gconfd-1 2105 dpaschal mem REG 3,2 12102 241022 /lib/libdl-2.2.5.so gconfd-1 2105 dpaschal mem REG 3,2 45415 241042 /lib/libnss_files-2.2.5.so gconfd-1 2105 dpaschal mem REG 3,2 499990 96496 /usr/lib/libxml.so.1.8.17 gconfd-1 2105 dpaschal mem REG 3,2 59778 96445 /usr/lib/libz.so.1.1.3 gconfd-1 2105 dpaschal mem REG 3,2 1401050 241003 /lib/i686/libc-2.2.5.so gconfd-1 2105 dpaschal 0u CHR 1,3 66338 /dev/null gconfd-1 2105 dpaschal 1u CHR 1,3 66338 /dev/null gconfd-1 2105 dpaschal 2u CHR 1,3 66338 /dev/null gconfd-1 2105 dpaschal 3u CHR 1,3 66338 /dev/null gconfd-1 2105 dpaschal 4u unix 0xc20cd0e0 10187 socket gconfd-1 2105 dpaschal 5u unix 0xc0b6ba60 10188 /tmp/orbit-dpaschal/orb-18907697101992247766 gconfd-1 2105 dpaschal 6w REG 3,2 265 164848 /home/dpaschal/.gconfd/lock/0t1018917774ut195181u500p2105r546168450k3221149756 (deleted) gconfd-1 2105 dpaschal 7w REG 3,2 265 116222 /home/dpaschal/.gconf/%gconf-xml-backend.lock/1t1018917774ut236861u500p2105r137935288k3221149372 (deleted) gconfd-1 2105 dpaschal 8r REG 3,2 876 421381 /home/dpaschal/.gconfd/saved_state.orig (deleted) gconfd-1 2105 dpaschal 9r FIFO 0,6 10198 pipe gconfd-1 2105 dpaschal 10w FIFO 0,6 10198 pipe gconfd-1 2105 dpaschal 11u unix 0xc0b6b060 10199 /tmp/orbit-dpaschal/orb-18907697101992247766 gconfd-1 2105 dpaschal 13u unix 0xc03ed0a0 10219 socket
Created attachment 55124 [details] output of "strace -o output -f gconftool-1 --shutdown" (hangs on last select)
Essentially it looks like gconfd isn't responding. Tracing gconfd could be interesting. Also the syslog stuff as described on http://www.gnome.org/projects/gconf/
This appears to be the same bug as 62920. In any case the resolution suggested by alexl in 62920 (Try upgrading to xinetd 2.3.4-0.8 from rawhide. The shut down portmap and xinetd and then start them again.) Works.
Hmm, bug should have been closed a good long time ago it seems.