weirdness going from runlevel 5 to 3

Allen Kistler an037-ooai8 at yahoo.com
Sat Apr 11 21:16:50 UTC 2009


Robert P. J. Day wrote:
> On Sat, 11 Apr 2009, François Cami wrote:
>> "Robert P. J. Day" <rpjday crashcourse ca> wrote:
>>>
>>>   for debugging purposes (hi, francois!), i wanted to drop out of my
>>> GNOME desktop to the command line in runlevel 3.  the first time i
>>> tried that ("init 3"), the whole display went dark and i couldn't find
>>> anything on any virtual console.  eventually, i powered down and up
>>> again just to get my system back.
>>>
>>>   i just tried "init 3" again from the desktop and my GNOME session is
>>> still chugging along happily even though:
>>>
>>> # runlevel
>>> 5 3
>>> #
>>
>> What happens if you :
>>
>> * add the DontZap option to your xorg.conf :
>>
>> Section "ServerFlags"
>>     Option "DontZap" "false"
>> EndSection
> 
>   i lied, i had time to test this, and it seems to solve the problem.
> "init 3" now takes me properly to runlevel 3, where i can log in as
> root and "init 5" gets me back to GNOME so we're good.  and on that
> note, i'm out of here for the day.  thanks for all the assistance.
> 
> rday
> -- 
> 
> p.s.  just to summarize all of the X stuff since yesterday, this is my
> xorg.conf at the moment on my gateway m-1626:
> 
>   Section "ServerFlags"
>         Option "DontZap" "false"
>   EndSection
> 
>   Section "Device"
>     Identifier "Videocard0"
>     Driver "radeon"
>   EndSection
> 
> and i have "nomodeset" in my grub.conf.

The only thing that DontZap gets you is the ability to use alt-ctrl-bksp 
to kill X, which was the second part of François' suggestion.

I think I saw this same trouble as failure to shut down or reboot 
yesterday.  I thought I saw the rc0 script getting killed before the 
machine had shutdown.  That is, the process that's supposed to kill all 
the other processes got killed.  At least that's how I interpret the 
following log:

Apr 10 14:52:03 localhost console-kit-daemon[2303]: 
GLib-GObject-WARNING: IA__g_object_get_valist: value location for 
`gchararray' passed as NULL
Apr 10 14:52:03 localhost console-kit-daemon[2303]: 
GLib-GObject-WARNING: IA__g_object_get_valist: value location for 
`gchararray' passed as NULL
Apr 10 14:52:03 localhost console-kit-daemon[2303]: 
GLib-GObject-WARNING: IA__g_object_get_valist: value location for 
`gchararray' passed as NULL
Apr 10 14:52:03 localhost console-kit-daemon[2303]: 
GLib-GObject-WARNING: IA__g_object_get_valist: value location for 
`gchararray' passed as NULL
Apr 10 14:52:03 localhost init: tty4 main process (2533) killed by TERM 
signal
Apr 10 14:52:03 localhost init: tty5 main process (2534) killed by TERM 
signal
Apr 10 14:52:03 localhost init: tty2 main process (2535) killed by TERM 
signal
Apr 10 14:52:03 localhost init: tty3 main process (2536) killed by TERM 
signal
Apr 10 14:52:03 localhost init: tty6 main process (2537) killed by TERM 
signal
Apr 10 14:52:03 localhost init: rc0 main process (3287) killed by TERM 
signal
Apr 10 14:52:03 localhost kernel: [drm] Resetting GPU
Apr 10 14:52:03 localhost init: rc0 main process (3298) killed by TERM 
signal
Apr 10 14:52:03 localhost smartd[2529]: smartd received signal 15: 
Terminated
Apr 10 14:52:03 localhost smartd[2529]: smartd is exiting (exit status 0)

... then shutdown stops, but the machine stays up.  Today I got:

Apr 11 15:57:26 localhost console-kit-daemon[2295]: 
GLib-GObject-WARNING: IA__g_object_get_valist: value location for 
`gchararray' passed as NULL
Apr 11 15:57:26 localhost console-kit-daemon[2295]: 
GLib-GObject-WARNING: IA__g_object_get_valist: value location for 
`gchararray' passed as NULL
Apr 11 15:57:26 localhost console-kit-daemon[2295]: 
GLib-GObject-WARNING: IA__g_object_get_valist: value location for 
`gchararray' passed as NULL
Apr 11 15:57:26 localhost console-kit-daemon[2295]: 
GLib-GObject-WARNING: IA__g_object_get_valist: value location for 
`gchararray' passed as NULL
Apr 11 15:57:26 localhost console-kit-daemon[2295]: 
GLib-GObject-WARNING: IA__g_object_get_valist: value location for 
`gchararray' passed as NULL
Apr 11 15:57:26 localhost gdm-simple-slave[2582]: GLib-GObject-CRITICAL: 
g_object_ref: assertion `G_IS_OBJECT (object)' failed
Apr 11 15:57:26 localhost init: tty4 main process (2524) killed by TERM 
signal
Apr 11 15:57:26 localhost init: tty5 main process (2527) killed by TERM 
signal
Apr 11 15:57:26 localhost init: tty2 main process (2536) killed by TERM 
signal
Apr 11 15:57:26 localhost init: tty3 main process (2540) killed by TERM 
signal
Apr 11 15:57:26 localhost init: tty6 main process (2541) killed by TERM 
signal
Apr 11 15:57:26 localhost init: rc6 main process (3648) killed by TERM 
signal
Apr 11 15:57:26 localhost init: rc6 main process (3656) killed by TERM 
signal
Apr 11 15:57:26 localhost gdm-simple-slave[2582]: GLib-GObject-WARNING: 
invalid unclassed pointer in cast to `GdmSession'
Apr 11 15:57:26 localhost gdm-simple-slave[2582]: CRITICAL: 
_gdm_session_conversation_stopped: assertion `GDM_IS_SESSION (session)' 
failed
Apr 11 15:57:26 localhost gdm[3666]: ******************* START 
**********************************
[+++ most of traceback omitted for brevity +++]
Apr 11 15:57:27 localhost gdm[3666]: The program is running.  Quit 
anyway (and detach it)? (y or n) [answered Y; input not from terminal]
Apr 11 15:57:27 localhost gdm[3666]: ******************* END 
**********************************

Leaving me with a blank desktop and a mouse, but nothing else, which is 
mostly the effect I saw yesterday.  The power button is set up to shut 
down, so that's how I can ultimately shut down.

I'll look for filed bugs and file one myself if I don't see any.




More information about the fedora-test-list mailing list