Oops in sys_getsockname

[Matt Galvin]

Binary package hint: dbus

Current feisty all up to date with -9 kernel and all. machine boots i can log into gnome but the desktop never loads it hangs. after dropping into recovery mode (nothing will work right in regular boot mode b/c of this crash) and doing some poking around I found it seems to have to do with dbus or avahi.

When I run:

/etc/init.d/dbus start

this is the output:

--- snip ---
 * Starting system message bus dbus
   ...done.
 * Starting Hardware abstraction layer hald
   ...done.
 * Starting DHCP D-Bus daemon dhcdbd
   ...done.
 * Starting network connection manager NetworkManager
   ...done.
 * Starting Avahi mDNS/DNS-SD Daemon: avahi-daemon
Timeout reached while wating for return value
Could not receive return value from daemon process.
   ...fail!
 * Starting network events dispatcher NetworkManagerDispatcher
   ...done.
 * Starting System Tools Backends system-tools-backends
   ...done.
--- snip---

in there it crashes hard. The whole system acts weird after that. After the crash sudo will not even work, it just hangs on every command i try to run. I could not seem to get the crash logged to a file nor could i find it in any of the logs anywhere so I am attaching a screen shot (camera) of what i could catch of the message. Seems to be one problem after another with feisty on this machine *sigh*. Please let me know if there is any other info I can provide. currently feisty has been and is still unusable on my desktop machine.

[Matt Galvin (mgalvin)]

thats on i386... gonna try amd64 now

[Matt Galvin (mgalvin)]

same deal on amd64

[Mark J. Bobak (markbobak)]

Confirmed. Same problem here on amd64.

[Scott Robinson (scott-ubuntu)]

See if you can shift-pageup and get crash information from higher up in the log.

[Matt Galvin (mgalvin)]

Shift + PageUp does work and I was able to collect screen shots of the entire crash. It appears to be NetworkManager causing the crash.

[Matt Galvin (mgalvin)]

(sorry about all the images)

[Paul Dufresne (paulduf)]

That's an Oop 10 (1010) bit 2 is 0 indicating it happened in kernel
according to: http://en.wikipedia.org/wiki/Linux_kernel_oops
So moving to linux-source 2.6.20 (pretty sure I have seen it in one of
the picture)

I don't know much about oops, but according to
http://sosdg.org/~coywolf/lxr/source/Documentation/oops-tracing.txt?v=2.6.16
in section 'Where is the oops' you could find interesting info on how to get the information without having to take a picture of each screen.

But anyway, I think you did a pretty good job at it!

Also, confirming on Mark J. Bobak's comment.

[Paul Dufresne (paulduf)]

Probably a duplicate of bug #78255 which is in progress.
I have asked there to verify that.

[Paul Dufresne (paulduf)]

Well, doing in a console:
dmesg > Matt_Galvin_dmesg.txt
and then attaching it to this bug would probably make it easier for people to
study your report.

[Kyle McMartin (kyle)]

This is indeed a duplicate of #78255. A recent upload of module-init-tools should enable you to boot and start Gnome by preventing these modules from loading. Unfortunately, it means your wireless does not work, but I'm trying to fix that.

Thanks for your patience,
  Kyle