Created attachment 136937 [details]
syslog excerpt.

[lilydev@haring lilypond]$ uname -a
Linux haring 2.6.17-1.2647.fc6 #1 SMP Wed Sep 13 12:51:50 EDT 2006 i686 i686
i386 GNU/Linux

From the syslog:

Sep 22 12:22:35 localhost kernel: Modules linked in: rfcomm hidp l2cap hci_usb
bluetooth button ipw3945(U) ieee80211(U) ieee80211_crypt(U) i915 drm sunrpc
snd_hda_intel snd_hda_codec cpufreq_ondemand nls_utf8 loop vfat fat video sbs
ibm_acpi i2c_ec dock battery asus_acpi ac ipv6 parport_pc lp parport
snd_seq_dummy snd_seq_oss snd_seq_midi_event snd_seq snd_seq_device snd_pcm_oss
joydev snd_mixer_oss snd_pcm sg snd_timer snd soundcore e1000 i2c_i801
snd_page_alloc pcspkr i2c_core serio_raw ide_cd cdrom dm_snapshot dm_zero
dm_mirror dm_mod ahci libata sd_mod scsi_mod ext3 jbd ehci_hcd ohci_hcd uhci_hcd
Sep 22 12:22:35 localhost kernel: EIP:    0060:[<f898bf57>]    Tainted: GF     VLI

Why the forced modules?  Can you recreate this without force-loaded modules?  I
know you said not reproducible but seeing force-loaded modules and a subsequent
pointer error makes me a little suspicious.

Thanks,
-Eric

oh, silly me, you're right.  This is with a 3rd party driver (intel 3945 wifi) -
so I guess it's due to that.  I completely forgot that I had it loaded at that
point (I almost never use wifi.)

does that make it a WONTFIX bug? 
 
 

Well, I hate to WONTFIX it but when there are unknowns like that in the kernel,
and -you- can't reproduce it, and it doesn't match other signatures that I've
seen, then there's not a lot to be done.  I think CANTFIX is appropriate, I'm
afraid.

I'll keep an eye out for any similar problems w/o a tainted kernel, and re-open
or dup this bug to anything that comes up later.  If you hit it again w/o those
module, feel free to re-open.

Thanks,

-Eric

Re-opening so I can dup it and relate it to the other eerily similar bug.

Duping to 208488 - even though it came second we've got more information there.

*** This bug has been marked as a duplicate of 208488 ***