lkml.org 
[lkml]   [1997]   [Apr]   [4]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
From
SubjectAbout reporting Ooopses (besides using ksymoops.cc)
Date
Bryn Paul Arnold Jones Wrote:

> I would have been nicer if you looked those addresses up with ksymoops
> (source in linux/scripts/ksymoops.cc, read for compiling info, dosn't need
> the -liostream here).
>
> Run like this: ksymoops /whereever/System.map < /whereever/oops.txt, and
> repost the oops, and the output. You'll need to stript the 'kernel: 's
> off it too.

About reporting oopses. I sent a properly formatted one a few days ago (with
all symbols resolved thanks to klog, and with the last asm instructions
before the oops, gotten as described above.

Besides posting it to the list when the bug didn't seem to be linked to an
alpha driver you're testing, what else can you do (in my case no one seemed
to be inspired by it, most likely because the info I supplied, even if as
complete and accurate as possible, didn't give much info on what could
possibly be wrong, and you don't start re-reading the whole network code
from line one just because someone posted a bug report of a kernel that
oopsed somewhere in the network code (more accurately [dev_close+52/184],
[dev_open+64/92], and [register_netdevice_notifier+26/52])....).
Of course, as I expected, I can't reproduce the bug (yet), and it's not
really an issue for me, but is there anything else I should have done, or
anything else I should do now?

(The answer would should be useful for the FAQ draft)

Thanks,
Marc

\
 
 \ /
  Last update: 2005-03-22 13:39    [from the cache]
©2003-2014 Jasper Spaans. Advertise on this site