Kerneloops
James Westby
james.westby at ubuntu.com
Mon Jul 6 22:57:02 UTC 2009
manoj.iyer at canonical.com wrote:
>
> James,
>
> I tested kerneloops & apport for Jaunty, it filed a bug at
> kernelopps.org and also filed a bug in Launchpad. IMHO tagging the
> launchpad bug such that it indicates that the bug was reported to
> kerneloops.org will be a good thing.
That's certainly possible.
Also I noticed that when the crash
> or oops happned again, apport will file the same bug again (or ask the
> user if they want to file a bug). This was a but annoying. Is there a
> mechanism to detect that this bug was already filed and avoid duplicate
> bugs being filed from the same machine ?
That's something I will have to look at.
> Is there any mechanism by which I can save the apport data from one
> machine and file the bug on kerneloops from a different machine ? This
> is specially useful in server setup that does not have network
> connection to the outside world.
This is one thing that integration with apport would bring, you could
just copy the .crash file to a machine that can talk to launchpad and
use apport-cli or similar to submit it from there.
Amit Kucheria wrote:
> As Manoj already pointed out, tagging the LP bug as reported upstream
> (via kerneloops) would be helpful. Is it possible to file the same bug
> automatically in bugzilla.kernel.org and LP and link them?
If the kernel bugzilla enables the LP plugin then my understanding is
that you can push bugs easily to it from LP. I'm reluctant to send
bugs there automatically though. I thought one point of kerneloops.org
was to avoid having bugs for all of these issues in bugzilla?
> Atleast for the kernel, the duplicate finder is relatively useless
> because bugs are so HW-dependent. I am fairly certain that we prefer
> people to file separate bugs that _we_ mark as duplicate.
That's good to know thanks. I'll not spend time on dupe-detection across
users then.
It sounds like you are happy for kerneloops reports to be filed as LP
bugs as well by default? If so then I will proceed with the
implementation.
Thanks,
James
More information about the kernel-team
mailing list