Re: 2.6.24 regression: pan hanging unkilleable and un-straceable

[Nick Piggin <nickpiggin@yahoo.com.au>]

On Sunday 27 January 2008 01:27, Pascal Terjan wrote:
> Nick Piggin <nickpiggin <at> yahoo.com.au> writes:
> > On Sunday 27 January 2008 00:29, Frederik Himpe wrote:
> > > I just succeeded to reproduce the problem with this patch. Does this
> > > smell like an XFS problem?
>
> I got the same issue using ext3
>
> > Possible. Though I think it is more likely to be a bug in the
> > new deadlock avoidance code in the generic buffered write path.
> > Dang... I wonder why this hasn't come up earlier. It looks like
> > pan's use of writev might be tickling it.
> >
> > How quickly can you reproduce this?
>
> When I was using pan daily one month ago, I got it twice over a week
>
> > Can you use strace to see what the hanging syscall looks like?
>
> I tried last week during 5 hours without luck, I can try again

Dang, I didn't see any reports of this earlier :(

Anyway, I sent a patch to fix it in the original thread (can you
reply-to-all please? just it a bit easier to keep threads together)

Thanks,
Nick