[lkml]   [2008]   [Jan]   [9]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
Messages in this thread
SubjectRe: Top 10 kernel oopses for the week ending January 5th, 2008
On Tuesday January 8, wrote:
> FWIW, I'm going to go through Arjan's collection and post blow-by-blow
> analysis of some of those suckers. Tonight, probably...
> Let's take e.g.

Thanks for that analysis.
> Humm... So we have kobj->parent containing crap. What about the caller?
> It's from drivers/md/md.c:
> static void delayed_delete(struct work_struct *ws)

This is a good argument for sticking "md_" at the from of all my
function names, even if they are static. I'm fairly sure I looked at
that trace:

> Call Trace:
> [<ffffffff803b37e9>] kobject_put+0x19/0x20
> [<ffffffff803b389b>] kobject_del+0x2b/0x40
> [<ffffffff804d7d50>] delayed_delete+0x0/0xb0
> [<ffffffff804d7db9>] delayed_delete+0x69/0xb0
> [<ffffffff80249775>] run_workqueue+0x175/0x210
> [<ffffffff8024a411>] worker_thread+0x71/0xb0
> [<ffffffff8024d9e0>] autoremove_wake_function+0x0/0x40
> [<ffffffff8024a3a0>] worker_thread+0x0/0xb0
> [<ffffffff8024d5fd>] kthread+0x4d/0x80
> [<ffffffff8020c4b8>] child_rip+0xa/0x12
> [<ffffffff8020bbcf>] restore_args+0x0/0x30
> [<ffffffff8024d5b0>] kthread+0x0/0x80
> [<ffffffff8020c4ae>] child_rip+0x0/0x12

but as it doesn't mention 'md' or 'nfs' I moved on. My bad.

> What guarantees that it doesn't happen before we get to callback? AFAICS,
> nothing whatsoever...

Yes, that's bad isn't it :-)

I think I should be using sysfs_schedule_callback here. That makes the
required 'get' and 'put' calls.... but it can fail with -ENOMEM. I
wonder what I do if -ENOMEM??? Maybe I'll just continue to roll my
one :-(


 \ /
  Last update: 2008-01-10 05:17    [W:0.113 / U:2.780 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site