Bug#638316: base: Task blocked for more than 120 seconds
Hi,
Roberto Dal Zilio <roberto.dalzilio@ilvillage.it> writes:
> [904313.840676] updatedb.mloc D ffff880001855c40 0 14983 14979
> 0x00000000
> [904313.840684] ffff880001855c40 0000000000000282 0000000800017880
> ffff8800032f5810
> [904313.840693] 0000000000000000 0000000000000000 000000000000f9e0
> ffff88002cf2dfd8
> [904313.840701] 0000000000015780 0000000000015780 ffff880000017880
> ffff880000017b78
> [904313.840708] Call Trace:
> [904313.840722] [<ffffffff8102cdcc>] ? pvclock_clocksource_read
> +0x3a/0x8b
> [904313.840731] [<ffffffff810b420d>] ? sync_page+0x0/0x46
> [904313.840737] [<ffffffff810b420d>] ? sync_page+0x0/0x46
> [904313.840745] [<ffffffff812fb25a>] ? io_schedule+0x73/0xb7
> [904313.840750] [<ffffffff810b424e>] ? sync_page+0x41/0x46
> [904313.840757] [<ffffffff812fc51a>] ? _spin_unlock_irqrestore+0xd/0xe
> [904313.840762] [<ffffffff812fb66a>] ? __wait_on_bit_lock+0x3f/0x84
> [904313.840767] [<ffffffff810b41da>] ? __lock_page+0x5d/0x63
> [904313.840773] [<ffffffff81064f48>] ? wake_bit_function+0x0/0x23
> [904313.840780] [<ffffffff810ca7e0>] ? lock_page+0x9/0x1f
> [904313.840786] [<ffffffff810ccd1e>] ? handle_mm_fault+0x563/0x80f
> [904313.840794] [<ffffffff8100e242>] ? check_events+0x12/0x20
> [904313.840799] [<ffffffff8100e22f>] ? xen_restore_fl_direct_end
> +0x0/0x1
> [904313.840806] [<ffffffff812fe9f6>] ? do_page_fault+0x2e0/0x2fc
> [904313.840811] [<ffffffff812fc895>] ? page_fault+0x25/0x30
Something is waiting for something here but unfortunately we can't trust
the call trace much since it just contains pointers from stack.
> I can't perform the above commands because I don't have Xen.org but
> Citrix XenServer 5.6.
That's a same. Doesn't XenServer support any kind of crash dumps? About
a year ago I was struggling with a similar "tasks get stuck" bug and was
only able to understand it after I got several crash dumps and could
compare them against the working situation.
-Timo
Reply to: