Re: [PATCH v4 00/32] ldisc patchset

From: Peter Hurley
Date: Sat Feb 23 2013 - 13:44:07 EST


On Sat, 2013-02-23 at 10:24 -0500, Sasha Levin wrote:
> On 02/22/2013 01:37 PM, Peter Hurley wrote:
> > On Thu, 2013-02-21 at 08:38 -0500, Peter Hurley wrote:
> >> On Thu, 2013-02-21 at 08:16 -0500, Sasha Levin wrote:
> >>> On 02/20/2013 03:02 PM, Peter Hurley wrote:
> >>>> Sasha and Dave, my trinity testbeds die in other areas right now;
> >>>> I would really appreciate if you would please re-test this series.
> >>>
> >>> Hi Peter,
> >>>
> >>> I saw this twice in overnight fuzzing:
> >>>
> >>> [ 1473.912280] =================================
> >>> [ 1473.913180] [ BUG: bad contention detected! ]
> >>> [ 1473.914071] 3.8.0-next-20130220-sasha-00038-g1ad55df-dirty #8 Tainted: G W
> >>> [ 1473.915684] ---------------------------------
> >>> [ 1473.916549] kworker/1:1/361 is trying to contend lock (&tty->ldisc_sem) at:
> >>> [ 1473.918031] [<ffffffff81c493df>] tty_ldisc_ref+0x1f/0x60
> >>> [ 1473.919060] but there are no locks held!
> >>
> >> Ahh, of course. That explains why the rwsem trylock doesn't track lock
> >> stats -- because by the time lock_contended() is called, up_write()
> >> could have just called lockdep_release(), so that it appears as if the
> >> lock has been released when in fact it has not but is about to.
> >>
> >> I'll just remove the lock contention test from the trylocks.
> >
> > Hi Sasha,
> >
> > Sorry for the delay. I was actually looking into if I could tickle
> > lockdep into just recording the lock contention without testing, but
> > unfortunately, changes to where lockdep stores the contention now
> > requires the lockdep state to have an existing owner.
> >
> > So here's the trivial patch:
>
> Hi Peter,
>
> After more fuzzing, I'm seeing this sort of hangs (which are new):
>
> [ 2644.723879] INFO: task trinity:17893 blocked for more than 120 seconds.
> [ 2644.727112] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
> [ 2644.731916] trinity D ffff8800a9c904a8 5192 17893 8043 0x00000000
> [ 2644.733517] ffff88006efb3a78 0000000000000002 ffff8800aa0c3b10 ffff8800bb3d7180
> [ 2644.739350] ffff880019103000 ffff880097a78000 ffff88006efb3a78 00000000001d7180
> [ 2644.741459] ffff880097a78000 ffff88006efb3fd8 00000000001d7180 00000000001d7180
> [ 2644.746590] Call Trace:
> [ 2644.747177] [<ffffffff83db9909>] __schedule+0x2e9/0x3b0
> [ 2644.748294] [<ffffffff83db9b35>] schedule+0x55/0x60
> [ 2644.752382] [<ffffffff83db9e83>] schedule_preempt_disabled+0x13/0x20
> [ 2644.753737] [<ffffffff83db7fdd>] __mutex_lock_common+0x34d/0x560
> [ 2644.759037] [<ffffffff81c40893>] ? ptmx_open+0x83/0x190
> [ 2644.760590] [<ffffffff83db84b5>] ? __mutex_unlock_slowpath+0x185/0x1e0
> [ 2644.762064] [<ffffffff81c40893>] ? ptmx_open+0x83/0x190
> [ 2644.768967] [<ffffffff83db831f>] mutex_lock_nested+0x3f/0x50
> [ 2644.770314] [<ffffffff81c40893>] ptmx_open+0x83/0x190
> [ 2644.771413] [<ffffffff812917ae>] chrdev_open+0x11e/0x190
> [ 2644.780456] [<ffffffff81291690>] ? cdev_put+0x30/0x30
> [ 2644.781421] [<ffffffff8128af59>] do_dentry_open+0x1f9/0x310
> [ 2644.785550] [<ffffffff8128b0bc>] finish_open+0x4c/0x70
> [ 2644.786724] [<ffffffff8129c3eb>] do_last+0x61b/0x810
> [ 2644.787676] [<ffffffff8129c699>] path_openat+0xb9/0x4d0
> [ 2644.791868] [<ffffffff812ac278>] ? __alloc_fd+0x1e8/0x200
> [ 2644.792817] [<ffffffff81185214>] ? lock_release_nested+0xb4/0xf0
> [ 2644.794010] [<ffffffff81185331>] ? __lock_release+0xe1/0x100
> [ 2644.797401] [<ffffffff8129cebd>] do_filp_open+0x3d/0xa0
> [ 2644.798467] [<ffffffff812ac278>] ? __alloc_fd+0x1e8/0x200
> [ 2644.799577] [<ffffffff8128c51b>] do_sys_open+0x12b/0x1d0
> [ 2644.804667] [<ffffffff8128c5dc>] sys_open+0x1c/0x20
> [ 2644.805542] [<ffffffff83dc49d8>] tracesys+0xe1/0xe6
> [ 2644.822807] 1 lock held by trinity/17893:
> [ 2644.823685] #0: (tty_mutex){+.+.+.}, at: [<ffffffff81c40893>] ptmx_open+0x83/0x190
>
> The mutex is 'tty_mutex' at drivers/tty/pty.c:701 .
>
> I didn't grab sysrq-t this time since it was an overnight run, but I'll
> try to grab one when it happens again.

Hi Sasha,

Can you please 'make drivers/tty/pty.lst' for this kernel config and
paste ptmx_open() here?

This report makes no sense: this stack trace shows this task waiting on
a mutex that is not owned.


--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/