Bug 70166 - possible recursive locking detected
Summary: possible recursive locking detected
Status: RESOLVED NOTOURBUG
Alias: None
Product: PulseAudio
Classification: Unclassified
Component: misc (show other bugs)
Version: unspecified
Hardware: Other All
: medium normal
Assignee: pulseaudio-bugs
QA Contact: pulseaudio-bugs
URL:
Whiteboard:
Keywords:
Depends on:
Blocks:
 
Reported: 2013-10-05 13:56 UTC by almos
Modified: 2013-10-28 11:00 UTC (History)
1 user (show)

See Also:
i915 platform:
i915 features:


Attachments

Description almos 2013-10-05 13:56:23 UTC
syslog:

Oct  5 15:48:53 fxa kernel: [  199.766955]
Oct  5 15:48:53 fxa kernel: [  199.766959] =============================================
Oct  5 15:48:53 fxa kernel: [  199.766961] [ INFO: possible recursive locking detected ]
Oct  5 15:48:53 fxa kernel: [  199.766963] 3.10.13-fxa #1 Not tainted
Oct  5 15:48:53 fxa kernel: [  199.766964] ---------------------------------------------
Oct  5 15:48:53 fxa kernel: [  199.766966] pulseaudio/5108 is trying to acquire lock:
Oct  5 15:48:53 fxa kernel: [  199.766967]  (&(&substream->self_group.lock)->rlock/1){......}, at: [<ffffffffa02a
d57a>] snd_pcm_action_group+0x64/0x1cd [snd_pcm]
Oct  5 15:48:53 fxa kernel: [  199.766978]
Oct  5 15:48:53 fxa kernel: [  199.766978] but task is already holding lock:
Oct  5 15:48:53 fxa kernel: [  199.766980]  (&(&substream->self_group.lock)->rlock/1){......}, at: [<ffffffffa02a
d57a>] snd_pcm_action_group+0x64/0x1cd [snd_pcm]
Oct  5 15:48:53 fxa kernel: [  199.766986]
Oct  5 15:48:53 fxa kernel: [  199.766986] other info that might help us debug this:
Oct  5 15:48:53 fxa kernel: [  199.766988]  Possible unsafe locking scenario:
Oct  5 15:48:53 fxa kernel: [  199.766988]
Oct  5 15:48:53 fxa kernel: [  199.766989]        CPU0
Oct  5 15:48:53 fxa kernel: [  199.766990]        ----
Oct  5 15:48:53 fxa kernel: [  199.766991]   lock(&(&substream->self_group.lock)->rlock/1);
Oct  5 15:48:53 fxa kernel: [  199.766993]   lock(&(&substream->self_group.lock)->rlock/1);
Oct  5 15:48:53 fxa kernel: [  199.766994]
Oct  5 15:48:53 fxa kernel: [  199.766994]  *** DEADLOCK ***
Oct  5 15:48:53 fxa kernel: [  199.766994]
Oct  5 15:48:53 fxa kernel: [  199.766996]  May be due to missing lock nesting notation
Oct  5 15:48:53 fxa kernel: [  199.766996]
Oct  5 15:48:53 fxa kernel: [  199.766998] 4 locks held by pulseaudio/5108:
Oct  5 15:48:53 fxa kernel: [  199.766998]  #0:  (snd_pcm_link_rwlock){......}, at: [<ffffffffa02acf46>] snd_pcm_
stream_lock_irq+0x14/0x23 [snd_pcm]
Oct  5 15:48:53 fxa kernel: [  199.767004]  #1:  (&(&substream->self_group.lock)->rlock){......}, at: [<ffffffffa
02acf52>] snd_pcm_stream_lock_irq+0x20/0x23 [snd_pcm]
Oct  5 15:48:53 fxa kernel: [  199.767009]  #2:  (&(&substream->group->lock)->rlock){......}, at: [<ffffffffa02ae
386>] snd_pcm_action+0x2b/0x85 [snd_pcm]
Oct  5 15:48:53 fxa kernel: [  199.767015]  #3:  (&(&substream->self_group.lock)->rlock/1){......}, at: [<fffffff
fa02ad57a>] snd_pcm_action_group+0x64/0x1cd [snd_pcm]
Oct  5 15:48:53 fxa kernel: [  199.767021]
Oct  5 15:48:53 fxa kernel: [  199.767021] stack backtrace:
Oct  5 15:48:53 fxa kernel: [  199.767023] CPU: 3 PID: 5108 Comm: pulseaudio Not tainted 3.10.13-fxa #1
Oct  5 15:48:53 fxa kernel: [  199.767024] Hardware name: Gigabyte Technology Co., Ltd. GA-990FXA-UD3/GA-990FXA-U
D3, BIOS F5 10/13/2011
Oct  5 15:48:53 fxa kernel: [  199.767026]  0000000000000000 ffff88041a27dc00 ffffffff81360159 ffff88041a27dca8
Oct  5 15:48:53 fxa kernel: [  199.767029]  ffffffff810789f6 ffff880400000001 ffff880400000000 ffffffff81c2b720
Oct  5 15:48:53 fxa kernel: [  199.767032]  0000000000000002 49724c12809424a1 ffff88041a248788 ffff88041a27dcf0
Oct  5 15:48:53 fxa kernel: [  199.767034] Call Trace:
Oct  5 15:48:53 fxa kernel: [  199.767039]  [<ffffffff81360159>] dump_stack+0x19/0x1b
Oct  5 15:48:53 fxa kernel: [  199.767043]  [<ffffffff810789f6>] __lock_acquire+0xbff/0xd92
Oct  5 15:48:53 fxa kernel: [  199.767045]  [<ffffffff8107817e>] ? __lock_acquire+0x387/0xd92
Oct  5 15:48:53 fxa kernel: [  199.767047]  [<ffffffff810783dd>] ? __lock_acquire+0x5e6/0xd92
Oct  5 15:48:53 fxa kernel: [  199.767049]  [<ffffffff81078ea6>] lock_acquire+0xa3/0x11a
Oct  5 15:48:53 fxa kernel: [  199.767054]  [<ffffffffa02ad57a>] ? snd_pcm_action_group+0x64/0x1cd [snd_pcm]
Oct  5 15:48:53 fxa kernel: [  199.767057]  [<ffffffff813636b5>] _raw_spin_lock_nested+0x36/0x42
Oct  5 15:48:53 fxa kernel: [  199.767062]  [<ffffffffa02ad57a>] ? snd_pcm_action_group+0x64/0x1cd [snd_pcm]
Oct  5 15:48:53 fxa kernel: [  199.767066]  [<ffffffffa02ad57a>] snd_pcm_action_group+0x64/0x1cd [snd_pcm]
Oct  5 15:48:53 fxa kernel: [  199.767071]  [<ffffffffa02ae3bc>] snd_pcm_action+0x61/0x85 [snd_pcm]
Oct  5 15:48:53 fxa kernel: [  199.767075]  [<ffffffffa02ae3f5>] snd_pcm_stop+0x15/0x17 [snd_pcm]
Oct  5 15:48:53 fxa kernel: [  199.767080]  [<ffffffffa02af1c4>] snd_pcm_drop+0x69/0x7f [snd_pcm]
Oct  5 15:48:53 fxa kernel: [  199.767084]  [<ffffffffa02b08a8>] snd_pcm_common_ioctl1+0x87e/0xad1 [snd_pcm]
Oct  5 15:48:53 fxa kernel: [  199.767086]  [<ffffffff8107817e>] ? __lock_acquire+0x387/0xd92
Oct  5 15:48:53 fxa kernel: [  199.767091]  [<ffffffffa02b0d0e>] snd_pcm_playback_ioctl1+0x213/0x230 [snd_pcm]
Oct  5 15:48:53 fxa kernel: [  199.767096]  [<ffffffffa02b0d55>] snd_pcm_playback_ioctl+0x2a/0x2e [snd_pcm]
Oct  5 15:48:53 fxa kernel: [  199.767099]  [<ffffffff81118465>] vfs_ioctl+0x18/0x34
Oct  5 15:48:53 fxa kernel: [  199.767101]  [<ffffffff81118cb2>] do_vfs_ioctl+0x38d/0x44b
Oct  5 15:48:53 fxa kernel: [  199.767104]  [<ffffffff81120777>] ? rcu_read_unlock+0x1c/0x2d
Oct  5 15:48:53 fxa kernel: [  199.767107]  [<ffffffff81118dc2>] SyS_ioctl+0x52/0x7d
Oct  5 15:48:53 fxa kernel: [  199.767109]  [<ffffffff8111b92a>] ? dput+0x3f/0x114
Oct  5 15:48:53 fxa kernel: [  199.767112]  [<ffffffff813695d6>] system_call_fastpath+0x1a/0x1f
Oct  5 15:48:53 fxa pulseaudio[5108]: [pulseaudio] sink.c: Default and alternate sample rates are the same.
Comment 1 Tanu Kaskinen 2013-10-28 10:57:07 UTC
I believe this is a locking issue in the kernel, not in PulseAudio, so I'm closing this bug. If you have a reason to believe otherwise, please reopen the bug.
Comment 2 Tanu Kaskinen 2013-10-28 11:00:02 UTC
Oh, and the reason I think this is a kernel bug is that the log message identifies the problematic lock as follows: (&(&substream->self_group.lock)->rlock/1)

That refers to an in-kernel lock, not any lock that is directly used by PulseAudio.


Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.