Bug 85941 - Regression: kernel 3.17 halts sometimes (with Call trace)
Summary: Regression: kernel 3.17 halts sometimes (with Call trace)
Status: RESOLVED OBSOLETE
Alias: None
Product: Other
Classification: Unclassified
Component: Other (show other bugs)
Hardware: x86-64 Linux
: P1 normal
Assignee: Robert Love
URL:
Keywords:
Depends on:
Blocks:
 
Reported: 2014-10-09 03:18 UTC by Dan
Modified: 2014-12-07 07:59 UTC (History)
1 user (show)

See Also:
Kernel Version: 3.17
Subsystem:
Regression: Yes
Bisected commit-id:


Attachments
full log from systemd journalctl, freeze around 11:51 - 11:52 (497.28 KB, text/plain)
2014-10-26 15:48 UTC, Peter Weber
Details
kernel log from sytstemd journalctl, freeze around 11:51 - 11:52 (84.90 KB, text/plain)
2014-10-26 15:49 UTC, Peter Weber
Details

Description Dan 2014-10-09 03:18:50 UTC
I upgraded from kernel 3.16 to 3.17 and sometimes when I watch a video on Youtube or play a audio file, the system freezes (no keyboard, no mouse) and I have to hard reset it. I found it in the log:

Oct  9 00:01:58 tux kernel: [33765.901479] INFO: rcu_preempt detected stalls on CPUs/tasks: { 3} (detected by 0, t=60002 jiffies, g=298657, c=298656, q=0)
Oct  9 00:01:58 tux kernel: [33765.901484] Task dump for CPU 3:
Oct  9 00:01:58 tux kernel: [33765.901485] plugin-containe R  running task        0  9568   1003 0x00080008
Oct  9 00:01:58 tux kernel: [33765.901487]  ffff880063b9ba90 ffffffff8139536f 000000000000a000 0000000000000286
Oct  9 00:01:58 tux kernel: [33765.901489]  ffff880063b9ba78 ffffffff8108aef0 ffff880063b9bab0 00000000ffffffff
Oct  9 00:01:58 tux kernel: [33765.901490]  0000000000000286 ffff880063b9baa0 ffffffff8108b46c 0000000000000286
Oct  9 00:01:58 tux kernel: [33765.901492] Call Trace:
Oct  9 00:01:58 tux kernel: [33765.901497]  [<ffffffff8139536f>] ? __schedule+0x27f/0x7d0
Oct  9 00:01:58 tux kernel: [33765.901500]  [<ffffffff8108aef0>] ? lock_hrtimer_base.isra.31+0x20/0x50
Oct  9 00:01:59 tux kernel: [33765.901502]  [<ffffffff8108b46c>] ? hrtimer_try_to_cancel+0x6c/0x80
Oct  9 00:01:59 tux kernel: [33765.901504]  [<ffffffff8139920b>] ? schedule_hrtimeout_range_clock+0xab/0x140
Oct  9 00:01:59 tux kernel: [33765.901506]  [<ffffffff810d82e9>] ? zone_statistics+0x79/0x90
Oct  9 00:01:59 tux kernel: [33765.901508]  [<ffffffff810d82e9>] ? zone_statistics+0x79/0x90
Oct  9 00:01:59 tux kernel: [33765.901509]  [<ffffffff810d82e9>] ? zone_statistics+0x79/0x90
Oct  9 00:01:59 tux kernel: [33765.901512]  [<ffffffff810c707b>] ? get_page_from_freelist+0x52b/0x900
Oct  9 00:01:59 tux kernel: [33765.901514]  [<ffffffff810c7731>] ? __alloc_pages_nodemask+0xe1/0x950
Oct  9 00:01:59 tux kernel: [33765.901515]  [<ffffffff810cb8a1>] ? release_pages+0xa1/0x1e0
Oct  9 00:01:59 tux kernel: [33765.901518]  [<ffffffff81069f40>] ? __enqueue_entity+0x70/0x80
Oct  9 00:01:59 tux kernel: [33765.901519]  [<ffffffff8106cba5>] ? enqueue_task_fair+0xd25/0xf60
Oct  9 00:01:59 tux kernel: [33765.901521]  [<ffffffff81063362>] ? check_preempt_curr+0x52/0x90
Oct  9 00:01:59 tux kernel: [33765.901522]  [<ffffffff810633b2>] ? ttwu_do_wakeup+0x12/0x90
Oct  9 00:01:59 tux kernel: [33765.901524]  [<ffffffff81063508>] ? ttwu_do_activate.constprop.75+0x58/0x60
Oct  9 00:01:59 tux kernel: [33765.901525]  [<ffffffff81065c9b>] ? try_to_wake_up+0xeb/0x340
Oct  9 00:01:59 tux kernel: [33765.901526]  [<ffffffff81065f3b>] ? wake_up_state+0xb/0x10
Oct  9 00:01:59 tux kernel: [33765.901528]  [<ffffffff8109a77f>] ? wake_futex+0x5f/0x80
Oct  9 00:01:59 tux kernel: [33765.901530]  [<ffffffff8109cf0d>] ? do_futex+0x86d/0xa30
Oct  9 00:01:59 tux kernel: [33765.901531]  [<ffffffff81087413>] ? rcu_eqs_enter+0x93/0xa0
Oct  9 00:01:59 tux kernel: [33765.901533]  [<ffffffff81087ebe>] ? rcu_user_enter+0xe/0x10
Oct  9 00:01:59 tux kernel: [33765.901535]  [<ffffffff810103ed>] ? syscall_trace_leave+0x5d/0x90
Oct  9 00:01:59 tux kernel: [33765.901537]  [<ffffffff8139a47d>] ? int_check_syscall_exit_work+0x34/0x3d

****

If you need more information, just ask. Thanks.
Comment 1 Dan 2014-10-26 02:47:34 UTC
Nobody?
Comment 2 Peter Weber 2014-10-26 14:35:44 UTC
Hi!
I've got at least four or five harde freezes during the last weeks, but wasn't able read anything from the log afterwards (turned off through power-button). Maybe this is a bug. How do you retrieve the log (journalctl? dmesg? ssh-remote?)?
Comment 3 Dan 2014-10-26 14:58:56 UTC
(In reply to Peter Weber from comment #2)
> Hi!
> I've got at least four or five harde freezes during the last weeks, but
> wasn't able read anything from the log afterwards (turned off through
> power-button). Maybe this is a bug. How do you retrieve the log (journalctl?
> dmesg? ssh-remote?)?

Hi Peter. I retrieved from the normal syslog file. Does the output help with something? If you need more information, just ask.
Comment 4 Peter Weber 2014-10-26 15:48:19 UTC
Created attachment 155211 [details]
full log from systemd journalctl, freeze around 11:51 - 11:52
Comment 5 Peter Weber 2014-10-26 15:49:28 UTC
Created attachment 155221 [details]
kernel log from sytstemd journalctl, freeze around 11:51 - 11:52
Comment 6 Peter Weber 2014-10-26 15:54:13 UTC
I've attached two log files, which look different than yours in general. I'm running Archlinux with a self compiled kernel (not much changed in last months, just usual make oldconfig). The last line from the full log, shows that the system is still running, systemd has noticed that I pressed the power button. It is a classic freeze of STDIO/STDOUT, therefore not keyboard input possible and the cursor has stopped blinking.

Maybe your kernel log-level differs (mine should be 4 i.e. the default).
Comment 7 Peter Weber 2014-10-26 15:55:25 UTC
Sorry for flooding. Yes, I've undocked/docked the laptop in my despair :)
Comment 8 Dan 2014-10-26 15:56:55 UTC
(In reply to Peter Weber from comment #6)

> Maybe your kernel log-level differs (mine should be 4 i.e. the default).

Maybe your case is another issue? My log has a "Call trace", but yours don't.
Comment 9 Peter Weber 2014-10-26 16:11:54 UTC
I'm afraid your right. Which wouldn't help either of us :-(
Comment 10 Dan 2014-12-07 07:59:04 UTC
This bug was fixed in recent kernels.

Note You need to log in before you can comment on or make changes to this bug.