Bug 66311 - [HSW ULT] Loop to test S4, system will call trace and network disconnect
Summary: [HSW ULT] Loop to test S4, system will call trace and network disconnect
Status: CLOSED MOVED
Alias: None
Product: Power Management
Classification: Unclassified
Component: Hibernation/Suspend (show other bugs)
Hardware: All Linux
: P1 normal
Assignee: Aaron Lu
URL:
Keywords:
Depends on:
Blocks:
 
Reported: 2013-12-02 06:18 UTC by Qingshuai Tian
Modified: 2014-03-12 03:00 UTC (History)
5 users (show)

See Also:
Kernel Version: 3.12.0rc3
Tree: Mainline
Regression: No


Attachments
Image of the monitor that shows Calltrace (1.43 MB, image/jpeg)
2013-12-02 06:20 UTC, Qingshuai Tian
Details
dmesg of latest nightly kernel(1be8f2 ) after s4 call trace (72.47 KB, text/plain)
2014-02-17 07:33 UTC, Qingshuai Tian
Details
dmesg after forty rounds of s4 with kmemleak=off set (118.95 KB, text/plain)
2014-02-19 03:29 UTC, Qingshuai Tian
Details

Description Qingshuai Tian 2013-12-02 06:18:24 UTC
Environment:
-------------------
Kernel: (drm-intel-next-queued)9d1cb9147dbe45f6e94dc796518ecf67cb64b359
Author: Paulo Zanoni <paulo.r.zanoni@intel.com>
Date:   Fri Nov 1 13:32:08 2013 -0200
         drm/i915: avoid unclaimed registers when capturing the error state

Description:
--------------------
When I test s4 in a loop , the machine will resume with call trace and the network disconnect, and sometimes it will also hang at the same time. It may happen at the 5 or 6 round of the loop ,or more like 24 round . I add the picture of the call trace in the attachment.

This machine is a new ultrabook.
Machin Info: Lenovo idealpad u330p
BIOS Version: 7CCN15WW
CPU: Intel(R) Core(TM) i3-4010U CPU @1.70GHZ

When I looped testing s4 with the "i915.modeset=0" module parameter setted, the machine will still call trace and hang at about round 21 of the loop.
Reproduce step:
----------------------
1.boot up machine 
2.echo 0 > /sys/class/rtc/rtc0/wakealarm ; 
  echo +30 > /sys/class/rtc/rtc0/wakealarm; 
3.echo disk > /sys/power/state
4.loop running step 2 and 3
Comment 1 Qingshuai Tian 2013-12-02 06:20:37 UTC
Created attachment 117111 [details]
Image of the monitor that shows Calltrace

This is the image of the monitor that shows the Calltrace.
Comment 2 Lan Tianyu 2013-12-16 07:29:42 UTC
hi, Could you check this bug on the v3.13-rc4?
Comment 3 Qingshuai Tian 2013-12-26 05:11:58 UTC
(In reply to Lan Tianyu from comment #2)
> hi, Could you check this bug on the v3.13-rc4?
When looped to test s4 with latest v3.13-rc4 kernel, the machine still would hang sporadically.
Comment 4 Lan Tianyu 2013-12-26 07:27:22 UTC
Could you follow the file to debug loop s4?

http://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git/tree/Documentation/power/basic-pm-debugging.txt
Comment 5 Aaron Lu 2014-01-27 08:16:56 UTC
ping, what's the status of this bug on newest upstream kernel?
Comment 6 Qingshuai Tian 2014-02-17 07:33:08 UTC
Created attachment 126371 [details]
dmesg of latest nightly kernel(1be8f2 ) after s4  call trace

With the lateset drm-intel-nightly kernel(1be8f2 ),I repeated s4 in a loop, the machine still will call trace or hang sporadically.
I caught the dmesg with call trace and added it in the attachment.
Comment 7 Aaron Lu 2014-02-18 06:34:23 UTC
The trace is about kmemleak detector... Please add kmemleak=off to kernel cmdline and see if you still have the issue.
Comment 8 Qingshuai Tian 2014-02-19 03:29:20 UTC
Created attachment 126721 [details]
dmesg after forty rounds of s4 with kmemleak=off set


I finished forty rounds of s4 with kmemleak=off set,and there was no Call trace or hang appeared again. I added the dmesg in the attachment. You can check it .
Comment 9 Aaron Lu 2014-03-12 02:59:51 UTC
Feel free to report this to kmemleak developer(I'm not sure who he is and if this is expected), I'll close this bug here.

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