Bug 8181 - suspend/resume hangs until keypress
suspend/resume hangs until keypress
Status: CLOSED CODE_FIX
Product: Power Management
Classification: Unclassified
Component: Other
i386 Linux
: P2 normal
Assigned To: power-management_other
:
Depends on:
Blocks:
  Show dependency treegraph
 
Reported: 2007-03-12 04:28 UTC by Tomas Janousek
Modified: 2007-03-17 07:55 UTC (History)
1 user (show)

See Also:
Kernel Version: 2.6.29-rc3-git20070311
Tree: Mainline
Regression: ---


Attachments
dmesg before suspend (23.86 KB, application/octet-stream)
2007-03-12 04:29 UTC, Tomas Janousek
Details
kernel config (54.02 KB, application/octet-stream)
2007-03-12 04:30 UTC, Tomas Janousek
Details
lspci/cpuinfo/lsmod (6.54 KB, application/octet-stream)
2007-03-12 04:34 UTC, Tomas Janousek
Details

Description Tomas Janousek 2007-03-12 04:28:08 UTC
Most recent kernel where this bug did *NOT* occur: 2.6.20
Distribution: debian testing/unstable
Hardware Environment: HP Compaq nx7300 (generally the same as nx7400)
Software Environment:
Problem Description:
I use hibernate and uswsusp. When I try to suspend to disk, it gets stuck on
'Suspending consoles' and does not continue until I press like 3 keys. It does
the same on resume.
Comment 1 Tomas Janousek 2007-03-12 04:29:19 UTC
Created attachment 10712 [details]
dmesg before suspend
Comment 2 Tomas Janousek 2007-03-12 04:30:31 UTC
Created attachment 10713 [details]
kernel config
Comment 3 Tomas Janousek 2007-03-12 04:34:05 UTC
Created attachment 10714 [details]
lspci/cpuinfo/lsmod
Comment 4 Milan Broz 2007-03-13 10:14:30 UTC
Seems the same as in thread here (I see it on Thinkpad X60s)
http://lkml.org/lkml/2007/3/11/112

2.6.21-rc3 with NO_HZ or CONFIG_TICK_ONESHOT, 

suspend trace:
enable_nonboot_cpus
 _cpu_up
  raw_notifier_callchain (CPU_UP_PREPARE)
    ...
    update_sched_domains
     detach_destroy_domains
       [waits here] --> synchronize_sched (==synchronize_rcu)
(some details in thread above)

Maybe related to
http://lkml.org/lkml/2007/3/13/3

Comment 5 Milan Broz 2007-03-17 02:31:14 UTC
Fixed for me by the patch here http://lkml.org/lkml/2007/3/16/406
Comment 6 Tomas Janousek 2007-03-17 07:55:19 UTC
For me as well.

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