ondemand cpu governor prevents resume from RAM on X60s

Bug #84455 reported by Gregory Oschwald
14
Affects Status Importance Assigned to Milestone
linux-meta (Ubuntu)
Fix Released
Undecided
Unassigned

Bug Description

I am experiencing the following bug on my Thinkpad X60s:

http://bugzilla.kernel.org/show_bug.cgi?id=6166

If I used the ondemand cpu frequency governor, my X60s will sometimes not resume from a suspend to RAM. This most frequently occurs when I have switched power states after suspending but before resuming. This bug has been bothering me for months and it was quite hard to track down. I chanced upon the above bug report, and this is exactly what I am experience. This happens on the 2.6.15, 2.6.17, and 2.6.19 Ubuntu kernels that I have tried. It continued to happen after a fresh reinstall of Edgy. Using the userspace governor seems to fix the problem.

Tags: kernel-oops
Revision history for this message
Gregory Oschwald (osch0001) wrote :
Download full text (3.3 KiB)

My speculated cause might be wrong, as it seems to not be resuming consistently again. This happens most often when I switch power states while the machine is suspend to RAM (e.g., detach the power cord and resume). If it doesn't resume properly, I can get it to resume by plugging it in and unplugging it until it resumes (usually after doing this 2 or 3 times). If I just leave the machine when it is in this state, it will not resume by itself.

Occasionally after forcing a resume by the power cycling method, I get messages like the following in an open terminal:

Message from syslogd@thinkpad at Sun Feb 11 10:36:53 2007 ...
thinkpad kernel: [17257086.340000] Oops: 0000 [#1]

Message from syslogd@thinkpad at Sun Feb 11 10:36:53 2007 ...
thinkpad kernel: [17257086.340000] SMP

Message from syslogd@thinkpad at Sun Feb 11 10:36:53 2007 ...
thinkpad kernel: [17257086.340000] CPU: 0

Message from syslogd@thinkpad at Sun Feb 11 10:36:53 2007 ...
thinkpad kernel: [17257086.340000] EIP is at 0x0

Message from syslogd@thinkpad at Sun Feb 11 10:36:53 2007 ...
thinkpad kernel: [17257086.340000] eax: dffdeb30 ebx: c2138b64 ecx: 00000000 edx: 00000081

Message from syslogd@thinkpad at Sun Feb 11 10:36:53 2007 ...
thinkpad kernel: [17257086.340000] esi: c8849d1c edi: dff94c40 ebp: f2e95d40 esp: dff07f40

Message from syslogd@thinkpad at Sun Feb 11 10:36:53 2007 ...
thinkpad kernel: [17257086.340000] ds: 007b es: 007b ss: 0068

Message from syslogd@thinkpad at Sun Feb 11 10:36:53 2007 ...
thinkpad kernel: [17257086.340000] Process kacpi_notify (pid: 17, threadinfo=dff06000 task=dff93030)

Message from syslogd@thinkpad at Sun Feb 11 10:36:53 2007 ...
thinkpad kernel: [17257086.340000] Stack: c0202612 f2e95d40 f2e95d4c c01fcc8d dff94c40 f2e95d48 c0132702 00000000

Message from syslogd@thinkpad at Sun Feb 11 10:36:53 2007 ...
thinkpad kernel: [17257086.340000] dff06000 00000000 dff94c4c dff94c60 00000292 c01fcc6f dff06000 dff94c4c

Message from syslogd@thinkpad at Sun Feb 11 10:36:53 2007 ...
thinkpad kernel: [17257086.340000] dff94c54 dff94c40 c01332e7 00000001 00000000 00000000 00010000 00000000

Message from syslogd@thinkpad at Sun Feb 11 10:36:53 2007 ...
thinkpad kernel: [17257086.340000] Call Trace:

Message from syslogd@thinkpad at Sun Feb 11 10:36:53 2007 ...
thinkpad kernel: [17257086.340000] <c0202612> acpi_ev_notify_dispatch+0x4c/0x55 <c01fcc8d> acpi_os_execute_deferred+0x1e/0x27

Message from syslogd@thinkpad at Sun Feb 11 10:36:53 2007 ...
thinkpad kernel: [17257086.340000] <c0132702> run_workqueue+0x72/0xf0 <c01fcc6f> acpi_os_execute_deferred+0x0/0x27

Message from syslogd@thinkpad at Sun Feb 11 10:36:53 2007 ...
thinkpad kernel: [17257086.340000] <c01332e7> worker_thread+0x117/0x140 <c011bde0> default_wake_function+0x0/0x10

Message from syslogd@thinkpad at Sun Feb 11 10:36:53 2007 ...
thinkpad kernel: [17257086.340000] <c01331d0> worker_thread+0x0/0x140 <c0135f8b> kthread+0xab/0xe0

Message from syslogd@thinkpad at Sun Feb 11 10:36:53 2007 ...
thinkpad kernel: [17257086.340000] <c0135ee0> kthread+0x0/0xe0 <c0101005> kernel_thread_helper+0x5/0x10

Message from syslogd@thinkpad at Sun Feb 11 10:36:...

Read more...

Revision history for this message
Gregory Oschwald (osch0001) wrote :

I don't know if this is related at all, but sometime after coming back from resume, cpu0 is using the userspace governor and cpu1 is using the performance governor. I haven't yet established if this is correlated with the above problems when coming out of resume.

Revision history for this message
Gregory Oschwald (osch0001) wrote :

The last comment is when forcing powernowd to run, as I have been testing the machine without using the "ondemand" governor per my original post.

Revision history for this message
Gregory Oschwald (osch0001) wrote :

I am closing this as I no longer experience the problem on Gutsy.

Changed in linux-meta:
status: New → Fix Released
Revision history for this message
Gregory Oschwald (osch0001) wrote :

Note: I meant on _Hardy_. I did experience the problem on Gutsy.

Revision history for this message
Muelli (ubuntu-bugs-auftrags-killer) wrote :
Revision history for this message
Muelli (ubuntu-bugs-auftrags-killer) wrote :

uh, Launchpad FAIL. I have no idea how this attachment has landed here. It's supposed to be to my just reported bug. I'm trying to attach it to the appropriate bug. Sorry for the noise.

To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Duplicates of this bug

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.