Skip to contentSkip to conversation

2.6.29-rc1: thinkpad problems during resume

DISCUSSION Pavel Machek 8 participants 26 messages Jan 13, 2009 – Jan 26, 2009

lore
Share & tools
lore
#0 lore raw

Hi!

Just when I thought long merge window means less problems...

...suspend seemed to work until last tuesday or so. Now it hanged two times; once so hard that power button powercycle did not result in booting machine. (thinkpad x60)

I'll do some more tests...
Pavel

Signature · 2 lines
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html
#1 lore raw
Quoted 9 lines · Pavel Machek

Hi!

Just when I thought long merge window means less problems...

...suspend seemed to work until last tuesday or so. Now it hanged two times; once so hard that power button powercycle did not result in booting machine. (thinkpad x60)

I'll do some more tests...

There are a couple of things you may want to revert, like 7503bfbae89eba07b46441a5d1594647f6b8ab7d and a0d4922da2e4ccb0973095d8d29f36f6b1b5f703 .

Please let me know if that helped.

Thanks,
Rafael

#2 lore raw

Am Dienstag 13 Januar 2009 schrieb Rafael J. Wysocki:

Quoted 9 lines · Pavel Machek
Quoted 8 lines

Hi!

Just when I thought long merge window means less problems...

...suspend seemed to work until last tuesday or so. Now it hanged two times; once so hard that power button powercycle did not result in booting machine. (thinkpad x60)

Quoted 7 lines
Quoted 1 line

I'll do some more tests...

There are a couple of things you may want to revert, like 7503bfbae89eba07b46441a5d1594647f6b8ab7d and a0d4922da2e4ccb0973095d8d29f36f6b1b5f703 .

Please let me know if that helped.

I tested rc1+ these two reverts yesterday on my T61p. Doesnt work. There must be at least one more patch that broke S2R on my thinkpad. In fact, it is cpu hotplug that is broken (unplug works).

I will try to bisect that as well, but currently I am busy trying to find a virtio regression....

Christian

#15 lore raw
In reply to Rafael J. Wysocki
Quoted 10 lines · Pavel Machek
Quoted 8 lines

Hi!

Just when I thought long merge window means less problems...

...suspend seemed to work until last tuesday or so. Now it hanged two times; once so hard that power button powercycle did not result in booting machine. (thinkpad x60)

Quoted 7 lines
Quoted 1 line

I'll do some more tests...

There are a couple of things you may want to revert, like 7503bfbae89eba07b46441a5d1594647f6b8ab7d and a0d4922da2e4ccb0973095d8d29f36f6b1b5f703 .

Please let me know if that helped.

Harwey told me they are already reverted in latest -git, so I tried that but no luck. I'll try disabling CONFIG_SMP now.

Actually, in one case I seen some oopses scroll by... looked timekeeping related, but certainly went too fast.
Pavel

Signature · 2 lines
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html
#3 lore raw
In reply to Christian Borntraeger
Quoted 16 lines

Am Dienstag 13 Januar 2009 schrieb Rafael J. Wysocki:

Quoted 15 lines · Pavel Machek
Quoted 9 lines

Hi!

Just when I thought long merge window means less problems...

...suspend seemed to work until last tuesday or so. Now it hanged two times; once so hard that power button powercycle did not result in booting machine. (thinkpad x60)

I'll do some more tests...

There are a couple of things you may want to revert, like 7503bfbae89eba07b46441a5d1594647f6b8ab7d and a0d4922da2e4ccb0973095d8d29f36f6b1b5f703 .

Quoted 5 lines
Quoted 1 line

Please let me know if that helped.

I tested rc1+ these two reverts yesterday on my T61p. Doesnt work. There must be at least one more patch that broke S2R on my thinkpad. In fact, it is cpu hotplug that is broken (unplug works).

I reproduced that: echo 0 > online works, echo 1 > online hangs. Will compile SMP-less kernel to verify that there are no more problems...
Pavel

Signature · 2 lines
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html
#4 lore raw
Quoted 22 lines · Pavel Machek
Quoted 16 lines

Am Dienstag 13 Januar 2009 schrieb Rafael J. Wysocki:

Quoted 15 lines · Pavel Machek
Quoted 9 lines

Hi!

Just when I thought long merge window means less problems...

...suspend seemed to work until last tuesday or so. Now it hanged two times; once so hard that power button powercycle did not result in booting machine. (thinkpad x60)

I'll do some more tests...

There are a couple of things you may want to revert, like 7503bfbae89eba07b46441a5d1594647f6b8ab7d and a0d4922da2e4ccb0973095d8d29f36f6b1b5f703 .

Quoted 5 lines
Quoted 1 line

Please let me know if that helped.

I tested rc1+ these two reverts yesterday on my T61p. Doesnt work. There must be at least one more patch that broke S2R on my thinkpad. In fact, it is cpu hotplug that is broken (unplug works).

Quoted 11 lines

I reproduced that: echo 0 > online works, echo 1 > online hangs. Will compile SMP-less kernel to verify that there are no more problems...
Pavel
--
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/

I've a X61 running 32-bit and it's able to suspend-to-ram.

Please see "[Bug #12422] 2.6.28-git can't resume from str" if it's fixes your problem. Try to pull Linus' latest tree first and you need apply only Rafael's latest patch.

Thanks,
Jeff.

#14 lore raw
In reply to Christian Borntraeger

Hi,

i´ve got the same problem with 2.6.29-rc1 and a thinkpad T61p. But there is something very strange with 2.6.29-rc1-git3 it works perfectly.

-----Ursprüngliche Nachricht-----
Von: [email protected] [mailto:[email protected]] Im Auftrag von Christian Borntraeger
Gesendet: Dienstag, 13. Januar 2009 13:43
An: Rafael J. Wysocki
Cc: Pavel Machek; kernel list
Betreff: Re: 2.6.29-rc1: thinkpad problems during resume

Am Dienstag 13 Januar 2009 schrieb Rafael J. Wysocki:

Quoted 9 lines · Pavel Machek
Quoted 8 lines

Hi!

Just when I thought long merge window means less problems...

...suspend seemed to work until last tuesday or so. Now it hanged two times; once so hard that power button powercycle did not result in booting machine. (thinkpad x60)

Quoted 7 lines
Quoted 1 line

I'll do some more tests...

There are a couple of things you may want to revert, like 7503bfbae89eba07b46441a5d1594647f6b8ab7d and a0d4922da2e4ccb0973095d8d29f36f6b1b5f703 .

Please let me know if that helped.

I tested rc1+ these two reverts yesterday on my T61p. Doesnt work. There must be at least one more patch that broke S2R on my thinkpad. In fact, it is cpu hotplug that is broken (unplug works).

I will try to bisect that as well, but currently I am busy trying to find a virtio regression....

Christian
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/

#16 lore raw
In reply to Pavel Machek
Quoted 16 lines · Pavel Machek
Quoted 16 lines · Pavel Machek
Quoted 9 lines

Hi!

Just when I thought long merge window means less problems...

...suspend seemed to work until last tuesday or so. Now it hanged two times; once so hard that power button powercycle did not result in booting machine. (thinkpad x60)

I'll do some more tests...

There are a couple of things you may want to revert, like 7503bfbae89eba07b46441a5d1594647f6b8ab7d and a0d4922da2e4ccb0973095d8d29f36f6b1b5f703 .

Quoted 7 lines
Quoted 1 line

Please let me know if that helped.

Harwey told me they are already reverted in latest -git, so I tried that but no luck. I'll try disabling CONFIG_SMP now.

Actually, in one case I seen some oopses scroll by... looked timekeeping related, but certainly went too fast.

Hm, I'm unable to reproduce this on my test boxes.

What hardware is there in your machine?

Rafael

#17 lore raw
Quoted 22 lines · Rafael J. Wysocki
Quoted 21 lines · Pavel Machek
Quoted 10 lines · Pavel Machek
Quoted 8 lines

Hi!

Just when I thought long merge window means less problems...

...suspend seemed to work until last tuesday or so. Now it hanged two times; once so hard that power button powercycle did not result in booting machine. (thinkpad x60)

Quoted 7 lines
Quoted 1 line

I'll do some more tests...

There are a couple of things you may want to revert, like 7503bfbae89eba07b46441a5d1594647f6b8ab7d and a0d4922da2e4ccb0973095d8d29f36f6b1b5f703 .

Please let me know if that helped.

Harwey told me they are already reverted in latest -git, so I tried that but no luck. I'll try disabling CONFIG_SMP now.

Quoted 6 lines
Quoted 2 lines

Actually, in one case I seen some oopses scroll by... looked timekeeping related, but certainly went too fast.

Hm, I'm unable to reproduce this on my test boxes.

What hardware is there in your machine?

Thinkpad x60... pretty standard intel machine.
Pavelp

Signature · 2 lines
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html
#18 lore raw
In reply to Pavel Machek
Quoted 9 lines · Pavel Machek

Hi!

Just when I thought long merge window means less problems...

...suspend seemed to work until last tuesday or so. Now it hanged two times; once so hard that power button powercycle did not result in booting machine. (thinkpad x60)

I'll do some more tests...

It resumes fine with CONFIG_SMP unset.
Pavel

Signature · 2 lines
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html
#19 lore raw
Quoted 1 line · Pavel Machek

resumes fine with CONFIG_SMP unset.

On my X60s, I had problem with s2ram, and I switch to the following ...

echo mem > /sys/power/state

Try and see if it works for you.

Thanks,
Jeff.

#5 lore raw
In reply to Jeff Chua

Hi!

Quoted 18 lines
Quoted 6 lines
Quoted 5 lines
Quoted 1 line

Please let me know if that helped.

I tested rc1+ these two reverts yesterday on my T61p. Doesnt work. There must be at least one more patch that broke S2R on my thinkpad. In fact, it is cpu hotplug that is broken (unplug works).

Quoted 11 lines

I reproduced that: echo 0 > online works, echo 1 > online hangs. Will compile SMP-less kernel to verify that there are no more problems...
Pavel
--
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/

Quoted 5 lines

I've a X61 running 32-bit and it's able to suspend-to-ram.

Please see "[Bug #12422] 2.6.28-git can't resume from str" if it's fixes your problem. Try to pull Linus' latest tree first and you need apply only Rafael's latest patch.

Which patch?

Anyway, it is CPU hotplug vs. cpufreq interaction:

bash D c024ed50 0 1780 1
f71ae6b8 00000082 00000004 c024ed50 f71ae6b8 00000001 00000000
00000000
c2010db4 f71ae6b8 f71166b8 f71ae910 c201b500 00000001 c0ab4500
f5d04dd0
ffff8659 00000000 f71aeb68 00000000 00000046 00000000 c077b5c9
c2010da4

Call Trace:
 [<c024ed50>] __lock_acquire+0x240/0xac0
 [<c077b5c9>] rwsem_down_failed_common+0x29/0x170
 [<c077b615>] rwsem_down_failed_common+0x75/0x170
 [<c077b72a>] rwsem_down_write_failed+0x1a/0x30
 [<c077b7a2>] call_rwsem_down_write_failed+0x6/0x8
 [<c077abd5>] down_write+0x45/0x50
 [<c05c0f7c>] lock_policy_rwsem_write+0x2c/0x60
 [<c05c0f7c>] lock_policy_rwsem_write+0x2c/0x60
 [<c05c12ea>] cpufreq_add_dev+0x1ca/0x4f0
 [<c05c0988>] cpufreq_cpu_get+0xa8/0xb0
 [<c05c0fdf>] cpufreq_update_policy+0x2f/0xf0
 [<c0778b54>] coretemp_device_add+0xb4/0xd0
 [<c0778c9f>] cpufreq_cpu_callback+0x6f/0x80
 [<c02455ec>] notifier_call_chain+0x3c/0x80
 [<c0245677>] raw_notifier_call_chain+0x17/0x20
 [<c0776e8a>] _cpu_up+0x10a/0x110
 [<c0776edb>] cpu_up+0x4b/0x70
 [<c074b629>] store_online+0x59/0x80
 [<c074b5d0>] store_online+0x0/0x80
 [<c04220ad>] sysdev_store+0x2d/0x40
 [<c02cc6d6>] sysfs_write_file+0xa6/0x100
 [<c028a913>] vfs_write+0xa3/0x140
 [<c02cc630>] sysfs_write_file+0x0/0x100
 [<c028af61>] sys_write+0x41/0x70
 [<c020323e>] syscall_call+0x7/0xb
Signature · 2 lines
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html
#6 lore raw
Quoted 1 line · Pavel Machek

Which patch?

Try this patch attached.

Jeff.

#20 lore raw
In reply to Jeff Chua
Quoted 8 lines · Jeff Chua
Quoted 1 line · Pavel Machek

resumes fine with CONFIG_SMP unset.

On my X60s, I had problem with s2ram, and I switch to the following ...

echo mem > /sys/power/state

Try and see if it works for you.

I tried that; unlike s2ram, this actually produces some oopses that can be viewed. Unfortunately, they scroll a bit too fast.

cpu hotplug vs. cpufreq seems to be responsible in my case, looking at the logs from "echo 1 > online".
Pavel

Signature · 2 lines
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html
#21 lore raw
Quoted 8 lines · Pavel Machek
Quoted 7 lines · Jeff Chua
Quoted 1 line · Pavel Machek

resumes fine with CONFIG_SMP unset.

On my X60s, I had problem with s2ram, and I switch to the following ...

echo mem > /sys/power/state

Quoted 7 lines
Quoted 1 line

Try and see if it works for you.

I tried that; unlike s2ram, this actually produces some oopses that can be viewed. Unfortunately, they scroll a bit too fast.

cpu hotplug vs. cpufreq seems to be responsible in my case, looking at the logs from "echo 1 > online".

There are three patches from Mike Travis that can help:
http://lkml.org/lkml/2009/1/16/377

Thanks,
Rafael

#22 lore raw
Quoted 12 lines · Rafael J. Wysocki
Quoted 11 lines · Pavel Machek
Quoted 7 lines · Jeff Chua
Quoted 1 line · Pavel Machek

resumes fine with CONFIG_SMP unset.

On my X60s, I had problem with s2ram, and I switch to the following ...

echo mem > /sys/power/state

Try and see if it works for you.

I tried that; unlike s2ram, this actually produces some oopses that can be viewed. Unfortunately, they scroll a bit too fast.

Quoted 8 lines
Quoted 2 lines

cpu hotplug vs. cpufreq seems to be responsible in my case, looking at the logs from "echo 1 > online".

There are three patches from Mike Travis that can help:
http://lkml.org/lkml/2009/1/16/377

Thanks,
Rafael

Note that these don't really fix anything, just replaces the current method of using set_cpus_allowed with the work_on_cpu call. If the laptop worked before the [bad] patch, then it should work again now with this one.

Thanks,
Mike

#23 lore raw
Quoted 19 lines · Mike Travis
Quoted 18 lines · Rafael J. Wysocki
Quoted 5 lines · Pavel Machek
Quoted 4 lines · Jeff Chua
Quoted 1 line · Pavel Machek

resumes fine with CONFIG_SMP unset.

On my X60s, I had problem with s2ram, and I switch to the following ...

Quoted 8 lines
Quoted 3 lines

echo mem > /sys/power/state

Try and see if it works for you.

I tried that; unlike s2ram, this actually produces some oopses that can be viewed. Unfortunately, they scroll a bit too fast.

cpu hotplug vs. cpufreq seems to be responsible in my case, looking at the logs from "echo 1 > online".

There are three patches from Mike Travis that can help:
http://lkml.org/lkml/2009/1/16/377

Quoted 7 lines
Quoted 2 lines

Thanks,
Rafael

Note that these don't really fix anything, just replaces the current method of using set_cpus_allowed with the work_on_cpu call. If the laptop worked before the [bad] patch, then it should work again now with this one.

Thanks for the clarification, I misread the changelogs.

OK, so we still have a cpufreq vs CPU hotplug problem, it seems. I can't reproduce it myself, though.

Thanks,
Rafael

#7 lore raw
In reply to Pavel Machek

Am Freitag 16 Januar 2009 schrieb Pavel Machek:

Quoted 10 lines

Hi!

Quoted 8 lines
Quoted 8 lines
Quoted 5 lines
Quoted 1 line

Please let me know if that helped.

I tested rc1+ these two reverts yesterday on my T61p. Doesnt work. There must be at least one more patch that broke S2R on my thinkpad. In fact, it is cpu hotplug that is broken (unplug works).

I reproduced that: echo 0 > online works, echo 1 > online hangs. Will compile SMP-less kernel to verify that there are no more problems...

I made some progress on my T61p. git from yesterday + a revert of commit e39ad415ac15116df213dfa2aa2a4f1b0857af9c

Author: Mike Travis <[email protected]>
Date:   Sun Jan 4 05:18:10 2009 -0800

cpumask: use work_on_cpu in acpi-cpufreq.c for read_measured_perf_ctrs

makes cpu hotplug working again. Resume from suspend to ram is still broken on my Thinkpad. Looks like there is yet another patch to find....

Ingo, Mike: reverting e39ad415ac15116df213dfa2aa2a4f1b0857af9c fixes cpu hotplug on my thinkpad. Any idea how to proceed? Should we revert that patch in Linus git until we know what was wrong?

#8 lore raw

Re: 2.6.29-rc1: [SOLVED] thinkpad problems during resume

Quoted 4 lines

commit e39ad415ac15116df213dfa2aa2a4f1b0857af9c

Author: Mike Travis <[email protected]>
Date:   Sun Jan 4 05:18:10 2009 -0800
Quoted 8 lines

cpumask: use work_on_cpu in acpi-cpufreq.c for read_measured_perf_ctrs

makes cpu hotplug working again. Resume from suspend to ram is still broken on my Thinkpad. Looks like there is yet another patch to find....

Ingo, Mike: reverting e39ad415ac15116df213dfa2aa2a4f1b0857af9c fixes cpu hotplug on my thinkpad. Any idea how to proceed? Should we revert that patch in Linus git until we know what was wrong?

I forgot to revert the usb patch.
Current status:
2.6.29-rc2 +
reverted e39ad415ac15116df213dfa2aa2a4f1b0857af9c (cpumask: use work_on_cpu in acpi-cpufreq.c for read_measured_perf_ctrs) +
reverted a0d4922da2e4ccb0973095d8d29f36f6b1b5f703 (USB: fix up suspend and resume for PCI host controllers)

has a working suspend to ram on my T61p. With any of these two patches not reverted resume hangs.

Christian

#9 lore raw

Re: 2.6.29-rc1: [SOLVED] thinkpad problems during resume

2009/1/17 Christian Borntraeger <[email protected]>:

Quoted 9 lines

I forgot to revert the usb patch.
Current status:
2.6.29-rc2 +
reverted e39ad415ac15116df213dfa2aa2a4f1b0857af9c (cpumask: use work_on_cpu in acpi-cpufreq.c for read_measured_perf_ctrs) +
reverted a0d4922da2e4ccb0973095d8d29f36f6b1b5f703 (USB: fix up suspend and resume for PCI host controllers)

Quoted 3 lines

has a working suspend to ram on my T61p. With any of these two patches not reverted resume hangs.

Great work. Thanks. On my HP Compaq nx6310 revert this patches also
helps: suspend to ram/disk works.

Thanks again :-)

Signature · 2 lines
Maciej Rutecki
http://www.maciek.unixy.pl
#10 lore raw

Re: 2.6.29-rc1: [SOLVED] thinkpad problems during resume

In reply to Christian Borntraeger
Quoted 21 lines · Christian Borntraeger
Quoted 4 lines

commit e39ad415ac15116df213dfa2aa2a4f1b0857af9c

Author: Mike Travis <[email protected]>
Date:   Sun Jan 4 05:18:10 2009 -0800
Quoted 8 lines

cpumask: use work_on_cpu in acpi-cpufreq.c for read_measured_perf_ctrs

makes cpu hotplug working again. Resume from suspend to ram is still broken on my Thinkpad. Looks like there is yet another patch to find....

Ingo, Mike: reverting e39ad415ac15116df213dfa2aa2a4f1b0857af9c fixes cpu hotplug on my thinkpad. Any idea how to proceed? Should we revert that patch in Linus git until we know what was wrong?

I forgot to revert the usb patch.
Current status:
2.6.29-rc2 +
reverted e39ad415ac15116df213dfa2aa2a4f1b0857af9c (cpumask: use work_on_cpu in acpi-cpufreq.c for read_measured_perf_ctrs) +
reverted a0d4922da2e4ccb0973095d8d29f36f6b1b5f703 (USB: fix up suspend and resume for PCI host controllers)

Quoted 4 lines

has a working suspend to ram on my T61p. With any of these two patches not reverted resume hangs.

Christian

Could you try one more experiment? With your working system, please cherry pick these commits and try it again?

68564a4... work_on_cpu: don't try to get_online_cpus() in work_on_cpu. e1d9ec6... work_on_cpu: Use our own workqueue.
6eb714c... cpufreq: use work_on_cpu in acpi-cpufreq.c for drv_read and drv_write

Quoted 1 line

From git://git.kernel.org/pub/scm/linux/kernel/git/travis/linux-2.6-cpus4096-for-ingo.git

If that works, then try and un-revert e39ad415ac and see if it's fixed as well?

(The underlying work_on_cpu() function was causing these problems, but the 1st two of the above should fix it.)

Thanks!
Mike

#11 lore raw

Re: 2.6.29-rc1: [SOLVED] thinkpad problems during resume

Quoted 8 lines

Could you try one more experiment? With your working system, please cherry pick these commits and try it again?

68564a4... work_on_cpu: don't try to get_online_cpus() in work_on_cpu. e1d9ec6... work_on_cpu: Use our own workqueue.
6eb714c... cpufreq: use work_on_cpu in acpi-cpufreq.c for drv_read and drv_write

Quoted 6 lines

From
git://git.kernel.org/pub/scm/linux/kernel/git/travis/linux-2.6-cpus4096-for
-ingo.git

If that works, then try and un-revert e39ad415ac and see if it's fixed as well?

Yes that works.

29-rc2 + your 3 patches cherry-picked + revert of a0d4922da2e4ccb0973095d8d29f36f6b1b5f703( usb problem) can suspend/resume. If you like, you can add an

to your three patches.

Christian

#12 lore raw

Re: 2.6.29-rc1: [SOLVED] thinkpad problems during resume

Quoted 15 lines · Christian Borntraeger
Quoted 8 lines

Could you try one more experiment? With your working system, please cherry pick these commits and try it again?

68564a4... work_on_cpu: don't try to get_online_cpus() in work_on_cpu. e1d9ec6... work_on_cpu: Use our own workqueue.
6eb714c... cpufreq: use work_on_cpu in acpi-cpufreq.c for drv_read and drv_write

Quoted 6 lines

From
git://git.kernel.org/pub/scm/linux/kernel/git/travis/linux-2.6-cpus4096-for
-ingo.git

If that works, then try and un-revert e39ad415ac and see if it's fixed as well?

Quoted 9 lines

Yes that works.

29-rc2 + your 3 patches cherry-picked + revert of a0d4922da2e4ccb0973095d8d29f36f6b1b5f703( usb problem) can suspend/resume. If you like, you can add an

to your three patches.

Thanks Christian!

#13 lore raw

Re: 2.6.29-rc1: [SOLVED] thinkpad problems during resume

Quoted 18 lines · Mike Travis
Quoted 17 lines · Christian Borntraeger
Quoted 8 lines

Could you try one more experiment? With your working system, please cherry pick these commits and try it again?

68564a4... work_on_cpu: don't try to get_online_cpus() in work_on_cpu. e1d9ec6... work_on_cpu: Use our own workqueue.
6eb714c... cpufreq: use work_on_cpu in acpi-cpufreq.c for drv_read and drv_write

Quoted 6 lines

From
git://git.kernel.org/pub/scm/linux/kernel/git/travis/linux-2.6-cpus4096-for
-ingo.git

If that works, then try and un-revert e39ad415ac and see if it's fixed as well?

Yes that works.

Quoted 8 lines
Quoted 7 lines

29-rc2 + your 3 patches cherry-picked + revert of a0d4922da2e4ccb0973095d8d29f36f6b1b5f703( usb problem) can suspend/resume. If you like, you can add an

to your three patches.

Thanks Christian!

Great.

Mike, could you please prepare a separate, linus/master based 'fixes' branch that cherry-picks these commits? (they overlap existing commits in cpus4096 which is a bit unfortunate but not that bad as long as it's done within reason - Git will sort it out.)

Ingo

#24 lore raw
In reply to Rafael J. Wysocki

Hi!

Quoted 11 lines
Quoted 10 lines
Quoted 5 lines

There are three patches from Mike Travis that can help:
http://lkml.org/lkml/2009/1/16/377

Thanks,
Rafael

Note that these don't really fix anything, just replaces the current method of using set_cpus_allowed with the work_on_cpu call. If the laptop worked before the [bad] patch, then it should work again now with this one.

Quoted 4 lines

Thanks for the clarification, I misread the changelogs.

OK, so we still have a cpufreq vs CPU hotplug problem, it seems. I can't reproduce it myself, though.

I kind of lost track here. Is everything fixed, or will there still be problems in -rc3?

I turned off CPUFREQ to get rid of this problem, so yes, I can reproduce it easily.

Signature · 2 lines
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html
#25 lore raw
Quoted 13 lines · Pavel Machek

Hi!

Quoted 11 lines
Quoted 10 lines
Quoted 5 lines

There are three patches from Mike Travis that can help:
http://lkml.org/lkml/2009/1/16/377

Thanks,
Rafael

Note that these don't really fix anything, just replaces the current method of using set_cpus_allowed with the work_on_cpu call. If the laptop worked before the [bad] patch, then it should work again now with this one.

Quoted 7 lines
Quoted 4 lines

Thanks for the clarification, I misread the changelogs.

OK, so we still have a cpufreq vs CPU hotplug problem, it seems. I can't reproduce it myself, though.

I kind of lost track here. Is everything fixed, or will there still be problems in -rc3?

That depends on which patches get merged before -rc3. There are quite a few of them. :-(

Thanks,
Rafael