Computer suspends immediately after resuming if power is unplugged while suspended

[Phil Sung]

Binary package hint: gnome-power-manager

(This is an up-to-date Karmic AMD64 on a Thinkpad X61s.)

Immediately after resuming from suspend, Ubuntu occasionally goes right back into suspend. Subsequent attempts to resume are still successful.

This started a few weeks ago.

I have only noticed this happening when AC power is unplugged between suspend and resume.

ProblemType: Bug
Architecture: amd64
Date: Sun Sep 6 14:48:16 2009
DistroRelease: Ubuntu 9.10
Package: gnome-power-manager 2.27.91-0ubuntu1
ProcEnviron:
 PATH=(custom, user)
 LANG=en_US.UTF-8
 SHELL=/bin/bash
ProcVersionSignature: Ubuntu 2.6.31-9.29-generic
SourcePackage: gnome-power-manager
Uname: Linux 2.6.31-9-generic x86_64

[Katsudon (katsudon)]

I can confirm this on Thinkpad X61s.

[Guillaume Giroux (gug-microflex)]

confirmed on thinkpad T61p running amd64 karmic with all updates too - as the reporter says, it seems to be related to AC power being unplugged before resuming.

[Sandro Mani (sandromani)]

Same here on a T400 as well as on a T60.

[Matt Zimmerman (mdz)]

Confirming, as there are multiple reports here and some anecdotal support in bug 261084 of a more recent regression in Karmic

[Matt Zimmerman (mdz)]

I've seen plenty of spurious suspends immediately after resume, but haven't been able to isolate the circumstances which trigger it. Connecting or disconnecting power while suspended doesn't seem to trigger the problem for me.

[Matt Zimmerman (mdz)]

I see this problem on a Dell mini 10v, but NOT on a ThinkPad T61, both running current Karmic

[Gunnar Thielebein (lorem-ipsum)]

I have this issue with a hp touchsmart tx2 running the latest karmic-amd64. Tested with connected and disconnected power cable.
Second resume is successful. Suspending with pm-suspend resumes correctly on first time.

[Martin Pitt (pitti)]

I tested this with the automatic suspend when the battery goes low. DK-power git head as a patch for it.

[Martin Pitt (pitti)]

Fix applied to Debian package git tree, and uploaded to unapproved.

[Steve Beaulac (steve-sagacity)]

Same situation on my MSI U120 running fresh (just updated) Karmic.

The system resumes perfectly everything seem to be working but when you
press any key on the keyboard the system goes back to suspend mode. Everything work after second resume.

Will be waiting for the fix

[Martin Pitt (pitti)]

I would highly appreciate if folks could add https://launchpad.net/~ubuntu-desktop/+archive/ppa and upgrade from this. It has a new dk-power (and nothing else) which hopefully fixes this. Please report back here with both positive and negative feedback. Thanks!

[Martin Pitt (pitti)]

If it does not work, please do

 killall gnome-power-manager
 gnome-power-manager --debug 2>&1 | tee ~/gpm.log

then reproduce the problem, press Control-C, and attach ~/gpm.log here.

Then press Alt+F2, run "gnome-power-manager" to get it back in normal shape.

Thanks!

[Martin Pitt (pitti)]

Sorry, not "--debug", but "--verbose":

 killall gnome-power-manager
 gnome-power-manager --verbose 2>&1 | tee ~/gpm.log

[Andy Whitcroft (apw)]

Ok. The versions from the above PPA do not see to work for me on my scenario; specifially if you suspend with AC power and then remove the power while suspended you get a double suspend:

  apw@penfold:~$ dpkg -l | grep kit-power
  ii devicekit-power 011-1ubuntu1 abstraction for power
  ii libdevkit-power-gobject1 011-1ubuntu1 abstraction for power

As per comment #13/#14 I have reproduced this with gnome-power-manager debugging turned on and am attaching the gpm.log, note that this does report two suspends:

  - suspending, reason: The lid has been closed on ac power.
  - suspending, reason: The lid has been closed, and the ac adapter removed (and gconf is okay).

[Martin Pitt (pitti)]

Thanks. I tested the inverse scenario when plugging in AC power during suspend, and g-p-m didn't notice quickly enough and still thought I'd run out of battery and thus suspended again. I confirmed that the pending dk-power upload fixes this.

Andy's scenario looks like g-p-m is getting a "bouncing" effect on the lid-close input event. It does not seem related to dk-power at all. So I open a new g-p-m task for this.

[Launchpad Janitor (janitor)]

This bug was fixed in the package devicekit-power - 011-1ubuntu1

---------------
devicekit-power (011-1ubuntu1) karmic; urgency=low

  * Add 00git-fix-ac-refresh.patch: Properly refresh update-time after AC
    changes. This avoids immediately suspending again when AC state changes
    during suspend. (LP: #425411)

 -- Martin Pitt <email address hidden> Thu, 15 Oct 2009 15:45:44 +0200

[Elie Charest (archiesteel)]

I get the same bug with devicekit-power updated to version 011-1ubuntu1 :-/

The gpm.log file also shows two suspending events, like Andy's.

Problem occurs under similar conditions: unplugging suspended computer then waking up (but not the reverse).

Current karmic running on a Dell Mini10v.

[Gunnar Thielebein (lorem-ipsum)]

I have the lid_up event twice, also with the ppa version of devicekit-power on hp touchsmart tx2.

[Gunnar Thielebein (lorem-ipsum)]

s/lid_up/lid_down regarding previous comment.

Perhaps acpi_listen is also of interest (2x open/close):

button/lid LID 00000080 00000007
button/lid LID 00000080 00000008
button/lid LID 00000080 00000009
button/lid LID 00000080 0000000a

[Lagos (artificial-music)]

Confirmed on Asus Eeepc 1000h.
If the power cable is unplugged during a suspended state, the system will go right back to suspend after login.

Basically.
1. plug system into ac power
2. close lid to suspend
3. disconnect power cord.
4. open lid to wake up.
5. system will wake up, ask for log in, and as soon as it is logged in, suspend again. Press any key to wake the system up after that and the problem will be gone.

[rupert (r-plumridge)]

Confirmed on a Dell XPS M1330 running Karmic updated as fully as possible.

[Martin-Éric Racine (q-funk)]

I get a similar problem on a Dell D430 as reported in bug #459082. However, it also happens if power if unplugged *before* going into suspend. Here, I always unplug the adapter *before* and give it a minute or two before closing the lid.

[Martin-Éric Racine (q-funk)]

I'll also point out that this bug is not amd64-specific. This D430 has a single-core CPU that is pure i386 and the issue exists too.

[Gavin Chappell (g-a-c)]

Also affects me on a Toshiba Portege A600 running the Karmic RC (been running and upgrading roughly since the beta release but haven't upgraded for a day or two). Attached are my gpm.log and a list of power related packages on my system.

[Gavin Chappell (g-a-c)]

Here's the second attachment, I couldn't see a way to attach two files to a single comment. I should also have told you I'm on the AMD64 release.

[Martin-Éric Racine (q-funk)]

At the risk of stating the obvious, this bug falls into the "if it ain't broken, don't fix it" category, as it was working flawlessly in Jaunty and during the early phase of Karmic development. Someone somewhere obviously made a drastic change (most probably upstream, but we shouldn't entirely discount changes in Debian/Ubuntu packaging, especially build-time patches) and it would be a good idea to figure out who, when and why, in order to avoid this ever happening again.

[Martin Pitt (pitti)]

Martin-Éric Racine [2009-10-27 11:27 -0000]:
> At the risk of stating the obvious, this bug falls into the "if it ain't
> broken, don't fix it" category, as it was working flawlessly in Jaunty
> and during the early phase of Karmic development. Someone somewhere
> obviously made a drastic change (most probably upstream, but we
> shouldn't entirely discount changes in Debian/Ubuntu packaging,
> especially build-time patches) and it would be a good idea to figure out
> who, when and why, in order to avoid this ever happening again.

Presumably in the change from hal to dk-power.

[flyman (fleitman68)]

Bump!
Computer suspends immediately after resuming _if_ power is unplugged while suspended.

[Max (mpflueger)]

I just upgraded to the release of Karmic on my Lenovo T400 and I am experiencing this bug.

[mnoe (matthias-noe)]

After upgrading from jaunty to karmic I am experiencing this bug on a Macbook Pro 5,5.

[Aviv Ben-Yosef (aviv-by)]

Same here on a Thinkpad X200s, latest karmic

[Martin Ling (martin-launchpad)]

Also getting this on a Dell Latitude D610 since upgrade to Karmic. I think it does only happen when the power has been unplugged whilst suspended. It also doesn't seem to happen if I just suspend, unplug, and wake up again. Every time it happens the laptop has been asleep in my bag for quite a while.

[O. Emmerson (oemmerson)]

I have been experiencing this problem since installing Ubuntu v9.10.
I am running a Lenovo 3000 n500 laptop.

[maiden30403 (maiden30403)]

Seem to be having the same problem on Asus N81-VG using Ubuntu 9.10 AMD64 with latest packages.

[Jason Gao (jasongao)]

I seem to be having the same problem on an X200 Tablet (similar hardware to the X200s), except that the whole thing locks up when it tries to suspend again. I even get as far as seeing the unlock prompt before it freezes.

[Aviv Ben-Yosef (aviv-by)]

Jason: that's happening on my X200s too, and documented in this bug: http://linux.byexamples.com/archives/297/how-to-ssh-without-password/

[Aviv Ben-Yosef (aviv-by)]

Bad link, I meant this: https://bugs.launchpad.net/ubuntu/+source/linux/+bug/479842

[Mark Cariaga (mzc)]

this also affects Acer aspire D250. if I can recall corrctly there was a patch/update for dk-power in between alpha and beta release, after that everything went downhill.

[Gunnar Thielebein (lorem-ipsum)]

> this also affects Acer aspire D250. if I can recall corrctly there was a patch/update for dk-power in between alpha and beta
> release, after that everything went downhill.

can you point to that patch? Was it Ubuntu specific or upstream?

[Sandro Mani (sandromani)]

The issue is not ubuntu-specific, getting the same on fedora 12 / rawhide.

[Mark Cariaga (mzc)]

Sorry Gunnar. I have switch back to 9.04 for my netbook. I cant
remember exactly which patch that was.

On Sat, 2009-11-21 at 12:01 +0000, Gunnar Thielebein wrote:
> > this also affects Acer aspire D250. if I can recall corrctly there was a patch/update for dk-power in between alpha and beta
> > release, after that everything went downhill.
>
> can you point to that patch? Was it Ubuntu specific or upstream?
>

[Mark Cariaga (mzc)]

@Martin Pitt

I'm attaching my gpm.log, hope its helps to visualise this bug

[Mark Cariaga (mzc)]

@Martin Pitt
I tried #12, it does not fix the problem. When I unplug AC, it thinks I'm closing the lid (based on settiing) but when I actually close the lid it does not suspend but just turn off lcd.

[Andrew Ivanov (aa.ivanov)]

I am experiencing this issue on a Toshiba Satellite A110 laptop.
A slight difference with the above reports that's worth mentioning is that I use uswsusp instead of the stock solution.

I have tried the following scenarios:
1. I tried invoking the s2ram command directly - works like charm, just as expected
2. I suspended the computer from the respective option in the shutdown menu - worked as expected
3. I suspended by closing the lid - on resume the machine auto-suspended. Second resume attempt was successful (g-p-m.log attached)

I have repeated step 3 with both AC power on and off. Makes no difference.

Looking at the log file I can't see any lid events to cause the second suspend...

Any idea where to look next?

[t3rmin (matt-thetrents)]

Seeing this same issue on an HP TX2525NR with 9.10 Netbook Remix.

[kenneth hsu (kennethhsu)]

This fixed the problem for me (running 2.6.31-15-generic on Karmic)

Run gconf-editor and go to

/ --> apps --> gnome-power-manager --> actions

and uncheck the box labeled "event_when_closed_battery".

This option is described to be:

"If the battery lid close event should occur (for example 'Suspend when lid closed on battery') when the lid is previously shut and the AC power disconnected at a later time."

[Mário Buči (mario-buci)]

Who would think that a small fix like that would help. Thanks kenneth hsu, it works! :)

[Gorka Navarrete (emrys)]

Kudos for Kenneth. Any insight about why the problem started now?

[Andrew Ivanov (aa.ivanov)]

Well, at least in my case the fix proposed by kenneth in comment #47 makes no difference :(

[Alexander Hunziker (alex-hunziker)]

Kenneth Hsu's workaround works for me as well.

[kv488 (bernhard-ott)]

Thanks to kenneth hsu!
Pulling the power chord doesn't trigger suspension on my Acer TM 380 (fix described in #47)
Hope the rest is going to be fixed soon (lid events) ...

[Mark Cariaga (mzc)]

hi,

in somes cases like mine, when you do "sudo devkit-power -d" and
noticed that the lid is always closed even if its not then the following
values needs to be changed also:

gnome-power-manager/buttons/lid_Ac and lid_batt - changes values to
nothing
 see screenshot
On Mon, 2009-11-30 at 11:46 +0000, kv488 wrote:
> Thanks to kenneth hsu!
> Pulling the power chord doesn't trigger suspension on my Acer TM 380 (fix described in #47)
> Hope the rest is going to be fixed soon (lid events) ...
>

[Katsudon (katsudon)]

The original bug report concerned the problem related to resuming if power unplugged while suspended and if I understand, it is still an issue for some users.

The problems related to closing lid might be related, but maybe it would make sense to keep it as a separate bug report or at least not to remove the original problem from the description of the bug. Another option would be to change the description of the bug to include both problems.

[Mark Cariaga (mzc)]

@katsudon
there are already several bugs refering to power management issues. after doing my own research, most of this bugs are just different combinations of issues with lids, suspend, and AC/batt. (power management)

my point is most likely if this bug is solve so will be other somewhat similar bugs. Hence, making another report may be just redundance.

[Andrew Ivanov (aa.ivanov)]

@Katsudon, comment 54

So far I have counted half a dozen changes in the summary of this bug (not counting other changes in other descriptors, flags, keywords, etc., etc, whatever). I find this continuous change of the scope of the bug quite confusing. Does it affect me? Well... last week it affected me, but this week someone redefined the bug (again) and I'm not affected by it as it is defined now... perhaps next week they'll redefine it again and then... who knows...

Furthermore a whole lot of different bug reports are marked as duplicate on this one:
448089 - System suspends instantly after resume from suspend
451433 - disconnecting AC power suspends laptop
457966 - immediate hibernation after resume
459082 - suspends again after suspend-on-lid
477147 - Acer Aspire one D250 suspends automatically when AC is plugged in or out

Based on the current definition of the scope of this bug, as you outlined it in your comment 54, please unlink bugs 448089, 457966, and 459082. They don't mention any issues with plugging in/out AC and thus they are unrelated to this bug. Since you want the double power/suspend/hibernate action on lid event tracked in another bug, I'd propose 448089, and the rest linked as duplicates on it.

Or I can file a new one as soon as I get back to my laptop, if you prefer it that way.

[Chris Conway (cconway)]

Kenneth Hsu's suggestion seemed to fix the problem on my Dell Inspiron 6400 (per #56, my specific problem as immediate suspend after resume), but it also seemed to cause a regression on Bug 368704 (which hadn't been a problem for me in several months).

[Todd Lyons (tlyons)]

Karmic on a Lenovo T60. I see the "immediate suspend after resume" issue as well, but it doesn't seem to have anything to do with power presence or power state change for me. It happens when the screen is blanked and I close the lid. Gnome-power-manager is configured with "Put display to sleep when inactive for: 10 minutes"), and I have my system set to suspend when the lid is closed. It properly suspends whether the screen is lit or blanked. However, if the screen is blanked when I close the lid, the next time I open it, the "immediate suspend after resume" issue will occur. If I make sure that the screen is not blanked when I close the lid, I cannot recall ever seeing the issue. I will do some experimenting to try and make sure that power state change does or does not have any bearing on the symptoms.

[Todd Lyons (tlyons)]

Blanked screen seems to be a red herring. My T60 with Karmic had the issue this morning when I turned it on. Procedure was: using computer last night, close lid (suspend), left plugged in overnight, unplug power this morning, get on bus, open lid. It resumed and then immediately suspended. My previous comment (#58) should be disregarded.

[Dan Maranville (likuidkewl)]

Has this been flagged as upstream or not? Can anyone confirm it does happen on a different distribution(as per comment #41)?

[Nikke (nmellegard)]

@Todd Lyons #59
Same thing happening with my T61+Karmic; whenever I suspended the computer with power supply attached and waking it on battery it would go straight back to suspend. This did not happen in Jaunty.

The procedure in #47 fixed the problem for me.

/N

[Gunnar Thielebein (lorem-ipsum)]

All given solutions did not worked on touchsmart tx2.
The former config I used was to suspend on ac_power and battery.
I recognised that there is a difference when the laptop was suspended via button and via closing the lid. Via button the resume went fine without any following suspend action.

So my current workaround is to disable the suspend on ac_power / battery - event. This forces myself to suspend via the menu in gpm icon.

You can get the menu if you flag gnome-power-manager/ui "show_actions_in_menu".

[Gunnar Thielebein (lorem-ipsum)]

This is in someway near to mcz's solution. I wonder if you still get the lid down event with your config?

[Todd Lyons (tlyons)]

@Nikke #61
Confirmed that the procedure in #47 fixed the problem for my Thinkpad T60 with Karmic. Thanks!

[Mark Cariaga (mzc)]

@Gunnar Thielebein:
This is to confirm that editing some gconf values on
gnome-power-management helps reduce the annoyance. But the fallback is
the system still does not suspend when I close my lid. This fallback may
be specific on my pc

On Thu, 2009-12-03 at 13:43 +0000, Gunnar Thielebein wrote:
> Gunnar Thielebein

[Ori Kremer (ori-kremer)]

When unplugging the AC the system directly suspends.
Changing the lid close operation to black screen is the workaround i use.
This is on LG S1 Express Dual after a fresh Karmic install.

[m4lte (malte-koester)]

I'm also experiencing from this bug.

[Martin Gollowitzer (gollo)]

I am also affected by this bug. When I suspend my Lenovo X61 by closing the lid and opening it again, the notebook will immediately suspend after resuming. If I send it to sleep via the power-off menu and then close the lid, it resumes normally. I also tried with debian testing (squeeze), which shows the same symptoms. I didn't plug in or unplug the power cord while the system was in sleep mode since I currently use the system without a battery so it's plugged in all the time.

[Chris Conway (cconway)]

Seeing this on a brand new Inspiron 1545.

[Tobias McNulty (tmcnulty1982)]

I have a similar problem on a Lenovo ThinkPad T400s. If I unplug the machine while it's asleep, it just freezes outright on resume. I can see the login screen, but the mouse and keyboard don't function, and the "I" bar (text entry cursor) is not blinking per usual.

The resolution described in comment #47 seems to fix the issue for me (though I'm not exactly sure what it does).

I'm running a fully up to date copy of Karmic.

[Tobias McNulty (tmcnulty1982)]

I should also say that on my T60 I see the behavior that everyone else describes (computer suspends itself immediately after resuming).

[Martin Pitt (pitti)]

Retitling back. One report per problem, mixing different symptoms in one report is hard to track.

[m4lte (malte-koester)]

#47 also solved the problem for me.

[Bart Samwel (bart-samwel)]

#47 works on my Samsung NC-10 as well. I don't think it's a "proper fix" though -- but it does indicate where the problem lies. In particular, the difference between a failing situation and an expected situation is mediated by the presence of a battery event immediately after resume. And apparently SOMETHING is reacting to that battery event in an incorrect way. Perhaps some low-battery auto-suspend functionality that triggers while battery capacity is not yet correctly known? Does anybody have any hints on how to determine what is responding to this "battery event"?

[Chris Conway (cconway)]

#47 doesn't work for me on my Inspiron 1545.

[Przemek K. (azrael)]

Another possibly related bug in Karmic: bug 492649
This bug has a patch included (debdiff).

[Przemek K. (azrael)]

One more bug with a similar description: bug 480492 - it might be a duplicate of this one or the previously mentioned one.

[Przemek K. (azrael)]

Similar bug in Red Hat: https://bugzilla.redhat.com/show_bug.cgi?id=475585

[Przemek K. (azrael)]

Here's the patch in review mode on another Gnome bug:
https://bugzilla.gnome.org/review?bug=604362&attachment=151235

https://bugzilla.gnome.org/show_bug.cgi?id=604362

[Przemek K. (azrael)]

Can anyone review the patch from bug 492649?
http://launchpadlibrarian.net/37770207/gpm-debdiff

It was based on my logs:
http://launchpadlibrarian.net/37743253/gnome-power-manager.out
(output of gnome-power-manager --verbose when reproducing this bug
[sleep, plug off power, resume, laptop goes automatically to sleep in
a few seconds, resume again] ran on my Macbook v2,1, Ubuntu 9.10
64-bit)

I've also posted some other logs in that bug that might interest you:
http://launchpadlibrarian.net/37809088/gpm-hpcompaq-unaffected.out
(Logs from my older laptop - HP Compaq NX6325.
Tried both suspending on AC + resuming on battery and vice versa, I
did the suspending by using the FUSA menu, not by closing the lid -
this wasn't triggering the bug)

http://launchpadlibrarian.net/37809294/gpm-hpcompaq-affected-lidclose.out
(Logs from the same laptop when suspending by closing the lid - bug triggered)

--
## Przemysław Kulczycki <<=>> Azrael Nightwalker ##
# Jabber/XMPP/Gtalk/Tlen ID: azrael[na]jabster.pl #
# (Co to jest? Zobacz na: http://jabberfaq.info ) #
## www: http://reksio.ftj.agh.edu.pl/~azrael/ #####

[Chow Loong Jin (hyperair)]

Setting the GNOME bug for GPM to BGO #604362, because that one has a patch which solves this issue. The same patch is in the debdiff from LP #492649.

Chris Choulson has mentioned that he'll look into including the patch into g-p-m 2.29.x. After that we can look into pushing it into Karmic as an SRU.

[Chow Loong Jin (hyperair)]

Er I meant Chris Coulson, sorry.

[Launchpad Janitor (janitor)]

This bug was fixed in the package gnome-power-manager - 2.29.1-0ubuntu1

---------------
gnome-power-manager (2.29.1-0ubuntu1) lucid; urgency=low

  [ Chris Coulson ]
  * New upstream version:
    - Switch to a single menu design, rather than a seporate left and
      right menu.
    - Show the sleep failure as a modal dialog box.
    - Sets some tooltips for the gnome-power-preferences buttons.
    - Throttle screensaver before suspend/hibernate.
    - Add a --device parameter to gnome-power-statistics so we can set
      the device focus.
    - When we click on a battery device in the context menu, show the
      gnome-power-statistics dialog.
    - Use gnome-settings-daemon popup code.
  * debian/control:
    - Don't build-depend on libhal-dev, it's not needed even with HAL
      support.
  * debian/rules:
    - Configure with --enable-hal. HAL support is now conditional, and
      disabled upstream by default.
  * Drop 01-segfault-in-gpm_disks_register.patch - the patch is not needed
    anymore.
  * 02-notify-osd-support.patch: Re-write, using code shared with
    gnome-settings-daemon
  * 90-autotools.patch - Autotools update for notify-osd patch

  [ Chow Loong Jin ]
  * debian/patches/09-fix-double-suspend.patch:
    - Manually check lid status in gpm_manager_client_changed_cb.
      Fixes double-suspend bug (LP: #425411)
 -- Chris Coulson <email address hidden> Tue, 19 Jan 2010 01:47:54 +0000

[Chow Loong Jin (hyperair)]

Here's the debdiff for Karmic SRU

[Martin Pitt (pitti)]

Sponsored, thanks hyperair!

[Martin Pitt (pitti)]

Accepted gnome-power-manager into karmic-proposed, the package will build now and be available in a few hours. Please test and give feedback here. See https://wiki.ubuntu.com/Testing/EnableProposed for documentation how to enable and use -proposed. Thank you in advance!

[Neil Woolford (neil-neilwoolford)]

Fixed on my Lenovo 3000 N200 using the new 2.28.1-0ubuntu1.1 from karmic-proposed. No apparent regressions.

[Max (mpflueger)]

Upgraded to gnome-power-manager 2.28.1-0ubuntu1.1 on my Lenovo T400.

My settings are AC: blank screen on lid close, Battery: suspend on lid close. Initially after the upgrade (and computer restart) the computer did not suspend on lid close while on battery. After changing my AC setting from blank screen to suspend and back everything seems to work.

[Tomas Pospisek (tpo-deb)]

Upgraded to gnome-power-manager 2.28.1-0ubuntu1.1 on HP 6710b. Seems to work until now.

[sam tygier (samtygier)]

2.28.1-0ubuntu1.1 fix seems good on Lenovo S12. thanks

[Przemek K. (azrael)]

2010/1/21 sam tygier <email address hidden>:
> 2.28.1-0ubuntu1.1 fix seems good on Lenovo S12. thanks

Works for me too - Macbook v2,1 amd64 Karmic

--
## Przemysław Kulczycki <<=>> Azrael Nightwalker ##
# Jabber/XMPP/Gtalk/Tlen ID: azrael[na]jabster.pl #
# (Co to jest? Zobacz na: http://jabberfaq.info ) #
## www: http://reksio.ftj.agh.edu.pl/~azrael/ #####

[Mark Cariaga (mzc)]

Also works on hp mini 1116nr. Anyone on aspire one that can confirm
if the bug is squished?

On Jan 21, 2010, at 1:52 PM, Przemysław Kulczycki <<email address hidden>
m> wrote:

> 2010/1/21 sam tygier <email address hidden>:
>> 2.28.1-0ubuntu1.1 fix seems good on Lenovo S12. thanks
>
> Works for me too - Macbook v2,1 amd64 Karmic
>
> --
> ## Przemysław Kulczycki <<=>> Azrael Nightwalker ##
> # Jabber/XMPP/Gtalk/Tlen ID: azrael[na]jabster.pl #
> # (Co to jest? Zobacz na: http://jabberfaq.info ) #
> ## www: http://reksio.ftj.agh.edu.pl/~azrael/ #####
>
> --
> Computer suspends immediately after resuming if power is unplugged
> while suspended
> https://bugs.launchpad.net/bugs/425411
> You received this bug notification because you are a direct subscriber
> of a duplicate bug.
>
> Status in DeviceKit based power management D-Bus backend: Unknown
> Status in Gnome Powermanager: Unknown
> Status in "devicekit-power" package in Ubuntu: Fix Released
> Status in "gnome-power-manager" package in Ubuntu: Fix Released
> Status in "devicekit-power" source package in Lucid: Fix Released
> Status in "gnome-power-manager" source package in Lucid: Fix
> Released
> Status in "devicekit-power" source package in Karmic: Fix Released
> Status in "gnome-power-manager" source package in Karmic: Fix
> Committed
>
> Bug description:
> Binary package hint: gnome-power-manager
>
> (This is an up-to-date Karmic AMD64 on a Thinkpad X61s.)
>
> Immediately after resuming from suspend, Ubuntu occasionally goes
> right back into suspend. Subsequent attempts to resume are still
> successful.
>
> This started a few weeks ago.
>
> I have only noticed this happening when AC power is unplugged
> between suspend and resume.
>
> ProblemType: Bug
> Architecture: amd64
> Date: Sun Sep 6 14:48:16 2009
> DistroRelease: Ubuntu 9.10
> Package: gnome-power-manager 2.27.91-0ubuntu1
> ProcEnviron:
> PATH=(custom, user)
> LANG=en_US.UTF-8
> SHELL=/bin/bash
> ProcVersionSignature: Ubuntu 2.6.31-9.29-generic
> SourcePackage: gnome-power-manager
> Uname: Linux 2.6.31-9-generic x86_64
>
> To unsubscribe from this bug, go to:
> https://bugs.launchpad.net/devicekit-power/+bug/425411/+subscribe

[Adam Stark (adstark1982)]

2.28.1-0ununtu1.1 does not seem to fix this problem on a Toshiba NB205: g-p-m.log follows. acpi_listen only reports one lid switch event

[Chris Coulson (chrisccoulson)]

Did you actually fully restart your session after installing the update?

[Mark (mark-wege)]

I have the same issue running Kubuntu with KDE. Would be nice if there could be a fix for Kubuntu too.

[Adam Stark (adstark1982)]

@Chris:
Yes. I've also restarted the computer since, and the problem persists. This is on Karmic.

[Adam Stark (adstark1982)]

For now I've put a hook in /etc/pm/sleep.d/ that will inhibit a suspend less than 15 seconds after the previous resume.

[Chow Loong Jin (hyperair)]

On Saturday 23,January,2010 09:29 AM, Adam Stark wrote:
> For now I've put a hook in /etc/pm/sleep.d/ that will inhibit a suspend
> less than 15 seconds after the previous resume.
>
Ah, I see what the problem is. It's the same problem, occurring in another
portion of the code. Blargh. I'll come up with another patch to fix this.

The problem is that gnome-power-manager tracks console changes as well, and
since you're not using KMS (I think you're not) your console changes to tty1 or
something else prior to suspend and then changes back when you're done. PM-utils
is responsible for changing it, but it is something necessary. I'll come up with
a patch that fixes this more globally.

--
Kind regards,
Chow Loong Jin (GPG: 0x8F02A411)
Ubuntu Contributing Developer

[Chow Loong Jin (hyperair)]

On Saturday 23,January,2010 09:29 AM, Adam Stark wrote:
> For now I've put a hook in /etc/pm/sleep.d/ that will inhibit a suspend
> less than 15 seconds after the previous resume.
>
Two patches are attached. The new 09-fix-double-suspend.patch that should fix
the issue pointed out by Adam Stark, and the new Karmic SRU debdiff (from
ubuntu1.1 to ubuntu1.2).

--
Kind regards,
Chow Loong Jin (GPG: 0x8F02A411)
Ubuntu Contributing Developer

[Chow Loong Jin (hyperair)]

The patch I submitted earlier has been committed upstream as commit
b10f242af345f01415d26e35edb147dd7177453b.

  affects ubuntu/gnome-power-manager
  status fixcommitted

--
Kind regards,
Chow Loong Jin (GPG: 0x8F02A411)
Ubuntu Contributing Developer

[Chris Coulson (chrisccoulson)]

Awesome work, and thanks for tracking this down hyperair! It's getting late here now, but if nobody has uploaded these by the time I get up tomorrow morning then I shall upload both of these for you.

[Launchpad Janitor (janitor)]

This bug was fixed in the package gnome-power-manager - 2.29.1-0ubuntu2

---------------
gnome-power-manager (2.29.1-0ubuntu2) lucid; urgency=low

  [ Chow Loong Jin ]
  * debian/patches/09-fix-double-suspend.patch:
    - Updated to make gpm_button_is_lid_closed return a current value
      from DeviceKit instead of its own locally stored (outdated) value.
      This fixes another resume race, and really fixes LP: #425411.
 -- Chris Coulson <email address hidden> Sun, 24 Jan 2010 18:53:10 +0000

[Martin Pitt (pitti)]

Accepted gnome-power-manager into karmic-proposed, the package will build now and be available in a few hours. Please test and give feedback here. See https://wiki.ubuntu.com/Testing/EnableProposed for documentation how to enable and use -proposed. Thank you in advance!

[Chris Conway (cconway)]

Just install 2.29.1-0ubuntu2 from ubuntu-proposed and rebooted. I've still got the same problem on my Inspiron 15n.

[Chow Loong Jin (hyperair)]

On Monday 25,January,2010 11:34 PM, Chris Conway wrote:
> Just install 2.29.1-0ubuntu2 from ubuntu-proposed and rebooted. I've
> still got the same problem on my Inspiron 15n.
>
Please test 2.28.1-0ubuntu1.3~hyper1 from my PPA at
https://launchpad.net/~hyperair/+archive/bugfixes. If that build is fine, then
I'll get it uploaded.

--
Kind regards,
Chow Loong Jin (GPG: 0x8F02A411)
Ubuntu Contributing Developer

[Novecento (alexander-doelz)]

I have same issue on a Acer Aspire One 150 w/ Ubuntu 9.10 UNR.
The new gnome-power-manager from comment #103 doesn't fix the problem here.
Comment #47 does fix the problem.

[Chow Loong Jin (hyperair)]

What about #105? I already know #103 doesn't fix the problem, after making two miserable uploads which don't fix the bug one after another.

[Chris Conway (cconway)]

Chow, I installed the package from your PPA, but it doesn't fix my problem.

On Mon, Jan 25, 2010 at 4:48 PM, Chow Loong Jin <email address hidden> wrote:
> What about #105? I already know #103 doesn't fix the problem, after
> making two miserable uploads which don't fix the bug one after another.
>
> --
> Computer suspends immediately after resuming if power is unplugged while suspended
> https://bugs.launchpad.net/bugs/425411
> You received this bug notification because you are a direct subscriber
> of the bug.
>

[Chow Loong Jin (hyperair)]

On Tuesday 26,January,2010 11:53 AM, Chris Conway wrote:
> Chow, I installed the package from your PPA, but it doesn't fix my
> problem.
I forgot to mention, you have to kill gnome-power-manager and start it again. If
the problem still persists after that, please post the output of
`gnome-power-manager --verbose`.

--
Kind regards,
Chow Loong Jin (GPG: 0x8F02A411)
Ubuntu Contributing Developer

[Novecento (alexander-doelz)]

I've tried #105 now but it doesn't fix it.

[Chris Conway (cconway)]

Same here, after a full reboot. A log is attached.

[Chow Loong Jin (hyperair)]

On Tuesday 26,January,2010 09:34 PM, Chris Conway wrote:
> Same here, after a full reboot. A log is attached.
>
> ** Attachment added: "gnome-power-manager log for: lid close, remove AC, lid open"
> http://launchpadlibrarian.net/38407533/gpm.log
>
Please get the libdevkit-power-gobject1 package from the same PPA before
testing. gnome-power-manager needs to be restarted (killall gnome-power-manager,
and run gnome-power-manager in a run dialog) after you've installed the new
gnome-power-manager and libdevkit-power-gobject1.

--
Kind regards,
Chow Loong Jin (GPG: 0x8F02A411)
Ubuntu Contributing Developer

[Chris Conway (cconway)]

Chow, That worked!

On Tue, Jan 26, 2010 at 8:48 AM, Chow Loong Jin <email address hidden> wrote:
> On Tuesday 26,January,2010 09:34 PM, Chris Conway wrote:
>> Same here, after a full reboot. A log is attached.
>>
>> ** Attachment added: "gnome-power-manager log for: lid close, remove AC, lid open"
>>    http://launchpadlibrarian.net/38407533/gpm.log
>>
> Please get the libdevkit-power-gobject1 package from the same PPA before
> testing. gnome-power-manager needs to be restarted (killall gnome-power-manager,
> and run gnome-power-manager in a run dialog) after you've installed the new
> gnome-power-manager and libdevkit-power-gobject1.
>
> --
> Kind regards,
> Chow Loong Jin (GPG: 0x8F02A411)
> Ubuntu Contributing Developer
>
>
> ** Changed in: devicekit-power (Ubuntu Karmic)
>       Status: Fix Released => In Progress
>
> ** Changed in: devicekit-power (Ubuntu Karmic)
>     Assignee: Martin Pitt (pitti) => Chow Loong Jin (hyperair)
>
> ** Changed in: devicekit-power (Ubuntu Lucid)
>       Status: Fix Released => In Progress
>
> ** Changed in: devicekit-power (Ubuntu Lucid)
>     Assignee: Martin Pitt (pitti) => Chow Loong Jin (hyperair)
>
> ** Changed in: gnome-power-manager (Ubuntu Karmic)
>       Status: Fix Committed => In Progress
>
> ** Changed in: gnome-power-manager (Ubuntu Lucid)
>       Status: Fix Released => In Progress
>
> ** Changed in: gnome-power-manager (Ubuntu Lucid)
>     Assignee: Martin Pitt (pitti) => Chow Loong Jin (hyperair)
>
> --
> Computer suspends immediately after resuming if power is unplugged while suspended
> https://bugs.launchpad.net/bugs/425411
> You received this bug notification because you are a direct subscriber
> of the bug.
>

[Launchpad Janitor (janitor)]

This bug was fixed in the package devicekit-power - 014-3

---------------
devicekit-power (014-3) unstable; urgency=low

  * Add 02-dkpclient-singleton.patch: Make DkpClient a singleton to avoid some
    weird race conditions. Patch taken from upstream git. (Closes: #567021,
    LP: #425411)
 -- Martin Pitt <email address hidden> Tue, 26 Jan 2010 20:26:03 +0100

[Martin Pitt (pitti)]

Accepted devicekit-power into karmic-proposed, the package will build now and be available in a few hours. Please test and give feedback here. See https://wiki.ubuntu.com/Testing/EnableProposed for documentation how to enable and use -proposed. Thank you in advance!

[Bastanteroma (bastanteroma)]

This may new a bug, but:

Since upgrading to Lucid a week ago, resuming by opening the lid leaves me with a blank screen. Everything else appears to be working (or music, internet connection, at least).

Suspending through the menu and then resuming with the power button works fine, as does suspending from tty1.

[Chow Loong Jin (hyperair)]

On Thursday 28,January,2010 06:41 AM, Bastanteroma wrote:
> This may new a bug, but:
>
> Since upgrading to Lucid a week ago, resuming by opening the lid leaves
> me with a blank screen. Everything else appears to be working (or music,
> internet connection, at least).
>
> Suspending through the menu and then resuming with the power button
> works fine, as does suspending from tty1.
>
Yes, that is a new bug. Please file it separately.

--
Kind regards,
Chow Loong Jin (GPG: 0x8F02A411)
Ubuntu Contributing Developer

[Chow Loong Jin (hyperair)]

And here's the final gnome-power-manager debdiff that will fix all double-suspend issues for Karmic!

[ethanay (ethan-y-us)]

This message doesn't directly contribute to the bug fix. However, as
an Ubuntu (and ex-M$) user (and unofficial Ubuntu family sysadmin), I
just want to thank you folks and others for the work you are doing in
making Ubuntu an increasingly solid distro. The level of transparency
is refreshing and amazing.

Thanks!

[Chow Loong Jin (hyperair)]

The next upload of gnome-power-manager to Debian will include this fix, then it'll be synced over.

[Martin Pitt (pitti)]

Thanks! I sponsored/accepted the new g-p-m into karmic-proposed. Please test again.

[Novecento (alexander-doelz)]

Hi Martin, It worked!

[glass.dimly (jmjohn)]

Installing devicekit-power from karmic proposed as suggested worked perfectly for me.

Make sure to pin and add the proposed repositories as suggested here:

https://wiki.ubuntu.com/Testing/EnableProposed

Then,

sudo aptitude install devicekit-power/karmic-proposed

Thanks for your work!

[Martin Ling (martin-launchpad)]

I have installed gnome-power-manager and devicekit-power from karmic-proposed and rebooted, but am still seeing this problem on a Dell Studio XPS 1645.

[Chow Loong Jin (hyperair)]

On Tuesday 02,February,2010 09:35 PM, Martin Ling wrote:
> I have installed gnome-power-manager and devicekit-power from karmic-
> proposed and rebooted, but am still seeing this problem on a Dell Studio
> XPS 1645.
>
There were actually multiple variants of this problem. I'm quite sure I got the
first two. Could you explain exactly what you do that can reproduce this bug?
Also please kill gnome-power-manager and post the output of `gnome-power-manager
--verbose` from a double-suspend round.

--
Kind regards,
Chow Loong Jin (GPG: 0x8F02A411)
Ubuntu Contributing Developer

[Launchpad Janitor (janitor)]

This bug was fixed in the package devicekit-power - 011-1ubuntu2

---------------
devicekit-power (011-1ubuntu2) karmic-proposed; urgency=low

  * debian/patches/02-dkpclient-singleton.patch:
    - Make DkpClient a singleton to avoid strange races that can
      arise. Patch pulled from upstream Git. (LP: #425411)
 -- Chow Loong Jin <email address hidden> Wed, 27 Jan 2010 02:02:00 +0800

[Launchpad Janitor (janitor)]

This bug was fixed in the package gnome-power-manager - 2.28.1-0ubuntu1.3

---------------
gnome-power-manager (2.28.1-0ubuntu1.3) karmic-proposed; urgency=low

  * debian/patches/09-fix-double-suspend.patch:
    + Inhibit consolekit events after resuming from suspend to fix a race
      condition regarding pm-utils, uswsusp, and chvt (LP: #425411)

gnome-power-manager (2.28.1-0ubuntu1.2) karmic-proposed; urgency=low

  * debian/patches/09-fix-double-suspend.patch:
    + Fix gpm_button_is_lid_closed to return an updated value from
      DeviceKit-Power instead of its own locally stored value which may be
      outdated (LP: #425411)

gnome-power-manager (2.28.1-0ubuntu1.1) karmic-proposed; urgency=low

  * debian/patches/09-fix-double-suspend.patch:
    + Manually check lid status in gpm_manager_client_changed_cb. Fixes
      double-suspend bug (LP: #425411)
 -- Chow Loong Jin <email address hidden> Thu, 28 Jan 2010 08:26:09 +0800

[Martin Ling (martin-launchpad)]

> There were actually multiple variants of this problem. I'm quite sure I got the
> first two. Could you explain exactly what you do that can reproduce this bug?

With AC power connected, I close the lid and the machine suspends. I then remove the power cable and open the lid. The machine resumes, but then immediately suspends again. If I resume it again by pressing the power button, it stays resumed.

> Also please kill gnome-power-manager and post the output of `gnome-power-manager
> --verbose` from a double-suspend round.

Attached.

[Chow Loong Jin (hyperair)]

On Wednesday 03,February,2010 01:14 AM, Martin Ling wrote:
>> There were actually multiple variants of this problem. I'm quite sure I got the
>> first two. Could you explain exactly what you do that can reproduce this bug?
>
> With AC power connected, I close the lid and the machine suspends. I
> then remove the power cable and open the lid. The machine resumes, but
> then immediately suspends again. If I resume it again by pressing the
> power button, it stays resumed.
>
>> Also please kill gnome-power-manager and post the output of `gnome-power-manager
>> --verbose` from a double-suspend round.

Did you install the updated libdevkit-power-gobject1 package (it comes from the
devicekit-power source package)? That's the particular package that contains the
fix for your problem. You'll need to restart gnome-power-manager after
installing it, of course.

--
Kind regards,
Chow Loong Jin (GPG: 0x8F02A411)
Ubuntu Contributing Developer

[Tomas Pospisek (tpo-deb)]

I am confirmin that after reverting the configuration change from

  https://bugs.launchpad.net/ubuntu/+source/devicekit-power/+bug/425411/comments/47

upgrading to:

linux-image-generic 2.6.31.19.32
devicekit-power 011-1ubuntu2
gnome-power-manager 2.28.1-0ubuntu1.3

and rebooting, suspending and waking up is working as it should. I.e.:

1) I close the lid while AC is plugged in -> the laptop suspends correctly
2) while suspended I unplug AC
3) I open the lid -> the laptop wakes up
4) and that's it, no more re-suspending, the laptop stays up awake

(HP Compaq 6710b here)

Thanks to Chow Loong Jin and Marin Pitt fixing and trackingall this stuff! Horray!

[Krister Swenson (thekswenson)]

The fix seems to work on the dell xps m1330.

[Launchpad Janitor (janitor)]

This bug was fixed in the package gnome-power-manager - 2.29.2-0ubuntu1

---------------
gnome-power-manager (2.29.2-0ubuntu1) lucid; urgency=low

  * New upstream release:
    - Move the power management preferences into the hardware section of the
      control center.
    - Use the name of 'Power' for the power management preferences capplet
    - Don't rely on the cached value of the lid status, to fix a double
      suspend issue (LP: #425411)
    - Add a flag to inhibit consolekit events just after we resumed
    - Don't automatically suspend if there are suspend inhibits
    - Do not exit if hal is not available
    - Only connect to HAL if there is no xrandr backlight hardware
    - Fix compile when using an ld that defaults to --as-needed
    - Enable the help action in gnome-power-statistics. Fixes #607005
    - Ensure the window is realized before we invalidate it. Fixes #604918
    - Don't show the user a sleep failed link pointing to the quirk site
    - Show the device name even when using UPower
    - Translation updates.
  * Drop patches accepted upstream:
    + 03-run-without-hal.patch
    + 04-dont-connect-to-hal-with-xrandr.patch
    + 09-fix-double-suspend.patch
  * Regenerate 90-autotools.patch.
 -- Martin Pitt <email address hidden> Fri, 05 Feb 2010 13:28:09 -0800

[glass.dimly (jmjohn)]

Fix worked for me! Thanks for all the work!

[David Hardstone (dhardstone)]

+ my thanks!!

[cement_head (andorjkiss)]

Hi,

  At some point in the last few (three?) weeks this bug was fixed for me using an HP6910p laptop (I posted my original bug here: https://bugs.launchpad.net/ubuntu/+source/linux/+bug/505479 ).

Thanks,
CH

[paul stoop (paul-stoop-nl)]

What does it mean: "Fix released"? I still have this problem, every day twice, as i use my laptop while commuting by train and charging at home and at work!

[Chow Loong Jin (hyperair)]

On Wednesday 10,February,2010 10:46 PM, paul stoop wrote:
> What does it mean: "Fix released"? I still have this problem, every day
> twice, as i use my laptop while commuting by train and charging at home
> and at work!
>
As shown by the previous few comments, this issue is fixed for most people. If
you can still see this problem, please check the versions of your
gnome-power-manager and libdevkit-power-gobject1 packages. You need
2.28.1-0ubuntu1.3 of gnome-power-manager and 011-1ubuntu2 of
libdevkit-power-gobject1.

You can check these by running this in a terminal:
apt-cache policy gnome-power-manager libdevkit-power-gobject1

If you do have the required versions mentioned above, then the bug fix was not
complete, and we will need you to follow the steps detailed in comments #13 and
#14 of this bug report.

--
Kind regards,
Chow Loong Jin (GPG: 0x8F02A411)
Ubuntu Contributing Developer

[Chris Conway (cconway)]

Paul, the bug is (thought to be) fixed in devicekit version 011-1ubuntu2 and gnome-power-manager version 2.28.1-0ubuntu1.3, both available in karmic-updates. Do you have the karmic-updates repository enabled? If so, you should post a follow-up more details on your system and the output from "gnome-power-manager --verbose" on a suspend, resume, re-suspend cycle.

[Erik Andrén (erik-andren)]

2010/2/10 paul stoop <email address hidden>:
> What does it mean: "Fix released"? I still have this problem, every day
> twice, as i use my laptop while commuting by train and charging at home
> and at work!
>

Paul,
Please make sure you've enabled the proposed repository.

Best regards,
Erik

> --
> Computer suspends immediately after resuming if power is unplugged while suspended
> https://bugs.launchpad.net/bugs/425411
> You received this bug notification because you are a direct subscriber
> of a duplicate bug.
>

[Celsius (celsius-netbel)]

Thanks ! The fix works on my dell latitude e6500

I had open bug 496525 which is a duplicate of this bug. Is there anything I have to do to close that bug or mark it as fixed ?

[paul stoop (paul-stoop-nl)]

Thanks all. My T61p is up to date and I can' t reproduce the bug right now. I'll pay close attention the next days is it shows up again. I think it only happens/happened after a significant battery discharge. cheers, Paul.

[David Tombs (dgtombs)]

@Celsius: Nope, you don't have to do anything. Thanks for the thought, though!

[paul stoop (paul-stoop-nl)]

After 10 days of being aware that this bug was fixed, I can confirm that it definitely seems to have been expelled from my laptop (lenovo T61p) as well. Congretulations, this was the last major pain in my neck since Karmic.
-p

[Mark (mark-wege)]

i can confirm it working with kde too.

[ethanay (ethan-y-us)]

Are there other "immediately resuspends" bugs that are currently open?

The power state problem this bug describes is now fixed on my XPS m1330, but I am still experiencing immediate resuspends on occasion. It seems to be after

a. locking or blanking the screen, closing the lid (triggers suspend), or
b. letting the computer suspend itself and

attempting resume after several hours have passed.

[Kieran Hogg (xerosis)]

I'm still seeing this on an up to date Kubuntu Karmic, without devicekit or gnome-power-manager installed. is the only solution to install these packages? Neither are installed by default on Kubuntu.

[Chow Loong Jin (hyperair)]

On Monday 08,March,2010 04:07 AM, Kieran Hogg wrote:
> I'm still seeing this on an up to date Kubuntu Karmic, without devicekit
> or gnome-power-manager installed. is the only solution to install these
> packages? Neither are installed by default on Kubuntu.
>
This bug belongs to gnome-power-manager/devicekit-power, and has been fixed.
What you are seeing is a different bug with similar symptoms. Neither
gnome-power-manager nor devicekit-power will fix your problem on Kubuntu.

--
Kind regards,
Chow Loong Jin (GPG: 0x8F02A411)
Ubuntu Developer

[ethanay (ethan-y-us)]

This was fixed with me for a while, but I'm seeing it again on an up-to-date Karmic laptop (Dell XPS m1330)...

1. Occasional, seemingly-random suspend immediately after resume
2. Suspend after resume if power state changes during suspend (e.g., laptop is unplugged)

both seem to happen after a threshold time is exceeded -- can't reproduce with a suspend and immediate resume. All cases involve laptop lid close triggering suspend.

[David Tombs (dgtombs)]

ethanay: I recommend reporting a new bug with "ubuntu-bug gnome-power-manager". This one has been fixed, you probably have a different issue.