guidance-power-manager and kpowersave stop working

[gpothier]

Both guidance-power-manager and kpowersave stop working after some time in Feisty (I am not running both at the same time, I just tried one and then the other). In particular, they stop reacting to the lid close event, which is supposed to suspend my laptop. Killing and restarting guidance of kpowersave temporarily solves the problem, until they stop working again.
It seems that this is not a HAL nor ACPI problem: the lid button events appear in "lshal -m" and in /var/log/acpid. And I can always suspend by pressing the suspend special key.
With guidance, the following lines appears in /var/log/acpid:
8<----
[Sat May 19 19:57:58 2007] received event "button/lid LID 00000080 00000008"
[Sat May 19 19:57:58 2007] notifying client 5026[0:0]
[Sat May 19 19:57:58 2007] notifying client 26392[106:110]
[Sat May 19 19:57:58 2007] notifying client 11430[0:0]
[Sat May 19 19:57:58 2007] executing action "/etc/acpi/fglrx-powermode.sh"
[Sat May 19 19:57:58 2007] BEGIN HANDLER MESSAGES
[Sat May 19 19:57:58 2007] END HANDLER MESSAGES
[Sat May 19 19:57:58 2007] action exited with status 0
[Sat May 19 19:57:58 2007] executing action "/etc/acpi/lid.sh"
[Sat May 19 19:57:58 2007] BEGIN HANDLER MESSAGES
ERROR: Couldn't attach to DCOP server!
[Sat May 19 19:57:58 2007] END HANDLER MESSAGES
[Sat May 19 19:57:58 2007] action exited with status 1
[Sat May 19 19:57:58 2007] completed event "button/lid LID 00000080 00000008"
8<----

With kpowersave nothing noteworthy appears.

How can I help debug that?

[Brian Murray (brian-murray)]

 affects ubuntu/kde-guidance

[Zak B. Elep (zakame)]

Can confirm on Kubuntu Feisty with latest updates as of this writing. Said workaround does not always work for me, however.

[gpothier (gpothier)]

This problem is still present in Gutsy. Is there something I can do to help resolve it?

[ErikZap (eam67)]

I can confirm this, still a problem in Gutsy. Strangely, sometimes suspend to RAM works when I press the lid btn. I noticed that failure of suspend to RAM is accompanied with executing xscreensaver (which does NOT happen when pressing suspend btn (FN F4) or suspending via KDE's logout menu). Also, the screen blanks but not completely if you look carefully.

I can also confirm that the DCOP error does not appear when suspending using the suspend btn/logout menu.

I found this thread http://www.mail-archive.com/newbie%40mandrivalinux.org/msg10825.html where ppl suggest to delete .dcop as a workaround. This is not a durable solutions, but i might try it as a work around.

[ErikZap (eam67)]

BTW, why are there two distinct scripts for lid.sh and sleepbtn.sh ? does that make any sense? shouldn't the same actions be executed?

[Adam Porter (alphapapa)]

I can confirm this on a Dell XPS M1330. Closing the lid the first two times successfully suspends to RAM and resumes. After the first two times, it doesn't suspend, and the DCOP errors appear. Restarting guidance or kpowersave seems to fix it, but only for one or two suspend/resume cycles.

[Adam Porter (alphapapa)]

Also, after the failure, I can suspend from the KDE Log Out dialog and it works.

[Adam Porter (alphapapa)]

Here is part of my acpid log after many suspend/resume attempts.
Notice how sometimes the script returns 1 and doesn't suspend, and
sometimes it returns 0 and works. I'm not completely sure if the DCOP
errors are the problem or not. The script that returns the error code
checks to see if kpowersave or other things are running...strangely,
though, it doesn't check for guidance, as far as I can tell.

Surely this is fixable. Please let me know how I can help debug this.

On Wed, Feb 13, 2008 at 12:30 AM, Adam Porter <email address hidden> wrote:
> I can confirm this on a Dell XPS M1330. Closing the lid the first two
> times successfully suspends to RAM and resumes. After the first two
> times, it doesn't suspend, and the DCOP errors appear. Restarting
> guidance or kpowersave seems to fix it, but only for one or two
> suspend/resume cycles.
>
>
>
> --
> guidance-power-manager and kpowersave stop working
> https://bugs.launchpad.net/bugs/116416
> You received this bug notification because you are a direct subscriber
> of the bug.
>

[Andreas Ntaflos (daff)]

This is still or again a problem in Hardy (Alpha 6, current as of today). When closing the lid of my Thinkpad T41p I get these messages in /var/log/acpid. No need to mention that putting the laptop to sleep doesn't work.

[Thu Mar 13 00:16:50 2008] received event "button/lid LID 00000080 00000002"
[Thu Mar 13 00:16:50 2008] notifying client 5165[107:116]
[Thu Mar 13 00:16:50 2008] notifying client 6194[0:0]
[Thu Mar 13 00:16:50 2008] executing action "/etc/acpi/lid.sh"
[Thu Mar 13 00:16:50 2008] BEGIN HANDLER MESSAGES
ERROR: Couldn't attach to DCOP server!
[Thu Mar 13 00:16:51 2008] END HANDLER MESSAGES
[Thu Mar 13 00:16:51 2008] action exited with status 1
[Thu Mar 13 00:16:51 2008] completed event "button/lid LID 00000080 00000002"

There is no /root/.dcop or any ~/.dcop for that matter. I don't know how to work around this bug or what causes it. I can provide more info if needed, just tell me what.

[Adam Porter (alphapapa)]

This really needs to be fixed before Hardy is released!

On Wed, Mar 12, 2008 at 6:51 PM, Andreas Ntaflos <email address hidden> wrote:
> This is still or again a problem in Hardy (Alpha 6, current as of
> today). When closing the lid of my Thinkpad T41p I get these messages in
> /var/log/acpid. No need to mention that putting the laptop to sleep
> doesn't work.
>
> [Thu Mar 13 00:16:50 2008] received event "button/lid LID 00000080 00000002"
> [Thu Mar 13 00:16:50 2008] notifying client 5165[107:116]
> [Thu Mar 13 00:16:50 2008] notifying client 6194[0:0]
> [Thu Mar 13 00:16:50 2008] executing action "/etc/acpi/lid.sh"
> [Thu Mar 13 00:16:50 2008] BEGIN HANDLER MESSAGES
> ERROR: Couldn't attach to DCOP server!
> [Thu Mar 13 00:16:51 2008] END HANDLER MESSAGES
> [Thu Mar 13 00:16:51 2008] action exited with status 1
> [Thu Mar 13 00:16:51 2008] completed event "button/lid LID 00000080 00000002"
>
> There is no /root/.dcop or any ~/.dcop for that matter. I don't know how
> to work around this bug or what causes it. I can provide more info if
> needed, just tell me what.
>
>
>
> --
> guidance-power-manager and kpowersave stop working
> https://bugs.launchpad.net/bugs/116416
> You received this bug notification because you are a direct subscriber
> of the bug.
>