acpid steals powerbutton events from powerdevil

[mrvanes]

Binary package hint: acpid

On my laptop (Dell Vostro 1400) when acpid is running, powerdevil doesn't see the powerbutton event. This means that I don't get the 'logout/restart/poweroff' dialog, but an immediate powerdown if I press the powerbutton. Ubuntu release: Jaunty.

Don't know if this a duplicate of https://bugs.launchpad.net/gnome-power/+bug/57872 but that one hasn't been touched for a while and is about gnome-power-manager instead of powerdevil (kde).

[Chris Coulson (chrisccoulson)]

Thank you for your bug report. Could you please do the following:

1) Stop acpid (sudo /etc/init.d/acpid stop).
2) Run "sudo cat /proc/acpi/event"
3) Press your power button and paste any terminal output it to this bug report.

Thanks

[mrvanes (mrvanes)]

# cat /proc/acpi/event
button/power PBTN 00000080 00000001
^C

There's a minor change in the original bug since recent updates:
I *do* see the logout/restart/poweroff dialog for a very short time now, I just don't have the choice...

So powerdevil probably sees the event, but acpid takes over after that.

[Chris Coulson (chrisccoulson)]

Thanks, that looks like an acpi-support bug

[Chris Coulson (chrisccoulson)]

I'll take a look at this

[Chris Coulson (chrisccoulson)]

Ok, someone else might have to take a look at it instead. I thought it would be a trivial fix, but I don't know enough about KDE to implement it ;)

[Michael Marley (mamarley)]

I know how to write bash scripts pretty well, so I will take a look at this.

[Michael Marley (mamarley)]

OK, I think I have this working now. Since (at least in KDE4) PowerDevil does not have its own daemon, there is no way to check to see if PowerDevil is running. Instead, my modified script also checks to see if kded4 (which I believe is specific to KDE4) is running, and if it is, the script exits, allowing PowerDevil to handle the event. I attached the script. It would be nice if somebody could also test it on a Gnome system to make sure it doesn't interfere with that.

[mrvanes (mrvanes)]

I can confirm that it fixes the problem! Thx.

[Michael Marley (mamarley)]

OK, this seems to be fixed now. Could someone with upload privileges upload it, please?

[mrvanes (mrvanes)]

Recent updates in Jaunty overwrote my manual changes (apparently) and the problem is back. Could this fix please be uploaded into the repository please?

[Michael Marley (mamarley)]

I would if I could, but I can't. Sorry. We will have to wait for someone who has the privileges to upload it.

[mrvanes (mrvanes)]

I understand, but how are they notified?

[Michael Marley (mamarley)]

I can try sending a message to Chris Coulson, who was previously working on the bug. I think he has upload privileges.

[Tom Chiverton (bugs-launchpad-net-falkensweb)]

I made the same patch in the 8.10 version of this bug, but there's been no sign anyone cares about fixing that either.

[Tom Chiverton (bugs-launchpad-net-falkensweb)]

Re, attached replacement script.
Based on this test not printing 'kde':
if pidof gnome-power-manager kpowersave > /dev/null || (pidof dcopserver > /dev/null && test -x /usr/bin/dcop && /usr/bin/dcop kded kded loadedModules | grep -q klaptopdaemon | pidof kded4 > /dev/null) ; then echo "kde"; fi
I think maybe the test should be like this (note brackets and double bars):
if pidof gnome-power-manager kpowersave > /dev/null ||
  (pidof dcopserver > /dev/null && test -x /usr/bin/dcop && /usr/bin/dcop kded kded loadedModules | grep -q klaptopdaemon) || pidof kded4 > /dev/null ; then
    exit
fi

[emikaadeo (emikaadeo)]

Yes, this script really solved the situation here.
Please devs, upload this.

[Malte S. Stretz (mss)]

This is the same as bug 268734 and both have a (slightly different) patch. Not sure which to mark as a dup of the other :)

[Malte S. Stretz (mss)]

Here is the above change as a patch (from Tom Chiverton on bug 288639).