Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

Bug#462723: hal does not give actual values of battery states with 2.6.24

0 views
Skip to first unread message

Didier Raboud

unread,
Jan 26, 2008, 9:00:13 PM1/26/08
to
Package: hal
Version: 0.5.10-5
Severity: important

--- Please enter the report below this line. ---
Hi,

I think that hal is not reporting what it should about battery states (at
least).

Testcase :
------------------------------------
$ date +%H:%M:%S &&
hal-get-property --udi /org/freedesktop/Hal/devices/computer_power_supply_0 --key
battery.charge_level.percentage
02:06:11
25
didier@Tamino:~$ sudo /etc/init.d/hal restart
Restarting Hardware abstraction layer: hald.
didier@Tamino:~$ date +%H:%M:%S &&
hal-get-property --udi /org/freedesktop/Hal/devices/computer_power_supply_0 --key
battery.charge_level.percentage
02:06:26
33
------------------------------------
So I won 8% of battery charging level in 15 seconds ! I can reproduce even
better results by waiting more.

The thing that makes this bug important is that (at least) kpowersave is based
on these results (current, percentage, rate, ...) and is the one responsible
for getting my computer into suspend when the battery gets too low. It can't
do that if it can't get accurate results...

Please ask if you need details.

Regards,

Didier

--- System information. ---
Architecture: amd64
Kernel: Linux 2.6.24-1-amd64

Debian Release: lenny/sid
700 testing mirror.switch.ch
700 stable mirror.switch.ch
600 unstable mirror.switch.ch
50 unstable mirror.switch.ch
50 unstable debian.netcologne.de
50 testing mirror.switch.ch
50 testing debian.netcologne.de
50 stable mirror.switch.ch
50 stable debian.netcologne.de
50 kernel-dists-trunk kernel-archive.buildserver.net
50 kernel-dists-sid kernel-archive.buildserver.net
50 kernel-dists-etch kernel-archive.buildserver.net
50 experimental mirror.switch.ch
50 experimental debian.netcologne.de

--- Package information. ---
Depends (Version) | Installed
==================================-+-==============
adduser | 3.105
dbus (>= 0.61) | 1.1.2-1
hal-info (>= 20070402) | 20071212-2
libc6 (>= 2.7-1) | 2.7-6
libdbus-1-3 (>= 1.1.1) | 1.1.2-1
libdbus-glib-1-2 (>= 0.74) | 0.74-1
libexpat1 (>= 1.95.8) | 1.95.8-4
libgcc1 (>= 1:4.2.1) | 1:4.3-20080116-1
libglib2.0-0 (>= 2.14.0) | 2.15.3-1
libhal-storage1 | 0.5.10-5
libhal1 (>= 0.5) | 0.5.10-5
libsmbios1 | 0.13.10-1
libstdc++6 (>= 4.2.1) | 4.3-20080116-1
libusb-0.1-4 (>= 2:0.1.12) | 2:0.1.12-9
libvolume-id0 (>= 0.113) | 0.114-2
lsb-base | 3.1-24
mount (>= 2.13) | 2.13-8
pciutils | 1:2.2.4~pre4-1
pm-utils | 0.99.2-3
udev (>= 0.065) | 0.114-2
usbutils | 0.73-5

signature.asc

Michael Biebl

unread,
Jan 26, 2008, 9:10:11 PM1/26/08
to
Didier Raboud wrote:
> Package: hal
> Version: 0.5.10-5
> Severity: important
>
> --- Please enter the report below this line. ---
> Hi,
>
> I think that hal is not reporting what it should about battery states (at
> least).
>

This is partly related to #455198 and the new power sysfs interface in
kernel 2.6.24.
I bet you have
CONFIG_ACPI_SYSFS_POWER=y
and
CONFIG_ACPI_PROCFS_POWER=n

Until this new sysfs interface is stable and fully functional within hal
I'd suggest to use
CONFIG_ACPI_SYSFS_POWER=n
CONFIG_ACPI_PROCFS_POWER=y

This is obviously not the real fix, but better than broken battery
values atm.

Cheers,
Michael

--
Why is it that all of the instruments seeking intelligent life in the
universe are pointed away from Earth?

signature.asc

Didier Raboud

unread,
Jan 26, 2008, 9:40:08 PM1/26/08
to
Le dimanche, 27 janvier 2008 03.01:25 Michael Biebl, vous avez écrit :
> Didier Raboud wrote:
> > Package: hal
> > Version: 0.5.10-5
> > Severity: important
> >
> > --- Please enter the report below this line. ---
> > Hi,
> >
> > I think that hal is not reporting what it should about battery states (at
> > least).
>
> This is partly related to #455198 and the new power sysfs interface in
> kernel 2.6.24.
> I bet you have
> CONFIG_ACPI_SYSFS_POWER=y
> and
> CONFIG_ACPI_PROCFS_POWER=n
>
> Until this new sysfs interface is stable and fully functional within hal
> I'd suggest to use
> CONFIG_ACPI_SYSFS_POWER=n
> CONFIG_ACPI_PROCFS_POWER=y
>
> This is obviously not the real fix, but better than broken battery
> values atm.
>
> Cheers,
> Michael

Well. The thing is that I'm using the kernel from debian-kernel team with this
entry in latest changelog :

linux-2.6 (2.6.24-1~experimental.1) UNRELEASED; urgency=low
(...)
  * [amd64, i386]: Enable ACPI_SYSFS_POWER and disable ACPI_PROCFS_POWER.

So this is visibly the default behavior which is about to land in
experimental.

This is still a bug in hal though. :)

Regards,

Didier

signature.asc

Michael Biebl

unread,
Jan 26, 2008, 9:50:08 PM1/26/08
to

Depends. A lot of information (on my system) is missing in the new /sys
interface (that's also why I prefer proc atm). Also, what you describe
is simply hal reporting the values provided by the kernel. So it might
actually be a kernel issue after all.

signature.asc

Mike Hommey

unread,
Jan 28, 2008, 2:10:12 AM1/28/08
to
On Sun, Jan 27, 2008 at 03:42:46AM +0100, Michael Biebl wrote:
>> This is still a bug in hal though. :)
>
> Depends. A lot of information (on my system) is missing in the new /sys
> interface (that's also why I prefer proc atm). Also, what you describe
> is simply hal reporting the values provided by the kernel. So it might
> actually be a kernel issue after all.

On mine, there is more than enough to have the current discharge rate
and to know the battery is discharging, yet hal doesn't know.

Mike

--
To UNSUBSCRIBE, email to debian-bugs-...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listm...@lists.debian.org

0 new messages