ACPI -correct without AC- measures a power of ~720 Watts when AC is plugged in

[Peter Waller]

I'm on gentoo, and I'm using gnome-extra/gnome-power-manager-2.27.92.

When plugged in, /proc/acpi/battery/BAT0/state reports:
  charging state:          charging
  present rate:            64645 mA

After unplugging AC:
  charging state:          discharging
  present rate:            4168 mA

And for a good minute after unplugging, the rate is still ridiculously high. This is making gnome-power-manager incorrectly report time-to-empty and time-to-full by a significant margin, enough to make the estimates useless (2 minutes to completely fill an empty battery?).

I guess I need to talk to the kernel ACPI guys to get it fixed there, but I was wondering if it would be possible to have an option in GPM to ignore these obviously ridiculous values.

Thoughts?

Thanks,

- Peter

_______________________________________________
gnome-power-manager-list mailing list
gnome-power-manager-list@...
http://mail.gnome.org/mailman/listinfo/gnome-power-manager-list

[Richard Hughes-6]

On 26/09/2009, Peter Waller <peter.waller@...> wrote:
> This is making gnome-power-manager incorrectly report time-to-empty and
> time-to-full by a significant margin, enough to make the estimates useless
> (2 minutes to completely fill an empty battery?).

Smeels like a kernel bug...

> I guess I need to talk to the kernel ACPI guys to get it fixed there, but I
> was wondering if it would be possible to have an option in GPM to ignore
> these obviously ridiculous values.

Well, it would be a DeviceKit-power feature request. Arguably, we
should just finish the time estimation work in DeviceKit-power and
turn it on by default. Then this sort of bug wouldn't happen.

Richard.
_______________________________________________
gnome-power-manager-list mailing list
gnome-power-manager-list@...
http://mail.gnome.org/mailman/listinfo/gnome-power-manager-list