upower gives incorrect data/misleading data

[am]

Binary package hint: upower

I am using lucid. I have a dell inspiron 6000

upower gives data that is inconsistent with acpi.

 > acpi
Battery 0: Charging, 27%, 01:25:50 until charged

this agrees with the bios, so I believe it to be correct.

> upower --dump
Device: /org/freedesktop/UPower/devices/line_power_AC
  native-path: /sys/devices/LNXSYSTM:00/LNXSYBUS:00/ACPI0003:00/power_supply/AC
  power supply: yes
  updated: Tue Jun 15 22:34:20 2010 (754 seconds ago)
  has history: no
  has statistics: no
  line-power
    online: yes

Device: /org/freedesktop/UPower/devices/battery_BAT0
  native-path: /sys/devices/LNXSYSTM:00/LNXSYBUS:00/PNP0C0A:00/power_supply/BAT0
  vendor: Sanyo
  model: DELL F51267
  serial: 1410
  power supply: yes
  updated: Tue Jun 15 22:46:25 2010 (29 seconds ago)
  has history: yes
  has statistics: yes
  battery
    present: yes
    rechargeable: yes
    state: charging
    energy: 14.8962 Wh
    energy-empty: 0 Wh
    energy-full: 79.92 Wh
    energy-full-design: 79.92 Wh
    energy-rate: 26.3514 W
    voltage: 11.64 V
    time to full: 2.5 hours
    percentage: 18.6389%
    capacity: 67.375%
    technology: lithium-ion
  History (charge):
    1276638385 18.639 charging
    1276638355 18.347 charging
    1276638325 18.042 charging
  History (rate):
    1276638385 26.351 charging
    1276638355 29.393 charging
    1276638325 27.728 charging

Daemon:
  daemon-version: 0.9.1
  can-suspend: yes
  can-hibernate yes
  on-battery: no
  on-low-battery: no
  lid-is-closed: no
  lid-is-present: yes

the incorrect stat is percentage. Note that it is 18%, i.e. different from what acpi reports. this battery is an older battery and only holds about 65% of the design capacity when full. So, energy-full seems to be wrong as well. I don't know what capacity is, but as I said the last full capacity is about 2/3 of the design capacity, so it could be that. In fact, 18%/67%= 27% agrees with acpi, so that is probably what capacity means.

Anyways, this is causing my gnome-power-manager to report the wrong charge as well.

[David Mack (barid42)]

Something similar is happening to my computer. Not only is the charge information inconsistent, the reporting technology of the battery also changes inexplicably from Li-ion to Nimh. I believe, in my case, these issues may be related. I have been looking into writing a udev rule to force the correct battery type, but I don't know if this will correct the charge information.

System: HP Pavilion dv5-1125nr
OS: Ubuntu Lucid Lynx 10.04, all updates applied (6-16-10)

Dave

[am (juniper1982)]

I didn't notice if mine gives inconsistent reporting of technology, but I can check. Yes, my impression is that upower is a bit buggy.

[David Mack (barid42)]

I've been digging further. It appears my issue may not be with upower.

First, my acpi data is consistent with upower --dump. Percentages match, etc.

Second, I have found notes from other users (even those using winblows) about battery information being incorrect with my BIOS.

Therefore, I believe my problem is with HP, not with upower. I recently updated to the latest (allegedly greatest) BIOS firmware version F.38, and my problems got worse. I assume that is the issue.

Sorry to put the blame where it didn't belong.

Also on a side note, I was able to create a udev rule to force the battery type to Li-ion, and that did not solve the problem, so it appears the BIOS is the root cause.

Next time, I'm buying a computer with Ubuntu out of the box. I've had it with shoehorning a decent OS into a crappy box.

Thanks for everything,

Dave

[Dmitry Gulevich (mitrg)]

I am on Alienware M15X, Ubuntu 10.10. UPower does not register that the lid closed after first suspend/resume, while acpi does:

> yes "upower --dump | grep lid-is-closed; cat /proc/acpi/button/lid/*/state; sleep 2" | sh

  lid-is-closed: yes
state: closed
  lid-is-closed: yes
state: closed

After suspend/resume:

  lid-is-closed: no
state: closed
  lid-is-closed: no
state: closed

acpi correctly says "state: closed", while upower says "lid-is-closed: no".

Tried to dig for 4 hours, but could not solve this bug...

[Launchpad Janitor (janitor)]

Status changed to 'Confirmed' because the bug affects multiple users.

[monochromec (monochromec)]

Same thing 12.10.

acpi shows 43% full, but according to upower the battery should be almost empty as "upower --dump" reports a percentage of 3.8%.