Bugzilla – Bug 146517
Battery information read wrong in 10.1 Beta 2
Last modified: 2006-02-11 08:24:33 UTC
I have a fresh install of OpenSUSE 10.1 Beta 2. It reads the battery information wrong. SUSE 10.0 works correctly here. I have one battery (should be BAT0) with design capacity of 4400mAh and voltage of 14.4V. This of course makes battery meter in KDE to show wrong information. Here's the information retrieved from /proc/acpi/battery: BAT0/alarm, info, state: present: no BAT1/alarm: alarm: unsupported BAT1/state: present: yes capacity state: ok charging state: charged present rate: unknown remaining capacity: 65280 mAh present voltage: 255 mV BAT1/ present: yes design capacity: 40960 mAh last full capacity: 40975 mAh battery technology: unknown design voltage: 53248 mV design capacity warning: 4097 mAh design capacity low: 2048 mAh capacity granularity 1: 30777 mAh capacity granularity 2: 8 mAh model number: 244II0 serial number: 00002 battery type: LiIon OEM info: Secondary-SK
Pavel, can you look at this one, please?
What kind of machine is that? Can you try to play with ec_burst (or how is it called) options? Does it work okay in 2.6.15?
the option is called ec_intr=[0,1] now IIRC.
Could you also possibly post the output from 10.0 for comparison? If you don't have it installed anymore, then you can either use just a 10.0 kernel or do the dump from a rescue system.
Created attachment 65879 [details] Battery information from working 10.0 I don't have 10.0 installed anymore, but I saved the battery info before upgrading. Here it is. I'll try the options later today.
ec_intr=0,1 did not make any difference. This is a Uniwill N244 laptop. I haven't tried with 2.6.15. Would you like me to?
Better wait a bit, there will come ACPI patches, not for Beta3, but shortly after, I can point you to a kernel then.
sorry, but your tarball only contains empty files.
It seems tar couldn't read the proc files. Sorry. I'm downloading 10.0 live dvd to get you the "correct" lines.
If it worked, I don't think exact values are strictly neccessary. 10.0 live dvd might still be nice in case we want to do experiment with various settings...
is this by chance a HP machine?
This is a Uniwill N244ii0 Intel Pentium-M laptop
This problem has vanished with Linux linux 2.6.16-rc2-git2-3-default #1 Mon Feb 6 23:30:07 UTC 2006 i686 i686 i386 GNU/Linux
ok. So even if we do not know for sure why it works now, we can consider this fixed ;-) If you see something like this again (even if it is spurious), please report. The reason i asked if this was on a hp is that i have seen frequent spurious mis-readings of acpi stuff (battery, temperature, other) on hps, but it might well be that it also happens on other machines.