Hi Norbert, I have the exact same issue on a dell latitude e5510 running the x86_64 kernel. And on a dell latitude 1220 running the i686 kernel. Didn't know about the hack, thanks. Regards, Wim On Tue, May 17, 2011 at 7:26 PM, Norbert Zeh <nzeh@cs.dal.ca> wrote:

Hi folks,

I've run into a weird problem with monitoring the status of my laptop battery. Since I assume I'm doing something wrong, I thought I'd ask for feedback here first before going and filing a bug report. Here goes:

When my machine boots, /sys/class/power_supply/BAT0 and all the files in it do not exist:

$ ls /sys/class/power_supply AC

Since my battery monitor uses this as the source of information, it obviously fails.

Now, as soon as I query the state of the battery through /proc/acpi/battery/BAT0, the missing files under /sys/class/power_supply get created:

$ cat /proc/acpi/battery/BAT0/state present: yes capacity state: ok charging state: discharging present rate: 1265 mA remaining capacity: 4234 mAh present voltage: 11340 mV

$ ls /sys/class/power_supply AC BAT0

This is not just a timeout issue. If I don't query /proc/acpi/battery/BAT0/state, /sys/class/power_supply/BAT0 does not get created even after a few hours of uptime.

Also, after suspending, the files remain present but cannot be read. Querying the battery through the /proc interface once again brings everything back to normal.

Now what's funny about this is that I read online that the /proc/acpi/battery interface is deprecated and that the /sys/... interface should be used instead. Yet, without querying the former, the latter doesn't even exist.

Did anybody run into similar issues? Does anybody have any ideas how to fix this "the proper way"? (My workaround right now is to include "cat /proc/acpi/battery/BAT0/state > /dev/null" in rc.local, but that's obviously a hack.)

Thanks for any pointers you may give.

Cheers, Norbert

PS: Machine is Dell Latitude E6510 running 64-bit ARCH.