linux-kernel - Kernel doesn't generate a "Discharging" battery event after unplugging AC power on Full battery

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [day] [month] [year] [list]

Message-ID: <119632.18101.qm@web26701.mail.ukl.yahoo.com>
Date:	Thu, 8 Oct 2009 20:30:58 +0000 (GMT)
From:	James Gallagher <jbgallagher2000@...oo.co.uk>
To:	linux-kernel@...r.kernel.org
Subject: Kernel doesn't generate a "Discharging" battery event after unplugging AC power on Full battery

(This applies to kernel 2.6.31.2 and older)

If I unplug AC power on an Acer Aspire One with fully charged battery, the kernel/udev events generated indicate the only change as POWER_SUPPLY_ONLINE=0 (from 1), the battery POWER_SUPPLY_STATUS is still indicating FULL, and no change in the battery environment, even though the event is labelled ACTION=change on both the mains (ACAD) and battery (BAT1) (see attached log of the 4 kernel/udev events generated after unplugging)

This causes DeviceKit-power to suspend polling, so the power manager never records the subsequent discharge of the battery (the battery state in /sys updates to "Discharging" about 10 seconds later). The redhat guys on bugzilla are insisting it's a kernel/udev bug, not a DeviceKit-power bug.

This also occurs on a Dell Inspiron 1520, but less predictably - ie sometimes the battery STATE=Discharging and sometimes =FULL.

So Questions?

1. Why does the kernel generate a change event for the battery when there is no change?
2. Why does the battery STATE indicate Full rather than Discharging?
3. Why does it take ~10 seconds for the battery state to change to "Discharging" in /sys, and why is there no accompanying udev event to indicate the status change?

See https://bugzilla.redhat.com/show_bug.cgi?id=521874 (bug has been re-assigned from DeviceKit-power to Kernel)

James Gallagher

(can replies be cc'ed to me, cheers)

View attachment "udev monitor.txt" of type "text/plain" (2162 bytes)