From: Thomas R. <tr...@su...> - 2005-02-09 18:57:07
|
Lorenzo Colitti wrote: > Hi, > > I am having problems with battery monitoring and sleep mode: if I plug > in or unplug the battery while the machine is asleep, when it wakes up > the change has not been registered and no further changes are logged. > > This means that I suspend the machine while on AC power and wake it up > on battery power, there is no way for me to keep track of how much > charge is left, which is a problem! > > As an example, I started with init=/bin/sh, with no battery and AC > plugged in, and did the following: > >> sh-3.00# modprobe battery >> ACPI: Battery Slot [C138] (battery absent) >> ACPI: Battery Slot [C137] (battery absent) >> sh-3.00# modprobe ac >> ACPI: AC Adapter [C135] (on-line) >> sh-3.00# echo 3 > /proc/acpi/sleep > > > While the machine was asleep, I plugged in the battery and removed AC > power, then turned the machine on. The kernel didn't notice that the > battery had been plugged in: > >> sh-3.00# cat /proc/acpi/battery/C138/info present: no >> sh-3.00# cat /proc/acpi/battery/C137/info >> present: no >> sh-3.00# cat /proc/acpi/ac_adapter/C135/state state: >> off-line > > > Unplugging and reinserting the battery and/or reloading the battery > module didn't help. :( > > The machine is an HP NC6000 on which ACPI mostly works (decompiling and > recompiling the DSDT yields only the warning "Unknown reserved name > (_WDG)"). I have ACPI debugging enabled in the kernel, but I don't know > how to use it. > > Any ideas? > I expect that your _WAK function does not work properly. Does S3 invoke the _WAK function after wake up? I remember some things like that concerning suspend to disk (no buttons, wrong AC/battery states), that vanished when using platform mode. The main differnce to shutdown mode is (correct me if I am wrong), that the BIOS _WAK function is called and the devices get initialised again. Thomas |