hybrid-graphics-linux team mailing list archive

[Dawid Lorenz <dawid@xxxxxxxxx>]

On 8 February 2011 22:09, Dawid Lorenz <dawid@xxxxxxxxx> wrote:

> I can confirm (after repeating the procedure several times) that there is
>> an improvement of 200mA on the battery discharging rate (reported by acpi)
>> with an scatter of 30 or 20 mA.
>> It would be great to confirm if this is the power consumption of the
>> nvidia 310m.
>>
>
> Thanks for performing these tests. I will need to take another look into
> the case and see if I get similar difference. I don't have much time for
> detailed testing lately...


Ha! I finally found the reason why I've initially had an impression that I
get no/little difference in battery life after making successful acpi call.
:)

It turned out that after each suspend/resume cycle (which I do quite a lot
and virtually never turn my laptop completely off, actually) acpi call has
to be made again in order to get battery life gain. So I have crafted quick
script to execute on each resume:

/etc/pm/sleep.d/05-nvdown

#!/bin/sh

case "$1" in
        resume|thaw)
            /etc/init.d/nvdown stop
            ;;
        *)
            ;;
esac

exit $?

(Note that I'm using Fedora, so YMMV)

And now after each suspend/resume cycle I get consistent rate of approx.
1000-1100 mA, instead of 1400-1500 mA. Yeah! :)

-- 
Dawid 'evad' Lorenz * http://dawid.lorenz.co

null:// I haven't lost my mind - it's backed up on disk somewhere