Re: powerbook g3
what about battery life, hard disk spin down monitor dimming detection of
inserted new batteries etc?
There is much work to be done here. All those who are interested send me
a note and I'll assemble a group together.
regardz
On Tue, 1 Jun 1999, Benjamin Herrenschmidt wrote:
> On Tue, Jun 1, 1999, Josh Huber <huberj@WPI.EDU> wrote:
>
> >> Just out of curiosity does anyone have MacOS X maybe they have power
> >> management there that we can we look at code from?
> >I think Paul mentioned that even though MacOS X server runs on the G3
> >powerbooks, it doesn't support sleep (am I wrong here?)
> >
> >As a side note, I got somewhere between 5-10 people come up to me at the
> Linux
> >expo and ask me lots of questions about Linux on the powerbooks...mostly
> people
> >who were considering getting one of the nice new slim ones instead of YAPC.
>
> Welcome to the team of people who watched the PowerBook reboot for hours
> trying to get it sleep ;-)
>
> I managed to grab some infos about the sleep process of those machines,
> but it still requires some ROM diassembling before we can figure it out
> completely.
>
> The main difference with older PowerBook models is that the CPU is
> completely powered off and reset on wakeup. The ROM reset vector looks
> for a structure with a magic word or some kind of checksum, I don't know
> for sure, at a fixed location in memory (I didn't figure out the location
> yet). This structure contains infos (registers, PC, ...) used for
> resuming the OS.
>
> So in order to get it work on linux, we must:
>
> - Figure out the gory details of the above process, including the MMU
> setup used by the ROM when looking for this structure
>
> - Hack the kernel so that the location used by this structure (probably
> an absolute physical address) is reserved. The structure is certainly
> somewhere in low memory, somwhere after the exception vectors, so in the
> middle of the kernel code.
>
> - Change via-pmu.c sleep code so that it doesn't try to hack the memory
> controller of previous machines when faced with a Grackle based machine,
> and add code so that Grackle goes to sleep too (I think NAP mode is used
> by MacOS).
>
> - Add the remaining sleep/wakeup notifiers to drivers that didn't need
> them until now (Paul added them to the drivers used by the 3400, but this
> doesnt't include atyfb for example).
>
> --
> Perso. e-mail: <mailto:bh40@calva.net>
> Work e-mail: <mailto:benh@mipsys.com>
> BenH. Web : <http://calvaweb.calvacom.fr/bh40/>
>
>
>
>
> --
> To UNSUBSCRIBE, email to debian-powerpc-request@lists.debian.org
> with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
>
>
Jambi.
--This message was never sent.
http://scw.ucsd.edu/~bigj
Reply to: