Suspend issues with GSM modem?

[Nick Guenther]

On Dec 17, 2007 6:57 AM, "Jørgen P. Tjernø" <jorgen at devsoft.no> wrote:
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
>
> I've recently left my Neo1973 silent and lonely, due to being swamped
> with projects and school. But smoe time ago I heard there was an issue
> with the GSM modem after resuming from suspend - and this prevents the
> batterytime to getting even close to bearable.
>
> Is this still present? What're the specifics of the issue, and are the
> OpenMoko-guys working on it? Is it a software issue, or an actual
> hardware defect?
>

http://wiki.openmoko.org/wiki/Qtopia_on_Neo_1973#Suspend_issues
---use u-boot to remove the console= boot variable, makes the console
stop interfering with gsmd (since, on the hardware level, they use the
same uart)

http://labs.trolltech.com/blogs/2007/12/04/neo-neuros-and-other-things/
"""The problem was that suspending would cause the modem to start
spitting out junk. and so would not respond to a modem wake up or be
able to make phone calls after resuming. The fix is a bit odd, really.
I just delete the /dev/ttySAC0 (modem) device and then create it again
right before I power up the modem. So now, that aspect of
suspending/resuming is working, at least on the one phone I have.
Well, for the first suspend anyway."""

You should note that suspending also makes gsmd stop listening to the
gsm modem, so it's either battery-life or phone. I don't know how they
are planning on fixing this, but it would have to be something similar
to ethernet wake-on-LAN.
As a temporary hack, I've got a USB charger (like
http://www.ladyada.net/make/mintyboost/, though I'm ashamed to admit
that mine is a pre-made Noma one) that I wander around with. I haven't
tested it all out in rugged conditions yet, but I imagine I could get
by for at least a day.

-Nick