--- In Eric Robishaw <> wrote:
>
> Charlie,
>
> Thanks, I had forgotten about the 7800ctl code for sleep.
>
> Question: using the AVR code to sleep, how would my app "know" when it's
> woken back up to initiate a reboot?
I looked at the schematic, and AFAIK the AVR just controls the +5V to the
marvell chip (Q4, FDC634P). So it's like I suspected: no interaction with
kernel power management at all, just one big hammer (well, power switch ...)
> I'm guessing that when I sleep, all my threads stop as well, so I suppose I
> would set a flag in the app that I'm going to sleep
> when the cpu wakes back up I guess my app is still running and gets another
> loop through the thread, checks the flag, if set, issues a reboot?
No, if you use the AVR, it will be just like pulling the power plug for <sleep>
specified seconds, then plugging it in again. Your monitoring thread or
process, when it detects a perilous temperature, needs to cleanly save state,
unmount disks, etc. and then finally tell the AVR to turn off the power. When
<sleep> seconds expire, the AVR should re-apply power, which will be a cold
boot as far as linux and your programs are concerned. The 200 microamp sleep
mode TS refers to has only the AVR running
Actually that should work OK, because I guess you will probably need to sleep
quite a while until ambient temps come down.
I may be nice if TS could build in a slight delay into the AVR sleep code, ie
delay say 5 seconds after telling the AVR to sleep before actually turning off
the power.
If you had access to the AVR code, you could hook up a temp sensor, and
re-apply power to the ARM chip as soon as the temp was safe.
regards ........ Charlie
>
> Does that make sense?
>
> Thanks
> Eric
>
>
>
>
> On Wed, Jun 3, 2009 at 8:05 AM, charliem_1216 <>wrote:
>
> >
> >
> > --- In <ts-7000%40yahoogroups.com>, Eric Robishaw
> > <eric@> wrote:
> > >
> > > No, not really... the problem is the ambient temperature. There's not
> > enough
> > > differential to diffuse heat away from something when the air you're
> > blowing
> > > over it is as hot as the object itself.
> > >
> > > I really need a sleep mode of some kind. Some PC motherboard BIOSs have
> > an
> > > automatic wakeup at a certain time of day, I was hoping to find something
> > > equivalent in the 7800 or something.
> >
> > Please read the 'rtc.txt' file in ./Documentation of any recent 2.6 kernel.
> > There is a good description of what's possible, along with a test program
> > rtc.c. Sounds like many functions can be handled using the generic handler
> > if the clock chip driver doesn't handle them. That doesn't address how you
> > enter hibernation though, only how you wake up. If the orion boards are
> > hooked up to the kernel PM suspend & resume, this could be a reasonable way
> > to go. A side benefit is that your code would (could) be somewhat generic
> > and work on x86 boards too.
> >
> > A second approach would be to use the AVR microcontroller that TS has
> > included in the TS-7800. In the utility ts7800ctl
> > (ftp://ftp.embeddedarm.com/ts-arm-sbc/ts-7800-linux/samples/ts7800ctl.c)
> > there is a sleep command that will 'send stop signal' to the marvel chip.
> > Max sleep is 8 * 65535 seconds in length, plenty long enough to wake up
> > after the sun goes down! I guess this is the mode referred to on the ts-7800
> > preliminary manual page: "Sleep mode uses 200 microamps". AFAIK, it would be
> > up to your pgm to monitor the temps, clean up your running processes, etc,
> > to get ready for sleep, and then issue the command to the AVR.
> >
> > I don't know how well, if at all, the AVR sleep is integrated with the
> > linux PM functions. My guess is not at all, and you should plan on a reboot
> > rather than a resume from suspend.
> >
> > regards, .......... Charlie
> >
> > >
> > > E
> > >
> > >
> > >
> > > On Tue, Jun 2, 2009 at 11:07 PM, Morgan Gangwere <0.fractalus@>wrote:
> > >
> > > >
> > > >
> > > > I'd assume fans/heatsinks are not an option at this point?
> > > >
> > > > --
> > > > Morgan gangwere
> > > >
> > > > "Space does not reflect society, it expresses it." -- Castells, M.,
> > > > Space of Flows, Space of Places: Materials for a Theory of Urbanism in
> > > > the Information Age, in The Cybercities Reader, S. Graham, Editor.
> > > > 2004, Routledge: London. p. 82-93.
> > > >
> > > >
> > >
> >
> >
> >
>
------------------------------------
Yahoo! Groups Links
<*> To visit your group on the web, go to:
http://groups.yahoo.com/group/ts-7000/
<*> Your email settings:
Individual Email | Traditional
<*> To change settings online go to:
http://groups.yahoo.com/group/ts-7000/join
(Yahoo! ID required)
<*> To change settings via email:
<*> To unsubscribe from this group, send an email to:
<*> Your use of Yahoo! Groups is subject to:
http://docs.yahoo.com/info/terms/
|