aux button interrupt to pmu?

[Carsten Haitzler (The Rasterman)]

On Fri, 06 Jun 2008 07:46:24 +0100 Andy Green <andy at openmoko.com> babbled:

> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
> 
> Somebody in the thread at some point said:
> | On Fri, 6 Jun 2008 00:36:21 -0300 Werner Almesberger <werner at openmoko.org>
> | babbled:
> |
> |> Carsten Haitzler wrote:
> |>> just checking- i'm sure you said this before - but the aux button has no
> |>> interrupt line to the pum - right?
> |> Correct.
> |>
> |>> (so it can't be used for wakeup on suspend)?
> |> It could be, since it's connected to the CPU, so it could wake the CPU
> |> directly. However, please bear in mind that AUX also changes whether
> |> we boot from NAND or from NOR, so using it to wake up from suspend,
> |> which involves a small part of the boot process, has a high risk of
> |> causing trouble in general, and especially as the u-boot in NAND and
> |> the u-boot in NOR diverge.
> |
> | so basically.. "no". is the answer to using the aux button for a reliable
> | wakeup from suspend (if we do zero-clock though this is another matter...)
> | :)
> 
> This is actually an interesting story I think we already learned lessons
> from about what not to do.
> 
> AUX comes to EINT5 and is a wake source for suspend just fine.
> 
> But we overload AUX button to select coming up in NOR; when we wake from
> suspend it is a normal reset for the CPU and if you have the AUX button
> down even for wake, we will wake up in NOR U-Boot.
> 
> Werner mentions "diverge[nce]" for NOR U-Boot, it is because it can't be
> updated due to this "write once" policy (you can rewrite it from a debug
> board only).
> 
> In fact there should be limited trouble it can cause since we jump to a
> wake address in a fixed CPU register that survives suspend, if NOR
> U-Boot can deal with wake at all it should be safe.  The main thing
> that'll be broken is the wake reason, since NOR image AFAIK is too old
> to have the patch.
> 
> 
> But... Carsten does it say in the spec we should wake on AUX press?

it doesnt say it must. it is being asked for. if aux can't tell us what woke
it... then we have a problem. it doesn't seem that its a functional wake button
(i tried aux to wake myself on my gta02 and nothing happened).

so unless the nor contents are changed... it isnt a fully functional wake
source - right? at this stage the spec is much more malleable than the
hardware. the aux for wake is new in the last few weeks.

-- 
Carsten Haitzler (The Rasterman) <raster at openmoko.org>