From: Andrew Poelstra <asp11_at_sfu.ca>
Date: Sun, 21 Nov 2010 15:33:14 -0800 (PST)

----- Original Message -----
> "Andrew Poelstra" <asp11_at_sfu.ca> schrieb:
>
> >For those who haven't heard of me, I haven't done much of anything
> >since the school year started. That's why.
> >
> >When I was working, my goal was to port Rockbox to the Cowon S9
> >platform. Unfortunately, a few days ago, my only S9 was run over
> >by a truck.
> >
> >Therefore, I am stopping my development and unsubscribing from
> >the list. Sorry guys and thanks for all the great work on this
> >project.
> >
> >
> >--
> >Andrew Poelstra
> >Email: asp11 at sfu.ca OR apoelstra at wpsoftware.net
> >Web: http://www.wpsoftware.net/andrew/
> >
> >nb Due to DNS issues ``wpsoftware.net'' does not work right
> > now. It may be a few weeks; Yahoo is screwing with me.
>
> it probably depend on the progress you have made and on the price, but
> there's the possibility of Rockbox funding you a new device.
> The last thing I heard was that you made tcctool(?) work with the s9.
>
> How far have you come?
>

A new player would be about $250, but I would not feel comfortable accepting
money for a new one. The long explanation:


I had not made it much farther --- using my SDCFG parameter I was able
to brick/unbrick the player freely (though I could not permanently program
it with tcctool; the recovery progress involved temporarily installing
the manufacturer firmware then using the "official" upgrade process to
make it permanent).

However, my SDCFG parameter didn't work for other people, presumably because
newer player models use a different SDRAM configuration. So nobody could
confirm my recovery process, even though it worked for me 100% of the time.

Also, when loading test code onto the player, I would often get CRC errors.
As near as I could tell, there was no rhyme or reason to this. It would simply
work or it wouldn't. Later on, the player itself developed NandFlash problems,
so that even when copying music with the manufacturer firmware, songs (and the
filesystem) would often be corrupted. I'm not sure how much this contributed to,
or was caused by, my programming failures.

Finally, I never did manage to get the screen to turn on with my test code. I
made some promising inroads dissembling the manufacturer firmware, but they
did a good job obfuscating it so that I was never sure what GPIO was used for
the backlight. I intended to build a JTAG programmer and do some debugging
through there but never found the time.

Without any means of communicating with the player, and without anyone else
who could successfully program it, I was pretty much stuck :).

Most of my dev time was spent searching the manufacturer code for either
display initialization or upgrade CRC code, and guessing at GPIO setups
in an attempt to turn on the backlight.


Andrew
Received on 2010-11-22