Re: [Gumstix-users] serial console error

[Søren S. C. <li...@ss...>]

Hi Gzp,

Sorry for the late response...

> I did not used the .ift/signGP files as I think it is needed only to update
> the NAND. But to update the NAND, first I have to get an uboot loaded (ex.
> from sdcard). But I could not get it working. Note: I got the original/new
> device as it had no working flash at all. Please correct me if I'm wrong and
> I've been using the wrong files all along. (Though, I've also tried an .ift
> to use as the MLO on the SD card from
> http://www.sakoman.com/feeds/omap3/glibc/images/overo/201006072220/ )
You are right - I was wrong about the .ift-files - Sorry for being too fast in my previous answer...

> The device is a Water in a TOBI extension board. (The OMAP has the sign:
> 9WA98 JW256, if the big black box in the middle of the COM module is the
> "CPU" ).
Actually what you can see is the combined NAND+LPDDR device. 
The OMAP is mounted between the PCB and this device, so you can only see it (being ~0.6mm high) from the side of the module :-)
For more information on PoP please have a look at: http://focus.ti.com/lit/an/spraav2/spraav2.pdf

> I've also tried to boot from uart in a similar way as it is written for
> beagleboard: http://elinux.org/BeagleBoardRecovery, but apart from getting
> the ASIC ID I did not managed to do anything.
Strange - Method is normally working for Overo as well...

> Than I've read more posts about omap3, and using a beagle-board mlo, I've
> managed to get some response from overo (similar to my previous post), but
> it won't start the u-boot either: http://elinux.org/BeagleBoardRecovery /
> http://beagleboard.googlecode.com/files/MLO_revc_v3
> Texas Instruments X-Loader 1.4.2 (Feb 19 2009 - 12:01:24)
> Reading boot sector
> Loading u-boot.bin from nand
> ECC Failed, page 0x00080300
> 48 0 8d 0 34 50 8d c0 e 10 a0 0 f 0 85 c0 d 0 a0 0 1e 8 0 c1 0 0 a0 0 0 0 a0
> c1 
Seems like stuff goes bad whenever you try to involve external memory (the PoP LPDDR)?
Have you ever seen any X-loader version load (and execute) U-boot successfully?

> And of course  I could not find the source for the MLO_revc_v3  :-(.
All MLO versions are more or less equal, so having (and being able to modify) the source of one of them should bring you in a position to at least load and start U-boot...

> Is it possible that the device is faulty in a way that, when the overo
> x-loader enables some periphery, the OMAP hangs, but when I use the loader
> of another device it does not enables those ports/pins/timings/etc. and my
> device can provide some response ?
Straight out of my mind my guess would be a defect OMAP/LPDDR-chip.
This could as well somehow explain why you got the device with no SW preloaded. 
Normally they are preloaded with an Angstrom image when shipped...

I'm unfortunately a little out of ideas as well - Best regards - Good luck
  Søren

---
SSC Solutions ApS - Denmark - www.ssc-solutions.dk