GTA01 SD card problem, kernel 2.6.24

[Andy Green]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Somebody in the thread at some point said:
> O Mighty Magicians of the OpenMoko Kernel!  ;)   I beg a moment of your
> time -- can someone please glance at:
> http://bugzilla.openmoko.org/cgi-bin/bugzilla/show_bug.cgi?id=1223
> and offer a quick suggestion on where to look to debug this?  I'm
> suspecting this is related to the work done to make the wifi card on the
> GTA02 function, but if someone can just confirm that to be reasonable,
> or not, it would be helpful.  Timing problem?  What's the debugging
> approach of choice for this?  (I've never messed with I2C, MMC drivers,
> or SD cards.)
> 
> I feel the need to tinker in the guts of something this weekend, and
> it's either the kernel or the plumbing in my house...  :-)

I'm glad you're thinking about looking at it, on GTA02 we no longer use
the CPU SD interface for memory cards so this is not getting tested on
my side anyway.

For debugging there is a kernel .config CONFIG_MMC_DEBUG which increases
verbosity that should be your first move.

If only the first byte is written, it sounds to me like either block
sizing is broken or DMA is broken as first ideas.  Also might affect it
if it is an SDHC card since they use block-based addressing and not
byte-based.

Post your verbose log here or on the Bugzilla I guess.

- -Andy
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.7 (GNU/Linux)
Comment: Using GnuPG with Fedora - http://enigmail.mozdev.org

iD8DBQFHtqS2OjLpvpq7dMoRAiJbAJ4ycsTsTTUfCWh0p0LwE8iqjVHpLwCdEStS
LpdYL2Uo5DxARwY46LdZ6RM=
=ZsfA
-----END PGP SIGNATURE-----