Ok, just checked in a temporary fix which re-enables those MMC cards.
On Sep 20, 2004, at 5:53 PM, Craig Hughes wrote:
> Looks like the new SD code in fact doesn't work with the MMC cards we
> sell on the gumstix store :)
> I'll take a look into that -- something about not parsing the CSD
> structure correctly...
From: David Farrell <davidjf2001@ya...> - 2004-09-21 13:06:48
I'll try it later tonight, did you find why the
clock was not set to 20MHZ? I'll check for that
too, but it is obvious. The read/writes are very
slow. I assume that mmcinit is still not
required? Do raw reads & write now work too?
The only check to see if 4-bit IO is present is
to check the CPU ID to see what CPU you have.
What is you opinion of the /dev name? mmcblk0p1?
I think the driver assumes a max of 8 partitions
so every 8th minor is a new mmc card? Don't
forget to fix the /dev table in buildroot.
BTW I have been testing with a SanDisk SD128
On another note, I have been doing more testing
with bluetooth, I changed the FIFO depth to 32
the read loop to 32. I no longer have extra
chars on my console. Bluetooth seems to run
better, but still fails, there is more work to
During the course of this I noticed ifconfig
metrics are now working, I assume this happened
somewhere along the way with the kernel update
and not any side effect from what I am doing.
--- Craig Hughes <craig@...> wrote:
> Ok, just checked in a temporary fix which
> re-enables those MMC cards.
> Revision 102.
> On Sep 20, 2004, at 5:53 PM, Craig Hughes
> > Looks like the new SD code in fact doesn't
> work with the MMC cards we
> > sell on the gumstix store :)
> > I'll take a look into that -- something about
> not parsing the CSD
> > structure correctly...
Get latest updates about Open Source Projects, Conferences and News.