From: Sean Luke <sean@cs...> - 2005-01-08 05:29:04
Or, heck, rshd and rcp. We'd like to be able to log into the gumstix
over a network. Any clues as to how to do so?
Also, any clues about how to set up networking over client USB to a
remote Linux box?
Last, has anyone gotten PAN working stably on the gumstix?
From: Craig R Hughes <craig@gu...> - 2005-01-09 06:00:47
Sean Luke wrote:
SL> Or, heck, rshd and rcp. We'd like to be able to log into the gumstix over a
SL> network. Any clues as to how to do so?
Out of the box, the gumstix is running an sshd/scpd -- username/password of
root/gumstix should work, if the 'stix has an IP address. It doesn't include a
ssh/scp *client* though on the 'stix, though I think a recent update of dropbear
added that -- I'll look into it.
SL> Also, any clues about how to set up networking over client USB to a remote
SL> Linux box?
Make sure the remote linux box has either kernel 2.6.10 on it, or else (2.6.8 or
2.6.9) with patches to the drivers/usb tree to support recognition of PXA
RNDIS/CDC gadgets. This patch was finally folded in with 2.6.10, but it's like
a 3 line patch if you want to roll it into an earlier kernel.
SL> Last, has anyone gotten PAN working stably on the gumstix?
Yes, I have. However, lots of other people have seen its flakiness persist even
after the fix I made. Then, last week, I had a gumstix bluetooth HCI link drop
while I was testing something else. First time I've seen that happen since
r234; I've got a little post-it note on that gumstix so I can go back to it and
try to reproduce/debug the problem. But I checked /proc/tty/drivers/PXA Serial
when the HCI link went down, and oddly it was reporting neither overflow errors
nor framing or parity errors, so I'm not sure how the HCI link could have
broken. But as I said, now that I've got a 'stix which reproduces the problem
(or maybe it's just the room I was in at the time, which I've also made a note
of), I'll have a much better chance of debugging than when it just wouldn't fail