Re: [coldsync-hackers] Patch: Palm m50x USB support (Linux)

[Koen Deforche <kdf@irule.be>]

Greg KH wrote:
> 
> On Mon, Jul 16, 2001 at 07:55:25PM +0200, Koen Deforche wrote:
> >
> > The first mystery is that the first 6B (which happen to be a header)
> > of communication on the ttyUSB1 pipe from palm to PC are missing!
> 
> They are missing because they are part of the USB discovery process that
> the palm USB protocol specifies.  Take a look at the visor.h file in the
> linux kernel source directory drivers/usb/serial for a description of
> these commands and responses.  They are not part of the data stream used
> by the sync process.

They aren't?

But in serial.c I don't see where these 6B are read; in visor_startup()
everything that happens is with usb_control_msg()'s, which isn't
supposed to touch the other USB pipes? Also, everything that the
palm is sending back there is printed out, and all the messages are
the same as what I see in the USBSnoopy log. Thus, to me, there is
really 6B missing somewhere. The communication on the ttyUSB1 pipe is
all in a packet format, which consists of 6B header + payload, where the
header specifies the length of the payload. It is this first header that
seems to be missing.
 
> Are you using the usb-ohci.o or usb-uhci.o driver?  It looks like a
> problem in there, as the visor driver never gets the data.

That is also my impression. It is the usb-uhci.o driver that I use.
 
> Of course, it could be a bug in the device, as a "close" command just
> got sent to the device, which might flush out the data from the pilot.

I also get this behaviour when I just disconnect the palm from the
cradle, or the cradle from the PC. Thus the last 64B data which is
observed must already have been sent from the palm to the PC. Still it
may be a bug in the palm where it does not acknowledge this urb and then
the disconnect may cause the usb-uhci driver to flush out this
urb? (I'm just gessing around here, I don't know how the USB urb
mechanism
works)
 
> Hard to tell which it is without a USB protocol analizer.  You wouldn't
> happen to have access to one, would you?

Your question sounds as if it is not really something that I can just
"download" :) Is it a hardware device?
 
> > One more thing I will try, is to find a way to get the naked USB data,
> > so that I can verify if the Palm actually sends it, or not. Do you
> > know where in the kernel that I should look for that?
> 
> Try turning on debugging in the host controller driver.  Each one does
> it differently.

I did this, but now there is such an overload of debugging information
that
syslogd drops data. Which makes the result quite useless. Anyone knows
a simple workaround to that?

cu,
koen.
This message was sent through the coldsync-hackers mailing list.  To remove
yourself from this mailing list, send a message to majordomo@thedotin.net
with the words "unsubscribe coldsync-hackers" in the message body.  For more
information on Coldsync, send mail to coldsync-hackers-owner@thedotin.net.