Hi Greg/Lee,

Thank you very much for the information and yesterday we also had this
thought. We are tried it working fine now. 

This is one of the good solutions.


Thanks again.


Shalini. 
 


-----Original Message-----
From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED] On Behalf Of Greg
Lee
Sent: Thursday, May 12, 2005 1:55 AM
To: linux-usb-devel@lists.sourceforge.net
Subject: RE: FW: [linux-usb-devel] Usb serial with power ON/OFF problem
using Linux 2.6.11.8

> I'm sorry, but I don't see the specs for your device on how 
> it works. Did you forget to attach them?
> 
> thanks,
> 
> greg k-h

Greg, I'm jumping in here with some information that I hope helps a bit.
I
presently work with several different kernels (2.4.17, 2.4.19, 2.4.21
and a
recent 2.6 kernel which I'm not sure of the version) and have
experienced
this problem as well when using usbserial.  First, let me define what I
see
as "the problem".

When you interface with mobile  phones via USB, they tend to "come and
go"
on the USB port due to various events that occur on the phone, the
primary
example being power cycling the phone which is something I have to do
regularly for my application.  The standard command to switch the power
state on the phone is "AT+CFUN=1" to power it on and "AT+CFUN=0" to
power
it off.  When you switch the power state the USB port typically
disappears
and then reappears several times over the course of 5 seconds or so
until
it settles down.  Naturally this should invalidate any file descriptors
that you have open for the serial device, however, this does not appear
to
work correctly.  The symptom that you see is if you write data to the
USB
device on a handle that was opened before switching the power state you
will get a kernel OPPS.  I work around this by closing the file handle
immediately after sending the AT+CFUN command, waiting 10 seconds and
then
attempt to reopen the port.

While I do have a workaround, occasionally my software is not able to
account for this case and I do end up with kernel OPPS which are caught
by
the system watchdog.  It "would be nice" if I did not have to account
for
this problem.  I have seen a couple of others on this mailing list
dealing
with similar problems in the past month as well.

Regards,
Greg Lee
[EMAIL PROTECTED]



-------------------------------------------------------
This SF.Net email is sponsored by Oracle Space Sweepstakes
Want to be the first software developer in space?
Enter now for the Oracle Space Sweepstakes!
http://ads.osdn.com/?ad_ids93&alloc_id281&op=ick
_______________________________________________
linux-usb-devel@lists.sourceforge.net
To unsubscribe, use the last form field at:
https://lists.sourceforge.net/lists/listinfo/linux-usb-devel


-------------------------------------------------------
This SF.Net email is sponsored by Oracle Space Sweepstakes
Want to be the first software developer in space?
Enter now for the Oracle Space Sweepstakes!
http://ads.osdn.com/?ad_ids93&alloc_id281&op=click
_______________________________________________
linux-usb-devel@lists.sourceforge.net
To unsubscribe, use the last form field at:
https://lists.sourceforge.net/lists/listinfo/linux-usb-devel

Reply via email to