Hi David,
> >>>I found out that usb_set_configuration() was removed from the USB API by
> >>>you. ...
> >
> > it is not exported anymore and that is my problem.
>
> Oh, sorry. That was so long ago! Nobody was using it
> outside of usbcore.
is there a problem with re-exporting it?
> >> Your firmware load
> >>isn't the "load then re-enumerate" type, right?
> >
> > It was my plan to do it this way. The firmware must be loaded in
> > configuration 1 and then you have to switch to configuration 2 for the
> > running state. No endpoints are changed. I don't yet know if it will
> > re-enumerate.
>
> Hmm, if it can't re-enumerate that'd seem like a bug in the firmware.
>
> Normally one expects that loading firmware makes some externally
> visible changes ... either device descriptor changes, or config
> descriptor changes, or both.
>
> What happens if you usb_reset_device() after loading the firmware?
> Ideally that'd turn up new descriptors and the device could be made
> to select the "right" config right away.
I can't tell you that right now, because the 2.4 version was different
and I haven't used the usb_set_configuration() call outside the probe()
routine. My plan is to do it this way:
probe()
switch (configuration)
case 1:
load_firmware()
schedule_work()
set_config(2)
case 2:
register_device()
If I use usb_set_configuration() I expect that the USB core will
disconnect the interface with the current configuration and probe
another one with the new configuration.
Regards
Marcel
-------------------------------------------------------
This SF.Net email is sponsored by BEA Weblogic Workshop
FREE Java Enterprise J2EE developer tools!
Get your free copy of BEA WebLogic Workshop 8.1 today.
http://ads.osdn.com/?ad_id=4721&alloc_id=10040&op=click
_______________________________________________
[EMAIL PROTECTED]
To unsubscribe, use the last form field at:
https://lists.sourceforge.net/lists/listinfo/linux-usb-devel
