[RFC] [PATCH] usbatm.[ch]: logic changes + error handling

Roman Kagan rkagan at mail.ru
Sat Apr 16 13:46:40 EDT 2005


On Fri, Apr 15, 2005 at 01:21:22PM +0200, castet.matthieu at free.fr wrote:
> Selon Roman Kagan <rkagan at mail.ru>:
> > OK so here goes another update of my patch to the current CVS, in case
> > it gets somebody interested.
> 
> For the iso part, I think we should get the size of packet of the selected
> alternativate on the endpoint : this will allow the user of useratm to choose
> the right alternative in it's bind without changing the iso packet size.

Do you want it different from how it's done in my patch?  There it
initializes the rx_channel.endpoint to a bulk pipe from subdriver's
->in.  Then the subdriver in its ->bind() is free to claim the right
interface, choose the appropriate altsetting, and then update
rx_channel.endpoint.  Later on usbatm_usb_probe(), in case the endpoint
is an iso pipe, uses usb_maxpacket() on the given pipe to obtain the iso
packet size, and adjusts other values accordingly.

> For errors produced by the modem, I haven't too much time to investigate, but an
> adi developper gave us a documentation [1] of a part of commands we could send
> to the modem and there is a paragraph were they talk about crc errors in some
> case.

Uff it's a 224 page document...  

Cheers,
  Roman.



More information about the Usbatm mailing list