[PATCH][RFC] usbatm.[ch]: cleanup and OAM F5 loopback

[matthieu castet]

Hi Duncan,

Duncan Sands wrote:
>>I finaly test it on another computer : it seem it is the modem.
> 
> 
> Similar problems are sometimes seen with the speedtouch modems.
> Upgrading to more recent firmware usually fixes it.
> 
yes, but for the speedtouch everything is done by the firmware, here the 
driver have some works to do. And without the spec it is very hard to 
fix it. I supppect setting correctly a timeout could improved the things 
[1], but the iso pipe works well so why don't use it...

cheers,

Matthieu

PS : I am using a quite recent firmware.

[1]
			pAdiSM->XferRate0 = pMsg->data;
                         pAdiSM->DownRate = (pMsg->data>>16) *32*1024;
                         pAdiSM->UpRate = (pMsg->data&0xFFFF) *32*1024;
			if (pAdiSM->FwRxTimeout != pAdiSM->XferRate0)
                         {
                             UInt32 threshold;
                             if (pAdiSM->DownRate <= 0x80 *32*1024)
                             {
                                 threshold = 0xB000;
                             }
                             else
                             {
                                 threshold = 0xF000;
                             }
                             if (pAdiSM->FwRxTimeout != threshold)
                             {
                                 pAdiSM->FwRxTimeout = threshold;
/**********************************************************************/
                                 /* We're getting an error -6912 back 
from the device                  */
                                 /* when we send this command. Waiting 
to hear back                    */
                                 /* from ADI about this, 'cause the NDIS 
code doesn't                  */
                                 /* actually send it. 
                                */
                                 /* eu_cmd_to_modemNoPB( ins, 
ADI_CMD_SET_TIMEOUT, threshold, 0, NULL );*/
                                 /* ADI says not to worry about this - 
hardware does                   */
                                 /* fine without having the driver set 
this value                      */
 
/**********************************************************************/
                             }
                         }