[PATCH] iso pipe support for usbatm

[matthieu castet]

matthieu castet wrote:
> matthieu castet wrote:
> 
>> Hi,
>>
>> Roman Kagan wrote:
>>
>>> On Sat, Mar 19, 2005 at 07:28:04PM +0100, matthieu castet wrote:
>>>
>>>>> BTW do you get -EILSEQ often?  Documentation/usb/error-codes.txt 
>>>>> reads:
>>>>>
>>>>> (*) Error codes like -EPROTO, -EILSEQ and -EOVERFLOW normally indicate
>>>>> hardware problems such as bad devices (including firmware) or cables.
>>>>>
>>>>
>>>> I think this is specific to via controler (someone told me that 
>>>> there windows driver have some trick for -EILSEQ error)
>>>
>>>
>>>
>>>
>>> And are there many of those errors?
>>>
>> When I loged the data, I saw there are EOVERFLOW errors. And because 
>> the errors  aren't catch by usbatm, the bad packet go through 
>> usbatm_extract_cells and produce the vpi/vci error.
>> I will try with a smaller size.
>>
> 
>         /* may not be in_interrupt() */
>         spin_lock_irqsave(&instance->receive_lock, flags);
>         list_add(&rcv->list, &instance->spare_receivers);
>         list_add_tail(&buf->list, &instance->filled_receive_buffers);
>         if (likely(!urb->status) ||
>                         /* via chipset produce lot's of crc error...*/
>                         (usb_pipeisoc(instance->rx_endpoint) && 
> urb->status == -EILSEQ))
>                 tasklet_schedule(&instance->receive_tasklet);
>         spin_unlock_irqrestore(&instance->receive_lock, flags);
> 
> 
> In case of error : status !=0 couldn't we add the buffer to the 
> spare_receive_buffers list ?
> 
or at least set buf->filled_cells to 0.

Matthieu