[RFC PATCH 1/4] USB: HCD: support giveback of URB in tasklet context

Ming Lei ming.lei at canonical.com
Mon Jun 10 05:51:14 EDT 2013


On Mon, Jun 10, 2013 at 5:31 PM, Oliver Neukum <oliver at neukum.org> wrote:
> On Monday 10 June 2013 17:23:46 Ming Lei wrote:
>> On Mon, Jun 10, 2013 at 4:43 PM, Oliver Neukum <oliver at neukum.org> wrote:
>> > On Sunday 09 June 2013 23:18:28 Ming Lei wrote:
>> >> 2), the biggest change is the situation in which usb_submit_urb() is called
>> >> in complete() callback, so the introduced tasklet schedule delay might be a
>> >> con, but it shouldn't be a big deal:
>> >>
>> >>         - control/bulk asynchronous transfer isn't sensitive to schedule
>> >>           delay
>> >
>> > That is debatable.Missing a frame boundary is expensive because the increased
>> > latency then translates into lower throughput.
>>
>> Suppose so, considered that bulk transfer will do large data block transfer, and
>> the extra frame or uFrame doesn't matter over the whole transfer time.
>
> That is not true for all use cases. Networking looks vulnerable.

> That is debatable.Missing a frame boundary is expensive because the increased
> latency then translates into lower throughput.

Missing uframe/frame boundary doesn't cause lower throughput since network
devices use bulk transfer, which is scheduled in hw aync queue and there should
always have pending transfers in the queue when submitting bulk URB to the
queue.


Thanks,
--
Ming Lei



More information about the linux-arm-kernel mailing list