Re: [Libusb-devel] isochronous IN transfer returns EPROTO (71) errors

[Alan S. <st...@ro...>]

On Mon, 6 Jun 2011, Glenn Howald wrote:

> Thank you both for your input.
> 
> I will look again when I get home tonight, but I am pretty sure the
> descriptor is returning 3060 bytes.
> 
> On Mon, Jun 6, 2011 at 12:26 PM, Tim Roberts <ti...@pr...> wrote:
> 
> > Glenn Howald wrote:
> > >
> > > I have set up a test program to communicate with my USB device using
> > > LIBUSB 1.0, and I am receiving the expected ioschronous transfers from
> > > the device interface to my host.
> > >
> > > I set up the buffer to be 32 3060 byte "multi-frame sets", but it
> > > seems that data is being truncated at 512 bytes
> > > (transfer->packet_desc[N].actual_length) and returning a EPROTO (-71)
> > > status (transfer->packet_desc[N].status) in the truncated "multi frame
> > > sets".

By the way, that combination (512 bytes of iso-IN data and -EPROTO
status for the same transaction) is very peculiar.  I can't think how
it could occur.

Alan Stern