On Thursday 22 February 2007 12:51 pm, Jeff Warren wrote:
> On Thu, 2007-02-22 at 12:29, David Brownell wrote:

> > On Linux, at least, you would set a flag in the URB to say that it must
> > terminated by a zero-length packet.  Your driver wouldn't have to check
> > anything, the underlying controller driver would ensure the transfer is
> > terminated by a "short" packet (in this case, a zero length packet).
> > 
> > - Dave
>       I see that code in write_fifo so that my linux gadget can write a
> ZLP,but I dont see it in read_fifo where it receives a ZLP from the
> host. 
>       I think you are talking about if the Linux gadget is sending to the
> host. My problem is in the case of the Linux gadget receiving (960 bytes
> in my previous example) from the host(Windows).

No, the host side *should* have the same mechanism.  ISTR at one point
seeing it described in MSFT docs, and suspecting that someone had once
made that part of the Linux API look like Windows, but maybe I'm wrong
and the MS-Windows driver interface is deficient in that respect.


-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys-and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
_______________________________________________
[email protected]
To unsubscribe, use the last form field at:
https://lists.sourceforge.net/lists/listinfo/linux-usb-devel

Reply via email to