Re: Zero-Byte Packet Questions

On Wednesday 07 March 2007 8:45 am, Pandita, Vikram wrote:
> 
> Alan
>  If the peripheral driver is to decide to send a ZLP or not based on the
> length of data transfer, then the API given to Gadget driver of req.zero
> is redundant.
> 
> My understanding is that the Gadget driver should respect the API and
> set the req.zero responsibly and hence the following fix suggested in
> Gadget Zero. Else the zero flag can be removed?

For *control transfers only* the zero flag should not be needed.
Some gadget drivers set it, primarily for paranoia.

For bulk transfers, see responses from both Alan and myself.


> Signed-off-by: Vikram Pandita <[email protected]>

NAK.


> ---
> diff -purN a/drivers/usb/gadget/zero.c b/drivers/usb/gadget/zero.c
> --- a/drivers/usb/gadget/zero.c	2007-02-02 18:33:34.000000000 +0530
> +++ b/drivers/usb/gadget/zero.c	2007-02-07 17:28:47.000000000 +0530
> @@ -1072,7 +1072,8 @@ unknown:
>  	/* respond with data transfer before status phase? */
>  	if (value >= 0) {
>  		req->length = value;
> -		req->zero = value < w_length;
> +		req->zero = value < w_length &&
> +				(value % gadget->ep0->maxpacket) == 0;
>  		value = usb_ep_queue (gadget->ep0, req, GFP_ATOMIC);
>  		if (value < 0) {
>  			DBG (dev, "ep_queue --> %d\n", value);
> 
> 
> 

-------------------------------------------------------------------------
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