Am Donnerstag, 13. März 2003 00:23 schrieb David Brownell:
> Oliver Neukum wrote:
> >>And as near as I can tell, nobody has ever needed usb_unlink_one()
> >>like we have today, except as a way to build usb_unlink_all().
> >
> > How about SCSI abort and network timeout ?
>
> SCSI abort applies to one request ... which often means
> dozens of queued urbs. Perhaps in the future there will
> be devices that can keep multiple requests in flight at
> once, but that'd be a ways off yet, with much higher end
> disks than I understand are yet available through USB.
It just needs an USB2.0 USB to SCSI bridge.
> Network timeouts ... if the link times out for one packet,
> there's no way it'd be accepting the next N in the queue.
> So the same thing applies: no reason not to kill them all.
True, but not all URBs may relate to output. Suffering a timeout
on output doesn't mean that all input URBs have to be canceled
or status reporting URBs have to be dropped.
You can hope that manufacturers sanely divide such traffic
among endpoints, but depending on vendors' sanity is dangerous.
Furthermore we already have an endpoint where killing all URBs
is definitely wrong, EP 0.
Is there really a large benefit in bringing this assumption
into a generic API?
Regards
Oliver
-------------------------------------------------------
This SF.net email is sponsored by:Crypto Challenge is now open!
Get cracking and register here for some mind boggling fun and
the chance of winning an Apple iPod:
http://ads.sourceforge.net/cgi-bin/redirect.pl?thaw0031en
_______________________________________________
[EMAIL PROTECTED]
To unsubscribe, use the last form field at:
https://lists.sourceforge.net/lists/listinfo/linux-usb-devel
