On Tue, May 14, 2002, David Brownell <[EMAIL PROTECTED]> wrote:
> > > Heh ... I knew there were indeed areas where we could agree!
> >
> > Who ever said we could be unreasonable? :)
>
> Those little birdies don't know what they're talking about
> this morning. That earthquake clearly shook their little
> hummingbird brains ... :)
>
>
> > Gotcha. Makes perfect sense.
> >
> > I'll make a patch for uhci.c to address both of these issues.
>
> I assume you mean removing urb->next support for ISO and
> handling urb->interval for ISO? Great!
I was going to remove urb->next completely. While it's mostly used for
ISO and I think the documentation mentions something along those lines
too, it's generic enough that it applied to any transfer, for uhci.c
atleast.
> Greg's suggestion a while back was that I see if anyone
> threw any particularly heavy fits at the notion of removing
> urb->next before actually removing it. So far, no ... but I'll
> still hold off on a patch to remove that. (And merge those
> extra error checks into usb_submit_urb.)
>
> If you're prepping "uhci.c" to work without it, then we
> will have three drivers ("ohci-hcd", "ehci-hcd", and
> then "uhci") that use that updated ISO model. It'll
> be a good moment to switch over to "ohci-hcd".
I'll work on making the changes to uhci-hcd. That way uhci.c can stay
similar across 2.4 and 2.5 until the point we remove the next field from
struct urb.
JE
_______________________________________________________________
Have big pipes? SourceForge.net is looking for download mirrors. We supply
the hardware. You get the recognition. Email Us: [EMAIL PROTECTED]
_______________________________________________
[EMAIL PROTECTED]
To unsubscribe, use the last form field at:
https://lists.sourceforge.net/lists/listinfo/linux-usb-devel
