With the device speed in dev->speed, it's no longer needed in the "generic" pipe data structure. Which means __create_pipe() gets simpler too.
But I see what you mean: the UHCI code has a hidden dependency on that layout, TD_CTRL_LS. (Sigh; the explicit ones are always more maintainable IMO.) I'll see about providing updates for those drivers. Has anyone else noticed how nasty the code to rip out fields from the pipe bitmask looks, when you read the ASM output? :) - Dave ----- Original Message ----- From: "Greg KH" <[EMAIL PROTECTED]> To: "David Brownell" <[EMAIL PROTECTED]> Cc: <[EMAIL PROTECTED]> Sent: Wednesday, January 16, 2002 8:36 AM Subject: Re: PATCH: usb-0114 (usb.h cleanup) > On Mon, Jan 14, 2002 at 05:09:06PM -0800, David Brownell wrote: > > > I'll send off another round of 2.5 USB patches after the next -pre > > > release happens (there's not much in the queue.) > > > > Not much in the queue? We shouldn't let that happen... :) > > > > Here's a minor cleanup to <include/usb.h> that deletes some > > stuff that hasn't really been appropriate for quite some time. > > > > It's against 2.5.2-pre11. > > Doesn't the other host controller drivers need to be changed do to your > change to __create_pipe(), much like you did for the OHCI driver in > this patch? > > Or am I not understanding what you are trying to cleanup here? :) > > thanks, > > greg k-h _______________________________________________ [EMAIL PROTECTED] To unsubscribe, use the last form field at: https://lists.sourceforge.net/lists/listinfo/linux-usb-devel
