Re: EHCI considered harmful?

From: Sean McNeil <sean_at_mcneil.com>
Date: Fri, 29 Oct 2004 14:26:24 -0700
On Fri, 2004-10-29 at 14:12, Julian Elischer wrote:
> Michael Nottebrock wrote:
> 
> >On Friday, 29. October 2004 20:29, Mike Tancsa wrote:
> >
> >>I think EHCI would only make it worse. 
> >
> >Yes.
> >
> >>The same 1.1 bugs would be there 
> >>that you mention, and then the ones added by EHCI.  i.e. it doesnt take any
> >>1.1 bugs away, just adds more.
> >>
> >
> >Exactly. And I still say it should go in. And you should file a PR about your 
> >ehci issues (oh, yeah, and we need USB maintainers to take them, too :().
> 
> I have been taking a sebatical from freeBSD after puting my marriage a 
> bit too close to the
>  line than I like.. :-)
>  but one of my next things to look at is teh USB code..
> I've already been in there a bit and will be getting in again with a few 
> others who have shown
> interest as soon as:
> 
> 1/ I've spent enough time at home with the kids/wife to keep that side 
> balanced..
> 2/ work settles down
> 3/ the 5.3 push is over.
> 
> We have about 6 people who have shown interest in USB and we should be 
> starting up a mailing
>  list soon (who's doing that?) which should help.
> In addition we need expertise in newbus, CAM/SIM to help us get those 
> sides of things in order.
> 
> It needs more than a small touchup..

What I am wondering about is why I get ehci in my kernel when I do not
ask for it:

server# grep ehci LINT
LINT:device             ehci

So there is a config directive available...

server# grep hci AMD64
AMD64:device            uhci            # UHCI PCI->USB interface
AMD64:device            ohci            # OHCI PCI->USB interface

So I have not specified it...

server# kldstat -v | grep ehci
                135 ehci/usb

Yet it is in my kernel?

Cheers,
Sean


Received on Fri Oct 29 2004 - 19:26:26 UTC

This archive was generated by hypermail 2.4.0 : Wed May 19 2021 - 11:38:20 UTC