On Mon, May 29, 2000, David Brownell <[EMAIL PROTECTED]> wrote:
> > > The closest I came to a solution is just modprobing "usb-ohci"
> > > and "usb-uhci" ... which autoloads usbcore because of the symbol
> > > dependencies. (Unless something changed recently enough that I
> > > missed it ... pre-install Johan mentioned hasn't been needed.)
> > > A "post-install" (?) can mount usbdevfs, and could also start
> > > a usb daemon.
> > >
> > > I don't quite know what system component would do the modprobes,
> > > or how to avoid the nasty messages for the HC that doesn't exist.
> > > And I seem to recall that usbcore stayed loaded even if no HC
> > > existed, though maybe it autocleans.
> >
> > I think we should what SCSI does for their host adapter. They modprobe
> > "scsi_hostadapter" and /etc/conf.modules aliases that to whatever their
> > host adapter is.
> >
> > We could modprobe "usb_hostcontroller" and the user can alias that to
> > uhci or whatever.
>
> Not that I've had occasion yet to look at that code, but why
> shouldn't the PCI code just directly modprobe the right driver?
>
> After all, the OHCI and UHCI (and EHCI :-) controllers will
> have different PCI identifiers.
This would require a large table of devices/classes and their drivers in
the kernel to know what to do.
This is probably best left in userspace.
The new PCI code has something like this, but the driver must register
itself first. It's the old chicken and the egg problem.
In fact, I was planning on creating a driver to do this (as well as a SCSI
driver) to load drivers and do other "magic" things. Very much like my
userspace driver binding/devfs patch for USB works. Haven't even started
yet tho.
JE
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
