On Fri, May 05, 2000, Dunlap, Randy <[EMAIL PROTECTED]> wrote:
> Johannes,
>
> A few questions about this patch:
>
> > > 1. USB PnP support (usbd etc.); permanence of usbdevfs?
> >
> > I've been involved in this quite a bit recently.
> >
> > First off, I've attached a patch to completely nuke
> > usbdevfs's VFS layer in
> > favor of using devfs for all of that stuff. Nodes are still
> > created, and you
> > use the same ioctl()'s, you just use a different name and you
> > don't have
> > to worry about mounting an extra filesystem to get it.
>
> Does this address Jim Gettys's issues about the permanence
> (guaranteed existence) of /proc/bus/usb/devices or some
> other way to guarantee X that it will receive connect/disconnect
> events? And not miss any of them?
It's permanent as long as people use devfs.
What's needed is a requirement from USB that something be mounted.
> Does this patch (using devfs) use some form of message passing
> to get events from the kernel to a usb daemon? And then how
> do events get from the usb daemon to X, for example?
There doesn't necessarily need to be a USB daemon. The interface is
identical to /proc/bus/usb except for it's rooted at /dev/usb and there's
no explicit mounting required.
> And with the same ioctl's as usbdevfs, working userspace drivers
> (like rio500 using usbdevfs) will still work?
Exactly. There's a couple of new features, but all existing applications
will continue to work.
JE
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
