I never saw a "first patch" either.
I won't apply it for now, but we do (collectively)
need to try to have devfs support in the USB
drivers, so I'd appreciate it if you (Greg) and
Nick could come to an agreement on it.
Like Greg, I haven't tried it (devfs) yet.
~Randy
> -----Original Message-----
> From: Greg KH [mailto:[EMAIL PROTECTED]]
>
> Thanks for sending this, I never got a copy of your first patch, and I
> haven't had the time to test devfs out much, but, I disagree
> with parts
> of this patch. Randy, please don't apply it.
>
> On Wed, Mar 29, 2000 at 03:56:27PM -0700, [EMAIL PROTECTED] wrote:
> > I've cleaned up my usb-serial patch a little so I'll
> submit it again.
> > This patch adds proper devfs support to the usb-serial
> driver. It creates
> > it's devices as </dev>/usb/serial/#. Also, new in the
> version: It tries to
> > create a /dev/pilot symlink if it detects a Visor is connected.
>
> I don't think that the driver should create the symlink to /dev/pilot.
> Do other drivers do this kind of thing? And there is nothing
> saying that
> HandSpring next week will not switch the ports around in the Visor (in
> fact that will probably happen soon anyway, what with more of
> the plugin
> modules coming out.) You need to look for the port that is a HotSync
> port if you want to do this sort of thing.
>
> And then there's the fact that I just split up usb-serial.c
> into a bunch
> of smaller files, this probably will not patch cleanly.
>
> > A couple of known issues: When syncing usb/serial/#
> devices might not get
> > cleaned up properly. I think this is because the Visor is
> disconnecting
> > before pilot-xfer closes the device. And, while it won't
> change an existing
> > pilot symlink, it will remove it.
>
> So if a user app has a node open, and it disappears (like the Visor
> does), and then the user app closes the node, devfs doesn't remove it?
> Is this a devfs problem, or a driver problem?
>
> Man I gotta try out devfs soon...
>
> greg k-h
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
