Re: [PATCH 2.6] Altix serial driver

[Russell King <rmk+lkml@arm.linux.org.uk>]

On Sun, Jun 27, 2004 at 05:33:11PM -0700, Chris Wedgwood wrote:
> On Sun, Jun 27, 2004 at 07:24:34PM -0500, Erik Jacobson wrote:
> > Maybe you can help me clear it up then.  When I feed serial core the
> > name ttyS with TTY_MAJOR and minor 64, the registration fails.  If I
> > disable 8250 in the kernel config, the registration works for us.
> 
> I'm not sure why the 8250 code reserves/registers ttyS0 on Altix when
> do such hardware exists.  I vaguely recall patching it to prevent this
> in a hacky way.
> 
> I would like to know why the 8250 code is registering a driver for
> hardware that doesn't exist and see that fixed.

It's the way its always been done, and the way the tty layer works.
You register a range of ttys that you're going to be driving, and
you own those ttys whether or not you actually have hardware for
them.

You can't say "ok, I have ttyS1 and ttyS3, I'll leave ttyS0 and ttyS2
available for someone else to use" because the tty layer just doesn't
work like that.  It has the notion of a single driver for a range of
non-overlapping ttys.

Yes, it would be nice to get rid of that limitation, but we're in a
stable kernel series and the tty layer doesn't have a maintainer to
do the work...

-- 
Russell King
 Linux kernel    2.6 ARM Linux   - http://www.arm.linux.org.uk/
 maintainer of:  2.6 PCMCIA      - http://pcmcia.arm.linux.org.uk/
                 2.6 Serial core