John D'Emic wrote:

> On Wed, 1 Sep 1999 seebs@plethora.net wrote:
>
> > In message <Pine.GSO.3.96.990901161848.24937E-100000@interport.net>, "John D'Em
> > ic" writes:
> > >Hmm, just made an additional 16 tty's, added them to /etc/tty's, SIGHUP'd
> > >inetd and init and still to no avail.
> >
> > Hmm.  What are the ttys (/dev/ttyXX)?  What's the "pty" line in your kernel
> > say?  Did you make corresponding /dev/ptyXX devices?
> >
> >
>
> the files created by MAKEDEV were /dev/ttyT0-9,a-f and /dev/ptyT0-9,a-f.
> i added the corresponding ttyTx entries in /etc/ttys.

Well from NetBSD 1.3.X to NetBSD 1.4.X minor numbers went for 8 bits
to 12 bits bits (256 minor numbers to 4096).  The MAKEDEV script creates
32 devices (master/slave pairs) in NetBSD 1.4.0.  To get all 12 bits, you
would need 3 hex digits (4 bits a hex nibble) and you probably break a
LOT of code that ties to allocate pty master/slave pairs.  I do not know
what System V did for EFT (Expanded Fundamental Types) back in SVR3.

I think the format should be /dev/ttyT[0-9,a-f][0-9,a-f][0-9,a-f] and
/dev/ttyT[0-9,a-f][0-9,a-f][0-9,a-f].  I am just kidding, okay?

I seem to remember W. R. Stevens having a C subroutine to allocate a
master/slave pty pair based on the master side being an exclusive
open, but I don't recall one in libc.  Is there such a creature?

--
robert pearson bobp@sevenofnine.erols.com     "I am unique and special,
                                               just like everyone else." ;-)