I've finally gotten around to upgrading my old i386 box from 1.3_BETA
to a reasonably recent 1.3D (less than one month old, anyway).  This
went very smoothly (I'd already upgraded my Sparc and VAX systems, so
I knew where the potential snags were), except that my SLIP link no
longer worked.  The slattach(8) process would just sit there and spin,
eating CPU, without forking and disconnecting.  A bit of debugging
showed that the problem was the asserting of DTR, thus:

	if (ioctl(fd, TIOCSDTR, 0) < 0)
		err(1, "TIOCSDTR");

For some weird reason, the call to ioctl(2) never returns; as stated
above, the process just spins in user space.  ifdeffing out this one
statement from the code gave me a slattach that works, so I'm happy,
but it does look that a bug that could do with some fixing.  :-)

>From /sys/arch/i386/conf/TIPCAT:

	com2 at isa0 port 0x3e8 irq 5
	pseudo-device sl 2

>From dmesg output:

	NetBSD 1.3D (TIPCAT) #0: Thu Apr  9 20:39:37 CEST 1998
	    root@tipcat.Hamartun.Priv.NO:/usr/src/sys/arch/i386/compile/TIPCAT
	com2 at isa0 port 0x3e8-0x3ef irq 5: ns16550a, working fifo
	biomask 840 netmask c40 ttymask c42

In /dev:

	crw-------  1 uucp  wheel  - 8, 2 Nov 24 21:40 /dev/tty02

>From /etc/ttys:

	tty02 "/usr/libexec/getty std.9600" unknown off softcar rtscts

>From /etc/rc.local:

	ifconfig sl0 192.168.3.1 192.168.3.2 netmask 0xffffff00
	slattach -h -s 19200 /dev/tty02

-tih
-- 
Popularity is the hallmark of mediocrity.  --Niles Crane, "Frasier"