10 Sep
2002
10 Sep
'02
12:03 a.m.
Hi -
A few days ago the "getty plays parity games" topic came up
again.
I've been thinking about it and looked at the kernel a bit more.
7e1 is rather firmly in place and I'm not sure turning the 'console'
(kl/dl) driver into a "LITOUT only" one is a 100% correct way to
go (the other drivers such as dz, dh, etc remain 7e1 output by
default).
Here's a patch for getty/main.c which I think will do what's
wanted without having to modify the kernel:
------------snip-------------
*** main.c.dist Mon Sep 9 16:52:24 2002
--- main.c Mon Sep 9 16:56:55 2002
***************
*** 384,391 ****
c = cc;
c |= partab[c&0177] & 0200;
! if (OP)
! c ^= 0200;
if (!UB) {
outbuf[obufcnt++] = c;
if (obufcnt >= OBUFSIZ)
--- 384,401 ----
c = cc;
c |= partab[c&0177] & 0200;
! /*
! * If "any" parity do nothing otherwise set even parity unless OP is
! * set. Since 'ap' is set in the "default" entry of /etc/gettytab
this
! * has the effect of disabling parity on output without having to change
! * the kernel.
! */
! if (!AP) {
! c |= partab[c & 0177] & 0200;
! if (OP)
! c ^= 0200;
! }
!
if (!UB) {
outbuf[obufcnt++] = c;
if (obufcnt >= OBUFSIZ)
--------------snip---------------
What this does is check the "AP" (AnyParity) flag from /etc/gettytab
and if NOT set then do the 'even' (or 'odd' if OP is set) parity.
Since 'ap' is present in the default line of /etc/gettytab the above
block effectively becomes a no-op unless /etc/gettytab is explicitly
set for 'ep' or 'op'.
If someone could test this and report back I'd appreciate it.
Thanks.
Steven Schultz
sms(a)2bsd.com