From owner-freebsd-commit  Wed Jun 14 02:52:58 1995
Return-Path: commit-owner
Received: (from majordom@localhost)
          by freefall.cdrom.com (8.6.10/8.6.6) id CAA28547
          for commit-outgoing; Wed, 14 Jun 1995 02:52:58 -0700
Received: (from majordom@localhost)
          by freefall.cdrom.com (8.6.10/8.6.6) id CAA28410
          for cvs-sys-outgoing; Wed, 14 Jun 1995 02:51:28 -0700
Received: from godzilla.zeta.org.au (godzilla.zeta.org.au [203.2.228.34])
          by freefall.cdrom.com (8.6.10/8.6.6) with ESMTP id CAA28353
          ; Wed, 14 Jun 1995 02:51:02 -0700
Received: (from bde@localhost) by godzilla.zeta.org.au (8.6.9/8.6.9) id TAA30366; Wed, 14 Jun 1995 19:45:24 +1000
Date: Wed, 14 Jun 1995 19:45:24 +1000
From: Bruce Evans <bde@zeta.org.au>
Message-Id: <199506140945.TAA30366@godzilla.zeta.org.au>
To: bde@freefall.cdrom.com, rgrimes@gndrsh.aac.dev.com
Subject: Re: cvs commit: src/sys/i386/isa isa.h
Cc: CVS-commiters@freefall.cdrom.com, cvs-sys@freefall.cdrom.com
Sender: commit-owner@FreeBSD.org
Precedence: bulk

>>   Remove bogus comments about com ports having i/o size 4.

>What was the reference document for this???  I was the one who

sio.c.  It accesses the 8 xx(x)50 registers at linear offsets from
the base port.

>put that comment in there, and it comes from no less than IBM
>documents itself.  The fact is PS/2 class machines only have 4
>addresses for COM ports, and can have a com port at 0x3fc!!

>>   The i/o sizes are unreliable because of split address ranges
>>   for many devices (VGA, wd).

>VGA???  Maybe if you consider the fact that the ``syscons'' driver
>should be split so that we don't have both the VGA ports and the
>Keyboard ports all in one driver (along with a ton of architecual
>reasons to split this out).

Splitting wouldn't help much.  Beeping belongs in the output part
of the driver.  VGAs have the follow scattered ports according to
an old book by Ferraro:

PC/AT card	Global Enable	46E8	w/o	# sometimes decoded as 2E8
PC/AT all	Card Enable	102	r/w	# I don't remember seeing this
Standard VGA			3C0-3CF
Standard VGA	Feature Control	3BA/3DA		# mono/color (nearby ports too?)
Standard VGA	CRTC		3B4-5/3D4-5	# mono/color (nearby ports too?)
...

Then there are individual cards, more modern cards (SVGAs) and older
cards (CGAs, HGCs).  3B0-3DF covers most of the addresses (not 102,
46E8, or 4BC4-5 (V7VGA Cache Control)).  3B0-3DF is the best that can be
done in isa.h.  It only overlaps LPT3.  In fact, isa.h already describes
this range in 3 pieces of size 16 (MDA/CGA/VGA) but nothing uses the
definitions.  Config files don't use the definitions because a combined
range is wanted and the keyboard and screen addresses can't both be
configured.  syscons doesn't use them - it prefers its own defines for
almost everything.

>Pretty much I created the current content of that file from official
>IBM documentation for the XT, AT, and PS/2 family.

IBM hasn't been setting PC standards since about 1990.  This is more clear
now than 1993.

Bruce