From owner-cvs-sys  Thu Apr  6 05:38:03 1995
Return-Path: cvs-sys-owner
Received: (from majordom@localhost)
          by freefall.cdrom.com (8.6.10/8.6.6) id FAA25306
          for cvs-sys-outgoing; Thu, 6 Apr 1995 05:38:03 -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 FAA25299
          ; Thu, 6 Apr 1995 05:37:46 -0700
Received: (from bde@localhost) by godzilla.zeta.org.au (8.6.9/8.6.9) id WAA20371; Thu, 6 Apr 1995 22:35:35 +1000
Date: Thu, 6 Apr 1995 22:35:35 +1000
From: Bruce Evans <bde@zeta.org.au>
Message-Id: <199504061235.WAA20371@godzilla.zeta.org.au>
To: ache@astral.msk.su, bde@zeta.org.au, phk@ref.tfs.com
Subject: Re: cvs commit: src/sys/i386/isa isa.c
Cc: CVS-commiters@freefall.cdrom.com, cvs-sys@freefall.cdrom.com
Sender: cvs-sys-owner@freebsd.org
Precedence: bulk

>>Sorry, it's wrong for another reason.  id_iobase == 0 is the default
>>magic value for all devices without i/o ports (all memory mapped
>>devices...). ...

>Does it wrong to say "on motherboard" for memory-mapped device?
>If does, what is correct line to say?

I think "on isa" is right if there is any isa memory or ports.

>From my boot messages:

	cyb0 at 0x0 irq 10 maddr 0xd4000
	 ports 8
	...
	pca0 not probed due to I/O address conflict with cyb0 at 0x0

They should be:

	cyb0 irq 10 maddr 0xd4000 msize 0x2000 on isa ports 8

Many bugs are evident:
"at 0 " and the pca0 conflict are because of a driver bug (cyb hasn't
been fully ported to FreeBSD and cyprobe() returns 1 instead of -1 for
success && no-ports).

"\n ports 8" is because cybattach() doesn't expect its caller to have
printed the \n.  This is not related to the recent \n breakage.

I forgot to specify the msize.  The driver didn't set it.  It defaults
to 0.  isa.c didn't catch this error and printed the maddr.  I think
some other config lines/drivers have the same problem.  Perhaps the
memory address range should be printed instead of the base and size.
This would save a little space, and the bug in the range 0xd400-0xd3ff
would be very obvious.

>>The pca driver is on the motherboard but its id_iobase is only 0
>>because it lies - see other mail.

>IO_TIMER1 and IO_PPI not specified directly
>in any config description now. Can we pick up one of them for pca?

Yes, IO_PPI would be best for now.

Bruce