From mmcg@heraclitus.cs.monash.edu.au  Mon Dec 16 01:09:23 1996
Received: from heraclitus.cs.monash.edu.au (heraclitus.cs.monash.edu.au [130.194.64.241])
          by freefall.freebsd.org (8.8.4/8.8.4) with ESMTP id BAA13795
          for <FreeBSD-gnats-submit@freebsd.org>; Mon, 16 Dec 1996 01:09:21 -0800 (PST)
Received: (from mmcg@localhost) by heraclitus.cs.monash.edu.au (8.7.5/8.7.3) id UAA03679; Mon, 16 Dec 1996 20:09:08 +1100 (EST)
Message-Id: <199612160909.UAA03679@heraclitus.cs.monash.edu.au>
Date: Mon, 16 Dec 1996 20:09:08 +1100 (EST)
From: Mike McGaughey <mmcg@heraclitus.cs.monash.edu.au>
Reply-To: mmcg@heraclitus.cs.monash.edu.au
To: FreeBSD-gnats-submit@freebsd.org
Subject: 2.1.6 fails to detect com2
X-Send-Pr-Version: 3.2

>Number:         2222
>Category:       kern
>Synopsis:       2.1.6 fails to detect com2
>Confidential:   no
>Severity:       serious
>Priority:       medium
>Responsible:    freebsd-bugs
>State:          closed
>Quarter:
>Keywords:
>Date-Required:
>Class:          sw-bug
>Submitter-Id:   current-users
>Arrival-Date:   Mon Dec 16 01:10:01 PST 1996
>Closed-Date:    Wed Dec 18 19:01:51 PST 1996
>Last-Modified:  Wed Dec 18 19:02:12 PST 1996
>Originator:     Mike McGaughey
>Release:        FreeBSD 2.1.6-RELEASE
>Organization:
Monash University 
>Environment:

2.1.6-R (actually 2.1.6.1).

(1) 486 DX33, 20M ram, 2x220M IDE, a bog standard `small' box.
(2) P120, 16M, 4G IDE disk, Mitsumi CDrom, sound blaster

Each has an (ahem) WD8003 ethernet card on IRQ 3.

>Description:

After the upgrade to 2.1.6, the probe for sio1 fails,
both using GENERIC and recompiled kernels.

The listed boxes were definitely fine under 2.1.0.  I'm not sure whether
they were working under 2.1.5 (I rarely see their console messages,
since I usually rlogin to these boxes).  It's unlikely
to be some bizarre simultaneous hardware failure (they're
ethernetted together, not laplinked).  Having the ether cards
configured to IRQ3 has never disturbed the serial probe before.

For one of the recompiled kernels (the P120):

FreeBSD 2.1.6-RELEASE #0: Mon Dec 16 17:18:07 EST 1996
    mmcg@mjolnir.cs.monash.edu.au:/usr/src/sys/compile/MJOLNIR
CPU: 120-MHz Pentium 735\\90 or 815\\100 (Pentium-class CPU)
[...]
sio0 at 0x3f8-0x3ff irq 4 on isa
sio0: type 16550A
sio1 not found at 0x2f8
sio2: disabled, not probed.

and - for the GENERIC kernel (on iapedus)

FreeBSD 2.1.6-RELEASE #0: Sat Nov 16 05:44:30  1996
    jkh@whisker.cdrom.com:/usr/src/sys/compile/GENERIC
CPU: i486DX (486-class CPU)
[...]
sio0 at 0x3f8-0x3ff irq 4 on isa
sio0: type 16450
sio1 not found at 0x2f8
sio2: disabled, not probed.

>How-To-Repeat:

Boot 2.1.6; dmesg.

>Fix:

Probably trivial - just thought I should mention it before you cut and ship
the 2.1.6 CDrom!

BTW: I found two other pr's reporting something like this - bin/1705 (`Com2
only detected when booting from floppy' - 2.0.5-R) and i386/2002 - failure
to detect com2 on a laptop, running 2.1.5-R.  A recurrent typo?

Give me a yell if you need any info - I'm quite happy to blow
either or both of those machines away.

Cheers,

    Mike.
>Release-Note:
>Audit-Trail:

From: Bruce Evans <bde@zeta.org.au>
To: FreeBSD-gnats-submit@freebsd.org, mmcg@heraclitus.cs.monash.edu.au
Cc:  Subject: Re: kern/2222: 2.1.6 fails to detect com2
Date: Mon, 16 Dec 1996 21:43:24 +1100

 >Each has an (ahem) WD8003 ethernet card on IRQ 3.
 >
 >>Description:
 >
 >After the upgrade to 2.1.6, the probe for sio1 fails,
 >both using GENERIC and recompiled kernels.
 >
 >The listed boxes were definitely fine under 2.1.0.  I'm not sure whether
 >they were working under 2.1.5 (I rarely see their console messages,
 >since I usually rlogin to these boxes).  It's unlikely
 >to be some bizarre simultaneous hardware failure (they're
 >ethernetted together, not laplinked).  Having the ether cards
 >configured to IRQ3 has never disturbed the serial probe before.
 
 The only relevant changes in sio in 2.1.6 are that ports aren't
 initialized unless they are configured (and enabled), and sio3 isn't
 enabled by default.  Now sio3 must be enabled manually if it exists and is
 on irq 3.  Otherwise, sio3 may conflict with sio1.  The conflict normally
 stops sio1 from working.  Enabling both sio1 and sio3 stops only the
 last one probed (usually sio3) from working (if it is on the same irq).
 
 I don't know why having the ethernet boards on the same irq doesn't
 cause similar problems.
 
 Bruce
State-Changed-From-To: open->closed 
State-Changed-By: gpalmer 
State-Changed-When: Wed Dec 18 19:01:51 PST 1996 
State-Changed-Why:  
Closed per customer request. 
>Unformatted:
