NetBSD Problem Report #216
From gnats Fri Apr 8 02:02:09 1994
Received: from aie8.aie.nl (root@aie8.aie.nl [193.78.1.40]) by sun-lamp.cs.berkeley.edu (8.6.8/8.6.6) with ESMTP id CAA29809 for <gnats-bugs@sun-lamp.cs.berkeley.edu>; Fri, 8 Apr 1994 02:02:03 -0700
Message-Id: <199404080902.LAA00411@aie8.aie.nl>
Date: Fri, 8 Apr 1994 11:02:29 +0200
From: sreiz@aie.nl
Reply-To: sreiz@aie.nl
To: gnats-bugs@sun-lamp.cs.berkeley.edu
Subject: system with 1542CF hangs occasionally
X-Send-Pr-Version: 3.2
>Number: 216
>Category: port-i386
>Synopsis: sd0(aha0:6:0): timed out, adapter not taking commands, frozen?!
>Confidential: no
>Severity: serious
>Priority: medium
>Responsible: gnats-admin (GNATS administrator)
>State: closed
>Class: sw-bug
>Submitter-Id: net
>Arrival-Date: Fri Apr 08 02:05:08 +0000 1994
>Closed-Date: Wed Dec 28 19:51:43 +0000 1994
>Last-Modified:
>Originator: Steven Reiz
>Release: NetBSD-current of Apr 6, 1994 (kernel and userland binaries)
>Organization:
A.I. Engineering, Amsterdam
>Environment:
486DX/33 ISA with 2 ESDI disks on WD1007V, Adaptec 1542CF with
one SCSI disk, a Conner CFP1060S 1.05 GB SCSI 2 on id 6.
System: NetBSD aie8.aie.nl 0.9a AIE_ALL#0 i386
>Description:
After on the average two days of uptime, with disk i/o
on the SCSI disk at least once every 10 minutes (news disk),
the machine produces the following lines:
sd0(aha0:6:0): timed out
sd0(aha0:6:0): timed out
adapter not taking commands.. frozen?!
and jumps into the debugger, after a reset everything is
(temporarily) ok again.
>How-To-Repeat:
The bug(?) is typically NOT triggered during heavy disk i/o,
but typically at night or during the weekend.
>Fix:
I've tried a lot of settings combinations in the 1542CF
bios (max 3.3 MB transfer, no sync neg., no fast SCSI),
doesn't make a difference, the SCSI driver of before the
last change (say jan., feb. 1994) had this problem as well.
It may well be a problem with the Conner disk, even though it
works for days just fine. However, I would really appreciate it
if the 1542 driver would try to reset the adapter and/or devices
instead of just jumping into the debugger. I am, of course,
available to test new code.
Timeouts can occur while the disk is doing error recovery. This
is usually okay. The real problem is that the aha driver was
not handling timeouts correctly, causing the kernel to eventually
panic. This is fixed. - mycroft
>Release-Note:
>Audit-Trail:
State-Changed-From-To: open->closed
State-Changed-By: mycroft
State-Changed-When: Wed Dec 28 11:51:43 PST 1994
State-Changed-Why:
I believe this is fixed. If you have further problems, let me know.
>Unformatted:
(Contact us)
$NetBSD: query-full-pr,v 1.51 2026/08/10 02:28:17 riastradh Exp $
$NetBSD: gnats_config.sh,v 1.10 2026/05/13 22:00:09 riastradh Exp $
Copyright © 1994-2026
The NetBSD Foundation, Inc. ALL RIGHTS RESERVED.