From owner-freebsd-hackers Wed Jan 15 09:28:24 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.4/8.8.4) id JAA12297 for hackers-outgoing; Wed, 15 Jan 1997 09:28:24 -0800 (PST) Received: from dg-rtp.dg.com (dg-rtp.rtp.dg.com [128.222.1.2]) by freefall.freebsd.org (8.8.4/8.8.4) with SMTP id JAA12292 for ; Wed, 15 Jan 1997 09:28:19 -0800 (PST) Received: by dg-rtp.dg.com (5.4R3.10/dg-rtp-v02) id AA06594; Wed, 15 Jan 1997 12:20:02 -0500 Received: from ponds by dg-rtp.dg.com.rtp.dg.com; Wed, 15 Jan 1997 12:20 EST Received: from lakes.water.net (lakes [10.0.0.3]) by ponds.water.net (8.8.3/8.7.3) with ESMTP id JAA22685 for ; Wed, 15 Jan 1997 09:44:32 -0500 (EST) Received: (from rivers@localhost) by lakes.water.net (8.8.3/8.6.9) id JAA04441 for freebsd-hackers@freebsd.org; Wed, 15 Jan 1997 09:48:04 -0500 (EST) Date: Wed, 15 Jan 1997 09:48:04 -0500 (EST) From: Thomas David Rivers Message-Id: <199701151448.JAA04441@lakes.water.net> To: ponds!freebsd.org!freebsd-hackers Subject: 2.1.6.1 - SCSI problems (infinite bus reset) w/Aha2940UW. Content-Type: text Sender: owner-hackers@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk OK - I reported this with 2.1.5-STABLE, and after moving things around, I've (finally) gotten 2.1.6.1 installed on the same machine. The "moving around" was made quite difficult by the problem I'm about to describe; as it made writting a QIC-150 tape impossible. Now - the details: 2.1.6.1 running on a 430VX-based PCI Pentium133 with 32Meg, and Aha 2940UW controller - and a Wangtek 5150ES QIC-150. When I write a tape; the writes go fine until the very end, after the final write. The tape begins it's rewind sequence and suddenly everything goes haywire. I get a bunch a SCSI bus reset messages, aborted SCBs, etc... These, of course, are not logged to /var/log/messages, since we can no longer write to the disk - as the SCSI bus appears to be in some sort of "infinite reset" mode. A reset is issued, all the devices respond, the tape momentarily stops rewinding and then picks up again... I reported this initially in 2.1.5-STABLE when some new SCSI code was committed to that branch. It's still in 2.1.6.1. [See the mail archives; the last time I reported it, I hastily scribbled the syslog messages - they're the same in 2.1.6.1.] I'm happy to try anything out; if someone wants to debug this. Let me close with "This worked fine in 2.1.5-RELEASE, and the hardware hasn't changed" (Where have you heard that before :-) ) I'd like to make sure this isn't a problem in 2.2. So, I'm willing to mess around with things; etc... to reach that goal. - Dave Rivers -