From owner-freebsd-hackers Mon Nov 4 21:20:42 1996 Return-Path: owner-hackers Received: (from root@localhost) by freefall.freebsd.org (8.7.5/8.7.3) id VAA04715 for hackers-outgoing; Mon, 4 Nov 1996 21:20:42 -0800 (PST) Received: from dg-rtp.dg.com (dg-rtp.rtp.dg.com [128.222.1.2]) by freefall.freebsd.org (8.7.5/8.7.3) with SMTP id VAA04687 for ; Mon, 4 Nov 1996 21:20:34 -0800 (PST) Received: by dg-rtp.dg.com (5.4R3.10/dg-rtp-v02) id AA00799; Tue, 5 Nov 1996 00:20:02 -0500 Received: from ponds by dg-rtp.dg.com.rtp.dg.com; Tue, 5 Nov 1996 00:20 EST Received: from lakes.water.net (lakes [10.0.0.3]) by ponds.water.net (8.7.5/8.7.3) with ESMTP id XAA06217 for ; Mon, 4 Nov 1996 23:13:15 -0500 (EST) Received: (from rivers@localhost) by lakes.water.net (8.7.5/8.6.9) id XAA00266 for freebsd-hackers@freefall.cdrom.com; Mon, 4 Nov 1996 23:14:10 -0500 (EST) Date: Mon, 4 Nov 1996 23:14:10 -0500 (EST) From: Thomas David Rivers Message-Id: <199611050414.XAA00266@lakes.water.net> To: ponds!freefall.cdrom.com!freebsd-hackers Subject: Another round of aha2940 (or perhaps generic SCSI) lock ups.. Content-Type: text Sender: owner-hackers@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk Ok - I've done everything Adaptec tech. support said to do. I've ensure proper termination; dropped the transfer times, disabled synchronous & wide negotation on the older devices, disconnection negotiation, etc... What this has done has let me reproduce the SCSI lock-up more readily. Now, every time I write to the tape device (a Wangtek 5150ES QIC-150), when I get to the end of the writing, and the tape begins to rewind; *whamo* - my SCSI bus locks up.... The messages, of course, are not logged in /var/log/messages - because nothing can be written out. I've jotted them down - from my paper notes they are: ahc0: Issued Channel A Bus Reset #1. n SCB's aborted sd0(ahc0:0:0) timed out in message out phase, SCSISGI==0xb4 sd0(ahc0:0:0) asserted ATN - device reset in message buffer sd0(ahc0:0:0) timed out in message out phase, SCSISGI==0xa4 This just repeats until I reboot the machine. I let it continue for about 5 minutes. The only variation was the number of SCBs aborted. [Note, if I do a reset - the aha2940 is unable to probe the SCSI bus for devices. I have to turn the machine off and back on.] I note that "Vladimir P. Frolov" has just reported the same problem; with an Adaptec 7880. I did not experience my problems until *after* I upgraded my kernel to 2.1.5-STABLE. - Dave Rivers -