From owner-freebsd-hackers Mon Mar 31 17:50:37 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id RAA27847 for hackers-outgoing; Mon, 31 Mar 1997 17:50:37 -0800 (PST) Received: from dg-rtp.dg.com (dg-rtp.rtp.dg.com [128.222.1.2]) by freefall.freebsd.org (8.8.5/8.8.5) with SMTP id RAA27842 for ; Mon, 31 Mar 1997 17:50:34 -0800 (PST) Received: by dg-rtp.dg.com (5.4R3.10/dg-rtp-v02) id AA00630; Mon, 31 Mar 1997 20:50:02 -0500 Received: from ponds by dg-rtp.dg.com.rtp.dg.com; Mon, 31 Mar 1997 20:50 EST Received: from lakes.water.net (lakes [10.0.0.3]) by ponds.water.net (8.8.3/8.7.3) with ESMTP id UAA06227 for ; Mon, 31 Mar 1997 20:38:28 -0500 (EST) Received: (from rivers@localhost) by lakes.water.net (8.8.3/8.6.9) id UAA00322 for freebsd-hackers@freefall.cdrom.com; Mon, 31 Mar 1997 20:44:07 -0500 (EST) Date: Mon, 31 Mar 1997 20:44:07 -0500 (EST) From: Thomas David Rivers Message-Id: <199704010144.UAA00322@lakes.water.net> To: ponds!freefall.cdrom.com!freebsd-hackers Subject: aha2940 problems on 2.1.7.1. Content-Type: text Sender: owner-hackers@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk Just to let everyone know what I've tried, in an attempt to diagnose the aha2940 problems in 2.1.7.1. I increased (to 20 minutes) all of the timeout parms in the scsi_scsi_cmd() calls in st.c; thinking that perhaps the write at the end of the time was timing out too soon; causing my problems. [Recall, my problem is that I can write to a Wangtek 5150ES (QIC-150) and fill the tape up; which locks down my 2.1.7.1 system completely - this is a 2.1.6.1 system with a 2.1.7.1 kernel.] However, this didn't help the problem, when the write that fills the tape up completes I get (on the console): sd0(ahc0:0:0): SCB 0x2 - timed out in command phase, SCSISIGI == 0x84 SEQADDR == 0x42 st0(ahc0:2:0): abort message in message buffer Note that ahc0:0:0 is my primary disk drive; apparently filling the tape up caused a SCSI bus reset which didn't do much for my primary disk drive. Also notice that I didn't get any I/O about the abort having completed... everything except ping'ing the machine has "gone south" at this point. Also, after a press-the-reset-button reboot (not a complete shutdown) I got as far as starting login and xdm when suddenly, I got: sd0(ahc0:0:0) SCB 0x0 - timed out in message out phasse, SCSISIGI == 0xa4 SEQADDR == 0x99 sd0(ahc0:0:0): abort message in message buffer sd0(ahc0:0:0): SCB 0 - Abort Completed. panic: Couldn't find next SCB A turn-off-the-machine/cold reboot seems to have gotten me working again... (phew) However; this means two things: 1) 2.1.7.1's AHA2940 support has a problem. 2) My idea about needing longer timeouts in st.c has nothing to do with the problem. I really need to back stuff up before an upgrade to 2.2.1 - does anyone have any suggestions? - Dave Rivers - p.s. Here's the pertinent dmesg from the last boot; to give everyone an idea about the devices I have: FreeBSD 2.1.7.1-RELEASE #1: Mon Mar 31 19:38:48 EST 1997 rivers@lakes.water.net:/usr/src/sys-2.1.7.1/compile/LAKES CPU: 133-MHz Pentium 735\\90 or 815\\100 (Pentium-class CPU) Origin = "GenuineIntel" Id = 0x52c Stepping=12 Features=0x1bf real memory = 33554432 (32768K bytes) avail memory = 30457856 (29744K bytes) Probing for devices on PCI bus 0: chip0 rev 1 on pci0:0 chip1 rev 1 on pci0:7:0 chip2 rev 0 on pci0:7:1 ahc0 rev 0 int a irq 15 on pci0:17 ahc0: aic7880 Wide Channel, SCSI Id=7, 16 SCBs ahc0 waiting for scsi devices to settle (ahc0:0:0): "HP C3323-300 4242" type 0 fixed SCSI 2 sd0(ahc0:0:0): Direct-Access 1003MB (2056008 512 byte sectors) (ahc0:1:0): "MICROP 1548-15MZ1077802 HZ2P" type 0 fixed SCSI 1 sd1(ahc0:1:0): Direct-Access 1635MB (3349512 512 byte sectors) (ahc0:2:0): "WANGTEK 5150ES SCSI FA23 08" type 1 removable SCSI 1 st0(ahc0:2:0): Sequential-Access drive offline (ahc0:3:0): "NEC CD-ROM DRIVE:400 1.0" type 5 removable SCSI 2 cd0(ahc0:3:0): CD-ROM cd present.[217422 x 2048 byte records]