Date: Thu, 15 Jan 1998 21:32:05 -0800 (PST) From: sherwink@ix.netcom.com To: freebsd-gnats-submit@FreeBSD.ORG Subject: kern/5508: SCSI Message sd0: COMMAND FAILED (4 28) @f04e1000 during make buildworld Message-ID: <199801160532.VAA27190@hub.freebsd.org>
next in thread | raw e-mail | index | archive | help
>Number: 5508 >Category: kern >Synopsis: SCSI Message sd0: COMMAND FAILED (4 28) @f04e1000 during make buildworld >Confidential: no >Severity: non-critical >Priority: medium >Responsible: freebsd-bugs >State: open >Quarter: >Keywords: >Date-Required: >Class: sw-bug >Submitter-Id: current-users >Arrival-Date: Thu Jan 15 21:40:01 PST 1998 >Last-Modified: >Originator: Sherwin Kaplan >Organization: >Release: 2.2.5 Stable as of 11/29/97 >Environment: FreeBSD bookend.ix.netcom.com 2.2.5-STABLE FreeBSD 2.2.5-STABLE #0: Wed Jan 14 05:07:17 PST 1998 sherwin@bookend.ix.netcom.com:/usr/src/sys/compile/SHERWIN i386 >Description: Approximately one hour into a "make buildworld", I received the following messages: Jan 13 23:49:11 bookend /kernel: assertion "cp" failed: file "../../pci/ncr.c", line 6191 Jan 13 23:49:12 bookend /kernel: sd0(ncr0:0:0): COMMAND FAILED (4 28) @f05c8800. Jan 13 23:49:12 bookend /kernel: assertion "cp" failed: file "../../pci/ncr.c", line 6191 Jan 13 23:49:12 bookend /kernel: sd0(ncr0:0:0): COMMAND FAILED (4 28) @f05c8800. The same messages were also displayed during "make buildworld" using both 3.0 Current 01/01/98 and 01/08/98, on the same machine, using a different slice of the same disk. In the case of the 01/01/98 system, there were 33 sets of messages, in the case of the 01/08/98 system, there were 3 sets of messages. The messages above are from the make of the 2.2.5-STABLE shown in the Environment box, running on the old 2.2.5-STABLE of 11/29/97. I have not tried running a make using this 2.2.5-STABLE of 01/14/98 system. The one time that I actually saw the messages displayed, instead of seeing them after the fact in /var/log/messages, there was very intense disk activity. Hardware is: Pentium Pro 150 Intel Venus Motherboard 32 MB memory Symbios 53c825 Quantum Atlas II, 2.2 GB Ultra SCSI >How-To-Repeat: Run a "make buildworld", it has happened 3 out of 3 times. The disk does pass a fsck, and the code generated does seem to work properly. >Fix: None Known. The error appears to indicate a ccb queue full, but I am not clear from looking at the code of ncr.c, as to why this occurs. Is this a problem with the code, or is my disk failing to respond? I am also not clear as to whether the command is retried. The application may retry, the SCSI code appears not to retry. I would appreciate any advice. >Audit-Trail: >Unformatted:
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199801160532.VAA27190>