From owner-freebsd-scsi Sun Nov 26 1:47:14 2000 Delivered-To: freebsd-scsi@freebsd.org Received: from funkthat.com (mg128-062.ricochet.net [204.179.128.62]) by hub.freebsd.org (Postfix) with ESMTP id 4EC6937B4CF for ; Sun, 26 Nov 2000 01:47:10 -0800 (PST) Received: (from jmg@localhost) by funkthat.com (8.9.3/8.8.7) id BAA02682; Sun, 26 Nov 2000 01:48:10 -0800 (PST) Message-ID: <20001126014805.22964@hydrogen.funkthat.com> Date: Sun, 26 Nov 2000 01:48:05 -0800 From: John-Mark Gurney To: freebsd-scsi@freebsd.org Subject: Re: ncr problems (and amd) References: <20001125142603.14261@hydrogen.funkthat.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Mailer: Mutt 0.69 In-Reply-To: <20001125142603.14261@hydrogen.funkthat.com>; from John-Mark Gurney on Sat, Nov 25, 2000 at 02:26:03PM -0800 Reply-To: John-Mark Gurney Organization: Cu Networking X-Operating-System: FreeBSD 3.4-RELEASE i386 X-PGP-Fingerprint: B7 EC EF F8 AE ED A7 31 96 7A 22 B3 D8 56 36 F4 X-Files: The truth is out there X-URL: http://resnet.uoregon.edu/~gurney_j/ Sender: owner-freebsd-scsi@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org John-Mark Gurney scribbled this message on Nov 25: looks like I left out an important part of the log: (da0:ncr0:0:0:0): phase change 2-3 10@038ee29c resid=4. (da0:ncr0:0:0:0): WRITE(10). CDB: 2a 0 0 3f 8a e0 0 0 60 0 (da0:ncr0:0:0:0): ILLEGAL REQUEST asc:24,0 (da0:ncr0:0:0:0): Invalid field in CDB which happened about at least six minutes before the following started (can tell by syslog messages to the console) > ncr0: timeout nccb=0xc0d92000 (skip) then later in the add (after a reboot), a similar event happened: ncr0: SCSI phase error fixup: CCB address mismatch (0xc120ac00 != 0xc11fca00) np->nccb = 0xc120ac00 (da2:ncr0:0:4:0): phase change 6-3 3@03efbdb4 resid=3. ncr0: queue empty. ncr0: SCSI phase error fixup: CCB already dequeued (0xc120ac00) ncr0: timeout nccb=0xc0f05800 (skip) ncr0: timeout nccb=0xc0f05a00 (skip) ncr0: timeout nccb=0xc0ecb800 (skip) ncr0: timeout nccb=0xc0ecbc00 (skip) ncr0: timeout nccb=0xc120ac00 (skip) which is kinda interesting, as it looks like the ccb that caused this isn't even valid (the head disaggress what virtual address it is at..) of course, I'm having troubles wrapping my head aroun the ncr driver as most of it isn't commented very well.. the last lockup had this on output: (pass1:ncr0:0:1:0): MSG_DISCONNECT received, but datapointer not saved: data=2274d90 save=e71026a0 goal=e71026c4. (pass1:ncr0:0:1:0): MSG_DISCONNECT received, but datapointer not saved: data=2274d90 save=e71026a0 goal=e71026c4. (pass2:ncr0:0:2:0): MSG_DISCONNECT received, but datapointer not saved: data=38bf390 save=e71026a0 goal=e71026c4. (pass1:ncr0:0:1:0): MSG_DISCONNECT received, but datapointer not saved: data=2274d90 save=e71026a0 goal=e71026c4. (pass2:ncr0:0:2:0): MSG_DISCONNECT received, but datapointer not saved: data=38bf190 save=e71026a0 goal=e71026c4. Nov 25 20:47:00 hydrogen su: jmg to root on /dev/ttyp6 [other unrelated activity eleted] Nov 25 22:04:07 hydrogen su: jmg to root on /dev/ttyp9 ncr0: SCSI phase error fixup: CCB address mismatch (0xc0f5c200 != 0xc0eff200) np->nccb = 0xc0f5c200 (da2:ncr0:0:4:0): phase change 6-3 3@049553b4 resid=3. ncr0: SCSI phase error fixup: CCB already dequeued (0xc0f5c200) ncr0: timeout nccb=0xc0eff400 (skip) hopefully someone can figure out why it enters the timeout and never completes... or at least provide some help in debugging this.. -- John-Mark Gurney Voice: +1 408 975 9651 Cu Networking "Thank God I'm an atheist, that'd just be confusing." -- cmc To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-scsi" in the body of the message