From owner-cvs-all Tue Oct 13 08:48:07 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id IAA07755 for cvs-all-outgoing; Tue, 13 Oct 1998 08:48:07 -0700 (PDT) (envelope-from owner-cvs-all@FreeBSD.ORG) Received: from panzer.plutotech.com (panzer.plutotech.com [206.168.67.125]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id IAA07749; Tue, 13 Oct 1998 08:48:04 -0700 (PDT) (envelope-from ken@panzer.plutotech.com) Received: (from ken@localhost) by panzer.plutotech.com (8.9.1/8.8.5) id JAA03036; Tue, 13 Oct 1998 09:47:37 -0600 (MDT) From: "Kenneth D. Merry" Message-Id: <199810131547.JAA03036@panzer.plutotech.com> Subject: Re: cvs commit: src/sys/cam/scsi scsi_da.c In-Reply-To: from Blaz Zupan at "Oct 13, 98 08:23:44 am" To: blaz@gold.amis.net (Blaz Zupan) Date: Tue, 13 Oct 1998 09:47:37 -0600 (MDT) Cc: cvs-committers@FreeBSD.ORG, cvs-all@FreeBSD.ORG X-Mailer: ELM [version 2.4ME+ PL28s (25)] MIME-Version: 1.0 Content-Type: multipart/mixed; boundary=ELM908293657-3005-0_ Content-Transfer-Encoding: 7bit Sender: owner-cvs-all@FreeBSD.ORG Precedence: bulk --ELM908293657-3005-0_ Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Blaz Zupan wrote... > > > Yes, one of my disks completely hangs so a CAM-ified kernel can not boot. > > > One disk is an old Quantum 540MB, the other is a 1.5GB NEC. I suspect the > > > Quantum, I'll be finding out which one is the culprit tomorrow. > > > > Then you've probably got a different problem. The cache sync doesn't > > happen until final close, or system shutdown. > > Well, the following patch works just fine: > > *** /sys/cam/scsi/scsi_da.c.orig Thu Sep 17 22:19:58 1998 > --- /sys/cam/scsi/scsi_da.c Thu Sep 17 22:22:57 1998 > *************** > *** 358,363 **** > --- 358,364 ---- > > dsclose(dev, fmt, softc->dk_slices); > > + #if 0 > ccb = cam_periph_getccb(periph, /*priority*/1); > > scsi_synchronize_cache(&ccb->csio, > *************** > *** 374,379 **** > --- 375,381 ---- > /*sense_flags*/0, &softc->device_stats); > > xpt_release_ccb(ccb); > + #endif > > if ((softc->flags & DA_FLAG_PACK_REMOVABLE) != 0) { > daprevent(periph, PR_ALLOW); > > > In fact, I'm writing this mail from a machine with the above patch. I was > told by Justin that you were working on a quirk entry which would enable > to selectively turn off the above for certain disks, and I believe the > patch in the subject is it. Or am I mistaken? My two disks are: You're correct. It must be that we get a close during the boot process. > I believe the Quantum is at fault, but will know for sure later today, > when I get home from work. Try applying the attached patch against revision 1.10 of scsi_da.c. If your Quantum drive is at fault, this should have the same effect as your patch above. Ken -- Kenneth Merry ken@plutotech.com --ELM908293657-3005-0_ Content-Type: text/plain; charset=ISO-8859-1 Content-Disposition: attachment; filename=scsi_da.quantum_patch.101398 Content-Description: scsi_da.quantum_patch.101398 Content-Transfer-Encoding: 7bit ==== //depot/cam/sys/cam/scsi/scsi_da.c#86 - /usr/home/ken/perforce/cam/sys/cam/scsi/scsi_da.c ==== *** /tmp/tmp.23963.0 Tue Oct 13 09:43:52 1998 --- /usr/home/ken/perforce/cam/sys/cam/scsi/scsi_da.c Tue Oct 13 09:25:12 1998 *************** *** 138,143 **** --- 138,151 ---- */ {T_DIRECT, SIP_MEDIA_FIXED, "MICROP", "2217*", "*"}, /*quirks*/ DA_Q_NO_SYNC_CACHE + }, + { + /* + * Doesn't like the synchronize cache command. + * Reported by: Blaz Zupan + */ + {T_DIRECT, SIP_MEDIA_FIXED, "QUANTUM", "LPS540*", "*"}, + /*quirks*/ DA_Q_NO_SYNC_CACHE } }; --ELM908293657-3005-0_-- To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe cvs-all" in the body of the message