From owner-freebsd-hardware Wed Jun 14 03:48:04 1995 Return-Path: hardware-owner Received: (from majordom@localhost) by freefall.cdrom.com (8.6.10/8.6.6) id DAA04601 for hardware-outgoing; Wed, 14 Jun 1995 03:48:04 -0700 Received: from avangard.master.nsk.su (small.master.nsk.su [193.124.45.26]) by freefall.cdrom.com (8.6.10/8.6.6) with ESMTP id DAA04308 for ; Wed, 14 Jun 1995 03:45:11 -0700 Received: (from root@localhost) by avangard.master.nsk.su (8.6.11/8.6.9) id RAA22014; Wed, 14 Jun 1995 17:39:04 +0700 To: dufault@hda.com Cc: freebsd-hardware@FreeBSD.ORG References: <199506140959.FAA14968@hda.com> Message-ID: Organization: Master Net Laboratory From: Serge Goncharov Date: Wed, 14 Jun 95 17:39:02 +0700 X-Mailer: BML [UNIX Beauty Mail v.1.39] Subject: Re: Pioneer DRM 604X CD ROM changer under FreeBSD 2.0.0-SNAP950322 Sender: hardware-owner@FreeBSD.ORG Precedence: bulk > Serge Goncharov writes: > > > > I have Pioneer DRM-604x drive. This is SCSI-2 CDROM changer (jukebox) > > for 6 CDs. It represents itself as SCSI ID 5 (for example) and 6 LUNs within > > it. It can be mounted as 6 different CDs and changer substitute corresponding > > disk at access query. > > That SCSI device was successfully configured and works under > > FreeBSD 1.1.5 by recompiling kernel with follow lines in config file: > > > > device cd0 > > device uk0 > > > > and devices /dev/cd[0-5] was made. All works fine, all LUNs are recognized, > > probed and configured during boot. No problem. > > > > Otherwise in SNAP950322. After adding > > device cd0 > > in config file, system recognize only one (first) CD, which have LUN=0. > > Seems to be only first (0) LUN are probed by default, and result of > > "scsi -f /dev/cd0a -p" command to reprobe whole device (all LUNs) is: > > > > "0:inappropriate ioctl for device: SCIOCADDR" > > "...." > > "6:inappropriate ioctl for device: SCIOCADDR" > > The cd device is rejecting this SCIOCADDR request to "pretend" it is a > different device; I think cd0c might accept it. > > > Looking through the code, I've found and add > > > > options NEW_SCSICONF > > > > in config. Now system probe _all_ CD LUNs and I can mount its, > > Apparently OLD_SCSICONF is broken, since there is a match in there for > your device saying to probe more luns. > > > but system dumps to core randomizely at "bio" (suppose, DMA operations) when > > I tries to execute some "ls -lR /cd's_mountpoint/cd[0-5]& ; du / &" processes. > > You mean it panics, not core dumps, right? And the stack backtrace is > always different? > Right. System always panics at disk operations and writes something like "Page not found". > -- > Peter Dufault Real Time Machine Control and Simulation > HD Associates, Inc. Voice: 508 433 6936 > dufault@hda.com Fax: 508 433 5267 > --- Serge