From owner-freebsd-scsi Mon Aug 28 20:36:57 2000 Delivered-To: freebsd-scsi@freebsd.org Received: from feral.com (feral.com [192.67.166.1]) by hub.freebsd.org (Postfix) with ESMTP id 1B0BD37B423 for ; Mon, 28 Aug 2000 20:36:55 -0700 (PDT) Received: from beppo.feral.com (beppo [192.67.166.79]) by feral.com (8.9.3/8.9.3) with ESMTP id UAA25441; Mon, 28 Aug 2000 20:36:45 -0700 Date: Mon, 28 Aug 2000 20:36:42 -0700 (PDT) From: Matthew Jacob Reply-To: mjacob@feral.com To: Greg Lehey Cc: Sam , freebsd-scsi@FreeBSD.ORG Subject: Re: "tape is now frozen" In-Reply-To: <20000829130040.Q11422@wantadilla.lemis.com> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-scsi@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org On Tue, 29 Aug 2000, Greg Lehey wrote: > On Tuesday, 8 August 2000 at 16:39:03 -0700, Matt Jacob wrote: > >> > >> Indeed, it looks like he omitted the message that we most wanted to > >> see. > >> > >> Let me repeat here that I find this *very* irritating. It happens, > >> for example, if I try to read a block which is too long. There's no > >> way to know the length of a tape block in advance, so this is > >> relatively easy to get, particularly with DDS-4 drives, and it > > > > This error should not occur if you're in variable block mode. If you > > set the drive in fixed block mode and read a block that's too large, > > the tape driver cannot know where the tape heads are located. It's > > that simple. > > It's not that simple. If it happens in the middle of the tape, I need > to rewind the bloody thing to recover. At least an fsf or bsf should > be sufficient. So you space a filemark. Where specfically on the tape are you given you've lost knowledge of where you are? The whole point of rewind, eom or offline is to bring the tape to a *known* place. Spacing one filemark is not sufficient. I thought for awhile about allowing the use of rdhpos/sethpos to allow for unfreezing. You're also, again, begging the question as to why it has occurred. It has occurred because there was an I/O error, or you're not using the tape correctly (fixed block mode and you don't issue the correct read size). I keep on getting pushback on this. Perhaps an 'unfreeze' ioctl, call it MTIOSUICIDE, is in order. Not. -matt To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-scsi" in the body of the message