Skip site navigation (1)Skip section navigation (2)
Date:      Sat, 20 Nov 1999 21:21:31 +0100 (MET)
From:      j@uriah.heep.sax.de (J Wunsch)
To:        freebsd-current@freebsd.org
Subject:   Re: FreeBSD 4.0 SCSI Tape Driver- Okay, Okay, you win....
Message-ID:  <199911202021.VAA11447@uriah.heep.sax.de>
References:  <Pine.BSF.4.10.9911160735130.98613-100000@beppo.feral.com> <Pine.BSF.4.10.9911160803220.98650-100000@beppo.feral.com>

next in thread | previous in thread | raw e-mail | index | archive | help
Matthew Jacob <mjacob@feral.com> wrote:

> Too many people have objected. I didn't make my case clearly enough,
> but because enough people of have raised issues, the default won't
> be changed.

Too bad.  I think your idea was absoultely right, and i'm rather tired
myself to have to `fix' it for every new tape drive we encounter.  All
the current tape drives do support a form of physical EOD marks, where
you can't read on once you've hit EOD.

The driver could easily have faked the double-filemark at EOT, so the
API wouldn't have changed at all.  (Just as you see the EOD from the
tape, finish the current transfer and return a "short read", and being
asked a second time, return a null read.  Upon further attempts,
return an error.)

Btw., Matt, couldn't modepage 0x10, field "EOD defined" be used in
order to decide whether a tape drive supports a physical EOD notation,
and thus doesn't require two filemarks at EOD?  If i read the standard
correctly, those fields should be set to 3 for a tape drive that
doesn't support physical EOD.  (I can't verify, i don't have any such
drive.)

-- 
cheers, J"org

joerg_wunsch@uriah.heep.sax.de -- http://www.sax.de/~joerg/ -- NIC: JW11-RIPE
Never trust an operating system you don't have sources for. ;-)


To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-current" in the body of the message




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199911202021.VAA11447>