Skip site navigation (1)Skip section navigation (2)
Date:      Tue, 4 Feb 1997 09:44:27 +0100
From:      j@uriah.heep.sax.de (J Wunsch)
To:        freebsd-current@freebsd.org
Subject:   Re: Good name for a dump(8) option?
Message-ID:  <Mutt.19970204094427.j@uriah.heep.sax.de>
In-Reply-To: <m0vrc6T-000utcC@nemesis.lonestar.org>; from Frank Durda IV on Feb 3, 1997 22:03:00 -0500
References:  <m0vrc6T-000utcC@nemesis.lonestar.org>

next in thread | previous in thread | raw e-mail | index | archive | help
As Frank Durda IV wrote:

> Actually, "e" is an excellent choice since the official name for that point
> on the tape is the "EOT" (End-Of-Tape), whether it be a reflective
> marker or punch-out in the media.

Hmm, it has been voted for "a" now (automatic EOF handling).

> A question though.  When you get the error back, what do you consider
> the state of the block just written or group of blocks just written?

I haven't even dealt with this... this magic used to be in dump all
the time, it's only that you normally didn't hit EOM unless you
specified a tape length longer than the actual length.

> Some drives will report the EOT and keep recording the current block
> then stop (returning errors for any additional blocks), some drives keep
> recording despite reporting EOT (and you can run the tape off the spindle
> if you aren't careful), and other drives will stop instantly and refuse to
> write further when the EOT is encountered.  Some even back the tape up and
> write two EOF marks over the block if it spans the EOT mark.

I think that's a matter of the driver.  dump will probably trust it
for whatever it reported to be written to the tape, when it returns
from the write(2) call.

-- 
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. ;-)



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?Mutt.19970204094427.j>