Mailing-List: contact cygwin-help AT cygwin DOT com; run by ezmlm
List-Subscribe: <mailto:cygwin-subscribe AT cygwin DOT com>
List-Archive: <http://sources.redhat.com/ml/cygwin/>
List-Post: <mailto:cygwin AT cygwin DOT com>
List-Help: <mailto:cygwin-help AT cygwin DOT com>, <http://sources.redhat.com/ml/#faqs>
Sender: cygwin-owner AT cygwin DOT com
Mail-Followup-To: cygwin AT cygwin DOT com
Delivered-To: mailing list cygwin AT cygwin DOT com
Date: Fri, 5 Mar 2004 11:19:47 +0100
From: Corinna Vinschen <corinna-cygwin AT cygwin DOT com>
To: cygwin AT cygwin DOT com
Subject: Re: read(): varblk tape records...(& Fix for : read())
Message-ID: <20040305101947.GB21805@cygbert.vinschen.de>
Reply-To: cygwin AT cygwin DOT com
Mail-Followup-To: cygwin AT cygwin DOT com
References: <Sea2-F42FZEoKz4r02V0005f529 AT hotmail DOT com>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <Sea2-F42FZEoKz4r02V0005f529@hotmail.com>
User-Agent: Mutt/1.4.2i

On Mar  5 05:14, Jason Winter wrote:
> Hi Corinna,
> 
> It turns out that your new fix (for read();) might (I'm not sure until the 
> nightly builds are working again) prevent the bug from happening with 
> var-blk records - but I think the 'bug' will still cause problems with 
> fixed-block records and maybe other filetypes.  The issue I'm having is: 
> raw_read() is setting devbufstart & devbufend, and raw_write() checks 
> devbufend for buffering (which isn't required with var-block records or 
> properly written fixed-block records in any case.)

What means "properly"?  It's not necessary with var-blocks, but it is
necessary with fixed blocks.  If somebody writes 1200 bytes to a tape
set to 512 byte blocksize, raw_write writes 2 blocks (1024 bytes) to
the tape and buffers the remaining 176 bytes for the next write.  How
should that work otherwise?

> My own feeling on this is, when reading tape-block data (fixed or variable 
> blocks) and the blocks are not being buffered in calls (ie. my test-harness 
> uses correctly sized buffers) then raw_read() shouldn't leave devbufstart 
> or devbufend non-zero.

I don't quite understand what you mean.  The first thing in raw_read()
is to call writebuf() which checks if devbuf is used as a write buffer
and if so, tries to write the data in the buffer onto the medium.  This
also sets devbufstart and devbufend to zero.  Am I missing something?

Unfortunately I can't run your testcase.  My DDS tape drive seems to be
broken.  It can read, but it behaves weird when trying to write :-(((


Corinna

-- 
Corinna Vinschen                  Please, send mails regarding Cygwin to
Cygwin Developer                                mailto:cygwin AT cygwin DOT com
Red Hat, Inc.

--
Unsubscribe info:      http://cygwin.com/ml/#unsubscribe-simple
Problem reports:       http://cygwin.com/problems.html
Documentation:         http://cygwin.com/docs.html
FAQ:                   http://cygwin.com/faq/