X-Spam-Check-By: sourceware.org
Message-Id: <6.1.2.0.2.20051123061647.01bab8d0@cic-mail.lanl.gov>
X-Sender: u085598 AT cic-mail DOT lanl DOT gov
Date: Wed, 23 Nov 2005 06:58:12 -0700
To: cygwin AT cygwin DOT com
From: Richard Kandarian <richard DOT kandarian AT lanl DOT gov>
Subject: Re: rcs 5.7 truncated one of my files
In-Reply-To: <438469F7.4010608@byu.net>
References: <6 DOT 1 DOT 2 DOT 0 DOT 2 DOT 20051122153707 DOT 01db9c28 AT cic-mail DOT lanl DOT gov>  <4383A666 DOT 3020708 AT ukf DOT net>  <438469F7 DOT 4010608 AT byu DOT net>
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"; format=flowed
Mailing-List: contact cygwin-help AT cygwin DOT com; run by ezmlm
Precedence: bulk
List-Unsubscribe: <mailto:cygwin-unsubscribe-archive-cygwin=delorie DOT com AT cygwin DOT com>
List-Subscribe: <mailto:cygwin-subscribe AT cygwin DOT com>
List-Archive: <http://sourceware.org/ml/cygwin/>
List-Post: <mailto:cygwin AT cygwin DOT com>
List-Help: <mailto:cygwin-help AT cygwin DOT com>, <http://sourceware.org/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

At 06:09 AM 11/23/2005, Eric Blake wrote:
>-----BEGIN PGP SIGNED MESSAGE-----
>Hash: SHA1
>
>According to Max Bowsher on 11/22/2005 4:14 PM:
> >
> > RCS has not been updated in years.
>
>Could it be that RCS is still using 32-bit offsets because it has not been
>recompiled against newer cygwin headers to turn on 64-bit offsets?  You
>can figure this out if the truncation is occuring on ,v files > 2 gig in size.
>
> > Can you find a way to reproduce the truncation?
> > If you can, then it ought to be possible to find the problem.
> > If you can't, this is likely going to remain an unsolved mystery.
>
>That is entirely true - without more details, especially a simple
>testcase, very few people are motivated enough to try to help you resolve
>the issue.
>
>- --
>Life is short - so eat dessert first!
>
>Eric Blake             ebb9 AT byu DOT net
>-----BEGIN PGP SIGNATURE-----
>Version: GnuPG v1.4.1 (Cygwin)
>Comment: Public key at home.comcast.net/~ericblake/eblake.gpg
>Comment: Using GnuPG with Thunderbird - http://enigmail.mozdev.org
>
>iD8DBQFDhGn284KuGfSFAYARAmu2AJ0Yfotdg2NbFfTQ+juPrr5lOmIcsQCeOEVO
>k2buKt2w1f3VXxCgA03NB00=
>=/0CJ
>-----END PGP SIGNATURE-----

Who wants to have anything to do with a > 2 gig Perl script? Note the other 
clue in the preceding question. I notice that rcs has to double the @ 
characters it finds in Perl scripts.

The new file
$ ll mkbsub.ph
-rw-r--r--  1 085598 None 34319 Nov 22 16:44 mkbsub.ph

The initial release rcs file:
$ ll RCS.bak/mkbsub.ph,v
-r--r--r--  1 085598 None 32993 Nov 22 15:12 RCS.bak/mkbsub.ph,v

The rcs file with the new file checked in
$ ll RCS/mkbsub.ph,v
-r--r--r--  1 085598 None 13515 Nov 22 16:27 RCS/mkbsub.ph,v

The new file checked out
$ ll mkbsub.ph
-rw-r--r--  1 085598 None 1024 Nov 23 06:23 mkbsub.ph

The part of RCS/mkbsub.ph,v that matches the truncation *exactly* but for 
the nine doubled @ characters. That size includes a final \n which is not 
in the truncated version.
$ ll junk
-rw-rw-rw-  1 085598 None 1034 Nov 23 06:33 junk

I think the 1024 size of that penultimate file, mkbsub.ph, is very 
interesting. I hadn't noticed that before.

The truncated file matches the beginning of the second version but does not 
match the beginning of the initial revision:

$ diff -u0 junk mkbsub.ph
--- junk        2005-11-23 06:42:11.001443900 -0700
+++ mkbsub.ph   2005-11-23 06:23:29.741722500 -0700
@@ -3,7 +3,3 @@
....

And junk, the equivalent portion of the initial revision, is 53 bytes 
bigger with one @ character more.

I'll see if I can come up with any theories which might help me produce an 
example I can share.

Thanks



Richard Kandarian
http://www.lanl.gov/cgi-bin/fonelink.pl/085598
Any opinions stated in this message are not expressed on behalf of any 
individual or entity other than me unless explicitly noted otherwise. My 
node in the Web: http://www.kandarian.com



--
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/