X-Authentication-Warning: delorie.com: mailnull set sender to djgpp-workers-bounces using -f
From: sandmann AT clio DOT rice DOT edu (Charles Sandmann)
Message-Id: <10201100020.AA19380@clio.rice.edu>
Subject: Re: Proposed lstat.c patch for Win2000/XP device bit fix
To: djgpp-workers AT delorie DOT com
Date: Wed, 9 Jan 2002 18:20:54 -0600 (CST)
Cc: ams AT ludd DOT luth DOT se
In-Reply-To: <200201092157.WAA24942@father.ludd.luth.se> from "Martin Str|mberg" at Jan 09, 2002 10:57:01 PM
X-Mailer: ELM [version 2.5 PL2]
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Reply-To: djgpp-workers AT delorie DOT com
Errors-To: nobody AT delorie DOT com
X-Mailing-List: djgpp-workers AT delorie DOT com
X-Unsubscribes-To: listserv AT delorie DOT com
Precedence: bulk

> According to Charles Sandmann:
> > strcmp is already used in the same procedure - so nothing new was added.
> > see lstat.c source for details.
> 
> Indeed. Here's a patch for that (manually cut-and-pasted, so will
> probably not apply cleanly and there some offset as well due to my GCC
> warning corrections):

It's certainly more efficient (but shouldn't it just be '2' for the 
second compare?) - if everyone agrees I'll make it a macro and use it
twice in the code.  

> -  if ( ! strcmp(ff_blk.lfn_magic,"LFN32") )
> +  if ( *(long *)(&(ff_blk.lfn_magic[0])) == 'L'+('F'<<8)+('N'<<16)+('3'<<24)
> +    && *(short *)(&(ff_blk.lfn_magic[4])) == 0+('2'<<8) )
>      {
> 
> What parameters should I give to cvs to get the v2.03 update branch?

Note: 2.03 uses strcmp an additional place in the FNCASE code - so you
can't easily eliminate strcmp there.

1) Create a completely separate working tree:
    cvs checkout -r v2_03_1 djgpp
2) Please don't commit anything unless we're sure we are building a
   new distribution (we want to keep the two consistent).  If we wait
   long enough we would need to branch v2_03_1 into v2_03_2 ;-P