Date: Tue, 10 Jul 2001 15:10:18 +0300 (IDT)
From: Eli Zaretskii <eliz AT is DOT elta DOT co DOT il>
X-Sender: eliz AT is
To: Martin Str|mberg <ams AT ludd DOT luth DOT se>
cc: djgpp-workers AT delorie DOT com
Subject: Re: glob buffer overflow fix
In-Reply-To: <200107101137.NAA01123@father.ludd.luth.se>
Message-ID: <Pine.SUN.3.91.1010710150040.20713A-100000@is>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
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


On Tue, 10 Jul 2001, Martin Str|mberg wrote:

> According to Mark E.:
> > This patch adds buffer overflow checks to the output buffer. The idea is that 
> > a filename in the output buffer can't be valid and won't match if its size is 
> > greater or equal to 2000 bytes (or whatever impossible length).
> 
> Any reason for not using MAXPATH (or PATH_MAX)?

You cannot limit the buffer at PATH_MAX because the string crunched by 
glob can include wildcard characters.  Consider this example:

  [abcdefghIJKLMNOP][xyzXYZ09876]*/[!-RSTUVrstuv_]*/*.[cShyl]

etc., etc.  So if you want a constant that is not entirely arbitrary, 
you need something like FILENAME_MAX * ((128 - 32 - 5)*1.5 + 2 + 1)
(I probably didn't get that count entirely accurate), which is 10 times 
more than 2000.