Rick Welykochy wrote:
>
> Darrell Burkey wrote:
>
> > There is also a documented exploit (see AUSCERT) with an older version of
> > bind8 with this exact symptom. I think it's a buffer overflow technique.
>
> Probably *IS* a buffer overflow exploit.
>
> Given that a v.large percentage of *nix exploits are indeed
> buffer overflows, perhaps its time to start an open source audit of
> all source code and look for the damn things. Many are quite obvious
> to spot, i.e. strcpy(), strcat(), etc, instead of strncpy().
There already is a Linux audit project with its own mailing list at:
[EMAIL PROTECTED]
> BTW: if the suspect code is re-writtento take advantage of "a little
> bit of C++", and replace all statis and dynamic string buffers with
> instances of the string class, the buffer overflow problem vanishes
> instanatly. Stroustrup has pointed this out countless times, but no-one
> in open source seems to be listening <sigh>.
To me, "A little bit of C++" sounds like "a little pregnant". It also
convienently ignores the huge amounts of C code which uses C++ keywords
as variable names. Its probably far easier to audit and correct the
offending C code.
In addition, most of the machines that get hacked are hacked by script
kiddies using well-known exploits that have usually been identified and
fixed months earlier. In this case writing a new version of a program in
C++ doesn't retrospectively fix machines already running with broken
software.
Erik
--
+-------------------------------------------------+
Erik de Castro Lopo [EMAIL PROTECTED]
+-------------------------------------------------+
"Fifty years of programming language research, and we
end up with C++ ???" --Richard A. O'Keefe
--
SLUG - Sydney Linux Users Group Mailing List - http://www.slug.org.au
To unsubscribe send email to [EMAIL PROTECTED] with
unsubscribe in the text
