Re: deny ktrace without read permissions?

[Nate Williams]

> The PR proposes (and the patch given earlier implements) tighter
> security on the premise that security in the presence of KTRACE
> should be at least as tight as without KTRACE.  It achieves this
> by requiring read permissions on an executable before it can be
> KTRACE'd.

As other have pointed out, if you're good enough to reverse engineer a
program from just it's syscall, you're probably good enough to stick in
a new shared library that allows you to 'reverse engineer' w/out
requiring KTRACE.

Again, security through obscurity is never a good solution, and this is
just that wrapped in different clothes.


Nate


To Unsubscribe: send mail to [EMAIL PROTECTED]
with "unsubscribe freebsd-hackers" in the body of the message