[patch] add modes to open call of ismounted

[Rob Landley]

On Friday 28 July 2006 11:04 pm, Rich Felker wrote:
> BTW, I accept in general your principle of not fixing stuff unless it
> actually has practical problems, as opposed to just theoretical
> correctness issues. BUT... security is _not_ a place you can apply
> that principle. A potential vuln _is_ exploitable unless you can
> _prove_ otherwise. To treat it any other way is utterly irresponsible.

I am fixing it, by removing the current entire contents of the e2fsprogs 
directory.  (I finally read up enough on e2fsck that I think I can clean up 
that one too.  There's an old paper on the subject that gives the design 
pretty clearly.)  The only question was whether it was worth patching -stable 
in the meantime.

And yes, "the suid bit" is a valid objection.  Not something an attacker can 
trigger except by luck, but still valid.

Rob
-- 
Never bet against the cheap plastic solution.