Messages in this thread | | | From | (Alan Cox) | Subject | Re: [PATCH] [SECURITY] suid procs exec'd with bad 0,1,2 fds | Date | Wed, 5 Aug 1998 08:22:14 +0100 (BST) |
| |
> It sure is, and the choice we're making is that people should fix the > applications instead of putting a hack into the mainstream kernel.
They are working on it. Given another 30 to 40 years it might happen as people learn how to handle such stuff. And odds on the final resulting code will _not_ be written in C by then
In the mean time "hacks" like that are saving ISP's millions of dollars.
Thats production reality versus technical perfection
> Because once the class of exploits is fixed in the applications, the > kernel hack no longer is relevant. And given this situation, the
Wrong. Nobody has or is likely to achieve a set of formal proofs for correctness of every random application. Thats also another reason why C is unlikely to remain a language of choice for "secure" applications.
And here secure means "any application that interacts directly or indirectly with untrusted data" - ie most of them.
> The golden rule is, if it can be fixed in userspace, make doing it > there the preferred solution if it makes sense. And here it makes > sense.
Definitely. Hence Im most interested to see if the "sane hardware" case of nonexec stack can be done by ld.so
Alan
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.altern.org/andrebalsa/doc/lkml-faq.html
| |