Re: [PATCH] exec: do not leave bprm->interp on stack

[Kees Cook]

On Sat, Oct 27, 2012 at 3:47 AM, P J P <ppandit@xxxxxxxxxx> wrote:
> +-- On Fri, 26 Oct 2012, Al Viro wrote --+
> | > not. Module alias could dodge this though, I guess.
> | "Could"?  Can you show a single module that would have name matching
> | binfmt-[0-9a-f]*?  In other words, are they ever loaded _not_ via an
> | alias?
>
>   I understand. I was wondering if alias information is accessible in the
> kernel via any routine, alike find_module().
>
> Just to get perspective about how many times request_module() would be called
> with the latest patch, in general installations(or distributions), how
> prevalent(in use) are binfmt-xxxx loadable modules?

Al showed a list of them earlier in the thread. I don't have any on
the various distros I checked.

The problem I see here is that we only want to do module loading in
the "no match" case. But that means that either we need to restart
with the original bprm, or we need to keep bprm changes off the stack.
Leading with a module load is going to wreck performance.

-Kees

-- 
Kees Cook
Chrome OS Security
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/