Re: Q: check_unsafe_exec() races (Was: [PATCH 2/4] fix setuidsometimes doesn't)

[Al Viro]

On Mon, Mar 30, 2009 at 03:32:35PM +0100, Hugh Dickins wrote:
> > Anyway, completely untested patchset is in
> > git.kernel.org/pub/scm/linux/kernel/git/viro/vfs-2.6.git/ execve-mess
> > (the last 9 changesets of it).
> > 
> > WARNING: that's *NOT* for merge at the moment; this is not a pull request.

> I'm looking now.  Your description of what you intended sounded good.
> 
> First impression is that there's lots of good cleanup in there, but
> it's all too mixed up and misordered at present - though we have friends
> who insist upon doing the cleanup first (and I usually like to work that
> way too), it's going to be tiresome when backporting to 2.6.29.stable
> (luckily not needed earlier, unless you've uncovered more on the way).

Yes, and worse than that; this was basically a code dump - the state at
the moment when I'd been going down.

> I think your new chroot_fs_refs() should have a path_get(new_root)
> just before each path_put(old_root)?

Not just before; at the time it replaces ->root or ->pwd.  And yes,
it's one of the brainos there.

Anyway, I've got that stuff to something reasonably-looking.  See the
same branch (rebased), just 5 changesets now.  Have fun.  Cleanups are
not part of that set.
--
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/