Messages in this thread | | | Date | Sun, 29 May 2011 19:23:56 +0100 | From | Al Viro <> | Subject | Re: [PATCH 3/5] v2 seccomp_filters: Enable ftrace-based system call filtering |
| |
On Sun, May 29, 2011 at 10:02:06AM -0700, Linus Torvalds wrote:
> And yes, it is absolutely acceptable. The only questions in my mind are > > - why haven't we done this long ago? > > - do we have the flag space? > > - should we do a O_NOMNT_PATH flag to do the same for mount-points? > > Some people worry about being confused by bind mounts etc. > > - do we think ".." is worthy of a flag too? > > or is that a "user space can damn well check that itself, even if > it would be absolutely trivial to check in the kernel too"? > > Whatever. I think the NOFOLLOW_PATH one is pretty much a no-brainer. > It's not like symlink worries are unusual.
It's not *quite* a no-brainer. Guys, please hold that one off for a while; we have more massage to do in the area and I *really* want to get atomic open work finished (== intents gone, revalidation vs mountpoints sanitized, etc.) before anything else is done to fs/namie.c. OK?
And as for .. - userland can bloody well check that on its own if it cares. Let's keep it simple, please - we already have things far too complicated in there for my taste.
| |