linux-kernel - Re: [PATCH] locks: change POSIX lock ownership on execve when files

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-ID: <1521314890.4064.12.camel@kernel.org>
Date:   Sat, 17 Mar 2018 15:28:10 -0400
From:   Jeff Layton <jlayton@...nel.org>
To:     Al Viro <viro@...IV.linux.org.uk>
Cc:     linux-fsdevel@...r.kernel.org, linux-kernel@...r.kernel.org,
        "J. Bruce Fields" <bfields@...ldses.org>,
        Thomas Gleixner <tglx@...utronix.de>,
        "Daniel P ." Berrangé <berrange@...hat.com>,
        Kate Stewart <kstewart@...uxfoundation.org>,
        Dan Williams <dan.j.williams@...el.com>,
        Philippe Ombredanne <pombredanne@...b.com>,
        Greg Kroah-Hartman <gregkh@...uxfoundation.org>
Subject: Re: [PATCH] locks: change POSIX lock ownership on execve when
 files_struct is displaced

On Sat, 2018-03-17 at 15:52 +0000, Al Viro wrote:
> On Sat, Mar 17, 2018 at 11:43:28AM -0400, Jeff Layton wrote:
> > On Sat, 2018-03-17 at 15:05 +0000, Al Viro wrote:
> > > On Sat, Mar 17, 2018 at 10:25:20AM -0400, Jeff Layton wrote:
> > > > From: Jeff Layton <jlayton@...hat.com>
> > > > 
> > > > POSIX mandates that open fds and their associated file locks should be
> > > > preserved across an execve. This works, unless the process is
> > > > multithreaded at the time that execve is called.
> > > > 
> > > > In that case, we'll end up unsharing the files_struct but the locks will
> > > > still have their fl_owner set to the address of the old one. Eventually,
> > > > when the other threads die and the last reference to the old
> > > > files_struct is put, any POSIX locks get torn down since it looks like
> > > > a close occurred on them.
> > > > 
> > > > The result is that all of your open files will be intact with none of
> > > > the locks you held before execve. The simple answer to this is "use OFD
> > > > locks", but this is a nasty surprise and it violates the spec.
> > > > 
> > > > On a successful execve, change ownership of any POSIX file_locks
> > > > associated with the old files_struct to the new one, if we ended up
> > > > swapping it out.
> > > 
> > > TBH, I don't like the way you implement that.  Why not simply use
> > > iterate_fd()?
> > 
> > Ahh, I wasn't aware of it. I copied the loop in change_lock_owners from
> > close_files. I'll have a look at iterate_fd().
> 
> BTW, if iterate_fd() turns out to be slower, it might make sense to have it
> look at the bitmap to skip unpopulated parts of descriptor table faster -
> other users might also benefit from that.

Thanks, I'll keep that in mind.

Full disclosure: I haven't done any performance testing with this. My
assumption is that threaded programs that execve without forking first
are rather rare. I don't know of a great way to confirm that though.

I made a small change to the v2 patch as well to use
struct files_struct * instead of fl_owner_t here. That gives us more
type safety and should prevent any problems if Bruce's patch to remove
fl_owner_t gets merged.

Thanks,
-- 
Jeff Layton <jlayton@...nel.org>