Sender: Tejun Heo <htejun@gmail.com>
Date:   Wed, 8 Dec 2021 09:06:06 -1000
From:   Tejun Heo <tj@kernel.org>
To:     "Eric W. Biederman" <ebiederm@xmission.com>
Cc:     Michal =?iso-8859-1?Q?Koutn=FD?= <mkoutny@suse.com>,
        linux-kernel@vger.kernel.org, Jens Axboe <axboe@kernel.dk>,
        Kees Cook <keescook@chromium.org>,
        Oleg Nesterov <oleg@redhat.com>,
        Peter Zijlstra <peterz@infradead.org>,
        Thomas Gleixner <tglx@linutronix.de>,
        Jim Newsome <jnewsome@torproject.org>,
        Alexey Gladkov <legion@kernel.org>, security@kernel.org,
        Andy Lutomirski <luto@amacapital.net>,
        Jann Horn <jannh@google.com>
Subject: Re: [PATCH] exit: Retain nsproxy for exit_task_work() work entries
Message-ID: <YbECHjMLPEHO0vqA@slm.duckdns.org>
References: <20211208180501.11969-1-mkoutny@suse.com>
 <87sfv3540t.fsf@email.froward.int.ebiederm.org>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <87sfv3540t.fsf@email.froward.int.ebiederm.org>
Precedence: bulk

Hello, Eric.

On Wed, Dec 08, 2021 at 12:45:54PM -0600, Eric W. Biederman wrote:
> Permission checks on files need to happen at open time, not at
> write time.  It is all too easy to confuse something that writes
> to stdout to write to a file of your choosing to make permission
> checking at write time a good idea.

Whether a given operation is allowed or not is dependent on the content of
the write. I don't see how it can be moved to open time. I'm not a fan of
the fs based interface but they're not real files and it has been like this
from the beginning.

> That said I can't quite tell if the test should be moved into
> cgroup_file_open or if there is a permission entry that would work.

It can't.

> I may be wrong but at first glance this looks like the cgroup code is
> going to need significant surgery to get the permission checks happening
> at open time where they belong.

There's no way to change that without changing the interface drastically.

> Please don't apply this patch.
> 
> exit_task_work running after exit_task_namespaces is the messenger
> that just told us about something ugly.

I don't see good alternatives here. You got any?

Thanks.

-- 
tejun