Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
Date:   Tue, 8 Jan 2019 12:21:49 -0600
From:   "Serge E. Hallyn" <serge@hallyn.com>
To:     Tycho Andersen <tycho@tycho.ws>
Cc:     "Serge E. Hallyn" <serge@hallyn.com>,
        Christian Brauner <christian@brauner.io>,
        linux-kernel@vger.kernel.org, linux-api@vger.kernel.org,
        luto@kernel.org, arnd@arndb.de, keescook@chromium.org,
        akpm@linux-foundation.org, jannh@google.com, oleg@redhat.com,
        cyphar@cyphar.com, viro@zeniv.linux.org.uk,
        linux-fsdevel@vger.kernel.org, dancol@google.com,
        timmurray@google.com, fweimer@redhat.com, tglx@linutronix.de,
        x86@kernel.org, ebiederm@xmission.com
Subject: Re: [PATCH v7 2/2] selftests: add tests for pidfd_send_signal()
Message-ID: <20190108182149.GA22762@mail.hallyn.com>
References: <20190102161654.9093-1-christian@brauner.io>
 <20190102161654.9093-2-christian@brauner.io>
 <20190108175306.GF29009@cisco>
 <20190108175415.GA22353@mail.hallyn.com>
 <20190108175843.GA26016@cisco>
 <20190108181742.GA22614@mail.hallyn.com>
 <20190108182023.GA28330@cisco>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20190108182023.GA28330@cisco>
User-Agent: Mutt/1.9.4 (2018-02-28)
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On Tue, Jan 08, 2019 at 11:20:23AM -0700, Tycho Andersen wrote:
> On Tue, Jan 08, 2019 at 12:17:42PM -0600, Serge E. Hallyn wrote:
> > On Tue, Jan 08, 2019 at 10:58:43AM -0700, Tycho Andersen wrote:
> > > On Tue, Jan 08, 2019 at 11:54:15AM -0600, Serge E. Hallyn wrote:
> > > > On Tue, Jan 08, 2019 at 10:53:06AM -0700, Tycho Andersen wrote:
> > > > > On Wed, Jan 02, 2019 at 05:16:54PM +0100, Christian Brauner wrote:
> > > > > > +			/*
> > > > > > +			 * Stop the child so we can inspect whether we have
> > > > > > +			 * recycled pid PID_RECYCLE.
> > > > > > +			 */
> > > > > > +			close(pipe_fds[0]);
> > > > > > +			ret = kill(recycled_pid, SIGSTOP);
> > > > > > +			close(pipe_fds[1]);
> > > > > > +			if (ret) {
> > > > > > +				(void)wait_for_pid(recycled_pid);
> > > > > > +				_exit(PIDFD_ERROR);
> > > > > > +			}
> > > > > 
> > > > > Sorry for being late to the party, but I wonder if this whole thing
> > > > > couldn't be simplified with /proc/sys/kenrel/ns_last_pid?
> > > > 
> > > > no, bc it's not namespaced :)
> > > 
> > > Huh? It looks like it is...
> > > 
> > > static int pid_ns_ctl_handler(struct ctl_table *table, int write,
> > >                 void __user *buffer, size_t *lenp, loff_t *ppos)
> > > {
> > >         struct pid_namespace *pid_ns = task_active_pid_ns(current);
> > >         struct ctl_table tmp = *table;
> > >         int ret, next;
> > > 
> > >         if (write && !ns_capable(pid_ns->user_ns, CAP_SYS_ADMIN))
> > >                 return -EPERM;
> > > 
> > >         ...
> > 
> > Oh - hah, but that's ns_last_pid.  You'd want pid_max.  And that one
> > is not namespaced.
> 
> Perhaps I'm misunderstanding, but isn't the point of all this code to
> get the same pid again? So can't we just fork(), kill(), then set
> ns_last_pid to pid-1, and fork() again to re-use?

Oh yeah that would work :)

I was stuck on the idea of just limiting the range of pids.