Re: [PATCH v4 1/2] pid: Replace pid bitmap implementation with IDR API

[Rik van Riel <riel@surriel.com>]

[-- Attachment #1: Type: text/plain, Size: 1333 bytes --]

On Tue, 2017-10-10 at 13:50 +0200, Oleg Nesterov wrote:
> On 10/09, Andrew Morton wrote:
> > 
> > > @@ -240,17 +230,11 @@ void zap_pid_ns_processes(struct
> > > pid_namespace *pid_ns)
> > >  	 *
> > >  	 */
> > >  	read_lock(&tasklist_lock);
> > > -	nr = next_pidmap(pid_ns, 1);
> > > -	while (nr > 0) {
> > > -		rcu_read_lock();
> > > -
> > > -		task = pid_task(find_vpid(nr), PIDTYPE_PID);
> > > +	nr = 2;
> > > +	idr_for_each_entry_continue(&pid_ns->idr, pid, nr) {
> > > +		task = pid_task(pid, PIDTYPE_PID);
> > >  		if (task && !__fatal_signal_pending(task))
> > >  			send_sig_info(SIGKILL, SEND_SIG_FORCED,
> > > task);
> > > -
> > > -		rcu_read_unlock();
> > > -
> > > -		nr = next_pidmap(pid_ns, nr);
> > >  	}
> > >  	read_unlock(&tasklist_lock);
> > 
> > Especially here.  I don't think pidmap_lock is held.  Is that IDR
> > iteration safe?
> 
> Yes, this doesn't look right, we need rcu_read_lock() or pidmap_lock.
> 
> And, we also need rcu_read_lock() for another reason, to protect
> "struct pid".

I think rcu_read_lock alone should do the trick, for both.

The IDR code specifically says that lookups are safe under just
the rcu_read_lock, and that only insertions and deletions need
a separate lock for synchronization.

Good catch.

-- 
All Rights Reversed.

[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 473 bytes --]