Re: [PATCH 1/2] cgroup: remove tasklist_lock from cgroup_attach_proc

[Mandeep Singh Baines <msb-F7+t8E8rja9g9hUCZPvPmw@public.gmane.org>]

Tejun Heo (tj-DgEjT+Ai2ygdnm+yROfE0A@public.gmane.org) wrote:
> Hello,
> 
> On Thu, Dec 22, 2011 at 02:26:55PM -0800, Mandeep Singh Baines wrote:
> > @@ -2283,14 +2261,26 @@ static int attach_task_by_pid(struct cgroup *cgrp, u64 pid, bool threadgroup)
> >  
> >  	threadgroup_lock(tsk);
> >  
> > -	if (threadgroup)
> > +	if (threadgroup) {
> > +		if (!thread_group_leader(tsk)) {
> > +			/*
> > +			 * a race with de_thread from another thread's exec()
> > +			 * may strip us of our leadership, if this happens,
> > +			 * there is no choice but to throw this task away and
> > +			 * try again (from cgroup_procs_write); this is
> > +			 * "double-double-toil-and-trouble-check locking".
> > +			 */
> > +			ret = -EAGAIN;
> > +			goto out_unlock_threadgroup;
> > +		}
> >  		ret = cgroup_attach_proc(cgrp, tsk);
> 
> Ummm... can't we just do tsk = tsk->leader after locking threadgroup?

You're right. Nice!

New series coming up.

> Why do we need to retry from the beginning?
> 
> Thanks.
> 
> -- 
> tejun