Re: [PATCH v1 1/1] waitid: Add support for waiting for the current process group

From: Oleg Nesterov <oleg@redhat.com>
To: Christian Brauner <christian.brauner@ubuntu.com>
Cc: linux-kernel@vger.kernel.org, libc-alpha@sourceware.org,
	alistair23@gmail.com, ebiederm@xmission.com, arnd@arndb.de,
	dalias@libc.org, torvalds@linux-foundation.org,
	adhemerval.zanella@linaro.org, fweimer@redhat.com,
	palmer@sifive.com, macro@wdc.com, zongbox@gmail.com,
	akpm@linux-foundation.org, viro@zeniv.linux.org.uk,
	hpa@zytor.com
Subject: Re: [PATCH v1 1/1] waitid: Add support for waiting for the current process group
Date: Wed, 14 Aug 2019 14:50:12 +0200	[thread overview]
Message-ID: <20190814125012.GB11595@redhat.com> (raw)
In-Reply-To: <20190814124551.hnt363g3blhuf2pv@wittgenstein>

On 08/14, Christian Brauner wrote:
>
> On Wed, Aug 14, 2019 at 02:29:10PM +0200, Oleg Nesterov wrote:
> > On 08/14, christian.brauner@ubuntu.com wrote:
> > >
> > >  	case P_PGID:
> > >  		type = PIDTYPE_PGID;
> > > -		if (upid <= 0)
> > > +		if (upid < 0)
> > >  			return -EINVAL;
> > > +
> > > +		if (upid == 0)
> > > +			pid = get_pid(task_pgrp(current));
> >
> > this needs rcu lock or tasklist_lock, this can race with another thread
> > doing sys_setpgid/setsid (see change_pid(PIDTYPE_PGID)).
>
> Oh, I naively assumed task_pgrp() would take an rcu lock...

but it would not help ;)

> though I think we should be fine with just rcu_read_lock().

Yes,

Oleg.