From mboxrd@z Thu Jan 1 00:00:00 1970 From: Oleg Nesterov Subject: Re: [PATCH 4/9] pid: Generalize task_active_pid_ns Date: Thu, 13 Dec 2007 20:07:27 +0300 Message-ID: <20071213170727.GA344@tv-sign.ru> References: <20071213160128.GA219@tv-sign.ru> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Return-path: Content-Disposition: inline In-Reply-To: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: containers-bounces-cunTk1MwBs9QetFLy7KEm3xJsTq8ys+cHZ5vskTnxNA@public.gmane.org Errors-To: containers-bounces-cunTk1MwBs9QetFLy7KEm3xJsTq8ys+cHZ5vskTnxNA@public.gmane.org To: "Eric W. Biederman" Cc: Linux Containers , Andrew Morton , Pavel Emelyanov List-Id: containers.vger.kernel.org On 12/13, Eric W. Biederman wrote: > > Oleg Nesterov writes: > > > On 12/12, Eric W. Biederman wrote: > >> > >> Currently task_active_pid_ns is not safe to call after a > >> task becomes a zombie and exit_task_namespaces is called, > >> as nsproxy becomes NULL. By reading the pid namespace from > >> the pid of the task we can trivially solve this problem > > > > Confused. If the task becomes a zombie, we can't assume it has a valid > > ->pids[].pid. The parent can release us as soon as exit_notify() drops > > tasklist. > > > What this allows me to do (as seen later in the patchset) is to send > to call pid_nr_ns and deliver a signal to a task group without caring > if the element of the task group I am talking to is a zombie or not. Yes I see, thanks Eric. I was confused by changelog, and I missed the subsequent changes which need task_active_pid_ns() of the reciever. Oleg.