From: Oleg Nesterov <oleg@redhat.com>
To: "Eric W. Biederman" <ebiederm@xmission.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>,
Thomas Gleixner <tglx@linutronix.de>,
Peter Zijlstra <peterz@infradead.org>,
Ingo Molnar <mingo@elte.hu>,
Christoph Hellwig <hch@infradead.org>,
Nick Piggin <npiggin@suse.de>,
Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
"Paul E. McKenney" <paulmck@linux.vnet.ibm.com>
Subject: Re: [rfc] "fair" rw spinlocks
Date: Wed, 9 Dec 2009 16:37:09 +0100 [thread overview]
Message-ID: <20091209153709.GA13192@redhat.com> (raw)
In-Reply-To: <m1aaxu305x.fsf@fess.ebiederm.org>
On 12/07, Eric W. Biederman wrote:
>
> Oleg Nesterov <oleg@redhat.com> writes:
>
> > On 12/05, Eric W. Biederman wrote:
> >>
> >> Atomically sending signal to every member of a process group, is the
> >> big fly in the ointment I am aware of. Last time I looked I could
> >> not see how to convert it rcu.
> >
> > I am not sure, but iirc we can do this lockless (under rcu_lock).
> > We need to modify pid_link to use list_entry and attach_pid() should
> > add the new task to the end. Of course we need more changes, but
> > (again iirc) this is not too hard.
>
> The problem is that even adding to the end of the list, we could run
> into a deleted entry and not see the new end of the list.
>
> Suppose when we start iterating the list we have:
>
> A -> B -> C -> D
>
> Then someone deletes some of the entries while we are iterating the list.
>
> A ->
> B' -> C' -> D'
>
> We will continue on traversing through the deleted entries.
>
> Then someone adds a new entry to the end of the list.
>
> A-> N
>
> Since we are at B', C' or D' we will never see the new entry on the
> end of the list.
Yes, but who can add the new entry?
Let's forget about setpgrp/etc for the moment, I think we have "races"
with or without tasklist. Say, setpgrp() can add the new process to the
already "killed" pgrp.
Then, I think the only important case is SIGKILL/SIGSTOP (or other
signals which can't be blockes/ignored). We must kill/stop the entire
pgrp, we must not race with fork() and miss a child.
In this case I _think_ rcu_read_lock() is enough,
rcu_read_lock()
list_for_each_entry_rcu(task, pid->tasks[PIDTYPE_PGID)
group_send_sig_info(sig, task);
rcu_read_unlock();
except group_send_sig_info() can race with mt-exec, but this is simple
to fix.
If we send a signal (not necessary SIGKILL) to a process P, we must see
all childs which were forked by P, both send_signal() and copy_process()
take the same ->siglock, we must see the result of list_add_tail_rcu().
And, after we sent SIGKILL/SIGSTOP, it can't fork the new child.
If list_for_each_entry() does not see the exited process P, this means
we see the result of list_del_rcu(). But this also means we must the
the result of the previous list_add_rcu().
IOW, fork+exit means list_add_rcu() + wmb() + list_del_rcu(), if we
don't see the new entry on list, we must see the new one, right?
(I am ignoring the case when list_for_each_entry_rcu() sees a process
P but lock_task_sighand(P) fails, I think this is the same as if we
we missed P)
Now suppose a signal is blocked/ignored or has a handler. In this case
we can miss a child, but I think this is OK, we can pretend the new
child was forked after kill_pgrp() completes. Say, this child C was
forked by some process P. We can miss C only if it was forked after
we already sent the signal to P.
However. I do not pretend the reasoning above is "complete", and
perhaps I missed something else.
> Additionally we have the other possibility that if a child is forking
> we send the signal to the parent after the child forks away but before
> the child joins whichever list we are walking, and we complete our
> traversal without seeing the child.
Not sure I understand... But afaics this case is covered above.
->siglock should serialize this, copy_process() does attach_pid()
under this lock.
Oleg.
next prev parent reply other threads:[~2009-12-09 15:43 UTC|newest]
Thread overview: 58+ messages / expand[flat|nested] mbox.gz Atom feed top
2009-11-23 14:54 Nick Piggin
2009-11-24 20:19 ` David Miller
2009-11-25 6:52 ` Nick Piggin
2009-11-25 8:49 ` Andi Kleen
2009-11-25 8:56 ` Nick Piggin
2009-11-24 20:47 ` Andi Kleen
2009-11-25 6:54 ` Nick Piggin
2009-11-25 8:48 ` Andi Kleen
2009-11-25 13:09 ` Arnd Bergmann
2009-11-28 2:07 ` Paul E. McKenney
2009-11-28 11:15 ` Andi Kleen
2009-11-28 15:20 ` Paul E. McKenney
2009-11-28 17:30 ` Linus Torvalds
2009-11-29 18:51 ` Paul E. McKenney
2009-11-30 7:57 ` Nick Piggin
2009-11-30 7:55 ` Nick Piggin
2009-11-30 15:22 ` Linus Torvalds
2009-11-30 15:40 ` Nick Piggin
2009-11-30 16:07 ` Linus Torvalds
2009-11-30 16:17 ` Nick Piggin
2009-11-30 16:39 ` Paul E. McKenney
2009-11-30 17:05 ` Linus Torvalds
2009-11-30 17:13 ` Nick Piggin
2009-11-30 17:18 ` Linus Torvalds
2009-12-01 17:03 ` Arnd Bergmann
2009-12-01 17:15 ` Linus Torvalds
2009-11-30 18:29 ` Paul E. McKenney
2009-11-30 16:20 ` Paul E. McKenney
2009-11-30 10:00 ` Christoph Hellwig
2009-11-30 15:52 ` Linus Torvalds
2009-11-30 17:46 ` Ingo Molnar
2009-11-30 21:12 ` Thomas Gleixner
2009-11-30 21:27 ` Peter Zijlstra
2009-11-30 22:02 ` Thomas Gleixner
2009-11-30 22:11 ` Linus Torvalds
2009-11-30 22:37 ` Thomas Gleixner
2009-11-30 22:49 ` Linus Torvalds
2009-12-01 17:37 ` [PATCH] audit: Call tty_audit_push_task() outside preempt disabled region Thomas Gleixner
2009-12-01 18:22 ` Oleg Nesterov
2009-12-01 19:53 ` Thomas Gleixner
2009-12-06 3:12 ` [rfc] "fair" rw spinlocks Eric W. Biederman
2009-12-07 18:18 ` Paul E. McKenney
2009-12-07 22:24 ` Eric W. Biederman
2009-12-07 22:35 ` Andi Kleen
2009-12-07 23:19 ` Eric W. Biederman
2009-12-08 1:39 ` Paul E. McKenney
2009-12-08 2:11 ` Eric W. Biederman
2009-12-08 2:37 ` Paul E. McKenney
2009-12-07 18:32 ` Oleg Nesterov
2009-12-07 20:38 ` Peter Zijlstra
2009-12-09 15:55 ` Oleg Nesterov
2009-12-07 22:10 ` Eric W. Biederman
2009-12-09 15:37 ` Oleg Nesterov [this message]
2009-12-10 3:36 ` Eric W. Biederman
2009-12-10 6:22 ` Paul E. McKenney
2009-12-10 10:31 ` Eric W. Biederman
2009-12-10 16:41 ` Paul E. McKenney
2009-12-01 19:01 ` Mathieu Desnoyers
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20091209153709.GA13192@redhat.com \
--to=oleg@redhat.com \
--cc=ebiederm@xmission.com \
--cc=hch@infradead.org \
--cc=linux-kernel@vger.kernel.org \
--cc=mingo@elte.hu \
--cc=npiggin@suse.de \
--cc=paulmck@linux.vnet.ibm.com \
--cc=peterz@infradead.org \
--cc=tglx@linutronix.de \
--cc=torvalds@linux-foundation.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox
Powered by JetHome