From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-7.0 required=3.0 tests=HEADER_FROM_DIFFERENT_DOMAINS, INCLUDES_PATCH,MAILING_LIST_MULTI,SIGNED_OFF_BY,SPF_HELO_NONE,SPF_PASS autolearn=ham autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id 4C726C433E0 for ; Tue, 23 Jun 2020 21:58:36 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by mail.kernel.org (Postfix) with ESMTP id 2F13E208A9 for ; Tue, 23 Jun 2020 21:58:36 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S2388635AbgFWV6e (ORCPT ); Tue, 23 Jun 2020 17:58:34 -0400 Received: from out02.mta.xmission.com ([166.70.13.232]:57574 "EHLO out02.mta.xmission.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S2387455AbgFWV6d (ORCPT ); Tue, 23 Jun 2020 17:58:33 -0400 Received: from in01.mta.xmission.com ([166.70.13.51]) by out02.mta.xmission.com with esmtps (TLS1.2:ECDHE_RSA_AES_128_GCM_SHA256:128) (Exim 4.90_1) (envelope-from ) id 1jnqw3-0004Nx-RL; Tue, 23 Jun 2020 15:58:31 -0600 Received: from ip68-227-160-95.om.om.cox.net ([68.227.160.95] helo=x220.xmission.com) by in01.mta.xmission.com with esmtpsa (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.87) (envelope-from ) id 1jnqw3-0003FG-0o; Tue, 23 Jun 2020 15:58:31 -0600 From: ebiederm@xmission.com (Eric W. Biederman) To: Cc: , Linus Torvalds , Oleg Nesterov , Jann Horn , Kees Cook , Bernd Edlinger References: <87pn9u6h8c.fsf@x220.int.ebiederm.org> <87r1u5laac.fsf@x220.int.ebiederm.org> Date: Tue, 23 Jun 2020 16:54:07 -0500 In-Reply-To: <87r1u5laac.fsf@x220.int.ebiederm.org> (Eric W. Biederman's message of "Tue, 23 Jun 2020 16:52:43 -0500") Message-ID: <87ftalla80.fsf_-_@x220.int.ebiederm.org> User-Agent: Gnus/5.13 (Gnus v5.13) Emacs/26.1 (gnu/linux) MIME-Version: 1.0 Content-Type: text/plain X-XM-SPF: eid=1jnqw3-0003FG-0o;;;mid=<87ftalla80.fsf_-_@x220.int.ebiederm.org>;;;hst=in01.mta.xmission.com;;;ip=68.227.160.95;;;frm=ebiederm@xmission.com;;;spf=neutral X-XM-AID: U2FsdGVkX192WU5omqc6II6syhD9LEG1XUFeakeHPL4= X-SA-Exim-Connect-IP: 68.227.160.95 X-SA-Exim-Mail-From: ebiederm@xmission.com Subject: [PATCH v2 2/6] exec: Lock more defensively in exec X-SA-Exim-Version: 4.2.1 (built Thu, 05 May 2016 13:38:54 -0600) X-SA-Exim-Scanned: Yes (on in01.mta.xmission.com) Sender: linux-kernel-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org When taking the task_list_lock in de_thread also take the siglock. This makes de_thread closer to fork the canonical place where these locks are taken. To complete the defensiveness always take siglock when clearing group_exit_task and notify_count. This gives now gives the guarantee that group_exit_task and notify_count are now always changed under siglock. As anything multi-threaded in exec is a rare and slow path I don't think we care if we take an extra lock in practice. The practical reason for doing this is to enable setting signal->flags along with group_exit_task so that the function signal_group_exit can be simplified. Signed-off-by: "Eric W. Biederman" --- fs/exec.c | 11 +++++++++-- 1 file changed, 9 insertions(+), 2 deletions(-) diff --git a/fs/exec.c b/fs/exec.c index e6e8a9a70327..33b5d9229c01 100644 --- a/fs/exec.c +++ b/fs/exec.c @@ -1171,6 +1171,7 @@ static int de_thread(struct task_struct *tsk) for (;;) { cgroup_threadgroup_change_begin(tsk); write_lock_irq(&tasklist_lock); + spin_lock(lock); /* * Do this under tasklist_lock to ensure that * exit_notify() can't miss ->group_exit_task @@ -1179,6 +1180,7 @@ static int de_thread(struct task_struct *tsk) if (likely(leader->exit_state)) break; __set_current_state(TASK_KILLABLE); + spin_unlock(lock); write_unlock_irq(&tasklist_lock); cgroup_threadgroup_change_end(tsk); schedule(); @@ -1234,14 +1236,17 @@ static int de_thread(struct task_struct *tsk) */ if (unlikely(leader->ptrace)) __wake_up_parent(leader, leader->parent); + spin_unlock(lock); write_unlock_irq(&tasklist_lock); cgroup_threadgroup_change_end(tsk); release_task(leader); } + spin_lock_irq(lock); sig->group_exit_task = NULL; sig->notify_count = 0; + spin_unlock_irq(lock); no_thread_group: /* we have changed execution domain */ @@ -1252,10 +1257,12 @@ static int de_thread(struct task_struct *tsk) killed: /* protects against exit_notify() and __exit_signal() */ - read_lock(&tasklist_lock); + read_lock_irq(&tasklist_lock); + spin_lock(lock); sig->group_exit_task = NULL; sig->notify_count = 0; - read_unlock(&tasklist_lock); + spin_unlock(lock); + read_unlock_irq(&tasklist_lock); return -EAGAIN; } -- 2.20.1