From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1756994Ab3FDOWR (ORCPT ); Tue, 4 Jun 2013 10:22:17 -0400 Received: from youngberry.canonical.com ([91.189.89.112]:46731 "EHLO youngberry.canonical.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1755676Ab3FDOIB (ORCPT ); Tue, 4 Jun 2013 10:08:01 -0400 From: Luis Henriques To: linux-kernel@vger.kernel.org, stable@vger.kernel.org, kernel-team@lists.ubuntu.com Cc: Trond Myklebust , Luis Henriques Subject: [PATCH 25/70] SUNRPC: Prevent an rpc_task wakeup race Date: Tue, 4 Jun 2013 15:06:41 +0100 Message-Id: <1370354846-14213-26-git-send-email-luis.henriques@canonical.com> X-Mailer: git-send-email 1.8.1.2 In-Reply-To: <1370354846-14213-1-git-send-email-luis.henriques@canonical.com> References: <1370354846-14213-1-git-send-email-luis.henriques@canonical.com> X-Extended-Stable: 3.5 Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org 3.5.7.14 -stable review patch. If anyone has any objections, please let me know. ------------------ From: Trond Myklebust commit a3c3cac5d31879cd9ae2de7874dc6544ca704aec upstream. The lockless RPC_IS_QUEUED() test in __rpc_execute means that we need to be careful about ordering the calls to rpc_test_and_set_running(task) and rpc_clear_queued(task). If we get the order wrong, then we may end up testing the RPC_TASK_RUNNING flag after __rpc_execute() has looped and changed the state of the rpc_task. Signed-off-by: Trond Myklebust Signed-off-by: Luis Henriques --- net/sunrpc/sched.c | 8 +++++++- 1 file changed, 7 insertions(+), 1 deletion(-) diff --git a/net/sunrpc/sched.c b/net/sunrpc/sched.c index 38cca32..c393017 100644 --- a/net/sunrpc/sched.c +++ b/net/sunrpc/sched.c @@ -305,11 +305,17 @@ EXPORT_SYMBOL_GPL(__rpc_wait_for_completion_task); * Note: If the task is ASYNC, and is being made runnable after sitting on an * rpc_wait_queue, this must be called with the queue spinlock held to protect * the wait queue operation. + * Note the ordering of rpc_test_and_set_running() and rpc_clear_queued(), + * which is needed to ensure that __rpc_execute() doesn't loop (due to the + * lockless RPC_IS_QUEUED() test) before we've had a chance to test + * the RPC_TASK_RUNNING flag. */ static void rpc_make_runnable(struct rpc_task *task) { + bool need_wakeup = !rpc_test_and_set_running(task); + rpc_clear_queued(task); - if (rpc_test_and_set_running(task)) + if (!need_wakeup) return; if (RPC_IS_ASYNC(task)) { INIT_WORK(&task->u.tk_work, rpc_async_schedule); -- 1.8.1.2