From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1751817AbaGVFBV (ORCPT ); Tue, 22 Jul 2014 01:01:21 -0400 Received: from cn.fujitsu.com ([59.151.112.132]:19244 "EHLO heian.cn.fujitsu.com" rhost-flags-OK-FAIL-OK-FAIL) by vger.kernel.org with ESMTP id S1751095AbaGVFBI (ORCPT ); Tue, 22 Jul 2014 01:01:08 -0400 X-IronPort-AV: E=Sophos;i="5.00,931,1396972800"; d="scan'208";a="33612898" From: Lai Jiangshan To: CC: Tejun Heo , Lai Jiangshan Subject: [PATCH 2/2 V2] workqueue: remove the argument @wakeup from worker_set_flags() Date: Tue, 22 Jul 2014 13:02:00 +0800 Message-ID: <1406005330-11864-2-git-send-email-laijs@cn.fujitsu.com> X-Mailer: git-send-email 1.7.4.4 In-Reply-To: <1406005330-11864-1-git-send-email-laijs@cn.fujitsu.com> References: <20140718225317.GC5739@htj.dyndns.org> <1406005330-11864-1-git-send-email-laijs@cn.fujitsu.com> MIME-Version: 1.0 Content-Type: text/plain X-Originating-IP: [10.167.226.103] Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org worker_set_flags() doesn't necessarily wake next worker and the @wakeup can be removed, the caller can use the following conbination instead when needed: worker_set_flags(); if (need_more_worker(pool)) wake_up_worker(pool); It reduces the machine-code and footprint for the process_one_work() if worker_set_flags() is really inlined. Signed-off-by: Lai Jiangshan --- kernel/workqueue.c | 27 ++++++++------------------- 1 files changed, 8 insertions(+), 19 deletions(-) diff --git a/kernel/workqueue.c b/kernel/workqueue.c index cd75689..40b1f00 100644 --- a/kernel/workqueue.c +++ b/kernel/workqueue.c @@ -867,35 +867,22 @@ struct task_struct *wq_worker_sleeping(struct task_struct *task, int cpu) * worker_set_flags - set worker flags and adjust nr_running accordingly * @worker: self * @flags: flags to set - * @wakeup: wakeup an idle worker if necessary * - * Set @flags in @worker->flags and adjust nr_running accordingly. If - * nr_running becomes zero and @wakeup is %true, an idle worker is - * woken up. + * Set @flags in @worker->flags and adjust nr_running accordingly. * * CONTEXT: * spin_lock_irq(pool->lock) */ -static inline void worker_set_flags(struct worker *worker, unsigned int flags, - bool wakeup) +static inline void worker_set_flags(struct worker *worker, unsigned int flags) { struct worker_pool *pool = worker->pool; WARN_ON_ONCE(worker->task != current); - /* - * If transitioning into NOT_RUNNING, adjust nr_running and - * wake up an idle worker as necessary if requested by - * @wakeup. - */ + /* If transitioning into NOT_RUNNING, adjust nr_running. */ if ((flags & WORKER_NOT_RUNNING) && !(worker->flags & WORKER_NOT_RUNNING)) { - if (wakeup) { - if (atomic_dec_and_test(&pool->nr_running) && - !list_empty(&pool->worklist)) - wake_up_worker(pool); - } else - atomic_dec(&pool->nr_running); + atomic_dec(&pool->nr_running); } worker->flags |= flags; @@ -2043,9 +2030,11 @@ __acquires(&pool->lock) /* * CPU intensive works don't participate in concurrency * management. They're the scheduler's responsibility. + * Another worker should be woken up after WORKER_CPU_INTENSIVE + * is set, see next code. */ if (unlikely(cpu_intensive)) - worker_set_flags(worker, WORKER_CPU_INTENSIVE, true); + worker_set_flags(worker, WORKER_CPU_INTENSIVE); /* * Wake up another worker if necessary. It is a no-op @@ -2210,7 +2199,7 @@ recheck: } } while (keep_working(pool)); - worker_set_flags(worker, WORKER_PREP, false); + worker_set_flags(worker, WORKER_PREP); sleep: /* * pool->lock is held and there's no work to process and no need to -- 1.7.4.4