Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1759999Ab3FNClb (ORCPT ); Thu, 13 Jun 2013 22:41:31 -0400 Received: from mail-pb0-f47.google.com ([209.85.160.47]:52357 "EHLO mail-pb0-f47.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1759888Ab3FNCl0 (ORCPT ); Thu, 13 Jun 2013 22:41:26 -0400 Date: Thu, 13 Jun 2013 19:41:22 -0700 From: Tejun Heo To: Tejun Heo Cc: Kent Overstreet , linux-kernel@vger.kernel.org, Rusty Russell , Oleg Nesterov , Christoph Lameter Subject: [PATCH v3 2/2] percpu-refcount: implement percpu_tryget() along with percpu_ref_kill_and_confirm() Message-ID: <20130614024122.GA7846@htj.dyndns.org> References: <20130612204532.GB15092@htj.dyndns.org> <20130612204627.GC15092@htj.dyndns.org> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20130612204627.GC15092@htj.dyndns.org> User-Agent: Mutt/1.5.21 (2010-09-15) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 6062 Lines: 171 >From dbece3a0f1ef0b19aff1cc6ed0942fec9ab98de1 Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Thu, 13 Jun 2013 19:23:53 -0700 Implement percpu_tryget() which stops giving out references once the percpu_ref is visible as killed. Because the refcnt is per-cpu, different CPUs will start to see a refcnt as killed at different points in time and tryget() may continue to succeed on subset of cpus for a while after percpu_ref_kill() returns. For use cases where it's necessary to know when all CPUs start to see the refcnt as dead, percpu_ref_kill_and_confirm() is added. The new function takes an extra argument @confirm_kill which is invoked when the refcnt is guaranteed to be viewed as killed on all CPUs. While this isn't the prettiest interface, it doesn't force synchronous wait and is much safer than requiring the caller to do its own call_rcu(). v2: Patch description rephrased to emphasize that tryget() may continue to succeed on some CPUs after kill() returns as suggested by Kent. v3: Function comment in percpu_ref_kill_and_confirm() updated warning people to not depend on the implied RCU grace period from the confirm callback as it's an implementation detail. Signed-off-by: Tejun Heo Slightly-Grumpily-Acked-by: Kent Overstreet --- Applied to percpu/for-3.11. Thanks! include/linux/percpu-refcount.h | 50 ++++++++++++++++++++++++++++++++++++++++- lib/percpu-refcount.c | 23 ++++++++++++++----- 2 files changed, 66 insertions(+), 7 deletions(-) diff --git a/include/linux/percpu-refcount.h b/include/linux/percpu-refcount.h index 6d843d6..dd2a086 100644 --- a/include/linux/percpu-refcount.h +++ b/include/linux/percpu-refcount.h @@ -63,13 +63,30 @@ struct percpu_ref { */ unsigned __percpu *pcpu_count; percpu_ref_func_t *release; + percpu_ref_func_t *confirm_kill; struct rcu_head rcu; }; int __must_check percpu_ref_init(struct percpu_ref *ref, percpu_ref_func_t *release); void percpu_ref_cancel_init(struct percpu_ref *ref); -void percpu_ref_kill(struct percpu_ref *ref); +void percpu_ref_kill_and_confirm(struct percpu_ref *ref, + percpu_ref_func_t *confirm_kill); + +/** + * percpu_ref_kill - drop the initial ref + * @ref: percpu_ref to kill + * + * Must be used to drop the initial ref on a percpu refcount; must be called + * precisely once before shutdown. + * + * Puts @ref in non percpu mode, then does a call_rcu() before gathering up the + * percpu counters and dropping the initial ref. + */ +static inline void percpu_ref_kill(struct percpu_ref *ref) +{ + return percpu_ref_kill_and_confirm(ref, NULL); +} #define PCPU_STATUS_BITS 2 #define PCPU_STATUS_MASK ((1 << PCPU_STATUS_BITS) - 1) @@ -101,6 +118,37 @@ static inline void percpu_ref_get(struct percpu_ref *ref) } /** + * percpu_ref_tryget - try to increment a percpu refcount + * @ref: percpu_ref to try-get + * + * Increment a percpu refcount unless it has already been killed. Returns + * %true on success; %false on failure. + * + * Completion of percpu_ref_kill() in itself doesn't guarantee that tryget + * will fail. For such guarantee, percpu_ref_kill_and_confirm() should be + * used. After the confirm_kill callback is invoked, it's guaranteed that + * no new reference will be given out by percpu_ref_tryget(). + */ +static inline bool percpu_ref_tryget(struct percpu_ref *ref) +{ + unsigned __percpu *pcpu_count; + int ret = false; + + rcu_read_lock(); + + pcpu_count = ACCESS_ONCE(ref->pcpu_count); + + if (likely(REF_STATUS(pcpu_count) == PCPU_REF_PTR)) { + __this_cpu_inc(*pcpu_count); + ret = true; + } + + rcu_read_unlock(); + + return ret; +} + +/** * percpu_ref_put - decrement a percpu refcount * @ref: percpu_ref to put * diff --git a/lib/percpu-refcount.c b/lib/percpu-refcount.c index ebeaac2..8bf9e71 100644 --- a/lib/percpu-refcount.c +++ b/lib/percpu-refcount.c @@ -118,6 +118,10 @@ static void percpu_ref_kill_rcu(struct rcu_head *rcu) atomic_add((int) count - PCPU_COUNT_BIAS, &ref->count); + /* @ref is viewed as dead on all CPUs, send out kill confirmation */ + if (ref->confirm_kill) + ref->confirm_kill(ref); + /* * Now we're in single atomic_t mode with a consistent refcount, so it's * safe to drop our initial ref: @@ -126,22 +130,29 @@ static void percpu_ref_kill_rcu(struct rcu_head *rcu) } /** - * percpu_ref_kill - safely drop initial ref + * percpu_ref_kill_and_confirm - drop the initial ref and schedule confirmation * @ref: percpu_ref to kill + * @confirm_kill: optional confirmation callback * - * Must be used to drop the initial ref on a percpu refcount; must be called - * precisely once before shutdown. + * Equivalent to percpu_ref_kill() but also schedules kill confirmation if + * @confirm_kill is not NULL. @confirm_kill, which may not block, will be + * called after @ref is seen as dead from all CPUs - all further + * invocations of percpu_ref_tryget() will fail. See percpu_ref_tryget() + * for more details. * - * Puts @ref in non percpu mode, then does a call_rcu() before gathering up the - * percpu counters and dropping the initial ref. + * Due to the way percpu_ref is implemented, @confirm_kill will be called + * after at least one full RCU grace period has passed but this is an + * implementation detail and callers must not depend on it. */ -void percpu_ref_kill(struct percpu_ref *ref) +void percpu_ref_kill_and_confirm(struct percpu_ref *ref, + percpu_ref_func_t *confirm_kill) { WARN_ONCE(REF_STATUS(ref->pcpu_count) == PCPU_REF_DEAD, "percpu_ref_kill() called more than once!\n"); ref->pcpu_count = (unsigned __percpu *) (((unsigned long) ref->pcpu_count)|PCPU_REF_DEAD); + ref->confirm_kill = confirm_kill; call_rcu(&ref->rcu, percpu_ref_kill_rcu); } -- 1.8.2.1 -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/