public inbox for linux-kernel@vger.kernel.org
 help / color / mirror / Atom feed
From: Tejun Heo <tj@kernel.org>
To: linux-kernel@vger.kernel.org, jeff@garzik.org, mingo@elte.hu,
	akpm@linux-foundation.org, jens.axboe@oracle.com,
	rusty@rustcorp.com.au, cl@linux-foundation.org,
	dhowells@redhat.com, arjan@linux.intel.com,
	torvalds@linux-foundation.org, avi@redhat.com,
	peterz@infradead.org, andi@firstfloor.org, fweisbec@gmail.com
Cc: Tejun Heo <tj@kernel.org>, Andi Kleen <ak@linux.intel.com>,
	Oleg Nesterov <oleg@redhat.com>
Subject: [PATCH 01/21] workqueue: fix race condition in schedule_on_each_cpu()
Date: Tue, 17 Nov 2009 02:15:06 +0900	[thread overview]
Message-ID: <1258391726-30264-2-git-send-email-tj@kernel.org> (raw)
In-Reply-To: <1258391726-30264-1-git-send-email-tj@kernel.org>

Commit 65a64464349883891e21e74af16c05d6e1eeb4e9 which allows
schedule_on_each_cpu() to be called from keventd added a race
condition.  schedule_on_each_cpu() may race with cpu hotplug and end
up executing the function twice on a cpu.

Fix it by moving direct execution into the section protected with
get/put_online_cpus().  While at it, update code such that direct
execution is done after works have been scheduled for all other cpus
and drop unnecessary cpu != orig test from flush loop.

Signed-off-by: Tejun Heo <tj@kernel.org>
Cc: Andi Kleen <ak@linux.intel.com>
Cc: Oleg Nesterov <oleg@redhat.com>
---
 kernel/workqueue.c |   28 +++++++++++++---------------
 1 files changed, 13 insertions(+), 15 deletions(-)

diff --git a/kernel/workqueue.c b/kernel/workqueue.c
index 1232814..67e526b 100644
--- a/kernel/workqueue.c
+++ b/kernel/workqueue.c
@@ -692,31 +692,29 @@ int schedule_on_each_cpu(work_func_t func)
 	if (!works)
 		return -ENOMEM;
 
+	get_online_cpus();
+
 	/*
-	 * when running in keventd don't schedule a work item on itself.
-	 * Can just call directly because the work queue is already bound.
-	 * This also is faster.
-	 * Make this a generic parameter for other workqueues?
+	 * When running in keventd don't schedule a work item on
+	 * itself.  Can just call directly because the work queue is
+	 * already bound.  This also is faster.
 	 */
-	if (current_is_keventd()) {
+	if (current_is_keventd())
 		orig = raw_smp_processor_id();
-		INIT_WORK(per_cpu_ptr(works, orig), func);
-		func(per_cpu_ptr(works, orig));
-	}
 
-	get_online_cpus();
 	for_each_online_cpu(cpu) {
 		struct work_struct *work = per_cpu_ptr(works, cpu);
 
-		if (cpu == orig)
-			continue;
 		INIT_WORK(work, func);
-		schedule_work_on(cpu, work);
-	}
-	for_each_online_cpu(cpu) {
 		if (cpu != orig)
-			flush_work(per_cpu_ptr(works, cpu));
+			schedule_work_on(cpu, work);
 	}
+	if (orig >= 0)
+		func(per_cpu_ptr(works, orig));
+
+	for_each_online_cpu(cpu)
+		flush_work(per_cpu_ptr(works, cpu));
+
 	put_online_cpus();
 	free_percpu(works);
 	return 0;
-- 
1.6.4.2


  reply	other threads:[~2009-11-16 17:17 UTC|newest]

Thread overview: 47+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2009-11-16 17:15 [PATCHSET] workqueue: prepare for concurrency managed workqueue Tejun Heo
2009-11-16 17:15 ` Tejun Heo [this message]
2009-11-16 23:04   ` [PATCH 01/21] workqueue: fix race condition in schedule_on_each_cpu() Frederic Weisbecker
2009-11-17  0:08     ` Tejun Heo
2009-11-17  7:04     ` Avi Kivity
2009-11-17 16:16       ` Tejun Heo
2009-11-16 17:15 ` [PATCH 02/21] sched, kvm: fix race condition involving sched_in_preempt_notifers Tejun Heo
2009-11-16 17:15 ` [PATCH 03/21] workqueue: Add debugobjects support Tejun Heo
2009-11-16 17:15 ` [PATCH 04/21] sched: implement scheduler notifiers Tejun Heo
2009-11-16 18:31   ` Avi Kivity
2009-11-16 18:43     ` Tejun Heo
2009-11-16 18:41   ` Peter Zijlstra
2009-11-16 18:54     ` Tejun Heo
2009-11-16 20:29       ` Peter Zijlstra
2009-11-17 16:16         ` Tejun Heo
2009-11-16 17:15 ` [PATCH 05/21] kvm: convert kvm to use new " Tejun Heo
2009-11-16 17:15 ` [PATCH 06/21] sched: drop preempt notifiers Tejun Heo
2009-11-16 17:15 ` [PATCH 07/21] sched: implement sched_notifier_wake_up_process() Tejun Heo
2009-11-16 17:15 ` [PATCH 08/21] scheduler: implement force_cpus_allowed_ptr() Tejun Heo
2009-11-17  5:14   ` Rusty Russell
2009-11-17  5:19     ` Tejun Heo
2009-11-16 17:15 ` [PATCH 09/21] acpi: use queue_work_on() instead of binding workqueue worker to cpu0 Tejun Heo
2009-11-16 17:15 ` [PATCH 10/21] stop_machine: reimplement without using workqueue Tejun Heo
2009-11-16 17:15 ` [PATCH 11/21] workqueue: misc/cosmetic updates Tejun Heo
2009-11-16 17:15 ` [PATCH 12/21] workqueue: merge feature parametesr into flags Tejun Heo
2009-11-16 17:15 ` [PATCH 13/21] workqueue: update cwq alignement and make one more flag bit available Tejun Heo
2009-11-16 17:15 ` [PATCH 14/21] workqueue: define both bit position and mask for work flags Tejun Heo
2009-11-16 17:15 ` [PATCH 15/21] workqueue: separate out process_one_work() Tejun Heo
2009-11-16 17:15 ` [PATCH 16/21] workqueue: temporarily disable workqueue tracing Tejun Heo
2009-11-16 17:15 ` [PATCH 17/21] workqueue: simple reimplementation of SINGLE_THREAD workqueue Tejun Heo
2009-11-17  0:47   ` Andy Walls
2009-11-17  5:23     ` Tejun Heo
2009-11-17 12:05       ` Andy Walls
2009-11-17 16:21         ` Tejun Heo
2009-11-17 16:26           ` Hi ... I want to introduce myself :) Setiajie &#20313;&#40251;&#26124;
2009-11-17 15:05       ` [PATCH 17/21] workqueue: simple reimplementation of SINGLE_THREAD workqueue Linus Torvalds
2009-11-17 16:12         ` Tejun Heo
2009-11-17 19:01           ` Linus Torvalds
2009-11-17 14:03   ` Johannes Berg
2009-11-17 16:24     ` Tejun Heo
2009-11-16 17:15 ` [PATCH 18/21] workqueue: reimplement workqueue flushing using color coded works Tejun Heo
2009-11-16 17:15 ` [PATCH 19/21] workqueue: introduce worker Tejun Heo
2009-11-17 11:39   ` Louis Rilling
2009-11-17 11:51     ` Louis Rilling
2009-11-17 16:25       ` Tejun Heo
2009-11-16 17:15 ` [PATCH 20/21] workqueue: reimplement work flushing using linked works Tejun Heo
2009-11-16 17:15 ` [PATCH 21/21] workqueue: reimplement workqueue freeze using cwq->frozen_works queue Tejun Heo

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=1258391726-30264-2-git-send-email-tj@kernel.org \
    --to=tj@kernel.org \
    --cc=ak@linux.intel.com \
    --cc=akpm@linux-foundation.org \
    --cc=andi@firstfloor.org \
    --cc=arjan@linux.intel.com \
    --cc=avi@redhat.com \
    --cc=cl@linux-foundation.org \
    --cc=dhowells@redhat.com \
    --cc=fweisbec@gmail.com \
    --cc=jeff@garzik.org \
    --cc=jens.axboe@oracle.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mingo@elte.hu \
    --cc=oleg@redhat.com \
    --cc=peterz@infradead.org \
    --cc=rusty@rustcorp.com.au \
    --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