From: Tejun Heo <tj@kernel.org>
To: Lai Jiangshan <laijs@cn.fujitsu.com>
Cc: linux-kernel@vger.kernel.org
Subject: [PATCH 4/4] workqueue: better define synchronization rule around rescuer->pool updates
Date: Tue, 19 Feb 2013 12:22:53 -0800 [thread overview]
Message-ID: <20130219202253.GX17414@htj.dyndns.org> (raw)
In-Reply-To: <1361203940-6300-9-git-send-email-laijs@cn.fujitsu.com>
>From 7bceeff75e5599f5c026797229d897f518dc028e Mon Sep 17 00:00:00 2001
From: Lai Jiangshan <laijs@cn.fujitsu.com>
Date: Tue, 19 Feb 2013 12:17:02 -0800
Rescuers visit different worker_pools to process work items from pools
under pressure. Currently, rescuer->pool is updated outside any
locking and when an outsider looks at a rescuer, there's no way to
tell when and whether rescuer->pool is gonna change. While this
doesn't currently cause any problem, it is nasty.
With recent worker_maybe_bind_and_lock() changes, we can move
rescuer->pool updates inside pool locks such that if rescuer->pool
equals a locked pool, it's guaranteed to stay that way until the pool
is unlocked.
Move rescuer->pool inside pool->lock.
This patch doesn't introduce any visible behavior difference.
tj: Updated the description.
Signed-off-by: Lai Jiangshan <laijs@cn.fujitsu.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
---
Applied to wq/for-3.10-tmp. Note that I'm gonna reabse the branch
once 3.9-rc1 is out.
Thanks.
kernel/workqueue.c | 3 ++-
kernel/workqueue_internal.h | 1 +
2 files changed, 3 insertions(+), 1 deletion(-)
diff --git a/kernel/workqueue.c b/kernel/workqueue.c
index 1770b09..0b1e6f2 100644
--- a/kernel/workqueue.c
+++ b/kernel/workqueue.c
@@ -2359,8 +2359,8 @@ repeat:
mayday_clear_cpu(cpu, wq->mayday_mask);
/* migrate to the target cpu if possible */
- rescuer->pool = pool;
worker_maybe_bind_and_lock(pool);
+ rescuer->pool = pool;
/*
* Slurp in all works issued via this workqueue and
@@ -2381,6 +2381,7 @@ repeat:
if (keep_working(pool))
wake_up_worker(pool);
+ rescuer->pool = NULL;
spin_unlock_irq(&pool->lock);
}
diff --git a/kernel/workqueue_internal.h b/kernel/workqueue_internal.h
index 0765026..f9c8877 100644
--- a/kernel/workqueue_internal.h
+++ b/kernel/workqueue_internal.h
@@ -32,6 +32,7 @@ struct worker {
struct list_head scheduled; /* L: scheduled works */
struct task_struct *task; /* I: worker task */
struct worker_pool *pool; /* I: the associated pool */
+ /* L: for rescuers */
/* 64 bytes boundary on 64bit, 32 on 32bit */
unsigned long last_active; /* L: last active timestamp */
unsigned int flags; /* X: flags */
--
1.8.1
next prev parent reply other threads:[~2013-02-19 20:22 UTC|newest]
Thread overview: 25+ messages / expand[flat|nested] mbox.gz Atom feed top
2013-02-18 16:12 [PATCH V2 00/15] workqueue: enhance locking and lookup worker via ID Lai Jiangshan
2013-02-18 16:12 ` [PATCH V2 01/15] workqueue: add lock_work_pool() Lai Jiangshan
2013-02-18 16:12 ` [PATCH V2 02/15] workqueue: allow more work_pool id space Lai Jiangshan
2013-02-19 20:19 ` [PATCH 1/4] workqueue: allow more off-queue flag space Tejun Heo
2013-02-18 16:12 ` [PATCH V2 03/15] workqueue: remname current worker->id to worker->id_in_pool Lai Jiangshan
2013-02-18 16:12 ` [PATCH V2 04/15] workqueue: add worker's global worker ID Lai Jiangshan
2013-02-18 16:12 ` [PATCH V2 05/15] workqueue: only set pool id when the work is running Lai Jiangshan
2013-02-18 16:12 ` [PATCH V2 06/15] workqueue: use current instead of worker->task in worker_maybe_bind_and_lock() Lai Jiangshan
2013-02-19 20:20 ` [PATCH 2/4] workqueue: use %current " Tejun Heo
2013-02-18 16:12 ` [PATCH V2 07/15] workqueue: change argument of worker_maybe_bind_and_lock() to pool Lai Jiangshan
2013-02-19 20:20 ` [PATCH 3/4] workqueue: change argument of worker_maybe_bind_and_lock() to @pool Tejun Heo
2013-02-18 16:12 ` [PATCH V2 08/15] workqueue: only change worker->pool with pool lock held Lai Jiangshan
2013-02-19 20:22 ` Tejun Heo [this message]
2013-02-18 16:12 ` [PATCH V2 09/15] workqueue: use worker id in work->data instead Lai Jiangshan
2013-02-18 16:12 ` [PATCH V2 10/15] workqueue: avoid unneeded calls to get_work_cwq() Lai Jiangshan
2013-02-18 16:12 ` [PATCH V2 11/15] workqueue: split work_flags to delayed_flags and color_flags in __queue_work() Lai Jiangshan
2013-02-18 16:12 ` [PATCH V2 12/15] workqueue: add extra flags to set_work_worker_and_keep_pending() Lai Jiangshan
2013-02-18 16:12 ` [PATCH V2 13/15] workqueue: also record worker in work->data if running&&queued Lai Jiangshan
2013-02-18 19:50 ` Tejun Heo
2013-02-19 15:04 ` Lai Jiangshan
2013-02-19 19:27 ` Tejun Heo
2013-02-19 20:24 ` Tejun Heo
2013-02-18 16:12 ` [PATCH V2 14/15] workqueue: convert busy hash to busy list Lai Jiangshan
2013-02-18 16:12 ` [PATCH V2 15/15] workqueue: queue worker to busy list outside process_one_work() Lai Jiangshan
2013-02-18 16:28 ` [PATCH V2 00/15] workqueue: enhance locking and lookup worker via ID Lai Jiangshan
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=20130219202253.GX17414@htj.dyndns.org \
--to=tj@kernel.org \
--cc=laijs@cn.fujitsu.com \
--cc=linux-kernel@vger.kernel.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;
as well as URLs for NNTP newsgroup(s).