From: Tejun Heo <tj@kernel.org>
To: linux-kernel@vger.kernel.org, laijs@cn.fujitsu.com
Cc: axboe@kernel.dk, jmoyer@redhat.com, zab@redhat.com,
Tejun Heo <tj@kernel.org>
Subject: [PATCH 03/31] workqueue: introduce kmem_cache for pool_workqueues
Date: Fri, 1 Mar 2013 19:23:54 -0800 [thread overview]
Message-ID: <1362194662-2344-4-git-send-email-tj@kernel.org> (raw)
In-Reply-To: <1362194662-2344-1-git-send-email-tj@kernel.org>
pool_workqueues need to be aligned to 1 << WORK_STRUCT_FLAG_BITS as
the lower bits of work->data are used for flags when they're pointing
to pool_workqueues.
Due to historical reasons, unbound pool_workqueues are allocated using
kzalloc() with sufficient buffer area for alignment and aligned
manually. The original pointer is stored at the end which free_pwqs()
retrieves when freeing it.
There's no reason for this hackery anymore. Set alignment of struct
pool_workqueue to 1 << WORK_STRUCT_FLAG_BITS, add kmem_cache for
pool_workqueues with proper alignment and replace the hacky alloc and
free implementation with plain kmem_cache_zalloc/free().
In case WORK_STRUCT_FLAG_BITS gets shrunk too much and makes fields of
pool_workqueues misaligned, trigger WARN if the alignment of struct
pool_workqueue becomes smaller than that of long long.
Note that assertion on IS_ALIGNED() is removed from alloc_pwqs(). We
already have another one in pwq init loop in __alloc_workqueue_key().
This patch doesn't introduce any visible behavior changes.
Signed-off-by: Tejun Heo <tj@kernel.org>
---
kernel/workqueue.c | 43 ++++++++++++-------------------------------
1 file changed, 12 insertions(+), 31 deletions(-)
diff --git a/kernel/workqueue.c b/kernel/workqueue.c
index 61f78ef..69f1268 100644
--- a/kernel/workqueue.c
+++ b/kernel/workqueue.c
@@ -169,7 +169,7 @@ struct pool_workqueue {
int nr_active; /* L: nr of active works */
int max_active; /* L: max active works */
struct list_head delayed_works; /* L: delayed works */
-};
+} __aligned(1 << WORK_STRUCT_FLAG_BITS);
/*
* Structure used to wait for workqueue flush.
@@ -233,6 +233,8 @@ struct workqueue_struct {
char name[]; /* I: workqueue name */
};
+static struct kmem_cache *pwq_cache;
+
struct workqueue_struct *system_wq __read_mostly;
EXPORT_SYMBOL_GPL(system_wq);
struct workqueue_struct *system_highpri_wq __read_mostly;
@@ -3098,34 +3100,11 @@ int keventd_up(void)
static int alloc_pwqs(struct workqueue_struct *wq)
{
- /*
- * pwqs are forced aligned according to WORK_STRUCT_FLAG_BITS.
- * Make sure that the alignment isn't lower than that of
- * unsigned long long.
- */
- const size_t size = sizeof(struct pool_workqueue);
- const size_t align = max_t(size_t, 1 << WORK_STRUCT_FLAG_BITS,
- __alignof__(unsigned long long));
-
if (!(wq->flags & WQ_UNBOUND))
- wq->pool_wq.pcpu = __alloc_percpu(size, align);
- else {
- void *ptr;
-
- /*
- * Allocate enough room to align pwq and put an extra
- * pointer at the end pointing back to the originally
- * allocated pointer which will be used for free.
- */
- ptr = kzalloc(size + align + sizeof(void *), GFP_KERNEL);
- if (ptr) {
- wq->pool_wq.single = PTR_ALIGN(ptr, align);
- *(void **)(wq->pool_wq.single + 1) = ptr;
- }
- }
+ wq->pool_wq.pcpu = alloc_percpu(struct pool_workqueue);
+ else
+ wq->pool_wq.single = kmem_cache_zalloc(pwq_cache, GFP_KERNEL);
- /* just in case, make sure it's actually aligned */
- BUG_ON(!IS_ALIGNED(wq->pool_wq.v, align));
return wq->pool_wq.v ? 0 : -ENOMEM;
}
@@ -3133,10 +3112,8 @@ static void free_pwqs(struct workqueue_struct *wq)
{
if (!(wq->flags & WQ_UNBOUND))
free_percpu(wq->pool_wq.pcpu);
- else if (wq->pool_wq.single) {
- /* the pointer to free is stored right after the pwq */
- kfree(*(void **)(wq->pool_wq.single + 1));
- }
+ else
+ kmem_cache_free(pwq_cache, wq->pool_wq.single);
}
static int wq_clamp_max_active(int max_active, unsigned int flags,
@@ -3737,6 +3714,10 @@ static int __init init_workqueues(void)
BUILD_BUG_ON((1LU << (BITS_PER_LONG - WORK_OFFQ_POOL_SHIFT)) <
WORK_CPU_END * NR_STD_WORKER_POOLS);
+ WARN_ON(__alignof__(struct pool_workqueue) < __alignof__(long long));
+
+ pwq_cache = KMEM_CACHE(pool_workqueue, SLAB_PANIC);
+
cpu_notifier(workqueue_cpu_up_callback, CPU_PRI_WORKQUEUE_UP);
hotcpu_notifier(workqueue_cpu_down_callback, CPU_PRI_WORKQUEUE_DOWN);
--
1.8.1.2
next prev parent reply other threads:[~2013-03-02 3:24 UTC|newest]
Thread overview: 77+ messages / expand[flat|nested] mbox.gz Atom feed top
2013-03-02 3:23 [PATCHSET wq/for-3.10-tmp] workqueue: implement workqueue with custom worker attributes Tejun Heo
2013-03-02 3:23 ` [PATCH 01/31] workqueue: make sanity checks less punshing using WARN_ON[_ONCE]()s Tejun Heo
2013-03-02 3:23 ` [PATCH 02/31] workqueue: make workqueue_lock irq-safe Tejun Heo
2013-03-02 3:23 ` Tejun Heo [this message]
2013-03-02 3:23 ` [PATCH 04/31] workqueue: add workqueue_struct->pwqs list Tejun Heo
2013-03-02 3:23 ` [PATCH 05/31] workqueue: replace for_each_pwq_cpu() with for_each_pwq() Tejun Heo
2013-03-02 3:23 ` [PATCH 06/31] workqueue: introduce for_each_pool() Tejun Heo
2013-03-02 3:23 ` [PATCH 07/31] workqueue: restructure pool / pool_workqueue iterations in freeze/thaw functions Tejun Heo
2013-03-10 10:09 ` Lai Jiangshan
2013-03-10 12:34 ` Tejun Heo
2013-03-02 3:23 ` [PATCH 08/31] workqueue: add wokrqueue_struct->maydays list to replace mayday cpu iterators Tejun Heo
2013-03-02 3:24 ` [PATCH 09/31] workqueue: consistently use int for @cpu variables Tejun Heo
2013-03-02 3:24 ` [PATCH 10/31] workqueue: remove workqueue_struct->pool_wq.single Tejun Heo
2013-03-02 3:24 ` [PATCH 11/31] workqueue: replace get_pwq() with explicit per_cpu_ptr() accesses and first_pwq() Tejun Heo
2013-03-02 3:24 ` [PATCH 12/31] workqueue: update synchronization rules on workqueue->pwqs Tejun Heo
2013-03-10 10:09 ` Lai Jiangshan
2013-03-10 12:38 ` Tejun Heo
2013-03-12 18:20 ` [PATCH v2 " Tejun Heo
2013-03-02 3:24 ` [PATCH 13/31] workqueue: update synchronization rules on worker_pool_idr Tejun Heo
2013-03-12 18:20 ` [PATCH v2 " Tejun Heo
2013-03-02 3:24 ` [PATCH 14/31] workqueue: replace POOL_MANAGING_WORKERS flag with worker_pool->manager_mutex Tejun Heo
2013-03-10 10:09 ` Lai Jiangshan
2013-03-10 12:46 ` Tejun Heo
2013-03-12 18:19 ` [PATCH v2 " Tejun Heo
2013-03-02 3:24 ` [PATCH 15/31] workqueue: separate out init_worker_pool() from init_workqueues() Tejun Heo
2013-03-02 3:24 ` [PATCH 16/31] workqueue: introduce workqueue_attrs Tejun Heo
2013-03-04 18:37 ` [PATCH v2 " Tejun Heo
2013-03-05 22:29 ` Ryan Mallon
2013-03-05 22:33 ` Tejun Heo
2013-03-05 22:34 ` Tejun Heo
2013-03-05 22:40 ` Ryan Mallon
2013-03-05 22:44 ` Tejun Heo
2013-03-05 23:20 ` Ryan Mallon
2013-03-05 23:28 ` Tejun Heo
2013-03-02 3:24 ` [PATCH 17/31] workqueue: implement attribute-based unbound worker_pool management Tejun Heo
2013-03-10 10:08 ` Lai Jiangshan
2013-03-10 12:58 ` Tejun Heo
2013-03-10 18:36 ` Tejun Heo
2013-03-12 18:21 ` [PATCH v2 " Tejun Heo
2013-03-02 3:24 ` [PATCH 18/31] workqueue: remove unbound_std_worker_pools[] and related helpers Tejun Heo
2013-03-02 3:24 ` [PATCH 19/31] workqueue: drop "std" from cpu_std_worker_pools and for_each_std_worker_pool() Tejun Heo
2013-03-02 3:24 ` [PATCH 20/31] workqueue: add pool ID to the names of unbound kworkers Tejun Heo
2013-03-02 3:24 ` [PATCH 21/31] workqueue: drop WQ_RESCUER and test workqueue->rescuer for NULL instead Tejun Heo
2013-03-02 3:24 ` [PATCH 22/31] workqueue: restructure __alloc_workqueue_key() Tejun Heo
2013-03-02 3:24 ` [PATCH 23/31] workqueue: implement get/put_pwq() Tejun Heo
2013-03-02 3:24 ` [PATCH 24/31] workqueue: prepare flush_workqueue() for dynamic creation and destrucion of unbound pool_workqueues Tejun Heo
2013-03-02 3:24 ` [PATCH 25/31] workqueue: perform non-reentrancy test when queueing to unbound workqueues too Tejun Heo
2013-03-02 3:24 ` [PATCH 26/31] workqueue: implement apply_workqueue_attrs() Tejun Heo
2013-03-02 3:24 ` [PATCH 27/31] workqueue: make it clear that WQ_DRAINING is an internal flag Tejun Heo
2013-03-02 3:24 ` [PATCH 28/31] workqueue: reject increasing max_active for ordered workqueues Tejun Heo
2013-03-04 18:30 ` [PATCH UPDATED 28/31] workqueue: reject adjusting max_active or applying attrs to " Tejun Heo
2013-03-02 3:24 ` [PATCH 29/31] cpumask: implement cpumask_parse() Tejun Heo
2013-03-02 3:24 ` [PATCH 30/31] driver/base: implement subsys_virtual_register() Tejun Heo
2013-03-02 18:17 ` Greg Kroah-Hartman
2013-03-02 20:26 ` Tejun Heo
2013-03-03 6:42 ` Kay Sievers
2013-03-05 20:43 ` Tejun Heo
2013-03-07 23:31 ` Greg Kroah-Hartman
2013-03-08 0:04 ` Kay Sievers
2013-03-10 11:57 ` Tejun Heo
2013-03-10 16:45 ` Greg Kroah-Hartman
2013-03-10 17:00 ` Kay Sievers
2013-03-10 17:24 ` Greg Kroah-Hartman
2013-03-10 17:50 ` Kay Sievers
2013-03-10 18:34 ` Tejun Heo
2013-03-12 18:40 ` Tejun Heo
2013-03-02 3:24 ` [PATCH 31/31] workqueue: implement sysfs interface for workqueues Tejun Heo
2013-03-04 18:30 ` [PATCH v2 " Tejun Heo
2013-03-05 20:41 ` [PATCHSET wq/for-3.10-tmp] workqueue: implement workqueue with custom worker attributes Tejun Heo
2013-03-10 10:34 ` Lai Jiangshan
2013-03-10 12:01 ` Tejun Heo
2013-03-11 15:24 ` Tejun Heo
2013-03-11 15:40 ` Lai Jiangshan
2013-03-11 15:42 ` Lai Jiangshan
2013-03-11 15:43 ` Tejun Heo
2013-03-12 18:10 ` Tejun Heo
2013-03-12 18:34 ` 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=1362194662-2344-4-git-send-email-tj@kernel.org \
--to=tj@kernel.org \
--cc=axboe@kernel.dk \
--cc=jmoyer@redhat.com \
--cc=laijs@cn.fujitsu.com \
--cc=linux-kernel@vger.kernel.org \
--cc=zab@redhat.com \
/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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.