From: Tejun Heo <tj@kernel.org>
To: torvalds@linux-foundation.org, awalls@radix.net,
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, avi@redhat.com,
peterz@infradead.org, johannes@sipsolutions.net,
andi@firstfloor.org
Cc: Tejun Heo <tj@kernel.org>
Subject: [PATCH 15/27] workqueue: update cwq alignement
Date: Fri, 18 Dec 2009 21:57:56 +0900 [thread overview]
Message-ID: <1261141088-2014-16-git-send-email-tj@kernel.org> (raw)
In-Reply-To: <1261141088-2014-1-git-send-email-tj@kernel.org>
work->data field is used for two purposes. It points to cwq it's
queued on and the lower bits are used for flags. Currently, two bits
are reserved which is always safe as 4 byte alignment is guaranteed on
every architecture. However, future changes will need more flag bits.
On SMP, the percpu allocator is capable of honoring larger alignment
(there are other users which depend on it) and larger alignment works
just fine. On UP, percpu allocator is a thin wrapper around
kzalloc/kfree() and don't honor alignment request.
This patch introduces WORK_STRUCT_FLAG_BITS and implements
alloc/free_cwqs() which guarantees (1 << WORK_STRUCT_FLAG_BITS)
alignment both on SMP and UP. On SMP, simply wrapping percpu
allocator is enouhg. On UP, extra space is allocated so that cwq can
be aligned and the original pointer can be stored after it which is
used in the free path.
While at it, as cwqs are now forced aligned, make sure the resulting
alignment is at least equal to or larger than that of long long.
Alignment problem on UP is reported by Michal Simek.
Signed-off-by: Tejun Heo <tj@kernel.org>
Cc: Christoph Lameter <cl@linux-foundation.org>
Cc: Ingo Molnar <mingo@elte.hu>
Reported-by: Michal Simek <michal.simek@petalogix.com>
---
include/linux/workqueue.h | 4 ++-
kernel/workqueue.c | 59 +++++++++++++++++++++++++++++++++++++++++----
2 files changed, 57 insertions(+), 6 deletions(-)
diff --git a/include/linux/workqueue.h b/include/linux/workqueue.h
index e51b5dc..011738a 100644
--- a/include/linux/workqueue.h
+++ b/include/linux/workqueue.h
@@ -29,7 +29,9 @@ enum {
WORK_STRUCT_PENDING = 1 << WORK_STRUCT_PENDING_BIT,
WORK_STRUCT_STATIC = 1 << WORK_STRUCT_STATIC_BIT,
- WORK_STRUCT_FLAG_MASK = 3UL,
+ WORK_STRUCT_FLAG_BITS = 2,
+
+ WORK_STRUCT_FLAG_MASK = (1UL << WORK_STRUCT_FLAG_BITS) - 1,
WORK_STRUCT_WQ_DATA_MASK = ~WORK_STRUCT_FLAG_MASK,
};
diff --git a/kernel/workqueue.c b/kernel/workqueue.c
index d29e069..14edfcd 100644
--- a/kernel/workqueue.c
+++ b/kernel/workqueue.c
@@ -46,7 +46,9 @@
/*
* The per-CPU workqueue (if single thread, we always use the first
- * possible cpu).
+ * possible cpu). The lower WORK_STRUCT_FLAG_BITS of
+ * work_struct->data are used for flags and thus cwqs need to be
+ * aligned at two's power of the number of flag bits.
*/
struct cpu_workqueue_struct {
@@ -58,7 +60,7 @@ struct cpu_workqueue_struct {
struct workqueue_struct *wq; /* I: the owning workqueue */
struct task_struct *thread;
-} ____cacheline_aligned;
+};
/*
* The externally visible workqueue abstraction is an array of
@@ -930,6 +932,44 @@ int current_is_keventd(void)
}
+static struct cpu_workqueue_struct *alloc_cwqs(void)
+{
+ const size_t size = sizeof(struct cpu_workqueue_struct);
+ const size_t align = 1 << WORK_STRUCT_FLAG_BITS;
+ struct cpu_workqueue_struct *cwqs;
+#ifndef CONFIG_SMP
+ void *ptr;
+
+ /*
+ * On UP, percpu allocator doesn't honor alignment parameter
+ * and simply uses arch-dependent default. Allocate enough
+ * room to align cwq and put an extra pointer at the end
+ * pointing back to the originally allocated pointer which
+ * will be used for free.
+ */
+ ptr = __alloc_percpu(size + align + sizeof(void *), 1);
+ cwqs = PTR_ALIGN(ptr, align);
+ *(void **)per_cpu_ptr(cwqs + 1, 0) = ptr;
+#else
+ /* On SMP, percpu allocator can do it itself */
+ cwqs = __alloc_percpu(size, align);
+#endif
+ /* just in case, make sure it's actually aligned */
+ BUG_ON(!IS_ALIGNED((unsigned long)cwqs, align));
+ return cwqs;
+}
+
+static void free_cwqs(struct cpu_workqueue_struct *cwqs)
+{
+#ifndef CONFIG_SMP
+ /* on UP, the pointer to free is stored right after the cwq */
+ if (cwqs)
+ free_percpu(*(void **)per_cpu_ptr(cwqs + 1, 0));
+#else
+ free_percpu(cwqs);
+#endif
+}
+
static int create_workqueue_thread(struct cpu_workqueue_struct *cwq, int cpu)
{
struct workqueue_struct *wq = cwq->wq;
@@ -975,7 +1015,7 @@ struct workqueue_struct *__create_workqueue_key(const char *name,
if (!wq)
goto err;
- wq->cpu_wq = alloc_percpu(struct cpu_workqueue_struct);
+ wq->cpu_wq = alloc_cwqs();
if (!wq->cpu_wq)
goto err;
@@ -994,6 +1034,7 @@ struct workqueue_struct *__create_workqueue_key(const char *name,
for_each_possible_cpu(cpu) {
struct cpu_workqueue_struct *cwq = get_cwq(cpu, wq);
+ BUG_ON((unsigned long)cwq & WORK_STRUCT_FLAG_MASK);
cwq->wq = wq;
spin_lock_init(&cwq->lock);
INIT_LIST_HEAD(&cwq->worklist);
@@ -1021,7 +1062,7 @@ struct workqueue_struct *__create_workqueue_key(const char *name,
return wq;
err:
if (wq) {
- free_percpu(wq->cpu_wq);
+ free_cwqs(wq->cpu_wq);
kfree(wq);
}
return NULL;
@@ -1072,7 +1113,7 @@ void destroy_workqueue(struct workqueue_struct *wq)
for_each_possible_cpu(cpu)
cleanup_workqueue_thread(get_cwq(cpu, wq));
- free_percpu(wq->cpu_wq);
+ free_cwqs(wq->cpu_wq);
kfree(wq);
}
EXPORT_SYMBOL_GPL(destroy_workqueue);
@@ -1159,6 +1200,14 @@ EXPORT_SYMBOL_GPL(work_on_cpu);
void __init init_workqueues(void)
{
+ /*
+ * cwqs are forced aligned according to WORK_STRUCT_FLAG_BITS.
+ * Make sure that the alignment isn't lower than that of
+ * unsigned long long.
+ */
+ BUILD_BUG_ON(__alignof__(struct cpu_workqueue_struct) <
+ __alignof__(unsigned long long));
+
singlethread_cpu = cpumask_first(cpu_possible_mask);
hotcpu_notifier(workqueue_cpu_callback, 0);
keventd_wq = create_workqueue("events");
--
1.6.4.2
next prev parent reply other threads:[~2009-12-18 12:59 UTC|newest]
Thread overview: 104+ messages / expand[flat|nested] mbox.gz Atom feed top
2009-12-18 12:57 Tejun Heo
2009-12-18 12:57 ` [PATCH 01/27] sched: rename preempt_notifiers to sched_notifiers and refactor implementation Tejun Heo
2009-12-18 12:57 ` [PATCH 02/27] sched: refactor try_to_wake_up() Tejun Heo
2009-12-18 12:57 ` [PATCH 03/27] sched: implement __set_cpus_allowed() Tejun Heo
2009-12-18 12:57 ` [PATCH 04/27] sched: make sched_notifiers unconditional Tejun Heo
2009-12-18 12:57 ` [PATCH 05/27] sched: add wakeup/sleep sched_notifiers and allow NULL notifier ops Tejun Heo
2009-12-18 12:57 ` [PATCH 06/27] sched: implement try_to_wake_up_local() Tejun Heo
2009-12-18 12:57 ` [PATCH 07/27] acpi: use queue_work_on() instead of binding workqueue worker to cpu0 Tejun Heo
2009-12-18 12:57 ` [PATCH 08/27] stop_machine: reimplement without using workqueue Tejun Heo
2009-12-18 12:57 ` [PATCH 09/27] workqueue: misc/cosmetic updates Tejun Heo
2009-12-18 12:57 ` [PATCH 10/27] workqueue: merge feature parameters into flags Tejun Heo
2009-12-18 12:57 ` [PATCH 11/27] workqueue: define both bit position and mask for work flags Tejun Heo
2009-12-18 12:57 ` [PATCH 12/27] workqueue: separate out process_one_work() Tejun Heo
2009-12-18 12:57 ` [PATCH 13/27] workqueue: temporarily disable workqueue tracing Tejun Heo
2009-12-18 12:57 ` [PATCH 14/27] workqueue: kill cpu_populated_map Tejun Heo
2009-12-18 12:57 ` Tejun Heo [this message]
2009-12-18 12:57 ` [PATCH 16/27] workqueue: reimplement workqueue flushing using color coded works Tejun Heo
2009-12-18 12:57 ` [PATCH 17/27] workqueue: introduce worker Tejun Heo
2009-12-18 12:57 ` [PATCH 18/27] workqueue: reimplement work flushing using linked works Tejun Heo
2009-12-18 12:58 ` [PATCH 19/27] workqueue: implement per-cwq active work limit Tejun Heo
2009-12-18 12:58 ` [PATCH 20/27] workqueue: reimplement workqueue freeze using max_active Tejun Heo
2009-12-18 12:58 ` [PATCH 21/27] workqueue: introduce global cwq and unify cwq locks Tejun Heo
2009-12-18 12:58 ` [PATCH 22/27] workqueue: implement worker states Tejun Heo
2009-12-18 12:58 ` [PATCH 23/27] workqueue: reimplement CPU hotplugging support using trustee Tejun Heo
2009-12-18 12:58 ` [PATCH 24/27] workqueue: make single thread workqueue shared worker pool friendly Tejun Heo
2009-12-18 12:58 ` [PATCH 25/27] workqueue: use shared worklist and pool all workers per cpu Tejun Heo
2009-12-18 12:58 ` [PATCH 26/27] workqueue: implement concurrency managed dynamic worker pool Tejun Heo
2009-12-18 12:58 ` [PATCH 27/27] workqueue: increase max_active of keventd and kill current_is_keventd() Tejun Heo
2009-12-18 13:00 ` SUBJ: [RFC PATCHSET] concurrency managed workqueue, take#2 Tejun Heo
2009-12-18 13:03 ` Tejun Heo
2009-12-18 13:45 ` workqueue thing Peter Zijlstra
2009-12-18 13:50 ` Andi Kleen
2009-12-18 15:01 ` Arjan van de Ven
2009-12-21 3:19 ` Tejun Heo
2009-12-21 9:17 ` Jens Axboe
2009-12-21 10:35 ` Peter Zijlstra
2009-12-21 11:09 ` Andi Kleen
2009-12-21 11:17 ` Arjan van de Ven
2009-12-21 11:33 ` Andi Kleen
2009-12-21 13:18 ` Tejun Heo
2009-12-21 11:11 ` Arjan van de Ven
2009-12-21 13:22 ` Tejun Heo
2009-12-21 13:53 ` Arjan van de Ven
2009-12-21 14:19 ` Tejun Heo
2009-12-21 15:19 ` Arjan van de Ven
2009-12-22 0:00 ` Tejun Heo
2009-12-22 11:10 ` Peter Zijlstra
2009-12-22 17:20 ` Linus Torvalds
2009-12-22 17:47 ` Peter Zijlstra
2009-12-22 18:07 ` Andi Kleen
2009-12-22 18:20 ` Peter Zijlstra
2009-12-23 8:17 ` Stijn Devriendt
2009-12-23 8:43 ` Peter Zijlstra
2009-12-23 9:01 ` Stijn Devriendt
2009-12-22 18:28 ` Linus Torvalds
2009-12-23 8:06 ` Johannes Berg
2009-12-23 3:37 ` Tejun Heo
2009-12-23 6:52 ` Herbert Xu
2009-12-23 8:00 ` Steffen Klassert
2009-12-23 8:01 ` [PATCH 0/2] Parallel crypto/IPsec v7 Steffen Klassert
2009-12-23 8:03 ` [PATCH 1/2] padata: generic parallelization/serialization interface Steffen Klassert
2009-12-23 8:04 ` [PATCH 2/2] crypto: pcrypt - Add pcrypt crypto parallelization wrapper Steffen Klassert
2010-01-07 5:39 ` [PATCH 0/2] Parallel crypto/IPsec v7 Herbert Xu
2010-01-16 9:44 ` David Miller
2009-12-18 15:30 ` workqueue thing Linus Torvalds
2009-12-18 15:39 ` Ingo Molnar
2009-12-18 15:39 ` Peter Zijlstra
2009-12-18 15:47 ` Linus Torvalds
2009-12-18 15:53 ` Peter Zijlstra
2009-12-21 3:04 ` Tejun Heo
2009-12-21 9:22 ` Peter Zijlstra
2009-12-21 13:30 ` Tejun Heo
2009-12-21 14:26 ` Peter Zijlstra
2009-12-21 23:50 ` Tejun Heo
2009-12-22 11:00 ` Peter Zijlstra
2009-12-22 11:03 ` Peter Zijlstra
2009-12-23 3:43 ` Tejun Heo
2009-12-22 11:04 ` Peter Zijlstra
2009-12-23 3:48 ` Tejun Heo
2009-12-22 11:06 ` Peter Zijlstra
2009-12-23 4:18 ` Tejun Heo
2009-12-23 4:42 ` Linus Torvalds
2009-12-23 6:02 ` Ingo Molnar
2009-12-23 6:13 ` Jeff Garzik
2009-12-23 7:53 ` Ingo Molnar
2009-12-23 8:41 ` Peter Zijlstra
2009-12-23 10:25 ` Jeff Garzik
2009-12-23 13:33 ` Stefan Richter
2009-12-23 14:20 ` Mark Brown
2009-12-23 7:09 ` Tejun Heo
2009-12-23 8:01 ` Ingo Molnar
2009-12-23 8:12 ` Ingo Molnar
2009-12-23 8:32 ` Tejun Heo
2009-12-23 8:42 ` Ingo Molnar
2009-12-23 8:27 ` Tejun Heo
2009-12-23 8:37 ` Ingo Molnar
2009-12-23 8:49 ` Tejun Heo
2009-12-23 8:49 ` Ingo Molnar
2009-12-23 9:03 ` Tejun Heo
2009-12-23 13:40 ` Stefan Richter
2009-12-23 13:43 ` Stefan Richter
2009-12-23 8:25 ` Arjan van de Ven
2009-12-23 13:00 ` Stefan Richter
2009-12-23 8:31 ` Stijn Devriendt
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=1261141088-2014-16-git-send-email-tj@kernel.org \
--to=tj@kernel.org \
--cc=akpm@linux-foundation.org \
--cc=andi@firstfloor.org \
--cc=arjan@linux.intel.com \
--cc=avi@redhat.com \
--cc=awalls@radix.net \
--cc=cl@linux-foundation.org \
--cc=dhowells@redhat.com \
--cc=jeff@garzik.org \
--cc=jens.axboe@oracle.com \
--cc=johannes@sipsolutions.net \
--cc=linux-kernel@vger.kernel.org \
--cc=mingo@elte.hu \
--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