linux-kernel - Re: [V2 PATCH] kthread_work: Make lockdep happy

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:	Tue, 21 Dec 2010 19:39:53 -0500
From:	Andy Walls <awalls@...metrocast.net>
To:	Yong Zhang <yong.zhang0@...il.com>
Cc:	Tejun Heo <tj@...nel.org>, linux-kernel@...r.kernel.org,
	nicolas.mailhot@...oste.net, Jarod Wilson <jarod@...hat.com>,
	Ingo Molnar <mingo@...hat.com>,
	Mauro Carvalho Chehab <mchehab@...hat.com>,
	Hans Verkuil <hverkuil@...all.nl>,
	Andrew Morton <akpm@...ux-foundation.org>
Subject: Re: [V2 PATCH] kthread_work: Make lockdep happy

On Tue, 2010-12-21 at 12:40 +0800, Yong Zhang wrote:
> From: Yong Zhang <yong.zhang0@...il.com>
> Subject: [V2 PATCH] kthread_work: Make lockdep happy
> 
> spinlock in kthread_worker and wait_queue_head in kthread_work
> both should be lockdep sensible.
> So change the interface to make it suiltable for CONFIG_LOCKDEP.
> 
> Reported-by: Nicolas <nicolas.mailhot@...oste.net>
> Signed-off-by: Yong Zhang <yong.zhang0@...il.com>
> Cc: Tejun Heo <tj@...nel.org>
> Cc: Andrew Morton <akpm@...ux-foundation.org>
> Cc: Andy Walls <awalls@...metrocast.net>
> ---
> Changes from V1:
>  *According to Tejun, kthread_worker could be defined on stack,
>   So introduce DEFINE_KTHREAD_WORKER_ONSTACK.
>  *Change wrong setting to kthread_work->task. Thanks Adny for
>   pointing it.
>  *including some minor issue.
> 
> BTW, only passed build.
> 
>  include/linux/kthread.h |   45 +++++++++++++++++++++++++++++++++++----------
>  kernel/kthread.c        |    9 +++++++++
>  2 files changed, 44 insertions(+), 10 deletions(-)

Yong,

I have tested your patch with the following patch on top of yours.  I
find the two patches together acceptable in testing in a machine with
both a PVR-350 card and a PVR-500 card installed.

Tested-by: Andy Walls <awalls@...metrocast.net>
Signed-off-by: Andy Walls <awalls@...metrocast.net>

Could you please author a [V3 PATCH] adding my changes?  Since the
majority of the change is your work, it should be attributed to you as
the author.


About my additional changes:

I needed the explicit EXPORT_SYMBOL_GPL(__init_kthread_worker),
otherwise the build would fail, because the ivtv module had an
unresolved symbol.

I added the lockdep class name to make it easier to identify the ivtv
module's kthread worker's lock class.  Now one can actually recognize
the lock class in lockdep output strings:

	# cat lockdep | grep itv
	ffffffffa0483a60 FD:    1 BD:    1 ......: (&itv->irq_worker)->lock

Thanks again.

Regards,
Andy


diff --git a/include/linux/kthread.h b/include/linux/kthread.h
index f8b3320..994564a 100644
--- a/include/linux/kthread.h
+++ b/include/linux/kthread.h
@@ -100,13 +100,15 @@ struct kthread_work {
 # define DEFINE_KTHREAD_WORK_ONSTACK(work, fn) DEFINE_KTHREAD_WORK(work, fn)
 #endif
 
-extern void __init_kthread_worker(struct kthread_worker *worker,
-					struct lock_class_key *key);
+void __init_kthread_worker(struct kthread_worker *worker,
+			   struct lock_class_key *key,
+			   const char *lock_class_name);
 
 #define init_kthread_worker(worker)					\
 	do {								\
 		static struct lock_class_key __key;			\
-		__init_kthread_worker((worker), &__key);		\
+		__init_kthread_worker((worker), &__key,			\
+				      "(" #worker ")->lock");		\
 	} while (0)
 
 #define init_kthread_work(work, fn)					\
diff --git a/kernel/kthread.c b/kernel/kthread.c
index f760587..4657ebd 100644
--- a/kernel/kthread.c
+++ b/kernel/kthread.c
@@ -266,13 +266,15 @@ int kthreadd(void *unused)
 }
 
 void __init_kthread_worker(struct kthread_worker *worker,
-				struct lock_class_key *key)
+			   struct lock_class_key *key,
+			   const char *lock_class_name)
 {
 	spin_lock_init(&worker->lock);
-	lockdep_set_class(&worker->lock, key);
+	lockdep_set_class_and_name(&worker->lock, key, lock_class_name);
 	INIT_LIST_HEAD(&worker->work_list);
 	worker->task = NULL;
 }
+EXPORT_SYMBOL_GPL(__init_kthread_worker);
 
 /**
  * kthread_worker_fn - kthread function to process kthread_worker


--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/