linux-fsdevel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
* [PATCH v2] wait: use wrapper functions
@ 2010-05-07  6:33 Changli Gao
  2010-05-07 10:12 ` Peter Zijlstra
       [not found] ` <1273214006-2979-1-git-send-email-xiaosuo-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>
  0 siblings, 2 replies; 4+ messages in thread
From: Changli Gao @ 2010-05-07  6:33 UTC (permalink / raw)
  To: Andrew Morton
  Cc: Alexander Viro, Paul Menage, Li Zefan, Ingo Molnar,
	Davide Libenzi, Peter Zijlstra, linux-fsdevel, linux-kernel,
	containers, Changli Gao

use wrapper functions

epoll should not touch flags in wait_queue_t. This patch introduces a new
function __add_wait_queue_excl(), for the users, who use wait queue as a
LIFO queue.

__add_wait_queue_tail_excl() is introduced too instead of
add_wait_queue_exclusive_locked(). remove_wait_queue_locked() is removed, as
it is a duplicate of __remove_wait_queue(), disliked by users, and with less
users.

Signed-off-by: Changli Gao <xiaosuo@gmail.com>
----
 fs/eventpoll.c       |    3 +--
 include/linux/wait.h |   35 +++++++++++++++--------------------
 kernel/cgroup.c      |    2 +-
 kernel/sched.c       |    3 +--
 4 files changed, 18 insertions(+), 25 deletions(-)
diff --git a/fs/eventpoll.c b/fs/eventpoll.c
index bd056a5..a6ba470 100644
--- a/fs/eventpoll.c
+++ b/fs/eventpoll.c
@@ -1140,8 +1140,7 @@ retry:
 		 * ep_poll_callback() when events will become available.
 		 */
 		init_waitqueue_entry(&wait, current);
-		wait.flags |= WQ_FLAG_EXCLUSIVE;
-		__add_wait_queue(&ep->wq, &wait);
+		__add_wait_queue_excl(&ep->wq, &wait);
 
 		for (;;) {
 			/*
diff --git a/include/linux/wait.h b/include/linux/wait.h
index a48e16b..4db4147 100644
--- a/include/linux/wait.h
+++ b/include/linux/wait.h
@@ -127,12 +127,26 @@ static inline void __add_wait_queue(wait_queue_head_t *head, wait_queue_t *new)
 /*
  * Used for wake-one threads:
  */
+static inline void __add_wait_queue_excl(wait_queue_head_t *q,
+					      wait_queue_t *wait)
+{
+	wait->flags |= WQ_FLAG_EXCLUSIVE;
+	__add_wait_queue(q, wait);
+}
+
 static inline void __add_wait_queue_tail(wait_queue_head_t *head,
-						wait_queue_t *new)
+					 wait_queue_t *new)
 {
 	list_add_tail(&new->task_list, &head->task_list);
 }
 
+static inline void __add_wait_queue_tail_excl(wait_queue_head_t *q,
+					      wait_queue_t *wait)
+{
+	wait->flags |= WQ_FLAG_EXCLUSIVE;
+	__add_wait_queue_tail(q, wait);
+}
+
 static inline void __remove_wait_queue(wait_queue_head_t *head,
 							wait_queue_t *old)
 {
@@ -404,25 +418,6 @@ do {									\
 })
 
 /*
- * Must be called with the spinlock in the wait_queue_head_t held.
- */
-static inline void add_wait_queue_exclusive_locked(wait_queue_head_t *q,
-						   wait_queue_t * wait)
-{
-	wait->flags |= WQ_FLAG_EXCLUSIVE;
-	__add_wait_queue_tail(q,  wait);
-}
-
-/*
- * Must be called with the spinlock in the wait_queue_head_t held.
- */
-static inline void remove_wait_queue_locked(wait_queue_head_t *q,
-					    wait_queue_t * wait)
-{
-	__remove_wait_queue(q,  wait);
-}
-
-/*
  * These are the old interfaces to sleep waiting for an event.
  * They are racy.  DO NOT use them, use the wait_event* interfaces above.
  * We plan to remove these interfaces.
diff --git a/kernel/cgroup.c b/kernel/cgroup.c
index e2769e1..4a07d05 100644
--- a/kernel/cgroup.c
+++ b/kernel/cgroup.c
@@ -3010,7 +3010,7 @@ static int cgroup_event_wake(wait_queue_t *wait, unsigned mode,
 	unsigned long flags = (unsigned long)key;
 
 	if (flags & POLLHUP) {
-		remove_wait_queue_locked(event->wqh, &event->wait);
+		__remove_wait_queue(event->wqh, &event->wait);
 		spin_lock(&cgrp->event_list_lock);
 		list_del(&event->list);
 		spin_unlock(&cgrp->event_list_lock);
diff --git a/kernel/sched.c b/kernel/sched.c
index 6af210a..c5c998d 100644
--- a/kernel/sched.c
+++ b/kernel/sched.c
@@ -4039,8 +4039,7 @@ do_wait_for_common(struct completion *x, long timeout, int state)
 	if (!x->done) {
 		DECLARE_WAITQUEUE(wait, current);
 
-		wait.flags |= WQ_FLAG_EXCLUSIVE;
-		__add_wait_queue_tail(&x->wait, &wait);
+		__add_wait_queue_tail_excl(&x->wait, &wait);
 		do {
 			if (signal_pending_state(state, current)) {
 				timeout = -ERESTARTSYS;

^ permalink raw reply related	[flat|nested] 4+ messages in thread

* Re: [PATCH v2] wait: use wrapper functions
  2010-05-07  6:33 [PATCH v2] wait: use wrapper functions Changli Gao
@ 2010-05-07 10:12 ` Peter Zijlstra
       [not found] ` <1273214006-2979-1-git-send-email-xiaosuo-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>
  1 sibling, 0 replies; 4+ messages in thread
From: Peter Zijlstra @ 2010-05-07 10:12 UTC (permalink / raw)
  To: Changli Gao
  Cc: Andrew Morton, Alexander Viro, Paul Menage, Li Zefan, Ingo Molnar,
	Davide Libenzi, linux-fsdevel, linux-kernel, containers

On Fri, 2010-05-07 at 14:33 +0800, Changli Gao wrote:
> use wrapper functions
> 
> epoll should not touch flags in wait_queue_t. This patch introduces a new
> function __add_wait_queue_excl(), for the users, who use wait queue as a
> LIFO queue.
> 
> __add_wait_queue_tail_excl() is introduced too instead of
> add_wait_queue_exclusive_locked(). remove_wait_queue_locked() is removed, as
> it is a duplicate of __remove_wait_queue(), disliked by users, and with less
> users.
> 
> Signed-off-by: Changli Gao <xiaosuo@gmail.com>

Looks good to me. I can feed it to Ingo if he doesn't pick it up
himself.

^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: [PATCH v2] wait: use wrapper functions
       [not found] ` <1273214006-2979-1-git-send-email-xiaosuo-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>
@ 2010-05-07 20:52   ` Andrew Morton
  2010-05-07 20:56     ` Peter Zijlstra
  0 siblings, 1 reply; 4+ messages in thread
From: Andrew Morton @ 2010-05-07 20:52 UTC (permalink / raw)
  To: Changli Gao
  Cc: containers-cunTk1MwBs9QetFLy7KEm3xJsTq8ys+cHZ5vskTnxNA,
	linux-kernel-u79uwXL29TY76Z2rM5mHXA, Alexander Viro,
	Davide Libenzi, linux-fsdevel-u79uwXL29TY76Z2rM5mHXA, Paul Menage,
	Ingo Molnar

On Fri,  7 May 2010 14:33:26 +0800
Changli Gao <xiaosuo-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org> wrote:

> epoll should not touch flags in wait_queue_t. This patch introduces a new
> function __add_wait_queue_excl(), for the users, who use wait queue as a
> LIFO queue.

__add_wait_queue_exclusive(), please.  Let's be consistent.

^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: [PATCH v2] wait: use wrapper functions
  2010-05-07 20:52   ` Andrew Morton
@ 2010-05-07 20:56     ` Peter Zijlstra
  0 siblings, 0 replies; 4+ messages in thread
From: Peter Zijlstra @ 2010-05-07 20:56 UTC (permalink / raw)
  To: Andrew Morton
  Cc: Changli Gao, Alexander Viro, Paul Menage, Li Zefan, Ingo Molnar,
	Davide Libenzi, linux-fsdevel, linux-kernel, containers

On Fri, 2010-05-07 at 13:52 -0700, Andrew Morton wrote:
> On Fri,  7 May 2010 14:33:26 +0800
> Changli Gao <xiaosuo@gmail.com> wrote:
> 
> > epoll should not touch flags in wait_queue_t. This patch introduces a new
> > function __add_wait_queue_excl(), for the users, who use wait queue as a
> > LIFO queue.
> 
> __add_wait_queue_exclusive(), please.  Let's be consistent.

done

^ permalink raw reply	[flat|nested] 4+ messages in thread

end of thread, other threads:[~2010-05-07 20:56 UTC | newest]

Thread overview: 4+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2010-05-07  6:33 [PATCH v2] wait: use wrapper functions Changli Gao
2010-05-07 10:12 ` Peter Zijlstra
     [not found] ` <1273214006-2979-1-git-send-email-xiaosuo-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>
2010-05-07 20:52   ` Andrew Morton
2010-05-07 20:56     ` Peter Zijlstra

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).