mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: Thomas Gleixner <tglx@linutronix.de>
To: LKML <linux-kernel@vger.kernel.org>
Cc: Steven Rostedt <rostedt@goodmis.org>,
	Peter Zijlstra <peterz@infradead.org>,
	Ingo Molnar <mingo@kernel.org>, Oleg Nesterov <oleg@redhat.com>
Subject: [patch V4 02/10] rtmutex: Simplify rtmutex_slowtrylock()
Date: Wed, 11 Jun 2014 18:44:04 -0000	[thread overview]
Message-ID: <20140611183853.029651737@linutronix.de> (raw)
In-Reply-To: <20140611182944.108526809@linutronix.de>

[-- Attachment #1: rtmutex-simplify-slow-trylock.patch --]
[-- Type: text/plain, Size: 2045 bytes --]

Oleg noticed that rtmutex_slowtrylock() has a pointless check for
rt_mutex_owner(lock) != current.

To avoid calling try_to_take_rtmutex() we really want to check whether
the lock has an owner at all or whether the trylock failed because the
owner is NULL, but the RT_MUTEX_HAS_WAITERS bit is set. This covers
the lock is owned by caller situation as well.

We can actually do this check lockless. trylock is taking a chance
whether we take lock->wait_lock to do the check or not.

Add comments to the function while at it.

Reported-by: Oleg Nesterov <oleg@redhat.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
---
 kernel/locking/rtmutex.c |   32 +++++++++++++++++++++-----------
 1 file changed, 21 insertions(+), 11 deletions(-)

Index: tip/kernel/locking/rtmutex.c
===================================================================
--- tip.orig/kernel/locking/rtmutex.c
+++ tip/kernel/locking/rtmutex.c
@@ -963,22 +963,32 @@ rt_mutex_slowlock(struct rt_mutex *lock,
 /*
  * Slow path try-lock function:
  */
-static inline int
-rt_mutex_slowtrylock(struct rt_mutex *lock)
+static inline int rt_mutex_slowtrylock(struct rt_mutex *lock)
 {
-	int ret = 0;
+	int ret;
 
+	/*
+	 * trylock is taking a chance. So we dont have to take
+	 * @lock->wait_lock to figure out whether @lock has a real or
+	 * if @lock owner is NULL and the RT_MUTEX_HAS_WAITERS bit is
+	 * set.
+	 */
+	if (rt_mutex_owner(lock))
+		return 0;
+
+	/*
+	 * The mutex has currently no owner. Lock the wait lock and
+	 * try to acquire the lock.
+	 */
 	raw_spin_lock(&lock->wait_lock);
 
-	if (likely(rt_mutex_owner(lock) != current)) {
+	ret = try_to_take_rt_mutex(lock, current, NULL);
 
-		ret = try_to_take_rt_mutex(lock, current, NULL);
-		/*
-		 * try_to_take_rt_mutex() sets the lock waiters
-		 * bit unconditionally. Clean this up.
-		 */
-		fixup_rt_mutex_waiters(lock);
-	}
+	/*
+	 * try_to_take_rt_mutex() sets the lock waiters bit
+	 * unconditionally. Clean this up.
+	 */
+	fixup_rt_mutex_waiters(lock);
 
 	raw_spin_unlock(&lock->wait_lock);
 



  reply	other threads:[~2014-06-11 18:48 UTC|newest]

Thread overview: 27+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2014-06-11 18:44 [patch V4 00/10] rtmutex: Code clarification and optimization Thomas Gleixner
2014-06-11 18:44 ` Thomas Gleixner [this message]
2014-06-12  3:34   ` [patch V4 02/10] rtmutex: Simplify rtmutex_slowtrylock() Lai Jiangshan
2014-06-13 15:58   ` Steven Rostedt
2014-06-11 18:44 ` [patch V4 01/10] rtmutex: Plug slow unlock race Thomas Gleixner
2014-06-13 15:41   ` Steven Rostedt
2014-06-16  8:06   ` [tip:locking/urgent] " tip-bot for Thomas Gleixner
2014-06-11 18:44 ` [patch V4 04/10] rtmutex: No need to keep task ref for lock owner check Thomas Gleixner
2014-06-13 16:30   ` Steven Rostedt
2014-06-11 18:44 ` [patch V4 03/10] rtmutex: Simplify and document try_to_take_rtmutex() Thomas Gleixner
2014-06-13 16:27   ` Steven Rostedt
2014-06-11 18:44 ` [patch V4 05/10] rtmutex: Clarify the boost/deboost part Thomas Gleixner
2014-06-13 16:35   ` Steven Rostedt
2014-06-11 18:44 ` [patch V4 06/10] rtmutex: Document pi chain walk Thomas Gleixner
2014-06-13 16:54   ` Steven Rostedt
2014-06-11 18:44 ` [patch V4 07/10] rtmutex: Simplify remove_waiter() Thomas Gleixner
2014-06-13 16:58   ` Steven Rostedt
2014-06-11 18:44 ` [patch V4 08/10] rtmutex: Confine deadlock logic to futex Thomas Gleixner
2014-06-13 17:10   ` Steven Rostedt
2014-06-11 18:44 ` [patch V4 09/10] rtmutex: Cleanup deadlock detector debug logic Thomas Gleixner
2014-06-13 17:19   ` Steven Rostedt
2014-06-13 19:43     ` Thomas Gleixner
2014-06-11 18:44 ` [patch V4 10/10] rtmutex: Avoid pointless requeueing in the deadlock detection chain walk Thomas Gleixner
2014-06-13 17:28   ` Steven Rostedt
2014-06-13 19:46     ` Thomas Gleixner
2014-06-22  8:45 ` [patch V4 00/10] rtmutex: Code clarification and optimization Ingo Molnar
2014-06-22  9:20   ` Thomas Gleixner

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=20140611183853.029651737@linutronix.de \
    --to=tglx@linutronix.de \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mingo@kernel.org \
    --cc=oleg@redhat.com \
    --cc=peterz@infradead.org \
    --cc=rostedt@goodmis.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

all inboxes | Powered by JetHome®