Message-ID: <1434861344.6341.14.camel@gmail.com>
Subject: Re: [PATCH v2] futex: lower the lock contention on the HB lock
 during wake up
From: Mike Galbraith <umgwanakikbuti@gmail.com>
To: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Cc: Davidlohr Bueso <dave@stgolabs.net>, Thomas Gleixner <tglx@linutronix.de>,
        Peter Zijlstra <peterz@infradead.org>, Ingo Molnar <mingo@redhat.com>,
        Steven Rostedt <rostedt@goodmis.org>,
        "Paul E. McKenney" <paulmck@linux.vnet.ibm.com>,
        linux-kernel@vger.kernel.org
Date: Sun, 21 Jun 2015 06:35:44 +0200
In-Reply-To: <558183FB.7060808@linutronix.de>
References: <1432056298-18738-1-git-send-email-dave@stgolabs.net>
	 <1432056298-18738-2-git-send-email-dave@stgolabs.net>
	 <20150616192911.GA19500@linutronix.de>
	 <1434484226.1903.19.camel@stgolabs.net>
	 <20150617083350.GA2433@linutronix.de> <1434550624.23509.18.camel@gmail.com>
	 <558183FB.7060808@linutronix.de>
Content-Type: text/plain; charset="UTF-8"
Mime-Version: 1.0
Content-Transfer-Encoding: 7bit
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 1960
Lines: 40

On Wed, 2015-06-17 at 16:28 +0200, Sebastian Andrzej Siewior wrote:
> On 06/17/2015 04:17 PM, Mike Galbraith wrote:
> > On Wed, 2015-06-17 at 10:33 +0200, Sebastian Andrzej Siewior wrote:
> >> wake_futex_pi() wakes the task before releasing the hash bucket lock
> >> (HB). The first thing the woken up task usually does is to acquire the
> >> lock which requires the HB lock. On SMP Systems this leads to blocking
> >> on the HB lock which is released by the owner shortly after.
> >> This patch rearranges the unlock path by first releasing the HB lock and
> >> then waking up the task.
> >>
> >> [bigeasy: redo ontop of lockless wake-queues]
> >> Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
> >> Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
> > 
> > 4.1-rc8-rt4 contains this via 4.0-rt4, and seems fine on my 64 core
> > DL980.  I ran a few iterations of futextests and stockfish, then mixed
> > two loops of futextest at different rt prios, with stockfish also rt,
> > and ltplight as tossed in as... crack filler.  Box is still doing that,
> > is way too busy, but not griping about it.  
> 
> There are two patches mostly doing the same thing. The patch posted
> here is a redo ontop of "lockless wake-queues". It does hb-unlock,
> wakeup, de-boost. The patch merged into -RT is the original approach
> not using "lockless wake-queues" and performing wakeup, hb-unlock,
> de-boost.
> 
> I plan to get into -RT the final solution once it hits upstream.

I plugged patch1 and tip version into rt and beat it, seems solid.

Converting the rest of rtmutex.c to use wake queues with ->save_state to
select wake function went less well.  Kernel does a good impersonation
of a working kernel until I beat it up, then it loses wakeups.  Hohum,
so much for yet another early morning tinker session.

	-Mike

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
Please read the FAQ at  http://www.tux.org/lkml/