From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <linux-kernel-owner@vger.kernel.org>
Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand
	id S1756175AbbCFUAJ (ORCPT <rfc822;w@1wt.eu>);
	Fri, 6 Mar 2015 15:00:09 -0500
Received: from cantor2.suse.de ([195.135.220.15]:34492 "EHLO mx2.suse.de"
	rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP
	id S1754233AbbCFUAG (ORCPT <rfc822;linux-kernel@vger.kernel.org>);
	Fri, 6 Mar 2015 15:00:06 -0500
Message-ID: <1425672000.19505.57.camel@stgolabs.net>
Subject: Re: sched: softlockups in multi_cpu_stop
From: Davidlohr Bueso <dave@stgolabs.net>
To: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Jason Low <jason.low2@hp.com>, Ingo Molnar <mingo@kernel.org>,
        Sasha Levin <sasha.levin@oracle.com>,
        Peter Zijlstra <peterz@infradead.org>,
        LKML <linux-kernel@vger.kernel.org>,
        Dave Jones <davej@codemonkey.org.uk>
Date: Fri, 06 Mar 2015 12:00:00 -0800
In-Reply-To: <1425671702.19505.54.camel@stgolabs.net>
References: <54F41516.6060608@oracle.com> <54F98F1F.3080107@oracle.com>
	 <20150306123233.GA9972@gmail.com> <1425662342.19505.41.camel@stgolabs.net>
	 <1425668223.2475.94.camel@j-VirtualBox>
	 <CA+55aFwZufdV5Q7Wm+b2F8KurtgXsJ_eNe9b6_TOSUhuW_GfSg@mail.gmail.com>
	 <1425669643.19505.46.camel@stgolabs.net>
	 <CA+55aFzuaUyAtbCAHrUZ1Prew1Dn4DJquH1LtCX_6A5fUK4Mqw@mail.gmail.com>
	 <1425671702.19505.54.camel@stgolabs.net>
Content-Type: text/plain; charset="UTF-8"
X-Mailer: Evolution 3.12.9 
Mime-Version: 1.0
Content-Transfer-Encoding: 7bit
Sender: linux-kernel-owner@vger.kernel.org
List-ID: <linux-kernel.vger.kernel.org>
X-Mailing-List: linux-kernel@vger.kernel.org

On Fri, 2015-03-06 at 11:55 -0800, Davidlohr Bueso wrote:
> On Fri, 2015-03-06 at 11:32 -0800, Linus Torvalds wrote:
> 
> > IOW, I wonder if we could special-case the common non-IO
> > fault-handling path something along the lines of:
> > 
> >  - look up the vma in the vma lookup cache
> 
> But you'd still need mmap_sem there to at least get the VMA's first
> value.

Incomplete msg... what I'm getting at is that one way or another
mmap_sem contention is easy to hit -- yes, avoiding it in IO paths is
obviously good for hold times. But I also think that it protects way too
much, ie making pthread intense workloads really suck as well.