Date: Wed, 29 Mar 2017 08:31:33 -0700
From: Davidlohr Bueso <dave@stgolabs.net>
To: Laurent Dufour <ldufour@linux.vnet.ibm.com>
Cc: "Kirill A. Shutemov" <kirill@shutemov.name>, mingo@kernel.org,
        peterz@infradead.org, akpm@linux-foundation.org, jack@suse.cz,
        kirill.shutemov@linux.intel.com, mhocko@suse.com,
        mgorman@techsingularity.net, linux-kernel@vger.kernel.org,
        Davidlohr Bueso <dbueso@suse.de>
Subject: Re: [PATCH 1/5] locking: Introduce range reader/writer lock
Message-ID: <20170329153133.GG27446@linux-80c1.suse>
References: <1488863010-13028-1-git-send-email-dave@stgolabs.net>
 <1488863010-13028-2-git-send-email-dave@stgolabs.net>
 <2f7628f4-58e1-22c4-ccbe-3106c15cb405@linux.vnet.ibm.com>
 <20170328163918.GA27446@linux-80c1.suse>
 <20170328165803.podjvgo5zim44gip@node.shutemov.name>
 <5df9746e-a671-ed3f-89a1-fb6fe6ec4f4e@linux.vnet.ibm.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii; format=flowed
Content-Disposition: inline
In-Reply-To: <5df9746e-a671-ed3f-89a1-fb6fe6ec4f4e@linux.vnet.ibm.com>
User-Agent: Mutt/1.5.24 (2015-08-30)
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 1122
Lines: 23

On Wed, 29 Mar 2017, Laurent Dufour wrote:

>On 28/03/2017 18:58, Kirill A. Shutemov wrote:
>> On Tue, Mar 28, 2017 at 09:39:18AM -0700, Davidlohr Bueso wrote:
>>> I'll wait to see if there are any more concerns and send a v2 with your corrections.
>>
>> Have you tried drop-in replacement of mmap_sem with full range lock?
>> It would be interesting to see performance implication for this.
>>
>
>I've a patch that replace the mmap_sem with a full range lock, it seems
>to work fine for x86 and ppc64 for now. I'll send it soon.
>But I didn't yet check for performance. What is the best way to that ?

I expect performance to take a measurable hit if we simply use full range
lock as a drop in replacement. My rwsem vs range lock measurements were
done with this in mind. We only win with range locks when improving the
level of parallelism. There are plenty of benchmarks out there that
stress address space concurrency. I've been using some from the mosbench
suite (metis, psearchy, dedup); Ingo also has a microbench that does
mmap()/munmap()/mremap() combos:

http://people.redhat.com/mingo/threaded-mmap-stresstest/