linux-kernel - Re: [PATCH v2 3/5] mm: use a dedicated workqueue for the free workers

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <20170322084103.GC2360@aaronlu.sh.intel.com>
Date:   Wed, 22 Mar 2017 16:41:04 +0800
From:   Aaron Lu <aaron.lu@...el.com>
To:     Minchan Kim <minchan@...nel.org>
Cc:     linux-mm@...ck.org, linux-kernel@...r.kernel.org,
        Dave Hansen <dave.hansen@...el.com>,
        Tim Chen <tim.c.chen@...el.com>,
        Andrew Morton <akpm@...ux-foundation.org>,
        Ying Huang <ying.huang@...el.com>
Subject: Re: [PATCH v2 3/5] mm: use a dedicated workqueue for the free workers

On Wed, Mar 22, 2017 at 03:33:35PM +0900, Minchan Kim wrote:
> Hi,
> 
> On Wed, Mar 15, 2017 at 05:00:02PM +0800, Aaron Lu wrote:
> > Introduce a workqueue for all the free workers so that user can fine
> > tune how many workers can be active through sysfs interface: max_active.
> > More workers will normally lead to better performance, but too many can
> > cause severe lock contention.
> 
> Let me ask a question.
> 
> How well can workqueue distribute the jobs in multiple CPU?

I would say it's good enough for my needs.
After all, it doesn't need many kworkers to achieve the 50% time
decrease: 2-4 kworkers for EP and 4-8 kworkers for EX are enough from
previous attched data.

> I don't ask about currency but parallelism.
> I guess benefit you are seeing comes from the parallelism and
> for your goal, unbound wq should spawn a thread per cpu and
> doing the work in every each CPU. does it work?

I don't think a unbound workqueue will spawn a thread per CPU, that
seems too much a cost to have a unbound workqueue.

My understanding of the unbound workqueue is that it will create a
thread pool for each node, versus each CPU as in the bound workqueue
case, and use threads from the thread pool(create threads if not enough)
to do the work.

I guess you want to ask if the unbound workqueue can spawn enough
threads to do the job? From the output of 'vmstat 1' during the free()
test, I can see some 70+ processes in runnable state when I didn't
set an upper limit for max_active of the workqueue.

Thanks,
Aaron