Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
Date:   Mon, 23 Jul 2018 13:28:49 -0700 (PDT)
From:   David Rientjes <rientjes@google.com>
To:     Matthew Wilcox <willy@infradead.org>
cc:     Yang Shi <yang.shi@linux.alibaba.com>,
        Andrew Morton <akpm@linux-foundation.org>,
        kirill@shutemov.name, hughd@google.com, aaron.lu@intel.com,
        linux-mm@kvack.org, linux-kernel@vger.kernel.org
Subject: Re: [PATCH] mm: thp: remove use_zero_page sysfs knob
In-Reply-To: <20180722035156.GA12125@bombadil.infradead.org>
Message-ID: <alpine.DEB.2.21.1807231323460.105582@chino.kir.corp.google.com>
References: <1532110430-115278-1-git-send-email-yang.shi@linux.alibaba.com> <20180720123243.6dfc95ba061cd06e05c0262e@linux-foundation.org> <alpine.DEB.2.21.1807201300290.224013@chino.kir.corp.google.com> <3238b5d2-fd89-a6be-0382-027a24a4d3ad@linux.alibaba.com>
 <alpine.DEB.2.21.1807201401390.231119@chino.kir.corp.google.com> <20180722035156.GA12125@bombadil.infradead.org>
User-Agent: Alpine 2.21 (DEB 202 2017-01-01)
MIME-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On Sat, 21 Jul 2018, Matthew Wilcox wrote:

> > The huge zero page can be reclaimed under memory pressure and, if it is, 
> > it is attempted to be allocted again with gfp flags that attempt memory 
> > compaction that can become expensive.  If we are constantly under memory 
> > pressure, it gets freed and reallocated millions of times always trying to 
> > compact memory both directly and by kicking kcompactd in the background.
> > 
> > It likely should also be per node.
> 
> Have you benchmarked making the non-huge zero page per-node?
> 

Not since we disable it :)  I will, though.  The more concerning issue for 
us, modulo CVE-2017-1000405, is the cpu cost of constantly directly 
compacting memory for allocating the hzp in real time after it has been 
reclaimed.  We've observed this happening tens or hundreds of thousands 
of times on some systems.  It will be 2MB per node on x86 if the data 
suggests we should make it NUMA aware, I don't think the cost is too high 
to leave it persistently available even under memory pressure if 
use_zero_page is enabled.