Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
IronPort-SDR: otUwOAszHYWRxGY+bfkDb+r+kJnozbTdyqLaFkYXt5+pGT8E4ysiYJTl1VhELzH7Z4UcwzBohT
 i6oPGt2qlYcQ==
IronPort-SDR: on8+Bxc11wN/xt986f7vAgSqLFvRKTLK70zFKK8TkcK1dLMvz+SQfPzAnK97FPsY7Z2ZFHm+Gz
 7AewwwY1uGYw==
From:   "Huang\, Ying" <ying.huang@intel.com>
To:     Linus Torvalds <torvalds@linux-foundation.org>
Cc:     huang ying <huang.ying.caritas@gmail.com>,
        Matthew Wilcox <willy@infradead.org>,
        Andrew Morton <akpm@linux-foundation.org>,
        Linux-MM <linux-mm@kvack.org>,
        Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
        Peter Xu <peterx@redhat.com>, Hugh Dickins <hughd@google.com>,
        Johannes Weiner <hannes@cmpxchg.org>,
        Mel Gorman <mgorman@suse.de>, Rik van Riel <riel@surriel.com>,
        Andrea Arcangeli <aarcange@redhat.com>,
        Michal Hocko <mhocko@kernel.org>,
        Dave Hansen <dave.hansen@intel.com>,
        Tim Chen <tim.c.chen@intel.com>
Subject: Re: [PATCH] mm: Free unused swap cache page in write protection fault handler
References: <20210113024241.179113-1-ying.huang@intel.com>
        <CAHk-=whn5kVxEitkC0AyzvWRyxbF91rMrO9ZG6JHBNYLckpDDw@mail.gmail.com>
        <CAC=cRTNpoqHpbtZM1uDvVQYQZVyRbfPh+Dirb=-Xgoh22u_W1A@mail.gmail.com>
        <20210113031142.GM35215@casper.infradead.org>
        <CAC=cRTNMN5Pmz2PC8p3N2-uJkqLUCW95wr=VWTZmodoECmj6nQ@mail.gmail.com>
        <CAHk-=wh3nL28QfkKV6gpVGjh5iPjoEZtEQYPDvwPgscm-yMAfA@mail.gmail.com>
Date:   Fri, 15 Jan 2021 16:47:37 +0800
In-Reply-To: <CAHk-=wh3nL28QfkKV6gpVGjh5iPjoEZtEQYPDvwPgscm-yMAfA@mail.gmail.com>
        (Linus Torvalds's message of "Wed, 13 Jan 2021 13:09:31 -0800")
Message-ID: <87sg72vb6e.fsf@yhuang-dev.intel.com>
User-Agent: Gnus/5.13 (Gnus v5.13) Emacs/26.3 (gnu/linux)
MIME-Version: 1.0
Content-Type: text/plain; charset=ascii
Precedence: bulk

Linus Torvalds <torvalds@linux-foundation.org> writes:

> On Tue, Jan 12, 2021 at 9:24 PM huang ying <huang.ying.caritas@gmail.com> wrote:
>> >
>> > Couldn't we just move it to the tail of the LRU list so it's reclaimed
>> > first?  Or is locking going to be a problem here?
>>
>> Yes.  That's a way to reduce the disturbance to the page reclaiming.
>> For LRU lock contention, is it sufficient to use another pagevec?
>
> I wonder if this is really worth it. I'd like to see numbers.
>
> Because in probably 99%+ of all cases, that LRU dance is only going to
> hurt and add extra locking overhead and dirty caches.
>
> So I'd like to see some numbers that it actually helps measurably in
> whatever paging-heavy case...

OK.  I will start from a simpler version and only use a pagevec if
there's measurable difference.

Best Regards,
Huang, Ying