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:   Fri, 10 May 2019 09:36:12 -0700
From:   Matthew Wilcox <willy@infradead.org>
To:     "Huang, Ying" <ying.huang@intel.com>
Cc:     Yang Shi <yang.shi@linux.alibaba.com>, hannes@cmpxchg.org,
        mhocko@suse.com, mgorman@techsingularity.net,
        kirill.shutemov@linux.intel.com, hughd@google.com,
        akpm@linux-foundation.org, linux-mm@kvack.org,
        linux-kernel@vger.kernel.org
Subject: Re: [PATCH] mm: vmscan: correct nr_reclaimed for THP
Message-ID: <20190510163612.GA23417@bombadil.infradead.org>
References: <1557447392-61607-1-git-send-email-yang.shi@linux.alibaba.com>
 <87y33fjbvr.fsf@yhuang-dev.intel.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <87y33fjbvr.fsf@yhuang-dev.intel.com>
User-Agent: Mutt/1.9.2 (2017-12-15)
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On Fri, May 10, 2019 at 10:12:40AM +0800, Huang, Ying wrote:
> > +		nr_reclaimed += (1 << compound_order(page));
> 
> How about to change this to
> 
> 
>         nr_reclaimed += hpage_nr_pages(page);

Please don't.  That embeds the knowledge that we can only swap out either 
normal pages or THP sized pages.  I'm trying to make the VM capable of 
supporting arbitrary-order pages, and this would be just one more place
to fix.

I'm sympathetic to the "self documenting" argument.  My current tree has
a patch in it:

    mm: Introduce compound_nr
    
    Replace 1 << compound_order(page) with compound_nr(page).  Minor
    improvements in readability.

It goes along with this patch:

    mm: Introduce page_size()

    It's unnecessarily hard to find out the size of a potentially huge page.
    Replace 'PAGE_SIZE << compound_order(page)' with page_size(page).

Better suggestions on naming gratefully received.  I'm more happy with 
page_size() than I am with compound_nr().  page_nr() gives the wrong
impression; page_count() isn't great either.