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;
Subject: Re: [PATCH] mm: fix a data race in put_page()
To:     Qian Cai <cai@lca.pw>
CC:     Marco Elver <elver@google.com>, Jan Kara <jack@suse.cz>,
        David Hildenbrand <david@redhat.com>,
        Andrew Morton <akpm@linux-foundation.org>,
        <ira.weiny@intel.com>, Dan Williams <dan.j.williams@intel.com>,
        Linux Memory Management List <linux-mm@kvack.org>,
        Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
        "Paul E. McKenney" <paulmck@kernel.org>
References: <5402183a-2372-b442-84d3-c28fb59fa7af@nvidia.com>
 <8602A57D-B420-489C-89CC-23D096014C47@lca.pw>
From:   John Hubbard <jhubbard@nvidia.com>
Message-ID: <1a179bea-fd71-7b53-34c5-895986c24931@nvidia.com>
Date:   Sat, 8 Feb 2020 23:12:48 -0800
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:68.0) Gecko/20100101
 Thunderbird/68.4.2
MIME-Version: 1.0
In-Reply-To: <8602A57D-B420-489C-89CC-23D096014C47@lca.pw>
Content-Type: text/plain; charset="utf-8"; format=flowed
Content-Language: en-US
Content-Transfer-Encoding: quoted-printable
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On 2/8/20 7:10 PM, Qian Cai wrote:
>=20
>=20
>> On Feb 8, 2020, at 8:44 PM, John Hubbard <jhubbard@nvidia.com> wrote:
>>
>> So it looks like we're probably stuck with having to annotate the code. =
Given
>> that, there is a balance between how many macros, and how much commentin=
g. For
>> example, if there is a single macro (data_race, for example), then we'll=
 need to
>> add comments for the various cases, explaining which data_race situation=
 is
>> happening.
>=20
> On the other hand, it is perfect fine of not commenting on each data_race=
() that most of times, people could run git blame to learn more details. Ac=
tually, no maintainers from various of subsystems asked for commenting so f=
ar.
>=20

Well, maybe I'm looking at this wrong. I was thinking that one should attem=
pt to
understand the code on the screen, and that's generally best--but here, may=
be
"data_race" is just something that means "tool cruft", really. So mentally =
we
would move toward visually filtering out the data_race "key word".

I really don't like it but at least there is a significant benefit from the=
 tool
that probably makes it worth the visual noise.

Blue sky thoughts for The Far Future: It would be nice if the tools got a l=
ot
better--maybe in the direction of C language extensions, even if only used =
in
this project at first.

thanks,
--=20
John Hubbard
NVIDIA

>>
>> That's still true, but to a lesser extent if more macros are added. In t=
his case,
>> I suspect that READ_BITS() makes the commenting easier and shorter. So I=
'd tentatively
>> lead towards adding it, but what do others on the list think?
>=20
> Even read bits could be dangerous from data races and confusing at best, =
so I am not really sure what the value of introducing this new macro. Peopl=
e who like to understand it correctly still need to read the commit logs.
>=20
> This flags->zonenum is such a special case that I don=E2=80=99t really se=
e it regularly for the last few weeks digging KCSAN reports, so even if it =
is worth adding READ_BITS(), there are more equally important macros need t=
o be added together to be useful initially. For example, HARMLESS_COUNTERS(=
), READ_SINGLE_BIT(), READ_IMMUTATABLE_BITS() etc which Linus said exactly =
wanted to avoid.
>=20