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;
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: quoted-printable
From:   Qian Cai <cai@lca.pw>
Mime-Version: 1.0 (1.0)
Date:   Tue, 7 Jul 2020 07:43:48 -0400
Subject: Re: [mm] 4e2c82a409: ltp.overcommit_memory01.fail
Message-Id: <AF8CFC10-7655-4664-974D-3632793B0710@lca.pw>
Cc:     Feng Tang <feng.tang@intel.com>,
        kernel test robot <rong.a.chen@intel.com>,
        Andrew Morton <akpm@linux-foundation.org>,
        Johannes Weiner <hannes@cmpxchg.org>,
        Matthew Wilcox <willy@infradead.org>,
        Mel Gorman <mgorman@suse.de>,
        Kees Cook <keescook@chromium.org>,
        Luis Chamberlain <mcgrof@kernel.org>,
        Iurii Zaikin <yzaikin@google.com>, andi.kleen@intel.com,
        tim.c.chen@intel.com, dave.hansen@intel.com, ying.huang@intel.com,
        linux-mm@kvack.org, linux-kernel@vger.kernel.org, lkp@lists.01.org
To:     Michal Hocko <mhocko@kernel.org>
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk


> On Jul 7, 2020, at 6:28 AM, Michal Hocko <mhocko@kernel.org> wrote:
>=20
> Would you have any examples? Because I find this highly unlikely.
> OVERCOMMIT_NEVER only works when virtual memory is not largerly
> overcommited wrt to real memory demand. And that tends to be more of
> an exception rather than a rule. "Modern" userspace (whatever that
> means) tends to be really hungry with virtual memory which is only used
> very sparsely.
>=20
> I would argue that either somebody is running an "OVERCOMMIT_NEVER"
> friendly SW and this is a permanent setting or this is not used at all.
> At least this is my experience.
>=20
> So I strongly suspect that LTP test failure is not something we should
> really lose sleep over. It would be nice to find a way to flush existing
> batches but I would rather see a real workload that would suffer from
> this imprecision.

I hear you many times that you really don=E2=80=99t care about those use cas=
es unless you hear exactly people are using in your world.

For example, when you said LTP oom tests are totally artificial last time an=
d how less you care about if they are failing, and I could only enjoy their e=
fficiencies to find many issues like race conditions and bad error accumulat=
ion handling etc that your =E2=80=9Creal world use cases=E2=80=9D are going t=
o take ages or no way to flag them.

There are just too many valid use cases in this wild world. The difference i=
s that I admit that I don=E2=80=99t know or even aware all the use cases, an=
d I don=E2=80=99t believe you do as well.

If a patchset broke the existing behaviors that written exactly in the spec,=
 it is then someone has to prove its innocent. For example, if nobody is goi=
ng to rely on something like this now and future, and then fix the spec and e=
xplain exactly nobody should be rely upon.=