Received-SPF: pass (google.com: domain of linux-kernel+bounces-45147-linux.lists.archive=gmail.com@vger.kernel.org designates 2604:1380:45e3:2400::1 as permitted sender) client-ip=2604:1380:45e3:2400::1;
Date: Tue, 30 Jan 2024 18:55:43 +0000
In-Reply-To: <CAKEwX=OwsEcSnw5V_zivdFzGLLVA_PC+Acv+5sozkCJU1NHpeg@mail.gmail.com>
Precedence: bulk
Mime-Version: 1.0
References: <20240129224542.162599-1-nphamcs@gmail.com> <20240129224542.162599-2-nphamcs@gmail.com>
 <ZbhKoJ5BcP_RhMt4@google.com> <CAKEwX=OwsEcSnw5V_zivdFzGLLVA_PC+Acv+5sozkCJU1NHpeg@mail.gmail.com>
Message-ID: <ZblGL3BgB3MAL57Z@google.com>
Subject: Re: [PATCH 1/3] selftests: zswap: add zswap selftest file to zswap
 maintainer entry
From: Yosry Ahmed <yosryahmed@google.com>
To: Nhat Pham <nphamcs@gmail.com>
Cc: akpm@linux-foundation.org, shuah@kernel.org, hannes@cmpxchg.org, 
	tj@kernel.org, lizefan.x@bytedance.com, linux-mm@kvack.org, 
	kernel-team@meta.com, linux-kernel@vger.kernel.org, cgroups@vger.kernel.org, 
	linux-kselftest@vger.kernel.org
Content-Type: text/plain; charset="utf-8"
Content-Transfer-Encoding: quoted-printable

On Tue, Jan 30, 2024 at 10:37:15AM -0800, Nhat Pham wrote:
> On Mon, Jan 29, 2024 at 5:02=E2=80=AFPM Yosry Ahmed <yosryahmed@google.co=
m> wrote:
> >
> > On Mon, Jan 29, 2024 at 02:45:40PM -0800, Nhat Pham wrote:
> > > Make it easier for contributors to find the zswap maintainers when th=
ey
> > > update the zswap tests.
> > >
> > > Signed-off-by: Nhat Pham <nphamcs@gmail.com>
> >
> > I guess I had to check the zswap tests at some point :)
>=20
> We sorely need more zswap tests :)
>=20
> I'm one of the offenders of adding new features without including
> tests, so no judging anyone of course, and admittedly zswap is quite
> intertwined with other parts of MM, so it's kinda hard to write
> unit-ish tests for zswap only. I often had to resort to scripting
> stress tests to iron out bugs.
>=20
> But there are still tests that we can write to verify public API
> (cgroup's zswap options come to mind), simple tests that cover crucial
> code paths, etc. that we should probably add in. At the very least
> this can be a quick/sanity check for developing and backporting
> patches into the production system.

Agreed. I am yet to take a close look at the existing tests tbh, but I
took a quick look before and there is room for improvement (and little
bandwidth to act on it).