Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.32 as permitted sender) client-ip=23.128.96.32;
Date:   Thu, 23 Nov 2023 12:07:40 +0700
From:   Bagas Sanjaya <bagasdotme@gmail.com>
To:     Chun Ng <chunn@nvidia.com>,
        Linux Kernel Mailing List <linux-kernel@vger.kernel.org>
Cc:     Linux Regressions <regressions@lists.linux.dev>,
        Andrew Morton <akpm@linux-foundation.org>,
        Linux Memory Management List <linux-mm@kvack.org>,
        "Liam R. Howlett" <Liam.Howlett@oracle.com>,
        Ankita Garg <ankitag@nvidia.com>
Subject: Re: [REGRESSION]: mmap performance regression starting with k-6.1
Message-ID: <ZV7eHE2Fxb75oRpG@archie.me>
References: <PH7PR12MB7937B0DF19E7E8539703D0E3D6BAA@PH7PR12MB7937.namprd12.prod.outlook.com>
MIME-Version: 1.0
Content-Type: multipart/signed; micalg=pgp-sha512;
        protocol="application/pgp-signature"; boundary="KIdW2yF7ZM5mmI7P"
Content-Disposition: inline
In-Reply-To: <PH7PR12MB7937B0DF19E7E8539703D0E3D6BAA@PH7PR12MB7937.namprd12.prod.outlook.com>
Precedence: bulk


--KIdW2yF7ZM5mmI7P
Content-Type: text/plain; charset=utf-8
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable

On Wed, Nov 22, 2023 at 08:03:19PM +0000, Chun Ng wrote:
> Hi,
>=20
> Recently I observed there is performance regression on system call mmap(.=
=2E). I tried both vanilla kernels and Raspberry Pi kernels on a Raspberry =
Pi 4 box and the results are pretty consistent among them.
>=20
> Bisection showed that the regression starts from k-6.1, and the latest va=
nilla k-6.7 is still showing the same regression.
>=20
> The test program calls mmap/munmap for a 4K page with MAP_ANON and MAP_PR=
IVATE flags, and ftrace is used to measure the time spent on the do_mmap(..=
) call.=C2=A0 Measured time of a sample run with different vanilla kernel v=
ersions are:
> k-5.10 and k-6.0: ~157us
> k-6.1: ~194us
> k-6.7: ~214us
> Results are pretty consistent across multiple runs with a small percentag=
e variance.=C2=A0 Ftrace shows that latency of mmap_region(...) has increas=
ed since k-6.1.=C2=A0=C2=A0An application that makes frequent mmap(..) call=
s the accumulated extra latency is very noticeable.=20
>=20
> Please find the ftrace results and kernel config files in this folder:
> https://drive.google.com/drive/folders/1qy8YTBqxu8Gdbs7IigYbSd4FXldId5sd?=
usp=3Ddrive_link
>=20
> The test program can be found in here:
> https://drive.google.com/file/d/1tG6_BbQMCHwfKebvAIAg_xqbM_lpPcuM/view?us=
p=3Dsharing
>=20
> Info on the testing environment:
> cpufreq_governor: performance
> Test machine: Raspberry Pi 4, 8GB DDR
> SCHED_FIFO with priority 99 for running the test program
>=20
> Vanilla kernels are not tainted. However on k-6.0 and k-6.7, I have to pa=
tch the drivers/clk/bcm/clk-raspberrypi.c file with the version in Raspberr=
y Pi kernel tree for the CPU frequency governor to work.
>=20

The next step is to find the commit that introduces your regression with
`git bisect`. If you haven't done so, see
Documentation/admin-guide/bug-bisect.rst for instructions.

Anyway, I'm adding this regression to regzbot:

#regzbot ^introduced: v6.0..v6.1

Thanks.

--=20
An old man doll... just what I always wanted! - Clara

--KIdW2yF7ZM5mmI7P
Content-Type: application/pgp-signature; name="signature.asc"

-----BEGIN PGP SIGNATURE-----

iHUEABYKAB0WIQSSYQ6Cy7oyFNCHrUH2uYlJVVFOowUCZV7eGAAKCRD2uYlJVVFO
o72FAQCK5rdafR/mKJWt3KzyR6K08xDUXWyyIuqjKnddfzjV1gD/RNJ2SAAIHKWN
re/3qbV0011KdIQjaBsiuk0g/iTkaAA=
=/6zc
-----END PGP SIGNATURE-----

--KIdW2yF7ZM5mmI7P--