Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Date:   Wed, 31 May 2023 11:31:26 +0200
From:   Lukasz Majewski <lukma@denx.de>
To:     Oleksij Rempel <o.rempel@pengutronix.de>
Cc:     Andrew Lunn <andrew@lunn.ch>, Russell King <linux@armlinux.org.uk>,
        Vivien Didelot <vivien.didelot@gmail.com>,
        Florian Fainelli <f.fainelli@gmail.com>,
        Vladimir Oltean <olteanv@gmail.com>,
        "David S. Miller" <davem@davemloft.net>,
        Jakub Kicinski <kuba@kernel.org>, netdev@vger.kernel.org,
        linux-kernel@vger.kernel.org
Subject: Re: [RFC] net: dsa: slave: Advertise correct EEE capabilities at
 slave PHY setup
Message-ID: <20230531113121.4dd392d9@wsk>
In-Reply-To: <ZHcBZ5hGTu7aBCsJ@pengutronix.de>
References: <20230530122621.2142192-1-lukma@denx.de>
        <32aa2c0f-e284-4c5e-ba13-a2ea7783c202@lunn.ch>
        <20230530154039.4552e08a@wsk>
        <ZHcBZ5hGTu7aBCsJ@pengutronix.de>
Organization: denx.de
MIME-Version: 1.0
Content-Type: multipart/signed; boundary="Sig_/p0ndmoiOlqD_T5Ryyql7.kd";
 protocol="application/pgp-signature"; micalg=pgp-sha512
Precedence: bulk

--Sig_/p0ndmoiOlqD_T5Ryyql7.kd
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: quoted-printable

Hi Oleksij,

> Hi Lukasz,
>=20
> On Tue, May 30, 2023 at 03:40:39PM +0200, Lukasz Majewski wrote:
> > Hi Andrew,
> >  =20
> > > On Tue, May 30, 2023 at 02:26:21PM +0200, Lukasz Majewski wrote: =20
> > > > One can disable in device tree advertising of EEE capabilities
> > > > of PHY when 'eee-broken-100tx' property is present in DTS.
> > > >=20
> > > > With DSA switch it also may happen that one would need to
> > > > disable EEE due to some network issues.   =20
> > >=20
> > > Is EEE actually broken in the MAC/PHY combination?
> > >  =20
> >=20
> > Problem is that when I connect on this project some non-manageable
> > switches (which by default have EEE enabled), then I observe very
> > rare and sporadic link loss and reconnection. =20
>=20
> The interesting question is, do other link partner or local system is
> broken?

I was able to reproduce customer issue on my local setup with this
switch.

Moreover, only two devices were connected to this switch - mv88e6071
and the host PC.

> In some cases, not proper tx-timer was triggering this kind of
> symptoms. And timer configuration may depend on the link speed. So,
> driver may be need to take care of this.

The speed is only 100 Mbps, full duplex (as this switch only supports
speed up till 100Mbps).=20

>=20
> > Disabling EEE solves the problem.
> >  =20
> > > You should not be using this DT option for configuration. It is
> > > there because there is some hardware which is truly broken, and
> > > needs EEE turned off. =20
> >=20
> > Yes, I do think that the above sentence sums up my use case. =20
>=20
> As Andrew already described, current linux kernel EEE support is not
> in the best shape, it is hard to see the difference between broken HW
> and SW.

Ok.

>=20
> > > If EEE does work, but you need to turn it off because of latency
> > > etc, then please use ethtool.
> > >  =20
> >=20
> > Yes, correct - it is possible to disable the EEE with=20
> >=20
> > ethtool --set-eee lan2 eee off
> >=20
> > However, as I've stated in the mail, I cannot re-enable EEE once
> > disabled with:
> >=20
> > ethtool --set-eee lan2 eee on
> >=20
> > ethtool --show-eee lan2
> > EEE Settings for lan2:
> >         EEE status: not supported
> >=20
> >=20
> > As the capability register shows value of 0. =20
>=20
> Some PHYs indeed have this issues:
> https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/d=
rivers/net/phy/micrel.c?h=3Dv6.4-rc4#n1402
>=20
> In case of your older kernel version, you will need to fake access to
> the EEE caps register.

Thanks for pointing this out - I will add this functionality when
re-enabling of EEE will be required via ethtool.

>=20
> Regards,
> Oleksij


Best regards,

Lukasz Majewski

--

DENX Software Engineering GmbH,      Managing Director: Erika Unter
HRB 165235 Munich, Office: Kirchenstr.5, D-82194 Groebenzell, Germany
Phone: (+49)-8142-66989-59 Fax: (+49)-8142-66989-80 Email: lukma@denx.de

--Sig_/p0ndmoiOlqD_T5Ryyql7.kd
Content-Type: application/pgp-signature
Content-Description: OpenPGP digital signature

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

iQEzBAEBCgAdFiEEgAyFJ+N6uu6+XupJAR8vZIA0zr0FAmR3E+4ACgkQAR8vZIA0
zr259wf/VvtYLeQbYYlr2YpitCq/igWjlShQTjsO2holZJmOHsBfQF5KoRPjcFbT
FEAn/X1C3PsVu2MkxdFUgTlnwZ7FnFkCtvo7o91oTwmoRAI+nkSkfVroA7/iGhj0
lU7EsIElUAK25cH2jOGwucvbCbCtvMjEF3g3M8wjwDrpd8DDaXQtUasiB439D5te
q9NyJ0NZCJcq6DwEWiDsH4KuNX91vdnpiGwxrph7quVngvLa9/3+W6/InTmtcm38
PLoCRqYQP5o56Tl1jHQfEc/JsJ114Jco8GRlvbmps7tWsr4lLyM3GNNAPRhAmJt8
0WUGOmEq4rmkzGTkwAWSpPmM48koEQ==
=NHQC
-----END PGP SIGNATURE-----

--Sig_/p0ndmoiOlqD_T5Ryyql7.kd--