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;
MIME-Version: 1.0
References: <20190802215419.313512-1-taoren@fb.com> <CA+h21hrOEape89MTqCUyGFt=f6ba7Q-2KcOsN_Vw2Qv8iq86jw@mail.gmail.com>
 <53e18a01-3d08-3023-374f-2c712c4ee9ea@fb.com> <20190804145152.GA6800@lunn.ch>
In-Reply-To: <20190804145152.GA6800@lunn.ch>
From:   Vladimir Oltean <olteanv@gmail.com>
Date:   Sun, 4 Aug 2019 18:59:44 +0300
Message-ID: <CA+h21hrUDaSxKpsy9TuWqwgaxKYaoXHyhgS=xSoAcPwxXzvrHg@mail.gmail.com>
Subject: Re: [PATCH net-next v3] net: phy: broadcom: add 1000Base-X support
 for BCM54616S
To:     Andrew Lunn <andrew@lunn.ch>
Cc:     Tao Ren <taoren@fb.com>, Florian Fainelli <f.fainelli@gmail.com>,
        Heiner Kallweit <hkallweit1@gmail.com>,
        "David S . Miller" <davem@davemloft.net>,
        Arun Parameswaran <arun.parameswaran@broadcom.com>,
        Justin Chen <justinpopo6@gmail.com>,
        netdev <netdev@vger.kernel.org>,
        lkml <linux-kernel@vger.kernel.org>,
        "openbmc@lists.ozlabs.org" <openbmc@lists.ozlabs.org>
Content-Type: text/plain; charset="UTF-8"
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On Sun, 4 Aug 2019 at 17:52, Andrew Lunn <andrew@lunn.ch> wrote:
>
> > > The patchset looks better now. But is it ok, I wonder, to keep
> > > PHY_BCM_FLAGS_MODE_1000BX in phydev->dev_flags, considering that
> > > phy_attach_direct is overwriting it?
> >
>
> > I checked ftgmac100 driver (used on my machine) and it calls
> > phy_connect_direct which passes phydev->dev_flags when calling
> > phy_attach_direct: that explains why the flag is not cleared in my
> > case.
>
> Yes, that is the way it is intended to be used. The MAC driver can
> pass flags to the PHY. It is a fragile API, since the MAC needs to
> know what PHY is being used, since the flags are driver specific.
>
> One option would be to modify the assignment in phy_attach_direct() to
> OR in the flags passed to it with flags which are already in
> phydev->dev_flags.
>
>         Andrew

Even if that were the case (patching phy_attach_direct to apply a
logical-or to dev_flags), it sounds fishy to me that the genphy code
is unable to determine that this PHY is running in 1000Base-X mode.

In my opinion it all boils down to this warning:

"PHY advertising (0,00000200,000062c0) more modes than genphy
supports, some modes not advertised".

You see, the 0x200 in the above advertising mask corresponds exactly
to this definition from ethtool.h:
    ETHTOOL_LINK_MODE_1000baseX_Full_BIT    = 41,

But it gets truncated and hence lost.

Regards,
-Vladimir