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:   Thu, 5 Jan 2023 19:43:42 +0200
From:   Vladimir Oltean <olteanv@gmail.com>
To:     "Russell King (Oracle)" <linux@armlinux.org.uk>
Cc:     Sean Anderson <sean.anderson@seco.com>,
        Andrew Lunn <andrew@lunn.ch>,
        Heiner Kallweit <hkallweit1@gmail.com>, netdev@vger.kernel.org,
        "David S . Miller" <davem@davemloft.net>,
        Paolo Abeni <pabeni@redhat.com>, linux-kernel@vger.kernel.org,
        Jakub Kicinski <kuba@kernel.org>,
        Eric Dumazet <edumazet@google.com>,
        Tim Harvey <tharvey@gateworks.com>
Subject: Re: [PATCH net-next v5 4/4] phy: aquantia: Determine rate adaptation
 support from registers
Message-ID: <20230105174342.jldjjisgzs6dmcpd@skbuf>
References: <20230103220511.3378316-1-sean.anderson@seco.com>
 <20230103220511.3378316-5-sean.anderson@seco.com>
 <20230105140421.bqd2aed6du5mtxn4@skbuf>
 <Y7bhctPZoyNnw1ay@shell.armlinux.org.uk>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <Y7bhctPZoyNnw1ay@shell.armlinux.org.uk>
Precedence: bulk

On Thu, Jan 05, 2023 at 02:40:50PM +0000, Russell King (Oracle) wrote:
> > If the PHY firmware uses a combination like this: 10GBASE-R/XFI for
> > media speeds of 10G, 5G, 2.5G (rate adapted), and SGMII for 1G, 100M
> > and 10M, a call to your implementation of
> > aqr107_get_rate_matching(PHY_INTERFACE_MODE_10GBASER) would return
> > RATE_MATCH_NONE, right? So only ETHTOOL_LINK_MODE_10000baseT_Full_BIT
> > would be advertised on the media side?
> 
> No, beause of the special condition in phylink that if it's a clause 45
> PHY and we use something like 10GBASE-R, we don't limit to just 10G
> speed, but try all interface modes - on the assumption that the PHY
> will switch its host interface.
> 
> RATE_MATCH_NONE doesn't state anything about whether the PHY operates
> in a single interface mode or not - with 10G PHYs (and thus clause 45
> PHYs) it seems very common from current observations for
> implementations to do this kind of host-interface switching.

So you mention commits
7642cc28fd37 ("net: phylink: fix PHY validation with rate adaption") and
df3f57ac9605 ("net: phylink: extend clause 45 PHY validation workaround").

IIUC, these allow the advertised capabilities to be more than 10G (based
on supported_interfaces), on the premise that it's possible for the PHY
to switch SERDES protocol to achieve lower speeds.

This does partly correct the last part of my question, but I believe
that the essence of it still remains. We won't make use of PAUSE rate
adaptation to support the speeds which aren't directly covered by the
supported_interfaces. Aren't we interpreting the PHY provisioning somewhat
too conservatively in this case, or do you believe that this is just an
academic concern?