Re: [PATCH net-next v1 1/1] net: phy: dp83td510: Add basic support for the DP83TD510 Ethernet PHY

From: Andrew Lunn <andrew@lunn.ch>
To: Oleksij Rempel <o.rempel@pengutronix.de>
Cc: Heiner Kallweit <hkallweit1@gmail.com>,
	"David S. Miller" <davem@davemloft.net>,
	Jakub Kicinski <kuba@kernel.org>, Dan Murphy <dmurphy@ti.com>,
	kernel@pengutronix.de, netdev@vger.kernel.org,
	linux-kernel@vger.kernel.org, David Jander <david@protonic.nl>,
	Russell King <linux@armlinux.org.uk>
Subject: Re: [PATCH net-next v1 1/1] net: phy: dp83td510: Add basic support for the DP83TD510 Ethernet PHY
Date: Fri, 23 Jul 2021 20:12:05 +0200	[thread overview]
Message-ID: <YPsGddTXtk/Hinmp@lunn.ch> (raw)
In-Reply-To: <20210723170848.lh3l62l7spcyphly@pengutronix.de>

On Fri, Jul 23, 2021 at 07:08:49PM +0200, Oleksij Rempel wrote:
> Hi Andrew,
> 
> On Fri, Jul 23, 2021 at 03:22:16PM +0200, Andrew Lunn wrote:
> > On Fri, Jul 23, 2021 at 12:42:18PM +0200, Oleksij Rempel wrote:
> > > The DP83TD510E is an ultra-low power Ethernet physical layer transceiver
> > > that supports 10M single pair cable.
> > > 
> > > This driver provides basic support for this chip:
> > > - link status
> > > - autoneg can be turned off
> > > - master/slave can be configured to be able to work without autoneg
> > > 
> > > This driver and PHY was tested with ASIX AX88772B USB Ethernet controller.
> > 
> > Hi Oleksij
> > 
> > There were patches flying around recently for another T1L PHY which
> > added new link modes. Please could you work together with that patch
> > to set the phydev features correctly to indicate this PHY is also a
> > T1L, and if it support 2.4v etc.
> 
> ACK, thx. I was not able to spend enough time to investigate all needed
> caps, so I decided to go mainline with limited functionality first.

Limited functionality is fine, but what is implemented should be
correct. And from what i see in the patch, it is hard to know if the
PHYs basic features are correctly determined. What does ethtool show?
Is 100BaseT being offered? Half duplex?

> voltage depends on the end application: cable length, safety requirements. I do
> not see how this can be chosen only on auto negotiation. We would need proper
> user space interface to let user/integrator set the limits.

I think we are talking at cross purposes here. As far as i understand
T1L supports the data signals to be 2.4Vpp as well as the usual
1Vpp. This is something which can be negotiated in the same way as
master/slave, duplex etc.

I suspect you are talking about the PoE aspects. That is outside the
scope for phylib. PoE in general is not really supported in the Linux
kernel, and probably needs a subsystem of its own.

   Andrew