Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::3:4 as permitted sender) client-ip=2620:137:e000::3:4;
From:   =?UTF-8?q?Ram=C3=B3n=20N=2ERodriguez?= 
        <ramon.nordin.rodriguez@ferroamp.se>
To:     Andrew Lunn <andrew@lunn.ch>,
        Heiner Kallweit <hkallweit1@gmail.com>,
        Russell King <linux@armlinux.org.uk>,
        "David S. Miller" <davem@davemloft.net>,
        Eric Dumazet <edumazet@google.com>,
        Jakub Kicinski <kuba@kernel.org>,
        Paolo Abeni <pabeni@redhat.com>, netdev@vger.kernel.org,
        linux-kernel@vger.kernel.org
Cc:     =?UTF-8?q?Ram=C3=B3n=20N=2ERodriguez?= 
        <ramon.nordin.rodriguez@ferroamp.se>
Subject: [PATCH 0/3] net: microchip_t1s: additional phy support and collision detect handling
Date:   Mon, 27 Nov 2023 11:40:42 +0100
Message-ID: <20231127104045.96722-1-ramon.nordin.rodriguez@ferroamp.se>
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
Precedence: bulk

Hi,

This patch series adds support for LAN867x Rev.C and sets the collision
enable bit conditionally on plca enable for LAN865X and LAN867X phys.

The addition of Rev.C support is pretty straight forward (but weird), it 
follows the recommended magic and unexplained steps in Microchip
Application Note 1699.

The second change, conditionally setting collision detection has a
significant impact on link performance. Having both PLCA and collision
detection enabled led to a lot of dropped packets in the setup I've been
developing on.
The datasheets recommends disabling collision detect when PLCA is
enabled.
Took me a couple of weeks to find said footnote, hoping this series can
save the next dev some headache.

Following is a brief description of my test/eval setup:

LAN867x Rev.C
This was tested with Microchips usb eval board, the only testing I did
was that I got the driver attached and got a link.
No performance/stress testing has been performed.

Collision detection
This has been tested on a setup where one ARM system with a LAN8650
mac-phy is daisy-chained to 8 mcus using lan8670 phys. Without the patch we
were limited to really short cables, about 1m per node, but we were
still getting a lot of connection drops.
With the patch we could increase the total cable length to at least 40M.
Electrical properties play in here, both cable types and how the
termination is designed, so these results might not be reproducible in
another setup.
The only testing that has been performed has been a best effort of
estimating dropped packets on the linux node, with low frequency traffic.

Ramón Nordin Rodriguez (3):
  net: microchip_t1s: refactor reset functionality
  net: microchip_t1s: add support for LAN867x Rev.C1
  net: microchip_t1s: conditional collision detection

 drivers/net/phy/microchip_t1s.c | 135 +++++++++++++++++++++++++++++++-
 1 file changed, 132 insertions(+), 3 deletions(-)

-- 
2.40.1