From: Fabio Estevam <festevam@gmail.com>
To: netdev <netdev@vger.kernel.org>, Andrew Lunn <andrew@lunn.ch>,
Florian Fainelli <f.fainelli@gmail.com>,
Russell King - ARM Linux <linux@armlinux.org.uk>,
Heiner Kallweit <hkallweit1@gmail.com>
Cc: Bruno Thomsen <bth@kamstrup.com>, Lars Alex Pedersen <laa@kamstrup.com>
Subject: Re: fec: micrel: Ethernet PHY type ID auto-detection issue
Date: Fri, 17 Jul 2020 12:52:08 -0300 [thread overview]
Message-ID: <CAOMZO5DtYDomD8FDCZDwYCSr2AwNT81Ay4==aDxXyBxtyvPiJA@mail.gmail.com> (raw)
In-Reply-To: <CAH+2xPCzrBgngz5cY9DDDjnFUBNa=NSH3VMchFcnoVbjSm3rEw@mail.gmail.com>
Hi Bruno,
Adding netdev and the Ethernet PHY folks.
On Fri, Jul 17, 2020 at 12:48 PM Bruno Thomsen <bruno.thomsen@gmail.com> wrote:
>
> Hi
>
> I have been having issues with Ethernet PHY type ID
> auto-detection when changing from the deprecated fec
> phy-reset-{gpios,duration,post-delay} properties to the
> modern mdio reset-{assert-us,deassert-us,gpios}
> properties in the device tree.
>
> Hardware:
> SoC: i.MX7D
> PHY: Mircrel KSZ8081RNB
> PHY address: 1
> RMII clock is generated by the Micrel PHY.
>
> Micrel pin wiring:
> CONFIG0: 4k75 PU
> CONFIG1: 1k PD
> PHYAD0: 4k75 PU
> PHYAD1: 1k PD
> PHYAD2/RXD1: 1k PD
> REXT: 6k49 PD
> INTRP: 1k PU
> XO+XI: 25MHz xtal
>
> Kernel trace from working (deprecated fec reset):
> kernel: Micrel KSZ8081 or KSZ8091 30be0000.ethernet-1:01: attached PHY driver
> [Micrel KSZ8081 or KSZ8091] (mii_bus:phy_addr=30be0000.ethernet-1:01, irq=POLL)
>
> This way has been working with the mainline kernel 5.2.x-5.7.8.
>
> Kernel error messages (modem mdio reset):
> mdio_bus 30be0000.ethernet-1: MDIO device at address 1 is missing.
> fec 30be0000.ethernet eth0: Unable to connect to phy
>
> When debugging it I end up with the get_phy_device() call not
> working inside of_mdiobus_register_phy().
>
> A known workaround with modern mdio properties is
> to add "ethernet-phy-id0022.1560" as compatible,
> but it would be nice to know the root cause of this issue,
> so I can enable auto PHY id detection again.
>
> I have inserted snippets of the device tree below.
>
> DT common pin mux:
> &iomuxc {
> pinctrl_enet1: enet1grp {
> fsl,pins = <
> MX7D_PAD_GPIO1_IO10__ENET1_MDIO
> 0x03 /* X2-48 */
> MX7D_PAD_GPIO1_IO11__ENET1_MDC
> 0x03 /* X2-46 */
> MX7D_PAD_ENET1_RGMII_TD0__ENET1_RGMII_TD0
> 0x71 /* X2-53 */
> MX7D_PAD_ENET1_RGMII_TD1__ENET1_RGMII_TD1
> 0x71 /* X2-55 */
> MX7D_PAD_ENET1_RGMII_TX_CTL__ENET1_RGMII_TX_CTL
> 0x71 /* X2-61 */
> MX7D_PAD_ENET1_RGMII_RD0__ENET1_RGMII_RD0
> 0x79 /* X2-56 */
> MX7D_PAD_ENET1_RGMII_RD1__ENET1_RGMII_RD1
> 0x79 /* X2-58 */
> MX7D_PAD_ENET1_RGMII_RX_CTL__ENET1_RGMII_RX_CTL
> 0x79 /* X2-64 */
> MX7D_PAD_ENET1_RGMII_RXC__ENET1_RX_ER
> 0x73 /* X2-52 */
> /* PHY reset: SION, 100kPU, SRE_FAST, DSE_X1 */
> MX7D_PAD_ENET1_COL__GPIO7_IO15
> 0x40000070 /* X1-96 */
> /* Clock from PHY to MAC */
> MX7D_PAD_GPIO1_IO12__CCM_ENET_REF_CLK1
> 0x40000073 /* X3-4 */
> >;
> };
> };
>
> DT working:
> &fec1 {
> pinctrl-names = "default";
> pinctrl-0 = <&pinctrl_enet1>;
> phy-mode = "rmii";
> phy-reset-gpios = <&gpio7 15 GPIO_ACTIVE_LOW>;
> phy-reset-duration = <100>;
> phy-reset-post-delay = <1000>;
> phy-handle = <ðphy>;
> status = "okay";
>
> mdio {
> #address-cells = <1>;
> #size-cells = <0>;
> ethphy: ethernet-phy@1 {
> compatible = "ethernet-phy-ieee802.3-c22";
> reg = <1>;
> };
> };
> };
>
> DT unable to detect PHY type ID:
> &fec1 {
> pinctrl-names = "default";
> pinctrl-0 = <&pinctrl_enet1>;
> phy-mode = "rmii";
> phy-handle = <ðphy>;
> status = "okay";
>
> mdio {
> #address-cells = <1>;
> #size-cells = <0>;
>
> ethphy: ethernet-phy@1 {
> compatible = "ethernet-phy-ieee802.3-c22";
> reg = <1>;
> max-speed = <100>;
> reset-assert-us = <100000>;
> reset-deassert-us = <1000000>;
> reset-gpios = <&gpio7 15 GPIO_ACTIVE_LOW>;
> };
> };
> };
>
> I had a suspicion that it might be caused by a missing interrupt
> signal as it runs in poll mode and not irq mode. But fixing that
> with the following additions did not fix phy id auto-detection.
>
> ethphy: ethernet-phy@1 {
> interrupt-parent = <&gpio1>;
> interrupts = <9 IRQ_TYPE_LEVEL_LOW>;
> };
>
> pinctrl_enet1: enet1grp {
> fsl,pins = <
> MX7D_PAD_GPIO1_IO09__GPIO1_IO9 0x40000078 /* X1-80 */
> >;
> };
>
> But with the mdio reset workaround from above it now uses irq.
>
> Kernel trace:
> kernel: Micrel KSZ8081 or KSZ8091 30be0000.ethernet-1:01: attached PHY driver
> [Micrel KSZ8081 or KSZ8091] (mii_bus:phy_addr=30be0000.ethernet-1:01, irq=66)
>
> Any suggestions that could help locate the root cause of the issue?
>
> Thanks in advance.
>
> /Bruno
next parent reply other threads:[~2020-07-17 15:52 UTC|newest]
Thread overview: 6+ messages / expand[flat|nested] mbox.gz Atom feed top
[not found] <CAH+2xPCzrBgngz5cY9DDDjnFUBNa=NSH3VMchFcnoVbjSm3rEw@mail.gmail.com>
2020-07-17 15:52 ` Fabio Estevam [this message]
2020-07-17 16:34 ` fec: micrel: Ethernet PHY type ID auto-detection issue Andrew Lunn
2020-07-22 10:59 ` Bruno Thomsen
2020-07-22 13:26 ` Andrew Lunn
2020-07-28 8:51 ` Bruno Thomsen
2020-07-22 13:42 ` Heiner Kallweit
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to='CAOMZO5DtYDomD8FDCZDwYCSr2AwNT81Ay4==aDxXyBxtyvPiJA@mail.gmail.com' \
--to=festevam@gmail.com \
--cc=andrew@lunn.ch \
--cc=bth@kamstrup.com \
--cc=f.fainelli@gmail.com \
--cc=hkallweit1@gmail.com \
--cc=laa@kamstrup.com \
--cc=linux@armlinux.org.uk \
--cc=netdev@vger.kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).