Ethernet PHY KSZ9031 is not working
gianluca
gianlucarenzi at eurek.it
Thu Mar 3 04:44:54 PST 2016
Hello list,
It's me again! ;-)
Actually I had tested 80% of the board with barebox.
- USB
The microUSB is working (as I can use serial Downloader)
USB Host-A has to be tested, but if I type:
>> usb
> USB: scanning bus for devices...
> Bus 001 Device 001: ID 0000:0000 EHCI Host Controller
> 1 USB Device(s) found
so I can suppose the host controller is working. So I have to check on
HOW to test the usb-pendisk, but it can be done later on... I have to
activate the +5V by a gpio, so I have to think where is the best place
to do that. At the C code level in the device_initcalls()?? Or using the
gpio pinmux features in the device tree?
- I2C (i2c0, i2c1, i2c2)
They all three are working fine.
- DDR3 Memory Controller
Memtest is passing all tests. So I can suppose the timings are quite
good enough to use the Dynamic RAM correctly. I will do a stress test
later...
- SDHC (mmc1, mmc2, mmc3)
> -- 2194000.usdhc
> `-- mmc1
This port is connected to a SDIO Wifi Module so I will test it with Linux
> -- 2198000.usdhc
> `-- mmc2
> `-- 0x00000000-0x1d9bfffff ( 7.4 GiB): /dev/mmc2
> `-- 0x00400000-0x00bfffff ( 8 MiB): /dev/mmc2.0
> `-- 0x00c00000-0x02bfffff ( 32 MiB): /dev/mmc2.1
> `-- 0x02c00000-0x1d9bfffff ( 7.4 GiB): /dev/mmc2.2
This port has a microSD 8Gb with 3 partitions.
> -- 219c000.usdhc
> `-- mmc3
> `-- 0x00000000-0x001fffff ( 2 MiB): /dev/mmc3.boot0
> `-- 0x00000000-0x001fffff ( 2 MiB): /dev/mmc3.boot1
> `-- 0x00000000-0xe4ffffff ( 3.6 GiB): /dev/mmc3
This port has an eMMC 4GB unpartioned. 2 boot partitions and a big one
for boot when it starts from internal boot.
- GPIOS
The gpios can be activated/deactivated on demand. As far as I see all
gpios I had on this board are working as expected.
- SERIAL
In this board I have 3 serial ports (only RX/TX connected)
> `-- 21ec000.serial
> `-- serial2
This is the uart3 as debug console and it works fine.
> `-- 21e8000.serial
> `-- serial1
> `-- 2000000.aips-bus
> `-- 2000000.spba-bus
> `-- 2020000.serial
> `-- serial0
How can I test them with barebox? Is it possible to set the baudrate,
and send a character with the 'echo' command? And as soon as I have the
/dev/ node can I read from it using the 'cat' command?
- FEC/PHY
This is the worst part of the game. :-(
It does not work at all.
I was looking on other boards with the same phy (KSZ9031 by Micrel) like
Variscite-imx6, but with no luck.
In board.c I setup ONLY the fixup for ksz9031 and assert/de-assert the
nRESET RGMII pin
> static int ksz9031rn_phy_fixup(struct phy_device *dev)
> {
> pr_info("%s Called\n", __FUNCTION__);
> /*
> * min rx data delay, max rx/tx clock delay,
> * min rx/tx control delay
> */
> phy_write_mmd_indirect(dev, 4, 2, 0);
> phy_write_mmd_indirect(dev, 5, 2, 0);
> phy_write_mmd_indirect(dev, 8, 2, 0x03ff);
>
> pr_info("%s Exit\n", __FUNCTION__);
> return 0;
> }
>
> static int ek360_enet_init(void)
> {
> pr_info("%s Called\n", __FUNCTION__);
> if (!of_machine_is_compatible("eurek,ek360"))
> return 0;
>
> /*
> * As this board has all configure pins (MODE0..3, PHYADD0..2, ...)
> * with a fixed state (pullup & pulldown resistors) there should no
> * need to have gpio working here for configuration.
> * We need only assert & de-assert the reset pin here
> */
> mxc_iomux_v3_setup_pad(MX6Q_PAD_EIM_D23__GPIO_3_23);
> gpio_direction_output(IMX_GPIO_NR(3,23), 0); /* assert nRESET */
> mdelay(50);
> gpio_direction_output(IMX_GPIO_NR(3,23), 1); /* de-assert nRESET */
> gpio_free(IMX_GPIO_NR(3,23));
>
> phy_register_fixup_for_uid(PHY_ID_KSZ9031, MICREL_PHY_ID_MASK,
> ksz9031rn_phy_fixup);
>
> pr_info("%s Exit\n", __FUNCTION__);
> return 0;
> }
> fs_initcall(ek360_enet_init);
>
and let the device-tree to set the correct pins and setup the fec:
> pinctrl_enet: enetgrp {
> fsl,pins = <
> MX6QDL_PAD_RGMII_RXC__RGMII_RXC 0x1b0b0
> MX6QDL_PAD_RGMII_RD0__RGMII_RD0 0x1b0b0
> MX6QDL_PAD_RGMII_RD1__RGMII_RD1 0x1b0b0
> MX6QDL_PAD_RGMII_RD2__RGMII_RD2 0x1b0b0
> MX6QDL_PAD_RGMII_RD3__RGMII_RD3 0x1b0b0
> MX6QDL_PAD_RGMII_RX_CTL__RGMII_RX_CTL 0x1b0b0
> MX6QDL_PAD_RGMII_TXC__RGMII_TXC 0x1b0b0
> MX6QDL_PAD_RGMII_TD0__RGMII_TD0 0x1b0b0
> MX6QDL_PAD_RGMII_TD1__RGMII_TD1 0x1b0b0
> MX6QDL_PAD_RGMII_TD2__RGMII_TD2 0x1b0b0
> MX6QDL_PAD_RGMII_TD3__RGMII_TD3 0x1b0b0
> MX6QDL_PAD_RGMII_TX_CTL__RGMII_TX_CTL 0x1b0b0
> MX6QDL_PAD_ENET_REF_CLK__ENET_TX_CLK 0x1b0b0
> MX6QDL_PAD_ENET_MDIO__ENET_MDIO 0x1b0b0
> MX6QDL_PAD_ENET_MDC__ENET_MDC 0x1b0b0
> MX6QDL_PAD_EIM_D23__GPIO3_IO23 0x80000000 /* KSZ9031 PHY Reset */
> >;
> };
> &fec {
> pinctrl-names = "default";
> pinctrl-0 = <&pinctrl_enet>;
> phy-mode = "rgmii";
> phy-reset-gpios = <&gpio3 23 GPIO_ACTIVE_HIGH>;
> phy-reset-duration = <100>;
> status = "okay";
> };
On reset/poweron I can see clearly some RJ-45 JackMag leds already lite
due to the fixed hardware configuration.
When Barebox takes control over, those pins are turned off, the device
tree information is telling me:
> `-- 2188000.ethernet
> `-- miibus0
> `-- phy0
> `-- 0x00000000-0x0000003f ( 64 Bytes): /dev/phy0
> `-- eth0
but when I try to bring up the ethernet connection here is the message:
> dhcp
> warning: No MAC address set. Using random address ae:05:f1:62:bd:d8
> ksz9031rn_phy_fixup Called
> ksz9031rn_phy_fixup Exit
> dhcp: Network is down
And all leds in the RJ-45 JackMag are still turned off.
I don't know where I can check if something goes wrong. May be a failure
on board, but at this time this board is the ONLY BOARD I can use.
Can somebody help me out????
Best Regards,
Gianluca
--
Eurek s.r.l. |
Electronic Engineering | http://www.eurek.it
via Celletta 8/B, 40026 Imola, Italy | Phone: +39-(0)542-609120
p.iva 00690621206 - c.f. 04020030377 | Fax: +39-(0)542-609212
More information about the barebox
mailing list