Hello @khan_misbah,
Thanks for reaching out to us. Regarding your question, the SerDes configuration in U-Boot, and therefore the SGMII configuration is managed by an environment variable, the hwconfig variable, for more information please check your BSP user manual, for example for BSP44 check section 7.2.2 SerDes configuration for SGMII in U-Boot in particular page 51:
If you have already check this information, please share the following so I may be able to offer you better support:
- BSP version
- hwconfig current value
Thanks in advance
Hello @khan_misbah,
I thanks for the information. It indeed seems that none of the Ethernet clocks are getting enabled, for example, here is the diff with my clocks, outlined in red are the important differences:
Please execute the following commands and share the output:
md 0x4033C000 0x4DF
md 0x460A0000 0x4DF
md 0x460A4000 0x4DF
md 0x460A8000 0x4DFThe commands are just a register dump of the GMAC, PFE_EMAC0, PFE_EMAC1 and PFE_EMAC2 base addresses, the 0x4DF is the size in 32bit words.
I am attaching the output of those commands in my RDB2 board as reference.
Please also let me know what changes did you require successfully run the verifclk command.
Thanks for all the information.
Hello @khan_misbah,
Thanks for the information, I compared the registers of my PFE1 using SGMII to your PFE2, I did not find much, however, what I did find are some differences in the Clause 45 registers, and following the BSP43 manual, section 7.3.1.1 GMAC U-Boot driver and checking the S32G2-EVB3 device tree for ATF/u-boot s32g2xxa-evb3.dts#L43 (and considering that GMAC and PFE interfaces have the same registers), can you try adding the following in &pfe_mdio2:
&pfe_mdio2 {
status = "okay";
/* TJA1104B */
pfe_mdio_b_phy2: ethernet-phy@2 {
compatible = "ethernet-phy-ieee802.3-c45";
reg = <27>;
};
};
You can also try configuring the CONFIG_PHY_NXP_C45_TJA11XX to your configuration with the menuconfig in u-boot:
Which adds support for the TJA11 family.
Since I do not have your board I cannot test the exact effects of the changes, however, please test each alone and then both.
If the problem persists, please also share the output of the dm tree command in u-boot, to compare it with my own.
By the way, I do understand that it is quite unusual that Linux works and u-boot does not, however, the device tree does not define the behavior of driver, it is only a descriptor, therefore Linux and u-boot have different behaviors.
Thanks in advance.
