RSS Bot Posted June 8 Posted June 8 Description Per Igor comment helios64 required work to sync with upstream changes. I took the liberty to keep more of upstream than in Armbian 6.8, namely: I kept nodes at their upstream location to remove unnecessary work when syncing to the upstream keep "ethernet0 = &gmac;" in aliases, to let the ethernet MAC retrieved from OTP in u-boot apply. This will restore this feature (the u-boot was still functional). U-boot retried ethernet MACs are in u-boot printenv output "ethaddr" and "eth1addr". Seems for a few years I had the Linux kernel autogenerated one but my config still had the correct one commented (I did not find why it broke back then). I removed the overclock disabling code as the overclock is not included in the dts so no need to disable it (will lower the sync to upstream burden) kept upstream fix 93b36e1d3748c352a70c69aa378715e6572e51d1 "arm64: dts: rockchip: Fix USB interface compatible string on kobol-helios64" from upstream in vcc3v0_sd node I kept "enable-active-high;" and "gpio = <&gpio0 RK_PA1 GPIO_ACTIVE_HIGH>;" sync to upstream from helios64::status to helios64:green:status for system_leds : status_led kept startup-delay-us = <2000000> on hdd nodes from upstream (I believe they should not be the same value for rail a and b of the HDD, one should do as in the helios64 u-boot code where one is started after the other to lower the current load) I also did not sync with upstream the differences of gpio pull config for "power > sdmmc0_pwr_h" and "pinctrl > pmic > pmic_int_l" I kept the upstream DTS model "Kobol Helios64" instead of the armbian "Helios64". Ethernet rename from armbian-hardware-optimization works. Documentation summary for feature / change [ ] short restore ethernet MACs for eth0 read from OTP via u-boot [ ] This will restore the original intent which is to have the eth0 mAC be defined from the mainboard. This might affect MAC-based DHCP definitions [ ] ip l to get the Linux-defined eth0/end0 MAC, printenv in u-boot to see ethaddr and eth1addr for eth0/end0 and eth1 How Has This Been Tested? [ ] Compile, installed, and reboot, checked assigned IP was from the dnsmasq dhcp-host definition for this ethernet MAC. Did a few other reboots. Checklist: [ x ] My code follows the style guidelines of this project [ x ] I have performed a self-review of my own code [ x ] My changes generate no new warnings View the full article
Recommended Posts