IRC #olimex 2020-11-06

[00:00:06] <nedko> *latter (works with rev.L)
[00:05:57] <nedko> maybe i have another G2 somewhere...
[00:43:35] <nedko> rgmii-id phy-mode in dts works with 5.4.74 (longterm)
[00:51:45] <nedko> after setting board revision in u-boot (olinuxino config write), G2 network works
[00:52:31] <nedko> probably same has to be done for rev.C, where test result was no network
[03:10:19] <jo0nas> nedko: with rev. C check if the board runs in master or slave mode
[03:11:18] <jo0nas> you can check that with ethtool v5.8 or newer
[03:13:49] <jo0nas> mainline linux and mainline u-boot uses one device-tree file for rev. A-E and another for rev. F-L
[03:16:26] <jo0nas> as far as I am aware, the difference to the _device-tree files_ is support or not for eMMC and SPI - but the default u-boot config also differ in that the one for A-E boards forces the Realtekdriver into master mode
[03:19:01] <jo0nas> ...which means that a) if you use the proper setup for your rev. C board then it must be connected to something that supports running in slave mode, and b) if you use the newer setup then will randomly work (when connection randomly ends in board being master, or at lower speeds than gigabit) and randomly fail (when board ends in gigabit slave mode)
[03:19:55] <jo0nas> ...unless some of that changes when switching device-tree file from rgmii to rgmii-id
[10:55:23] <leon-anavi> hi
[18:27:42] <nedko> the A20-OLinuXino-buster-minimal-20201105-110358.img so far works great...
[18:27:46] <nedko> with G2
[18:28:27] <nedko> :D
[18:33:28] <nedko> i made root autologin on ttyS0, until i make ssh work over uart transport :D
[18:34:29] <nedko> so with few modifications the resulting node feels really good!
[18:34:31] <nedko> ACTION happy
[18:35:00] <braunr> :)
[18:37:15] <nedko> also, /etc/issue now includes kernel version, for convenience when looking into boot log :)