| rrq | this CS968 box seems to have a "dual function" setup for OTG and its eth0, and I can't figure out the CONFIG_* settings to make that work with v6.1 | 01:50 |
|---|---|---|
| gnarface | rrq: do you have a working kernel? you could start by copying the config from that... | 01:52 |
| rrq | I have a working v5.10 for it and have tried that path; I get v6.1 started the wirte dmesg but no eth0 | 01:53 |
| rrq | mmm "the wirte" = "and writing" | 01:53 |
| gnarface | no extra firmware packages are involved? | 01:54 |
| gnarface | if it's not a firmware version mismatch, the only other idea i have would be to start doing a string search in menuconfig for the device name, maybe they just changed the actual flags' spellings | 01:55 |
| rrq | not for this... it needs something for wlan0 which I don't have | 01:55 |
| rrq | and I don't have the monitor working either ... don't need those | 01:55 |
| gnarface | if you type "/" it should give you a prompt to do string search through the fieldd values | 01:55 |
| gnarface | it's not super obvious but then you can actually cursor-highlight the search result entries and hit enter to jump right to them in the menu | 01:56 |
| gnarface | (been compiling linux kernels for almost 30 years and i just learned this a couple weeks ago) | 01:56 |
| rrq | yes. I use "ncurses" so it's F8 ... hmm jumping directly would help a bit | 01:57 |
| rrq | I can compare the dmesg from 5.10 and 6.1 ... and found it deviates where 5.10 flips the OTG port and installs echi | 01:57 |
| rrq | while 6.1 flips the OTG port and complains about the port status | 01:58 |
| rrq | something (I don't know enough there) | 01:58 |
| rrq | it appears to be due to udevd starting, but I'm not sure bout that (either) | 01:59 |
| rrq | same root fs (daedalus) an in face 5.10 is happy starting with the 6.1 initrd | 02:00 |
| gnarface | ah, a change to udev scripts could certainly break device node creation even if the kernel is loading the driver right | 02:00 |
| rrq | (5.10 has the drivers built-in apparently) | 02:01 |
| gnarface | udev still needs to work right | 02:01 |
| gnarface | (or you need to call mknod manually, i think that still works...) | 02:02 |
| rrq | ok.. hmm I guess I can mess with the initrd init to get some userspace dump into syslog... I do get dmesg, so it pivots and boots well | 02:06 |
| rrq | but I don't have any comms channel for interactive | 02:07 |
| rrq | hmmm maybe some rc.local scripting can force that | 02:08 |
| rrq | thanks | 02:11 |
| gnarface | ? | 02:12 |
| rrq | "thanks for inspiring a thought of another approach..." :) | 02:14 |
| gnarface | ah | 02:14 |
| gnarface | well, good luck with it | 02:15 |
| gnarface | let us know if you figure anything out | 02:15 |
| gnarface | i can't really provide any specific insight for that particular piece of hardware, but maybe some other lurker here can | 02:15 |
Generated by irclog2html.py 2.17.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!