| freemangordon | Wizzup: https://github.com/raspberrypi/linux/issues/5535 | 09:22 |
|---|---|---|
| freemangordon | I think we shall do the same | 09:22 |
| Wizzup | ah, we don't depend on these? | 10:33 |
| Wizzup | https://github.com/maemo-leste/bugtracker/issues/755 | 10:43 |
| freemangordon | also, the wifi issue on d4(first connection attempt always fails) seems to be way worse on allwinner | 10:45 |
| freemangordon | I'll inverstigate | 10:45 |
| freemangordon | but basically, it is almost impossible to connect from wifi menu | 10:45 |
| freemangordon | wpa_supplicant in shell never fails | 10:46 |
| Wizzup | even with latest code? | 10:49 |
| Wizzup | I think some drivers aren't happy to associate while still scanning | 10:50 |
| freemangordon | yes, latest 0.1.16+m7 | 10:57 |
| freemangordon | Wizzup: btw, do you know how to stop udev(?) from renaming usb0? | 11:00 |
| Wizzup | yes | 11:18 |
| Wizzup | this should not happen btw | 11:18 |
| freemangordon | umm, this is on PC | 11:19 |
| freemangordon | on device it is ok | 11:19 |
| Wizzup | ah | 11:19 |
| Wizzup | ok | 11:19 |
| Wizzup | yeah so you can boot with certain params like net.ifnames=0 | 11:19 |
| freemangordon | ah, I see | 11:20 |
| Wizzup | but there's also a track of symlinking a certain file to /dev/null iirc, but that name has changed many times | 11:20 |
| freemangordon | lemme see if I can stop NM from trying to manage it | 11:20 |
| Wizzup | https://unix.stackexchange.com/questions/760436/how-to-cancel-the-kernels-rename-of-the-network-interface | 11:20 |
| Wizzup | https://askubuntu.com/questions/1333034/how-do-i-stop-eth0-being-renamed-ens3 | 11:21 |
| freemangordon | thanks | 11:21 |
| freemangordon | well, I am not going to change grub parameters and unfortunately cannot use udev rule either, as devices change their usm mac on every boot | 11:22 |
| freemangordon | I just set all ens* devices to be ignored by NM | 11:23 |
| Wizzup | kdir -p /etc/systemd/network && ln -s /dev/null /etc/systemd/network/99-default.link | 11:23 |
| Wizzup | per the stack overflow | 11:23 |
| freemangordon | ok, lets focus on wifi issue | 11:23 |
| Wizzup | these are poor wifi driver problems | 11:26 |
| Wizzup | we can decide to not connect while scanning if you want | 11:27 |
| freemangordon | I am not sure it is driver issue, but anyway, lets get some logs first | 11:28 |
| Wizzup | on the droid4 the driver is 100% the problem | 11:28 |
| Wizzup | it times out while associating within a millisecond | 11:28 |
| Wizzup | the allwinner devices also have this notoriously poor wifi chip driver (at least at the time) | 11:29 |
| freemangordon | no, here it is different | 11:30 |
| Wizzup | ok (back in 15 mins or so) | 11:31 |
| freemangordon | ok | 11:31 |
| sicelo | freemangordon: could still have used udev rule. there are lots of properties you could do the matching by | 11:32 |
| sicelo | Anyway back to the wifi issue, if the issue is with connecting while scanning, who does the connecting? if it's us, could we not immediately stop scanning just before doing the connecting? | 11:33 |
| freemangordon | see https://pastebin.com/HF2EtqN2 | 11:33 |
| freemangordon | this happens right after https://pastebin.com/HF2EtqN2 | 11:34 |
| freemangordon | oops | 11:34 |
| freemangordon | this happens right after : | 11:34 |
| freemangordon | Dec 30 12:29:56 localhost icd2 0.99.1[3801]: libicd-network-wpasupplicant: on_scan_done. params: (true,) | 11:34 |
| freemangordon | Dec 30 12:29:57 localhost icd2 0.99.1[3801]: libicd-network-wpasupplicant: SCAN DONE | 11:34 |
| freemangordon | so, it is *not* scanning | 11:34 |
| sicelo | yeah | 11:36 |
| mkf | freemangordon: did you calibrated your screen? | 13:48 |
| freemangordon | mkf: no | 13:53 |
| freemangordon | but I took firmware from android | 13:54 |
| freemangordon | maybe it already has the correct calibration | 13:54 |
| freemangordon | also, if you use wrong firmware, I think you may have calibration issues | 13:54 |
| freemangordon | hmm, what the? samsung evo 128 A2 -> read speed 10MB/s | 13:55 |
| freemangordon | sequential ;) | 13:55 |
| freemangordon | at times drops to 5MB/s | 13:56 |
| freemangordon | no wonder that tablet feels like snail | 13:56 |
| arno11 | 5-10MB/s read ? indeed not good | 14:00 |
| arno11 | maybe because of sd bus, no ? samsung evo works quite well generally | 14:01 |
| freemangordon | no | 14:20 |
| freemangordon | this is in PC | 14:20 |
| freemangordon | tried directly in my laptop sdcard reader | 14:20 |
| freemangordon | in the same device, sandisk extreme pro a2 does stable 50 MB/s | 14:21 |
| freemangordon | something is wrong with the card | 14:21 |
| arno11 | ah ok yeah | 14:26 |
| freemangordon | lets hope it will remain alive until I create an image :) | 14:26 |
| freemangordon | arno11: actually, it seems to average @ 6MB/s | 14:44 |
| arno11 | seems really slow indeed | 14:49 |
| freemangordon | ok, seems it passed the broken part as @ 11% speed increased to 18MB/s. this is in USB2 card reader | 14:50 |
| mkf | freemangordon: oh, does fw has it's own calibration? | 14:59 |
| freemangordon | I think it has factory calibration, but can;t be sure | 14:59 |
| freemangordon | like, it somehow matches the display resolution or something | 15:00 |
| freemangordon | but, you should use the correct firmware for the TS | 15:00 |
| freemangordon | so, what is wring with android on your device? | 15:00 |
| freemangordon | *wrong | 15:00 |
| freemangordon | even if it does not boot, perhaps you can mount android partitions under linux and take whatever needed from there | 15:01 |
| freemangordon | esp script0.bin | 15:01 |
| freemangordon | without it I doubt you will be able to properly create fdt for your tablet | 15:02 |
| mkf | there is a script.bin i've found on internet that roughly matches my device | 15:13 |
| freemangordon | well, your TS is 'roughly' working :) | 15:15 |
| freemangordon | so, what is with android on your device? | 15:15 |
| * freemangordon will never ever buy samsung sdcard again | 15:16 | |
| arno11 | freemangordon: is it evo or evo plus ? | 15:18 |
| freemangordon | I think evo | 15:19 |
| freemangordon | RN it is in the reader and can't check | 15:20 |
| freemangordon | but, for sure is marked A2 | 15:20 |
| arno11 | ok | 15:20 |
| freemangordon | will chek in 4 hours and 30 minutes, when image creations should be complete :) | 15:21 |
| freemangordon | *check | 15:21 |
| arno11 | seems really wrong :( | 15:22 |
| freemangordon | oh, read speed dropped to 2MB/s | 15:22 |
| freemangordon | yeah, that card is totally busted | 15:23 |
| arno11 | (my evo plus 128 is quite fast on both pc-ubuntu and n900) | 15:23 |
| arno11 | yeah i think so | 15:23 |
| arno11 | 2MB/s is like a very old class 1 sd | 15:24 |
| freemangordon | well, it was just for a while, most of the time it keeps @ 6.5 | 15:24 |
| arno11 | ok but even 6.5 is quite slow | 15:25 |
| freemangordon | sure | 15:25 |
| freemangordon | it should be 18 | 15:25 |
| freemangordon | (card reader limit) | 15:26 |
| arno11 | yup | 15:26 |
| freemangordon | I wonder if it wouldn't have been faster to just copy the FS | 15:27 |
| Wizzup | it usually is, but can be more complicated | 15:48 |
| * sicelo wonders what would come out of benchmarking his cards ... shivers, and skips | 17:14 | |
| freemangordon | arno11: evo plus | 17:43 |
| arno11 | ugh | 17:46 |
| arno11 | seems defective | 17:49 |
| freemangordon | mhm | 18:03 |
| freemangordon | Wizzup: any clue what is theoretical mmc speed on allwinner a33? like olimex one | 19:58 |
| Wizzup | not sure | 20:46 |
Generated by irclog2html.py 2.17.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!