| donihalim | I still doesn't understand what happen when the system boot, the system often freeze after booting to the desktop that i had to force shutdown by removing the battery | 14:23 |
|---|---|---|
| donihalim | This makes rebooting the device quiete frustating | 14:23 |
| arno11 | oh | 14:25 |
| arno11 | donihalim: please elaborate | 14:25 |
| arno11 | with chimaera and kernel 6.6 and sd swap ? | 14:26 |
| arno11 | *chimaera -devel | 14:29 |
| donihalim | Ah sorry, so this is what happen: i powered on the phone and it boot normally, since i have display timeout set to 30s, after the phone boot to the hildon desktop the display will turned of after 30s | 14:30 |
| donihalim | After that the system often freeze, touching the screen or using the lock button doesnt not turned on the display | 14:31 |
| donihalim | This often happen everytime i reboot the phone or powered on the phone from shutdown state | 14:32 |
| donihalim | Yes im using chimaera -devel and kernel 6.6 | 14:32 |
| donihalim | To workaround this i had to touch the screen after the initial 30s display timeout | 14:33 |
| arno11 | hmm yeah the screen time out could be a problem | 14:34 |
| arno11 | i usually disable it btw | 14:34 |
| arno11 | or set it to 2min | 14:35 |
| arno11 | when you say 'reboot', you mean you reboot the device from command line ? | 14:36 |
| donihalim | Either rebooting from cli org poweroff the phone via power menu | 14:37 |
| arno11 | ok | 14:37 |
| arno11 | did you disable trackers or ? | 14:37 |
| donihalim | Yes i disable tracker and apt worker, following the wiki | 14:38 |
| arno11 | ok | 14:39 |
| arno11 | and sure your fstab is ok ? | 14:40 |
| arno11 | and do you use some custom scripts on boot ? | 14:42 |
| arno11 | anyway, worth a try to set screen time out to 0 or 2 min | 14:47 |
| Wizzup | I don't think it makes sense to disable apt worker | 14:47 |
| arno11 | it is just useful for first boots, need to update wiki | 14:48 |
| donihalim | No custom script and here's fstab https://0x0.st/8QPk.txt | 14:49 |
| donihalim | /swap is being created when dist-upgrade to -devel | 14:49 |
| arno11 | ok | 14:50 |
| donihalim | Wizzup: should i activate it again? How? | 14:50 |
| arno11 | so as i said try to set time out to 0 or 2 min | 14:50 |
| arno11 | for apt worker, same command but with different value | 14:51 |
| arno11 | let me check | 14:51 |
| donihalim | I've set backlight timeout to 2min, see if still causing freeze | 14:52 |
| arno11 | ok | 14:52 |
| arno11 | for apt worker you can set check_interval to 10080 (once a week) | 14:53 |
| arno11 | or 1440 for every day (original value) | 14:53 |
| arno11 | *using the command from wiki | 14:54 |
| donihalim | Will try that, thanks | 14:54 |
| arno11 | np | 14:54 |
| arno11 | Wizzup: on daedalus, no need to deactivate ap worker for sure | 14:56 |
| arno11 | *apt | 14:56 |
| arno11 | but from chimaera img it is useful as many important optimisations are only in devel | 14:57 |
| arno11 | and device is very slooow and absolutely not usable if apt worker start | 14:58 |
| arno11 | anyway, no need now with new imgs | 14:58 |
| sicelo | i still haven't got my libseat problem resolved, so still have never got a successful boot of daedalus. | 15:06 |
| sicelo | Wizzup: any ideas? | 15:06 |
| sicelo | i haven't been able to spend any time on it tbh ... maybe if i do, i can figure it out | 15:06 |
| donihalim | arno11: setting backlight timeout to 2min doesnt solve the issue, i've tried rebooting several time, now the phon freeze again | 15:08 |
| donihalim | Time doesnt change, display not turned off even after 2 min | 15:08 |
| arno11 | weird | 15:08 |
| arno11 | btw you didn't encounter same issue on daedalus, right ? | 15:09 |
| sicelo | i've chrooted to it now... doing apt-upgrade. maybe it'll do a miracle | 15:10 |
| donihalim | arno11: it happen on daedalus aswell | 15:11 |
| arno11 | oh | 15:11 |
| donihalim | Maybe hardware problem? | 15:11 |
| arno11 | really don't know tbh | 15:11 |
| arno11 | checking your Xorg log could be useful | 15:12 |
| arno11 | sicelo: i doubt apt-upgrade could be useful as the img is almost up to date with chimaera -devel | 15:16 |
| sicelo | The following packages will be upgraded: conversations hildon-application-manager hildon-application-manager-settings-standard hildon-input-method hildon-meta hildon-meta-core hildon-meta-n900 libclockcore0-0 libgnutls30 libhildon-im-ui3 libjavascriptcoregtk-4.1-0 libopenh264-7 libtasn1-6 libwebkit2gtk-4.1-0 ofono ofono-scripts openssh-client openssh-server openssh-sftp-server qt-input-maemo | 15:17 |
| sicelo | The following packages will be DOWNGRADED: gtk2-engines-pixbuf libcal1 libcityinfo0-0 libclutter-0.8-0 libgail-common libgail18 libgtk2.0-0 libgtk2.0-bin libgtk2.0-common libiphb0 maemo-keyring | 15:17 |
| sicelo | 20 upgraded, 0 newly installed, 11 downgraded, 0 to remove and 0 not upgraded. | 15:17 |
| sicelo | that downgrade is fine/acceptable? | 15:17 |
| arno11 | yes | 15:17 |
| arno11 | (versioning was wrong, that's why it says downgrade) | 15:19 |
| arno11 | donihalim: ah, i forgot something...let me check | 15:28 |
| arno11 | donihalim: ls /etc/sysctl.d | 15:33 |
| arno11 | do you see n900-perf.conf in red ? | 15:33 |
| donihalim | Let me check | 15:35 |
| Wizzup | arno11: did we not push the no parallel apt change to chimaera? | 15:40 |
| Wizzup | sicelo: libseat problem on what device | 15:40 |
| arno11 | Wizzup: i think we pushed it | 15:41 |
| sicelo | Wizzup: N900 with daedalus image. https://paste.debian.net/1362745/ | 15:41 |
| Wizzup | sicelo: I think there's more problems going on there | 15:42 |
| Wizzup | it looks to be loading swrast | 15:42 |
| Wizzup | [ 72.767] (EE) AIGLX error: dlopen of /usr/lib/arm-linux-gnueabihf/dri/omap_dri.so failed (/usr/lib/arm-linux-gnueabihf/dri/omap_dri.so: cannot open shared object file: No such file or directory) | 15:43 |
| Wizzup | [ 72.767] (EE) AIGLX error: unable to load driver omap | 15:43 |
| sicelo | i think that part was also the same on arno11's N900. | 15:44 |
| arno11 | yes | 15:44 |
| arno11 | and everything is fine | 15:44 |
| arno11 | but could explain qt5 troubles | 15:44 |
| arno11 | *raster vs gpu | 15:45 |
| donihalim | arno11: it's linked to n900-perf.conf.leste | 15:45 |
| arno11 | ok so that's fine | 15:46 |
| arno11 | and what is the content of the file btw ? | 15:46 |
| donihalim | arno11: here https://0x0.st/8QZg.txt | 15:49 |
| arno11 | Wizzup: there is no omap_dri.so indeed | 15:49 |
| sicelo | having chrooted into the sd card, got an excuse to also build upstream iio-sensor-proxy already. so i can now spare my N900 ;-) | 15:49 |
| arno11 | but i supposed it is pvr_dri.so instead, or ? | 15:49 |
| arno11 | donihalim: oh | 15:50 |
| arno11 | so there is a real problem lol | 15:50 |
| arno11 | Wizzup: some important commits are missing in -devel for n900 | 15:51 |
| donihalim | whats the content supposed to be? | 15:51 |
| arno11 | donihalim: i think you should comment all lines in this file and reboot :) | 15:52 |
| Wizzup | arno11: -devel being chimaera-devel yeah | 15:52 |
| Wizzup | sicelo: let me check my n900 | 15:52 |
| arno11 | this file is supposed to be removed or empty | 15:52 |
| arno11 | Wizzup: yep, old vm tweaks are still there | 15:53 |
| donihalim | Should remove the link or? | 15:53 |
| arno11 | so no surprise if donihalim device is freezing after boot | 15:53 |
| arno11 | donihalim: just comment all lines | 15:53 |
| arno11 | *# | 15:53 |
| Wizzup | sicelo: yes, ok, I see, the omap error I also have but it seems to be ok | 15:54 |
| Wizzup | I don't have this libseat issue, I wonder if it is a race condition somehow | 15:54 |
| Wizzup | I guess I should flash the latest n900 image | 15:54 |
| donihalim | Glad it's not hardware issue :) | 15:55 |
| arno11 | yes indeed :) | 15:55 |
| donihalim | I'll rebooting several time, i'll report back | 15:56 |
| arno11 | i need to check one more thing because maybe another file is problematic | 15:56 |
| arno11 | donihalim: do you have 'leste-config-n900' in /etc/init.d/ ? | 15:58 |
| arno11 | if yes, delete it or move it to $HOME | 15:59 |
| donihalim | Yes it present in /etc/init.d | 16:01 |
| donihalim | I'll move it to home | 16:02 |
| arno11 | ok so that is the real culprit | 16:02 |
| arno11 | ok | 16:02 |
| arno11 | it should work fine now on next reboot | 16:02 |
| donihalim | Nice, many thanks arno11 :) | 16:04 |
| arno11 | Wizzup: commits eaf6d22 is missing in chimaera, don't know why | 16:05 |
| sicelo | chimaera, or daedalus? also what repo? | 16:06 |
| arno11 | (it reverts problematic nr_request and vm stuff on boot) | 16:06 |
| arno11 | for daedalus it seems ok | 16:06 |
| arno11 | so that's weird | 16:07 |
| arno11 | leste-config repo | 16:08 |
| Wizzup | arno11: ok, I can fix that today, so I just need to check what is missing in chimaera? | 16:11 |
| arno11 | yes, in leste config | 16:11 |
| arno11 | seems ok in daedalus as i've never seen those files | 16:12 |
| arno11 | donihalim: np, but many thx to YOU for the feedback | 16:13 |
| sicelo | i asked about my libseat issue in the libseat room, and they said it's failing during shutdown ... but i'm booting :-p | 16:15 |
| sicelo | so that suggests that the libseat errors are not the real culprit. i need to find who/what initiates a shutdown | 16:15 |
| sicelo | oh i'm dumb ... i think it's upower :-D | 16:16 |
| arno11 | oh | 16:16 |
| Wizzup | that's quite possible | 16:18 |
| Wizzup | I dist-upgraded so I still have the old upower | 16:18 |
| Wizzup | you can -x upowerd and see if it works | 16:18 |
| arno11 | but why no issue on my device ? (daedalus same img) | 16:19 |
| sicelo | arno11: my device has broken USB, so % readout in /sys/class is always 0% | 16:20 |
| arno11 | ah ok | 16:20 |
| sicelo | it can happen to you as well if you remove the battery for a long time, since the fuel gauge will get reset. if you try and boot again, i'm sure you'll have same experience as i, assuming you don't connect to a charger | 16:21 |
| arno11 | ok i see | 16:21 |
| sicelo | upower from leste/chimaera made some estimations based on voltage | 16:24 |
| sicelo | and yay, i'm in Hildon now | 16:26 |
| Wizzup | great | 16:27 |
| arno11 | yeah cool | 16:28 |
| arno11 | welcome in daedalus ! | 16:29 |
| sicelo | with my subpar SD card, can't do anything yet. | 16:31 |
| sicelo | arno11: so (1) swap, (2) that /etc/init.d/leste-config-n900, what else? | 16:31 |
| arno11 | you have also leste-config-n900 in daedalus ? | 16:32 |
| sicelo | no idea. haven't looked yet. i'm just asking for what fixes i need to make in order to be able to use this ... | 16:32 |
| sicelo | my card isn't of the best quality, so yeah | 16:33 |
| Wizzup | I think it's just swap | 16:34 |
| arno11 | ok, so you should not have troubles with leste config files but yes you need sd swap | 16:34 |
| arno11 | if your sd is slow, you should block tracker3 or at least add .nomedia in $HOME | 16:36 |
| arno11 | and use 500MHz min freq | 16:36 |
| arno11 | that's it | 16:37 |
| Wizzup | I don't block tracer3 on my n900 fwiw | 16:37 |
| arno11 | yes i know but your card is fast :) | 16:38 |
| sicelo | i didn't have it blocked in chimaera, nor played with frequencies | 16:38 |
| arno11 | that's just suggestions | 16:39 |
| sicelo | still struggling to get to terminal to even do expandcard :-D | 16:39 |
| arno11 | that's normal | 16:39 |
| arno11 | first boot is a pita with slow sd | 16:39 |
| arno11 | should work fine on next reboot | 16:40 |
| sicelo | oh wait ... maybe it's so dogslow because it's still trying to show that first boot thing ... still hasn't shown btw, and that's 14 min after :-D | 16:40 |
| arno11 | omg | 16:40 |
| arno11 | yeah that's the problem | 16:40 |
| Wizzup | top should probably show what it is doing | 16:41 |
| arno11 | btw on next boots the only very slow thing should be apt and ham (if the card is not fast enough) | 16:42 |
| sicelo | thank God for lp5523 ... at least the white LED gives consolation | 16:45 |
| Wizzup | :) | 16:45 |
| arno11 | lol | 16:45 |
| sicelo | I'm more excited to see my iio-sensor-proxy working in Leste ... so screen will now get blanked during a phone call | 17:45 |
| sicelo | I'll submit MR later today hopefully | 17:47 |
| sicelo | if time permits, next stop should be upower ... we need to fix our stack | 17:48 |
| Wizzup | thank you for all the energy/efforts :) | 17:54 |
| arno11 | cool @iio-sensor, will save at least 50mA | 18:30 |
| arno11 | during calls | 18:30 |
| arno11 | 70mA saved in fact with brightness @50 | 18:37 |
| arno11 | (on n900 calls actually burn around 350mA !) | 18:39 |
| sicelo | oh wow! didn't realize that | 18:42 |
| Wizzup | the cpu routing part of it probably causes a bunch of that | 18:47 |
| Wizzup | pipewire might make some of that a bit better | 18:47 |
| sicelo | i have given pipewire some thought for sure ... but i guest we will need to write a pipewire component for cmtspeech :-/ | 18:50 |
| sicelo | unless we go cmtspeech -> pulse -> pipewire | 18:50 |
| arno11 | cmtspeech -> pulse -> pipewire should be easy, there is module for that iirc | 20:10 |
| arno11 | but anyway, i doubt we can improve cpu usage, just my opinion | 20:10 |
| arno11 | unless we find a way to use dsp chip... | 20:11 |
| sicelo | i really wish so too (dsp) | 20:28 |
| arno11 | oh now i remember what fmg said about cmt_pulse cpu usage: some neon code is missing for resampling 8000 -> 48000Hz iirc | 20:39 |
| arno11 | and 48000 to 8000 in the other way | 20:40 |
| arno11 | sicelo: yeah definitely @dsp | 20:41 |
| sicelo | @ neon code ... i think it's in one of the repositories. have never looked at it closely though | 20:46 |
| arno11 | yes indeed iirc. the answer is probably in irc logs | 20:47 |
| arno11 | if nobody has spare time for that, maybe AI could help | 20:50 |
| arno11 | i mean for neon code | 20:51 |
| sicelo | dsc_: have you ever used the calendar in Leste? any idea where one can look if they wanted to fix the layout/display issue, namely, the week display on the left is 6 rows, but the actual days/dates are 5 rows.. | 20:57 |
| dsc_ | sicelo: my life is so boring I dont need a calendar | 21:12 |
| dsc_ | afaik Wizzup uses it quite a bit | 21:12 |
| dsc_ | do you have a link to the repo? | 21:12 |
| sicelo | dsc_: heh, i figured :-D | 21:21 |
| sicelo | https://github.com/maemo-leste/qalendar/ | 21:21 |
| dsc_ | sicelo: its widgets, so .ui files | 21:28 |
| dsc_ | you can open them in Qt Designer | 21:28 |
| dsc_ | you'll have to find the widget/.ui file thats responsible | 21:28 |
| dsc_ | sometimes you can just grep if there is known text | 21:29 |
| sicelo | i have looked at it before, but will try again sometime soon | 21:29 |
| Wizzup | yeah I use qalendar quite a bit | 22:36 |
| Wizzup | with syncevo | 22:36 |
| dsc_ | freemangordon: around? | 22:50 |
| dsc_ | nvm fixed it | 22:51 |
Generated by irclog2html.py 2.17.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!