| freemangordon | Wizzup: kernel repo does not seem migrated to git.maemo.org | 07:13 |
|---|---|---|
| dsc_ | also missing jib | 07:15 |
| freemangordon | oh, it is in forks :) | 07:35 |
| Wizzup | freemangordon: yeah it is in forks | 08:36 |
| Wizzup | dsc_: yes extras aren't yet, maybe I should do that today | 08:36 |
| Wizzup | freemangordon: seemed more appropriate | 08:36 |
| freemangordon | ok, I was just not used to | 08:43 |
| Wizzup | yeah | 08:43 |
| Wizzup | there's 4-5 repos that I moved this way | 08:43 |
| Wizzup | the rest are as expected | 08:43 |
| freemangordon | ok | 08:49 |
| Wizzup | mostly just the linux repos, and gtk | 08:50 |
| Wizzup | gtk (2) I also moved to forks | 08:50 |
| freemangordon | Wizzup: hmm, is there a way to tell openrc to reboot if a service fails? | 10:37 |
| Wizzup | reboot the os? | 10:43 |
| freemangordon | yeah | 10:44 |
| freemangordon | ok, AI helped me | 10:44 |
| Wizzup | freemangordon1: what was the solution | 11:24 |
| freemangordon1 | will see, lemme test it a bit and will push | 11:24 |
| freemangordon | hmm, wait, will not work in our openrc | 11:27 |
| freemangordon | maybe in trixie | 11:27 |
| freemangordon | otherwise there is --retry_failed option to supervise-daemon | 11:27 |
| Wizzup | what are you looking for | 11:35 |
| Wizzup | trixie has openrc start-stop-daemon btw, whereas bookworm does not | 11:35 |
| dsc_ | Wizzup: hi | 11:37 |
| dsc_ | To git.maemo.org:leste/qt-input-maemo.git ! [remote rejected] master -> master (pre-receive hook declined) | 11:38 |
| dsc_ | error: failed to push some refs to 'git.maemo.org:leste/qt-input-maemo.git' | 11:38 |
| dsc_ | oh | 11:38 |
| dsc_ | remote: Forgejo: User permission denied for writing. | 11:38 |
| Wizzup | dsc_: force push to master fails? let me check\ | 11:40 |
| dsc_ | its not force | 11:40 |
| Wizzup | dsc_: let me see | 11:40 |
| Wizzup | you're using this yeah? git@git.maemo.org:leste/qt-input-maemo.git | 11:40 |
| dsc_ | yeah | 11:41 |
| Wizzup | does pull work? | 11:41 |
| Wizzup | oh nvm | 11:41 |
| Wizzup | I get it now | 11:41 |
| Wizzup | try now | 11:41 |
| dsc_ | thx worked | 11:41 |
| Wizzup | freemangordon: we should figure out how set up perms/teams | 11:41 |
| Wizzup | freemangordon: like, do we just give folks push access to core, or do we stick to per repo mostly? I am fine with either | 11:43 |
| freemangordon | Wizzup: I think we shale create at least one team for those that work on core repos and give them push access. for the others it is better 'per repo' | 12:33 |
| Wizzup | ok | 12:33 |
| freemangordon | it is another question who those members will be :) | 12:33 |
| Wizzup | I think we're at that point | 12:34 |
| Wizzup | we can deal with it later in any case, there is a group of admins (you, me) | 12:34 |
| Wizzup | and I made a developers 'team' but we're not using it atm | 12:34 |
| Wizzup | I'm ok with principle of least priv / doing it specifically per repo | 12:34 |
| freemangordon | mhm | 12:35 |
| freemangordon | make sense | 12:35 |
| Wizzup | ok, so should I just move over the extras repos? I'm wondering how to let the extras people know | 12:35 |
| Wizzup | I guess I can make a news post about it, separate from the daedalus one | 12:35 |
| freemangordon | don;t they have e-mails? | 12:35 |
| Wizzup | maybe, but github doesn't work around that necessarily | 12:39 |
| freemangordon | github? I meant to send emails to them on behalf of "maemo leste" | 12:41 |
| freemangordon | or "maemo leste admins' or whatever | 12:41 |
| freemangordon | informing the repo has moved | 12:41 |
| Wizzup | I understand, but not every users email is exposed | 12:42 |
| Wizzup | I think making an issue in the repos could work too | 12:42 |
| freemangordon | or making them read-only | 12:44 |
| Wizzup | damn we have 122 extras repos? :D | 12:44 |
| freemangordon | ugh :) | 12:44 |
| Wizzup | I'll just migrate them now | 12:45 |
| crab | hi all, is maemo-leste / devuan still on beowulf or am i behind the times? | 12:52 |
| crab | it looks like im 2 releases behind the times | 12:52 |
| freemangordon | it is on chimaera and we will move to daedalus as soon as we fix the few remaining issues | 12:55 |
| crab | im going to try dist-upgrading to chimaera | 12:56 |
| crab | we'll see what happens... | 12:56 |
| Wizzup | yeah definitely not on beowulf :) | 13:00 |
| freemangordon | arno11: please upgrade kernel and pull latest procmngd, it should work now | 13:47 |
| arno11 | freemangordon: ok, let's go | 14:50 |
| arno11 | last procmngd pulled, upgrading kernel... | 14:58 |
| arno11 | reboot | 15:00 |
| crab | *sigh* | 15:13 |
| arno11 | weird, linux-image-omap on my device is 6.6.58 but uname -r still returns 6.6.53 | 15:14 |
| * crab has totally forgotten how the bootloader works, what the bootloader is called, where the boot loader config is held, the layout of all the storage, whether i need to even do anything if this dist-upgrade actually works | 15:14 | |
| crab | "cat > /etc/bootmenu.d/30-maemo-leste.item" that kind of thing is giving me intense deja vu | 15:14 |
| crab | but i havent yet found a partition with anything like that on it! :P | 15:15 |
| arno11 | freemangordon: ^^^ | 15:25 |
| arno11 | something seems wrong with kernel upgrade | 15:26 |
| arno11 | freemangordon: well, so i can't test procmngd ATM. my kernel is still 6.6.53 after upgrade (linux-image-omap 6.6.58) | 15:57 |
| arno11 | Wizzup: ^ any hint/tips ? | 15:58 |
| Wizzup | extras repos are migrated, will set up push mirror | 16:42 |
| freemangordon | arno11: are you an daedalus? | 16:48 |
| freemangordon | ugh, wait | 16:49 |
| freemangordon | it is the same here | 16:49 |
| freemangordon | but you should be ok re procmngd | 16:49 |
| freemangordon | Wizzup: there is something wring with kernel version | 16:49 |
| freemangordon | *wrong | 16:50 |
| freemangordon | changelog says 58, but that's not true | 16:50 |
| freemangordon | ieah, I wonder where this .58 came from | 16:52 |
| freemangordon | *yeah | 16:52 |
| freemangordon | arno11: so, you *can* test | 16:53 |
| Wizzup | freemangordon: did you tag? | 16:54 |
| Wizzup | push mirror set up too | 16:59 |
| freemangordon | yes, I did | 17:16 |
| freemangordon | it is just that changelog lies | 17:16 |
| freemangordon | see maemo 6.6.y history | 17:16 |
| freemangordon | latest upstream commit is https://git.maemo.org/leste-upstream-forks/droid4-linux/commit/4ad9fa5c30edc19acf05b2960dd686c29cbe75a2 | 17:17 |
| Wizzup | I am confused, but it sounds like you got it ;) | 17:19 |
| Wizzup | I normally add an extra .[number] | 17:19 |
| Wizzup | to prevent this issue | 17:19 |
| freemangordon | Wizzup: umm, ok, lemme try to explain: | 17:41 |
| freemangordon | version in changelog is 6.6.58.4-1+4m7 | 17:41 |
| freemangordon | but *kernel* version is 6.6.53 | 17:42 |
| Wizzup | right, why don't we use 6.6.53.4? | 17:43 |
| Wizzup | or whatever | 17:43 |
| freemangordon | well, don't as me ;) https://git.maemo.org/leste-upstream-forks/droid4-linux/commits/branch/maemo/daedalus | 17:47 |
| freemangordon | *ask | 17:47 |
| Wizzup | looks like I made the mistake too | 17:50 |
| Wizzup | :D | 17:50 |
| Wizzup | I'm confused now | 17:50 |
| freemangordon | the first bad commit is https://git.maemo.org/leste-upstream-forks/droid4-linux/commit/dfabfe295d358697a0201240eeff8fe0acf73f28, me and you just increased the subversion | 18:09 |
| arno11 | freemangordon: '@i can try', ok but it didn't work this afternoon. let me try again | 19:11 |
| arno11 | hmm, if i stop procmngd and start it again, it says 'supervise-daemon: fopen `/run/procmngd.pid': No such file or directory' | 19:15 |
| freemangordon | arno11: I tested it on d4 and can confirm that it works | 19:15 |
| arno11 | ah | 19:15 |
| freemangordon | also, check syslog for messages | 19:15 |
| arno11 | syslog returns nothing unfortunately | 19:16 |
| arno11 | in fact i can't get any useful log. i can just confirm that the daemonis running | 19:16 |
| arno11 | *is | 19:17 |
| arno11 | let me check again | 19:17 |
| freemangordon | ok, stop the service and run the binary by hand, as rootg | 19:17 |
| freemangordon | *root | 19:17 |
| arno11 | ok | 19:17 |
| freemangordon | /usr/sbin/procmngd -l debug | 19:17 |
| arno11 | well, sure you don't load any additional module i'm not aware of or something ? | 19:25 |
| freemangordon | yes, what is needed is built-in | 19:30 |
| freemangordon | so, could you provide the output of the above command? | 19:31 |
| freemangordon | also, make sure the service is stopped | 19:31 |
| arno11 | it just returns Logging to stderr | 19:31 |
| arno11 | procmngd started | 19:31 |
| freemangordon | what is the result of dpkg -l | grep omap | 19:32 |
| arno11 | 6.6.58.4-1+4m7 | 19:33 |
| freemangordon | that's correct | 19:33 |
| freemangordon | what is the version of procmngd? | 19:34 |
| arno11 | 1.0.0 | 19:35 |
| freemangordon | so, while "/usr/sbin/procmngd -l debug" is running, what happens if you open another ssh session? | 19:36 |
| freemangordon | don't you see stuff like: | 19:36 |
| freemangordon | EXEC(8110): | 19:36 |
| freemangordon | exe: /usr/bin/dpkg-deb | 19:36 |
| freemangordon | cmdline: dpkg-deb | 19:36 |
| freemangordon | ? | 19:36 |
| arno11 | yeah i see them | 19:36 |
| freemangordon | so, everything works fine | 19:37 |
| freemangordon | what do you have in /etc/procmngd.d/ ? | 19:37 |
| arno11 | the same json you provided yesterday | 19:37 |
| arno11 | in a .cfg file | 19:38 |
| freemangordon | I don;t remember, please pastebin the output | 19:38 |
| arno11 | { | 19:40 |
| arno11 | "/usr/bin/top": { | 19:40 |
| arno11 | "nice": 10 | 19:40 |
| arno11 | } | 19:40 |
| arno11 | } | 19:40 |
| freemangordon | ok, stop the binary and start it with -l info: | 19:41 |
| freemangordon | /usr/sbin/procmngd -l info | 19:41 |
| arno11 | ok | 19:41 |
| freemangordon | and start top in another session | 19:41 |
| arno11 | ah it fails parsing the json | 19:42 |
| arno11 | let me check | 19:42 |
| arno11 | it skips the config file | 19:42 |
| freemangordon | fix it | 19:43 |
| Wizzup | you can use something like json_pp < foo.json to tset | 19:44 |
| Wizzup | test* | 19:44 |
| arno11 | it s ok now, let's try nicing something | 19:47 |
| arno11 | yeah, now it works :) the problem was just the json | 19:51 |
| arno11 | diner time, but this night i will try to 'nice' H-D and other stuff on boot and let you know guys | 19:53 |
| freemangordon | Wizzup: does it make sense to implement ionice support as well? | 19:53 |
| freemangordon | arno11: I am not ure it is a good idea to renice h-d, we'd better use cgroups for it | 19:54 |
| arno11 | no worries it already works with hack | 19:54 |
| freemangordon | see https://paste.debian.net/1368072/ | 19:55 |
| arno11 | but sure, better use cgroup | 19:55 |
| freemangordon | this is all that is 'niced' on fremantle | 19:55 |
| arno11 | ok ty | 19:57 |
| freemangordon | I have to implement command line matching as well | 19:58 |
| arno11 | freemangordon: ok i set cfg files based on your fremantle nice script (excepting for PA as it is already -11 by default). need to reboot, let's see. | 21:49 |
| freemangordon | keep in mind that re-nicing mafw-dbus-wrapper is not that simple | 21:52 |
| freemangordon | because the are 3 of it :) | 21:52 |
| freemangordon | tat's why command line mathing has to be implementd | 21:52 |
| freemangordon | to match only mafw-gst-renderer | 21:52 |
| sicelo | making MR's to some projects is always 'fun' ... for some reason i have easier time with kernel than the userspace stuff :p | 22:05 |
| sicelo | so with upower, i initially made one big commit that had the changes -- i think it was ok because while it spanned across 4 files, it was all really one logical change | 22:05 |
| sicelo | anyway, they said to split it to one commit per affected file, which i did. but i only gave a full description for the first commit in the 'series' and now they want full descriptions on all of the 4 commits. i don't know what to write :-D | 22:07 |
| sicelo | since it's still one thing that i've done, heh. not sure if i should copy the description from the other commits into the remaining 3 | 22:07 |
| sicelo | anyhow, at least it seems they're happy with the code itself, and with some luck, the thing could get merged and even a release tagged soon | 22:09 |
| arno11 | freemangordon: yeah @mafw | 22:18 |
| freemangordon | I am on it | 22:19 |
| arno11 | ok | 22:19 |
| arno11 | so to resume: | 22:20 |
| arno11 | we already have xorg and mce renicing on boot, PA is already @-11, mafw needs work, and i can't find a way to 'nice' dsme and tonegend | 22:22 |
| arno11 | with the few processes i m able to nice, things semm a bit slower | 22:22 |
| arno11 | *seem | 22:22 |
| sicelo | not so 'nice' :-D | 22:23 |
| arno11 | the best results i got with leste during the last year was: H-D -8, Xorg -8, mce -5 | 22:23 |
| arno11 | (that's it, no other nicing) | 22:24 |
| arno11 | sicelo: lol | 22:24 |
| arno11 | but anyway, the most important things are that procmngd works fine and we can use it with no issue on n900: low mem usage and no 'visible' cpu usage | 22:28 |
| arno11 | well, as soon as i renice H-D to -8 again, things are very responsive again | 22:37 |
| freemangordon | dsme should already be at -5 | 22:58 |
| freemangordon | we also need to renice dbus | 22:58 |
| sicelo | interesting ... someone submitted a patch to ofono using a tesla.com address. wonder if that could hint at tesla using ofono in their stack (as opposed to ModemManager) | 23:03 |
| freemangordon | oh, if that's the case I guess I can ask for a profit share :D | 23:05 |
| freemangordon | you too | 23:05 |
| sicelo | hehe | 23:06 |
| sicelo | btw, maybe you can have a quick look at this one if you have a brief moment - https://paste.debian.net/hidden/f2c3205c/ ... does my patch look correct/valid? | 23:07 |
| freemangordon | sorry, this will need time I don't have now | 23:10 |
| freemangordon | maybe tomorrow morning | 23:10 |
| sicelo | sure. no rush. i actually did this about 2 months ago already :p | 23:11 |
| sicelo | i've actually just submitted it to the list after all. it seems correct (asked AI to verify it) | 23:54 |
| freemangordon | cool | 23:54 |
| freemangordon | arno11: please pull latest code and maybe integrate https://paste.debian.net/1368129/ with your current config | 23:55 |
Generated by irclog2html.py 2.17.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!