| Wizzup | https://git.maemo.org/explore/organizations | 00:01 |
|---|---|---|
| Wizzup | freemangordon: leste, leste-upstream-forks, leste-extras ok? (organisations) | 00:01 |
| freemangordon | mkf: re dts freq - because in fex that device has freq enabled that are higher than defaults in dts that's upstream | 07:31 |
| freemangordon | so it was running underclocked | 07:32 |
| freemangordon | https://paste.debian.net/1365329/ | 07:34 |
| sicelo | arno11: upower reports discharge rate as 215mW on my idling N900. is it typical? | 08:28 |
| sicelo | ah, looks good ... seems to be around 55mA. i guess it's acceptable | 08:29 |
| mkf | freemangordon: could you either send your kernel build or see how much power does you a33 sips (i test via sensors command, but there might be other ways) | 09:18 |
| sicelo | Wizzup: arno11: regarding the 'wrong | 09:56 |
| sicelo | 'wrong' batteries and chargers on N900, https://github.com/maemo-leste/leste-config/pull/58 | 09:56 |
| sicelo | freemangordon: Wizzup: how would i debug status-area-applet-battery since it's an .so? my changes don't seem to have any effect, for some reason i can't fathom | 10:14 |
| sicelo | specifically, i'd like to see why batmon.c isn't doing what i'd expect | 10:15 |
| Wizzup | you cna use maemo-invoker / maemo-summone and gdb | 11:08 |
| Wizzup | https://leste.maemo.org/Debugging#Dealing_with_Maemo_Launcher_.2F_Maemo_Invoker | 11:08 |
| Wizzup | it specifically mentions status menu :) | 11:08 |
| sicelo | ty! i was search for that since i do remember reading it before. but my google-fu didn't help | 11:11 |
| arno11 | sicelo: yeah, 55mA is acceptable. however, it usually fluctuates a bit between 44 and 56mA | 11:14 |
| Wizzup | we need to work on our developer docs :) | 11:14 |
| Wizzup | freemangordon: any objection to me working on migrating all core repos today? nothing will really change initially, but it'd be nice to push to both repos if there are additional changes | 11:15 |
| Wizzup | we can also do say extra pkgs first[1;5D[1;5D[1;5D | 11:15 |
| sicelo | Wizzup: mmm, my lack of gdb-fu is catching up with me. i can start h-s-m as indicated on that page, but how do i actually gain access to what status-area-applet-battery:batmon.c is doing? e.g. to set a breakpoint so i can inspect variables at a given point in the execution | 11:25 |
| mkf | you could put a bp at open() i guess? | 11:36 |
| mkf | and write/read | 11:36 |
| mkf | but that might be too low level, applet might not be that low level | 11:37 |
| Wizzup | sicelo: get dbgsym installed, then you can do break source_file.c:lineno | 11:40 |
| Wizzup | or break <function name> | 11:40 |
| sicelo | thanks. seems DISPLAY=:0 gdb --symbols=/usr/lib/debug/.build-id/82/906f43adfb5d228516e676db2ce4a5f26688fe.debug --readnow --args maemo-summoner /usr/bin/hildon-status-menu.launch works | 11:50 |
| sicelo | oh nice, at least my mce stuff is working nicely. once capacity_level reported "Low" the device just started shutting down :p | 11:56 |
| sicelo | mmm, or not? i wonder what did shut it down, because it's not upower, and i didn't configure mce.conf for that on my device yet | 12:00 |
| Wizzup | magic :) | 12:01 |
| Wizzup | I suspect /var/log/ will contain some info | 12:01 |
| sicelo | 2025-03-26T12:55:47.471253+02:00 localhost systemui[2813]: systemui: shutdown_ind from DSME, quitting | 12:08 |
| sicelo | interesting. wonder where/how dsme got that from | 12:08 |
| sicelo | ah, mce.ini:CriticalVoltage | 12:10 |
| mkf | cpufreq works. | 12:24 |
| mkf | but power drain still remains unsolved. | 12:24 |
| freemangordon | mkf: don't expect too much from allwinner | 12:50 |
| freemangordon | I'd say 300 mA idle is ok | 12:50 |
| freemangordon | lemme check what mine is doing | 12:50 |
| freemangordon | there is a reason I replaced original 2300mAh battery with 3x1800 mAh :D | 12:52 |
| sicelo | heh, 3x? i.e. you stacked them? | 12:53 |
| freemangordon | connected in parallel | 12:53 |
| freemangordon | POWER_SUPPLY_CURRENT_NOW=-388000 | 12:54 |
| freemangordon | that's more or less normal | 12:55 |
| mkf | eh | 12:55 |
| sicelo | yeah, SoC issue ... same issue with Pinephone and Librem5 | 12:56 |
| freemangordon | mhm | 12:56 |
| mkf | so allwinner should last less then n900 | 12:57 |
| freemangordon | times less | 12:57 |
| mkf | even though it's battery can 5 times as large? | 12:57 |
| freemangordon | mhm | 12:57 |
| mkf | sux. :( | 12:57 |
| sicelo | and ... bull's eye! the status menu stuff works good now :-) | 12:57 |
| freemangordon | :) | 12:57 |
| mkf | congrats. | 12:57 |
| sicelo | will update the code, and submit MRs | 12:57 |
| freemangordon | sicelo: did you see my comment? | 12:58 |
| sicelo | yes, i did. thanks. there were lots of other problems too, which the compiler complained about | 12:58 |
| freemangordon | ok | 12:58 |
| arno11 | sicelo: oh great @h-s-m ! | 13:01 |
| mkf | wait tho. | 13:01 |
| mkf | this isn't sipping 300mA | 13:01 |
| mkf | it's sipping 1mA | 13:01 |
| mkf | *1A | 13:01 |
| mkf | something is clearly wrong... | 13:01 |
| arno11 | sicelo: ty btw for blacklist, now battery applet is ok and orange led as well | 13:02 |
| arno11 | mkf: wow @1A | 13:02 |
| mkf | freemangordon: idle as in screen off? | 13:03 |
| mkf | or screen on and doing nothing? | 13:03 |
| sicelo | arno11: ah, so that's why the last applet and led work didn't help you ... guess it was working for me because i'd done the blacklist correctly, but submitted garbage :p | 13:07 |
| arno11 | sicelo: :P | 13:26 |
| arno11 | sicelo: hm it doesn t work each time i plug on charger: sometimes charging notif appears immediately and it starts charging (with led working), and sometimes it starts charging after few sec with no notif and no led | 13:31 |
| arno11 | if i unplug/plug charger, it works again (100% of time apparently) | 13:33 |
| arno11 | hm not 100% | 13:35 |
| arno11 | maybe 60% of time | 13:36 |
| sicelo | mmm, maybe our patches aren't all in place | 13:37 |
| arno11 | yeah maybe | 13:41 |
| sicelo | because on my previous (chimaera) image, it was working 100% | 13:42 |
| sicelo | will re-test after i'm done with this upower work | 13:42 |
| arno11 | ok | 13:43 |
| donihalim | friends laugh at me for using n900, until they receive whatsapp message from it | 13:50 |
| arno11 | hehe :) | 13:56 |
| sicelo | freemangordon: Wizzup: MR's are up now, https://github.com/maemo-leste/status-area-applet-battery/pull/8/files | 14:04 |
| sicelo | this one doesn't really depend on the patched upower, since it just looks at the 'voltage' property of UPower, which is available in any version. | 14:06 |
| mkf | what do they do? | 14:06 |
| sicelo | our long term plan is to drop the patched upower that we have ... which estimates the percentage from voltage if the fuel gauge does not report it | 14:07 |
| sicelo | so we're trying to push some changes upstream ( https://gitlab.freedesktop.org/upower/upower/-/merge_requests/259/ ). since we will not be estimating in upower anymore, we need alternative for (1) ensuring device shuts down at appropriate time, and (2) we display some reasonable icon for the user. | 14:09 |
| sicelo | this patch does (2) above | 14:09 |
| mkf | yay | 14:09 |
| sicelo | freemangordon: Wizzup: for mce, https://github.com/maemo-leste/mce/pull/63/ ... | 14:10 |
| sicelo | mkf: this one does (1) from the list above :-) | 14:10 |
| mkf | awesome | 14:11 |
| Wizzup | any ideas on how to best coordinate the migration to git.maemo.org, so that we don't have to manually try to keep thing in sync? | 14:18 |
| Wizzup | I'd like to leave the gh repos open / as they are, but now sure how to make sure people go to the right place then | 14:19 |
| Wizzup | (or in case we want to switch back, then it's nice to leave them open too) | 14:19 |
| mkf | maybe a git hook? | 14:21 |
| sicelo | i think GH supports setting repos as archive? | 14:22 |
| Wizzup | I don't really want to archive them | 14:24 |
| Wizzup | we can have git.maemo.org push any changes to them | 14:24 |
| Wizzup | but like for new pull requests, issues etc, it's better to have it on git.maemo.org | 14:24 |
| Wizzup | I think there's value in keeping the repos synced | 14:25 |
| Wizzup | I don't think we can do it when it is an arhive | 14:25 |
| mkf | you can't undo a archive iirc | 14:26 |
| mkf | and yes, i think if leste or maemo.org ever dies, having sources in github would be helpful for future people | 14:26 |
| sicelo | ok. should be possible to disable creation of new MR's and Issues | 14:26 |
| mkf | (it's really hard to find stuff about cordia or other early n900 related projects) | 14:27 |
| Wizzup | I mean I can just at least attempt to move all repos over now, as long as the folks with push access keep it in mind? | 14:38 |
| Wizzup | any objections to me starting with this today? | 14:42 |
| sicelo | maybe check if fmg also agrees :-) | 14:44 |
| Wizzup | freemangordon: ^ | 14:46 |
| Wizzup | in any case I can start by just syncing all the repos, this will take hours (not my time, computer time), and then we can re-clone some if they changed | 14:46 |
| arno11 | sicelo: Wizzup: is it normal we use upstream upower in daedalus and not in chimaera -devel ? | 15:01 |
| sicelo | well, i'd say yes :-) | 15:03 |
| arno11 | ok | 15:03 |
| sicelo | for one, in chimaera, we would have no hope to even run upstream upower, due to glib and gudev versions | 15:03 |
| arno11 | ah ok | 15:04 |
| sicelo | even in daedalus, i found that to build 1.90.7, i had to downgrade gudev requirement by two versions. upstream actually wants 238, but daedalus only has 237 | 15:05 |
| sicelo | s/two versions/one version/ | 15:06 |
| arno11 | ok | 15:07 |
| sicelo | i find the N900 too addictive :p | 15:12 |
| sicelo | i had wanted to spend time on the L5 port, but N900 took over | 15:13 |
| freemangordon | Wizzup: is short, do you ask me if I am ok with start using git.maemo.org instead of github? | 15:13 |
| arno11 | sicelo: i perfectly know that feeling unfortunately :) | 15:16 |
| Wizzup | freemangordon: yes but I also want to find a good time to do it | 15:20 |
| Wizzup | so that I don't sync everything, then we push to github and wonder if any code is lost | 15:20 |
| arno11 | Wizzup: sicelo: @calls issue, PA returns 'memblock.c : pull full, and killed at the begining of the call. so i suppose it is low on memory but why ? | 15:56 |
| arno11 | ah it says sink.c: Cannot update sample spec, SINK_IS_RUNNING, will keep using s16le and 48000 Hz | 15:58 |
| arno11 | and module-suspend-on-idle.c: Sink alsa_output.platform-n900-audio.Voice_Call__hw_RX51_0__sink becomes busy, resuming. | 16:00 |
| arno11 | weird that the issue only happens with ringtone, not with other kind of sound | 16:12 |
| gnarface | maybe it's something specific about the format of that ringtone file? | 16:13 |
| gnarface | maybe it's a weird format? | 16:13 |
| arno11 | that's the original one | 16:13 |
| arno11 | but we always had issues with nokia ringtones btw | 16:14 |
| gnarface | well i presume it's not S16_LE or 48000 Hz | 16:14 |
| gnarface | and/or | 16:14 |
| Wizzup | https://pulseaudio-discuss.freedesktop.narkive.com/tmXaf4Wq/lots-of-null-sink-memblock-c-pool-full | 16:14 |
| arno11 | ty | 16:15 |
| Wizzup | there is no solution in there :) | 16:15 |
| arno11 | ok, fair enough, vibrator works fine | 16:15 |
| arno11 | :) | 16:16 |
| Wizzup | no, I mean, we do want to solve this | 16:18 |
| arno11 | yes ofc lol | 16:18 |
| Wizzup | didn't we increase the amount of ... memfds or so that pa could open? | 16:18 |
| Wizzup | something with shm at least | 16:18 |
| Wizzup | I wonder if that maybe got lost somehow | 16:18 |
| Wizzup | I can't wait to switch to pipewire | 16:19 |
| Wizzup | lol | 16:19 |
| arno11 | yeah i tried to modify shm but same result. i didn t try to decrease it btw | 16:19 |
| Wizzup | ok | 16:19 |
| arno11 | yeah @pipewire | 16:19 |
| Wizzup | freemangordon: I'll add droid4-linux and see how that goes, since it's a huge repo | 16:25 |
| freemangordon | ok | 17:05 |
| freemangordon | otherwise I am fine | 17:05 |
| Wizzup | ok | 17:09 |
| mkfx | freemangordon: sorry to mention you again, can you specify how you tested your sunxi? With screen off? | 17:12 |
| freemangordon | mhm | 17:17 |
| freemangordon | through ssh | 17:17 |
| freemangordon | if by "tested" you mean getting power usage | 17:17 |
| mkfx | Could you also test with screen on, wifi on, max backlight? | 17:18 |
| freemangordon | sec | 17:18 |
| freemangordon | -935000 | 17:19 |
| mkfx | So 1A | 17:20 |
| mkfx | Very well. I thought something was wrong with my dtb | 17:21 |
| freemangordon | yeah, allwinner is a hungry beast | 17:26 |
| Wizzup | I'd be very happy if we can get some sunxi/allwinner images going again | 17:27 |
| freemangordon | me too | 17:27 |
| Wizzup | I have so much sunxi hardware and I'd still like to a lime2 or olinuxino64 into a car navigation system :) | 17:27 |
| Wizzup | liek to turn* | 17:27 |
| freemangordon | heh | 17:27 |
| freemangordon | unfortunately the tablet I have here has no gps | 17:31 |
| Wizzup | I was planning to hook up a usb modem to the lime2 or olinuxino | 17:32 |
| freemangordon | how precise that would be? | 17:33 |
| freemangordon | or, you mean modem with gps? | 17:33 |
| Wizzup | yes | 17:33 |
| Wizzup | almost all of them have gps | 17:33 |
| freemangordon | really? | 17:33 |
| Wizzup | sure, the ones you have all do gps | 17:33 |
| Wizzup | the quectel ones | 17:34 |
| freemangordon | do they? ok, cool. | 17:34 |
| Wizzup | EG25-G supports Qualcomm IZat location technology Gen8C Lite (GPS, GLONASS, BeiDou/Compass, Galileo and QZSS) and the optional multi-constellation GNSS receiver greatly simplifies product design and provides quicker, more accurate and more dependable positioning. | 17:34 |
| Wizzup | glonass being the russian one, galileo being the european one | 17:35 |
| freemangordon | I have ec25-e | 17:35 |
| freemangordon | yeah, seems it supports it | 17:35 |
| Wizzup | yeah, and they work if you have a few of the udev rules in place | 17:37 |
| Wizzup | I think I showed this off on openfest with the lime2's some years ago | 17:37 |
| freemangordon | what do I need udev rules for? ofono supports location on qmi devices | 17:39 |
| Wizzup | gpsd device name | 17:39 |
| Wizzup | I don't think we've ever used ofono for location | 17:39 |
| Wizzup | https://git.maemo.org/leste-upstream-forks/droid4-linux loading the branches take a bit of time | 18:24 |
| Wizzup | yeah, the linux repos might cause some problems | 18:28 |
| Wizzup | I could add some more cpus I suppose | 18:38 |
| mkfx | Wizzup: sounds good. | 18:39 |
| mkfx | Pretty much all features of my q8 work, it's just too hot/uses too much power. | 18:39 |
| mkfx | I might go for a13 support next | 18:39 |
| Wizzup | as I understood it (years ago though) a13 was usually a less supported soc than the a10/a20/a33/a64 | 18:41 |
| sicelo | while Leste isn't really bullying | 18:46 |
| sicelo | mkf while Leste isn't built around that (for good reason), maybe you can look at suspend to ram when idle | 18:47 |
| mkfx | (re: a13) | 18:48 |
| mkf | sure. | 19:03 |
| mkf | hi | 21:13 |
| sicelo | hey | 21:14 |
| sicelo | freemangordon: thanks for review. am i correct to say that in addition to the suggested cchange, i should also add `g_free(upowbat.supply_level);` after https://github.com/maemo-leste/mce/blob/8cb9bfa401ceba27899dbbabeba86e9df5cebaa2/src/modules/battery-upower.c#L553 ? | 21:16 |
| mkf | i think my last message didnt get here, did it? | 21:16 |
| sicelo | it didn't | 21:16 |
| mkf | other than pine phone, n900 and motorollas there are no well supported targets? | 21:17 |
| sicelo | yes. there is a poco f1 port, but it's not really maintained afaict. | 21:18 |
| sicelo | there should also be Librem5 by now, but as you can see, i'm busy wasting time on N900 | 21:18 |
| * sicelo hides | 21:18 | |
| mkf | hm, i can't find any motorola, pinephone or liberm in my country | 21:19 |
| sicelo | anyway, anyone can submit a port and i'm quite sure it'll be accepted, as long as said person is available/willing to maintain it | 21:20 |
| mkf | i might port it to lg v20, since that seems like a nice phone and i already have one. | 21:20 |
| mkf | i just need to root it i guess. | 21:20 |
| sicelo | mkf: pick any other phone that is well supported by mainline, e.g. Samsung A3/A5 2015, Poco F1 (you can pick up the port), Oneplus 6/6T, Pixel 3a, | 21:22 |
| sicelo | interesting that lg v20 ... | 21:24 |
| sicelo | seems kernel for it would be https://gitlab.com/msm8996-mainline/linux ... but it has weird screen. do you even have bootloader unlocked on it? seems only US variant can be unlocked | 21:29 |
| mkf | it can be unlocked thru hacks for all varients, only us varient can be unlocked offically | 21:31 |
| sicelo | ok | 21:33 |
| arno11 | part #maemo-leste | 23:36 |
| arno11 | lol | 23:36 |
Generated by irclog2html.py 2.17.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!