libera/#maemo-leste/ Thursday, 2024-06-27

arno11nice, do you see gpu_unai ?00:01
Wizzuparno11: yes00:12
arno11cool, so it works, i'll send you stuff tomorrow00:13
Wizzupuvos: https://phoenix.maemo.org/job/droid4-linux-binaries/architecture=armhf,label=armhf/195/00:32
Wizzupmake[6]: *** No rule to make target 'debian/linux-image/usr/lib/linux-image-omap/omap4-droid3-xt862.dts', needed by '__dtbs_install'.  Stop.00:32
uvosit seams to be looking in the wrong place00:44
uvosall the dts are now in arch/arm/boot/dts/ti/omap/ instead of in arch/arm/boot/dts00:44
Wizzupok, want me to fix it or shall I?00:45
Wizzuperr00:45
Wizzuplol00:45
Wizzupwant me to fix it or will you?00:45
uvosi mean i gues, if that is the issue00:46
uvosbut then again i dont know how it finds the other dtbs00:46
uvossince those also changed location00:46
uvosso im not sure what to change atm00:46
uvosmaybe somehow you are not asking for ti to be built instead00:47
uvosbtw this is not how we version the kernel00:47
uvosversion should be 6.6.8.000:47
uvosthe 0 being the version we increment when we make code changes00:47
Wizzupk, it's experimental so not the biggest issue I think00:48
uvosanyhow where is the decision what dtbs to built even made00:49
uvosin the packageing files00:49
WizzupI don't think it's looking in the wrong place fwiw00:49
Wizzupscripts/package/builddeb00:50
WizzupI think the droid3 dtb just isn't being built00:50
WizzupI think I fixed this issue before00:51
Wizzuplol00:52
Wizzup    omap4-droid4-xt894.dtb \00:52
Wizzup    omap4-droid3-xt862.dts \00:52
Wizzup    omap4-duovero-parlor.dtb \00:52
Wizzupsee the problem yet?00:52
Wizzupthe problem is in 29885d11320666d987fc1028adcc9058f9769fc7 / arch: arm: dts: omap4-droid3-xt862: move into ti/omap00:53
uvosheh00:53
uvosok00:53
WizzupI'll fix it00:53
Wizzupso what version do you want in the changelog?00:53
uvosex 6.1.67.1-100:54
Wizzupit's 6.6.x00:54
uvos6.1.67 is the upstream version .1 is the second code revision of our patches00:54
Wizzupbut I don't know what stable release you rebased upon00:55
uvosits just so we can bump the version without it going out of sync with upstream00:55
uvoshttps://github.com/maemo-leste/droid4-linux/blob/6852f3753c06ae2ad1eef2e1f25441a89b802bcf/Makefile#L400:56
Wizzup6.6.29.0- then?00:57
Wizzup6.6.29.0-1 then?00:57
uvosjup00:58
WizzupNAME = Hurr durr I'ma ninja sloth00:59
WizzupI forgot about this one00:59
uvosdem do be silly00:59
uvosbest one is still 3.11.0 linux for workgroups00:59
Wizzuphehe01:00
Wizzupok so the hope is that apart from d4 audio calls, this 6.6 works ok?01:00
uvosi mean i have it on my d401:01
uvosand yes01:01
uvosworks fine besides that01:01
uvosat least on d401:01
uvosts issue on n900 might be because of your patches as you say01:01
uvoswhich are included01:01
WizzupI think we already included them before, but I am not sure01:02
Wizzupit's building now, I might make a future experimental build with the regmap writable01:02
uvossure01:02
arno11Wizzup: btw most arm64 devices are probably powerful enough to run almost all emulators through RetroArch (even if it is a pita to configure)11:46
arno11at least diejuse ran generic pcsx easely in chroot iirc11:47
Wizzuparno11: do you use OMP on the n900?14:49
arno11Wizzup: nope i don't like it (even on fremantle)15:35
arno11i prefer mplayer/smplayer15:36
arno11and audacious for music15:36
arno11i don't like the way omp works with mafw stuff15:37
Wizzuphow do you notice that in the user experience?15:39
Wizzupin any case I'm using it now and it works ok for me, but my n900 in general is pretty slow still15:39
arno11ah ok. from a user pov, mafw is too slow for n90015:40
arno11and the fact that you can't really choose what is really playable or not is a problem for me15:41
arno11btw we can have a look on what is still slowing down your n90015:49
Wizzupwhat do you mean, what is playable or not?16:01
Wizzuparno11: appreciate the help in checking, I'm currently deep into some work thing16:01
Wizzupbut I tested the latest 6.6 on my n900 and it works ok16:01
Wizzup(still playing music with it)16:01
Wizzupdidn't test call audio, but it's uvos latest branch now16:01
Wizzupso you might want to give it a swing16:01
arno11ah ok cool16:01
arno11playable or not: i meant lot of audio/video formats are not tracked iirc16:04
Wizzupif they are not we shoul dtry to fix that over time16:06
WizzupI personally really like OMP16:06
arno11ok, btw i forgot to mention that n900 is very slow/unusable when mafw tracker is running (ofc if there is lot of media)16:08
arno11same in fremantle if i let my leste sd card in the slot btw16:09
arno11bbl16:15
Wizzupyes, all devices will be of course17:13
Wizzupbut it doesn't run often at all17:13
arno11yes indeed17:56
arno11i forgot the main issue with omp on N900: cpu and power usage17:57
arno11power usage is 2-3 times higher than audacious for music17:58
arno11around 100mA screen off with audacious, around 200 with omp on N900 iirc18:03
arno11and cpu usage makes multitasking difficult18:05
arno11i wonder why there is so much diff18:06
arno11anyway we all have different favorite apps for different reasons :)18:08
dsc_nioCat: welcome :)18:23
nioCatmeow18:24
kona_Wizzup: i haven't tried this yet on my 900 but might consider configuring any background activity to be nice(1) and starting any cpu-bound jobs with nice as well.19:54
kona_may help with the issues arno11 was noting about responsiveness.19:54
kona_it helped a bit on my n810 back in the day19:55
kona_for system daemons/services/whatever might consider making that default behavior19:55
konaarno11: re: performance with background tasks running, might try renice -n 10 -g ${pgrp_pid}20:42
konaif you do that for any troublesome background task/service/daemon and use the process group leader pid, it should make foreground stuff feel more responsive.20:44
arno11yes i already made lot of tests with nice/renice and chrd20:47
arno11it makes a diff for stuff like phone calls, audio and other stuff. but for heavy daemons like trackers, that's another story.20:49
arno11imo nice stuff is already not too bad on leste and that's tricky to be sure there is not too much impact on other process/daemons when we modify priorities20:54
arno11but again that's just my user opinion20:55
arno11btw if we implement new transitions and no tracker in $HOME, there is almost no slowness anymore excepting for very few things21:00
arno11(i mean on n900)21:03
WizzupI do agree tracker should not follow $HOME21:20
WizzupI forgot what we decided was the best way to solve it21:20
Wizzupyou were looking at the gconf schema, did we just want to change the defaults xml?21:20
arno11i was looking to gconf for apt-worker and indeed we could change the default xml21:25
arno11for gnome tracker, still don't know what could be the best way21:26
arno11and if it should be apply on all devices or just n90021:27
Wizzupwe build gnome tracker s owe cna change the default xml21:39
arno11ah ok21:41
arno11cool21:41
sicelokona: still have the N810, and working?21:59
konasicelo: so, i should have two of them, both working but one of them has a cracked screen.22:39
konabut i haven't seen them since moving in 2015, so idk :)22:39
sicelo:-D22:43
siceloi think they were beautiful devices22:43
WizzupI also have them here in some drawers22:44
Wizzupbut I already think I'm masochistic enough just to work on the n900 :D22:44
konai think they felt better in the hand than the n900, but the n900 has better specs for sure.22:44

Generated by irclog2html.py 2.17.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!