| Yardanico | hi everyone, some logs about my video playback issue - https://gist.github.com/Yardanico/56af4ccaffc643ae0528b1586ef3b94a . also - where can I change what does minimal brightness mean for my device (it's not really minimal on my device) | 08:01 |
|---|---|---|
| Yardanico | And how can I fix " to /dev/cpuset/system-background/tasks: No space left on device" spam in dmesg? :) | 08:03 |
| Yardanico | maybe I should just use gms-omx and libav for video output like redmi 4x and addison do? | 08:23 |
| r0kk3rz | its an option | 08:23 |
| r0kk3rz | but if you want things like camera to work | 08:24 |
| r0kk3rz | then fixing droidmedia is how you do that | 08:24 |
| Yardanico | camera already works | 08:24 |
| Yardanico | everything major except video playback works | 08:24 |
| r0kk3rz | hmm ok | 08:24 |
| r0kk3rz | that shouldnt be that bad then | 08:25 |
| birdzhang | Yardanico: same issue with redmi 5 plus https://github.com/orgs/Sailfish-On-Vince/teams/sailor/discussions/14 | 08:34 |
| Yardanico | birdzhang, ohh, but it seems like discussion is private :) | 08:36 |
| birdzhang | oh no | 08:37 |
| birdzhang | https://github.com/Sailfish-On-Vince/issues/issues/8#issuecomment-412447275 | 08:37 |
| Yardanico | but yeah, our devices are very close because you have sd 625 too | 08:38 |
| birdzhang | yeah | 08:38 |
| vknecht | not sure it's, do you have a [sparse]/etc/gst-droid/gstdroidcodec.conf file ? eg. https://github.com/vknecht/droid-config-idol3/blob/master/sparse/etc/gst-droid/gstdroidcodec.conf | 08:45 |
| vknecht | *it's the problem | 08:45 |
| Yardanico | vknecht, no I don't have that file, should I have it? | 08:46 |
| * vknecht thinks so | 08:47 | |
| vknecht | try it on device to check ;-) | 08:47 |
| Yardanico | ok I will | 08:48 |
| vknecht | I think it was mostly for web videos, but probably can't make things worse | 08:48 |
| Yardanico | well for me any videos can't play including videos in gallery shot by my own device | 08:48 |
| vknecht | using strace on the player could give some insights | 08:49 |
| Yardanico | nah, sadly gstdroidcodec didn't work, same issue | 08:50 |
| Yardanico | developer for my device (who creates custom roms based on android) hit the same problem almost 9 months ago but he solved it by using updated "venus" firmware | 08:51 |
| Yardanico | idk how to use updated firmware without treble, hmm | 08:51 |
| Yardanico | birdzhang, does that look similar to your log in dmesg while you're trying to play a video? https://gist.github.com/Razziell/ed7358938d61c29ec8e1a902f6e802d9 | 08:52 |
| birdzhang | Yardanico: no, on redmi 5 plus no dmesg output when play video | 08:53 |
| Yardanico | oh, strange | 08:53 |
| birdzhang | this file /etc/gst-droid/gstdroidcodec.conf is needed, i think | 08:54 |
| birdzhang | did you reboot? | 08:54 |
| Yardanico | yes | 08:55 |
| Yardanico | and mido port doesn't have gstdroidcodec | 08:55 |
| Yardanico | and videos work fine on his port | 08:55 |
| r0kk3rz | you mean the config file :P | 08:55 |
| Yardanico | there's only config files for camera here - https://github.com/piggz/droid-config-mido/tree/master/sparse/etc/gst-droid | 08:55 |
| r0kk3rz | is piggz using gst-omx? | 08:56 |
| Yardanico | r0kk3rz, no | 08:56 |
| Yardanico | he only uses gst-droid | 08:56 |
| Yardanico | https://github.com/piggz/droid-config-mido/commit/10cae897f3286261b4654c19ecbca1f1847a69b1 | 08:57 |
| r0kk3rz | thats not to say you dont need codec quirks, but they might be the wrong ones | 08:58 |
| r0kk3rz | you probably want to show logs to abranson sometime | 08:58 |
| Yardanico | also is it ok (and maybe I can somehow remove these spammy messages): "Service 'ims_rtp_daemon' (pid 7353) exited with status 2" sometimes status is 0 instead of 2 | 08:59 |
| Yardanico | droid-hal-init spams with these every 10 sec | 09:00 |
| Yardanico | and sometimes imsdatadaemon too: "Service 'imsdatadaemon' (pid 8400) exited with status 0" | 09:00 |
| birdzhang | you can disable ims_rtp_daemon | 09:03 |
| birdzhang | and imsdatadaemon | 09:04 |
| Yardanico | oh really? I thought they were related to mobile data/ | 09:04 |
| Yardanico | thanks | 09:04 |
| birdzhang | https://github.com/Sailfish-On-Vince/device_xiaomi_vince/blob/master/rootdir/init.target.rc#L202-L207 | 09:04 |
| r0kk3rz | yeah we dont use all the android services, so its a good idea to check what it is first | 09:04 |
| Yardanico | birdzhang, thanks a lot | 09:04 |
| Yardanico | less spam in dmesg is always better | 09:04 |
| abranson | Yardanico: there doesn't seem to be much wrong in the logcat there. maybe try adding symlinks for those media_codec files it's complaining about? | 09:06 |
| abranson | otherwise could be some vendor specific thing | 09:06 |
| Yardanico | ah you mean /data/misc/media/media_codecs_profiling_results.xml ? | 09:07 |
| Yardanico | I can try | 09:07 |
| abranson | yeah, i thought there was another in there somewhere... | 09:07 |
| abranson | it's certainly locating the h264 codec and creating it. not sure what param it's failing with for the ERROR: UnsupportedSetting(0x80001019) | 09:08 |
| Yardanico | LOL | 10:28 |
| Yardanico | how did I not notice this eariler? https://github.com/piggz/droid-config-mido/blob/master/sparse/usr/share/csd/settings.d/hw-settings.ini.txt | 10:28 |
| Yardanico | it has .txt in it and I imported configs from mido and that's why auto brighness didn't work | 10:28 |
| mal | Yardanico: maybe that explain why piggz had issues with sensors | 10:33 |
| Yardanico | mal, xD | 10:33 |
| mal | @adampigg why is this .txt https://github.com/piggz/droid-config-mido/blob/master/sparse/usr/share/csd/settings.d/hw-settings.ini.txt ? | 10:33 |
| Yardanico | maybe he forgot to rename it in git repo but did it locally | 10:34 |
| mal | he builds in OBS so it uses the one in git | 10:35 |
| Yardanico | oh lol | 10:35 |
| mal | probably some issue when saving a file through web browser, sometimes those ass .txt by default | 10:36 |
| mal | *add | 10:36 |
| T4 | <adampigg> Mal yeah,.i noticed that before, but my local install i copied it to .ini so all working for me....if i can get my laptop online i will fix while i remember | 10:48 |
| Yardanico | does anyone know why I get "Device model: Redmi 4 Prime (markw / markw)" in `ssu s` and is this fine? | 11:13 |
| r0kk3rz | yeah thats fine | 11:16 |
| mal | @adampigg wondering if you even need the other config after that? | 11:21 |
| Yardanico | ok, not really good news, my phone has different variations and they have different touchscreens (focaltach or atmel), so some people told me that their atmel touchscreen doesn't work at all | 11:56 |
| Yardanico | I think this might be the issue with different /dev/input/event path? | 11:56 |
| r0kk3rz | use a udev rule to symlink it to something known | 11:59 |
| r0kk3rz | /dev/touchscreen is a good place | 11:59 |
| r0kk3rz | for eg. this looks for an input device with the name "clearpad" - https://github.com/r0kk3rz/droid-config-scorpion_windy/blob/hybris-aosp6/sparse/lib/udev/rules.d/61-touchscreen.rules#L2https://github.com/r0kk3rz/droid-config-scorpion_windy/blob/hybris-aosp6/sparse/lib/udev/rules.d/61-touchscreen.rules#L2 | 12:01 |
| mal | yep, that is the proper way to handle different touchscreens | 12:03 |
| Yardanico | r0kk3rz, how do I check which name my touchscreen has? :) | 12:06 |
| Yardanico | so your is /dev/input/clearpad ? | 12:06 |
| Yardanico | ah I see "udevadm info --export-db" | 12:08 |
| mal | udevadm is the one for those, or evdev_trace -i might also give some of the needed information | 12:10 |
| r0kk3rz | my problem was that depending on the boot it might appear as /dev/input/event2 or /event3 | 12:11 |
| r0kk3rz | so there should be extra attributes that you can pick out in your case | 12:11 |
| r0kk3rz | probably name is enough | 12:11 |
| mal | usually name should be enough | 12:11 |
| Yardanico | ok, thanks | 12:12 |
| mal | Yardanico: so you have to ask the other testers to get you udevadm output | 12:12 |
| Yardanico | mal, yeah I'll connect to the PC of one of the testers (TeamViewer) and I'll telnet to the phone via putty | 12:13 |
| Yardanico | so I can check myself | 12:13 |
| mal | that is quite practical way also | 12:13 |
| Yardanico | if udevadm says "E: OF_NAME=ftech" I can check if it's "ftech" via ATTRS{name}, right? | 12:17 |
| r0kk3rz | do a udevadm info -a -n /dev/input/eventX to check | 12:19 |
| Yardanico | oh it's ATTRS{name}=="ft5346" | 12:22 |
| r0kk3rz | there you go | 12:22 |
| r0kk3rz | presumably the other one will be atsomethingsomething | 12:23 |
| Yardanico | ok nice udev rule worked on my device | 12:26 |
| T4 | <adampigg> Mal, other config? | 12:37 |
| mal | @adampigg https://github.com/piggz/droid-config-mido/blob/master/sparse/etc/sensorfw/sensord.conf.d/60-sensor-selection-overrides.conf | 12:52 |
| mal | maybe that is still needed for some things | 12:52 |
| T4 | <adampigg> K | 12:59 |
| mal | @adampigg just wondering because not all need that, so I would assume there must have been something wrong and that something could have been the naming | 13:00 |
| Yardanico | About my video issue - maybe I'll need to use omx and libav as well? like redmi 4x did | 13:00 |
| T4 | <adampigg> K | 13:01 |
| T4 | <adampigg> Doubt it,.surely video hw is same as mido? | 13:01 |
| Yardanico | well, same chipset | 13:02 |
| Yardanico | my device's kernel (unofficial) is based on TheScarastic's kernel | 13:02 |
| Yardanico | https://github.com/karevn/android_device_xiaomi_markw#device-configuration-for-xiaomi-redmi-4-markw (almost) all hardware component names | 13:04 |
| Yardanico | birdzhang_pc, please ping me if you will be able to solve video issue :) | 13:33 |
| birdzhang_pc | Yardanico: sure | 13:33 |
| Yardanico | r0kk3rz, it's atmel-maxtouch :) finally | 14:51 |
| Yardanico | (name of the device in udev) | 14:52 |
| Yardanico | pketo, hello, I'd want to request Jolla Store access for my phone. Builds for my phone are not on OBS yet but people here on IRC told me that it's not required. I checked "ssu s" on my phone and on the phone of one of testers, device UID is unique, persistent across reboots, and it's equal to the first IMEI (phone has two sim slots). Device model shown by `ssu s`: "Device model: Redmi 4 Prime (markw / markw)" | 15:27 |
| Yardanico | all sources are here - https://github.com/SailfishMarkw | 15:27 |
| Yardanico | I can show screenshots from my phone and another phone with ssu s output | 15:29 |
| vknecht | Yardanico, relax :-) | 15:29 |
| Yardanico | vknecht, :D | 15:29 |
| r0kk3rz | Yardanico: do you have a correctly filled out 27b/6 ? | 15:30 |
| Yardanico | r0kk3rz, XD | 15:30 |
| vknecht | r0kk3rz, I thought that one was only for source-release requests... | 15:31 |
| pketo | Yardanico: Added. And thank you, that was clear request with complete information :) | 15:50 |
| Yardanico | pketo, thanks for quick response :) | 15:51 |
| Yardanico | btw, what's the preferred way to set a CPU governor in Sailfish OS ports? | 16:11 |
| * vknecht launches suspend test and crosses fingers | 16:47 | |
| Yardanico | guys is this a typo? https://github.com/piggz/droid-config-mido/blob/master/sparse/usr/bin/droid/android-links.sh#L2 I wondered why android-links didn't work for me | 17:16 |
| Yardanico | I added a space before "];" and it worked | 17:16 |
| vknecht | Yardanico, I'd say you did the right thing, provided this addition is necessary ;-) | 17:21 |
| vknecht | Yardanico, for cpu governors, afaik it's not an usual tweaked part (no entry in sfos cheat-sheet)... tried cpufreq ? | 17:24 |
| Yardanico | vknecht, the thing is that the phone boots with "performance" governor by default | 17:25 |
| Yardanico | and that's not good for battery life | 17:25 |
| vknecht | though I read quite some pestering against big.little cpus ;-) | 17:25 |
| Yardanico | sd 625 is not really big.LITTLE | 17:25 |
| Yardanico | all cores are equal | 17:25 |
| Yardanico | sledges, hello, can you create nemo:devel:hw:xiaomi:markw repo on OBS for me? I want to start moving things to OBS | 17:26 |
| Yardanico | oh, I need to register first and I can't | 17:27 |
| vknecht | I've never messed with that, but there certainly is something good for battery to gain | 17:28 |
| vknecht | some even do undervolting and stuff like that | 17:28 |
| vknecht | but it's a two-edged sword :-) | 17:29 |
| Yardanico | well performance governor is really bad for mobile phones :) | 17:30 |
| vknecht | I think I read it was on-demand by default on sfos, but couldn't garantee that | 17:30 |
| vknecht | tried the irc archives with some good grep keywords ? :-) | 17:31 |
| Yardanico | vknecht, no :) | 17:37 |
| r0kk3rz | Yardanico: theres usually a prop that gets set and triggers some android init script | 17:40 |
| Yardanico | r0kk3rz, ah I see there's a file https://github.com/SailfishMarkw/android_device_xiaomi_markw/blob/hybris-14.1/rootdir/etc/init.qcom.post_boot.sh | 17:42 |
| vknecht | Yardanico, for OBS you have to request a Bugzilla account for https://bugs.merproject.org/ (you can do the request in this channel) | 17:43 |
| Yardanico | r0kk3rz, yeah thanks I found out what was the problem | 17:50 |
| vknecht | I wonder if the 40 iterations come from running a second time in same session, running a 3rd time gives 20, 4th time as 40, etc. | 19:01 |
| r0kk3rz | 40 iterations? | 19:03 |
| vknecht | in CSD suspend test | 19:04 |
| r0kk3rz | oh ok | 19:08 |
| vknecht | r0kk3rz, it was a followup (missed mentioning CSD) of http://merproject.org/logs/%23sailfishos-porters/%23sailfishos-porters.2018-08-12.log.html#t2018-08-12T16:16:46 | 19:14 |
| vknecht | any thought about this test ? :-) | 19:15 |
| vknecht | thought/experience | 19:15 |
| r0kk3rz | i think its a thing that exists yes | 19:17 |
| vknecht | hehe, ok :-) | 19:19 |
| kimmoli | meh. my mw build fails, libhybris says: autoreconf: 'configure.ac' or 'configure.in' is required | 20:32 |
| kimmoli | am i missing something ... | 20:32 |
| r0kk3rz | like a configure.ac maybe? | 20:41 |
| r0kk3rz | hmm submodule is inited? | 20:42 |
| kimmoli | should be | 20:48 |
| r0kk3rz | yes, it should | 20:49 |
| r0kk3rz | worse case nuke the mw/libhybris folder | 20:50 |
| kimmoli | lets see libhybris submodule was dangling | 20:56 |
| kimmoli | just want to play with test_gps and raw nmea data | 20:57 |
| kimmoli | oo success on that | 21:00 |
| kimmoli | Thanks r0kk3rz being my rubber duck | 21:00 |
| r0kk3rz | quack! | 21:01 |
| kimmoli | oh time flies. need also to hack cheapo-chinese bike rear light (8x8 led, acceleration sensor, animations are stupid. need to replace the cpu with something that i can reprogram) | 21:02 |
| kimmoli | hmh maybe with this https://oshpark.com/shared_projects/KhRpkfcU | 21:05 |
| Umeaboy | Hi! | 22:15 |
| Umeaboy | Have I made my droid-hal-version repo correct? | 22:16 |
| Umeaboy | https://github.com/sailinggalaxians/droid-hal-version | 22:16 |
| mal | is tat supposed to be the device specific repo? that is wrong | 22:18 |
| mal | it should look something like this https://github.com/mlehtima/droid-hal-version-fp2-sibon | 22:18 |
| mal | even the naming of your repo is wrong | 22:19 |
| mal | you didn't seem to understand the submodule, why didn't you just follow the hadk pdf instructions? | 22:20 |
| T4 | <NotKit> demonaxsh, do you still have problems with Bluetooth? | 22:26 |
| T4 | <NotKit> currently on the go, but please check Ubuntu Tablet - https://github.com/ubports/kernel_bq_m10 and mine https://github.com/gemian/gemini-linux-kernel-3.18/commits/master/drivers/misc/mediatek/connectivity | 22:28 |
| T4 | <NotKit> it's needed to port older driver which can work with BlueZ as opposed to BlueDroid | 22:28 |
| Umeaboy | mal: How do I add the droid-hal-version @ 573d50d into my repo? I cloned the sibon repo and made the changes and pushed without a problem, but the diretory is not there. | 23:45 |
| Umeaboy | https://github.com/sailinggalaxians/droid-hal-version-sony-maple is the right naming of the repo. | 23:46 |
Generated by irclog2html.py 2.17.1 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!