Sunday, 2018-07-22

octonezdI am failing to build gstreamer, pack_sourec_droidmedia-localbuild.sh always throws "Please build droidmedia per HADK instructions"09:41
malhow did you build the android side things?09:56
octonezdusing make09:57
octonezdI found solution to that, build 32-bit libdroidmedia09:58
malin the beginning of chapter 13.2 there are quite clear instructions, make sure you have latest hadk pdf (2.2.0.1)09:58
octonezdNow I got different error, build_packages says tar not found09:58
octonezdI did it by that instructions, it didn't help. I found https://bitrvmpd.github.io/sailfish-santoni/Troubleshooting.html and fixed that error with latest thing09:59
octonezdSo now I am stuck on + tar -xvf droidmedia-0.20180709.0+2+gc4ea759.tgz /var/tmp/rpm-tmp.qJZ24C: line 37: tar: command not found09:59
malso install tar? it says to install it in chapter 4.209:59
octonezdI already have tar10:00
octonezdand if I type tar --help, it outputs tar --help.10:00
malso then install it to the target also sb2 -t $VENDOR-$DEVICE-$PORT_ARCH -R -m sdk-install zypper in tar10:00
octonezdIt worked! Now I get libminisf.so no such file10:01
maloctonezd: just to be sure, you used this to build the android side stuff: make -j4 $(external/droidmedia/detect_build_targets.sh $PORT_ARCH)10:01
octonezdYep, that commmand10:01
octonezdExcept -j8, not -j410:01
malthat doesn't matter10:01
malok, that should have built 32-bit automatically10:02
octonezdSo, what about libminisf.so not found?10:02
octonezdNevermind, found in FAQ.10:02
malNOO10:03
octonezd?10:03
malthat should have been built, you are doing something wrong10:03
octonezdI doing everything by guide10:03
maland the hadk pdf version is correct?10:03
mal2.2.0.110:03
octonezd2.2.0.110:03
octonezddevice arch is arm64, PORT_ARCH is armv7hl10:04
malhmm, then the detection script is broken10:05
malbecause it should have found what to use https://github.com/sailfishos/droidmedia/blob/master/detect_build_targets.sh10:05
octonezdAh, I think I get what was my current problem, should I redo hybris-hal package before, yep?10:05
octonezdcat lunch_arch says not found10:06
malok, then that is wrong10:07
malso then just build those manually10:08
octonezdShould I redo build_packages.sh -d after that?10:09
maljust follow the instructions, you first have to package droidmedia so it will remove the conflicting files10:10
maloctonezd: also do chapter 13.3 also after doing that 13.210:10
octonezdwell, that is quite weird10:10
malwhat?10:10
octonezdI am still getting minifs not found, but make clearly says that [100% 10/10] Install: /home/octo/hadk/out/target/product/s2/system/lib/libminisf.so10:11
malyou didn't run make libminisf_32?10:11
octonezdI did10:11
malah, you always need to build all of those in that script10:12
malnot just one10:12
octonezdAnd there is out/target/product/s2/system/lib/libminisf.so10:12
malyou are mixing minisf with libminisf10:12
malnot the same thing10:12
octonezdIt says it can't stat out/target/product/*/system/lib/libminisf.so10:12
maljust run make libdroidmedia_32 minimediaservice minisfservice libminisf_3210:13
maloctonezd: stop for a moment so I can explain10:13
octonezdok10:13
maloctonezd: the pack_sourec_droidmedia-localbuild.sh can only be run ONCE after each build of make libdroidmedia_32 minimediaservice minisfservice libminisf_3210:13
malit will move the files built from out10:14
octonezdOh, I get it now. Thanks!10:14
malso if you want to rerun pack script then you always rerun make libdroidmedia_32 minimediaservice minisfservice libminisf_32 before that10:14
octonezdBuilding successful now10:15
maloctonezd: so you will probably have the same issue of make in 13.3 so check external/audioflingerglue/detect_build_targets.sh when you are in that chapter to see what you need to build10:18
octonezdHmm10:51
octonezdAfter installing gstreamer my phone stucks at logo and then shuts off with red light10:51
octonezdOkaaay, it booted for some reason after multiple reboots10:57
octonezdHow can I check if gstreamer is really installed?10:58
elroszypper se -s gstreamer1.0-droid. 'i' indicates installed package11:03
neothefoxhttps://imgur.com/a/tCBLz7h hey everyone, any idea where to start with this one?11:56
elrosneothefox: what android base?12:01
neothefoxelros, LinageOS 14.112:01
elrosI assume you didn't create needed symlinks described in faq12:02
neothefoxelros, let me check12:03
malneothefox: pastebin (or similar services) are much better for sharing text content12:05
neothefoxmal, the copy-paste is broken for me in this telnet, when I'll get to ssh it would be ok12:07
octonezdI stuck at droid-camres, I moved the file, but the camera settings only have 16:9(?Mpix) 4:3(?Mpix) in both of them12:22
octonezdAlso, what about getting sound to work?12:24
maldid the file look ok12:31
octonezdYeah, seems pretty much fine12:31
maldepends on what you mean by sound, do all audio fail or just mp3s12:32
malas I have said many times logs logs logs12:32
octonezdNone of audio seems to work12:32
octonezdNo sound when clicking on things in menus, no sound when trying ringtones12:33
octonezdAlso my phone uses type-c to 3.5, audio over headphones wont work12:34
neothefoxugh, the tar bug is back12:55
neothefoxand here I thought that a new recovery solved it12:55
pseudodevoctonezd: what device?13:10
pseudodevalso, can any one help me with infinite loading screen on lock screen?13:12
pseudodevi dont have telnet so cant check13:12
malthat is caused by ofono13:12
octonezdLeEco Le 2, SD version13:14
maloctonezd: you probably might need something like this for audio, assuming journalctl shows audio_policy.conf parsing error from pulseaudio, if you need it then you need to copy the audio_policy.conf from here https://github.com/LineageOS/android_device_leeco_s2/blob/49e673fb08c9f6e823561f27a0cf4df06444b5b1/audio/audio_policy.conf13:56
maloops forgot the other link https://github.com/mer-hybris/droid-config-sony-nile/commit/6849858c82a68e27fbc314bc7cf3d8e2360b834c13:56
maloctonezd: so add that but use the correct audio_policy.con for your which I linked13:56
*** Nokius_ is now known as Nokius14:01
_svenmal: any idea what could be wrong if the device doesn't change from speaker output to headphone when i plug in a headphone? The gemini plays sound on both, headphone and speaker together after plugin a headphone.14:13
_svenmal: 1. https://pastebin.ubuntu.com/p/3mMB4Xnvp3/ pacmd info output from X; 2. https://pastebin.ubuntu.com/p/vHTKQBxCdb/ pacmd info output from X after plugin a Headphone; 3. https://pastebin.ubuntu.com/p/9ThKYjTxVH/ diff between both files 4. https://pastebin.ubuntu.com/p/HqKcVBP5C8/ pacmd info output from gemini before and after plugin a headphone. There are no changes on the gemini14:15
mal_sven: have you checked faq about headphone detection14:23
neothefoxso, I've created the symlinks and I've rebuild everything, but this is still the same https://pastebin.com/PB3hn1KP14:23
malthat's not a log, get dmesg and journalctl14:23
neothefoxmal, of what unit? or just the kernel itself? It's pretty long in its entirety14:25
malneothefox: just the damn command I told and get the WHOLE output14:26
neothefoxok, give me a second to fish it out with scp14:28
neothefoxmal, kernel log - https://pastebin.com/NQ2K5iWb , journal - https://pastebin.com/3ktJ9r6N14:37
malneothefox: you haven't installed droid-hal-$DEVICE-detritus package, you need to add that to patterns like this https://github.com/mlehtima/droid-config-fp2-sibon/blob/master/patterns/jolla-hw-adaptation-fp2-sibon.yaml#L514:41
malneothefox: I assuem you used straggler_files in droid-hal-$DEVICE.spec14:42
neothefoxmal, thanks again! yup, I've added all extra files to straggler_files14:42
malof course remember to use the correct codename in patterns14:43
neothefoxby the way mal, you always know exactly what to do, are you involved in sailfish development directly?14:48
malI'm just a random porter who has done this way too much14:48
neothefoxmal, that's really cool how fast you can navigate these issues14:51
malmost issues are very common14:52
_svenmal: thanks for the hint, i checked the faq, evdev_trace doesn't show SW_HEADPHONE_INSERT* SW_MICROPHONE_INSERT* https://pastebin.ubuntu.com/p/qSHvFzCKNQ/ it only show me SW_LID for /dev/input/event015:17
mal_sven: so check the uevent method described in faq15:19
_svenmal: /sys/class/switch/h2w changes from 0 to 2 after plugin the headphone so i created the accessories.ini in /etc/ohm/plugins.d and added the 2 lines recommended in the faq15:19
_svendo i have to reboot or restarting a service?15:21
malmaybe restarting ohmd might be enough, not sure15:21
_sventhe problem is not like the faq points to. The audio is routed to headphone, but at the same time its routed to speaker. i want to stop the speaker after plugin the headphone.15:23
mal_sven: some device route the audio to both by default15:25
_svenmal: pacmd info show me that module-udev-detect is not loaded on gemini but it may neccessary because its loaded on the X. If i load it manually using "pactl load-module module-udev-detect" then the gemini reboots15:27
maludev detect is needed for usb audio I think15:27
_svenwhats the trick do disable manually the speaker? is there a /sys/class switch i can use?15:32
_svenmal: strange, i rebootet, now (after creating the /etc/ohm/accessoires.ini) when i plug in the headphone the speaker is off like i wanted and sound comes only from headphone. but after unplug the headphone there is still no sound from speaker.15:39
mal_sven: maybe it needs some quirk, I'll have a look a bit later today15:40
_svenmal: ok thanks15:42
mal_sven: just to be sure, that is geminipda, not the other gemini?15:43
_svenwhat other gemini? its the geminipda yes.15:44
mal_sven: xiaomi mi 5 has codename gemini15:44
mal_sven: so it would be better to use geminipda to be clear15:44
_svenok15:45
neothefoxmal, doesn't look like much changed https://pastebin.com/5tMkbj6a https://pastebin.com/aFa519Vi15:45
malneothefox: get new dmesg after reboot15:47
malneothefox: that is missing all of the important parts15:47
neothefoxmal, https://pastebin.com/C6W9TSMy15:55
malneothefox: so what did you rebuild after adding the detritus package to patterns?16:00
neothefoxmal, everything, the whole repo and then the fs, also when I use zypper to check if this package is installed it shows that it is16:01
malneothefox: hmm, the log still says "droid-hal-init: Failed to initialize property area" which is usually caused by missing detritus packages and the /property_context file that with it16:03
neothefoxhttps://pastebin.com/4q46bqhS16:03
malso you have the /property_contexts file on the device?16:03
neothefoxmal, I do, and it it's not empty16:05
neothefoxhttps://pastebin.com/81pzmX5T16:06
neothefoxcould this be selinux issue?16:07
malneothefox: did you disable audit in kernel config?16:08
neothefoxmal, https://github.com/NeoTheFox/android_kernel_samsung_exynos5410/blob/500b5dc254636badc37161f6cc184cb7006d67da/arch/arm/configs/lineageos_ja3gxx_defconfig#L3436 I've left them undefined since selinux doesn't work with this LOS14.1, if that can be related I would have to rebuild it and debug it further because they caused errors during build16:10
malwe don't want to disable audit I think16:11
neothefoxmal, the config checker has this to say Comment says: Required by hybris, SELinux needs to be disabled. Leave as not set, if you have unset AUDIT (read more about the CONFIG_AUDIT flag)16:13
malwe prefer to disable selinux via commandline and leave audit as it is16:14
malusually disabling audit disables selinux completely but can cause other issues, some kernels won't even build without audit16:14
neothefoxwell, lets see what happens if I disable it16:22
neothefoxthe boot and recovery haven't failed with USE_AUDIT=n16:28
_svenmal: i have to correct me. unplugging the headphone does not stop speaker sound but sound level is very very low. its not the same level as before plugin the headphone.16:28
neothefoxWelp, it booted into GUI16:51
neothefoxhttps://imgur.com/a/HCpgbn916:52
neothefoxthe touchscreen doesn't work, but it's most likely that I've put a wrong event devices as a touchscreen16:53
neothefoxis there a better way than bruteforce to find out?16:55
malneothefox: bruteforce is always there, for evdev_trace you need to install a package16:55
malneothefox: set the input device in https://github.com/mer-hybris/droid-config-sony-nile/blob/master/sparse/var/lib/environment/compositor/droid-hal-device.conf16:56
neothefoxmal, installing a package seems like a better option :)16:56
malneothefox: you probably should adjust pixel ratio to 2 or something like that in droid-config-$DEVICE.spec17:01
neothefoxmal, is it too small?17:02
malyes17:05
neothefoxmal, there is no mce_tools package, am I missing a repo?17:06
_svenneothefox mce-tools?17:07
neothefox_sven, the one where evdev_trace should be17:08
_svenyes17:08
_svenmce-tools not mce_tools17:09
neothefoxok, nvm, I've taken a look at /sys, sec_touchscreen is event117:10
neothefoxby the way, I have an original Jolla Phone, is there a way for me to somehow yank aliendalvik package from there for personal usage? It's touch is dead17:36
neothefoxhttps://imgur.com/a/19Bsa5I18:17
malneothefox: that's an exynos based device?18:18
neothefoxmal, yup18:19
malok, good that the graphics worked so easily18:20
neothefoxyeah, wifi works, camera doesn't and I am about to test the modem18:20
malfor phone calls, camera etc you need to do chapters 13.2 and 13.3 from hadk pdf18:22
malfor build issues check todays log from this channel, there are instructions18:22
neothefoxmal, not yet, I haven't got to these chapters yet, I think it's a good time to commit my configs18:23
malyes18:23
neothefoxwhile building gstreamer - /var/tmp/rpm-tmp.7TcfFa: line 37: tar: command not found - what? tar is installed18:43
pseudodevneothefox:  sb2 -t $VENDOR-$DEVICE-armv7hl -R -m sdk-install zypper in tar18:46
neothefoxpseudodev, ah, got it18:47
neothefoxthank you18:47
pseudodevneothefox: anytime :)18:47
vknecht http://merproject.org/logs/%23sailfishos-porters/%23sailfishos-porters.2018-03-10.log.html#t2018-03-10T10:20:12 : got an idea where that weird third camera comes from : there's an "eye verify" feature for unlocking phone, and it seems it sets up that 3rd cam (checked by renaming some specific libmmcamera*eyeverify.so)18:49
neothefoxhow long a droid-camres script usually runs? It takes a long time for me20:03
steeeve_lbt: can i have that obs repo? ._.20:04
malvknecht: quite interesting20:09
malneothefox: few seconds max20:09
malneothefox: just curious, go to external/droidmedia in your sources and check git log that it's up to date with this https://github.com/sailfishos/droidmedia/commits/master if it's behind then run repo sync external/droidmedia20:11
neothefoxmal, it's in sync with the repo20:14
malhmm, ok20:18
malneothefox: did you first run mk-cam-conf?20:18
maldid that work?20:19
neothefoxmal, yup, it worked20:21
neothefoxhttps://pastebin.com/f17bT2RC I am having this issue with RIL now20:21
neothefoxthe camera works somewhat so it's not a high priority right now20:21
neothefoxand here is the camera conf https://github.com/NeoTheFox/droid-config-ja3gxx/commit/f8e219e17e6b77090440fae054708a301f929bb020:22
malok, then you need to get logs, this time /usr/libexec/droid-hybris/system/bin/logcat, preferrably after reboot and then try to run droid-camres, btw, which user were you when running that, it might have an effect20:23
neothefoxmal, nemo20:23
malneothefox: some devices need some additional configuration of ofono to get things working20:24
malneothefox: for RIL issue get logcat -b radio (of course the same path to logcat as before)20:28
neothefoxmal, https://pastebin.com/ZAx0j21320:28
neothefoxfore some reason -b radio doesn't include RIL stuff, I've tried it from the guide, let me try it again20:29
malneothefox: monich is the best person to help with ofono and RIL related issues20:29
malneothefox: odd that the logcat has very little messages from camera20:30
malah, somehow missed those, there were some20:31
neothefoxmal, https://paste.ee/p/6lDlI here is the -b radio20:31
monichneothefox: Fatal signal 11 (SIGSEGV), code 1, fault addr 0x2 in tid 736 (rild) - that sounds bad (rild is what ofono is talking to)20:32
neothefoxmonich, which is strange since /system/bin/rild keeps running20:33
malneothefox: it might restart, check the process id with ps aux | grep rild20:34
neothefoxmal, yeah, the PID keeps changing20:35
steeeve_Ofono can be also journalded20:38
neothefoxsteeeve_, https://pastebin.com/NaV7zBEB this is what it says20:39
neothefoxon repeat20:39
steeeve_Is that a samsung device?20:40
neothefoxsteeeve_, yup, i950020:40
steeeve_Try this https://gitlab.com/Marco01/droid-config-a5ultexx/blob/master/sparse/etc/ofono/ril_subscription.conf20:42
steeeve_If you're lucky and it uses same ril it will work20:43
malI was just about to say the same thing, except for another device20:43
steeeve_Yaa20:43
malherolte has same config20:43
malneothefox: so copy that file to your device to /etc/ofono/ril_subscription.conf20:44
neothefoxmal, it worked! you guys are a dream team20:50
neothefoxso I should add it under the hybris/droid-configs/sparse/etc/ofono/ril/subscription.conf?20:51
steeeve_Esketitt20:51
steeeve_Yes20:52
steeeve_neothefox: do calls work?20:53
neothefoxsteeeve_, let me install the glue...20:54
steeeve_K20:55
steeeve_However i meant if they worked, if you can actually call, no matter the audio20:56
steeeve_Because probably they wont, but maybe yes ^^20:57
neothefoxsteeeve_, no, they work20:58
steeeve_Niice20:58
neothefoxIt can dial, it can recieve, but audio as a whole doesn't work yet20:58
malneothefox: maybe you need a similar fix I gave for another porter earlier today20:59
malfor audio20:59
steeeve_Oh, nice install glue and test20:59
malneothefox: give me a link to your device repo20:59
malsteeeve_: he said audio doesn't work at all on the device20:59
neothefoxmal, which one? https://github.com/NeoTheFox?tab=repositories21:00
steeeve_mal: ahh, rip21:00
neothefoxsteeeve_, yeah, it was installed, but the audio doesn't work at all21:00
malneothefox: I mean the lineage device repo21:00
neothefoxmal, https://github.com/exynos5410/android_device_samsung_ja3gxx21:00
malneothefox: but if were able to call and the app didn't hang or something then after fixing audio calls should work21:00
steeeve_Exynos :/21:01
neothefoxmal, this is great news! That means that Sailfish is going to get another mostly working device you can use daily (knock on the wood)21:01
malneothefox: what do you see in journalctl from pulseaudio21:03
malneothefox: does the device have /system/etc/audio_policy.conf?21:05
neothefoxmal, yes it does have audio_policy.conf, and I can't find the pulseaudio log in journalctl, isn't pulse meant to be run as a user?21:06
steeeve_Yea21:07
neothefoxhttps://paste.ee/p/q4rK8 here is the content of audio_policy21:09
malneothefox: systemctl status pulseaudio (run as nemo)21:10
neothefoxmal, not found, inactive(dead)21:11
neothefoxaccording to pactl an instance is working21:12
malhmm, does volume control seem to do anything?21:12
neothefoxwell the slider is moving21:13
neothefoxhmm, the SINK1 is in SUSPENDED state21:14
maldoes it show anything about droid in pactl?21:14
steeeve_Pulseaudio is not even running wtf21:15
neothefoxmal, https://paste.ee/p/ue3aR the primary sink is there21:15
neothefoxsteeeve_, it runs as a child of systemd21:15
steeeve_Ah21:15
neothefoxwhile I am playing a rington the primary sink changes state to RUNNING21:18
neothefoxand it says that its output is speaker, so as it should be21:19
malneothefox: but no audio?21:19
neothefoxmal, yup, also if headphones are plugged it doesn't switch the output21:19
malneothefox: headphone instructionsa are in faq but that doesn't help much is audio isn't working21:23
neothefoxmal, well it might be that only speaker doesn't21:23
malneothefox: this is probably not going to help but try anyway add file /etc/pulse/arm_droid_card_custom.pa with a line containing: load-module module-droid-card rate=48000 quirks=+no_hw_volume21:26
neothefoxmal, yup, that did nothing sadly21:33
steeeve_at least in lineage audio works?21:34
neothefoxyea21:36
steeeve_..21:36
neothefoxit worked the last time I've checked21:36
malneothefox: need to investigate more maybe tomorrow21:37
neothefoxhttps://github.com/exynos5410/android_device_samsung_ja3gxx/blob/cm-14.1/audio/audioworkaround/src/org/cyanogenmod/audioworkaround/AudioWorkaround.java21:37
neothefoxthere is this workaround for android in the repos tho21:37
neothefoxmal, ok, thanks for all the help, see you!21:38
malneothefox: another samsung device had this https://gitlab.com/Marco01/droid-config-a5ultexx/blob/master/sparse/etc/pulse/arm_droid_card_custom.pa21:38
malbut that quirk should only affect call audio21:38
maltbr: it seems the sfe symlink in devaamo is not working anymore21:40
steeeve_Mine <321:40
steeeve_Yes, he probably doesnt need that21:41
malwe'll see once audio is working21:41
steeeve_And it is only to switch to speaker on calls21:42
malok21:42
neothefoxyeah, that quirk is not working either21:46
steeeve_If that workaround is needed to get audio, you'll probably need to modify pulseaudio module sources21:47
malwe might need another quirk but we'll see21:48
steeeve_Yeah21:49
malbut that workaround also mentions only in call21:49
steeeve_The commit says 'temp. fix for audio on boot"21:51
malok21:51
steeeve_Also, the mediaplayer stuff and the  is suspicious21:51
neothefoxok, that's gotta be it for today, again, thanks to everyone for help!22:00
*** steeeve_ is now known as steeeve22:13

Generated by irclog2html.py 2.17.1 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!