Tuesday, 2019-06-18

*** ahjolinna__ is now known as ahjolinna00:36
liuxstill can not launch camera, log pasted here  https://paste.ubuntu.com/p/yHQnm6VSz5/  and  https://paste.ubuntu.com/p/dqZkTcCnYR/. who has met similar issue and got solution?00:40
vknechtliux, https://github.com/sailfishos/droidmedia/pull/1104:45
*** OhYash1 is now known as ohyash07:10
Mister_Magisterhello i don't have sms notification sound07:28
Mister_Magisteri'm aware that it's pretty cheap way of asking for help i just don't have idea what to write xd07:28
masha11Hello. After adding symlink to the firmware, my device (griffin, cm-14.1) goes into a bootloop. I made mask user@100000.service. How do I get to start GUI? Lipstick fails.11:32
malyou need to provide logs as usual11:32
masha11journalctl: https://pastebin.com/fTxNYgK4, dmesg: https://pastebin.com/9gVu1aGX, EGL_PLATFORM=hwcomposer lipstick --platform hwcomposer: https://pastebin.com/fNPn5Zua strace lipstick: https://pastebin.com/d3zFKExm, systeb/bin/logcat: https://pastebin.com/kNLQN66311:33
mal@adampigg I doubt you native PRs will be included to qtbase and mesa, first of all mesa would break existing OBS projects because if we have two specs then the naming must match one of those11:42
T4<adampigg> mal, currently, there  is only mesa-llvmpipe ... is a real mesa not desirable?11:44
liuxvknecht, thank you, the patch is for CM12.1, maybe I can modify it for my CM11.011:44
mal@adampigg one issue is the maintenance, keeping that working during future updates is a problem if there are official devices to use for testing11:47
T4<adampigg> mal, would official devices using mesa not be using a native mesa?11:48
T4<adampigg> llvmpipe for emu11:49
T4<adampigg> other devices use hybriswith egl?11:49
T4<adampigg> mesa is a bonus for native devices, and, the only current native devices are pretty much mine ;)11:50
T4<adampigg> but we can always keep it as a seperate repo for the native common builds if that is pre3ferable11:50
malmesa is mostly used for build dependency and emulator11:51
T4<eugenio_g7> That reminds me that I still have to upstream the sensorfw iio changes11:51
malanyway that needs some discussion11:52
mal@adampigg maybe a topic to next community irc meeting11:52
T4<adampigg> mal, sure..as mesa isnt really used on a device level, its not that significant11:52
T4<adampigg> mal, my trouble is making the meetings!11:53
T4<adampigg> @eugenio_g7 pull your finger out :P11:53
T4<adampigg> i need to send a config fix for the iio sensor frequency service for us11:53
malI think the sensorfw iio support is a bit limited currently11:53
T4<eugenio_g7> I will :P should be pretty safe11:53
T4<eugenio_g7> And shouldn't break things for other ports11:54
T4<eugenio_g7> Iio support is less than ideal right now, and is somewhat broken on upstream mer11:54
masha11mal, take a look at my logs when you have time?12:03
malno time right now12:09
T4<elros34> @masha11:  try  test_hwcomposer and minimer. These are easier to debug then lipstick12:16
malit's obvious in those logs that some android side things are still broken, probably some of those need to be fixed12:17
masha11@elros34: EGL_PLATFORM=hwcomposer strace test_hwcomposer: https://pastebin.com/QHhSnUhB EGL_PLATFORM=hwcomposer test_hwcomposer, /system/bin/logcat: https://pastebin.com/t8w30Ztr. Minimer has not tried yet.12:48
masha11After starting test_hwcomposer the device restarts.12:49
mal@adampigg btw, the qtbase changes wouldn't work very well, that forces gbm requirement which is provided only by the new mesa spec which wouldn't normally be used12:55
T4<adampigg> mal, true...thats why we dicuss these things :)12:57
T4<dashinfantry> https://twitter.com/JollaHQ/status/1140932431735922689?s=0912:59
liuxmy camera got work after patched in minimediaservice,  the photo mode works fine, but the video recorder does not work. log pasted here: https://paste.ubuntu.com/p/zMcdYfFFnS/13:13
Mister_Magistertypical xd13:16
T4<elros34> masha11: I would start with missing libsubsystem_control.so. Do you have that file somewhere? Does it show the same error in android and so on.13:16
liuxplease take a look at my log if anyone can help, thanks13:31
masha11@elros34: I copied it manually from /home/ubuntu/hadk/vendor/motorola/griffin/proprietary/vendor/lib/libsubsystem_control.so to /system/vendor/lib on the device, but the error remains.13:39
malso it wasn't on device? why would that be, it should probably be in vendor or system partitions13:40
masha11mal: The device did not have this file.13:41
masha11Also, there were no following files that I had to add manually:libsdm-color.so, libsdm-disp-apis.so, libsdmextension.so,libscalar.so, libsdm-diag.so, libqti_performance.so,perfd,qcom-system-daemon.13:43
malare you sure all needed partition are mounted and you have correct lineage base installed?13:45
malit sounds very odd that you would have those in vendor repo but not on device, maybe those are not really needed if those are not on device13:46
masha11mal: I think that I have the correct base. How to know that all partitions are correctly mounted?13:46
masha11mal: I had such errors in test_hwcomposer: library "libsdmextension.so" not found, library "libsdmextension.so" not found, library "libsdm-color.so" not found, library "libsdmextension.so" not found, library "libsdmextension.so" not found, library "libsdm-disp-apis.so" not found. so I tried to add them manually.13:50
T4<elros34> systemctl | grep -i 'droid mount' also df should gives you overview about partitions13:52
malsometimes not all libs are on all devices and you can get "errors" which are not really errors13:54
masha11mal: Ok, I reinstall without copying these files.13:57
malfirst priority is to check mounts and symlinks14:01
liuxmal, could you please help me with camera issue when you hava time? journal log : https://paste.ubuntu.com/p/KFqnKjtV6Z/,  logcat log: https://paste.ubuntu.com/p/zMcdYfFFnS/14:02
malprobably not today14:06
masha11@elros34: my droid mount: https://pastebin.com/dpKy4L9M14:06
Mister_Magistermasha11: is that motorola?14:07
masha11and df: https://pastebin.com/rzdLDMaX14:07
masha11@Mister_Magister: yes14:07
Mister_Magistertfw you can recognize motorola just by looking at partitions xd14:08
Mister_Magistermasha11: just small hint persist has to be there14:09
masha11 @Mister_Magiste: You also ported the motorola device?14:11
masha11why partitions customize and persist are not mounted?14:12
T4<elros34> are you soure you didn't have  /system/vendor/lib/libsubsystem_control.so: https://github.com/TheMuppets/proprietary_vendor_motorola/blob/cm-14.1/griffin/griffin-vendor.mk#L101314:12
T4<elros34> for partitions use systemctl status persist.mount14:15
masha11@elros34: right. there was no file on the device, there was a file on the computer.14:15
T4<elros34> so maybe you flashed too old image14:16
masha11@elros34: systemctl status persist.mount and customize.mount: https://pastebin.com/6STJyJJX14:17
masha11@elros34: Sorry, how to understand what image I'm flashing, old or not?14:23
liuxmal, It seems something wrong with AudioPolicyService on CM11.0, and incorrect config for camera. I'll dig it myself. thank you anyway14:24
masha11Thanks, I will continue tomorrow.14:25
liuxIt hangs when press to record vedio and log printed : Sailfish lipstick[1843]: shared memfd open() failed: Function not implemented Sailfish pulseaudio[1791]: Failed to read from stream. (err -22)15:29
malyou should probably focus on messages in logcat15:31
liuxgot OMXCodec and alsa_pcm error when hanging recod video. logcat pasted here: https://paste.ubuntu.com/p/PT8kN8XrKP/16:00
T4<jnrwrt> I was able to compile the rootfs and install it. … But it doesn't load the system and I see this error in the whole dmesg through Telnet: … [   21.904104] power_supply max77843-fuelgauge: driver failed to report `status' property: -2216:35
malas mentioned so many times, give whole logs, do not try to guess what the issue is16:37
T4<jnrwrt> This is the last_kmsg.log … https://pastebin.com/kzgrbZmx … And if I type 'dmesg > my.log' and them 'head my.log' or 'cat my.log'. … just I can see in whole log: … [ 21.904104] power_supply max77843-fuelgauge: driver failed to report `status' property: -22 … Thanks16:59
Mister_Magistermal: or anyone did you have problem while building droid-hal? https://paste.opensuse.org/view//2784531817:52
Mister_Magisterok it fixed itself nvm17:53
Mister_Magisternope it's back when building everything17:56
Mister_Magisternuke target?17:56
malMister_Magister: reinstall target or delete manually the rpm cache, if I remember correctly that was caused by rpm package update17:57
Mister_Magisterokay thanks17:57
Mister_Magisterrm /var/lib/rpm/?17:58
malcan't remember the exact path, there was some other command needed also to recreate it17:59
malthat issue was quite long time ago, maybe it's easier to reinstall the target17:59
Mister_Magisteri agree it was faster18:02
T4Pratieq was added by: Pratieq20:00
T4<adampigg> @eugenio_g7 pRs awaiting20:09
piggzmal: another idea for meeting .... get someone to create a web service for searching the logs20:17
malpiggz: probably would be a good idea20:17
piggzmal: downloading over a years worth of logs atm so i follow etiquete and not just ask u :D20:18
malpiggz: btw, did you ever check the content of the drivers package of mesa, for arm build most of the files have size 020:18
piggzthere is a comment in the spec about that maybe20:18
piggzmal: yes....20:19
piggzbash-3.2$ cat /home/piggz/mer2/android/droid.fxtecpro1/hybris/droid-configs.log20:19
piggzwarning: Macro expanded in comment on line 157: %{name} contains ssu.ini file which is needed to build kickstarts20:19
piggzPackage 'ssu-kickstart-configuration' not found.20:19
piggzNOTICE: No tags describe the HEAD, will not fix package version.20:19
piggzwarning: Macro expanded in comment on line 157: %{name} contains ssu.ini file which is needed to build kickstarts20:19
piggzerror: Failed build dependencies:20:19
piggz        pkgconfig(android-headers) is needed by droid-config-t5-1-1.armv7hl20:19
piggz        ssu-kickstart-configuration is needed by droid-config-t5-1-1.armv7hl20:19
piggzBuilding target platforms: armv7hl-meego-linux20:19
piggzBuilding for target armv7hl-meego-linux20:19
piggzsorry not that!20:19
piggz(thats my error).....20:19
piggz# Remove empty file created by build20:19
piggzrm -rf %{buildroot}/%{_libdir}/dri/kms_swrast_dri.so20:19
piggzis in the spec from somewhere? fedora?20:19
malpiggz: yes, actually after enabling drm/kms that is not empty20:20
malpiggz: I did that myself20:20
malI think20:20
malpiggz: I removed that so it wouldn't break anything, just in case20:20
malpiggz: anyway something odd is happening in the build20:21
maleven some x86 libs are empty20:22
Mister_Magisterpiggz: whatcha doin20:24
malwe are making common repo for native ports20:25
Mister_Magisternative ports?20:25
Mister_Magisterthat doesn't look like native port to me20:26
malMister_Magister: not that20:26
malfor example codename latte20:26
Mister_Magistermeanwhile i'm struggling with property service20:27
malwhich device?20:27
maland base20:27
Mister_Magisterdoes it matter xd20:27
Mister_Magisterbase 9.020:27
Mister_Magistermal: u okay with lil pm?20:28
maldo you have the selinux_stubs patch20:29
Mister_Magisteri'm not sure20:30
Mister_Magisterit's third time i hear about it :P20:30
malmight also be that you are missing some system/core patches20:31
Mister_Magistermal: yeap i do have stubs patch20:32
Mister_Magistermal: stracing droid-hal-init isn't really saying much20:32
Mister_Magisterspamming restart_syscall(<... resuming interrupted restart_syscall ...>) = -1 EEXIST (File exists)20:32
Mister_Magistermal: any tips on debugging it?20:35
malneed to see if I can find the patches, maybe not today as it's quite late and I feel quite stressed already20:38
Mister_Magistermal: yeah sure take it easy20:38
Mister_MagisterTheKit: was patching it so dunno what are we missing20:38
eugeniopiggz, have you pushed the fixed commit on the PR? :)20:52
eugeniodroid-configs PR I mean20:52
piggzeugenio: well, u could look at the PR, but yes, changed filename and made flavour devel20:55
eugenioyeah I actually looked, it's the same as before here20:55
piggzalso another one iio fix20:55
eugenioyeah, merged that20:55
eugenioI think the force-push didn't go through20:55
eugenioGitHub should show that you force-pushed something after yesterday's review20:56
eugeniobut there is nothing in the PR conversation tab20:56
piggzeugenio: ah, i ballsed my repo with an erroneous force push20:59
piggzeugenio: try now21:01
eugeniomerged :)21:03
malpiggz: ah, wondering if the issue is because of we didn't enable libdrm code for those21:03
piggzmal: thats what i was thinking21:03
malso next up patching21:04
malso next up patching libdrm21:04
eugeniomal, how do you think we should handle the community-native-common repository? I was thinking on putting the feature in a subpackage inside community-adaptation21:05
maleugenio: one possibility or to the upcoming native config submodule21:07
malbut that might not be the best idea because it would cause issues for hypothetical official native devices, although those are quite unlikely21:08
eugenioI would love to eventually see an official device without hybris21:09
maleugenio: a tablet would be in the realm of possibility, a phone still quite far21:10
eugeniomal, for the native config submodule you mean piggz's dhc fork ('native' branch)?21:10
maleugenio: a new one, the fork probably still has many useless things from droid adaptations21:10
piggzmal: jolla should officially support the pinephone tho21:11
eugenioI guess someone could try native-izing the tbj :D21:11
malpiggz: when are those actually going to be made, wasn't it only in devboard level now?21:12
maleugenio: that might be possible21:12
piggzmal: yes, i have the devboard, abranson is jealous .... towards the end of the year?21:12
malwe need to see what kind of submodule would be best for each repo21:14
malI need to try qmi backend of ofono on mainline fp2 to see how it works21:16
piggzmal: how much truth is there to reports of SMS sometimes failing?  i find it odd that such problems still exist21:41
malpiggz: not sure21:43
malpiggz: quite many of those on tjc it seems21:56
malpiggz: should we enable the old freedreno support i.e. for downstream kernels21:59
T4<adampigg> @mal [piggz: should we enable the old freedreno supp …], Cant hurt22:02

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