| deathmist | just wanted to drop by and confirm that the new 3.2 Torronsuo build targets seem to function properly before going to sleep :) https://i.imgur.com/mV7TJon.png | 00:35 |
|---|---|---|
| *** birdzhang1 is now known as birdzhang | 12:39 | |
| deathmist | welp, just got audio working on hybris-16.0 using the "ln -s /system/lib/vndk-28/libaudioroute.so /usr/libexec/droid-hybris/system/lib" method I tried before that failed :DD will definitely have to backtrack now a bit and see if any other symlinks I added helped to achieve this! | 12:42 |
| deathmist | currently I have a weird issue with the vibration working on some boots but not others, keyboard taps seem to be unaffected however which is even weirder. it can be seen (not felt) on pulley menus and home screen actions. even weirder all haptics work in "Components Qt5" | 12:50 |
| rinigus | when debugging hybris, can we use HYBRIS_INFO? I have inserted it in few places but don't see any output in journal nor logcat. is there anything else that needs to be enabled? | 13:00 |
| rinigus | or should I use something else for it? | 13:00 |
| T42 | <hacker12455> Any idea why syncing hybris 14.1 repos fails on this? | 14:35 |
| T42 | <hacker12455> Fwd from hacker12455: Checking out project mer-hybris/android_bionic | 14:35 |
| T42 | <hacker12455> error: Cannot checkout mer-hybris/android_bionic: GitError: mer-hybris/android_bionic rev-list ('^ef7432cc40d58b7359a2b6f5b6f9f5f5ea52510a', 'HEAD', '--'): fatal: bad object HEAD | 14:35 |
| T42 | <adampigg> Rinigus: does it work? | 14:51 |
| rinigus | adampigg: had to hunt it down where its loaded from. in my case, its in the library on vendor's image. so, all testing takes time - have to reflash. going to test soon (here it means today) | 14:53 |
| rinigus | adampigg: btw, you had also mount-sd@sdX services running. where do we kill udisk2 desire to mount them? | 14:54 |
| mal | z@hacker12455 try without --fetch-submodules | 15:03 |
| T42 | <hacker12455> Still the same error | 15:04 |
| mal | @hacker12455 did you do repo init also now or sometime before? | 15:09 |
| T42 | <hacker12455> Did repo init before as I moved bases | 15:09 |
| T42 | <hacker12455> But the previous hadk folder was deleted | 15:10 |
| T42 | <hacker12455> This is in a new one. | 15:10 |
| mal | hmm, that commit hash in that error message matches the latest one here https://github.com/mer-hybris/android_bionic/commits/hybris-14.1 | 15:11 |
| T42 | <Lukapanio> What dose it mean | 15:20 |
| T42 | <adampigg> Rinigus: i use a bind mount to override the /vendor file, and ship the .so in /usr/libexec/droid-hybris/vendor.... | 15:30 |
| rinigus | adampigg: as we are shipping with vendor image (its compiled from aosp), I can compile it correctly in vendor. but something got messed up and the last image is missing my patch. have to grep where it is | 15:31 |
| T42 | <hacker12455> mal: sync works fine now | 15:51 |
| rinigus | adampigg: will have to make new image, I missed one step :( | 15:53 |
| rinigus | mal: eglplatformcommon_eglQueryString max string length: 1060 | 15:53 |
| rinigus | mal: waylandws_eglQueryString max string: 1133 | 15:54 |
| rinigus | mal: so, your theory regarding 1024 being short is confirmed | 15:54 |
| vknecht | so, 2048 should be enough for everyone ? :-) | 16:00 |
| mal | vknecht: rinigus this reminds me of the old memory quote :) | 16:03 |
| rinigus | exactly | 16:03 |
| vknecht | hehe, a false one I heard, but funny nevertheless ;-) | 16:03 |
| mal | ok, it's difficult to know which ones are real and which not | 16:05 |
| mal | rinigus: vknecht should we just add 2048 or make the code have dynamic memory, latter might need some thinking | 16:07 |
| rinigus | mal: its interesting the way this code is used - local static var gets returned. I presume that dynamic part is already there outside this function | 16:08 |
| rinigus | mal: in this respect, setting to 2048 shouldn't be major issue | 16:09 |
| vknecht | bits of progress on pov showing metada, not sure how localized values came in since yesterday, tho... https://framapic.org/gallery#j0W9M0U7m3P4/Z6PbudSfZyDx.jpg,vQCY6uLcgp6l/5kWytcIWLz6D.jpg | 16:15 |
| vknecht | not nitpick needed for now on line elision(?)/wrapping ;-) | 16:17 |
| vknecht | tried enabling text elide/wrapping but it's a mess, cuts lines not needing to | 16:19 |
| vknecht | pleaz allow quillmetadata in harbour ;-) | 16:24 |
| vknecht | (using libexif directly through QLibrary atm) | 16:25 |
| rinigus | can't get rid of udisk2 trying to mount sda5 and sda2 using an extra rule in /etc/udev/rules.d : https://dpaste.de/aTbq | 16:29 |
| rinigus | when checking with `udevadm info --name /dev/sda2 --query all` all seem to be applied. but still systemd mount is tried | 16:30 |
| mal | rinigus: maybe create a RP to upstream libhybris with 2048 | 16:38 |
| mal | *PR | 16:38 |
| rinigus | mal: will do | 16:45 |
| mal | rinigus: not sure how long that will help if the number of extensions increases this fast | 16:46 |
| rinigus | yeah, no idea. let's hope for some time | 16:47 |
| deathmist | I thought the "mode=ask" usb-moded config https://git.io/JegoJ changed the default mode to "Always ask" but on my images it still is always defaults to "Developer mode" or am I interpreting it wrong? :p | 16:47 |
| vknecht | deathmist, see may commit here, there was some changes... couldn't tell details tho... https://git.merproject.org/mer-core/usb-moded/commits/master | 16:49 |
| T42 | <Lukapanio> error: Bad exit status from /var/tmp/rpm-tmp.BFq9fE (%build) | 18:11 |
| T42 | <Lukapanio> RPM build errors: | 18:11 |
| T42 | <Lukapanio> Bad exit status from /var/tmp/rpm-tmp.BFq9fE (%build) | 18:11 |
| T42 | <Lukapanio> when running rpm/dhd/helpers/build_packages.sh | 18:11 |
| T42 | <Lukapanio> error: Bad exit status from /var/tmp/rpm-tmp.q6Ly0e (%build) | 18:14 |
| T42 | <Lukapanio> RPM build errors: | 18:14 |
| T42 | <Lukapanio> Bad exit status from /var/tmp/rpm-tmp.q6Ly0e (%build) | 18:14 |
| T42 | <Lukapanio> in log | 18:14 |
| T42 | <Lukapanio> @elros34 any idea? | 18:15 |
| deathmist | @Lukapanio that tells absolutely nothing, what package did it fail on? pastebin full build log instead of picking useless temporary file mentions | 18:16 |
| deathmist | and if it worked before, what did you do now that it doesn't work anymore? | 18:17 |
| T42 | <Lukapanio> https://bpaste.net/show/TBUPY | 18:21 |
| T42 | <Lukapanio> full log | 18:24 |
| deathmist | yes I can see, why is CONFIG_DUMMY=y? it should be =n in your defconfig, I'm assuming that's why building hal packaging fails | 18:25 |
| T42 | <Lukapanio> ouch sorry | 18:25 |
| deathmist | also CONFIG_SECURITY_SELINUX_BOOTPARAM should be =y too if you want to disable SELinux on kernel cmdline using "selinux=0" | 18:26 |
| T42 | <Lukapanio> hmmm i haveCONFIG_SELINUX_BOOTPARAM=y | 18:29 |
| deathmist | then I assume your device normally boots without SELinux too? "CONFIG_SECURITY_SELINUX=y" should be present in there too if SELinux is enabled | 18:32 |
| T42 | <Lukapanio> yep my device can boot with CONFIG_SECURITY_SELINUX=y and without it | 18:33 |
| deathmist | have you booted SFOS yet? I'm not sure if SECURITY_SELINUX should be =y if you boot without it already | 18:34 |
| T42 | <Lukapanio> nohavnt boot yet but ubuntu touch boots ok without CONFIG_SECURITY_SELINUX=y | 18:35 |
| deathmist | sure, you can probably continue without enabling SECURITY_SELINUX then if it was already not enabled. you may want to enable CONFIG_AUTOFS4_FS and CONFIG_FANOTIFY tho at least still | 18:41 |
| deathmist | vknecht: btw I found out that usb-moded.ini is deprecated and gets removed if found ;) I fixed it by simply renaming the file to "80-use-ask-mode.ini" so it loads after "70-use-developer-mode.ini" that was already present on my device and has "mode=developer_mode" in it | 18:49 |
| vknecht | thanks, good to know :-) | 18:50 |
| T42 | <DylanVanAssche> spirroin: DBus monitor doesn't show any DBus events from mce when pressing the volume keys. | 18:52 |
| T42 | <DylanVanAssche> I used this monitor command: `dbus-monitor --system type=signal sender="com.nokia.mce" interface="com.nokia.mce.signal" path="/com/nokia/mce/signal"` | 18:52 |
| T42 | <DylanVanAssche> Other mce DBus events as display blanking are coming through. | 18:52 |
| rinigus | mal: PR submitted https://github.com/libhybris/libhybris/pull/427 | 18:57 |
| mal | rinigus: should the 2048 be moved to a header as a define? | 19:10 |
| mal | and for the other value define-2? | 19:11 |
| rinigus | mal: no idea. I don't even know where -2 is coming from. right now I kept it minimal | 19:18 |
| rinigus | piggz: your idea worked. here is the implementation that I used https://github.com/sailfishos-sony-tama/device-sony-common/commit/62fa1e6e0eab38dac16a9f2f14ee3bc6194f417e | 19:19 |
| rinigus | piggz: based on your patch with the loop coming out during discussion with Marijn from sony's dev channel | 19:22 |
| T42 | <hacker12455> Fwd from hacker12455: * Building rpm/droid-hal-version-cedric.spec | 19:47 |
| T42 | <hacker12455> No provider of 'ngfd-plugin-droid-vibrator' found. | 19:47 |
| T42 | <hacker12455> No provider of 'qt5-feedback-haptics-droid-vibrator' found. | 19:47 |
| T42 | <hacker12455> NOTICE: No tags describe the HEAD, will not fix package version. | 19:47 |
| T42 | <hacker12455> error: Failed build dependencies: | 19:47 |
| T42 | <hacker12455> ngfd-plugin-droid-vibrator is needed by droid-hal-version-cedric-0.0.1-1.armv7hl | 19:47 |
| T42 | <hacker12455> qt5-feedback-haptics-droid-vibrator is needed by droid-hal-version-cedric-0.0.1-1.armv7hl | 19:47 |
| T42 | <hacker12455> Building target platforms: armv7hl-meego-linux | 19:47 |
| T42 | <hacker12455> Building for target armv7hl-meego-linux | 19:47 |
| T42 | <hacker12455> * Check /home/hello/hadk/hybris/droid-hal-version-cedric.log for full log. | 19:47 |
| T42 | <hacker12455> How to fix that? | 19:47 |
| T42 | <Lukapanio> and log | 19:47 |
| T42 | <hacker12455> well | 19:48 |
| T42 | <hacker12455> log contains this | 19:48 |
| T42 | <hacker12455> PlatformSDK bash-3.2$ cat /home/hello/hadk/hybris/droid-hal-version-cedric.log | 19:48 |
| T42 | <hacker12455> No provider of 'ngfd-plugin-droid-vibrator' found. | 19:48 |
| T42 | <hacker12455> No provider of 'qt5-feedback-haptics-droid-vibrator' found. | 19:48 |
| T42 | <hacker12455> NOTICE: No tags describe the HEAD, will not fix package version. | 19:48 |
| T42 | <hacker12455> error: Failed build dependencies: | 19:48 |
| T42 | <hacker12455> ngfd-plugin-droid-vibrator is needed by droid-hal-version-cedric-0.0.1-1.armv7hl | 19:48 |
| T42 | <hacker12455> qt5-feedback-haptics-droid-vibrator is needed by droid-hal-version-cedric-0.0.1-1.armv7hl | 19:48 |
| T42 | <hacker12455> Building target platforms: armv7hl-meego-linux | 19:48 |
| T42 | <hacker12455> Building for target armv7hl-meego-linux | 19:48 |
| T42 | <Lukapanio> may be skip them for now | 19:49 |
| T42 | <hacker12455> how | 19:50 |
| T42 | <hacker12455> build_packages doesn't give me an option to skip | 19:50 |
| T42 | <hacker12455> It did before I think | 19:50 |
| T42 | <Lukapanio> try now | 19:50 |
| T42 | <hacker12455> I could comment out that | 19:50 |
| T42 | <hacker12455> @Lukapanio [try now], what did you do? | 19:50 |
| T42 | <Lukapanio> @hacker12455 [I could comment out that], i donr that | 19:50 |
| T42 | <hacker12455> ok runninf | 19:51 |
| T42 | <Lukapanio> in yaml file | 19:51 |
| T42 | <hacker12455> *running | 19:51 |
| deathmist | idk which one of you has the issue but you need "%define have_vibrator_native 1" in your dvd repo spec file, replace "%define have_vibrator 1" entirely with it | 19:54 |
| deathmist | could broken RTC HW clock cause issues for mobile/cellular data? that's still broken for me on hybris-16.0 (maybe unrelated but SMS and calls seem to already work). here's a snippet of what happens when I toggle the setting https://pastebin.com/bw2Nz8u5 | 20:14 |
| *** OhYash1 is now known as ohyash | 20:42 | |
| vknecht | grmpf, not much metadata from Qt about vids... https://framapic.org/FKnki3Dq7cwe/J91Q5QbkLQkC | 21:14 |
Generated by irclog2html.py 2.17.1 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!
