| T42 | <adampigg> Change version of mapplauncherd, 4.2 has a commit that is later reverted, see dontbeevil repo (re @b100dian: still bisecting what...) | 05:50 |
|---|---|---|
| T42 | <adampigg> Ah, elros replied too :) | 05:51 |
| T42 | <b100dian> Thank you | 07:45 |
| kquote03[m] | Hello | 11:44 |
| kquote03[m] | I'm following the PDF, currently trying to build HL | 11:44 |
| kquote03[m] | s/HL/HAL/ | 11:45 |
| kquote03[m] | and i get the following error... (full message at https://matrix.org/_matrix/media/r0/download/matrix.org/fUkKqxroHWJCtltoSXsApQwF) | 11:45 |
| Thaodan | Try to log what defines these import includes | 12:09 |
| Thaodan | *export | 12:09 |
| Thaodan | also please post full logs on a paste services | 12:09 |
| poetaster | anyone an idea how I debug /org/nemomobile/lipstick/screenshot rebooting volla? | 12:12 |
| poetaster | no log entries on reboot | 12:12 |
| T42 | <elros34> IIRC it use grabWindow() method. I can give you a workaround if you want:) | 12:13 |
| kquote03[m] | <Thaodan> "Try to log what defines these..." <- Mind if I ask where are logs located (and whether if I need to enable logging or not? Thank you | 12:16 |
| Thaodan | the logs that you get why run make | 12:18 |
| kquote03[m] | Ah just console output then? | 12:18 |
| Thaodan | logs are in $ANDROID_ROOT/ | 12:19 |
| Thaodan | logs are in $ANDROID_ROOT/out | 12:19 |
| kquote03[m] | soong.log, does this sound right? | 12:19 |
| Thaodan | soong is just one stage, there is build.trace and verbose.log | 12:25 |
| Thaodan | see here for more | 12:25 |
| Thaodan | https://android.googlesource.com/platform/build/+/HEAD/Usage.txt#43 | 12:25 |
| kquote03[m] | thanks | 12:25 |
| Thaodan | Most of this stuff related to building since Android side of hybris is more an Android thing like building AOSP | 12:27 |
| kquote03[m] | <Thaodan> "soong is just one stage, there..." <- Hey I have build.trace (http://hastebin.com/mubocafipu.json) but not verbose.log (http://hastebin.com/ugojeqiyav.css) | 12:35 |
| T42 | <elros34> kquote03[m] do you use correct branch for android_hardware_samsung? | 12:43 |
| T42 | <elros34> poetaster might be worth to try https://github.com/mer-hybris/droid-config-f5121/blob/76757c61805e5cc04db2bfc6f8d9847bf1b7f97b/sparse/var/lib/environment/nemo/99-qtscenegraph.conf, I think it might be during screenshot capture | 13:16 |
| T42 | <elros34> used* | 13:16 |
| poetaster | elros34, thanks | 13:21 |
| poetaster | elros34: QT_OPENGL_NO_BGRA=1 is the value I have? | 13:23 |
| T42 | <elros34> ee? Do you have that file already? | 13:29 |
| poetaster | elros34: yes that file is on the phone | 13:43 |
| T42 | <elros34> ok then try to remove it. I use some workaround for screenshot because glReadpixels fails for me so it might works for you too. Unfortunately you would have to build patched lipstick-qt5: https://github.com/elros34/droid-hal-moto_msm8960_jbbl/blob/master/patches/lipstick.patch | 13:45 |
| poetaster | ok, I'll take a look. | 13:46 |
| T42 | <elros34> remember: remove is not same as QT_OPENGL_NO_BGRA=0 | 13:47 |
| poetaster | yes. but # presumably is :) | 13:49 |
| poetaster | strange. I and Pasik2 on the forums also get, after the first screenshot reboot, the sim pin preference isn't saved. | 13:52 |
| T42 | <elros34> maybe some file is corrupted because of unexpected reboot | 13:56 |
| poetaster | elros34: yeah, likely. | 13:56 |
| poetaster | elros34: although, it is the identical behaviour for Pasik2. identical corruption? possible. | 13:57 |
| T42 | <elros34> no idea where these settings are saved but I remember some path for setting which cause issues in past: /var/lib/ofono | 13:58 |
| poetaster | elros34: rings a bell | 13:58 |
| kquote03[m] | <T42> "<elros34> kquote03 do you use..." <- I'll check again, but I'm certain I did. | 14:08 |
| T42 | <elros34> check also external/dtc, it should have libfdt | 14:16 |
| T42 | <Josedelavega> does hybris 16.0 support dtboimage? | 15:07 |
| Thaodan | which kind of support? | 15:20 |
| Thaodan | dhd can create a subpackage for it but there is no need to directly support it in kernel | 15:21 |
| T42 | <Josedelavega> so i should try using 17.1 instead? , im having some issues related to a media codec | 15:29 |
| Thaodan | you can use 1 | 15:56 |
| Thaodan | 16.0 too dhd is independent of that | 15:57 |
| T42 | <adampigg> mal: had to add some extras to initrd image creation https://github.com/HelloVolla/droid-hal-img-boot-volla-yggdrasil/commit/bcc0c26a73dc8a814d40128c62ce825d788deb67 | 18:35 |
| Thaodan | Interesting there was something like this for tama | 19:12 |
| Thaodan | https://github.com/sailfishos-sony-tama/main/blob/9e38c9035c5c704082ece3f3280022bf89e4b764/flashing.md#flashing-files | 19:13 |
| Thaodan | Maybe the way you adjusted the vbmeta image helps in case of tama too? | 19:14 |
| Thaodan | cc: rinigus | 19:14 |
| rinigus | Thaodan: thanks for pinging | 19:15 |
| rinigus | I've cut some corners and used the same approach as by many for AOSP ROMs - asked users to get empty vbmeta separately. I guess piggz solved it correctly and has vbmeta generated for the image | 19:17 |
| Thaodan | I just wonder why some port need it and why some don't. | 19:18 |
| kquote03[m] | It seems that my problem was indeed with branches, specifically my local_manifests file | 19:18 |
| Thaodan | Like I didn't need it on kumano, for me the selinux rules were wrong in the userdebug build I had of Android and the hybris Android build. | 19:19 |
| kquote03[m] | Is there a way to easily write the $DEVICE.xml? I've seen some that have 10s of sources, but the guide only has 2 | 19:19 |
| Thaodan | Is that some lineage thing? AOSP just has one Manifest. | 19:20 |
| kquote03[m] | Yea I'm dealing with Lineage. But I've also seen some Sony manifests that have 5 or so sources | 19:21 |
| rinigus | Thaodan: but is vbmeta.img needed on AOSP kumano? for tama AOSP it is a part of installed images | 19:22 |
| Thaodan | rinigus: The partition is there but it is not build. I guess it can be ignored. | 19:24 |
| rinigus | Thaodan: strange, but if it works :) . how is kumano doing? have you had a chance to work on it? | 19:25 |
| Thaodan | https://github.com/sonyxperiadev/device-sony-kumano/search?q=dtbo | 19:25 |
| Thaodan | Kumano has dtbo but I had to do nothing with it but I tried the vbmeta trick as you had while testing so maybe that already did something. | 19:26 |
| Thaodan | rinigus: Was vbmeta created for you while building? | 19:26 |
| kquote03[m] | can someone send their $DEVICE.xml so I can try to sanity check? | 19:26 |
| Thaodan | I guess some platform just don't need to flash it e.g. seine doesnt' need it as you can see | 19:27 |
| Thaodan | *doesn't need to modify | 19:27 |
| rinigus | Thaodan: yes, it was built as far as I remember. just a sec, let me check | 19:27 |
| rinigus | I am building it (vbmetaimage) together with systemimage vendorimage. then once in a port lifetime get it from there and add as a separate file in github. | 19:29 |
| T42 | <elros34> kquote03[m] start with device repo, then add any device specific repos which you find in lineage.dependencies file | 19:48 |
| T42 | <elros34> if something will be missing then your build will fail | 19:48 |
| kquote03[m] | T42: lineage.dependencies file EXISTS??? thanks for letting me know, you rock! | 19:49 |
| T42 | <elros34> I am not sure I understand you correctly but why it woudn't? | 19:50 |
| kquote03[m] | i just didnt know. Im a total newb | 19:50 |
| kquote03[m] | i REALLY feel out of my depth, but eh, its fun | 19:51 |
| T42 | <eugenio_g7> avbtool make_vbmeta_image --flags 2 --padding_size 4096 --set_hashtree_disabled_flag --output should be enough to create a suitable vbmeta btw | 19:51 |
| T42 | <eugenio_g7> (also tested fine on kumano) | 19:51 |
| piggz | rinigus: Thaodan: its not quite upstreamable, probably needs a flag to enable/disable, and i also made changes to hybris-boot to package avbtool | 20:28 |
| Thaodan | piggz: avbtool would need to go to a different package | 20:44 |
| piggz | Thaodan: why not the -tools sub package? | 20:45 |
| Thaodan | https://github.com/sailfishos/android-tools-hadk | 20:45 |
| Thaodan | Ah that's the one I meant | 20:45 |
| piggz | we already package mkbootimg etc there | 20:45 |
| Thaodan | I thought you mean adding them to hybris-boot | 20:45 |
| piggz | sorry, meant d-h-d subpackage | 20:45 |
| Thaodan | piggz: Makes sense, that's what was about to suggest | 20:45 |
| Thaodan | We clearly have some duplication there | 20:47 |
| Thaodan | android-tools-hadk also has mkbootimhg | 20:47 |
| Thaodan | *android-tools-hadk also has mkbootimg | 20:47 |
| T42 | <b100dian> oof.. got my install booting again. hopefully I can check that ngfd now Thaodan, sorry for the delay | 22:11 |
| T42 | <b100dian> it seems that this cheeky named commit https://github.com/sailfishos-on-tucana/droid-config-tucana/commit/720f2335f0e14d3c186abcbaf28cea977e199bd5 was preventing audioserver to start on first boot after installation which made the device reboot in 60 seconds like https://pastebin.ubuntu.com/p/PnvmVtZVtp/ | 22:15 |
| T42 | <b100dian> The commit was the 'fix' for the race that made sound unavailable in calls if I want proximity sensor:) | 22:16 |
| T42 | <b100dian> that, coupled with mapplauncher fresh install error (which wasn't visible in OTA upgrade) cost me 2 days:P | 22:16 |
| Thaodan | @b100dian: no Problem I tried to work on that my self but didn't got the time before finishing it. | 22:27 |
| T42 | <b100dian> was there another port that said that call audio is not available on first boot? I remember reading that somewhere | 22:27 |
| T42 | <b100dian> Thaodan: can't wait to try that other plugin. I mean I try now:) | 22:27 |
| Thaodan | b100dian: What do you logs say? | 22:29 |
| T42 | <b100dian> I'm just staring to modify droid-configs | 22:29 |
| Thaodan | ok | 22:30 |
| T42 | <b100dian> or I can just zypper it | 22:30 |
| T42 | <b100dian> hmm libngf-qt5 is installed | 22:31 |
| T42 | <b100dian> but I don't have other vibra than: menus, calls, alarms, plugging in | 22:32 |
| T42 | <b100dian> should I get rid of qt5-feedback-haptics-ffmemless first | 22:32 |
| Thaodan | yes | 22:39 |
| Thaodan | only the ngfd plugin is needed that will access ffmemless | 22:40 |
| T42 | <b100dian> libngf-qt5 right? getting rif of qt5...memless requires droid-config change for me. | 22:41 |
| Thaodan | yes | 22:41 |
| Thaodan | that's normal | 22:41 |
| Thaodan | see git changelog of droid-configs repo | 22:42 |
| T42 | <elros34> use rpm to ignore deps | 22:45 |
| T42 | <b100dian> I modified droid-configs after all | 23:07 |
| T42 | <b100dian> but I don't get any extra vibrations in qt | 23:07 |
| T42 | <b100dian> I did like https://github.com/mer-hybris/droid-hal-configs/commit/d5a927170b382e75139fab5774bd49c7d8733919 | 23:07 |
| T42 | <b100dian> any other config that needs to be done for the ngf-qt5 plugin? | 23:07 |
| T42 | <b100dian> (haptics tests in components app don't do anything as before Thaodan) | 23:08 |
| Thaodan | did you remove the old qt plugin? | 23:14 |
Generated by irclog2html.py 2.17.1 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!