**** BEGIN LOGGING AT Thu Feb 15 03:00:02 2018 Feb 15 06:50:07 Herrie: https://paste2.org/gvVgm8h8 Feb 15 07:35:10 Morning! Feb 15 07:35:14 Tofe: What's that? Feb 15 08:04:08 Tofe: I guess it woud be helpful to have output of all *test* binaries so I know which outputs are normal Feb 15 08:07:12 I.e. I see some errors in mine but not sure those are normal" Feb 15 08:36:47 Herrie: That's Hammerhead's LuneOS "ps axf" Feb 15 08:37:15 ... and I realize that's not what you asked for. Never do anything before the first coffee ! Feb 15 08:38:20 Tofe: LOL Feb 15 08:38:33 Just I'd like to understand if some of the "errors" or failures I am getting are normal or not Feb 15 08:40:23 I.e. some of the sensor*_test give me "funny" outputs Feb 15 08:50:01 Tofe: Had a long night of sleep, could sleep till 6 today, luxury LOL :P Feb 15 09:06:01 :) Feb 15 09:53:03 Tofe: On https://github.com/shr-distribution/meta-smartphone/tree/pyro/meta-android/recipes-core/android-system/android-system I guess we can remove init.rc completely? Feb 15 09:53:33 https://github.com/shr-distribution/meta-smartphone/blob/pyro/meta-android/recipes-core/android-system/android-system/lxc-config#L26 should be updated to remove the hal-droid ones and add the lib64 ones? Feb 15 09:56:49 On https://github.com/shr-distribution/meta-smartphone/blob/pyro/meta-android/recipes-core/android-system/android-system/20-remove-services I think we can delete lines 4, 10, 11 and maybe 12, 13? Feb 15 09:57:15 Not really sure what the mediaserver does on Android side? Feb 15 09:58:56 Not sure about the sensorservice and surfaceflinger either... You know those better I guess Feb 15 10:06:59 Herrie|Laptop: things that are not even pointed by the recipe can be removed, yes, it'll be a little cleanup Feb 15 10:08:08 And the lxc.environment can be removed too completely yes Feb 15 10:08:20 (it's even possible it gets in the way instead of helping) Feb 15 10:09:27 For the ubuntu stuff removal, we'll just have to confirm it's actually not there anymore with Halium Feb 15 10:19:59 Tofe: If I look in the image, I don't have ubuntuappmanager, upstart-property-watcher and surfaceflinger Feb 15 10:20:31 And no ubuntu_chroot Feb 15 10:20:49 So those should be able to go for sure Feb 15 10:24:07 Question now is if we need sensorservice & mediaserver Feb 15 10:30:49 Herrie: at this point, let's just try to use directly Halium's init.rc Feb 15 10:31:01 so ignore completely 20-remove Feb 15 10:31:29 We'll have to validate it for N4,N5 and TP of course Feb 15 10:36:07 Tofe: Feb 15 10:36:09 OK Feb 15 10:36:18 Let me try that for Mido first see what it does Feb 15 10:40:02 Tofe: For udev rules, sometimes I see ACTION=="add" and sometimes without Feb 15 10:40:06 You know what's the difference? Feb 15 10:41:01 I.e. the one I generated with the command has the ACTION=="add"' in there, but the ones from Piggz doesn't have: https://github.com/piggz/droid-config-mido/blob/master/sparse/lib/udev/rules.d/999-droid-fm.rules and https://github.com/piggz/droid-config-mido/blob/master/sparse/lib/udev/rules.d/999-droid-vidc.rules Feb 15 10:52:47 Tofe: Seems our lxc-config is very similar to Halium's: https://github.com/shr-distribution/meta-smartphone/blob/pyro/meta-android/recipes-core/android-system/android-system/lxc-config#L26 and https://github.com/Halium/lxc-android/blob/9ca601468de119b1245756fe49b640202ac8ca77/var/lib/lxc/android/config Feb 15 10:54:04 Only differences are that we have lxc.environment, lxc.hook.post-stop which Halium doesn't. Halium has lxc.init_cmd = /init lxc.aa_profile = unconfined and lxc.autodev = 0 which we don't have Feb 15 11:37:11 JaMa: Hi :) Feb 15 11:42:52 Tofe: I don't seem to have adverse things on Mido when removing the 20-remove and removing the lxc.environment Feb 15 12:17:11 Herrie|Laptop: hi Feb 15 12:21:16 JaMa: I'll PR some meta-smartphone stuff later :) Feb 15 12:21:33 Almost got LuneOS to run on my Mido finally :) Feb 15 12:21:45 oh Feb 15 12:21:52 "almost"? Feb 15 12:21:53 Just gstreamer is now segfaulting preventing UI from coming up.... Feb 15 12:22:18 Herrie|Pre3: ok :) Feb 15 12:23:27 bshah: well test_hwcomposer works, just ui fails on libgstreamer stuff Feb 15 12:24:59 Herrie|Pre3: for udev, I don't know, maybe that's the default action Feb 15 13:21:44 Tofe: OK Feb 15 13:26:37 Tofe: Could you do a strace luna-next to see what happens when you start luna-next? Feb 15 13:26:54 I have a few missing bits in my strace, not sure these are critical and the cause for the gstreamer issue Feb 15 13:27:28 I have these missing: https://bpaste.net/show/e4a166b30e65 Feb 15 13:27:40 LAst 2 files I copied to the location where it expects them, but still no luck Feb 15 13:56:39 bshah: You around? Feb 15 14:09:45 yo Feb 15 14:11:20 In my udev rules I seem to have a lot of users/groups that don't seem to exist. You know where these are defined/created? Feb 15 14:11:42 I tried searching Halium, UBPorts & LineageOS GitHub but little luck Feb 15 14:11:42 it's somewhere in yur rootfs creation I believe Feb 15 14:11:52 Tofe: might know Feb 15 14:12:12 Where you do it in Halium? Feb 15 14:13:21 in halium I do e.g. This :https://github.com/Halium/rootfs-builder/blob/ubuntu/customization/hooks/01-setup_user_group.chroot Feb 15 14:13:39 this is done at rootfs creation level Feb 15 14:16:19 Ok that helps :) Feb 15 14:20:25 Herrie|Laptop: this is probably something that is done at the Yocto level Feb 15 14:20:36 But I don't know where Feb 15 14:20:42 Tofe: I think I have a clue ;) Feb 15 14:22:38 Seems to come from base_passwd I guess Feb 15 14:22:44 We have this patch in meta-android: https://github.com/shr-distribution/meta-smartphone/blob/pyro/meta-android/recipes-core/base-passwd/base-passwd/change-android-group-ids.patch Feb 15 14:24:14 Which comes from https://github.com/openembedded/openembedded-core/blob/pyro/meta/recipes-core/base-passwd/base-passwd_3.5.29.bb which gets the following https://launchpad.net/debian/+archive/primary/+files/base-passwd_3.5.29.tar.gz Feb 15 14:27:05 ok, yes, that must be it Feb 15 14:27:24 So I guess we need to do some patching there :P Feb 15 14:28:13 Additionally we might be able to do it here: https://github.com/shr-distribution/meta-smartphone/blob/pyro/meta-android/recipes-core/android-system/android-system_1.0.bb#L37 Feb 15 14:29:09 That would create the users, but not the groups I guess Feb 15 14:38:54 Tofe: We might want to have a look at this as well? https://github.com/shr-distribution/meta-smartphone/blob/pyro/meta-android/recipes-core/android-system-image/android-system-image.inc#L43 Feb 15 14:39:10 References hal-droid Feb 15 14:50:36 Herrie|Laptop: I'll go a general cleanup soon Feb 15 14:58:43 Tofe: Seems needed Feb 15 14:58:56 Not sure if these users cause issues or not Feb 15 14:59:16 Just going through logs now and fixing possible issues I see Feb 15 16:55:16 Seems there's a pretty extensive list here we could use: https://github.com/LineageOS/android_device_motorola_addison/blob/cm-14.1/include/private/android_filesystem_config.h#L45 Feb 15 16:55:28 Even Halium guys are missing quite some it seems Feb 15 16:56:02 mer-hybris has some more: https://github.com/mer-hybris/hybris-initrd/blob/937e1a21214355a7b5cff33e90f47a43081ed9b1/etc/passwd Feb 15 16:56:19 Question is of course if we need all, but we can have a look & see for the targets which ones are required Feb 15 16:56:46 That should be fairly easy to grep from LineageOS, AOSP and CAF repos Feb 15 18:05:18 Tofe: I guess I could do a quick test where I simply add the missing users from my journalctl to see if that improves anything on logcat & luna-next side Feb 15 18:35:04 Herrie|Pre3: https://paste2.org/pdVkpztJ test_sensors output on N5 Feb 15 18:35:41 But that's no surprise, it has always crashed. Feb 15 18:38:07 and strace luna-next: https://paste2.org/XZk6wB4v Feb 15 19:00:26 Tofe: Thnx Feb 15 20:33:30 Tofe: You get QTI Sensors Module and 19 sensors ;) I don't get those ;) Feb 15 21:21:06 oh. Feb 15 21:39:41 I somehow managed to kill logcat :P Feb 15 21:39:46 Time for a new image it seems Feb 15 21:39:57 THat's for tomorrow Feb 15 21:40:06 Also got Tenderloin building for overnight Feb 15 21:40:19 Tomorrow at home so some time to play hopefully Feb 15 21:41:23 gn8 **** ENDING LOGGING AT Fri Feb 16 03:00:01 2018