**** BEGIN LOGGING AT Tue Jan 30 03:00:02 2018 Jan 30 05:41:40 (Photo, 480x800) https://irc.ubports.com/PVVBYRDj/file_3943.jpg Jan 30 05:50:34 @maharudra108, See the pin message, post logs not screenshots, use paste.ubuntu.com or something similar like that Jan 30 05:56:20 @zoomer296, Then create it, and replace [manufacturer] and [device] Jan 30 06:30:35 @NSA_AMERICA, Or termbin! Jan 30 10:02:35 how can I add additional kernel commandline parameters to the BoardConfig in order that hybris-boot picks them up? Is seams I need to set "audit=0 selinux=0" Jan 30 10:04:14 @zoomer296, If the whole folder is missing, you need to update your tree. The repo has been added a few days ago Jan 30 10:33:28 ZeroPointEnergy: BOARD_KERNEL_CMDLINE or something like that env var Jan 30 10:33:35 just += additional items Jan 30 10:35:49 ok, thx Jan 30 12:14:03 @ErnyTech, How is your port going? Jan 30 12:14:30 @unknown, Now i can't develop Jan 30 12:15:24 @ErnyTech, Are you making progress? … What does already work? Jan 30 12:16:18 @unknown, System boot but can't complete access to hardware because the vendor blobs are arm64 only Jan 30 12:17:10 @ErnyTech, Okay so it is a problem with arm64 in general? Jan 30 12:17:38 @unknown, It's a problem for some new arm64 devices Jan 30 12:18:56 I'm trying fix it with build a arm64 rootfs Jan 30 12:19:26 @ErnyTech, He can fix the problem on all devices Jan 30 12:30:52 @ErnyTech, Great 🤠 Jan 30 12:30:59 @unknown, most of arm64 devices usually provide 32-bit vendor blobs Jan 30 12:31:29 @NotKit, Okay Jan 30 12:49:44 @NotKit, Yes but why use armhf rootfs when we can use a aarch64 Jan 30 14:28:33 SerialNumber: GNU/Linux Device on rndis0 10.15.19.82 ----- FINALLY!! Jan 30 14:28:46 ZeroPointEnergy: congrats :) Jan 30 14:28:51 thx :-D Jan 30 14:30:02 there where some additional settings missing which mer-checker did not catch. went trough the list here: https://github.com/systemd/systemd/blob/master/README Jan 30 14:31:20 But seriously, systemd is a pain to debug if you have no shell. The tools just don't work if there is no daemon running Jan 30 14:58:02 Guys i've got a problem, rsyslog is flooding 2 files in /var/log/syslog and /var/log/kern.log and i haven't got space in roots.img ... What can i do? Jan 30 15:42:49 you have space left on the device but only the rootfs is too small you could just pull it to your machine with adb, resize it with resize2fs and then push it to the phone again Jan 30 15:43:35 well no Jan 30 15:44:04 I'd fix what is flooding syslog Jan 30 15:44:19 if it is kernel message read the code and see why it cries or is unhappy and fix it Jan 30 15:44:33 if it's some binary blob, I would try to kill it with plain fire Jan 30 15:59:33 and if plan fire wont do, try magic fire Jan 30 16:00:11 :P Jan 30 16:00:28 your device is on now fire.. :P Jan 30 16:16:02 Anyone please help me on adding mount points. I can't understand what really i have to do with fstab.* Jan 30 16:23:45 @maharudra108, check what mointpoints are there, add them to fixupmountpoint script , Jan 30 16:24:08 with they same format as they are discribed in docs Jan 30 16:26:13 Yeah but there mount points isn't described by block it is in the format of emc@android .... Jan 30 17:44:32 Progromizd was added by: Progromizd Jan 30 17:46:36 hello! I'm stuck on: … "First, you’ll want to find the repositories for your device on LineageOS’s GitHub organization." … - what if there is no repo for my device_name 5.1 ? How to be in this situation? … thanks! Jan 30 17:48:00 What is the codename of the device you want to port? Jan 30 17:48:34 Maybe no LineageOS port exists, then porting Halium will not be easily possible Jan 30 17:51:06 @JBBgameich DOOGEE X5PRO (with 5.1 currently installed) Jan 30 17:52:02 5.1 stock ROM or Cyanogenmod? Jan 30 17:52:12 stock Jan 30 17:52:17 rooted Jan 30 17:52:40 Does it have lineage or anyother custom rom? Jan 30 17:53:55 Oh, it's a MediaTek device? Jan 30 17:54:25 But here is a device-tree for it: https://github.com/nE0sIghT/android_device_doogee_x5pro Jan 30 17:54:32 But for LineageOS 14.1 Jan 30 17:55:01 Here is one for 12.1 (android 5): https://github.com/nE0sIghT/android_device_doogee_x5pro/tree/cm-12.1 Jan 30 17:55:45 @JBBgameich unfortunatelly, yes, MTK, and its support both Cyanogen 12.1 and LOS 14.1 Jan 30 17:56:08 @Progromizd is it 32-bit or 64-bit? Jan 30 17:57:14 https://github.com/nE0sIghT/android_device_doogee_x5pro/blob/cm-14.1/prebuilt/kernel - that's the problem Jan 30 17:57:36 I wonder why he didn't use https://github.com/SnowCat6/DOOGEE-kernel-3.18 Jan 30 18:03:52 @NotKit 64bit … current kernel v3.4.5 … Thanks for link, I'll check it Jan 30 18:04:26 you need to build custom kernel, since Halium requires kernel different configuration from normal Android builds Jan 30 18:13:11 @NotKit so, I firstly need to compile kernel 3.18 (from link above) and than think about what to do next, right? Jan 30 18:13:32 no, you need to modify device tree to build kernel from source Jan 30 18:13:40 instead of using prebuilt one Jan 30 18:13:43 no, you need to change BoardConfid.mk to build the kernel in the Halium tree Jan 30 18:14:16 @bshah, I've already resized rootfs by adding 1gb... Everytime the phone turnes on rsyslogd start to occupies 180% of CPU and if i cat syslog file i see i cycling dmesg... Sorry for my bad english Jan 30 18:14:30 @Progromizd, So, first clone the Halium tree, as the porting guide says and add the kernel and device tree. Then modify the device tree to build the kernel (look at other devices for inspiration) Jan 30 18:18:12 @JBBgameich, Are steps for mediatek device different? Jan 30 18:20:56 https://hastebin.com/uqelogegiv.sql Jan 30 18:23:06 @maharudra108, The device uses a prebuilt kernel. You need to find a source replacement for it. Jan 30 18:26:12 @JBBgameich, Means? Jan 30 18:26:19 I have kernel source Jan 30 18:26:32 But the device tree tries to use a prebuilt kernel Jan 30 18:26:39 change BoardConfig.mk then Jan 30 18:26:46 Okay Jan 30 18:27:55 @JBBgameich, In device/ Jan 30 18:28:10 device/v$endor/$codename Jan 30 18:28:21 Okay Jan 30 18:30:48 What to change there Jan 30 18:32:28 add at least … ```TARGET_KERNEL_CONFIG := $NAME_OF_DEFCONFIG … TARGET_KERNEL_SOURCE := kernel/$VENDOR/$CODENAME``` Jan 30 18:32:32 ld/core/base_rules.mk:316: warning: igno … ring old commands for target `/root/prithvi … /halium/out/target/product/a106/system/vend … or/lib/mediadrm/libdrmclearkeyplugin.so' … Starting build with ninja … ninja: Entering directory `.' … ninja: error: '/root/prithvi/halium/out/tar … get/product/a106/android-boot.img', needed … by '/root/prithvi/halium/out/target/product … /a106/obj/PACKAGING Jan 30 18:32:33 tes/lineage_a106-target_files-5a24fffc41.zi … p', missing and no known rule to make it … make: *** [build/core/ninja.mk:152: ninja_w … rapper] Error 1 … make: Leaving directory '/root/prithvi/hali … um' … #### make failed to build some targets (14 … seconds) #### Jan 30 18:33:13 Sorry i pasted log here Jan 30 18:38:54 I guess there is still some prebuilt kernel stuff in BoarConfig.mk Jan 30 18:40:48 Look at an other device tre and how kernel stuff is done there: … https://github.com/Halium/android_device_lge_hammerhead/blob/ubp-5.1/BoardConfig.mk#L110 Jan 30 18:41:47 Or better this one: … https://github.com/Halium/android_device_lge_bullhead/blob/halium-7.1/BoardConfig.mk#L35 Jan 30 18:45:33 @NotKit executing cmd: … repo init -u https://github.com/Halium/android -b halium-5.1 … I see … curl: (22) The requested URL returned error: 404 Not Found … Server does not provide clone.bundle; ignoring. … - Is it normal? Jan 30 18:45:59 Yes, that's normal and shouldn't be a problem Jan 30 18:46:13 it's normal, but I think 5.1 branch isn't properly tested/maintained Jan 30 18:46:44 I think it should work, but if you can port 7.1, maybe better do that Jan 30 18:47:57 @JBBgameich , yes, thank you for tip Jan 30 19:02:43 @NotKit how and where "add the kernel and device tree" ? … Since Android 7.1 assume more memory consumption, I'd like to build for 5.1 … sorry for primitive questions Jan 30 19:03:33 Halium's android is very minimal, only for hardware adaption. I don't think there will be any memory differences Jan 30 19:22:57 @Progromizd to your local_manifest.xml Jan 30 19:23:03 it's probably mentioned in Halium docs Jan 30 19:24:09 https://docs.halium.org/en/latest/porting/get-sources.html#adding-your-device-specific-source Jan 30 19:25:56 as for building kernel, check https://github.com/danielhk/android_device_lenovo_aio_otfp/blob/lineage-15.0/BoardConfig.mk#L53 as example Jan 30 19:29:42 @NotKit, yes, mentioned. … "Navigate into your Halium directory and edit the file..." … means "create the file if it doesnt exists" ? Since my local halium dir contains only 1 hidden '.repo' folder with 2 folders inside Jan 30 19:29:57 yes, create it Jan 30 19:30:28 @Progromizd, Did repo sync fail? Jan 30 19:30:59 I think local_manifest.xml isn't created by default? Jan 30 19:33:22 Docs say to create a new manifest in halium/devices/manifests I think. Jan 30 19:33:40 And if repo sync didn't succeed, the folder doesn't exist Jan 30 19:34:03 so before adding device specific repos, the tree needs to be fully downloaded and checked out Jan 30 19:35:03 ah, ok, this is different from HADK then Jan 30 19:35:40 the script in halium/devices copies it to .repo/local_manifests then Jan 30 19:35:45 so in the end it's the same Jan 30 19:36:21 @NotKit by default there was a file named .. /halium/.repo/manifests/default.xml, containing a lot of config-lines. … after repo init -u https://github.com/Halium/android -b halium-7.1 … I'v started … repo sync. … now waiting until download finish. Jan 30 19:36:57 ok, wait then Jan 30 19:48:38 is netd.rc disabled in normal Halium builds? Jan 30 20:08:29 @NotKit 1.3GB downloaded and continue. I don't know the final size of data Jan 30 20:09:00 I think the docs say you should have 20GB free storage Jan 30 20:09:07 So there's still a lot to download Jan 30 20:27:17 @NotKit is it necessary to install Andorid 7.1 onto device before in order to port using "halium-7.1 based on LineageOS 14.1" ? Jan 30 20:27:41 better yes Jan 30 20:28:20 but it depends on device Jan 30 20:28:47 if Android 7.1 doesn't flash new firmware to other partitions (I doubt unofficial port does this), it shouldn't be required Jan 30 20:31:23 (Document) https://irc.ubports.com/qKzVE7gk/file_3946.mp4 Jan 30 20:33:52 @NotKit if truth, DOOGE_X5PRO (x64, 16GB, 2GB RAM) suports Android up to 8.0 … Early, on other resources, I sow that it was possible to install Ubuntu Touch with Android in DualBoot mode Jan 30 20:34:28 no idea if someone ever did this Jan 30 20:38:29 @NotKit I have not tested, gues, maybe only specific devices is supported: … https://wiki.ubuntu.com/Touch/DualBootInstallation … After download finished (tomorrow), I'll probably back to continue an experiment. … Thank you very much for your support! Jan 30 20:41:40 that page is really old Jan 30 20:47:33 hello, I am trying to port halium on tenshi (bq aquaris U Plus): I used hybris-recovery where I can connect via telnet but not via ssh. I ran echo "continue" >/init-ctl/stdin device rebooted and I can access via telnet again. /target is populated correctly Jan 30 20:47:57 Any tips to understand why it doesn't boot to rootfs? Jan 30 20:48:11 any recent changes in rootfs? Jan 30 20:50:37 any suggestion how to move from initrd to /target and start the init system? Jan 30 21:13:36 (Photo, 1280x303) https://irc.ubports.com/IWCDwjFY/file_3948.jpg Jan 30 21:13:43 there's an issue with docs. how could that work if the tree is not yet synced? Jan 30 21:14:07 someone removed the "sync" part completely from the docs at some point -_- Jan 30 21:15:51 i will find a person to blame now Jan 30 21:16:43 `git blame` Jan 30 21:17:03 https://github.com/Halium/docs/commit/a071b1bec89d6fc25972a058aecaabe5cf53914b#diff-6bfe11ac64e0b29ecde918f60834cb55 Jan 30 21:17:06 there's that commit Jan 30 21:17:37 i mean it makes no sense. `repo init` leaves a user with a small `.repo` folder. there's no `./halium/devices/setup` whatsoever Jan 30 21:17:56 (Photo, 184x172) https://irc.ubports.com/tQKrFhbo/file_3950.jpg Jan 30 21:19:48 Yeah, repo sync needs to be run one time after repo init Jan 30 21:20:16 thank you for confirming that. i needed to make sure i'm sane :P Jan 30 21:21:14 who would fix that? Jan 30 21:21:19 (please not me) Jan 30 21:21:31 @Ellenjott I guess :) Jan 30 21:21:42 affirmative. thanks ;) Jan 30 21:22:53 Sorry!! I'll fix it Jan 30 21:30:35 https://github.com/Halium/docs/pull/65 Jan 30 21:30:46 so I got Halium container to start fine on Gemini, but for some reason modem service doesn't get activated Jan 30 21:32:23 @vanyasem, See https://github.com/Halium/docs/pull/65 Jan 30 21:34:58 @Ellenjott, @vanyasem I think it would be a good idea to directly merge the fix Jan 30 21:35:12 it's merged Jan 30 21:35:14 +1 Jan 30 21:35:20 so what's the problem Jan 30 21:35:37 i merged it Jan 30 21:36:01 (Photo, 1280x401) https://irc.ubports.com/y0IP6LtL/file_3952.jpg Jan 30 21:40:15 it seems like there's a new UBports rootfs out https://ci.ubports.com/job/xenial-rootfs-armhf/ Jan 30 21:40:34 has anybody tried it? what's new ? Jan 30 21:43:56 @bhushanshah what was the solution for missing /dev/socket/property_service? Jan 30 21:47:10 @da Jan 30 21:47:26 has dalton left this chat Jan 30 21:47:36 Hmm Jan 30 21:47:50 alright Jan 30 21:47:59 telegram bugged on me and refused to @ you Jan 30 21:48:16 please clarify the situation with all the different rootfs for UBports on Jenkins Jan 30 21:48:35 https://ci.ubports.com/blue/organizations/jenkins/xenial-7.1-caf-sudoku-rootfs/activity/ … https://ci.ubports.com/blue/organizations/jenkins/xenial-rootfs-armhf/activity/ … https://ci.ubports.com/blue/organizations/jenkins/xenial-rootfs-arm64/activity/ … i can see 3 different rootfs Jan 30 21:48:56 Well Jan 30 21:49:01 which one is meant to be used with halium? all of them? Jan 30 21:49:04 One says caf, one says arm64, and the other has none Jan 30 21:49:19 But actually now just use the armhf one for caf as well. We fixed it. :D Jan 30 21:49:28 https://docs.ubports.com/en/latest/porting/installing-16-04.html#download-the-rootfs … docs only mention the normal one Jan 30 21:49:32 Yep Jan 30 21:49:42 I changed that on purpose Jan 30 21:49:59 which one should i use for CAF aarch 64 HTC 10? … and which one for non-caf nexus 5x ? (forgot the architecture) Jan 30 21:50:09 ok, found https://github.com/Halium/android_system_core/pull/1 Jan 30 21:50:10 just the normal one? Jan 30 21:50:33 You might try arm64... but I don't think it works yet Jan 30 21:50:37 Just use armhf regular Jan 30 21:50:47 ok.. thank you Jan 30 22:00:10 @UniversalSuperBox, so i should use one image for all devices atm, right? Jan 30 22:00:22 no matter arm or aarch, caf or not? Jan 30 22:00:24 am i right? Jan 30 22:02:50 @JBBgameich why have you made your script so compicated over time lol. it's a big pain to patch it to work with ubports now Jan 30 22:05:13 It should work with ubuntu touch already Jan 30 22:05:46 At least once Ubuntu Touch uses the new initramfs-tools based initrd Jan 30 22:06:31 But the hook for ubuntu touch (symlinks etc.) Is already there Jan 30 22:07:27 @JBBgameich, i don't think so, i once had a working version of your script, but never commited it Jan 30 22:08:10 (Photo, 1280x332) https://irc.ubports.com/5ltBURa4/file_3954.jpg Jan 30 22:08:19 that's what i had to add at the end of the script Jan 30 22:08:29 and i also added some stuff in between the lines as well Jan 30 22:09:30 (Photo, 1197x1280) https://irc.ubports.com/BM6WsFyD/file_3956.jpg Jan 30 22:09:54 That's the same code adapted for running on the host Jan 30 22:09:55 i guess i am looking at the wrong file Jan 30 22:10:32 functions/post-install.sh Jan 30 22:12:29 i was looking at the wrong file in fact Jan 30 22:12:40 i will test that stuff tomorrow :P thank you Jan 30 22:12:59 @NSA_AMERICA JBB already commited patches to the script, stop bugging me about that Jan 30 22:23:24 docs are broken even more Jan 30 22:24:06 (Photo, 800x130) https://irc.ubports.com/W7W1anoQ/file_3958.jpg docs say to place the device XML to manifests when it should be local_manifests Jan 30 22:24:44 @Ellenjott also your responsibility btw Jan 30 22:25:10 so on MediaTek: … * /fstab.{SoC} has to be present … * /nvram needs to be mounted by mount_all or modem service won't get launched Jan 30 22:29:21 I guess either nvram_daemon or other closed service tries to ask /nvdata status either from vold or from other source and fails even though it's mounted Jan 31 01:07:22 i can't get hwcomposer to work on Nexus 5x on reference halium rootfs Jan 31 01:07:38 @vanyasem any chance you have experimental libhardware patch by Marius.? Jan 31 01:07:50 tried 2 system.img's. one from images.plasma, second built myself from the tree Jan 31 01:07:59 ```~# EGL_PLATFORM=hwcomposer test_hwcomposer … library "libmm-qdcm.so" not found … width: 1080 height: 1920 … test_hwcomposer: test_hwcomposer.cpp:301: int main(int, char**): Assertion `eglGetError() == EGL_SUCCESS' failed. … Aborted … root@localhost:~# strace EGL_PLATFORM=hwcomposer test_hwcomposer … strace: Can't stat 'EGL_PLATFORM=hwcomposer': No such file or directory``` Jan 31 01:08:02 @bhushanshah, ? Jan 31 01:08:22 @vanyasem, i should've pasted that to pastebin Jan 31 01:08:25 sorry Jan 31 01:08:39 Please give me mount output Jan 31 01:09:28 https://paste.ubuntu.com/26492435/ Jan 31 01:10:43 the weird thing, plasma boots successfully on that system.img Jan 31 01:10:53 there is something fundamentally broken in the reference rootfs Jan 31 01:11:02 @vanyasem, and i can run test_hwcomposer from plasma Jan 31 01:13:37 @vanyasem, I see vendor not mounted Jan 31 01:18:19 Hmmm Jan 31 01:18:29 *yawns* Jan 31 01:18:58