**** BEGIN LOGGING AT Fri Dec 01 15:21:58 2017 Dec 01 15:23:06 bshah[m], thanks Dec 01 15:23:30 ka6sox: can you give me link? Dec 01 15:24:12 let me look now and give that to you...brb Dec 01 15:24:24 Okay Dec 01 15:25:18 http://logs.nslu2-linux.org/livelogs/halium.txt for Live Logs Dec 01 15:25:24 and Dec 01 15:25:40 http://logs.nslu2-linux.org/livelogs/halium/ for previous logs Dec 01 15:25:45 (by UTC day) Dec 01 15:26:15 What is the command to build halium libhybris? I wanna do some tests :D Dec 01 15:29:03 cristo: basically, get the source for libhybris using apt source libhybris and then use dpkg-buildpackage to build it Dec 01 15:29:42 Ahh, and it will build with my modifications, right? Dec 01 15:30:09 You can modify it later on yeah Dec 01 15:30:27 Nice, ill play with that flag Dec 01 15:30:32 Thx f;r help Dec 01 15:39:34 ka6sox: I'd assume logs will show up 24 hours later? currently it's 404 :) Dec 01 16:04:22 the file should be open and happening..let me see if it was created with wrong permissions. Dec 01 16:04:53 bshah: now i install all the packages in .. ? Dec 01 16:06:25 got it Dec 01 16:20:49 still dlopen failed. i'll ask in libhybris, to see if there's a way, or maybe textrels are supported Dec 01 16:34:00 bshah: i've sorted the halium logs now, you should be able to see today's & it'll be archived tonight as ka6sox described Dec 01 16:34:36 novaldex: Okay, thanks :-) **** ENDING LOGGING AT Sat Dec 02 02:59:59 2017 **** BEGIN LOGGING AT Sat Dec 02 02:59:59 2017 Dec 02 11:10:24 Faqeerilallah was added by: Faqeerilallah Dec 02 20:49:47 Hello. What roms can i use with halium 5.1 img? Dec 02 21:08:55 Is there anything I can do to help with the OP5 tree? Dec 02 21:09:15 Or is it just about there? Dec 02 21:28:37 @UniversalSuperBox hm I halted the boot process and chrooted into the ubuntu rootfs Dec 02 21:28:44 I could dist-upgrade everything.. Dec 02 21:28:50 but it would add systemd.. Dec 02 21:29:19 Anyone can suggest me a rootfs for 5.1? Dec 02 21:29:24 Compatible? Dec 02 22:00:26 I don't know which rootfses support 5.1 Dec 02 22:00:36 Go porting 7.1 instead Dec 02 22:01:35 Some Plasma Mobile rootfs and some UBports rootfs already worked on 5.1 Dec 02 22:01:45 And LuneOS with some changes Dec 02 22:02:44 Well... Dec 02 22:02:56 Luneos works? Dec 02 22:03:01 What changes? Dec 02 22:03:16 You have to use their initrd instead of hybris-boot Dec 02 22:03:43 but better ask the LuneOS devs Dec 02 22:03:45 Ah Dec 02 22:03:54 Where? Dec 02 22:12:04 Not sure, maybe they have an IRC channel Dec 02 22:13:42 I take it nobody is here who knows where I can help out with Halium in regards to cheeseburger? Not even point me to somewhere I can see what I can do with little programming knowledge? Dec 02 22:14:56 I can ping @mariogrip and he might reply Dec 02 22:15:25 Here is the manifest for cheeseburger, the sourcecode is in the halium organization … https://raw.githubusercontent.com/JBBgameich/halium-devices/halium-7.1/cheeseburger/manifest.xml Dec 02 22:15:50 @JBBgameich, And that's built similar to Android, right Dec 02 22:15:53 Halium-devices Dec 02 22:15:55 Novel! Dec 02 22:15:57 ? Dec 02 22:16:01 Yes Dec 02 22:16:08 @c_smith, Hay, it has a big problem that android has changed to configfs that makes it hard to get debugging working on it Dec 02 22:16:08 Luneos irc channel has 1 user -_- Dec 02 22:16:44 Then it's probably not the right one Dec 02 22:16:45 Then I know what I'm doing today, gonna hammer on the cheeseburger tree and see where I get Dec 02 22:17:22 @Iandore[m], What's Luneos? Sounds interesting... Dec 02 22:17:38 A WebOS continuation Dec 02 22:17:51 Got it Dec 02 22:17:55 Continuation of HP WebOS, a continuation of Palm WebOS Dec 02 22:18:00 C_smith an os Dec 02 22:18:18 There, brought it home Dec 02 22:18:19 WebOS was neat when I tried in grouper years and years ago Dec 02 22:26:47 Bhushan Shah: is halium libhybris built with https://github.com/libhybris/libhybris/commit/54dd4749706334882f9c404fca01a19f01325d07 ? It seems promising ^^ Dec 02 22:27:21 cristo: yes Dec 02 22:27:43 Oh Dec 02 22:27:45 Kk Dec 02 22:49:48 For halium 5.1 caf, normal rootfs will work? **** ENDING LOGGING AT Sun Dec 03 03:00:01 2017 **** BEGIN LOGGING AT Sun Dec 03 03:00:02 2017 Dec 03 04:59:36 @bhushanshah, How do get the rild log ? is it from logcat ? Dec 03 05:10:00 thanks Dec 03 05:17:17 test Dec 03 05:17:23 hm Dec 03 05:17:42 wonder why previous message didn't get relay to IRC Dec 03 05:18:10 ah it was from / Dec 03 05:18:19 yep Dec 03 07:06:26 Anyone have success with halium on devices with a 4.4 kernel? Dec 03 07:11:50 @UniversalSuperBox, Is MarioGrip still AWOL? Dec 03 07:34:33 Hello. Dec 03 07:35:05 How I can download arm-eabi-4.7 toolchain? Dec 03 07:35:32 I can not find any documents Dec 03 07:37:28 there is following in manifest.xml Dec 03 07:37:30 Dec 03 07:37:32 Dec 03 07:37:39 possibly try adding one for 4.7 Dec 03 07:37:43 Or what different arm-eabi and arm-none-eabi? Dec 03 09:11:52 > For halium 5.1 caf, normal rootfs will work? Dec 03 09:11:53 ^ Dec 03 09:19:59 @i_love_halium_on> , I had to use non-caf plasma mobile with my caf hybris-boot and system.img Dec 03 09:20:37 And it worked? Dec 03 09:23:53 Rk585 was added by: Rk585 Dec 03 09:24:43 Fwd from akshat2K2: https://hastebin.com/quhidecowe.vbs Dec 03 09:24:45 Help Dec 03 09:29:55 Ive seen a commit about that somewhere Dec 03 09:32:09 Ah no, my bad, it was similar :D Dec 03 09:32:38 Help plZ Dec 03 09:32:55 Dunno Dec 03 09:33:25 "please click on my vbs link" isn't something that'll work here, lad Dec 03 10:01:40 Tofe: despite .vbs instructions it's just normal webpage :P Dec 03 10:03:39 bshah: well, yes, but it's not tempting :) Dec 03 10:03:48 ikr Dec 03 10:32:40 @akshat2K2, No one knows about it Dec 03 10:33:16 @bshah, Tf it's a bot ? Dec 03 10:34:18 Yes Dec 03 10:34:18 It's IRC bridge Dec 03 10:34:47 @akshat2K2, Is it a arm64 device? Mine is arm64 but has only 32 bit Android, so I'm not sure if anybody has ever compiled for an arm64 device yet. Dec 03 10:44:25 @bshah, So it's copy msgs from there ? Dec 03 10:44:36 yes Dec 03 12:59:15 @JBBgameich, Yes arm64 Dec 03 13:11:18 akshat2K2: I built 7.1 based Halium with gcc 6.3 yesterday :) Dec 03 13:11:40 akshat2K2: That was for Redmi Note 4 (mido) target Dec 03 13:11:53 Had some issues with n-linker on libhybris but was easy to sort Dec 03 13:47:55 U didn't got that errors ? Dec 03 13:48:44 Can u tell the fix Dec 03 13:53:04 @akshat2K2 can you link me your device tree sorces? Dec 03 13:53:09 Ok Dec 03 13:53:31 https://github.com/RiteshSaxena/device_xiaomi_land Dec 03 13:58:04 Where is the halium for oneplus 3? Dec 03 13:58:17 alshat2k2: https://github.com/libhybris/libhybris/pull/365 Dec 03 14:12:22 This should fix my error ? Dec 03 14:12:39 No your issue and bug Herrie have is totally different Dec 03 14:12:47 Ohk Dec 03 14:13:48 akshat2K2: Any reason why you're trying 5.1 based and not 7.1 based? Dec 03 14:16:31 Sir I am making 7. 1 not 5.1 Dec 03 14:17:01 How would I go upgrading a 5.1 based AOSP source to 7.1 Dec 03 14:17:58 I have an Amazon otterx (Kindle Fire 1st Gen with FOSS bootloader) and the dude has posted some experimental 7.1 builds that seems to have everything working but hasn't pushed his changed to github Dec 03 14:20:23 @bhushanshah, Sir got any clue ? Dec 03 14:22:05 I did tell you already what can be tried in other channel Dec 03 14:22:54 LOCAL_MULTILIB := 32 maybe Dec 03 14:23:18 OK let me trt Dec 03 14:24:38 you should add this whereever LOCAL_MODULE is defined Dec 03 14:26:14 Ok Dec 03 16:42:54 ankaos was added by: ankaos Dec 03 17:46:33 @mariogrip @UniversalSuperBox what uports ppa should be preferred? Dec 03 17:48:56 hm with mir 0.26 the android platform stopped being able to create a gralloc module .. Dec 03 17:54:45 I think repo.ubports.com is the official repository now, not the launchpad ppa. Dec 03 18:19:53 @JBBgameich, yeah we are kinda switching to this Dec 03 18:20:06 since we also got now more build hardwarepower we dont even need lp builds anymore Dec 03 19:19:47 sure there are a bunch of source to pick from .. Dec 03 19:19:53 a mir 0.28 test.. Dec 03 19:20:13 another one marked as being for debug / test.. Dec 03 20:15:08 Lo sdrogo Dec 04 02:12:32 Where can I find the halium status for nexus 5? Dec 04 02:16:29 https://github.com/Halium/projectmanagement/issues/8 Dec 04 02:16:44 it's a reference device **** ENDING LOGGING AT Mon Dec 04 03:00:01 2017 **** BEGIN LOGGING AT Mon Dec 04 03:00:02 2017 Dec 04 13:50:15 sopernam was added by: sopernam Dec 04 13:50:19 h Dec 04 14:16:52 hey how to pull device teee and kernel ? Dec 04 14:16:54 (Photo, 720x1280) https://irc.ubports.com/Hkp54ve1/file_2563.jpg Dec 04 14:18:37 kel Dec 04 14:18:39 Kek Dec 04 14:21:36 Where can i find reference 5.1 rootfs? Dec 04 14:22:54 cristo: it's same roots Dec 04 14:23:30 @LLJYY, 😭 Dec 04 14:23:48 ? Dec 04 14:23:55 @sopernam, Are you really going to build Halium on a phone?.. Dec 04 14:24:50 @ilyaishere, SSH to a server perhaps.. Dec 04 14:25:04 Hope so :) Dec 04 14:25:14 it is a SSH Dec 04 14:25:26 it's definitely not on his phobe Dec 04 14:25:32 lol Dec 04 14:25:49 What device do you use, @sopernam ? Dec 04 14:25:55 Ah yeah JuiceSSH Dec 04 14:26:00 He uses a sumire Dec 04 14:26:02 Z5 Dec 04 14:26:18 (Photo, 720x1280) https://irc.ubports.com/vQTglQW1/file_2565.jpg Dec 04 14:26:44 @LLJYY, ty ❤️ Dec 04 14:26:46 @bhushanshah, no termux Dec 04 14:27:00 @sopernam, Yeah, that's why I asked Dec 04 14:27:20 I really know a person who does some builds with Termux... Dec 04 14:27:52 You need to find device tree and blob repositories on the web yourself... Dec 04 14:27:59 how can i pull device tree and kernel ? Dec 04 14:28:05 Same for the kernel Dec 04 14:28:19 Just Google a bit Dec 04 14:29:21 https://github.com/LineageOS/android_device_sony_sumire Dec 04 14:29:24 http://docs.halium.org/en/latest/porting/get-sources.html#adding-your-device-specific-source Dec 04 14:32:27 https://github.com/LineageOS/android_kernel_sony_msm8994/commits/cm-14.1 Dec 04 14:33:16 For the correct repos, just do `git clone ` Dec 04 14:33:32 Later, you can create a local manifest Dec 04 14:33:47 😭 Dec 04 14:48:44 (Photo, 720x1280) https://irc.ubports.com/ZNNhxcN0/file_2567.jpg Dec 04 14:50:25 Where did you put the device repository Dec 04 14:51:13 i dont know ? Dec 04 14:51:31 Well then how are you expecting to do something that requires the device repository? Dec 04 14:51:51 https://docs.halium.org/en/latest/porting/get-sources.html#adding-your-device-specific-source Dec 04 14:59:20 @UniversalSuperBox, I have no idea how to do it Dec 04 14:59:29 Well, read doc? Dec 04 14:59:34 It mentions how to do it Dec 04 15:00:34 > cristo: it's same roots Dec 04 15:00:34 Yes but the link? :P Dec 04 15:00:49 cristo: docs.halium.org :P Dec 04 15:01:04 If you're lazy I'm double lazy :P Dec 04 15:01:11 hides Dec 04 15:02:22 Aww Dec 04 15:03:30 Not lazy. I didnt know there were the link in docs as jt's for 3-4 months that i havent read the docs ;) Dec 04 15:10:05 @bhushanshah, i dont understand Dec 04 15:10:07 :( Dec 04 15:10:33 @sopernam, Have u ever built a ROM ? Dec 04 15:10:43 @bhushanshah, Sir that fix didn't work Dec 04 15:11:33 @sopernam … … you careate some Dec 04 15:11:34 manifest Dec 04 15:11:40 (this one is for bullhead) Dec 04 15:11:42 @akshat2K2, no thats first Dec 04 15:11:52 @bhushanshah, thank youuuuu Dec 04 15:11:54 @sopernam, Sir.. Dec 04 15:12:01 Build los first Dec 04 15:12:11 Ubtouch is not for u then Dec 04 15:12:27 Well, building los is not requirement for halium.. Dec 04 15:12:40 @bhushanshah, Atleast he should have knowledge Dec 04 15:12:46 @akshat2K2, I've no idea how to further fix this sorry Dec 04 15:12:51 @akshat2K2, I didn't. Dec 04 15:13:04 @bhushanshah, OK anyway thnz for your help Dec 04 15:13:10 @UniversalSuperBox, ;-; Dec 04 15:13:16 https://forum.xda-developers.com/showthread.php?t=2329228, "Understanding the Manifest" @sopernam Dec 04 15:13:51 @bhushanshah, Stick it here? .repo/local_manifest/sony_sumire.xml Dec 04 15:14:02 And "Understanding local manifests" Dec 04 15:14:12 @sopernam, Works for me Dec 04 15:14:21 But not that one exactly Dec 04 15:14:30 That's for the Nexus 5X Dec 04 15:14:33 We're not making this for you Dec 04 15:14:56 Who will look into supporting arm64 Dec 04 15:15:08 ;-; Dec 04 15:15:29 Can you umm have some patience please? Dec 04 15:16:14 HAHA LO SDROGOO Dec 04 15:16:17 For now, please report bug on https://github.com/halium/projectmanagement/issues/ so we have that on our todo list Dec 04 15:17:36 @bhushanshah, OK sir Dec 04 15:18:03 (also no need to call me or anyone sir) Dec 04 15:18:04 :P Dec 04 15:18:14 Ok Dec 04 15:19:05 @akshat2K2 arm-v8a is (almost) fully compatible with v7 systems Dec 04 15:19:25 so in theory, these 32-bit builds work, and work for all Dec 04 15:20:00 only issue is device level stuff for which you can't find blobs or get it to work ( like gps in my case ) Dec 04 15:21:29 Dec 04 15:21:34 is that true ? Dec 04 15:22:14 No, for you the remote will be los Dec 04 15:22:22 And also branch is wrong Dec 04 15:22:24 @trivial_inanity, So what to do Dec 04 15:22:50 um; what issues are you getting? I dont remember having any on my msm8939 Dec 04 15:22:56 ( although it's been months since i've built ) Dec 04 15:23:13 Somehow busybox is not building with some arm64 reloccation errors Dec 04 15:23:14 Dec 04 15:23:22 @sopernam, now @bhushanshah Dec 04 15:23:28 no? Dec 04 15:23:46 😭 Dec 04 15:24:19 Dec 04 15:24:26 @sopernam replace … `remote="hal"` with `remote="los"` Dec 04 15:24:29 oh well Dec 04 15:24:39 @trivial_inanity, thank you Dec 04 15:25:04 @bhushanshah, @trivial_inanity Dec 04 15:25:31 what is the true ? … "halium/android_device_sony_sumire" or "LineageOS/android_device_sony_sumire" Dec 04 15:25:34 @bhushanshah, wut wut wut !!! Dec 04 15:25:58 @sopernam, https://github.com/LineageOS/android_device_sony_sumire Dec 04 15:26:22 if you are using los remote, just use : android_device_sony_sumire Dec 04 15:26:25 @bhushanshah is that a toolchain / build system issue or ? Dec 04 15:26:43 @akshat2K2, @trivial_inanity see this is the error Dec 04 15:27:09 Unable to locate package libncurses5-dev:i386 … Unable to locate package libx11-dev:i386 … Unable to locate package libreadline6-dev:i386 … Unable to locate package libgl1-mesa-glx:i386 … Unable to locate package zlib1g-dev:i386 Dec 04 15:27:12 @trivial_inanity I am not entirely sure... I need to investigate but at this moment I am too busy with work stuff Dec 04 15:27:32 @shenoy13, Iam getting this error in finding these packages, should I just skip them? Dec 04 15:27:40 In initial setup Dec 04 15:27:42 @shenoy13, didn't we include package list in wiki ? Dec 04 15:27:57 @trivial_inanity, From wiki I copy pasted Dec 04 15:28:05 @bhushanshah, as far as I can tell "halium/android_device_sony_sumire" is true Dec 04 15:28:19 :/ Dec 04 15:28:26 you didn't read my message Dec 04 15:28:28 at all Dec 04 15:28:31 :( Dec 04 15:28:46 @akshat2K2 have you set your toolchain someplace in device tree ? not sure why it's using aarch64 toolchain in the first palce. Dec 04 15:28:55 check your boardconfig.mk files Dec 04 15:29:16 lol @bhushanshah nice take on this Dec 04 15:29:37 @shenoy13 what's your linux distro? Dec 04 15:29:38 @shenoy13 Which ubuntu/debian distribution you are using? Dec 04 15:29:49 Ubuntu 16lts Dec 04 15:29:56 @bhushanshah is thah true … Dec 04 15:30:26 @so Dec 04 15:30:31 remove revision Dec 04 15:32:31 @bhushanshah would be easier to introduce @sopernam to XML :P Dec 04 15:34:56 @bhushanshah now? … Dec 04 15:35:01 @bhushanshah, Ubuntu 16.04 Dec 04 15:35:47 umm works for me Dec 04 15:40:49 @shenoy13, Looks like you skipped the sudo dpkg --add-architecture i386 step. However this packages aren't needed for most devices anyway. Dec 04 15:43:30 (Photo, 530x269) https://irc.ubports.com/ifT4eZ2w/file_2569.jpg Dec 04 15:43:51 @sopernam, why is not sync device tree Dec 04 15:43:57 Looks fine Dec 04 15:44:20 @JBBgameich, ty Dec 04 15:44:47 It will sync the device tree too, it already syncs the kernel Dec 04 15:45:03 @JBBgameich, Ok Dec 04 15:45:33 @JBBgameich, I haven't skipped dpkg --adds architecture i386 Dec 04 15:45:42 thank you for your help Dec 04 15:45:58 @shenoy13, Have you done apt update afterwards? Dec 04 15:46:26 @JBBgameich, Oops no Dec 04 15:47:22 (Photo, 1280x419) https://irc.ubports.com/5ZKANMCO/file_2571.jpg Dec 04 15:48:07 Install plasma mobile Dec 04 15:48:07 Ubuntu sux Dec 04 15:48:23 @sopernam, You are missing the repositories from the lineage.dependencies file Dec 04 15:48:36 kitakami-common Dec 04 15:48:38 @JBBgameich, how i can add ? Dec 04 15:49:05 @sopernam, Search in lineage git Dec 04 15:49:07 Lel Dec 04 15:49:16 [ … { … "repository": "android_device_sony_kitakami-common", … "target_path": "device/sony/kitakami-common" … } … ] Dec 04 15:49:27 Open the manifest again and add the additional repositories like before. Dec 04 15:50:28 @Rk585, Can we not try to discourage someone who is trying to learn? This seems to be common thing now a days in the group.. trolling people who are new to something Dec 04 15:50:36 And I am not happy about it at all Dec 04 15:51:25 @JBBgameich, how 😭😭😭 Dec 04 15:52:01 @bhushanshah, 👍🏼 Dec 04 15:52:09 (Photo, 1280x494) https://irc.ubports.com/kEjvVC10/file_2573.jpg before some developer help me Dec 04 15:52:27 It is literally mentioned in the documentation on how to transform entries in the lineageos.dependencies to local_manifest, please read documentation Dec 04 15:52:51 @sopernam, First time building ? Dec 04 15:52:53 Nice Dec 04 15:53:16 english is so bad, i do not understand Dec 04 15:53:35 @Rk585, no i didnt make this Dec 04 15:53:38 @sopernam, Dec 04 15:54:22 Something like that Dec 04 15:57:54 fatal: error parsing manifest /home/ghost/halium/.repo/local_manifests/sony_sumire.xml: mismatched tag: line 7, column 2 Dec 04 15:58:04 Dec 04 15:58:47 Add a / before the > after los Dec 04 15:59:37 oh 😅 Dec 04 16:00:29 thank youu Dec 04 16:10:06 (Photo, 1280x693) https://irc.ubports.com/PIC0YZ0G/file_2575.jpg why is stuck in there Dec 04 16:10:17 ? Dec 04 16:10:20 yes Dec 04 16:11:45 (Photo, 1280x693) https://irc.ubports.com/THTEx0IM/file_2575.jpg why is stuck in there Dec 04 16:15:20 Something is still downloading, probably Dec 04 16:16:02 @UniversalSuperBox, this is toooooooo big Dec 04 16:16:13 20GB, as the documentation says Dec 04 16:16:13 or more Dec 04 16:16:31 😳 Dec 04 16:16:57 only qcom_display Dec 04 16:17:14 No, it's probably still working on one that came before Dec 04 16:18:14 oh okay im waiting... Dec 04 16:29:31 (Photo, 1280x145) https://irc.ubports.com/sVIDfkZ3/file_2577.jpg Dec 04 16:29:54 i am still waiting Dec 04 16:32:26 apologies ... your bytes will soon be deliverd Dec 04 16:33:39 Halium is even very small compared to other Android based roms Dec 04 17:30:38 what's this https://paste.pound-python.org/show/5YHeEPKvEHs5lyMvpkAL/ ' Dec 04 17:30:40 ? Dec 04 17:30:44 5.1 Dec 04 17:40:54 @sopernam, still waitinggggg....... Dec 04 18:12:34 hi should i add kernel resources to manifest ? Dec 04 18:19:31 no Dec 04 18:33:53 Why no? The kernel repository should also be added to the manifest. But I thought you already added it ... Dec 04 18:35:17 @JBBgameich, kernel isnt mentioned halium guide Dec 04 18:36:10 @sopernam, You are already downloading it in this picture Dec 04 18:36:30 android_kernel_sony_msm8994 Dec 04 18:38:29 I completely forgot what halium is for Dec 04 18:38:31 😂 Dec 04 18:39:12 Runing Gnu / Linux on android devices by using Android's drivers (as long as we don't have native linux drivers) Dec 04 18:40:30 Oh I see Dec 04 18:40:34 Ok, I'll work on it Dec 04 18:40:37 On day Dec 04 18:40:42 Sony Xperia XZ Dec 04 18:56:59 (Photo, 1200x330) https://irc.ubports.com/Dq64N0be/file_2579.jpg Dec 04 18:57:46 ghost@instance-1:~/halium$ ls .repo/local_manifest/ … sony_sumire.xml Dec 04 18:58:00 should i rename lineage_sumire ? Dec 04 18:58:50 @LLJYY, i hate sony politics Dec 04 18:58:51 no, the name there is not important really Dec 04 18:59:20 is there something in device/sony/sumire? Dec 04 18:59:50 @sopernam, it says permission denied .repo/local_manifest/ Dec 04 19:00:29 @JBBgameich, only google Dec 04 19:00:47 Then something is wrong with repo / your manifest .. Dec 04 19:00:58 name="android_device_kitakami-common" remote="los" /> … Dec 04 19:02:10 phhusson was added by: phhusson Dec 04 19:03:03 `ls -la .repo/local_manifests` Dec 04 19:05:44 ghost@instance-1:~/halium/.repo$ ls … local_manifest manifest.xml manifests manifests.git project-objects project.list projects repo … bad mistakes Dec 04 19:05:48 :( :D Dec 04 19:06:15 i have fixed it Dec 04 19:06:25 👍 Dec 04 19:23:01 (Photo, 1062x506) https://irc.ubports.com/S0sec6KO/file_2583.jpg Dec 04 19:23:02 stuck in %98 Dec 04 19:23:03 my friend says add —depth=1 git clone Dec 04 19:23:04 where is git clone ? Dec 04 19:35:39 (Photo, 1280x630) https://irc.ubports.com/KOFv4Bf6/file_2585.jpg Dec 04 19:35:40 Fwd from developerfect: Guys help Dec 04 19:35:41 Fwd from developerfect: Stucked %99 Dec 04 19:36:13 please help Dec 04 19:36:56 Fwd from vanyasem: Ctrl+C and start over? Dec 04 19:36:57 Fwd from vanyasem: with `--depth=1`? Dec 04 19:37:20 @vanyasem, repo sync -c —depth=1 Dec 04 19:37:22 ? Dec 04 19:37:29 Never trust repo's percentage Dec 04 19:37:31 or at least check with iotop/htop if it's really stuck Dec 04 19:37:37 or just doing smth in the background Dec 04 19:37:42 @JBBgameich, because of this Dec 04 19:37:44 It will most likely continue again Dec 04 19:38:07 Usage: repo sync [...] … main.py: error: no such option: —depth Dec 04 19:39:27 that's an argument for repo init Dec 04 19:39:30 not for repo sync Dec 04 19:40:01 with sync you can use `-j` argument to specify more parallel threads if you have a beefy interner Dec 04 19:40:38 @vanyasem, Cloud Computing. has fast int. speed Dec 04 19:40:56 then -j10 will be appropriate :P Dec 04 19:41:55 ghost@instance-1:~/halium$ sudo repo init -u https://github.com/Halium/android -b halium-7.1 —depth=1 … remote: Total 0 (delta 0), reused 0 (delta 0), pack-reused 0 … Your identity is: orkun <> … If you want to change this, please re-run 'repo init' with —config-name … repo has been initialized in /home/ghost/halium Dec 04 19:42:06 but same Dec 04 19:42:13 it worked :) Dec 04 19:42:17 resync the tree Dec 04 19:42:24 and be `patient` Dec 04 19:42:29 go for a walk maybew Dec 04 19:42:37 same stuck in %99 Dec 04 19:42:48 is it stuck? Dec 04 19:42:54 or is it doing something in the BG? Dec 04 19:42:59 check with iotop/htop Dec 04 19:43:54 (Photo, 1280x693) https://irc.ubports.com/NWrQIs74/file_2587.jpg Dec 04 19:44:05 0.0 cpu Dec 04 19:44:18 and IO? Dec 04 19:44:21 It shouldn't use cpu much anyway, it downloads stuff Dec 04 19:44:30 why are you running repo as root? Dec 04 19:45:19 @JBBgameich, great question Dec 04 19:45:42 i was tried root and not root but same results Dec 04 19:47:07 delete everything and start from scratch. Dec 04 19:47:19 and then `go for a walk`. Dec 04 19:47:34 * vanyasem leaves Dec 04 19:47:36 @vanyasem, i have tried Dec 04 19:48:00 5 times Dec 04 20:24:15 hi Dec 04 20:24:50 @sopernam, Try deleting only the repo you're stuck with Dec 04 20:25:35 I reset the server, the result is still the same Dec 04 20:30:17 DuceMiaLuce was added by: DuceMiaLuce Dec 04 21:46:46 I'm trying to build for nexus 4 (mako) but i seem to be stuck is there anyone who can point me in the right direction ^ Dec 04 21:47:05 external/e2fsprogs/e2fsck/problem.c:856:30: warning: missing field 'second_code' initializer [-Wmissing-field-initializers] Dec 04 21:47:08 PROMPT_CLEAR, PR_PREEN_OK }, Dec 04 21:47:11 ^ Dec 04 21:47:13 external/e2fsprogs/e2fsck/problem.c:861:30: warning: missing field 'second_code' initializer [-Wmissing-field-initializers] Dec 04 21:47:16 PROMPT_CLEAR, PR_PREEN_OK }, Dec 04 21:47:19 ^ Dec 04 21:47:21 external/e2fsprogs/e2fsck/problem.c:866:30: warning: missing field 'second_code' initializer [-Wmissing-field-initializers] Dec 04 21:47:24 PROMPT_CLEAR, PR_PREEN_OK }, Dec 04 21:47:27 ^ Dec 04 21:47:29 external/e2fsprogs/e2fsck/problem.c:871:18: warning: missing field 'second_code' initializer [-Wmissing-field-initializers] Dec 04 22:18:21 First, please use pastebin or a similar service to post your errors revertive Dec 04 22:18:48 Next, I'm not able to see an error in the posted output. Was there more? Dec 04 23:03:52 @UniversalSuperBox, pastebin forbidden my country Dec 04 23:04:43 (Photo, 1280x693) https://irc.ubports.com/bekHEN0u/file_2589.jpg final stage Dec 04 23:08:23 sopernam: This is what I needed for LuneOS, you might need some as well: https://github.com/webOS-ports/android/commit/e0fd3668d1b5ca37b5ddc9ce7caf70d2d61ed364 Dec 04 23:08:36 From my default.xml **** ENDING LOGGING AT Tue Dec 05 03:00:01 2017 **** BEGIN LOGGING AT Tue Dec 05 03:00:01 2017 Dec 05 05:18:40 (Photo, 1123x1280) https://irc.ubports.com/Os1xE6Qq/file_2591.jpg Dec 05 05:19:59 nice Dec 05 05:31:25 @sopernam, Build completed? Dec 05 05:35:45 no Dec 05 05:36:03 @sopernam, i had error Dec 05 05:36:07 did your repo sync properly? Dec 05 05:36:43 yes Dec 05 05:37:17 when you run source build/envsetup.sh, how many lines output you got? Dec 05 05:37:53 (Photo, 1280x288) https://irc.ubports.com/5Wn4IWV4/file_2594.jpg Dec 05 05:38:15 iam getting only 4 lines Dec 05 05:38:27 including device/ line is not present Dec 05 05:38:38 any idea about the error? Dec 05 05:38:51 i have no idea Dec 05 05:38:57 ok Dec 05 05:39:16 @shenoy13, it is not necessary to have the device/ line present Dec 05 05:41:56 @shenoy13: does lunch command works for you? and do you have any specific error? Dec 05 05:44:39 no i did not try lunch yet Dec 05 05:44:57 (Photo, 690x677) https://irc.ubports.com/Olf9V6s5/file_2596.jpg Dec 05 05:45:22 here at the arrow mark i used source build/envsetup.sh Dec 05 05:46:13 @shenoy13, im using breakfast Dec 05 05:46:25 yes breakfast Dec 05 05:47:09 breakfast and what code name? Dec 05 05:47:34 sumire Dec 05 05:47:41 not appear Dec 05 05:47:55 is the maniffest wrong ? Dec 05 05:48:37 breakfast [codename] Dec 05 05:49:19 @shenoy13, same result Dec 05 05:49:26 (Photo, 720x1280) https://irc.ubports.com/bB0rorVw/file_2598.jpg Dec 05 05:50:09 Now use mka hybris-boot systemimage Dec 05 05:50:34 @JBBgameich, there is no out file Dec 05 05:50:43 we have got error Dec 05 05:51:30 (Photo, 676x768) https://irc.ubports.com/9LhVNqmN/file_2600.jpg Iam getting this error Dec 05 05:52:23 you need to add vendor_leeco to manifest Dec 05 05:52:33 @sopernam, what error you are getting Dec 05 05:53:32 @bhushanshah, Where in manifest? Dec 05 05:53:44 I mean I just followed halium guide Dec 05 05:53:45 local_manifest Dec 05 05:53:46 @shenoy13, yes you must have vendor Dec 05 05:53:59 @bhushanshah, wait Dec 05 05:54:13 (Photo, 1280x693) https://irc.ubports.com/7AMlRDaF/file_2602.jpg Dec 05 05:54:54 ?xml version="1.0" encoding="UTF-8"?> Dec 05 05:55:03 this is my manifest Dec 05 05:55:48 Add … Dec 05 05:55:56 @sopernam Dec 05 05:56:02 @bhushanshah, thank yu Dec 05 05:56:33 @shenoy13 you need to add … Dec 05 05:56:36 @sopernam, Didn't even start ? Dec 05 05:56:45 @Rk585, no Dec 05 05:56:51 @bhushanshah, Tf LeEco ? Dec 05 05:56:58 @sopernam, Bish wtaf Dec 05 05:57:00 @bhushanshah, ok, thanks Dec 05 05:57:13 I was ril related errors Dec 05 05:57:18 @Rk585, What? Dec 05 05:57:22 @shenoy13, Which phone Dec 05 05:57:33 @Rk585, Leeco x2 Dec 05 05:57:34 I was building for Le 2 Dec 05 05:57:40 @shenoy13, Le max 2 ? Dec 05 05:57:46 @Rk585, did you complete? Dec 05 05:57:52 @Rk585, yes Dec 05 05:58:00 @shenoy13, Ril related shits came up Dec 05 05:58:22 Just git clone things after syncing Dec 05 05:58:27 That will do Dec 05 05:58:37 @Rk585, Why? Dec 05 05:58:45 Aaah Dec 05 06:00:36 @bhushanshah, Do I need to resync after adding this? Dec 05 06:00:55 @shenoy13 just sync the vendor/leeco Dec 05 06:01:04 `repo sync vendor/leeco` Dec 05 06:01:13 ok Dec 05 06:08:14 (Photo, 621x81) https://irc.ubports.com/GrjO7jcs/file_2604.jpg Dec 05 06:08:26 Now iam getting this does not exist Dec 05 06:08:34 well Dec 05 06:08:37 that is not error Dec 05 06:08:46 ok Dec 05 06:08:47 you can just run breakfast and it will work Dec 05 06:11:21 (Photo, 668x380) https://irc.ubports.com/qaApKdoe/file_2606.jpg Dec 05 06:12:37 Iam getting similar error as earlier for "breakfast" Dec 05 06:12:41 no well Dec 05 06:12:46 breakfast Dec 05 06:12:57 ok Dec 05 06:18:41 (Photo, 692x402) https://irc.ubports.com/niEf7p8z/file_2608.jpg Dec 05 06:19:40 Fwd from bhushanshah: Add … Dec 05 06:20:03 ok Dec 05 06:36:17 (Photo, 720x1280) https://irc.ubports.com/vxoXzfRs/file_2610.jpg Dec 05 06:37:15 😭 Dec 05 06:38:04 bhushanshah: Might be worth to add some of these repos to the default.xml? Dec 05 06:38:31 bhushanshan: I needed these for mido on 7.1 to build: https://github.com/webOS-ports/android/commit/e0fd3668d1b5ca37b5ddc9ce7caf70d2d61ed364 Dec 05 06:38:36 I can PR if needed? Dec 05 06:40:00 Herrie: I guess makes sense.. (only issue is it will break already existing manifests, given entry will be declared twice) Dec 05 06:40:04 @bhushanshah Dec 05 06:40:32 @sopernam ? Dec 05 06:40:36 @sopernam, . Dec 05 06:40:44 i have got this error Dec 05 06:40:51 well Dec 05 06:41:01 you now need to move on from this command Dec 05 06:41:04 this is not error Dec 05 06:41:10 run mka hybris-boot Dec 05 06:42:06 should i not check for kernel config and mount up points before "mka hybris"? Dec 05 06:42:22 @bhushanshah, thank you❤️ Dec 05 06:43:06 @shenoy13 well, yes, you can do that and then mka hybris-boot or well, at least verify that build environment is setup properly and then fix up mountpoints Dec 05 06:43:15 end result will be same Dec 05 06:43:20 ok Dec 05 06:43:30 (Photo, 720x1280) https://irc.ubports.com/BURuCB0y/file_2612.jpg Dec 05 06:45:09 you need to install imagemagick Dec 05 06:47:03 Starting build with ninja ninja: Entering directory `.' ninja: error: '/home/ghost/halium/out/target/product/sumire/kernel', needed by '/home/ghost/halium/out/target/product/sumire/obj/ROOT/hybris-boot_intermediates/hybris-boot.img', missing and no known rule to make it build/core/ninja.mk:151: recipe for target 'ninja_wrapper' failed … make: *** [ninja_wrapper] Error 1 make: Leaving d Dec 05 06:47:04 '/home/ghost/halium' #### make failed to build some targets (13 seconds) #### Dec 05 06:47:25 wut Dec 05 06:48:45 @sopernam, Iam getting similar output Dec 05 06:50:15 (Photo, 1247x1280) https://irc.ubports.com/P0EgXJA1/file_2614.jpg Dec 05 06:51:00 Have you cloned kernel? Dec 05 06:51:06 @bhushanshah, yes Dec 05 06:51:20 @bhushanshah, how to install this? Dec 05 06:51:22 @bhushanshah, how can i check ? Dec 05 06:51:35 @shenoy13, sudo apt install imagemagick Dec 05 06:53:21 @sopernam, Yes now here. Dec 05 06:53:47 @shenoy13, are you got this error? Dec 05 06:54:23 (Photo, 672x140) https://irc.ubports.com/Xh0Oj9gl/file_2616.jpg Dec 05 06:54:24 @bhushanshah, Lol the first device I see with that requirement xd Dec 05 06:55:17 @sopernam, Not exactly. Dec 05 06:55:18 @sopernam kernel/sony/msm8994 dir exists? Dec 05 06:55:40 @shenoy13, remove doze from the Android.mk completely Dec 05 06:56:39 find: ‘device/*/generic’: No such file or directory … find: ‘device/unknown’: No such file or directory … find: ‘device/android’: No such file or directory Dec 05 06:57:06 @bhushanshah, How to do this? Dec 05 06:57:18 ls Dec 05 06:57:43 (Photo, 1280x1033) https://irc.ubports.com/aXHxWwe1/file_2618.jpg Dec 05 06:58:20 should i rename this folder msm8994 Dec 05 06:59:04 @shenoy13, https://github.com/LineageOS/android_device_leeco_msm8996-common/blob/cm-14.1/msm8996.mk#L150 comment this two line out Dec 05 06:59:25 @sopernam yes.. Dec 05 07:00:40 (Photo, 1269x1280) https://irc.ubports.com/C0Hk2ZXm/file_2620.jpg then Dec 05 07:01:27 did you run breakfast first? Dec 05 07:02:53 @bhushanshah, thank you Dec 05 07:03:04 everything will be good Dec 05 07:35:00 Iam not being able to find Android.mk file, any idea where to find it? Dec 05 07:36:47 Not android.mk but this file https://github.com/LineageOS/android_device_leeco_msm8996-common/blob/cm-14.1/msm8996.mk#L150 Dec 05 07:52:53 (Photo, 1033x105) https://irc.ubports.com/nYd5mKCN/file_2624.jpg Dec 05 07:53:01 I replaced earlier one. Dec 05 07:53:07 Now Iam getting this Dec 05 07:53:39 @shenoy13, I mean I commented out Dec 05 08:23:09 @sopernam, oh you are building a 7.1 image for sumire .. Dec 05 08:23:39 I do the same .. I could push a patched kernel and config when I get home.. Dec 05 08:47:19 hi im break build and how can i restart Dec 05 08:48:11 (Photo, 868x1280) https://irc.ubports.com/mF5vR2Dg/file_2626.jpg Dec 05 08:48:31 @bhushanshah Dec 05 09:18:48 (Photo, 1280x291) https://irc.ubports.com/qrQyzwRg/file_2628.jpg Dec 05 09:18:56 It seems to be stuck here for me Dec 05 09:19:20 Not moving forward Dec 05 09:20:22 @shenoy13, U got error Dec 05 09:20:37 @sopernam, Fixed mountpoints ? Dec 05 09:21:12 People don't want to rear the guide properly -_- Dec 05 09:21:28 @akshat2K2, I know I didnt do the fix mount points Dec 05 09:22:46 But this is crude build to verify whether build environment is setup properly as Bhushan Shah suggested earlier. Dec 05 09:23:04 Ohk Dec 05 09:23:36 So does fixing mount points makes it good now? Dec 05 09:24:51 @Rk585, how Dec 05 09:25:04 how to fix mountpoint? Dec 05 09:25:55 Add for ur device in the specific file mentioned in guide Dec 05 09:27:09 @sopernam, Guides r there Dec 05 09:27:12 In their git Dec 05 09:29:41 http://docs.halium.org/en/latest/porting/build-sources.html#include-your-device-in-fixup-mountpoints Dec 05 09:31:32 @bhushanshah is my build looking fine everything apart from mount points? Dec 05 09:31:43 Yes Dec 05 09:31:44 @shenoy13, this Dec 05 09:42:52 bshah: you Herrie's PR proposal: I propose we PR it nevertheless, and then we discuss within the PR. IMHO, it would make little sense to refuse any evolution of the main manifest; if someone syncs its manifests with upstream, then merges are to be handled there Dec 05 09:43:00 s/you/about/ Dec 05 09:44:01 Tofe: sure... I'm not against PR for record :-) Dec 05 09:44:11 bshah[m]: I know :) Dec 05 11:21:39 So u should just copy paste my fstab.qcom to fixup-mountpoints Dec 05 11:24:09 Dead asf Dec 05 12:17:35 (Photo, 1280x719) https://irc.ubports.com/tvOyuN8Z/file_2633.jpg Dec 05 12:19:00 How to configure kernel for halium? Dec 05 12:19:14 point it towards your actual defconfig, not just the folder Dec 05 12:19:23 check the BoardConfig.mk in your device tree to figure out the exact file Dec 05 12:21:00 eg TARGET_KERNEL_CONFIG := lineage_x2_defconfig Dec 05 12:23:40 (Photo, 553x125) https://irc.ubports.com/pG7fVmc8/file_2635.jpg Dec 05 12:23:58 I see these many boardconfig.mk files Dec 05 12:30:48 Should I run the command as "./mer_verify_kernel_config lineage_x2_defconfig" Dec 05 12:30:50 ? Dec 05 12:31:08 @shenoy13, the one for your device...so leeco/x2 Dec 05 12:31:44 @shenoy13, `mer_verify_kernel_config kernel/leeco/msm8996/arch/arm/configs/lineage_x2_defconfig` Dec 05 12:31:57 @Lyokanthrope, Thank you Dec 05 12:36:10 (Photo, 1280x446) https://irc.ubports.com/RslEwHXA/file_2637.jpg Dec 05 12:36:15 I have got the output Dec 05 12:36:25 @shenoy13, Are you on windows? Dec 05 12:36:46 Iam using cloud Dec 05 12:37:00 @shenoy13, How to configure required items? Dec 05 12:37:17 just edit your defconfig (it's just a text file) and set those parameters Dec 05 12:37:19 @technicalbird, seriously someone compiles in that mess of operating system? Dec 05 12:37:51 @ErnyTech, using cloud Dec 05 12:38:18 @Lyokanthrope, Ok, thank you. Should I edit optional files also? Dec 05 12:38:27 @ErnyTech, Unfortunately it runs Ubuntu, suse on top of it(user space). Dec 05 12:38:31 And both "arm" and "arm64" configs? Dec 05 12:38:37 Is your device arm64? Dec 05 12:38:45 @Lyokanthrope, yes Dec 05 12:38:50 then yes Dec 05 12:39:06 (Document) https://irc.ubports.com/gaVDiGKF/file_2638.mp4 Dec 05 12:39:07 @shenoy13, do the non-optional first, build and make sure it works, then try the optionals one-by-one Dec 05 12:39:23 @Lyokanthrope, ok Dec 05 12:54:36 When I find few for example "CONFIG_SECURITY_SELINUX_BOOTPARAM" in the kernel config file, it says "not found" Dec 05 12:54:50 so should I edit this manually? Dec 05 12:55:12 That's what I've usually done is just manually add them. Dec 05 12:55:25 ok Dec 05 12:55:50 when 2 values are present "y,m" which value to give? Dec 05 12:56:02 for what in particular? Dec 05 12:57:31 for example, Dec 05 12:57:34 WARNING: CONFIG_NETFILTER_XT_MATCH_NFACCT is invalid … It is unset … Allowed values : y, m, ! Dec 05 12:57:43 here 2 allowed values are present Dec 05 12:57:58 I usually just set that to y Dec 05 12:58:59 @shenoy13 which kernel version is it ? Dec 05 12:59:27 3.18 Dec 05 12:59:57 @Lyokanthrope, ok Dec 05 13:01:35 @Lyokanthrope, We can add it anywhere in the config right? Dec 05 13:01:49 I mean not necessary at the end of the file Dec 05 13:02:06 it shouldn't matter yeah, but it's best to keep 'new' entries somewhere where you know it is just in case Dec 05 13:03:15 @Lyokanthrope, ok. Dec 05 13:23:55 if something is to be left empty, should we just put = and leave? Dec 05 13:23:59 for example Dec 05 13:24:13 CONFIG_UEVENT_HELPER_PATH= Dec 05 13:24:28 @shenoy13, allowed value is empty Dec 05 13:25:26 Bhushan Shah: multirom pr? :p Dec 05 13:26:38 I mean https://github.com/DeadSquirrel01/Team-Win-Recovery-Project/commit/92fe3259f7281dbb499602721fd371566e13f9d9 eh Dec 05 13:30:22 @i_love_halium_on, why remove sailfish? Dec 05 13:31:01 Because it's useless and will use halium Dec 05 13:31:21 (Photo, 1280x103) https://irc.ubports.com/CzUdvkQ8/file_2640.jpg Dec 05 13:31:33 Should I have this "y" or "n" Dec 05 13:31:44 allowed values: n, ! Dec 05 13:31:50 set it to n or unset it Dec 05 13:31:54 because comment says it can be left allowed as "n" also Dec 05 13:32:15 @Lyokanthrope, ok Dec 05 13:32:25 depends on kernel Dec 05 13:32:39 but try that first Dec 05 13:32:49 But if u like it you can readd sfos lines and change 2 variables Dec 05 13:58:04 (Photo, 1280x719) https://irc.ubports.com/19ShOF9C/file_2644.jpg Dec 05 13:58:33 (Photo, 1280x75) https://irc.ubports.com/jhcV5wc4/file_2646.jpg Dec 05 13:58:47 I have made mer kernel and mount points properly Dec 05 13:59:06 is config_autofs4_fs set? Dec 05 13:59:34 @Lyokanthrope, merkernel shows no errors or warnings Dec 05 13:59:39 on arm64 Dec 05 14:00:01 and there is no lineage_x2_defconfig in arm kernel Dec 05 14:00:56 Double check that is set though, that script sometimes acts up Dec 05 14:01:10 ok Dec 05 14:01:42 No errors Dec 05 14:02:01 It immediately takes to next line after mer kernel check Dec 05 14:02:06 Which means no error Dec 05 14:03:18 @Lyokanthrope, This is not found in the kernel Dec 05 14:04:11 Consider running this check on the resulting config, not the defconfig Dec 05 14:07:03 @ilyaishere, Resulting kernel means which one? Dec 05 14:07:12 @shenoy13, I Kean config Dec 05 14:07:37 @shenoy13, I mean resulting config means which one? Dec 05 14:07:47 I don't remember the exact location Dec 05 14:08:08 Somewhere in your out folder Dec 05 14:08:20 Related to KERNEL_OBJ Dec 05 14:08:38 @ilyaishere, OK I will check Dec 05 14:08:39 May be compressed to config.gz or similar Dec 05 14:13:40 @shenoy13, I mean config Dec 05 14:13:41 @ilyaishere, Which resulting config? Dec 05 14:14:27 Fwd from ilyaishere: I don't remember the exact location Dec 05 14:14:28 Fwd from ilyaishere: Somewhere in your out folder Dec 05 14:14:29 Fwd from ilyaishere: Related to KERNEL_OBJ Dec 05 14:14:30 Fwd from ilyaishere: May be compressed to config.gz or similar Dec 05 14:43:47 for 5.1 reference rootfs do i need egl patch for hwcomposer or not? Dec 05 14:49:42 nope Dec 05 14:50:09 ( in case you're talking about the patch that adds "CAF" compatbility ) Dec 05 14:53:02 yay Dec 05 15:00:11 (Photo, 1280x407) https://irc.ubports.com/jWHFdG0I/file_2649.jpg Dec 05 15:00:26 (Photo, 1176x236) https://irc.ubports.com/HuXzgcSv/file_2651.jpg Dec 05 15:00:35 Anyone got any idea about these errors? Dec 05 15:01:04 can you give a full log on pastebin or something? Dec 05 15:01:32 ok Dec 05 15:27:51 https://pastebin.com/tgLpe2EW Dec 05 15:28:01 @shenoy13, Here paste bin log Dec 05 15:31:03 try setting CONFIG_EXT4_ENCRYPTION to n Dec 05 15:31:20 @Lyokanthrope, Ok Dec 05 15:48:55 Still error persists. Dec 05 15:49:08 Maybe I will do a clean build some time Dec 05 15:51:11 But I have one doubt Dec 05 15:51:12 https://pastebin.com/Gpq725wR Dec 05 15:51:21 in this fstab, Dec 05 15:52:12 there are 2 instances of cache and one more thing. Dec 05 15:52:34 and yeild same result in "readlink" Dec 05 15:52:59 Should I have 2 seperate instance in the fixup-mountpoints? Dec 05 16:05:50 No Dec 05 16:06:26 That's just so the os can mount those as ext4 or f2fs Dec 05 16:09:58 @Lyokanthrope, Ok, I currently have only one instance for both instance in fstab. Dec 05 16:11:28 I think there is problem with initial manifest Dec 05 16:11:46 coz I can see the defconfig of another device from the same manufacturer Dec 05 16:12:12 in the kernel/leeco/msm8996/arch/arm64/ folder Dec 05 16:13:53 (Document) https://irc.ubports.com/JhESfBLY/file_2661.PNG Dec 05 16:13:53 like here Dec 05 16:14:22 lineage_x2 and lineage_z11 Dec 05 16:14:25 @shenoy13, no, that's fine Dec 05 16:14:27 x2 is what should be relevant for your device Dec 05 16:14:45 @shenoy13, Both devices share the same kernel Dec 05 16:14:52 @JBBgameich, ok Dec 05 16:17:43 https://pastebin.com/aRsHcTd4 Dec 05 16:18:00 does this manifest look fine? Dec 05 16:18:55 you can remove line 4 probably Dec 05 16:21:13 @JBBgameich, But vendor blobs from muppets is required right? Dec 05 16:22:16 they are, but you added them twice. Once in line 4, and once in line 6. The difference is just that line 6 has the correct location and line 4 not Dec 05 16:23:26 ok thanks Dec 05 17:02:11 im getting these: https://paste.pound-python.org/show/Pd8JB6pYagP0QUoFx73o/ in logcat in 5.1 and these https://paste.pound-python.org/show/RGrVce4cukFSYilbJ4lJ/ in dmesg. i tried both reference rootfs and pm normal rootfs for my caf device. Anyone knows why it fails? Dec 05 17:26:10 anyone has an idea? do i need something for 5.1 caf? or something in my kernel or tree? Dec 05 17:53:07 Should jack work with audio patch? Im tired of 5.1 port :/ Dec 05 18:09:06 hi Dec 05 18:09:12 (Document) https://irc.ubports.com/ZMFXuE63/file_2664 Dec 05 18:09:32 @sopernam, which blocks should i add Dec 05 18:12:39 Blocks corresponding to these only Dec 05 18:14:31 @shenoy13, r u sure ? Dec 05 18:15:01 @sopernam, Yes in the guide it's clear about this Dec 05 18:15:41 For just booting the device boot usedata and cache should be enough. But better add a few too many than missing some you need later Dec 05 18:16:22 there is another sony devices yuga Dec 05 18:16:40 do i do it same ? Dec 05 18:16:45 Yeah, that's the Xperia Z Dec 05 18:17:02 But it's partitions are most likely not the same Dec 05 18:17:35 It has a different SoC so they are probably different Dec 05 18:17:51 @JBBgameich, yeah i know Dec 05 18:20:11 I asked it whether i should to similar to it Dec 05 18:20:47 I will write the mount points for my phone Dec 05 18:21:49 @sopernam, Not necessary Dec 05 18:22:10 For example, my device and one plus 3 has same chipset Dec 05 18:22:18 But still lot of difference Dec 05 18:22:37 But yeah, if you need to know how the format of the mount points script works, look at yuga. Dec 05 18:24:08 is halium compatible fingerprint Dec 05 18:24:40 Fingerprint sensor? Dec 05 18:28:24 yes Dec 05 18:28:47 I'm not sure if libhybris already has a module for that Dec 05 18:29:06 It could be done, but would maybe require additional work Dec 05 18:29:39 And it depends on the target rootfs (OS) if it supports fingerprint unlocking at all. Dec 05 19:32:02 Guys, one question... Halium 5.1 builded with lineage os 12.1 and kernel 3.4 can work? Dec 05 19:36:51 @JBBgameich, The BQ E5 Ubuntu Edition (or maybe the pro 5, I can't keep track) has a working fingerprint sensor. I don't know how. Dec 05 19:39:41 I wonder if it's as simple as using libfprint Dec 05 19:39:53 Nothing is simple Dec 05 19:45:54 @Mattia, Can someone answer to my question? Dec 05 19:45:55 yes Dec 05 19:46:35 Becuase a friend said this: … that i am not sure … lxc is pretty broken on 3.4 kernels … that's why i wanted to move to 3.10 kernle … that would bring us both anbox and halium based roms which use lxc Dec 05 19:47:30 That could be the case. 3.4 will need a couple patches (which I don't have) to make the Android container work, but otherwise should be okay Dec 05 19:47:35 Newer is always better, of course Dec 05 19:48:03 So if i try probably doesn't boot. Correct? Dec 05 19:48:17 Can't say for sure Dec 05 19:48:27 Ok thanks for the information! Dec 05 19:55:01 @shenoy13, what is diag ? Dec 05 19:55:20 and what is DDR Dec 05 19:56:43 3.4 kernel should work with LXC we use it on LuneOS too Dec 05 19:56:53 YOu need some patches to kernel but nothing too shocking Dec 05 19:57:31 See commits from 13th of October on https://github.com/Halium/android_kernel_lge_hammerhead/commits/halium-5.1 for example Dec 05 19:57:40 That works for us in LuneOS with recent LXC Dec 05 19:58:47 With recent I mean 2.0.8 Dec 05 20:08:08 hey Dec 05 20:08:17 hey @all Dec 05 20:08:58 `make mkbootimg` Dec 05 20:09:40 As mentioned http://docs.halium.org/en/latest/porting/build-sources.html#building-the-system-img-and-hybris-boot-img Dec 05 20:09:46 ninja: error: unknown target 'bootimg', did you mean 'mkbootimg'? Dec 05 20:10:00 sorry Dec 05 20:10:01 :D Dec 05 20:11:12 (Photo, 1280x693) https://irc.ubports.com/nW4pyUCc/file_2666.jpg Dec 05 20:11:15 finally :D Dec 05 20:11:22 @Mattia, It will boot, but the container won't start. SSH debugging should work though. A few people already got LXC working on 3.4 kernels. At least I know it started on the Sony Xperia Z Dec 05 20:16:28 hey what is kernel module not enabled Dec 05 20:17:01 In your kernel config, CONFIG_MODULES is set to n most likely Dec 05 20:17:27 Have you already adapted the kernel config? mer-verify-kernel-config should tell you what you need to change Dec 05 20:18:21 @sopernam, shoıld kernel module is enabled ? Dec 05 20:18:46 I'm not sure, I think it was enabled per default on my device Dec 05 20:19:08 Is it necessary to do it? Dec 05 20:19:52 (Photo, 1280x693) https://irc.ubports.com/LjEHJ4KI/file_2668.jpg is it normal ? Dec 05 20:20:13 yeah Dec 05 20:20:18 Yeah Dec 05 20:20:28 I'm pretty sure it is. As long as there are only warnings and no errors, everything is fine. Dec 05 20:21:00 😅 Dec 05 21:14:54 Thanks @JBBgameich and Herrie, i will try Dec 05 21:20:27 (Photo, 1280x693) https://irc.ubports.com/VRSBX5kf/file_2672.jpg Dec 05 21:20:30 finally Dec 05 21:20:55 then what i do? Dec 05 21:21:11 Keep reading the docs Dec 05 21:21:16 On to the next page! Dec 05 21:21:25 Flash it, and fix the bugs that appear basically Dec 05 21:22:30 But follow the docs, not just flash something somewhere randomly Dec 05 21:26:04 how can i disable rild Dec 05 22:18:31 i'm trying to port halium to my Nexus 4, i have successfully create/flashed hybris-boot.img and the reference rootfs, but i'm stuck for some reason it can't find a valid RAM Disk Image and root device at startup (pastebin.com/iSyiBEw4) can someone point me in the right direction to fix this? Dec 05 22:25:41 i have tryed supplying root=mmcblk0p21 (/System) and root=mmcblk0p23 (/Data) but it still can't find an root device Dec 05 22:48:39 hi Dec 05 22:48:56 (Photo, 653x155) https://irc.ubports.com/Gzf7I45g/file_2674.jpg which file i need Dec 05 22:49:14 which file pull ? Dec 05 23:00:07 hi Dec 05 23:00:13 how to make rootfs ? Dec 05 23:00:37 You can just download it Dec 05 23:00:45 okay Dec 05 23:02:25 where do i copy this file _ Dec 05 23:03:01 @bhushanshah, We need use "halium-install" script in TWRP or Fastboot? Dec 05 23:04:29 @bhushanshah, how can i use this file basically Dec 05 23:05:30 Look in the install documentatiom Dec 05 23:06:06 @UniversalSuperBox, there is no guide Dec 05 23:06:12 There sure is Dec 05 23:06:24 https://docs.halium.org/en/latest/porting/install-build/index.html Dec 05 23:07:17 @UniversalSuperBox, I can not find the exact answer here. … We need use "halium-install" script in TWRP or Fastboot? Dec 05 23:08:11 On the reference rootfs page it says "You can use the halium-install script as below, when device is connected in recovery mode" Dec 05 23:08:18 Recovery is TWRP Dec 05 23:08:29 what is rootfs tarball ? Dec 05 23:09:17 It's a .tar.gz file Dec 05 23:10:07 @UniversalSuperBox, Thank you. That was good. Dec 05 23:10:43 We can flash hybris boot with twrp? Dec 05 23:12:19 No, use fastboot. Dec 05 23:25:32 [rider .android]# adb devices … List of devices attached … CB5A2BFDEC unauthorized Dec 05 23:25:40 how can i fix this error Dec 05 23:35:51 ramdisk/sbin/adbd takr from twrp Dec 05 23:35:56 and open default.prop Dec 05 23:36:07 disable all secures Dec 05 23:36:19 make adb debuggable 1 Dec 05 23:36:29 makr usb config adb Dec 05 23:45:02 android-tools-fsutils no packages for arch linux Dec 05 23:46:55 find ppa in internet Dec 05 23:46:59 add ppa to sources Dec 05 23:47:20 and make update for packages Dec 05 23:47:23 install again Dec 05 23:48:11 @MacTavishOT, im using ARCH Dec 06 00:10:42 (Photo, 1280x411) https://irc.ubports.com/Bet1Qx0S/file_2676.jpg Dec 06 00:10:52 (Document) https://irc.ubports.com/FfmggaR3/file_2677 Dec 06 00:10:58 i've got this error **** ENDING LOGGING AT Wed Dec 06 02:59:59 2017 **** BEGIN LOGGING AT Wed Dec 06 03:00:00 2017 Dec 06 03:26:46 @sopernam, Size is 160M .. Dec 06 03:40:39 change this line in install script to bigger size.. Dec 06 03:40:40 Fwd from MacTavishOT: make_ext4fs -l 160M Dec 06 03:41:43 I don't know ideal size for rootfs exactly Dec 06 05:43:13 @Satyam2345, I used 210mb Dec 06 06:18:11 That's system image, not rootfs, and it can vary in size Dec 06 08:36:02 Fwd from sopernam: hey Dec 06 08:36:03 Fwd from sopernam: WARNING: CONFIG_NETFILTER_NETLINK_ACCT is invalid … It is unset … Allowed values : y, m, ! Dec 06 08:36:04 Fwd from sopernam: how can add it ? Dec 06 09:14:38 Find it in your defconfig Dec 06 10:32:11 (Photo, 816x508) https://irc.ubports.com/RtF2MR2z/file_2688.jpg what is wrong ? Dec 06 10:36:45 @bhushanshah Dec 06 10:37:35 Can you show me your local_manifest? Dec 06 10:37:47 I think your kernel is at wrong location Dec 06 10:38:16 (Document) https://irc.ubports.com/we4yTkMm/file_2689.xml Dec 06 10:39:12 yeah Dec 06 10:39:23 `` Dec 06 10:40:05 can you rename `kernel/sony/sumire` directory to `kernel/sony/msm8994` ? Dec 06 10:40:15 oh Dec 06 10:40:19 i remember Dec 06 10:40:22 :D Dec 06 10:40:26 ty Dec 06 10:40:28 and also make same changes in local_manifest.xml Dec 06 10:41:36 Please don't forget @sopernam Dec 06 10:42:52 @bshah[m], i ö fix it Dec 06 10:57:55 how can i enable kernel module Dec 06 10:57:56 ? Dec 06 10:59:54 (Photo, 1280x693) https://irc.ubports.com/90GvSjVr/file_2691.jpg @bhushanshah Dec 06 11:00:42 CONFIG_MODULES Dec 06 11:01:28 @bshah[m], How activate? Dec 06 11:01:50 make that option =y in kernel config well Dec 06 11:02:57 (Sticker, 512x512) https://irc.ubports.com/CLw9yQJk/file_2692 Dec 06 11:03:03 👌 Dec 06 11:04:16 parduscix was added by: parduscix Dec 06 11:08:23 Any idea about this error https://pastebin.com/PjBGvSbz Dec 06 11:09:10 @shenoy13, pastebin has forbidden in my country sorry :( Dec 06 11:10:21 ASN.1 crypto/asymmetric_keys/x509-asn1.c … Extracted 203 tokens … Extracted 14 types … Extracted 12 actions … Pass 1 … Pass 2 … ASN.1 crypto/asymmetric_keys/x509_rsakey-asn1.c … Extracted 16 tokens … Extracted 1 types … Extracted 1 actions … Pass 1 … Pass 2 … CC crypto/asymmetric_keys/x509_public_key.o … CC fs/ext4/hash.o … ASN.1 crypto/asymmetric_keys/pkcs7-asn1.c … Extracted 371 tokens … Ext Dec 06 11:10:22 Pass 1 … Pass 2 … CC crypto/asymmetric_keys/pkcs7_trust.o … CC fs/ext4/resize.o … CC crypto/asymmetric_keys/pkcs7_verify.o … CC crypto/asymmetric_keys/pkcs7_key_type.o … LD crypto/asymmetric_keys/asymmetric_keys.o … CC crypto/asymmetric_keys/x509-asn1.o … CC crypto/asymmetric_keys/x509_rsakey-asn1.o … CC crypto/asymmetric_keys/x509_cert_parser.o … CC fs/ext4/extents.o … CC Dec 06 11:10:23 crypto/asymmetric_keys/pkcs7-asn1.o … CC crypto/asymmetric_keys/pkcs7_parser.o … LD crypto/asymmetric_keys/pkcs7_test_key.o … LD crypto/asymmetric_keys/x509_key_parser.o … LD crypto/asymmetric_keys/pkcs7_message.o … LD crypto/asymmetric_keys/built-in.o … CC crypto/hash_info.o … CC crypto/ablk_helper.o … LD crypto/crypto.o … LD crypto/crypto_algapi.o … LD crypto/crypto_blk Dec 06 11:10:24 crypto/crypto_hash.o … LD crypto/cryptomgr.o … LD crypto/built-in.o … CC block/bio.o … CC fs/ext4/ext4_jbd2.o … CC fs/ext4/migrate.o … CC block/elevator.o … CC fs/ext4/mballoc.o … CC block/blk-core.o … CC block/blk-tag.o … CC fs/ext4/block_validity.o … CC block/blk-sysfs.o … CC fs/ext4/move_extent.o … CC block/blk-flush.o … CC fs/ext4/m Dec 06 11:10:25 fs/ext4/indirect.o … CC block/blk-ioc.o … CC block/blk-map.o … CC fs/ext4/extents_status.o … CC block/blk-exec.o … CC block/blk-merge.o … CC fs/ext4/xattr.o … CC block/blk-softirq.o … CC fs/ext4/xattr_user.o … CC block/blk-timeout.o … CC fs/ext4/xattr_trusted.o … CC block/blk-iopoll.o … CC fs/ext4/inline.o … CC block/blk-lib.o … CC Dec 06 11:10:26 fs/ext4/readpage.o … CC fs/ext4/acl.o … CC block/blk-mq-tag.o … CC fs/ext4/xattr_security.o … CC block/blk-mq-sysfs.o … CC fs/ext4/crypto_policy.o … CC block/blk-mq-cpu.o … CC fs/ext4/crypto.o … CC block/blk-mq-cpumap.o … CC fs/ext4/crypto_key.o … CC block/ioctl.o … /home/vignesh1308/halium2/kernel/leeco/msm8996/fs/ext4/crypto_key.c: In function '_ext4_get_encryption_in Dec 06 11:10:27 /home/vignesh1308/halium2/kernel/leeco/msm8996/fs/ext4/crypto_key.c:228:29: error: invalid application of 'sizeof' to incomplete type 'struct ext4_encryption_key' … if (ukp->datalen != sizeof(struct ext4_encryption_key)) { … ^ … /home/vignesh1308/halium2/kernel/leeco/msm8996/fs/ext4/crypto_key.c:236:16: error: dereferencing pointer to incomplete type … if (master_key->size != EXT4_AES_256_XTS_KEY_SIZE) { … Dec 06 11:10:28 from /home/vignesh1308/halium2/kernel/leeco/msm8996/include/linux/kernel.h:13:0, … from /home/vignesh1308/halium2/kernel/leeco/msm8996/include/linux/list.h:8, … from /home/vignesh1308/halium2/kernel/leeco/msm8996/include/linux/key.h:19, … from /home/vignesh1308/halium2/kernel/leeco/msm8996/include/keys/encrypted-type.h:18, … from /home/vignesh1308/halium2/kernel/leeco/msm8996/fs/ext4/c Dec 06 11:10:29 /home/vignesh1308/halium2/kernel/leeco/msm8996/fs/ext4/crypto_key.c:239:18: error: dereferencing pointer to incomplete type … master_key->size); … ^ … /home/vignesh1308/halium2/kernel/leeco/msm8996/include/linux/printk.h:288:17: note: in definition of macro 'printk_once' … printk(fmt, ##__VA_ARGS__); \ … ^ … /home/vignesh1308/halium2/kernel/leeco/msm8996/fs/ext4/crypto_key.c:249:50: error: dereferenci Dec 06 11:10:30 type … res = ext4_derive_key_aes(ctx.nonce, master_key->raw, Dec 06 11:10:32 ^ … /home/vignesh1308/halium2/kernel/leeco/msm8996/scripts/Makefile.build:257: recipe for target 'fs/ext4/crypto_key.o' failed … make[3]: * [fs/ext4/crypto_key.o] Error 1 … /home/vignesh1308/halium2/kernel/leeco/msm8996/scripts/Makefile.build:402: recipe for target 'fs/ext4' failed … make[2]: * [fs/ext4] Error 2 … /home/vignesh1308/halium2/kernel/leeco/msm8996/Makefile:946: recipe for target 'fs' failed … make[1]: * [fs] Error 2 … mak Dec 06 11:10:33 unfinished jobs.... … CC block/genhd.o … CC block/scsi_ioctl.o … CC block/partition-generic.o … CC block/ioprio.o … CC block/partitions/check.o … CC block/bounce.o … CC block/partitions/msdos.o … CC block/bsg.o … CC block/partitions/efi.o … CC block/blk-cgroup.o … LD block/partitions/built-in.o … CC block/noop-iosched.o … CC block/deadline-iosch Dec 06 11:10:34 CC block/test-iosched.o … CC block/bfq-iosched.o … CC block/compat_ioctl.o … LD block/built-in.o … make[1]: Leaving directory '/home/vignesh1308/halium2/out/target/product/x2/obj/KERNEL_OBJ' … Makefile:145: recipe for target 'sub-make' failed … make: * [sub-make] Error 2 … make: Leaving directory '/home/vignesh1308/halium2/kernel/leeco/msm8996' … [ 57% 1233/2152] target C++: libc++abi <= external/libcxxabi/src/cxa_dem Dec 06 11:10:35 stopped: subcommand failed. … build/core/ninja.mk:151: recipe for target 'ninja_wrapper' failed … make: * [ninja_wrapper] Error 1 … make: Leaving directory '/home/vignesh1308/halium2' … #### make failed to build some targets (04:30 (mm:ss)) #### Dec 06 11:10:39 Erm Dec 06 11:10:44 Don't do it again Dec 06 11:10:51 sorry Dec 06 11:10:55 @bhushanshah, What? Dec 06 11:11:09 Never paste more then 4 lines in chat Dec 06 11:11:44 http://haydo.esy.es/workspace/paste/paste/1512558037.txt Dec 06 11:11:53 @bhushanshah, Have you got any idea about the error? Dec 06 11:12:48 @parduscix, @sopernam Dec 06 11:13:20 try disabling : CONFIG_EXT4_FS_ENCRYPTION @shenoy13 Dec 06 11:13:34 @bhushanshah, ok Dec 06 11:14:18 @bhushanshah, This is not found Dec 06 11:14:36 Any way I will create and initialize to 'n' Dec 06 11:22:25 (Document) https://irc.ubports.com/Mw8uDe0X/file_2693 Dec 06 11:22:33 im getting this error Dec 06 11:25:37 @sopernam, @ubports_bot Dec 06 11:29:33 hm I didnt have that linker error in my branch Dec 06 11:31:18 :( Dec 06 11:32:15 maybe compare kernel config? @anpok can you paste your kernel config? Dec 06 11:32:21 and you are also porting to sumire? Dec 06 11:32:26 yes Dec 06 11:32:35 I wanted to push my stuff to ubports yesteday Dec 06 11:32:42 ah.. how far is your port? Dec 06 11:32:45 but yeah .. reality affected the plans Dec 06 11:33:00 booting into rootfs after usc starts up Dec 06 11:33:03 then nothing happens Dec 06 11:33:06 no sshd starts .. Dec 06 11:33:18 And halium reference rootfs + hybris-boot works? Dec 06 11:33:34 I think that's easiest method to at least verify that hwcomposer etc bits are working Dec 06 11:33:39 no didnt try that Dec 06 11:33:47 but I made fix mountpoints.sh changes.. Dec 06 11:33:50 maybe worth trying :) Dec 06 11:34:06 but I need 60s sleep inside the init scripts:) Dec 06 11:34:09 because of things.. Dec 06 11:34:31 also.. yay (not!) for duplicated work :-( Dec 06 11:34:53 What should I do now? Dec 06 11:35:24 @sopernam I guess.. best wait for @anpox … to upload his changes somewhere.. and then you can base it on top of it Dec 06 11:35:25 wait a few minutes .. heading home to my workstation Dec 06 11:35:45 oaky Dec 06 11:52:16 https://pastebin.com/2nNYdxfp Dec 06 11:52:32 Error about mkbooting Dec 06 11:53:41 @shenoy13, Any idea about this one? Dec 06 11:54:08 `mka mkbootimg` Dec 06 11:54:18 @bhushanshah, ok Dec 06 12:11:36 hi … What is the situation? Dec 06 12:12:04 Patience :) Dec 06 12:12:10 :) Dec 06 12:46:00 Tuboil was added by: Tuboil Dec 06 12:46:55 here I am Dec 06 12:47:55 hi Dec 06 12:48:13 did you fix it ? Dec 06 12:49:40 hehe Dec 06 12:49:49 :) Dec 06 12:50:39 ? :D Dec 06 12:51:37 Just generated the fork now cleaning up the diff a bit Dec 06 12:51:40 then pushing Dec 06 12:53:38 Davide was added by: Davide Dec 06 12:53:49 oops I might have pushed the wrong buttons and github goes the long way round to fork the lineageos kernel repo Dec 06 12:54:20 :D Dec 06 12:58:26 https://github.com/ubports/android_kernel_sony_msm8994 .. it will go there Dec 06 12:58:41 but I made other changes too.. Dec 06 12:58:56 (Photo, 465x167) https://irc.ubports.com/49Hnm5nX/file_2695.jpg Dec 06 12:59:17 I have config mouldes enabled.. Dec 06 12:59:19 I think Dec 06 12:59:29 @anpok, empty Dec 06 12:59:35 yeah still forking Dec 06 13:00:32 @anpok, Do I need to change the kernel line in the manifest? Dec 06 13:01:03 ... you would have too .. but the sync should be fast because we both based on lineagos 14.1 Dec 06 13:01:45 Hi, I am Davide from Planet Computers. I would like to experiment porting halium on our Gemini device, based on Mediatek X27 chip. I have a question - we don't have lineageos working on it (yet) and our code is currently based on android aosp 7.1.1. Can I port halium using that? Dec 06 13:03:20 nice hardware Dec 06 13:04:37 would be really nice if you had open drivers for that ;) Dec 06 13:05:04 try it! Dec 06 13:07:17 https://pastebin.com/EzYwTa6E Dec 06 13:07:38 @shenoy13, Mer kernel check shows no error in the config Dec 06 13:07:57 @Davide, to answer your question, i've seen ports based on other android ROM's so it's at least theoretically possible Dec 06 13:08:15 but i think bsah or marius are the ones to really comment on it Dec 06 13:08:16 But If I add CONFIG_MODULES=y in my config, build fails Dec 06 13:09:02 How to enable kernel modules? Dec 06 13:09:13 for instance, cheeseburger on current halium repo is XOS based (even though that probably only boots) Dec 06 13:11:05 @sopernam https://github.com/ubports/android_kernel_sony_msm8994-1/tree/halium-7.1 Dec 06 13:11:14 -1 ... becaue I did it again.. Dec 06 13:11:29 the def config passes the ubports kernel config check.. Dec 06 13:11:49 https://pastebin.com/9EawKeLJ Dec 06 13:11:53 Thanks Rogier. I tried following the guide for lineageos but got stuck quite soon. I can see the device when I do 'lunch' but it does not show up with 'breakfast'... Dec 06 13:11:59 @anpok, how can i check this Dec 06 13:12:12 sorry Dec 06 13:12:13 @shenoy13, This is shown when I add CONFIG_MODULES to my defconfig Dec 06 13:12:16 sync Dec 06 13:12:35 Any idea about this error? Dec 06 13:12:48 @sopernam you can do it manually .. by adding that as a remote and git pulling the changes Dec 06 13:12:53 or change the local manifest Dec 06 13:13:10 @anpok, how 😭 Dec 06 13:13:11 i am not sure whether repo sync will remove the other repo or just get the diff Dec 06 13:13:47 change Dec 06 13:13:48 into Dec 06 13:14:48 Dec 06 13:15:04 going through the other repos now.. Dec 06 13:16:00 @anpok, ty Dec 06 13:17:14 hm I remoed a jni wrapper library from android_device_sony_sumire device.mk Dec 06 13:19:16 @anpok, fatal: repository 'https://github.com/android_kernel_sony_msm8994-1/' not found Dec 06 13:20:02 @Davide, try "breakfast " Dec 06 13:20:35 change kernel parameter to ubports/android_kernel_sony_msm8994-1 Dec 06 13:20:52 Dec 06 13:20:54 i.e Dec 06 13:21:04 ty Dec 06 13:22:10 Thanks Rogier, I'll try that Dec 06 13:36:03 @anpok did you edit kernel/sony/msm8994/arch/arm64/configs/kitakami_sumire_defconfig Dec 06 13:44:24 ywa Dec 06 13:44:26 yes Dec 06 13:44:36 hold on Dec 06 13:44:37 arm64 Dec 06 13:44:40 hmm nay Dec 06 13:45:37 oh sorry Dec 06 13:46:12 the config arch/arm64/configs/kitakami_sumire_defconfig is the one I use right now Dec 06 13:46:30 for a second I thought there was another one in the arch/arm directory Dec 06 13:59:11 (Document) https://irc.ubports.com/gIty0w0M/file_2697 Dec 06 14:04:38 same error Dec 06 14:04:46 😐 Dec 06 14:05:59 @anpok Dec 06 14:15:52 I used old configs. Dec 06 14:20:28 @sopernam, Even Iam stuck at same error Dec 06 14:20:44 Maybe problem with halium and arm 64 architecture Dec 06 14:21:24 @shenoy13, Your problem is not error btw Dec 06 14:21:32 It's just warning Dec 06 14:21:43 Feel free to ignore it Dec 06 14:22:27 @bhushanshah, If I edit defconfig to have CONFIG_MODULES, then it will show same error as the above got by supernam Dec 06 14:22:38 @bhushanshah, Ok Dec 06 14:22:49 Don't have config_modules Dec 06 14:22:53 You don't need it Dec 06 14:23:21 @sopernam, . Dec 06 14:24:56 @sopernam, Just remove config_modules from defconfig Dec 06 14:25:05 It would compile successfully Dec 06 14:26:12 but not boot Dec 06 14:34:56 (Photo, 915x220) https://irc.ubports.com/H0SPZXQm/file_2701.jpg Dec 06 14:35:10 Iam getting this error for mka-systemimage Dec 06 14:46:50 @sopernam, hm? whats the current status now? Dec 06 14:46:57 I was afk .. Dec 06 14:48:13 @sopernam, @anpok Dec 06 14:48:31 Fwd from sopernam: same error Dec 06 14:48:32 Fwd from sopernam: I used old configs. Dec 06 14:49:03 Guys it is correct?? Because it says that it can't find remote los … Dec 06 14:49:46 Are you building for halium 5? Dec 06 14:51:29 Yes, but i resolved, thanks Dec 06 15:02:11 @sopernam, ok seems to two missing flags.. I will see why they are not enable Dec 06 15:04:23 so the config I have has FIPS disabled and HW_RANDOM_MSM enabed Dec 06 15:04:25 enabled.. Dec 06 15:04:40 then msm_rng.c should be linked to the kernel Dec 06 15:04:58 and that translation unit defined g_fips140_status.. Dec 06 15:05:37 @anpok, how ? Dec 06 15:17:48 msm_rng.o is not in the list of translation units Dec 06 15:18:13 so CONFIG_HW_RANDOM_MSM=y is not true? Dec 06 15:20:07 im fix it using mer kernel check Dec 06 15:20:28 i will try 3th build Dec 06 15:21:05 what is inside out/target/product/sumire/obj/KERNEL_OBJ/.config ? Dec 06 15:21:19 .. because that is the actual one used during the last build Dec 06 15:21:48 (Document) https://irc.ubports.com/0TVQVLiY/file_2703 Dec 06 15:22:08 @anpok, yes its true Dec 06 15:22:44 could you look intside ${your_build}/out/target/product/sumire/obj/KERNEL_OBJ/.config too? Dec 06 15:26:32 I am using the config of your repo right now. I checked and corrected it. I'm doing a new build now. Dec 06 15:26:37 please wait Dec 06 15:28:25 I am actively poliing this connection Dec 06 15:29:30 (Photo, 1280x720) https://irc.ubports.com/k0DpZT08/file_2705.jpg Dec 06 15:32:45 (Photo, 1280x720) https://irc.ubports.com/48571FB0/file_2707.jpg Dec 06 15:32:57 thank you so much @anpok Dec 06 15:44:57 @sopernam, Now the real work starts. Think you just flash it and it works? Lets see... I ask you again in a few months ^^ Dec 06 16:33:35 @sopernam, You are appreciated your patience... You really struggled to compile all the stuff.. Dec 06 16:35:41 (Photo, 816x549) https://irc.ubports.com/GPuC9g3G/file_2713.jpg Dec 06 16:36:34 @technicalbird, thank you ☺️ Dec 06 16:56:11 Meg Dec 06 16:56:13 Meh Dec 06 17:41:42 (Video, 52s)https://irc.ubports.com/z2HjcXt8/file_2714.mp4 Dec 06 17:41:54 its boot but crashing Dec 06 17:45:25 @sopernam, Thats called bootloop and its normal at beginning. No need to send a video, nothing exciting Dec 06 17:45:38 @Flohack, Yes Dec 06 17:45:55 @Flohack, But my friends want video Dec 06 18:11:55 Hi now what should i? Dec 06 18:19:12 (Video, 31s)https://irc.ubports.com/skTjTQ3V/file_2718 Dec 06 18:22:34 @sopernam, is it rebooting? Dec 06 18:23:18 Yes Dec 06 18:23:49 @vanyasem can help you. He helped me too on same issue. Dec 06 18:24:12 You summoned me Dec 06 18:24:18 so what is the issue Dec 06 18:24:26 @sopernam, this Dec 06 18:24:27 @TheImpulson, wow he's free software activist Dec 06 18:24:57 don't worry, i Dec 06 18:25:35 so are there any logs? Dec 06 18:25:47 have you tried booting the recovery image? Dec 06 18:25:54 @vanyasem, No Dec 06 18:26:03 do it! Dec 06 18:26:07 Okay Dec 06 18:26:10 mka hybris-recovery in your tree Dec 06 18:26:17 Ty Dec 06 18:26:21 and flash hybris-recovery.img as your BOOT image Dec 06 18:26:45 it's not a recovery! it's just a modified boot image that doesn't reboot on error Dec 06 18:27:01 Okay Dec 06 18:27:02 so we can telnet in and see what went wromg Dec 06 18:36:34 @vanyasem, I flashed it stuck on brand logo Dec 06 18:36:50 i am sorry but what did you expect Dec 06 18:36:58 naked women? Dec 06 18:37:06 😂 Dec 06 18:37:07 tbqh, a beautiful... nvm. Dec 06 18:37:10 😂 Dec 06 18:37:12 @sopernam, that's intended ;) Dec 06 18:37:16 now telnet in as per docs Dec 06 18:37:34 @vanyasem, Im flashed hybris-recovery to recovery block Dec 06 18:37:47 why would you do that. Dec 06 18:37:54 i have literally told you not to Dec 06 18:38:05 now reflash TWRP and do everything from scratch Dec 06 18:38:26 @vanyasem, Okayy Dec 06 18:39:04 @vanyasem, Then Dec 06 18:39:30 flash hybris-recovery to boot and boot the device Dec 06 18:39:46 if it doesn't reboot, it means it worked, so we will be able to telnet in Dec 06 18:39:58 Oh okay Dec 06 18:40:49 @vanyasem, Same error Dec 06 18:41:08 does it reboot with the recovery image flashed on boot? Dec 06 18:41:44 @vanyasem, Yes Dec 06 18:41:55 then you messed something really badly Dec 06 18:41:59 is /data encrypted? Dec 06 18:42:03 No Dec 06 18:42:08 are you sure? Dec 06 18:42:13 Yes Dec 06 18:42:31 Im sure Dec 06 18:42:38 have you added your device to fixup-mountpoint? Dec 06 18:43:01 Yes Dec 06 18:43:48 were they pick up by the compiler? Dec 06 18:45:35 @vanyasem, 🤔 Dec 06 18:46:57 @sopernam, halium/hybris-boot/Android.mk Dec 06 18:47:09 @shenoy13, See this file Dec 06 18:47:21 Okay Dec 06 18:49:36 (Photo, 628x69) https://irc.ubports.com/ne2507ul/file_2720.jpg Dec 06 18:50:29 @vanyasem Dec 06 18:50:45 you have just highlighted the code from the makefile Dec 06 18:50:56 im asking if they were set during the build process Dec 06 18:51:03 why would you show me the damn makefile Dec 06 18:51:17 (Document) https://irc.ubports.com/GGiwQUf6/file_2721.mk Dec 06 18:52:03 WHY WOULD YOU SEND ME THE MAKEFILE Dec 06 18:52:17 i have the exact same one, it's not related to the issue Dec 06 18:52:23 please close it and forget it's location Dec 06 18:52:33 now build your systemimage again Dec 06 18:52:44 @vanyasem, and look for that output in the terminal Dec 06 18:52:55 then send me the paths it showed you Dec 06 18:53:37 @vanyasem, i have closed terminal Dec 06 18:53:47 open it then ? Dec 06 18:54:00 and rebuild systemimage as i tould you? Dec 06 18:54:08 @vanyasem, okay Dec 06 18:54:43 Sopernam adam sana söylediğim gibi build ettin mi diyor? @sopernam Dec 06 18:55:00 I like how Orkun is triggering everyone in this group Dec 06 18:55:03 HAHAHA Dec 06 18:55:24 @sopernam, Bish Dec 06 18:56:11 😐🔪 Dec 06 18:57:54 @LLJYY be respect Dec 06 18:58:23 @sopernam, I'm not respect Dec 06 18:58:28 My name is Jeff Dec 06 18:59:00 😕 Dec 06 19:01:48 @vanyasem, You didnt tell me something about how to build Dec 06 19:02:06 have you read the documentation? Dec 06 19:02:12 yes Dec 06 19:02:20 @LLJYY, i hate you Dec 06 19:02:29 have you read the section about building systemimage? Dec 06 19:03:18 @vanyasem, https://docs.halium.org/en/latest/porting/build-sources.html … yes im read Dec 06 19:03:39 so why should i tell you how to build systemimage if it's already in the docs? Dec 06 19:04:30 @vanyasem, mka systemimage Dec 06 19:04:34 He want help! We want teach and learn halium system. Dec 06 19:04:47 pastebin.com/iSyiBEw4 trying to port to nexus 4, but it can't find a valid RAM Disk Image, and Root device, can someone point me in the right direction on how to fix this? Dec 06 19:04:49 well i can't babysit him Dec 06 19:05:21 i can help for sure, but not walk through every step Dec 06 19:09:00 What do you expect from a rookie? to get you into the space? Dec 06 19:09:59 * vanyasem leaves Dec 06 19:10:24 😶 Dec 06 20:01:24 re .. Dec 06 20:02:15 so @sopernam so the next problem with z5 is that for some reason the lineage kernel takes up to uptime == 65s to finally detect the partition layout of the device Dec 06 20:02:25 that means any earlier attempt to mount a partition fails Dec 06 20:02:47 I avoided the problem by adding a 60sec sleep into the init-ctl of the intramfs Dec 06 20:02:52 initramfs Dec 06 20:03:18 .. untill I step through the platform init of the kernel and figure out what blocks the kernel .. Dec 06 20:03:37 @sopernam lookinto hybris-boot Dec 06 20:03:43 there should be an initramfs directory Dec 06 20:03:51 you need to make a surgery there.. Dec 06 20:04:19 @anpok, okay Dec 06 20:04:50 @anpok, the problem is not related to the screen drivers? Dec 06 20:12:06 @anpok, Thank you for your help. Now I'm taking this job. I will continue on the weekend Dec 06 20:14:00 Because I have my exam for two days Dec 06 20:15:02 I note what you said. Dec 06 20:26:02 @developerfect, hm why? Dec 06 20:34:17 @Davide, Just look further at the page ... I would really love to see hallium working there.. Please try.. and.. yeah if you have spare devices.. ... Dec 06 20:52:25 @Davide, You need a device tree, kernel source and vendor blobs. The AOSP device tree could work with minor adaptions I think (I'm not completely sure though) Dec 06 21:36:23 thanks - I would love too to see halium running on our device! At the moment I tried following the guide and I ended up with lots of "warning: multiple rules generate ..." and a dependency cycle error Dec 06 21:37:15 ninja: error: dependency cycle: /Code/halium/out/target/common/obj/JAVA_LIBRARIES/CustomPropInterface_intermediates/classes.dex -> /Code/halium/out/target/common/obj/JAVA_LIBRARIES/CustomPropInterface_intermediates/with-local/classes.dex -> vendor/mediatek/proprietary/frameworks/base/custom/interface/com/mediatek/custom/CustomPropInterface.java -> /Code/halium/out/target/common/obj/JAVA_LIBRARIES/CustomPropInterface_intermediates/classes.dex Dec 06 21:37:58 You need to remove the Java and apk parts from the vendor and device tree Makefiles Dec 06 21:38:19 Halium doesn't include any Java or dalvikvm parts Dec 06 21:39:00 Thanks - I'll try that Dec 06 21:40:35 pitelpan was added by: iosifidis Dec 06 22:48:28 I think the system now sees the kernel: Dec 06 22:48:32 No private recovery resources for TARGET_DEVICE aeon6797_6m_n Dec 06 22:49:13 it complained about the config file being missing so I added one, but now it complains about not knowing what to do with it: ninja: error: 'kernel/eastaeon/aeon6797_6m_n/arch/arm64/configs/aeon6797_6m_n_debug_deconfig', needed by '/Code/halium/out/target/product/aeon6797_6m_n/obj/KERNEL_OBJ/.config', missing and no known rule to make it **** ENDING LOGGING AT Thu Dec 07 03:00:01 2017 **** BEGIN LOGGING AT Thu Dec 07 03:00:02 2017 Dec 07 05:49:16 The config needs to be in kernel/eastaeon/aeon6797_6m_n/arch/arm64/configs/aeon6797_6m_n_debug_defconfig Dec 07 06:07:51 @anpok, should i edit hybris-boot.img or hybris-boot directory ? Dec 07 06:24:27 (Photo, 1280x443) https://irc.ubports.com/tHpunblB/file_2724.jpg Dec 07 06:56:29 I need to edit init-script imo. @anpok Dec 07 09:07:37 Buğra Uygur was added by: Buğra Uygur Dec 07 11:07:31 @JBBgameich - my mistake, I fixed that, now the build process starts but I get an error while compiling the kernel: /Code/halium/kernel/eastaeon/aeon6797_6m_n/scripts/Makefile.headersinst:55: *** Missing UAPI file /Code/halium/kernel/eastaeon/aeon6797_6m_n/include/uapi/linux/autoconf.h. Stop. Dec 07 11:41:45 Keshav was added by: Keshav Dec 07 11:46:24 Sa Dec 07 11:50:44 @Buğra Uygur, only english Dec 07 12:10:23 you need to learn for exams Dec 07 12:12:47 @Buğra Uygur, as Dec 07 12:15:24 @sopernam, yes hybris boot seems to assemble the initramfs during build .. so you can edit the init-script directly Dec 07 12:15:39 add a sleep 60s prior to mounting .. Dec 07 12:16:21 also take a look at the Readme.md about debugging the system startup Dec 07 12:16:33 using fasboot boot ... Dec 07 12:17:21 @anpok, okay Dec 07 12:18:19 @anpox: FWIW, this 60s sleep failure sounds *strange* have you got dmesg or anything? Dec 07 12:18:29 is it same for android ROM? Dec 07 12:19:39 @bshah, no Dec 07 12:20:46 dsmeg last kernel log? where i found this ? Dec 07 12:21:12 the mmc driverreports the paritions 65s after starting up .. consistently Dec 07 12:22:20 @anpok can you provide me dmesg when you get time? Dec 07 12:24:21 >[ 63.833283] FFU: mmc0: Firmware catalog was not found /emmc/catalog.txt -12 … <6>[ 63.833298] mmc0: new HS400 MMC card at address 0001 … <6>[ 63.833759] mmcblk0: mmc0:0001 032G74 29.1 GiB … <6>[ 63.833887] mmcblk0rpmb: mmc0:0001 032G74 partition 3 4.00 MiB … <6>[ 63.838968] mmcblk0: p1 p2 p3 p4 p5 p6 p7 p8 p9 p10 p11 p12 p13 p14 p15 p16 p17 p18 p19 p20 p21 p22 p23 p24 p25 p26 p27 p28 p29 p30 p31 p32 p33 p34 p35 p36 p37 p38 p39 p40 Dec 07 12:24:48 this if I recall correctly is with the copyleft kernel from the sony repositories.. Dec 07 12:25:20 maybe that is due to /emmc/catalog.txt not found in in the rootfs Dec 07 12:26:57 something to dig further into when I have time again.. Dec 07 12:32:48 @anpok, it might be worth while to ask in ##linux-msm or some qcom related channels, because I for one sure remember that there is some bug in 4.x kernel which makes it have 60s timeout Dec 07 12:43:13 http://paste.ubuntu.com/26132309/ this is the full log from the ubports recovery image that I built with a different kernel tree .. but yeah behavior is essentially the same Dec 07 12:43:33 it might be related to some encrypted partitions ... Dec 07 12:44:10 oh it's 3.10 kernel hm Dec 07 12:46:31 patched beyond recognition Dec 07 12:47:07 now what i do ? Dec 07 12:47:31 understand how the mmc driver init is triggered Dec 07 12:47:36 anpok: I suggest to disable verity.. (fastboot oem disable-verity or something like that) and retry Dec 07 12:47:51 I think it's verity which is screwing up things maybe Dec 07 12:47:54 verity? Dec 07 12:49:22 https://source.android.com/security/verifiedboot/dm-verity Dec 07 12:49:33 not exactly sure how to disable it though) Dec 07 12:50:39 'fastboot oem disable_dm_verity' duckduckgo says Dec 07 12:52:32 When we flash any root patch in stock kernel the same error appears. Dec 07 12:52:45 (Photo, 1280x637) https://irc.ubports.com/wQhMwTuf/file_2767.jpg Dec 07 12:53:25 hm it doesnt like the commands.. Dec 07 12:53:36 maybe there is a kernel option .. Dec 07 13:05:10 (Photo, 1080x1920) https://irc.ubports.com/1Syp0K2M/file_2769.jpg Dec 07 13:05:21 I got this error becuase i have to remove all android references? Dec 07 13:14:21 how can i disable it ? Dec 07 14:06:42 @Mattia, No ok i'm stupid ahahh i resolved it Dec 07 14:09:05 @sopernam, verity? I think it's a kernel config option Dec 07 14:09:11 CONFIG_DM_VERITY Dec 07 14:15:22 CONFIG_DM_VERITY=n ? Dec 07 14:19:26 yeah Dec 07 14:19:48 ☺️ Dec 07 15:01:41 use the source ... Dec 07 15:02:03 ... still at work .. cant do anythng... you could learnhow the init script works Dec 07 15:02:15 and add a temporary workaround for now.. Dec 07 15:02:51 or go to the kernel source and search for emmc/catalog.txt .. Dec 07 15:02:57 learn how that part works Dec 07 15:03:13 then make mr against the kernel Dec 07 15:05:38 @anpok, me? Dec 07 15:05:48 yeah Dec 07 15:06:15 the purpose of this here is to enable a community of people to solve hallium related problems Dec 07 15:07:04 sheltered accommodation... erhm.. learning Dec 07 15:09:33 okay Dec 07 19:06:19 Bhushan Shah: does jack work with your audio patch? Dec 07 19:17:00 @i_love_halium_on, only aux not working. but headset working Dec 08 00:42:10 (Document) https://irc.ubports.com/lVCluKYD/file_2773.jpg Dec 08 00:42:15 It's halium 5.1... what i have to falsh now? System and boot or i have to use the installer script form github? **** ENDING LOGGING AT Fri Dec 08 03:00:02 2017 **** BEGIN LOGGING AT Fri Dec 08 03:00:02 2017 Dec 08 09:35:38 new pm rootfs graphichs doesn't work- test_hwcomposer does. Did someone break something or it's my problem? Dec 08 09:36:20 i_love_halium_on: take previous pm rootfs Dec 08 09:36:22 not last one Dec 08 09:37:39 Take previous rootfs not latest one (seems my message from irc didn't get relayed) Dec 08 09:38:46 and i cant update it? Dec 08 09:39:08 no Dec 08 09:39:32 (too many things broken.. :<) Dec 08 09:40:51 kk :/ Dec 08 09:43:02 halium doesn't boot when data is encrypted i assume? Dec 08 09:56:27 hmmno, i get an error from halium_install Dec 08 09:56:43 Reserved block group size: 15 … error: ext4_allocate_best_fit_partial: failed to allocate 19 blocks, out of space? Dec 08 09:57:00 from make_ext4fs Dec 08 09:57:15 problem is i have plenty of free space on my local disk Dec 08 09:57:20 anyone any idea? Dec 08 09:57:53 It's possibly about space on device Dec 08 09:58:55 i think that has 80G free Dec 08 10:03:56 hmm that rootfs has a timestamp from june Dec 08 10:04:00 that's probably wrong Dec 08 10:04:13 No it's not Dec 08 10:04:16 :) Dec 08 10:05:20 eh ok Dec 08 10:06:22 i can't even find workdir and tmpdir Dec 08 10:06:24 i must be loosing it Dec 08 10:08:01 hmmno it's there Dec 08 10:08:15 should be plenty of space - 3G+ raw file though? Dec 08 10:08:53 Do you have full install log? Dec 08 10:09:43 yeah Dec 08 10:09:51 pastebin? Dec 08 10:10:04 Yes please Dec 08 10:11:11 https://pastebin.com/RTCkL9Bm Dec 08 10:12:38 Is your device umm encrypted? Dec 08 10:13:16 presumably\ Dec 08 10:13:39 Best try to remove encryption and retry Dec 08 10:13:52 hmm that's a factory reset Dec 08 10:18:26 hmm won't even let me do it Dec 08 10:22:09 did it, didn't help though Dec 08 10:23:45 You can try @JBBgameich fork of Halium-scripts Dec 08 10:24:07 It I think fixes this error by creating rootfs locally and then pushing it to device Dec 08 10:28:22 In halium-install.sh there is a parameter set to 180Mb i think. Try to inrease it Dec 08 10:29:28 convert_android_img() … { … if file $SYSIMG | grep -v ": Linux rev 1.0 ext4" >/dev/null; then … simg2img $SYSIMG $WORKDIR/system.img.raw … mkdir $TMPMOUNT … mount -t ext4 -o loop $WORKDIR/system.img.raw $TMPMOUNT … make_ext4fs -l 160M $WORKDIR/system.img $TMPMOUNT … SYSIMAGE=$WORKDIR/system.img … else … SYSIMAGE=$SYSIMG … fi … } Dec 08 10:29:52 In this part of halium install there is 160M increase it Dec 08 10:30:07 yeah i tried that first - it gave me a different number of blocks it couldn't allocate (both more or less size) Dec 08 10:30:24 Set to 200M or 210M Dec 08 10:35:05 jbbgameich version hangs at setting unix password here Dec 08 10:35:43 or is this supposed to take very long? Dec 08 10:36:29 Well it does ask you for password no? Dec 08 10:36:34 yeah Dec 08 10:36:53 it then shows what i type in plaintext (?) and nothing seemingly happens after enter Dec 08 10:40:06 No idea Dec 08 10:40:20 @JBBgameich ^ or @vanyasem Dec 08 10:40:33 chroot exists on device, but i don't see passwd file update there Dec 08 10:46:53 what was the repo to add and the package to install and the pulseaudio module to fix audio? Dec 08 10:47:31 i_love_halium_on: it is no longer needed to add extra repo, just upgrade pulseaudio Dec 08 10:48:04 (package got moved to xenial-updates so yay) Dec 08 10:49:38 ah ok Dec 08 10:53:25 @rogieroudshoorn, no Dec 08 10:56:47 bshah: if i try apt upgrade pulseaudio it wants to upgrade everything Dec 08 10:57:01 that's why you don't apt upgrade Dec 08 10:57:03 which will break graphics Dec 08 10:57:08 but apt install Dec 08 10:57:35 ah :D Dec 08 10:57:37 legit Dec 08 11:04:53 how can i test audio without graphics? because my compiled libhybris with textrel flag broke it Dec 08 11:05:14 paplay Dec 08 11:05:43 aww, and whats the audio module to add in default.pa? Dec 08 11:05:53 module-droid-card-22 Dec 08 11:06:01 thanks Dec 08 16:13:04 vasuratanpara was added by: vasuratanpara Dec 08 20:46:51 @anpok, I didn't find this Dec 08 20:47:16 find | grep emmc … ./arch/arm/boot/dts/qcom/mdm9640-emmc-cdp.dts Dec 08 20:47:51 find | grep catalog … ./fs/hfs/catalog.c … ./fs/hfsplus/catalog.c Dec 08 22:33:54 im trying dm verity disable. **** ENDING LOGGING AT Sat Dec 09 03:00:00 2017 **** BEGIN LOGGING AT Sat Dec 09 03:00:01 2017 Dec 09 06:00:27 (Document) https://irc.ubports.com/8t9KcxoR/file_2780.log Dec 09 06:01:22 is that halium kernel log ? Dec 09 06:02:09 samavar@grayleshy-virtualbox Dec 09 06:02:19 is this your username/hostname? Dec 09 06:02:59 though it looks like recovery dmesg Dec 09 06:03:59 (Photo, 720x1280) https://irc.ubports.com/90z8OIUz/file_2782.jpg Dec 09 06:04:15 bshah yep you are right Dec 09 06:04:23 this in in fact recovery dmesg Dec 09 06:04:58 😂 Dec 09 06:04:58 ok Dec 09 06:05:51 😒 Dec 09 06:21:38 how can i get this : Dec 09 06:36:35 were you finally able to telnet in? Dec 09 06:40:08 i will take that as a no Dec 09 06:51:23 @vanyasem, no Dec 09 06:51:34 i cant boot Dec 09 06:52:41 have you uploaded your modified tree to github already? Dec 09 06:52:41 also manifests? Dec 09 07:04:09 @vanyasem, I did not edit the kernels Dec 09 07:04:39 haven't you edited the configuration Dec 09 07:05:07 @vanyasem, where is ? Dec 09 07:05:20 kernel config Dec 09 07:05:33 have you? Dec 09 07:05:44 @vanyasem, yes is have Dec 09 07:06:39 why won't you commit it to your fork of the devices kernel? Dec 09 07:06:39 have you PRed your fixup-mountpoints? Dec 09 07:06:52 @vanyasem, pred ? Dec 09 07:06:53 have you edited your deviecs fstab file to represent the changes? Dec 09 07:07:05 @sopernam, sent a PULL REQUEST to upstream Dec 09 07:07:23 @vanyasem, ues Dec 09 07:07:34 everything is ok Dec 09 07:07:43 looks like you've made a lot of changes right? Dec 09 07:07:53 so make a freaking repo that i can look at Dec 09 07:08:16 okay Dec 09 07:08:23 i will make it Dec 09 07:08:25 wait Dec 09 07:08:30 but don't just squash everyrthing Dec 09 07:08:59 in 1 commit Dec 09 07:08:59 separate reference trees with your changes please Dec 09 07:10:39 okay Dec 09 07:38:55 @vanyasem, https://github.com/ubports/android_kernel_sony_msm8994-1/tree/halium-7.1 … @anpok edited for me Dec 09 07:39:18 and the device tree? Dec 09 07:39:41 wait Dec 09 07:40:13 not in a hurry, im here to help Dec 09 07:40:43 take your time Dec 09 07:41:45 https://github.com/LineageOS/android_device_sony_sumire Dec 09 07:46:33 (Document) https://irc.ubports.com/06hnqWYL/file_2783 Dec 09 07:59:33 @mariogrip Dec 09 12:28:38 @sopernam integrated that change... but why did you enable HCI UART and the other two CGROUP related options? Dec 09 12:48:46 oh mer kerne config required that Dec 09 13:49:43 @anpok, I might have changed it when I checked the mer kernel conrol. Dec 09 14:43:15 will look into the uart hci thing but I think the other two are not necessary.. Dec 09 15:00:10 you'll need hci uart if you want bt to work :) Dec 09 15:29:26 🤔 Dec 09 17:39:22 looks like bt dongle support Dec 09 18:28:52 Can i use mm linker on n in libhybris for 7.1 libhybris? Dec 09 18:29:00 Halium 7.1* Dec 09 18:34:27 anpok: sony devices tend to use broadcomm wifi+bt chips, which are connected to the main SoC vie high speed uart Dec 09 19:02:23 Anyone that ported 5.1 caf with graphics can help me? Dec 09 19:03:28 I can't fix the graphics problem. I think that'd be better than of fixing textrels in 7.1 Dec 09 19:15:07 This is the error i think it has to be fixed: Dec 09 19:15:24 mdss_mdp_overlay_pipe_setup: exceeded max mixer supported resolution Dec 09 19:35:38 I got a similar problem: you have to add a special caf repo and update Dec 09 19:36:28 What caf repo? Dec 09 19:37:03 Bhushan Shah, [08.11.17 14:37] … [Forwarded from [ matrix ]] … Bhushan Shah: So now what you will have to do is, add http://repo.halium.org/caf (http://repo.halium.org/caf) repository in apt sources, apt update, and apt upgrade Dec 09 19:37:34 I was looking for the reply bhushah gave to me :) Dec 09 19:37:43 In 5.1? They told me reference rootfs has caf support Dec 09 19:38:05 Wtf Dec 09 19:38:14 in my case it was 7.1... but try anyway Dec 09 19:38:26 Sure ill try Dec 09 19:38:48 in my case, it worked Dec 09 19:39:53 reference rootfs is just for testing... Dec 09 19:40:57 Yes, in fact it's missing caf repos Dec 09 19:40:58 Yes, but before running pm it's legit to test ;) Dec 09 19:47:15 repo.halium.org has 5.1 libhybris,too? Dec 09 20:01:29 I don't know, but I guess it has Dec 09 20:04:11 Ok, thanks Dec 09 21:19:27 https://pastebin.com/bFvqygKM <- Could someone have a look @ my last_kmsg? And maybe point me in the right direction? Dec 09 21:26:29 Hm.. found it, INITRD: 0x8120000+0x000a1bf9 overlaps in-use memory region - disabling initrd Dec 09 22:48:48 Fix ya dts Dec 09 23:14:43 @sopernam the problem is caused by the mmc driver asking to load firmware binaries in the init routine .. that is done by injecting and waiting for a handler that would look inside the fs to find a specific file .. the flag to disable is CONFIG_MMC_FFU Dec 09 23:15:02 is the halium install script really doing the right thing? Dec 09 23:15:56 I thought the android system img should end up in var lib lxc somewhere Dec 09 23:18:25 It's mounted to there I guess **** ENDING LOGGING AT Sun Dec 10 03:00:02 2017 **** BEGIN LOGGING AT Sun Dec 10 03:00:02 2017 Dec 10 05:49:23 So, when Halium is ready on the OP5, would it theoretically be possible to run KDE soon it if one chooses not to run Ubuntu Touch? Dec 10 09:06:12 @c_smith, You mean Plasma? Dec 10 09:08:54 @Flohack, Yes, plasma mobile if memory serves Dec 10 09:09:43 Yes @bhushanshah has the details but thats exactly the point of Halium, to make the OS more independent frmo the HAL and to use a common HAL vor various OSes Dec 10 09:42:32 @bhushanshah should I really use halium-install when I run hybris-boot? Dec 10 09:43:20 Yes, halium-install should work with the reference rootfs anf hybris-boot Dec 10 09:45:09 ih Dec 10 09:45:57 could it be that the makefiles of ubports-boot interfere with hybris-boot ? Dec 10 09:46:20 yes Dec 10 09:46:26 and they do Dec 10 09:46:29 that's an issue Dec 10 09:46:46 remove ubports boot before making hybris-boot Dec 10 09:51:17 kaushal7007 was added by: kaushal7007 Dec 10 09:52:47 I added halium caf repo and apt update, upgrade. now EGL_PLATFORM=hwcomposer test_hwcomposer segfaults but not test_hwcomposer Dec 10 17:40:51 Have any one ported (Ubuntu Touch) for Xiaomi Mi Redmi note 4? Dec 10 17:50:38 @vasuratanpara, no. Dec 10 17:51:18 it was started Dec 10 17:51:27 @vasuratanpara, It's in progress Dec 10 17:51:36 @vanyasem, by Herrie on IRC Dec 10 17:51:41 but it's not finished afaik Dec 10 17:52:36 Did herrie mention his github? Dec 10 17:53:38 https://github.com/Herrie82 Dec 10 17:54:03 https://github.com/Herrie82/android_device_xiaomi_mido … Herrie82/android_kernel_xiaomi_msm8953 Dec 10 17:54:35 (FYI port by Herrie is not tested yet) Dec 10 17:55:19 @bhushanshah, true Dec 10 17:55:28 that's why I answered no in the first place Dec 10 18:34:13 Thanks but any idea how to install it on device? Dec 10 18:35:49 @vanyasem, Ok thank you @vanyasem Dec 10 18:47:40 hm so it boots into the hybris rootfs .. now there is no adb but and usb rndis interface Dec 10 18:48:18 Confused.. there is rndis interface right? Dec 10 18:48:58 English being... English.. 😛 Dec 10 18:54:48 @anpok ^ Dec 10 18:57:54 so it is beyond the initramfs.. I assume Dec 10 18:58:33 Well rndis might be from the initramfs as well Dec 10 18:58:56 but after the udhcp and telnet was turned after? Dec 10 18:59:17 In that case it's rootfs Dec 10 19:00:12 Do you have an IP on rndis? Dec 10 19:00:27 http://docs.halium.org/en/latest/porting/debug-build/logging-in.html Dec 10 19:00:33 yeah, that Dec 10 19:00:58 ah Dec 10 19:01:05 ok will try that Dec 10 19:51:46 [ 212.249509] IPC_RTR: msm_ipc_router_bind: pm-service Do not have permission Dec 10 19:51:54 dmesg got spammed with that Dec 10 19:51:57 Meh. Dec 10 19:52:06 [ 212.089679] systemd-journald[651]: /dev/kmsg buffer overrun, some messages lost. Dec 10 19:52:14 What else isn't working? Dec 10 20:52:40 02-11 17:19:22.315 9 19 E NetlinkListener: recvmsg failed (I/O error Dec 10 20:52:56 02-11 17:17:49.081 3038 3038 I ta_qmi_service: Starting TA QMI service … 02-11 17:17:49.082 3038 3038 E ta_qmi_service: Unable to register TA QMI service! Dec 10 20:53:02 lots of those messages in logcat Dec 10 20:53:58 http://paste.ubuntu.com/26158988/ Dec 10 21:02:52 What isn't working? Dec 10 21:03:06 Well... What do you want to make work? Dec 10 21:13:02 ubports .. just tried halium rootfs to see what is wrong Dec 10 21:21:06 On the Redmi Note 4, managed to hard brick on first attempt.... Forgot to update boot partition to 021 and overwritten aboot on 019 Dec 10 21:21:47 So needed some surgery but it's alive again. Didn't have much time in past days to look further. It fails hard on my LuneOS boot but I think I might need to tweak a thing here or there Dec 10 21:45:28 wifi device is detected vibrator worked..but no graphics so far Dec 10 21:57:00 hm none of the qcom selinux poilcy files were installed into the system.img Dec 10 22:26:43 No graphics? Dec 10 22:27:00 Is it stuck on the bootloader or Dec 10 22:36:59 it still shows the vendor boot logo right now .. no one has claimed the screen yet Dec 10 22:37:34 Is it CAF device? Dec 10 22:43:01 `cat .cache/upstart/unity-system-compositor` Dec 10 22:44:57 Maybe it was `cat .cache/upstart/unity8` and `sudo unity-system-compositor` Dec 10 22:46:08 oh I am currently running the halium rootfs.. hoping to get a clue about what went wrong when booing ubports.. Dec 10 22:46:24 You won't get too many Dec 10 22:46:31 so in the best scenario of the ubports attempt I got unity system compositor to start Dec 10 22:46:47 but no sshd Dec 10 22:46:51 so no idea to look further Dec 10 22:47:05 Did you use new rootstock? Dec 10 22:47:22 with the sshd setup scripts.. Dec 10 22:47:53 @bhushanshah, how would I know Dec 10 22:48:09 It's in the UBports docs! Dec 10 22:48:12 :D Dec 10 22:49:07 https://docs.ubports.com/en/latest/porting/installing-16-04.html#choose-a-rootfs Dec 10 22:51:15 The dhcp doesn't work under ubports, you do need to set the IP using ip or ifconfig Dec 10 22:57:20 @bhushanshah, yes.. Dec 10 22:57:39 Okay in that case Dec 10 22:58:00 add http://repo.halium.org/caf (http://repo.halium.org/caf) repository in apt sources, apt update, and apt upgrade Dec 10 22:58:55 That should fix graphics on reference rootfs most likely Dec 10 23:00:34 @anpok, About this (it's related to paranoid network): see the bullhead kernel on GitHub, last 4 commits.. apply those changes Dec 10 23:01:31 "QTI_BSP" Dec 10 23:01:43 Are there TI based devices with halium support? Dec 10 23:01:58 I suppose no Dec 10 23:02:43 What's QTI? Dec 10 23:03:35 Qualcomm Technology.... Dec 10 23:03:42 Ahh **** ENDING LOGGING AT Mon Dec 11 03:00:01 2017 **** BEGIN LOGGING AT Mon Dec 11 03:00:02 2017 Dec 11 06:44:13 @bhushanshah hm test_hwcomposer only prints GL|ES version and the screen size.. then exists Dec 11 06:44:35 time to strace \o/ and logcat :) Dec 11 06:45:38 logcat is occupied by a broadcom daemon trying to find a bluetooth daemon Dec 11 06:45:51 grep hwc :) Dec 11 06:46:15 2-11 06:43:35.364 0 2502 I Adreno : IsValidNativeBuffer: Buffer has a NULL handle … 12-11 06:43:35.364 0 2502 I Adreno : IsValidNativeBuffer: Buffer has a NULL handle … 12-11 06:43:35.365 0 2502 I Adreno : IsValidNativeBuffer: Buffer has a NULL handle … 12-11 06:43:35.366 0 2502 I Adreno : DequeueBuffer: Dequeued Buffer is not valid … 12-11 06:43:35.367 0 2502 I Adreno : IsValidNativeBuffer: Buffer has a NULL handl Dec 11 06:46:16 06:43:35.367 0 2502 I Adreno : IsValidNativeBuffer: Buffer has a NULL handle Dec 11 06:46:36 that's it? Dec 11 06:46:51 and Dec 11 06:46:53 2-11 06:46:40.929 0 2570 E qdhwcomposer: getPartialUpdatePref: Failed to open sysfd node: sys/class/graphics/fb0/dyn_pu … 12-11 06:46:40.930 0 2570 I qdhwcomposer: Successfully Loaded perf hint API's … 12-11 06:46:40.930 0 2570 I qdhwcomposer: PerfHintWindow = 50 … 12-11 06:46:40.930 0 2570 I qdhwcomposer: Initializing Qualcomm Hardware Composer … 12-11 06:46:40.930 0 2570 I qdhwcomposer: MDP version: 500 … 12-11 06:4 Dec 11 06:46:54 D qdhwcomposer: hwc_setPowerMode: Setting mode 2 on display: 0 … 12-11 06:46:40.930 0 2570 D qdhwcomposer: hwc_setPowerMode: Done setting mode 2 on display 0 … 12-11 06:46:40.930 0 2570 D qdhwcomposer: hwc_getDisplayAttributes disp = 0, width = 1080 … 12-11 06:46:40.930 0 2570 D qdhwcomposer: hwc_getDisplayAttributes disp = 0, height = 1920 … 12-11 06:46:40.930 0 2570 E qdhwcomposer: Unknown display attribute 0 Dec 11 06:47:07 sounds good so far.. Dec 11 06:47:34 I guess you should grep -v the other broadcom spam and try to get a proper logcat Dec 11 06:47:41 also strace can be helpful Dec 11 06:55:11 http://paste.ubuntu.com/26162014/ Dec 11 06:56:26 can you provide strace output? Dec 11 06:56:44 (with EGL_PLATFORM=hwcomposer eported) Dec 11 06:58:31 ah what is the default egl platform? Dec 11 06:58:44 so thats working too Dec 11 06:59:30 as in you see spiral on screen? Dec 11 06:59:33 \o/ Dec 11 06:59:35 yes Dec 11 06:59:47 default platform would be null I believe Dec 11 07:00:24 So I think now that hwcomposer is working I can think only issue is ubports side Dec 11 07:01:17 hm and bt needs fixing and there is a bunch of suspicious messes in syslog Dec 11 07:01:27 1484.514308] init: Starting service 'uim'... … [ 1484.517623] init: couldn't write 873 to /dev/cpuset/system-background/tasks: No space left on device … [ 1484.553085] init: Service 'uim' (pid 873) exited with status 255 … [ 1484.553155] init: Service 'uim' (pid 873) killing any children in process group … [ Dec 11 07:04:11 is there a repository or place to upload repo manifest configs for the various ported devices? Dec 11 07:06:17 not really, but JBBgameich have been collecting manifests to his github Dec 11 07:06:36 maybe it should be forked/moved to halium Dec 11 09:30:30 @anpok, increase logbuf. you can get it to 2M iirc Dec 11 16:53:01 @bshah, I could move it into the Halium org right now, should I? Dec 11 16:53:13 please Dec 11 16:53:16 :) Dec 11 16:54:53 It's here now: https://github.com/Halium/halium-devices. Unfortunately, I can't rename it to just "devices" anymore because I have not enough rights in the new repository Dec 11 16:55:49 It's fine Dec 11 19:43:25 nice **** ENDING LOGGING AT Tue Dec 12 03:00:01 2017 **** BEGIN LOGGING AT Tue Dec 12 03:00:01 2017 Dec 12 19:14:32 Guys, i've got a problem. … Seems that fixupmountpoints script doesn't work Dec 12 19:14:35 + mount /dev/block/bootdevice/by-name/userdata /data … mount: mounting /dev/block/bootdevice/by-name/userdata on /data failed: No such file or directory … + mount /data/rootfs.img /target … mount: mounting /data/rootfs.img on /target failed: No such file or directory … + mkdir -p /target/data … + mount --bind /data/ /target/data … mount: mounting /data/ on /target/data failed: Invalid argument … + mount … rootfs on / type rootfs (rw) Dec 12 19:14:36 devpts (rw,relatime,mode=600) … sysfs on /sys type sysfs (rw,relatime) … proc on /proc type proc (rw,relatime) … + DBG_REASON= Dec 12 19:43:07 Everytime i compile halium i get this: Dec 12 19:43:09 halium/hybris-boot/Android.mk:67: ***************** /boot appears to live on /dev/block/bootdevice/by-name/boot … halium/hybris-boot/Android.mk:68: ***************** /data appears to live on /dev/block/bootdevice/by-name/userdata Dec 12 19:43:24 Good Dec 12 19:46:58 It isn't correct Dec 12 19:47:16 In that case, your device probably isn't in fixup-mountpoints Dec 12 19:53:08 yes … "cancro") … sed -i \ … -e 's block/platform/msm_sdcc.1/by-name/DDR mmcblk0p4 ' \ … -e 's block/platform/msm_sdcc.1/by-name/aboot mmcblk0p7 ' \ … -e 's block/platform/msm_sdcc.1/by-name/bk1 mmcblk0p8 ' \ … -e 's block/platform/msm_sdcc.1/by-name/bk2 mmcblk0p11 ' \ … -e 's block/platform/msm_sdcc.1/by-name/bk3 mmcblk0p15 ' \ … -e 's block/platform/msm_s Dec 12 19:53:09 mmcblk0p17 ' \ … -e 's block/platform/msm_sdcc.1/by-name/bk5 mmcblk0p18 ' \ … -e 's block/platform/msm_sdcc.1/by-name/boot mmcblk0p19 ' \ … -e 's block/platform/msm_sdcc.1/by-name/cache mmcblk0p24 ' \ … -e 's block/platform/msm_sdcc.1/by-name/dbi mmcblk0p6 ' \ … -e 's block/platform/msm_sdcc.1/by-name/fsc mmcblk0p14 ' \ … -e 's block/platform/msm_sdcc.1/by-name/fsg mmcblk0p16 ' Dec 12 19:53:10 block/platform/msm_sdcc.1/by-name/logo mmcblk0p10 ' \ … -e 's block/platform/msm_sdcc.1/by-name/misc mmcblk0p9 ' \ … -e 's block/platform/msm_sdcc.1/by-name/modem mmcblk0p22 ' \ … -e 's block/platform/msm_sdcc.1/by-name/modemst1 mmcblk0p12 ' \ … -e 's block/platform/msm_sdcc.1/by-name/modemst2 mmcblk0p13 ' \ … -e 's block/platform/msm_sdcc.1/by-name/persist mmcblk0p21 ' \ … -e Dec 12 19:53:11 block/platform/msm_sdcc.1/by-name/recovery mmcblk0p20 ' \ … -e 's block/platform/msm_sdcc.1/by-name/rpm mmcblk0p2 ' \ … -e 's block/platform/msm_sdcc.1/by-name/sbl1 mmcblk0p1 ' \ … -e 's block/platform/msm_sdcc.1/by-name/ssd mmcblk0p5 ' \ … -e 's block/platform/msm_sdcc.1/by-name/system mmcblk0p23 ' \ … -e 's block/platform/msm_sdcc.1/by-name/tz mmcblk0p3 ' \ … -e 's Dec 12 19:53:12 block/platform/msm_sdcc.1/by-name/userdata mmcblk0p25 ' \ … "$@" … ;; Dec 12 19:53:51 cancro is a dst :B Dec 12 19:54:10 don't google it Dec 12 19:54:28 I haven't understand, sorry Dec 12 19:54:37 I mean, it should be hard to have a device named "cancro". Dec 12 19:55:00 ahahhah, but it is its real name Dec 12 19:55:02 Because everytime you google it, you'll be exposed to some weird pictures Dec 12 19:55:13 sad bro Dec 12 20:03:31 @jersobh, shit Dec 12 20:03:37 why have i ddgoed it Dec 12 20:03:43 i regret it Dec 12 20:08:46 I told you not to google it Dec 12 20:09:00 Why the heck would a device be named as a dst? Dec 12 20:09:11 I mean, std Dec 12 20:35:22 Can someone help me? Dec 12 20:40:27 Boh Dec 12 21:10:26 @Mattia, Please, someone Can help me? Dec 12 21:12:12 Failed to determine wether /sys/fs/cgroup is a mount point: Too many levels of symbolic links <- does someone know how to fix this? Dec 12 21:12:15 after that comes: Failed to mount cgroup at /sys/fs/cgroup/systemd: No such file or directory... Freezing execution Dec 12 21:55:35 @Mattia, have you ran breakfast codename after making changes to fixup mountpoints? Dec 12 22:00:19 @vanyasem, Yes, but i have the same problem Dec 12 22:00:30 after running breakfast? Dec 12 22:00:44 are you sure that your devices codename is really what you specified? Dec 12 22:00:50 double check that Dec 12 22:01:06 are you sure that is the only entry with that name in fixup mountpoints? Dec 12 22:01:07 Mi3 codename is "cancro" Dec 12 22:01:15 Yes Dec 12 22:01:16 are you sure your tree is up to date? Dec 12 22:02:11 I have downloaded all repositories One week ago Dec 12 22:02:29 have you tried cleaning your build? Dec 12 22:03:02 `make clean` Dec 12 22:04:41 I do this command every time i compile it **** ENDING LOGGING AT Wed Dec 13 03:00:01 2017 **** BEGIN LOGGING AT Wed Dec 13 03:00:03 2017 Dec 13 03:03:36 щр Dec 13 03:03:37 oh Dec 13 03:03:48 i the devices fstab right? Dec 13 03:04:00 i suggest you to check it as well Dec 13 03:04:07 it was wrong for me on PME Dec 13 05:46:37 What is pme? Dec 13 06:14:54 my device's codename Dec 13 06:15:00 doesn't really matter in that case Dec 13 06:15:17 just take a look at fstab file and make sure entries there match the ones in your fixup-mounpoints Dec 13 09:59:34 Ok, when i'll get home i'll check. Thanks Dec 13 12:23:01 ivan Dec 13 12:23:45 Have you examined my mistake? Dec 13 12:26:11 oh I totally forgot about you, been busy this week. sorry, I will definitely do so @sopernam ! Dec 13 12:36:06 @sopernam, I have pushed all branches related to kitakami/sumire to ubports .. and had halium tests running successfully a few days ago Dec 13 12:39:14 thus also the fix mountpoint changes.. and changes to the system startup scripts inside the android services container.. Dec 13 12:43:03 @anpok, Are you telling me to rebuild the kernel? Dec 13 13:58:19 nope Dec 13 14:00:05 there were bunch of other changes in the android_device_* repositories that you might have already applied or not .. Dec 13 14:00:21 not sure where you are in the process.. Dec 13 15:19:26 @vanyasem Here you are the fstab which i have modified Dec 13 15:19:30 (Document) https://irc.ubports.com/o3ddI13G/file_2851.qcom Dec 13 15:23:18 is it correct? Dec 13 15:23:35 No Dec 13 15:37:58 @vanyasem, I think😂 Dec 13 15:52:40 @Mattia, where we use this file ? Dec 13 15:53:04 During mount of partitions Dec 13 15:53:26 @vanyasem … I've resolved the problem but now i got this: … + mkdir /data … + mkdir /target … + mount /dev/mmcblk0p25 /data … + mount --bind /data//.stowaways/sailfishos /target … mount: mounting /data//.stowaways/sailfishos on /target failed: No such file or directory … + mkdir /target/data … + mount --bind /data/ /target/data … + mount … rootfs on / type rootfs (rw) … devpts on /dev/pts type devpts (rw,relatime,mode=600) … s Dec 13 15:53:27 (rw,relatime) … proc on /proc type proc (rw,relatime) … /dev/mmcblk0p25 on /data type ext4 (rw,relatime,data=ordered) … /dev/mmcblk0p25 on /target/data type ext4 (rw,relatime,data=ordered) … double slash after data and .stowaways doesn't exist Dec 13 15:54:26 @Mattia, how Dec 13 15:54:43 @sopernam, Same old Fixup Mount points Dec 13 15:56:17 @sopernam, Not sure if I told you but I figured out what caused the delay in partition detection. Solved it by disabling the feature instead of fixing the driver. Dec 13 15:57:18 @Mattia, ls -la /dev/block/bootdevice/by-name Dec 13 15:57:56 @anpok, how can i disable it ? Dec 13 15:58:50 https://github.com/ubports/android_kernel_sony_msm8994-1/commit/63630285694687b2566a9b52a78198bcf0ae6d88 Dec 13 16:01:12 @Mattia, Are you sure you are using right hybris-boot repo? Dec 13 16:01:28 @bhushanshah, +1 on this question Dec 13 16:01:32 why does it say Sailfish os Dec 13 16:01:59 @anpok, thank you i will test it when i go home Dec 13 16:02:02 ☺️ Dec 13 16:02:22 https://github.com/mer-hybris/hybris-boot … I'm using this Dec 13 16:02:52 That's wrong Dec 13 16:02:58 For halium use halium fork Dec 13 16:03:21 GitHub.com/Halium/hybris-boot Dec 13 16:03:52 OMG Thanks everyone! Dec 13 16:04:47 Why were you using the sailfish version? When you clone the halium tree using repo, you get the correct one already Dec 13 16:08:55 @JBBgameich, I don't know why... What a fail Dec 13 16:41:29 Now it can't found rootfs.img … + mount /dev/mmcblk0p25 /data … + mount /data/rootfs.img /target … mount: mounting /data/rootfs.img on /target failed: No such file or directory … + mkdir -p /target/data … + mount --bind /data/ /target/data … + mount Dec 13 16:43:32 If i try this? … https://github.com/JBBgameich/halium-install Dec 13 17:04:29 Have you already installed a rootfs? The documentation contains a list to the official installation script. I wrote mine because the official one didn't work on my device, but it's probably a good Idea to try the official one first. Dec 13 17:05:02 https://github.com/halium/halium-scripts Dec 13 17:10:26 With default install script i can't see rootfs.img in /data … with your script, during the adb push, it stop, everytime, at 75% more or less Dec 13 17:11:08 can you fsck /data? Dec 13 17:11:30 I remember this being a problem when using adb push Dec 13 17:12:15 ok, 1 sec Dec 13 17:14:08 I havn't got that command (adb, telnet) Dec 13 17:18:41 oh wait .. maybe the command is called different Dec 13 17:19:07 `e2fsck` Dec 13 17:20:32 @JBBgameich, ~ # e2fsck /data … e2fsck 1.42.9 (28-Dec-2013) … e2fsck: Is a directory while trying to open /data … The superblock could not be read or does not describe a correct ext2 … filesystem. If the device is valid and it really contains an ext2 … filesystem (and not swap or ufs or something else), then the superblock … is corrupt, and you might try running e2fsck with an alternate superblock: … e2fsck -b 8193 Dec 13 17:20:58 `e2fsck /dev/mmcblk0p25` Dec 13 17:21:51 @JBBgameich, ~ # e2fsck /dev/block/mmcblk0p25 … e2fsck 1.42.9 (28-Dec-2013) … data: recovering journal … data has been mounted 22 times without being checked, check forced. … Pass 1: Checking inodes, blocks, and sizes … Pass 2: Checking directory structure … Pass 3: Checking directory connectivity … Pass 4: Checking reference counts … Pass 5: Checking group summary information … data: 29917/835584 files (6.4% non-contiguous), 1379346 Dec 13 17:21:55 Done Dec 13 17:22:07 now try installing again Dec 13 17:26:54 No, crashing continously at 50+% Dec 13 17:27:41 rootfs.img is 2.1Gb is it Normal? Dec 13 17:28:18 I have one other Idea ... can you comment out the flash and clean step in the script and copy the created rootfs.img and system.img to the device manually? Dec 13 17:28:42 Some TWRP builds are really strange ... Dec 13 17:29:11 How can i copy that file manually to the phone? I have got halium installed (telnet) Dec 13 17:29:58 Copy it to an sdcard and move them to /data in recovery for example Dec 13 17:30:37 Ok i'll try. Thanks! Dec 13 17:58:35 https://pastebin.com/Y1NWASdi <- systemd fails to mount cgroup, no kernel panic though it just hangs Dec 13 17:59:53 Check hammerhead kernel on halium GitHub, it have a patch from torvalds cherry-picked.. Dec 13 18:00:16 You need to cherry-pick it Dec 13 18:00:23 ill have a look at the hammerhead kernel Dec 13 18:00:35 Which kernel version? Dec 13 18:01:31 3.4.0 (nexus 4 - mako) Dec 13 18:01:50 Yeah it'll need that patch Dec 13 18:03:00 https://github.com/Halium/android_kernel_lge_hammerhead/commit/25437b2a54dd619a96e268ecaf303b089aa785e4 Dec 13 18:04:21 thanks Dec 13 18:37:22 just got banned on this channel (IRC) because of joining 3 times? i don't get it Dec 13 18:38:32 kllp: possibly you got redirected to ##halium-unregistered Dec 13 18:38:54 #halium: Cannot join channel (+b) - you are banned Dec 13 18:39:07 Uhh Dec 13 18:39:37 I'll try to unban you tomorrow if it's fine.. I am on mobile right now Dec 13 18:39:57 yeah thanks Dec 13 18:44:08 @JBBgameich Script finally worked but now i got this: … + mkdir /data … + mkdir /target … + mount /dev/mmcblk0p25 /data … + mount /data/rootfs.img /target … mount: mounting /dev/block/loop0 on /target failed: Invalid argument … + mkdir -p /target/data … + mount --bind /data/ /target/data … + mount Dec 13 19:03:48 Can you mount /data/rootfs.img in recovery? Dec 13 19:06:54 No "invalid argument" Dec 13 19:10:56 What does file /data/rootfs.img say? Dec 13 19:16:21 In which sense? What command i have to do? Dec 13 19:17:19 `file /data/rootfs.img` in recovery mode Dec 13 19:17:38 file command doesn't exist Dec 13 22:06:53 hi Dec 13 22:07:35 (Photo, 1280x720) https://irc.ubports.com/1B0nC6zu/file_2864.jpg @anpok ehy im gettin this error Dec 13 22:13:19 @sopernam, you made changes to fixup-mountpoints but there are also changes in Github Dec 13 22:13:35 its a conflict resoultion you need now Dec 13 22:13:49 either your stash your changes or commit them Dec 13 22:14:08 @Mattia, @Flohack What do you think of this? Dec 13 22:15:37 @Mattia, No idea sry ^^ Dec 13 22:16:02 @Flohack, It's ok thanks Dec 13 22:16:41 @Flohack, i deleted it then problem solved Dec 13 22:16:49 yeah or this Dec 13 22:17:14 how can i fix it other vay ? Dec 13 22:18:33 @sopernam, before you start work, make git pull to always stay in sync with Github Dec 13 22:18:49 commit your stuff often Dec 13 22:19:11 But you will eventually have conflicts, and you need to learn how to resolve. Sometimes people change the same stuff in a file Dec 13 22:19:47 ty Dec 13 22:24:31 do I have to rebuild the system after the change in kernel configuration? Dec 13 22:33:59 @sopernam, if you change the kernel config also look to remove the kernel from the out directory otherwise your changes might not get updated Dec 13 22:37:27 @Flohack, is it enough to remove hybris-boot.img with boot.img for this operation? Dec 13 22:37:47 No actually you need to remove the kernel config file just a sec Dec 13 22:38:47 Its in $OUT/obj/KERNEL_OBJ/.config Dec 13 22:38:53 maybe also remo .config.old dunno Dec 13 22:39:09 @Flohack, thank you Dec 13 22:41:00 @Flohack, already renamed .config.old Dec 13 22:46:06 fsociety@buildbot ~ $ md5sum .config.old … 456986ff7f4276cf9d3952fd37a1a1a8 .config.old … fsociety@buildbot ~ $ md5sum config … 456986ff7f4276cf9d3952fd37a1a1a8 config … its same file Dec 13 23:14:54 @anpok im stuck in brand logo Dec 13 23:18:27 iSerial 3 GNU/Linux Device on usb0 10.15.19.82 <- finally succes \o/ Dec 13 23:27:03 halium is alive on Nexus 4 (MAKO), still need to fix some things though :P Dec 13 23:28:07 yes its working 🤓 Dec 13 23:28:42 i tried fastboot boot hybris-boot.img and its working Dec 13 23:34:11 vibration / lights are working the rest i didn't test for now, time to sleep Dec 13 23:39:22 (Photo, 1280x960) https://irc.ubports.com/LLEr5ViS/file_2866.jpg is it worked ? Dec 13 23:42:24 how can i test it ? Dec 13 23:43:38 lsusb -vd 0fce: | grep -i serial Dec 13 23:44:15 if it outputs: "iSerial 3 GNU/Linux Device on usb0 10.15.19.82" it booted atleast Dec 13 23:44:52 Right Dec 13 23:48:25 iSerial 3 GNU/Linux Device on rndis0 10.15.19.82 Dec 13 23:49:00 Grats. Move on in the docs. Dec 13 23:49:12 ☺️ Dec 13 23:49:42 https://docs.halium.org/en/latest/porting/debug-build/logging-in.html Dec 13 23:50:20 Ok time too feed my child and sleep, Goodnight everyone Dec 13 23:56:46 goodnight Dec 14 00:03:45 ~ # chroot /a /bin/bash … chroot: can't execute '/bin/bash': No such file or directory **** ENDING LOGGING AT Thu Dec 14 03:00:00 2017 **** BEGIN LOGGING AT Thu Dec 14 03:00:01 2017 Dec 14 05:11:37 im tried /sbin/sh /sbin/ash /system/bin/sh /system/xbin/sh but im getting same error Dec 14 05:19:07 kllp[m]: hey, I don't see you banned here.. can you try to re-join so I can see what's happening? Dec 14 10:00:04 bshah: thank you, it worked Dec 14 10:00:15 I did nothing :'D Dec 14 10:00:33 :) Dec 14 10:00:55 kllp: as said previously, I think given requirement of having identified here, I am sure you must've been redirected to #halium-unregistered Dec 14 10:01:11 how can i contribute to halium without owning a device to test with? Dec 14 10:03:54 @sopernam, :( Dec 14 10:43:19 hey is there anybody ? Dec 14 11:26:20 yep? Dec 14 11:26:50 @sopernam, ? Dec 14 11:29:28 What exact commands did you run? Dec 14 11:29:42 chroot /a /bin/bash Dec 14 11:29:48 well... Dec 14 11:29:50 before that? Dec 14 11:30:00 chroot: can't execute '/bin/bash': No such file or directory Dec 14 11:30:23 before doing chroot what command did you run? Dec 14 11:30:31 did you mount rootfs.img? Dec 14 11:30:34 what was output of it? Dec 14 11:30:41 mkdir /a … mount /data/system.img /a Dec 14 11:30:46 @bhushanshah, no Dec 14 11:30:53 well... Dec 14 11:31:29 well, if you are using halium-install script from halium repo, you need to follow instructions for halium and not ubports Dec 14 11:31:44 which is to mount rootfs.img Dec 14 11:32:05 @bhushanshah, https://docs.halium.org/en/latest/porting/debug-build/logging-in.html Dec 14 11:32:25 @bhushanshah, im not mount rootfs.img ? Dec 14 11:32:59 that instructions needs updating Dec 14 11:33:05 mount /data/rootfs.img Dec 14 11:33:14 and not /data/system.img Dec 14 11:33:23 ty :) Dec 14 11:52:32 (Photo, 1280x719) https://irc.ubports.com/VOIw6ITF/file_2872.jpg im getting this error. cable unpullugged Dec 14 11:52:56 ? Dec 14 11:53:03 telnet Dec 14 11:53:14 ? Dec 14 11:54:10 ip link set USBNETWORK address 02:01:02:03:04:08 … Cannot find device "USBNETWORK" Dec 14 11:55:18 (Photo, 735x484) https://irc.ubports.com/HiKGk26j/file_2874.jpg Dec 14 11:55:22 stuck Dec 14 11:55:36 Do you even read instructions? Dec 14 11:55:37 :/ Dec 14 11:55:42 2nd point Dec 14 11:56:01 See first and 2nd point in that doc Dec 14 11:57:27 iSerial 3 GNU/Linux Device on rndis0 10.15.19.82 … iSerial 1 0000:00:1d.0 … iSerial 3 20100201396000000 … iSerial 1 0000:00:1a.0 Dec 14 11:59:53 USBNETWORK needs to be replaced with the device name which you can find using dmesg Dec 14 12:22:21 @revertive, how ? Dec 14 12:22:29 cat /var/lib/lxc/android/rootfs/ueventd*.rc|grep ^/dev|sed -e 's/^\/dev\///'|awk '{printf "ACTION==\"add\", KERNEL==\"%s\", OWNER=\"%s\", GROUP=\"%s\", MODE=\"%s\"\n",$1,$3,$4,$2}' | sed -e 's/\r//' >/etc/udev/rules.d/70-[codename].rules ?? Dec 14 12:24:41 replace [codename] with your device name Dec 14 12:25:50 "dmesg |grep rndis_host" to see the device name it problably start with enp0......... Dec 14 12:26:08 which device are you trying to boot? **** BEGIN LOGGING AT Thu Dec 14 12:51:56 2017 Dec 14 12:52:49 @revertive, sumire Dec 14 12:57:22 @revertive, enp0s29u1u3 Dec 14 13:00:58 @Mattia, @bhushanshah Can you help me? Have you got some ideas? Dec 14 13:01:48 Bit busy with $work at moment Dec 14 13:01:53 Will try to help later Dec 14 13:02:12 Ok, i'll wait. Thanks Dec 14 13:03:57 @sopernam, replace [codename] with sumire, and USBNETWORK with enp0s29u1u3 Dec 14 13:05:32 @revertive, i have no idea for how to do this ? Dec 14 13:07:14 USBNETWORK with enp0s29u1u3 ? Dec 14 13:09:15 ip address add 10.15.19.1 dev enp0s29u1u3 && ip route add 10.15.19.82 dev enp0s29u1u3 Dec 14 13:09:38 ssh root@10.15.19.82 Dec 14 13:10:13 yhank you so much Dec 14 13:11:13 your welcome Dec 14 13:16:18 (Document) https://irc.ubports.com/aqDxoA2y/file_2875 Dec 14 13:55:21 it worked fine inside the halium rootfs Dec 14 15:00:57 Nothing... … + mount -o loop -t ext4 /data/rootfs.img /target … mount: mounting /dev/block/loop0 on /target failed: Invalid argument Dec 14 15:02:19 Try `-t ext2`? Dec 14 15:03:01 It's generally read-only, so there's no need for the journal. Dec 14 15:03:38 ~ # mount -o loop -t ext2 data/rootfs.img target/ … mount: mounting /dev/block/loop0 on target/ failed: Invalid argument … ~ # Dec 14 15:04:32 @anpok, ty so much ☺️ Dec 14 15:18:44 mount /data/rootfs.img /target? worked for me Dec 14 15:30:48 hm now with the same device most recent configuration .. yesterdays build of a car rootfs using @UniversalSuperBox rootstock script from the xenial-caf-7.1 branch and the rootstock install script from the other xenial branch .. the sshd is up ( at least nmap thinks so) but it does not want me to log in with phablet - it just hangs Dec 14 15:31:49 ssh -vvvv Dec 14 15:32:16 `ssh --if-you-so-much-as-breathe-print-to-stderr` Dec 14 15:32:45 OpenSSH_7.5p1 Ubuntu-10, OpenSSL 1.0.2g 1 Mar 2016 … debug1: Reading configuration data /home/andreas/.ssh/config … debug1: /home/andreas/.ssh/config line 6: Applying options for * … debug1: Reading configuration data /etc/ssh/ssh_config … debug1: /etc/ssh/ssh_config line 19: Applying options for * … debug2: resolving "10.15.19.82" port 22 … debug2: ssh_connect_direct: needpriv 0 … debug1: Connecting to 10.15.19.82 [10.15.19.82] port 22. Dec 14 15:32:46 established. … debug1: identity file /home/andreas/.ssh/id_rsa type 1 … debug1: key_load_public: No such file or directory … debug1: identity file /home/andreas/.ssh/id_rsa-cert type -1 … debug1: key_load_public: No such file or directory … debug1: identity file /home/andreas/.ssh/id_dsa type -1 … debug1: key_load_public: No such file or directory … debug1: identity file /home/andreas/.ssh/id_dsa-cert type -1 … debug1: key_load_public: No such Dec 14 15:32:48 identity file /home/andreas/.ssh/id_ecdsa type -1 … debug1: key_load_public: No such file or directory … debug1: identity file /home/andreas/.ssh/id_ecdsa-cert type -1 … debug1: key_load_public: No such file or directory … debug1: identity file /home/andreas/.ssh/id_ed25519 type -1 … debug1: key_load_public: No such file or directory … debug1: identity file /home/andreas/.ssh/id_ed25519-cert type -1 … debug1: Enabling compatibility mode for proto Dec 14 15:32:49 version string SSH-2.0-OpenSSH_7.5p1 Ubuntu-1 Dec 14 15:32:58 RIP IRC. Dec 14 15:33:41 Hmm Dec 14 15:33:48 Doesn't tell what's wrong Dec 14 15:34:18 @anpok, So, you never get the password prompt? Dec 14 15:34:36 yes Dec 14 15:37:29 Is your network reconnecting? Dec 14 15:39:44 not really .. nm just always wanted to run dhcp client but after I reconfigured it with a fixed address and set the routes.. Dec 14 15:39:52 it stayed like that since yestersay Dec 14 15:43:23 guess I will try again later and halt it during the init script in the initramfs Dec 14 16:00:52 Dmesg said this: I found this: <4>[ 5.976374] EXT4-fs (mmcblk0p25): warning: maximal mount count reached, running e2fsck is recommended … <6>[ 5.977437] EXT4-fs (mmcblk0p25): mounted filesystem with ordered data mode. Opts: (null) … <3>[ 6.006668] EXT4-fs (loop0): couldn't mount RDWR because of unsupported optional features (400) Dec 14 16:27:55 different topic .. sony Xperia X ... the one supported by sailfish .. are there open bits that we could use to have an easy hallium port? Dec 14 16:31:36 I guess you can use the LineageOS or Sony AOSP device sources for that device just fine. It's a little bit easier to port because sailfishOS already added the fixup mountpoints for hybris-boot. We already have them in the Halium version Dec 14 18:01:32 Ok i used an older version of e2fsck and now works but i got this in diagnosis.log: … ~ # cat diagnosis.log … CONFIG_AUTOFS4_FS=y not found in /proc/config.gz … CONFIG_DEVTMPFS_MOUNT=y not found in /proc/config.gz … CONFIG_DEVTMPFS=y not found in /proc/config.gz … CONFIG_VT=y not found in /proc/config.gz Dec 14 18:06:26 enable those in your kernel config file Dec 14 18:06:47 /arch/arm/configs/ Dec 14 18:39:31 library "libis_compat_layer.so" not found <- test_input Dec 14 19:00:14 @revertive> /arch/arm/configs/ What are you building after changing the defconfig? Dec 14 19:01:41 @UniversalSuperBox, hybris-boot and systemimage after a clean Dec 14 19:06:15 Do I need to have a phone to tinker with in order to help halium progress? Dec 14 19:08:37 You can probably help with documentation e.g. Dec 14 20:57:54 Resolved but now i can't connect to phone with telnet or ssh... Any solution? Dec 14 21:08:23 Mattia: sparati **** ENDING LOGGING AT Fri Dec 15 03:00:02 2017 **** BEGIN LOGGING AT Fri Dec 15 03:00:02 2017 Dec 15 06:12:55 gouster4 was added by: gouster4 Dec 15 06:14:17 Hi. Where i can download sources? Dec 15 06:18:16 @gouster4, hello. Sources for what( Dec 15 06:19:55 for Halium. I want to try build it for my Xperia XZ Dec 15 06:20:35 GitHub.com/halium Dec 15 06:20:47 but please read docs at Halium.org first Dec 15 06:21:28 Thanx. Dec 15 10:59:30 @JBBgameich when you get time can you go through packages added in the Dec 15 11:00:04 I am not good Debian packager.. so might be it's crappy job done Dec 15 11:31:23 root@localhost:~# test_glesv2 … Segmentation fault Dec 15 11:31:29 root@localhost:~# test_egl … Segmentation fault Dec 15 11:54:09 these tests are broken Dec 15 11:54:14 they are supposed to fail Dec 15 11:55:45 does hwcomposer work Dec 15 12:34:39 @bhushanshah, After a quick look, everything seems fine there! The only thing you could do is updating to Standards-Version 4.1.1, but as it is not yet the default in dh_make, it's not a problem. Dec 15 13:17:32 @JBBgameich welp Dec 15 13:18:45 @JBBgameich I've a debian package, and I've some patches... and somehow patchces are not getting applied at all.. wut Dec 15 13:23:15 dpkg-buildpackage should do everything if … * the patch is in the debian/patches/series file … * and quilt is installed Dec 15 13:23:40 hm.. Dec 15 13:23:50 I guess something messed up with local setup Dec 15 13:23:56 let me try to build on CI and see Dec 15 13:24:10 but sometimes removing the .pc folder inside the package helps Dec 15 13:24:35 It's probably not recommended to do it but I use this "fix" often too Dec 15 13:24:47 ah ofcourse Dec 15 13:24:49 hmm Dec 15 13:31:17 this is going to fail badly on CI as chroot was not clean :D Dec 15 13:31:23 but let's try neverthless Dec 15 13:32:58 What are you using on the CI? pbuilder or something else? Dec 15 13:33:54 @bhushanshah, debian packaging is hell, isn't it? Dec 15 13:34:05 custom stack that powers nci, but essentially dpkg-buildpackage in docker Dec 15 13:36:53 `Undefined subroutine &main::catfile called at /usr/bin/uscan line 3783.` Dec 15 13:36:54 wtf Dec 15 13:37:00 @JBBgameich any idea? Dec 15 13:37:22 what are you using uscan for at all? Dec 15 13:37:42 to fetch orig tarball Dec 15 13:37:55 (it's not me using but CI using) Dec 15 13:38:32 ah ... that's good, but maybe the watch file is wrong somehow or your uscan version has a bug... Dec 15 13:39:19 see latest package I pushed in halium-packaging Dec 15 13:44:07 I'll try building it locally Dec 15 13:45:44 Works completely fine ... Dec 15 13:47:28 Hmm Dec 15 13:47:36 What uscan version? Dec 15 13:47:58 Can you try xenial pbuilder/chroot Dec 15 13:49:33 @vanyasem, what sould i do? Dec 15 13:49:55 wait for them to get fixed? Dec 15 13:50:10 there is nothing you can really do :P ubless you know how to code Dec 15 13:51:54 @vanyasem, 😂 Dec 15 13:51:59 @vanyasem, eta ? Dec 15 13:55:56 @bhushanshah, But I can't test uscan in there, right? Dec 15 13:56:13 @bhushanshah, 2.17.11 Dec 15 14:18:08 (Sticker, 512x489) https://irc.ubports.com/EuS0RIIg/file_2883 Dec 15 14:42:20 @vanyasem, 😂 Dec 15 14:54:38 how git sources work with Debian packages? Dec 15 15:07:17 (Sticker, 512x512) https://irc.ubports.com/iIN6AyQL/file_2884.webp Dec 15 15:12:36 what do you mean? The source is pulled as tarball using uscan for this package Dec 15 15:16:29 @JBBgameich so it's like "use uscan to download source", then build as usual? Seems to be different from https://wiki.debian.org/PackagingWithGit#Upstream_import_methods Dec 15 15:19:19 https://github.com/halium-packaging/libhybris/tree/halium-7.1/ doesn't have debian/watch file though Dec 15 15:20:20 Because upsteam doesn't release the source tarball Dec 15 15:23:22 hm, so the magic is done in https://github.com/halium-packaging/libhybris/blob/b73f165ba14189360fb7ec99722a913af19c0392/debian/rules#L69 Dec 15 15:25:00 If you manually want to do it yes Dec 15 15:25:20 Otherwise currently CI builds orig tarball Dec 15 15:26:42 @vanyasem, дааа... Dec 15 15:44:46 @NotKit, You can use gbp, uscan, or just download the tarball yourself. Dec 15 15:45:36 In the end, you always get a dsc source package which is what debian uses for it's own build servers. You don't have to maintain the package in git Dec 15 17:03:34 @NotKit, сегодня пончики с дамой купил, очень вкусно Dec 15 17:37:43 E Mi Sono lnnamorato ma di Tuommarito Dec 15 17:45:03 cristo: any reason you are spamming here? Dec 15 17:57:34 @sopernam, @anpok Dec 15 18:24:51 Bhushan Shah: im not spamming here, why? Dec 15 18:25:11 > E Mi Sono lnnamorato ma di Tuommarito Dec 15 18:25:11 ^ Dec 15 18:25:23 Thats not spam Dec 15 18:25:32 ? Dec 15 18:25:42 ? Dec 15 18:26:27 Well it's totally unrelated sentence that o don't understand, so explain me what's it then? Dec 15 18:27:07 It means that i got some stuff on my device... Dec 15 18:28:22 Can you please communicate in English next time? Dec 15 18:28:30 Uh Dec 15 19:52:52 strange now I have the same problems like a few weeks ago Dec 15 19:53:17 the kernel parameters are not the way I expect them Dec 15 19:55:00 as if the kernel replaces the parameters passed to mkbootimage with other parameters.. Dec 15 20:02:40 Some bootloaders do Dec 15 20:08:37 i noticed that as well Dec 15 20:08:45 but it turned out to be a non-issue for me Dec 15 20:09:01 i.e. something else was broken, not the kernel Dec 15 20:11:38 should i wait? Dec 15 20:11:49 wait for what? Dec 15 20:12:05 please ask questions that make sense Dec 15 20:12:09 @sopernam, . Dec 15 20:12:26 well it's broken Dec 15 20:12:50 what do you want to do instead of waiting? Dec 15 20:13:14 what am I waiting for? Dec 15 20:13:41 you are waiting in line for a new pairs of adidas shoes? Dec 15 20:14:07 i am sorry for such an answer, but the question is pretty silly to me Dec 15 20:14:21 you are waiting for the tests to get fixed in halium 7.1 Dec 15 20:15:24 is the halium 7.1 unstable? Dec 15 20:15:45 halium is in developement and not ready for daily usage overall. Dec 15 20:16:05 so all halium trees are pretty much unstalbe more or less, even on reference devices Dec 15 20:16:10 it's work in progress Dec 15 20:16:47 thank you for answers. 😊 Dec 15 20:21:54 i am pretty sure it's in the docs Dec 15 21:22:05 retro486 was added by: retro486 Dec 15 21:27:03 Hello! I'm trying to follow the halium docs on my nexus 4 with los 14.1. I've gotten as far as getting telnet to work, but the step with chroot after booting into recovery complains that bash isn't found. Same message if I try variations of /bin and /sbin and sh instead of bash... I'm using TWRP; does chroot assume a different recovery environment? Dec 15 21:34:51 Actually, re-reading the steps is that a typo? Shouldn't I be chrooting into rootfs.img not system.img? Dec 15 21:38:09 @retro486, `chroot /path/to/chroot /bin/bash` Dec 15 21:38:22 ofc it is not found, as you have not specified it's path Dec 15 21:38:28 you need to do so explicitly Dec 15 21:39:20 it's not that much related to Halium, it's just basics of working with chroot. I faced that issue myself and found a solution on the internet in a few minutes Dec 15 21:40:38 Right, and I did and found that answer as well but /bin/bash doesn't exist, nor any variant of /bin, /sbin, or using sh instead of bash... Does your TWRP have bash in /bin/bash? That was my original question before I asked about using rootfs vs system Dec 15 21:41:10 it's not related to your recovery Dec 15 21:41:21 you specify a path to bash INSIDE your chroot Dec 15 21:41:53 if you don't have bin/bash in your chroot, then it's corrupt Dec 15 21:42:02 Ah Dec 15 21:42:09 @retro486, and it shouldn't Dec 15 21:42:14 because system.img is an Android image Dec 15 21:42:22 you need to chroot into rootfs.img Dec 15 21:42:39 Ok, that much works for me. The docs say to use system.img, hence my confusion Dec 15 21:42:54 which docs? Dec 15 21:43:02 can you send a screenshot? Dec 15 21:43:09 https://docs.halium.org/en/latest/porting/debug-build/logging-in.html Dec 15 21:43:21 First code quote Dec 15 21:43:56 just a sec, i need to find a person to blame for that Dec 15 21:44:08 Hah 👍 Dec 15 21:44:11 gotcha Dec 15 21:44:22 @UniversalSuperBox you made Halium docs incorrect Dec 15 21:44:49 i guess you just copied docs from installing UBports 16.04 and forgot to change images' names Dec 15 21:45:07 UBports uses system.img, but halium uses rootfs.img Dec 15 21:45:22 I didn't know that Dec 15 21:45:33 well i blame you for that :P Dec 15 21:45:40 I thought that maybe we could all agree on at least one thing Dec 15 21:45:44 :P Dec 15 21:45:57 @UniversalSuperBox, UBports doesn't follow Halium standarts Dec 15 21:46:00 Great, that means I can change the UBports documentation to say `ubuntu.img` instead. Dec 15 21:46:15 @UniversalSuperBox, ubuntu.img is symlinked to system.img Dec 15 21:46:23 UBports is a big hack around halium Dec 15 21:46:26 Right, but system.img is the old name Dec 15 21:46:31 in fact, android image is inside the rootfs Dec 15 21:46:39 when for halium they are separate Dec 15 21:46:52 system.img for the android system, and rootfs.img for the rootfs image Dec 15 21:46:55 Shouldn't be that hard to change if the opportunity presents itself Dec 15 21:46:59 @vanyasem, makes sense, doesn't it? Dec 15 21:47:26 If you wanted to dig into the initramfs for a while, you could probably do it Dec 15 21:47:46 @UniversalSuperBox, what? Dec 15 21:47:54 that's how it currently works in Halium Dec 15 21:47:55 That's where everything is mounted Dec 15 21:48:02 Right, I mean the UBports one. Dec 15 21:48:28 plz fix the docs Dec 15 21:48:34 people already get confused over it Dec 15 21:48:37 I'm at work, mister Dec 15 21:48:42 But I will Dec 15 21:48:56 I'm still not sure if it wouldn't be easier to fix upstart compatibility with hybris-boot than maintaining the Ubports initrd ... Dec 15 21:49:19 @UniversalSuperBox, no hurry, take your time Dec 15 21:49:23 I could take a lookie there, too Dec 15 21:53:10 No promises, I am a stupid. Dec 16 00:56:24 lxc-start 20171207030306.430 ERROR lxc_utils - utils.c:mkdir_p:254 - Invalid argument - failed to create directory '/sys/fs/cgroup/net_cls//lxc/android' Dec 16 00:57:29 the lxc android container failed to start because it cant access /sys/fs/cgroup/net_cls/lxc/, any idea how i can fix this Dec 16 02:44:46 If my kernel complains "systemd[1]: Failed to determine whether /sys/fs/cgroup is a mount point: Too many levels of symbolic links" will that prevent the reference Halium rootfs from loading? Dec 16 02:45:30 I see a lot of posts talking that it can prevent lxc from functioning; I can't seem to find info on if the halium reference rootfs uses lxc. Dec 16 02:56:35 @mariogrip Is there anything specific, and byte-sized (as I'm not really much of a programmer yet, All I know is a couple things in Python and a bunch of Bash) that I can help with for the OP5? **** ENDING LOGGING AT Sat Dec 16 03:00:01 2017 **** BEGIN LOGGING AT Sat Dec 16 03:00:03 2017 Dec 16 08:40:48 @retro486, it uses LXC to run Android container Dec 16 10:59:02 * kllp[m] sent a long message: kllp[m]_2017-12-16_10:59:01.txt Dec 16 11:00:30 Hybriss was added by: Hybriss Dec 16 11:00:54 kutlay1653 was added by: kutlay1653 Dec 16 11:52:14 "If you’re not sure of your remote, jump down to Remotes." After reading this part I still don't get what remote to choose. Dec 16 11:55:16 if your device tree is on http://github.com/lineageos/ then choose "los" Dec 16 11:59:00 Thank you. May I add this to the docs? Dec 16 12:00:46 There is already a table containing that information in the docs Dec 16 12:09:06 The table is about available remotes, not about which one to choose. Dec 16 12:19:39 @UniversalSuperBox, I first thought it was because I missed CMDLINE_EXTEND .. thus the proposal to the kernel conf check Dec 16 13:07:58 @JBBgameich, Sorry for dumb question, but when is lineage avaible for my device, then is porting of halium simpler? Dec 16 13:26:36 A LineageOS port is nearly required for porting Halium Dec 16 14:13:39 is there a way to build libhybris compat layer with halium-7.1, i can only find it for halium-5.1? Dec 16 14:27:23 why do you want to set them Dec 16 14:27:29 you can just ignore the error Dec 16 14:27:37 Okay, good Dec 16 14:41:35 @revertive, libhybris is built against android headers from Halium 5.1 or 7.1, or do I misunderstand your question somehow? Dec 16 14:41:57 I think he means libcompat Dec 16 14:42:14 test_camera, test_input and test_media give "Library *-compat_layer.so not found, segmentation fault" Dec 16 14:43:19 And no one really uses libcompat any more (to my knowledge), so maybe those tests are broken? Dec 16 14:45:12 oke is guess the tests are broken then, ill see if i can test the hardware another way thanks Dec 16 14:45:27 A real OS is a great tester. :) Dec 16 14:47:41 ill try a readl OS then ;) and see what works Dec 16 14:59:37 @JBBgameich, Is possible to build halium on android 8 (oreo)? Because for XZ is only lineage 15.0 avaible. Dec 16 15:00:10 @revertive, @mariogrip 🙃 Dec 16 15:00:25 @gouster4, Our tree is still lineageOS 14.1 based, so I guess it's not possible if you don't want to update the whole tree Dec 16 15:01:30 (revertive: my last message might come out of context but I pinged mariogrip about libcompat on telegram side) Dec 16 15:03:49 @gouster4, 😏 Dec 16 15:04:47 @gouster4, no Dec 16 15:05:07 Halium 5.1 and Halium 7.1, does that say something? Dec 16 15:09:48 @sopernam, Why? Dec 16 15:09:53 @gouster4, This is the same as saying. "is it possible to build oreo from nuga kernel source?" Dec 16 15:10:04 exactly :) Dec 16 15:10:08 it's nougat though Dec 16 15:10:09 @vanyasem, Yeah, only for lollipop and marschmellow Dec 16 15:10:11 @gouster4, because i have xperia z5. Dec 16 15:11:04 and our phone's development stopped. Dec 16 15:11:28 @sopernam, Because of releasing x series? Dec 16 15:11:34 yes Dec 16 15:12:17 this is absurd. Dec 16 15:12:54 @sopernam, Oh. Sorry for that. Dec 16 15:13:56 @gouster4, I know that feeling. Before XZ i had SE Xperia neo. Dec 16 15:14:22 I'm not really angry with the x series coming out. I'm pissed off by the developers. Dec 16 15:14:57 they are starting to support the latest phones. Dec 16 16:08:53 haha Dec 16 16:09:08 lol gouster4 Dec 16 17:00:10 I used the one linked in the docs: http://bshah.in/halium/halium-rootfs-20170630-151006.tar.gz Dec 16 17:04:18 Hmm the reference rootfs uses systemd and the ubports one upstart Dec 16 17:04:52 The reference rootfs is tested and works Dec 16 17:08:51 Huh. I guess I'm not sure why mine is seemingly hanging once i let the init continue Dec 16 17:09:24 The usb network device is retained, but manually assigning an IP doesn't seem to get ssh working Dec 16 17:09:41 Is the rootfs mounted read only? it doesn't seem to generate any logs either... Dec 16 17:10:11 This is probably getting into generic linux debugging questions now... :) Dec 16 17:10:33 But I'm testing the docs hence why I'm asking here Dec 16 17:11:57 The initial debug init (with telnet on ip 192.168.2.15 and port 23) works fine; just once I "continue" from there it unmounts everything then seemingly hangs when trying to load the rootfs. Dec 16 17:14:50 @retro486 welcome to the cloub Dec 16 17:15:24 So this is a common sticking point? Dec 16 17:32:41 Hello Dec 16 17:32:54 I'm stucked in this; Dec 16 17:33:10 (Photo, 977x963) https://irc.ubports.com/zfdxaVyz/file_2899.jpg Dec 16 17:35:28 It's a file that should exist in your device tree Dec 16 17:35:34 For example mako: https://github.com/LineageOS/android_device_moto_shamu Dec 16 17:35:58 you can see lineage.dependencies there. Your device should have one as well, for lineage os 14.1 anyway Dec 16 17:40:51 Hello. When I try to build the systemimage, I get an error saying that the signapk.jar is missing and that there is not rule to build it (even though there is a folder signapk in build/tools). Dec 16 17:40:51 Does anybody know where I made a mistake? Dec 16 17:41:49 (Photo, 1180x730) https://irc.ubports.com/uLFEDE0T/file_2901.jpg Dec 16 17:42:01 what can I do now? Dec 16 17:44:31 Follow the instructions under "Dependicies". they list an example of how to interpret the repository name. The remote should be either the lineageos (if you didn't fork the repos) or your own remote (if you did fork the repos). Dec 16 17:46:04 @Dorni[m], You need to comment out all java and apk parts in your vendor/manufacturer/codename makefiles Dec 16 17:49:41 @retro486, at least for me :( Dec 16 17:50:50 @JBBgameich: Thank you, just did. Let's see how far we come now Dec 16 18:09:36 @anpok, I don't know about you but I have zero knowledge on systemd so I'm finding this article helpful in understanding, but I'm not yet at a point to tell why the hybris-init isn't loading systemd: https://www.digitalocean.com/community/tutorials/how-to-use-systemctl-to-manage-systemd-services-and-units Dec 16 18:10:36 and of course https://wiki.debian.org/systemd Dec 16 18:20:45 Thank you, this time it built without problems. So, is there any way to create a flashable zip-file for Android recoveries or is manual flashing (and thus replacing everything Android) the only way atm? Cause I'd love to dual-boot with MultiRom Dec 16 18:26:04 So I updated halium/hybris-boot/init-script to check for /target/lib/systemd/systemd (in addition to the various inits) and now it's not bootlooping and I can ping it after manually assigning an ip address. No ssh or telnet but the screen dimmed slightly and it's not bootlooping anymore. Dec 16 18:27:22 No adb either. I have to run but I'll check what state it's in a bit Dec 16 18:27:39 It should not be required to change hybris-boot at all Dec 16 18:27:57 But if you have ubports-boot in the same tree, hybris-boot will brake Dec 16 18:28:39 I followed the directions on hybris and that's what it took to get the reference rootfs to do anything... Dec 16 18:28:56 I did a full wipe/format from twrp after I tested with ubports rootfs so i twould be blank Dec 16 18:29:28 @retro486, (Changing the init-script shouldn't be neccesary) Dec 16 18:29:39 Do you have ubports-boot in your halium source tree? Dec 16 18:31:57 nope Dec 16 18:32:13 just hybris-boot, audioflingerglu, and libhybris Dec 16 18:43:20 This is on halium-7.1 as well. I don’t know I mentioned that. Dec 16 18:54:03 Hey alright, yeah systemd is now in fact loading and I see messages in /proc/last_kmsg Dec 16 18:54:17 And I see a nice explicit lxc error Dec 16 19:17:15 Same here lxc@android.service failed together with lxc-net.service and isc-dhcp-server.service Dec 16 19:18:01 systemctl status lxc@android Dec 16 19:20:58 https://pastebin.com/yNed91XC Dec 16 19:23:04 HOLY WOW Dec 16 19:23:35 I got hybris-boot to start Upstart on an Ubuntu Touch image Dec 16 19:23:46 @UniversalSuperBox, HOW Dec 16 19:23:55 THAT IS A GAME CHANGER Dec 16 19:24:02 d????????? ? ? ? ? ? rootfs <- output of ls -al /var/lib/lxc/android, i guess that's my problem :p Dec 16 19:24:06 It was actually pretty easy Dec 16 19:25:40 @UniversalSuperBox, I am waiting for some kind of guidelines Dec 16 19:25:48 1. Install Ubuntu Touch with rootstock … 2. Move ubuntu.img to rootfs.img … 3. mount rootfs.img and delete the data/ symlink … 4. Edit `init-script` to remove ` --log-target=kmsg` on init call … 5. rebuild hybris-recovery … 6. Flash … 7. Boot Dec 16 19:26:06 There are other issues here, but it did boot. Dec 16 19:30:42 For example, /android wasn't mounted Dec 16 19:30:57 Correctly Dec 16 19:31:44 @UniversalSuperBox, rootstock as in ubports rootstock or halium rootstock? Dec 16 19:31:49 UBports Dec 16 19:31:59 where is the part that you place your android image? Dec 16 19:32:12 you need system.img alongside rootfs.img Dec 16 19:32:17 Rootstock stuffs it in the rootfs for you Dec 16 19:32:37 yes, but hybris-boot looks for data/system.img Dec 16 19:32:41 It looks like mounting that is handled when you hit the rootfs Dec 16 19:32:42 how does that even work Dec 16 19:32:48 @UniversalSuperBox, is it? Dec 16 19:33:24 Looks like it Dec 16 19:33:32 The boat loader doesn't try to find a system.img ever Dec 16 19:33:42 then that's it, great Dec 16 19:33:55 ``` mount $DATA_PARTITION /data … mount /data/rootfs.img /target``` Dec 16 19:33:58 @UniversalSuperBox, why have we build ubports-boot then Dec 16 19:34:01 That's pretty much its most important work Dec 16 19:34:12 @vanyasem, if everything basically works with hybris-boot Dec 16 19:34:14 @vanyasem, Because it was easier than what I'm doing right now Dec 16 19:34:21 And basically everything doesn't work, I'm missing a mount Dec 16 19:34:29 @UniversalSuperBox, oh, that's it Dec 16 19:34:32 it's broken Dec 16 19:34:33 :P Dec 16 19:35:20 It started Upstart. I call that a success Dec 16 19:35:26 Missing mount should be easy-peasey Dec 16 19:35:36 Hey, you got something with a UBports rootfs around? Dec 16 19:37:03 not really, only Plasma Mobile Dec 16 19:37:08 bah Dec 16 19:37:18 i can flash it on my HTC 10 really quick Dec 16 19:37:22 if you want me to Dec 16 19:37:32 grabs Fairphone Dec 16 19:37:43 @vanyasem, you will have to send me a virtual emoji heart though Dec 16 19:37:48 ❤️ Dec 16 19:37:55 @UniversalSuperBox, has marius fixed it Dec 16 19:38:04 Mine wasn't broken Dec 16 19:38:24 that was a joke basically :P Dec 16 19:38:32 Ohhhhhh Dec 16 19:38:33 Duh Dec 16 19:38:44 * vanyasem grabs HTC 10 Dec 16 19:40:30 `none 4.0K 0 4.0K 0% /android` Dec 16 19:40:32 Uh Dec 16 19:40:35 Now I'm scared Dec 16 19:44:42 LXC@android: Failed to create /sys/fs/cgroup/net_cls//lxc/android :s Dec 16 20:02:31 @UniversalSuperBox installing ubports Dec 16 20:02:55 The Android mount is goint tobe sticky. That actually is set up by the initrd Dec 16 20:03:40 huh? Dec 16 20:04:32 ubports-boot mounts `/cache, /data, /firmware, /persist, /system` at `/android` Dec 16 20:14:35 in Halium, that's the job of LXC scripts Dec 16 20:15:11 https://github.com/Halium/lxc-android/blob/master/var/lib/lxc/android/pre-start.sh Dec 16 20:15:45 but doesn't 16.04 use systemd? Dec 16 20:17:42 Not from UBports Dec 16 20:18:06 Canonical didn't finish making Unity compatible and there were some missing units as well Dec 16 20:18:30 so that cool name approach keeps upstart as well, righ? Dec 16 20:18:39 can't we just symlink them @UniversalSuperBox? Dec 16 20:20:29 Right Dec 16 20:20:38 @vanyasem, Well, they aren't mounted anywhere Dec 16 20:25:30 wut Dec 16 20:25:36 are they completely missing Dec 16 20:25:53 how do you work with kernel trees, git freezes when i type `git status` Dec 16 20:26:04 Own a faster computer Dec 16 20:26:33 isn't 8 threads 4.2 Ghz each enough? Dec 16 20:26:40 HDD? Dec 16 20:27:13 my cores are not overloaded. htop says they are barely used Dec 16 20:27:19 yep, hdd. that might be it Dec 16 20:27:24 It is Dec 16 20:27:42 `git status` takes 9 seconds on my machine Dec 16 20:27:53 iotop says that my hdd is nuked right now Dec 16 20:28:09 daymn Dec 16 20:28:22 still not finished -_- Dec 16 20:29:16 i'll copy it to my ssd then Dec 16 20:29:23 let's see if it'll get better Dec 16 20:29:33 oh git status finished. Dec 16 20:29:41 4 minutes of my time Dec 16 20:30:01 Hmm, I wonder what actually mounts the /android bits Dec 16 20:30:23 oh now it takes 2 seconds Dec 16 20:30:37 i guess i never ran git status before and it needed to collect some info or something like that Dec 16 20:30:42 SERIOUSLY? Dec 16 20:30:49 I did `grep -r` on my Fairphone and it crashed Dec 16 20:32:04 Wait. Interesting. It always does it in the same place. Dec 16 20:32:24 (Photo, 1280x425) https://irc.ubports.com/B5S4VjLQ/file_2903.jpg er what Dec 16 20:32:30 why does it list unchanged files Dec 16 20:32:43 they aren't renamed or anything. they are just unchanged Dec 16 20:32:49 Permissions? Dec 16 20:32:50 @vanyasem, just do 'git status' on intrested pieses: e.g., 'git status arch'. No? Dec 16 20:33:14 @vanyasem, didn't know that was possible. but i wanted a whole overview because of that Dec 16 20:33:48 i can't stage them btw, they all just disappeared Dec 16 20:33:53 seems like a weird git bug Dec 16 20:34:09 Or Atom Dec 16 20:34:15 that's not atom Dec 16 20:34:48 also i have done that from my terminal, as i just told you. i opened gitkraken to illustrate that for a screenshot Dec 16 20:34:57 Ah Dec 16 20:35:07 a'ight, going back to irc Dec 16 20:53:00 oh im stupid Dec 16 20:53:03 Melp Dec 16 20:53:21 ofc git status got frozen, i am currently downloading stuff on this hard drive in the background Dec 16 21:00:56 Any kernel hackers awake? Dec 16 21:02:06 I've continued investigating how I'm able to make the Fairphone die. I can read last_kmsg but it seems horribly corrupted. Wondering if someone who has looked at a lot of these logs could decipher it. Dec 16 21:04:44 Oop, found the start Dec 16 21:04:54 ```[ 20.457667] Kernel paN)" - not synCing: Fatal exceptiol … Y 20.55773'Y CPU0: stoppinf``` Dec 16 21:04:57 ._. Dec 16 21:13:40 is that really how it was written lol Dec 16 21:13:46 Yes Dec 16 21:13:59 kernel pan, something revolutionary Dec 16 21:14:15 Something is very wrong, though I think it's in the ram console rather than the panic Dec 16 21:14:20 Since it happens on normal reboots too Dec 16 21:14:39 does is crash instantly? Dec 16 21:14:52 No, this is the fairphone Dec 16 21:14:55 if not, try to run a script that will save dmesg in a loop to some file Dec 16 21:15:06 worked for me when i was debugging htc 10 Dec 16 21:15:19 Oh, in that case yeah it crashes as soon as I try to read a procfile. Dec 16 21:15:35 too bad, i have nothing else to suggest Dec 16 21:15:59 install gentoo maybe Dec 16 21:16:01 but not sure Dec 16 21:16:09 * vanyasem runs Dec 16 21:16:27 'not syncing' implies "It's too unsafe to write out any open files" Dec 16 21:17:30 Hey, I got a normal one! Dec 16 21:18:46 send it here Dec 16 21:18:52 i want to look at it Dec 16 21:19:14 http://paste.ubuntu.com/26196996/ Dec 16 21:19:17 Almost normal Dec 16 21:28:43 dalton Dec 16 21:28:53 That's my name Dec 16 21:28:54 i have finally build my halium tree for the famous nexus 5x Dec 16 21:28:59 built* Dec 16 21:29:04 yay Dec 16 21:29:23 just one month since i promised to do so "tomorrow" Dec 16 21:30:22 oops nop Dec 16 21:30:26 build failed Dec 16 21:58:33 oh, that's arch having python symlinked to python3 by default Dec 16 22:31:42 @UniversalSuperBox, any news? Dec 16 22:31:51 On? Dec 16 22:32:09 on your issue with mounts Dec 16 22:32:17 mmmno Dec 16 22:32:19 you were trying to run UBports with hybris-boot Dec 16 22:32:24 oh, well Dec 16 22:32:35 I have just finished building my tree :P Dec 16 22:48:43 vanyasem: If you really want something to do, figure out what mounts /android/* in Ubuntu Touch Dec 16 22:48:56 I don't think it's in the initrd Dec 16 23:27:21 @UniversalSuperBox, https://superuser.com/q/613702 Dec 16 23:44:03 @HenryOne, The Fairphone kernel is pretty bad, so I'd be willing to bet they just screwed up the ram console. Dec 17 02:17:41 Is the fstab supposed to be in the system.img somewhere? Dec 17 02:34:59 Oh, duh. I see what I did wrong. The fstab is in the android ramdisk, which is in `[system.img]/boot/` Dec 17 02:52:07 @bhushanshah I think I'm getting closer to getting this boat loaded. However, I don't get the `/dev/block/bootdevice` mount points so the mount-android script doesn't work. Isn't this what fixup-mountpoints is for? Dec 17 02:52:43 How do I either get `/dev/block/bootdevice` or get an fstab that's been edited by fixup-mountpoints? **** ENDING LOGGING AT Sun Dec 17 03:00:00 2017 **** BEGIN LOGGING AT Sun Dec 17 03:00:01 2017 Dec 17 03:24:52 @UniversalSuperBox, I don't get why you need /dev/block/bootdevice symlink? Dec 17 03:25:21 Because that's what's in my `fstab.qcom` file in the boot.img. Dec 17 03:25:42 Actually it contains `/dev/block/bootdevice/by-name` Dec 17 03:27:51 Anyway, there is code in mount-android to guess partition without block symlink Dec 17 03:29:02 Even though it's all `by-name`? Dec 17 03:29:14 Yes Dec 17 03:30:28 Aha Dec 17 03:30:46 https://github.com/Halium/lxc-android/blob/master/usr/lib/lxc-android/mount-android#L11 Dec 17 03:30:51 This loop Dec 17 03:31:05 (shamelessly taken from ubports-boot) Dec 17 03:31:18 I like it better out of the initrd Dec 17 03:31:29 Me too Dec 17 03:32:04 Apparently I can only boot with hybris-boot once per rootfs Dec 17 03:32:20 Past that I get `ssh_exchange_identification: read: Connection reset by peer` Dec 17 03:32:23 I blame upstart. Dec 17 03:33:52 ssh host keys are getting reset? Dec 17 03:34:32 I would expect a different error if that was the case Dec 17 03:35:20 Yeah, removing it from known_hosts doesn't help Dec 17 03:50:38 The script still seems to be ignoring all of my mounts. Dec 17 03:51:29 Via elite debug (echo) skills, it probably doesn't like the $path it's getting Dec 17 03:54:33 Aha, I was right. Not only is `/dev/block` a symlink to `/dev/`, it also doesn't have `/dev/disk` available. Dec 17 03:54:48 UT expects everything to be handed to it Dec 17 03:54:53 Did udev run? Dec 17 03:55:18 It didn't. Dec 17 03:55:39 but why did it not... Dec 17 03:55:47 That's what creates /dev/disk symlink Dec 17 03:56:08 In ubp it runs udev after the container Dec 17 03:56:15 So you need to fix deps there Dec 17 03:56:33 And since the mounts aren't there, no container. Thanks for the lead. Dec 17 04:01:00 Okay, noted. I'm going to head off for the night. Thanks for the help. Dec 17 04:01:13 Tomorrow I get to dig in to upstart scripts Dec 17 04:01:38 Surprisingly sane, I wouldn't be upset if it had won over systemd. Dec 17 04:02:19 (Document) https://irc.ubports.com/5OpZUbwo/file_2909.mp4 Dec 17 04:02:28 Really. Dec 17 04:06:37 I mean, don't get me wrong, I like units. But Upstart configs aren't as bad as the others Dec 17 04:08:52 I don't disagree on units being confusing at a times Dec 17 09:14:26 I try configure a build for sony z3c. During my search for the file specified in TARGET_KERNEL_CONFIG I got to an include statement that is missing in my source tree ('include device/sony/msm8974-common/BoardConfigCommon.mk'). The strange thing is that this was no dependency in 'lineageos.dependencies'. Dec 17 09:20:19 kllp: it is possible that a) lineageos.dependencies is out of date and b) deps are nested.. i.e your device repo includes some other repo and it includes this repo Dec 17 09:22:14 you are right! there was another 'lineage.dependencies' inside a dependency Dec 17 09:24:14 z3c as-in what's your device codename? Dec 17 09:47:46 'aries' or in newer roms z3c Dec 17 11:39:52 ahh something went wrong with rootstock-install ... for ubports .. system.img and ubuntu.ing in the data partition are two different files Dec 17 11:52:22 My device won't boot correctly, the diagnosis.log file says `No /proc/config.gz. Enable CONFIG_IKCONFIG and CONFIG_IKCONFIG_PROC`, even though I added these two in the defconfig file Dec 17 12:08:26 Seems like I was missing some entries in the fstab. Let's see if it works now Dec 17 12:30:02 Nope that wasn't it. Still giving the same error Dec 17 12:41:56 hello, I did some progress porting halium for our Gemini device. I now compiled hybris-boot and systemimage. The device itself refuses to boot - "X refuses to boot. See /diagnosis.log (in initrd only)". I tried connecting via RNDIS but can't get DHCP to work. Is there a default static IP associated to the device? Also I see that the serial cable works, but there is no shell running. Any suggestions on how to get the diagnosis file out of the device? Dec 17 12:43:18 you need to telnet into it. if you can't get DHCP to work, assign a static ip yourself Dec 17 12:43:27 https://github.com/JBBgameich/halium-install/blob/master/connect-telnet.sh Dec 17 12:43:39 this script telnets into your device using a static ip Dec 17 12:51:31 @vanyasem thanks - I can see the device now. telnet port is not open, but ssh port is. What is the login/password? Dec 17 12:51:56 if your ssh port is open, then you booted into rootfs Dec 17 12:52:04 then it has not crashed in initrd Dec 17 12:52:05 (which is good) Dec 17 12:52:14 exactly Dec 17 12:52:20 cool Dec 17 12:52:31 so there is not need to look for diagnosis.log, you were actually able to boot the system Dec 17 12:52:44 There is also a script for ssh: https://github.com/JBBgameich/halium-install/blob/master/connect-ssh.sh Dec 17 12:56:57 my device is visible under 'usb0' network on Linux host, using ip 192.168.2.48, and SSH asks for a password - I tried obvios ones (i.e. root/halium) without success Dec 17 12:57:24 set it ip address of 10.15.19.100 and then ssh root@10.15.19.82 Dec 17 12:57:43 If you used the install script from the halium-scripts repository, you still need to set the password in recovery I guess Dec 17 12:58:06 yep, chroot into the rootfs from twrp and run `passwd` Dec 17 12:58:19 JBB's script does that for you Dec 17 13:00:23 thanks - will set the password in the root system then Dec 17 13:01:51 i broke my kernel tree somehow Dec 17 13:01:54 it refuses to build Dec 17 13:02:52 https://paste.ubuntu.com/26201246/ Dec 17 13:02:56 can someone take a look at that Dec 17 13:05:39 oh i know why lol Dec 17 13:05:53 @bhushanshah how have you sorted your kernel via menuconfig? Dec 17 13:06:13 it uses my host arch (x86) for that Dec 17 13:06:30 i specified KCONFIG location, but still no luch Dec 17 13:06:58 There was env var somewhere for that Dec 17 13:07:05 I don't exactly remember Dec 17 13:08:24 i couldn't find it yesterday Dec 17 13:08:28 and thus broke my config Dec 17 13:11:49 if you recall it, please, ping me :P i would appreciate that a lot Dec 17 13:19:15 @bhushanshah, `ARCH=arm`, i guess Dec 17 13:19:29 I think Dec 17 13:19:39 `arm64` in my case Dec 17 13:30:52 @vanyasem make menuconfig KCONFIG_CONFIG=blablalbalabala Dec 17 13:30:58 that's what i've done Dec 17 13:31:09 it broke after that Dec 17 13:31:22 rip Dec 17 13:31:46 https://github.com/Halium/android_kernel_htc_msm8996/commit/48c0d432f7f46bc596df13e704d0c5874e075367 Dec 17 13:31:47 this commit Dec 17 13:32:08 i'll probably revert it and start from scratch Dec 17 13:54:40 @vanyasem so in short you didn't test the commits before pushing :P Dec 17 13:55:07 @bhushanshah, tss Dec 17 14:01:16 @vanyasem, yep worked after reversting Dec 17 14:01:29 i pushed unstable untested code that broke the tree Dec 17 14:01:30 blame me Dec 17 14:21:14 Davide what way did you go around having only MediaTek AOSP tree? by porting Halium changes to AOSP tree or by trying to build Halium tree with AOSP device tree? Dec 17 14:28:57 @NotKit I have built the Halium tree by copying the device tree and a minimum set of drivers. However I had to cheat a bit to get it to compile both hybris-boot and systemimage, so I don't know exactly what to expect, hopefully I will get an idea in a bit Dec 17 14:59:08 Slowly getting there, but now systemd is freezing execution. Anybody got an idea? Dec 17 14:59:09 https://paste2.org/ZE42wxVt Dec 17 15:00:13 See hammerhead kernel on halium GitHub, there is one commit by torvalds in halium-5.1 branch Dec 17 15:00:20 Backport that patch Dec 17 15:00:51 great, I will look into that, thank you Dec 17 15:01:34 https://github.com/Halium/android_kernel_lge_hammerhead/commit/25437b2a54dd619a96e268ecaf303b089aa785e4 Dec 17 15:01:35 this commit Dec 17 15:09:08 Davide from what I remember, MediaTek has some AOSP parts patched, for example xlog symbols in bionic, which might be needed to get blobs to work later Dec 17 15:13:55 https://github.com/danielhk/android_device_lenovo_aio_otfp/blob/cm-14.1/platform_patches/system_core_liblog.patch Dec 17 15:19:09 Thanks, now it does not freeze anymore. But I still don't see an iSerial listing. Or is it just taking such a long time? Dec 17 15:50:25 Just flashed the `hybris-recovery.img` but still get no feedback. last_kmsg says no error, systemd started normally Dec 17 15:54:18 @Dorni[m], are you checking dmesg? Dec 17 15:55:44 yes, I check what I can get from /proc/last_kmsg Dec 17 15:56:51 last entries are from systemd, last one is `systemd[1]: Started Read required files in advance.` Dec 17 15:57:51 @Dorni[m], I mean how do you know that there is no iSerial listing? Dec 17 15:58:03 have you checked your desktop's dmesg? Dec 17 15:58:51 I have used the command that outputs the lsusb list Dec 17 15:59:33 My desktops dmesg does not list any usb network devices Dec 17 16:03:28 @vanyasem did you verify fix works? :P Dec 17 16:04:58 @bhushanshah, I verified that it builds Dec 17 16:05:37 Read: 'not really' Dec 17 16:05:39 "works" Dec 17 16:06:13 @UniversalSuperBox, 😂 Dec 17 16:06:53 Well I am having a bad day... I had got a hardware accelerated video working earlier but now it's not working Dec 17 16:06:59 and I don't know what's wrong Dec 17 16:15:05 Mh.. now it won't even start systemd. I have some `avc: denied` errors, then an emergency remount Dec 17 16:15:34 @bhushanshah, Such is life Dec 17 16:47:27 I think I underestimated just how much work is put in by the initrd in an Ubuntu Touch system. Dec 17 16:57:41 I don't get it. Systemd seems to load without problems, but I cannot get a usb connection 😕 anything else I can try? Dec 17 16:58:11 From chroot disable lxc@android service Dec 17 16:58:15 And then try to boot Dec 17 17:06:04 Can I do that via adb? Dec 17 17:06:04 okay found it Dec 17 17:12:32 Nope, it still doesn't work. Last messages in kmsg are from systemd initializing and that's it Dec 17 17:27:00 I can't find the kernel configuration. There is no 'arch/arm/configs' directory in my source tree. TARGET_KERNEL_CONFIG := lineageos_shinano_aries_defconfig Dec 17 17:27:49 its in `kernel///` Dec 17 17:28:20 if you have no kernel directory, you need to add it to your local manifest file and run repo sync again Dec 17 17:30:52 halium@gemini - I managed to telnet to the device and I found a few kernel configuration issues in diagnosis.log. I will try to apply those changes and will report back Dec 17 18:13:05 I love halium on my nexus Dec 17 18:17:00 Building the system.img and hybris-boot.img fails due to 'ERROR: fstab not found'. I did not edit 'fixup-mountpoints' as the entry was already there. Dec 17 18:18:43 * kllp[m] sent a long message: kllp[m]_2017-12-17_18:18:43.txt Dec 17 18:20:36 You forgot to run lunch/breakfast command Dec 17 18:31:43 Im downloading reference rootfs for tests Dec 17 18:36:06 'breakfast z3c' gives ** Don't have a product spec for: 'lineage_z3c' Dec 17 18:36:06 ** Do you have the right repo manifest? Dec 17 18:36:31 [[device/sony/z3c/lineage.mk]]: "vendor/sony/msm8974-common/msm8974-common-vendor.mk" does not exist. Stop. Dec 17 18:47:21 You need an additional Android_device_sony_msm8974-common repository Dec 17 18:47:36 No sorry Dec 17 18:47:57 It's a vendor repository Dec 17 18:48:01 if I recompile the kernel with the suggestions I found in diagnosis.log I get a kernel panic when trying to do switch_root: [ 10.278260] <1>-(1)[1:switch_root]Kernel panic - not syncing: Attempted to kill init! exitcode=0x00000100 Dec 17 18:59:04 do you have any rootfs installed on device? Dec 17 19:00:14 I suppose it would exit earlier otherwise though Dec 17 19:04:25 please check if mer-kernel-check is fine with current config for systemd-required options, then you can boot hybris-recovery.img instead, which will stop boot process before switching root, so you can examine if rootfs is mounted correctly and is in read-write mode Dec 17 19:36:22 Found an IRC log (#sailfishos-porters) where someone got the same problem but I can't find a solution http://www.merproject.org/logs/%23sailfishos-porters/%23sailfishos-porters.2015-09-13.log.html Dec 17 19:37:05 I installed caf libhybris in reference rootfs, and when i run EGL_PLAtFORM=hwcomposer test_hwcomposer i get a segfould. anyone knows why? Dec 17 19:42:47 provide strace maybe? Dec 17 19:46:46 vanyasem: https://paste.pound-python.org/show/sEJ7KbJejL04NB51XqtI/ Dec 17 19:47:42 and in dmesg: https://paste.pound-python.org/show/1h6wmslfkx6BIV6tDUak/ Dec 17 20:34:11 Should the vendors folder be part of my local manifest file? Dec 17 20:45:27 vendors folder? if by that you mean vendors repo, then yes Dec 17 20:45:41 Fwd from bhushanshah: @sopernam … Dec 17 20:45:42 something like this as local manifest Dec 17 20:45:47 here's an example for bullhead Dec 17 20:46:31 omg pound python pastebin Dec 17 20:46:35 never heard of it before Dec 17 20:54:58 "Vendor blobs go in the vendor/ folder in your Halium source tree." Dec 17 21:52:48 aki237 was added by: aki237 Dec 17 22:01:49 guys I see an error : … system/core/fastboot/usb_linux.c: In function ‘find_usb_device.constprop’: … system/core/fastboot/usb_linux.c:147:63: error: ‘%s’ directive output may be truncated writing up to 255 bytes into a region of size 252 [-Werror=format-truncation=] … snprintf(info.device_path, sizeof(info.device_path), "usb:%s", sysfs_name); … ^~ … system/core/fastb Dec 17 22:01:50 ‘snprintf’ output between 5 and 260 bytes into a destination of size 256 … snprintf(info.device_path, sizeof(info.device_path), "usb:%s", sysfs_name); … ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ … system/core/fastboot/usb_linux.c:161:40: error: ‘%s’ directive output may be truncated writing up to 255 bytes into a region of size 59 [-Werror=format-truncation=] … "/sys/bus/usb/devices/% Dec 17 22:01:51 ^~ … system/core/fastboot/usb_linux.c:160:9: note: ‘snprintf’ output between 29 and 284 bytes into a destination of size 80 … snprintf(path, sizeof(path), … ^~~~~~~~~~~~~~~~~~~~~~~~~~~~ … "/sys/bus/usb/devices/%s/serial", sysfs_name); … ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ … cc1: all warnings being treated as errors Dec 17 22:02:33 gcc : 7.2.1 Dec 17 22:08:08 probably the wrong group to ask kernel development related questions ( that are not related to halium ), but … 1. why are you using gcc 7.2.1? … 2. does it compile with the default compiler in android tree? (usually gcc 4.9) … 3. if you are certain your warning isn't the end of the world, you can nuke gcc-wrapper.py if exists ( see example here: https://git.sphere.ly/simonsimons34/android_kernel_htc_k2/commit/0a16b05ca950719e9469f9bc3f3 Dec 17 22:09:17 @trivial_inanity, I dont think it is kernel build warning Dec 17 22:09:42 ouch. Dec 17 22:09:47 point 1 still applies though Dec 17 22:09:58 1 and 2 Dec 17 22:10:18 if it compiles with 4.9 it shouldn't be the end of the world warning Dec 17 22:10:53 if you've explicitly set the toolchain in your device tree, please remove it and try again Dec 17 22:11:44 No it uses system compiler (In Arch Linux... Now downloading a docker image of Ubuntu....) Dec 17 22:11:55 Which ubuntu is good for build? 14.04? Dec 17 22:19:48 hmm, I've compiled halium with arch just fine Dec 17 22:20:09 is this 5.1 or 7.1 ? Dec 17 22:20:21 gcc or android? Dec 17 22:20:36 gcc : 7.2.1 and cm-12.1 Dec 17 22:27:40 @trivial_inanity, ? Dec 17 22:28:03 oh, lollipop Dec 17 22:28:21 here: https://github.com/z3ntu/halium-5.1-build-env Dec 17 22:28:31 use this docker image Dec 17 22:30:34 With your help I successfully build 'hybris-boot.img'. Building 'system.img' fails because 'out/host/linux-x86/framework/signapk.jar' is missing. Do you know where I can get it from? Dec 17 22:32:09 Remove all apk related parts from the vendor Makefiles Dec 17 22:32:20 @trivial_inanity, Thanks Dec 17 22:51:10 This error now : … ```make: *** No rule to make target `/var/android/out/target/product/ham/obj/SHARED_LIBRARIES/libcryptfs_hw_intermediates/export_includes', needed by `/var/android/out/target/product/ham/obj/EXECUTABLES/vold_intermediates/import_includes'. Stop.``` Dec 17 22:57:32 It's a tough task to check every vendor makefile for apk parts. Dec 17 23:46:57 any idea about this error : … ```make: *** No rule to make target `libqmi', needed by `/var/android/out/target/product/ham/obj/EXECUTABLES/wpa_supplicant_intermediates/config.o'. Stop.``` Dec 17 23:57:23 I believe Halium doesn't use Android's wpa supplicant Dec 17 23:57:29 so you can just safely remove it Dec 17 23:57:36 From where? Dec 17 23:57:49 from an appropriate makefile Dec 17 23:58:09 run a grep -r search to figure it out Dec 17 23:58:58 from … ```device/$VENDOR/$DEVICE``` … makefiles? Dec 17 23:59:53 i am not sure if it's in the device or vendor tree Dec 17 23:59:58 figure it out Dec 18 00:01:29 I had this error before. I think I have to add a new configuration for my device somewhere. … ```ham``` Dec 18 00:01:59 That error comes when the vendor repo is not included in that config. Dec 18 00:03:46 well you don't need wpa supplicant anyways, so you'd better remove it Dec 18 00:04:06 from where any idea Dec 18 00:04:30 run a search and figure it out... Dec 18 00:04:51 or manually look for networking-related makefiles Dec 18 02:33:21 Which image should I be trying? vivid? Dec 18 02:47:28 image for what? Dec 18 02:47:30 ugh Dec 18 02:47:47 my bad. For flashing Dec 18 02:47:56 what do you want to flash? Dec 18 02:47:57 vivid or xenial? Dec 18 02:48:16 I have built ubp-5.1 Dec 18 02:48:30 Which is compatible? Dec 18 02:48:47 do you know this is a supergroup about halium and not ubports? Dec 18 02:49:05 have you built a Halium 5.1 tree or ubp-5.1 tree? Dec 18 02:49:10 Oh ok sorry. Dec 18 02:49:14 ubp-5.1 Dec 18 02:49:42 why do you ask questions about it in halium supergroup? -_- Dec 18 02:49:48 ask in @ubports maybe? Dec 18 02:50:02 i should say vivid probably, as xenial is built for halium Dec 18 02:50:11 @vanyasem, ok Dec 18 02:50:38 just follow the official docs for ubports, ok? Dec 18 02:50:56 @vanyasem, Sure thanks Dec 18 02:51:16 i am pretty sure it's specified in there, although i have no idea what they say, as again, this is a halium-specific chat Dec 18 02:52:50 i am so frustrated right now Dec 18 02:53:46 @vanyasem I'm real sorry. The guys over there said any questions on porting should be asked here. Thats why.. Now I'm clear Dec 18 02:55:09 that's generally a good idea, as the porting process is simmilar. but ubports specific parts belong in @ubports chat. actually, ubp-5.1 is deprecated and you should not really port to it Dec 18 02:55:54 newer xenial builds are based on Halium and that's what we're working on here. you can ask questions about xenial builds here and i will personally answer any of your questions Dec 18 02:56:41 @vanyasem, you're working on something that will get obsolete soon, so you're probably just wasting your time **** ENDING LOGGING AT Mon Dec 18 03:00:01 2017 **** BEGIN LOGGING AT Mon Dec 18 03:00:02 2017 Dec 18 03:02:06 what is your device btw? @aki237 Dec 18 03:14:33 Oh sorry, I was away @aki237 Dec 18 03:14:37 Zuk Z1 (ham) Dec 18 03:15:10 I didn't realize you were building a 5.1 tree. However, you should be able to 'rebase' that on top of Halium-5.1 and find yourself in the exact same situation Dec 18 03:15:16 Since Halium-5.1 is a fork of ubp-5.1 Dec 18 03:15:27 oh Dec 18 03:15:54 Were you on the wiki to find that information? Dec 18 03:16:12 which? Dec 18 03:16:16 wiki.ubports.com Dec 18 03:17:25 Yes Dec 18 03:17:34 Okay, I'll delete that now. Dec 18 03:19:35 Oh no, the wiki has a merge conflict again Dec 18 03:22:14 Well, for now there's a warning at the top of https://wiki.ubports.com/wiki/UBports-Development-Information which has been there for about a month. I'm not able to edit the rest. Dec 18 05:16:31 Guys I'm getting a weird error in Kernel build in halium-7.1 … ```/ubport/halium/kernel/cyanogen/msm8974/kernel/user_namespace.c: In function 'create_user_ns': … /ubport/halium/kernel/cyanogen/msm8974/kernel/user_namespace.c:36:2: error: implicit declaration of function 'proc_alloc_inum' [-Werror=implicit-function-declaration] … ret = proc_alloc_inum(&ns->proc_inum); … ^ … /ubport/halium/kernel/cyanogen/msm8974/kernel/user_namespace.c: I Dec 18 05:16:32 'free_user_ns_work': … /ubport/halium/kernel/cyanogen/msm8974/kernel/user_namespace.c:83:2: error: implicit declaration of function 'proc_free_inum' [-Werror=implicit-function-declaration] … proc_free_inum(ns->proc_inum);``` Dec 18 05:17:14 As fas as seen it points to CONFIG_USER_NS setting in kernel configuration Dec 18 05:18:14 Don't paste more then 4 lines directly in chat Dec 18 05:18:16 Use pastebin Dec 18 05:18:24 Will do Dec 18 05:18:47 Also you don't need CONFIG_USER_NS configuration Dec 18 05:19:14 ok Dec 18 05:19:21 We will be run priviliged android container so USER_NS is not so useful Dec 18 05:19:41 ok Dec 18 05:36:41 Do we need any of the XT configs : http://paste.ubuntu.com/26206942/ Dec 18 05:37:07 gcc : 7.2.1 Dec 18 12:20:10 * kllp[m] sent a long message: kllp[m]_2017-12-18_12:20:09.txt Dec 18 12:56:44 Halium for 8.0 Oreo? Will be available in future? Dec 18 12:58:52 We'd need to update the whole tree. That has already been done once from halium 5.1 to 7.1, but I don't think anyone already works on 8.0 because there are not many devices that require 8.0 Dec 18 13:01:01 So maybe in future. If oreo will be not skiped, and halium updated to android 9.0. Dec 18 16:04:29 Veth pair device: enabled, not loaded <- anyone knows why lxc-checkconfig says enabled, not loaded? Dec 18 16:04:43 ignore it :) Dec 18 16:08:59 oke Dec 18 16:09:46 CONFIG_NF_NAT_IPV4: missing <- can't be found in make menuconfig is this gonna be a problem? Dec 18 16:10:29 revertive: in general, for lxc-checkconfig, only 1st section is useful Dec 18 16:10:49 err 1st two I mean Dec 18 16:11:38 oke thanks, thats all enabled so thats good. Dec 18 16:13:44 trying a clean build and see if lxc@android start without errors this time Dec 18 17:39:31 lxc@android still fails to create directory /sys/fs/cgroup/net_cls//lxc/android Dec 18 17:39:42 anyone know where i should look Dec 18 19:46:01 @UniversalSuperBox, btw, can we maybemake the init-script detect if the rootfs uses systemd and set the log-target only if it's using systemd? Dec 18 20:24:55 aki237: For newer GCC you might need some patches. Depends on your kernel version Dec 18 20:24:55 probably by checking if /sbin/init is symlink to ../lib/systemd/systemd Dec 18 20:25:03 For 3.4 and older kernels you migth need some Dec 18 20:25:05 or maybe simply if /lib/systemd/systemd exists Dec 18 20:25:06 We needed for our LuneOS Dec 18 20:25:20 On newer kernels like 3.18 they don't seem needed Dec 18 20:26:47 @NotKit, I had the same idea in the meantime, I'm working on it :) Dec 19 00:32:28 Any ideas for this error , gcc : 7.2.1 : http://paste.ubuntu.com/26206942/ Dec 19 01:35:29 have you switched to halium 5.1 branch? Dec 19 01:35:48 halium-7.1 Dec 19 01:36:00 oh, you on halium 7.1 now Dec 19 01:36:05 and is your device tree also 7.1? Dec 19 01:36:26 just to make sure Dec 19 01:36:44 also, what is the reason you're using a custom gcc? Dec 19 01:37:15 That's the gcc that came with the system Dec 19 01:37:46 isn't gcc bundled with halium tree? 🤔 Dec 19 01:38:21 nope, it's not. sorry Dec 19 01:38:41 I didn't override any options. I assumed it used system compiler thats why I mentioned the system compiler version Dec 19 01:39:02 is gcc used at all? Dec 19 01:39:13 @vanyasem, You're right Dec 19 01:39:14 i believe GNU make is the compiler used heer Dec 19 01:39:32 Make is not compiler Dec 19 01:39:33 Make is an automation for other compilers Dec 19 01:39:34 make is the build tool. Dec 19 01:39:46 and I think 7.1 uses ninja Dec 19 01:39:54 i should not try to assist people at 4 AM Dec 19 01:40:20 @bhushanshah, which version is it? Dec 19 01:40:28 some system components compiles correctly. But this happens when trying to compile the kernel. Dec 19 01:40:45 @aki237, does it compile with the default config? Dec 19 01:40:49 @vanyasem, There are multiple, iirc Dec 19 01:40:53 This happened before for me when building the ubp-5.1 branch. Dec 19 01:41:01 @vanyasem, maybe it's some option that's broken in your kernel Dec 19 01:41:18 @aki237, have you switched from 5.1 branch to 7.1 without changing your device treE? Dec 19 01:41:25 @vanyasem, Without the libhybris config... I havent checked Dec 19 01:41:34 @vanyasem, are you building halium 7.1 with android 5.1 sources Dec 19 01:41:47 @vanyasem, started from scratch ... no rebasing or nothing Dec 19 01:42:04 @vanyasem, halium-7.1 with cm{lineage}-14.1 Dec 19 01:42:12 @aki237, right Dec 19 01:42:20 try building with the stock kernel config Dec 19 01:42:41 Without the libhybris compatibility changes? Dec 19 01:42:45 and if it builds, then it's some option that's you've enabled with mer-config-checker and that's broken in your kernel tree Dec 19 01:42:48 Will do Dec 19 01:42:57 @aki237, yes. vanilla config from lineage repos Dec 19 01:43:21 @aki237, I was advised to remove the "_XT_" options from the default configuration Dec 19 01:43:26 sometimes vendor break certain functionality because it was never used for the device Dec 19 01:43:32 and so it was never tested Dec 19 01:43:59 if it builds with the default config, then it's not your issue and we'll need to figure out which option is faulty Dec 19 01:44:07 I thought 7.1 uses systemd unlike 5.1 so the kernel has to be re-configured. Dec 19 01:44:14 ok Dec 19 01:44:17 @vanyasem, and then either fix it or disable altogether Dec 19 01:44:20 @aki237, that is right Dec 19 01:44:41 5.1 uses systemd as well if im not mistaken Dec 19 01:44:55 don't get confused between UBports versions and Halium versions Dec 19 01:45:01 @vanyasem, @bhushanshah am i right? Dec 19 01:45:13 Yes Dec 19 01:45:14 I think you're missing the separation here Dec 19 01:45:32 Or well at least halium does Dec 19 01:46:23 the fact that ubports is not following halium standarts properly confuses new people Dec 19 01:46:34 😒 Dec 19 01:46:41 Halium does not care what the init system is Dec 19 01:46:53 But Dec 19 01:47:15 Halium does provide systemd units that can be used by the rootfs to start the Android container Dec 19 01:47:21 @UniversalSuperBox, it's built with systemd in mind though Dec 19 01:47:43 @UniversalSuperBox, i would love to screw it over with openRC Dec 19 01:47:50 or runit for that matter Dec 19 01:47:58 You can do whatever you want Dec 19 01:48:17 i know that, i'm just sharing my thoughts for future Dec 19 01:48:33 maybe someone has already done something related or has something to say Dec 19 01:49:49 @vanyasem, I've already did this with sailfish OS Dec 19 01:50:04 Pain in the butt Dec 19 01:52:33 lets start the "move away from systemd, halium" chanting here as well :p Dec 19 01:53:00 we have a whole group dedicated for that purpose, @ubports_ot Dec 19 01:53:08 i remember at least 2 rants about systemd there Dec 19 01:53:24 i was that gentoo user fighting for openRC btw Dec 19 01:53:40 You can do whatever you want Dec 19 01:53:54 I'm today going to rant about systemd race condition I am going to debug Dec 19 01:54:20 btw @bhushanshah i find kde bug reporter so confusing Dec 19 01:54:35 OT Dec 19 01:54:43 err Dec 19 01:56:34 bshah, take that opportunity and nuke systemd /s Dec 19 01:57:18 So, I am that only weird dude who likes systemd? Damn. Dec 19 01:57:25 No, you are not. Dec 19 02:05:46 Will we be needing CONFIG_AUDIT in the kernel configs? Dec 19 02:06:45 i think it's not so much people not liking systemd as android vendors shipping fucking old frankenstein kernels that don't support it Dec 19 02:08:59 doesn't systemd work on kernels as late as 2.6 ? (gentoo wiki tells me this) Dec 19 02:09:27 i think you're referring to glibc (?) Dec 19 02:09:46 which requires kernel 3.2 Dec 19 02:09:55 > glibc v2.24 Dec 19 02:09:59 @aki237, Is it not working when you build with it? Dec 19 02:10:29 while -some- android devices are still running either 3.0.x or 2.6.x Dec 19 02:11:35 @UniversalSuperBox, Building... Will check Dec 19 02:13:31 @UniversalSuperBox, mer's perl script is saying me to set it to n. but check-config is saying me to set it to y. Dec 19 02:14:07 mer kernel config check needs updating to check for kernel version Dec 19 02:14:20 ok check-config it is. Dec 19 02:31:52 @bhushanshah, @bhushanshah I think we need NS ... I get this error http://paste.ubuntu.com/26212450/ Dec 19 02:33:34 @aki237, And for this http://paste.ubuntu.com/26212455/ patch fixed it Dec 19 02:52:05 does the halium use ccache? Dec 19 02:52:12 You can if you want to Dec 19 02:52:19 By default? Dec 19 02:52:23 No Dec 19 02:52:28 Unless you have it exported **** ENDING LOGGING AT Tue Dec 19 03:00:01 2017 **** BEGIN LOGGING AT Tue Dec 19 03:00:02 2017 Dec 19 03:03:14 ok Dec 19 05:09:37 http://paste.ubuntu.com/26212450/ error : even though CONFIG_PID_NS is switched on. Dec 19 05:10:28 Hmm your kernel vendor backported half feature Dec 19 05:10:37 And thought it would just work Dec 19 05:10:40 Because why not? Dec 19 05:10:49 @bhushanshah, ? Dec 19 05:11:05 What's kernel version? Dec 19 05:11:13 3.4.0 Dec 19 05:11:47 I suggest to take a look at hammerhead kernel on halium GitHub (halium-5.1 branch) Dec 19 05:12:08 There are 3 to 4 commits about finishing implementation of pidns Dec 19 05:12:22 Try those commits and see if it works Dec 19 05:12:27 ok Dec 19 05:12:32 If not I'll tell you what commits to revert Dec 19 05:13:24 The changes should be in … ```fs/proc/namespaces.c``` … right? Dec 19 05:14:06 I think more files are affected Dec 19 05:14:13 Check git log Dec 19 05:14:19 Ok Dec 19 05:24:19 @bhushanshah I found 2 commits : https://github.com/Halium/android_kernel_lge_hammerhead/commit/bd221854de33b75db7a7fa01cb34274b62a7cbf8 & https://github.com/Halium/android_kernel_lge_hammerhead/commit/5754614eb43dea44a99e54898e3b83d4d96d8b83 Dec 19 05:26:13 Yeah that should do it I believe Dec 19 05:26:41 Do I have to pick the second one?? That is meant for user namespace... Dec 19 05:27:03 I don't think is needed Dec 19 05:28:06 Btw what device is this? Dec 19 05:28:30 Zuk Z1 (ham, Soc : 801, similar to bacon and hammerhead) Dec 19 05:28:41 Okay Dec 19 05:29:45 Another thing the pid ns implementation is done in the later versions right. In some google groups thread it is said that namespaces is not implemented for the earlier 3.4.xx versions Dec 19 05:30:09 Yeah Dec 19 05:30:35 It's like vendor cherry-pick some patches and not whole feature Dec 19 05:30:41 Sony have also done same Dec 19 05:31:17 So just the verision is LTS but runs code that is from newer releases.. Dec 19 05:31:27 Yeah Dec 19 05:31:45 https://github.com/LNJ2/android_kernel_sony_apq8064/commits/halium-7.1 see 3 reverts at top Dec 19 05:31:55 Ah Dec 19 05:35:33 Kernel compiles fine... Thanks @bhushanshah Dec 19 05:36:39 Yay Dec 19 05:37:25 in … ```mka systemimage``` … : signapk.jar needed by package.apk. Dec 19 05:37:32 ninja build stops Dec 19 05:38:13 For that you need to remove that package from Android.mk Dec 19 05:38:23 in device tree? Dec 19 05:38:24 We don't build java/apk stuff Dec 19 05:38:28 Yeah Dec 19 05:38:29 Ya Dec 19 05:41:03 it is in a txt file (propreitary-files-qc.txt) Dec 19 05:41:13 Should I have to remove all the apk and jar files? Dec 19 09:19:37 I don't think you have to remove it from the text file, but from /vendor/$vendor/$codename/$codename-vendor.mk Dec 19 09:19:54 @JBBgameich, figured Dec 19 09:21:45 I tried to fix the upstart issue: … https://github.com/Halium/hybris-boot/pull/21 Dec 19 09:25:17 Guys I have cloned https://github.com/LineageOS/android_vendor_qcom_opensource_cryptfs_hw in … ```vendor/qcom/opensource/cryptfs/hw``` … but compiler wasn't informed about this path : … ```system/vold/cryptfs.c:74:10: fatal error: 'cryptfs_hw.h' file not found``` Dec 19 09:25:29 um, nuke encryption ? Dec 19 09:25:38 we don't support encrypted data yet Dec 19 09:25:44 Oh ok Dec 19 09:25:55 what should I attack? Dec 19 09:25:56 At least not encrypted using android :) Dec 19 09:26:02 you'll probably have to fix your fstab also, not sure how to go about it Dec 19 09:26:03 Oh ok Dec 19 09:26:18 find your maintainer's "fix encryption" commit :p Dec 19 09:26:53 Or you can just add path in build/core/main.mk and call it a day Dec 19 09:27:10 doesn't harm to build and include it. Dec 19 09:27:58 @bhushanshah, How to? Dec 19 09:28:10 -include ?? Dec 19 09:28:30 Though wait Dec 19 09:28:36 ok Dec 19 09:28:44 vendor should already be in that subdirs Dec 19 09:29:31 vendor/qcom/opensource/cryptfs/hw/ Dec 19 09:30:47 I'm on move right now so difficult to see code , but can you paste full error line? Dec 19 09:31:33 http://paste.ubuntu.com/26214043/ Dec 19 09:31:49 I think it expects some other include directory. Dec 19 09:33:25 I'm adding an include dir in that Android.mk Dec 19 09:33:39 of system/vold/ Dec 19 09:34:20 Yeah that did it Dec 19 09:52:22 halium-install script tries to mount /cache and /data, but /data is not in my /proc/mounts and I can't mount it manually (I supposed it is encrypted). Any ideas on how to do that? Can't I just burn the system.img generated by mka in the /system partition? Dec 19 09:54:17 if the /data partition is encrypted, halium will not be able to -normally- mount the /data partition and hence, the switch_root into the target os will -not- work Dec 19 09:54:43 format (twrp > format /data) your data partition Dec 19 09:55:46 that's another thing, I don't have twrp at the moment. Anyway, I assume ext4? Dec 19 10:09:53 well, depends on whatever filesystem works ( mostly ext4, or f2fs ) Dec 19 10:10:15 just run `mount` via adb and grep for `userdata` Dec 19 10:10:41 ohwait sorry, `/data` not `userdata` Dec 19 10:18:29 I'm not sure what I have to do here. I hacked my way through and I ended up in a userdata partition which is now ext2 and after halium-install contains rootfs.img and system.img Dec 19 10:18:53 🙈 Dec 19 10:19:03 well mounting wont fail at least Dec 19 10:19:06 but I can't get it to boot. Can I flash those 2 manually? Dec 19 10:19:29 rootfs.img and system.img are not flashed Dec 19 10:19:37 They are just put in /data Dec 19 10:20:39 the boot.img contains busybox which does the following: … 1. try to get rndis/ethernet working over USB for telnet ( for debug ) … 2. mount your userdata partition … 3. mount rootfs.img and system.img from your userdata partition somewhere … 4. execute switch_root to switch to your rootfs Dec 19 10:21:02 this is the 1000 feet overview of what it does Dec 19 10:21:31 @trivial_inanity, It doesn't mount system.img but accurate Dec 19 10:21:58 bshah should we document this and telnet debugging in the porting doc? Dec 19 10:22:19 well not that it isn't in /init in the ramdisk yet still Dec 19 10:22:21 It is documented already Dec 19 10:22:32 oops, nvm then 🙈 Dec 19 10:22:38 (at least telnet debugging) Dec 19 10:22:46 But not overview Dec 19 10:22:52 So PR welcome :p Dec 19 10:22:58 overview is very interesting, thanks Dec 19 10:31:00 is the kernel panic reason still the same? Dec 19 10:31:23 you can debug rootfs mounting with hybris-recovery, which pauses boot process before switch Dec 19 10:31:50 it's actually boot.img, not recovery Dec 19 10:33:09 @NotKit correct, I need to move the code on a unit with serial console enabled and will report shortly Dec 19 10:34:42 (Photo, 596x386) https://irc.ubports.com/0K0VpV0o/file_2952.jpg Dec 19 10:34:51 I have flashed the built image and halium-reference-image : I get this in nautilus while booting Dec 19 10:35:54 kllp[m]: Thanks for creating issue Dec 19 10:35:56 @aki237 please check https://docs.halium.org/en/latest/porting/debug-build/early-init.html#debugging-via-telnet Dec 19 10:36:08 for your error you need to add external/elfutils in device manifest Dec 19 10:38:01 In diagnosis.log … ```CONFIG_DUMMY=y found in /proc/config.gz, must be disabled``` Dec 19 10:39:27 you need to edit kernel defconfig to disable it Dec 19 10:39:37 ok Dec 19 10:42:51 kllp[m]: Actually.. best do it in ummm actual halium manifest Dec 19 10:42:54 doesn't hurt Dec 19 10:49:24 @NotKit, Just mka hybris-boot is enough right? Dec 19 10:49:36 yes Dec 19 10:51:32 make: Leaving directory 'BUILDDIR/kernel/sony/msm8974' Dec 19 10:51:32 Kernel Modules not enabled Dec 19 10:51:33 ninja: build stopped: subcommand failed. Dec 19 10:52:42 That's not error Dec 19 10:52:52 Error is somewhere else Dec 19 11:03:23 I have modified my defconfig but the make is not rebuilding the kernel Dec 19 11:07:03 is it the right one? you can try removing out/target/product/$DEVICE/obj/KERNEL_OBJ/.config, then do build and check if new one gets created and has the changes you want Dec 19 11:07:37 @NotKit, So that change doesn't require whole kernel rebuild? Dec 19 11:12:55 Do i need lineage to build halium, or aosp is enought? Dec 19 11:13:29 @aki237, it shouldn't require full rebuild, but build system may handle it in a weird way Dec 19 11:13:47 @NotKit, @NotKit now the phone is in a boot loop Dec 19 11:14:27 * kllp[m] sent a long message: kllp[m]_2017-12-19_11:14:27.txt Dec 19 11:15:15 @aki237 mka hybris-recovery and try to boot with it instead, telnet to device, then check if rootfs is mounted correctly before continuing Dec 19 11:15:25 ok Dec 19 11:15:54 kllp[m]: Umm, looks like python2 and python3 mixup Dec 19 11:16:56 So Arch related. Ubuntu is on py2 right? Dec 19 11:17:34 kllp[m]: ah you're on arch.. I suggest to create a python2 virtualenv and use it before you do source build/envsetup.sh Dec 19 11:18:01 but agian, what you are seeing is not actual error Dec 19 11:18:03 :P Dec 19 11:18:17 actual error is somewhere after that and previous error you showed me Dec 19 11:18:50 @klip: I had the same problem, fixed it like this: https://wiki.archlinux.org/index.php/Python#Dealing_with_version_problem_in_build_scripts Dec 19 11:19:06 @gouster4, In most cases AOSP should work Afaik Dec 19 11:19:53 Dorni[m]: that's ... meh solution, use virtualenv Dec 19 11:20:23 @NotKit, in recovery I'm able to telnet to the device and the mounts are like : http://paste.ubuntu.com/26214571/ Dec 19 11:22:03 is is /target fine? Dec 19 11:23:22 @NotKit, fine in the sense there is a data folder and and I'm able to acces the sdcard through media/0 Dec 19 11:23:40 no, I mean, does it contain normal Linux rootfs? Dec 19 11:23:49 (Halium reference rootfs) Dec 19 11:24:31 Nope Dec 19 11:24:43 then something is wrong with your rootfs install Dec 19 11:25:37 there is a 2G rootfs.img in /target/data/ Dec 19 11:26:11 @aki237, please pastebin output of mount Dec 19 11:26:41 bshah : http://paste.ubuntu.com/26214571/ Dec 19 11:27:15 it seems to be mounted on /target, but /target is empty besides /data folder? Dec 19 11:27:50 yes Dec 19 11:28:48 can rootfs.img be empty for some reason? try mounting it yourself somewhere in initramfs or recovery Dec 19 11:31:34 Yes rootfs is a empty 2Gb partition img with a data folder and lost+found (ext2) Dec 19 11:32:31 I pulled it locally and inspected it. Dec 19 11:34:53 Should I build the systemimage again? Dec 19 11:35:18 Not build Dec 19 11:35:28 But install is where things went sideways Dec 19 11:36:25 Another thing while I was having android (oreo) messed up the sdcard permissions I wasn't able to write to the card at all. Dec 19 11:37:04 @aki237, Oreo messud up the permissions by changing the fs to sdcardfs. Dec 19 11:37:31 Will that be of a deal here? Dec 19 11:40:25 bshah, what exactly is different between the system.img and a normal lineage compilation? stuff in frameworks/base, frameworks/av, system/core, etc? Dec 19 11:47:52 @NotKit I did whatever halium install manually now target has a rootfs stuff. Dec 19 11:48:52 good, does it still boot loop or boots into rootfs? Dec 19 11:49:05 flashing the old boot Dec 19 11:52:08 @NotKit, It doens't boot loop but stuck at boot logo not telnet interface Dec 19 11:53:43 @NotKit, Btw does halium rootfs contain any ui? Dec 19 11:53:59 How would one know the system is up and running? Dec 19 11:54:34 no UI, but you should be able to bring up network interface and connect via SSH Dec 19 11:54:56 @aki237, This is expected? Dec 19 11:55:32 yes, but usb network adapter should be there Dec 19 11:55:42 ? Dec 19 11:56:20 network interface I mean Dec 19 11:56:28 nope Dec 19 12:03:38 It is stuck at boot logo and no network interface yet Dec 19 12:04:13 it won't boot then, hold the power to turn it off Dec 19 12:05:08 does Halium reference rootfs makes any on disk logs, or systemd is in volatile mode by default? Dec 19 12:05:20 volatile Dec 19 12:06:23 * kllp[m] sent a long message: kllp[m]_2017-12-19_12:06:22.txt Dec 19 12:06:59 ... selinux.. did I mention I hate it? Dec 19 12:07:10 kllp[m]: can you point me to your device tree? Dec 19 12:07:37 not online yet Dec 19 12:07:50 well lineage one is fine Dec 19 12:08:13 https://github.com/LineageOS/android_device_sony_z3c Dec 19 12:13:43 ``` Dec 19 12:13:44 Error: could not load context file from BUILDDIR/out/target/product/z3c/obj/ETC/property_contexts_intermediates/property_contexts.tmp Dec 19 12:13:45 ``` Dec 19 12:25:17 @NotKit, Any idea for how to proceed? Dec 19 12:27:44 I would try to edit /etc/systemd/journald.conf in rootfs and temporary set Storage=persisent Dec 19 12:28:08 then check if systemd journal gets created after failed boot Dec 19 12:28:26 does it require android's system.img to start the usb ssh? Dec 19 12:28:32 no Dec 19 12:28:37 ok Dec 19 12:28:51 @NotKit, Where in rootfs itself? Dec 19 12:29:20 yes Dec 19 12:29:25 fine. Dec 19 12:33:20 bshah: Do you think the error is caused by selinux? Dec 19 12:33:36 kllp[m]: yes, but bit occupied right now to point you to solution Dec 19 12:33:42 give me 15 mins Dec 19 12:35:21 of course Dec 19 12:35:41 @NotKit, Where will it be stored? in rootfs.img? Dec 19 12:37:58 yes Dec 19 12:39:43 Nope Dec 19 12:42:16 This is not a kernel driver problem right. cuz we had rndis and telnet functional in ordinary hybris-boot(during an error) and hybris-recovery Dec 19 12:44:10 it might be that systemd fails to operate properly or it can't reenable rndis networking for some reason Dec 19 12:44:52 how do I enable the led indicator through /sys interface? Dec 19 12:48:12 @aki237, have you tried masking/disabling lxc@android service through chroot? Dec 19 12:48:19 can you fetch /proc/last_kmsg? Dec 19 12:48:51 bshah : nope. if you mean rootfs.img/proc/last_kmesg no Dec 19 12:48:56 @bshah, How? Dec 19 12:49:31 @aki237 /proc/last_kmsg after you boot into halium -> reboot into recovery -> get /proc/last_kmsg Dec 19 12:50:59 @bshah> @aki237 /proc/last_kmsg after you boot into halium -> reboot into recovery -, halium-boot.img then {twrp or halium-recovery.img}? Dec 19 12:51:09 twrp Dec 19 12:51:28 Then no file found in /proc Dec 19 12:51:37 hm Dec 19 12:54:43 bshah : wait I'll try again Dec 19 12:55:23 I want to make a service that runs after systemd loads to test it Dec 19 12:55:38 echo 255 > /sys/..../led/brightness Dec 19 12:57:54 Just atleast to test the working of systemd Dec 19 13:02:01 @aki237, Atleast after the sys and tmpfs mounts. Dec 19 13:17:51 How to get the kernel parameters? Dec 19 13:17:58 /proc/cmdline Dec 19 13:19:10 How to proceed further in hybris-recovery.img? (after telnet) Dec 19 13:19:37 echo "continue" thingie mentioned there Dec 19 13:29:15 bshah : after continue, there is an ethernet interface but it's not running the dhcpd... (not assigning me any address) Dec 19 13:30:16 it sounds great I think Dec 19 13:30:38 read guide and assign it IP address in range of 10.15.19.* Dec 19 13:30:51 then ssh root@10.15.19.82 Dec 19 13:33:14 I did that no pings even Dec 19 13:33:51 stranger things Dec 19 13:33:54 :P Dec 19 13:34:44 huh Dec 19 13:39:56 if there is ethernet interface it means the system is alive Dec 19 13:40:28 @NotKit this is in hybris-recovery after continue Dec 19 13:40:42 not in normal hybris-boot Dec 19 13:41:34 weird, hybris-recovery disables rndis when continuing Dec 19 13:46:43 This is the cmdline of the non-debug boot-image : … ```console=tty60,115200,n8 androidboot.hardware=qcom user_debug=31 msm_rtb.filter=0x3b7 ehci-hcd.p``` Dec 19 13:55:45 how to clean and build the kernel again from start? Dec 19 14:00:33 rm -rf KERNEL_OBJ, but unlikely it will help Dec 19 14:01:06 Thinking of building the kernel again.... Dec 19 14:04:33 if the kernel panics then it'll go into a boot loop right else it is running fine? Dec 19 14:13:02 not necessary, but it should reboot on panic Dec 19 14:13:05 newer kernels have pstore Dec 19 14:13:22 does pstore work after hard shutdown? Dec 19 14:13:50 device doesn't reboot Dec 19 14:13:56 @trivial_inanity, newer than 3.4.0 ? :P Dec 19 14:20:16 Which command does the kernel run /init or /linuxrc Dec 19 14:20:49 it firsts starts /init in initramfs Dec 19 14:21:05 which is hybris-boot script Dec 19 14:21:17 Hybris-boot starts /sbin/init from the rootfs then Dec 19 14:21:46 can I change the init to do something? Dec 19 14:29:51 Whats the difference between init in hybris-{boot, recovery}.img ? Dec 19 14:32:12 um, unpack the images and diff ramdisk/init ? Dec 19 14:34:19 Can I edit that init? Dec 19 14:35:30 the initrd script is in hybris-boot/init-script, you can edit it, it's just a shell script Dec 19 14:35:40 ok Dec 19 14:51:59 the var/log is still volatile. no idea how to debug the system Dec 19 15:11:12 Anyone can help me with this: vanyasem: https://paste.pound-python.org/show/sEJ7KbJejL04NB51XqtI/ ? Dec 19 15:11:27 Test_hwcomposer segfault in reference rootfs with caf repos Dec 19 15:13:42 are you sure the libhybris is from caf repo? Dec 19 15:14:00 I think Dec 19 15:14:25 I added the /caf repo, apt update and upgrade Dec 19 15:15:15 apt-cache policy libhybris please Dec 19 15:15:58 Aww, ill boot thaat, brb Dec 19 15:20:45 bshah: https://paste.pound-python.org/show/uUHgRrf31BakZIuhGGJO/ Dec 19 15:21:00 okay Dec 19 15:21:04 that's goo Dec 19 15:21:09 *good Dec 19 15:21:29 ideas :D? Dec 19 15:22:37 theres a service 'display.qservice' died in logcat Dec 19 15:22:54 and some Failed on audit_set_pid with error: Protocol not supported Dec 19 15:23:14 yeah I would take a look at logcat messages when test is running Dec 19 15:23:28 (start telnet in one ssh session and then run test in other one) Dec 19 15:24:25 i get them https://paste.pound-python.org/show/E9OTZVMYAw7UDGS35NOt/ Dec 19 15:24:55 umm Dec 19 15:25:04 it instantly segfaults, after saying OpenGL ES 2.0 (OpenGL ES 3.0 V@100.0 AU@05.00.02.031.018 SKARAJ Dec 19 15:25:21 sounds like auditd/selinux mess Dec 19 15:25:53 selinux is fully disabled Dec 19 15:26:06 if i dont disable it in kernel, systemd doesnt start Dec 19 15:29:44 i could try to disable selinux in my z3c build, too :) Dec 19 15:35:12 also what's this test_hwcomposer: unhandled page fault (11) at 0x00000013, code 0x017 ? Dec 19 15:35:20 is it related to selinux? Dec 19 15:35:45 5.1 doesn't have selinux removal commit as 7.1 afaik Dec 19 15:52:04 @aki237, Any idea for this? Dec 19 15:57:04 you could get network interface though, couldn't you? Dec 19 15:57:20 in hybris-recovery. Dec 19 15:57:25 Yes Dec 19 15:57:32 (after continue) Dec 19 17:35:28 good news - I managed to do some steps forward. I now boot into halium, I can see the message "Welcome to Ubuntu 16.04.2 LTS!" in the logs Dec 19 17:35:59 @Davide, wow Dec 19 17:36:10 ...is that correct? Dec 19 17:37:57 @Davide, how can you get log? Dec 19 17:38:05 via serial console Dec 19 17:38:11 ty Dec 19 17:38:25 You probably don't have that hardware for your device Dec 19 17:38:46 however the unit crash soon after that Dec 19 17:39:19 [ 11.595019] <0>.(0)[1:systemd]Unable to handle kernel paging request at virtual address ffffff7fffffdead Dec 19 17:39:48 That looks like a debug statement Dec 19 17:39:59 By the memory address, that is Dec 19 17:42:18 @Davide, do you see LXC starting in logs? Dec 19 17:42:25 yes Dec 19 17:42:41 [ OK ] Created slice system-lxc.slice. Dec 19 17:43:40 not sure if it's the cause, but you can try to disable it first Dec 19 17:43:57 last action seems to be this one Dec 19 17:44:23 @Davide, Does usb tethering not work? Dec 19 17:45:21 using recovery image it usb tethering works, but I have boot loops using normal boot Dec 19 17:46:44 systemctl mask lxc@android.service in chroot Dec 19 17:47:08 the last action seems to be [ 11.570905] <2>.(2)[1:systemd]systemd[1]: Starting Remount Root and Kernel File Systems... Dec 19 17:47:58 [ 11.593559] <0>.(0)[1:systemd]BUG: failure at /Code/halium/kernel/.../kernel/sched/rt.c:111/init_rt_rq()! Dec 19 17:48:41 just curious, did you use similar kernel config as for Debian on the device? Dec 19 17:49:10 reference Halium rootfs is basically Ubuntu 16.04 + LXC container + libhybris Dec 19 17:49:24 @NotKit yes, and I have systemd running on Debian Dec 19 17:55:08 so Android LXC container is pot even started at this stage, puzzles me why it would crash Dec 19 17:55:46 Do you have full dmesg? Dec 19 17:59:09 i'll collect it Dec 19 18:07:16 When I run the `mer_verify_kernel_config` script on the .config file in the out dir after building the kernel, it says that some values are unset, even though I set them in the corresponding defconfig. Anything I overlooked? Dec 19 18:11:46 depends on exact values, some options it checks might be completely missing in newer kernels Dec 19 18:12:48 Thank goodness, there's finally an official TWRP build for my device! Dec 19 18:13:36 There are 4 values he says are missing. 3 of them seem optional, the fourtht is `CONFIG_SECURITY_SELINUX_BOOTPARAM`. Though it shouldn't matter as `CONFIG_AUDIT` is set to n Dec 19 18:14:04 @UniversalSuperBox, RIP rootstock issues Dec 19 18:14:06 I just thought this could be the cause of my non-booting on my phone. Time to continue looking Dec 19 18:14:23 @bhushanshah, Somehow I still think that their Busybox is going to be broken Dec 19 18:17:34 It worked Dec 19 18:37:55 got a full dmesg log here https://pastebin.com/L62WuzZ2 Dec 19 18:50:08 Davide was Debian kernel compiled by toolchain from MTK AOSP tree? Dec 19 18:51:15 I would test if it crashes the same way in Ubuntu or not Dec 19 18:51:59 yes, I used standard MTK toolchain Dec 19 18:52:29 unless they changed format from MTK6795, https://github.com/vianney/bootimgtool can be used to quickly unpack/repack boot.img Dec 19 18:52:44 you can check hybris-boot initramfs with your working Debian kernel Dec 19 18:52:59 CONFIG_DUMMY check can be skipped with "echo continue > /init-ctl/stdin" in telnet Dec 19 18:55:56 i can try that, but I don't have CONFIG_VT in debian kernel, would that work anyway? Dec 19 18:57:41 it should be required to start systemd if Debian runs with it Dec 19 18:58:11 I will give it a try Dec 19 19:31:52 Could somebody take a look at my kmsg log? I suspect selinux due to the avc denied messages, but I don't know if that is the problem Dec 19 19:31:52 https://paste2.org/5OxyfzpF Dec 19 19:36:14 hm but isnt avc about fine grained power management? Dec 19 19:39:24 are you sure that the values do not occur twice in your config? Dec 19 19:39:40 When I searched the errors ('avc denied'), it turned up several pages about selinux Dec 19 19:39:42 https://wiki.gentoo.org/wiki/SELinux/Tutorials/Where_to_find_SELinux_permission_denial_details#Disecting_the_AVC_denial Dec 19 19:40:33 @anpok did your PR on ubports-boot work correctly when you used it? Dec 19 19:40:59 the initramfs? Dec 19 19:41:02 Yeah Dec 19 19:41:55 it got me into the init of the ubuntu.img Dec 19 19:42:34 where it hangs somewhere Dec 19 19:42:38 Ah. For whatever reason, udev doesn't seem to be firing correctly once I hit the system Dec 19 19:43:45 oh? Dec 19 19:44:01 Yeah. I'm just not sure if it's me or the initramfs. Dec 19 19:44:42 because atm I cannot get beyond the point of usc starting Dec 19 19:44:48 it just starts and sits there.. Dec 19 19:44:57 You have graphics? Dec 19 19:45:11 will try without my change.. since I dont need the extra waits anyore Dec 19 19:45:13 yes Dec 19 19:45:18 well blacks Dec 19 19:45:23 .. it blanks the screen Dec 19 19:45:34 Anything in its logs? Dec 19 19:45:35 and I have hwc / egl exxtensions in the logs Dec 19 19:45:47 but no indication of a unity8 session being started Dec 19 19:46:07 Do you have anything in /firmware? Dec 19 19:46:34 no ssh connection .. Dec 19 19:48:04 Ah, yep, it is the new initramfs that's doing it Dec 19 19:48:39 Unity welcomes me to the world when I use the old one Dec 19 19:49:00 Just checked, there are no duplicate entries in the defconfig Dec 19 19:49:41 @UniversalSuperBox tzfingerprint files.. Dec 19 19:49:43 I don't get it. It cannot be that hard to build for the Nexus 5... Dec 19 19:50:39 ```$ status lightdm … status: Unknown job: lightdm``` Dec 19 19:50:40 wat Dec 19 19:52:59 While I was sitting here trying to troubleshoot it, Unity was running the whole time. Dec 19 19:53:06 It had just put the screen to sleep. -_- Dec 19 19:55:02 I thought it was mounts, but that doesn't seem to be it. Dec 19 19:56:29 so the theory would be that one of the initramfs tools is linked to the wrong binary and thus does nonsense Dec 19 19:56:50 Doesn't everything just come from busybox? Dec 19 19:59:49 it does in hybris-boot Dec 19 20:02:50 @UniversalSuperBox odd .. most bins are not busybox https://paste.ubuntu.com/26216988/ Dec 19 20:03:26 I linked all of the 6bfa53ceed8f7991609159652d4356cc to chgrp .. andleft the copies of busybox alone Dec 19 20:03:47 which is only chroot egrep mount sh sleep .. Dec 19 20:05:19 hm I switch to ubports-boot master build ubports-boot again.. the outcome did not change Dec 19 20:07:45 I think I figured out why ubports-boot breaks hybris-boot though Dec 19 20:07:50 We used the same variable names Dec 19 20:13:37 @anpok, Indeed Dec 19 20:14:48 wait Dec 19 20:16:00 Yeah, most of them are busybox Dec 19 20:16:31 ```# chgrp … BusyBox v1.22.1 (Ubuntu 1:1.22.0-9ubuntu1) multi-call binary.``` Dec 19 20:19:28 hmm If I unzip the resulting binary Dec 19 20:19:38 and run cpio -i < blub .. Dec 19 20:20:20 then the resulting md5sums of the linked files are the same.. Dec 19 20:20:21 The md5sums match in the old rootfs. Something tells me we're chasing a red herring Dec 19 20:20:36 ayyy Dec 19 20:20:41 are the relative links broken? Dec 19 20:21:04 I'll have to get it. One moment. Dec 19 20:25:46 https://paste.ubuntu.com/26217068/ Dec 19 20:25:55 Looks ok ¯\_(ツ)_/¯ Dec 19 20:26:01 hm interesting the bigger files like chgrp are statically linked .. while the other ones pointing to a dynamically linked busybox Dec 19 20:26:31 lxc@android wants to make an 'android' directory in /sys/fs/cgroup/net-cls/lxc/ but it fails can someone point me in a direction so i could see if i can fix this? Dec 19 20:28:38 @anpok, Oh, I see. Are you going to try to change those links to the other busybox? Dec 19 20:29:37 maybe there is a reason for that? Dec 19 20:29:51 Oh, what am I saying. That's how it is. Dec 19 20:32:17 So .. my understanding so far is that the resulting initramfs produces the same tool binaries as those found in the initramfs.gz that is currently inside the repository Dec 19 20:32:42 Then it has to be something I'm doing wrong. Dec 19 20:35:35 hm the existing links in /sbin are absolute Dec 19 20:35:48 It werks? Dec 19 20:35:50 Hmmm Dec 19 20:35:54 Wipe it and see why! Dec 19 20:39:49 After a wipe, `/firmware`\ is still mounted correctly Dec 19 20:41:46 I don't know what I did wrong Dec 19 20:42:15 so you dont think there is a problem? Dec 19 20:42:41 I can only assume it was me. I'll rebuilt the initrd Dec 19 20:44:53 It was me! I hadn't flashed the initrd back to the new one Dec 19 20:45:11 Now `/firmware` is a broken link again Dec 19 20:53:11 hmm why did I change that into links in the first place? Dec 19 20:55:02 nono Dec 19 20:57:03 no I was asking why i changed the binaries inside the repository to be links .. there was no technical reason.. Dec 19 20:57:15 Git can't handle hardlinks Dec 19 21:00:36 Okay, getting closer. It's definitely something with the mount_android_paritions() function Dec 19 21:09:24 ```# ls /dev/disk … ls: /dev/disk: No such file or directory``` Dec 19 21:09:29 And there's the problem @anpok Dec 19 21:09:53 Maybe. :P Dec 19 21:18:13 hm I made change .. Dec 19 21:18:25 now using absolute links in sbin Dec 19 21:18:44 I'll rebuild and see Dec 19 21:35:51 It all comes back to udev, doesn't it Dec 19 21:36:20 lib/udev/rules.d/60-persistent-storage.rules isn't being fired for some reason Dec 19 21:53:42 I figured it out Dec 19 22:11:01 @anpok any chance I can get you to test that change? Dec 19 22:12:31 where is the change? Dec 19 22:12:39 https://github.com/ubports/ubports-boot/pull/7 Dec 19 22:12:43 I pushed it to your branch. :P Dec 19 22:13:19 oh thx .. will try now Dec 19 22:16:00 (Photo, 2560x1920) https://irc.ubports.com/CmdxcqlT/file_2968.jpg Dec 19 22:16:04 It works for me. :) Dec 19 22:26:55 it does not not work .. I see usc and sshd starting .. but still fail to connect via ssh and the usc-spinner is not being launched Dec 19 22:27:06 @UniversalSuperBox, is there turkish 😂 Dec 19 22:27:29 @anpok, So... same as with the prebuilt? Dec 19 22:28:17 yes Dec 19 22:28:25 @sopernam, What's its native name? Dec 19 22:28:30 @anpok, Cool. I ship it. Dec 19 22:31:35 @UniversalSuperBox, @UniversalSuperBox what is this phone? It seems a Nexus Dec 19 22:31:46 @Mattia, Moto G5 Plus Dec 19 22:32:02 As evidenced by the giant "moto" logo above the screen. :P Dec 19 22:32:12 which I hate Dec 19 22:33:48 (Photo, 2560x1920) https://irc.ubports.com/OwVu0EfP/file_2970.jpg Dec 19 22:33:54 @sopernam, yes. Dec 20 02:20:50 Is the ueventdrc* files are created after the lxc is started? **** ENDING LOGGING AT Wed Dec 20 03:00:02 2017 **** BEGIN LOGGING AT Wed Dec 20 03:00:02 2017 Dec 20 03:38:38 @NotKit after the continue, my lsusb yields … ```Bus 001 Device 107: ID 18d1:d001 Google Inc. Nexus 4 (fastboot)``` … . But fastboot devices returns nothing. Dec 20 04:11:44 Guys I made a small shell script in /target/init-debug. After reboot when I ls in /target I get this … ```ls: ./init-debug-stderrout: I/O error``` … and the root fs contents Dec 20 05:00:07 Guys I edited the init script so that it wont stop telnetd and Dec 20 05:00:13 udhcpcd Dec 20 05:01:53 I'm unable to call any binaries but … ```cd ``` … shows me all the directories.... Dec 20 05:02:01 I think systemd is running Dec 20 05:02:25 source etc/environment Dec 20 05:04:17 @bhushanshah, etc... not found. I think some how I have to change the … ```/proc/[pid]/root``` … of the current shell Dec 20 05:04:31 Which is impossible without the busybox Dec 20 05:05:01 or other utilities Dec 20 05:05:38 I tried LD_LIBRARY_PATH and running the /proc/1/cwd/bin/ls... no luck Dec 20 05:09:38 @bhushanshah This is the output of ls in /proc https://paste.ubuntu.com/26219201/ Dec 20 05:10:19 That's umm useless shell Dec 20 05:10:57 I just want to see whether the systemd is running if so I'll add a telnetd service in rootfs and connect to it then. Dec 20 05:10:58 Would be lovely to figure out if a) systemd started b) if not why it crashed Dec 20 05:11:14 If it crashes it would have restarted right? Dec 20 05:11:33 Yeah it would boot loop Dec 20 05:11:42 mine isnt bootlooping. Dec 20 05:12:01 I suggest to figure out/research why you don't get /proc/last_kmsg Dec 20 05:12:21 Because it is possible that you might be missing some kernel option Dec 20 05:12:25 If I can rewite shell and use that .. Dec 20 05:12:29 HELL YEAH Dec 20 05:13:01 I'll write a custom shell and test it... I'll add some inbuilt functions to atleast read a file. Dec 20 05:13:13 Okay Dec 20 05:33:48 How can I change the root directory of the current shell process? Dec 20 05:36:38 Which file should I read to get the log? Dec 20 05:36:56 in /dev, /sys, /proc/, etc., Dec 20 05:39:28 Guys can any one post the output of /proc/1/cmdline Dec 20 05:40:51 Can do in bit, not on laptop right now Dec 20 05:41:06 Thx,... I'll wait Dec 20 05:48:07 YES systemd is running I think... Dec 20 05:48:22 catting the /proc/1/comm gives me systemd Dec 20 05:53:54 @bhushanshah : this is the output of /proc/1/status https://paste.ubuntu.com/26219349/ Dec 20 06:01:39 Where will the android lxc be mounted? Dec 20 06:07:38 what should the /etc/fstab contains in the rootfs? Dec 20 06:11:59 fstab is mostly empty Dec 20 06:18:06 @aki237 cat /proc/1/cmdline gives this : /sbin/init--log-target=kmsg Dec 20 06:18:36 no spaces between the /sbin/init and —log... ? Dec 20 06:18:39 and fstab is empty Dec 20 06:18:43 no no spaces Dec 20 06:18:46 Same here. Dec 20 06:18:53 But well Dec 20 06:19:06 Atleast systemd is running I think. Dec 20 06:19:19 As mentioned above, I would investigate why /proc/last_kmsg is not there Dec 20 06:19:22 I didnt kick off any of the services. Dec 20 06:20:19 It seems the initrd doesn't store the last_kmesg I think Dec 20 06:20:59 It's not initrd function at all Dec 20 06:21:05 Oh Dec 20 06:21:07 it's function from kernel itself Dec 20 06:21:14 where does it store it anyway? Dec 20 06:21:26 On memory itself Dec 20 06:21:39 no no. across boots.. Dec 20 06:21:55 Yes, on memory itself on soft-reboots Dec 20 06:22:01 on hard-reboot it doesn't survive Dec 20 06:22:09 at least for 3.4 kernels it's there Dec 20 06:22:25 hard-reboot in the sense, long press power button? Dec 20 06:22:31 Can you check if you have CONFIG_ANDROID_RAM_CONSOLE or something like that turned on? Dec 20 06:22:41 yes it is on Dec 20 06:22:56 hard-reboot as-in poweroff device completely and start it again Dec 20 06:23:57 @aki237, This reboot the device Dec 20 06:24:53 yes, but that is soft reboot Dec 20 06:25:08 it doesn't "power off" device completely Dec 20 06:25:18 I think that is hard reset Dec 20 06:25:28 in my case. it powers off and then turns on Dec 20 06:25:45 Can you try this? Dec 20 06:25:49 Boot into recovery Dec 20 06:25:57 > adb shell reboot recovery Dec 20 06:25:59 ok Dec 20 06:26:03 ok Dec 20 06:26:08 and then in recovery check if you have /proc/last_kmsg Dec 20 06:26:58 Just say me this. How will I reboot the device without reboot command... you know my writing someting to a device Dec 20 06:27:03 or something like that Dec 20 06:27:22 no idea really :P Dec 20 06:27:35 ok Dec 20 06:27:36 I always just reboot with reboot command that's it Dec 20 06:28:34 No file or directory Dec 20 06:28:50 hmm Dec 20 06:29:02 @bhushanshah, recovery -> adb reboot rec... -> cat /proc/last_kmsg no file or directory Dec 20 06:29:27 it makes debugging really hard :/ Dec 20 06:29:44 Have you managed to get journalctl on persistent storage? Dec 20 06:29:50 I can run the cat /proc/kmsg in background and continue then can see something Dec 20 06:29:59 aki237: Are you testing 5.1 or 7.1? Dec 20 06:30:01 nothing in var/log Dec 20 06:30:04 7.1 Dec 20 06:30:14 Ah there's no /proc/last_kmsg there Dec 20 06:30:25 Try /sys/fs/pstore/console-ramoops or /sys/fs/pstore/console-ramoops-0 Dec 20 06:30:30 Herrie: well for 3.4 kernel there should be Dec 20 06:30:35 At lesat that's my understanding Dec 20 06:30:44 no /sys/fs/pstore Dec 20 06:30:53 aki237: Ah OK :S Dec 20 06:31:38 aki237: Issue I had when I got no log at all was that I had Android 6 on my Mido, but was trying Halium/Android 7 kernel Dec 20 06:31:43 And that didn't work Dec 20 06:31:56 Once I had Android 7 on my target I magically started to get logs Dec 20 06:32:46 Took me a while to figure this out :P Dec 20 06:32:58 AAAHHHH Atlast go the error... Dec 20 06:33:29 @aki237, This did it Dec 20 06:33:43 ```<27>[ 98.912151] systemd[1]: Failed to determine whether /sys/fs/cgroup is a mount point: Too many levels of symbolic links … <27>[ 98.912318] systemd[1]: Failed to mount cgroup at /sys/fs/cgroup/systemd: No such file or directory … <24>[ 98.912718] systemd[1]: Freezing execution.``` Dec 20 06:33:53 yay Dec 20 06:35:31 https://github.com/Halium/android_kernel_lge_hammerhead/commit/25437b2a54dd619a96e268ecaf303b089aa785e4.patch Dec 20 06:35:34 apply this patch Dec 20 06:35:41 and hopefully things will magically work Dec 20 06:36:01 Implementing cgroups? Dec 20 06:36:02 I think you can scrap changes in init-script in hybris-boot Dec 20 06:36:22 no, there is bug, and that patch fixes it Dec 20 06:36:23 @bhushanshah, ? Dec 20 06:37:26 Man that happened to me when I was trying to read using shell inbuild read and while from /proc/x/fd.. I just thought it was symlinking to the initrd root. Dec 20 06:38:04 @bhushanshah, The changes you have made to not kill telene/udhcpd Dec 20 06:38:19 Change them back? Dec 20 06:38:21 Sure. Dec 20 06:49:48 aki237: Yeah we had this issue on some of our LuneOS target and we patched it Dec 20 06:50:15 Oh.. Lune OS.... Googling... Dec 20 06:51:35 Don't remember 100% sure how, could be these: https://github.com/Halium/android_kernel_lge_hammerhead/commit/bd221854de33b75db7a7fa01cb34274b62a7cbf8 and https://github.com/Halium/android_kernel_lge_hammerhead/commit/5754614eb43dea44a99e54898e3b83d4d96d8b83 Dec 20 06:51:59 Or this systemd patch: https://github.com/webOS-ports/meta-webos-ports/blob/pyro/meta-luneos/recipes-core/systemd/systemd/fd_fdinfo_mnt_id_disablefdinfostat.patch ? Dec 20 06:52:12 https://github.com/Halium/android_kernel_lge_hammerhead/commit/25437b2a54dd619a96e268ecaf303b089aa785e4.patch <- Dec 20 06:52:18 but yes, last one Dec 20 06:52:25 @Herrie, I did these two Dec 20 06:52:59 Herrie: actually the patch I pointed is better solution to the last one you pointed Dec 20 06:53:03 your solution is hack :p Dec 20 06:53:38 fd_fdinfo_mnt_id_disablefdinfostat.patch should be replaced with 25437b2a54dd619a96e268ecaf303b089aa785e4.patch Dec 20 06:53:59 bshah: Will discuss with Tofe. Dec 20 06:54:18 bshah: Getting newer systemd to work is difficult at times Dec 20 06:54:30 So we're happy when stuff works and didn't look at the kernel source :P Dec 20 06:54:48 I don't disagree :p Dec 20 06:55:13 bshah : the patch you mentioned... full kernel build or obj build? Dec 20 06:55:15 nevermind Dec 20 06:55:18 It's a lot easier to find something to patch in systemd compared to kernel tree ;) Dec 20 06:56:19 Herrie : but when people want to change the init.... (because I hear people like nuking systemd :P) Dec 20 06:58:11 bshah : I have did the udev rules step in recovery(twrp) itself.. I hope it's ok? Dec 20 06:58:45 udev rules is not something to worry about now Dec 20 06:58:52 ok Dec 20 07:02:31 WOHOOOOO 10.15.19.82 Dec 20 07:14:15 well Dec 20 07:14:24 first check `systemctl status lxc@android` Dec 20 07:17:34 failed Dec 20 07:18:18 any more log? Dec 20 07:19:32 lxc fail log after restart https://paste.ubuntu.com/26219642/ Dec 20 07:20:05 note: never manually restart lxc container Dec 20 07:20:11 ok Dec 20 07:20:21 The first didnt give me any logs. Dec 20 07:20:50 can you check lxc-checkconfig? Dec 20 07:21:01 and fix any problems in first two sections of it? Dec 20 07:21:06 user name space missing. Dec 20 07:21:18 that is not issue, other then that?/ Dec 20 07:21:37 https://paste.ubuntu.com/26219649/ Dec 20 07:23:03 (Photo, 677x890) https://irc.ubports.com/Tm3aayo5/file_2982.jpg With colors Dec 20 07:23:11 should work :/ Dec 20 07:23:30 But, can you do a reboot, and then provide me systemctl status output? Dec 20 07:27:19 this had may other things similar to no permission (omitted it) ... https://paste.ubuntu.com/26219666/ Dec 20 07:29:53 ymm Dec 20 07:30:06 you tried to restart lxc@android again? Dec 20 07:30:15 no Dec 20 07:30:27 Will try the after rebooting and send you the log Dec 20 07:30:52 yeah, please reboot Dec 20 07:33:10 Now it reboots. Dec 20 07:33:32 apparently target (rootfs) is empty Dec 20 07:34:04 :P Dec 20 07:34:23 It happened because you did reboot, I think best would be to re-install (halium-install) Dec 20 07:34:30 Yes Dec 20 07:35:14 @bhushanshah, reboot the lxc or system reboot? Dec 20 07:35:40 btw I'm still using hybris-recovery.img if that is of any use Dec 20 07:38:53 system reboot Dec 20 07:39:11 never ever restart android lxc, android doesn't like it Dec 20 07:39:17 ok Dec 20 07:46:17 https://paste.ubuntu.com/26219723/ Dec 20 07:46:36 Thats all journalctl gave me systemctl status was even smaller that this Dec 20 07:47:16 output is not useful, can you just give me `systemctl status` output? Dec 20 07:47:53 https://paste.ubuntu.com/26219732/ Dec 20 07:48:27 ummmm just `systemctl status` Dec 20 07:48:29 not lxc@android Dec 20 07:48:30 oh ok Dec 20 07:49:19 https://paste.ubuntu.com/26219737/ Dec 20 07:50:56 hmm Dec 20 07:51:04 @UniversalSuperBox, tr turkish Dec 20 07:52:04 @aki237 is selinux enabled by any chance? Dec 20 07:52:22 can you give full dmesg or journalctl? Dec 20 07:54:00 (Document) https://irc.ubports.com/RHDsYT4t/file_2983 Dec 20 07:54:06 (Document) https://irc.ubports.com/CO4AZ55L/file_2984 Dec 20 07:55:50 Damn it I think selinux is enabled. Dec 20 07:56:25 @bhushanshah https://paste.ubuntu.com/26219755/ Dec 20 07:58:30 Even my kernel config has it. I thought the check-config script would have weeded it out. Dec 20 07:58:56 CONFIG*SELINUX* should be removed right from the kernel? Dec 20 07:59:05 check-config script just checks it doesn't remove it Dec 20 07:59:09 oh Dec 20 07:59:16 @aki237, ? Dec 20 08:00:29 Do we need TOMOYO? Dec 20 08:00:34 yeah, well you can remove it or just pass androidboot.selinux=permissive in kernel parameter Dec 20 08:00:40 or selinux=0 Dec 20 08:00:41 ok Dec 20 08:00:52 I'll remove it... That is a pain in the butt. Dec 20 08:01:46 What are the CONFIG_SECURITY_* should I remove? there is TOMOYO, YAMA, SMACK Dec 20 08:02:24 they can stay Dec 20 08:02:27 ok Dec 20 08:02:56 you can follow the hammerhead kernel config Dec 20 08:03:01 ok Dec 20 08:32:12 ah selinux=enforcing is in my boardconfig Dec 20 08:32:38 @anpok, @bhushanshah Should I remove this too? Dec 20 08:32:56 also I see filters in makefiles that would filter out selinux=permissive and =disable Dec 20 08:34:14 @bhushanshah, still the same error. Dec 20 08:39:39 Hmm.. not entirely sure what to do at this moment tbh Dec 20 08:39:46 ok Dec 20 08:40:01 I am out of ideas at moment, will try to help you later when I get free Dec 20 08:40:05 ok Dec 20 08:40:09 Thanks anyway Dec 20 08:52:05 apparmour denies dhcpd from opening /etc/hybris.../dhcpd.conf.. is there anyway I can fix that context (selinux :?) ? Dec 20 10:03:35 Guys when trying to power on the wifi (through /dev/wl...) it says bad address after some time. Dec 20 11:49:31 @bhushanshah There are some device specific rules found under /usr/lib/lxc-android/ like … ```70-bacon.rules``` … is that necessary for the lxc to start? Dec 20 11:49:57 no Dec 20 11:50:04 huh Dec 20 11:50:05 they are not necessary Dec 20 11:50:18 They are needed for tests but not container Dec 20 11:50:20 Also Dec 20 11:50:26 Wifi won't work without container Dec 20 11:50:35 ok Thats a relief. Dec 20 11:53:35 Is there anyway we can update the lxc version? Dec 20 11:53:38 @bhushanshah, Works here xD Dec 20 11:54:01 @aki237, why though? :p Dec 20 11:54:14 Just asking. Dec 20 11:56:19 @aki237, probably enabling xenial backports Dec 20 11:56:38 oh Dec 20 12:01:51 @bhushanshah can I run the lxc's Exec command from command line? Dec 20 12:03:19 yes you can but you need to disable lxc@android service first and then reboot and then run it Dec 20 12:03:31 it's important to disable lxc@android first Dec 20 12:03:45 I can run only once per boot? Dec 20 12:05:01 yes Dec 20 12:05:17 Is that due to the pre-start hook script or what? Dec 20 12:06:37 yes and no Dec 20 12:06:51 well, it's both pre-start script and android not likeing being started twice Dec 20 12:07:33 Another thing android is being run in lxc why does it delete my rootfs? Dec 20 12:07:59 Well.... as said you are not supposed to run it twice Dec 20 12:08:10 ok Dec 20 12:08:54 Ah in pre-start script it deletes the rootfs and extracts from system.img which removes the rootfs.img too.. Dec 20 12:09:04 Chicken in the egg situation. Dec 20 12:10:01 yep, well essentially it cleans rootfs -> extracts rootfs -> bind mounts /data -> start container -> fails Dec 20 12:10:06 but now when you restart Dec 20 12:10:10 it doesn't unmount Dec 20 12:10:15 yeah. Dec 20 12:10:31 should be fixed easily ... but mmeh Dec 20 12:10:36 it doesn't work further anyway Dec 20 12:11:10 Devs meant it be called once and called right... Bulls eye? Dec 20 12:11:36 yeah Dec 20 12:11:50 Should be fixed. Dec 20 12:11:51 basically this is undefined behaviour :p Dec 20 12:12:06 or well more or less.. unsupported behavior Dec 20 12:12:18 just for failsafe. Dec 20 12:12:35 yeah... I will try to fix it Dec 20 12:13:20 I was in lxc irc they seem to say that this seems to be a kernel bug. Can you look at this and say if you've handled this before? https://paste.ubuntu.com/26220788/ Dec 20 12:13:31 @bhushanshah, I can help Dec 20 12:14:14 Do you have iRC log? or something? Dec 20 12:14:26 I do. Dec 20 12:17:58 https://paste.ubuntu.com/26220857/plain/ Dec 20 12:18:14 It's not being logged. I have to copy from weechat Dec 20 12:19:52 hm okay Dec 20 12:20:35 possibly what you can try is check hammerhead kernel and see log/history of pidns implementation and see if you are missing any fixes Dec 20 12:20:46 ok Dec 20 12:42:20 @bhushanshah, @bhushanshah Apparentely that is related to this last patch Dec 20 12:46:04 Maybe maybe not? Dec 20 12:46:32 Why you think so? Dec 20 12:47:09 The kernel oops says so in the function : … ```proc_ns_follow_link``` Dec 20 12:48:31 well surely that patch fixes at least systemd startup so there maybe some other patch you are missing Dec 20 12:49:57 Yep. Atleast able to find the point of fault 😅 Dec 20 13:21:39 If I edit a source file in kernel will it build everything or just build the obj of that modified source and relink it? Dec 20 14:37:13 @bhushanshah I got where the oops occur : https://github.com/LineageOS/android_kernel_cyanogen_msm8974/blob/cm-14.1/fs/proc/namespaces.c#L127 Dec 20 14:38:43 Which can be fixed by first commit https://kernel.opensuse.org/cgit/kernel/log/?h=v3.8.5&showmsg=1&qt=grep&q=nd_jump_link which inturn needs the second Dec 20 18:45:47 does anyone know how to get Cell signal strength info (dBm) in UT phones? Dec 20 23:16:19 where and how in the ubports rootfs is configured that datapartition/system-data and user-data is used as overlay? Dec 20 23:17:02 @anpok, All in the initrd Dec 20 23:17:11 That's what makes ubports-boot really nice. :/ Dec 20 23:18:14 Nothing is mounted after you hit `upstart`. Can't break the overlay because if you're in the system, the overlay probably worked right. Dec 20 23:18:32 Check out the `touch` script in ubports-boot Dec 20 23:18:40 ah Dec 20 23:18:56 It seems to me that some init scripts try to create /var/run Dec 20 23:19:08 and I dont see that in system-data/ Dec 20 23:19:23 nor mounted as tmpfs .. Dec 20 23:21:15 or rather write to var/run ... Dec 20 23:25:20 Nope, because root changes to /root when you hit the end of init. Dec 20 23:26:02 So that /var/run is outside of the running system. It doesn't exist. Dec 20 23:33:18 i.e. dbus.conf creates a pid file in /var/run/dbus/ Dec 21 00:15:41 hm but there is also a mounted-var script that tries to migrate var run over to run .. ok no idea what could be wrong there.. Dec 21 00:24:48 I have an idea Dec 21 00:25:09 Does it just `mv /var/run /run`? Dec 21 00:28:32 it cleares /var/run and creates symlink to /run inside var Dec 21 00:29:04 assuming that /var is writable at that point in time - but I am not sure if it is triggerd .. Dec 21 00:29:18 'start on mounted MOUNTPOINT=/var' Dec 21 00:34:21 ... hmm I think not much Dec 21 00:34:26 is working Dec 21 00:34:55 7.1? Dec 21 00:35:00 yes .. Dec 21 00:35:20 repowerd does not start ... most or no dbus service seems to be claimable .. Dec 21 00:35:47 lighttdm fails for various reasons .. usc is started but loses the connection to lightdm Dec 21 00:36:05 sshd starts but gets stuck once it tries to handle a session Dec 21 00:45:16 @anpok, You have no /var/run? **** ENDING LOGGING AT Thu Dec 21 03:00:00 2017 **** BEGIN LOGGING AT Thu Dec 21 03:00:01 2017 Dec 21 04:46:04 Can any one point me to halium-7.0 branch of android_kernel_lge_hammerhead (Nexus 5 kernel tree) Dec 21 04:50:29 There is not Dec 21 04:50:36 ok Dec 21 04:53:08 @bhushanshah https://paste.ubuntu.com/26225178/ the kernel code that causes oops with the debug output : https://paste.ubuntu.com/26225180/ Dec 21 05:09:24 @bhushanshah namespaces source code is device independant right? Dec 21 05:09:36 Yeah it is Dec 21 05:10:10 Half assed back ports.. namespace contains functions that is not found in the source tree even in 3.4 Dec 21 05:10:55 Ooh wait Dec 21 05:11:06 You did 3 reverts? Dec 21 05:11:14 ? Dec 21 05:11:21 Okay most likely not Dec 21 05:12:02 https://github.com/LNJ2/android_kernel_sony_apq8064/commits/halium-7.1 Dec 21 05:12:20 Chances are that you will need 3 reverts in this git history Dec 21 05:12:42 revert (vfs, proc, proc)?? Dec 21 05:12:50 Yeah Dec 21 05:12:53 Thanks. Dec 21 05:13:17 Though that reverts were build fixes so shrug Dec 21 05:13:24 Am I the crazy guy debugging using printk? Dec 21 05:13:42 I hate when vendors like to backport things without fixing shit Dec 21 05:14:01 Just because they dont use that in the config? Dec 21 05:14:25 Yeah Dec 21 05:14:33 They never test make allconfig Dec 21 05:14:48 Yeah. Dec 21 05:21:11 github sucks.... unable to search code in forked repos. Dec 21 05:21:30 git grep is your friend :p Dec 21 05:21:54 @bhushanshah Thats the point... I dont want to clone the whole source tree. Dec 21 05:24:14 Oh well Dec 21 05:24:20 OpenSuSE's git browse is awesome and elixir of free electrons too. Dec 21 05:25:22 @bhushanshah, why would they Dec 21 05:25:37 all they need is to ship some crap that will get obsolete in a year Dec 21 05:25:56 @vanyasem, Well.. they should before push. Dec 21 05:26:09 @vanyasem, they don't even test hardware properly most of the time Dec 21 05:26:33 like... htc 10's screen was making noise sounds because the `cable was overloaded` Dec 21 05:26:36 how absurd is that? Dec 21 05:26:55 @vanyasem, What...? Dec 21 05:27:02 they released an update lowering the maximum brightness, how you like that fix? Dec 21 05:27:14 now led gets less power and the cable is not overloaded Dec 21 05:27:33 couldn't they test their shit before shipping it? Dec 21 05:41:20 @bhushanshah, Were these commits reverted in hammerhead or It didnt have that junk beforehand? Dec 21 05:44:46 Hammerhead kernel was sane Dec 21 05:45:07 Because it was maintained by google? Dec 21 05:45:22 I guess... Dec 21 05:55:56 Morning! Dec 21 05:56:20 Quick question: Do we need Doze to be build and included in Android bits for Halium? Dec 21 05:58:11 This is for a OnePlus X target ;) Dec 21 06:03:51 @bhushanshah Atleast the lxc process is not getting dumped now. it gives me some log... Dec 21 06:04:02 (After those reverts) Dec 21 06:05:56 @bhushanshah lxc's /var/log/lxc/android.log give me this https://paste.ubuntu.com/26225497/ Dec 21 06:10:16 Herrie: not sure my last message made it or not but answer is no Dec 21 06:13:10 @aki237 lxc-start 20171220113913.995 ERROR lxc_start - start.c:preserve_ns:153 - No such file or directory - Kernel does not support attaching to mnt namespaces. Dec 21 06:13:24 Damn. Dec 21 06:13:51 (which looks like lie) Dec 21 06:14:05 actually that is not error Dec 21 06:14:14 The reverts you said removed the mount namespace handling I think Dec 21 06:14:21 (at least not fatal) Dec 21 06:14:36 @aki237 can you grab dmesg? Dec 21 06:14:45 https://github.com/LNJ2/android_kernel_sony_apq8064/commit/b3bbbf6ce0bd76076e4e02bc7a2de90ea523535e#diff-3b35875fb9a81612b098b440bddcf5a8 Dec 21 06:14:55 bshah : sure Dec 21 06:19:34 bshah : https://paste.ee/p/oQfA2 Dec 21 06:27:39 [ 14.941117] systemd[1]: lxc@android.service: Main process exited, code=exited, status=255/n/a Dec 21 06:27:41 [ 14.978583] systemd[1]: lxc@android.service: Unit entered failed state. Dec 21 06:27:43 [ 14.978639] systemd[1]: lxc@android.service: Failed with result 'exit-code'. Dec 21 06:27:51 umm you don't have lxc@android service disabled? Dec 21 06:28:21 bshah? Dec 21 06:28:54 the output you gave me above was manually starting lxc android container right? Dec 21 06:29:08 No Dec 21 06:29:20 systemd service. Dec 21 06:29:23 https://paste.ubuntu.com/26225565/ Dec 21 06:29:37 Should I have to mask it and run manually? Dec 21 06:30:07 hm no, that's fine.. bust still no useful on what exactly went wrong :/ Dec 21 06:31:35 bshah : log of lxc (/var/log/lxc/android.log) https://paste.ubuntu.com/26225574/ Dec 21 06:33:09 on line 90 it seems is working? Dec 21 06:33:59 Line 122 is where everything goes down south Dec 21 06:37:17 yeah but it's so verbose of it to tell which process crashed /s Dec 21 06:38:11 bshsh : how about this : https://github.com/lxc/lxc/issues/1026 Dec 21 06:39:45 Can you try to set apparmor=0 on cmdline? Dec 21 06:39:59 Where? Dec 21 06:40:07 kernel cmdline Dec 21 06:40:18 How to? Dec 21 06:40:30 heh Dec 21 06:40:43 BOARD_KERNEL_CMDLINE or something in BoardConfig.mk Dec 21 06:41:11 Oh ok... I thought I have to set somewhere in /boot. (non existent) Dec 21 06:44:57 bshah : which apparmour profile does the container use? Dec 21 06:45:10 unconfined? Dec 21 06:45:38 yeah Dec 21 06:45:52 well, apparmor is not present/backported in your kernel to begin with Dec 21 06:46:43 It didn't say anything about incomplete support. Dec 21 06:50:59 bshah : there is security/apparmour in the kernel tree. Dec 21 06:52:18 @UniversalSuperBox no - no idea.. repowerd seems to not start due toa an abi mismatch in ubuntu application api (iirc it uses that as abstraction library to read sensor status) but no idea yet why the rest has problems during start Dec 21 06:52:58 bshah : apparmour_status says module is loaded... but … ```Could not open /sys/kernel/security/apparmor/profiles: No such file or directory``` Dec 21 06:53:23 well, tried with apparmor=0? Dec 21 06:53:38 mka' ing Dec 21 06:55:35 @bshah, If apparmour is switched off the android container is not confined right? Dec 21 06:55:47 yeah Dec 21 06:55:52 well.. Dec 21 06:55:59 android container is not confined anyway Dec 21 06:56:09 it uses unconfined profile if you see config Dec 21 06:56:16 but anyway with apparmour it runs unconfined... Dec 21 06:56:19 Saw it. Dec 21 06:56:39 What is the need of apparmour then? other system level auditing? Dec 21 06:57:33 yeah, for instance ubuntu touch uses it Dec 21 06:59:48 bshah : file hybris-boot.img doesn't give apparmour=0 in the cmdline. Dec 21 07:00:51 Fine fine. Dec 21 07:03:28 bshah can you send me the ls of the /etc/apparmour.d/lxc/ directory Dec 21 07:03:55 bshah same error with apparmour=0 Dec 21 07:04:08 slightly later... don't have access to halium device at moment Dec 21 07:04:13 ok Dec 21 07:04:42 should I try removing lxc.aa_profile? Dec 21 07:04:56 with apparmour=0 Dec 21 07:08:44 @aki237, and lxc.aa_allow_incomplete = 1 started the container.... Wohooo Dec 21 07:13:55 bshah : lxc starts up fine but the init fails and systemd restarts the container and thus deleting my rootfs.img and system.img. 😫 Dec 21 07:14:36 Logs or didn't happen Dec 21 07:14:46 :p Dec 21 07:16:57 Aw maaan. Dec 21 07:21:16 Serves me right to mess around with config options. Dec 21 07:26:55 @bhushanshah, Please....😭😭😭 Dec 21 07:37:09 (Document) https://irc.ubports.com/YoM7LWI1/file_2996 Dec 21 07:37:24 @aki237, @bhushanshah logs... Dec 21 07:38:16 journal alone. Dec 21 07:38:37 before I could get dmesg, system reboot. Dec 21 07:41:00 I see Android services starting at least Dec 21 07:41:14 But I see them stopping too :P Dec 21 07:41:23 Details Dec 21 07:41:26 😛😂 Dec 21 07:42:04 Not sure why journalctl have repeated entries Dec 21 07:42:38 Most of messages are repeated twice in succession Dec 21 07:42:57 @aki237, apparently systemd starts the lxc as it fails..... Just an inch away from bulls eye Dec 21 07:45:22 Systemd should not restart container Dec 21 07:45:44 Then it must be android's init's failsafe Dec 21 07:46:12 I think what's happening is Android init is mounting rootfs read-only Dec 21 07:46:29 *remounting Dec 21 07:46:31 you mean rootfs.img? Dec 21 07:47:10 Feb 11 16:28:31 localhost.localdomain init: critical process 'lmkd' exited 4 times in 4 minutes; rebooting into recovery mode Dec 21 07:47:11 Feb 11 16:28:31 localhost.localdomain init: couldn't write 172 to /dev/cpuset/foreground/tasks: No space left on device Dec 21 07:47:13 Feb 11 16:28:31 localhost.localdomain android_reboot: Failed to open sysrq-trigger. Dec 21 07:47:30 I saw that too. Dec 21 07:48:01 can you mask lxc@android and then reboot and give me output of mount? Dec 21 07:49:50 and another thing before when I restarted the container manually it deleted the /data/* (in twrp recovery's reference) but now the /data (in recovery again) is populated with app, anr, system etc (andorid's rootfs) with no files Dec 21 07:50:04 yeah, that's fine Dec 21 07:50:12 it doesn't actually delete stuff Dec 21 07:50:18 ok Dec 21 07:50:23 since /data is mounted read-only Dec 21 07:50:29 ok Dec 21 07:50:36 ... I should patch out this stuff from android init Dec 21 07:50:41 (failsafe-reboot mode) Dec 21 07:51:54 @bshah, mask from twrp? Dec 21 07:52:11 um yeah Dec 21 07:52:20 ok Dec 21 07:52:33 mount /data/rootfs.img somewhere -> chroot -> source /etc/environment -> systemctl disable lxc@android Dec 21 07:52:50 bshah : oh come on.... got it got it Dec 21 07:56:29 bshah : https://paste.ubuntu.com/26225838/ (apparmour=0) Dec 21 07:58:35 first real error I see in log is : Feb 11 16:28:11 localhost.localdomain init: error opening block device /dev/block/bootdevice/by-name/recovery: No such file or directory Dec 21 07:58:47 maybe this needs to be killed from recovery Dec 21 07:58:53 err Dec 21 07:58:56 fstab I meant Dec 21 07:59:31 then only everything goes south Dec 21 07:59:43 ok Dec 21 08:00:04 also you need to kill init.*usb*.rc import from your init.yourdevice.rc Dec 21 08:00:38 bshah : I modified the pre-start script of lxc : https://paste.ubuntu.com/26225846/ Dec 21 08:00:56 bshah : ok Dec 21 08:01:35 in system.img? Dec 21 08:02:14 hm? Dec 21 08:02:28 @bshah, ? Dec 21 08:02:49 it is part of device tree and you would need to rebuild the systemimage yes Dec 21 08:03:07 Thats what... for mka systemimage. Dec 21 08:05:01 What does init*usb* do? Dec 21 08:06:59 they setup usb stuff. which we don't want container to do Dec 21 08:07:06 ok Dec 21 08:07:14 What else should I have to exclude Dec 21 08:08:17 just usb Dec 21 08:08:34 What about mounts (recovery) Dec 21 08:09:35 just remove the recovery mount from fstab or comment it out Dec 21 08:09:50 fstab.qcom... got it Dec 21 08:10:19 init.qcom.rc? Dec 21 08:12:05 bshah : init.qcom.power.rc too? (I saw some log messages unable to acces and control the cpu freq. Dec 21 08:12:17 no keep init.qcom.power.rc Dec 21 08:12:28 ok Dec 21 08:15:14 bshah : This was how my /tmp/android looked like : https://paste.ubuntu.com/26225888/ Dec 21 08:15:59 does init.usb.rc gets created by init.qcom.usb.rc during boot? Dec 21 08:18:06 no, but mustbe something else installing it Dec 21 08:18:12 well... installing it is fine Dec 21 08:18:18 just remove it's import init.usb.rc Dec 21 08:18:28 and import init.qcom.usb.rc Dec 21 08:18:39 In my init.qcom.rc only init.qcom.usb.rc is imported. Dec 21 08:18:49 removed it Dec 21 08:18:59 @bshah, I didn't see any such import. Dec 21 08:19:07 okay, then it's fine Dec 21 08:19:19 maybe init.usb.rc comes form system/core and is already commented Dec 21 08:19:28 I removed init.qcom.usb.rc Dec 21 08:19:33 okay Dec 21 08:19:38 so now what's status? Dec 21 08:19:46 mkaing Dec 21 08:20:31 only pushing system.img is enuf right? Dec 21 08:21:49 yeah, just make sure though system.img is not sparse Dec 21 08:21:58 ok Dec 21 08:22:02 there is replace-android-image script in halium-script that does it Dec 21 08:22:25 We just need to run convert_android_img before adb push Dec 21 08:23:37 yeah Dec 21 08:23:46 bshah : in twrp... should I mask the service? or allow it burst? Dec 21 08:24:13 better keep it masked.. easier for debugging I suppose? Dec 21 08:24:26 actually Dec 21 08:24:27 It's not even running. Dec 21 08:24:30 we want to start container Dec 21 08:24:34 so yeah remove mask Dec 21 08:24:38 ok Dec 21 08:25:55 bshah : how about this : https://paste.ubuntu.com/26225846/ (modified pre-start script) Dec 21 08:26:10 can you open PR on halium/lxc-android repo? Dec 21 08:26:19 That looks good? Dec 21 08:26:26 yeaha Dec 21 08:26:36 ok Dec 21 08:30:05 (Document) https://irc.ubports.com/FEmmd2aw/file_2997 Dec 21 08:30:22 @aki237, bshah : journal now. Dec 21 08:31:16 Apparently everything is getting killed. Dec 21 08:32:03 @aki237, aaaaand reboot. Dec 21 08:32:46 in twrp : it deleted everything.... Dec 21 08:32:51 :( Dec 21 08:32:52 (in data) Dec 21 08:33:08 I thought it was expected. Dec 21 08:36:04 ah, do you still have selinux=0 in cmdline? Dec 21 08:36:23 Will check Dec 21 08:37:07 no Dec 21 08:37:28 But selinux is not built with the kernel so what effect does it make. Dec 21 08:38:12 I am not entirely sure what's wrong though.. selinux was just random guess Dec 21 08:38:13 so ignore Dec 21 08:38:30 bshah : still will try. Dec 21 08:42:10 bshah : https://github.com/Halium/lxc-android/blob/master/var/lib/lxc/android/pre-start.sh#L15 should be unmounted if the lxc is called second time right? Dec 21 08:42:22 so it wont get removed. Dec 21 08:50:39 bshah : how about this script : https://paste.ubuntu.com/26226024/ (change is it'll unmount any mounts in $LXC_ROOT_PATH if it was started again) Dec 21 08:51:20 umount -R actually doesn't work Dec 21 08:51:24 (been there done that( Dec 21 08:51:25 ? Dec 21 08:51:28 *) Dec 21 08:51:33 Damn Dec 21 08:53:12 But does umount induvidually work? Dec 21 08:53:26 if so we can write the similar funciotn to do that. Dec 21 08:56:02 well thing is, I would prefer to just straight out exit if container is started again Dec 21 08:56:21 so exit -255 is fine? Dec 21 08:56:28 yeah Dec 21 08:56:45 I'll just echo a message and exit Dec 21 08:57:03 How about unmount and exit? Dec 21 08:57:39 it will prevent debugging in case things are e.g not mounted correctly Dec 21 08:57:48 ok Dec 21 08:57:53 just exit it is Dec 21 09:22:53 bshah : I made a pull request https://github.com/Halium/lxc-android/pull/3 Dec 21 09:22:58 thanks Dec 21 09:23:00 will merge Dec 21 09:23:06 ok Dec 21 10:34:55 bshah : the new script seems to work... https://paste.ubuntu.com/26226349/ Dec 21 10:35:26 yay Dec 21 10:40:46 @aki237, bshah : I haven't formed the udev rules. May be due to that? Dec 21 10:41:28 I doubt it but adding udev rules doesn't hurt Dec 21 10:41:33 ok Dec 21 10:47:30 @bhushanshah it seems systemd restarts lxc@android Dec 21 10:47:49 after it fails it starts the container. But the script killed the container. Dec 21 10:49:38 no it's not systemd Dec 21 10:49:44 it's android init restarting itself Dec 21 10:49:56 See this https://paste.ubuntu.com/26226397/ Dec 21 10:55:19 @bhushanshah It seems the permission denied is by apparmour : https://github.com/lxc/lxd/issues/2661 Dec 21 11:21:23 What is netmgrd? Dec 21 11:24:23 bshah: atlast lxc is up and running Dec 21 11:24:38 wee Dec 21 11:24:43 what did oyu do? Dec 21 11:25:28 bshah : you'll kill me...😅 apparmour vs apparmor. Dec 21 11:25:38 apparmour=0 Dec 21 11:25:42 lolz Dec 21 11:25:55 what is netmgrd? Dec 21 11:26:17 That exits with status 0 repeatedly Dec 21 11:26:29 umm not sure Dec 21 11:26:34 ok Dec 21 11:29:15 @UniversalSuperBox there are Read-only File system errors in some of the upstrart log files Dec 21 11:31:59 @anpok: remount root maybe? Dec 21 11:33:36 WOOO wlan working. Dec 21 11:42:13 bshah : netmgrd is responsible for RIL. It's crashing. Dec 21 11:42:31 most likely due to paranoid_network thingie Dec 21 11:42:49 have you enabled paranoid_network kernel config option or is it disabled? Dec 21 11:43:04 in either case, try to flip it and see if it works Dec 21 11:43:50 disabled. Dec 21 11:45:09 what should the test_hwcomposer output? Dec 21 11:45:23 it should show soemthing on screen and not crash Dec 21 11:45:48 http://blog.bshah.in/images/test_hwcomposer.jpg Dec 21 11:46:02 but it will crash most likely I am sure Dec 21 11:46:08 It didn't show anything on screen. Dec 21 11:46:18 what's output? Dec 21 11:46:38 Resolution, OpenGLES version, stop Dec 21 11:46:54 test_egl : segfault Dec 21 11:47:05 test_egl_configs > puts something. Dec 21 11:47:23 test_glesv2 fail. Dec 21 11:47:28 *segfault Dec 21 11:47:55 for test_hwcomposer Dec 21 11:48:12 try export EGL_PLATFORM=hwcomposer test_hwcomposer Dec 21 11:48:26 https://paste.ubuntu.com/26226759/ Dec 21 11:48:46 same for both cases? Dec 21 11:50:11 After EGL_PLATFORM, Resolution, Opengl version, and still running Dec 21 11:50:17 But nothing on screen Dec 21 11:51:33 time to debug.. :) Dec 21 11:51:39 can you strace it? Dec 21 11:51:46 Will do. Dec 21 11:51:50 (with env var exported) Dec 21 11:53:41 ```connect(36, {sa_family=AF_LOCAL, sun_path="/dev/socket/property_service"}, 31) = -1 ENOENT (No such file or directory)``` Dec 21 11:55:09 ignore Dec 21 11:55:17 bshah ok Dec 21 11:55:47 best would be to have whole log, so is easier to know what's up Dec 21 11:56:21 bshah : is 2M enough? Dec 21 11:57:13 well, first 200ish lines would be really useful Dec 21 11:57:21 ok Dec 21 11:59:17 access /etc/ld.so.nohwcap ENOENT Dec 21 12:00:06 not problem Dec 21 12:00:28 yes.. it is trying to linker cache. Dec 21 12:01:03 the actual problem will be highlighted much below that lines.. so yeah best give log and not individual lines Dec 21 12:01:04 :p Dec 21 12:01:19 :P Dec 21 12:02:09 (Document) https://irc.ubports.com/Nyk80b9z/file_3001 Dec 21 12:03:06 Lots of EBADF Dec 21 12:03:17 alright Dec 21 12:03:37 so I think, what you need is umm, add a caf repo, apt update and apt upgrade Dec 21 12:03:38 Oh I have attached it in telegram. Dec 21 12:04:02 the test_hwcomposer is not compatible? Dec 21 12:04:22 can you add http://repo.halium.org/caf in the sources.list -> apt update and apt upgrade? Dec 21 12:12:23 only main right? Dec 21 12:12:42 yeah Dec 21 12:13:16 wifi is very slow. Dec 21 12:15:33 No space left on device. Dec 21 12:15:45 urgh Dec 21 12:15:52 ? Dec 21 12:15:59 okay, in that case you can probably resize rootfs Dec 21 12:16:03 from adb shell Dec 21 12:16:06 ok Dec 21 12:16:10 (in recovery0 Dec 21 12:16:11 twrp? Dec 21 12:16:14 Thats ok Dec 21 12:16:24 to 5? Dec 21 12:16:45 e2fsck -yf /data/rootfs.img; resize2fs -f /data/rootfs.img 4G Dec 21 12:16:59 4G is fine? Dec 21 12:17:04 yeah will be enough Dec 21 12:32:24 bshah : PARANOID_NETWORK is y.. now ping says socket : permission denied. Dec 21 12:32:46 Meh Dec 21 12:33:20 ? Dec 21 12:33:32 But but...is netmgr thingie fixed? Dec 21 12:33:55 Nope...😂😂 Dec 21 12:34:04 :-( Dec 21 12:34:16 So can be switched off right? Dec 21 12:34:24 Yeah Dec 21 12:36:16 @bhushanshah for future reference https://elinux.org/Android_Security#Paranoid_network-ing Dec 21 12:38:43 netmgrd : librmnetctl.so not found from logcats Dec 21 12:44:03 @bhushanshah We dont need surface flinger right? Dec 21 12:46:45 @bhushanshah here is the logcat https://paste.ubuntu.com/26227021/ Dec 21 13:00:43 Can anyone check whether you have librmnetctl.so is in the lxc or /system/lib/ ? Dec 21 13:00:57 if so can you check the source where it came from? Dec 21 13:41:42 @bhushanshah the netmgrd issue is solved. but still apt says not space left on device after resize Dec 21 13:42:27 hm wat Dec 21 13:42:29 df -h please? Dec 21 13:42:55 bshah : https://paste.ubuntu.com/26227288/ Dec 21 13:45:16 this looks like resize didn't work? Dec 21 13:45:40 4Gb right? Dec 21 13:45:45 loop0 Dec 21 13:45:49 if my calculation skills are right rootfs is still 2GB? Dec 21 13:46:19 or it is possible that it is filled up with journalctl crap Dec 21 13:46:27 (/var/log/journal/ Dec 21 13:46:29 0 Dec 21 13:48:05 delete both? Dec 21 13:48:22 yeah I suppose Dec 21 13:48:58 I am curious though why it filled up... what's pattern in dmesg? Dec 21 13:49:11 something which keeps looping and looping and looping Dec 21 13:49:25 I think the battery (qpnp loops regularly) Dec 21 13:49:45 meh Dec 21 13:50:10 maybe grep and kill those msessages from kernel Dec 21 13:50:25 I was about to delete all the printk in the actual driver. Dec 21 14:01:12 bshah : in twrp, the mounter rootfs image is only 15% used. Dec 21 14:07:32 @bhushanshah, I tried to build some of the packages and I noticed a few things: … In droidmedia, a space is missing between your email adress and the date (https://github.com/halium-packaging/droidmedia/blob/master/debian/changelog#L5). And it looks like you do not compile anything in the package (is that intended?), so it's architecture independent. You can probably change the architecture to all here Dec 21 14:07:33 (https://github.com/halium-packaging/droidmedia/blob/master/debian/control#L12) Dec 21 14:09:08 Also, you don't need to override it if you use the automatic configure thing then anyway: https://github.com/halium-packaging/gst-droid/blob/master/debian/rules#L9 Dec 21 14:13:06 @bhushanshah is this an error of libc? https://unix.stackexchange.com/questions/195293/bind-mounting-and-df-output Dec 21 14:23:05 @UniversalSuperBox Ok no idea yet... trying to make sense out of the problem i just made another attempt with a vivid rootfs and now I get a proper dmesg Dec 21 14:47:00 @JBBgameich, for last PR (version number) it is kinda late.. given umm 1.x version is in repo already Dec 21 14:47:09 I don't want to add pin Dec 21 14:47:20 also I used this version frop spec file IIRc Dec 21 14:47:49 bshah : is mount namespace enabled in hammerhead? Dec 21 14:48:01 I am not sure Dec 21 14:48:18 @bshah, We could use an epoch again ... Dec 21 14:48:43 I hate epoches.. but umm, whatever :/ Dec 21 14:48:44 It would make sense to use the upstream version numbers instead of inventing our own Dec 21 14:48:50 okay Dec 21 14:49:13 bshah : it seems mount namespace is required. else the lxc mounts are done in a same global namespace thus bind mounts are considered to be the part of the current's process's mounts. Dec 21 14:49:43 @JBBgameich, I used version from upstream btw: https://github.com/sailfishos/gst-droid/blob/master/rpm/gst-droid.spec#L6 Dec 21 14:49:51 and didn't invent them :P Dec 21 14:50:15 oh, they don't use their own version numbers as well 😆 Ok, then it's probably ok Dec 21 14:50:16 but yeah, whatever you think is best.. if you downgrade version add epoch please Dec 21 14:51:29 @aki237: can you try with mount namespaces enabled? Dec 21 14:51:45 mkaing Dec 21 14:52:37 https://www.xkcd.com/303/ Dec 21 14:52:58 @bshah, 🤣🤣 Dec 21 14:53:34 Btw, I cross-compiled the libhybris package today. It's much faster than building with qemu-user-static like I did before. just adding `--host-arch=armhf` did the trick Dec 21 14:54:11 not necessary that it would work.. I've faced multiple issues with cross-compiled libhybris IIRc Dec 21 14:54:33 oh, does it make problems on the device then? Dec 21 14:55:06 yeah I don't remember details but in initial days of plasma mobile we had multiple issues with it Dec 21 14:55:46 Cross compiling was improved a lot this year (in debian), so maybe it would work now. Anyway, probably not on xenial Dec 21 14:56:12 but yeah what I am telling is from vivid times ^^ Dec 21 14:56:15 :P Dec 21 14:57:04 the libhybris package contains extra "hacks" to compile on vivid at all even 😆 Dec 21 14:57:42 reminds me we should "cleanup" that things Dec 21 14:58:39 I already worked on that a little bit: https://github.com/debian-pm/libhybris-packaging Dec 21 14:59:00 But I guess it will have merge conflicts with the halium version now Dec 21 15:13:27 bshah : https://groups.google.com/a/chromium.org/forum/#!topic/chromium-os-reviews/AVvM4Keea8U does this look like my problem? Dec 21 15:14:32 I am not entirely sure tbh.. your kernel is weird Dec 21 15:14:42 I am not even sure what exact version it is based on Dec 21 15:14:56 it claims 3.4.0 but have commits from 3.4.5+ Dec 21 15:14:58 o_o Dec 21 15:15:26 so I'd say it's mostly try and error from your end maybe? Dec 21 15:15:41 But the message from google groups seems like it is stating my problem Dec 21 16:31:51 how was the pulseaudio droid module called? Dec 21 16:37:14 i found itt Dec 21 16:49:32 is github down? Dec 21 16:49:40 Seems so Dec 21 16:49:46 no Dec 21 16:49:54 working fine here Dec 21 16:49:59 Or at least terribly slow for me Dec 21 16:50:31 https://status.github.com/messages says fine. Dec 21 16:53:45 @bhushanshah PR : https://github.com/Halium/lxc-android/pull/3 Dec 21 16:54:11 Yes I know about that PR well Dec 21 16:54:16 It's on my to-do list Dec 21 16:54:27 cool Dec 21 16:55:03 @bhushanshah can you send me the output of df , mount if possible Dec 21 16:55:17 I am already off at moment Dec 21 16:55:28 I mean well off from work machine Dec 21 16:55:34 Whenevr possible. Dec 21 16:55:51 Sure it it'll be tomorrow morning Dec 21 16:56:05 Or someone here can provide it from working device Dec 21 16:56:09 ^^ Dec 21 16:56:19 Ah yes. Dec 21 17:06:31 o_O oh I killed my ipv4 network Dec 21 17:07:18 Hello, how's there in future? Dec 21 17:07:21 😛 Dec 21 17:07:37 😂 Dec 21 18:35:01 Guys when the lxc is disabled the disk usage is file. but when enabled it fills up the rootfs and errors occur (no space left on device) Dec 21 18:35:33 I have to run a fsck on the rootfs.img in twrp to restore the normal size. Dec 21 18:36:37 Any ideas how to prevent this. Stackoverflow says to kill any process that keeps a large file open which is then deleted. ( … ```lsof +L1``` … ) Dec 21 18:44:19 Vince was added by: Vince Dec 21 20:38:03 @bhushanshah apparently rsyslogd was hogging up all the space in /var/log/ Dec 21 20:52:27 I'm continueing porting, current state is that the android container is starting and `test_vibrator` works. … test_hwcomposer doesn't work on the reference rootfs and also not with the libhybris n linker (my own build). … Strace output is here: … https://paste.kde.org/p7wv3tkw1 … Does anybody have an Idea? Dec 21 20:53:13 Does logcat run in reference? Dec 21 20:53:21 Maybe watch that while you try to run test_hwcomposer as well Dec 21 20:53:31 That can give some handy clues Dec 21 20:53:51 We have systemd in rootfs. But why rsyslogd is also installed? Dec 21 20:54:37 @UniversalSuperBox, dmesg shows … `[ 1673.801772] binder: 1588:1588 transaction failed 29189, size 0-0` … btw Dec 21 20:55:23 logcat just says `Unable to open connection to debuggerd: Connection refused` Dec 21 20:55:30 pffft Dec 21 20:55:35 I hope that is not the fault of my rootfs build Dec 21 20:56:00 Maybe you can't get a connection from inside reference... I've honestly not tried Dec 21 20:57:16 (It's my own rootfs build I'm using currently) Dec 21 20:57:34 `` … ├─lxc … │ └─android … │ ├─1079 /init —second-stage … │ ├─1081 /sbin/ueventd … │ ├─1109 /system/bin/logd … │ └─1116 /system/bin/vdc —wait cryptfs init_user0 … ``` … is android even fully started? Dec 21 20:59:46 (Photo, 784x979) https://irc.ubports.com/5EDISE9b/file_3033.jpg Dec 21 20:59:55 Comparing with this one .... it's not Dec 21 21:00:04 That'll do it Dec 21 21:00:27 But how do I find out what the problem is? Dec 21 21:00:50 Is anything spamming dmesg? Dec 21 21:01:40 Here are some init related errors: … ```[ 180.805493] init: init second stage started! … [ 180.807987] init: Running restorecon... … [ 180.808097] init: waitpid failed: No child processes … [ 180.808522] init: (Loading properties from /default.prop took 0.00s.) … [ 180.811302] init: (Parsing /init.environ.rc took 0.00s.) … [ 180.813963] init: /init.qcom.rc: 654: user requires 1 argument … [ 180.815603] init: /init.qcom.rc: 124 Dec 21 21:01:41 'load_all_props' … [ 180.815801] init: could not import file 'init.qcom.power.rc' from '/init.qcom.rc': No such file or directory … [ 180.822765] init: (Parsing init.qcom.usb.rc took 0.01s.) … [ 180.822994] init: ignored duplicate definition of service 'per_mgr'init: (Parsing init.target.rc took 0.00s.) Dec 21 21:01:42 could not import file 'init.bq.rc' from '/init.qcom.rc': No such file or directory … [ 180.823109] init: (Parsing /init.qcom.rc took 0.01s.) … [ 180.824214] init: (Parsing /init.cm.rc took 0.00s.)``` Dec 21 21:01:47 (Sticker, 512x300) https://irc.ubports.com/D1il98x5/file_3034 Dec 21 21:01:57 Sorry, IRC... Dec 21 21:02:03 https://paste.kde.org/pqxzbwbxw Dec 21 21:03:44 So, understand that pretty much everything I say is an educated guess Dec 21 21:03:47 Anything after that? Dec 21 21:04:12 sometimes the guesses aren't even educated Dec 21 21:04:24 https://paste.kde.org/po4ysfi8p Dec 21 21:04:38 Any Idea helps me ... Dec 21 21:05:23 `mmcblk0p30` is the oem partition Dec 21 21:05:33 @JBBgameich Which device is this? Dec 21 21:05:39 device tree. Dec 21 21:05:53 https://github.com/jbbgameich/android_device_bq_chaozu Dec 21 21:06:02 Well, you can take vold out of inits Dec 21 21:06:58 Have you commented ot this line? https://github.com/JBBgameich/android_device_bq_chaozu/blob/cm-14.1/rootdir/init.qcom.rc#L29 Dec 21 21:07:52 init.qcom*usb* Dec 21 21:07:54 I'm exactly using the version from github Dec 21 21:08:17 Oh.. I was advised to disable the usb init service. Dec 21 21:08:30 I can try doing that as well ... Dec 21 21:08:45 I love this Dec 21 21:08:55 "starting debuggerd" … "debuggerd killed by signal 6" Dec 21 21:09:00 can you give me the output of ls /var/lib/lxc/android/rootfs/ Dec 21 21:09:14 @UniversalSuperBox, Irony.. Dec 21 21:09:29 sure … ```acct d etc init.cm.rc init.qcom.syspart_fixup.sh init.usb.rc persist seapp_contexts storage ueventd.rc … bugreports data file_contexts.bin init.environ.rc init.qcom.usb.rc init.zygote32.rc proc selinux_version sys vendor … cache default.prop firmware Dec 21 21:09:30 init.qcom.early_boot.sh init.rc loggy.sh property_contexts sepolicy system … charger dev fstab.qcom init.qcom.rc init.target.rc mnt sbin service_contexts tombstones … config dsp init init.qcom.sh init.usb.configfs.rc oem sdcard socket ueventd.qcom.rc``` Dec 21 21:09:31 So, that's why you can't use logcat, and it's probably the first place to look. Dec 21 21:09:34 oops Dec 21 21:09:49 https://paste.kde.org/pllytroca Dec 21 21:09:50 angry shaking Dec 21 21:09:52 :P Dec 21 21:10:05 I should use the pastebin more 😆 Dec 21 21:10:58 @JBBgameich I didn't see any init.bq.rc in that directory... Dec 21 21:11:13 Probably system build problem Dec 21 21:11:24 Device tree broken again ....? Dec 21 21:12:09 This device tree has never been really tested (only on the lite version of this phone) ... I'm just copying things over from tenshi (the pro version) Dec 21 21:13:56 The other device doesn't even have init.bq.rc Dec 21 21:14:52 Maybe I should just comment it out Dec 21 21:15:14 No No Dec 21 21:19:23 also what about init.qcom.power.rc? It's missing too Dec 21 21:19:32 I'm checking. Dec 21 21:20:41 Go to your out directory. Dec 21 21:20:53 in your system Dec 21 21:21:02 *build computer Dec 21 21:21:05 I'm there Dec 21 21:21:34 do you see a android-ramdisk.img? Dec 21 21:21:46 run … ```file android...``` Dec 21 21:21:55 and paste the output Dec 21 21:22:18 ```android-ramdisk.img: gzip compressed data, from Unix``` Dec 21 21:22:40 good. now create a new directory. and cd into it Dec 21 21:22:49 where? Dec 21 21:23:09 in out ... just some temp directory. Dec 21 21:23:16 some name. Dec 21 21:23:27 ok Dec 21 21:23:37 run this … ```cat ../android-ramdisk.img | gzip -d | cpio -i``` Dec 21 21:24:08 ok finished Dec 21 21:24:20 ls output Dec 21 21:24:47 https://paste.kde.org/pf1e5nzjj Dec 21 21:25:30 but wait, I did some changes to add the missing files in the meantime and rebuilt Dec 21 21:25:41 Whaaaa ... Dec 21 21:25:44 Oh ok Dec 21 21:26:01 That's probably the reason why init.bq.rc is there Dec 21 21:26:18 The system.img in your build pc is fine apparently something happened to the ramdisk in the lxc. Dec 21 21:26:24 ok Dec 21 21:26:35 replace the image and try again Dec 21 21:27:59 shutdown is a little bit tricky, as it waits for the container to timeout. Will take a few seconds to reboot the device properly Dec 21 21:28:12 Yes mine too. Dec 21 21:32:21 ok, it fixed the two missing file errors from dmesg Dec 21 21:32:29 ok Dec 21 21:32:38 oh, init.qcom.power.rc is still missing ... Dec 21 21:33:47 @JBBgameich Just add a line in https://github.com/JBBgameich/android_device_bq_chaozu/blob/cm-14.1/device.mk#L91 … ```init.qcom.power.rc \``` Dec 21 21:34:11 and mka systemimage Dec 21 21:41:19 @aki237, To me it looks like that won't change anything unless I add the files to rootdir/Android.mk Dec 21 21:42:13 Ok Dec 21 21:42:20 Now it is included? Dec 21 21:42:45 I need to rebuild again Dec 21 21:42:58 Yeah Dec 21 21:47:25 I'm installing now Dec 21 21:50:31 ok, we fixed missing file errors now Dec 21 21:50:39 ok Dec 21 21:50:51 we still have … ```[ 18.240631] init: /init.qcom.rc: 654: user requires 1 argument … [ 18.242251] init: /init.qcom.rc: 1245: invalid keyword 'load_all_props' … [ 18.242976] init: (Parsing init.qcom.power.rc took 0.00s.) … [ 18.243185] init: ignored duplicate definition of service 'per_mgr'init: (Parsing init.target.rc took 0.00s.)``` Dec 21 21:51:04 ignore Dec 21 21:51:47 android still doesn't fully start Dec 21 21:51:55 journal? Dec 21 21:52:13 ```journalctl | grep "localdomain init" Dec 21 21:53:04 https://paste.kde.org/p3h8swe6c Dec 21 21:53:40 logcat Dec 21 21:53:49 It's not working Dec 21 21:53:58 because debuggerd is not started Dec 21 21:54:03 cat /proc/cmdline Dec 21 21:54:37 https://paste.kde.org/pqpyywgzk Dec 21 21:55:46 systemctl status still says: https://paste.kde.org/ Dec 21 21:56:56 ls /sys/fs Dec 21 21:57:12 @JBBgameich, You, uh... forgot something Dec 21 21:57:15 `cgroup ext4 fuse` Dec 21 21:58:02 ```aa-enabled``` Dec 21 21:58:19 is that a command? Dec 21 21:58:23 yes Dec 21 21:58:30 -bash: aa-enabled: command not found Dec 21 21:59:01 . /etc/environment Dec 21 21:59:26 halium ref rootfs right Dec 21 22:00:15 no ... My own build Dec 21 22:00:43 debian buster Dec 21 22:02:17 ```zgrep CONFIG_SECURITY /proc/config.gz``` Dec 21 22:02:38 ```# CONFIG_SECURITY_NETWORK_XFRM is not set … # CONFIG_SECURITY_PATH is not set … # CONFIG_SECURITY_SMACK is not set … # CONFIG_SECURITY_TOMOYO is not set … # CONFIG_SECURITY_APPARMOR is not set … # CONFIG_SECURITY_YAMA is not set``` Dec 21 22:03:06 ok Dec 21 22:03:35 cat /var/log/lxc/andorid.log Dec 21 22:04:36 https://paste.kde.org/pqm52eyj2 Dec 21 22:04:49 There was only `cat /var/lib/lxc/android/android.log` Dec 21 22:05:15 changes for systems Dec 21 22:06:46 I have lxc 2.0.9 Dec 21 22:07:07 Should I change back to the reference rootfs or is it something unrelated probably? Dec 21 22:07:26 Nah it should work Dec 21 22:15:25 @aki237, Can you give me more of this Dec 21 22:17:13 and ls /var/lib/lxc/android/rootfs/system/** Dec 21 22:19:00 @aki237, That was the full output of the command Dec 21 22:19:18 thats it.. huh Dec 21 22:19:43 https://paste.kde.org/pyawnzuox Dec 21 22:20:19 The log ends with vold and debuggerd are suddenly killed Dec 21 22:23:03 I had a look at a log of another device, the init should do much more than it does, it just stops on this device Dec 21 22:23:18 cat var...lxc../system/etc/init/debuggerd.. Dec 21 22:24:27 ```service debuggerd /system/bin/debuggerd … group root readproc … writepid /dev/cpuset/system-background/tasks``` Dec 21 22:24:36 `cat /var/lib/lxc/android/rootfs/system/etc/init/debuggerd.rc` Dec 21 22:24:47 yeah yeah Dec 21 22:24:58 cat /var/../lxc/config Dec 21 22:25:28 https://paste.kde.org/pxrvwc6if Dec 21 22:25:35 Disable init_user0 call Dec 21 22:25:47 It is somewhere in system/core Dec 21 22:26:14 @bhushanshah Graphics work. Wohoo Dec 21 22:26:15 Where did you find it? Dec 21 22:26:34 @bhushanshah Or where can I disable it? Dec 21 22:26:44 @bhushanshah, . Dec 21 22:26:52 ah ... Dec 21 22:28:37 `system/core/init/builtins.cpp`? Dec 21 22:29:17 I *think* that's it Dec 21 22:29:22 Let me check Dec 21 22:29:59 At line 1186 is sort of a "table" Dec 21 22:30:19 Should I remove the row containing init_user0? Dec 21 22:30:24 halium successfully compiled for oneplus2 😊🎉🎉🎉 Dec 21 22:30:37 @JBBgameich, Umm no Dec 21 22:30:59 Not this.. Dec 21 22:31:15 Somewhere in init.*.rc Dec 21 22:33:42 Can't find it ... Dec 21 22:33:50 Question: y'all don't mind if I just lurk, do ya? Can't seem to do anything useful for the OP5 port Dec 21 22:34:14 @c_smith, Sure, feel free to lurk Dec 21 22:34:20 Cool Dec 21 22:34:45 TBH lurking is how I caught on to a lot of basics in other projects Dec 21 22:34:54 @bhushanshah, In the device tree, right? Dec 21 22:35:33 https://android.googlesource.com/platform/system/core/+/5949745534e11a877ec839931af76c569f78e058/rootdir/init.rc#423 Dec 21 22:35:45 thanks :) Dec 21 22:38:05 @bhushanshah test_hwcomposer works after installing libhybris from caf Dec 21 22:38:22 Yeah, was expected Dec 21 22:38:41 other gl tests still say segfault Dec 21 22:38:53 I probably need to rebuild now? Dec 21 22:39:00 Some are broken well Dec 21 22:39:04 Yes jbb Dec 21 22:39:47 so next I should try ubuntu? Dec 21 22:40:15 I suppose.. yeah Dec 21 22:40:22 Wohoo Dec 21 22:41:41 Just change the rootfs image? Dec 21 22:41:59 ie., tar.gz? Dec 21 22:47:29 No Dec 21 22:47:37 i got it... Dec 21 22:47:46 Unfortunately ubports needs more tinkering Dec 21 22:47:58 im at the wiki Dec 21 22:48:00 docs.ubports.com Dec 21 22:48:11 sorry *docs. Dec 21 22:48:35 Big ol' Porting Information in the sidebar, under Platform. https://docs.ubports.com/en/latest/porting/introduction.html Dec 21 22:49:04 Is that not a good placement? Dec 21 22:49:25 https://paste.kde.org/pdriadqmu Dec 21 22:49:41 Bring it up, put it down Dec 21 22:49:43 more is started .. and killed Dec 21 22:49:45 It's the Android way Dec 21 22:50:37 I don't know why it does that, but if you guys can figure it out I'd love to know. Dec 21 22:51:25 @JBBgameich do you have all dmesg message with init:? Dec 21 22:51:36 Error is likely above what you posted Dec 21 22:53:38 @bhushanshah why is syslog installed.. we have journal right. Dec 21 22:53:46 What is the difference? Dec 21 22:53:50 (Document) https://irc.ubports.com/zKqCXkX0/file_3049 Dec 21 22:54:07 @aki237, Probably ubuntu does that for compatibility reasons or whatever Dec 21 22:54:25 https://paste.kde.org/plypcisii Dec 21 22:59:58 The problem is that the android init tries to restart itself Dec 21 23:03:22 https://github.com/LineageOS/android_system_vold/blob/cm-14.1/Android.mk#L128 Dec 21 23:03:53 Possibly stupid hack but try to comment this Dec 21 23:03:59 oh is vold.rc maybe missing? Dec 21 23:04:17 @bhushanshah, I'll try :) Dec 21 23:04:21 No, issue is it's present Dec 21 23:05:09 And while at it also uncomment https://android.googlesource.com/platform/system/core/+/5949745534e11a877ec839931af76c569f78e058/rootdir/init.rc#438 Dec 21 23:09:16 I'm at https://docs.ubports.com/en/latest/porting/building-ubports-boot.html at mka stage. ubports-boot target not found Dec 21 23:10:22 ```# Specific projects for Halium … subdirs += \ … halium/hybris-boot \ … halium/ubports-boot \``` Dec 21 23:10:33 Add ubports-boot to build/core/main.mk Dec 21 23:10:47 ok Dec 21 23:12:14 can I run mka systemimage and boot simultaneously? Dec 21 23:12:35 `mka systemimage hybris-boot` Dec 21 23:12:44 parallel? Dec 21 23:13:19 It won't be parallel most likely Dec 21 23:13:24 Huh Dec 21 23:14:26 Well hybris-boot builds kernel so it'll be built first and then syayen.img Dec 21 23:15:03 As I need code aurora sources, I have to choose CAF ubuntu touch right? Dec 21 23:15:55 Yeah Dec 21 23:16:35 the kernel is more or less the same configuration right? Dec 21 23:17:25 more or less Dec 21 23:22:17 Is it ok to disable syslogd? Dec 21 23:24:48 unable to change the password in twrp chroot. Dec 21 23:25:27 What happens? Dec 21 23:25:43 ```passwd: Authentication token manipulation error … passwd: password unchanged``` Dec 21 23:26:10 The strangest errors Dec 21 23:26:25 really Dec 21 23:27:00 `ls -l etc/shadow` Dec 21 23:27:25 -rw-r----- Dec 21 23:27:38 root:shadow Dec 21 23:27:58 Is the image full already? Dec 21 23:28:43 ```Filesystem 1K-blocks Used Available Use% Mounted on … - 2015824 1412192 501232 74% /``` Dec 21 23:28:55 Maybe something here? https://askubuntu.com/questions/57620/getting-an-authentication-token-manipulation-error-when-trying-to-change-my-us Dec 21 23:29:45 New dmesg with all the hacks included: https://paste.kde.org/p524mn8sb Dec 21 23:34:04 init.qcom.early_boot.sh Dec 21 23:34:15 What's in that file? Dec 21 23:35:14 https://github.com/JBBgameich/android_device_bq_chaozu/blob/cm-14.1/rootdir/init.qcom.early_boot.sh Dec 21 23:35:42 @UniversalSuperBox, Thanks... pam refresh did it Dec 21 23:35:53 👍 Dec 21 23:37:26 [   31.531732] init: Starting service 'vold'... Dec 21 23:37:29 Y tho Dec 21 23:39:04 Port 22 conn refused. Dec 21 23:39:10 ping is fine. Dec 21 23:39:36 @JBBgameich at least container doesn't reboot itself right? Dec 21 23:39:49 It's displayed as started Dec 21 23:40:29 Well yes Dec 21 23:40:37 But what's process tree Dec 21 23:40:38 ? Dec 21 23:40:48 @aki237, Where did you get your rootstock from? Dec 21 23:41:00 tool? AUR Dec 21 23:41:19 No, the `rootstock-touch-install` script Dec 21 23:41:25 yes Dec 21 23:41:28 AUR. Dec 21 23:41:33 That's why you're having problems. Dec 21 23:41:41 ok Dec 21 23:41:42 @bhushanshah, ``` │ └─android … │ ├─1057 /init --second-stage … │ ├─1059 /sbin/ueventd … │ └─1087 /system/bin/logd``` Dec 21 23:42:03 :-| Dec 21 23:42:33 @aki237, https://docs.ubports.com/en/latest/porting/installing-16-04.html#install-system-img-and-rootfs Dec 21 23:43:43 Oh I used this https://github.com/ubports/rootstock-ng/ Dec 21 23:44:35 Right, that one doesn't have the changes to enable SSH. Alternatively, let me pick the line here... Dec 21 23:44:47 Ok Dec 21 23:44:58 do_shell "sed -i 's/PasswordAuthentication=no/PasswordAuthentication=yes/g' /data/recovery/system/etc/init/ssh.override" Dec 21 23:45:03 I can do int manually by mounting the image right? Dec 21 23:45:04 I, uh Dec 21 23:45:05 Didn't want that Dec 21 23:45:10 https://github.com/UniversalSuperBox/rootstock-ng/blob/b9127fe8854243886433c0ff7ca4b3c276534c92/rootstock-touch-install#L271 Dec 21 23:45:14 That line and the next one Dec 21 23:45:21 oh Dec 21 23:45:27 @anpok maybe that's what happened for you as well Dec 21 23:48:10 offtopic... Guesstimate how many flash/writes can the device withstand? Dec 21 23:48:45 Lots Dec 21 23:48:53 hope so Dec 21 23:50:07 storage controller will dead before you will do max write cycles Dec 21 23:50:21 I think Dec 21 23:50:55 As on many usb flash drives Dec 21 23:51:21 @UniversalSuperBox I do have that version.. Dec 21 23:51:26 okay. Dec 21 23:51:40 but there should be an entry in shadow after that right? Dec 21 23:51:57 Probably Dec 21 23:52:41 lxc@android status gives : … ```Failed to get properties: No such interface ''``` … and the device reboots after some time. (I did disable all the radio based services) Dec 21 23:53:29 it remained untouched.. Dec 22 00:00:20 Guys Have I disabled all radio based services? https://paste.ubuntu.com/26230082/ Dec 22 00:00:54 Ubports doesn't use lxc@android service Dec 22 00:01:03 Or systemd for that matter Dec 22 00:01:10 upstatrt? Dec 22 00:01:19 @UniversalSuperBox can give you more info on what to do next tbh Dec 22 00:08:09 how to see the android contiainer logs in upstart? Dec 22 00:25:55 phone reboot after sometime (no unity thingie yet only boot logo) Dec 22 00:46:47 unity-compositor doesn't seem to start even though the android conatiner is running fine : https://paste.ubuntu.com/26230230/ Dec 22 00:47:25 Can you get its unity-system-compositor.log? Dec 22 00:47:34 @UniversalSuperBox This is the unity-system-com... https://paste.ubuntu.com/26230231/ Dec 22 00:47:38 wow Dec 22 00:47:57 I was on the way. Dec 22 00:48:00 Can you do `sudo unity-system-compositor`? Dec 22 00:48:27 Seg fault. Dec 22 00:48:34 strace? Dec 22 00:48:36 cooool Dec 22 00:48:53 😂 Dec 22 00:50:28 @UniversalSuperBox another thing the /var/log/syslog fills up rapidly and fills up my rootfs.img space. I have to truncate every time it's about reach the limit … ```: > /proc//fd/{5,6}``` Dec 22 00:50:36 Is there any other way? Dec 22 00:50:53 The image should be read-only on Ubuntu Touch Dec 22 00:51:32 Then why does it reboot? (android service interference??) Dec 22 00:51:41 very rarely it doesn't reboot. Dec 22 00:51:44 If lightdm fails to start 5 times, it will reboot Dec 22 00:51:51 Oh ok Dec 22 00:51:53 Now. Dec 22 00:51:54 You should be able to break that by doing `stop lightdm` Dec 22 00:52:23 Otherwise no android interference with ofonod? Dec 22 00:52:36 It would probably reboot faster if that was the case Dec 22 00:52:46 Like, not get far enough to hit lightdm Dec 22 00:52:51 😂😂 Dec 22 00:53:15 Because I think that the watchdog is actually in Android and ofono touches it the wrong way, then panic Dec 22 00:53:29 Everything I say is an educated guess Dec 22 00:53:30 ok Dec 22 00:53:59 There has to be a god for a blind...😜 Dec 22 00:54:10 We're all blind Dec 22 00:54:20 Eh. Dec 22 00:54:23 Oh, on the phone, can you do `ls /android`? Dec 22 00:54:33 yes Dec 22 00:54:48 boot, cache, firmware etc.... only folders Dec 22 00:54:58 But firmware is mounted. That's good. Dec 22 00:55:16 yup Dec 22 00:55:19 Could I get you to grab the snippet of logcat that comes out when you run `unity-system-compositor`? Dec 22 00:55:30 ahm. Dec 22 00:55:40 debuggerd is not running I guess. Dec 22 00:56:00 Nah nah it is fine. Dec 22 00:56:01 Sounds like the container is not as happy as you think it is Dec 22 00:56:15 @UniversalSuperBox, No no it is really happy Dec 22 00:58:57 (Document) https://irc.ubports.com/U3ZdnwWL/file_3053 Dec 22 00:59:07 wew Dec 22 00:59:24 My bad... I'll paste it. Dec 22 00:59:35 I can open this Dec 22 00:59:40 ok Dec 22 00:59:41 Gedit is surprisingly competent Dec 22 00:59:54 01-03 11:18:56.609 0 29551 E libEGL : validate_display:99 error 3008 (EGL_BAD_DISPLAY) Dec 22 00:59:57 emacs😎 Dec 22 00:59:59 yes Dec 22 01:00:02 Hmm... Hand-off! @mariogrip Dec 22 01:00:58 is this mir or hybris tests? Dec 22 01:01:02 Mir Dec 22 01:02:10 do you have mir logs? Dec 22 01:02:35 @aki237, This is the unity...log Dec 22 01:02:45 @aki237, lightdm.log Dec 22 01:02:57 @aki237, yeah, the unity log is perfect Dec 22 01:03:36 perfect in the sense? expected errors or no errors? Dec 22 01:04:20 It segfaults btw Dec 22 01:04:36 ah, is this 5.1 or 7.1? Dec 22 01:04:38 @aki237, You said bootlogo... The Ubuntu one or your phone's? Dec 22 01:04:40 7.1 Dec 22 01:04:51 @UniversalSuperBox, Phone's Dec 22 01:04:54 k Dec 22 01:04:56 oh? so it shows ubuntu logo? Dec 22 01:05:05 ah! have you added udev rules? Dec 22 01:05:08 No No vendor logo (Zuk Z1) Dec 22 01:05:19 I'll check... ran the script. Dec 22 01:05:50 Are you using the caf rootfs? Dec 22 01:05:58 yes caf. Dec 22 01:06:17 do you have wifi? Dec 22 01:06:37 (asking so you can install mir-demos) Dec 22 01:06:46 This is my udev 70-ham.rules Dec 22 01:06:51 I have network. Dec 22 01:06:54 usb tethering. Dec 22 01:08:21 can you try Dec 22 01:08:26 ```echo 1 > /dev/wcnss_wlan … echo sta > /sys/module/wlan/parameters/fwpath``` Dec 22 01:08:36 then do `nmcli d` Dec 22 01:08:44 If that doesn't work, i have an idea Dec 22 01:09:03 might be broadcom, or missing the fw Dec 22 01:09:24 No qcom only it worked fine in the halium rootfs. Dec 22 01:09:41 now I have internet in the phone. It is downloading the mir-demos Dec 22 01:09:51 installed mir-demos. Dec 22 01:10:12 @aki237, https://paste.ubuntu.com/26230298/ Dec 22 01:11:30 @mariogrip, write error , operation not permitted. Dec 22 01:11:52 Oh, you didn't have wi-fi yet? Dec 22 01:12:04 I had it in halium. here this error. Dec 22 01:12:05 You might consider setting CONFIG_FW_LOADER_USER_HELPER to y Dec 22 01:12:15 ok Dec 22 01:12:23 It worked for me, at least Dec 22 01:12:34 starring. Dec 22 01:12:41 thanks self for making good notes Dec 22 01:21:36 @UniversalSuperBox Have installed the mir-demos. Then Dec 22 01:21:41 Well Dec 22 01:22:02 First let's run `mir-demo-standalone-render-to-fb` Dec 22 01:22:05 Maybe that's its name Dec 22 01:22:11 It's close, at least. Probably. Dec 22 01:23:09 https://paste.ubuntu.com/26230365/ Dec 22 01:23:28 Hmm, what was the correct command Dec 22 01:23:30 notes Dec 22 01:24:06 Can you run `mir_demo_server` itself? Dec 22 01:24:12 We dont need any gl for fb output right? Dec 22 01:24:18 Nope Dec 22 01:24:34 Aha Dec 22 01:24:45 It should say "using platform: android" or something similar in there somewhere Dec 22 01:24:55 I, uh... don't know how it determines its platform Dec 22 01:26:29 @UniversalSuperBox https://paste.ubuntu.com/26230371/ backtrace like for the mighty crash Dec 22 01:27:35 Exception while creating platform. That's no doubt related to the EGL error from earlier Dec 22 01:27:58 ok Dec 22 01:28:04 I'm guessing that Mir is trying to ask the display driver to do something with the wrong parameters, like an incorrect resolution Dec 22 01:28:14 Don't know how to fix it, though... Dec 22 01:43:07 Chill Dec 22 01:43:24 I havent slept in 18 hrs.. better hit the bricks. Dec 22 01:43:40 Yeah, take a good swing at it tomorrow Dec 22 01:44:27 Morning here😃 Dec 22 02:52:18 @bhushanshah would you be okay with removing the reboot command from halium-install? It's really just an inconvenience. **** ENDING LOGGING AT Fri Dec 22 03:00:01 2017 **** BEGIN LOGGING AT Fri Dec 22 03:00:02 2017 Dec 22 03:28:59 @UniversalSuperBox, Sure Dec 22 03:34:20 @bhushanshah, Noted, will get done. Dec 22 03:35:55 How do you feel about changing from the temporary folders to a location on /data? Problem is, on my device the RAM fills up due to the files and then tar gets killed Dec 22 03:36:23 Which is funny since the device has 4GB RAM. 😡 Dec 22 03:36:29 Thought your script or @vanyasem script fork does it already? Dec 22 03:36:47 It does, but that's not in halium-install Dec 22 03:36:57 Upstream Dec 22 03:37:15 Ah will have to look at PRs I guess Dec 22 03:37:48 There is a trade-off, of course. If the script doesn't finish correctly, /data/recovery can be left behind. That's avoided using temp mounts. Dec 22 03:39:19 @bhushanshah, Oh, the PR was on PM's rootstock fork Dec 22 03:40:46 There are some 6 month old doodads on Halium-install thoigh Dec 22 03:41:45 Hmmmm Dec 22 03:41:50 Will look into it today Dec 22 03:42:38 That raises a question... What is the purpose of halium-install? Should it be recommended (an RFC "should") to all Halium distros, or is it just for the reference rootfs? Dec 22 03:43:15 Recommended Dec 22 03:43:23 Does PM use it? Dec 22 03:43:37 If you are using reference rootfs spec e.g Dec 22 03:43:53 PM is using it, yes Dec 22 03:44:14 Oh... Well why does that fork exist? Dec 22 03:44:16 But fork in pm-flashtool have some extra goodies which I plan to upsteam soon Dec 22 03:44:22 Aha Dec 22 03:47:06 I got ubports-boot to boot the standard file layout of Halium today. Next is seeing why we can't use halium-install. Dec 22 03:49:25 Honestly, I'm warming up to ubports-boot now that it can be tinkered with locally. It's really nice to know that things are almost definitely in the right place once you hit your chosen `init` Dec 22 03:52:28 ubports-boot is complex code ... Dec 22 03:53:23 Once you know that everything relevant is in `touch`, it isn't as scary. Dec 22 03:54:14 And it feels a lot less complex than trying to mount overlays when you are inside the root that you're mounting on Dec 22 03:54:37 That's a recipe for disaster. Like doing your own open heart surgery. Dec 22 03:55:01 Overlays? Dec 22 03:55:13 Have you checked out the UT file layout? Dec 22 03:55:32 Ah yeah I know that Dec 22 03:55:40 Read-only rootfs Dec 22 03:55:49 And some writable paths Dec 22 03:55:54 Yeah Dec 22 03:56:30 I've had PoC for that on the hybris-boot Dec 22 03:56:44 What does it involve? Dec 22 03:56:56 Right now the boatloader is pretty dum Dec 22 03:57:33 (In what it achieves compared to what UT has, that is) Dec 22 03:57:37 I don't remember full details but I basically bind mounted dirs inside /target Dec 22 03:58:05 Does that involve udev? Dec 22 03:58:21 No why would it need udev? Dec 22 03:58:44 That's how ubports-boot finds the partitions so well Dec 22 03:59:15 No need for any fixup-mountpoints Dec 22 03:59:22 Well shrug ... Thats why I don't like ubports boot.. it includes mini libhybris inside you know? Dec 22 03:59:51 At least libhybris-common Dec 22 04:00:40 Looking at some of the other libs here, it seems that we could handle encrypted root. Dec 22 04:01:01 As much as I like concept of one initrd to rule all devices, it makes initrd too much cluttered Dec 22 04:01:58 But well yeah.. as long as you can get ubports-boot to boot normal rootfs and not mess with hybris-boot we can use it completely fine I guess Dec 22 04:03:01 So... What all goes in to hybris-boot? Dec 22 04:03:12 Just the system's busybox, Dec 22 04:03:20 Yes that's it Dec 22 04:03:35 I do like that simplicity Dec 22 04:04:59 Side note, have you gotten its bootsplash functionality working on any device? Dec 22 04:06:20 Ah yes.. it is umm worked on hammerhead for me earlier then it stopped working Dec 22 04:06:32 I haven't investigated yet why Dec 22 04:06:40 Framebuffers Dec 22 04:07:05 Android devices can't even get the simplest of things working Dec 22 04:10:46 Hmm. Maybe a majority of ubports-boot's work could be done in /sbin/preinit Dec 22 04:21:11 Here's an interesting question... How does the boat loader handle charge boots? Dec 22 04:21:55 Good question.. I haven't investigated it though Dec 22 04:22:23 Him. Dec 22 04:22:34 Good job, autocorrect Dec 22 04:24:09 Why have we separated the system.ing from the rootfs? I can see that complicating the 'production mode' proposed by jbb. Dec 22 04:28:27 Also, what is the need for fixing up every possible mountpoint when only userdata (and system in proposed production mode) is needed by hybris-boot? Dec 22 04:29:13 I know, I have a lot of questions. I just want to be super sure that I use my time in the best possible way for all Halium distros. Dec 22 04:36:08 @UniversalSuperBox, Generally, userdata is enough, but it's present there just in case in future someone want to use some partition in initrd like recovery e.g Dec 22 04:37:12 @UniversalSuperBox, Reason for separate system.img was to support multiboot.. so one system.img multiple rootfs Dec 22 04:46:15 @bhushanshah, i was playing with something related recently Dec 22 04:46:20 really loved that concept Dec 22 11:07:41 in theory .. if the rootfs img was more compact and the persistent volatile data more organzies one could easily have multiple versions of the same core rootfs and safely upgrade and downgrade between them Dec 22 11:11:43 and in that the system.img might be something that we hardly upgrade.. Dec 22 14:12:40 Ubuntu Touch already does deltas between the current ubuntu.img and the new one. The separated model would mean two (possibly three) deltas between them. Dec 22 17:46:29 @anpok, What do you mean with "Persistant volatile storage more organized"? Dec 22 18:37:55 system-data user-data .. android-data Dec 22 18:38:17 How would you prefer to see it? Dec 22 18:38:58 Ubuntu Touch can already switch between images and, as long as there hasn't been a huge change between them, use the data as it was before Dec 22 18:39:05 That's why you can downgrade from daily to stable Dec 22 18:39:37 Heck, I was able to take a phone from 15.04 to 16.04 and back again with no consequences Dec 22 18:40:17 Oh I think for ubports it is fine right now.. I just thought if we would have different roofs .. or rather rootfs from different sources .. i.e. add plasma.. Dec 22 18:40:40 Oof, that's where it becomes really nice to have something before init handling the mounts Dec 22 18:40:57 Because we could just mount a different directory as the overlay Dec 22 18:41:00 yes Dec 22 18:42:28 Oh no, it's like competing solutions all have their given upsides and downsides! Dec 22 18:44:02 How does PostMarket handle their pre-boot stuff? Dec 22 18:44:18 It's really interesting Dec 22 18:44:40 Is this s/interesting/hacky/ or s/interesting/beautiful/? Dec 22 18:44:47 Beautiful Dec 22 18:45:02 searches documentation Dec 22 18:45:06 It can do on screen keyboard for full disk integration Dec 22 18:45:12 And some nice goodies Dec 22 18:45:18 Look at their wiki Dec 22 18:45:20 But how? Dec 22 18:45:28 Not how to look at their wiki Dec 22 18:45:35 It's two initrd concept Dec 22 18:45:54 There is one really minimal initrd in the boot.img Dec 22 18:46:07 And then there are two sub partition of rootfs Dec 22 18:47:06 Ahh, I was thinking of something a little more monolithic. Initrd -> chains to something like 'Halium.img' -> chains to real OS Dec 22 18:47:44 PostMarket is against Libhybris, no? Dec 22 18:47:49 Some interesting parts is, it's flashing mechanism.. Dec 22 18:48:06 Yeah they mostly try to use opensource drivers Dec 22 18:48:48 I managed to boot plasma mobile on pmos today Dec 22 18:48:54 :> Dec 22 18:49:35 Oh, it builds the OS for your phone before it flashes? Dec 22 18:49:49 Yes Dec 22 18:49:55 Er, okay Dec 22 18:50:30 Well it's flashing mechanism is flexible enough to abstract out though Dec 22 18:50:43 I am highly interested in using it for plasma mobile Dec 22 18:50:58 Do you see it as a replacement for Halium? Dec 22 18:51:25 Not replacement for halium Android stack but just from infrastructure PoV Dec 22 18:51:35 ? Dec 22 18:51:40 I don't quite follow Dec 22 18:52:35 Well given bugs in rootstock and things.. I would love to replace it with something more robust which is used by pmos Dec 22 18:52:52 Without the actual 'alpine Linux' part? Dec 22 18:53:01 (parts include flashing mechanism and initial bootup) Dec 22 18:53:03 Yes Dec 22 18:53:15 looks at initrd work Dec 22 18:53:27 Ever feel like everything is a moving target, always? Dec 22 18:54:14 Well my idea is really long term Dec 22 18:54:20 Don't worry Dec 22 18:54:28 Anyway, you said there are two init stages. Do you have an example prebuilt image that contains this? Dec 22 18:55:04 I've it yes but at moment i'm on mobile Dec 22 18:55:10 Aha Dec 22 18:55:35 That could be a really interesting compromise that lets us all have our own partition layouts but One(tm) Halium Dec 22 18:55:44 s/partition/file/ Dec 22 18:56:28 trying (and failing) to get rid of insomnia atm Dec 22 18:56:50 Airplane Mode is fun Dec 22 19:08:59 hm looks like a single moving piece Dec 22 19:11:33 Hmm? Dec 22 19:20:34 nevermind .. just a wrong premature statement Dec 22 19:20:51 on the postmarket os wiki Dec 22 19:37:15 @UniversalSuperBox what are the differences between plain xenial and xenial 7.1 caf sudoku? Dec 22 19:37:53 One is built with CAF headers, the other is not Dec 22 19:38:03 7.1 CAF, specifically Dec 22 19:38:58 The regular image can be used on 5.1, 5.1-caf, and 7.1 Dec 22 19:39:06 CAF is only needed for 7.1-caf devices **** ENDING LOGGING AT Sat Dec 23 03:00:02 2017 **** BEGIN LOGGING AT Sat Dec 23 03:00:03 2017 Dec 23 07:12:09 mvaisakh was added by: mvaisakh Dec 23 08:04:44 @UniversalSuperBox Hello... Dec 23 08:05:58 This error when trying to run any GL application … ```libEGL : validate_display:99 error 3008 (EGL_BAD_DISPLAY)``` Dec 23 08:14:17 what EGL platform is set to? Dec 23 08:14:36 You mean EGL_PLATFORM? Dec 23 08:14:43 yes Dec 23 08:14:52 should be EGL_PLATFORM=wayland after compositor is running Dec 23 08:15:03 ok Dec 23 08:15:42 none. when I run text_hwcomposer it says segfault Dec 23 08:16:14 *test. With platform as hwcomposer Dec 23 08:16:18 no, EGL_PLATFORM=hwcomposer test_hwcomposer Dec 23 08:16:27 yes I did Dec 23 08:16:35 so yoy don't have kwin or anything running yet, yes? Dec 23 08:16:43 nope nothing Dec 23 08:17:00 even lightdm is sleeping. No nothing Dec 23 08:18:03 did you use -caf or non-caf rootfs? Dec 23 08:18:09 caf Dec 23 08:18:35 https://paste.ubuntu.com/26237309/ is the logcat Dec 23 08:19:08 is it of test_hwcomposer or Mir? Dec 23 08:19:15 https://paste.ubuntu.com/26237310/ unity-system-compositor Dec 23 08:19:27 @NotKit, unity-system-compositor Dec 23 08:19:48 I suppose try to get test_hwcomposer first Dec 23 08:19:56 ok Dec 23 08:20:36 anything meaningful in strace of it? Dec 23 08:21:54 final few lines of strace of test_hwcompopser (platform EGL_PLATFORM=hwcomposer) https://paste.ubuntu.com/26237314/ Dec 23 08:22:40 and in logcat straight after running it? Dec 23 08:23:36 in logcat right after running it : https://paste.ubuntu.com/26237320/ Dec 23 08:24:25 you may want to run EGL_PLATFORM=hwcomposer gdb test_hwcomposer Dec 23 08:24:35 ok Dec 23 08:24:44 after it crashes, do "bt" Dec 23 08:25:26 https://paste.ubuntu.com/26237326/ Dec 23 08:35:00 should I be trying non-caf rootfs? Dec 23 08:36:45 the weird thing is that it crashes in linker Dec 23 08:36:50 If it helps SurfaceFlinger is not running. Dec 23 08:37:25 can you try to install libc6-dbg and run it again? Dec 23 08:37:59 also EGL_PLATFORM=hwcomposer valgrind test_hwcomposer sometimes might give more clues Dec 23 08:38:06 ok Dec 23 08:45:50 libc6-dbg even reduced the bt Dec 23 08:47:17 @NotKit https://paste.ubuntu.com/26237404/ valgrind Dec 23 08:49:54 crashes too early in valgrind Dec 23 08:50:02 ok Dec 23 08:50:03 btw, is your build based on 5.1 or 7.1 tree? Dec 23 08:50:07 7.1 Dec 23 08:50:08 I wonder why mm linker Dec 23 08:50:32 Isn't that required in 7.1? Dec 23 08:50:41 but now there is n linker Dec 23 08:51:00 what do you get in gcc with symbols installed? Dec 23 08:51:09 ? Dec 23 08:51:26 Fwd from NotKit: you may want to run EGL_PLATFORM=hwcomposer gdb test_hwcomposer Dec 23 08:51:27 Fwd from NotKit: after it crashes, do "bt" Dec 23 08:51:28 Fwd from NotKit: can you try to install libc6-dbg and run it again? Dec 23 08:51:37 *gdb, sorry Dec 23 08:51:53 gdb-dbg? Dec 23 08:53:04 @NotKit after lib6-dbg https://paste.ubuntu.com/26237424/ Dec 23 08:53:54 check pid with "info prog" Dec 23 08:54:16 then cat /proc/$pid/maps Dec 23 08:55:25 (don't exit gdb, of course) Dec 23 08:57:21 @NotKit mmaps : https://paste.ubuntu.com/26237440/ Dec 23 08:58:55 0xb4e875a4 Dec 23 08:59:32 ? Dec 23 09:01:03 the crash probably happens in that library, but it's vendor blob Dec 23 09:02:53 @NotKit the hwcomposer ran fine in the halium rootfs. Dec 23 09:02:58 Why now? Dec 23 09:03:02 Any ideas. Dec 23 09:03:14 so you switched to Ubuntu rootfs? Dec 23 09:03:21 maybe different libhybris versions Dec 23 09:03:27 Ok Dec 23 09:03:29 Is Android container running? lxc-info may tell Dec 23 09:04:11 i think it should be, as logcat is working Dec 23 09:04:13 Running. Dec 23 09:04:34 And is it full functional? Dec 23 09:04:47 How would I test it? Dec 23 09:04:50 Can you give pstree for the Android init process? Dec 23 09:06:25 @bhushanshah https://paste.ubuntu.com/26237471/ Dec 23 09:06:59 Looks fine .. Dec 23 09:07:45 It's the time I suppose I hand over torch to @UniversalSuperBox or @mariogrip .. I don't know much about ubports Dec 23 09:08:37 @bhushanshah just in case, does latest Halium rootfs use Nougat linker merged upstream, or still MM one? Dec 23 09:10:00 It still uses MM, their libhybris is bit outdated I think Dec 23 09:10:19 Should I update it? Dec 23 09:10:41 Won't fix stuff, but will break more Dec 23 09:10:46 Damn. Dec 23 09:10:54 That said Dec 23 09:11:06 Do you see something in dmesg regarding firmware? Dec 23 09:11:11 Can I install stuff and entend from halium footfs. Dec 23 09:11:20 grep for fw/firmware Dec 23 09:11:50 @aki237, It won't work that way well, given ubports have libhybris fork with Mir patches Dec 23 09:11:55 Grep => empty Dec 23 09:12:03 ok Dec 23 09:12:22 When you run dmesg -w and start test_hwcomposer Dec 23 09:12:28 What do you see in dmesg? Dec 23 09:12:34 ok Dec 23 09:13:15 ```[ 4940.117653] mdss_fb_release_all: AD shutdown cleanup failed ret = -22``` Dec 23 09:13:25 That was the different one. Dec 23 09:29:32 @bhushanshah What linker does halium rootfs use? mm or n? Dec 23 09:29:45 n Dec 23 09:30:48 so ubports is older than halium? Dec 23 09:31:55 yes Dec 23 09:32:07 But I doubt linker makes diference here Dec 23 09:32:19 given your device is not that new Dec 23 09:32:31 (i.e came pre-installed with android 7 I mean) Dec 23 09:32:43 no almost any application that uses android service, segfaults at the end. Dec 23 09:32:58 @bhushanshah, 5.1 Dec 23 09:59:06 @aki237, @mariogrip any idea, happens when accessing the framebuffer device. (test_hwcomposer) Dec 23 10:59:26 i finally realized what was wrong with the halium install script on my device - system.img was bigger then the size specified to make_ext4fs Dec 23 10:59:28 dammit Dec 23 17:33:46 Hi. When building Halium, and i will build on aosp, and on for example ressurection remix, do i will see some advantage of RR? or it will be same? Dec 23 17:45:34 @gouster4, You should avoid RR if possible .. Dec 23 17:46:43 Using RR tree will provide no additional benifit to halium port Dec 23 19:23:11 XavierXX was added by: XavierXX Dec 23 19:33:33 you can't easily build on AOSP though Dec 23 19:33:44 unless you port LineageOS to device or Halium patches to AOSP tree Dec 23 19:34:37 Although you can maybe build with the AOSP devce tree in the Halium LineageOS based tree Dec 23 19:47:26 So is the oneplus one supported by halium currently I think not right? ;( Dec 23 19:55:39 AstonishedByTheLackOfNeko was added by: AstonishedByTheLackOfNeko Dec 23 19:56:04 Hi guys I have an experimental aosp branch up for oneplus 2 … Does it boot? Who knows but it's certainly compiling Dec 23 20:23:11 the same … oneplus2 … I have compiled Halium for oneplus2 yesterday … not yet tested Dec 23 20:25:12 which aosp branch do you have? Dec 23 20:28:18 why AOSP? Dec 23 20:28:44 I had Halium booting to UI on OP2 Dec 23 20:29:56 nice :D … (not AOSP for me in fact... ) … I got sources for halium github Dec 23 20:31:54 Is there any difficulty for halium system.img build and ubports-boot.img build to a usable UI? Dec 23 20:32:12 Sorry for my poor english Dec 23 20:33:18 better ask @UniversalSuperBox about UBPorts Dec 23 20:36:29 ok I will ask … and for pure halium? any problem? Dec 23 20:39:55 Waldbursche was added by: Waldbursche Dec 23 20:40:08 Vince you can check this channel logs to find out :) Dec 23 20:40:55 lots of them at first, but they are solvable with enough dedication Dec 23 21:04:46 @Vince, HaruhiOS Oreo bringup branches are fully aosp Dec 23 21:05:19 I think the only issue I ever saw was the audio HAL Dec 23 21:05:30 The rest of the source is totally untouched Dec 23 21:44:45 @NotKit, ok I will see the for the help :) Dec 23 21:45:05 *thanks for the help Dec 23 22:44:58 petya230 was added by: petya230 Dec 23 22:45:15 Hi! :D Dec 23 22:45:20 hi **** ENDING LOGGING AT Sun Dec 24 03:00:06 2017 **** BEGIN LOGGING AT Sun Dec 24 03:00:07 2017 Dec 24 09:12:47 @JBBgameich, If pure aosp will be problem, i will be not able to build, or i can buld it and cannot get it workin on phone? Dec 24 09:36:00 fatal: missing blob object 'a1a9a0372e9cdad8d75d7953ae20d35c258231a9' … error: https://github.com/Halium/android_frameworks_base did not send all necessary objects … error: Cannot fetch Halium/android_frameworks_base … warn: --force-broken, continuing to sync … Fetching projects: 100% (200/200) … error: Exited sync due to fetch errors Dec 24 09:36:01 wtf? Dec 24 13:10:09 arpitjindal1997 was added by: arpitjindal1997 Dec 24 14:04:44 hell0, pulseaudio gives me these errors https://paste.pound-python.org/show/JNKG9K42UgyiX0VhcWk5/ Audio hal starts,cuz i can see it in logcat Dec 24 14:29:30 i fixed that by removing .config/pulse Dec 24 14:29:59 now, what's this: "FIXME: Setting AUDIO_DEVICE_IN_BUILTIN_MIC as initial device." in pulseaudio -vvvvvvvvvvvvvv ? Dec 24 14:35:35 i noticed a tfa98xx: ERROR vendor/samsung/common/frameworks/audio/tfa_amp/TFA9895/s_integration/lxI2c.c:lxI2cInit:137: Can't open i2c bus:/dev/i2c-1 in logcat Dec 24 14:35:39 rules* Dec 24 14:35:40 ill try to add it in udev ruled Dec 24 15:29:53 I got an audio from a nice man on boot Dec 24 15:29:58 seems working :D Dec 24 15:30:23 Yay Dec 24 15:31:06 (How can i disable that audio -_- ?) It scared me D: Dec 24 15:31:59 See plasma-phone script in /usr/bin Dec 24 15:33:07 oh, got it Dec 24 15:33:46 now, if i try to make the volume lower (or mute it) it returns to maximum Dec 24 15:35:24 and alsamixer says operation not permitted :/ Dec 24 15:36:08 oh, but after installed alsa-utils i can control it Dec 24 15:40:43 edit: the bar seems bugged like the brightness one was in the past Dec 24 15:47:51 Does camera work? Dec 24 15:48:28 Locally yes Dec 24 15:48:56 Ill package it next year Dec 24 15:49:54 ah ok, thanks Dec 24 16:38:02 can i upgrade rootfs? Dec 24 17:26:58 @bhushanshah does PM 'use' SELinux? Dec 24 17:35:13 No Dec 24 17:35:34 Do you plan to? Dec 24 17:35:51 I'm not quite sure Dec 24 17:36:10 You can take it as : maybe Dec 24 17:36:20 Okay, that presents a problem Dec 24 17:36:30 I know Dec 24 17:36:59 Can you load and activate a LSM after the kernel has started? Dec 24 17:37:30 No idea Dec 24 17:41:25 Otherwise we might need to look at how you can start another kernel from the already running kernel Dec 24 17:41:47 Or replacing the boot.img with another bootloader... yo dawg Dec 24 18:12:09 selinux sux Dec 24 18:25:58 Does ril work if you're so pro to port it? Dec 24 18:26:22 :( Dec 24 18:26:25 SELinux is awesome Dec 24 18:26:32 it's not even hard to work with Dec 24 18:26:52 Lol, #surro guy Dec 24 18:28:34 https://pastebin.com/r2ScRY34, im stuck trying to start the android lxc container. Does anyone know how to fix this error? Dec 24 18:29:20 Son_Goku: depends on context, when you're trying to run android in container it's pain Dec 24 18:29:40 For desktop usage... Sure it's awesome Dec 24 18:29:43 what type of container technology? Dec 24 18:29:58 usually, Android in container leads to MCS labeling Dec 24 18:30:04 Melp Dec 24 18:30:16 I thinl ure missing kernel configs cgroup Dec 24 18:30:41 Son_Goku: lxc Dec 24 18:31:06 iirc, LXC defaults to setting MCS labels relative to the host Dec 24 18:31:33 the main issue with SELinux (and indeed, all MAC systems) is that you can't instantiate a child set of policies in a container Dec 24 18:31:35 revertive: you need to disable res_ctlr or something like that option in kernel config Dec 24 18:31:46 i'm trying to start the rootfs build from https://github.com/Halium/rootfs-builder/tree/ubuntu Dec 24 18:32:12 oke ill try to find it in my kernel config Dec 24 18:32:34 oh blech Dec 24 18:32:36 the ubuntu kernel Dec 24 18:32:41 revertive: check hammerhead kernel halium-5.1 branch, there is commit named "some options needed for lxc" or something like that Dec 24 18:33:02 Son_Goku: worse.. Android kernel Dec 24 18:33:06 ugh Dec 24 18:33:22 bshah[m]: thanks, i'll look for the commit Dec 24 18:33:44 this is why I'm glad that distros like Fedora and openSUSE are working to make us less reliant on Android kernels for ARM hardware Dec 24 18:35:31 Is it legit to upgrade or it will trigger my r00tfs? Dec 24 18:37:11 cristo: I'm not sure.. maybe kwin is up-to-date or not... Most likely not though Dec 24 18:38:05 Hmm, OK, il backup and try. If it trigger rootfs il restore the old one Dec 24 19:13:16 > Does ril work if you're so pro to port it? Dec 24 19:13:17 ? Dec 24 20:23:43 AHA Dec 24 20:23:56 Never have I been more excited to see "GNU/Linux Device on rndis0 10.15.19.82" Dec 24 20:25:06 woohoo lxc@android container started, now i've got light/vibrator and for the first time sensors :p, now i've gotta find out why hwcomposer is missing libGLESv2S3D_adreno.so Dec 24 20:25:17 That's a red herring Dec 24 20:26:11 @UniversalSuperBox, ? Dec 24 20:26:29 @bhushanshah, Just got ubports-boot to stop clashing with hybris-boot when they're in the same tree Dec 24 20:26:39 Yay Dec 24 20:27:02 What was issue? Dec 24 20:27:55 Marius used some of the same variable names Dec 24 20:28:01 All of the same variable names, rather Dec 24 20:28:07 So the environment got polluted really fast Dec 24 20:29:03 EGL_PLATFORM=hwcomposer test_hwcomposer, give me a rotating spiral so thats somewhat working \o/ Dec 24 20:29:13 Wow Dec 24 20:29:28 Awesome. UBports kept booting too Dec 24 20:39:09 Okay, if some of you Halium party people have some time, I'd appreciate if you put ubports-boot into your source tree (as specified in https://docs.ubports.com/en/latest/porting/building-ubports-boot.html) and rebuilt hybris-boot to make sure it's not broken. You'll need to pull https://github.com/UniversalSuperBox/ubports-boot/tree/stop-halium-clash Dec 24 20:41:57 You'll also need to change from line 648 in `build/core/main.mk` so it looks like this: https://paste.ubuntu.com/26247503/ Dec 24 20:45:57 * revertive building Dec 24 20:51:42 UniversalSuperBox: it builds just fine Dec 24 20:52:19 both hybris-boot and ubports-boot Dec 24 20:54:48 Does hybris-boot still boot? Dec 24 20:55:27 The problem was that the UBports initrd was getting shoved into hybris-boot Dec 24 20:55:48 yup, boots just fine here Dec 24 20:56:07 \o/ Dec 24 21:02:36 it's a good night before christmas, hybris-boot and ubports-boot build together, and finely have a semi working halium build for Google Nexus 4 Dec 24 21:05:05 hm.. plasma mobile give me audio but a black screen but he the powerbutton dims the backlight so it's something Dec 24 21:08:02 There's a few more things before I'm really happy with ubports-boot Dec 24 21:08:09 But that was the most important onr Dec 24 21:36:50 Universal....: no Dec 24 21:40:05 k Dec 24 21:49:35 Interesting... Dec 24 22:03:29 ninja: error: '/home/petya230/halium/out/host/linux-x86/framework/signapk.jar', needed by '/home/petya230/halium/out/target/product/vns/obj/APPS/gnss_supl20service_hisi_intermediates/package.apk', missing and no known rule to make it Dec 24 22:03:36 i hate errors.. :l Dec 24 22:04:08 how can i solve this? Dec 24 22:21:33 Remove any APKs from makefiles Dec 24 22:21:36 And jars Dec 24 22:21:57 @UniversalSuperBox i've seen your pr, i will test it and leave my comments later today Dec 24 22:22:10 that will probably be tomorrow for you Dec 24 22:38:30 Works for me Dec 24 23:00:34 i was the one to report it in the first place, i have to double-check that it's fixed Dec 24 23:00:44 Yep, that's why I mentioned you Dec 25 00:25:58 Merry Christmas!! Dec 25 01:09:40 Merry 71 **** ENDING LOGGING AT Mon Dec 25 03:00:01 2017 **** BEGIN LOGGING AT Mon Dec 25 03:00:01 2017 Dec 25 03:36:21 Guys can I install plasma mobile from ref rootfs? Dec 25 03:41:33 Yeah Dec 25 03:41:47 Erm actually no Dec 25 03:42:19 Instead get latest CAF rootfs from images.plasma-mobile.org Dec 25 03:42:29 systemd? Dec 25 03:43:51 ? Dec 25 03:44:17 The rootfs what is it based on? Dec 25 03:44:29 Ah, systemd yes Dec 25 03:44:44 Ubuntu or ? Dec 25 03:44:58 Ubuntu 16.04 Dec 25 03:45:07 Ok Dec 25 03:45:47 Wayland? Dec 25 03:46:26 yes Dec 25 03:48:36 I have to boot with hybris boot right? Dec 25 03:49:04 correct Dec 25 04:02:21 Apparently I didn't completely fix it. hybris-boot won't tell me where it thinks `/data` is when I build it, and ubports-boot tells me twice. Maybe I'll just resume work on booting Touch out of hybris-boot. Dec 25 04:02:45 ubports-boot gets the partition wrong btw. Dec 25 04:02:49 @UniversalSuperBox, 7.1? Dec 25 04:02:55 Yeah Dec 25 04:03:28 @UniversalSuperBox, does that mean I don't have to test your pr Dec 25 04:03:57 You can try, maybe I've screwed up my tree Dec 25 04:35:07 what is the default login account in plasma? (phablet in ubports) root login is disabled. Dec 25 04:36:36 phablet also Dec 25 04:36:52 you have to set the password for yourself though Dec 25 04:39:08 Any kernel hackers in the audience... Are apparmor or selinux activated simply by loading in their kernel modules? That is, could we ship both apparmor and selinux as modules outside of boot.img and allow the OS to load its desired one in in its early boot process, Dec 25 04:39:51 Without modifying the kernel cmdline, since that's impossible for us Dec 25 04:40:01 After shipping the image Dec 25 04:40:11 selinux can be built in and disabled by default Dec 25 04:40:19 then activated on runtime Dec 25 04:40:27 yes Dec 25 04:40:46 That's handy. Maybe apparmor has something similar. Dec 25 04:40:56 guys, in systemd, is rsyslogd needed? Dec 25 04:40:59 Do you have a page of documentation describing that? Dec 25 04:41:19 That is filling up my varlog very fast and no space left on device eventually Dec 25 04:41:47 Well, that would be happening without rsyalogd I suppose. Dec 25 04:42:36 Sounds like you have some android services that need to be fixed or stopped. Dec 25 04:42:54 Nah that is the case. rsyslogd's 5th and 6th fd corresponding to varlog/{syslog,kern.log} Dec 25 04:42:56 Oh ok Dec 25 04:43:19 But is it needed? if journal is running for systemd. Dec 25 04:43:24 @UniversalSuperBox http://eckucukoglu.com/tag/setenforce/ Dec 25 04:43:25 If you feel it necessary, disable it and see what happens. Dec 25 04:43:26 maybe this Dec 25 04:43:34 It stops the logging. Dec 25 04:47:36 @NotKit, Interesting. I wonder if selinux is still loaded in that case. My reading on LSM seems to tell me that only one module may be loaded at a time, otherwise Bad Things happen. Dec 25 04:48:50 Also, it says that relabeling must take place before re-enabling, else the kernel will crash. That seems concerning. And probably means that modularizing the LSMs for Halium won't work either. Dec 25 04:54:26 From http://wiki.apparmor.net/index.php/FAQ, " While technically possible to stack two LSMs, both LSMs must cooperate." Dec 25 04:55:17 That can read a lot of ways... But I'd like to read it as "if one LSM rolls over while the other actually does some work, the kernel will not panic" Dec 25 05:18:49 Guys I have flashed the plasma mobile rootfs using halium-install. hwcomoser works. Dec 25 05:19:02 but no dm running, only boot logo Dec 25 05:19:21 Have you added udev rules? Dec 25 05:19:27 Yes Dec 25 05:19:42 And test_egl* segfaults. Dec 25 05:19:58 hm Dec 25 05:20:20 I do wonder if CAF rootfs is functional at this point Dec 25 05:20:41 Yeah. Same happened for ubuntu too Dec 25 05:20:52 atleast here, the hwcomposer works. Dec 25 05:20:55 and wifi too Dec 25 05:21:02 similar to halium rootfs. Dec 25 05:21:19 Can you check if kwin_wayland process is running? Dec 25 05:21:31 And also logcat please Dec 25 05:21:32 ok Dec 25 05:23:10 logcat for test_egl with platform hwcomposer https://paste.ubuntu.com/26249546/ Dec 25 05:23:33 test_egl is broken test Dec 25 05:23:39 Ok Dec 25 05:25:40 @bhushanshah kwin_wayland isn't running but, when run it from shell no segfaults. :) Dec 25 05:26:18 How do you run it from shell? (CLI options?) Dec 25 05:26:28 nope. Dec 25 05:27:05 just kwin_wayland. given I have setup some needed env. $(dbus-launch) and some needed for QT Dec 25 05:27:21 @aki237, and boot logo is flushed. Dec 25 05:27:27 It won't really fly though Dec 25 05:27:37 ? Dec 25 05:27:57 Well running kwin_wayland from shell won't do much useful Dec 25 05:28:15 It needs to be run from inside logind session Dec 25 05:28:31 ok Dec 25 05:28:41 What is the logind service here anyway? Dec 25 05:28:47 kdm? Dec 25 05:29:07 can you give me journalctl | grep simplelogin Dec 25 05:29:45 https://paste.ubuntu.com/26249570/ Dec 25 05:29:53 Damn. Dec 25 05:30:23 Interesting Dec 25 05:30:34 Does test_egl_configs pass? Dec 25 05:30:40 will see Dec 25 05:30:55 That works fine. Dec 25 05:30:56 With EGL_PLATFORM env set to hwcomposer Dec 25 05:31:27 yes Dec 25 05:31:43 Hmmm Dec 25 05:31:45 no segfaults or anything. Dec 25 05:32:08 But it prints lot of text, right? Dec 25 05:32:16 (different egl configs) Dec 25 05:32:21 Yes Dec 25 05:32:32 Okay Dec 25 05:32:37 Let's try this Dec 25 05:32:49 ok Dec 25 05:32:59 `kwin_wayland --xwayland --hwcomposer` Dec 25 05:33:08 And see if it segfaults? Dec 25 05:33:16 logcat says EGL_BAD_PARAMERTER while running kwinwrapper. Dec 25 05:33:25 (unset EGL_PLATFORM) Dec 25 05:33:57 @bhushanshah, I didn't do this. Just waited for sometime for kwinwrapper to work it works. Dec 25 05:35:04 (running as root.. fine?) Dec 25 05:36:26 No, you should run it as ohab Dec 25 05:36:31 *phablet Dec 25 05:37:08 @bhushanshah running as phab.. segfault. root : https://paste.ubuntu.com/26249597/ waiting at clean up Dec 25 05:37:25 If it works fine with root user and not phablet, I believe udev rules are at play Dec 25 05:37:35 ok Dec 25 05:37:41 I'll check and try again Dec 25 05:38:48 What's permission of /dev/kgsl-3d0? Dec 25 05:39:29 @bhushanshah oops. udev was the problem. Dec 25 05:39:44 70-ham{garbage}.rules Dec 25 05:39:55 Whats the default password? Dec 25 05:41:36 What you set for phablet Dec 25 05:41:59 alphanum.... here num keyboard only Dec 25 05:42:26 Time to reset 😛 Dec 25 05:42:36 I have ssh. Dec 25 05:42:53 scalig is bad. Dec 25 05:43:58 (Document) https://irc.ubports.com/2qIeuz7j/file_3165.jpg Dec 25 05:44:45 Ugh Dec 25 05:45:19 What's resolution and DPI? Dec 25 05:46:27 1080p, 401 ppi... Dec 25 05:46:46 Okay Dec 25 05:47:09 @bhushanshah, ? flash again? or passwd? Dec 25 05:47:18 Password Dec 25 05:48:06 reset to 1234. I have entered 1234 in the box. no login. where is the enter key?😭 Dec 25 05:48:54 Wtf Dec 25 05:49:02 O.o Dec 25 05:49:56 From ssh can you kill kscreenlocker_greet process? Dec 25 05:50:44 Now I get a unlock button...😂😂😂 Dec 25 06:03:03 O.o plasma is so away from mobile. Dec 25 06:03:46 Sure, Plasma mobile is not ready for end users at all Dec 25 06:03:57 Is there anything else. Dec 25 06:04:14 ? Dec 25 06:04:22 I should wait for a newer utouch based on halium rootfs. Dec 25 06:04:47 I wonder if your problem with ubports rootfs was also udev rules? Dec 25 06:05:26 That I checked. There almost any program ran from root which relied on the android lxc, segfaulted no matter what. Dec 25 06:06:59 I utouch also like plasma mobile? Dec 25 06:07:27 I believe it's more user ready then plasma mobile Dec 25 06:07:34 huh Dec 25 06:08:11 why wont we extend the halium rootfs to a utouch rootfs? plasma mobile is more or less like it. Dec 25 06:42:09 @bhushanshah can I run gnome shell in plasma environment (--replace) Dec 25 06:42:43 No Dec 25 06:42:52 ok Dec 25 09:27:48 @UniversalSuperBox, Which makefiles? Dec 25 09:52:01 @petya230 like this https://github.com/TheMuppets/proprietary_vendor_zuk/search?utf8=%E2%9C%93&q=apk&type= Dec 25 09:52:34 oh, okay! Dec 25 10:02:46 @aki237, https://github.com/OpenKirin/android_device_huawei_vns/search?utf8=%E2%9C%93&q=.apk&type= Dec 25 10:02:47 om. Dec 25 10:03:25 i dont have problems with doze, coz i commented this :D Dec 25 10:03:30 sooo Dec 25 10:03:46 gnss_supl20service_hisi_intermediates Dec 25 10:03:51 >< Dec 25 13:33:03 okay, i solved this. :D … But --> ninja: error: '/home/petya230/halium/out/target/product/vns/obj_arm/EXECUTABLES/updater_intermediates/updater', needed by '/home/petya230/halium/out/target/product/vns/obj/PACKAGING/target_files_intermediates/lineage_vns-target_files-78ea6a3818.zip', missing and no known rule to make it Dec 25 13:34:25 @UniversalSuperBox works for me .. Dec 25 13:38:20 but do you have an idea why the list of mounted filesystems differ sometimes: Dec 25 13:38:24 https://paste.ubuntu.com/26251163/ Dec 25 13:38:37 this is at break=bottom Dec 25 13:39:44 there is especially no /etc/fstab.. Dec 25 13:40:34 out of curiosity, @anpok what's your "rootfs" ? Dec 25 13:41:02 xenial sudoku caf armhf Dec 25 13:43:30 interesting Dec 25 17:51:52 Melp Dec 25 17:52:01 Plasma mobile fo life Dec 25 18:11:38 meh bootmode is charger .. init does something elese.. Dec 25 18:13:03 I still do not see how the actual kernel cmd line is assembled.. some parts from the boot image are forwarded.. others filtered out... and a lot of stuff is added.. Dec 25 18:55:10 @UniversalSuperBox It seems that in some cases the system adds androidboot.mode=charger then there is no /etc/fstab as tmpfs and the directory overlays do not happen Dec 25 18:59:22 Oh. Dec 25 18:59:33 Yeah, on a charger boot it boots Android Dec 25 19:01:17 at least that explains the differences in how init mounted stuff Dec 25 19:01:26 do we need that mode in init? Dec 25 19:02:24 (it does not solve my issues though) Dec 25 19:24:39 @anpok, Yes Dec 25 19:25:17 Otherwise the charger plugin boots into the OS, which is undesirable Dec 25 19:25:48 Hybris-boot needs to gain that capability Dec 25 19:26:04 Or maybe a boot shim that we get... Dec 25 19:29:04 Yeah... Maybe that'll be the best idea. Dec 25 19:30:41 I see.. I thought that part is handled without the boot.img Dec 25 19:31:18 ubports-boot takes the Linux kernel approach Dec 25 19:31:47 Monolith Dec 25 19:33:41 Everything is in place before Upstart even knows its name. Dec 25 19:34:07 hm .. but that means we can ubports-ify the charging animation Dec 25 19:34:27 Android can handle its charging better Dec 25 19:34:32 :) Dec 25 19:35:22 SurfaceFlinger or whatever is used in the charging animation comes up faster Dec 25 19:40:22 I'm still thinking on that boot shim idea... Keeps us using hybris-boot, doesn't require rearchitecting the boot order Dec 25 19:44:58 Something like what Bhushan was saying that postmarket does. Chain from hybris-boot, to shim that runs udev and mounts everything, into UT Dec 25 23:58:59 Is there any danger of different devices of the same model having different `/dev/block/mmcblk0pXX` nodes for a single partition? Dec 25 23:59:26 I doubt it Dec 25 23:59:31 Good Dec 26 00:00:49 In that case, would there be any harm in replacing the fstab in the built system.img with one that's fixup-ed? Dec 26 00:02:52 Um possibly.. it may break the service mounting Android partitions depending on how fragile code I may have written Dec 26 00:03:10 In any case, try and see? Dec 26 00:03:36 Hey yeah, works for me Dec 26 00:10:09 bshah, I'll just be over here. Messages are taking 1min+ to send right now Dec 26 00:10:28 UniSuperAlt: same for me Dec 26 00:10:54 What's last communication you heard from me? Dec 26 00:12:03 >Okay, now curious question, why you want to do that? Dec 26 00:12:37 Okay Dec 26 00:13:04 After that I wrote: ah classic chicken egg issue Dec 26 00:13:12 Yep. Dependency hell. Dec 26 00:13:26 Circular, whatever Dec 26 00:13:30 Dalton Durst: Dec 26 00:13:30 But those can't be mounted unless udev has put block devices in the by-name and such places Dec 26 00:13:31 And last I heard this from you Dec 26 00:13:42 That's correct Dec 26 00:14:36 I'm also thinking in terms of having a possible multiboot or standardized boot situation. Halium could provide a sanitary fstab to the linux distro, in case they don't have fancy udev rules. Dec 26 00:14:49 Somehow Dec 26 00:18:17 UniSuperAlt: I've also stumbled upon huge startup delay in the bullhead due to this circular dep fun Dec 26 00:18:37 For Ubuntu Touch or...? Dec 26 00:18:49 Plasma mobile Dec 26 00:19:02 Ah. What do you think is wrong in your case? Dec 26 00:19:42 Well it's container blocking udev rules Dec 26 00:20:05 Ahh. So it's not just us, then. Dec 26 00:20:08 That's a relief. Dec 26 00:20:12 then udev service timeouts and then kwin starts Dec 26 00:21:06 The problem in our case is that Android mounts `can't` be done until after the container starts Dec 26 00:21:10 Because we don't know where they are Dec 26 00:21:24 At least you can boot. :P Dec 26 00:22:14 In any case lets discuss this on projectmanagement issue.. I need to to ask for bit.. can you open issue? Dec 26 00:22:34 You bet. About what, exactly? Dec 26 00:23:11 About circular dep of udev and lxc Dec 26 00:24:41 On it Dec 26 00:32:00 bshah[m], why does udev need to start after the container anyway? Dec 26 00:32:28 I have an idea (devices aren't ready until after the container brings them up), but not sure if that's right Dec 26 00:33:38 That's what plasma mobile is doing Dec 26 00:33:47 Right, but why? Dec 26 00:33:52 Why not just swap the order Dec 26 00:34:04 I figure there's a reason, but not what the reason it Dec 26 00:34:08 is* Dec 26 00:34:59 I mean, swapping order is what reference rootfs and plasma mobile doing Dec 26 00:35:23 But it takes forever then for udevadm to settle Dec 26 00:35:38 Solution for it is to how utouch does it Dec 26 00:35:49 Which brings it's own set of problems Dec 26 00:36:52 Ohh Dec 26 00:40:37 https://github.com/Halium/projectmanagement/issues/33 **** ENDING LOGGING AT Tue Dec 26 03:00:00 2017 **** BEGIN LOGGING AT Tue Dec 26 03:00:01 2017 Dec 26 03:16:50 @bhushanshah I'm still in the process of testing this, but would you be opposed to adding a switch (possibly as a file in the rootfs.img) that will cause hybris-boot to remount `/target` as read-only before executing init? Dec 26 03:20:37 ubports-boot handles this in a fail-closed configuration, where the presence of a file causes it to remount the image as rw. I personally think that'd be the better solution. However, if Ubuntu Touch is the weird one with a read-only root, I guess there'd be a majority rule there. Dec 26 03:23:32 @UniversalSuperBox, I don't think I've issues there Dec 26 03:23:48 Does PM do a read-only root? Dec 26 03:24:02 No Dec 26 03:29:53 Darn Dec 26 04:31:15 What happened? Dec 26 08:50:44 SashkaT was added by: SashkaT Dec 26 08:51:29 @SashkaT, welcome :P Dec 26 08:51:57 Let's make an impression that there are only Russians in this chat :P Dec 26 08:52:39 водка Dec 26 08:53:06 Enough 😂 Dec 26 08:54:51 @SashkaT, is your phone on snapdragon? Dec 26 08:55:12 No. Dec 26 08:55:29 which SOC then Dec 26 08:55:41 it sucks though, everything except snapdragon is a pain to work with Dec 26 08:55:45 My CPU is Qualcom MSM8974 … (Krait 400) Dec 26 08:56:10 isn't this Qualcomm Snapdragon Dec 26 08:56:24 Oh Dec 26 08:56:36 come on 😂 Dec 26 08:56:49 @SashkaT, it should be relatively easy to port to in that case Dec 26 08:57:34 Yeah Dec 26 08:57:56 It's Qualcomm snapdragon 800 Dec 26 08:57:59 read the docs at halium.org, the ones at the UBports website are outdated Dec 26 08:58:18 I will be more than glad to help you if you encounter any issues, but be sure to read the docs at least 4 times Dec 26 08:58:49 Of couse Dec 26 09:29:30 @SashkaT, Which device? Dec 26 09:29:41 hlte Dec 26 09:29:50 Samsung galaxy note 3 LTE Dec 26 09:29:56 SM-N9005 Dec 26 09:30:33 Qcom SD800? Dec 26 09:30:41 Yeah Dec 26 09:30:43 @SashkaT, YES Dec 26 09:31:15 isn't this the same chipset used in n5 Dec 26 09:31:25 Yes Dec 26 09:31:27 good luck with hlte, no fastboot + boot partition is tiny Dec 26 09:31:43 all samsung phones have no fastboot afaik Dec 26 09:31:49 yeah Dec 26 09:31:50 there is hemidal that works sometimes Dec 26 09:31:55 for gnu/linux and odin for win Dec 26 09:32:20 i tried porting ubports to my old note 3 ages back and didn't really get anywhere because ubuntu's ramdisk bloated the hell out of the boot image Dec 26 09:32:38 ubports or halium though Dec 26 09:32:59 maybe Sasha will be more luch with Halium :) Dec 26 09:33:00 We can resize the boot in odin i think Dec 26 09:33:01 This was before halium was really established so Dec 26 09:33:31 i didn't want to try repartitioning because that's pretty risky and basically means no-one will be able to use it lol Dec 26 10:06:19 It's not risky, trust me Dec 26 10:16:55 The Ubuntu initrd is about 50MB big iirc, hybris-boot is only busybox basically, so should be much smaller Dec 26 10:22:29 @JBBgameich, exactly what I was talking about Dec 26 12:18:57 * caspermeijn[m] sent a long message: caspermeijn[m]_2017-12-26_12:18:57.txt Dec 26 12:26:17 if you message is long, pastebin it Dec 26 12:26:43 i am not comfortable at all with opening a link for your long message Dec 26 12:26:50 it's a sepolicy error though Dec 26 12:27:08 so remove selinux flags from the kernel, or remove it from the tree altogether Dec 26 12:28:05 summart: set SELINUX-related flags to N in defconfig or look for makefiles that define SELINUX and comment them out Dec 26 12:29:25 Sorry for the long message, I will pastebin next time... Dec 26 12:29:31 Thanks, I will try that Dec 26 12:29:43 we all make mistakes, it's okay ;) Dec 26 13:09:31 @van Dec 26 13:10:22 @vanyasam My build now continues, I had to remove a reference to sepolicy.mk Dec 26 13:10:59 I hope this doesn't break stuff during runtime, I will see... Dec 26 19:11:31 I'm trying to install halium (after months again) but the halium-install script fails: … `error: ext4_allocate_best_fit_partial: failed to allocate 48 blocks, out of space?` … But I've got my maximum space available (formatted /system and /data before) … Full log is here: https://paste.kde.org/phzgm7rdh Dec 26 20:48:59 Con te, partiróóó, paesiii che non ho mai veduto e vissuto con teee adesso si li vivróóó Dec 26 20:49:05 Grande bocc!! Dec 26 21:03:58 Please speak English Dec 26 21:18:45 @Ellenjott, It's not about free space on device, it's about your system.img size Dec 26 21:19:05 You need to increase system image size in the script Dec 26 21:19:48 There's something like `make_ext4fs -l 150M`, change the size there Dec 26 21:25:25 @ilyaishere Ahh thanks :) Dec 26 21:26:33 Fwd from bhushanshah: adb shell resize2fs -f /data/rootfs.img 1024000 Dec 26 21:26:49 you can post resize it Dec 26 21:26:55 after the creation of the image Dec 26 21:26:57 if you want to Dec 26 21:37:26 It's not about rootfs Dec 26 21:43:47 oops, sorry then. haven't really read the messages Dec 26 21:48:40 https://github.com/Halium/halium-scripts/pull/4/files Dec 26 21:48:44 This should fix that in the future Dec 26 21:49:51 Realistically, though, I think JBB's script is the much better flash tool that we should support in the future. Dec 26 21:50:13 Even if it is a little slower Dec 26 22:39:37 @UniversalSuperBox, agreed Dec 26 22:43:19 yo Dec 26 22:52:41 IGNNE was added by: IGNNE Dec 26 22:55:06 @IGNNE, welcome Dec 26 22:55:23 welcome Dec 26 22:55:40 quick question: how's development going? Dec 26 22:55:57 Hi, thx Dec 26 22:56:56 Actually, I am trying to port to Sony's Z5C... Dec 26 22:58:05 i mean the os itself Dec 26 22:58:25 what exactly is the current difference to ubuntu touch? Dec 26 22:58:45 Me? Dec 26 22:58:51 Well, Halium isn't really an OS in itself... Dec 26 23:00:36 hm... why am I +b on irc? Telegram kinda sucks if you just bricked your hpone :P Dec 26 23:01:49 +b? Dec 26 23:03:18 (Photo, 327x23) https://irc.ubports.com/74ybOYfD/file_3225.jpg Dec 26 23:03:34 oh Dec 26 23:03:37 is there a reason? Dec 26 23:04:18 afaik not, maybe just reboot my router Dec 26 23:14:02 mount Dec 26 23:21:23 Pro-tip: If your init isn't executable, it won't run Dec 26 23:21:38 Pro-tip: If you try to redirect your init's output to a read-only filesystem, it won't work Dec 26 23:21:54 I am my own worst enemy. ;_; Dec 26 23:26:16 👍 Dec 26 23:26:57 Btw, why does my dev not have a "block/bootdevice"? Dec 26 23:47:05 wait... weird question... who tf actually mounts /dev? My kernel shouldn't (anymore), but when the init script tries, it is already mounted... Dec 26 23:47:20 ls Dec 27 00:13:41 Are you in the rootfs or not Dec 27 00:13:52 @Lyokanthrope, It would be possible to make a barebones lk that automatically goes to fastboot mode and have that in the boot partition (make it the kernel of an Android bootimg, no ramdisk), and use that for temporary fastboot purposes. Dec 27 00:19:01 @UniversalSuperBox, Sadly, no, that's why I am cursing my dev :/ Dec 27 00:20:17 hybris-recovery with telnet -> no /dev/block/bootdevice Dec 27 00:23:33 No, that's created by udev Dec 27 00:23:38 That's why fixup-mountpoints exists Dec 27 00:25:12 But the init script wants to mount it, iirc? Dec 27 00:25:24 No, it just wants to mount the userdata partition Dec 27 00:25:33 let's see, what was its name in the script Dec 27 00:25:49 `echo $DATA_PARTITION` Dec 27 00:26:20 through telnet, please Dec 27 00:27:17 one sec, I just flashed lineage boot back Dec 27 00:31:12 ~ # cat init.log | grep DATA_PARTITION … + DATA_PARTITION=/dev/block/bootdevice/by-name/userdata Dec 27 00:31:59 Have you included your device in fixup-mountpoints? Dec 27 00:32:56 Pretty shure I did, suzuran should the same layout as sumire (normal Z5), but let me check once more Dec 27 00:33:36 If that's the case, do what Z00L did. … `"Z00L"|"Z00LD"` Dec 27 00:34:03 So you'd have `"sumire"|"suzuran" Dec 27 00:34:28 pretty shure I did this Dec 27 00:35:54 ... ;; … "sumire" | "suzuran") … sed -i \ … -e 's block/platform/soc.0/by-name/DDR mmcblk0p7 ' \ Dec 27 00:36:37 Try chopping off the spaces Dec 27 00:44:55 Nope, same result. Actually, is there somewhere to look for the results of fixup-mountpoints? Dec 27 00:45:01 ls Dec 27 00:45:14 In your build output, it'll say `/data appears to live on ...` Dec 27 00:45:20 Wait Dec 27 00:45:25 Do you have ubports-boot in the tree? Dec 27 00:45:57 erm... should I? :-P Dec 27 00:46:04 No, if you don't that's good Dec 27 00:46:08 Because it would cause exactly this Dec 27 00:46:43 At least not on purpose. Only Halium itself plus some Repos from Lineage Dec 27 00:46:51 Grand Dec 27 00:47:04 So what does it say in that `/data appears to live on` stuff? Dec 27 00:47:32 Hey Dec 27 00:47:36 May I interrupt? Dec 27 00:47:44 Well, you already have. :P Dec 27 00:47:58 Obvious question, obvious answer Dec 27 00:48:10 I saw you forked Ubuntu Touch Dec 27 00:48:23 Yes, ubuntu-touch.io Dec 27 00:49:01 What exactly is currently the huge difference Dec 27 00:49:15 between what? Dec 27 00:49:16 And I've got a device officially supported by Ubuntu Touch, would it be hard to port? Dec 27 00:49:32 Halium and Ubuntu Touch Dec 27 00:49:45 halium/hybris-boot/Android.mk:67: ***************** /boot appears to live on /dev/block/bootdevice/by-name/boot … halium/hybris-boot/Android.mk:68: ***************** /data appears to live on /dev/block/bootdevice/by-name/userdata Dec 27 00:50:11 Oh, if it's officially supported you can probably install it using the UBports installer. Just grab the AppImage from https://github.com/ubports/ubports-installer/releases/tag/0.1.9-beta and run it as `sudo whatever.AppImage` Dec 27 00:50:45 Maybe you'd be better served by joining #ubports Dec 27 00:50:50 I actually just tried out Ubuntu Touch Dec 27 00:50:54 Didn't like it Dec 27 00:51:24 erm... Halium is not an OS Dec 27 00:51:27 But since I don't get a /dev/block/bootdevice at all, I don't get a .[...]/by-name/userdata Dec 27 00:51:32 @IGNNE, Right. Dec 27 00:51:37 But my question is, 1. I didn't exactly understand the overall mission, is this just like project treble? Dec 27 00:52:02 If fixup-mountpoints was firing correctly, that would instead be `/dev/block/mmcblkxxxx` Dec 27 00:52:24 And 2. If it's the boilerplate, what *is* different from ubuntu touch than just removing the UI? Dec 27 00:53:02 Ah, ok. let me check again, i bet there is some stupid typo Dec 27 00:53:31 @guenstig_werben, um Dec 27 00:54:06 I get im asking the wrong questions ^^ Dec 27 00:54:18 Halium provides hardware compatibility and nothing more. It is not a distribution itself, but a set of components that other distributions can work in Dec 27 00:55:36 Then, how can one actually build something with it. Dec 27 00:55:37 Plus, what about project treble? Dec 27 00:56:01 That's due to Android using a different C Standard Library (bionic) than almost every Linux distribution (glibc). They're essentially different languages. We need libhybris, which translates between bionic and glibc Dec 27 00:56:06 Without it, we can't use Android drivers Dec 27 00:56:13 And without Android drivers, we can't be on Android phones Dec 27 00:56:45 Yeah I encountered that clib problem..... Dec 27 00:57:15 I mean project treble is coming only for the newer phones probably, but still, Google has a little bit more power 🙂 Dec 27 00:57:24 hold on, I found it Dec 27 00:57:52 Treble might make it so that Halium doesn't need to be rebuilt for every device Dec 27 00:57:56 But we'll still need Treble Dec 27 00:57:58 Bah Dec 27 00:58:02 We'll still need Halium Dec 27 00:58:14 For some reason I don't remeber, the /dev/block/whatever/by-name changed in some release, or between lineage and stock, idk Dec 27 00:58:41 So this is to stick a random Linux on top of it, so it runs on the android hardware? Dec 27 00:58:44 Lett#s just rebuild this Dec 27 00:58:58 @IGNNE, ohhhh, that is annoying Dec 27 00:59:31 sudo fastboot reboot Dec 27 00:59:37 lol Dec 27 00:59:58 Please please don't do unnecessary sudo Dec 27 01:00:00 I really have to check where my keyboard points to :( Dec 27 01:00:43 @guenstig_werben, The distro needs to make a few concessions and include some parts, but yeah. Dec 27 01:00:48 On my quick+dirty build machine I am not a member of the right groups, and I'm to lazy to fix it. Dec 27 01:00:59 For example, the compositor needs support for Libhybris since it acts as a display driver Dec 27 01:01:03 Among many other things Dec 27 01:02:31 And if, pardon the stupid question, I just wanted to run a simple, say arch, without actually caring about screen, touch, etc. (well, only wifi would be good to have). Dec 27 01:02:39 How would I go about building the whole thing? Dec 27 01:02:43 Woohoo, it's mounted. Still stuck in a bootloop, but hey, a different bootloop! Thx a lot Dec 27 01:02:52 I guess putting in the binary blob for wifi shouldn't be that hard Dec 27 01:03:40 The wifi hardware is still controlled by the Android container Dec 27 01:03:51 I think there was a project to get an Arch rootfs out Dec 27 01:03:52 https://github.com/Halium/rootfs-builder/tree/archlinux/customization Dec 27 01:04:39 Did that ever work out? Dec 27 01:04:48 ¯\_(ツ)_/¯ Dec 27 01:05:10 Oh we Dec 27 01:05:13 *we Dec 27 01:05:15 *well Dec 27 01:11:40 Well, also no last_kmesg - it does not want me to get too happy apparently Dec 27 01:11:52 There's another place that could be Dec 27 01:14:17 Nothing. And yes, I have pstore compiled in Dec 27 01:14:26 :-/ Dec 27 01:14:44 Can you get back into hybris-recovery? Dec 27 01:14:59 yes Dec 27 01:15:53 Okay. Dec 27 01:15:55 `ls /target` Dec 27 01:16:08 data lost+found Dec 27 01:16:16 Your rootfs isn't there Dec 27 01:16:32 There's nothing in it Dec 27 01:16:56 Best bet is to reinstall it using `rootstock`. Dec 27 01:18:00 Is your `userdata` partition F2FS? Dec 27 01:18:09 ext4 Dec 27 01:18:14 guud Dec 27 01:20:19 Oh, I know what happened Dec 27 01:20:47 please don't say it inculdes rebuilding system Dec 27 01:20:49 A reinstall will fix it Dec 27 01:20:56 Just do it from recovery using the Halium scripts Dec 27 01:21:10 It happens from time to time, the image gets corrupted and throws a hissy fit. Dec 27 01:21:37 "Happens from time to time" sounds reassuring Dec 27 01:22:19 Dangers of mounting a large image file as your root Dec 27 01:23:38 Crap, I knew I should not have done make clean Dec 27 01:23:48 :( Dec 27 01:39:30 Yay, I've god a rootfs! Dec 27 01:47:20 Well, thank you for your help, at least now I have a running system to play with. Dec 27 01:47:27 Yay! :) Dec 27 02:10:17 @Ellenjott re: last comment, do you have udev rules in place? Dec 27 02:10:31 (on GitHub that is) Dec 27 02:10:48 I'm not sure .. Dec 27 02:10:51 so probably not Dec 27 02:11:09 Best fix it Dec 27 02:11:14 :-) Dec 27 02:11:24 I'm alive! Dec 27 02:11:25 kinda Dec 27 02:15:45 I've done this: http://docs.halium.org/en/latest/porting/debug-build/index.html?highlight=udev#add-udev-rules … ... but I see no change after reboot Dec 27 02:15:59 hey but my led is now green Dec 27 02:16:17 So ok it seems so there is a difference at least Dec 27 02:16:35 I have determined something Dec 27 02:16:43 upstart ordering ist not as good as systemd Dec 27 02:16:44 green led? O.o Dec 27 02:17:33 no idea what that means Dec 27 02:17:36 @Ellenjott, Can you manually start kwin_wayland --xwayland --hwcomposer as phablet user? Dec 27 02:18:10 QStandardPaths: XDG_RUNTIME_DIR not set, defaulting to '/tmp/runtime-phablet' … error: XDG_RUNTIME_DIR not set in the environment … FATAL ERROR: could not create Wayland server Dec 27 02:18:26 Hm Dec 27 02:18:56 Are you ssh'd in as root? Dec 27 02:19:21 connected as root and then `su phablet` Dec 27 02:19:30 It won't work Dec 27 02:19:38 What should I do then? Dec 27 02:19:39 ssh as phablet Dec 27 02:19:43 Ok Dec 27 02:19:58 passwd? Dec 27 02:20:20 Set it Dec 27 02:20:29 :) Dec 27 02:20:35 The one you set it? (Pro-tip: set numeric password) Dec 27 02:20:57 (This pro tip is brought to you by the "we couldn't afford anything but a numpad" society) Dec 27 02:20:58 :P Dec 27 02:21:13 *shrug* Dec 27 02:21:39 So now I get: … ```unable to lock lockfile /run/user/32011/wayland-0.lock, maybe another compositor is running … FATAL ERROR: could not create Wayland server``` Dec 27 02:21:43 I would love to know if PAM supported pattern unlock e.g Dec 27 02:21:59 Is kwin_wayland process running? Dec 27 02:22:18 @bhushanshah, It doesn't. PIN unlock in Ubuntu Touch is a hack. Dec 27 02:22:23 phablet 1092 0.2 1.2 132812 23320 tty2 Sl+ 02:14 0:01 kwin_wayland —xwayland —libinput —lockscreen —inputmethod maliit-server plasma-phone Dec 27 02:22:30 Yes it is Dec 27 02:22:39 Hm Dec 27 02:22:54 Basically, "set a password but the keyboard presses enter after you type four characters" Dec 27 02:23:08 journalctl | grep simplelogin Dec 27 02:23:28 @UniversalSuperBox, Not pin, but pattern Dec 27 02:23:58 (Photo, 1091x217) https://irc.ubports.com/N8uMFzL5/file_3233.jpg Dec 27 02:24:25 That's it? Strange Dec 27 02:24:49 What's permission for e.g /dev/kgsl-3d0 Dec 27 02:25:35 crw-rw-rw- 1 system system 240, 0 Dec 27 02:15 /dev/kgsl-3d0 Dec 27 02:26:49 kwin is stuck somehow and I don't know why/where Dec 27 02:27:25 Oh Dec 27 02:27:26 :/ Dec 27 02:27:56 Can you kill this kwin process and then run it as phablet? Dec 27 02:28:10 Ok Dec 27 02:28:23 (basically I want slightly more verbose output of what's going on) Dec 27 02:29:08 unable to lock lockfile /run/user/32011/wayland-0.lock, maybe another compositor is running … FATAL ERROR: could not create Wayland server Dec 27 02:29:26 but we get the same output ... :( Dec 27 02:29:33 Um sure kwin_wayland process is killed? Dec 27 02:30:20 oh kill hasn't done anything :( Dec 27 02:30:33 `kill -9` Dec 27 02:30:49 ah Dec 27 02:30:58 `alias decimate="kill -9` Dec 27 02:31:03 (kwin doesn't like to get kill signal from normal user, use root) Dec 27 02:31:11 kglobalaccel-runtime: Failed to register service org.kde.kglobalaccel Dec 27 02:31:23 Ignore Dec 27 02:31:27 @Ellenjott, only output Dec 27 02:31:32 :/ Dec 27 02:31:47 Alright let's enable debug output Dec 27 02:32:01 How do I do this? Dec 27 02:32:12 I'm not experienced with starting kwin manually Dec 27 02:34:33 https://gist.github.com/bhush9/7c92672602245770821af8e5dda9cba3 Dec 27 02:35:05 Where do I have to put it into? Dec 27 02:35:07 Can you paste it in ~/.config/QtProject/qtlogging.ini? Dec 27 02:35:10 Ok Dec 27 02:35:14 Would it be undesirable for the init-stderrout file to go to `/target/data/` rather than `/target`? Dec 27 02:35:17 (for phablet user) Dec 27 02:35:38 (Please say no, read-only rootfs's don't work if you say yes) Dec 27 02:35:52 With that in mind, I don't know why I asked the question Dec 27 02:36:26 @UniversalSuperBox, I don't see any issue with it Dec 27 02:37:15 kwin_core: Platform does not support input, enforcing libinput support Dec 27 02:37:22 kwin_core: Init of kglobalaccel failed Dec 27 02:37:40 That's it? Dec 27 02:37:41 (Photo, 997x407) https://irc.ubports.com/xT3gJUPY/file_3235.jpg Dec 27 02:37:53 Yes :/ Dec 27 02:38:34 Does test_hwconposer work as phablet user? Dec 27 02:39:00 -bash: test_hwconposer: command not found Dec 27 02:39:06 pm rootfs :( Dec 27 02:39:12 `test_hwcomposer` Dec 27 02:39:21 lol Dec 27 02:39:55 It's running ... and it's happening nothing ... Dec 27 02:41:02 export EGL_PLATFORM=hwcomposer Dec 27 02:41:32 no change, no output ... Dec 27 02:42:00 I'm still getting on udev rules Dec 27 02:42:05 dvogel said the tests never worked for him ... but probably only in the reference rootfs Dec 27 02:42:30 Where did you put udev rules? Dec 27 02:43:17 https://paste.kde.org/psqrkksv7 Dec 27 02:43:35 @Ellenjott, -> they seem correct to me Dec 27 02:43:49 Right Dec 27 02:44:50 Can you run logcat from another terminal and then test_hwcomposer and get output in logcat? Dec 27 02:45:00 Ok Dec 27 02:45:29 -su: logcat: command not found ? Dec 27 02:46:10 root@localhost:~# /system/bin/logcat … logcat read failureroot@localhost:~# Dec 27 02:46:22 :( Dec 27 02:46:41 Now I'm going to ask important question Dec 27 02:46:51 Is android container running? Dec 27 02:47:01 systemctl status lxc@android Dec 27 02:47:08 Yes it is Dec 27 02:47:27 O.o Dec 27 02:47:39 (Photo, 907x231) https://irc.ubports.com/SUvgEBdf/file_3237.jpg Dec 27 02:47:46 Or wait .. theres an error Dec 27 02:47:58 Dec 27 02:15:12 localhost.localdomain lxc-start[1314]: lxc-start: start.c: preserve_ns: 153 No such file or directory - Kernel does not support attaching to mnt namespaces. Dec 27 02:48:20 oh no not the kernel config again ... Dec 27 02:48:25 Ouput of `systemctl status` please Dec 27 02:48:40 That's not fatal error Dec 27 02:49:08 https://paste.kde.org/plkcs8r9b Dec 27 02:49:37 State: starting Dec 27 02:49:39 still ... Dec 27 02:49:48 feels like "still compiling" Dec 27 02:49:55 Nope Android container is not started fully Dec 27 02:50:03 dmesg please Dec 27 02:50:49 https://paste.kde.org/pl4wjpj1p Dec 27 02:51:02 binder: 15191:15191 transaction failed 29189, size 0-0 Dec 27 02:51:19 msm_fb_pan_display_ex: no pan display for fb2! Dec 27 02:52:31 [   44.419830] charger: --------------- STARTING CHARGER MODE --------------- Dec 27 02:52:44 waitaminute Dec 27 02:52:46 Sure... Android sure Dec 27 02:53:07 That might mean that Android is trying to boot into its slimmed-down mode Dec 27 02:53:45 Seems so it tries to load the charging animation … charger: Could not open animation description at /res/values/charger/animation.txt Dec 27 02:54:42 damn it, Android Dec 27 02:55:20 Thus even test_lights don't work anymore now ... Dec 27 02:55:25 I suggest to kill charger service from init.devicename.rc Dec 27 02:55:53 How do I enter the lxc again? Dec 27 02:55:58 The boot type is determined earlier than that, is it not? Dec 27 02:56:30 @Ellenjott, Enter lxc? Dec 27 02:56:50 Well I mean edit it out from device tree locally Dec 27 02:56:59 Ah ok :) Dec 27 02:57:28 @UniversalSuperBox yeah , maybe bootmode=charger in cmdline Dec 27 02:58:05 @Ellenjott do you have that in kernel cmdline? (cat /proc/cmdline) Dec 27 02:58:49 I've got this androidboot.mode=charger Dec 27 02:59:01 # cat /proc/cmdline … androidboot.hardware=qcom msm_rtb.filter=0x3F ehci-hcd.park=3 androidboot.emmc=true androidboot.bootloader=s1 oemandroidboot.s1boot=S1_Boot_Lagan_1.1_10 androidboot.serialno=CB5A1WNNDL ta_info=1,16,256 startup=0x00000004 warmboot=0x00000000 oemandroidboot.imei=3597750501946400 oemandroidboot.babe1325=313236372d31363931 oemandroidboot.babe1326=33 androidboot.baseband=mdm androidboot.mode=charger Dec 27 02:59:04 There you go.. **** ENDING LOGGING AT Wed Dec 27 03:00:00 2017 **** BEGIN LOGGING AT Wed Dec 27 03:00:01 2017 Dec 27 03:00:36 @Ellenjott I *think* a clean reboot from bootloader might fix this Dec 27 03:00:49 Ok Dec 27 03:03:45 IT'S WORKING!!! Dec 27 03:04:01 I CAN SEE PLASMA MOBILE Dec 27 03:05:38 (Photo, 960x1280) https://irc.ubports.com/0Jrcx40a/file_3239.jpg Dec 27 03:05:52 But the touch screen isn't working Dec 27 03:07:15 Only the power button is working Dec 27 03:08:32 You need to fix up udev rules Dec 27 03:08:58 What's missing there? Dec 27 03:09:13 Delete current rules and rerun the command? Dec 27 03:09:18 No no Dec 27 03:10:01 Your device marks touchscreen as button(!!!) somehow Dec 27 03:10:19 No the touchscreen isn't working at all Dec 27 03:10:38 but I can use the hardware button on the side Dec 27 03:10:42 Can you figure out which of the /dev/input/event* is touchscreen by using evtest Dec 27 03:11:18 ... have to install evtest first Dec 27 03:14:19 dvogel had same issue.. I am trying to find exact line he put in udev rules Dec 27 03:14:54 (Photo, 609x201) https://irc.ubports.com/BdHTxEAZ/file_3241.jpg Dec 27 03:15:22 I think it's event1 Dec 27 03:15:32 Works Dec 27 03:15:42 getting output on touching the screen Dec 27 03:15:48 Right Dec 27 03:15:53 So wait a moment Dec 27 03:17:50 ACTION=="add|change", KERNEL=="event1" ENV{ID_INPUT_TOUCHSCREEN}="1" Dec 27 03:18:03 You need something like this in udev rule Dec 27 03:18:10 (totally untested Dec 27 03:20:17 Do I need to change something in it? Dec 27 03:22:31 No, just add this line and reboot and try Dec 27 03:24:03 WORKS! :) Dec 27 03:25:14 But which PIN/Password do I have to enter? - phablet? (then I have to change it again to make it unsafer) Dec 27 03:25:54 Yeah phablet Dec 27 03:25:56 Yeah!!! :) Dec 27 03:26:03 Working! Dec 27 03:26:24 (Photo, 960x1280) https://irc.ubports.com/RVO8SghS/file_3243.jpg Dec 27 03:27:17 emh .. but now I can't do anything? Dec 27 03:27:27 again only the hardware button is working :( Dec 27 03:27:35 (also it'll be better to move Plasma mobile topic to @plasmamobile group..) Dec 27 03:27:45 @Ellenjott, Wut Dec 27 03:28:10 udevadm info /dev/input/event1 Dec 27 03:28:36 # udevadm info /dev/input/event1 … P: /devices/virtual/input/input1/event1 … N: input/event1 … E: DEVNAME=/dev/input/event1 … E: DEVPATH=/devices/virtual/input/input1/event1 … E: ID_INPUT=1 … E: ID_INPUT_KEY=1 … E: ID_INPUT_TOUCHSCREEN=1 … E: MAJOR=13 … E: MINOR=65 … E: SUBSYSTEM=input … E: USEC_INITIALIZED=8382297 Dec 27 03:29:16 Okay, let's edit the udev rule bit Dec 27 03:29:58 Bhushan Shah: … ACTION=="add|change", KERNEL=="event1" ENV{ID_INPUT_TOUCHSCREEN}="1", ENV{ID_INPUT_KEY}="1" Dec 27 03:30:23 Err actually Dec 27 03:30:35 Edited and fixed Dec 27 03:30:51 (basically no 1 in key env) Dec 27 03:34:00 again ... I can't do anything after logging in ... Dec 27 03:34:15 only locking / unlocking screen Dec 27 03:34:27 Hmm Dec 27 03:34:55 Wonder if plasmashell process is stuck Dec 27 03:42:29 @bhushanshah, Yep .. it's crashing ... Dec 27 08:49:27 Cl1608Ho was added by: Cl1608Ho Dec 27 09:31:15 @IGNNE in what way where the mount points changed/ Dec 27 09:59:38 hey. i'm currently checking out the source and while doing repo sync i'm getting a lot of 404 errors, is that a problem? Dec 27 10:01:40 @Cl1608Ho, canonical servers are down sometimes Dec 27 10:02:08 lol Dec 27 10:02:45 yeah it seem so Dec 27 10:02:49 but is it a problem? Dec 27 10:04:02 @IGNNE also you might want to have a look at https://github.com/ubports/android_kernel_sony_msm8994 .. Dec 27 10:10:07 I had to make changes to the device kitakami and sumire repos from lineageos and pushed them to ubports.. with that halium tests were running fne Dec 27 10:10:32 @Cl1608Ho, yes Dec 27 10:11:41 @IGNNE I am now still working on the apparmor merge - there seems to be something missing or wrong. At least ubuntu touch does not boot properly Dec 27 10:12:41 within halium basic services like wifi .. usb charging \o/ .. egl glesv2 were working Dec 27 10:13:05 @anpok, 5.1 tree? Dec 27 10:13:16 7.1 Dec 27 10:13:24 egl glesv2 should not really work in 7.1 tree Dec 27 10:13:40 the hwccomposer test Dec 27 10:13:50 oh, that's a totally different thing :P Dec 27 10:13:53 ok then Dec 27 10:14:09 but it does draw with glesv2? Dec 27 10:18:26 yeah Dec 27 10:18:49 but the tests are different, and the test_glesv2 or whatever it's named doesn't work for some reason Dec 27 10:18:56 that's why i got surprised Dec 27 10:21:19 ok Dec 27 10:32:10 @Cl1608Ho, The "404 Server does not provide clone.bundle" errors are normal and are not a problem Dec 27 10:32:28 thanks :) Dec 27 10:32:55 apart from that, are you serious with ~25gb source? Dec 27 10:58:49 @anpok, Yup, same here, apart from graphics. I'm still struck at the Sony logo. Maybe I'll have to take a look at your kernel, currently I am using plain Lineage Dec 27 11:03:08 I had the same problem once with an old version of twrp, so maybe a different kernel version, idk. My current config has the same mountpoints as lineage Dec 27 11:09:04 https://github.com/ubports/android_kernel_sony_msm8994/commit/63630285694687b2566a9b52a78198bcf0ae6d88 disabling CONFIG_MMC_FFU might help Dec 27 11:09:57 MMC FFU enables a back ground task that blocks the mmc init routine Dec 27 11:10:50 that task times out after 60s and until then the kernel sees no other partitions.. Dec 27 11:23:59 @anpok Just for fun, which base kernel did you use? 3.10.84 caf? Dec 27 11:31:29 it is based on this one https://github.com/LineageOS/android_kernel_sony_msm8994/commits/cm-14.1 Dec 27 11:33:35 That *is* strange - why do I get different mountpoints, if we have the same partition layout, the same kernel and probably also very similar userspace? Somehow this feels wrong Dec 27 11:37:03 ok I will look again Dec 27 11:37:15 I dont rely on that atm because I am using ubports boot Dec 27 12:51:31 fatal: This operation must be run in a work tree … Traceback (most recent call last): … File "/media/clemens/Elements/halium/.repo/repo/main.py", line 531, in … _Main(sys.argv[1:]) … File "/media/clemens/Elements/halium/.repo/repo/main.py", line 507, in _Main … result = repo._Run(argv) or 0 … File "/media/clemens/Elements/halium/.repo/repo/main.py", line 180, in _Run … result = cmd.Execute(copts, cargs) … Dec 27 12:51:32 "/media/clemens/Elements/halium/.repo/repo/subcmds/sync.py", line 821, in Execute … project.Sync_LocalHalf(syncbuf, force_sync=opt.force_sync) … File "/media/clemens/Elements/halium/.repo/repo/project.py", line 1327, in Sync_LocalHalf … self._InitWorkTree(force_sync=force_sync) … File "/media/clemens/Elements/halium/.repo/repo/project.py", line 2486, in _InitWorkTree … raise GitError("cannot initialize work tree") … error.GitError: Dec 27 12:51:50 i'm getting this error when doing repo sync -c -j10, at the end after downloading everything, in the middle of checking out Dec 27 12:52:09 how to resolve it ? :P Dec 27 12:52:41 try in an another folder Dec 27 12:52:48 and see if the issue persists Dec 27 12:54:35 downloading everything again you mean? Dec 27 12:54:56 yes Dec 27 12:54:59 ... Dec 27 12:55:09 getting this messages before, do they have anyhting to do with that? Dec 27 12:55:10 Shared project android_hardware_qcom_media found, disabling pruning. … Shared project android_hardware_qcom_media found, disabling pruning. … Shared project android_hardware_qcom_media found, disabling pruning. … Shared project android_hardware_qcom_media found, disabling pruning. … Shared project android_hardware_qcom_media found, disabling pruning. … Shared project android_hardware_qcom_media found, disabling pruning. … Shared project Dec 27 12:55:11 android_hardware_qcom_media found, disabling pruning. … Shared project android_hardware_qcom_media found, disabling pruning. … Shared project android_hardware_qcom_media found, disabling pruning. … Shared project android_hardware_qcom_media found, disabling pruning. … Shared project android_hardware_qcom_bt found, disabling pruning. … Shared project android_hardware_qcom_bt found, disabling pruning. … Shared project android_hardware_qcom_wlan found Dec 27 12:55:12 Shared project android_hardware_qcom_wlan found, disabling pruning. … Shared project android_hardware_qcom_audio found, disabling pruning. … Shared project android_hardware_qcom_audio found, disabling pruning. … Shared project android_hardware_qcom_audio found, disabling pruning. … Shared project android_hardware_qcom_audio found, disabling pruning. … Shared project android_hardware_qcom_audio found, disabling pruning. … Shared project android_hard Dec 27 12:55:13 disabling pruning. … Shared project android_hardware_qcom_audio found, disabling pruning. … Shared project android_hardware_qcom_audio found, disabling pruning. … Shared project android_hardware_qcom_audio found, disabling pruning. … Shared project android_hardware_qcom_audio found, disabling pruning. … Shared project android_hardware_ril found, disabling pruning. … Shared project android_hardware_ril found, disabling pruning. … Shared project an Dec 27 12:55:14 found, disabling pruning. … Shared project android_hardware_qcom_display found, disabling pruning. … Shared project android_hardware_qcom_display found, disabling pruning. … Shared project android_hardware_qcom_display found, disabling pruning. … Shared project android_hardware_qcom_display found, disabling pruning. … Shared project android_hardware_qcom_display found, disabling pruning. … Shared project android_hardware_qcom_display found, disabli Dec 27 12:55:15 android_hardware_qcom_display found, disabling pruning. … Shared project android_hardware_qcom_display found, disabling pruning. … Shared project android_hardware_qcom_display found, disabling pruning. Dec 27 13:24:40 Btw... Please please, if you want to show more then 4 lines use pastebin Dec 27 13:25:47 SPAAM Dec 27 13:25:49 @Cl1608Ho, no Dec 27 13:25:51 that is ok Dec 27 13:26:01 also yes, use pastebin Dec 27 13:26:11 there are hundreds of different ones to choose from Dec 27 13:30:02 Yeh sorry Dec 27 13:56:12 Optimus9650 was added by: Optimus9650 Dec 27 14:00:05 i identified the faulty repo and deleted it. now i think its working Dec 27 14:00:56 I am on arch... How do i add i386 architecture? Dec 27 14:01:19 For most devices, it's probably not needed Dec 27 14:01:32 use the multilib repo if necesssary Dec 27 14:01:58 i guess it's not needed anymore because multilib is merged with the normal packages a couple of days ago Dec 27 14:02:11 just try a build and install additional packages if they are necessary Dec 27 14:02:17 good :) Dec 27 14:05:35 Libc6-dev is not present in any arch repo or yaourt Dec 27 14:07:55 Nvm... Would it be good option if i use ubuntu in live mode? Dec 27 14:08:05 use arch Dec 27 14:08:21 @vanyasem, I am on arch... Dec 27 14:08:22 I use arch, @bhushanshah also does. everything is fine Dec 27 14:09:24 But .. it's dependencies are not there Dec 27 14:09:29 😕 Dec 27 14:09:32 they are :) Dec 27 14:10:08 just install the meta package `lineage-devel` from AUR Dec 27 14:10:14 I don't remember it's exact name Dec 27 14:10:35 Got it Dec 27 14:11:02 also `qemu-arch-static` for jbb's script to work (also from AUR) Dec 27 14:11:03 @vanyasem, it includes all the deps Dec 27 14:11:24 @Optimus9650, ping me if you encounter any more issues, you are welcome Dec 27 14:11:48 lineage devel includes more than needed though, we are yet to write docs on that one Dec 27 14:12:20 My device has 4.4 to 8.0 Dec 27 14:12:27 Which is the most stable? Dec 27 14:13:27 for Halium? Dec 27 14:13:36 Yup... Dec 27 14:13:37 5.1 I would say Dec 27 14:13:52 @vanyasem, this Dec 27 14:14:00 I'd say 7.1 makes most sense for porting Dec 27 14:14:24 5.1 doesn't have volte support... Dec 27 14:16:02 does ofono support volte though? Dec 27 14:17:05 it does support LTE at least I believe Dec 27 14:17:42 agreed, but at the same time 5.1 is more stable and has more things working Dec 27 14:17:44 oh I'm going underground in subway. cya Dec 27 14:26:21 "repository name" Dec 27 14:26:28 What do i have to will there Dec 27 14:26:29 Hm, I cannot make the lxc container working: … lxc_setup_devpts: 1409 failed to bind mount "/dev/pts/ptmx" to "/dev/ptmx" … -> /dev/ptmx already exists, but there is no /dev/pts/ptmx Dec 27 14:26:30 ? Dec 27 14:27:20 @Optimus9650, ????? Dec 27 14:27:32 what? Dec 27 14:27:56 @Optimus9650, ?? Dec 27 14:28:25 @IGNNE, Enable CONFIG_DEVPTS_MULTIPLE_INSTANCES Dec 27 14:29:03 Of course, had that before. Thx Dec 27 14:29:06 @Optimus9650, are you filling your repo manifest? Dec 27 14:29:22 Yup Dec 27 14:29:25 have you read the docs at Halium.org carefully? Dec 27 14:29:30 please read them again :P Dec 27 14:31:01 @vanyasem, I am on that page only... 😅 Dec 27 14:31:17 What stuff i have to fill there Dec 27 14:32:42 repo names for your device trees Dec 27 14:33:51 vendor (from themuppets) , device and kernel from lineage (or whatever rom you're basing onto) Dec 27 14:34:39 it's well explained there Dec 27 14:34:47 what part don't you understand exactly? Dec 27 14:40:24 @vanyasem, Repository name would be android_device_manufactu_device Dec 27 14:40:25 ?? Dec 27 14:42:52 yes Dec 27 14:42:53 Btw.. huge noob alert from my side... I am first time getting into these kind of stuff Dec 27 14:42:56 problably Dec 27 14:48:22 @Optimus9650, you need 3 trees Dec 27 14:48:24 device vendor and kernel Dec 27 14:48:49 what is your device codename? open themuppets repo and look for your device. then open lineage website and once again look for your device Dec 27 14:49:05 then copy the repo names and paste them into the manifest at right places Dec 27 14:52:19 https://gist.github.com/vanyasem/64fa7fdc0df556561c50a6048a982ce1 Dec 27 14:52:22 take a look at my repo manifest for HTC 10 (codename PME) Dec 27 14:52:31 (Photo, 1280x720) https://irc.ubports.com/lELPKdZQ/file_3253.jpg Dec 27 15:04:00 I am confused as f****..... Pulling my hair 🤯 Dec 27 15:08:41 <-- add this to .repo/local_manifest/roomservice.xml Dec 27 15:09:15 Woohooo, hwcomposer runs on Suzuran! Dec 27 15:09:22 nice Dec 27 15:09:51 @revertive> no Dec 27 15:10:19 It's an IRC bridge Dec 27 15:10:39 posting the messages from IRC here and from Telegram to IRC Dec 27 15:11:09 no i'm not a bot ;) Dec 27 15:15:03 Lol, funny Dec 27 15:19:46 Mr_Doge: melp? Dec 27 15:20:05 heyy... what is those .. remote=[thingy] Dec 27 15:20:08 ? Dec 27 15:20:13 *are Dec 27 15:20:20 Wrong answer Dec 27 15:20:38 Again Dec 27 15:20:48 Halium melp? Dec 27 15:21:00 umm yes Dec 27 15:21:03 ? Dec 27 15:21:48 On what? Dec 27 15:22:51 writing manifest.xml Dec 27 15:23:18 here is the link to dependency file Dec 27 15:23:23 https://github.com/LineageOS/android_device_wingtech_wt88047/blob/cm-14.1/lineage.dependencies Dec 27 15:23:51 Go writin it, then Dec 27 15:33:19 what do i have to fill in remote?? Dec 27 15:33:33 Mr_Doge: for lineageos, los Dec 27 15:33:43 yup Dec 27 15:34:13 Mr_Doge: see, example e.g : https://github.com/Halium/halium-devices/blob/halium-7.1/FP2/manifest.xml Dec 27 15:49:22 help me with the vendor blobs Dec 27 16:07:50 what device? Dec 27 16:16:54 @revertive, Nvm.. figured it... I didn't even read the whole page .. and started asking questions... Dec 27 16:17:11 🤦‍♂🤦‍♂ Dec 27 16:18:59 @Optimus9650, I told you to read it to the end several times.. Dec 27 16:19:57 😬 Dec 27 16:20:13 @vanyasem, It was easy... Once i read the whole page... Dec 27 18:19:14 Guys what was the command parameter for repo sync to just download affected branches Dec 27 18:19:17 not full git repos? Dec 27 18:25:50 -c Dec 27 18:25:57 I think Dec 27 18:49:35 @Flohack, --depth=1 Dec 27 18:54:17 @Flohack, `repo sync -c` Dec 27 18:54:49 which one now ^^ Dec 27 18:55:07 --depth=1 in repo init, later only repo sync -c Dec 27 19:06:11 I knew ittt Dec 27 19:58:34 Does somebody know how I can put my udev rules into the device tree? Dec 27 20:02:58 So probably I shouldn't add them to the device tree, instead probably here: https://github.com/Halium/lxc-android/tree/master/usr/lib/lxc-android Dec 27 20:16:20 Done syncing .. phhew 😤 Dec 27 20:16:47 Suggest me a distribution Dec 27 20:18:41 Halium rootfs? Have you already built hybris-boot and the systemimage already? Dec 27 20:19:36 First build the halium base, then debug with the reference rootfs and then you can think about a final rootfs to use Dec 27 20:20:24 Halium is stable? Dec 27 20:20:40 is not it useful? Dec 27 20:21:17 Halium consists of Android libraries, LXC and Libhybris Dec 27 20:21:50 what is your most successful device? via port? Dec 27 20:21:57 Is Linux stable? Quite Dec 27 20:22:10 Is Android stable? Yeah, to a degree Dec 27 20:22:12 Etc. Dec 27 20:22:26 > Does somebody know how I can put my udev rules into the device tree? Dec 27 20:22:27 Cp? Dec 27 20:22:42 @developerfect, Personally, I'm stuck for now, ask guys that have Qualcomm chipsets Dec 27 20:23:59 @i_love_halium_on, It seems so they are all collected in the lxc-android project Dec 27 20:25:32 Ah you wanted to copy it to rootfs directly not manually in the device with adb Dec 27 20:54:28 Woohoo got wifi working \o/ Dec 27 21:11:27 got input (touchscreen and buttons) / Wifi / Screen / Lights / Sensors / Vibrator working on Nexus 4 almost there :P Dec 27 21:12:07 anyone started working on Halium for/with Project Treble? Dec 27 21:20:11 Phh? Dec 27 21:20:17 French boi? Dec 27 21:20:25 uh, yes? Dec 27 21:20:46 Oh, tu me peux dire comme cluoire le twrp sur mon nexus? Dec 27 21:21:16 ffs Dec 27 21:21:52 Hmm non? Dec 27 21:43:33 How can I disable the recovery.img generation in mka systemimage? Dec 27 21:43:40 Somehow mkbootimg failed .. Dec 27 21:46:02 Build/core/something.mk Dec 27 21:50:52 error: /media/lnj/lnjXdata/halium/7.1/out/target/product/yuga/recovery.img too large (17139712 > [17301504 - 270336]) Dec 27 22:14:17 @Ellenjott, @bhushanshah I'd like to discuss how we can do this, but I'm going to be driving for >1hr right now Dec 27 22:42:59 (Document) https://irc.ubports.com/IEVRIy02/file_3264.png Dec 27 22:43:02 this satisfying moment Dec 27 22:45:34 Still going rough on cheeseburger, eh? Dec 27 22:49:12 Well Dec 27 22:49:15 That command fials Dec 27 22:49:17 Fails Dec 27 22:50:48 do i need to recursively fetch dependencies? Dec 27 22:54:42 oh i forget to include the binary blobs Dec 27 23:11:05 seriously.. a png .. containing breakfast bacon? Dec 27 23:11:18 why not a short mpeg2 video? Dec 27 23:11:24 :P Dec 27 23:29:37 ok so Dec 27 23:29:55 i'm running the verify kernel config script and i'm getting huge amount of warnings Dec 27 23:30:24 shall i just change the values accorindgly or modify them somewhere else or what? Dec 27 23:39:10 besides, trying to build it, it seems that https://github.com/Halium/android_build/tree/halium-7.1/core is missing at least java_library.mk Dec 27 23:39:34 Java libraries should not be built Dec 27 23:39:39 comment them out if a problem appears Dec 27 23:39:52 ok... where exactly? Dec 27 23:40:16 in the files that require them? Dec 27 23:40:23 probably in the vendor blobs *.mk Dec 27 23:40:34 just comment out the include thing Dec 27 23:40:43 and the file is pretty pointless then but it will work? Dec 27 23:41:00 great :) what aobut those kernel parameters? Dec 27 23:41:39 just do what mer-verify-kernel-configs wants Dec 27 23:41:54 yeah but how, just edit the source file? Dec 27 23:42:06 the one that is being checked Dec 27 23:42:31 modify the defconfig that mer-very-kernelconfig checked Dec 27 23:43:31 ok Dec 27 23:43:47 do i have to rerun the mkbootimg afterwards? Dec 27 23:44:14 you have to rebuild. Just run mka hybris-boot again Dec 27 23:44:35 didnt go that far yet :P Dec 27 23:44:47 lets see :) Dec 27 23:45:29 what would the value ! mean? Dec 27 23:45:42 Not set Dec 27 23:46:08 ah Dec 27 23:46:12 thx Dec 27 23:48:25 WARNING: CONFIG_AUTOFS4_FS is invalid … It is unset … Allowed values : y, m, ! Dec 27 23:48:27 is this a spell error? Dec 27 23:50:03 why should it be? You need to set it to y or m Dec 27 23:50:17 and which one? Dec 27 23:50:22 and what does which mean :P Dec 27 23:50:36 just try setting it to y Dec 27 23:51:02 invalid does only mean it's not how mer-verify-config wants it to be, it's no spelling error Dec 27 23:51:16 yeah Dec 27 23:51:25 i just thought its a spelling error and maybe someone mistyped m and n Dec 27 23:56:55 ok mer is good now ;) Dec 28 00:03:02 @UniversalSuperBox ah we could get those rules during the build process .. but we need them inside the rootfs .. right? Dec 28 00:03:11 yes Dec 28 00:03:15 or initramfs? Dec 28 00:03:21 absolutely no Dec 28 00:03:25 i already rebuilt it Dec 28 00:04:35 @anpok, We could add the lxc-android repository to the repo manifest and add the rules to there while building. Then you'd have to make a PR and your changes would be in the next rootfs build Dec 28 00:04:46 but that's not really fast Dec 28 00:04:58 No, because we have a lot of OSs depending on this Dec 28 00:05:02 Some of them aren't Ubuntus Dec 28 00:05:24 this ==? the lxc-android config .. or the initramfs Dec 28 00:05:51 Including every device's rules in the rootfs isn't scalable Dec 28 00:05:58 Just like fixup-mountpoints isn't scalable Dec 28 00:06:12 it would be nice if that would come 'with' the system.img Dec 28 00:06:18 Exactly Dec 28 00:06:29 what about a per device config file in the device tree? Dec 28 00:07:03 or lets say a per-system file for each system that wants to build on top of us? Dec 28 00:07:24 It needs to be given per-device, but from Halium. Dec 28 00:07:34 fixup-mountpoints already shows you why Dec 28 00:08:20 btw ive seen a ? under progress for the bacon Dec 28 00:08:25 is anyone actually working on it? Dec 28 00:19:56 @UniversalSuperBox as long as we do not ship the android container with its system image in some sort of package format capable of doing all that we should just put an udev rules file into userdata and have the initramfs init process copy it Dec 28 00:20:26 that means /lib/udev/rules.d needs to be writable? Dec 28 00:23:16 why exactly i this taking ages? Dec 28 00:23:17 [ 99% 2063/2067] Making initramfs : /media/clemens/Elements/halium/out/target/product/bacon/obj/ROOT/hybris-boot_intermediates/boot-initramfs.gz Dec 28 00:24:38 @anpok, I don't want the initrd modifying the rootfs. :P Dec 28 00:27:46 uhm Dec 28 00:27:51 anyone here for first aid? Dec 28 00:27:52 ninja: error: '/media/clemens/Elements/halium/out/host/linux-x86/framework/signapk.jar', needed by '/media/clemens/Elements/halium/out/target/product/bacon/obj/APPS/TimeService_intermediates/package.apk', missing and no known rule to make it Dec 28 00:28:01 Remove it from the makefile that makes it Dec 28 00:28:25 again a java lib? Dec 28 00:28:31 y Dec 28 00:28:37 remove the whole ninja wrapper? Dec 28 00:29:01 no Dec 28 00:29:13 remove it from $device-vendor.mk in the vendor blobs Dec 28 00:29:37 ah Dec 28 00:30:21 cant find int in there Dec 28 00:30:35 look at the other makefiles in therre Dec 28 00:30:43 It's called TimeService Dec 28 00:30:51 cant find it anywhere in the oneplus/bacon dir Dec 28 00:30:55 or: grep cant find it :P Dec 28 00:31:21 so what Dec 28 00:31:21 Might be in your vendor files then Dec 28 00:31:22 remove the whole apk? Dec 28 00:31:29 i was checking the vendor files Dec 28 00:31:32 i found timeservice.apk Dec 28 00:32:05 and i can uncomment libTimeService in the bacon-vendor.mk Dec 28 00:32:43 would that work? Dec 28 00:33:12 There's something that copies and signs that apk Dec 28 00:34:04 yeah Dec 28 00:34:13 i believe its called from within bacon-vendor.mk Dec 28 00:34:18 but can i just remove that apk? Dec 28 00:35:26 btw. wanted to comment that out … vendor/oneplus/bacon/bacon-vendor.mk was modified, regenerating... Dec 28 00:36:04 yeah that's fine, ninja just generates new build files from your changes Dec 28 00:36:38 i think i made the wrong changes Dec 28 00:36:40 oh Dec 28 00:36:41 well Dec 28 00:36:48 shall i jsut exclude that apk? Dec 28 00:37:27 Exclude all APKs Dec 28 00:38:05 Just make similar changes to this commit: https://github.com/JBBgameich/proprietary_vendor_bq/commit/5469b289cc167c17c6554aa5574f8a5d6b658dd2 Dec 28 00:38:44 i think i did Dec 28 00:39:46 tho you probably should rather change the device tree and let the script then regenerate everything else, idk Dec 28 00:40:03 i hope building systemimage is a little bit faster Dec 28 00:40:04 I did that too :) Dec 28 00:41:43 yey Dec 28 00:41:44 * Cl1608Ho no jobserver pipe, prefix recipe command with '+' **** ENDING LOGGING AT Thu Dec 28 03:00:02 2017 **** BEGIN LOGGING AT Thu Dec 28 03:00:03 2017 Dec 28 07:45:04 Anyone? Dec 28 08:40:41 stupid me, i accidentally modified ninja.mk :P … ninja: error: '/media/clemens/Elements/halium/out/target/product/bacon/obj/SHARED_LIBRARIES/libcryptfs_hw_intermediates/export_includes', needed by '/media/clemens/Elements/halium/out/target/product/bacon/obj/EXECUTABLES/vold_intermediates/import_includes', missing and no known rule to make it Dec 28 09:50:51 (Photo, 1280x720) https://irc.ubports.com/W82T6EBl/file_3279.jpg How can i fix this Dec 28 09:53:19 Here is the manifest file … goo.gl/cSLTvH Dec 28 09:57:34 did you check out the binary blobs? Dec 28 09:58:12 yeah you did, did you do a repo sync afterwards? Dec 28 09:58:21 I did Dec 28 09:58:51 ah i have your error Dec 28 09:59:08 … remove the /wt88047 from the path Dec 28 09:59:31 (its usually a hint to take whats in the name, remove the first word and replace _ by / :P Dec 28 10:01:44 So this was the building adventure mentioned of halium page Dec 28 10:01:47 😂 Dec 28 10:08:00 @Cl1608Ho, Fixed the vendor blob error ... Still the same error Dec 28 10:08:54 Repo sync... Dec 28 10:09:01 Did Dec 28 10:09:31 Could you pastebin both the manifest and the output? Dec 28 10:09:38 Looking at actual Screenshots is painful... Dec 28 10:09:50 Wait Dec 28 10:10:47 Oh sorry Dec 28 10:10:48 @Cl1608Ho, +1 Dec 28 10:11:38 I actually can't remember whether I needed to do breakfast bacon or breakfast oneplus_bacon in order for it to work. You might also try `breakfast wingtech_wt88047` Dec 28 10:11:55 breakfast bacon Dec 28 10:12:00 just codename Dec 28 10:12:05 Yeah but idk Dec 28 10:12:07 Tried both Dec 28 10:12:09 Messed around Dec 28 10:12:13 In the end it was working :P Dec 28 10:12:16 `lunch lineage_bacon-userdebug` Dec 28 10:12:27 yeah, this for 5.1 trees Dec 28 10:12:29 Ehm. Im having breakfast, not lunch Dec 28 10:12:39 @vanyasem, not for 7.1 Dec 28 10:12:42 A little bit early for lunch, right? Dec 28 10:12:51 @vanyasem, I think it works in 7.1 too? Is it deprecated? Dec 28 10:13:02 deprecated and doesn't work with newer devices Dec 28 10:13:10 Well let's see what error @Optimus9650 has Dec 28 10:13:19 if a device was updated from 5.1 to 7.1, then it might work Dec 28 10:13:31 if it's a new device, then it won't Dec 28 10:13:53 it's in the docs btw Dec 28 10:14:07 https://pastebin.com/FH3NpWHk … Updated manifest Dec 28 10:15:14 looks ok Dec 28 10:15:32 https://pastebin.com/aL3Ef7Qn … Output Dec 28 10:16:11 @Optimus9650, Android version? Dec 28 10:16:16 5.1 or 7.1? Dec 28 10:16:20 7.1 Dec 28 10:17:11 what is your default python version? Dec 28 10:17:23 run `python` and tell me Dec 28 10:17:46 3.6.3 Dec 28 10:17:55 that's why it doesn't work Dec 28 10:17:57 ohhh Dec 28 10:18:41 Arch devs made 3.6 default because they are not too wise Dec 28 10:19:07 you need to make a local override for it in user bin or make Dec 28 10:19:12 @vanyasem, Dude ! 😡😂 Dec 28 10:19:28 that's how it is Dec 28 10:19:38 I'm on Arch as well and I encountered that issue too Dec 28 10:19:55 And how to that... Dec 28 10:21:47 @Optimus9650, `sudo ln -s /bin/python2 /usr/local/bin/python` Dec 28 10:21:51 should do it Dec 28 10:24:52 can anyone help me with this … system/vold/cryptfs.c:74:10: fatal error: 'cryptfs_hw.h' file not found Dec 28 10:24:57 beware though, some apps that were packaged by arch maintainers might stop working after that, so you may want to remove the override when you finish with halium (`ponysay` breaks) Dec 28 10:25:02 @vanyasem, `mkdir ~/bin … ln -s /usr/bin/python ~/bin/python … echo 'export PATH=~/bin:$PATH' >> ~/.bashrc … ` Dec 28 10:25:22 @trivial_inanity, also an option, but i prefer the other way Dec 28 10:25:29 @Cl1608Ho, just a sec Dec 28 10:25:48 great :) (is it a dependency on my system?) Dec 28 10:25:59 @Cl1608Ho, add ` ` to your manifest Dec 28 10:26:25 @trivial_inanity, also, that won't work Dec 28 10:26:41 ok Dec 28 10:26:45 you're overriding python3 with python3. as `/usr/bin/python` is python3 Dec 28 10:26:46 @vanyasem, already did, otherwise it would fail earlier Dec 28 10:27:04 oh wait Dec 28 10:27:10 is it cryptfs_hw or cryptfs/hw? Dec 28 10:27:17 as i wrote Dec 28 10:27:26 cause the latter also fixed errors earlier :) Dec 28 10:27:58 Same error .. after overriding with python2.7 Dec 28 10:28:00 btw earlier yuo state one should fork all the stuff one is adding there, is that right or not? (i mean.... is there a necessity of later changing stuff?) Dec 28 10:28:25 @Optimus9650, reboot? and run `python` and tell me which version it is Dec 28 10:28:51 @Cl1608Ho, if you change something, you need to fork upstream and push commits in your local repo Dec 28 10:29:08 yeah, but is there a need for that usually? Dec 28 10:29:26 device and kernel trees are usually forked Dec 28 10:29:39 i guess you've already modified a kernel config Dec 28 10:29:42 haven't you? Dec 28 10:29:58 yeah i have Dec 28 10:30:07 so push it to your fork then? Dec 28 10:30:08 but didnt push a thing Dec 28 10:30:12 how? Dec 28 10:30:24 read about using git, that's basics Dec 28 10:30:33 ehm Dec 28 10:30:36 i know git Dec 28 10:30:43 i just dont know the whole setup with repo and 200 repos Dec 28 10:30:51 cd to the kernel repo Dec 28 10:30:58 i dont know in which state my working tree is :) Dec 28 10:31:05 i.e. `kernel/htc/msm8996` from the root of the tree Dec 28 10:31:08 btw, does repo sync revert local cahnges? Dec 28 10:31:15 and then run usual `git commit` etc Dec 28 10:31:19 ok Dec 28 10:31:21 pushing too? Dec 28 10:31:27 @Cl1608Ho, sometimes, if you don't commit and push... Dec 28 10:31:32 ... Dec 28 10:31:37 something like repo sync after pushing? Dec 28 10:31:39 if you do commit, repo will load your changes instead Dec 28 10:31:50 yeh Dec 28 10:31:57 @Cl1608Ho, just usual git stuff, repo just clones a lot of repos really quick Dec 28 10:32:03 Python output 2.7 Dec 28 10:32:08 they are just usual git repos Dec 28 10:32:15 @Optimus9650, have you rebooted? Dec 28 10:32:38 @vanyasem, be sure to update your repo manifest Dec 28 10:32:43 to point to your local fork Dec 28 10:32:47 @vanyasem, Yes Dec 28 10:33:08 And the same error still Dec 28 10:33:21 @vanyasem, https://gist.github.com/vanyasem/64fa7fdc0df556561c50a6048a982ce1/f65ccd44d9cacf386628622f417441f28711bb35 something like this (old commit, outdated) Dec 28 10:34:47 @Optimus9650, have you tried syncing you repos again? Dec 28 10:35:34 @vanyasem, Yes Dec 28 10:35:38 I did Dec 28 10:35:39 @Optimus9650, oh Dec 28 10:35:41 i know why Dec 28 10:35:45 your repo is incorrect Dec 28 10:36:04 no i don't know why Dec 28 10:36:38 😐 Dec 28 10:36:50 have you set up your enviroment before running it? Dec 28 10:36:57 `. build/envsetup.sh` Dec 28 10:37:15 git also tracks file permissions, right? Dec 28 10:37:42 @vanyasem, Yes Dec 28 10:37:46 @Cl1608Ho, no Dec 28 10:37:51 @Cl1608Ho, https://stackoverflow.com/questions/3207728/retaining-file-permissions-with-git Dec 28 10:38:16 well git covers the execution bit Dec 28 10:38:30 that is true Dec 28 10:38:35 yeah Dec 28 10:38:36 well Dec 28 10:38:37 but not the permissions themselve Dec 28 10:38:45 is there an option to set git config for a lot of repos, but not global? Dec 28 10:39:15 chroot maybe? Dec 28 10:39:19 cause the repo is on an external ntfs drive and idk if its just wrong configuration, but my system sets 777 on any file :P Dec 28 10:39:27 @Cl1608Ho, it's okay Dec 28 10:39:32 i work on NTFS as well Dec 28 10:39:34 no issues Dec 28 10:39:47 apart from chkdsk deleting half of my files from time to time Dec 28 10:39:51 well i dont wanna commit the executional bit for every file Dec 28 10:40:07 don't worry about it Dec 28 10:40:09 git status still shows all files as changed Dec 28 10:40:17 that's an issue Dec 28 10:40:27 wow Dec 28 10:40:29 try to add them all and then remove Dec 28 10:40:41 they should be gone afterwards Dec 28 10:40:56 the remove is git checkout -- *? Dec 28 10:41:33 So i am stuck? Dec 28 10:41:38 @Optimus9650, yes Dec 28 10:41:51 sorry Dec 28 10:42:01 🤨 Dec 28 10:42:23 the error is exactly the same? Dec 28 10:42:24 @Optimus9650, white square, wow Dec 28 10:42:32 @Cl1608Ho, Yes Dec 28 10:42:59 @vanyasem, Below 7.1? Dec 28 10:43:30 On my desktop, I don't really use smartphones Dec 28 10:43:47 @vanyasem, It's a new emoji Dec 28 10:43:53 i think i know the thign Dec 28 10:43:55 let me check Dec 28 10:47:27 try adding this to your manifest … Dec 28 10:47:39 and repo sync and stuff Dec 28 10:47:47 wait.. Dec 28 10:47:56 @Cl1608Ho, oh, nice idea Dec 28 10:47:57 should work Dec 28 10:48:13 @vanyasem maybe should be stated in the docs that one should recursively crawl for dependencies.... Dec 28 10:48:42 lineage.dependencies in the dependencies' repos.... Dec 28 10:51:35 @Cl1608Ho, didnt work Dec 28 10:51:49 😞 Dec 28 10:52:48 hm Dec 28 10:54:25 "Don't have a product spec for: 'lineage_wt88047" Dec 28 10:55:46 hey who invited continuing builds? Dec 28 10:55:47 [ 0% 5/3713] Target buildinfo: /media/clemens/Elements/halium/out/target/product/bacon/obj/ETC/system_build_prop_intermediates/build.prop Dec 28 10:55:58 thanks that its not 5/6500 again :) Dec 28 10:57:25 well according to my error debugging... the script is unable to find wt88047 in github.com/lineageos Dec 28 10:57:44 which isnt problematic Dec 28 10:58:26 Device wt88047 not found. Attempting to retrieve device repository from LineageOS Github (http://github.com/LineageOS). … Found repository: android_device_wingtech_wt88047 Dec 28 10:58:27 and also doesnt make any sense at all Dec 28 10:58:45 exactly Dec 28 11:00:12 it does Dec 28 11:00:17 because it just uses githubs search Dec 28 11:00:25 or kindof Dec 28 11:01:48 so when you guys were bulding ... was the the output any similar to mine? Dec 28 11:02:36 I think revision is not required, it's already defined in remote. Dec 28 11:03:11 so should i remove the the revision="xxx" ? Dec 28 11:03:18 from the manifest Dec 28 11:03:43 might be good Dec 28 11:03:57 at least modify it to "refs/heads/cm-14.1" Dec 28 11:04:12 for all Dec 28 11:04:12 ? Dec 28 11:04:48 have you tried performing a clean sync Dec 28 11:04:50 @Optimus9650 Dec 28 11:05:15 @vanyasem, it is a clean sync Dec 28 11:05:30 i mean re-sync everything agaibn Dec 28 11:05:46 dude ..😨 Dec 28 11:06:50 no dont Dec 28 11:07:04 @Optimus9650 dont sync everything again Dec 28 11:07:22 Has anyone got an error stating `mkbootimg: Command not found` ? Dec 28 11:07:41 why are you runing mkbootimg Dec 28 11:07:51 run `mka mkbootimg` first Dec 28 11:07:57 @JBBgameich, exactly Dec 28 11:08:33 @Optimus9650, Just try to remove the "revision" and see Dec 28 11:10:06 @shenoy13, i replaced every revision with this... didnt work... Dec 28 11:10:17 @shenoy13, now trying this one Dec 28 11:13:30 @shenoy13, and again... same erro Dec 28 11:14:26 do i have to do mka bootimg and mka hybris-boot again after i added some dependencies? (those two were working fine yesterday, mka systemimage failed because of missing repos, i added them and now mka systemimage worked. do i have to re-run the other ones? Dec 28 11:14:35 can you copy paste the error again? :P Dec 28 11:14:52 @Cl1608Ho, me ? Dec 28 11:15:55 [abhishek@OPTIMUS halium]$ breakfast wt88047 … including vendor/cm/vendorsetup.sh … build/core/product_config.mk:249: * Can not locate config makefile for product "lineage_wt88047". Stop. … build/core/product_config.mk:249: * Can not locate config makefile for product "cm_wt88047". Stop. … build/core/product_config.mk:249: * Can not locate config makefile for product "lineage_wt88047". Stop. … Device wt88047 not found. Attempting to Dec 28 11:15:56 repository from LineageOS Github (http://github.com/LineageOS). … Found repository: android_device_wingtech_wt88047 … Traceback (most recent call last): … File "vendor/cm/build/tools/roomservice.py", line 263, in … default_revision = get_default_revision() … File "vendor/cm/build/tools/roomservice.py", line 117, in get_default_revision … return r.replace('refs/heads/', '').replace('refs/tags/', '') … AttributeError: 'NoneType Dec 28 11:15:57 'replace' … build/core/product_config.mk:249: * Can not locate config makefile for product "lineage_wt88047". Stop. … build/core/product_config.mk:249: * Can not locate config makefile for product "lineage_wt88047". Stop. … Don't have a product spec for: 'lineage_wt88047' … ** Do you have the right repo manifest? … [abhishek@OPTIMUS halium]$ Dec 28 11:16:24 Is this halium 5 or 7? Dec 28 11:16:29 7 Dec 28 11:16:44 Can you point me to device tree? Dec 28 11:17:13 here is the manifest Dec 28 11:17:14 https://gist.github.com/abhishek9650/0a33682de9c2307dd8694ea2168cb578 Dec 28 11:17:17 (also please don't paste more then 4 lines here, use pastebin) Dec 28 11:17:46 @Optimus9650, he used it before, idk what happened now Dec 28 11:19:12 Try lunch -eng Dec 28 11:20:10 @bhushanshah, isnt that for 5.1 ? Dec 28 11:20:18 No Dec 28 11:22:07 @bhushanshah, https://hastebin.com/fowaxozeyo.rb Dec 28 11:23:18 is device repository synced correctly? Dec 28 11:23:26 yes Dec 28 11:24:04 Weird, anyway I need to leave right now will try to help later in hour or so Dec 28 11:24:34 #### make completed successfully (29:08 (mm:ss)) #### Dec 28 11:24:36 woho Dec 28 11:25:05 @bhushanshah, okay..... by the time i will clean sync it again... Dec 28 11:25:06 testing now Dec 28 11:25:46 @Optimus9650, Do this lunch lineage_wt88407-eng Dec 28 11:26:34 http://bshah.in/halium/halium-rootfs-20170630-151006.tar.gz … is this still the latest rootfs? Dec 28 11:26:50 @bhushanshah Dec 28 11:26:57 @mvaisakh, https://hastebin.com/uwubaninox.coffeescript Dec 28 11:27:50 You have the device tree synced in device/wingtech/wt88407? Dec 28 11:28:43 wait... what... Dec 28 11:29:22 Okay so the device tree is present in lineage git right? Dec 28 11:29:27 [abhishek@OPTIMUS halium]$ cd device/ … [abhishek@OPTIMUS device]$ ls … google Dec 28 11:29:39 (Sticker, 392x512) https://irc.ubports.com/hpDKA48V/file_3283 Dec 28 11:29:51 Do this Dec 28 11:30:26 what is the apprx size of your halium folder? Dec 28 11:31:04 @Optimus9650, Around 20 gb Dec 28 11:31:21 I'm basically Android Rom developer Dec 28 11:31:35 Not Halium Dec 28 11:31:38 and mine is 16.2 gb .. did i miss something? Dec 28 11:31:46 @Optimus9650 can you give me output of ls -l device? Dec 28 11:31:51 Nope Dec 28 11:32:37 [abhishek@OPTIMUS device]$ ls -l … total 4 … drwxr-xr-x 3 abhishek users 4096 Dec 28 01:31 google Dec 28 11:33:11 Are you really sure that you synced device tree? Dec 28 11:34:04 Shared project android_hardware_qcom_wlan found, disabling pruning. … Syncing work tree: 100% (200/200), done. Dec 28 11:34:16 @Optimus9650, do this … `git clone https://github.com/LineageOS/android_device_wingtech_wt88047 -b cm-14.1 device/wingtech/wt88047` Dec 28 11:34:52 @mvaisakh, I'd rather fix local manifest properly Dec 28 11:35:07 Lol true that Dec 28 11:35:30 @Optimus9650, Again, device tree .. not just repo.. Dec 28 11:36:11 Wouldn't it be good to just "diff" and see if it is synced properly? Dec 28 11:36:14 @mvaisakh, now i am doing this Dec 28 11:37:17 @Optimus9650 can you run `source build/envsetup.sh` and post the output here? Dec 28 11:38:19 @trivial_inanity, https://hastebin.com/sulenawiho.rb Dec 28 11:41:54 You're probably missing an `add_lunch_combo`. … Create a file named `vendorsetup.sh` in the vendor tree ( in `vendor/wingtech/wt88047` according to the manifest above with the following content: … ```add_lunch_combo lineage_wt88047-userdebug``` Dec 28 11:42:14 @trivial_inanity, Well.. he's missing whole device repo Dec 28 11:42:22 @bhushanshah, yes Dec 28 11:42:23 @Optimus9650, . Dec 28 11:42:38 and this is embarrasing Dec 28 11:43:04 since device folder is completely empty Dec 28 11:43:24 i told you to re-sync already Dec 28 11:43:28 @Optimus9650, Did you clone it or add it to the manifest? Dec 28 11:43:50 * trivial_inanity will leave you be in peace then :] Dec 28 11:44:34 @vanyasem, resyncing .. went completely fine .. maybe sommething wrong with manifest Dec 28 11:44:49 *must Dec 28 11:45:10 try to start from scratch, delete everything and start again Dec 28 11:45:11 🤦‍♂️ Dec 28 11:45:13 as i told you Dec 28 11:45:24 @vanyasem, doing that Dec 28 11:45:37 Well you don't need to delete .repo folder Dec 28 11:45:43 You can just clean worktree Dec 28 11:45:47 @bhushanshah, yep Dec 28 11:45:48 And resynx Dec 28 11:45:58 So it won't download everything again Dec 28 11:47:44 @Optimus9650 from your source directory, can you run the following command? `find .repo/local_manifests/` Dec 28 11:48:25 yes i can Dec 28 11:48:37 post output ;) Dec 28 11:48:59 [abhishek@OPTIMUS halium]$ find .repo/local_manifests/ … .repo/local_manifests/ Dec 28 11:49:34 where is your manifest?? Dec 28 11:49:51 you manifest is not there Dec 28 11:50:39 [abhishek@OPTIMUS .repo]$ ls … local_manifests manifests manifests.git manifest.xml project.list project-objects projects repo Dec 28 11:50:56 please read the damn docs Dec 28 11:51:15 your local manifest should be in .repo/local_manifests/$VENDOR_$CODENAME Dec 28 11:51:24 @Optimus9650, ```$ find .repo/local_manifests … .repo/local_manifests … .repo/local_manifests/htc_pme.xml``` Dec 28 11:51:27 it is MISSING Dec 28 11:51:34 where is your local manifest? Dec 28 11:51:39 please read the docs properly Dec 28 11:52:13 fuck this... i will back soon.... by fixing every damn mistake i did... Dec 28 11:52:25 please don't swear Dec 28 11:52:26 coz i am not giving up on this Dec 28 11:52:29 sorry Dec 28 11:52:36 for the swear Dec 28 11:54:01 maybe i should do sudo rm -rf /* ... maybe that will solve my lot of problems😂😂 Dec 28 11:54:08 brb Dec 28 12:28:33 Quick question: https://docs.halium.org/en/latest/porting/debug-build/usbnet.html Dec 28 12:28:40 How can I set that up if telnet is never exposed? Dec 28 12:28:45 Or is it just not needed.... Dec 28 12:34:41 hybris-recovery always exposes telnet for debugging. You can build it using `mka hybris-recovery`, and it should be flashed to the boot partition just as hybris-boot. Dec 28 12:39:51 Sorry Dec 28 12:40:00 I was about to say the backup works fine Dec 28 12:40:29 I was about to say the device boots fine, and ssh is exposed Dec 28 14:33:03 i'm getting logcat read failure, ping gives me socket: Permission denied.... Dec 28 14:35:52 I remember someone having that issue before Dec 28 14:35:56 i forgot the fix though Dec 28 14:42:57 ok Dec 28 14:48:57 root@localhost:~# echo 1 > /dev/wcnss_wlan … -bash: /dev/wcnss_wlan: Bad address Dec 28 14:57:04 try to search this chat Dec 28 14:57:08 it usually helps Dec 28 14:57:22 is the container already started fully? Dec 28 14:57:28 You can essentially skip the echo 1 part Dec 28 14:57:46 @bhushanshah, yep Dec 28 14:57:50 Ah also, is container started? Dec 28 14:58:03 i meant that to the logcat issue Dec 28 14:58:44 I believe someone should edit docs to clarify that echo 1 is only needed if echo sta doesn't work Dec 28 14:59:33 (PRs welcome 😛) Dec 28 14:59:41 root@localhost:~# echo sta > /sys/module/wlan/parameters/fwpath … -bash: echo: write error: No such device … and dmesg … [ 1090.472518] wlan: loading driver v3.0.10.085 … [ 1100.503394] wlan: [F :HDD] hdd_driver_init: WCNSS driver not ready Dec 28 14:59:46 i'm looking already... Dec 28 15:00:00 @Cl1608Ho, run `systemctl status lxc@android` Dec 28 15:00:03 and pastebin output Dec 28 15:00:21 also run `lxc-checkconfig` and post output Dec 28 15:00:33 full log or short one? Dec 28 15:00:51 huh? Dec 28 15:02:12 https://paste.ubuntu.com/26272014/ Dec 28 15:02:23 @Cl1608Ho, `Active: failed` Dec 28 15:02:30 your container failed to start Dec 28 15:02:35 @vanyasem, now post this Dec 28 15:02:43 its below in that paste Dec 28 15:02:55 oh Dec 28 15:03:33 @Cl1608Ho, fix everything except user namespaces Dec 28 15:03:48 how? :P Dec 28 15:03:58 modify kernel config Dec 28 15:04:02 ah Dec 28 15:04:08 what do i need to recompile? Dec 28 15:04:15 hybris-boot Dec 28 15:04:23 run checkconfig (i suggest you to also use ubport's checkconfig) Dec 28 15:04:51 (Document) https://irc.ubports.com/LqyaFyfv/file_3287 Dec 28 15:04:52 here it is Dec 28 15:05:03 i have no idea from what repo it is taken Dec 28 15:05:06 that' Dec 28 15:05:15 https://github.com/janimo/phablet-porting-scripts … Probably from here Dec 28 15:05:26 @vanyasem, that's the newest one Dec 28 15:05:34 @JBBgameich, the one in this repo is old as hell Dec 28 15:05:47 it was modified by Marius for Xenial Dec 28 15:05:58 @vanyasem, anyways the executable is here so Dec 28 15:06:10 UBports could fork the repo and include the latest changes maybe Dec 28 15:06:29 exactly. but i still have no idea where is that repo Dec 28 15:06:31 that was my point Dec 28 15:06:46 we'll get this person confused now. we both understand that :P Dec 28 15:09:36 hm i'm getting more and more weird problems Dec 28 15:09:43 @Cl1608Ho, huh Dec 28 15:09:47 running source build/envsetup.sh and my path variable is unchanged Dec 28 15:10:01 try in a new shell maybe? Dec 28 15:11:11 that *is* in a new shell Dec 28 15:11:13 thats the problem Dec 28 15:11:18 try an another one? Dec 28 15:11:19 reboot? Dec 28 15:11:24 :P Dec 28 15:11:28 how can i help you with you local computer's problems? Dec 28 15:11:48 yeah Dec 28 15:11:49 i know Dec 28 15:11:52 if it doesn't error, then that's an issue with your GNU/Linux setup Dec 28 15:12:39 Are you using bash as shell? Dec 28 15:12:46 yes Dec 28 15:14:08 @JBBgameich, that should not matter i believe. all scripts have appropriate shebangs in them Dec 28 15:14:29 i learned that word from your git repo btw, thank you Dec 28 15:14:52 sheband sounds really weird to me, sounds like a kind of pornography... Dec 28 15:15:08 I learned it from the codacy website Dec 28 15:18:52 it worked as soon as i unplugged and turned off my phone Dec 28 15:18:53 strange Dec 28 15:19:30 @JBBgameich, there are a lot of synonyms though Dec 28 15:19:34 i searched on wiki Dec 28 15:21:05 CONFIG_DNS_RESOLVER is neither enabled nor disabled in the config file … what to do with those? (~100 of those?) Dec 28 15:21:10 shall i enable or disable Dec 28 15:21:29 i fixed all the repo errors Dec 28 15:21:32 you can just run it with `-w` option (or whatever it is) Dec 28 15:21:34 https://hastebin.com/tahukisoji.rb Dec 28 15:21:44 @vanyasem, it will fix most errors for you Dec 28 15:21:46 here is the error Dec 28 15:21:47 automatically Dec 28 15:21:57 @vanyasem, then fix the remaining errors by hand Dec 28 15:22:03 god bless you ^^ Dec 28 15:22:19 @Optimus9650, please run that `find` command again Dec 28 15:22:24 and paste it's output Dec 28 15:22:53 I love how you change your pastebin every time you paste something ^^ Dec 28 15:23:29 😛 Dec 28 15:23:43 @trivial_inanity, this Dec 28 15:24:25 @vanyasem, @Cl1608Ho ? Dec 28 15:24:29 [abhishek@OPTIMUS halium]$ find .repo/local_manifests/ … .repo/local_manifests/ … .repo/local_manifests/wingtech_wt88047.xml Dec 28 15:25:52 o Dec 28 15:25:56 now run `ls device -l` Dec 28 15:26:02 @Optimus9650 Dec 28 15:26:25 $ ls device -l … total 12 … drwxr-xr-x 3 abhishek users 4096 Dec 28 20:33 cyanogen … drwxr-xr-x 3 abhishek users 4096 Dec 28 20:33 google … drwxr-xr-x 3 abhishek users 4096 Dec 28 20:33 wingtech Dec 28 15:26:33 oh I haven't looked at your rrrir Dec 28 15:26:34 config-checker and mer say different things for a few config values, which to choose? Dec 28 15:26:44 @Optimus9650, you're missing sepolicy Dec 28 15:26:48 in your manifest Dec 28 15:26:57 either add it or disable from makefiles Dec 28 15:27:05 @Cl1608Ho trust mer-check Dec 28 15:27:09 :P Dec 28 15:27:34 @vanyasem, `` Dec 28 15:27:52 k :) Dec 28 15:27:57 @bhushanshah, they conflict on just one option, I tested that Dec 28 15:27:58 @vanyasem generally ubports config-check is not useful for halium as-it-is Dec 28 15:28:01 so just recompile the boot checker? Dec 28 15:28:07 for me it was 3 i think Dec 28 15:28:24 anyways prioritize mer check Dec 28 15:29:18 @vanyasem, @Optimus9650 add this to your local manifest Dec 28 15:30:22 rebuild just the hybris-boot? Dec 28 15:30:26 @vanyasem, did ..now syncing Dec 28 15:30:39 @Cl1608Ho, yes Dec 28 15:31:23 I always rebuild systemimage just in case, but don't listen to me Dec 28 15:33:20 :P Dec 28 15:38:03 @vanyasem i am at mir kernel config check Dec 28 15:38:18 here is a todo note what that means Dec 28 15:38:50 @Optimus9650, what? Dec 28 15:39:16 also grep command for path to kernel config .. returned nothing Dec 28 15:40:15 @Optimus9650, Checked path? Dec 28 15:40:55 @TheImpulson, grep "TARGET_KERNEL_CONFIG" device///BoardConfig.mk Dec 28 15:41:02 returned nothing Dec 28 15:41:14 locate it manually then Dec 28 15:41:56 is your device 32 bit or 64 bit Dec 28 15:42:13 32 Dec 28 15:42:50 just a sec, my hdd is overloaded Dec 28 15:43:18 sure.. btw device supports 64bit ... but i am building 32 Dec 28 15:43:26 @Optimus9650, what? Dec 28 15:43:37 what it your device's processor architecture? Dec 28 15:43:42 arm or arm64? Dec 28 15:43:45 i am warming up Dec 28 15:44:12 @Optimus9650, How is that possible? Dec 28 15:44:24 @TheImpulson, that just makes no sense :P Dec 28 15:44:31 @Optimus9650, 32 bit can run fine on 64 bit Dec 28 15:44:49 dude ... my device supports .. arm64... but i am building 32.. coz lineage built for my device is only is 32 Dec 28 15:44:57 @bhushanshah, this guy gets it Dec 28 15:45:06 that's the thing, lineage is 32-bit Dec 28 15:45:11 `kernel/$VENDOR/$BOARD/arch/arm/configs/$CODENAME_defconfig` Dec 28 15:45:17 your config should be there Dec 28 15:45:24 if not, look in `kernel/$VENDOR/$BOARD/arch/arm64/configs/$CODENAME_defconfig` Dec 28 15:48:48 were you able to locate it? Dec 28 15:50:58 @vanyasem, what you think Dec 28 15:51:05 😆 Dec 28 15:51:26 post `ls -l` output from that folder to pastebin Dec 28 15:52:23 nvm found it Dec 28 15:52:43 alrighty, now fix the config Dec 28 15:52:51 were you able to fix your selinux error? Dec 28 15:55:24 @vanyasem, https://paste.ee/p/6dn0y Dec 28 15:55:49 that was not addressed to you Dec 28 15:55:55 fix your selinux issues first Dec 28 15:56:19 Fwd from vanyasem: you're missing sepolicy Dec 28 15:56:20 Fwd from vanyasem: in your manifest Dec 28 15:56:21 Fwd from vanyasem: either add it or disable from makefiles Dec 28 15:56:22 Fwd from vanyasem: `` Dec 28 15:56:51 (Photo, 524x40) https://irc.ubports.com/OWUE5okN/file_3292.jpg Dec 28 15:56:55 there is your defconfig though Dec 28 15:57:04 you will need it in a moment Dec 28 15:57:16 write the path to it, so you won't have to look it up again Dec 28 15:58:15 hol up Dec 28 16:01:27 huh? Dec 28 16:02:17 right now i have gazillion errors at my face Dec 28 16:02:50 paste them so we will figure them out? Dec 28 16:03:04 it's not too early to give up, nobody said that porting it easy Dec 28 16:03:43 i have huge respect for rom devs Dec 28 16:03:45 now Dec 28 16:04:04 that is just the basic part :P Dec 28 16:05:19 https://hastebin.com/miharisixu.swift Dec 28 16:05:44 could you please not use hastebin Dec 28 16:05:53 it doesn't work with js disabled Dec 28 16:06:13 (Photo, 1024x524) https://irc.ubports.com/vF5rEfVs/file_3294.jpg Dec 28 16:06:20 that's what i see.. an empty page Dec 28 16:06:28 why you have js disabled? Dec 28 16:06:37 @Optimus9650, that's not a ton of issues Dec 28 16:06:38 privacy reasons? Dec 28 16:06:41 just a few Dec 28 16:06:42 fix them Dec 28 16:06:49 one by one Dec 28 16:07:03 @Optimus9650, yes. i am a free software activist afterall Dec 28 16:07:09 😱 Dec 28 16:11:55 @vanyasem, how Dec 28 16:12:14 @JBBgameich, That was my problem with your script in halium-devices: It executes both, but lunch after breakfast and with my harpia it doesn't work because its Android 6/7 Dec 28 16:12:18 @Optimus9650, open your defconfig and edit it according to what the log says Dec 28 16:12:24 it's IN THE DOCS Dec 28 16:12:26 read the docs Dec 28 16:12:31 read the docs read the docs read the docs Dec 28 16:13:25 @Ellenjott, I could change it to "if lunch fails use breakfast" or other way arorund Dec 28 16:14:41 https://paste.ubuntu.com/26272424/ Dec 28 16:14:46 is this really a warning thats failing the compilation? Dec 28 16:23:48 @vanyasem, log says Dec 28 16:24:02 WARNING: CONFIG_NETFILTER_XT_MATCH_NFACCT is invalid … It is unset … Allowed values : y, m, ! Dec 28 16:24:32 but ther is nothing like CONFIG_NETFILTER_XT_MATCH_NFACCT in the doc Dec 28 16:25:02 Have you read the documentation? Dec 28 16:25:24 please read it carefully Dec 28 16:25:45 Asking for help is fine, but first read documentation carefully please Dec 28 16:28:12 what are these allowed values.. y = enabled? Dec 28 16:28:56 @Optimus9650, yes. n is disabled. isn't that obvious Dec 28 16:29:27 @bhushanshah we should really make a requirement for porting to Halium Dec 28 16:29:32 that what is m and ! Dec 28 16:29:34 one requirement: "Install Gentoo" Dec 28 16:29:47 @Optimus9650, that's not related to Halium, that's just Linux kernel Dec 28 16:29:49 @vanyasem, 😂😂 Dec 28 16:30:09 install gentoo, you'll have to deal with it and you'll have a better understanding of what's going on Dec 28 16:30:50 i am first time getting into these kind of stuff.. porting .. Dec 28 16:31:13 that is NOT porting Dec 28 16:31:16 that is the linux kernel Dec 28 16:31:47 i had to start from somewhere .. so here i am Dec 28 16:31:54 (Sticker, 314x512) https://irc.ubports.com/70c0qmHc/file_3298 Dec 28 16:32:05 Agree. I started in much the same place @vanyasem Dec 28 16:32:25 We aren't making a "cool kids club" for only those with preexisting knowledge Dec 28 16:33:11 Search for ` CONFIG_NETFILTER_XT_MATCH_NFACCT` in the kernel config you where checking with mer-verify-kernel-config, and change it to y. mer-verify-kernel-config basically tells you what you need to change Dec 28 16:35:05 Change it from `# CONFIG_NETFILTER_XT_MATCH_NFACCT is not set` … to `CONFIG_NETFILTER_XT_MATCH_NFACCT=y` Dec 28 16:43:40 what if mer saying CONFIG_CGROUP_FREEZER is invalid.. but is already y in the config Dec 28 16:44:24 Maybe it shouldn't be set to y but to n Dec 28 16:44:30 what does it say? Dec 28 16:45:03 allowed values are y & ! Dec 28 16:45:29 is, CONFIG_CGROUP enabled first? Dec 28 16:45:55 yes Dec 28 16:46:08 Hmm Dec 28 16:46:41 jmnemonix was added by: jmnemonix Dec 28 16:48:00 ybotd was added by: ybotd Dec 28 16:49:50 tuna_mert was added by: tuna_mert Dec 28 16:50:45 Maybe delete the link, @JBBgameich Dec 28 17:08:41 hi … i am a developer curently thinking about to get my hands on an ubuntu touch device. whether to port my htc one m9 or buy a core device. Dec 28 17:09:13 i am not sure about the time i could spend .. Dec 28 17:11:12 so i am researching a bit right now Dec 28 17:14:07 @jmnemonix, build it... thats what i am doin right now ... Dec 28 17:34:26 can anyone have a look at my config + errors Dec 28 17:34:30 config https://paste.ee/p/TIwV7 Dec 28 17:35:43 errors https://paste.ee/p/PEteP Dec 28 17:38:04 ```Can't open kernel/wingtech/msm8916/arch/arm/configs/lineageos_wt88047_defconfig: No such file or directory``` Dec 28 17:48:06 @JBBgameich, thanks dude Dec 28 17:48:12 woohooo ... Dec 28 17:48:14 done Dec 28 17:48:24 all errors Dec 28 17:49:12 @vanyasem, lfs ... Dec 28 17:49:49 @anpok, only if you have 30 yrs Dec 28 17:49:54 😂 Dec 28 17:52:38 @anpok, gentoo is faster and still gives you the needed knowledge Dec 28 17:52:46 i'm yet still to find time to read lfs Dec 28 17:52:52 and i use gentoo :P Dec 28 18:07:24 why not build your own os ..on top of magenta Dec 28 18:07:26 ? Dec 28 18:13:25 Quick question, is Halium supposed to write to any partition other that boot and data? Dec 28 18:15:03 No, you only need /data and /boot Dec 28 18:15:40 Ok, so I'll consider it an error to mess up my /persist ;-) Dec 28 18:16:36 I don't know how this can even happen, but hey, you learn something new every day. Dec 28 18:16:40 Though, Android can write to any partition though Dec 28 18:16:59 @IGNNE by any chance did you restart lxc container? Dec 28 18:17:28 Well, that might be a reason, if android iself mounts them. Dec 28 18:17:48 @bhushanshah Sort of. Dec 28 18:18:05 Note that you should not restart lxc container in one session, Android doesn't like it at all Dec 28 18:18:33 I force-killed it for--- reasons :-P Dec 28 18:18:58 well you should never do that Dec 28 18:19:03 your android image is probably corrupt now Dec 28 18:19:05 If container failed to start, always reboot Dec 28 18:21:36 @vanyasem, I thought of this and reflashed a backup, but I was slightly suprised when my plain Lineage image started to get errors. Dec 28 18:25:21 Does ubuntu touch on xenial require a minimum version of apparmor or is there also a maximum version? Dec 28 18:46:03 uablak was added by: uablak Dec 28 18:47:02 furylions was added by: furylions Dec 28 18:48:42 welcome @uablak @furylions if you re new Dec 28 18:51:04 Rygebin was added by: developerfect Dec 28 18:51:05 Ty bro Dec 28 18:54:13 I have a question about Halium. … What is the difference between Halium and mer hybris? Dec 28 18:55:35 I mean they both look similar looking from above 🤔 Dec 28 18:58:12 mer-hybris can only be used for porting mer and derivates, Halium can be used by any distribution. Halium uses a Linux Container to run Android in, and mer-hybris just runs it inside the host system. Halium had a 7.1 tree earlier than mer-hybris, but currently both have one I think Dec 28 19:00:04 @JBBgameich, Thanks for the info👍🏼 Dec 28 19:00:50 And Halium generally tries to use the latest upstream libhybris version, while mer-hybris often uses quite old versions Dec 28 19:01:08 But Halium and mer-hybris also share many components, like hybris-boot Dec 28 19:01:18 Finally got device sensors back. It is obvious, but maybe one should increase the font size for "Make a full backup image" Dec 28 19:01:40 Like,. 150px red and flashing Dec 28 19:01:58 When writing docs, you expect the user to give a little Dec 28 19:02:06 Like, you know, reading the warnings and heeding them Dec 28 19:09:31 @Optimus9650, Thx bro Dec 28 20:16:28 tydell was added by: tydell Dec 28 20:57:42 Got plasma-mobile working on my Nexus 4 \o/ Dec 28 21:17:45 DylanVanAssche was added by: DylanVanAssche Dec 28 21:25:54 shutdown -h now Dec 28 21:39:19 As I asked in the UBports super group: Dec 28 21:42:14 Hi, I currently have an Oneplus X device running Sailfish OS using Libhydris (Android 5.1 based). Halium should be running quickly then from what I heard in the UBports super group. But is the Halium 5.1 branch good enough for Ubuntu Touch or is it advised to start with a new port based on Halium Android 7.1? Dec 28 21:48:53 From what i heard is that Halium-5.1 is more stable and Halium-7.1 supports more features, both should be fine i guess Dec 28 21:54:26 No, 5.1 sux Dec 28 21:54:52 Better 7.1 Dec 28 21:56:12 Either works. Dec 28 21:57:25 Yes but 5.1 is orcodio outdated Dec 28 22:28:49 I'm searching for people who own one of these devices: … * Motorola Moto G3 Turbo (merlin) … * Motorola Moto E 2015 LTE (surnia) … * Motorola Moto G 2015 (osprey) … * Motorola Moto X Play (lux) … They are very similar (from the porting side of view) to my Moto G4 Play (harpia) I'm porting. It would be easy to also port them. I'd just need a tester. 😃 Dec 28 22:32:18 DylanVanAssche: Oneplus X = onyx right? Dec 28 22:32:29 I got 7.1 building for that more or less for LuneOS Dec 28 22:32:42 Which uses most of Halium bits as basis Dec 28 22:32:58 So pure Halium port should be fairly straight forward I guess Dec 29 00:50:12 I'm getting a problem with the vendor files, it tries to sign apks … How can I disable apks at all, without modifying the vendor repository. … I saw that it sets a var like LOCAL_MODULE_CLASS := APPS, can I just disable somewhere to install these targets? Dec 29 00:50:32 That's a great question Dec 29 00:50:37 And one that hasn't been answered yet Dec 29 00:50:48 :D Dec 29 00:51:06 So do I have to fork the vendor repo? Dec 29 00:51:22 Or should I provide a signapk.jar or so? Dec 29 00:51:24 I'd recommend not, but definitely add your stake in the relevant issue on projectmanagement Dec 29 00:51:40 Ok Dec 29 00:51:57 Just comment out those parts of the vendor files for now Dec 29 00:52:02 But don't fork, it's dubious Dec 29 00:52:13 Will do that ... **** ENDING LOGGING AT Fri Dec 29 03:00:02 2017 **** BEGIN LOGGING AT Fri Dec 29 03:00:03 2017 Dec 29 03:29:02 I've got a better question than what I've tried asking in the past that'd likely get me in over my head: if I'm not a developer and know nothing about the languages used in Halium,(most I know is a smattering of Python and enough Bash to automate things), what can I do to help? Dec 29 05:16:31 You could port your mobile :) Dec 29 05:17:03 Fwd from bhushanshah: Documentation can be found at https://docs.halium.org Dec 29 06:01:35 Ellenjott: What we do in LuneOS is to provide patches that are applied before the build Dec 29 06:01:50 Like: https://github.com/webOS-ports/android_luneos_hal/pull/3 Dec 29 06:02:46 And we use this as build script: https://bpaste.net/show/bda907a60853 Dec 29 06:32:32 @Herrie, Yes, it's codename is onyx. I saw on the Halium Github that the onyx kernel was already there :) Dec 29 06:41:14 DylanVanAssche: This is what we use: https://github.com/webos-ports/android/tree/luneos-halium-7.1 it's specified in LuneOS targets Dec 29 06:41:45 Together with the link about for android_luneos_hal which holds the patches Dec 29 06:41:51 WE don't use hybris-boot currently Dec 29 06:42:04 For the rest our manifest is 1:1 the Halium one ;) Dec 29 06:46:00 I'm not so familiar with porting (first time :p ) but I will take a look at the repos and compare it with the Halium docs to see what needs to be done. Is hybris-boot an issue for Ubuntu Touch? Dec 29 06:56:21 DylanVanAssche: Also not sure ;) Also my first port ;) Dec 29 06:56:35 Just thought to share what we've done already Dec 29 06:56:41 SAves you reinventing the wheel Dec 29 06:56:54 You need to update defconfig probably still etc Dec 29 06:57:09 We just use the small created output for Android in our further build Dec 29 07:11:46 Herrie how far you gone.. (mido) Dec 29 07:14:45 Satyam2345: Well I got a minimal Android, just having problems with booting due to partition layout currently. Need to add lvm to our initramfs Dec 29 07:15:02 But that's mainly due to LuneOS specific things, might be different for hybris-boot Dec 29 07:15:21 So your halium part is done right? Dec 29 07:21:33 Satyam2345: Mainly yes, however not the rootfs part for Halium Dec 29 07:21:41 Just the bare minimal Android build Dec 29 09:21:42 i am at fixing mount points stuff, Dec 29 09:22:08 if codename has to start with 4 spaces Dec 29 09:22:51 then what about sed ? tab or spaces? Dec 29 09:36:18 @Optimus9650, Just check how is it aligned in previous device mountpoint Dec 29 09:44:09 Justin Verel was added by: Justin Verel Dec 29 09:49:40 @shenoy13, and what to do if there are two different entry for the same thing Dec 29 09:50:58 @shenoy13, https://paste.ee/p/TCTzh like this Dec 29 10:10:10 @Optimus9650, Only one entry is enough Dec 29 10:10:23 cool .. done.. Dec 29 10:15:11 running mka mkbootimg , failed Dec 29 10:15:15 anyone Dec 29 10:15:17 ? Dec 29 10:17:54 What is the error? Dec 29 10:19:06 https://paste.ee/p/pFpcp Dec 29 10:19:19 @shenoy13, 👆 Dec 29 10:20:47 You are missing a dependency, something like libtinfo5 Dec 29 10:23:01 Alex was added by: Alex Dec 29 10:23:31 Is Halium ported to Zuk Z2? Dec 29 10:23:33 @JBBgameich, fixed... Dec 29 10:23:47 @Alex, i think not Dec 29 11:26:14 https://paste.ee/p/GyK35 .. can any one look at he errors at while making mkbootimg Dec 29 11:30:10 @Optimus9650, It shows make completed Dec 29 11:31:07 thanks Dec 29 11:31:13 @Optimus9650 do you have a twitter account Dec 29 11:31:23 @vanyasem, yeah i do Dec 29 11:31:38 pm me a link to it plz, i want to sub Dec 29 11:41:11 error while making hybris boot... Dec 29 11:41:12 https://paste.ee/p/k0pK2 Dec 29 11:57:34 So uh, would devices using the A/B partition layout like the Pixels and the Essential phone face any issues with Halium? Dec 29 11:59:00 https://source.android.com/devices/tech/ota/ab_updates Dec 29 13:52:04 About the apk signing problem: … I think we need to add another custom TARGET_BUILD_VARIANT (such as user, userdebug, eng) … Then we could disable apks/apps totally Dec 29 14:23:51 @Optimus9650, any one.. ?? Dec 29 14:24:46 @Optimus9650, Me Dec 29 14:25:06 @Cl1608Ho, no need to ask Dec 29 14:25:15 Check out the DTB tool and put it in the right directory . Let me check github for a minute and I'll give you Dec 29 14:25:20 Had the same error with a different repo Dec 29 14:28:13 The repo I find is empty. … Can you run `mka --trace hybris-boot`, it might give us another clue Dec 29 14:35:06 @Cl1608Ho, https://paste.ee/p/lDlLM error out Dec 29 14:37:03 Just in case, do you have `android-tools-mkbootimg` installed on your computer? Dec 29 14:37:14 (at least that's the package for Ubuntu) Dec 29 14:38:44 make mkbootimg Dec 29 14:39:06 *make -j 10 mkbootimg Dec 29 14:39:17 You might wanna speed stuff a little bit up ^^ Dec 29 14:39:56 `make mkbootimg -j$(nproc)` Dec 29 14:40:12 You might wanna use the correct number of threads :) Dec 29 14:40:17 and then hybris-boot .. with make? Dec 29 14:40:29 Whatever ^ Dec 29 14:41:19 "mka bootimg" was successful Dec 29 14:41:25 last time Dec 29 14:41:48 do i need to do that again .. with make? Dec 29 14:44:04 @UniversalSuperBox, make hybris-boot spit out same error Dec 29 14:44:26 @Cl1608Ho, ^ Dec 29 14:45:05 i am on arch.. 🙃 Dec 29 14:45:20 Then whatever is continuing the program Dec 29 14:46:32 I'm talking about these tools: https://github.com/xiaolu/mkbootimg_tools Dec 29 15:11:51 in a huge nutshell ... what do i need to do now? Dec 29 15:34:40 nvm figuret it out... Dec 29 15:34:59 What was it? Dec 29 15:35:06 dtb tools Dec 29 15:35:14 ;) Dec 29 15:35:38 all i needed to do is just read the error Dec 29 15:36:12 Yeah but I mean I got the same error with a different package but that was a missing project in the manifest Dec 29 15:38:32 script was looking for dtb tools.. which was missing in the /out/host/..../bin folder ... so cloned that repo which you provided in that same folder Dec 29 15:41:40 @Cl1608Ho, now it needs some java files in the framework folder .. "signapk.jar" Dec 29 15:42:18 Now delete the part in vendor/ that's requiring it Dec 29 15:44:11 @JBBgameich, Look at this. Dec 29 15:44:24 @JBBgameich, And this Dec 29 15:50:43 (Photo, 865x547) https://irc.ubports.com/0a0jg7ED/file_3326.jpg this? Dec 29 15:50:59 @Optimus9650, what is your desktop enviroment Dec 29 15:51:02 Yep Dec 29 15:51:14 @vanyasem, deepin Dec 29 15:51:44 @Optimus9650 btw I meant take a look at those and the surrounding messages Dec 29 15:51:51 About the subject matter Dec 29 15:54:04 I've booted hybris-recovery, when I continue booting, the kernel panics (it reboots with boot reason = kernel panic) Dec 29 15:54:19 How can I get the log of the kernel panic? Dec 29 15:54:31 Boot into twrp immediately after the reboot Dec 29 15:54:41 And check the log file Dec 29 15:55:22 adb shell cat /proc/last_kmsg Dec 29 15:55:24 Yeah Dec 29 15:55:31 ok Dec 29 15:56:41 I'm getting`cat: can't open '/proc/last_kmsg': No such file or directory` Dec 29 15:56:56 Hm Dec 29 15:57:07 Anyone of those who wrote the docs here? :P Dec 29 15:57:32 @JBBgameich said last_kmsg is deprecated somehow Dec 29 15:57:41 @Cl1608Ho, make file is too big.. what do i need to grep? Dec 29 15:57:50 Jar? Dec 29 15:57:54 Java? Dec 29 15:57:55 Apk? Dec 29 15:57:59 Idk Dec 29 15:58:12 Prolly the thing that the log message tells you it can't build Dec 29 15:58:18 @Ellenjott, it doesn't exist on newer kernels Dec 29 15:58:26 (afaik) Dec 29 15:58:42 @JBBgameich, so probably including 3.10 Dec 29 15:59:05 @JBBgameich, We should update the docs Dec 29 15:59:07 It's in ramoops, maybe Dec 29 16:00:01 (Photo, 541x681) https://irc.ubports.com/pmP1euFh/file_3328.jpg Dec 29 16:01:13 @Optimus9650, Do you've got a msmXYZ-common folder or so? In my case it wasn't in my devices folder Dec 29 16:01:41 cat: can't open '/sys/fs/pstore/console-ramoops': No such file or directory Dec 29 16:01:52 Guess you don't have that then Dec 29 16:01:59 :( Dec 29 16:02:16 Are there any consoles in that folder? Dec 29 16:02:42 There's only /sys/fs/pstore/annotate-ramoops Dec 29 16:02:46 @Ellenjott, yep .. in the cm folder Dec 29 16:02:49 Hmm Dec 29 16:02:56 So it's not panicking, at least? Dec 29 16:03:10 It said it was panicking Dec 29 16:04:53 ```# cat /proc/bootinfo … [...] … Last boot reason: reboot``` Dec 29 16:05:18 ok now its not .. Dec 29 16:05:50 while making system image.. frameworks error.. now it it telling me to remove ims_blablabla from the vender make file... but ims is needed for volte.. if i remove that will the volte work properly? Dec 29 16:06:25 *is Dec 29 16:07:10 Now I continued from telnet, it reboots (and I get telnet again), and now it says … `Last boot reason: kernel_panic` Dec 29 16:07:46 @Optimus9650, @Cl1608Ho Dec 29 16:07:48 @Optimus9650, I think you can safely remove the ims apk/jar parts Dec 29 16:08:02 @Ellenjott, Can you build and deploy hybris-recovery? Dec 29 16:08:16 @UniversalSuperBox, I'm already in hybris-recovery Dec 29 16:08:26 So that happens after the continue? Dec 29 16:08:29 because hybris-boot gave a boot loop Dec 29 16:08:33 @UniversalSuperBox, yeah Dec 29 16:08:38 Okay, `ls /target` Dec 29 16:09:00 (Photo, 226x56) https://irc.ubports.com/nf23PenT/file_3330.jpg Dec 29 16:09:09 Your image has become corrupted Dec 29 16:09:15 Redeploy with rootstock Dec 29 16:09:20 Ok Dec 29 16:09:58 @Ellenjott, can i just comment stuff from the makefile? Dec 29 16:10:04 @Optimus9650, YEp Dec 29 16:15:09 @Optimus9650, Halium probably doesn't support volte Dec 29 16:15:43 ohh.. will it be added soon? Dec 29 16:15:54 Will what be added soon? Dec 29 16:17:31 i think shenoy texted... "halium probably dont support volte" Dec 29 16:17:39 If it's a JAR or APK, you don't need it Dec 29 16:18:09 Anything they do will be controlled by ofono or similar in the operating system that uses Halium Dec 29 16:18:54 Just like you don't need SurfaceFlinger. The operating system provides a display manager with Libhybris support. Dec 29 16:19:29 @Optimus9650, volte == Voice over LTE ?? Dec 29 16:19:38 @jmnemonix, yes Dec 29 16:19:39 Ysah Dec 29 16:20:04 it has spit out a new error ..https://paste.ee/p/Gtl5S Dec 29 16:20:42 ```. build/envsetup.sh … [breakfast/lunch]``` Dec 29 16:21:05 Something in the environment is likely outdated Dec 29 16:22:39 hmmm ??🙄🤔 Dec 29 16:22:48 Run lunch again. Dec 29 16:23:29 @Optimus9650, Yes actually sailfish doesn't support volte. So I thought even halium might not support volte🤔 Dec 29 16:23:43 Again, that'd be ofono's job Dec 29 16:23:49 So if ofono doesn't support it, it isn't supported Dec 29 16:25:02 @UniversalSuperBox, Means it is up to the rootfs and not halium? (supporting volte) Dec 29 16:25:08 Yes Dec 29 16:25:23 @UniversalSuperBox, Ok Dec 29 16:26:11 @UniversalSuperBox, isnt lunch for 5.1 ? Dec 29 16:26:19 Use whatever you use Dec 29 16:26:25 Breakfast, lunch, whatever Dec 29 16:26:31 Just rebuild the environment variables Dec 29 16:26:44 did .. breakfast device Dec 29 16:27:37 @UniversalSuperBox, didnt work Dec 29 16:29:45 Yay I'm into ssh and lxc is already running :) Dec 29 16:30:54 and i am stuck at making systemimage 😵 Dec 29 16:31:26 @Optimus9650, I also had the problem with the APKs Dec 29 16:31:59 @Ellenjott, apk problemms are solved.. its a new error.. 😥 Dec 29 16:32:12 @Optimus9650, this Dec 29 16:32:42 I had to do this to fix the libcryptfs_hw error: … https://github.com/LNJ2/android_device_motorola_msm8916-common/commit/fd9b7aec97f1949566be969217c8bebe34350e9b Dec 29 16:33:29 Actually it's just not finding it, its in device/qcom/common or so. But I just disabled it for now Dec 29 16:35:30 Or just add this "TARGET_CRYPTFS_HW_PATH := device/qcom/common/cryptfs_hw" Dec 29 16:35:40 In board configuration.mk Dec 29 16:46:38 I don't think we use vold anyway, so no need to worry about their encryption Dec 29 16:46:46 @bhushanshah can confirm that Dec 29 16:50:00 https://paste.kde.org/p9d9ykby3 … Now the container is failing, though Dec 29 16:50:50 @shenoy13, this path doesnt exist .. so adding that in /device/blablabla/device/boardconfig.mk would be useless Dec 29 16:50:56 @Ellenjott, PM rootfs? Dec 29 16:51:07 @UniversalSuperBox, reference rootfs Dec 29 16:51:24 failed to create pty Dec 29 16:51:30 dat's strange. Dec 29 16:51:33 What does lxc-checkconfig say? Dec 29 16:51:36 Just to be sure Dec 29 16:52:10 @JBBgameich, many settings are missing :( Dec 29 16:52:23 Most things do not matter Dec 29 16:52:28 https://paste.kde.org/poihnnmxg Dec 29 16:52:49 Multiple /dev/pts instances: missing Dec 29 16:52:57 That could be the problem Dec 29 16:53:11 But which config is it? Dec 29 16:53:46 CONFIG_DEVPTS_MULTIPLE_INSTANCES Dec 29 16:53:48 `CONFIG_DEVPTS_MULTIPLE_INSTANCES=y` Dec 29 16:54:17 thx Dec 29 17:14:02 com.dsi.ant.antradio_library.xml error ... anyone? Dec 29 17:14:14 @Optimus9650, comment out Dec 29 17:14:24 antradio is not rquired Dec 29 17:14:25 from where? Dec 29 17:14:43 BoardConfig and msmxxxx.mk Dec 29 17:14:48 Check both Dec 29 17:16:09 not there in any of them Dec 29 17:16:28 @Optimus9650, Give me your tree link Dec 29 17:16:41 My lxc@android doesn't start, it isn't able to create some cgroup directory. This feels like basic stuff, so I don't understand why this doesn't work. Do you have any pointers for me? Output of `journalctl -u lxc@android` https://paste.ubuntu.com/26279404/ Dec 29 17:16:48 It should be in permissions head BTW Dec 29 17:17:56 https://github.com/theimpulson/android_device_asus_msm8916-common/blob/8186b70cf513e12dd1e04a9f6b8d8b6506d8cd6e/msm8916.mk#L50 Dec 29 17:18:04 See, the same is on mine Dec 29 17:18:56 that message just dissappeared Dec 29 17:19:31 LOL Dec 29 17:20:03 (Photo, 800x600) https://irc.ubports.com/filSP2Sk/file_3334.jpg Dec 29 17:20:13 happening to me for the secont time Dec 29 17:20:30 @caspermeijn[m], Maybe` CONFIG_NET_CLS_CGROUP=y` is missing Dec 29 17:20:45 @Optimus9650, ? Dec 29 17:20:53 What should I see? Dec 29 17:21:16 Oh! I deleted it Dec 29 17:21:29 😛 Dec 29 17:21:30 😂 Wrong message Dec 29 17:21:41 @JBBgameich, Check this. Dec 29 17:22:01 and permissions head too on msmxxxx.mk Dec 29 17:22:50 @JBBgameich CONFIG_NET_CLS_CGROUP is set in both my defconfig as the genereate kernel config Dec 29 17:24:28 @TheImpulson, https://github.com/LineageOS/android_device_wingtech_wt88047 here is the device tree Dec 29 17:25:00 @Optimus9650, Kernel Dec 29 17:25:17 https://github.com/LineageOS/android_kernel_wingtech_msm8916 Dec 29 17:27:40 You have a common tree too? Dec 29 17:27:52 yes Dec 29 17:27:58 Give me link Dec 29 17:28:15 and BTW add this in your defconfig CONFIG_NET_CLS_CGROUP=y. You are missing it. Dec 29 17:29:45 @Optimus9650, Yes if that is the case. Dec 29 17:31:01 @TheImpulson, mer check utility passed it Dec 29 17:31:35 @Optimus9650, Then, it is in your device tree. Did you checked common tree too for that line? Dec 29 17:33:48 @TheImpulson, confused 😐😶 Dec 29 17:34:05 @Optimus9650, Common tree link please Dec 29 17:34:25 @vanyasem you here? Dec 29 17:34:59 yes Dec 29 17:35:13 @Optimus9650, . Dec 29 17:35:18 Help him Dec 29 17:35:23 I am sorry but I won't help with that Dec 29 17:35:29 have a nice day Dec 29 17:36:37 @Optimus9650, https://github.com/theimpulson/android_device_asus_msm8916-common/blob/8186b70cf513e12dd1e04a9f6b8d8b6506d8cd6e/msm8916.mk#L47 Dec 29 17:36:59 This is an example of my common tree. Search same in yours. Dec 29 17:42:18 @TheImpulson, there is no common tree present.... i am still learning the terminologies Dec 29 18:18:15 BlueKenny was added by: BlueKenny Dec 29 18:21:10 (Photo, 1280x330) https://irc.ubports.com/P8sSGPgu/file_3336.jpg Can somehow explain me where to find dtbToolCM? Dec 29 18:21:28 @BlueKenny, Use the search function Dec 29 18:22:01 @Cl1608Ho, Both Google and telegram. … But here you go Dec 29 18:22:46 @Optimus9650, https://github.com/CyanogenMod/android_device_cyanogen_msm8916-common/blob/cm-14.1/product/ant.mk here Dec 29 18:22:48 Besides, cause you're on Ubuntu, you get it by installing `android-tools-mkbootimg` Dec 29 18:23:21 @shenoy13, Removing this ant.mk should do Dec 29 18:23:57 @shenoy13, already fixed it.. but thanks anyway ☺️ Dec 29 18:24:07 @Optimus9650, Ok Dec 29 18:53:31 General question, do warnings mmatter while making images? Dec 29 18:57:46 Most warnings won't break your build Dec 29 18:59:13 @JBBgameich I figured out how to make it work, I had to disable these options for Linux 3.4: CONFIG_NET_CLS_CGROUP=n Dec 29 18:59:14 CONFIG_NETPRIO_CGROUP=n Dec 29 19:42:12 cleaning up … umount: can't umount /cache/system/: No such file or directory .. while flashing rootfs Dec 29 19:42:32 Any errors before that? Dec 29 19:42:57 @UniversalSuperBox, nope Dec 29 19:43:43 wait.. new error Dec 29 19:45:20 @UniversalSuperBox, https://paste.ee/p/H96Yg .. while .. adnroid tools are installed .. it is still giving me this error Dec 29 19:45:34 *android Dec 29 19:45:43 Are both of those specific packages installed Dec 29 19:46:35 yes.. since .. i was able to use them before.. also adb detects my phone.. server is up and running perfectly Dec 29 19:46:47 `make_ext4fs` in terminal? Dec 29 19:48:05 nothing found Dec 29 19:48:13 Distro? Dec 29 19:48:18 arch Dec 29 19:48:26 You need to install make_ext4fs Dec 29 19:48:35 I don't know what package it comes in Dec 29 19:48:37 But you need it Dec 29 19:49:30 @Lyokanthrope, Guess noone had an answer for this lol Dec 29 19:49:46 No one has owned a phone that expensive for porting Dec 29 19:50:07 @UniversalSuperBox, 😂 Dec 29 19:54:43 Woo I can be the first Dec 29 20:08:14 does anyone know similar package for android-tools-fsutils is present in pacman or aur Dec 29 20:26:00 @Optimus9650, nvm done.. downloaded the deb package, extracted .. put files to their respective folder .. done Dec 29 20:38:35 cat: can't open '/proc/last_kmsg': No such file or directory Dec 29 20:38:50 I guess you don't have one, then Dec 29 20:39:11 yes.. and idk. why Dec 29 20:39:21 Because the kernel doesn't have that built in Dec 29 20:39:22 @Optimus9650, try /sys/fs/pstore/console-ramoops Dec 29 20:39:27 What's actually going wrong? Dec 29 20:40:48 @Lyokanthrope, pstore directory is also empty Dec 29 20:41:03 But what is the phone doing? Dec 29 20:41:14 in the twrp mode Dec 29 20:41:26 But what is the phone doing that's wrong? Dec 29 20:42:17 stuck at boot logo Dec 29 20:42:31 *splash screen Dec 29 20:42:32 What comes up in `dmesg` while it's booting? Dec 29 20:43:34 @UniversalSuperBox, exactly nothing.. it is not moving forward from the boot logo Dec 29 20:43:41 *spalsh screen Dec 29 20:43:45 In the dmesg on your computer Dec 29 20:43:50 Which the phone is plugged into via usb Dec 29 20:47:11 too much logs Dec 29 20:47:12 Hi Dec 29 20:47:15 mer kernel check is a verification for halium? Dec 29 20:47:32 @Optimus9650, Just the end. Just the parts where it reports the devices that it's detecting via USB Dec 29 20:47:43 yes Dec 29 20:47:53 Okay... what does it say Dec 29 20:48:12 just the last line android_work: android_work: sent uevent USB_STATE=CONFIGURED Dec 29 20:48:21 Not on the device Dec 29 20:48:29 Your computer Dec 29 20:48:31 When you boot the phone Dec 29 20:49:34 @sealofnova, Yes, it tells you which kernel configs need to be in place Dec 29 20:49:47 @UniversalSuperBox, Only for halium? Dec 29 20:50:03 Yeah. There are a few different ones for ubuntu touch yet. Dec 29 20:50:47 But Halium can run both the reference rootfs and Plasma Mobile Dec 29 20:50:56 Halium with Ubuntu Touch modifications can run Ubuntu Touch Dec 29 20:51:11 Is there a special mer kernel check for android? is it necessary for building android? Dec 29 20:51:27 What? Dec 29 20:52:25 for some weird reason it is detected as ethernet cable by my pc Dec 29 20:52:33 That's correct Dec 29 20:52:41 Im trying compile viper os. Do I need to use mer kernel check for this? Dec 29 20:52:43 Have you read the "logging in" section of the docs? Dec 29 20:52:49 I don't know what viper os is. Dec 29 20:53:03 @UniversalSuperBox, Its aosp based. Dec 29 20:53:13 We are not an Android support group Dec 29 20:53:18 @sealofnova, it is not for that.. Dec 29 20:54:29 @UniversalSuperBox what do i need to d now? Dec 29 20:54:38 You read the documentation Dec 29 20:54:47 http://docs.halium.org/en/latest/porting/debug-build/index.html Dec 29 20:55:06 Yeah i know but I used it to build halium. I am confused. Sorry Dec 29 20:55:54 Well, if you're using their tree to build Halium then yes, you use the mer tools to verify your kernel Dec 29 20:56:33 Ty Dec 29 20:56:49 @UniversalSuperBox, But that will most likely not work in the end, because Halium uses patched versions of many android components which are not available in your tree most likely. Dec 29 20:57:10 Is the device tree that different so you can't use it with the halium-7.1 tree? Dec 29 21:12:43 now device is getting detected as new neworking interface...but pc is failing to connect, [not mentioned in docs] Dec 29 21:13:39 what does `dmesg | tail` (on the host) say when you connect the device? Dec 29 21:14:55 @JBBgameich, https://paste.ee/p/7nlfk Dec 29 21:16:18 Maybe the important thing is not in this part of the dmesg, but if you see GNU / Linux, your device already booted into SSH mode Dec 29 21:18:28 You can try this … `https://raw.githubusercontent.com/JBBgameich/halium-install/master/connect-telnet.sh` … or this … `https://raw.githubusercontent.com/JBBgameich/halium-install/master/connect-ssh.sh` … script to connect to the device, depending on in which state the device is Dec 29 22:01:48 @mariogrip you around? Dec 29 22:47:08 Hi, I'm trying to get basic software rendering running on my suzuran. I have a /dev/fb0 which works f.e. with mplayer, but applications with Xorg/fbdev only show up *after* I close them... Dec 29 22:47:19 Got this line from dmesg Dec 29 22:47:21 mdss_mdp_display_commit: ctl 0 error displaying frame Dec 29 22:47:33 and this line from Xorg log: Dec 29 22:47:39 FBIOPUTCMAP: Invalid argument Dec 29 22:49:19 I have no idea how Wayland works, is it worth a try and do they have smth like xeyes (so I don't have to download gnome-* to test)? Dec 29 23:50:48 User was added by: User Dec 30 00:40:01 @User, welcome Dec 30 01:38:19 SyntaxError820 was added by: SyntaxError820 Dec 30 02:55:54 @SyntaxError820, welcome **** ENDING LOGGING AT Sat Dec 30 03:00:01 2017 **** BEGIN LOGGING AT Sat Dec 30 03:00:02 2017 Dec 30 05:34:14 what is the root password for connecting to the device using ssh Dec 30 05:34:47 @Optimus9650, The same you setup Dec 30 05:35:20 when did i setup? Dec 30 05:35:30 😶 Dec 30 05:36:11 As I remember, you need to setup password first. Dec 30 05:40:23 @vanyasem, @Optimus9650 Dec 30 05:46:34 @TheImpulson, chroot: can't execute '/system/bin/sh': No such file or directory Dec 30 05:50:04 cant chroot Dec 30 06:06:31 @bhushanshah ? Dec 30 06:15:46 @vanyasem Dec 30 06:19:59 Patience? Dec 30 06:21:16 cant wait .. to hands on ub touch Dec 30 06:33:37 Anyway, add /bin/bash in chroot Dec 30 06:52:55 mount: /data/recovery/system: failed to setup loop device for /data/rootfs.img Dec 30 06:59:35 nvm figured it out Dec 30 07:15:35 while making hybris-recovery.img Dec 30 07:15:39 https://paste.ee/p/gkswu Dec 30 08:24:01 @Optimus9650, anyone? Dec 30 09:28:32 > cant wait .. to hands on ub touch Dec 30 09:28:32 Haha Dec 30 09:29:01 You forgot to lunch, boi Dec 30 09:50:06 @i_love_halium_on, My diet plan is different Dec 30 11:08:58 Mr_Doge was added by: Mr_Doge Dec 30 11:41:05 Hello everyone Dec 30 13:39:24 Can someone explain me if lunch command Dec 30 13:39:34 *the Dec 30 13:41:29 what seems to be the issue Dec 30 13:42:27 If it's needed if we are building for a 14.1 device Dec 30 13:42:49 if breakfast works, use breakfast Dec 30 13:43:10 @vanyasem, Thanks Dec 30 15:22:15 Hello guys how are you going to Dec 30 15:30:44 @Buğra Uygur, Lemme guess , not a native English speaker?... Btw hello Dec 30 16:00:11 By far everything is going such smooth, i doubt if i am doing right Dec 30 16:00:18 😂 Dec 30 16:51:18 while debugging ping -c 2 192.168.2.15 returns failure Dec 30 16:51:26 telnet Dec 30 17:12:30 [abhishek@OPTIMUS ~]$ ip address show dev enp0s20f0u3 … 13: enp0s20f0u3: mtu 1500 qdisc fq_codel state UNKNOWN group default qlen 1000 … link/ether c2:aa:54:8f:2e:61 brd ff:ff:ff:ff:ff:ff … inet6 fe80::a2d2:97f0:a9fe:86a7/64 scope link noprefixroute … valid_lft forever preferred_lft forever Dec 30 17:33:35 anyone? Dec 30 17:34:09 You're probably out of the init, then Dec 30 17:34:15 Go to the next page Dec 30 17:35:01 i was unable to connec to telnet .. was that alright? Dec 30 17:35:26 What appears in `dmesg` when you plug in the phone while it boots Dec 30 17:36:07 rndis_host 1-3:1.0 enp0s20f0u3: renamed from usb0 … [43905.016193] IPv6: ADDRCONF(NETDEV_UP): enp0s20f0u3: link is not ready Dec 30 17:36:15 dmesg | tail Dec 30 17:36:43 ```[ 2886.227852] usb 1-6: Manufacturer: Mer Boat Loader … [ 2886.227853] usb 1-6: SerialNumber: Mer Debug setting up (DONE_SWITCH=no)``` Dec 30 17:36:46 There's somthing like this Dec 30 17:37:36 @UniversalSuperBox, no Dec 30 17:37:48 It doesn't exactly say that Dec 30 17:37:51 @Mr_Doge is there any reason you are using two different accounts? Dec 30 17:37:54 It say the manufacturer and serial number Dec 30 17:38:08 @bhushanshah, yes.. Dec 30 17:38:19 Mind explaining? Dec 30 17:39:29 @bhushanshah, on device .. its the previous account.. Dec 30 17:39:42 Okay Dec 30 17:39:43 @bhushanshah, this one is on pc Dec 30 17:41:43 Alright, @bhushanshah I'm about to ask what's probably a stupid question. How do I get e2fsck into busybox Dec 30 17:41:51 I am searching for this question as well Dec 30 17:43:15 You enable that somewhere in it's defconfig Dec 30 17:43:25 Must not be in menuconfig Dec 30 17:43:55 @UniversalSuperBox isn't it 'fsck' ? Dec 30 17:44:06 not e2fsck, must do the same thing though. Dec 30 17:44:09 But I don't have `fsck.auto` or `fsck.ext4` so I'm missing something else Dec 30 17:44:34 I'm not entirely sure how to get it though Dec 30 17:44:43 `google` Dec 30 17:44:49 It's faling me Dec 30 17:45:24 @UniversalSuperBox, no nothing like that Dec 30 17:45:38 Is your phone even plugged in Dec 30 17:45:48 yes Dec 30 17:46:51 How about this? http://docs.halium.org/en/latest/porting/debug-build/early-init.html#determining-if-this-is-needed Dec 30 17:49:04 @UniversalSuperBox, script notices no changes Dec 30 17:50:53 and thats weird Dec 30 17:56:55 @UniversalSuperBox, got it.. it needed hybris-recovery.img .. Dec 30 18:00:00 Update: Busybox doesn't have e2fsprogs. Or support for ext4. Dec 30 18:00:18 It will happily call them if you provide them, though. :P Dec 30 18:49:18 @UniversalSuperBox i got into telnet.. added nameserver from telnet console... still cant ping debian.org Dec 30 18:49:37 No, because you don't have an internet connection? Dec 30 18:49:40 What are you trying to do? Dec 30 18:49:51 you need to set up internet sharing to the telnet interface Dec 30 18:50:02 i remember doing it once, over google Dec 30 18:50:12 (but it should be easier to setup wlan with halium rootfs lol) Dec 30 18:50:15 Right, but why? Dec 30 18:50:29 Why forward to telnet, that is Dec 30 18:50:30 @UniversalSuperBox ... because. Dec 30 18:50:49 how else am I gonna scp files over to the rootfs? :p Dec 30 18:51:08 jk, i used it to upload some logs over to hastebin.com Dec 30 18:51:24 ( or was it ix.io ) Dec 30 18:51:38 eitherway, internet on the rootfs is something that you must have at all costs. /s Dec 30 18:51:48 http://docs.halium.org/en/latest/porting/debug-build/usbnet.html .. i am at the last part of this Dec 30 18:51:56 That's not from telnet Dec 30 18:52:10 But... Why you want to setup internet? Dec 30 18:52:34 If you're in hybris-recovery, it's because you think that your phone isn't booting, right? Dec 30 18:52:50 You need to get out of hybris-recovery into the system first, then get internet access. Dec 30 18:52:56 https://www.youtube.com/watch?v=9nazm3_OXac Dec 30 18:58:35 Guys. Where is sepolicy in kernel? Dec 30 18:59:02 My boot ramdisk sepolicy file wrong... Dec 30 19:04:39 My kernel source need edit. Dec 30 19:07:04 what do I have to add to get ubports recovery into a halium 7.1 build tree Dec 30 19:12:07 We aren't there yet Dec 30 19:14:44 I found out why fsck is disabled in our android busybox Dec 30 19:14:52 Because it doesn't build. -_- Dec 30 19:20:04 \o/ Dec 30 19:20:26 umm stupid question.. how do i get a working system Dec 30 19:20:46 With the PRO SKILLLLZZZZ Dec 30 19:21:48 @i_love_halium_on, why some nubz skillz wont work? Dec 30 19:22:17 Because they aren't PRO SKILZZZZ Dec 30 19:22:29 And halium needs android dev SKILLZZ Dec 30 19:22:52 ESPEcially if you have textrel libs Dec 30 19:23:00 where can i buy some pro skillz .. are they on amazon ? :P Dec 30 19:23:26 Googleh give you SKILLZZ Dec 30 19:23:33 READ&LEARN Dec 30 19:25:03 @i_love_halium_on, wanna trade some security skillzz for andro dev skillz Dec 30 19:25:23 Wat? Dec 30 19:26:04 * i_love_halium_on melp Dec 30 19:26:18 nvm :P Dec 30 19:26:23 Kk Dec 30 19:28:16 after installing the boot .img from fastboot and root fs from the script .. where should device boot into.. i mean what should appear on screen Dec 30 19:28:48 Nothing Dec 30 19:28:51 Just nothing Dec 30 19:29:05 Halium is CLI not GUI Dec 30 19:29:26 a blank screen or just boot logo? Dec 30 19:29:34 Nothing Dec 30 19:29:36 splash screen Dec 30 19:29:45 Just nothing Dec 30 19:29:49 @i_love_halium_on, +1 Dec 30 19:30:20 blank screen .. right? Dec 30 19:30:56 If you want to ping me it's @i_love_halium_on_my_nexus write it all, bois! Dec 30 19:31:04 > blank screen .. right? Dec 30 19:31:05 Nothing Dec 30 19:31:20 Just nothing Dec 30 19:46:17 @Mr_Doge, Have you flashed hybris-boot.img or boot.img? Dec 30 19:46:38 @JBBgameich, yes Dec 30 19:51:30 @Mr_Doge, depends Dec 30 19:51:36 I have an HTC logo on my phone Dec 30 19:51:42 with a fully working Halium build Dec 30 19:51:59 it turns black after I run test_hwcomposer for the first time Dec 30 19:52:37 so boot/splash logo is also fine. don't get too concerned about it, just check desktop's dmesg (or ip link) Dec 30 19:53:35 spends 4 hours trying to figure out how he's going to get e2fsprogs into the tree Dec 30 19:53:37 it is Dec 30 19:54:27 and after those 4 hours you will solve it? Dec 30 19:54:44 Yeah. :) Dec 30 19:54:48 It was always there Dec 30 19:54:58 ooooh Dec 30 19:55:00 But now I know that there's a binary that's going to be statically linked Dec 30 19:55:39 So HOPEFULLY we'll be able to do e2fsck from hybris-boot Dec 30 19:55:41 hm thats the reason why we had static busybox binaries? Dec 30 19:55:43 That was the goal Dec 30 19:55:52 Static busybox stays. I just wanted e2fsck Dec 30 19:56:17 ```~ # e2fsck … Usage: e2fsck``` Dec 30 19:56:18 YEAH! Dec 30 19:57:00 @vanyasem, from which console it is needed to be run? Dec 30 19:57:15 You SSH into the phone from your computer Dec 30 19:57:24 @Mr_Doge, read the docs Dec 30 19:57:25 plz Dec 30 19:57:29 http://docs.halium.org/en/latest/porting/debug-build/logging-in.html Dec 30 19:57:38 @vanyasem, telnet/ssh section Dec 30 19:57:45 Not telnet Dec 30 19:59:39 Well, this change means that the boot.img is larger than 8MB, which might cause issues with very old devices Dec 30 19:59:43 ¯\_(ツ)_/¯ Dec 30 20:01:49 I currently update the apparmor in the sony kernel... maybe that helps Dec 30 20:02:05 (Photo, 752x186) https://irc.ubports.com/Ht0ur4k7/file_3384.jpg just couldnt get pass this thing Dec 30 20:11:16 @vanyasem, while ssh-ing ssh: connect to host 10.15.19.82 port 22: No route to host Dec 30 20:11:49 You haven't set the IP on the new network interface Dec 30 20:12:12 "You should see that you have a new network interface again. Assign it an IP of 10.15.19.100" Dec 30 20:12:24 In the screenshot you posted Dec 30 20:12:49 script.. i am using script.. Dec 30 20:12:59 There is no script Dec 30 20:13:18 there is.. someone provided me yesterday Dec 30 20:13:19 @UniversalSuperBox, there is by jbb's Dec 30 20:13:28 don't lie Dalton 😂 Dec 30 20:13:33 Sure, that's to install the rootfs on the system Dec 30 20:13:38 no Dec 30 20:13:46 (Document) https://irc.ubports.com/UPrNq835/file_3386.sh Dec 30 20:13:52 @vanyasem, he has telnet connect and ssh connect in the same repo Dec 30 20:14:05 Okay, that's cool Dec 30 20:14:55 script sets the ip.. still it couldnt ssh Dec 30 20:22:30 because? Dec 30 20:22:38 @Mr_Doge, here Dec 30 20:22:50 ah sorry Dec 30 20:36:23 This is why we do fsck in the initrd, folks Dec 30 20:56:10 This is why we cannot have any nice things! Dec 30 21:55:59 @mr_doge dead meme Dec 30 21:56:02 Rip Dec 30 23:38:53 Henrik was added by: Henrik Dec 31 01:04:45 Non ci credo Dec 31 01:04:51 \o/ Dec 31 01:09:16 Oddio siii Dec 31 01:09:25 Halium installato Dec 31 01:09:32 Sono felicissimoo Dec 31 01:33:05 Why you banned him ? Lol Dec 31 01:33:24 He was happy cuz he installed halium haha Dec 31 01:33:25 Melp Dec 31 01:33:54 (And yes, im spying with a tg account) **** ENDING LOGGING AT Sun Dec 31 03:00:02 2017 **** BEGIN LOGGING AT Sun Dec 31 03:00:02 2017 Dec 31 06:36:13 @i_love_halium_on, much wow, such alive Dec 31 06:36:20 @i_love_halium_on, who? Dec 31 07:02:41 root@localhost:~# test_hwcomposter … -bash: test_hwcomposter: command not found Dec 31 07:08:13 @Mr_Doge, Not composter Dec 31 07:08:18 Composer Dec 31 07:13:53 @shenoy13, so after installing the system, getting the ssh console .. what is the next step? Dec 31 07:14:04 *system.img Dec 31 07:19:20 @vanyasem Dec 31 07:20:08 @Mr_Doge, Check dmesg and systemctl to check for errors Dec 31 07:22:05 @Mr_Doge, Don't needlessly ping people, if you need help keep patience.. this is warning, if I ever see needlessly pinging people without any need I'll ban you Dec 31 07:22:36 @Mr_Doge, Next step is fixing container Dec 31 07:22:57 Check systemctl status lxc@android Dec 31 07:30:14 Hello. Dec 31 07:30:50 After making halium system and boot. I'm maked rootfs. Now flashed Dec 31 07:31:03 But telnet says : connection refused Dec 31 07:31:29 In rootfs telnet won't work but ssh Dec 31 07:32:05 I'm tried with ssh. Still same Dec 31 07:32:09 http://docs.halium.org/en/latest/porting/debug-build/logging-in.html Dec 31 07:32:21 You need to give it different ip Dec 31 07:33:21 @Mr_Doge, @developerfect use this script .. to connect to ssh.. Dec 31 07:38:41 @Mr_Doge, Thanks. I'm trying now. Dec 31 09:14:58 systemctl status lxc@android Dec 31 09:15:01 https://paste.ubuntu.com/26291941/ Dec 31 09:15:49 failed Dec 31 09:31:26 M3lp Dec 31 09:32:48 @i_love_halium_on, ? Dec 31 09:38:06 ? Dec 31 09:50:43 @i_love_halium_on, He did not understand your ironical comment ^^ Dec 31 09:54:56 what does it say? Dec 31 09:55:20 probably mean whelp in leetspeak ^^ Dec 31 09:57:19 guys I am also porting a bit today Dec 31 09:57:40 List of devices attached … Mer offline Dec 31 09:57:44 good sign? xd Dec 31 10:13:49 So I found the instructions for hybris-recovery. But it does not explain how to work with it Dec 31 10:15:22 Hybris-recovery is only a hybris-boot image which always boots in debug mode (telnet) Dec 31 10:15:34 Ok good lets try this Dec 31 10:16:24 @Mr_Doge, can anyone have a look at it.. Dec 31 10:16:38 @Mr_Doge, Sry I looked but no idea whats wrong Dec 31 10:17:06 @JBBgameich, Hmm or, as in my case just crashes ^^ Dec 31 10:17:35 lxc@android is failing, but dont know why :( Dec 31 10:19:36 @Mr_Doge, The system image is missing required files or can't be mounted correctly. And it seems like config_multiple_devpts_instances is missing in the kernel config Dec 31 10:20:31 @JBBgameich, thanks... Dec 31 10:23:41 @JBBgameich, Plus my last_kmsg is always overwritten by recovery boot, I cant gain anything from there meh. But smth is happening, it takes areound 30secs or more before it reboots. And the screen dims, so kernel seems to load Dec 31 10:27:16 Guys SSH still not allowed password Dec 31 10:27:26 Connection Refused Dec 31 10:27:46 That means that sshd is not running Dec 31 10:27:52 has nothing to do with password Dec 31 10:29:13 @developerfect, did you use the script which i provided? Dec 31 10:31:10 @Mr_Doge, Yes Dec 31 10:31:14 Still problemly Dec 31 10:31:35 @developerfect, what was the output after running the script? Dec 31 10:34:35 @Mr_Doge, I will try and report the result Dec 31 10:55:13 did you set the password ? Dec 31 10:55:21 for the user ? Dec 31 11:08:16 @Mr_Doge, config_multiple_devpts_instances adding this to kernel config didnt work Dec 31 11:31:48 The Android init system is still missing in your system image Dec 31 11:31:55 So it can't start Dec 31 11:42:51 @JBBgameich, will you guide me how to fix that? Dec 31 11:47:03 @Mr_Doge, Is /system mounted? Not partition but /data/system.img?