**** BEGIN LOGGING AT Sat Jun 27 02:59:57 2020 Jun 27 05:46:20 morning! Jun 27 05:53:30 mne says hello to his mentors Jun 27 05:58:10 Today i've pretty much free time. Planning to troubleshoot luneos kernel for land, if u guys are available :) Jun 27 06:11:57 BT40: I won't be much available this weekend unfortunately... I'm helping a friend move from his flat Jun 27 07:08:06 okay, np. awaiting If herrie can pull alone, then its fine. otherwise we'll plan next week. Jun 27 07:58:09 BT40 I am around Jun 27 07:58:18 Just having breakfast now Jun 27 08:15:13 Herrie, so how to proceed next? last kerned didn'd activated adb. so you are the guide for me now. Jun 27 08:16:29 since if kernel needs to be recompiled, can u please check whether calling option is activated in it or not? if not, we can enable that one also. Jun 27 08:19:44 BT40: Calls are to be fixed on LuneOS & Halium side Jun 27 08:19:55 Kernel for adb we can do seperately Jun 27 08:20:07 okay. Jun 27 08:26:18 BT40: Let me see where we were with the kernel Jun 27 08:39:20 OK I checked the flags on my mido where I know ADB works, let me apply the same flags for Land now Jun 27 08:43:34 OK building new kernel for you Jun 27 08:53:06 BT40: New kernel where ADB hopefully works: Image.gz-dtb--3.18.31+gitr5+7ee70c4f04-r0.2-land-0-0.fastboot on your dropbox Jun 27 08:53:40 ok. will test and report in an hour Jun 27 08:55:08 BT40: You can simply do: "fastboot boot Image.gz-dtb--3.18.31+gitr5+7ee70c4f04-r0.2-land-0-0.fastboot" to boot it or "fastboot flash boot Image.gz-dtb--3.18.31+gitr5+7ee70c4f04-r0.2-land-0-0.fastboot" to flash it to your boot partition Jun 27 08:56:55 sure Jun 27 08:57:10 its lunch time here :) Jun 27 08:58:03 Late lunch :P Jun 27 08:58:11 Or that's normal lunch time where you are? Jun 27 08:59:02 yes its normal. we generally take breakfast in morning 9am, lunch at 2pm, dinner at 8 or 9pm. :) Jun 27 08:59:20 now its 2.30pm in my country Jun 27 09:01:15 Ah OK. Here it's 7-8, 12-13 and around 18-19 usually ;) Jun 27 09:13:49 :) Jun 27 09:57:05 it is stuck on flastboot command, no progress, o failure Jun 27 09:57:09 < waiting for any device > Jun 27 09:57:35 when i run adb devices command, it shows connected device Jun 27 09:57:55 it seems i am in recovery mode, need to switch to fastboot Jun 27 09:58:03 Yes Jun 27 09:58:10 forgot after so long :) Jun 27 09:58:12 Hold the volume down + power Jun 27 09:58:26 Then you'll get the "Chinese Android doctor" image Jun 27 09:58:30 At least on my Mido ;) Jun 27 09:58:42 yes flashed Jun 27 10:02:07 stuck into bootloop Jun 27 10:02:39 Hmmz strange Jun 27 10:02:43 It booted OK before right? Jun 27 10:02:46 Just no ADB Jun 27 10:03:02 yeah, with previous kernel booted ok Jun 27 10:03:13 gitr3 was prev kernel Jun 27 10:03:23 like this one is gitr4 Jun 27 10:03:29 gitr5 sorry Jun 27 10:04:39 gitr4 is missing :) Jun 27 10:05:00 Let me see Jun 27 10:05:05 I might have that one somewhere still Jun 27 10:05:46 Found uploading now: Image.gz-dtb--3.18.31+gitr4+143a033636-r0.2-land-0-0.fastboot Jun 27 10:12:27 flashed. gitr4 booted fine Jun 27 10:12:31 i m in system Jun 27 10:13:07 adb not working though Jun 27 10:22:14 Hmmz weird Jun 27 10:22:35 Well I did change quite some things on my builder in between Jun 27 10:23:00 I should get it back to previous state and then build kernel again Jun 27 10:23:08 another thing Jun 27 10:23:25 devivice charging also stopped working after this update Jun 27 10:24:09 (even when switch off, charging not work ) so something is going on with usb Jun 27 10:26:21 Kernel and other things we flash shouldn't have influence when device is turned off Jun 27 10:26:36 It's either cable or port that is having issues Jun 27 10:26:48 I'd try some other cables first Jun 27 10:33:14 i also want to learn how u config kernel. can u please share? :) I'd love to experiment Jun 27 10:43:30 Well it's trial and error mainly Jun 27 10:43:34 is it like we try to hit and try based on experience plus online searches? These options CONFIG_DIAG_OVER_USB=y CONFIG_USB_FUNCTIONFS=n Jun 27 10:43:39 It's on my github Jun 27 10:44:22 ok browsing that Jun 27 10:44:53 https://github.com/Herrie82/android_kernel_xiaomi_msm8937/commits/halium-7.1 Jun 27 10:45:38 I basically update the various flags, point the linux-xiaomi-land.bb to the new SRCREV for the commit Jun 27 10:46:25 Then run "MACHINE=land bb virtual/kernel" which will make the new fastboot file Jun 27 10:47:09 android_kernel_xiaomi_msm8937/arch/arm64/configs/ Jun 27 10:47:19 There are various defconfigs here Jun 27 10:47:33 The one I changed Jun 27 10:47:34 do kernel select only one or merge all? Jun 27 10:48:10 it's specified which one it uses Jun 27 10:49:01 ok, so it mean only one of these is active for my device. Rest are just for other devices Jun 27 10:50:17 Yes Jun 27 10:50:23 You need: arch/arm64/configs/lineageos_land_defconfig Jun 27 10:51:29 in which file it is mentioned the path to this config file? i wanna do is, create a new fork config file from yours instead of messing that one, redirect kernel to that config. if anything goes wrong which is most probably to happen :), we can switch back to your main one. Jun 27 10:53:42 You can fork my repo, change whatever you want in your fork Jun 27 10:54:08 Just point you linux-xiaomi-land.bb to your repo instead of mine Jun 27 11:03:11 in meta-smartphones/recepies-kernel/linux-xiaomi-land_git.bb I hope this is the file u are refering to. ..... In this file, this line points to git source: SRC_URI = " \ Jun 27 11:03:11 git://github.com/kskarthik/android_kernel_xiaomi_msm8937.git;branch=halium-7.1 \ Right? Jun 27 11:04:00 do_configure_prepend() { Jun 27 11:04:00 cp -v -f ${S}/arch/arm64/configs/lineageos_land_defconfig ${WORKDIR}/defconfig Jun 27 11:04:00 } This line says which defconfig to use.....right? Jun 27 11:05:03 SRCREV = "aad1958e0b7ffd00ca2af58ad3f25578d241f233" This line states the commit number. Right? Jun 27 11:22:10 Yes Jun 27 11:22:31 I have it pointing to my herrie82 fork and latest commit for example Jun 27 11:26:07 So you basically just update SRC_URI and SRCREV that's all Jun 27 11:28:23 i have forked your one. But github has forked all branches which have nothing to do with luneos. So i wanna fork halium one only Jun 27 11:28:34 any work around? Jun 27 11:28:56 because when i sync it with local one, it consumes lot of disk and data Jun 27 11:29:11 so i just want your branch only Jun 27 11:29:21 with a particulat commit only Jun 27 11:32:42 On my github, i have manually deleted other branches from my forked one. I hope this is not illegal. Jun 27 11:33:24 also i am planning to your one as default instead of original. Is it okay? Jun 27 11:34:23 Github is giving this warning: changing default branch can have unintended consequences that can affect new pull requests and clones. Jun 27 11:43:48 No thats all fine Jun 27 11:44:06 Just delete ones you don't need Jun 27 11:44:27 The branches I Mean Jun 27 12:21:50 Herrie, one thing i wanna ask. We did halium part just for kernel image right? Cant we use default xiaomi kernel instead of halium one? Xiaomi has provided open source code and i have that one Jun 27 12:22:41 halium one might have influence of ubuntu touch which could be causing conflicts. Is it possible? Jun 27 13:46:58 BT40: unlikely Jun 27 13:47:05 This kernel is fine Jun 27 13:49:19 one more info i wanna give, if it is useful Jun 27 13:49:35 when i type command lsusb in terminal Jun 27 13:50:41 in recovery mode, where adb works, it detects fone as nexus .............. In luneos mode, where adb is not supported, nothing is listed. I mean, one device count is reduced Jun 27 13:51:55 and in fastboot mode, device is detected as google inc. Jun 27 13:52:13 BT40: This kernel originates from Xiaomi too. I doubt there are any UBPorts/UBTouch specific changes to the source code Jun 27 13:52:27 Only change might be some defconfig changes I would say Jun 27 13:52:55 Earlier i had doubt that it might be possible that ubuntu not detecting adb of luneos, now it is cleared that even that usb connection is not detected. so not an ubuntu issue Jun 27 13:52:59 UBPorts kernel patches aren't breaking LuneOS in generla as well Jun 27 13:53:41 Systems are quite similar, main difference is that UB Touch uses AppArmor which LuneOS doesn't use Jun 27 13:53:55 But that shouldn't influence adb really Jun 27 13:58:38 got one log which might be useful Herrie Jun 27 13:58:42 11:50:41 land init: Not starting service for keychord 1 because ADB is disabled Jun 27 13:59:08 it is from phone kernel log which we reviewed a few weeks ago Jun 27 14:08:15 It could be something in the android_device_xiaomi_land repo somehow Jun 27 14:24:02 Herrie, is this the main reason that adb is not working? or some other thing also needs to be discovered? Jun 27 14:24:19 Hard to say Jun 27 14:24:33 Keychord 1 doesn't ring a bell to me to be honest Jun 27 14:25:29 can u please share your journal file? i want to take a look how good working os file should look like Jun 27 14:44:33 Let me flash something on my mido Jun 27 14:44:39 Was trying to get Halium 9.0 working there Jun 27 14:44:44 Also having ADB issues :P Jun 27 15:06:41 Well mine isn't great because I flashed LinageOS 16 and now my LuneOS doesn't boot and log is quite poluted Jun 27 15:06:50 Let me see if I can restore Xiaomi Andorid 7 Jun 27 15:18:08 BT40: This is from Mido with LuneOS booting and ADB working: https://paste.ubuntu.com/p/wZ9W2d6PQY/ Jun 27 15:19:31 nowhere it is mentioned adb. strange Jun 27 15:19:57 according to documentation, adb daemon is must for adb working Jun 27 15:20:17 no where adb client server hint is therein logs Jun 27 15:20:51 how did u resolve adb issues as you earlier had? Jun 27 15:21:52 we need to reach this: android_usb gadget: android_usb ready Jun 27 15:22:39 I saw some hint in Halium somewhere... THere's a PR pending for ADB there Jun 27 15:22:43 Let me see if I can find it Jun 27 15:23:38 Ah it's this one, doesn't seem too relevant: https://github.com/Halium/android_system_core/pull/11/files Jun 27 15:26:12 BT40: I could try to recreate the Halium bits for Land, could be there's something obvious wrong in https://github.com/SRIDHARDV/android_device_xiaomi_land/commit/41065730b8cae2051f9337e48b3e64be8f1cce3c but nothing from the first glance Jun 27 15:26:25 Question is: Does ADB work on CM 14.1 build for Land? Jun 27 15:28:35 Could re-do the kernel/defconfig from scratch Jun 27 15:33:33 Taking cm14.1 as basis Jun 27 15:33:53 Maybe the patches somehow interfere but they really shouldn't normally Jun 27 15:38:28 there are several builds of CM14 forr my device. One was not even installing. Now the one that boots, gps doesnot work. CM is not reliable in case of my device Jun 27 15:39:00 so thats why i wanna try stock kernel, if u guys allow and support Jun 27 15:42:07 comparing ours with stock kernel, following things are missing from config file: https://pastebin.com/khZENDgt Jun 27 15:42:30 i'll add in config file and give it a try atleast Jun 27 15:59:21 Yeah why not Jun 27 15:59:40 You can always try stock kernel of course Jun 27 15:59:49 Just it would require some patches Jun 27 16:01:03 Herrie, halium also has its own kernel. so do i also need to make changes there? Or bitbake takes the care of that? Jun 27 16:01:18 i mean there will be two kernels, which will be prefered Jun 27 16:01:28 halium one or bitbake one? Jun 27 16:01:32 BT40: We don't really care about the kernel part in Halium Jun 27 16:01:50 okay Jun 27 16:02:09 Most important for us is that the kernel in Halium doesn't break the Halium build and the config options are more or less the same Jun 27 16:02:46 Ideally the config is identical but for me it worked with devices where they were different Jun 27 16:03:36 will compile kernel and give feedback Jun 27 17:28:54 compilation done, flashing now Jun 27 17:29:02 booted Jun 27 17:31:37 adb not worked, neither detected through lsusb Jun 27 17:32:03 charging works through usb cable when attached to computer Jun 27 18:18:49 Well when flashing works I would expect the cable is OK Jun 27 18:33:57 tried kernel compilation with only defconfig instead of land_defconfig, compilation failed with some errors. Jun 27 18:33:59 :) **** ENDING LOGGING AT Sun Jun 28 02:59:57 2020