**** BEGIN LOGGING AT Thu Jan 25 03:00:06 2018 Jan 25 12:30:08 hey bshah, has plasma mobile looked at anything for tablets? Jan 25 12:30:42 I suppose yes.. well it works quite well on "tablet-ish" form factor Jan 25 12:30:55 for instance on my convertibles Jan 25 12:31:04 -s Jan 25 12:31:28 r0kk3rz: see : https://imgur.com/a/lFGc2 Jan 25 12:32:25 maybe i'll have the check it out Jan 25 12:32:44 r0kk3rz: which device you want to run it on? Jan 25 12:33:11 i have a sony z3 tablet, but i have my own sfos adaptation Jan 25 12:33:35 aries codename? Jan 25 12:33:42 scorpion Jan 25 12:34:01 aries is the phone afaik Jan 25 12:34:04 ah Jan 25 12:34:27 is very similar anyway Jan 25 12:35:21 if you manage to make halium build for it, it should be much easier to run on it I believe Jan 25 13:28:30 @mariogrip would you add canal also for vivid for device titan? and how can i update system.img and boot.img ? Jan 25 13:36:47 Sergobot was added by: Sergobot Jan 25 13:43:38 Hi everybody! I'm porting Halium to Pixel XL and currently stuck at fixup-mountpoints step. Do I need to flash LOS 14.1 or stock (but rooted) Android 8 is okay for that? Jan 25 13:44:17 For fixup-mountpoints I don't think you need to flash the LOS 14.1 Jan 25 13:44:33 What's problem with fixup-mountpoints? Jan 25 13:44:57 hmm, sometimes mountpoints change with kernel upgrades though Jan 25 13:45:15 @bhushanshah, No problem, just that question for now :) Jan 25 13:45:37 @r0kk3rz, Hmm, I'll then flash it just to be sure. Thanks Jan 25 13:58:56 @Sergobot, just search the repo of lineage of your device for mountpoints no need to flash Jan 25 14:00:54 Hm, why don't docs suggest this way? Jan 25 14:02:05 @Sergobot, maybe they didnt look at it that way! Anyway it wouldnt come in the device if its not there in the source. Jan 25 14:02:29 Yeah, makes sense Jan 25 14:05:20 So, I have found this … https://github.com/LineageOS/android_device_google_marlin/blob/cm-14.1/recovery.fstab … But it doesn't include real nodes (except /sdcard) Jan 25 14:12:59 They seem real? Jan 25 14:14:51 adb shell ls -la /dev/block/boot-device/by-name Jan 25 14:14:57 @Sergobot, adb shell ls -la /dev/block/boot-device/by-name Jan 25 14:15:28 @petya230, Doesn't it require root? Jan 25 14:15:57 @Sergobot, I think those mountpoints are fine Jan 25 14:16:03 Go with them for now Jan 25 14:16:11 @ilyaishere, Probably. But there's no actual nodes Jan 25 14:16:21 What actual nodes? Jan 25 14:16:38 like `mmcblk0p1` Jan 25 14:17:14 That's fine, I think Jan 25 14:17:21 @Sergobot, you mean like these https://github.com/LineageOS/android_device_google_marlin/blob/accbb7ebdf00137ab7c5ab3b17aab358f9807493/fstab.common Jan 25 14:18:01 He means |mmcblk*` Jan 25 14:18:07 oh ok Jan 25 14:18:38 Look at T00F in the script Jan 25 14:18:48 I think I used by-name Jan 25 14:19:33 @ilyaishere, Yes Jan 25 14:19:46 Then you can go with them :) Jan 25 14:19:54 Okay then, I'll use those Jan 25 14:20:02 Thanks Jan 25 14:20:10 I have an impression that there are only Russians in this chat πŸ™ˆ Jan 25 14:20:23 hey Jan 25 14:20:25 im not russian Jan 25 14:20:34 thank you for ruining the joke Jan 25 14:20:42 πŸ˜‚πŸ˜˜ Jan 25 14:20:56 im waiting for the joke Jan 25 14:21:00 😁 Jan 25 14:21:05 LolπŸ˜‚ Jan 25 14:21:07 oh screw you Jan 25 14:21:23 :c Jan 25 14:21:25 @vanyasem, *cough* Jan 25 14:21:41 soo Jan 25 14:21:46 i have new problem Jan 25 14:21:57 my halium not boot, but, not reboot... Jan 25 14:22:04 but i dont have telnet xDD Jan 25 14:22:06 ? Jan 25 14:22:14 if i install hybris-recovery Jan 25 14:22:17 @bhushanshah, look at the pr on GitHub, NSA packaged all halium deps for Arch Jan 25 14:22:18 same :c Jan 25 14:22:32 @vanyasem, i want you to review this :P Jan 25 14:22:43 so i dont know, what i need to doo Jan 25 14:22:44 @vanyasem, Well if you reviewed merge it Jan 25 14:22:51 You don't need to wait for my review Jan 25 14:22:56 AUR (en) - halium-devel … https://aur.archlinux.org/packages/halium-devel/ Jan 25 14:22:57 Β―\_(ツ)_/Β― Jan 25 14:23:14 @bhushanshah, you rejected mariogrips PR for a reason Jan 25 14:23:31 so I need you to review this PR to see if it's okay for your Jan 25 14:23:57 @petya230, Possibly.. you're in rootfs stage? you get USB interface? Jan 25 14:24:06 @bhushanshah, nothing. Jan 25 14:24:20 maybe the roofs is booted Jan 25 14:24:26 but no usb Jan 25 14:24:32 i mean telnet Jan 25 14:24:43 is ssh exposed Jan 25 14:24:54 you don't get telnet in the rootfs Jan 25 14:25:22 but its dont have ip Jan 25 14:25:24 i cant ssh Jan 25 14:25:25 :c Jan 25 14:25:32 and its not work as "hotspot" Jan 25 14:28:05 so.. i cant get log :c Jan 25 14:29:54 @vanyasem @NSA_AMERICA last comment I added.. once it is resolved feel free to merge Jan 25 14:44:48 @petya230, you need to assign ip Jan 25 14:45:22 http://docs.halium.org/en/latest/porting/debug-build/logging-in.html Jan 25 14:45:25 see last para Jan 25 14:47:42 "Once your device is booted, you should see that you have a new network interface again" Jan 25 14:47:45 i dont see it.. Jan 25 14:47:53 its not make a network interface Jan 25 14:48:05 oh Jan 25 14:48:33 do you have last_kmsg or console ramoops? Jan 25 14:50:12 i dont have last_kmsg coz 4.x kernel Jan 25 14:50:35 i dont have console ramoops too, because i not have "emergency reboot" Jan 25 14:50:36 :l Jan 25 14:52:15 now, i cant work without logs.. :c Jan 25 14:52:50 maybe revert last things you changed Jan 25 14:52:51 :p Jan 25 14:53:00 ah Jan 25 14:53:07 i dont know, what my last step Jan 25 14:53:08 xD Jan 25 15:18:21 @bhushanshah, @bhushanshah for some reason @vanyasem thinks we need to guide user through everything (specifically arch) anyway i removed any speculations regarding user having any problem😐 ,also that pgp key hasnt changed for over a year so we can that let that stay, i willing to keep maintaining that meta package... add me to halium on github 😏 Jan 25 15:19:22 @NSA_AMERICA, Sure, gpg key have not changed for year but that doens't mean it won't change.. it can be compromised and change/revoked in one day Jan 25 15:19:40 It's basically information we can't control Jan 25 15:19:56 or don't control Jan 25 15:21:27 well i am in favour of removing that... it is just ivan keep bugging Jan 25 15:21:40 :p Jan 25 15:21:48 @vanyasem ? Jan 25 15:21:49 :P Jan 25 15:22:43 @bhushanshah, i spent my entire day reviewing changes to the pr and half of them were nonse (such as adding some weird french repo to pacman sources). everything's public on github ^^ Jan 25 15:23:21 Β―\_(ツ)_/Β― Jan 25 15:23:54 it's not called bugging, it's called "reviewing a PR" @NSA_AMERICA ;) Jan 25 15:26:24 @vanyasem, you made me do that, it was more simple in the first version, which was to let the user find the packages, and intall them with his own prefered method , if he use pacaur,yaourt or pamac ... well instead of removing the gpg stuff completely , just add a note? Jan 25 15:26:53 @NSA_AMERICA, just give a link to arch wiki https://wiki.archlinux.org/index.php/GnuPG Jan 25 15:27:20 @NSA_AMERICA, there was already PR #9 that was very simple. and it was not merged for reasons ;) Jan 25 15:27:45 for reasons i also quoted on the PR Jan 25 15:28:08 that pr had same gpg stuff too -__- Jan 25 15:28:18 was it meged though? Jan 25 15:28:31 no, for reasons. please, stop arguing. you're bugging me now Jan 25 15:28:55 its me bugging? πŸ˜‚ Jan 25 15:29:59 @bhushanshah is a seperate note okay for pgp ? Jan 25 15:30:22 pgp note is not needed at all tbh Jan 25 15:30:47 Arch users do know what installing from AUR entails Jan 25 15:31:05 and it's not in scope of halium documentation at all to explain that or duplicate stuff Jan 25 15:31:06 @vanyasem read that πŸ‘† Jan 25 15:31:18 He can read it Jan 25 15:31:30 Just update PR and be done with it, will you? Jan 25 15:31:31 @NSA_AMERICA, do you really think we all should think exactly the same? -_- Jan 25 15:37:41 bhushan go merge it fast , before ivan reviews itπŸ˜‚ Jan 25 15:38:44 question Jan 25 15:38:50 halium has bootanimation??? Jan 25 15:38:57 no Jan 25 15:39:18 but Jan 25 15:39:19 [ 99% 19165/19173] Building bootanimation … [ 99% 19166/19173] build /home/petya230/halium/out/target/product/hi6250/obj/ETC/bootanimation.zip_intermediates/bootanimation.zip … [ 99% 19167/19173] Install: /home/petya230/halium/out/target/product/hi6250/system/media/bootanimation.zip Jan 25 15:39:24 What is this? xDDD Jan 25 15:40:11 @petya230, you can ignore it Jan 25 15:42:04 kk Jan 25 15:47:36 Saurabh Pradhan was added by: hekdhbykodh Jan 25 16:11:08 Guys Jan 25 16:11:09 [ 73.543243s][pid:236,cpu0,ts_thread:0][E/TS] load fw data from bootimage error Jan 25 16:11:11 nope Jan 25 16:11:13 nope this Jan 25 16:11:22 http://termbin.com/w14k Jan 25 16:11:56 question Jan 25 16:12:00 halium 64bit? Jan 25 16:15:30 @ilyaishere, You need to twpr the device and then you can run adb shell as root Jan 25 16:17:10 @petya230, . Jan 25 16:17:31 no there is no arm64 rootfs at moment Jan 25 16:17:42 oh Jan 25 16:17:50 maybe this is the problem.. Jan 25 16:18:23 At this point, run readlink -f the leftmost part of each fstab entry (such as /dev/.../system and it will point to the correct mmcblk0pX where X is a number Jan 25 16:18:56 What you guys think about dependencies for fedora πŸ€”πŸ™ƒ Jan 25 16:19:45 we need 64bit rootfs Jan 25 16:19:46 xD Jan 25 16:20:05 I will be working on it this weekend Jan 25 16:20:43 :O Jan 25 16:20:44 yey Jan 25 16:20:52 @petya230, beware that 64-bit libhybris is not tested much Jan 25 16:20:59 :/ Jan 25 16:21:02 if you can, better use 32-bit rootfs for now Jan 25 16:21:08 kernel error Jan 25 16:21:16 i cant compile in 32bit Jan 25 16:21:27 you don't need to compile kernel or Android parts in 32-bit Jan 25 16:21:47 hm Jan 25 16:21:52 http://termbin.com/w14k Jan 25 16:22:04 so not this the problem? Jan 25 16:22:35 [ 73.543243s][2018:01:25 16:07:08][pid:236,cpu0,ts_thread:0][E/TS] Firmware image ts/venus_synaptics_VENS11140_ebbg.img not available Jan 25 16:22:59 p9lite have lots of version Jan 25 16:23:06 its dont needed for my phone Jan 25 16:23:42 I don't see it starting systemd though Jan 25 16:23:54 mer-kernel-check was fine with your config? Jan 25 16:24:30 CONFIG_SCHED_DEBUG … CONFIG_AUDIT Jan 25 16:24:32 these Jan 25 16:24:38 i cant enable Jan 25 16:24:39 MeticulusDev was added by: MeticulusDev Jan 25 16:24:44 that's fine then Jan 25 16:25:00 WARNING: CONFIG_SCHED_DEBUG is invalid … It is unset … Allowed values : y, ! … WARNING: CONFIG_AUDIT is invalid … Value is: y … Allowed values : n, ! Jan 25 16:25:21 second: i cant turn off xD coz not cimpile.. Jan 25 16:25:22 compile* Jan 25 16:25:37 Hello, I am helping @petya230 to try to make halium for huawei/honor hi6250 devices. πŸ˜ƒ Jan 25 16:28:40 @NotKit, so? xD … Any ide? Jan 25 16:28:43 idea* Jan 25 16:29:55 well, not many, you can build and flash hybris-recovery (which is boot.img actually), and it's going to start telnet and wait before switching to rootfs Jan 25 16:30:07 so you can make sure /target is mounted properly and is r/w Jan 25 16:30:13 Im an new but can I ask where is the source code for halium init. I seem like init attempts to mount 1 mmcblock and one in loop fails then emergency reboot... Jan 25 16:30:46 https://github.com/Halium/hybris-boot/blob/master/init-script Jan 25 16:30:54 before it switches to rootfs and starts systemd Jan 25 16:31:51 @NotKit, so, i need to flash hybris-reco, and boot to reco right? Jan 25 16:32:06 flash hybris-recovery to boot and boot normally Jan 25 16:32:18 to boot Jan 25 16:32:19 ? Jan 25 16:32:20 lol Jan 25 16:32:21 kk Jan 25 16:32:32 It's not a recovery, it just has more debugging stuff enabled Jan 25 16:32:41 oh, i see Jan 25 16:32:42 bad naming, actually Jan 25 16:32:46 The name is maybe confusing, but it has to be flashed to boot Jan 25 16:35:27 okay, om Jan 25 16:35:30 its reboot.xD Jan 25 16:35:33 I assume that our failure is happening here: https://github.com/Halium/hybris-boot/blob/master/init-script#L147 Jan 25 16:42:05 our failure says that these 2 can not be mounted due to incompatible features in ext3. Can this be a result of including selinux or that our TWRP, as a matter of formatting the data partition, includes selinux? Jan 25 16:45:21 ive seen that happen, but it eventually mounts Jan 25 16:46:24 @MeticulusDev, I had your problem... I've resolved it with an e2fsck-static downgrade Jan 25 16:50:17 @Mattia990, Are you saying that I should use e2fsck to "downgrade" the structure of the filesystem or that I should "downgrade" to a different static version of e2fsck? As I understand it e2fschk preforms it's check on the filesystem while unmounted. May I also ask where e2fsck is located in the runtime halium eco system? Jan 25 17:10:39 @MeticulusDev, . Jan 25 17:29:52 @MeticulusDev, I have Ubuntu, i download e2fsck debs and i installed with dpkg -i "file name.deb" Jan 25 17:32:18 (Document) https://irc.ubports.com/0Cf2Argz/file_3908.png Jan 25 17:32:18 You are aware that the build used the e2fsck in out/host/linux-x86/bin Jan 25 17:40:33 Mmm i don't know. I installed the deb of e2fsck (2015) Jan 25 17:44:40 @MeticulusDev, i like this mate ._. Jan 25 17:45:37 https://hastebin.com/lazulizore.go - I wonder what kernel option can make systemd that slow Jan 25 17:49:05 @NotKit, Got full dmesg? Jan 25 17:49:17 it's really flooded Jan 25 17:50:10 Want initial 1 minute worth log Jan 25 17:58:42 (Document) https://irc.ubports.com/K8hfrDLZ/file_3909.txt Jan 25 18:02:57 Going through 1.4mb log on phone is totally fun Jan 25 18:03:04 (not) Jan 25 18:04:52 @NotKit for starters disabling trace printk would help lot Jan 25 18:05:05 @Sergobot, So, I have installed stock Android 7.1.2 onto my Pixel XL and rooted it Jan 25 18:05:12 But here's one little problem Jan 25 18:05:28 there's no `/dev/block/bootdevice/by-name/system` Jan 25 18:05:55 Instead, I have `/dev/block/bootdevice/by-name/system_a` and `/dev/block/bootdevice/by-name/system_b` Jan 25 18:06:14 Which point to `/dev/block/sda33` and `34` respectively Jan 25 18:07:34 @Sergobot, check fstab for mount points. I think this is what is known in android world as A/B setup. Perhaps system_b mounts @ vendor? Jan 25 18:08:08 @Sergobot, You don't need to add system partition in fixup-mountpoint Jan 25 18:08:48 @bhushanshah, the bad thing is that drivers need to be patched for that, they didn't care to use proper loglevel or debug functions Jan 25 18:08:55 @bhushanshah, I was wondering about this. why do we bother with all those partitions, isn't the only one we need userdata? Jan 25 18:09:19 There're two so-called "slots" that have separate system, boot, modem, and others Jan 25 18:09:32 (Photo, 1280x78) https://irc.ubports.com/g8kJJmqE/file_3911.jpg Jan 25 18:10:04 @bhushanshah, two vendor partitions too Jan 25 18:10:48 @Sergobot, You just need userdata and boot partition Jan 25 18:11:02 Okay, two boot paritions too Ρ…D Jan 25 18:11:30 And I think for a/b system you can use either of them Jan 25 18:11:58 Tbh this is first port to a/b partitioned device Jan 25 18:12:22 So we need to research and figure out Jan 25 18:12:43 @Sergobot: It does bring the combinatorial complexity down to level where trail and error seems a reasonable tactic :) Jan 25 18:12:53 trial Jan 25 18:13:15 For now, I guess, we could just hardcode for either a or b partition Jan 25 18:13:19 Just for testing Jan 25 18:13:34 @doniks, Well we just need userdata and boot yes but people add all mountable partition so that if in future we decide to use other partition we can use them Jan 25 18:14:08 @bhushanshah. ah, ok, fair enough. thanks Jan 25 18:15:52 @NotKit, Good luck I should say 🀣 anyway will check log from laptop tomorrow Jan 25 18:15:54 I think this came from SailfishOS, as all partitions mount is handled by systemd Jan 25 18:16:22 so fstab is parsed, run through fixup-mountpoints and converted to .mount services Jan 25 18:16:29 On mobile it's terrible to decipher Jan 25 18:16:44 http://termbin.com/ruzd Jan 25 18:16:48 guys? >< Jan 25 18:18:24 @petya230, I can't find anything in there about a booting Halium device Jan 25 18:18:53 sooo.. halium not want to boot Jan 25 18:19:04 kernel? Jan 25 18:19:21 I can't even see hybris-boot in debug mode? Jan 25 18:19:34 its hybris-recovery Jan 25 18:19:49 Or wait is it the last kmsg? Jan 25 18:19:55 yap, its the last Jan 25 18:20:01 Ah ok Jan 25 18:23:17 @petya230, first try fix *.rc files problem, disable anything that cause error Jan 25 18:23:34 [ 11.917083s][pid:1,cpu5,init]init: could not import file '/init.chip.usb.rc' from '/init.recovery.huawei.rc': No such file or directory Jan 25 18:23:56 This looks like it's recovery image of your device Jan 25 18:24:12 And not hybris boot or hybris recovery Jan 25 18:26:42 @petya230, om.. wrong dmesg Jan 25 18:29:29 http://termbin.com/pu1e Jan 25 18:29:34 this is the "good" xD Jan 25 18:33:14 (Document) https://irc.ubports.com/pJUU7hdb/file_3913.png Jan 25 18:33:38 [ 14.628540s][pid:446,cpu4,mount]EXT4-fs (mmcblk0p49): couldn't mount as ext3 due to feature incompatibilities … [ 14.629577s][pid:446,cpu4,mount]EXT4-fs (mmcblk0p49): couldn't mount as ext2 due to feature incompatibilities … [ 14.637115s][pid:452,cpu6,mount]EXT4-fs (loop0): couldn't mount as ext3 due to feature incompatibilities … [ 14.637298s][pid:452,cpu6,mount]EXT4-fs (loop0): couldn't mount as ext2 due to feature incompatibilitie Jan 25 18:33:39 meh Jan 25 18:33:59 That's not issue tbh . Just warning Jan 25 18:34:08 @Sconio, Wrong group Jan 25 18:34:10 hm Jan 25 18:34:22 @petya230, so, what is the error? xD Jan 25 18:35:07 pid:234,cpu0,ts_thread:0][E/TS] Firmware image ts/venus_synaptics_VENS11140_ebbg.img not available [ 73.509857s][pid:234,cpu0,ts_thread:0][E/TS] load fw data from bootimage error [ 76.244995s]sysrq: SysRq : Emergency Remount R/O Jan 25 18:35:11 This Jan 25 18:36:12 hm Jan 25 18:37:09 @bhushanshah, dev said this is no bad."Touch screen work just fine without that firmware.." Jan 25 18:37:23 @Sergobot, I have briefly researched how that dual-slot thing works. Device boots the pre-selected slot (dual-boot can be enabled through manually setting the bootable slot using TWRP or fastboot), so there's only one set of partitions mounted at a time. I think, it should be possible to find out the slot during halium installation. Not sure how exactly to do it but I will research more tomorrow. Jan 25 18:37:59 On the other hand, the slot could be predefined at compile time. Jan 25 19:01:37 hidden00000 was added by: hidden00000 Jan 25 19:09:45 @Sergobot, Apparently the method of booting completely changes as well. Is it correct that we can't modify the bootimg to change its behavior? Jan 25 19:10:01 Well... either way, there's no recovery partition anyway Jan 25 19:16:47 @petya230 That error could be caused by MMC_FFU Jan 25 19:17:36 @UniversalSuperBox, Totally unsure of this, I'm a complete beginner in tampering with devices' partitions and boot especially. But the method does not reallly change. Choosing the bootable slot is done at a lower level of bootloader, I don't think we can change that. Jan 25 19:17:56 In other kernels I have found code in mmc init that tries to find firmware files in specific partitions before advertising the partitions to the rest of the system Jan 25 19:18:11 based on timing.. and the reaction.. that might be the caser here Jan 25 19:18:22 @UniversalSuperBox, And even though there's no separate recovery partition, there's an official port of TWRP that works just fine Jan 25 19:18:32 if you cannot find that option, look for the path inside drivers.. Jan 25 19:18:54 or error message.. Jan 25 19:21:57 @anpok, :o … and I need to turn off in kernel? Jan 25 19:22:01 Or where? Jan 25 19:22:33 either get the file in place .. or disable the code that seems to block the startup for a minute.. Jan 25 19:22:44 yes in kernel Jan 25 19:22:47 Kk Jan 25 19:22:49 Thanks! Jan 25 19:23:04 I am not sure what boot device means here.. (i mean which partition actually..) Jan 25 19:23:23 but the error message should be distinctive enough Jan 25 19:25:14 That branch is untested ... Jan 25 19:25:33 oh. Jan 25 19:25:35 xD Jan 25 19:25:37 Are you sure you want the flash-system-partition branch? Jan 25 19:25:58 i tried 3 branch. … master … ext4 … system Jan 25 19:26:11 master should be just fine for porting Jan 25 19:26:27 oh, kk Jan 25 19:26:31 i will use it Jan 25 19:26:36 The other branch are only work in progress stuff, or merged into master already Jan 25 19:26:41 @anpok, which error message? Jan 25 19:27:27 Firmware image ts/venus_synaptics_VENS11140_ebbg.img not available [ 73.509857s][pid:234,cpu0,ts_thread:0][E/TS] load fw data from bootimage error Jan 25 19:28:48 ._. Jan 25 19:29:37 or what is rebooting the system? Jan 25 19:30:00 yap Jan 25 19:30:02 its reboot Jan 25 19:30:07 and not boot Jan 25 19:31:26 and we dont find the problem.. :D Jan 25 19:31:34 now i try with what u say Jan 25 19:31:40 make all again Jan 25 19:32:03 maybe a coincidence .. but try to find that code.. maybe you can also pin down the sources for the follow up errors messages Jan 25 19:33:24 hmmm Jan 25 19:33:41 grep Jan 25 20:21:30 help … what is this.. Jan 25 20:21:48 ```BuildImage: in_dir = /home/petya230/halium/out/target/product/hi6250/system, out_file = /home/petya230/halium/out/target/product/hi6250/obj/PACKAGING/systemimage_intermediates/system.img … **fs type is not ext4** … Running: mkuserimg.sh -s /home/petya230/halium/out/target/product/hi6250/system /home/petya230/halium/out/target/product/hi6250/obj/PACKAGING/systemimage_intermediates/system.img ext4 system 2080374784 -D Jan 25 20:21:49 /home/petya230/halium/out/target/product/hi6250/system -L system /home/petya230/halium/out/target/product/hi6250/root/file_contexts.bin … make_ext4fs -s -T -1 -S /home/petya230/halium/out/target/product/hi6250/root/file_contexts.bin -L system -l 2080374784 -a system /home/petya230/halium/out/target/product/hi6250/obj/PACKAGING/systemimage_intermediates/system.img /home/petya230/halium/out/target/product/hi6250/system /home/petya230/halium/out/target/product Jan 25 20:21:55 fs type is not ext4 Jan 25 20:21:58 Read the pin Jan 25 20:22:17 meh. Jan 25 20:22:56 https://pastebin.com/jeaEL1mE Jan 25 20:23:32 now its good? Jan 25 20:39:01 hmm. Everything seams to work in the hybris boot init script, I can watch the output right until "exec switch_root /target /sbin/init" and then it stops. There is no output from the switch_root or systemd in /target/data/init-stderrout and SSH does not come up. How do I debug this? Jan 25 20:40:20 @ZeroPointEnergy, Do you get USB interface? Jan 25 20:40:35 nope Jan 25 20:40:39 not even that Jan 25 20:40:49 What kernel version? Jan 25 20:42:04 3.18.14 Jan 25 20:44:41 ZeroPointEnergy: last_kmsg? Jan 25 20:45:25 That kernel version would be using ramoops I believe Jan 25 20:45:40 let me check the kernel messages Jan 25 20:46:17 @bhushanshah, And ramoops is useless. :P Jan 25 20:47:48 was just full of unreadable stuff. Let be cycle it and try it again Jan 25 20:51:48 ok, nice, there are some systemd messages in there Jan 25 20:54:53 "power_supply max77854-fuelgauge: driver filed to report `status' property" Jan 25 20:56:24 Pastebin full log if you can.. this might be probably just false alarm Jan 25 21:01:47 https://pastebin.com/5VbVCbj1 Jan 25 21:02:18 the previous messages where all the same Jan 25 21:02:42 Uhhhhh Jan 25 21:02:45 Okay Jan 25 21:03:16 Let's try to disable lxc@android service Jan 25 21:03:51 ok Jan 25 21:04:15 And well I need to go off now..it's 2:30 am.. Jan 25 21:04:37 Hopefully someone will be able to help you Jan 25 21:05:30 no problem, I have a way to see some output, that's like 100% improvement already :-) Jan 25 21:12:29 ZeroPointEnergy: you know how to disable lxc? Jan 25 21:14:20 Yes, thanks to your notes :-) Jan 25 21:14:48 glad to hear they are of use Jan 25 21:15:01 curious to see your new log Jan 25 21:19:03 it is still the same. I try to disable the driver who goes nuts in the logs now, maybe that will let me see whats actually going on Jan 25 21:19:32 sounds good! Jan 25 21:39:38 heh, bootloops without that driver Jan 25 21:47:04 you said: hmm. Everything seams to work in the hybris boot init script, I can watch the output right until "exec switch_root /target /sbin/init" and then it stops. There is no output from the switch_root or systemd in /target/data/init-stderrout and SSH does not come up. How do I debug this? Jan 25 21:47:21 can you show the dmesg BEFORE switch_root? Jan 25 21:49:38 they are no longer in the kmsg unfortunatly Jan 25 21:49:53 will try to somehow get them Jan 25 21:54:54 what did you mean with "I can watch the output" Jan 25 22:06:54 Over /proc/kmsg, but that seams very limited and it depends on the time I interrupt the device what messages I see Jan 25 22:10:43 and in telnet, the output of dmesg is that more complete? Jan 25 22:11:19 I could imagine that the /proc/last_kmsg has some limit and if some driver floods the log then maybe you lose the begin Jan 25 22:13:52 yes, that was more complete, will have to force a telnet session. I try to modify the init script now to dump it onto /data/ before systemd takes over and for systemd I try --log-target=console and then pipe it to a file Jan 25 22:16:16 sounds good. Jan 25 22:16:35 I find the end of the log you posted last quite peculiar Jan 25 22:16:53 all the while it is full of noise, but at the very end Jan 25 22:17:30 it prints out a block Jan 25 22:17:55 why? how? does it do that right before it reboots? Jan 25 22:18:38 and secondly, there are some "verbose sounding" elements on that kernel command line it seems to be printing last Jan 25 22:18:38 if I press the reset combination it triggers a kernel bug, I think it is this Jan 25 22:19:02 ah ok Jan 25 22:19:44 ok and on that kernel commandline there is loglevel=4 androidboot.debug_level=0x4f4c Jan 25 22:20:01 maybe even sec_tima_log=0x200000@0x92e02000 sec_avc_log=0x40000@0x93004000 sec_summary_log=0x200000@0x93046000 auto_summary_log=0x1000@0x93248000 Jan 25 22:20:36 I'm not claiming I know what they do, but maybe you can get rid of those will make your kernel less chatty Jan 25 22:21:40 true, good idea Jan 25 22:23:07 ok, loglevel seems to be normal linux and 4 (KERN_WARNING) doesn't sound overly verbose https://www.kernel.org/doc/html/v4.10/admin-guide/kernel-parameters.html Jan 25 22:23:38 but yeah, maybe one of them does make a difference Jan 25 23:18:03 Getting "realloc: Out of memory" when running the halium-install script. Do you know which memory? **** ENDING LOGGING AT Fri Jan 26 03:00:02 2018