**** BEGIN LOGGING AT Sat Jan 13 03:00:01 2018 Jan 13 07:42:51 https://paste.ubuntu.com/26376682/ Jan 13 07:43:08 @SALPHA8506, aprrox 1 min lights on in display Jan 13 07:45:10 @SALPHA8506, what it means? Jan 13 07:54:33 all the files in /vendor/lib64/...... , so i think it search the file in /vendor/lib/...... Jan 13 07:56:19 @SALPHA8506, can we change code for serching file from lib64 lnstead of lib.... Jan 13 07:58:37 exp--> proprietary/vendor/lib64/libsdm-diag.so:system/vendor/lib64/libsdm-diag.so not in lib Jan 13 08:07:03 @SALPHA8506, No because the rootfs use a 32bit lib Jan 13 08:07:44 You have two ways: … -Find 32bit lib and add to vendor … -Use a aarch64 rootfs Jan 13 08:09:16 link Jan 13 08:09:36 rootfs64 Jan 13 08:09:54 @SALPHA8506, Arco linux armi Jan 13 08:12:16 @ErnyTech, is halium rootfs 64 bit avaliable..? Jan 13 08:14:44 @SALPHA8506, No it's armhf Jan 13 08:14:58 But you can remake it Jan 13 08:19:34 The ubuntu branch of rootfs-builder is very similar than halium rootfs https://github.com/Halium/rootfs-builder/tree/ubuntu Jan 13 08:21:43 But libhybris are not available for aarch64 in ubuntu repo Jan 13 08:23:55 you should compile all .. it's long you can not find armhf libraries? Jan 13 08:59:11 external/speex/Android.mk:56: TODOArm64: enable neon in libspeex … what it mean?? To do ARM64 Jan 13 09:00:11 @SALPHA8506, just ignore it Jan 13 09:00:16 it's not an error Jan 13 09:46:20 @SALPHA8506, what about this Jan 13 09:47:24 all the files exist in lib64 and it search in lib Jan 13 10:27:25 (Video, 0s)https://irc.ubports.com/rSSJ90dj/file_3749 Jan 13 10:27:39 udev rules for ubports: shall they go to /etc/udev or to var/lib thingie Jan 13 10:27:42 :) Jan 13 10:31:26 @Flohack, ur device blocked in bootanmation ? Jan 13 10:31:33 no udev rules now Jan 13 10:31:37 Creating them Jan 13 10:33:13 hmm the container worked well under Halium, but with UT it seems not to start correctly Jan 13 10:34:00 Hey flo .. how you managed to put ut in 10mb boot partition?.. did you resize it? Jan 13 10:34:19 @Mr_Doge, No, I disabled some stuff in the kernel config, plus switched to LZMA compression Jan 13 10:35:28 Oh cool Jan 13 10:35:51 hmm but the gui wont come up... Jan 13 10:36:01 it comes!!!!! Jan 13 10:36:07 its dead slow Jan 13 10:37:47 @SALPHA8506, Now i try build arm64 rootfs based from ubuntu 17.10 … then I send you the link so as to try, should also solve my problem Jan 13 10:39:47 It crashes a bit ^^^ Jan 13 10:47:11 I got no OSK Jan 13 10:47:17 Any idea how to debug maliit Jan 13 10:50:04 maliit-server keeps crashing, its busy in top and also apport ^^ Jan 13 10:55:08 And whats wrong with systemd: … root@ubuntu-phablet:/var/log# systemctl status lxc@android … Failed to get properties: No such interface '' Jan 13 10:55:18 this worked in Halium root fs Jan 13 10:56:51 UBports is using Upstart Jan 13 10:56:55 instead of systemd Jan 13 10:57:00 keyboard work for you florian Jan 13 10:57:30 ahhh Jan 13 10:57:41 damn you are right ^^ 16.04 is now all upstart Jan 13 10:58:51 @Walid, How? Jan 13 11:07:28 I got no /sys/module/wlan Jan 13 11:08:43 I got /dev/wcnss_wlan but ist permission denied ^^ Jan 13 11:08:48 is this a missing udev rule? Jan 13 11:16:57 i think you must modif your init.rc to start wifi Jan 13 11:18:07 oki Jan 13 11:18:13 somthing like this … service init_wlan_bt /system/bin/sh /system/etc/init_wlan_bt.sh … class main … user root … group root … oneshot … service wcnss-service /system/bin/wcnss_service … class main … user root … group root … oneshot Jan 13 11:18:37 (Document) https://irc.ubports.com/t2khj3t0/file_3750.sh Jan 13 11:19:00 cool =) where to put this Jan 13 11:19:33 init.qcom.rc or similar Jan 13 11:19:49 Oh jesus! ok will do Jan 13 11:20:13 a systemimage rebuild is enough? Jan 13 11:21:13 yes Jan 13 11:22:58 this will affect only android ramdisk Jan 13 11:28:51 hmm can I manually edit this also? Where is init.rc inside systemimage at the end? Jan 13 11:29:56 hmm android-ramdisk probably Jan 13 11:30:43 @walid can I have your kernel git? Jan 13 11:30:56 need to fix smth so I cant build systemimage right now Jan 13 11:40:39 https://github.com/walidham/ubuntu_touch_msm8226_kernel Jan 13 11:41:05 but not last work Jan 13 11:50:45 got it. Just bringing in apparmor 3 Jan 13 11:56:22 just replace apparmor folder Jan 13 11:58:10 hmm yes I was thinking like this, but it got a few dependencies into main kernel too Jan 13 11:58:17 strange that worked before Jan 13 12:03:16 what errors you got ? Jan 13 12:04:24 change dst.h file in include/linux i think, use file from my kernel Jan 13 12:04:58 hmm a lot of stuff also for task struct etc. basically its missing no_new_privs variable Jan 13 12:05:02 and a constant Jan 13 12:08:32 pastebin error maybe i can help you, i know most famous error of kernel compilation for msm8226 Jan 13 12:09:12 No, it seems the kernel 3.4 I got is somehow different from yours maybe ;) - but I try to fix it, then I can show you Jan 13 12:14:51 good luck :) Jan 13 12:21:47 @UniversalSuperBox hi when run command to create udev rules i got this error … -bash: /usr/lib/lxc-android-config/70-titan.rules: Permission denied Jan 13 12:22:18 you ran as sudo? Jan 13 12:22:51 yes Jan 13 12:23:11 writeable rootfs? Jan 13 12:23:25 sudo cat /var/lib/lxc/android/rootfs/ueventd*.rc|sudo grep ^/dev|sudo sed -e 's/^\/dev\///'|sudo awk '{printf "ACTION==\"add\", KERNEL==\"%s\", OWNER=\"%s\", GROUP=\"%s\", MODE=\"%s\"\n",$1,$3,$4,$2}' | sudo sed -e 's/\r//' >/usr/lib/lxc-android-config/70-titan.rules Jan 13 12:23:34 do it in a sudo -i shell Jan 13 12:23:40 yes i ran sudo mount -o remount,rw / Jan 13 12:23:41 its different ^^ Jan 13 12:26:07 I get a compile error now for kvfree Jan 13 12:26:12 with your apparmor directory Jan 13 12:26:37 yes its known error Jan 13 12:26:38 ../../../../../../kernel/samsung/s3ve3g/security/apparmor/apparmorfs.c: In function 'aa_simple_write_to_buffer': … ../../../../../../kernel/samsung/s3ve3g/security/apparmor/apparmorfs.c:110:3: error: implicit declaration of function 'kvfree' [-Werror=implicit-function-declaration] … kvfree(data); Jan 13 12:29:23 in my kernel its declared in another file, Jan 13 12:29:33 let me remember it Jan 13 12:34:14 https://github.com/ubports/android_kernel_moto_shamu/commit/83f949a8de673fe45499d1741da8654831a5afae Jan 13 12:34:19 use this patch Jan 13 12:35:27 ok let me try Jan 13 12:44:36 I am back to error: 'LSM_UNSAFE_NO_NEW_PRIVS' undeclared (first use in this function) Jan 13 12:44:41 The same as in mako kernel Jan 13 12:45:46 there is smth in backports.h what is strange Jan 13 12:45:55 this was commented out actually Jan 13 12:47:45 strange, for backport just copy folder will not work you must apply backport Jan 13 12:48:13 https://forums.ubports.com/topic/602/how-to-backport-kernel-to-make-bluetooth-work Jan 13 12:50:13 Now I got lsm error ^^ Jan 13 12:50:32 ../../../../../../kernel/samsung/s3ve3g/security/apparmor/lsm.c:1146:2: warning: initialization from incompatible pointer type [enabled by default] … error, forbidden warning: lsm.c:1146 Jan 13 12:55:09 change you KBUILD_FLAGS something like this (in Makefile of ur kernel) … KBUILD_CFLAGS := -Wall -Wundef -Wstrict-prototypes -Wno-trigraphs \ … -fno-strict-aliasing -fno-common \ … -Werror-implicit-function-declaration \ … -Wno-format-security -Wno-sizeof-pointer-memaccess \ … -fno-delete-null-pointer-checks -Wno-maybe-uninitialized Jan 13 13:12:40 Distanger was added by: Distanger Jan 13 13:34:12 @Walid, I fixed it by removing const keyword from the declaration in apparmor its ok ^^ Jan 13 13:34:40 Well I will stop for today. Pretty good progress, except that booting is very slow Jan 13 13:34:51 I can ssh into it, and before the container starts it takes ages Jan 13 16:58:11 hey everyone Jan 13 17:00:21 I had to re-do my build system cause the HDD crashed, got everything to build except for systemimage which is trying to build an apk, cant seem to find the original call for it, any ideas? Jan 13 17:03:00 @bhushanshah i enabled CONFIG_FW_LOADER_USER_HELPER but error still show, is it depend to another flag ? maybe CONFIG_FW_LOADER !!! Jan 13 17:03:19 That sounds right Jan 13 17:03:28 @Tyg3rpro, Remove all apks from the makefiles in vendor/$vendor/$device Jan 13 17:09:07 ok, Ill check to make sure I did that Jan 13 17:11:44 ok I checked, all the files in my vendor folder have no menton of an APK Jan 13 17:14:10 nvm, found it in the vendor folder of one of the dependencies Jan 13 17:14:33 Hey Dalton, is there any patch for firmware_class error in the sample kernel Jan 13 17:16:09 wat Jan 13 17:17:07 note to self, check all dependy vendor folders Jan 13 17:22:31 @UniversalSuperBox, Sorry took me time to go up there to find the paste bin Jan 13 17:22:36 https://paste.ubuntu.com/26324230/ Jan 13 17:22:53 I've never seen that error before in my life Jan 13 17:24:24 Well few kuid errors were fixed.. by your patch and few by Config_user_ns=n .... This is the last error Jan 13 17:24:30 Which is bugging me. Jan 13 17:24:42 isnt a forbidden error typically caused by lack of permissions? Jan 13 17:26:37 maybe a chmod on the file that caused the error? Jan 13 17:54:11 Tyg3rpro, just a shot in the dark, but Jan 13 17:54:13 https://github.com/sultanqasim/android_kernel_motorola_msm8916/blob/squid_oreo/drivers/base/firmware_class.c#L1063 Jan 13 17:54:44 this looks like it has smt to do with Jan 13 17:54:45 CONFIG_FW_LOADER_USER_HELPER Jan 13 17:54:49 do you have that defined? Jan 13 17:55:22 oh, sorry, you just said you did Jan 13 18:00:44 uh no, got the posting users mixed up, haha Jan 13 18:01:19 @Mr_Doge, you posted the error about fw_load_from_user_helper Jan 13 18:01:31 does defining CONFIG_FW_LOADER_USER_HELPER help? Jan 13 18:14:18 lol, its ok doniks Jan 13 18:20:13 ok, got a compiled image for halium again, willing to bet it wont get very far, I am hoping it at least outputs console stuff to the screen though Jan 13 18:23:08 Console output on the screen is quite unlikely. But you will get ssh connection to debug and view logs Jan 13 18:24:04 @doniks, No.. it was just enabled in the defconfig, Jan 13 18:27:18 last time i tried, it got stuck with something after pre-boot, never got to a point of getting ssh Jan 13 18:27:52 log never showed any errors or anything and I would leave it overnight and nothing would happen Jan 13 18:32:20 @Mr_Doge can you show Jan 13 18:32:22 pm_runtime: fail to wake up Jan 13 18:32:29 ugh not that, hang on Jan 13 18:32:39 grep FW_LOADER out/target/product/deb/obj/KERNEL_OBJ/.config Jan 13 18:32:41 that Jan 13 18:32:58 replace deb with your device name Jan 13 18:35:26 I dont have tree with me.... Give me an hour or so Jan 13 18:35:32 anyone know if there is work being done to port a asus tf201? Jan 13 18:36:03 the tf300 is similar hardware Jan 13 18:38:41 @Mr_Doge, sure, my thinking is this: line 1226 says fw_load_from_user_helper(fw, desc, timeout) and around line 1192 you see the three types which matches the definition in line 1051, which SHOULD be active with CONFIG_FW_LOADER_USER_HELPER , in line 1065 there is the alternative version which seems to NOT fit with 1266 Jan 13 18:38:49 Hope I'm making sense Jan 13 18:39:31 Well, not yet, but hopefully when you get a change to actually look at the code :) Jan 13 18:42:23 @doniks, https://github.com/abhishek9650/android_kernel_wingtech_msm8916/blob/Halium-7.1/drivers/base/firmware_class.c have a look... Jan 13 18:43:56 Walid, how is performance of Ubuntu Touch on titan? Jan 13 18:45:00 @Mr_Doge, exactly. Looks the same than the other link I dug up Jan 13 18:47:03 @UniversalSuperBox, with 15.04 work well, all features : camera, wifi, video.... i use it for daily Jan 13 18:48:05 with 16.04, halium-5.1 boot, graphic ok, wifi ok, ril ok, but no keyboard (when i try put passwd lock screen for example Jan 13 18:48:38 with 16.04 halium-7.1 no graphic and reboot every 60 seconds, ssh work Jan 13 19:16:55 reboot every 60 seconds sounds like this: https://github.com/ubports/docs.ubports.com/pull/67/files#diff-39dfecc00ea9bda5336efc08e3f79b9fL69 Jan 13 19:29:04 Walid: ubports rootfs sux, try pm Jan 13 19:33:30 yes PM work well with halium-7.1 its boot, graphics work, ssh, wifi, touch screen, keyboard Jan 13 19:34:52 I know, thats because ubports rootfs sux Jan 13 19:35:02 If pm works youre done Jan 13 19:35:42 anybody tried getting sailfishos to work on halium? Jan 13 19:35:53 No Jan 13 19:42:15 @i_love_halium_on, This. Jan 13 19:45:41 @trivial_inanity, How do I find out if the listed device is really the phone, or some wifi card? Jan 13 19:48:47 strange error … [ 73.599250,0] pil-q6v5-mss fc880000.qcom,mss: mba: Failed to locate mba.mdt … but i verified /firmware/image partition is mounted and contains mba.mdt Jan 13 20:01:02 so my issue in halium-7.1 is /firmware not well symlinked, anyone have any idea how to fix it, maybe in pre-start.sh Jan 13 20:03:11 @JBBgameich, `-tc` sorts to latest inode creation (iirc) Jan 13 20:07:43 ah thanks. That could be better than what I'm using currently, but I'm still looking for a way to identify the device without relying on the time when the device was attached Jan 13 20:08:22 Uh, the device gets added on a reboot? Jan 13 20:09:07 You mean when the device is rebooted? Yeah, it's added then Jan 13 20:39:59 (Photo, 960x1280) https://irc.ubports.com/2Da3H30q/file_3755.jpg Jan 13 20:40:08 I'm happy as... As... Jan 13 20:40:12 AAAAAAAAAAAAA Jan 13 20:40:16 wooo :D Jan 13 20:40:18 (Document) https://irc.ubports.com/pDvaqUAU/file_3757 Jan 13 20:40:30 @ilyaishere congrats! поздравляю! Jan 13 20:40:39 congratulations! Jan 13 20:41:05 What was the fix? I followed your issue on libhybris/libhybris Jan 13 20:41:26 @NotKit is genius Jan 13 20:42:09 I didn't do anything actually Jan 13 20:42:12 I manually specified the library path to libGLESv2 with an env variable Jan 13 20:42:17 what were your problems? Jan 13 20:42:30 He pointed me to the right place in the sources... Jan 13 20:49:11 https://github.com/abhishek9650/docs/commit/99eb97c80ce0d086942dfd2daa07965a27eae072 Jan 13 20:49:34 added dependencies for arch Jan 13 20:54:07 @ErnyTech, A long, long path through blaming Intel, PowerVR, Libhybris and masons altogether Jan 13 20:54:53 I had to manually specify an env variable to point to the GLESv2 lib Jan 13 20:55:14 @ilyaishere, how? Jan 13 21:00:16 KuranKaname was added by: KuranKaname Jan 13 21:00:21 @KuranKaname, welcome Jan 13 21:00:27 Hello there! Jan 13 21:00:34 Better ask here 🏃‍♂🏃‍♂🏃‍♂ Jan 13 21:00:40 Guess so Jan 13 21:01:29 @ErnyTech, LIBGLESv2=/full/path/to/libGLESv2-proprietary.so Jan 13 21:01:41 Guys I'm trying to get halium on the le max 2. It build but when I try to boot it, it doesn't get past the boot logo, only reboots in a loop. This is the log https://hastebin.com/ohanicefal.md Jan 13 21:02:47 @shenoy13 was trying lolz Jan 13 21:03:34 I know Jan 13 21:04:10 I'm just intrigued that you're here Jan 13 21:04:27 @ilyaishere, useful Jan 13 21:04:53 @Rk585, Yes no success.. Did sfos instead Jan 13 21:05:22 Does the cpu gov matter? Jan 13 21:05:33 Sed Jan 13 21:05:53 @KuranKaname: audit=n in defconfig, looks like selinux is active Jan 13 21:06:21 @revertive, CONFIG_AUDIT=n Jan 13 21:06:24 It's already like that Jan 13 21:07:56 If it makes a difference, I compiled the rom with uber 6 Jan 13 21:08:23 add "audit=0 selinux=0" to your cmdline and check if selinux is disabled with this Jan 13 21:09:08 @revertive, It had androidboot.selinux=permissive but I'll add those too, thanks Jan 13 21:10:30 But does cpu target and cpu gov affect this? I have both cpu variants on kryo. I'm compiling with a tree that's not mine as mine has blu_active gov as default, is it fine with it? Jan 13 21:11:00 @KuranKaname, As I saw … <4>[ 0.001666] /cpus/cpu@0: Unknown CPU type … <4>[ 0.001674] /cpus/cpu@1: Unknown CPU type … <4>[ 0.001681] /cpus/cpu@100: Unknown CPU type … <4>[ 0.001689] /cpus/cpu@101: Unknown CPU type Jan 13 21:12:27 @KuranKaname, are you sure the log is from your kernel? Jan 13 21:12:31 Linux version 3.18.20-perf-g5f9775f (letv@ubuntu) (gcc version 4.9.x-google 20140827 (prerelease) (GCC) ) #1 SMP PREEMPT Tue Jun 7 02:42:21 CST 2016 Jan 13 21:12:39 Jun 2016 Jan 13 21:12:45 is this not your recovery? Jan 13 21:13:34 @doniks, The log is last_kmsg grabbed with adb. I was a bit baffled by that too Jan 13 21:13:51 @KuranKaname, delete them and reboot Jan 13 21:14:21 Alright Jan 13 21:14:59 do you get any output at all via the lsusb iSerial? Jan 13 21:17:00 @doniks, Let me try Jan 13 21:23:05 @doniks, iSerial 3 (error) … iSerial 1 0000:02:00.0 … iSerial 1 0000:08:00.3 Jan 13 21:24:27 @KuranKaname check if your kernel configs have this … CONFIG_ANDROID_RAM_CONSOLE=y … CONFIG_ANDROID_RAM_CONSOLE_ENABLE_VERBOSE=y Jan 13 21:24:28 Deleted old kmsg Jan 13 21:24:38 @ErnyTech, Alright Jan 13 21:24:42 New kmsg is the same https://hastebin.com/ebesalefuh.md Jan 13 21:24:48 can you get telnet? Jan 13 21:25:08 @doniks, Nope, just throws "error" then reboots Jan 13 21:25:26 @KuranKaname, Also, kernel is 3.18.20. It should be 3.18.87 Jan 13 21:25:49 try hybris-recovery.img instead of hybris-boot.img that should get you into telnet Jan 13 21:25:59 Alright Jan 13 21:26:14 @KuranKaname, add it if missing+ Jan 13 21:26:56 I'll check on that and also build again Jan 13 21:27:16 @KuranKaname, it missing? Jan 13 21:29:40 @ErnyTech, Yep Jan 13 21:29:55 Someone also pointed out "atmel_mxt_ts 12-004a: Failed to initialize device" Jan 13 21:30:31 @KuranKaname, ok just add it for generate real kmsg Jan 13 21:31:34 Ok, I'll build again Jan 13 21:31:56 I'm quite convinced nothing you see in the log you posted helps you with your halium port, because I think it's from your recovery Jan 13 21:32:06 for some people the last_kmsg thing seems to not work Jan 13 21:32:53 @doniks, last_kmsg depends on the configs I posted above Jan 13 21:33:21 @ernytech sku sku Jan 13 21:33:33 ah ok, good to know Jan 13 21:49:07 i think my device merges the recovery stuff with the last_kmsg Jan 13 21:57:57 @Tigerpro, tigerpro in theory it should not do it Jan 13 22:02:52 When you reply to an IRC user, they are mentioned by the bot Jan 13 22:04:37 @KuranKaname: check is there is a log in /sys/fs/pstore, it could be your kernel isn't using last_kmsg Jan 13 22:05:20 @revertive, I'll check there thank you. New build is at 99% atm Jan 13 22:38:08 Make sure printk is set to y and a few other variables can be important. Otherwise if nothing else helps try enabling framebuffer console Jan 13 22:38:25 Which can be a pain, too Jan 13 22:38:35 @revertive, Nothing there. last_kmsg is the same deal. hybris_recovery doesnt work: https://hastebin.com/uxoqecexom.cpp Jan 13 22:39:09 @Flohack, CONFIG_PRINTK=y Jan 13 22:43:52 Hmm it seems that maybe the kernel is not started at all. How fast is the reboot happening Jan 13 22:44:34 Try abootimg -i hybris-....img and paste the result Jan 13 22:44:45 @Flohack, Doesn't get past the boot logo Jan 13 22:45:26 @KuranKaname, Thats no indication, most early boots dont show anything on the screen Jan 13 22:45:34 Early ports sry Jan 13 22:46:12 @Flohack, In time it's like 6 seconds. Then reboots and loops this Jan 13 22:46:33 Graphics might be the last you get working Jan 13 22:46:58 First I'd like it to not reboot Jan 13 22:47:42 @KuranKaname, Hm maybe the boot image has wrong offsets. As i said before show us abootimg -i output Jan 13 22:47:53 Doing it now Jan 13 22:48:23 @Flohack, https://hastebin.com/ejexohibac.md Jan 13 22:48:40 Can you compare it with e.g. Twrp recovery? Jan 13 22:49:31 @Flohack, https://hastebin.com/gowifoyese.md Jan 13 22:50:26 Jesus twrp kernel is huge Jan 13 22:50:49 Offsets are the same it seems Jan 13 22:50:50 Ye Jan 13 22:50:54 So is cmdline Jan 13 22:51:34 can you show it for hybris-recovery? Jan 13 22:51:40 You base on a cm/ lineage kernel? Jan 13 22:52:18 @doniks, https://hastebin.com/ikimosevaj.md Jan 13 22:52:23 @Flohack, Yes, it is Jan 13 22:52:40 Upstreamed to the latest version + latest caf tag and featured added Jan 13 22:52:44 On a normal rom it boots Jan 13 22:53:34 Hmm lineage kernels should boot in general. Early crash can be that it cant mount root fs Jan 13 22:54:38 Why would that be tho Jan 13 22:55:01 I could try to change device and vendor trees or another kernel Jan 13 22:55:33 I don't understand how hybris-recovery can work LESS than hybris-boot Jan 13 22:55:43 does hybris-recovery also bootloop? Jan 13 22:58:09 @doniks, Yes Jan 13 22:58:16 And telnet gives the same error Jan 13 22:58:17 @doniks, My hybris recovery never started network xD Jan 13 22:58:30 But boot is fine Jan 13 22:58:59 Sorry sleeping time here good night Jan 13 22:59:26 but isn't the ONLY difference that it adds this kernel param Jan 13 22:59:29 bootmode=debug Jan 13 22:59:42 @Flohack, Good night Jan 13 23:00:02 @KuranKaname, you could try to add Jan 13 23:00:04 bootmode=debug Jan 13 23:00:08 at the end of the line Jan 13 23:00:15 HYBRIS_BOOTIMG_COMMANDLINE := Jan 13 23:00:41 in halium/hybris-boot/Android.mk Jan 13 23:00:51 then rebuild hybris-boot and boot again Jan 13 23:01:01 NOW it really should go into telnet and not bootloop Jan 13 23:01:02 But wouldn't that make it the same with the normal boot img? Jan 13 23:02:01 https://github.com/Halium/hybris-boot/blob/master/init-script#L317 Jan 13 23:02:12 at this point it should turn on telnet Jan 13 23:02:19 if that param is present Jan 13 23:02:47 or one line below if you have set ALWAYSDEBUG := 1 in Android.mk Jan 13 23:03:27 I always thought normal and recovery are only different in that way, but that assumption doesn't hold if you say that recovery bootloops Jan 13 23:04:19 Ok brb Jan 13 23:08:31 @doniks, HYBRIS_BOOTIMG_COMMANDLINE := … HYBRIS_RECOVERYIMG_COMMANDLINE := bootmode=debug Jan 13 23:09:13 yeah Jan 13 23:09:19 copy the param o the normal one Jan 13 23:09:48 to Jan 13 23:10:02 do you know what I mean? Jan 13 23:10:09 Yeah I did Jan 13 23:10:52 @doniks, Also added this HYBRIS_B_ALWAYSDEBUG := 1 Jan 13 23:11:04 you might as well Jan 13 23:11:12 I think it doesn't matter Jan 13 23:11:17 either one would do the trick Jan 13 23:16:17 I'm trying to build UBports Jan 13 23:16:19 https://docs.ubports.com/en/latest/porting/building-ubports-boot.html#include-in-build-system Jan 13 23:16:27 it says make it look like: Jan 13 23:16:35 subdirs += \ Jan 13 23:16:37 halium/hybris-boot \ Jan 13 23:16:39 halium/ubports-boot Jan 13 23:16:44 currently it looks like this for me: Jan 13 23:16:52 subdirs += \ Jan 13 23:16:54 halium/hybris-boot \ Jan 13 23:16:56 halium/droidmedia Jan 13 23:17:09 Put ubports-boot at the end of the list Jan 13 23:17:13 does anybody know whether they ACUTALLY mean that I should remove droidmedia Jan 13 23:17:19 no Jan 13 23:17:22 ah ok Jan 13 23:17:56 thx Jan 13 23:19:49 how does one activate the framebuffer console? Jan 13 23:20:44 It's unlikely that the framebuffer console works on your device Jan 13 23:21:25 note 4? ill set console=tty1 just to see what happens Jan 13 23:21:49 or is it tty0? Jan 13 23:21:57 It's different for any device Jan 13 23:21:59 @Tigerpro, my personal attempt was to do this: https://forum.xda-developers.com/showpost.php?p=10075806&postcount=3 Jan 13 23:22:08 then stare at kernel errors for days Jan 13 23:22:12 then give up :( Jan 13 23:22:36 itd give me some insight to see where my device hangs at least Jan 13 23:23:40 last_kmsg doesn't work for you? Jan 13 23:25:16 it doesnt show where the issue is from what I can tell Jan 13 23:25:49 mhm, but framebuffer console wouldn't show you more Jan 13 23:25:56 just on a smaller screen Jan 13 23:26:36 well it might tell me if theres a kernel panic that doesnt seem to be getting logged Jan 13 23:26:55 can you post the current one Jan 13 23:27:33 there is a template to install a rootfs without a PC Jan 13 23:28:03 I gotta put my latest compile on my device first, I had a hdd crash (which may have been a factor) and just compiled today Jan 13 23:28:27 sure Jan 13 23:28:34 @KuranKaname, how's it going? Jan 13 23:29:41 I am thinking whatever was causing the issue (assuming I have an actual issue) might be getting lost in the log, if its even logging Jan 13 23:29:59 @doniks, Just booted Jan 13 23:30:01 Telnet works Jan 13 23:30:11 tadaaa Jan 13 23:30:12 Welcome to the Mer/SailfishOS Boat loader debug init system. Jan 13 23:30:30 is rootfs mounted? Jan 13 23:30:43 @doniks, Where to check that? Jan 13 23:30:52 post the output of Jan 13 23:30:54 mount Jan 13 23:30:57 it could just be the RNDIS isnt initializing, either way, it should help me get past my current problem and grats KuranKaname Jan 13 23:31:08 @Tigerpro, Thanks! Jan 13 23:31:17 @doniks, rootfs on / type rootfs (rw,size=1820008k,nr_inodes=455002) … devtmpfs on /dev type devtmpfs (rw,relatime,size=1820008k,nr_inodes=455002,mode=755) … devpts on /dev/pts type devpts (rw,relatime,mode=600) … sysfs on /sys type sysfs (rw,relatime) … proc on /proc type proc (rw,relatime) … /dev/block/sda9 on /data type ext4 (rw,relatime,data=ordered) … /dev/block/loop0 on /target type ext2 (rw,relatime) … /dev/block/sda9 on /ta Jan 13 23:31:18 (rw,relatime,data=ordered) Jan 13 23:31:32 i figured out rootfs seems to get mounted in /target Jan 13 23:31:33 ... dev block sda9 Jan 13 23:31:36 uhhh with some pastebin service please Jan 13 23:31:45 Sure Jan 13 23:31:56 @doniks, https://hastebin.com/ewohoxuwip.hs Jan 13 23:32:00 I thought it was rather small Jan 13 23:32:12 loop0 on /target Jan 13 23:32:12 @UniversalSuperBox, What's with that? Jan 13 23:32:18 ls /target Jan 13 23:32:20 ? Jan 13 23:32:27 @KuranKaname, Your manufacturer though they were funny Jan 13 23:32:41 @doniks, https://hastebin.com/pihivatavo.cs Jan 13 23:32:44 @UniversalSuperBox, lol Jan 13 23:32:48 well my rootfs was appearing in target when I last messed with this Jan 13 23:32:52 looks normal Jan 13 23:33:26 which rootfs is it? Jan 13 23:34:02 maybe disabling the lxc container Jan 13 23:34:03 I tried halium, ubports, and plasma, all appeared in /target in pre-boot Jan 13 23:34:07 how did that work again? Jan 13 23:34:14 @doniks, halium-rootfs-20170630-151006 Jan 13 23:34:52 iirc @doniks it made no difference from what I could tell Jan 13 23:36:08 om I gotta go, gotta make dinner Jan 13 23:36:41 What should I try? Jan 13 23:38:20 @KuranKaname, try this https://hastebin.com/icoyetesic.rb Jan 13 23:38:31 I think (tm) :) Jan 13 23:38:37 is half from memory :) Jan 13 23:40:26 @doniks, iSerial 3 GNU/Linux Device on rndis0 10.15.19.82 Jan 13 23:40:31 So it seems it booted Jan 13 23:40:35 nice Jan 13 23:40:50 yes that is the init system INSIDE the rootfs now talking Jan 13 23:40:55 you can get ssh access Jan 13 23:41:15 Any way to make that changes permanent? Jan 13 23:41:37 well, for the time being that change IS permanent Jan 13 23:41:43 Alright Jan 13 23:41:58 however, you want to find out why starting the lxc container resulted in a bootloop Jan 13 23:42:12 without the lxc container you're not gonna have much fun with your port Jan 13 23:42:17 well, you have ssh :) Jan 13 23:42:26 so, SOME fun can be had Jan 13 23:43:03 but, I'm kinda reaching the end of my knowledge also, I'm afraid Jan 13 23:43:09 I guess you can ssh in Jan 13 23:43:15 and try to start the lxc Jan 13 23:43:23 @doniks, I bet Jan 13 23:43:35 but likely that will just reboot, without necessarily telling you more about WHY Jan 13 23:44:04 do you have ssh? Jan 13 23:45:26 Getting it now Jan 13 23:47:29 It doesn't want to recognize adb now Jan 13 23:47:30 lol Jan 13 23:47:51 yeah I think it doesn't do adb Jan 13 23:48:09 Doesn't work even in recovery Jan 13 23:48:15 oh Jan 13 23:48:26 I'll reboot Jan 13 23:48:29 strange Jan 13 23:48:51 you can also set the password or ssh key in the chroot from telnet Jan 13 23:55:35 @doniks, Thanks for you help man. Tho I can't seem to be able to get ssh running Jan 13 23:55:39 It doesn't give any output Jan 13 23:56:03 you are back at Jan 13 23:56:10 iSerial 3 GNU/Linux Device on rndis0 10.15.19.82 Jan 13 23:56:12 ? Jan 13 23:56:17 Yeah Jan 13 23:56:25 ssh root@10.15.19.82 Jan 13 23:56:28 Just hangs Jan 13 23:56:57 yeah it will look like this that's ok for now Jan 13 23:57:21 ssh: connect to host 10.15.19.82 port 22: Connection timed out Jan 13 23:57:53 ip addr Jan 13 23:58:43 @doniks, 5: enp2s0f0u4: mtu 1500 qdisc pfifo_fast state UNKNOWN group default qlen 1000 … link/ether 4e:70:d1:1a:bc:5b brd ff:ff:ff:ff:ff:ff Jan 13 23:59:15 ok Jan 13 23:59:17 good Jan 13 23:59:23 now you need to do this again Jan 13 23:59:25 https://docs.halium.org/en/latest/porting/debug-build/early-init.html#debugging-via-telnet Jan 13 23:59:30 but now for the 10.... numbers Jan 14 00:03:44 smt like this: Jan 14 00:03:46 sudo ip address add 10.15.19.1 dev enp0s20f0u7 Jan 14 00:03:48 ip address show dev enp0s20f0u7 Jan 14 00:03:50 sudo ip route add 10.15.19.82 dev enp0s20f0u7 Jan 14 00:03:52 ping -c 2 10.15.19.82 Jan 14 00:04:03 but using enp2s0f0u4 Jan 14 00:10:24 @doniks, If I do that I get https://hastebin.com/dofuruvuto.vbs Jan 14 00:10:37 But If i do: sudo ip address add 10.15.19.82 dev enp0s20f0u7 Jan 14 00:10:42 It let's me try to connect Jan 14 00:10:49 But Permission denied, please try again. Jan 14 00:11:25 well, the WARNING: REMOTE HOST IDENTIFICATION HAS CHANGED! Jan 14 00:11:29 is not a problem Jan 14 00:11:49 I guess previously you had some attempt of connecting to a port Jan 14 00:12:10 but now there is a different ssh server key in the rootfs Jan 14 00:12:27 Did this: ssh-keygen -f "/root/.ssh/known_hosts" -R 10.15.19.82 Jan 14 00:12:31 good Jan 14 00:12:33 But it just hangs again Jan 14 00:12:51 ssh root@10.15.19.82 Jan 14 00:13:09 nvm Jan 14 00:13:11 Got it Jan 14 00:13:13 I'm in Jan 14 00:13:15 cool Jan 14 00:13:51 @doniks, https://hastebin.com/zuyuremanu.rb Jan 14 00:15:14 yeah dunno Jan 14 00:15:27 oh maybe here we find what caused the reboot Jan 14 00:15:31 lets do this first: Jan 14 00:15:34 test_lights Jan 14 00:15:36 I got a dmesg Jan 14 00:15:38 EGL_PLATFORM=hwcomposer test_hwcomposer Jan 14 00:15:54 @doniks, It light up the notifications led Jan 14 00:16:00 :) Jan 14 00:16:17 @doniks, This makes it flash Jan 14 00:16:54 (Document) https://irc.ubports.com/pDvaqUAU/file_3757 Jan 14 00:16:57 like this? Jan 14 00:17:29 @doniks, Nop I'm still at the boot logo Jan 14 00:17:35 It made the notifications led flash Jan 14 00:17:42 really? Jan 14 00:17:53 test_hwcomposer made the led flash? Jan 14 00:18:01 Yeah Jan 14 00:18:05 weirdd Jan 14 00:18:19 ok, no spinning spiral? Jan 14 00:18:29 Nope, only boot logo Jan 14 00:18:33 ha Jan 14 00:18:50 yeah ok, i guess then the lxc check you posted is next Jan 14 00:18:59 oh you could also try wifi first Jan 14 00:19:57 Logcat is messed up Jan 14 00:20:03 this is mine: https://paste.ubuntu.com/26381532/ Jan 14 00:21:06 @doniks, Should I enable control groups? Jan 14 00:21:12 Also, this is the logcat Jan 14 00:21:13 https://hastebin.com/guxiqofebe.bash Jan 14 00:21:52 @KuranKaname, Actually they are enabled Jan 14 00:21:56 But I don't get … Cgroup v1 mount points: … /sys/fs/cgroup/systemd … /sys/fs/cgroup/hugetlb … /sys/fs/cgroup/devices Jan 14 00:23:05 ugh that IS messed up Jan 14 00:23:17 I ran Jan 14 00:23:21 @KuranKaname, I don't get any of those like yours Jan 14 00:23:22 lxc-checkconfig on my laptop Jan 14 00:23:32 maybe it has a bit of a different version/layout Jan 14 00:23:39 not sure Jan 14 00:23:46 Oh ok Jan 14 00:23:55 from here on it's a bit unchartered territory for me Jan 14 00:23:56 @KuranKaname, This doesn't really look fine tho Jan 14 00:24:18 I guess some amount of kernel config enabling and retrying is the next step Jan 14 00:24:46 Yeah Jan 14 00:24:48 Thank you for your help Jan 14 00:25:40 oh wait haha Jan 14 00:25:53 I posted the lxc check of my laptops config Jan 14 00:25:56 hang on Jan 14 00:26:54 https://paste.ubuntu.com/26381556/ Jan 14 00:28:18 Oh alright Jan 14 00:28:34 But still I don't have any mount points Jan 14 00:29:25 lxc-checkconfig output is kernel version dependant Jan 14 00:29:38 (or possibly lxc version dependant) Jan 14 00:30:02 In any case you should fix all the missing things in your lxc-checkconfig except user namespaces Jan 14 00:30:12 Alright Jan 14 00:30:14 Thank you Jan 14 00:31:39 @bhushanshah, really? all of them? I have plenty missing but the container starts for me Jan 14 00:32:06 Well at least top two sections needs to be fixed Jan 14 00:32:51 (in this one https://hastebin.com/zuyuremanu.rb) Jan 14 00:33:24 btw, bhushanshah, I think I added this : Jan 14 00:33:26 User namespace: missing Jan 14 00:33:39 (dammit, not that, wait) Jan 14 00:33:59 cgroup_disable=memory Jan 14 00:34:05 based on a tip from you Jan 14 00:34:11 but I can't remember why Jan 14 00:34:14 do you know? Jan 14 00:34:25 I'll have to grep logs Jan 14 00:34:35 Which I can't do till Tuesday Jan 14 00:34:38 could it have anything to do with that KuranKaname is working on Jan 14 00:34:43 oh never mind Jan 14 00:35:01 Not really no.. your issue was different Jan 14 00:35:03 I just thought if off the top of you head it might help Jan 14 00:35:42 @KuranKaname you need to enable CONFIG_DEVPTS_MULTIPLE_INSTANCES Jan 14 00:36:01 That is bare minimum you'll need to get container started imo Jan 14 00:36:26 @bhushanshah, Ok will do that Jan 14 00:36:35 @doniks, Do I need this too? Jan 14 00:36:40 No Jan 14 00:37:13 doniks: you're up for quick test regarding plasma? Jan 14 00:37:25 @bhushanshah, Oke. Thanks Jan 14 00:40:01 sure Jan 14 00:40:18 but you're having your weekendrest remember :) Jan 14 00:40:33 I did a bit of image bisecting Jan 14 00:41:11 pm-rootfs-20171113-152515.tar.gz worked ok Jan 14 00:41:22 pm-rootfs-20171210-140255.tar.gz froze Jan 14 00:41:37 but the lock screen itself is always ok Jan 14 00:41:48 and if there is a kwallet dialog then THAT one is also ok Jan 14 00:42:08 So 1210 is first broken image? Jan 14 00:42:09 "ok" means - reacts to touch input Jan 14 00:42:14 yes Jan 14 00:42:31 Uhhh okay meh Jan 14 00:44:08 Anyway from non working image, if you "disable" network-manager systemd service and reboot.. does it work properly? Jan 14 00:44:47 I can try Jan 14 00:44:56 And yes re: kwallet, I know, it's just that plasmashell is broken.. nothing else Jan 14 00:44:58 need to flash 1210 again first Jan 14 00:45:08 And oh also Jan 14 00:45:44 have you tried waiting long enough? I know this is not solution lol but there might be dbus time out at work maybe Jan 14 00:46:18 well depends on the definition of enough :) Jan 14 00:46:28 I waited until the lock screen returns Jan 14 00:46:33 which is again fine Jan 14 00:46:36 unlock Jan 14 00:46:38 not fine Jan 14 00:46:50 Okay that is enough Jan 14 00:47:22 120(?) seconds is default dbus timeout iirc Jan 14 00:47:32 coincidentially I think 1210 is the image that stopped saying "oh it's gorgeous" Jan 14 00:47:52 :) I really hope that this is unrelated and we don't need to reintroduce that message ;) Jan 14 00:48:03 yeah I waited 120 sec Jan 14 00:48:36 1210 is also image which shipped proper pulseaudio config hmm Jan 14 00:50:27 1210 image also had too many changes to bisect really Jan 14 00:50:28 fml Jan 14 00:52:37 oh btw have you seen this: Jan 14 00:52:38 Due to leg surgery and recovery leave I didn't really work during that period iirc... And changes piled up and now it's just impossible to debug Jan 14 00:52:39 https://github.com/Halium/hybris-boot/pull/34 Jan 14 00:52:53 It's a bit of a wild mix of things, but curious to hear what you think Jan 14 00:52:58 Yes I got email.. still unread though Jan 14 00:53:04 sure Jan 14 00:53:28 logging initrd to kernel log Jan 14 00:56:18 ok, pm is now frozen for more that 120 sec .... Jan 14 00:56:38 systemctl stop networking.service Jan 14 00:56:41 this? Jan 14 00:57:01 systemctl stop NetworkManager.service Jan 14 00:57:03 or that? Jan 14 00:57:51 Not stop but disable network manager Jan 14 00:58:32 systemclt disable Networkmanager ; systemclt mask NetworkManager ; reboot Jan 14 00:58:34 this? Jan 14 00:59:36 Yeah Jan 14 00:59:54 Except *systemctl 😛 Jan 14 01:00:31 =) Jan 14 01:00:34 rebooting Jan 14 01:12:10 no, just as frozen as before, also with disabled NM Jan 14 01:13:10 Okay Jan 14 01:14:02 Let's continue this on Tuesday.. need to go off now Jan 14 01:14:05 :D Jan 14 01:14:14 sure Jan 14 01:14:29 I mean, we continue another time Jan 14 01:14:38 I'm probably not around all week next week Jan 14 01:15:04 Ok sure Jan 14 01:15:12 @KuranKaname, you still around? hows it going? **** ENDING LOGGING AT Sun Jan 14 03:00:03 2018