**** BEGIN LOGGING AT Wed Feb 14 03:00:02 2018 Feb 14 04:55:53 I've fixed the remaining kernel errors, but now telnet no longer works Feb 14 04:55:58 I get this error: https://hastebin.com/ebehijuhic.vbs Feb 14 04:56:12 It says "link is not ready" Feb 14 04:56:33 anyone know why? Feb 14 04:57:12 "[33107.636075] usb 1-1.4: SerialNumber: GNU/Linux Device on usb0 10.15.19.82" Feb 14 04:57:17 congrats, now you're in rootfs Feb 14 04:57:33 assign it IP of 10.15.19.100 and then ssh as root@10.15.19.82 Feb 14 05:06:13 What do you mean by assign it an IP? Feb 14 05:06:34 The guide doesn't say much Feb 14 05:07:32 with the "ip" or "ifconfig" command Feb 14 05:08:24 Oh, whoops Feb 14 05:11:58 Got past it, thanks Feb 14 05:12:10 Working through the tests now Feb 14 05:12:51 it might be too early for tests maybe Feb 14 05:12:56 check container status first Feb 14 05:13:03 systemctl status lxc@android Feb 14 05:14:28 Well, it failed Feb 14 05:14:40 https://hastebin.com/esetuparaj.sql Feb 14 05:14:56 alright Feb 14 05:15:08 run "lxc-checkconfig" Feb 14 05:16:59 that will tell you what kernel config options are missing, you need to fix all of them except User namespaces Feb 14 05:17:05 https://hastebin.com/akitagisub.sql Feb 14 05:17:23 only the required? Feb 14 05:17:30 Or the missing as well? Feb 14 05:18:01 missing as well Feb 14 05:18:12 Okay, thanks Feb 14 05:18:25 Btw, what device it is? and what kernel version? Feb 14 05:20:19 jfltetmo (Galaxy S4 T-Mobile) on kernel 3.4.112 Feb 14 05:20:30 Using Halium 7.1 Feb 14 05:20:34 Okay Feb 14 05:51:28 bshah: Found this PR from you that migth solve my hwcomposer issue on mido ;) https://github.com/Halium/android_system_core/pull/1 Feb 14 05:51:48 I.. don't really think that PR touches hwcomposer tbh Feb 14 05:51:51 Since I'm not having android_property_service that might cause the hwcomposer issue according to the sfos guys Feb 14 05:52:00 Well worth a try at least Feb 14 05:52:31 It won't harm having it, since i'm sure it's needed elsewhere too ;) Feb 14 05:52:45 yepyep.. like calls e.g ;) Feb 14 05:53:14 So putting a new image together now :) Feb 14 05:53:14 I'm holding on merging it given it haven't recieved enough testing ye Feb 14 05:53:17 *yet Feb 14 05:53:54 OK Feb 14 05:59:38 @bshah, I tested it with MTK Feb 14 06:00:02 .. and? Feb 14 06:00:04 :P Feb 14 06:16:00 bshah: Success! Feb 14 06:16:22 Well I got test_hwcomposer to show spiral and then reboot :P Feb 14 06:16:28 But that's progress! Feb 14 06:16:54 hurray Feb 14 06:17:16 bshah: It could also be my firmware that I need to fix up on this new image Feb 14 06:17:25 Yeah seems it's bootlooping in general after x time Feb 14 06:17:29 THat's the firmware Feb 14 06:18:11 It's a matter of pushing that across Feb 14 06:18:18 And fixing it properly going forward Feb 14 06:23:28 But that was solved by putting firmware in teh right place previously Feb 14 06:31:35 Guys, I think I finally solved the error I was stuck :) Feb 14 06:32:48 I isntaled and used arm-linux-gnueabihf-gcc instead of default gcc, and I now have another different error Feb 14 06:34:55 I hope I do not have to see it anymore Feb 14 06:40:40 When building to test changes in lxc-checkconfig, do I need to rebuild both the system image and the boot image? Or just the boot image? Feb 14 06:41:56 guys, I get 'fatal error: zconf.hash.c: No such file or directory … #include "zconf.hash.c"' … Should I comment that? Feb 14 06:42:26 I mean, is that something important? Feb 14 06:43:33 Pablo Alejandro Do a "mka clean" and you should be good Feb 14 06:43:40 That always works for me Feb 14 06:44:38 I'll try that, but however I did make clean in builddir and make mrproper in kernel directory, thanks Feb 14 06:46:57 oh, but that did the job, thanks :) Feb 14 06:47:28 No problem :) Feb 14 10:12:02 Texasfoldsem was added by: Texasfoldsem Feb 14 10:20:38 You guys are a blessing and I want to talk finances you need in turn just to have a chance to show my greatest asset, which is concepts and ideas . This skill out very much a help to many in the past and is still a constant brain mover; luckily because my personal feats have shrunk due to disabilities. I want to do all I can to Feb 14 10:20:38 drop the current crap tactics of Google and a list of future ones that makes me desire to rename them GoogleGestapo. Either way I wish you guys the best and hope I can help in a financial way and tiny idea via email possibility. I have a list of improvements on Android already. Feb 14 10:28:59 Text me on telegram or six8280two53four7 Feb 14 12:34:35 test_gps sais "GPS interface not found :(". And if I strace it seams that it can not find the gps.default.so lib. There is only a lib in the lib64 path and if I link that "/android/system/lib64/hw/gps.default.so" is 64-bit instead of 32-bit". Do I have to do something special to get the 32bit library compiled in the systems.img? Feb 14 12:39:09 ZeroPointEnergy: I do remember that there is option for building android in dual mode (32 bit+64 bit) but I don't know it off-hand Feb 14 12:39:41 ZeroPointEnergy: but that said, don't you have /system/lib/hw/ ? Feb 14 12:41:36 Yes I do. Most libraries seam to be there, except the gps library which is only in lib64 Feb 14 12:46:54 is gps library blob? Feb 14 12:47:08 i.e provided by TheMuppets? Feb 14 12:50:23 Yes, you are right. And there is only a 64bit version Feb 14 12:52:00 bad luck then.. Feb 14 12:52:02 :/ Feb 14 12:52:30 so unless at some point we build arm64 rootfs for halium, I doubt gps test would work with armhf Feb 14 12:54:56 there seam to be similar issues with other tests test_camera => "library "libcamera_compat_layer.so" not found" Feb 14 12:55:13 so you think a 64bit rootfs would help? Feb 14 12:56:05 how was the 32bit rootfs built? Feb 14 12:56:49 about test_camera it is broken anyway Feb 14 12:57:00 (not related to 64 bit0 Feb 14 12:57:09 what you can try is, now ubports rootfs btw Feb 14 12:57:34 since hwcomposer works, ubports should work Feb 14 12:57:39 and as well plasma mobile :) Feb 14 12:58:11 hwcomposer does not work unfortunatly :-b Feb 14 12:58:55 I erm confusing you with someone else it looks? Feb 14 12:59:31 test_hwcomposer: test_hwcomposer.cpp:307: int main(int, char**): Assertion `eglGetError() == EGL_SUCCESS' failed. Feb 14 13:00:31 no error in /system/bin/logcat Feb 14 13:00:59 time to fix it first :) Feb 14 13:01:06 can you strace test_hwcomposer? Feb 14 13:03:05 https://paste.ubuntu.com/p/BhGNt3HrtH/ Feb 14 13:04:00 there should be something in logcat at least (even if no error but log) Feb 14 13:04:12 can you grab that info/ Feb 14 13:04:18 also systemctl status output please Feb 14 13:05:07 logcat https://paste.ubuntu.com/p/Tz9sww7Z7m/ Feb 14 13:06:41 systemctl https://paste.ubuntu.com/p/g4cWXtXsnQ/ Feb 14 13:13:52 @ZeroPointEnergy, Have you applied udev rules? Feb 14 13:16:55 yes, but it complains about some missing users and groups when they get applied Feb 14 13:19:41 Will reboot and see if I can extract the log Feb 14 13:26:07 this are the logs from the udev : https://paste.ubuntu.com/p/vRwzvBV5DF/ Feb 14 14:01:31 any hints how I have to debug this? Feb 14 14:10:02 test_egl_configs seams to work Feb 14 14:42:01 there are also some messages in dmsg if I run the test https://paste.ubuntu.com/p/C8h7JvC7s8/ Feb 14 20:52:42 ZeroPointEnergy: Not 100% sure what your issue is but I had hwcomposer_test failing as well yesterday Feb 14 20:52:48 While all seemed OK otherwise Feb 14 20:53:16 Turned out to be the missing android_property_service in my case with 7.1 on Xiaomi Redmi Note 4x (mido) Feb 14 20:54:33 I applied https://github.com/Halium/android_system_core/pull/1 and magically stuff started to work at my end for test_hwcomposer! Feb 14 23:09:41 I'm working through clearing the missing parts of lxc-checkconfig and I am stuck at here: https://pastebin.com/f206KGPm Feb 14 23:09:59 I can't seem to find the missing ones in my kernel Feb 14 23:10:30 My kernel source: https://github.com/Qiangong2/halium_kernel_samsung_jf Feb 14 23:12:08 What config would VLAN be under? Feb 14 23:12:23 and packet and netlink_diag Feb 15 00:22:58 As I'm trying to debug my device, it keeps disconnecting after about 30 seconds of connecting Feb 15 00:23:07 *over ssh Feb 15 00:31:08 I'm trying to debug my device too. I'm booting into the halium rootfs. It doesn't kernel panic (I don't see anything appear in /sys/fs/pstore/, I do have that kernel feature enabled), and ssh nor telnet work. Feb 15 00:31:24 I am not sure where to go at this point Feb 15 00:42:22 Just out of curiosity, which device are you working on? Feb 15 00:44:56 zte axon 7. You? Feb 15 00:45:35 Galaxy S4 T-Mobile Feb 15 00:53:17 my device does respond to ping requests but ssh and telnet don't work... I don't see any messages in journalctl when I boot into TWRP and chroot into rootfs.img Feb 15 00:54:15 Is the filesystem not mounted r/w at that point? I thought it was Feb 15 00:55:08 It should be, Feb 15 00:55:17 but if you're using halium-boot you need to create the right file Feb 15 00:55:23 the write file, if you will Feb 15 00:55:28 (sorry that was terrible) Feb 15 00:55:34 >_> :) Feb 15 00:56:43 I added … echo "Starting telnet" >> /testlog … sync … to my telnet sysvinit style startup script. It's added to run levels 2-5 Feb 15 00:56:48 but no file appears Feb 15 00:56:59 Do you have `/data/.writable_image` Feb 15 00:57:23 I don't think so.... Feb 15 00:57:35 Did I send you this? :) https://github.com/Halium/projectmanagement/issues/40 Feb 15 00:58:49 .... yes, you did. I had stopped looking at it when you told me to change a few files in the makefile to build it instead. I didn't realize there were more special install instructions.... Feb 15 01:04:25 ! Feb 15 01:04:48 Ok, still no ssh and telnet, but I do get that "testlog" file I create! Feb 15 01:05:53 thanks for sending that link again. Feb 15 01:26:19 https://gist.github.com/rockybulwinkle/58aa15b9360ba8381c859740310b9175 Feb 15 01:26:54 This is fine Feb 15 01:55:47 When I run test_egl, the only output I have is: "Starting test (EGL_PLATFORM=(null))" Feb 15 01:55:58 And nothing appears on screen Feb 15 01:56:02 is that normal? Feb 15 01:56:21 oh, never mind Feb 15 01:56:29 it says Segmentation Fault Feb 15 02:00:53 should I be concerned about the "no ifupdown configuration found" for rndis0, or does it get set 'up' somewhere else? Feb 15 02:07:48 when I did an "ip addr" inside my telnet init script it showed it was down. This is the USB ethernet right? Who sets it up? Feb 15 02:43:42 uhhhh so. I forgot how IP's work. I'm ssh'd in now. Feb 15 02:43:55 *facepalm* **** ENDING LOGGING AT Thu Feb 15 03:00:01 2018