**** BEGIN LOGGING AT Sun Jan 08 03:00:01 2017 Jan 08 08:33:57 monring Jan 08 08:40:54 nizovn: ping ;) Jan 08 08:42:03 morning Jan 08 08:42:06 Herrie: pong Jan 08 08:43:25 i may reply slowly because i'm currently compiling PC system update Jan 08 08:48:53 nizovn: OK, that's always a good thing to do over the holidays :D Jan 08 08:49:07 Was just trying to build Morty image for Tenderloin and ran into the following: Jan 08 08:49:46 https://bpaste.net/show/59ac70478e5f Jan 08 08:50:12 hehe Jan 08 09:27:55 Herrie: could you test https://github.com/nizovn/meta-smartphone/commit/998e500ba1f311a7affd793df87c6580e5790762 ? Jan 08 09:55:21 nizovn: Testing :) Jan 08 10:02:54 Another error now: https://bpaste.net/show/1c08b6aabb74 Jan 08 10:07:16 could you show directory name in /home/herrie/LuneOS/webos-ports-env/webos-ports/tmp-glibc/work/tenderloin-webos-linux-gnueabi/linux-hp-tenderloin/3.0.101-20161117-94-r0/image/lib/modules Jan 08 10:10:23 3.0.101-gc31ea94 Jan 08 10:10:40 ok, seems it's another issue Jan 08 10:20:35 could you please show /home/herrie/LuneOS/webos-ports-env/webos-ports/tmp-glibc/work/tenderloin-webos-linux-gnueabi/linux-hp-tenderloin/3.0.101-20161117-94-r0/temp/run.do_deploy ? Jan 08 10:30:33 nizovn: https://bpaste.net/show/47f7496a66ab Jan 08 10:56:16 Herrie: could you try this: https://bpaste.net/show/b176d5814f86 ? Jan 08 10:57:14 just remove ? after KERNEL_IMAGE_BASE_NAME Jan 08 11:25:43 Yup that seems to do the trick Jan 08 11:25:47 It's continuing build now Jan 08 11:25:51 ok good Jan 08 11:26:07 So far Mako & Maguro are bootlooping & qemux86 has funny image :S Jan 08 11:26:15 So it'll need some work still probably :S Jan 08 11:26:58 The bootlooping is likely systemd... systemd always gives headaches on Yocto updates it seems :S Jan 08 11:27:20 yes, maybe Jan 08 11:28:22 I Jan 08 11:28:33 I'll try to downgrade to 229 as a start Jan 08 11:28:50 Tenderloing should be almost done so will test that one Jan 08 11:28:56 ok Jan 08 12:12:39 Hmmz Tenderloin also bootlooping Jan 08 12:12:52 I'll try to downgrade systemd see if that works around it Jan 08 12:15:02 you can also see the output of klog in webOS Jan 08 12:15:45 nizovn: How? Jan 08 12:16:03 novaldex|away: You around? Jan 08 12:17:13 Herrie: just execute klog in novaterm/etc Jan 08 12:19:28 https://bpaste.net/show/313f84058985 Jan 08 12:21:32 Hmmz some kernel panic :S Jan 08 12:45:36 Herrie: can you share uImage, seem we should check initramfs Jan 08 12:50:10 nizovn: Let me upload it Jan 08 12:51:26 http://herrie.org/uImage-morty Jan 08 12:57:38 mkimage has strange output: Jan 08 12:57:40 GP Header: Size 27051956 LoadAddr 9ef46b0c Jan 08 13:48:58 Herrie: do you compile it on windows? it looks like \r\n sequence is stripped Jan 08 13:59:39 Herrie: could you show output of: MACHINE=tenderloin bitbake-layers show-recipes "u-boot*" Jan 08 15:21:03 nizovn: I compile it on my Linux box, then copy to my windows box and upload from there Jan 08 15:21:07 I think it's systemd Jan 08 15:21:16 Mako stopped bootlooping with systemd downgrade Jan 08 15:21:29 Just finished tenderloin build let me try that now Jan 08 15:22:42 ok Jan 08 15:26:09 systemd was causing issues in the past too on Mako when we updated Yocto Jan 08 15:37:47 Hmmz still bootlooping Jan 08 15:38:15 I always do: Build on native Ubuntu 16.04, then copy files to my Windows machine & update from there... Been doing this for a year without issues Jan 08 15:38:26 It could be the file I ftp-ed got messed up somehow? Jan 08 15:40:13 http://www.herrie.org/uImage-morty.zip Jan 08 15:45:48 this uImage has correct mkimage information at least Jan 08 15:52:42 OK Jan 08 15:52:50 I guess my FTP client messed it up while uploading then Jan 08 16:16:40 this may be related https://lists.yoctoproject.org/pipermail/yocto/2016-January/028241.html Jan 08 16:22:22 nizovn: Hmmz that would be hard.... Jan 08 16:41:19 I think we already went glibc 2.24 with Krogoth Jan 08 16:43:10 seems no, i think we should switch back to 2.23 Jan 08 16:59:51 OK will try that Jan 08 17:00:45 you'll have to rebuild everything... Jan 08 17:19:47 I know, nice task for overnight :P Jan 08 18:16:50 nizovn: There might be another way? " --enable-kernel=2.6.32 This tells Glibc to compile the library with support for 2.6.32 and later Linux kernels. Workarounds for older kernels are not enabled." Jan 08 18:17:25 not sure Jan 08 18:36:53 i afraid this won't work if --enable-kernel is older than supported Jan 08 18:43:54 Not sure what version TP is on? Jan 08 18:51:10 3.0.101 Jan 08 18:59:57 So that should work theoretically Jan 08 19:03:18 choorp: Hi ;) Jan 08 19:03:34 Heya Herrie Jan 08 19:03:52 Long time no see here ;) Jan 08 19:04:20 It has been a while. hasn't it? Jan 08 19:05:18 Yup ;) Jan 08 19:06:06 Your webos.rocks is nice :D Jan 08 19:07:12 Thanks! It's definitely a fun project. Still much to do though Jan 08 19:08:00 Still buggy on some browsers too. :/ Jan 08 19:08:44 We could use your JS knowledge on a lot of other things :P Jan 08 19:09:27 What devices you have laying around? Jan 08 19:10:21 Pixel XL, iPhone 6S, iPad, Pre 2, FirefoxOS phones... Jan 08 19:11:38 Hmmz nothing we have a port for currently.... But we recently updated the N4 + Galaxy Nexus to CM 12.1 based builds Jan 08 19:12:01 And out porting approach got easier compared to previously, so that makes it "easier" for other targets too Jan 08 19:12:09 You have a RPi by any chance? Jan 08 19:12:55 I do have a Pi 2, but it's currently a retro gaming device in the living room Jan 08 19:13:10 Ah ok :) Jan 08 19:13:19 Andolamin has an almost working port on the Pi2 ;) Jan 08 19:13:28 And there's always VirtualBox to toy with of course Jan 08 19:14:16 Oh! I did get my old Nexus 4 back from my brother when I upgraded him to a Nexus 6P. Half the touchscreen is dead though. :( Jan 08 19:16:31 That's a pity ;) But VirtualBox works pretty well for most things and lots of Enyo stuff goes in browser too :P Jan 08 19:18:52 That's true! VB is good enough for most things Jan 08 19:24:22 We got phone calls, SMS, data + modern browser. Jan 08 19:24:35 Now just trying to get IM up and running but that needs some Enyo love :P Jan 08 19:45:36 nizovn: Gonna give it a try with setting https://github.com/openembedded/openembedded-core/blob/morty/meta/conf/bitbake.conf#L407 to 2.6.32 but I expect that might give other problems. Alternatively I could try setting it in a .bbappend for https://github.com/openembedded/openembedded-core/blob/morty/meta/recipes-core/glibc/glibc_2.24.bb#L72 Jan 08 19:45:41 Will see what happens Jan 08 19:46:16 There is a 3.4 kernel for TP though Jan 08 19:46:22 Not sure how stable it is though Jan 08 19:50:36 ok Jan 08 19:51:22 I expect havoc with setting OLDEST_KERNEL_VERSION but maybe not :s Jan 08 20:34:56 Morning Jan 08 20:35:16 I see you have some fun with bootloops Jan 08 20:37:00 Herrie: to debug that, if it's due to systemd, you can create /var/log/journal directory from a recovery adb shell, do a bootloop, return to the recovery, and pull /data/luneos/var/log/journal/journal.db to analyze it Jan 08 20:50:55 Tofe: morning, OK will give that a go, but probably only tomorrow night Jan 08 20:51:15 With downgrading systemd I got rid of the bootloop but got segfaults elsewhere :P Jan 08 20:52:05 mmh ok,,, not good Jan 08 20:52:10 Jan 08 18:15:39 mako systemd[1]: voicecall-manager.service: Main process exited, code=dumped, status=11/SEGV Jan 08 20:52:13 Jan 08 18:15:39 mako systemd[1]: luna-sysservice.service: Main process exited, code=dumped, status=11/SEGV Jan 08 20:52:33 Jan 08 18:15:47 mako systemd[1]: maliit-server.service: Main process exited, code=dumped, status=11/SEGV Jan 08 20:52:41 So that might not be the right approach Jan 08 20:53:05 But at least it started to boot, just stuck @ LuneOS logo **** ENDING LOGGING AT Mon Jan 09 03:00:00 2017