**** BEGIN LOGGING AT Mon Dec 23 03:00:01 2019
**** BEGIN LOGGING AT Mon Dec 23 08:31:31 2019
Dec 23 09:19:22 <Herrie> Tofe: Seems I forgot a few still: https://github.com/webOS-ports/imlibpurpleservice/pull/9, https://github.com/webOS-ports/certmgrd/pull/1 and https://github.com/webOS-ports/storaged/pull/4 so far
Dec 23 09:20:58 <Herrie> And https://github.com/webOS-ports/luna-next/pull/143
Dec 23 09:39:01 <Tofe> on it :)
Dec 23 09:42:11 <Tofe> novaldex|away: bonaire is up, but bitbake server seems down again
Dec 23 10:01:24 <Herrie> Tofe: Thnx
Dec 23 10:03:02 <Tofe> Herrie: I have been looking at the bad performances of our camera app lately
Dec 23 10:03:20 <Tofe> and the funky preferences popup thing is costly
Dec 23 10:03:44 <Tofe> so maybe we could go back to the previous implementation
Dec 23 10:05:49 <Tofe> Herrie: https://imgur.com/a/5CkF2Mz
Dec 23 13:37:55 <Herrie> Tofe: Why not?
Dec 23 13:38:01 <Herrie> If it helps performance
Dec 23 13:38:11 <Tofe> bok
Dec 23 13:38:13 <Tofe> ok
Dec 23 17:22:27 <Hacker1245> Tofe: Herrie told me to ask you for help regarding my LuneOS port not booting https://bpaste.net/show/FSQ6O
Dec 23 18:37:34 <Tofe> Hacker1245: a crash in ext4_xattr_security_get in the kernel... that's new too...
Dec 23 18:37:53 <Tofe> what's your defconfig ?
Dec 23 18:39:22 <Hacker1245> https://github.com/Hacker1245/android_kernel_motorola_msm8937-1/blob/lineage-15.1-64/arch/arm64/configs/cedric_defconfig
Dec 23 18:41:56 <Tofe> looks like you've run into a known bug: https://lore.kernel.org/lkml/20180813002409epcas1p2d0c7cf768c533de9348a994e614fc6ea~KSd-6p8Bi1658316583epcas1p2a@epcas1p2.samsung.com/
Dec 23 18:42:23 <Tofe> either backport the 5 commits like suggested, or apply the hacky patch one-liner
Dec 23 18:42:53 <Hacker1245> I'll take the patch ;)
Dec 23 18:43:04 <Tofe> :)
Dec 23 18:46:24 <Hacker1245> patched
Dec 23 19:00:49 <Herrie> Hacker1245: New files are on ftp
Dec 23 19:01:29 <Herrie> Just the .bin and . Fastboot
Dec 23 19:01:36 <Hacker1245> Gonna test tomorrow
Dec 23 19:02:12 <Herrie> Hopefully those will improve things
Dec 23 19:02:24 <Herrie> Tofe: You back or still traveling?
Dec 23 19:02:42 <Tofe> Herrie: still travelling for about a week
Dec 23 19:03:29 <Tofe> too bad that the build servers are acting weird just during this period
Dec 23 19:03:29 <Herrie> Tofe: OK :)
Dec 23 19:03:46 <Herrie> Tofe: I'm happy to build something for you
Dec 23 19:04:05 <Tofe> he, like a pinephone testing build ?... :0
Dec 23 19:04:24 <Tofe> s/:O/:)/
Dec 23 19:04:44 <Herrie> Hacker1245: Fastboot normally I'd say If you use fastboot
Dec 23 19:04:58 <Herrie> Tofe: Will kick that off later tonight
Dec 23 19:05:06 <Tofe> thanks !
Dec 23 19:05:29 <Hacker1245> Herrie: just fastboot boot it as a regular boot.img?
Dec 23 19:06:39 <Herrie> Hacker1245: Yeah Fastboot boot image_xyZ.fastboot
Dec 23 19:20:08 <Hacker1245> you know what, I'll try it now
Dec 23 19:27:11 <Hacker1245> btw should there be usb network?
Dec 23 19:28:02 <Hacker1245> this time the screen went dark
Dec 23 19:30:14 <Hacker1245> Tofe: https://paste.ubuntu.com/p/CsppxKTH2b/
Dec 23 19:31:36 <Hacker1245> note that we used a LineageOS 15.1 64 bit kernel
Dec 23 19:38:14 <Hacker1245> maybe it's because of enabled audit?
Dec 23 19:43:34 <Herrie> Hacker1245: This looks better for sure
Dec 23 19:43:41 <Herrie> Could be audit yes
Dec 23 19:44:20 <Hacker1245> Herrie: could you try to disable it and rebuild?
Dec 23 19:44:47 <Hacker1245> also what device tree is being used?
Dec 23 19:49:16 <Hacker1245> although LineageOS launches
Dec 23 19:49:54 <Herrie> You can do that in fastboot
Dec 23 19:50:06 <Herrie> You can pass cmdline I think
Dec 23 19:50:30 <Hacker1245> yeah, which option is audit on cmdline?
Dec 23 19:50:33 <Hacker1245> audit-0?
Dec 23 19:50:39 <Hacker1245> audit=0
Dec 23 19:52:18 <Herrie> I'm putting children to bed can have a look in a bit
Dec 23 20:15:56 <Herrie> Hacker1245: We use what's in Halium-devices manifest.xml for Cedric (except for Kernel)
Dec 23 20:16:10 <Hacker1245> Alright
Dec 23 20:19:27 <Hacker1245> added audit=0, rebooting
Dec 23 20:23:41 <Hacker1245> screen is still black
Dec 23 20:30:49 <Hacker1245> https://paste.ubuntu.com/p/J5xg2cqVWJ/
Dec 23 20:32:54 <Hacker1245> maybe I should try this? https://wiki.merproject.org/wiki/Adaptations/faq-hadk#Notes_on_LOS14.1_Porting
Dec 23 20:34:32 <Hacker1245> also https://paste.ubuntu.com/p/J5xg2cqVWJ/
Dec 23 20:40:40 <Herrie> Hacker1245: You might want to try "androidboot.selinux=permissive --"
Dec 23 20:40:43 <Herrie> On your cmdline
Dec 23 20:41:00 <Hacker1245> on it
Dec 23 20:42:30 <Hacker1245> display is still dark so far...
Dec 23 20:44:23 <Herrie> Can you adb shell into it?
Dec 23 20:44:39 <Hacker1245> yep
Dec 23 20:45:16 <Hacker1245> sh-5.0#
Dec 23 20:45:40 <Herrie> Can you output journalctl to a file? I.e. journalctl > /tmp/mylog.txt
Dec 23 20:45:46 <Herrie> And then adb pull /tmp/mylog.txt
Dec 23 20:46:11 <Hacker1245> pasting
Dec 23 20:46:38 <Hacker1245> oh wifi service is failing
Dec 23 20:46:57 <Herrie> Well wifi should be a worry for later really
Dec 23 20:47:02 <Herrie> That shouldn't cause UI not to come up
Dec 23 20:47:15 <Herrie> But you might be able to run the various libhybris tests
Dec 23 20:47:19 <Herrie> Like test_hwcomposer
Dec 23 20:47:43 <Hacker1245> https://paste.ubuntu.com/p/htNF8y9jRx/
Dec 23 20:48:01 <Hacker1245> sh-5.0# test_lights
Dec 23 20:48:01 <Hacker1245> library "libhardware.so" not found
Dec 23 20:48:01 <Hacker1245> Segmentation fault
Dec 23 20:48:40 <Herrie> Try export EGL_PLATFORM=hwcomposer
Dec 23 20:48:44 <Herrie> And then test_hwcomposer
Dec 23 20:48:55 <Hacker1245> sh-5.0# EGL_PLATFORM=hwcomposer test_hwcomposer
Dec 23 20:48:55 <Hacker1245> library "libGLESv2.so" not found
Dec 23 20:48:55 <Hacker1245> library "libhardware.so" not found
Dec 23 20:48:55 <Hacker1245> Segmentation fault
Dec 23 20:49:44 <Hacker1245> test_vibrator does not work too
Dec 23 20:49:59 <Herrie> Hmmz those wifi & BT is where I need to pick your patches, sorry forgot those
Dec 23 20:50:11 <Hacker1245> it's ok
Dec 23 20:50:25 <Hacker1245> I want to get the basic Halium stuff working first
Dec 23 20:50:56 <Herrie> Well I would expect libGLESv2.so to be somewhere in the image
Dec 23 20:51:01 <Herrie> Could be it just cannot find it
Dec 23 20:51:09 <Herrie> Which is LD library path issue probably
Dec 23 20:51:31 <Herrie> Can you see if the 2 .so files are in the image?
Dec 23 20:53:29 <Hacker1245> I can check if they are on the device
Dec 23 20:53:42 <Herrie> Yes
Dec 23 20:53:59 <Herrie> I'll check image here
Dec 23 20:54:35 <Hacker1245> libhardware is in /system/lib
Dec 23 20:54:49 <Hacker1245> libGLES too
Dec 23 20:55:42 <Hacker1245> here's what is in /lib https://paste.ubuntu.com/p/TSpZgTd83C/
Dec 23 20:57:11 <Herrie> Try "export LD_LIBRARY_PATH=/system/lib"
Dec 23 20:57:17 <Herrie> And then test_hwcomposer again
Dec 23 20:57:29 <Herrie> Or first see what's set as LD_LIBRARY_PATH maybe
Dec 23 20:57:47 <Hacker1245> with env?
Dec 23 20:59:00 <Hacker1245> it doesn't appear there
Dec 23 20:59:38 <Herrie> echo $LD_LIBRARY_PATH
Dec 23 20:59:41 <Herrie> Might also work
Dec 23 21:00:03 <Hacker1245> nothing
Dec 23 21:00:09 <Herrie> OK then set it with export ;)
Dec 23 21:00:40 <Hacker1245> same thing
Dec 23 21:00:44 <Hacker1245> still errors out
Dec 23 21:02:53 <Hacker1245> maybe I'll symlink them to /lib?
Dec 23 21:03:45 <Hacker1245> how can I remount as rw?
Dec 23 21:04:11 <Herrie> And if I you try "export LD_LIBRARY_PATH=:/system/lib:/system/lib64"
Dec 23 21:04:49 <Hacker1245> same thing
Dec 23 21:05:53 <Herrie> Hmmz still complaining about the 2 missing .so's ?
Dec 23 21:06:06 <Hacker1245> yep
Dec 23 21:06:17 <Hacker1245> sh-5.0# EGL_PLATFORM=hwcomposer test_hwcomposer
Dec 23 21:06:17 <Hacker1245> library "libGLESv2.so" not found
Dec 23 21:06:17 <Hacker1245> library "libhardware.so" not found
Dec 23 21:06:17 <Hacker1245> Segmentation fault
Dec 23 21:06:17 <Hacker1245> sh-5.0# env
Dec 23 21:06:18 <Hacker1245> DBUS_SESSION_US_PID=1132
Dec 23 21:06:20 <Hacker1245> PWD=/lib
Dec 23 21:06:22 <Hacker1245> LANG=C
Dec 23 21:06:24 <Hacker1245> INVOCATION_ID=a3616def29274889b42c8c74974e2d9a
Dec 23 21:06:26 <Hacker1245> SHLVL=1
Dec 23 21:06:28 <Hacker1245> LD_LIBRARY_PATH=:/system/lib:/system/lib64
Dec 23 21:06:30 <Hacker1245> JOURNAL_STREAM=7:26554
Dec 23 21:06:32 <Hacker1245> PATH=/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin
Dec 23 21:06:36 <Hacker1245> DBUS_SESSION_BUS_ADDRESS=unix:abstract=/tmp/dbus-9GUddDRO4f,guid=89efca41de9dab5407a35a575dd5b325
Dec 23 21:06:38 <Hacker1245> OLDPWD=/system/lib
Dec 23 21:06:40 <Hacker1245> _=/usr/bin/env
Dec 23 21:09:20 <Herrie> And the vibrator test, also the same?
Dec 23 21:09:34 <Hacker1245> Yep
Dec 23 21:10:08 <Herrie> Anything you can see in logcat?
Dec 23 21:10:33 <Hacker1245> umm
Dec 23 21:10:33 <Hacker1245> sh-5.0# /system/bin/logcat
Dec 23 21:10:34 <Hacker1245> linker: normalize_path - invalid input: "", the input path should be absolute
Dec 23 21:10:34 <Hacker1245> linker: Warning: unable to normalize ""
Dec 23 21:10:34 <Hacker1245> WARNING: linker: Warning: unable to normalize ""
Dec 23 21:10:36 <Hacker1245> logcat read failure
Dec 23 21:13:54 <Herrie> Ah that might be a selinux permission issue as per https://github.com/bytedeco/javacv/issues/1127
Dec 23 21:14:20 <Herrie> You're actually pretty close in getting things to start, because I see a lot of LuneOS bits in the log already
Dec 23 21:14:28 <Herrie> Which means kernel in general is pretty OK and rootfs too
Dec 23 21:14:35 <Herrie> It's probably something small somewhere
Dec 23 21:14:47 <Hacker1245> but where though
Dec 23 21:15:02 <Herrie> You said the Halium bits were 32 bits for device?
Dec 23 21:15:09 <Herrie> Maybe I should try the 32 bits kernel instead?
Dec 23 21:15:09 <Hacker1245> Yes
Dec 23 21:15:24 <Hacker1245> Maybe I should try the 32 bits kernel instead? yeah that would probably be best
Dec 23 21:15:31 <Hacker1245> if you can get it to compile that is
Dec 23 21:15:42 <Hacker1245> use the halium kernel
Dec 23 21:15:53 <Herrie> Can you add the updated defconfig & patch to https://github.com/Hacker1245/android_kernel_motorola_msm8937-1/commits/lineage-15.1-32 ?
Dec 23 21:16:06 <Hacker1245> I didn't add it?
Dec 23 21:16:16 <Herrie> Only to the -64 one
Dec 23 21:16:27 <Hacker1245> Oh alright
Dec 23 21:16:35 <Herrie> Though the kernel in principle seems OK-ish actually
Dec 23 21:16:40 <Hacker1245> could I just use the defconfig from halium?
Dec 23 21:17:11 <Herrie> I'm not sure that one was checked with the verify_kernel script tbh
Dec 23 21:17:25 <Hacker1245> I think it was
Dec 23 21:17:37 <Hacker1245> since it works with sfos, UT and the reference rootfs
Dec 23 21:20:19 <Herrie> OK let me try that quickly again then
Dec 23 21:20:28 <Hacker1245> adding config
Dec 23 21:23:06 <Hacker1245> added
Dec 23 21:23:18 <Hacker1245> hmm security_d_instantiate(entry, inode); appears three times in the kernel source
Dec 23 21:23:21 <Hacker1245> where do I apply it?
Dec 23 21:23:27 <Hacker1245> https://github.com/Hacker1245/android_kernel_motorola_msm8937-1/blob/lineage-15.1-32/fs/dcache.c
Dec 23 21:26:30 <Herrie> Might not be needed on 32 bits I guess
Dec 23 21:26:43 <Hacker1245> alright let's omit it
Dec 23 21:26:51 <Hacker1245> you can use the kernel source
Dec 23 21:27:35 <Herrie> Trying the Halium one now, will then try yours
Dec 23 21:31:33 <Herrie> Halium one fialed, trying other one
Dec 23 21:41:59 <Herrie> Also failing
Dec 23 21:42:17 <Hacker1245> the hell
Dec 23 21:42:22 <Herrie> Let me try without -dtb for KERNEL_IMAGE_TYPE
Dec 23 21:42:31 <Herrie> KERNEL_IMAGETYPE
Dec 23 21:42:34 <Hacker1245> so there is something with the 32 bit kernel...
Dec 23 21:43:38 <Herrie> Same
Dec 23 21:43:45 <Hacker1245> the hell
Dec 23 21:43:48 <Herrie> Well or our buildchain with newer GCC
Dec 23 21:44:04 <Hacker1245> maybe
Dec 23 21:44:19 <Herrie> Newer 3.18 seems OK
Dec 23 21:44:29 <Herrie> Sometimes kernel failing can be because of small things
Dec 23 21:44:58 <Herrie> I'm getting "msm_rq_stats.c:148:3: error: implicit declaration of function 'cpufreq_get_policy'; did you mean 'cpufreq_next_valid'? [-Werror=implicit-function-declaration]"
Dec 23 21:45:22 <Hacker1245> had the same thing
Dec 23 21:47:04 <Herrie> But in general the 64 bits kernel looks pretty OK, except for it not being able to find the libs
Dec 23 21:47:18 <Herrie> Once we tackle that it should be behaving a lot better
Dec 23 21:47:25 <Hacker1245> Yeah
Dec 23 21:47:25 <Herrie> bshah might have a clue too
Dec 23 21:48:08 <Herrie> He's more Plasma Mobile, but he knows his way around this low level Linux stuff and how to solve it
Dec 23 21:48:12 <Herrie> bshah: ^
Dec 23 21:53:14 <Hacker1245> btw do you want anything from my device?
Dec 23 21:53:20 <Herrie> Hacker1245: New Image.gz--3.18.113+gitr1+36559983e2-r0.6-cedric-0-0.fastboot on FTP
Dec 23 21:53:21 <Hacker1245> because it's still running adb
Dec 23 21:53:29 <Hacker1245> 32 bit?
Dec 23 21:53:32 <Herrie> Only change is the "androidboot.selinux=permissive --"
Dec 23 21:53:34 <Herrie> No 64
Dec 23 21:53:38 <Herrie> Couldn't get 32 to compile
Dec 23 21:53:41 <Hacker1245> ok
Dec 23 21:53:44 <Herrie> But our image is 64 in general as well
Dec 23 21:55:30 <Hacker1245> hmm I don't see the image
Dec 23 21:55:49 <Hacker1245> now I got it
Dec 23 21:57:37 <Hacker1245> alright back in adb
Dec 23 21:57:58 <Herrie> journalctl would be good again
Dec 23 21:58:07 <Herrie> And the console-ramoops
Dec 23 21:58:16 <Herrie> Curious if test_hwcomposer, logcat etc now will behave
Dec 23 21:59:13 <Hacker1245> journalctl
Dec 23 21:59:14 <Hacker1245> https://paste.ubuntu.com/p/pqJtvvxmyN/
Dec 23 21:59:51 <Hacker1245> hmm
Dec 23 21:59:52 <Hacker1245> sh-5.0# /system/bin/logcat
Dec 23 21:59:52 <Hacker1245> logcat read failure
Dec 23 22:00:16 <Hacker1245> sh-5.0# test_lights
Dec 23 22:00:16 <Hacker1245> library "libhardware.so" not found
Dec 23 22:00:16 <Hacker1245> Segmentation fault
Dec 23 22:00:29 <Hacker1245> that's without exporting libs
Dec 23 22:00:41 <Hacker1245> trying that now
Dec 23 22:00:58 <Hacker1245> sh-5.0# /system/bin/logcat
Dec 23 22:00:58 <Hacker1245> linker: normalize_path - invalid input: "", the input path should be absolute
Dec 23 22:00:58 <Hacker1245> linker: Warning: unable to normalize ""
Dec 23 22:00:58 <Hacker1245> WARNING: linker: Warning: unable to normalize ""
Dec 23 22:00:58 <Hacker1245> logcat read failure
Dec 23 22:16:53 <Hacker1245> Herrie: anything else to test?
Dec 23 22:17:51 <Herrie> No, I have some ideas for tomorrow morning though
Dec 23 22:18:14 <Herrie> Play more with audit and check various qcom flags for libhybris
Dec 23 22:18:50 <Hacker1245> Alright
Dec 23 22:18:56 <Hacker1245> restoring backup
**** ENDING LOGGING AT Tue Dec 24 02:59:57 2019