**** BEGIN LOGGING AT Fri Feb 09 03:00:02 2018
Feb 09 06:24:46 <bshah> Herrie: great!
Feb 09 06:25:14 <bshah> so for test_hwcomposer libtinyxml.so thing might be false alarm
Feb 09 06:25:36 <bshah> can you grab logcat while test_hwcomposer is running?
Feb 09 06:39:19 <Herrie> bshah: It fails right away
Feb 09 06:39:25 <Herrie> THere's an error log
Feb 09 06:39:38 <Herrie> Just on the way to office shortly, let me see if I can pull it
Feb 09 06:39:43 <Herrie> test_audio also fails
Feb 09 06:40:57 <Herrie> bshah: Output for both: https://bpaste.net/show/71eea3e0a2bd
Feb 09 06:46:19 <bshah> hm, it looks like.. you're missing libhybris fix?
Feb 09 06:46:30 <bshah> or LD_LIBRARY_PATH is worng?
Feb 09 06:46:37 <bshah> (arm64 rootfs right?
Feb 09 06:48:34 <Herrie> bshah: Yeah this is Yocto aarch64 rootfs
Feb 09 06:48:43 <Herrie> Let me check libhybris
Feb 09 06:49:31 <Herrie> https://bpaste.net/show/ee255fdd6f54
Feb 09 06:50:21 <Herrie> Latest commit frm upstream libhybris we build
Feb 09 06:56:13 <bshah> hmm 3cda04985dab5f46d4c0f2067d2aab61362ee4b7 is fix you want
Feb 09 06:56:23 <bshah> so I guess LD_LIBRARY_PATH is wrong
Feb 09 06:56:24 <Herrie> Seems my LD_LIBRARY_PATH might be empty
Feb 09 06:56:31 <Herrie> What should it be?
Feb 09 06:56:35 <bshah> and points ot /system/lib instead of /system/lib64
Feb 09 06:56:38 <Herrie> I can manually set it
Feb 09 06:57:30 <Herrie> same issue
Feb 09 06:57:38 <bshah> hm
Feb 09 06:57:49 <bshah> do you have logcat?
Feb 09 06:58:35 <Herrie> https://bpaste.net/show/54d796a7757e
Feb 09 06:59:45 <bshah> android logcat I mean.. not log :)
Feb 09 07:04:08 <Herrie> https://bpaste.net/show/b8dfbf47d49f
Feb 09 07:04:14 <Herrie> Never used that before :P
Feb 09 07:08:11 <bshah> hmm lots of spam
Feb 09 07:08:44 <bshah> Herrie: can you start logcat in one session and then in another ssh/shell session start hwcomposer test and then record new output?
Feb 09 07:14:05 <Herrie> https://bpaste.net/show/c6a0719d21bd
Feb 09 07:14:21 <Herrie> Should be in bottom couple of 100 lines
Feb 09 07:21:42 <bshah> relevant part : https://bpaste.net/show/fb04555e9e9b
Feb 09 08:23:41 <Herrie|Laptop> bshah: I do have a libsdm-color.so in /system/vendor/lib64
Feb 09 08:23:58 <Herrie|Laptop> Should it be in /system/lib64 instead?
Feb 09 08:29:17 <Herrie|Laptop> bshah: We don't have this file on Hammerhead so I cannot verify it there :P
Feb 09 08:54:07 <Herrie|Laptop> bshah: I'm not sure if the library is the critical issue here or the display.
Feb 09 08:54:15 <Herrie|Laptop> I saw the following pending in libhyrbis for merge: https://github.com/libhybris/libhybris/pull/363/files
Feb 09 08:54:18 <Herrie|Laptop> Might that solve it?
Feb 09 08:55:11 <bshah> sdm-color is not really issue
Feb 09 08:55:50 <bshah> but issue is : HWDevice::Init: Primary Display Device should be present
Feb 09 08:55:52 <bshah> this parth
Feb 09 08:57:07 <Herrie|Laptop> bshah: Any thing I could check?
Feb 09 08:57:12 <Herrie|Laptop> Could be some missing udev rule or something?
Feb 09 08:57:26 <bshah> can you give dmesg?
Feb 09 08:58:56 <Herrie|Laptop> https://bpaste.net/show/99e6f75ef22c
Feb 09 09:01:43 <bshah> hm
Feb 09 09:01:47 <bshah> it looks almost normal
Feb 09 09:01:59 <bshah> there are some stuff failing but doesn't look related to graphics
Feb 09 09:03:19 <Herrie|Laptop> bshah: Well it could be some of my luna-next config stuff needs tweaking: https://github.com/webOS-ports/meta-webos-ports/blob/pyro/meta-luneui/recipes-luneui/luna-next/luna-next-conf/mido/environment.conf
Feb 09 09:04:12 <Herrie|Laptop> Though it seems these values are identical across our targets so far
Feb 09 09:04:17 <Herrie|Laptop> We don't have many though :P
Feb 09 09:04:21 <Herrie|Laptop> So might not be a good reference
Feb 09 09:07:08 <bshah> either of them doesn't affect test_hwcocmposer tbh
Feb 09 09:09:37 <Herrie|Laptop> bshah: Let me recompile libhybris with explicit LD_LIBRARY_PATH defined and also add the https://github.com/libhybris/libhybris/pull/363/files
Feb 09 09:09:42 <Herrie|Laptop> See if that changes anything
Feb 09 09:09:47 <bshah> okay
Feb 09 09:10:15 <Herrie|Laptop> Worth a try at least
Feb 09 09:30:35 <Herrie|Laptop> Nope same it seems
Feb 09 09:30:51 <Herrie|Laptop> Seems that the messages aren't that uncommon... Android 6 boot on same device gives: https://gist.github.com/HighwayStar/6cc40ca56b2f4e886cc5d7adfa37ef27
Feb 09 09:31:01 <Herrie|Laptop> Also has the "HWDevice::GetHWPanelInfoByNode: Failed to open msm_fb_panel_info node device node 2"
Feb 09 11:02:02 <Herrie|Laptop> bshah: Most of my tests looks pretty decent actually
Feb 09 11:02:07 <Herrie|Laptop> GPS seems to work, led seems to work
Feb 09 11:02:25 <bshah> Herrie|Laptop: question
Feb 09 11:02:34 <bshah> is your libhybris built with QTI_BSP?
Feb 09 11:02:46 <Herrie|Laptop> bshah: Don't think so
Feb 09 11:02:49 <Herrie|Laptop> Let me double check
Feb 09 11:03:43 <Herrie|Laptop> Nope
Feb 09 11:03:50 <bshah> time to do it then :)
Feb 09 11:03:56 <Herrie|Laptop> Seems only for Tenderloin we use some hacks i.e. https://github.com/shr-distribution/meta-smartphone/blob/pyro/meta-hp/recipes-core/libhybris/libhybris_%25.bbappend
Feb 09 11:05:29 <bshah> similarly you have to add -DQTI_BSP -DQCOM_HARDWARE IIRC
Feb 09 11:06:55 <Herrie|Laptop> Let me try that
Feb 09 11:07:04 <Herrie|Laptop> These things aren't very well documented it seems ;)
Feb 09 11:16:52 <Herrie|Laptop> Ok different error now: https://bpaste.net/show/3870f18ed3c6
Feb 09 11:17:16 <bshah> LD_LIBRARY pls
Feb 09 11:17:26 <bshah> to the lib64
Feb 09 11:18:55 <Herrie|Laptop> The /system/lib64 or the /var/lib/lxc/android/rootfs/system/lib64/
Feb 09 11:27:03 <bshah> /system/lib64
Feb 09 11:27:17 <Herrie|Laptop> When I do export LD_LIBRARY_PATH=/system/lib64 I get the same error on test_hwcomposer
Feb 09 11:28:34 <bshah> and also what if you export /vendor/lib64:/system/lib64 ?
Feb 09 11:28:47 <bshah> or maybe build libhybris with that ld_library path
Feb 09 11:33:06 <Herrie|Laptop> https://bpaste.net/show/36b75df876b8
Feb 09 11:33:21 <Herrie|Laptop> bshah: It's alreayd build with system/lib64
Feb 09 11:33:58 <bshah> dunno maybe strace and see where it tries to find file
Feb 09 11:43:24 <Herrie|Laptop> https://bpaste.net/show/aad57b8c1dc0
Feb 09 11:44:47 <bshah> see L60-61
Feb 09 11:53:21 <Herrie|Laptop> Well I have these: https://bpaste.net/show/40755c174d96
Feb 09 12:33:00 <Herrie> bshah: I wonder what's the difference between the .so and .so.2?
Feb 09 12:33:08 <Herrie> And why I would have only the regular .so
Feb 09 14:55:56 <Herrie|Laptop> bshah: OK seems my new libhybris got broken because I had lib64sssssssssssssss in the path. Pushing a new IPK doesn't seem to solve that
Feb 09 14:56:02 <Herrie|Laptop> So I'll put a whole new image
Feb 09 14:56:05 <Herrie|Laptop> That should sort it
Feb 09 14:56:12 <bshah> okay
Feb 09 14:56:29 <Herrie|Laptop> I had that in my "--with-default-hybris-ld-library-path=/system/lib64ssssssssssss"
Feb 09 14:56:35 <Herrie|Laptop> I guess my ctrl+s didn't work :P
Feb 09 14:56:40 <Herrie|Laptop> VNC LOL
Feb 09 14:57:07 <Herrie|Laptop> Anyway even if I would add all the paths it would find all .so's however would fail to find libhardware.so
Feb 09 14:57:20 <Herrie|Laptop> So a new image should sort that really
Feb 09 14:57:35 <Herrie|Laptop> Just need to reapply my WLAN fixes I guess
Feb 09 14:57:54 <Herrie|Laptop> I nastily moved some files around to /lib/firmware to avoid the reboot due to wcnss
Feb 09 14:58:01 <Herrie|Laptop> Need to find a proper solution for that still
Feb 09 15:55:48 <Herrie|Laptop> bshah: Turned out I had the new IPK on wrong path of the FTP :P
Feb 09 15:55:52 <Herrie|Laptop> THat didn't help
Feb 09 15:56:04 <Herrie|Laptop> Now reinstalled the correct IPK and that error is gone ;)
Feb 09 16:06:08 <Herrie|Laptop> bshah: New strace looks a lot better https://bpaste.net/show/cf2c90f5d99d
Feb 09 16:06:32 <bshah> Herrie|Laptop: interested in android logcat
Feb 09 16:08:09 <Herrie|Laptop> https://bpaste.net/show/a0e6f7445b9e
Feb 09 16:09:02 <bshah> still output looks more or less same :/
Feb 09 16:12:17 <Herrie|Laptop> Is the libsdmextension.so a real problem or not?
Feb 09 16:15:18 <Herrie|Laptop>  On the libhybris bits: How to now when to set the flags?
Feb 09 16:15:25 <Herrie|Laptop> We have some more targets in the works ;)
Feb 09 16:16:02 <bshah> umm basically you grep for QCOM_HARDWARE and QTI_BSP in device tree
Feb 09 16:16:09 <bshah> if it matches you set that flag
Feb 09 16:16:11 <bshah> basically
Feb 09 16:16:13 <Herrie|Laptop> Ah ok :)
Feb 09 16:16:18 <Herrie|Laptop> That makes sense :)
Feb 09 16:16:35 <bshah> basically any non-nexus qualcomm device..  if you want to be lazy
Feb 09 16:16:37 <bshah> :
Feb 09 16:16:39 <bshah> :P
Feb 09 16:16:46 <Herrie|Laptop> That test_audio doesn't work is "normal" right?
Feb 09 16:17:19 <bshah> it is broken test by design fwiw
Feb 09 16:17:35 <bshah> it never worked for me in any halium 7 target
Feb 09 16:17:40 <Herrie|Laptop> OK
Feb 09 16:17:49 <Herrie|Laptop> My test_sensors tells me it cannot ifnd any
Feb 09 16:17:58 <Herrie|Laptop> Any suggestions on how to remedy that?
Feb 09 16:18:13 <Herrie|Laptop> Because it seems the sensorfw test does find some
Feb 09 16:18:20 <bshah> Herrie|Laptop: can you give me output of "systemctl status"? I think your android init is still not completel
Feb 09 16:18:22 <Herrie|Laptop> I read you had a similar problem
Feb 09 16:19:22 <Herrie|Laptop> https://bpaste.net/show/a560bf27f17f
Feb 09 16:22:21 <bshah> hmm
Feb 09 16:22:30 <bshah> everything is fine but something is not correct
Feb 09 16:22:32 <bshah> :(
Feb 09 16:22:41 <bshah> don't you hate that?
Feb 09 16:27:07 <Herrie> bshah: From LuneOS sideit looks pretty OK
Feb 09 16:27:21 <Herrie> Biggest issue seems hwcomposer
Feb 09 16:28:21 <Herrie> Some pulseaudio(-modules-droid) but nothing too major
Feb 09 16:30:47 <bshah> maybe tinyxml is actual issue?
Feb 09 16:33:02 <Herrie> I have a libtinyxml.so.2.6.2 in usr/lib
Feb 09 16:33:29 <Herrie> On Hammerhead 5.1 I have the same
Feb 09 16:33:41 <Herrie> And that one worksfine
Feb 09 16:36:43 <Herrie> Could udev rules have any impact on the hwcomposer?
Feb 09 16:38:38 <linux_unix-10> Are there any ports based on Halium?
Feb 09 16:44:12 <Herrie> bshah: We have this for Hammerhead for example: https://github.com/shr-distribution/meta-smartphone/blob/pyro/meta-lg/recipes-core/udev/udev-extraconf/hammerhead.rules#L2
Feb 09 16:44:31 <Herrie> linux_unix-10: Yeah Mako, Hammerhead, Tenderloin and WIP for Athene, Mido & Onyx
Feb 09 16:47:36 <linux_unix-10> That's great!  I've made an initial build of Halium for Angler (Nexus 6P) and I'm looking around for distributions to put on top of it.
Feb 09 16:48:14 <linux_unix-10> Are there any guides or instructions for porting LuneOS atop Halium?
Feb 09 16:52:15 <Herrie> linux_unix-10: Not really but it's pretty straight forward
Feb 09 16:52:41 <Herrie> 6P = Huawei one?
Feb 09 16:52:50 <linux_unix-10> Yep.
Feb 09 16:52:54 <Herrie> Ah OK
Feb 09 16:52:59 <Herrie> Which kernel version is it on?
Feb 09 16:53:32 <linux_unix-10> 3.10
Feb 09 16:53:42 <Herrie> OK that's decent in general
Feb 09 16:55:19 <linux_unix-10> Is there a rootfs for Halium (like Plasma Mobile)?
Feb 09 16:55:36 <linux_unix-10> Just something to start from.
Feb 09 17:00:50 <Herrie> No, basically you can build any other target
Feb 09 17:01:42 <Herrie> Gimme a few minutes
Feb 09 17:04:53 <Herrie|TP> You can basically follow the steps up to Building on http://webos-ports.org/wiki/Build_for_Mako
Feb 09 17:05:23 <Herrie|TP> Just don't do the last step (MACHINE=mako etc)
Feb 09 17:05:44 <Herrie|TP> You'd then need to create a new layer in meta-smartphone for huawei angler
Feb 09 17:06:01 <Herrie|TP> similar to https://github.com/shr-distribution/meta-smartphone/tree/pyro/meta-xiaomi
Feb 09 17:06:13 <Herrie|TP> I can do some legwork if you want
Feb 09 17:07:53 <Herrie|TP> Ah wait you could actually start with the Android bits. Those are quite similar to Halium. we don't use hybris-boot for the rest we actually use the same bits from Halium for Android part
Feb 09 17:09:34 <Herrie|TP> The meta-smartphone layer might look scary at first but it's not that hard in general
Feb 09 17:10:03 <Herrie|TP> I can help with putting it together
Feb 09 17:10:45 <linux_unix-10> Wait a minute, the build guide says that there should be 70 GB of storage available! :O
Feb 09 17:11:02 <linux_unix-10> Why does it need that much disk space?
Feb 09 17:11:08 <Herrie|TP> The tricky part is that we use newer gcc (6.3) at the moment for kernel, so there's usually a standard set of patches needed to be applied. But these are straight forward
Feb 09 17:11:30 <Herrie|TP> Ehm Android repo sync, then qtwebengine (chromium)
Feb 09 17:11:43 <Herrie|TP> It builds everything from scratch :-)
Feb 09 17:12:52 <linux_unix-10> Oof...
Feb 09 17:14:35 <linux_unix-10> I understand. Maybe I'll get back to this when I get the storage on my main machine doubled either next week or after 2 months.
Feb 09 17:15:25 <linux_unix-10> Halium took up a lot of the remaining disk space! :D
Feb 09 17:16:48 <Herrie|TP> Well I could put a stab at the Android bits and provide the output
Feb 09 17:16:59 <Herrie|TP> That's then used by meta-smartphone
Feb 09 17:17:37 <Herrie|TP> You don't really need the Android bits once you have them ready. You can just pick up the 100 or so mb tar.gz :-P
Feb 09 17:18:23 <Herrie|TP> If you have kernel, device & vendor blob repo links I'll have a look
Feb 09 17:19:58 <linux_unix-10> Halium-modified kernel repo: https://github.com/win8linux/android_kernel_huawei_angler/tree/halium-7.1
Feb 09 17:20:23 <linux_unix-10> Device repo: https://github.com/LineageOS/android_device_huawei_angler
Feb 09 17:20:53 <linux_unix-10> Vendor repo: https://github.com/win8linux/proprietary_vendor_huawei/tree/halium-7.1
Feb 09 17:21:51 <linux_unix-10> Note that Halium does not recommend using vendor repos from TheMuppets (legal stuff).
Feb 09 17:22:15 <linux_unix-10> Patching them is fine, though. ;-)
Feb 09 17:23:06 <linux_unix-10> *forked vendor repos
Feb 09 17:42:20 <Herrie> OK I'll have a look if I cann put something together in the next few days
Feb 09 18:26:54 <Herrie> elvispre: I might have some clues for your image
Feb 09 18:30:21 <Herrie> linux_unix-10: You're Halium build is 7.1 based right?
Feb 09 18:31:40 <Herrie> Ah I see in your repos it's 7.1
Feb 09 18:39:11 <Herrie> Tofe: Was there an easy way to cleanup sstate-cache a bit.. Mine is close to 250GB now :P
Feb 09 18:40:28 <Tofe> a moment, I may have something for you
Feb 09 18:41:04 <Tofe> something like this: ../scripts/sstate-cache-management.sh --remove-duplicated --cache-dir=/media/LuneOS/testing/webos-ports/sstate-cache --extra-archs=i586,x86_64,armv7a-neon,mako,all
Feb 09 18:41:43 <Tofe> adapt to your archs and dirs, of course
Feb 09 18:42:01 <Tofe> it's started from the webos-ports dir
Feb 09 18:42:39 <Herrie> Tofe: Thnx
Feb 09 18:42:43 <Herrie> Will do that in a bit
Feb 09 18:42:55 <Herrie> Let me quickly put a meta-smartphone layer for linux_unix-10 together ;)
Feb 09 18:43:22 <Tofe> :) just basic routine for you now
Feb 09 18:44:16 <Herrie> Tofe: Well ;) Yes a bit
Feb 09 18:44:21 <Herrie> mido seems to be almost there
Feb 09 18:44:26 <Herrie> hwcomposer still acting up & audio
Feb 09 18:44:34 <Herrie> But RIL works, led works, gps works
Feb 09 18:44:40 <Herrie> So it's not in a bad state
Feb 09 18:44:47 <Herrie> PulseAudio still complaining a bit
Feb 09 18:44:50 <Tofe> Herrie: no reboot loop now ?
Feb 09 18:45:02 <Tofe> I didn't read all your discussion with bshah
Feb 09 18:45:07 <Herrie> Tofe: No, I put everything in /lib/firmware/
Feb 09 18:45:13 <Tofe> ok great
Feb 09 18:45:20 <Herrie> Tofe: Well basically needed to have some flags for libhybris
Feb 09 18:45:30 <Herrie> And WLAN started to work
Feb 09 18:45:48 <Tofe> very good !
Feb 09 18:46:05 <Tofe> so graphics & audio still to fix
Feb 09 18:46:35 <Tofe> I'll look at the end of your discussion, in case I can see something
Feb 09 18:49:18 <Herrie> Tofe: well test
Feb 09 18:49:52 <Herrie> test_sensors doesn't see anything either
Feb 09 18:50:11 <Herrie> But qt5 sensor tests work half
Feb 09 18:53:12 <Herrie> linux_unix-10: Do you have the partition layout for me?
Feb 09 18:53:21 <Tofe> test_sensors has always been a bit unstable anyway
Feb 09 18:54:37 <Herrie> Well the others (sensor*) are randomly working
Feb 09 18:54:46 <Herrie> Some do, some half
Feb 09 18:55:17 <Herrie> evtest works ok too it seems
Feb 09 18:56:07 <Tofe> did you change the LD_LIBRARY_PATH environment variable before starting test_hwcomposer?
Feb 09 18:56:40 <Herrie> Tofe: I put it back in recipe and manually added aa well
Feb 09 18:56:59 <Tofe> no no, I mean when you were in adb
Feb 09 18:57:11 <Tofe> ah that's what you mean with manually
Feb 09 18:57:15 <Herrie> Yeah put there too
Feb 09 18:58:01 <Tofe> ok, that was not necessarily a good idea: our native glibc binaries shouldn't look in /system/lib64, only the android linker should do that (i.e. libhybris)
Feb 09 18:58:08 <Herrie> Basically reverted https://github.com/webOS-ports/meta-webos-ports/commit/77002056bef820d0597e58a8c209d6b914d97341 and updated path
Feb 09 18:58:58 <Herrie> Tofe: could I test with fbdev instead of hwcomposer somehow?
Feb 09 18:59:33 <Tofe> I don't think that'll work: usually it's at least partially deactivated by the kernel config, to be able to use hwcomposer
Feb 09 19:00:31 <Tofe> And your commit did look good, I think
Feb 09 19:03:59 <Tofe> You don't have any "/dev/socket/property_service" ? Mmmh is that normal ? Let me check on N5
Feb 09 19:05:09 <Tofe> I have one on N5
Feb 09 19:05:50 <Herrie> Tofe: Well it should theoretically work without
Feb 09 19:05:52 <Herrie> Let me try
Feb 09 19:06:01 <Herrie> Tofe: I noticed that one too
Feb 09 19:06:39 <Tofe> well if a property was set by the lxc container, the test_hwcomposer program won't see it
Feb 09 19:06:41 <Herrie> Tofe: Ah is that why my echo of $LD_LIBRARY_PATH was empty?
Feb 09 19:07:06 <Tofe> no, no, that's fine, it can be empty
Feb 09 19:07:11 <Herrie> Ah ok
Feb 09 19:07:14 <Herrie> Because it was
Feb 09 19:07:35 <Herrie> But I can see libhybris picked up the one I passed with --with-default-hybris-ld-library-path
Feb 09 19:07:46 <Herrie> Let me try without --with-default-hybris-ld-library-path
Feb 09 19:07:52 <Herrie> That should also work anyway
Feb 09 19:07:58 <Tofe> I think too
Feb 09 19:08:35 <Herrie> That was the point of https://github.com/libhybris/libhybris/commit/3cda04985dab5f46d4c0f2067d2aab61362ee4b7 at least my understanding of it
Feb 09 19:08:47 <Tofe> But setting LD_LIBRARY_PATH in the shell with /system/lib64 before starting a "native" program which uses libhybris will just confuse the glibc library loader
Feb 09 19:09:15 <Herrie> Tofe: OK
Feb 09 19:09:42 <Herrie> So either I don't set --with-default-hybris-ld-library-path at all and it should work, or I should all the possible paths with libraries in --with-default-hybris-ld-library-path ?
Feb 09 19:10:34 <Tofe> Basically yes. But aarch64 should be well handled by default by libhybris now
Feb 09 19:17:16 <Tofe> android-property-service is started on your device, but the device node isn't created, that's strange
Feb 09 19:19:50 <Herrie> This is hwcomposer strace output https://bpaste.net/show/e112dc605bdd with new libhybris without the --with-default-hybris-ld-library-path
Feb 09 19:20:49 <Herrie> Tofe: android-property-service is our code right?
Feb 09 19:20:57 <Herrie> Could be it needs updating for 7.1 somehow?
Feb 09 19:21:45 <Tofe> I don't know
Feb 09 19:21:53 <Tofe> It was pretty generic I think
Feb 09 19:21:57 <Tofe> Let me see
Feb 09 19:22:15 <Herrie> https://github.com/webOS-ports/android-property-service/tree/master/src
Feb 09 19:22:18 <Herrie> Should be I'd say
Feb 09 19:23:23 <Tofe> yes it's generic
Feb 09 19:23:44 <Herrie> Anything I could try?
Feb 09 19:23:56 <Herrie> Some luna-send like we do in Settings to get some properties to see if it returns something/
Feb 09 19:25:37 <Tofe> I was wrong, it's not that which creates the /dev/socket/android_properties file
Feb 09 19:26:09 <Herrie> Tofe: luna-send works OK
Feb 09 19:26:41 <Herrie> FWIW: https://bpaste.net/show/ced2b058ae2f
Feb 09 19:28:08 <Tofe> mmh. ok. maybe you're right, and it's no big deal if that file isn't there
Feb 09 19:28:16 <Herrie> Well maybe it is?
Feb 09 19:28:21 <Herrie> I'm not sure what it pulls frm there
Feb 09 19:28:21 <Tofe> :p
Feb 09 19:28:36 <Tofe> it's mirroring the one of the lxc container
Feb 09 19:30:32 <Herrie> Not sure if test_hwcomposer pulls some value frm it which it uses
Feb 09 19:31:20 <Tofe> hard to know... usually it uses it to know the device model and so on, and libhybris fallback on build.prop if needed, which is fine in most cases
Feb 09 19:32:00 <Tofe> It's probably not the root cause of test_hwcomposer failing
Feb 09 19:32:30 <Tofe> But we'll have to clarify that for other small issues, where it might play a role
Feb 09 21:05:07 <Herrie> linux_unix-10: Seems we need some gcc6 patch
Feb 09 21:05:16 <Herrie> Let me fork, patch & propose PR
Feb 09 21:08:35 <Herrie> linux_unix-10: Some initial work on meta-huawei layer for meta-smartphone
Feb 09 21:08:36 <Herrie> https://github.com/Herrie82/meta-smartphone/commit/6bbb3ee37a0cab9c7fbd7a616f7b72e7cc56ff70
Feb 09 21:08:47 <Herrie> Will need some polishing for sure, but it should be pretty OK already
Feb 09 22:20:45 <Herrie> linux_unix-10: These are the ones so far. I'll continue tomorrow: https://github.com/Herrie82/android_kernel_huawei_angler/commits/cm-14.1
**** ENDING LOGGING AT Sat Feb 10 03:00:01 2018