Wednesday, 2019-10-30

« Tuesday, 2019-10-29 Index Thursday, 2019-10-31 »
Thaodanmal: either fstab or a mount unit02:17
rinigusmal: good question. I'll have to check if i missed something (init or other scripts)06:04
rinigusmal: is selinuxfs mounted on xperia 10?06:05
T42<Lukapanio> so @hacker12455  @elros34  whats next?07:16
T42<hacker12455> Try to find out why the system mount unit does not mount I guess07:19
T42<birdzhang> @Lukapanio show your device tree github url07:20
T42<birdzhang> make sure your fstab.qcom contains /system and /vendor mount point or you can use this way https://github.com/sailfishos-oneplus5/droid-config-cheeseburger/commit/cf257e0cbe693fb7331c38e4b406582aa9a9e5ae07:30
T42<Lukapanio> @birdzhang [make sure your fstab.qcom contains /system and …], There is no vendor07:32
T42<Lukapanio> @birdzhang [make sure your fstab.qcom contains /system and …], Thanks will try. Can i try on device?07:34
T42<birdzhang> yeah07:35
T42<Lukapanio> Good thanks07:35
T42<birdzhang> you can find "What=/dev/ " from your fixup-mountpoint07:36
T42<Lukapanio> I know07:36
T42<birdzhang> okay07:37
T42<Lukapanio> mmcblk0p53for me07:37
T42<Lukapanio> trying to boot07:41
T42<Lukapanio> damn nothing new looks bad,but few more minutes and will get journal07:42
T42<birdzhang> you can check via "df -h", or systemctl status vendor.mount07:44
T42<Lukapanio> @birdzhang [you can check via "df -h", or systemctl status …], Failed to get D-Bus connection: Operation not permitted07:46
T42<Lukapanio> may be problem is that it cant be executed in debug shell07:46
T42<birdzhang> switch to root "su - root"07:47
T42<Lukapanio> @birdzhang [switch to root "su - root"], [root@(none) ~]# systemctl status system.mount07:47
T42<Lukapanio> Failed to get D-Bus connection: Operation not permitted07:47
T42<Lukapanio> [root@(none) ~]#07:47
T42<Lukapanio> must it work without systemd?07:48
T42<Lukapanio> @birdzhang  can i do mount /dev/mmcblk0p53 /system08:01
T42<Lukapanio> for test08:01
T42<birdzhang> yes you can08:01
T42<Lukapanio> ok thanks08:06
T42<Lukapanio> now echo "continue" >/init-ctl/stdin08:06
T42<Lukapanio> still same08:18
T42<Lukapanio> here is log https://bpaste.net/show/IOJEU08:30
T42<Lukapanio> but seemsnothing new in it08:30
T42<hacker12455> Hmm08:34
T42<hacker12455> Sailfish systemd[1]: data.mount: Unit is bound to inactive unit dev-mmcblk0p54.device. Stopping, too.08:34
T42<Lukapanio> hmmm08:37
T42<elros34> how about showing your changes: cat /lib/systemd/system/system.mount. Mountng it manually then ls -al /system08:39
T42<Lukapanio> https://bpaste.net/show/H2OV208:40
T42<Lukapanio> @elros34  https://bpaste.net/show/V2QWK08:41
T42<adampigg> mal: so, the older mido builds do not disable selinux on the cmdline.08:42
T42<adampigg> they start in permisive mode08:44
T42<adampigg> could a new systemd force into enforcing mode?08:44
T42<adampigg> and thats why it breaks?08:44
T42<elros34> @Lukapanio so now ln -s /dev/null /etc/systemd/system/droid-hal-init.service, continue booting and wait for a while. Then strace -f /sbin/droid-hal-init08:44
T42<Lukapanio> unable todo so08:45
T42<Lukapanio> after echo "continue" >/init-ctl/stdin no usb connection @elros3408:46
T42<elros34> straightaway or after few seconds?08:46
T42<Lukapanio> maybe 1-2 second08:47
T42<Lukapanio> sometimes if i fast i can even exit08:47
T42<adampigg> mal: changelog for systemd says08:47
T42<adampigg> enable audit support, enable selinux support08:47
T42<adampigg> and lots of config changes for the sony devices to disable selinux08:47
T42<elros34> @Lakkapanio even with masked droid-hal-init? Enable persistent logging in journal and get log from previous boot. Something must disable it.08:49
T42<elros34> or wait maybe usb-moded can be started before droid-hal-init so mask it08:51
T42<Lukapanio> @elros34 [@Lakkapanio even with masked droid-hal-init? E …], https://bpaste.net/show/NDO4M08:54
T42<hacker12455> Did you mask usb-moded?08:58
T42<hacker12455> @Lukapanio09:00
T42<elros34> system is mounted now...09:01
T42<Lukapanio> @hacker12455 [Did you mask usb-moded?], not yet09:01
T42<Lukapanio> @elros34 [system is mounted now...], thats good09:01
Mister_Magisterlbt: obs search throws 500ise please fix ;-;09:01
T42<hacker12455> Hmm, maybe you can try to start droid-hal-init?09:02
T42<Lukapanio> when i will have usb09:05
T42<elros34> so mask usb-moded.service maybe also user@100000.service09:05
T42<Lukapanio> okmasked09:07
T42<Lukapanio> bothe09:07
T42<Lukapanio> hmm usb net is ok09:08
T42<Lukapanio> https://bpaste.net/show/UXNUO09:09
T42<hacker12455> Is droid-hal-init in /sbin?09:10
T42<Lukapanio> sh-3.2# ls /sbin/droid*09:11
T42<Lukapanio> /sbin/droid-hal-init09:11
T42<Lukapanio> so yes it is09:11
T42<birdzhang> @Lukapanio i have the same issue, no idea how to resolve09:11
T42<birdzhang> http://www.merproject.org/logs/%23sailfishos-porters/%23sailfishos-porters.2019-06-18.log.html#t2019-06-18T20:32:4209:13
T42<birdzhang> Mister_Magister: do you have some idea?09:13
Mister_Magisterwhy are you pinging me09:13
T42<birdzhang> search from mer log, you have this before09:14
T42<elros34> @Lakapanio are you sure it's masked? droid-hal-init can be started only once. Maybe also execute before strace: touch /dev/.coldboot_done; export LD_LIBRARY_PATH=09:14
Mister_Magisterbirdzhang: first, you need 64bit strace09:15
T42<birdzhang> Mister_Magister: thank you09:16
Mister_Magisterbirdzhang: https://forum.xda-developers.com/nexus-9/development/useful-64-bit-aarch64-binaries-busybox-t2931373/amp/09:16
Mister_Magistereven have a link09:16
T42<hacker12455> @elros34 should be masked09:16
T42<hacker12455> 01 02:49:14 Sailfish systemd[1]: droid-hal-init.service: Cannot add dependency job, ignoring: Unit droid-hal-init.service is masked.09:16
T42<birdzhang> thanks a lot ;)09:17
T42<Lukapanio> thanks09:17
T42<hacker12455> Hmm @Lukapanio you are using a 64 bit kernel?09:19
T42<Lukapanio> yep09:20
T42<hacker12455> Try the strace from the link maybe?09:21
T42<elros34> try strace from system partition, I think /system/xbin/strace09:21
T42<Lukapanio> https://bpaste.net/show/E6GH409:24
T42<hacker12455> Can't find context files09:27
T42<Lukapanio> yep09:30
T42<Lukapanio> but why it is not generated09:30
T42<elros34> try make modules | grep context in HABUILD09:30
T42<Lukapanio> @elros34 [try make modules | grep context in HABUILD], https://bpaste.net/show/QWZQG09:36
T42<elros34> so try to make at least property_contexts plat_property_contexts. I don't know which one are important09:38
T42<Lukapanio> https://bpaste.net/show/LYYYW09:53
T42<Lukapanio> allthat i can find09:53
T42<elros34> grep is not for searching files. Use find or gnome-search-tool in out/ directory09:57
T42<elros34> building droid-hal should show new unpackaged files now09:57
Mister_Magisterfind / -name \* -exec sh -c "rm -rf {}" \;09:58
T42<Lukapanio> lol09:58
Mister_Magisteri kinda mastered find :P09:59
T42<Lukapanio> @elros34 [building droid-hal should show new unpackaged …], rpm/dhd/helpers/build_packages.sh now dont return any10:10
T42<Lukapanio> i done something wrong @elros3410:12
T42<elros34> when you make plat_property_contexts for the first time it will print at then end  path to out/target/product/$DEVICE/root/ so I am not sure what your searching for10:17
T42<Lukapanio> ouch sorry wasnt built becouse of java 1.710:19
T42<Lukapanio> instead of 1.810:19
T42<Lukapanio> @elros34  @hacker12455  now i do have /home/hello/hadk/out/target/product/cedric/system/etc/selinux/plat_property_contexts10:29
T42<Lukapanio> copy that to device?/10:30
T42<elros34> hmm if it's not in root dir then maybe it will be better to create symlink in device: ln -s /system/etc/selinux/plat_property_context  and in droid-config/sparse/10:34
T42<Lukapanio> ok10:35
T42<Lukapanio> for now will just copy to device10:35
T42<Lukapanio> so that fixed problem10:37
T42<Lukapanio> but now i am spamed with newfstatat(AT_FDCWD, "/dev/.coldboot_done", 0x7fd72e05f0, 0) = -1 ENOENT (No such file or directory)10:37
T42<Lukapanio> hmmm10:39
T42<Lukapanio> need to symlink /system/vendor to /vendor10:39
Mister_Magister@Lukapanio do you have system-as-root?10:41
T42<Lukapanio> dont think so i think ramdisk is root10:42
Mister_Magisterhoho something in matrix broke?10:42
T42<elros34> check /usr/bin/droid/droid-hal-startup.sh for every command you need to exe before strace10:43
Mister_Magister@Lukapanio what is on system partition? Is there system folder?10:43
T42<Lukapanio> no system folder in systempartition10:43
Mister_Magisterah carry on10:44
T42<elros34> wouldn't system as a root also use skip_initramfs?10:44
Mister_Magisterwhat cpu is that btw10:44
Mister_Magisterso?10:44
Mister_Magisterwe patched that10:44
T42<Lukapanio> @hacker12455  https://bpaste.net/show/J3WOO11:15
T42<Lukapanio> so @elros34  whats next?11:23
T42<elros34> it's only part of the logs so I don't know what fails know anyway you can try test_hwconposer https://github.com/mer-hybris/hadk-faq#graphics11:25
T42<Lukapanio> i know i run much of tests11:25
T42<Lukapanio> sadly passed only test_lights11:25
T42<Lukapanio> will strace hwcomposer now11:26
T42<Lukapanio> @elros34 [it's only part of the logs so I don't know wha …], https://bpaste.net/show/GFYIS11:29
T42<Lukapanio> looks not too good11:29
T42<adampigg> mal: ah11:32
T42<Lukapanio> https://bpaste.net/show/OG5DI here with EGL_PLATFORM=hwcomposer11:32
T42<adampigg> in my mido kernel, i have config_SECURITY_SELINUX_DEVELOP which is the permissive mode config...11:32
T42<adampigg> the systemd change must have turned it on and broken dhi11:32
T42<elros34> @Lukapanio there is also test_hwc211:33
T42<Lukapanio> @elros34 [@Lukapanio there is also test_hwc2], yep here it is https://bpaste.net/show/O64KA11:36
T42<Lukapanio> still openat(AT_FDCWD, "/property_contexts", O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory) is here11:42
T42<elros34> it might be not needed, check new journal maybe something still fails, unmask user@10000011:44
mal@adampigg interesting11:49
T42<adampigg> mal: yup, new systemd is compiled with selinux enabled. its in the changelog11:49
T42<adampigg> so, its not enough to have it in permissive mode, it must be turned off11:50
T42<Lukapanio> @elros34 [it might be not needed, check new journal mayb …], ok unmasked it and unmaked droid-hal-init11:50
T42<Lukapanio> and i still have usb11:51
T42<Lukapanio> i must ake jornalctl?11:51
T42<Lukapanio> this spammes dmesg Service 'fps_hal' (pid 4747) exited with status 011:51
T42<Lukapanio> [  195.956782] droid-hal-init: Sending signal 9 to service 'fps_hal' (pid 4747) process group... HYBRIS: killing PID instead of process group.11:51
T42<Lukapanio> ouch it seems to be fingerprint related11:52
T42<Lukapanio> will ignore for now11:52
T42<adampigg> mal: i wonder if you can have selinux enabled, and set dhi to be unconfined in the systemd unit11:55
mal@adampigg that would need some patching maybe11:56
mal@adampigg if you look at the sony branch in hybris-patches you can see the patch which allows enabling selinux11:58
T42<Lukapanio> so @elros34  https://paste.ubuntu.com/p/SCqvzt7pbt/ here is journal12:08
T42<Lukapanio> any idea how to fix?12:20
T42<hacker12455> Hmm12:20
T42<hacker12455> Jul 25 19:02:23 Sailfish lipstick[4558]: libminisf is incompatible or missing. Can not possibly start the SurfaceFlinger service. If you're experiencing troubles with media try updating droidmedia (and/or this plugin).== hwcomposer module ==12:20
T42<Lukapanio> idk but dont think it is releted to hwcompser much12:24
T42<hacker12455> I think it may be related12:25
T42<hacker12455> Not sure12:26
T42<Lukapanio> idk12:26
T42<Lukapanio> @elros34  can it be problem?12:26
T42<hacker12455> https://wiki.merproject.org/wiki/Adaptations/faq-hadk#lipstick_segfaults.2Fno_display12:26
T42<hacker12455> https://wiki.merproject.org/wiki/Adaptations/faq-hadk#libminisf.so_not_found12:27
rinigusmal: only place in android *.rc files that was referring to selinux fs was operation 'chmod 0444 /sys/fs/selinux/policy'12:27
rinigusmal: as we use systemd, I wonder if we need to have selinuxfs mount file as well?12:28
rinigusmal: maybe in android init its mounted automatically?12:28
rinigus^ just a guess12:28
rinigus... and repeating earlier question - do we have selinuxfs mounted on any port with selinux enabled?12:29
T42<Lukapanio> @hacker12455  seems like hwcomposer must work without libminisf but not shure12:33
T42<elros34> @Lukapanio worth to try but follow hadk's droidmedia instead linked faq by hacker12455. Looks like hwcomposer 2 is used but your earlier strace was without EGL_PLATFORM also first check droid-hal-$DEVICE.log whether QCOM_BSP is found12:34
malnot sue if hwc2 needs that define12:35
mal*sure12:35
T42<elros34> you mean QCOM_BSP or EGL_PLATFORM?12:35
T42<Lukapanio> + grep -q '^TARGET_USES_QCOM_BSP := true' ./device/motorola/cedric/BoardConfig.mk12:36
T42<Lukapanio> + '[' 8 -ge 5 ']'12:36
T42<Lukapanio> + grep -q '^BOARD_USES_QCOM_HARDWARE := true' ./device/motorola/cedric/BoardConfig.mk12:36
T42<Lukapanio> + echo Found QCOM_BSP12:36
T42<Lukapanio> Found QCOM_BSP12:36
T42<Lukapanio> from droid-hal-cedric.log12:36
mal@elros34 I was talking about QCOM_BSP, but based on that above it was found automatically12:37
T42<Lukapanio> https://bpaste.net/show/D2LVC here is output of EGL_PLATFORM=hwcomposer strace test_hwc2 @elros3412:40
T42<elros34> it fails with "failed to get hwcomposer service" so I would try to build droidmedia. Also /usr/libexec/droid-hybris/system/bin/logcat is useful when android side fails12:44
T42<Lukapanio> any idea?12:58
T42<Lukapanio> @elros34 [it fails with "failed to get hwcomposer servic …], 13.2  GStreamer v1.0?13:09
T42<Lukapanio> in hadk13:09
T42<Lukapanio> ninja: error: '/home/hello/hadk/out/target/common/docs/test-api-stubs-timestamp', needed by '/home/hello/hadk/out/target/common/obj/JAVA_LIBRARIES/android_test_stubs_current_intermediates/classes.jar', missing and no known rule to make it13:23
T42<Lukapanio> comment all jars and apk?13:25
T42<Lukapanio> ninja: error: '/home/hello/hadk/out/target/common/obj/JAVA_LIBRARIES/framework_intermediates/classes.jar', needed by '/home/hello/hadk/out/target/common/obj/JAVA_LIBRARIES/android_test_stubs_current_intermediates/classes.jar', missing and no known rule to make it13:29
T42<Lukapanio> 13:28:19 ninja failed with: exit status 113:29
T42<Lukapanio> will make -j4 libdroidmedia minimediaservice minisfservice libminisf13:30
T42<Lukapanio> now13:30
T42<elros34> that is wrong make command, I told you to use hadk instruction...13:41
T42<Lukapanio> @elros34 [that is wrong make command, I told you to use …], 13.2  GStreamer v1.0?13:53
T42<elros34> yes13:55
T42<Lukapanio> @elros34  i have problems withthat like ninja: error: '/home/hello/hadk/out/target/common/obj/JAVA_LIBRARIES/framework_intermediates/classes.jar', needed by '/home/hello/hadk/out/target/common/obj/JAVA_LIBRARIES/android_test_stubs_current_intermediates/classes.jar', missing and no known rule to make it14:01
T42<Lukapanio> у меня есть порт сейлфиша. раюотает: ничего, кроме свода и от части юсбетоди14:06
T42<Lukapanio> @hacker12455  ideas?14:07
T42<hacker12455> Hmm14:08
T42<Lukapanio> рфмутщ швуф14:38
T42<Lukapanio> have no idea*14:38
T42<Lukapanio> what to do now14:38
T42<Lukapanio> @elros34  what can i do?15:28
T42<hacker12455> hmm15:30
T42<hacker12455> libcore/: Don’t includeJavaLibrary.mk, as Java won’t be available15:30
T42<elros34> we don't build java so you need to figure out based on build log  what needs android_test_stubs_current and all dependecies and get rid of it15:30
T42<hacker12455> where is the build log?15:31
T42<hacker12455> droid-hal-cedric?15:32
T42<elros34> I mean make output15:37
T42<hacker12455> make -j$(nproc --all) $(external/droidmedia/detect_build_targets.sh \$PORT_ARCH$(gettargetarch))15:50
T42<hacker12455> ?15:50
T42<Lukapanio> hm thereis no home/hello/hadk/out/target/common/obj/JAVA_LIBRARIES/android_test_stubs_current_intermediates/classes.jar16:03
T42<Lukapanio> idk how to finde thato16:14
Thaodanlook at hybris patches how to minimize the build process16:26
Thaodanhybris hal doesn't require java, only droid system16:26
Thaodan(during building at least)16:27
T42<hacker12455> This?16:33
T42<hacker12455> https://github.com/mer-hybris/hybris-patches16:33
Thaodanyes16:35
Thaodanthe patches in build make those that you need for the build process16:36
Thaodanwhat is your base?16:36
T42<hacker12455> @Lukapanio16:36
T42<hacker12455> 1516:37
T42<hacker12455> Fwd from Lukapanio: my base is los15.116:37
Thaodanthan you at the patches applied to the Android repos at mer hybris they are like the one from hybris patches but for Android 816:39
T42<hacker12455> where are they?16:40
Thaodanat the same Organisation named something like platform_<name>16:42
T42<hacker12455> can't seem to find the,16:46
malThaodan: it's possible something was commented out from manifest which is needed on some device, assuming everything else was done correctly16:46
T42<elros34> @Lukapanio I hope you didn't just copy/paste make command from hadk because it will fail (not sure what would be built then). How about you provide make output from the first time it fails.17:01
T42<hacker12455> failing of which command?17:03
T42<hacker12455> make -j$(nproc --all) $(external/droidmedia/detect_build_targets.sh \$PORT_ARCH$(gettargetarch))17:03
T42<hacker12455> that one?17:03
T42<elros34> yes, from first execution of that17:04
T42<Lukapanio> @elros34 [@Lukapanio I hope you didn't just copy/paste m …], i tryed to copy paste too it failes but make failes i tryed make -j4 $(external/droidmedia/detect_build_targets.sh $PORT_ARCH)17:04
T42<hacker12455> @elros34 [yes, from first execution of that], ninja: error: '/home/hello/hadk/out/target/common/obj/JAVA_LIBRARIES/framework_intermediates/classes.jar', needed by '/home/hello/hadk/out/target/common/obj/JAVA_LIBRARIES/android_test_stubs_current_intermediates/classes.jar', missing and no known rule to make it17:04
T42<hacker12455> 15:14:45 ninja failed with: exit status 117:04
T42<elros34> whole log17:05
T42<hacker12455> https://bpaste.net/show/LWIC417:06
T42<elros34> It clearly fails: Could not determine android architecture. Please pass it as 2nd argument using gettargetarch17:07
T42<hacker12455> hmm now the error is different17:07
T42<hacker12455> https://bpaste.net/show/Y3BSM17:07
T42<Lukapanio> @hacker12455 [hmm now the error is different], damn why17:08
T42<elros34> because you COPY/PASTE it17:08
T42<hacker12455> oh17:08
T42<elros34> this one is correct: $(external/droidmedia/detect_build_targets.sh $PORT_ARCH $(gettargetarch))17:09
T42<elros34> you should not copy/paste any 2 lines commands from hadk17:09
T42<hacker12455> Okay now it passed17:10
T42<Lukapanio> yay17:10
T42<hacker12455> @elros34 https://bpaste.net/show/VXSVO17:10
T42<hacker12455> should it build this fast?17:11
T42<elros34> no, it should print at the end about libminisf, minimediaservice ..17:16
T42<hacker12455> hmm, what is the correct command then?17:16
T42<hacker12455> or why would it fail?17:17
T42<hacker12455> to print the stuff out17:17
T42<Lukapanio> can it be due to i run  make libdroidmedia_32 minimediaservice minisfservice libminisf_3217:17
T42<Lukapanio> before17:17
mal@elros34 there is now also an easier way to build those if hybris-boot has been updated recently https://github.com/mer-hybris/hybris-boot/commit/7022fa5f4905d5bfb25b7c3ac6be3d277c3fe0c717:23
T42<elros34> yes, I saw it nice addition17:23
malsimilar nice things to build_packages.sh17:24
malfor gst-droid etc17:24
T42<hacker12455> hmm17:31
T42<hacker12455> does it matter where I put the packages in the hw-adaptation pattern?17:31
T42<hacker12455> so far I uncommented gst-droid, added droidmedia, audioflingerglue and the glue module for pulseaudio17:33
T42<hacker12455> running build_packages with the -g switch17:35
T42<hacker12455> anything else I should add to the patterns?17:36
T42<hacker12455> ok so it failed on audioflingerglue17:38
T42<hacker12455> I assume I can omit that for now right?17:38
T42<hacker12455> After commenting audiflingerglue and the module it compiled fine17:42
rinigusevening! disclaimer - I still have selinux enabled on xperia xz2; haven't had a chance to rebuild all without18:24
rinigusmal: I . have . gui!18:24
rinigusin addition to fixing android version flags using mkbootimg, as discussed yesterday, I had to mount selinuxfs on /sys/fs/selinux18:25
rinigus... now will have to find a place where to set it to mount; currently using a script for it.18:27
rinigushow do you figure out which /dev/input/eventX is a correct one?18:27
T42<hacker12455> well you could cat the events18:28
T42<hacker12455> and touch the screen18:28
T42<elros34> rinigus just remove :/dev/input/eventX so it will auto detect it18:30
riniguselros34: thanks, will do!18:30
rinigushacker12455: that should work as well...18:33
T42<DylanVanAssche> rinigus: evdev_trace provides also some interesting information.18:35
rinigusfor some reason auto-detection didn't work, but cat was good enough. now I have touch as well18:37
vknecht🎉 🎉 🎉18:45
rinigusvknecht: early days still ...18:47
vknechtstill good after days of frustration and dull screen, isn't it ? :-)18:48
vknechtdays or, weeks ?18:48
rinigusstartup wizard is not starting and complains "Fatal: the server doesn't advertise the android_wlegl global!Aborted"18:50
riniguslogcat https://dpaste.de/EL0w ; journal https://dpaste.de/GMUa18:51
rinigusvknecht: weeks, I guess. but not too many :)18:51
rinigusfrom logs I can see last hits in 2017...18:52
rinigusand https://github.com/libhybris/libhybris/commit/c7f87904345f19e05a0b9170d9b3898e92e9a910 was supposed to help18:54
rinigusdisclaimer: I am still with selinux enabled18:55
T42<hacker12455> logs from @Lukapanio after adding droidmedia19:15
T42<hacker12455> Fwd from Lukapanio: https://pastebin.com/ftM1W9VL19:15
T42<hacker12455> Fwd from Lukapanio: journal19:15
T42<hacker12455> Fwd from Lukapanio: https://bpaste.net/show/ZE3M219:15
T42<hacker12455> Fwd from Lukapanio: strace19:15
T42<Lukapanio> so @elros34 whats next?19:35
T42<elros34> maybe provide logcat? you could try to run /system/bin/surfaceflinger19:40
T42<Lukapanio> damn something broken19:47
T42<hacker12455> what is broken19:50
T42<hacker12455> logcat?19:50
T42<Lukapanio> https://paste.ubuntu.com/p/gnPs33zWJM/19:52
T42<Lukapanio> @elros3419:52
T42<Lukapanio> so some premission problems19:52
T42<hacker12455> can you check dmesg after running test_hwcomposer?19:53
T42<elros34> that is not the logcat command I gave you earlier19:54
T42<hacker12455> @Lukapanio /usr/libexec/droid-hybris/system/bin/logcat19:56
T42<elros34> are you sure you have uevent*rc in out dir?19:57
T42<Lukapanio> @elros34  sorry https://paste.ubuntu.com/p/VRD7xNkDHk/19:57
T42<hacker12455> @elros34 directly in the out dir or in the root directory in there?19:58
T42<elros34> somwhere there so its found during droid-hal build19:59
T42<Lukapanio> hmmm sh-3.2# ls /ueventd.rc19:59
T42<Lukapanio> /ueventd.rc19:59
T42<Lukapanio> on device19:59
T42<hacker12455> @elros34 ueventd.rc is in out/root20:00
T42<elros34> examine it, also you should try to reboot and get logcat again so it contains more logs from early boot20:02
T42<hacker12455> what exactly should we look at in the file?20:03
T42<elros34> just make sure it contains some rules20:04
T42<hacker12455> it does20:04
T42<hacker12455> https://bpaste.net/show/5XU3A20:05
T42<hacker12455> maybe something is missing there?20:07
malrinigus: nice that you finally got UI20:28
malrinigus: is it possible that you have so much egl extensions that the fix I added is not enough?20:29
rinigusmal: you and many on this channel helped, thanks a lot!20:31
rinigusmal: I don't know regarding egl extensions. is there any way to count them?20:32
rinigus(also possible its related to selinux, still running with that)20:32
malrinigus: add debug printing to libhybris and see what it shows20:34
rinigusmal: is there any option for that or just plugin printf?20:35
malrinigus: can't see any direct printing, unless Qt side prints it somewhere20:37
rinigusmal: will add. looking at your patch, is it just to have longer strings to get unique names for extensions?20:46
malrinigus: yes, those are in a string and we use snprintf so if the source is too long it will not have the string we add to it21:09
rinigusmal: thanks, then making string longer makes sense - 2048 should do for 2 years :) . I'll continue with that tomorrow21:19
malrinigus: it would have been safer to prepend our change to it, but then it would not show the issue this easily21:24
piggzmal: if usb-moded needs config for a new usb ne device name, would connman need something similar? it seems like the firewall isnt applying correctly on the pro121:32
malhmm21:34
piggzmal: cant connect via ssh, unless i clear the iptables21:35
malpiggz: strange21:38
piggzmal: what does connmanctl technologies display for you?21:39
malpiggz: as root?21:42
malpiggz: I have gadget, cellular, wifi, gps, and bluetooth21:43
piggzi only have wifi, bt and gps21:43
piggz(no sim card in)21:43
piggzso, gadget is missing for me21:43
malpiggz: check how different your usb defconfig stuff is compared to for example xa2 or x1021:49
« Tuesday, 2019-10-29 Index Thursday, 2019-10-31 »

Generated by irclog2html.py 2.17.1 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!