T42 | <DarknessHiddenorg> https://pastebin.com/vabrUEvg anyone familiar with this issues? | 09:27 |
---|---|---|
mal | which android base is that? | 10:09 |
T42 | <DarknessHiddenorg> 18.1 (re @SailfishFreenodeIRCBridgeBot: <mal>which android b...) | 10:20 |
T42 | <DarknessHiddenorg> 11 | 10:20 |
mal | @DarknessHiddenorg wondering if you need this patch https://pastebin.com/55feRBJS | 10:52 |
mal | to system/core | 10:52 |
T42 | <elros34> apexd-bootstrap prints that this device does not support updatable APEX so you definitely need this patch. Additionally you are not supposed to disable selinux. Build latest release 4.5.0.21 | 10:58 |
T42 | <DarknessHiddenorg> im disable selinux via kernel cmdline (re @elros34: apexd-bootstrap prin...) | 11:12 |
T42 | <DarknessHiddenorg> and btw isee dsme or something in systemlog | 11:12 |
T42 | <elros34> good dsme should be started | 11:19 |
mal | @DarknessHiddenorg selinux should not be disabled from kernel cmdline, that is said is faq | 11:43 |
T42 | <Walid> sorry for late response, i checked binder-list there is no omx or configstore service, only keymaster (re @SailfishFreenodeIRCBridgeBot: <mal>also checking b...) | 11:57 |
T42 | <DarknessHiddenorg> about the patch should i do clean build or just rebuild? (re @SailfishFreenodeIRCBridgeBot: <mal>to system/core) | 12:05 |
mal | rebuild is enough | 12:09 |
T42 | <elros34> @walid you still didn't show where these services crash or even start | 12:15 |
T42 | <Walid> the crash is just | 14:49 |
T42 | <Walid> droid-hal-init: Could not find 'android.hardware.configstore@1.0::ISurfaceFlingerConfigs/default' for ctl.interface_start | 14:49 |
T42 | <Walid> and solve it byran this | 14:49 |
T42 | <Walid> setprop ctl.start "vendor.media.omx" | 14:50 |
T42 | <Walid> setprop ctl.start "vendor.configstore-hal" | 14:50 |
T42 | <Walid> so setprop resovle the , but must ran manually | 14:50 |
mal | you need to check if something in .rc files or .sh scripts called from .rc files should be doing something similar | 14:54 |
T42 | <elros34> this is not crash, so show whole log | 15:02 |
T42 | <Walid> dmesg ; https://pastebin.ubuntu.com/p/rSwfvqYzSF/ | 15:14 |
T42 | <Walid> journal : https://pastebin.ubuntu.com/p/ZCC6hMC3mk/ | 15:14 |
T42 | <elros34> what's that /usr/libexec/droid-hybris/system/etc/init/android.hardware.media.omx@1.0-service.rc? | 15:16 |
T42 | <elros34> there is mistake in this file | 15:17 |
T42 | <elros34> btw android services logs to logcat | 15:18 |
T42 | <elros34> grep for "vendor.media.omx" in other rc/sh files | 15:20 |
T42 | <DarknessHiddenorg> https://pastebin.com/cJqPuGVp no gui and rndis not work it just show and gone | 15:23 |
T42 | <elros34> @walid I do not see "starting service 'vendor.media.omx" anywere. | 15:23 |
T42 | <DarknessHiddenorg> https://pastebin.com/3eS168Y5 log in my compute dmesg | 15:25 |
T42 | <elros34> looks like you are not booting hybris-boot.img at all | 15:28 |
T42 | <DarknessHiddenorg> this in device log (re @DarknessHiddenorg: https://pastebin.com...) | 15:29 |
T42 | <elros34> its empty | 15:30 |
T42 | <DarknessHiddenorg> https://pastebin.com/tSwQrcHK (re @elros34: its empty) | 15:31 |
T42 | <DarknessHiddenorg> https://irc.thaodan.de/.imgstore/504cdc66/console_ramoops | 15:32 |
T42 | <elros34> again don't disable selinux and next time use init_enter_debug2 to get into device | 15:37 |
T42 | <DarknessHiddenorg> how to do that? (re @elros34: again don't disable ...) | 15:38 |
T42 | <elros34> follow hadk-hot | 15:39 |
T42 | <DarknessHiddenorg> where is it? (re @elros34: follow hadk-hot) | 15:48 |
T42 | <elros34> @walid maybe this is because of apex failure you have | 15:48 |
T42 | <Walid> yes thats what i think, apexd fail (re @elros34: @walid maybe this is...) | 15:51 |
T42 | <elros34> so nothing in logcat? | 15:57 |
T42 | <elros34> if I get it right, inside this function https://android.googlesource.com/platform/system/apex/+/refs/heads/android10-release/apexd/apex_database.cpp#282 | 16:03 |
T42 | <Walid> this is logcat i found interresting error | 16:48 |
T42 | <Walid> https://pastebin.ubuntu.com/p/6pS4KKm2Y6/ | 16:48 |
T42 | <Walid> SELinux : Unknown class service_manager | 16:48 |
T42 | <Walid> in halium they have patch | 16:48 |
T42 | <Walid> https://github.com/Halium/hybris-patches/blob/halium-11.0/frameworks/native/0005-halium-disable-SELinux-checks-in-ServiceManager.patch | 16:48 |
mal | @Walid you are sure you have the selinux configs correctly added? | 17:03 |
T42 | <Walid> # zcat /proc/config.gz |grep SELINUX | 17:06 |
T42 | <Walid> https://pastebin.ubuntu.com/p/2skncTwdKd/ | 17:06 |
T42 | <elros34> in /etc/selinux | 17:07 |
T42 | <Walid> yes | 17:08 |
T42 | <Walid> # ls /etc/selinux/ | 17:08 |
T42 | <Walid> config minimum | 17:08 |
T42 | <Walid> ls /etc/selinux/minimum/ | 17:08 |
T42 | <Walid> contexts policy | 17:08 |
T42 | <Walid> all files is there | 17:09 |
mal | and the folders and files under that minimum folder? pastebin output of "ls -lR /etc/selinux" | 17:10 |
T42 | <Walid> but in cmdline i found duplicatte params | 17:10 |
T42 | <Walid> onsole=tty0 audit=0 selinux=1 androidboot.selinux=permissive........androidboot.selinux=enforcing androidboot.security_mode=1526595585... | 17:10 |
T42 | <Walid> https://pastebin.ubuntu.com/p/Y7z4NSHRYC/ | 17:11 |
mal | so you didn't follow instructions in faq | 17:11 |
mal | is says to copy the content of the files to which those symlinks point to | 17:12 |
T42 | <Walid> oh ok sorry i will recopy it | 17:12 |
T42 | <Walid> same thing | 18:09 |
T42 | <Walid> https://pastebin.ubuntu.com/p/2hWJ874bTS/ | 18:09 |
mal | but you don't have anything about "SELinux : Unknown class service_manager" anymore | 18:13 |
T42 | <Walid> yes | 18:14 |
mal | did you try what I suggested, grepping .rc and .sh files in android folders for the service names, you can find the service name of those in the .rc files which define the services | 18:15 |
mal | there is one possible thing which I saw on some device, some devices add services to one class we don't start by default | 18:16 |
T42 | <Walid> i found thervices omx and configstore under /vendor/etc/init/ | 18:17 |
T42 | <Walid> /vendor/etc/init/android.hardware.media.omx@1.0-service.rc | 18:17 |
T42 | <Walid> but there is no setprop | 18:18 |
T42 | <Walid> so in hadk there this:" Apply patches as on hybris-16 but from the hybris-17.1 branch", when i check this branch there is no system/etc/init.*.rc, only in hybris-17.1-old | 18:31 |
T42 | <elros34> but didn't we already figure out this is most likely because not working apex? I suggest you to debug that function in apexd sources, check mountpoints, for me it looks like it fails because some parsing issue and data is invalid | 18:38 |
T42 | <elros34> btw do you have minimedia* service running? | 18:45 |
T42 | <elros34> 'pgrep -af mini' | 18:46 |
mal | does that device have product partition, I see that "E apexd : Failed to walk /product/apex : Can't open /product/apex for reading : No such file or directory" not sure if that is a problem or not | 18:50 |
T42 | <Walid> no there is no process running (re @elros34: 'pgrep -af mini') | 19:09 |
T42 | <Walid> in fstab there is no product mount , but this folder exist and apex is missing (re @SailfishFreenodeIRCBridgeBot: <mal>does that devic...) | 19:10 |
T42 | <elros34> did you build correct sfos release and droidmedia version? | 19:10 |
T42 | <elros34> and droid-hal submodule? | 19:11 |
T42 | <elros34> about product: do you have /system/product? | 19:13 |
T42 | <Walid> in hadk-env i have this :export RELEASE=4.5.0.21 (commit 956949b6e96cd5fec7a85ff7a8fdd81156db39f9 (HEAD, tag: 0.20230605.1, m/hybris-17.1, github/master) | 19:18 |
T42 | <Walid> | 19:18 |
T42 | <Walid> droidmedia : (HEAD detached at 956949b) | 19:18 |
T42 | <Walid> /rpm/dhd : * (HEAD detached at 960d6af) | 19:18 |
T42 | <Walid> droid-configs/droid-configs-device : * (HEAD detached at 414d96c) | 19:18 |
T42 | <Walid> in hadk-env i have this :export RELEASE=4.5.0.21 | 19:18 |
T42 | <Walid> | 19:18 |
T42 | <Walid> droidmedia : (HEAD detached at 956949b) (commit 956949b6e96cd5fec7a85ff7a8fdd81156db39f9 (HEAD, tag: 0.20230605.1, m/hybris-17.1, github/master) | 19:18 |
T42 | <Walid> /rpm/dhd : * (HEAD detached at 960d6af) | 19:18 |
T42 | <Walid> droid-configs/droid-configs-device : * (HEAD detached at 414d96c) (edited) | 19:18 |
T42 | <Walid> yes i have (re @elros34: about product: do yo...) | 19:18 |
T42 | <Walid> ls -al /system/product/ | 19:19 |
T42 | <Walid> https://pastebin.ubuntu.com/p/ZsKc7FdjpT/ | 19:19 |
T42 | <elros34> so do you have minimedia*rc in /usr/libexec/droid-hybris/*** | 19:23 |
T42 | <elros34> is it parsed during boot? | 19:24 |
T42 | <Walid> there is no minimedia*.rc, i just have this hybris-17.1-old | 19:30 |
T42 | <Walid> disabled_services.rc ecclist.rc hwservicemanager.rc hybris_extras.rc ofono-ril-ecclists.rc servicemanager.rc vndservicemanager.rc (re @elros34: so do you have minim...) | 19:30 |
T42 | <Walid> there is no minimedia*.rc, i just have this from hybris-17.1-old | 19:31 |
T42 | <Walid> disabled_services.rc ecclist.rc hwservicemanager.rc hybris_extras.rc ofono-ril-ecclists.rc servicemanager.rc vndservicemanager.rc (edited) (re @elros34: so do you have minim...) | 19:31 |
T42 | <elros34> why hybris-17.1-old? 'rpm --info -ql droidmedia' | 19:34 |
T42 | <Walid> https://pastebin.ubuntu.com/p/bvCBRZNzhR/ | 19:38 |
T42 | <Walid> https://github.com/sailfishos-oneplus5/droid-config-cheeseburger/tree/hybris-17.1.old/sparse/usr/libexec/droid-hybris/system/etc/init | 19:39 |
mal | @elros34 the mini*rc files are not there because dhd 960d6af didn't yet have the commit | 19:39 |
T42 | <Walid> /sparse/usr/libexec/droid-hybris/system/etc/init there is two branch hybris-17.1 which empty and hybris-17.1-old contains rc files | 19:40 |
T42 | <elros34> mal ah ok | 19:40 |
mal | aren't the services in one of the .rc files | 19:41 |
mal | was it servicemanager.rc | 19:41 |
T42 | <Walid> servicemanager.rc:service minimedia /usr/libexec/droid-hybris/system/bin/minimediaservice | 19:42 |
T42 | <elros34> I dont full get why do you even copy anything from that onplus5 project? | 19:46 |
T42 | <elros34> this is example project for hybris-16 not 17 | 19:46 |
T42 | <elros34> on 17 you set android_version_major which takes care of correct sparse | 19:47 |
T42 | <Walid> i just copy system/etc/init/*.rc files | 19:51 |
T42 | <Walid> from hybris-17.1-old | 19:52 |
T42 | <elros34> did you set android_version_major? I still think not everything starts because of missing apex so maybe it's not worth to search why minimedia does not start | 20:02 |
T42 | <Walid> yes i added it, (re @elros34: did you set android_...) | 20:05 |
T42 | <Walid> so i added symlink to /apex | 20:07 |
T42 | <Walid> com.android.media -> /system/apex/com.android.media | 20:07 |
T42 | <Walid> com.android.media.swcodec -> /system/apex/com.android.media.swcodec | 20:07 |
T42 | <Walid> audio work | 20:08 |
T42 | <elros34> can you show /proc/mounts? | 20:12 |
T42 | <Walid> https://pastebin.ubuntu.com/p/CpRFxcjkJ9/ (re @elros34: can you show /proc/m...) | 20:23 |
T42 | <Walid> when i added the two symlink, there is no omx error but only configstore, and to get gui just ran setprop | 20:25 |
T42 | <elros34> /proc/mounts would be more useful because that is what apex code use and fail | 20:25 |
T42 | <Walid> bash: /proc/mounts: Permission denied | 20:26 |
T42 | <Walid> i ran devel-su /su | 20:26 |
T42 | <Walid> i'm root | 20:26 |
T42 | <elros34> cat /proc/mounts, you dont execute sysfs | 20:26 |
T42 | <Walid> yes | 20:31 |
T42 | <Walid> https://pastebin.ubuntu.com/p/ydwtt22grT/ | 20:31 |
T42 | <elros34> have you tried starting apex manually? Not sure if its possible | 20:36 |
T42 | <Walid> core dump (wait for service manager and others error, (re @elros34: have you tried start...) | 20:43 |
T42 | <elros34> but does it prints regular scanning messages either in terminal or in logcat? | 20:43 |
T42 | <elros34> I wonder if double "/apex/com.android.runtime@1" in mounts could cause this issue | 20:46 |
T42 | <Walid> https://pastebin.ubuntu.com/p/887TkMVmqZ/ | 20:50 |
T42 | <elros34> ok so you can try to 'strace -f -s512 ' it | 20:52 |
T42 | <elros34> assuming this logcat snippet is from the time you run apexd manually not some leftover from boot or service restarting | 20:54 |
T42 | <Walid> https://pastebin.ubuntu.com/p/nQcz5gqcSy/ | 20:55 |
T42 | <Walid> can this cause apexd to strat | 21:01 |
T42 | <Walid> Waiting for service 'vold' | 21:01 |
T42 | <elros34> not sure what needs vold but it's probably disabled in disabled_services.rc | 21:02 |
T42 | <elros34> anything change if you run "umount /apex/com.android.runtime@1"? | 21:05 |
T42 | <Walid> umount: /apex/com.android.runtime@1: target is busy. | 21:06 |
T42 | <elros34> what findmnt shows about that runtime@1? | 21:09 |
T42 | <Walid> https://pastebin.ubuntu.com/p/yP4wp7M9qw/ | 21:14 |
T42 | <elros34> I think apexd-bootstrap mounts it twice for some reason | 21:19 |
T42 | <elros34> you could try https://paste.opensuse.org/pastes/48e4b4ae4f7d to see if it automagically change something in mount output and fix apexd | 21:45 |
T42 | <Umeaman> When will hybris-20.0 be released? My phone is currently on Android 13 and it has rollback protection. | 21:47 |
mal | well hybris-20.0 is possible but just needs some additional work to port, like using custom branches in some places | 22:33 |
Generated by irclog2html.py 2.17.1 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!