| Mister_Magister | poetaster: and i have 12 core xeon processor :D | 02:02 |
|---|---|---|
| Mister_Magister | not to mention that it takes ages for new sfos versions to be available on obs | 02:02 |
| T42 | <rodrisola> @elros34 https://pastebin.com/fbtVisH2 | 09:03 |
| T42 | <rodrisola> After struggling with the equipment, I realized that an update, returned the data partition as dynamic, and did not get to mount, what I did was downgrade firmware, I followed your advice and went steps back, now the phone starts me up and Modify the init_debug as you told me to run, at last I was able to automatically mount the system and adjacent partitions, but I had to put a sleep 2 in droid-init-early.sh, now it mounts everything, and | 09:03 |
| T42 | <rodrisola> Journalctl | 09:20 |
| T42 | <elros34> @rodrisola did you apply hybris patches? "Oct 27 22:36:54 mi11lite ueventd: ueventd started!". Still something is not mounted as it should because droid-hal-init fails for the first time, you have bunch of failing systemd mount units so I will not even guess which one is correct | 09:57 |
| T42 | <elros34> you talk about /init-debug but I see kernel reports mounted partitions several seconds later so I am not sure what you are doing. Maybe you should wait for somebody who did all of this for hybris-18 and get from him full step by step instruction | 10:05 |
| T42 | <elros34> Does 'grep ueventd /init.rc' return commented out '#start ueventd'? | 10:10 |
| T42 | <rodrisola> I think they are my attempts to mount sustem vendor odm ... etc ... so I'll put a mask on them, so as not to cause an error; on the other hand if I apply the patches but from a branch of android11, it seemed the most appropriate; now I will try the grep command that you commented | 11:18 |
| T42 | <rodrisola> https://pastebin.com/4FGBuTzA | 11:32 |
| T42 | <elros34> is init.rc in different place now? check with rpm -ql droid-hal-<device codename> or grep for ueventd in other places | 11:35 |
| T42 | <rodrisola> i try rpm -ql this a log | 11:46 |
| T42 | <rodrisola> https://pastebin.com/A3QC6CnQ | 11:46 |
| T42 | <elros34> so it's in different place, does it contains commented out ueventd? | 11:51 |
| mal | which android base is that? | 11:59 |
| T42 | <rodrisola> 18.1 | 12:02 |
| T42 | <rodrisola> cat init.rc | 12:05 |
| T42 | <rodrisola> https://pastebin.com/0Bxg1wc2 | 12:05 |
| T42 | <elros34> so looks ike it's disabled but still ueventd its started according to your log. | 12:10 |
| T42 | <elros34> maybe first focus on getting droid-hal-init working for the first time instead after few restarts | 12:15 |
| T42 | <rodrisola> journalctl https://pastebin.com/TBc7J4Vy | 12:15 |
| mal | 18.1 uses init.rc from /usr/libexec/droid-hybris/system/etc/init/hw/init.rc | 12:17 |
| mal | btw, 18.1 still has issues with init | 12:18 |
| T42 | <rodrisola> so I create an ln -s? | 12:19 |
| T42 | <elros34> mal: yes I notice it, any idea why ueventd is started even when start ueventd is commented out? | 12:20 |
| T42 | <elros34> @rod | 12:20 |
| T42 | <elros34> @rodrisola you should fix your mount points, how long you will mess with them? | 12:20 |
| T42 | <elros34> and to not symlink init.rc | 12:20 |
| T42 | <elros34> do not* | 12:21 |
| mal | what symlink are you talking about? | 12:23 |
| T42 | <rodrisola> They are already created, @ elros34, because I make it load in the init_debug, and from there my service loads before local_fs.target | 12:23 |
| T42 | <rodrisola> about create in ln -s /*/*/*/droid/init.rc /init.rc | 12:25 |
| T42 | <elros34> no they are not: from your log: https://pastebin.com/TBc7J4Vy, line 219, 607, 616, 619 and so on, all of them are after line 15 "Welcome to Sailfish OS 4.3.0.12 (Suomenlinna)" so none of them are mounted in init-debug. SO read your log once again and try to understand it | 12:27 |
| mal | https://github.com/mer-hybris/hybris-patches/blob/hybris-18.1/system/core/0019-hybris-Use-services-from-usr-libexec-droid-hybris-sy.patch#L26 just to not that reading from /usr/libexec/droid-hybris/system/etc/init/hw/init.rc is the way | 12:30 |
| mal | *note | 12:30 |
| T42 | <elros34> @rodrisola and some of them are before line 543 when droid-hal-init starts | 12:34 |
| T42 | <elros34> obviously I should say after instead before | 12:36 |
| T42 | <rodrisola> mmm @elros34, i have this in my init_debug /sbin/parse-android-dynparts /dev/sda25 | sh | 12:43 |
| T42 | <rodrisola> after in my mount-android-dynpart.service, one scrip to exec to mount | 12:44 |
| T42 | <elros34> then execute that script just "after /sbin/parse-android-dynparts /dev/sda25" instead. Is it really in /sbin? Here it suggest /usr/bin/: https://github.com/sailfish-on-nabu/parse-android-dynparts/blob/master/rpm/parse-android-dynparts.spec#L43 | 12:46 |
| T42 | <rodrisola> mal : i apply this patch of the hybris-18.1 | 12:47 |
| T42 | <rodrisola> yes, but I edited it to install in bin | 12:49 |
| T42 | <rodrisola> these lines are in my init-debug | 13:16 |
| T42 | <rodrisola> /sbin/parse-android-dynparts /dev/sda25 | sh | 13:16 |
| T42 | <rodrisola> /usr/sbin/mount-android-dynpart.sh | 13:16 |
| T42 | <rodrisola> this is dmesg.log | 13:17 |
| T42 | <rodrisola> https://pastebin.com/wHpLfER1 | 13:17 |
| T42 | <rodrisola> this is my init.log | 13:24 |
| T42 | <rodrisola> https://pastebin.com/A5d8VpWc | 13:24 |
| T42 | <rodrisola> If you realize at the end mount the partitions, and as you can see I put findmnt so that it can be seen in the log, that the partitions are mounted | 13:25 |
| T42 | <elros34> looks good so now grep for ueventd in all init*rc files you can find on device, something starts it but AFAIK it should be not used unless something changed for hybris-18 | 13:34 |
| T42 | <elros34> Just before ueventd is started there are some suppressed messages so also add printk.devkmsg=on to kernel cmdline | 13:39 |
| T42 | <rodrisola> https://pastebin.com/zutk6mCj @elros34, this is a grep | 15:32 |
| T42 | <rodrisola> https://irc.thaodan.de/.imgstore/080de081/dmesg.log | 16:05 |
| T42 | <elros34> so what did you change? logs at least contains something useful now | 16:09 |
| T42 | <elros34> it was clear that init.rc have disabled ueventd from your previous paste but there are more init*rc files than /usr/libexec/droid-hybris/system/etc/init/hw/, like in /system, Look at your dmesg, it contains path to .rc files. I have no idea what else might start it: "droid-hal-init: starting service 'ueventd'." | 16:11 |
| mal | the issue seems to be apex | 16:26 |
| mal | those apex mounts fail which causes failure | 16:27 |
| T42 | <rodrisola> Elros, [01/04/22 11:09] | 18:00 |
| T42 | <rodrisola> so what did you change? logs at least contains something useful now | 18:00 |
| T42 | <rodrisola> Elros, [01/04/22 11:11] | 18:00 |
| T42 | <rodrisola> it was clear that init.rc have disabled ueventd from your previous paste but there are more init * rc files than / usr / libexec / droid-hybris / system / etc / init / hw /, like in / system, Look at your dmesg , it contains path to .rc files. I have no idea what else might start it: "droid-hal-init: starting service 'ueventd'." | 18:00 |
| T42 | <rodrisola> T42, [01/04/22 11:26] | 18:00 |
| T42 | <rodrisola> <mal> the issue seems to be apex | 18:01 |
| T42 | <rodrisola> T42, [01/04/22 11:27] | 18:01 |
| T42 | <rodrisola> <mal> those apex mounts fail which causes failure | 18:01 |
| T42 | <rodrisola> I added cmd in the kernel, what @elros34 told me and now it seems to be more complete, the detail is that when I start the droid-hal-init the phone restarts in fastboot mode. I have to restart it manually, that means that the droid-hal-init is executed, what I do to obtain data from the dmesg is to put it to sleep early so that it can execute; mal: what do you mean by apex? | 18:01 |
| T42 | <rodrisola> I added cmd in the kernel, what @elros34 told me and now it seems to be more complete, the detail is that when I start the droid-hal-init the phone restarts in fastboot mode. I have to restart it manually, that means that the droid-hal-init is executed, what I do to obtain data from the dmesg is to put it to sleep early so that it can execute; mal: what do you mean by apex? | 18:01 |
| T42 | <elros34> did you grep -r ueventd in /system/etc/init/ and /vendor/etc/init/? | 18:07 |
| mal | android init does some apex related things at boot and those need to succeed otherwise device will reboot | 18:50 |
Generated by irclog2html.py 2.17.1 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!