| nightishaman | https://pastebin.com/ShMQCUMg | 00:04 |
|---|---|---|
| nightishaman | i changed into defaultuser using su and ran lipstick with the correct options manually | 00:04 |
| nightishaman | this is what happens | 00:04 |
| mal | probably the android side process are not ok because linkerconfig was failing earlier | 00:11 |
| nightishaman | yeah but why does it fail | 00:34 |
| nightishaman | for some reason it also waits 2 minutes before continuing with booting here: https://pastebin.com/vDv96nk5 | 01:08 |
| mal | I wonder if all modules in system side have been loaded? I don't mean vendor_boot but the modules in /lib/modules in roofs | 01:16 |
| mal | *rootfs | 01:16 |
| nightishaman | according to a previous log it detected the modules and loaded them in droid-hal-init | 01:17 |
| nightishaman | because stock os has an init.rc for that | 01:17 |
| mal | could you run "getprop vendor.all.modules.ready" | 01:18 |
| nightishaman | the reason it waits 2 mins is because it runs udev-settle and that timeouts in 2 minutes | 01:19 |
| nightishaman | i just found that out | 01:19 |
| mal | one thing is that loading modules in sfos side might affect that | 01:20 |
| nightishaman | its not set | 01:21 |
| mal | that could affect how android side works | 01:21 |
| mal | if android side i.e. droid-hal-init thinks module loading was not fully done then it will not continue startup properly | 01:22 |
| mal | on fp6 which is also android 16, I copied the module lists from the other locations, vendor_dkml and system_dlkm and where ever you might have those (excluding vendor_boot) and added those to sparse/etc/modules.load.d/modules.conf which just contain module names as list (with path or .ko file extension) | 01:24 |
| mal | now with modules loaded in sfos side I added /usr/libexec/droid-hybris/system/etc/init/modules_load.rc with content | 01:25 |
| mal | on early-init setprop vendor.all.modules.ready 1 | 01:25 |
| mal | sorry | 01:25 |
| mal | this content https://pastebin.com/Tffu2RT1 | 01:25 |
| nightishaman | vendor.all.modules.ready seems to be set nowhere tho | 01:26 |
| mal | hmm | 01:26 |
| nightishaman | i cant find an rc file that does it | 01:27 |
| nightishaman | only vendor.common.modules.ready is set | 01:27 |
| mal | ok, then you have different name for it | 01:27 |
| mal | and is that set now? | 01:27 |
| nightishaman | yes vendor.common.modules.ready was set | 01:27 |
| mal | ok | 01:27 |
| nightishaman | setting vendor.all.modules.ready also didnt do anything | 01:27 |
| nightishaman | i tried setprop and let it settle | 01:27 |
| nightishaman | the only mention of vendor.all.modules.ready is that the haptics processor wont enable before it is set | 01:30 |
| nightishaman | so i might add that actually to make sure | 01:30 |
| nightishaman | it seems to be stuck restarting vendor.citadeld and vendor.keymint-trusty, the droid-hal-init | 01:33 |
| mal | what do you see in logcat? | 01:33 |
| mal | one option is to try to strace droid-hal-init | 01:34 |
| mal | nightishaman: wait, I see in that journal it's waiting for some device nodes to appear, are you sure none of the .mount services in /usr/lib/systemd/system/ contain some path which contain "by-name" or something | 01:36 |
| nightishaman | mal: some do, mnt-vendor-modem_img.mount for example | 01:38 |
| mal | fix those | 01:38 |
| nightishaman | and how | 01:40 |
| nightishaman | should i use the direct sda id? | 01:40 |
| mal | yes, the actual device node like is used in those that were correctly handled by fixup-mountpoints | 01:41 |
| nightishaman | it was only that one | 01:44 |
| mal | some other fails also, maybe different reason | 01:46 |
| mal | mnt-vendor-efs_backup.mount | 01:46 |
| mal | also mnt-vendor-persist.mount fails | 01:46 |
| nightishaman | mnt-vendor-persist already uses /dev/sda1 | 01:48 |
| mal | mount: /mnt/vendor/persist: wrong fs type, bad option, bad superblock on /dev/sda1, missing codepage or helper program, or other error. | 01:49 |
| mal | check that the partition is correct | 01:49 |
| nightishaman | it is not mentioned in by-name, idk whether sda1 is the partition | 01:52 |
| nightishaman | nvm, im just blind, the partition is correct | 01:53 |
| nightishaman | it seems to be mounted as f2fs by the mount point but its either f2fs or ext4 according to fstab | 01:54 |
| nightishaman | it is an ext4 filesystem | 01:58 |
| nightishaman | so the mountpoint is wrong | 01:58 |
| nightishaman | https://pastebin.com/UECQgZpz | 02:03 |
| nightishaman | fixed the mountpoints | 02:03 |
| nightishaman | no progress | 02:03 |
| nightishaman | https://pastebin.com/MjtsdGgL | 02:10 |
| nightishaman | logcat shows this | 02:10 |
| nightishaman | for keystore2 libandroidicu.so is missing | 02:11 |
| nightishaman | i can only find it in apexes | 02:11 |
| nightishaman | also libtrusty: tipc_connect: cannot open tipc device "/dev/trusty-ipc-dev0": No such file or directory | 02:12 |
| mal | you fixed modem_img mount incorrectly, there should be no block/ in the path, just /dev/sda29 | 02:13 |
| nightishaman | i fixed it now just /dev/sda29 | 02:20 |
| nightishaman | udevadm settle still waits for 2 minutes | 02:20 |
| nightishaman | im really confused | 02:24 |
| nightishaman | https://pastebin.com/NnvYmURt | 02:24 |
| nightishaman | keystore2 starts fine earlier but later it doesnt work anymore | 02:25 |
| nightishaman | but it also has the problem that the keymint trusty doesnt start | 02:25 |
| nightishaman | its waiting for it | 02:25 |
| mal | you should try to strace the droid-hal-init process using strace -p pid_of_droid-hal-init, check the pid to use from ps aux | 02:27 |
| nightishaman | https://pastebin.com/i2U5NcZP | 02:36 |
| nightishaman | https://pastebin.com/fh7DCsp8 | 03:01 |
| nightishaman | my trusty doesnt seem to start properly | 03:01 |
| nightishaman | okay, problem, trusty-virtio is loaded before trusty-ipc | 03:35 |
| nightishaman | question is why does android recovery load it in the correct order, but my load_kernel_modules doesnt | 03:36 |
| T42 | <Mister_Magister> modules.dep idk | 07:31 |
| mal | nightishaman: what does you module loading use? | 12:03 |
| mal | *your | 12:03 |
| mal | insmod or modprobe? | 12:03 |
| Mister_Magister | bro ims.apk is goldmine, it decompiles co clearly | 13:54 |
| Mister_Magister | mal: yop, callstatechanged is 26 not 24 as in the plugin, i was correct with my assumption, 25 is call state changed in 1.1, 24 is onSsacStatusChangedIndication on 1.0 | 13:57 |
| Mister_Magister | wait now if i correct the indication id's and request id's and response id's, i should be able to use 1.2 interface | 13:58 |
| Mister_Magister | basically it has onVopsStatusChangedIndication and onSsacStatusChangedIndication on 1.0 additionally | 14:19 |
| Mister_Magister | I think i'll bump it to 1.5 regardless | 14:19 |
| Mister_Magister | and yeah request of 1.1 hangup is id 45 not 41 | 14:24 |
| Mister_Magister | even cancelModifyResponse is wrong on 1.0 | 14:27 |
| Mister_Magister | maybe that'll fix rejecting a call | 14:27 |
| Mister_Magister | cool beans there's no 1.1 hangup response | 14:28 |
| T42 | <Mister_Magister> theres literally no 1.1 or 1.2 responses | 14:31 |
| Mister_Magister | I think sending sms on 1.2 works now but i neex other end to confirm | 15:10 |
| Mister_Magister | at least messages app said it's sent which is new | 15:11 |
| Mister_Magister | yo i just fixed sms send and receive on volte hell yeah | 15:23 |
| Mister_Magister | still cannot reject a call | 15:23 |
| T42 | <nightishaman> mal: I modified the first stage Android init to do the kernel module loading and then exit immediately after that, since the order of loading the way Android init does it is important. | 15:34 |
| mal | loading of which modules? | 15:35 |
| mal | all? | 15:35 |
| T42 | <nightishaman> Everything from vendor kernel boot | 15:35 |
| T42 | <nightishaman> It’s run in the init ramdisk | 15:35 |
| mal | that is different from what I was talking about | 15:36 |
| mal | I meant the other modules which are loaded later in the boot from rootfs | 15:36 |
| T42 | <nightishaman> Id have to check that, but the thing currently is, trusty-virtio loads before trusty-ipc even tho trusty-virtio has a soft dependency on trusty-ipc. In Android recovery it loads correctly, but in mine it doesn’t which causes the keymint to fail | 15:39 |
| Mister_Magister | mal: i enabled the not implemented hangup but switched it from 1.2 interface to 1.2 and it liked that | 15:47 |
| Mister_Magister | everything works now | 15:47 |
| mal | @nightishaman you didn't say what you use for module loading, modprobe or insmod? | 15:55 |
| T42 | <the_hiktor> good day, i was building mw packages but it fails | 15:59 |
| T42 | <the_hiktor> https://paste.opensuse.org/pastes/2e4bddbf7363 | 15:59 |
| T42 | <the_hiktor> here's the logs | 15:59 |
| T42 | <the_hiktor> https://paste.opensuse.org/pastes/08424efff4d6 | 15:59 |
| mal | are you building against 5.0? | 16:04 |
| T42 | <the_hiktor> wdym ? | 16:05 |
| T42 | <the_hiktor> still dumb to understand that 🥲 | 16:06 |
| T42 | <the_hiktor> if you could explain in more noob friendly lamguage 🥲 | 16:07 |
| Mister_Magister | I'm not a noob and i'm also confused, but I think he means if you're buliding on 5.0 target | 16:08 |
| Mister_Magister | as in if your target is 5.0 sfos | 16:08 |
| T42 | <the_hiktor> yeah latest | 16:09 |
| Mister_Magister | latest is not necessarily 5.0 so verify it | 16:09 |
| T42 | <the_hiktor> i just updated to 5.0.0.77 yesterday so yeah | 16:10 |
| mal | maybe go to folder hybris/mw/libhybris | 16:10 |
| mal | and check with git diff if it's latest revision | 16:10 |
| mal | and that the submodule has been updated also | 16:10 |
| T42 | <the_hiktor> that folder doesnt exist | 16:13 |
| Mister_Magister | maybe it's in external | 16:15 |
| Mister_Magister | try external/libhybris | 16:15 |
| T42 | <elros34> What version you have when you run this: "sb2 -t $VENDOR-$DEVICE-$PORT_ARCH -m sdk-install -R ssu re", simillar for tooling | 16:17 |
| T42 | <the_hiktor> Mister_Magister git diff outputs nothing | 16:19 |
| T42 | <the_hiktor> @elros34 Device release is currently: 5.0.0.62 | 16:19 |
| T42 | <elros34> same for tooling? | 16:20 |
| mal | one thing to try is to clean the submodule by going to external/libhybris/libhybris and running "git clean -x -f -d" | 16:21 |
| mal | in case there are some things left from earlier build | 16:22 |
| T42 | <the_hiktor> @elros34 idk but i did update tooling and target manually yesterday cuz i got a new laptop | 16:23 |
| T42 | <elros34> then check is same command just use target name like -t `SailfishOS-5.0.0` | 16:23 |
| mal | did you also update platform sdk? | 16:25 |
| T42 | <Mister_Magister> @b100dian I've successfully disabled denoising and sharpening! thanks! | 16:55 |
| Mister_Magister | mal: are you willing to help me debug video recording on camera2 api? | 16:57 |
| Mister_Magister | I open sensorfw file and i see glorious "Copyright (c) 2025 Jollyboys Ltd." :3333 | 17:08 |
| T42 | <b100dian> Mister_Magister: good. What about exposure, is it applied at camera startup? | 17:08 |
| Mister_Magister | stay jolly | 17:08 |
| Mister_Magister | @b100dian i have no clue what you mean | 17:08 |
| T42 | <b100dian> If you change the exposure, the slider value is persisted for the next time you use the camera. But is the value applied at the next startup? | 17:08 |
| T42 | <b100dian> rinigus: hmm maybe we should switch the lpac env wrapping to something like lpac-gbinder package and ship that instead? | 17:11 |
| Mister_Magister | @b100dian no it is not | 17:12 |
| T42 | <b100dian> hope that you can unsee it now:D | 17:17 |
| Mister_Magister | what do you mean? | 17:17 |
| T42 | <b100dian> I mean I hope it isn't bothersome | 17:19 |
| T42 | <b100dian> I hoped that the new camera2 branch doesn't have this. I have a hack but it is very dirty, and not sure it applies there. | 17:20 |
| Mister_Magister | no i couldn't care less, in fact, the fact that it persists annoys me | 17:20 |
| Mister_Magister | because it's something you adjust when you're taking a picture so defaulting to neutral is good thing | 17:21 |
| T42 | <b100dian> might mean that you have sane defaults. Sake didn't | 17:21 |
| Mister_Magister | btw i added to sake file needed for my ports list, i hope you dont mind, you can fill in the rest of the data if you feel like it | 17:22 |
| T42 | <b100dian> where are you documenting what file and what contents are needed? | 17:23 |
| Mister_Magister | on the website | 17:23 |
| Mister_Magister | portslist.verdanditeam.com | 17:23 |
| Mister_Magister | the example file has all possible values and all of them are optinoal | 17:23 |
| T42 | <b100dian> I see it now. I thought you mentioned also in a forum message. | 17:24 |
| T42 | <b100dian> btw, it is very funny what's broken on sake, `BROKEN=Cannot unlock bootloader` | 17:25 |
| T42 | <b100dian> Like, all the things work but you can't have it | 17:25 |
| Mister_Magister | :P | 17:25 |
| T42 | <b100dian> (no, not all the things work) | 17:25 |
| Mister_Magister | if all things work please add that | 17:25 |
| Mister_Magister | :) | 17:25 |
| Mister_Magister | didn't know what works so i didn't fill that field in | 17:25 |
| Mister_Magister | maybe someone can try the new bootloader exploit on sake | 17:25 |
| Mister_Magister | there are max 5 elements on broken/working list | 17:27 |
| T42 | <b100dian> is there anything newer? not that I can try it | 17:27 |
| Mister_Magister | what do you mean? | 17:27 |
| T42 | <the_hiktor> @elros34 sorry but i dont know what do you want exactly , if you wanna know if targets and tooling are using the same version then yes i installed the same version | 17:27 |
| T42 | <the_hiktor> but nvm problem solved alredy | 17:28 |
| T42 | <the_hiktor> did what mal suggested | 17:28 |
| T42 | <b100dian> Mister_Magister: I mean a new exploit for the bootloader | 17:28 |
| Mister_Magister | well i just heard there's some exploit but i didn't inquire further | 17:29 |
| Mister_Magister | as my sake is unlocked :P | 17:29 |
| T42 | <b100dian> anyway, even if it were, the number of people that would use that to install probably not worh it.. it was a learning project and that's it | 17:29 |
| T42 | <b100dian> I would've changed the linege base to a newer one if it would've benn still unlockable | 17:30 |
| Mister_Magister | would be cool if asus wasn't ass and sus | 17:30 |
| Mister_Magister | it's quite smol | 17:30 |
| rinigus | b100dian: its not really binder only. more a wrapping and making it possible to have configs | 18:04 |
| Mister_Magister | rinigus: I've fixed my qti issues :333 | 18:05 |
| rinigus | maybe, while lpac is still rather new, we can push this wrapper to be standard. when other ports will start to support esims | 18:05 |
| Mister_Magister | ims.apk was goldmine | 18:05 |
| rinigus | it is. great source for proper data | 18:05 |
| Mister_Magister | i set correct indication and request ids (most of response ids from 1.1 and 1.2 were straight up missing tho idk where people get them from) | 18:06 |
| Mister_Magister | and i enabled commented out hang up implementation but sent it on 1.0 interface and now I can reject incoming call (tho my american user cannot for some reason so maybe i need to fix it on 1.2 interface) | 18:06 |
| Mister_Magister | but can send and receive sms and that's huge win | 18:07 |
| Mister_Magister | currently trying to add moto gestures to sensorfw but my changes made it completely poop the bed | 18:08 |
| rinigus | it is :) . these apk's are great source of information. just have to decompose and take time to study the code | 18:08 |
| Mister_Magister | not decompose :P | 18:09 |
| rinigus | :) | 18:11 |
| Mister_Magister | rinigus: do you perchance have made any ports? | 18:11 |
| rinigus | Mister_Magister: tama and with @b100dian nagara | 18:11 |
| Mister_Magister | rinigus: would be really cool if you added a file for my ports list :) | 18:12 |
| Mister_Magister | just https://github.com/VerdandiTeam/droid-config-pipa/blob/master/.portslistinfo file like this in droid-configs, if you don't feel like filling something, you can skip the line | 18:12 |
| Mister_Magister | would help me out :) | 18:13 |
| rinigus | we have for these ports proper docs and corresponding sites. those are not related to droid-config though but have separate repos for it. or should I stuff it to features? so, please add to the .portslistinfo URL that users can get over to. looking at the file, multiple authors supported? without emails? | 18:15 |
| Mister_Magister | whoa whoa slow down, my ports list just collects info from all the ports under nemo:testing:hw, and all of them has droid-configs, so if you add the file in droid-configs it will be automatically picked up :) currently just one author is allowed but i can fix that, same with email. I guess i can add url to respective like thread or site | 18:17 |
| Mister_Magister | give me like 5 minutes, maybe 15 | 18:17 |
| Mister_Magister | rinigus: feel free to add file with author and email separated by , and with URL= field so that i have something to test on :P | 18:19 |
Generated by irclog2html.py 2.17.1 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!