| deathmist | I'll check the IRC logs once I wake up for possible answers, but gonna go sleep now so cya! | 00:05 |
|---|---|---|
| mal | deathmist: have you verified that the build generates ueventd.rc to out/ ? some builds didn't do that and it has caused all sorts of issues | 09:20 |
| mal | deathmist: also verify that you have some fstab file in out/ | 09:21 |
| deathmist | aha! there is no .rc files in out/ | 09:29 |
| deathmist | should I rebuild hybris-hal? | 09:29 |
| deathmist | I ran 'make clean' because I was solving issues with the defconfig, but didn't remember to 'make hybris-hal' again, just hybris-boot and recovery | 09:33 |
| deathmist | I'll start with that and report in a bit if I get further with another boot image | 09:34 |
| mal | deathmist: hybris-hal won't help, run make modules and check the output for ueventd.rc or something like that | 09:41 |
| mal | then you need to manually run make some_target to build the file | 09:41 |
| mal | deathmist: how did you search for the file? try with: find out/ -name ueventd* | 09:42 |
| deathmist | I searched the same otherwise but just "*.rc", nothing with "ueventd*" either after 'make modules' | 09:56 |
| deathmist | also what does 'some_target' refer to? just 'cheeseburger' in my case? | 09:58 |
| mal | no, it's something you need to find out, that was just placeholder for the real thing that is unknown | 10:02 |
| mal | so do you find ueventd* in device/ folder somewhere? | 10:02 |
| deathmist | yes, there is 'device/oneplus/msm8998-common/init/vendor/ueventd.rc' | 10:03 |
| mal | try copying it manually to out/target/product/$DEVICE/root/ueventd.rc | 10:10 |
| mal | and then "build_packages.sh -d" and rebuild image | 10:11 |
| r0kk3rz | mal: what is the issue with the property area thing? | 10:24 |
| deathmist | even after placing the file in 2 more directories it still won't run build_packages https://pastebin.com/40b7gnuj | 10:32 |
| deathmist | you sure I don't need hybris-hal built? | 10:32 |
| T4 | <DSstill01> guyz i noticed that sailfish camera app crashes when taking videos from front camera (can take videos from front one). I tried to take videos from android (aliendalvik) and it works, same for advanced camera, which can take videos from main camera, but it doesnt save them. So, it seems a problem of sailfishos camera. Any known way to fi | 10:50 |
| T4 | x that or im the only one with this problem? or i dunno | 10:50 |
| T4 | <DSstill01> [Edit] guyz i noticed that sailfish camera app crashes when taking videos from main camera (can take videos from front one). I tried to take videos from android (aliendalvik) and it works, same for advanced camera, which can take videos from main camera, but it doesnt save them. So, it seems a problem of sailfishos camera. Any known way | 10:50 |
| T4 | to fix that or im the only one with this problem? or i dunno | 10:50 |
| mal | r0kk3rz: not sure | 10:52 |
| mal | deathmist: do you have any fstab in out/? | 10:53 |
| r0kk3rz | mal: hmm ok, several people have had it, and they all fixed it, but i dont know what the fix was | 10:53 |
| vknecht | @DSstill01: sure advancedcam doesn't save the vids ? check files in ~nemo/Videos/AdvancedCamera/ or jolla-gallery ; for now it simply doesn't show vids in its integrated gallery | 10:53 |
| mal | deathmist: I assumed you had hybris-hal build already, no need to rebuild it after this change | 10:54 |
| T4 | <DSstill01> ah ok, let me check | 10:54 |
| vknecht | otherwise I'd say doublecheck your hw cam config file | 10:54 |
| r0kk3rz | or, crazy idea, get the logs and see what the error is | 10:54 |
| mal | r0kk3rz: I'm also not sure if they used the ugly system/core hack or some other way | 10:54 |
| deathmist | I mentioned I ran 'make clean', so guess I | 10:55 |
| deathmist | am building that now then | 10:55 |
| T4 | <DSstill01> i recently made some changes to camerawrapper. it's weird that it doesnt take videos from main camera only on 1 app | 10:55 |
| mal | deathmist: so the finding of ueventd was useless if you didn't even have anything built | 10:55 |
| vknecht | @DSstill01: you're asking for trouble with your fiddlings all around ;-) | 10:56 |
| mal | deathmist: why would you even run make clean? | 10:56 |
| T4 | <DSstill01> :/ | 10:57 |
| deathmist | idk... I'm going to let it run since it takes around an hour (doesn't for some reason work with ccache properly) | 10:57 |
| T4 | <DSstill01> i confirm that advanced camera can take and save videos. so, why sailfishcamera crashes? | 10:57 |
| vknecht | double check etc/dconf/db/vendor.d/jolla-camera-hw.txt | 10:58 |
| T4 | <DSstill01> is that still needed? | 10:59 |
| vknecht | and maybe undo your camerawrapper changes (whatever this wrapper is) | 10:59 |
| T4 | <DSstill01> camerawrapper changes are fine | 10:59 |
| vknecht | it's the gstdroidcam files which aren't needed anymore, if I'm not mistaken | 10:59 |
| mal | @DSstill01 of course jolla-camera-hw.txt is needed, the older gst-droid camera config is not | 11:00 |
| T4 | <DSstill01> ive removed iso and zsl stuff to fix missing to fix an error about missing symbol | 11:01 |
| T4 | <DSstill01> @mal [@DSstill01 of course jolla-camera-hw.txt is ne …], aww ok..ill check it,then | 11:01 |
| mal | what error? | 11:07 |
| T4 | <DSstill01> _ZN7android16CameraParameters12KEY_ISO_MODEE missing symbol | 11:12 |
| T4 | <DSstill01> it's samsung stuff. probably it didnt took its header, but i dont care | 11:13 |
| vknecht | does the Xperia X has the same problem than XA2 with downgrading from Pie to Oreo ? | 11:17 |
| T4 | <DSstill01> i modified jolla-camera config but it still crashes. i removed the iso modes, other ideas :/? | 11:21 |
| mal | are resolutions correct? did you try gdb or GST_DEBUG | 11:22 |
| mal | vknecht: I haven't heard of any but cannot be sure, about the xa2 issue https://github.com/sonyxperiadev/kernel/pull/1996 | 11:23 |
| T4 | <DSstill01> gst_debug=9 or less? | 11:25 |
| T4 | <DSstill01> i meant 4 | 11:26 |
| *** ahjolinna_ is now known as ahjolinna | 11:28 | |
| T4 | <DSstill01> https://bpaste.net/show/b389c7d4f719 this is the full one | 11:29 |
| vknecht | thank you ; seems X doesn't use synaptic dsx ; glad it's solved or about to for XA2 :-) | 11:30 |
| T4 | <elros34> I lose telnet connection several times and have to run ip address/route add * few times until it is stable. Any ideas how to fix it? | 11:37 |
| *** OhYash1 is now known as ohyash | 11:51 | |
| vknecht | maybe host problem with usb3 or ModemManager service ? | 12:01 |
| T4 | <elros34> I use usb 2.I have only this issue with one device with kernel 5 | 12:03 |
| Mister_Magister | mal: are there any options to optimize gui smotthness? i get stuttering a lot | 12:26 |
| mal | Mister_Magister: do you have this line https://github.com/mer-hybris/droid-config-sony-nile/blob/master/sparse/var/lib/environment/compositor/droid-hal-device-sony-nile.conf#L3 | 12:38 |
| Mister_Magister | nope | 12:53 |
| Mister_Magister | i have QPA_HWC_BUFFER_COUNT=3 | 12:54 |
| mal | Mister_Magister: so try adding that | 12:56 |
| Mister_Magister | yeah yeah i'm doing it | 12:58 |
| deathmist | mal: okay finally got back to a usable working tree; I have 2 ueventd.rc files to look at in out/ | 13:12 |
| deathmist | also 3 fstab files | 13:23 |
| mal | ok, then the issue is something else, maybe you should try the system/core hack | 13:28 |
| Mister_Magister | can sfos ramdisk mount lvm partition? | 13:43 |
| deathmist | mal: ok how would one proceed with that? | 13:46 |
| mal | not sure if it's in faq, I'm not at my computer so can't help much | 13:51 |
| masya_ | I get "tar: short read" while unpack sfos tar.bz2 to device via adb. What could be the problem? | 15:42 |
| masya_ | Xiaomi Mi6X (wayne), LOS 15.1 | 15:43 |
| Nebrassy[m] | masya_: which twrp | 15:44 |
| masya_ | 3.3.1.0 | 15:45 |
| Nebrassy[m] | official? | 15:45 |
| masya_ | https://dl.twrp.me/wayne/ | 15:46 |
| masya_ | yes | 15:46 |
| T4 | <Harsh18262> Did u try to flash after copying it to the device | 15:46 |
| Nebrassy[m] | try using this one https://mega.nz/#!jNJjGKJI!wXDvrNMQydnaUpeXcTnTWUOznKrC5BDG9BqoMCCzh9Q | 15:47 |
| *** OhYash1 is now known as ohyash | 17:15 | |
| deathmist | could master branch of libhybris be the culprit of my "droid-hal-init: Failed to initialize property area" issue and /vendor not being mounted due to some authorization issues? | 17:32 |
| deathmist | they do have a "android8-initial-update" branch which I could possibly try as well | 17:33 |
| mal | depends on what android base you are using | 17:34 |
| mal | deathmist: master branch is the one that should be usually used, except hybris-12.1 might be broken with that | 17:34 |
| deathmist | android base is los-15.1 (8.1) | 17:35 |
| mal | deathmist: android9-initial-update branch is very much obsolote and should actually be removed | 17:35 |
| deathmist | oh ok, glad I didn't try it then | 17:35 |
| mal | deathmist: as a very ugly hack for now try to comment out this line https://github.com/mer-hybris/android_system_core/blob/hybris-15.1/init/property_service.cpp#L75 | 17:42 |
| mal | then make hybris-hal and build_packages.sh -d and new image | 17:42 |
| mal | the real issue is something else, not sure what | 17:43 |
| deathmist | ok will do, I saw that somewhere else too :p | 17:43 |
| mal | pretty it's because some files are missing from device / | 17:44 |
| mal | *pretty sure | 17:44 |
| mal | what do you see in device / | 17:44 |
| deathmist | there are 'common lineage oneplus oppo qcom' directories, also do you mean straggler files & detritus? I've not added anything to staggler since it didn't mention anything about unpackaged files | 17:46 |
| mal | well the issue can be something already in make hybris-hal | 17:58 |
| mal | it might not be building something we need | 17:58 |
| mal | there has been issues in 15.1 with that | 17:58 |
| mal | we minimize build too much | 17:58 |
| vknecht | how come XA2 can use latest dcd submodule ? because it use more recent than 3.0.3.9 ? or some build infra specific ? | 18:26 |
| mal | vknecht: it requires some changes in other packages which will probably come with next sfos release | 18:33 |
| * vknecht hopes it come soon ;-) | 18:34 | |
| mal | is there something you want from those changes? | 18:35 |
| vknecht | not sure yet, I'll do some local tests for now | 18:38 |
| mal | are you using the upgrade-3.0.3 branch as suggested in faq | 18:39 |
| vknecht | yep, but then had other problem ; have to check again, or perhaps just change the ssu package reference locally | 18:42 |
| vknecht | also maybe I made a mistake | 18:43 |
| Mister_Magister | mal: is there hybris-boot repo for sfosx? | 18:43 |
| mal | Mister_Magister: ? | 18:52 |
| Mister_Magister | xperias have everything in lvm and i would like to try that | 18:53 |
| mal | Mister_Magister: it might be possible to use the sfos build for xperia x instructions and adapt the repos | 18:58 |
| Mister_Magister | hmm | 18:58 |
| Mister_Magister | xperias have totally different ramdisk so i assumed they have separate hybris-boot | 18:59 |
| mal | of course | 19:01 |
| mal | I mean different ramdisk but not different hybris-boot | 19:02 |
| mal | as an example https://github.com/mer-hybris/droid-hal-img-boot-f5121' | 19:02 |
| mal | that builds the boot image | 19:02 |
| mal | the final one | 19:02 |
| mal | you can see in some adaptation repo that it has define about custom boot image | 19:03 |
| Mister_Magister | nice thanks! | 19:17 |
| deathmist | mal: after commenting out the line droid-hal-init.service is running! though /vendor still cannot be mounted (it's the same for /persist & /cache) journalctl https://pastebin.com/RG10cz4L & systemctl https://pastebin.com/qBrpLDAB where sdf6 is /vendor | 19:44 |
| mal | deathmist: have you checked which partition it tried to mount in the systemd mount services in /lib/systemd/system/ | 20:18 |
| mal | or log | 20:18 |
| deathmist | for sure it's the right partition; it uses the mount-sd@sdf6.service mounting and fails with "Error mounting /dev/sdf6", but above also says "Error opening current controlling terminal for the process (`/dev/tty'): No such device or address" | 21:12 |
| deathmist | I read up on this a bit and found this: "udisks2 requires an authentication agent (and the fallback agent fails because that requires a tty)." | 21:13 |
| deathmist | if I'm getting this correctly /usr/bin/pkttyagent is not running and thus tty fallback agent fails | 21:14 |
| T4 | <elros34> udisks2 shouldn't even touch your internal partitions, they are handled by systemd mount units (/lib/systemd/system/*.mount) created by dhd | 21:17 |
| mal | deathmist: the services I meant are the ones with filetype .mount | 21:23 |
| mal | you are not interested in sdcard mounting yet | 21:24 |
| vknecht | this is not the droid you're looking for... | 21:39 |
| deathmist | I'll get back to this tomorrow, thanks for the help! | 22:19 |
Generated by irclog2html.py 2.17.1 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!