Tuesday, 2022-06-28

UmeaboyHi! The errornous commit that broke syncing of the external repo is still valid.02:30
UmeaboyAny workaround?02:30
UmeaboyI just use git branch hybris-18.1-update and resync?02:30
UmeaboyOr is there a more mature solution?02:41
UmeaboyNope. I'm still failing to run breakfast $DEVICE due to not finding common.mk and here's the content of $DEVICE.xml:03:44
UmeaboyI did delete the roomservice.xml, but I still end up with the same result.03:44
UmeaboyI have applied the patches as well.03:46
T42<Umeaman> This manifest breakfast is driving me nuts! No matter how I edit it and remove roomservice.xml it still says duplicate path.04:20
T42<Umeaman> Can the part about the manifest not be changed so it's OK to use the roomservice.xml with the extra repos? This would save a lot of hassle.04:22
T42<Umeaman> At least for me.04:22
T42<edp_17> Thaodan : Regarding restart of services and raceconditions. Yes, restarting them after lipstick is a hack but I couldn't workout how to solve it properly.06:34
T42<elros34> didn't you figure out that at least for ngfd it was about wrong permission or I mixup something? So either create udev rule to change that or modify init*rc file. Which one depends on when chown/chmod from init*rc is executed: after or before ngfd starts.06:53
T42<elros34> this is what I use for example: https://github.com/elros34/droid-config-moto_msm8960_jbbl/blob/master/sparse/lib/udev/rules.d/99-volume-media-keys.rules or06:56
T42<Spidey24Z> hi09:23
T42<Spidey24Z> https://pastebin.ubuntu.com/p/g5pXDcwnDV/09:24
T42<Spidey24Z> can i use strip -N to remove this symbol ?09:24
T42<edp_17> @Spidey24Z : This is Waydroid, right? :) I've got this too. Tried to patchelf it but didn't work.10:26
T42<edp_17> @elros34 : You said "according to your logs ngfd starts about 4s after droid-hal-init finished". So permission of /sys/class/timed_output/vibrator/enable should be okay. How about if I amend ngfd service and introducing a pre-start script that sets the permission?10:35
T42<edp_17> Hm, that wouldn't work as ngfd is a user service, so it doesn't have permission to set permission of sys/class/timed_output/vibrator/enable, would it?10:38
T42<nephros> Isn't that udev's job?12:24
mal@edp_17 if the permission of /sys/class/timed_output/vibrator/enable is wrong maybe check that you have the patch in system/core repo in android side12:37
malif that android side has incorrect group then it might change the permissions after udev12:38
T42<b100dian> Thaodan: maybe you want to chime in here? https://forum.sailfishos.org/t/sfos-support-for-new-xperia-devices/5908/6413:37
T42<edp_17> mal : what patch are you referring to in system/core?14:05
T42<edp_17> The permission of sys/class/timed_output/vibrator/enable is correct because once lipstick started and I restart ngfd, it just works. Or is it possible that permission is wrong before that?14:05
T42<edp_17> Thanodan : What does this do? https://github.com/Thaodan/droid-config-sony-edo/blob/master/sparse/etc/usb-moded/usb-moded-configfs-lena.ini14:07
T42<edp_17> @elros34 , @nephros : I am not sure what rule I should add for this into udev. Can you help me out please?14:09
T42<Haibierron> Forwarded from Successful Trader: When you invest with a great company all you are going to get is a positive result and a successful withdrawal. I'm so happy right now, I invested £1,000 and got £10,500 in return in just 24hours of trade.  If you are looking for a legit platform to invest your money or you want to be successful like me too, contact Mr @Pierre_Relekom to start https://t.me/pierreinvestmentplatform14:47
T42<elros34> @edp_17 right if it's user service then it will not work but this way you can check what permissions it has at the time when ngfd starts. I do not remember excatly but wasn't action which contains chown/chmod for vibrator executed later?. For udev rule run 'udevadm info -a  /sys/class/timed_output/vibrator/'17:12
T42<nephros> udev rules generated from android rc/ini files are in /lib/udev/rules.d/99*droid*.rules17:15
T42<elros34> could be but there is no rule for timed_output in ueventd for my device17:19
T42<Umeaman> Guys, can the wrongful commit be corrected?17:29
T42<elros34> git commit --amend17:30
T42<elros34> but try to avoid it if you work with somebody else on project17:30
Thaodanb100dian: Replied18:20
T42<b100dian> Awesome!19:21
T42<Umeaman> I cannot build hybris-hal unless that commit can either be undone or fixed.19:27
T42<elros34> what commit? Even if there is anything which stops you from building you can still make local changes19:31
T42<edp_17> @Umeaman : Fix that locally and push into your own github. (This is what I'd do.)19:46
T42<Umeaman> The one regarding clang mentioned earlier. I think the commit numbers aren't correct. (re @elros34: what commit? Even if...)19:48
T42<elros34> no idea about clang related commit, quick search in channel logs also do not shows anything like that today19:51
T42<Umeaman> Not today. About a week ago I think.19:54
T42<elros34> so you expect us to remember it or search week ago msg?19:55
T42<Umeaman> No. I just thought you where aware of the failing to use repo sync --fetch-submodules due to 4 repos not working. (re @elros34: so you expect us to ...)19:57
T42<elros34> you mean for android apps repos? How is that clang related?19:58
T42<Umeaman> 2 of the repos are for Etar and the other 2 is for external pybind clang.19:59
T42<elros34> yeah so why don't you ignore it and try make hybris-hal like you were advised?19:59
T42<Umeaman> Because breakfast $DEVICE after running source build/envsetup.sh shows that the common.mk file is missing, but if I add the proper repo to fix it I end up with duplicate path. If I then delete the roomservice.xml it still gets recreated and the problem continues to haunt me.20:07
T42<Umeaman> I need to figure out how to make a working $DEVICE.xml that won't add a repo that already seems to exist in manifest.xml (I know it's symlinked) and my problem still persists.20:09
T42<elros34> what repo exists in manifest.xml? you need to be more specific. DO you have all device specific repos cloned in correct places?20:13
Thaodanumeaman: you can delist repos20:42
Thaodanread the git-repo manual20:42
T42<Umeaman> Thaodan: Wouldn't be better if we could use the roomservice.xml that breakfast seems to make and add proper repos in it to avoid duplicates?22:03
T42<Umeaman> Or is that not an option?22:03
T42<Umeaman> After all, we use AOSP/Lineage OS as base.22:03
Thaodanumeaman: I don't know the lineage commands I usually use plan aosp/caf.22:04
ThaodanBut sony delists some repos from google to use their own with Sony patches that's why I know how remove-projects from a manifest.22:04
T42<Umeaman> Okey. I suppose I could switch to use AOSP to make it simpler.22:05
T42<Umeaman> I have to solve this somehow.22:36
ThaodanYou can remove projects in your $device.xml and add new.22:39

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