Thursday, 2019-10-31

T42<Lukapanio> @elros34  sorry something is wrong there here is new logcat06:25
T42<Lukapanio> 07-25 15:33:31.746  2910  2910 W /system/bin/hwservicemanager: getTransport: Cannot find entry in either framework or device manifest.06:27
T42<Lukapanio> 07-25 15:33:31.746  3856  3856 F HwcComposer: failed to get hwcomposer service06:27
T42<elros34> Not starting sensorservice could also be important: ServiceManager: Waiting for service sensorservice...08:06
riniguswould like to fix one annoying issue - marking boot successful is failing with 'bootcontrolhal: update_slot_attribute: Failed to get disk info for xbl_a' in logcat.08:16
riniguswhen running bootctl mark-boot-successful I get 'Error marking as having booted successfully: Operation not permitted'08:16
rinigusas a debug, I was trying to get current slot and was getting 0 as a response to bootctl get-current-slot08:18
T42<Lukapanio> @elros34 [Not starting sensorservice could also be impor …], but musnt hwc work without that?08:21
T42<elros34> sorry I don't know, try to search in channel logs
rinigusor use
rinigusre slot successful: I suspect it fails at
T42<Lukapanio> so after reboot and starting surfaceflinger i have bootanimation o screen09:02
T42<Lukapanio> but tests still fails09:02
T42<Lukapanio> will strace later09:02
T42<Lukapanio> lol power button workesok09:06
T42<hacker12455> I think it worked even before Android did09:34
T42<Lukapanio> now it turns screen of/on09:59
T42<Lukapanio> off*09:59
T42<Lukapanio> @elros34  can i have boot animation without hwcomposer?10:03
T42<elros34> I don't know what do you mean. Boot animation is from android's surfaceflinger, in sfos we don't even have such a thing. Some moto devices on older android needed lipstick hack: start surfaceflinger for a few seconds and stop it, then lipstick works.10:10
T42<Lukapanio> hmmmm how do surfaceflinger dispalys it?10:12
riniguspiggz: from the logs sounds like you worked on bootctl issue (see my messages above). did you managed to resolve it for pro1?10:12
T42<elros34> @Lukapanio this is wrong:
T42<elros34> upload your changes so others can check them10:20
T42<Lukapanio> @elros34 [@Lukapanio this is wrong: …], what is bad there? i have same but lastest line is LIPSTICK_OPTIONS=-plugin evdevtouch:/dev/input/event2 -plugin evdevkeyboard:keymap=/usr/share/qt5/keymaps/droid.qma10:28
T42<elros34> yes, that env variable. Like mentioned in hadk "WRITE˓→ALL IN ONE LINE(:"11:35
T42<Lukapanio> I ServiceManager: service 'display.qservice' died12:41
T42<Lukapanio> hmmm found that12:46
T42<Lukapanio> omg, it's alive, alive!!1    14:0212:47
T42<Lukapanio> m4h0n3y    rebuilded hwcomposer.msm8960 and can run minimer at last :D    14:0212:47
T42<Lukapanio> m4h0n3y    with that patch    14:0212:47
T42<Lukapanio> m4h0n3y    and after reboot it asks me what language to use! lipstick started :D12:47
T42<hacker12455> What patch though?12:57
T42<Lukapanio> already built will test soon12:58
T42<hacker12455> Alright12:59
T42<Lukapanio> "/system/vendor/lib/hw/" is 64-bit instead of 32-bit13:45
T42<hacker12455> Maybe it's a problem with some parts of the 64 bit Oreo ROM being 32 bit?14:07
Mister_Magistereverything in lib is 32bit just so you know15:02
malunless something went wrong during build15:04
T42<Lukapanio> no ibuild that lib manualy15:05
T42<Lukapanio> @Lukapanio [omg, it's alive, alive!!1    14:0215:06
T42<Lukapanio> m4h0n3y …], here15:06
Mister_Magister@Lukapanio just so you know telegram replies in irc don't work really15:06
T42<Lukapanio> ouch sorry15:06
T42<Lukapanio> i found :15:06
T42<Lukapanio> omg, it's alive, alive!!1 14:0215:06
T42<Lukapanio> m4h0n3y rebuilded hwcomposer.msm8960 and can run minimer at last :D 14:0215:06
T42<Lukapanio> m4h0n3y with that patch 14:0215:06
T42<Lukapanio> m4h0n3y and after reboot it asks me what language to use! lipstick started :D15:06
T42<Lukapanio> @elros34  how to force start lipstick16:08
T42<Lukapanio> ?16:08
T42<elros34> usually it restarts itself on crash unless you masked user@100000.service. as root: systemctl-user restart lipstick16:14
T42<Lukapanio> so some updates17:44
T42<Lukapanio> here surfaceflinger is working17:45
T42<Lukapanio> and los bootanimation on screen17:45
T42<Lukapanio> here i stopped it and nothing on screen17:45
T42<Lukapanio> @hacker1245517:45
T42<DylanVanAssche> mal: Do you have any suggestion for debugging MCE volume keys?18:58
T42<DylanVanAssche> They are reported as /dev/event/input1 and work if you cat them.18:58
T42<DylanVanAssche> However, MCE doesn't seem to detect the volume keys pressing.18:58
deathmisthow bad would the overhead from using an overlayfs on both /system and /vendor be since I can't bind mount a few files before DHI?19:11
vknecht@DylanVanAssche: do they send KEY_VOLUMEDOWN / KEY_VOLUMEUP ?19:12
T42<DylanVanAssche> Yes19:12
T42<DylanVanAssche> vknecht:
vknechtboth on the same /dev/input/event1 ? not that I know much, but here I have that for jack, otherwise each are received on a distinct eventX19:15
T42<DylanVanAssche> Yes, they are.19:15
vknechtand "evdev_trace -t | grep VOLUME" shows them ?19:23
vknecht(need mce-tools package installed)19:24
mal@DylanVanAssche is audio working well?19:25
T42<DylanVanAssche> vknecht: Sure :)
T42<DylanVanAssche> mal: Yes, I have sound and can control it using alsamixer19:25
vknechtah, here's the man :-)19:26
malwhich device?19:26
T42<DylanVanAssche> mal: Mainline Linux on the PinePhone/PineTab :D19:26
maldeathmist: wondering how the mount service I gave is not working, I tested that myself and it mounted the file19:26
T42<DylanVanAssche> vknecht: hahha19:26
* vknecht goes back to his listview problems19:27
T42<DylanVanAssche> MCE notices the volume key press events (tklock notification: mce_volume_key), but that's it19:27
deathmistmal: omg I think I've been a massive idiot before, just tried it one more time... and it actually works!19:55
deathmistand just checked in TWRP too, no "hybris" in the file but SFOS boots still :)19:56
maldeathmist: yep :)20:01
maldeathmist: such mistakes happen, good that it work, much easier with the mount service20:02
wiktorek140its normal that *.mount for device are not generated?20:44
malwiktorek140: depends on android base, sometimes 15.1 or 16.0 hybris-hal builds don't copy ueventd*.rc and fstab* to out/ which causes that, first find those in out/ and then check output of "make modules" for those and if some of those exist in the output try building those targets with make and then check out/ again20:51
wiktorek140mal: thanks, i build on 15.1 base so i will check files20:54
wiktorek140may it be because fstab is copied to vendor (in system)?20:56
malthe build minimizing is not working correctly in all situation and sometimes those are not copied21:05
wiktorek140mal: i need to build it somehow manually or there are some trick that can be used?21:08
malwiktorek140: manually, check the output of make modules like I said21:35
wiktorek140mal: in theory it is in modules21:43
malwiktorek140: do you see anything about eventd or fstab in output of make modules22:02
wiktorek140mal: eventd not, fstab it is there22:04
maltry make with the name of the fstab target22:04
malwiktorek140: did you find any ueventd*.rc files in out/22:05
wiktorek140mal: yea, ueventd*.rc file are in out/22:06
wiktorek140doing it right now22:06
malwiktorek140: if you can check also on device in /system and /vendor what ueventd*.rc files those have22:07
wiktorek140mal: ueventd.rc in / and its only one22:09
wiktorek140building image with new hybris-hal22:13
malwiktorek140: did the fstab appear in out/22:16
wiktorek140mal: yea, now it is in system/vendor/etc/22:17
malhopefully that helps22:18
wiktorek140mal: mount files created. but hang on debug boot22:27
wiktorek140droid-hal-init: Failed to initialize property area22:28
maldo you have *contexts files from out/..../root in straggler files22:28
wiktorek140mal: no, nothing tell that files are required22:30
malwiktorek140: there is a bug that currently causes the unpackaged files to not show in droid-hal build22:32
wiktorek140mal: add all *context manualy?22:35
malwiktorek140: yes, those that are in that root folder22:35
malcheck also what else you have there, show ls -l of that folder22:35
T42<elros34> apply this change to /usr/lib/rpm/check-files in your platform target so it will show unpackaged files22:36
mal@elros34 I was trying to find that but couldn't yet, maybe make a pull request to dhd22:37
T42<elros34> to dhd? sorry I don't get it's rpm issue22:38
mal@elros34 oh, the fix was there, I thought it was some issue in droid-hal-device repo22:58
malwiktorek140: so add all *contexts files and also /vendor22:59
T42<elros34> it's broken even in sailfish sdk23:00
wiktorek140mal: vendor i have23:01

Generated by 2.17.1 by Marius Gedminas - find it at!