Wednesday, 2022-02-23

T42<HengYeDev> run at same time or stop surfaceflinger first? stop surfacefling than run minimer has same.result (re @elros34: you could see if sta...)01:00
T42<LSolrac> Hello. Out of curiousity, does anyone know why I get this message? I've a device repo on the manifest and it's cloned locally as well : https://irc.thaodan.de/.imgstore/7e9c442c/file_3329.jpg05:32
T42<edp_17> @elros34 : This is what I have in that file: https://paste.ubuntu.com/p/8bwgBRtDNw/ (re @elros34: check whether your h...)06:44
T42<edp_17> The treltexx (Note4) with sfos4.3 shows very similar behaviour. After flashing the image, intial setup goes smoothly, device works. Until I start the camera that crashes the device to reboot (or if I reboot it manually) then the device reboots, shows the lock screen then reboots. Then, it keeps doing this.06:47
T42<edp_17> Can this be a libhybris problem?06:47
T42<edp_17> On the treltexx I spotted something else. If the device is on charger, it doesn't crash-reboots, even if I start the camera! But keeps rebooting if I reboot. :(07:06
T42<edp_17> Which libhybris should be used for building 4.3? Latest master or upgrade-4.3.0 branch?08:07
T42<edp_17> Do I need this fix form building 4.3? https://github.com/mlehtima/droid-config-fp2-sibon/commit/1f6720efc02219ab7eabc185c8d07b44ebcfd42113:09
T42<edp_17> *for13:09
T42<elros34> yes13:45
T42<edp_17> Thanks. What can be the result if I use the incorrect one (with or without -hal)? Will the build fail or will cause issue on device?13:54
T42<elros34> you may end up with binder version, just different backend.13:57
T42<edp_17> What does that mean?13:57
T42<elros34> it just use different way/code to communicate with low level components /dev/hwbinder or something like that13:59
T42<HengYeDev> @elros34 Since the problem is caused by timeout in a vsync function, and surfaceflinger says "Faking VSYNC due to driver stall",14:00
T42<HengYeDev> any idea of what's going wrong? should i try the old surfaceflinger plugin?14:00
T42<elros34> @edp_17 you can see many binder plugins in mer-hybris repo for android >= 814:01
T42<edp_17> @elros34 : Can an incorrect binder/backend version cause the issue I have on the treltexx and n7000? (initial setup goes fine, then after the first reboot it keeps rebooting)14:03
T42<HengYeDev> if i run surfaceflinger and minimer at the same time I get  Failed to open vsync attribute at /sys/devices/platform/148c0000.decon_t/vsync14:04
T42<HengYeDev> That file doesn't exist14:04
ThaodanDon't run it?14:08
ThaodanWe don't run surfaceflinger14:08
T42<HengYeDev> i'm trying to debug why lipstick and minimer cause the device to reboot14:09
T42<HengYeDev> the error message makes it seem like its trying to access that file which doesn't exist, which may be leading to the vsync timeout and consequently kernel oops at https://github.com/sailfish-on-gta4xlwifi/android_kernel_samsung_gta4xl/blob/hybris-17.1/drivers/video/fbdev/exynos/dpu20/decon_core.c#L230714:11
T42<HengYeDev> the only decon directory is 148b0000.decon_f14:11
T42<elros34> @edp_17 unlikely, these packages are not critical. You need to figure out reboot reason from logs14:11
T42<edp_17> @elros34 : Okay. The problem is that on the treltexx, the reboot happens so quickly that I don't have time to check any logs. If I am lucky and can get in before the reboot triggered, by the time it starts outputting the logs on screen, it reboots. Plus, I am not sure where to look, journal, logcat, dmesg?14:15
T42<elros34> @HengYeDev sorry I do not have any good ideas except mentioned bootanim service. If that shows something then old qpa plugin might work. Exynos hwservice seems to be closed source so not sure what is going on. Some clever jolla devs might give you some better ideas14:15
T42<edp_17> What I am trying to do now is masking lipstick and/or user@100000 to see whether that stops the reboot. But first I need to restore the old system then upgrade it again. Time consuming. :(14:17
T42<elros34> @edp_17 persistent journal, then /sys/fs/pstore or dmesg and if possible logcat. Logs shouldne be hard to get, just stop booting at some point and run needed commands and wait14:17
T42<elros34> @HengYeDev you may compare your current output with lineage-17 output and you will see which errors are really critical. You can compare fbdev/exynos/dpu20 commits  with lineage-18 version for some fixes14:20
T42<edp_17> @elros34 : How to stop booting? Creating the /init_enter_debug2 file?14:29
T42<elros34> or mask user@100000, depends on what cause the issue14:42
T42<edp_17> I'll try both to see which one works. 👍14:53
T42<edp_17> @elros34 : Now I got a different problem. Restore 4.1, then removed patchmanager (with all patches) then upgraded to 4.2 in command line. After the reboot, the device stuck on the yamui splash screen as lipstick's failed to start: https://paste.ubuntu.com/p/wJks2GHgVK/15:06
T42<edp_17> Yesterday I did this upgrade the same way and it worked (at that time I didn't remove patchmanager).15:07
T42<elros34> why didn't you flash sfos, it would be faster?  undefined symbol: _ZN18LipstickCompositor16openUrlRequestedERK4QUrlRKN13ContentAction6ActionERK5QListIS4_E suggest some wrong packages versions15:09
T42<elros34> did you disable user repos while upgrading?15:10
T42<edp_17> Yes, chum and mousetracker15:10
T42<edp_17> And all openrepos one15:11
T42<edp_17> Here's why I tried this rather than flashing the image. After I OTAd 4.2 to 4.3, I got the bootloop problem. I though, should remove patchmanager from 4.2 then try the OTA. Therefore, I restored 4.2 but couldn't remove patchmanager. So I restored 4.1 to try from earlier (remove patchmanager, OTA to 4.2, then OTA to 4.3)15:16
T42<edp_17> Then I got this.15:16
T42<elros34> Remove patchmanager now15:16
T42<edp_17> I've done that before OTAd to 4.2.15:16
T42<elros34> it's still there: "Failed to start Patchmanager service"15:18
T42<edp_17> I've removed it with pkcon!15:19
T42<edp_17> zypper says No matching items found.15:19
T42<edp_17> sudo pkcon remove patchmanager: Package not found: patchmanager15:20
T42<elros34> IMHO pkcon is not something you should even touch. zypper se -s patchmanager15:21
T42<edp_17> I got this: https://paste.ubuntu.com/p/X95VhfpznH/15:22
T42<edp_17> How do you mean not touch pkcon? I always install a local rpm with that.15:22
T42<elros34> you choice. I try to avoid software which upgrades all packages on every action. LOoks like your upgrade wasn't finished, try again.15:23
T42<edp_17> Yeah, this is what I am thinking of. Going back to 4.1 and start again. Thanks.15:25
T42<elros34> no, just do this now and remove these leftovers from patchmanager if package is really removed15:27
T42<edp_17> Ops, I thought the "try again" meant to start over. I've already wiped the system and the restore (4.1) is running. It'll take a wile until I get to the same position where I am now. Will come back with the result later.15:30
T42<edp_17> On OBS (n7000) I have two packages uresolvable: qt5-qpa-hwcomposer-plugin and qtscenegraph-adaptation-droid. I don't know how to fix them. Any idea? https://paste.ubuntu.com/p/j5bNMhdxvJ/15:58
T42<elros34> missing "Project Config"16:47
T42<edp_17> @elros34 : Thnak you.16:55
Thaodan@edp_17: Enable persistant journal19:21
Thaodanhttps://github.com/mer-hybris/hadk-faq#persistent-journalctl19:37
T42<edp_17> Thanks for your support guys! The mistery has solved on my treltexx. The previous restores/upgrades have left a mess in the internal storage and caused those zypper issues. Factory resets weren't enough but when I reformatted everything except the external sd card, then I was able to remove patchmanager and OTA upgrade to 4.2. Now I have a working 4.2 version with waydroid and patchmanager 3. Tomorrow, I'll try to OTA upgrade to 4.3 and solve21:27
T42<HengYeDev> does hwservicemanager: getTransport: Cannot find entry android.hardware.graphics.mapper@3.0::IMapper/default in either framework or device manifest. seem like a critical error? should i look in lineage logs to see if it is?21:39

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