| *** poetaster is now known as Guest903 | 09:18 | |
| T42 | <adampigg> @elros34 mal, ill certainly rebuild kernel without latest patch and test it | 09:23 |
|---|---|---|
| T42 | <edp_17> What problem that change is supposed to solve? | 09:44 |
| T42 | <adampigg> @edp_17 its quite specific, related to bluetooth being reloaded on the pinephoblne causing a flood of dbus messages | 10:03 |
| T42 | <elros34> @edp_17 run 'dbus-monitor --system --profile "path=/net/connman/technology/bluetooth,interface=net.connman.Technology"' and count GetProperties on every bluetooth on/off. If it's more than 10 for every cycle and increases then you are affected otherwise it's probably irrelevant for every device | 10:34 |
| T42 | <Ayman_rgab> me 2 only | 10:39 |
| T42 | <elros34> I guess not Get but SetProperty? | 10:40 |
| T42 | <adampigg> @elros34 is it triggerable by just restarting bluez? | 10:42 |
| T42 | <Ayman_rgab> `mc 1670150339.398386 565 :1.57 net.connman /net/connman/technology/bluetooth net.connman.Technology SetProperty` | 10:42 |
| T42 | <Ayman_rgab> `sig 1670150339.530307 311 :1.7 <none> /net/connman/technology/bluetooth net.connman.Technology PropertyChanged` | 10:42 |
| T42 | <elros34> @Ayman_rgab that is not relevant, we are talking about GetProperty only | 10:42 |
| T42 | <elros34> @adampigg for me only by writting 0/1 to /sys/module/hci_smd/parameters/hcismd_set | 10:44 |
| T42 | <adampigg> so, that takes down the device full on your android port? | 10:44 |
| T42 | <adampigg> s/fully | 10:44 |
| T42 | <elros34> at least removes it from rfkill list and that is enough to trigger technologyremoved/added | 10:46 |
| T42 | <elros34> II do not think it disables device completely | 10:47 |
| T42 | <adampigg> well, sounds like its close enough to what the PP does to trigger the issue anyway | 10:48 |
| T42 | <elros34> yes | 10:48 |
| T42 | <elros34> I wonder if that is what caused reboot my device if I try to run hci-smd initialization more than once in past but I have never bother to analyze it previously. For sure I had few reboots when trying to debug this issue but could be also related to gdb | 10:51 |
| T42 | <adampigg> my pinephone has an uptime of 11 hours now and still wakes instantly, so atleast the kernel fix works | 10:53 |
| T42 | <edp_17> @adampigg , @elros34 : Thanks for explaining! I'll try that command on my devices. | 11:37 |
| voidanix[m] | why is it that flashing a new kernel ala sfos X does not actually flash it? | 13:26 |
| voidanix[m] | flashing hybris-boot in flashboot+installing droid-hal*-img-boot with zypper works 1 out of 10 times, maybe | 13:27 |
| voidanix[m] | *fastboot, oops | 13:28 |
| T42 | <b100dian> I'm planning to re-enable vendor.audio-hal-2.0 and audioservice for my port so that the Virtual (Ultrasound) proximity sensing works again. | 21:26 |
| T42 | <b100dian> I notice that fakeaudiopolicy and droidmedia only initi parts of the things the audioserver does | 21:27 |
| T42 | <b100dian> SoundTriggerHwService::instantiate(); // skipped. Maybe uninportant, sounds like Bixby/hey google thingie | 21:27 |
| T42 | <b100dian> AudioFlinger::instantiate(); // This however may be important | 21:28 |
| T42 | <b100dian> The thing is (I've yet to see what config files vendor.audio-hal-2-0 uses) I want to start them both, and it may be that AudioFlinger is harming audio in calls when they are both enabled (and proximity works, yay..erm) | 21:29 |
| T42 | <b100dian> If I only start vendor.audio-hal-2-0 without audioserver I get a kernel oops in 60 seconds, > adsp subsystem failure reason: err_qdi.c:450:EF:sensor_process:1:dog_virtual_user.c:240:USER-PD DOG detects stalled initialization, triage with IMAGE OWNER. | 21:33 |
| piggz_ | mal: spiiroin: anyway for mce/dsme to monitor a gpio and inhibit the late-suspen policy? i think whats happeiing now is, modem wakes cpu using gpio on ring, no wakelocks so cpu autosleeps after 1 second == no reaction from ofono/userspace from incoming call from susepnd state | 21:52 |
| piggz_ | funny, i just checked for events, and we get KEY_WAKEUP ... grepped the logs, and we discussed this in 2021 :D | 22:03 |
| piggz_ | seems ive already coded a solution and forgot about it :D https://github.com/sailfish-on-dontbeevil/mce-1/commit/3ed9b5c5422d4ca6dec9ac2dfb055f8f97271c1a | 22:50 |
| deathmist | mal: @slavamon: if you don't mind me asking how was https://github.com/mer-hybris/libgbinder/pull/109 discovered as needed? interestingly it seems that encode_ptr_cookie() was broken for nearly 2 years previously :p | 23:04 |
| mal | deathmist: not sure | 23:05 |
Generated by irclog2html.py 2.17.1 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!