| *** phlixi is now known as Guest9300 | 01:41 | |
| *** phlixi_ is now known as phlixi | 01:41 | |
| T42_ | <edp_17> Hi all. How can I read light sensor value on XA2? | 12:21 |
|---|---|---|
| mal | for what use? | 12:21 |
| T42_ | <edp_17> mal: experiment. :) | 12:46 |
| T42_ | <edp_17> mal: I'd like to trigger an event (execute a script) at certain levels (below or above). | 12:48 |
| mal | qt is easiest, dbus of course if also possible | 12:56 |
| mal | *is also possible | 12:56 |
| T42_ | <Lukapanio> Hi mal, agin my vk ot. I got vk to somewhat wotk, but VK_KHR_surface seems to be a big issue for linux apps | 13:03 |
| T42_ | <Lukapanio> is there any wrapper to use VK_KHR_android_surface? | 13:04 |
| mal | how did you get it working? | 13:06 |
| mal | @Lukapanio libhybris wraps those surfaces so to outside is looks like it uses wayland surface but it's really using that android surface within libhybris | 13:08 |
| T42_ | <Lukapanio> just did clean setup without mesa junk (re @SailfishFreenodeIRCBridgeBot: <mal>how did you get...) | 13:09 |
| T42_ | <Lukapanio> well vkcube dont agree with that (re @SailfishFreenodeIRCBridgeBot: <mal>@Lukapanio libh...) | 13:09 |
| T42_ | <Lukapanio> vkEnumerateInstanceExtensionProperties failed to find the VK_KHR_xcb_surface extension. | 13:10 |
| T42_ | <Lukapanio> | 13:10 |
| T42_ | <Lukapanio> Do you have a compatible Vulkan installable client driver (ICD) installed? | 13:10 |
| T42_ | <Lukapanio> Please look at the Getting Started guide for additional information. | 13:10 |
| T42_ | <Lukapanio> vulkaninfo is fine tho | 13:10 |
| T42_ | <Lukapanio> ahh ok vkube wayland is better | 13:13 |
| T42_ | <Lukapanio> vkcube-wayland | 13:13 |
| T42_ | <Lukapanio> library "libadreno_app_profiles.so" not found | 13:13 |
| T42_ | <Lukapanio> Selected GPU 0: Adreno (TM) 650, type: IntegratedGpu | 13:13 |
| T42_ | <Lukapanio> vkcube-wayland: ./cube/cube.c:1369: demo_prepare_buffers: Assertion `!err' failed. | 13:13 |
| T42_ | <Lukapanio> Aborted | 13:13 |
| mal | are you using wayland or xcb, libhybris only supports waylan | 13:15 |
| T42_ | <Lukapanio> yes sorry used wrong vkcbe initially (re @SailfishFreenodeIRCBridgeBot: <mal>are you using w...) | 13:15 |
| T42_ | <Lukapanio> as you can see already tried wayland one | 13:15 |
| mal | yeah | 13:16 |
| mal | but is wayland the default on your platform | 13:16 |
| T42_ | <Lukapanio> yes i am running wayland kde | 13:16 |
| mal | how does it fail now? any more information than that error | 13:18 |
| T42_ | <Lukapanio> seems like eperm on some ioctl | 13:21 |
| T42_ | <Lukapanio> 06-12 02:20:29.917 24376 6600 W Adreno-GSL: <gsl_ldd_control:553>: ioctl fd 6 code 0xc0200948 (IOCTL_KGSL_GPUOBJ_IMPORT) failed: errno 1 Operation not permitted | 13:21 |
| T42_ | <Lukapanio> 06-12 02:20:29.917 24376 6600 I AdrenoVK-0: MapGrallocBuffer: Map memory failed! | 13:21 |
| T42_ | <Lukapanio> witch is odd considering that i did run vkcube as root (re @SailfishFreenodeIRCBridgeBot: <mal>how does it fai...) | 13:21 |
| T42_ | <Walid> @edp_17 sorry for ping, dose audio call work in hammaerhead (nexus 5), bullhead still blocked, audio work when change rington, but no audio in call, also camera can't understand how start fix it | 14:38 |
| T42_ | <elros34> have you tried debug anything about it? You can see edp changes in his repo https://github.com/edp17/droid-config-hammerhead | 14:45 |
| T42_ | <Walid> is pulseaudio-modules-droid-hidl requied only for newer device ? its included in pattern inc file | 14:47 |
| T42_ | <Walid> Requires: pulseaudio-modules-droid-hidl | 14:47 |
| T42_ | <elros34> this might give some idea based on your android version https://github.com/droidian/audiosystem-passthrough | 15:54 |
| T42_ | <edp_17> Walid : Hi, yes, audio call works in Hammerhead (Nexus 5). What android base are you using? | 16:01 |
| T42_ | <Walid> android 9.0 pie (re @edp_17: Walid : Hi, yes, aud...) | 16:09 |
| T42_ | <elros34> so according to linked github you need pulseaudio-modules-droid-hidl and audiosystem-passthrough (maybe -dummy-af too I am not sure) | 16:28 |
| T42_ | <elros34> obviously use mer-hybris one, I copied wrong url by accident | 16:35 |
| T42_ | <edp_17> mal: can you give me exaple for reading the light sensor value with qt and/or dbus? (I'd like to use it in command line.) | 17:50 |
| mal | @edp_17 qml or c++? | 18:20 |
| T42_ | <edp_17> mal: I guess the c++ can be used better in command line than qml. | 18:34 |
| T42_ | <edp_17> mal: can you please give me both? :) | 18:34 |
| mal | https://github.com/Bytewerk/Messwerk/blob/master/src/light.cpp | 18:46 |
| mal | qml is probably trivial using qt documentation | 18:46 |
| T42_ | <edp_17> mal: Thanks. Isn't that possible with a shell script? | 19:23 |
| mal | well the dbus interface is quite annoying to use | 19:37 |
| Tofe | I'm currently looking at hammerhead and sensorfw using iioadapter for ALS, and I think there's something incorrect or missing in the way the device is detected | 19:45 |
| Tofe | basically for hammerhead we have this attibutes: https://paste.tofe.me/p/65c854.txt but sensorfw will look for "illuminance.*raw$", and there's no such thing. However, there is an "intensity.*raw$", so I wonder if updating the driver is needed | 19:47 |
| Tofe | I mean, update sensorfw's driver. But maybe the issue is in the kernel's driver, though I'm doubtful | 19:48 |
| Tofe | ok, after reading some documentation in the kernel, it looks like I mixed things up. intensity isn't in lux, contrary to illuminance. So it cannot be taken as-is by sensorfw, or the output will be incorrect. | 20:02 |
| mal | sensorfw iio support is a bit work in progress, I have some pathces for it but not related to that specific issue | 20:02 |
| mal | Tofe: what does sensorfw print? | 20:05 |
| Tofe | mal: not much, but I didn't activate debug output; one moment, let me see how to do that | 20:08 |
| mal | @adampigg how does als look in pinephone does it have the raw one? | 20:09 |
| T42_ | <adampigg> mal: you have one! | 20:10 |
| mal | too lazy to boot it :) | 20:11 |
| mal | it's in another room | 20:11 |
| T42_ | <adampigg> its booted, what am i looking for? | 20:12 |
| Tofe | on hammerhead I did "udevadm info --attribute-walk /sys/bus/i2c/drivers/tsl2772/2-0039/iio:device0" to get my output | 20:13 |
| Tofe | of course the driver and iio device will vary slightly | 20:13 |
| Tofe | here's a debug output for sensorfw on hammerhead: https://paste.tofe.me/p/NgJGFx.log (with following .conf: https://paste.tofe.me/p/8q2mA2.ini ) | 20:15 |
| Tofe | ^ I didn't read it yet, discovering now | 20:16 |
| T42_ | <adampigg> [defaultuser@PinePhone ~]$ cat /sys/bus/iio/devices/iio:device1/ | 20:16 |
| T42_ | <adampigg> dev in_illuminance_scale in_proximity_raw power/ | 20:16 |
| T42_ | <adampigg> events/ in_illuminance_scale_available in_proximity_scale subsystem/ | 20:16 |
| T42_ | <adampigg> in_illuminance_integration_time in_proximity_integration_time in_proximity_scale_available uevent | 20:16 |
| T42_ | <adampigg> in_illuminance_integration_time_available in_proximity_integration_time_available name waiting_for_supplier | 20:16 |
| T42_ | <adampigg> in_illuminance_raw in_proximity_nearlevel of_node/ | 20:16 |
| T42_ | <adampigg> yes, has _raw | 20:16 |
| Tofe | yup | 20:18 |
| Tofe | I guess the kernel driver tls2772 is a bit incomplete, and doesn't provide it. | 20:19 |
| Tofe | but still, in sensorfw the line https://github.com/sailfishos/sensorfw/blob/master/adaptors/iioadaptor/iioadaptor.cpp#L223 puts "ok2" to true a bit too soon, as we don't know yet if we'll find a raw channel | 20:23 |
| mal | I wonder about this in that debug output "alsadaptor" bool SysfsAdaptor::addPath(const QString&, int) "/sys/bus/i2c/drivers/tsl2772/2-0039/iio:device0/in_intensity0_raw" | 20:24 |
| Tofe | ah, this looks like a previous path I put in my config file, but I removed it before restarting sensorfw, so I don't know why it still appears | 20:24 |
| Tofe | let me double check | 20:25 |
| Tofe | ok, my bad, I messed up my logs: https://paste.tofe.me/p/tgy7BY.log | 20:30 |
| Tofe | but wait, it's still there ? wtf | 20:32 |
| Tofe | ooh it created a user config ".config/Unknown\ Organization/sensorfwd.conf"... | 20:38 |
| Tofe | ok, this is the clean one: https://paste.tofe.me/p/ziOgfm.log | 20:39 |
| mal | is that mainline kernel? or something else? | 20:40 |
| Tofe | mal it's a near mainlline, picked up from postmarketos: github.com/msm8974-mainline/linux.git branch=qcom-msm8974-6.9.y | 20:42 |
| mal | Tofe: how does that illuminance input value change? | 20:47 |
| Tofe | it's the last read value, varies typically from 0 to 15 or so | 20:48 |
| Tofe | therefore I think have the _raw value is probably a trivial patch | 20:49 |
| Tofe | having* | 20:49 |
| Tofe | but I'll have some sleep, and continue this tomorrow | 20:50 |
| T42_ | <edp_17> Tofe: What exactly are you working on the hammerhead? | 21:34 |
Generated by irclog2html.py 2.17.1 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!