*** Herrie|2 is now known as Herrie | 02:44 | |
*** Sellerie1 is now known as Sellerie | 02:44 | |
T42 | <simonschmeisser> mal: the hw-settings.ini on device is the same as here https://github.com/ghosalmartin/droid-config-bullhead/blob/master/sparse/usr/share/csd/settings.d/hw-settings.ini | 08:00 |
---|---|---|
T42 | <simonschmeisser> test_sensors: https://paste.ubuntu.com/p/b2C62tt7N7/ | 08:14 |
mal | @simonschmeisser the hw-settings.ini is really in correct location? | 12:02 |
T42 | <simonschmeisser> mal: I think so: -rw-r--r-- 1 root root 1396 Nov 23 18:32 /usr/share/csd/settings.d/hw-settings.ini | 18:53 |
T42 | <linusdan> @edp_17 : Happy new year :) | 19:23 |
T42 | <linusdan> By experiment. I want to see if it works outside the base of hybris-12.1. For now it still didn't work, I'm stuck on the samsung logo 😂 | 19:23 |
mal | @simonschmeisser does journalctl have any messages from sensorfw | 19:29 |
T42 | <simonschmeisser> mal: actually I couldn't find anything from sensorfw in the joutnalctl: https://paste.ubuntu.com/p/tNTsys2DDh/ | 20:15 |
mal | @simonschmeisser what happens if you do systemctl restart sensorfwd? | 20:19 |
T42 | <edp_17> @linusdan [@edp_17 : Happy new year :) | 20:20 |
T42 | <edp_17> By experiment. I w …], I see. If you can ssh or telnet into the device to grab some logs, there is hope. 😉 | 20:20 |
T42 | <simonschmeisser> mal: now there is a few lines in journal: https://paste.ubuntu.com/p/WBWJcyMTP2/ | 20:25 |
mal | so which sensors work now? | 20:26 |
T42 | <simonschmeisser> light doesnt, proximity sometimes switches but seems more like random, gyro pass, accel has gx at 1.4, compass fails | 20:30 |
mal | add this kind of config for proximity https://github.com/mer-hybris/droid-config-sony-ganges-pie/blob/master/sparse/etc/mce/60-proximity-sensor-ganges.conf | 20:38 |
mal | how bright light have you shown to light sensor? | 20:38 |
T42 | <simonschmeisser> mal: good point, it's somewhat dark in here ... using a 700lm headlamp I can make it show 18304 but the values are somewhat sticky and do not update? | 20:44 |
T42 | <simonschmeisser> current light value I mean | 20:45 |
T42 | <simonschmeisser> 20hybris-features.ini has ProximitySensor=false and LightSensor=false does that matter? | 20:50 |
T42 | <simonschmeisser> mal: there are a few more of these in the logs: Jan 03 21:46:12 Sailfish sensorfwd[6699]: Invalid interval requested for node ' "proximityadaptor" ' by session ' 15 ': 0 | 20:53 |
mal | @simonschmeisser do not otherwise touch mce configs related to sensors | 20:55 |
mal | only the file I mentioned | 20:55 |
T42 | <simonschmeisser> sure | 20:55 |
mal | it's normal to see some "Invalid interval requested for node" messages | 20:56 |
T42 | <simonschmeisser> now proximitysensor "works" worse, doesn't flip anymore at all | 20:57 |
T42 | <elros34> if you have light sensor reported by evdev_trace (if it's iio device then in sysfs) then you could at least check whether it reports sane values | 20:58 |
mal | yeah, check if there are sensors in "evdev_trace -i" output | 20:58 |
mal | also some strange devices needed power toggling via sysfs to make sensors work properly | 20:59 |
mal | not sure if that could be needed | 20:59 |
T42 | <simonschmeisser> what package is that in? | 21:00 |
T42 | <elros34> mce-tools | 21:02 |
T42 | <simonschmeisser> mal: proximity is there but at 0 https://paste.ubuntu.com/p/MFkGn7yqw5/ | 21:06 |
T42 | <simonschmeisser> cannot identify light sensor | 21:06 |
T42 | <elros34> check how proximity behaves evdev_trace -t 1 | 21:10 |
T42 | <elros34> it might be proxmity and als sensor in one | 21:12 |
T42 | <simonschmeisser> does not give any output | 21:13 |
T42 | <elros34> search in /sysfs for 6180, it might have parameters/enable or something similar | 21:18 |
T42 | <simonschmeisser> @elros34 ls /sys/class/input/event1/device/ -> capabilities event1 id modalias name phys power properties subsystem uevent uniq | 21:30 |
T42 | <elros34> something more like /sys/module/*6180*/parameters | 21:32 |
T42 | <simonschmeisser> ah ok | 21:34 |
T42 | <elros34> according to driver this should be enable_ps_sensor property | 21:35 |
T42 | <simonschmeisser> ls /sys/module/stmvl6180/ -> uevent version | 21:36 |
T42 | <elros34> what about /sys/kernel/range/enable_ps_sensor? | 21:37 |
T42 | <simonschmeisser> that exists | 21:40 |
T42 | <simonschmeisser> echo 1 > ? | 21:40 |
T42 | <elros34> sure | 21:41 |
T42 | <simonschmeisser> now evdev_trace -t 1 spits out lots of stuff | 21:42 |
T42 | <simonschmeisser> but the csd test does not change | 21:43 |
T42 | <simonschmeisser> sometimes it includes a value for distance | 21:44 |
T42 | <simonschmeisser> https://paste.ubuntu.com/p/9BQ2vYGQ7k/ | 21:46 |
T42 | <elros34> the only one you should care is ABS_DISTANCE or ABS_HAT1X. There is evdev adaptor in sensorfw which can be adapted for that driver as last resort but I not sure whether it will be work correctly | 21:52 |
T42 | <elros34> do you have state: OPEN -> CLOSED messages from mce now? | 21:54 |
T42 | <simonschmeisser> Jan 03 22:42:59 Sailfish mce[742]: modules/proximity.c: mp_datapipe_set_proximity_sensor_actual(): state: OPEN -> UNDEF | 21:56 |
T42 | <simonschmeisser> only one | 21:56 |
T42 | <simonschmeisser> HAT1X is the sensor on the back for autofocus | 21:59 |
T42 | <simonschmeisser> "infra"red | 22:00 |
T42 | <elros34> are you sure? that would mean we debug wrong sensor | 22:01 |
T42 | <simonschmeisser> yes, also DISTANCE is on the back | 22:02 |
T42 | <simonschmeisser> seems like cm vs mm | 22:02 |
mal | based on test_sensors output the proximity sensor is "RPR0521 Proximity Sensor" | 22:06 |
T42 | <elros34> so I guess something like that: https://android.googlesource.com/device/google/contexthub/+/refs/heads/master/firmware/os/drivers/rohm_rpr0521/rohm_rpr0521.c. Maybe you will find something in logcat then | 22:31 |
T42 | <simonschmeisser> @elros34 I get this line a couple of times but it's not that informative https://android.googlesource.com/device/google/contexthub/+/refs/heads/master/firmware/os/drivers/rohm_rpr0521/rohm_rpr0521.c#521 (offset=6) | 22:50 |
T42 | <simonschmeisser> I think I give up for now, will rather try to work on the pinephone camera once more tomorrow | 23:00 |
T42 | <agnesmarain> Fwd from Telegram: https://t.me/joinchat/TXL9L4WT1AAT_khG | 23:20 |
Generated by irclog2html.py 2.17.1 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!