Unable to determine AccelerometerOrientation

[bgaff]

For some reason screen rotation stopped working suddenly for me, if I query the devices via: watch -n 0.1 cat /sys/bus/iio/devices/iio\:device*/*raw* /dev/iio* I will see the devices change with my rotation. I tried rebuilding iio-sensor-proxy from master and still it's unable to determine the orientation. Hardware is a Dell Inspiron 13 7000.

$ uname -r
4.4.0-22-generic

$ sudo gdbus introspect --system --dest net.hadess.SensorProxy --object-path /net/hadess/SensorProxy
node /net/hadess/SensorProxy {
  interface org.freedesktop.DBus.Properties {
    methods:ien
      Get(in  s interface_name,
          in  s property_name,
          out v value);
      GetAll(in  s interface_name,
             out a{sv} properties);
      Set(in  s interface_name,
          in  s property_name,
          in  v value);
    signals:
      PropertiesChanged(s interface_name,
                        a{sv} changed_properties,
                        as invalidated_properties);
    properties:
  };
  interface org.freedesktop.DBus.Introspectable {
    methods:
      Introspect(out s xml_data);
    signals:
    properties:
  };
  interface org.freedesktop.DBus.Peer {
    methods:
      Ping();
      GetMachineId(out s machine_uuid);
    signals:
    properties:
  };
  interface net.hadess.SensorProxy {
    methods:
      ClaimAccelerometer();
      ReleaseAccelerometer();
      ClaimLight();
      ReleaseLight();
    signals:
    properties:
      readonly b HasAccelerometer = true;
      readonly s AccelerometerOrientation = 'undefined';
      readonly b HasAmbientLight = true;
      readonly s LightLevelUnit = 'lux';
      readonly d LightLevel = 0.0;
  };
  node Compass {
  };
};

[bgaff]

The output from running iio-proxy-sensor directly via GDB:

(gdb) r
Starting program: /usr/local/iio-sensor-proxy-master/sbin/iio-sensor-proxy 
[Thread debugging using libthread_db enabled]
Using host libthread_db library "/lib/x86_64-linux-gnu/libthread_db.so.1".
[New Thread 0x7ffff5912700 (LWP 11689)]
[New Thread 0x7ffff5111700 (LWP 11690)]
[New Thread 0x7ffff4910700 (LWP 11691)]
** (process:11685): DEBUG: Found als at /sys/devices/pci0000:00/0000:00:14.0/usb1/1-4/1-4:1.0/0003:0483:91D1.0002/HID-SENSOR-200041.7.auto/iio:device3
** (process:11685): DEBUG: Found device /sys/devices/pci0000:00/0000:00:14.0/usb1/1-4/1-4:1.0/0003:0483:91D1.0002/HID-SENSOR-200041.7.auto/iio:device3 of type ambient light sensor at IIO Buffer Light sensor
** (process:11685): DEBUG: Found accel_3d at /sys/devices/pci0000:00/0000:00:14.0/usb1/1-4/1-4:1.0/0003:0483:91D1.0002/HID-SENSOR-200073.2.auto/iio:device2
** (process:11685): DEBUG: Found device /sys/devices/pci0000:00/0000:00:14.0/usb1/1-4/1-4:1.0/0003:0483:91D1.0002/HID-SENSOR-200073.2.auto/iio:device2 of type accelerometer at IIO Buffer accelerometer
** (process:11685): DEBUG: Found magn_3d at /sys/devices/pci0000:00/0000:00:14.0/usb1/1-4/1-4:1.0/0003:0483:91D1.0002/HID-SENSOR-200083.4.auto/iio:device1
** (process:11685): DEBUG: Found device /sys/devices/pci0000:00/0000:00:14.0/usb1/1-4/1-4:1.0/0003:0483:91D1.0002/HID-SENSOR-200083.4.auto/iio:device1 of type compass at IIO Buffer Compass
** (process:11685): DEBUG: Found associated trigger at /sys/devices/pci0000:00/0000:00:14.0/usb1/1-4/1-4:1.0/0003:0483:91D1.0002/HID-SENSOR-200073.2.auto/trigger2
** (process:11685): DEBUG: Got type for in_accel_x: is signed: 1, bytes: 4, bits_used: 32, shift: 0, mask: 0x0, be: 0
** (process:11685): DEBUG: Got type for in_accel_y: is signed: 1, bytes: 4, bits_used: 32, shift: 0, mask: 0x0, be: 0
** (process:11685): DEBUG: Got type for in_accel_z: is signed: 1, bytes: 4, bits_used: 32, shift: 0, mask: 0x0, be: 0
** (process:11685): DEBUG: Built channel array for in_accel_x: is signed: 1, bytes: 4, bits_used: 32, shift: 0, mask: 0x0, be: 0
** (process:11685): DEBUG: Built channel array for in_accel_y: is signed: 1, bytes: 4, bits_used: 32, shift: 0, mask: 0x0, be: 0
** (process:11685): DEBUG: Built channel array for in_accel_z: is signed: 1, bytes: 4, bits_used: 32, shift: 0, mask: 0x0, be: 0
** (process:11685): DEBUG: Found associated trigger at /sys/devices/pci0000:00/0000:00:14.0/usb1/1-4/1-4:1.0/0003:0483:91D1.0002/HID-SENSOR-200041.7.auto/trigger3
** (process:11685): DEBUG: Got type for in_intensity_both: is signed: 1, bytes: 4, bits_used: 32, shift: 0, mask: 0x0, be: 0
** (process:11685): DEBUG: Built channel array for in_intensity_both: is signed: 1, bytes: 4, bits_used: 32, shift: 0, mask: 0x0, be: 0
** (process:11685): DEBUG: Found associated trigger at /sys/devices/pci0000:00/0000:00:14.0/usb1/1-4/1-4:1.0/0003:0483:91D1.0002/HID-SENSOR-200083.4.auto/trigger1
** (process:11685): DEBUG: Got type for in_magn_x: is signed: 1, bytes: 4, bits_used: 32, shift: 0, mask: 0x0, be: 0
** (process:11685): DEBUG: Got type for in_magn_y: is signed: 1, bytes: 4, bits_used: 32, shift: 0, mask: 0x0, be: 0
** (process:11685): DEBUG: Got type for in_magn_z: is signed: 1, bytes: 4, bits_used: 32, shift: 0, mask: 0x0, be: 0
** (process:11685): DEBUG: Got type for in_rot_from_north_magnetic_tilt_comp: is signed: 1, bytes: 4, bits_used: 32, shift: 0, mask: 0x0, be: 0
** (process:11685): DEBUG: Built channel array for in_magn_x: is signed: 1, bytes: 4, bits_used: 32, shift: 0, mask: 0x0, be: 0
** (process:11685): DEBUG: Built channel array for in_magn_y: is signed: 1, bytes: 4, bits_used: 32, shift: 0, mask: 0x0, be: 0
** (process:11685): DEBUG: Built channel array for in_magn_z: is signed: 1, bytes: 4, bits_used: 32, shift: 0, mask: 0x0, be: 0
** (process:11685): DEBUG: Built channel array for in_rot_from_north_magnetic_tilt_comp: is signed: 1, bytes: 4, bits_used: 32, shift: 0, mask: 0x0, be: 0
** (process:11685): DEBUG: Handling driver refcounting method 'ClaimLight' for ambient light sensor device
** (process:11685): DEBUG: Handling driver refcounting method 'ClaimLight' for ambient light sensor device
** (process:11685): DEBUG: Handling driver refcounting method 'ClaimAccelerometer' for accelerometer device
** (process:11685): DEBUG: Handling driver refcounting method 'ClaimAccelerometer' for accelerometer device
** (process:11685): DEBUG: No new data available
** (process:11685): DEBUG: No new data available
** (process:11685): DEBUG: No new data available

[bgaff]

Like I said, this was working about a day ago, nothing changed (as far as I can tell) on my system. I even tried the only other kernel version available on my system and it had the same problem. Any suggestions on what/where to check to figure out what could have changed or broke?

[bgaff]

After a little more digging I see that people suggested that things will work after a suspend, sure enough I tried suspending and now it's able to determine orientation. This seems to suggest a kernel bug. Also note that it's now correctly determining the light level as well.

$ sudo gdbus introspect --system --dest net.hadess.SensorProxy --object-path /net/hadess/SensorProxy
node /net/hadess/SensorProxy {
  interface org.freedesktop.DBus.Properties {
    methods:
      Get(in  s interface_name,
          in  s property_name,
          out v value);
      GetAll(in  s interface_name,
             out a{sv} properties);
      Set(in  s interface_name,
          in  s property_name,
          in  v value);
    signals:
      PropertiesChanged(s interface_name,
                        a{sv} changed_properties,
                        as invalidated_properties);
    properties:
  };
  interface org.freedesktop.DBus.Introspectable {
    methods:
      Introspect(out s xml_data);
    signals:
    properties:
  };
  interface org.freedesktop.DBus.Peer {
    methods:
      Ping();
      GetMachineId(out s machine_uuid);
    signals:
    properties:
  };
  interface net.hadess.SensorProxy {
    methods:
      ClaimAccelerometer();
      ReleaseAccelerometer();
      ClaimLight();
      ReleaseLight();
    signals:
    properties:
      readonly b HasAccelerometer = true;
      readonly s AccelerometerOrientation = 'normal';
      readonly b HasAmbientLight = true;
      readonly s LightLevelUnit = 'lux';
      readonly d LightLevel = 71.0;
  };
  node Compass {
  };
};

[nacc]

@bgaff I'm still seeing this issue (Ubuntu 16.04). S/R seems to make no difference (although I believe it used to...). Are you still seeing successful rotation with current kernels?

[nacc]

@bgaff FYI, I just built and booted into 4.7-rc5+ (0b295dd5b8eb755bc7b57220d93705994090e954 is my current tip HEAD) and screen rotation worked fine, regardless of suspend & resume.

[kekx]

@nacc Hm, I just installed the 4.7-rc7 kernel from the Fedora rawhide repo and it still doesn't work without suspend & resume. Using a Surface Pro 2 btw which is tested according to README.

What might be interesting for @hadess though is that after S&R and while running iio-server-proxy with debug output, as soon as I clip on my type cover, iio-server-proxy seems to lose all the sensors and I have to S&R again to regain functionality. Maybe attaching devices mixes up the iio devices or something.

[kekx]

Found a workaround for now: I simply added Type=idle to the service file to make it start later in the boot process. It never worked before directly after boot, now it does.

[bgaff]

Yah on 4.7-rc5 it still doesn't properly work, can you either provide more details?

[hadess]

Worked-around in master, see #116 for details.

[bgaff]

Thanks!