Skip to main content

Troubleshooting

example_camera_capture reports Device or resource busy​

Cause: frame_service is exclusively using /dev/video0.

Solution:

/etc/init.d/S52frame_service stop
./build/bin/example_camera_capture
/etc/init.d/S52frame_service start

For daily screenshots and frame testing, use:

frame_service_cli screenshot --out /tmp/screenshot.bmp

screenshot tool fails​

Check:

/etc/init.d/S52frame_service status
frame_service_cli --socket /run/frame_service/frame_service.sock health
ls -l /run/frame_service/frame_service.sock

Common causes:

  • frame_service is not running;
  • [hid].frame_socket path in agent.toml is inconsistent;
  • HDMI input is not synced;
  • RK628D or TC358743 HDMI subdevice status is abnormal.

Frame Service keeps restarting​

View logs:

tail -f /var/log/frame_service/frame_service.log

Check:

  • Whether /oem/usr/bin/frame_service exists and is executable;
  • Whether /dev/video0 and an RK628D or TC358743 subdevice exist;
  • Whether EDID / HDMI signal is normal;
  • Whether other processes are using /dev/video0.

Find the HDMI bridge subdevice without assuming a fixed index. RK628D reports rk628-csi; TC358743 reports tc358743:

for f in /sys/class/video4linux/v4l-subdev*/name; do echo "$f: $(cat "$f")"; done

If the matching rk628-csi or tc358743 node was not selected, set FRAME_SERVICE_SUBDEV=/dev/v4l-subdevX in /etc/aiden_frame_service.conf to that node and restart S52frame_service.

Agent Web UI won't open​

Check:

/etc/init.d/S53agent status
tail -f /userdata/agent/log/agent.log
netstat -lntp | grep 8080

Verify:

  • The daemon is listening on port 8080; both text and stt modes serve the Web UI;
  • agent.toml TOML syntax is correct;
  • Model configuration and API key are available;
  • Firewall, USB network, or Wi-Fi IP is correct.

Wi-Fi connection is unstable​

Solution:

  • Connect an external 2.4 GHz antenna, antenna connector uses IPEX1 generation;
  • Or try switching Wi-Fi channel to avoid heavily interfered channels.

Voice mode has no sound or cannot record audio​

Check:

/etc/init.d/S53audio_service status
audio_service_cli --socket /run/audio_service/audio_service.sock health
audio_service_cli --socket /run/audio_service/audio_service.sock get-volume
amixer sget 'DAC HPMIX'
amixer sget 'DAC LINEOUT'

Recommendations:

  • First run scripts/setup_audio_volume.sh;
  • Confirm [audio].socket path matches the service;
  • Use record-stream and play-stream to verify recording/playback separately;
  • Check if ffmpeg exists when TTS fails.

RKNN VAD inference fails​

First run helper self-test directly on the board:

/oem/usr/bin/rknn_vad --model /oem/usr/model/silero_vad_6_2_encoder_rv1106_w8a8_v1.rknn --weights /oem/usr/model/silero_vad_6_2_lstm_decoder_weights.bin --self-test
/oem/usr/bin/cpu_vad --weights /oem/usr/model/silero_vad_6_2_lstm_decoder_weights.bin --self-test

On success, it will output P <probability>. Current RV1106 helper uses RKNN zero-copy IO; if it outputs rknn_set_io_mem failed, rknn_run failed, or old helper outputs rknn_inputs_set failed, check:

  • Whether /oem/usr/lib/librknnmrt.so version matches the model;
  • Whether silero_vad_6_2_encoder_rv1106_w8a8_v1.rknn is the encoder model re-converted for RV1106 target;
  • Whether input/output tensor type, size, scale, zero-point in helper logs are normal.

HID input is ineffective​

Check:

ls -l /dev/hidg*
mount | grep configfs
lsmod | grep -E 'dwc2|libcomposite'

Try reinitializing:

sudo ./build/bin/example_usb_hid cleanup
sudo ./build/bin/example_usb_hid setup composite

For iOS target devices, confirm AssistiveTouch is enabled.

Cannot capture screen from iPhone 16e​

Cause: iPhone 16e's USB-C port has compatibility issues.

Solution:

  • Currently unable to support screen capture from iPhone 16e;
  • Switch to other compatible models for testing.

Board reboots frequently after connecting phone​

Cause: Insufficient power supply.

Solution:

  • Switch to a USB hub with better power supply capability;
  • Or use a USB hub with external power support.

HTTP Tool API access exception​

  • Set NO_PROXY for device private IP / USB network adapter address;
  • First access GET /api/tools to confirm service is reachable;
  • When a tool invocation fails, check is_error and output in the response from POST /api/tools/{tool_name};
  • Separate transport failure from tool failure judgement.

opkg update prints nothing, and no package can be found​

# opkg update
# opkg install htop
error: opkg_prepare_url_for_install: Couldn't find anything to satisfy 'htop'.

opkg update prints one Downloading line per configured source, so silence means no source is configured. Two different causes produce it:

cat /etc/opkg/*.conf | grep -E '^\s*(src|dist)'   # any sources at all?
/etc/init.d/S22opt status                          # is /opt actually bound?

Every src line lives in /opt/etc/opkg/userfeeds.conf, so an unbound /opt looks exactly like "no source was ever configured". Check both before writing a source into a file that cannot be read.

Note this is not opkg refusing to start. /etc/opkg/90-userfeeds.conf is a symlink into /opt, and when it dangles uClibc-ng's glob() silently drops it — opkg never tries to open it and runs normally. The protection against installing into the A/B rootfs comes from the read-only seal S22opt applies to /opt, plus the fact that no source is reachable.

Recover the /opt mount​

opt=bound is the only healthy status. Other states require mount recovery before running opkg:

  • opt=sealed: /userdata was unavailable or the bind mount failed, so S22opt protected the rootfs with a read-only tmpfs. Confirm that /userdata is mounted, fix the reported cause, then run /etc/init.d/S22opt restart.
  • opt=wrong-source: something other than /userdata/opt is mounted at /opt. Do not run opkg or unmount it blindly; stop processes using that mount, remove the configuration that created it, and reboot.
  • opt=unbound: run /etc/init.d/S22opt restart and use its error output to diagnose the missing /userdata mount, directory creation failure, or bind failure.

After recovery, require a successful status check before continuing:

/etc/init.d/S22opt status
# opt=bound source=/userdata/opt

Configure a package feed​

Package sources belong only in /opt/etc/opkg/userfeeds.conf. The init script seeds the Entware source when it creates this file for the first time, but it never changes an existing file, including an empty one. To enable or change a feed, first verify that /opt is bound, then edit the file so it contains one source declaration per feed:

src/gz entware https://bin.entware.net/armv7sf-k3.2

Avoid declaring the same source more than once. After saving the file, refresh the package lists and confirm that packages are visible:

opkg update
opkg list | head

Run plain opkg commands without -f; supplying a standalone configuration file bypasses the fixed fragments under /etc/opkg that keep package files, metadata, caches, and locks in their intended locations.

Recover opkg metadata​

If an interrupted opkg update leaves only the downloaded package lists or cache inconsistent, remove those regenerable files and download them again:

rm -f /opt/var/lib/opkg/lists/* /opt/var/cache/opkg/*
opkg update

Do not use that procedure for a truncated /opt/var/lib/opkg/status file. The status file, /opt/var/lib/opkg/info, and the installed package files form one state and must come from the same backup. Stop programs running from /opt, unmount /opt, move the current /userdata/opt aside instead of overwriting it, restore the complete backed-up tree at /userdata/opt, and start S22opt again. Preserve file ownership, modes, and symlinks during the restore. The S22opt stop action deliberately leaves /opt mounted because package processes may still be running; it is not a substitute for stopping those processes and running umount /opt before the replacement.

If no consistent backup exists, reset the whole package root rather than mixing old package files with a new database. Stop /opt programs, unmount /opt, move /userdata/opt to a recovery path, and run /etc/init.d/S22opt start. The init script creates a clean package layout and feed file; then run opkg update and reinstall the required packages. If /opt is busy, identify and stop its users or reboot into a maintenance state, then repeat the unmount check before replacing or restoring the package root.

Docker build fails​

Check:

docker buildx version
docker info

Recommended for Apple Silicon:

colima start --vm-type vz --vz-rosetta
./build.sh

Confirm not using --arch x86_64 to start Colima VM.