Inside the container we have v4.15.0. Even though the hardware is mounted in, we can not communicate with the hardware. This is likely because of the version mismatch with the driver on the host.
hailo@jpmeijers-desktop:/local/workspace/tappas/apps/h8/gstreamer/raspberrypi/sa
nity_pipeline$ hailortcli --version
HailoRT-CLI version 4.15.0
hailo@jpmeijers-desktop:/local/workspace/tappas/apps/h8/gstreamer/raspberrypi/sa
nity_pipeline$ hailortcli fw-control identify
[HailoRT] [error] Failed query driver info, errno 25
[HailoRT] [error] CHECK_SUCCESS_AS_EXPECTED failed with status=HAILO_DRIVER_FAIL(36)
[HailoRT] [error] CHECK_EXPECTED failed with status=HAILO_DRIVER_FAIL(36)
[HailoRT] [error] CHECK_EXPECTED failed with status=HAILO_DRIVER_FAIL(36)
[HailoRT CLI] [error] CHECK_EXPECTED failed with status=HAILO_DRIVER_FAIL(36)
hailo@jpmeijers-desktop:/local/workspace/tappas/apps/h8/gstreamer/raspberrypi/sa
nity_pipeline$ ls /dev/hailo*
/dev/hailo0
On the host we see driver 4.18.0 is being used.
$ hailortcli fw-control identify
Executing on device: 0000:01:00.0
Identifying board
Control Protocol Version: 2
Firmware Version: 4.18.0 (release,app,extended context switch buffer)
Logger Version: 0
Board Name: Hailo-8
Device Architecture: HAILO8L
Serial Number: HLDDLBB241602838
Part Number: HM21LB1C2LAE
Product Name: HAILO-8L AI ACC M.2 B+M KEY MODULE EXT TMP
Please update the tappas container for aarch64 with v4.18.0.
My main goal is to get the Raspberry Pi AI Kit in a working state for our industrial customers that use Ubuntu 24.04 Desktop and Server on Raspberry Pi 5’s. Currently with a custom Docker container I can get some examples to work, but it still needs manual tweaking at runtime. Unfortunately that is not a solution for industrial use.
Your official container will solve this, but we need to somehow pin the container version to the correct driver version on the host.
Pull requests are welcomed, I was putting it together before my trip, and there are a few items that I need to clarify next time I get to run through it. Let me know if it helps you!
Sorry, I forgot to update the post. Our R&D team has decided to discontinue providing the TAPPAS Docker for aarch64 to allocate resources for other developments. However, you can still install TAPPAS on your Arm-based system from source via our GitHub repository.