Fatal error on MSS CPU

[f-alemauro]

Hello guys,

I have this setup:

Raspberry PI 4 with POE Hat
Raspian buster
depthai v. 2.13.3.0
OAK-1 (on USB3) powered by USB

Every 2 to 5 minutes our service crash and this is the system log I have

Jul 15 12:30:36 easycounter python3[25278]: [14442C1051E395D000] [70.961] [system] [critical] Fatal error. Please report to developers. Log: 'Fatal error on MSS CPU: trap: 00, address: 00000000' '0'
Jul 15 12:30:37 easycounter kernel: [252103.450196] usb 1-1.1: new high-speed USB device number 57 using xhci_hcd
Jul 15 12:30:38 easycounter kernel: [252103.600999] usb 1-1.1: New USB device found, idVendor=03e7, idProduct=2485, bcdDevice= 0.01
Jul 15 12:30:38 easycounter kernel: [252103.601017] usb 1-1.1: New USB device strings: Mfr=1, Product=2, SerialNumber=3
Jul 15 12:30:38 easycounter kernel: [252103.601031] usb 1-1.1: Product: Movidius MyriadX
Jul 15 12:30:38 easycounter kernel: [252103.601045] usb 1-1.1: Manufacturer: Movidius Ltd.
Jul 15 12:30:38 easycounter kernel: [252103.601058] usb 1-1.1: SerialNumber: 03e72485
Jul 15 12:30:38 easycounter mtp-probe: checking bus 1, device 57: "/sys/devices/platform/scb/fd500000.pcie/pci0000:00/0000:00:00.0/0000:01:00.0/usb1/1-1/1-1.1"
Jul 15 12:30:38 easycounter mtp-probe: bus: 1, device: 57 was not an MTP device
Jul 15 12:30:38 easycounter mtp-probe: checking bus 1, device 57: "/sys/devices/platform/scb/fd500000.pcie/pci0000:00/0000:00:00.0/0000:01:00.0/usb1/1-1/1-1.1"
Jul 15 12:30:38 easycounter mtp-probe: bus: 1, device: 57 was not an MTP device
Jul 15 12:30:38 easycounter python3[25278]: Stack trace (most recent call last) in thread 25349:
Jul 15 12:30:38 easycounter python3[25278]: #3    Object "/home/brsadm/bs4-ai-core/venv/lib/python3.7/site-packages/depthai.cpython-37m-arm-linux-gnueabihf.so", at 0xb58e1937, in
Jul 15 12:30:38 easycounter python3[25278]: #2    Object "/home/brsadm/bs4-ai-core/venv/lib/python3.7/site-packages/depthai.cpython-37m-arm-linux-gnueabihf.so", at 0xb580ceaf, in
Jul 15 12:30:38 easycounter python3[25278]: #1    Object "/lib/arm-linux-gnueabihf/libc.so.6", at 0xb6ced11f, in
Jul 15 12:30:38 easycounter python3[25278]: #0    Object "/home/brsadm/bs4-ai-core/venv/lib/python3.7/site-packages/depthai.cpython-37m-arm-linux-gnueabihf.so", at 0xb598a5db, in backward::SignalHandling::sig_handler(int, siginfo_t*, void*)
Jul 15 12:30:38 easycounter python3[25278]: Bus error (Invalid address alignment [0x786f627b])
Jul 15 12:30:38 easycounter kernel: [252103.930502] usb 2-1: USB disconnect, device number 49
Jul 15 12:30:38 easycounter kernel: [252103.930939] xhci_hcd 0000:01:00.0: WARN Set TR Deq Ptr cmd failed due to incorrect slot or ep state.
Jul 15 12:30:38 easycounter kernel: [252103.932731] Alignment trap: not handling instruction e1932f9f at [<b580ceac>]
Jul 15 12:30:38 easycounter kernel: [252103.932741] 8<--- cut here ---
Jul 15 12:30:38 easycounter kernel: [252103.932756] Unhandled fault: alignment exception (0x221) at 0x786f627b
Jul 15 12:30:38 easycounter kernel: [252103.932768] pgd = e2b9fabf
Jul 15 12:30:38 easycounter kernel: [252103.932779] [786f627b] *pgd=00000000
Jul 15 12:30:38 easycounter systemd[1]: easycounter.service: Main process exited, code=killed, status=7/BUS
Jul 15 12:30:38 easycounter systemd[1]: easycounter.service: Failed with result 'signal'.

I've read this issues https://github.com/luxonis/depthai-core/issues/467 and applied this patch

script.setProcessor(dai.ProcessorType.LEON_CSS)

Doing test right now. Could this solve the problem? Can you point me to a documentation about LEON CSS and LEON MSS?

Thank you!

[Erol444]

Hi @f-alemauro , I would suggest using the latest depthai version (2.17), as since 2.13.3 we have added a lot of bugfixes. And currenty, LEON CSS runs a lot more stable and doesn't cause sporadic crashes. LEON cores docs can be found here. Thanks, Erik

[f-alemauro]

Has the image_manip_refactor branch been merged into the 2.17?

[themarpe]

@f-alemauro Yes, it was https://github.com/luxonis/depthai-python/releases/tag/v2.17.0.0