imx8qxp unhandled vpu interrupt

[angelo_d]

Hi all,

on imx8qxpmek, kernel  imx_5.4.70_2.3.0

i am occasionally getting

[ 2.377023] 000: irq 464: nobody cared (try booting with the "irqpoll" option)
[ 2.377032] 000: CPU: 0 PID: 0 Comm: swapper/0 Not tainted 5.4.70-rt40+gb9144d02 #1
[ 2.377039] 000: Hardware name: Freescale i.MX8QXP MEK (DT)
[ 2.377043] 000: Call trace:
[ 2.377045] 000: dump_backtrace+0x0/0x140
[ 2.377057] 000: show_stack+0x14/0x20
[ 2.377062] 000: dump_stack+0xb0/0x10c
[ 2.377070] 000: __report_bad_irq+0x48/0xd8
[ 2.377077] 000: note_interrupt+0x2bc/0x374
[ 2.377081] 000: handle_irq_event_percpu+0xa4/0xb0
[ 2.377089] 000: handle_irq_event+0x74/0x118

464 (496) is VPU_INT_0 interrupt. 

Could i safely ignore the info ?

Regards,

angelo

[angelo_d]

Well issue is related to rt patch i applied. Removing rt patch solves.

[angelo_d]

Hi, thanks,

sure, seems not creating any issue.

If it may be useful, did some further investigation:

the irq numbers are not significant (they are not matching RM values), so the "unhandled" is not related to VPU, but, required from 

i2c-16 37230000.i2c cm4 LPI2C using irq 464

So, since LPI2C M4 is also visible from the application processor, and there is no fw loaded for M4, very likely, linux signals it as unhandled.

[Bio_TICFSL]

Hello spectrum,

Try to burn again the sdcard in order to see if the problem go, otherwise if this not causing any other problem you can continue as it. i saw this issue before but with the burn out again the problem is solved.

Regards