Irq error on Xavier

Autonomous Machines Jetson & Embedded Systems Jetson AGX Xavier
1

Hi nvidia,

I have a problem about irq on my custom board:

[    1.754397] irq 103: nobody cared (try booting with the "irqpoll" option)
[    1.754515] CPU: 0 PID: 0 Comm: swapper/0 Not tainted 4.9.140-tegra #210
[    1.754582] Hardware name: Jetson-AGX (DT)
[    1.754638] Call trace:
[    1.754712] [<ffffff800808bdb8>] dump_backtrace+0x0/0x198
[    1.754797] [<ffffff800808c37c>] show_stack+0x24/0x30
[    1.754874] [<ffffff800845d820>] dump_stack+0x98/0xc0
[    1.754950] [<ffffff80081261d4>] __report_bad_irq+0x3c/0xf8
[    1.755031] [<ffffff8008126638>] note_interrupt+0x2c8/0x318
[    1.755119] [<ffffff80081234d0>] handle_irq_event_percpu+0x50/0x60
[    1.755200] [<ffffff8008123530>] handle_irq_event+0x50/0x80
[    1.755281] [<ffffff80081272f8>] handle_fasteoi_irq+0xc8/0x1b8
[    1.755383] [<ffffff800812224c>] generic_handle_irq+0x34/0x50
[    1.755485] [<ffffff8008122930>] __handle_domain_irq+0x68/0xc0
[    1.755554] [<ffffff8008080d44>] gic_handle_irq+0x5c/0xb0
[    1.755637] [<ffffff8008082be8>] el1_irq+0xe8/0x18c
[    1.755721] [<ffffff80080bb298>] irq_exit+0xd0/0x118
[    1.755786] [<ffffff8008122934>] __handle_domain_irq+0x6c/0xc0
[    1.755884] [<ffffff8008080d44>] gic_handle_irq+0x5c/0xb0
[    1.755972] [<ffffff8008082be8>] el1_irq+0xe8/0x18c
[    1.756063] [<ffffff8008085d20>] arch_cpu_idle+0x30/0x1b0
[    1.756129] [<ffffff8008112d28>] cpu_startup_entry+0x140/0x200
[    1.756222] [<ffffff8008f87934>] rest_init+0x84/0x90
[    1.756289] [<ffffff8009640b64>] start_kernel+0x370/0x384
[    1.756355] [<ffffff8009640204>] __primary_switched+0x80/0x94
[    1.756420] handlers:
[    1.756467] [<ffffff8008c7bb58>] tegra_mcerr_hard_irq threaded [<ffffff8008c7bd08>] tegra_mcerr_thread
[    1.756590] Disabling IRQ #103

I’ve found the interrupt response function in platform/tegra/mc/mcerr.c, but I was not sure the pin number of this interrupt from the codes.
Can you tell me this pin number so that I can check the hardware problem.

Thanks.

3

Can you share complete log to analysis.

4

Looks like a spurious or unhandled irq #71 related to nvgpu (gv11b).
Please check if any changes done and try once after disabling those to see if error still comes. Also, please share complete logs to check if any more errors.

5

Hi Sumitg,

I didn’t modify anything about nvgpu. I used l4t-r32.4.2.
Please see the attachment for the complete log.
irq_error.log (110.5 KB)

Thanks.

6

Does the default kernel Image/dtb have the same problem?

7

Hi Shane,

I had not enough tests with the default kernel Image/dtb, because this problem is accidental.
I judged that this is a hardware problem. However, I’ve not found the GPIO number corresponding to the interrupt from the codes. I want to know which GPIO signal to detect.

8

Hi 295839633,

Have you clarified the HW cause and reworked to fix it?
Any result can be shared or need further support?

10

Hi kayccc,

We finally decided to ignore it, because it had not caused any problems on the use of my clients.

Thank you.