Nvgpu_cic_mon_report_err_safety_services:92 [ERR] Error reporting is not supported in this platform

Autonomous Machines Jetson & Embedded Systems Jetson AGX Orin
, ,
1

Hi NV team，

When we use G-Streamer && CUDA to process related video data, the following error message may appear：

Jun 25 01:14:35 cntjitpy-igvak-9-dcu2 kernel: [ 2074.933337] nvgpu: 17000000.ga10b nvgpu_cic_mon_report_err_safety_services:92   [ERR]  Error reporting is not supported in this platform
Jun 25 01:14:35 cntjitpy-igvak-9-dcu2 kernel: [ 2074.934038] nvgpu: 17000000.ga10b nvgpu_cic_mon_report_err_safety_services:55   [ERR]  Error reporting is not supported in this platform
Jun 25 01:14:35 cntjitpy-igvak-9-dcu2 kernel: [ 2074.946288] nvgpu: 17000000.ga10b gv11b_mm_mmu_fault_handle_buf_valid_entry:525  [ERR]  page fault error: err_type = 0x8, fault_status = 0x200
Jun 25 01:14:35 cntjitpy-igvak-9-dcu2 kernel: [ 2074.946298] nvgpu: 17000000.ga10b      gv11b_fb_mmu_fault_info_dump:294  [ERR]  [MMU FAULT] mmu engine id:  65, ch id:  460, fault addr: 0x204b94000, fault addr aperture: 0, fault type: invalid pte, access type: virt atomic weak, 
Jun 25 01:14:35 cntjitpy-igvak-9-dcu2 kernel: [ 2074.946300] nvgpu: 17000000.ga10b      gv11b_fb_mmu_fault_info_dump:307  [ERR]  [MMU FAULT] protected mode: 0, client type: gpc, client id:  t1_5, gpc id if client type is gpc: 1, 
Jun 25 01:14:35 cntjitpy-igvak-9-dcu2 kernel: [ 2074.946308] nvgpu: 17000000.ga10b                nvgpu_rc_mmu_fault:352  [ERR]  mmu fault id=12 id_type=1 act_eng_bitmask=00000001
Jun 25 01:14:35 cntjitpy-igvak-9-dcu2 kernel: [ 2074.946340] nvgpu: 17000000.ga10b       nvgpu_tsg_set_ctx_mmu_error:648  [ERR]  TSG 12 generated a mmu fault
Jun 25 01:14:35 cntjitpy-igvak-9-dcu2 kernel: [ 2074.946349] nvgpu: 17000000.ga10b     nvgpu_set_err_notifier_locked:149  [ERR]  error notifier set to 31 for ch 460
Jun 25 01:14:35 cntjitpy-igvak-9-dcu2 kernel: [ 2074.959363] nvgpu: 17000000.ga10b nvgpu_gr_intr_handle_sm_exception:365  [ERR]  sm machine check err. gpc_id(1), tpc_id(2), offset(36864)
Jun 25 01:14:36 cntjitpy-igvak-9-dcu2 kernel: [ 2074.973076] __ga10b__ Channel Status - chip ga10b
Jun 25 01:14:36 cntjitpy-igvak-9-dcu2 kernel: [ 2075.057388] __ga10b__ ---------------------------
Jun 25 01:14:36 cntjitpy-igvak-9-dcu2 kernel: [ 2075.062317] __ga10b__ 429-ga10b, TSG: 17, pid 18984, refs: 2, deterministic: yes, domain name: (default)
Jun 25 01:14:36 cntjitpy-igvak-9-dcu2 kernel: [ 2075.067243] __ga10b__ channel status:  in use idle not busy
Jun 25 01:14:36 cntjitpy-igvak-9-dcu2 kernel: [ 2075.077187] __ga10b__ RAMFC: TOP: 000000000000 PUT: 000201384020 GET: 000201384020 FETCH: 000000000000 HEADER: 21540300 COUNT: 00000000 SEMAPHORE: addr 000000000000 payload 0000000000000000 execute 00000000
Jun 25 01:14:36 cntjitpy-igvak-9-dcu2 kernel: [ 2075.083017] __ga10b__  
Jun 25 01:14:36 cntjitpy-igvak-9-dcu2 kernel: [ 2075.102158] __ga10b__ 430-ga10b, TSG: 17, pid 18984, refs: 2, deterministic: yes, domain name: (default)
Jun 25 01:14:36 cntjitpy-igvak-9-dcu2 kernel: [ 2075.104711] __ga10b__ channel status:  in use idle not busy
Jun 25 01:14:36 cntjitpy-igvak-9-dcu2 kernel: [ 2075.114614] __ga10b__ RAMFC: TOP: 000000000000 PUT: 000201284020 GET: 000201284020 FETCH: 000000000000 HEADER: 21540300 COUNT: 00000000 SEMAPHORE: addr 000000000000 payload 0000000000000000 execute 00000000

Log files of two recurring issues：
Jun_25_01:14:35.txt (170.3 KB)
Jun_28_06:31:40.txt (290.3 KB)

Please analyze this issue.

thank you !

4

Hi NV，

Additional information：
Orin paltform，R35.3.1

5

Hi,
Please apply the setting for a try:
Jetson/L4T/TRT Customized Example - eLinux.org

If the issue is still there, please share the test app and steps. So that we can set up and try to replicate the issue on developer kit.

7

Hi DaneLLL,

How can I set the default value of CUDA_DEVICE_MAX_CONNECTIONS to 32 in the following method without setting environment variables?

Long delays when submitting several cudaMemcpy

Please try to increase the computing channel

$ export CUDA_DEVICE_MAX_CONNECTIONS=32

A document can be found here:

8

Hi,
This is the only aaproach. If you don’t want to have a global environment variable, you can configure it along with executing the CUDA application:

$ CUDA_DEVICE_MAX_CONNECTIONS=32 ./CUDA_application_binary
9

Hi DaneLLL,

Can the same be achieved by setting some configuration parameters?

10

Hi DaneLLL,

We executed $ export CUDA_DEVICE_MAX_CONNECTIONS=32,
and the problem still reappeared,
log:

kern_0725-1200.log (32.7 MB)

11

Hi,
We would need to reproduce it on Orin developer kit and check. Please try developer kit with Jetpack 6.0GA and see if the issue is present. If yes, please share the test sample and steps. We will set up developer kit to test.

12

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.