[VTUNE] sep5.service failed: 236/SMACK_PROCESS_LABEL

Bram_S_ · ‎04-27-2023

After installing intel-oneapi-vtune (2023.1.0-44286) on Ubuntu 23.04:

$ sudo systemctl status sep5
× sep5.service - systemd script to load sep5 driver at boot time
     Loaded: loaded (/lib/systemd/system/sep5.service; enabled; preset: enabled)
     Active: failed (Result: exit-code) since Thu 2023-04-27 12:50:19 PDT; 6min ago
    Process: 261391 ExecStart=/usr/local/sbin/sep5-load.sh start (code=exited, status=236/SMACK_PROCESS_LABEL)
   Main PID: 261391 (code=exited, status=236/SMACK_PROCESS_LABEL)
        CPU: 62ms

Apr 27 12:50:19 deca systemd[1]: Starting sep5.service - systemd script to load sep5 driver at boot time...
Apr 27 12:50:19 deca sep5-load.sh[261391]: Loading the sep5 driver:
Apr 27 12:50:19 deca groupmod[261445]: group changed in /etc/group (group vtune/1002)
Apr 27 12:50:19 deca sep5-load.sh[261392]: The sep5 driver is already loaded!  Use ./rmmod-sep to unload it.
Apr 27 12:50:19 deca systemd[1]: sep5.service: Main process exited, code=exited, status=236/SMACK_PROCESS_LABEL
Apr 27 12:50:19 deca systemd[1]: sep5.service: Failed with result 'exit-code'.
Apr 27 12:50:19 deca systemd[1]: Failed to start sep5.service - systemd script to load sep5 driver at boot time.
$

OS: Ubuntu 23.04

CPU: i12600k

Bram_S_ · ‎04-28-2023

Rebooting gave me a working sep5 service.

But the vtune-self-test.sh crashes my kernel in 50% or so of its runs.

From my system log:

Apr 28 13:08:51 deca kernel: BUG: scheduling while atomic: sep/25740/0x00000002

So there is something wrong in the sep5 module.

The loading of the sep5 module seems to be fine:

Apr 28 13:11:02 deca kernel: vtsspp: PERFMONv5: fixed events: 3, generic counters: 6
Apr 28 13:11:02 deca kernel: vtsspp: Kernel: KPTI detected
Apr 28 13:11:02 deca kernel: vtsspp: Kernel: KASLR detected
Apr 28 13:11:02 deca kernel: vtsspp: Driver has been loaded
Apr 28 13:11:02 deca sep5-load.sh[1123]: Checking for socwatch driver ... not detected.
Apr 28 13:11:02 deca sep5-load.sh[2413]: Executing: insmod ./socwatch/drivers/socwatch2_15-x32_64-6.2.0-20-genericsmp.ko
Apr 28 13:11:02 deca kernel: socwatch2_15: -----------------------------------------
Apr 28 13:11:02 deca kernel: socwatch2_15: OK: LOADED SoC Watch Driver
Apr 28 13:11:02 deca kernel: socwatch2_15: -----------------------------------------
Apr 28 13:11:03 deca sep5-load.sh[2413]: Setting group ownership of device file to group "vtune" ... done.
Apr 28 13:11:03 deca sep5-load.sh[2413]: Setting file permissions of device file to "660" ... done.
Apr 28 13:11:03 deca sep5-load.sh[2413]: The socwatch2_15-x32_64-6.2.0-20-genericsmp driver has been successfully loaded.
Apr 28 13:11:03 deca sep5-load.sh[2413]: NOTE:
Apr 28 13:11:03 deca sep5-load.sh[2413]: The driver is accessible only to users under the group 'vtune'.
Apr 28 13:11:03 deca sep5-load.sh[2413]: Please add the users to the group 'vtune' to use the tool.
Apr 28 13:11:03 deca sep5-load.sh[2413]: To change driver access group, reinstall the driver using -g <desired_group> option.
Apr 28 13:11:03 deca sep5-load.sh[1123]: NOTE:
Apr 28 13:11:03 deca sep5-load.sh[1123]: The driver is accessible only to users under the group vtune.
Apr 28 13:11:03 deca sep5-load.sh[1123]: Please add the users to the group vtune to use the tool.
Apr 28 13:11:03 deca sep5-load.sh[1123]: To change driver access group, reload the driver using -g <desired_group> option.
Apr 28 13:11:03 deca systemd[1]: Finished sep5.service - systemd script to load sep5 driver at boot time.

OS: Ubuntu 23.04

CPU: Intel Alderlake 12600k

GPU: Intel AlderLake-S GT1

vtune: 2023.1.0-44286

AlekhyaV_Intel · ‎05-02-2023

Hi,

Thank you for posting in Intel Communities. We tried reproducing your issue. Please find our observations below:

Issue 1: Failed sep5 service

We got the same error i.e. ERROR: sep5 driver "sep5-x32_64-6.2.0-20-genericsmp.ko" while building drivers. So, we tried to set the driver access group ownership to determine which set of users can perform the collection on the system. You may set your own group during installation in the Advanced options or change it manually after installation by executing: ./boot-script -–group <your_group> from the <install-dir>/sepdk/src directory. I have tried to set the permissions to root with the below command:

./boot-script --group root

Issue 2: vtune-self-test.sh crashes my kernel in 50% or so of its runs

We tried running vtune-self-checker.sh script from <vtune-installed-directory>/bin64 folder, we didn't observe any crash on kernel. It worked fine and we got the expected results as follows:

./vtune-self-checker.sh

Here's my kernel version: 6.2.0-20-generic

OS: Ubuntu 23.04

CPU: Intel Alder Lake-P

Could you please share the crash log and kernel version so that we can debug further?

Regards,

Alekhya

Bram_S_ · ‎05-02-2023

There is no way to get the logs from vtune, because my machine hard freezes, and I can only get out of that with powercycle or hard-reset.

The kernel does capture the kernel crash in the logs, fortunately.


May 02 09:19:31 deca kernel: vtsspp: Kernel: KPTI disabled
May 02 09:19:31 deca kernel: vtsspp: Transport per-process size: 8192 pages
May 02 09:19:31 deca kernel: vtsspp: PEBS: record size: 0xc8, mask: 0xf000000ff, counters: 12
May 02 09:19:31 deca kernel: vtsspp: PEBS: record size: 0xc8, mask: 0x70000003f, counters: 9
May 02 09:19:31 deca kernel: vtsspp: PMU: uploading 18 core events
May 02 09:19:31 deca kernel: vtsspp: PMU: uploading 4 atom events
May 02 09:19:31 deca kernel: vtsspp: Registered NMI handler
May 02 09:19:31 deca kernel: vtsspp: Registered 'sched_process_fork' tracepoint
May 02 09:19:31 deca kernel: vtsspp: Registered 'sched_process_exec' tracepoint
May 02 09:19:31 deca kernel: vtsspp: Registered 'sched_process_exit' tracepoint
May 02 09:19:31 deca kernel: vtsspp: Registered 'sched_switch' tracepoint
May 02 09:19:31 deca kernel: vtsspp: Registered 'vma_store' tracepoint
May 02 09:19:31 deca kernel: vtsspp: Trace flags: 0x872e
May 02 09:19:31 deca kernel: vtsspp: Active CPU mask: 0-15
May 02 09:19:31 deca kernel: vtsspp: Time source: TSC
May 02 09:19:31 deca kernel: vtsspp: Collection started
May 02 09:19:31 deca kernel: vtsspp: Attached to 'matrix' (pid: 23619)
May 02 09:19:35 deca kernel: vtsspp: Collection stopped
May 02 09:19:35 deca kernel: vtsspp: Memory used: 34MB
May 02 09:19:35 deca kernel: vtsspp: Collection duration: 4 sec
May 02 09:19:35 deca kernel: vtsspp: Transport shutdown: 31 msec
May 02 09:19:40 deca kernel: vtsspp: Kernel: KPTI disabled
May 02 09:19:40 deca kernel: vtsspp: Transport per-process size: 8192 pages
May 02 09:19:40 deca kernel: vtsspp: PEBS: record size: 0xc8, mask: 0xf000000ff, counters: 12
May 02 09:19:40 deca kernel: vtsspp: PEBS: record size: 0xc8, mask: 0x70000003f, counters: 9
May 02 09:19:40 deca kernel: vtsspp: PMU: uploading 6 core events
May 02 09:19:40 deca kernel: vtsspp: PMU: uploading 4 atom events
May 02 09:19:40 deca kernel: vtsspp: Registered NMI handler
May 02 09:19:40 deca kernel: vtsspp: Registered 'sched_process_fork' tracepoint
May 02 09:19:40 deca kernel: vtsspp: Registered 'sched_process_exec' tracepoint
May 02 09:19:40 deca kernel: vtsspp: Registered 'sched_process_exit' tracepoint
May 02 09:19:40 deca kernel: vtsspp: Registered 'sched_switch' tracepoint
May 02 09:19:40 deca kernel: vtsspp: Registered 'vma_store' tracepoint
May 02 09:19:40 deca kernel: vtsspp: Trace flags: 0x832f
May 02 09:19:40 deca kernel: vtsspp: Active CPU mask: 0-15
May 02 09:19:40 deca kernel: vtsspp: Time source: TSC
May 02 09:19:40 deca kernel: vtsspp: Collection started
May 02 09:19:40 deca kernel: vtsspp: Attached to 'matrix' (pid: 23747)
May 02 09:19:43 deca kernel: vtsspp: Collection stopped
May 02 09:19:43 deca kernel: vtsspp: Memory used: 34MB
May 02 09:19:43 deca kernel: vtsspp: Collection duration: 4 sec
May 02 09:19:43 deca kernel: vtsspp: Transport shutdown: 27 msec
May 02 09:19:50 deca kernel: BUG: scheduling while atomic: matrix.dpcpp/23977/0x00000002
May 02 09:19:50 deca kernel: Modules linked in: tls nvme_fabrics socwatch2_15(OE) vtsspp(OE) sep5(OE) socperf3(OE) snd_seq_dummy snd_hrtimer pax(OE) binfmt_misc snd_hda_codec_realtek snd_hda_codec_generic snd_sof_pci_intel_tg>
May 02 09:19:50 deca kernel:  polyval_clmulni videobuf2_vmalloc videobuf2_memops polyval_generic cec ghash_clmulni_intel snd_rawmidi videobuf2_v4l2 sha512_ssse3 rc_core videodev snd_seq aesni_intel drm_kms_helper crypto_simd >
May 02 09:19:50 deca kernel: CPU: 4 PID: 23977 Comm: matrix.dpcpp Tainted: G           OE      6.2.0-20-generic #20-Ubuntu
May 02 09:19:50 deca kernel: Hardware name: ASUS System Product Name/PRIME Z690M-PLUS D4, BIOS 2212 12/13/2022
May 02 09:19:50 deca kernel: Call Trace:
May 02 09:19:50 deca kernel:  <TASK>
May 02 09:19:50 deca kernel:  dump_stack_lvl+0x48/0x70
May 02 09:19:50 deca kernel:  dump_stack+0x10/0x20
May 02 09:19:50 deca kernel:  __schedule_bug+0x64/0x80
May 02 09:19:50 deca kernel:  __schedule+0x4f6/0x610
May 02 09:19:50 deca kernel:  schedule+0x63/0x110
May 02 09:19:50 deca kernel:  schedule_preempt_disabled+0x15/0x30
May 02 09:19:50 deca kernel:  rwsem_down_read_slowpath+0x284/0x4d0
May 02 09:19:50 deca kernel:  down_read+0x48/0xc0
May 02 09:19:50 deca kernel:  UTILITY_down_read_mm+0x12/0x20 [sep5]
May 02 09:19:50 deca kernel:  linuxos_Exec_Unmap_Notify+0xbd/0x180 [sep5]
May 02 09:19:50 deca kernel:  ? __vm_munmap+0x1/0x1a0
May 02 09:19:50 deca kernel:  kprobe_ftrace_handler+0x113/0x1e0
May 02 09:19:50 deca kernel:  ? __vm_munmap+0x5/0x1a0
May 02 09:19:50 deca kernel:  0xffffffffc04de0f7
May 02 09:19:50 deca kernel: RIP: 0010:__vm_munmap+0x1/0x1a0
May 02 09:19:50 deca kernel: Code: c9 31 f6 31 ff 45 31 c0 45 31 c9 c3 cc cc cc cc 66 0f 1f 84 00 00 00 00 00 90 90 90 90 90 90 90 90 90 90 90 90 90 90 90 90 e8 <ab> 44 71 24 55 48 89 e5 41 57 41 56 41 89 d6 41 55 4c 8d 6d 80>
May 02 09:19:50 deca kernel: RSP: 0018:ffff9b00477f7ea8 EFLAGS: 00000202 ORIG_RAX: 0000000000000000
May 02 09:19:50 deca kernel: RAX: ffffffff9bdc9d30 RBX: ffff9b00477f7f58 RCX: 0000000000000000
May 02 09:19:50 deca kernel: RDX: 0000000000000001 RSI: 0000000000001000 RDI: 00007f553559c000
May 02 09:19:50 deca kernel: RBP: ffff9b00477f7eb0 R08: 0000000000000000 R09: 0000000000000000
May 02 09:19:50 deca kernel: R10: 0000000000000000 R11: 0000000000000000 R12: 0000000000000000
May 02 09:19:50 deca kernel: R13: 0000000000000000 R14: 0000000000000000 R15: 0000000000000000
May 02 09:19:50 deca kernel:  ? __pfx___x64_sys_munmap+0x10/0x10
May 02 09:19:50 deca kernel:  ? __vm_munmap+0x5/0x1a0
May 02 09:19:50 deca kernel:  ? __x64_sys_munmap+0x1b/0x30
May 02 09:19:50 deca kernel:  ? __vm_munmap+0x5/0x1a0
May 02 09:19:50 deca kernel:  ? __x64_sys_munmap+0x1b/0x30
May 02 09:19:50 deca kernel:  do_syscall_64+0x58/0x90
May 02 09:19:50 deca kernel:  ? do_syscall_64+0x67/0x90
May 02 09:19:50 deca kernel:  ? __task_pid_nr_ns+0x6c/0xc0
May 02 09:19:50 deca kernel:  ? __task_pid_nr_ns+0x6c/0xc0
May 02 09:19:50 deca kernel:  ? exit_to_user_mode_prepare+0x30/0xb0
May 02 09:19:50 deca kernel:  ? syscall_exit_to_user_mode+0x29/0x50
May 02 09:19:50 deca kernel:  ? do_syscall_64+0x67/0x90
May 02 09:19:50 deca kernel:  ? do_syscall_64+0x67/0x90
May 02 09:19:50 deca kernel:  entry_SYSCALL_64_after_hwframe+0x72/0xdc
May 02 09:19:50 deca kernel: RIP: 0033:0x7f5539915a8b
May 02 09:19:50 deca kernel: Code: 8b 15 91 03 0e 00 f7 d8 64 89 02 48 c7 c0 ff ff ff ff c3 66 2e 0f 1f 84 00 00 00 00 00 66 90 ff 25 7a e5 a9 04 00 00 00 0f 05 <48> 3d 01 f0 ff ff 73 01 c3 48 8b 0d 5d 03 0e 00 f7 d8 64 89 01>
May 02 09:19:50 deca kernel: RSP: 002b:00007ffecf3ec8b8 EFLAGS: 00000206 ORIG_RAX: 000000000000000b
May 02 09:19:50 deca kernel: RAX: ffffffffffffffda RBX: 00000000024f10b0 RCX: 00007f5539915a8b
May 02 09:19:50 deca kernel: RDX: 00007f553559b000 RSI: 0000000000001000 RDI: 00007f553559c000
May 02 09:19:50 deca kernel: RBP: 00007ffecf3ec950 R08: 00007f55399f6c88 R09: 00007f553559b000
May 02 09:19:50 deca kernel: R10: 0000000000000000 R11: 0000000000000206 R12: 00007ffecf3eca10
May 02 09:19:50 deca kernel: R13: 00000000024f0ff0 R14: 00000000024c1e20 R15: 0000000000001000
May 02 09:19:50 deca kernel:  </TASK>

lfz1 · ‎03-08-2024

I met the similar problem, my initial problem is:

$ sudo systemctl status sep5
* sep5.service - systemd script to load sep5 driver at boot time
Loaded: loaded (/lib/systemd/system/sep5.service; enabled; vendor preset: enabled)
Active: failed (Result: exit-code) since Mon 2024-03-04 22:58:15 CST; 4 days ago
Main PID: 18825 (code=exited, status=236/SMACK_PROCESS_LABEL)

Mar 04 22:58:15 usr1 systemd[1]: Starting systemd script to load sep5 driver at boot time...
Mar 04 22:58:15 usr1 sep5-load.sh[18825]: Loading the sep5 driver:
Mar 04 22:58:15 usr1 groupmod[18884]: group changed in /etc/group (group vtune/1005)
Mar 04 22:58:15 usr1 sep5-load.sh[18826]: The sep5 driver is already loaded! Use ./rmmod-sep to unload it.
Mar 04 22:58:15 usr1 systemd[1]: sep5.service: Main process exited, code=exited, status=236/SMACK_PROCESS_LABEL
Mar 04 22:58:15 usr1 systemd[1]: sep5.service: Failed with result 'exit-code'.
Mar 04 22:58:15 usr1 systemd[1]: Failed to start systemd script to load sep5 driver at boot time.

Then I tried $./boot-script --group root, it said:

Configuring sep5 boot script with the following options:
driver files = /opt/intel/oneapi/vtune/2024.0/sepdk/src/.
driver group = root
Creating systemd load script /usr/local/sbin/sep5-load.sh ...
done.
Creating systemd boot config file /lib/systemd/system/sep5.service ...
done.
Configuring autoload of sep5.service service ...
Job for sep5.service failed because the control process exited with error code.
See "systemctl status sep5.service" and "journalctl -xe" for details.
WARNING: systemctl start returned error 1 ...
done.

As instructed, the result shows:

* sep5.service - systemd script to load sep5 driver at boot time
Loaded: loaded (/lib/systemd/system/sep5.service; enabled; vendor preset: enabled)
Active: failed (Result: exit-code) since Sat 2024-03-09 10:47:09 CST; 36s ago
Process: 11290 ExecStart=/usr/local/sbin/sep5-load.sh start (code=exited, status=236/SMACK_PROCESS_LABEL)
Main PID: 11290 (code=exited, status=236/SMACK_PROCESS_LABEL)

-- Defined-By: systemd
-- Support: http://www.ubuntu.com/support
--
-- An ExecStart= process belonging to unit sep5.service has exited.
--
-- The process' exit code is 'exited' and its exit status is 236.
Mar 09 10:47:09 tank-PR2715P-MS-S212 systemd[1]: sep5.service: Failed with result 'exit-code'.
-- Subject: Unit failed
-- Defined-By: systemd
-- Support: http://www.ubuntu.com/support
--
-- The unit sep5.service has entered the 'failed' state with result 'exit-code'.
Mar 09 10:47:09 tank-PR2715P-MS-S212 systemd[1]: Failed to start systemd script to load sep5 driver at boot time.
-- Subject: A start job for unit sep5.service has failed
-- Defined-By: systemd
-- Support: http://www.ubuntu.com/support
--
-- A start job for unit sep5.service has finished with a failure.
--
-- The job identifier is 976233 and the job result is failed.

OS: Ubuntu 20.04

CPU: Intel(R) Xeon(R) Silver 4215 CPU @ 2.50GHz

vtune: 2024.0.1.46

lfz1 · ‎03-08-2024

"initial"hhhhh, I means original problem

Bram_S_ · ‎05-02-2023

NOTE: From that log, I see that matrix.dpcpp is the trigger of the kernel crash:

May 02 09:19:50 deca kernel: BUG: scheduling while atomic: matrix.dpcpp/23977/0x00000002

AlekhyaV_Intel · ‎05-03-2023

Hi,

Thank you for sharing your log files. As your system's operating system and kernel is not in the lists of supported operating systems and supported kernels, it may/may not work.

Could you please try the same in any supported operating systems(https://www.intel.com/content/www/us/en/developer/articles/system-requirements/vtune-profiler-system-requirements.html) and check if the issue still persists?

Regards,

Alekhya

AlekhyaV_Intel · ‎05-14-2023

Hi,

We have not heard back from you. Could you please give us an update regarding this issue?

Regards,

Alekhya

AlekhyaV_Intel · ‎05-17-2023

Hi,

We have not heard back from you, so we will close this thread. If you need any further assistance, please post a new question as this thread will no longer be monitored by Intel.

Regards,

Alekhya

IonutNechita · ‎03-07-2024

Hi,

This issue is reproduced on Ubuntu 22.04 LTS.

No have error in dmesg, but service sep5.service is failed with: code=exited, status=236/SMACK_PROCESS_LABEL)

Kernel version: Linux 5.15.0-1050-intel-iotg #56-Ubuntu SMP Tue Feb 20 14:19:47 UTC 2024 x86_64 x86_64 x86_64 GNU/Linux

Vtune version: intel-oneapi-vtune 2024.0.1-11

sudo journalctl -b --no-pager --no-hostname | grep "sep5"
mar 07 12:16:04 kernel: sep5_43: Driver loading... sym_lookup_func_addr=ffffffff9139b4d0
mar 07 12:16:04 kernel: sep5_43: [load] [UTILITY_Driver_Log_Init@1193]: Initialized driver log using contiguous physical memory.
mar 07 12:16:04 kernel: sep5_43: [load] [lwpmu_Load@7469]: Major number is 505
mar 07 12:16:04 kernel: sep5_43: [load] [lwpmu_Load@7477]: Detected 16 total CPUs and 16 active CPUs.
mar 07 12:16:04 kernel: sep5_43: [warning] [lwpmudrv_Detect_PMT_Endpoints@6589]: Address of PMT function is invalid
mar 07 12:16:04 kernel: sep5_43: [load] [lwpmu_Load@7868]: PMU collection driver v5.43.14  has been loaded.
mar 07 12:16:04 kernel: sep5_43: [load] [lwpmu_Load@7878]: NMI will be used for handling PMU interrupts.
mar 07 12:16:04 kernel: sep5_43: [load] [PMU_LIST_Initialize@652]: PMU check enabled! F6.M9a.S3 index=47 drv_type=PUBLIC arch_pmu_info_used=no
mar 07 12:16:04 kernel: sep5_43: [load] [PMU_LIST_Build_PCI_List@729]: No PCI list information detected!
mar 07 12:16:07 systemd[1]: Starting systemd script to load sep5 driver at boot time...
mar 07 12:16:07 sep5-load.sh[11314]: Loading the sep5 driver:
mar 07 12:16:07 sep5-load.sh[11315]: The sep5 driver is already loaded!  Use ./rmmod-sep to unload it.
mar 07 12:16:07 systemd[1]: sep5.service: Main process exited, code=exited, status=236/SMACK_PROCESS_LABEL
mar 07 12:16:07 systemd[1]: sep5.service: Failed with result 'exit-code'.
mar 07 12:16:07 systemd[1]: Failed to start systemd script to load sep5 driver at boot time.
mar 07 12:16:43 sudo[11970]: inechita : TTY=pts/0 ; PWD=/home/inechita ; USER=root ; COMMAND=/usr/bin/systemctl restart sep5.service
mar 07 12:16:43 systemd[1]: Starting systemd script to load sep5 driver at boot time...
mar 07 12:16:43 sep5-load.sh[11974]: Loading the sep5 driver:
mar 07 12:16:43 sep5-load.sh[11975]: The sep5 driver is already loaded!  Use ./rmmod-sep to unload it.
mar 07 12:16:43 systemd[1]: sep5.service: Main process exited, code=exited, status=236/SMACK_PROCESS_LABEL
mar 07 12:16:43 systemd[1]: sep5.service: Failed with result 'exit-code'.
mar 07 12:16:43 systemd[1]: Failed to start systemd script to load sep5 driver at boot time.
mar 07 12:16:48 sudo[12074]: inechita : TTY=pts/0 ; PWD=/home/inechita ; USER=root ; COMMAND=/usr/bin/systemctl status sep5.service

yuzhang3_intel · ‎03-07-2024

Can you run /opt/intel/oneapi/vtune/latest/sepdk/src/insmod-sep -q and post the output?

IonutNechita · ‎03-07-2024

Hi @yuzhang3_intel ,

I have this:

 /opt/intel/oneapi/vtune/latest/sepdk/src/insmod-sep -q
pax driver is loaded and owned by group "vtune" with file permissions "660".
socperf3 driver is loaded and owned by group "vtune" with file permissions "660".
sep5 driver is loaded and owned by group "vtune" with file permissions "660".

socwatch2_15 driver is loaded and owned by group "vtune" with file permissions "660".

vtsspp driver is loaded and owned by group "vtune" with file permissions "660".

yuzhang3_intel · ‎03-07-2024

It looks vtune drivers are ready. Please run self-check script:

$ /opt/intel/oneapi/vtune/latest/bin64/vtune-self-checker.sh

IonutNechita · ‎03-07-2024

Hi @yuzhang3_intel ,

I have this:

/opt/intel/oneapi/vtune/latest/bin64/vtune-self-checker.sh
Intel(R) VTune(TM) Profiler Self Check Utility
Copyright (C) 2009 Intel Corporation. All rights reserved.
Build Number: 627177

HW event-based analysis (counting mode) (Intel driver)
Example of analysis types: Performance Snapshot
    Collection: Ok
    Finalization: Ok...
    Report: Ok

Instrumentation based analysis check
Example of analysis types: Hotspots and Threading with user-mode sampling
    Collection: Ok
    Finalization: Ok...
    Report: Ok

HW event-based analysis check (Intel driver)
Example of analysis types: Hotspots with HW event-based sampling, HPC Performance Characterization, etc.
    Collection: Ok
vtune: Warning: To enable hardware event-based sampling, VTune Profiler has disabled the NMI watchdog timer. The watchdog timer will be re-enabled after collection completes.
    Finalization: Ok...
vtune: Warning: Cannot locate debugging information for the Linux kernel. Source-level analysis will not be possible. Function-level analysis will be limited to kernel symbol tables. See the Enabling Linux Kernel Analysis topic in the product online help for instructions.

    Report: Ok

HW event-based analysis check (Intel driver)
Example of analysis types: Microarchitecture Exploration
    Collection: Ok
vtune: Warning: To enable hardware event-based sampling, VTune Profiler has disabled the NMI watchdog timer. The watchdog timer will be re-enabled after collection completes.
    Finalization: Ok...
vtune: Warning: The result contains a lot of raw data. Finalization may take a long time to complete.

vtune: Warning: Cannot locate debugging information for the Linux kernel. Source-level analysis will not be possible. Function-level analysis will be limited to kernel symbol tables. See the Enabling Linux Kernel Analysis topic in the product online help for instructions.

    Report: Ok

HW event-based analysis with uncore events (Intel driver)
Example of analysis types: Memory Access
    Collection: Ok
vtune: Warning: To enable hardware event-based sampling, VTune Profiler has disabled the NMI watchdog timer. The watchdog timer will be re-enabled after collection completes.
    Finalization: Ok...
vtune: Warning: The result contains a lot of raw data. Finalization may take a long time to complete.

vtune: Warning: Cannot locate debugging information for the Linux kernel. Source-level analysis will not be possible. Function-level analysis will be limited to kernel symbol tables. See the Enabling Linux Kernel Analysis topic in the product online help for instructions.

    Report: Ok

HW event-based analysis with stacks (Intel driver)
Example of analysis types: Hotspots with HW event-based sampling and call stacks
    Collection: Ok
vtune: Warning: To enable hardware event-based sampling, VTune Profiler has disabled the NMI watchdog timer. The watchdog timer will be re-enabled after collection completes.
    Finalization: Ok...
vtune: Warning: Cannot locate debugging information for the Linux kernel. Source-level analysis will not be possible. Function-level analysis will be limited to kernel symbol tables. See the Enabling Linux Kernel Analysis topic in the product online help for instructions.

    Report: Ok

HW event-based analysis with context switches (Intel driver)
Example of analysis types: Threading with HW event-based sampling
    Collection: Ok
vtune: Warning: To enable hardware event-based sampling, VTune Profiler has disabled the NMI watchdog timer. The watchdog timer will be re-enabled after collection completes.
    Finalization: Ok...
vtune: Warning: Cannot locate debugging information for the Linux kernel. Source-level analysis will not be possible. Function-level analysis will be limited to kernel symbol tables. See the Enabling Linux Kernel Analysis topic in the product online help for instructions.

    Report: Ok

Checking DPC++ application as prerequisite for GPU analyses: Ok

GPU HW event-based analysis with runtime tracing
Example of analysis types: GPU Compute/Media Hotspots (characterization mode)
    Collection: Fail
vtune: Error: Cannot collect GPU hardware metrics because neither libigdmd.so nor libmd.so was found. Make sure you have installed Metrics Discovery API library from https://github.com/intel/metrics-discovery. See Error Message: Cannot Collect GPU Hardware Metrics help topic for more details.

GPU software event-based analysis with runtime tracing
Example of analysis types: GPU Compute/Media Hotspots (source analysis mode)
    Collection: Fail
vtune: Error: [Instrumentation Engine]: GTPin:  IGC-GTPin interface version 4142833765 is not supported
vtune: Warning: [Instrumentation Engine]: GTPin:  GTPin didn't find any kernels... Exiting without doing anything.

The check observed a product failure on your system.
Review errors in the output above to fix a problem or contact Intel technical support.

The system is ready for the following analyses:
* Performance Snapshot
* Hotspots and Threading with user-mode sampling
* Hotspots with HW event-based sampling, HPC Performance Characterization, etc.
* Microarchitecture Exploration
* Memory Access
* Hotspots with HW event-based sampling and call stacks
* Threading with HW event-based sampling

The following analyses have failed on the system:
* GPU Compute/Media Hotspots (characterization mode)
* GPU Compute/Media Hotspots (source analysis mode)

Log location: /tmp/vtune-tmp-root/self-checker-2024.03.07_15.38.56/log.txt

IonutNechita · ‎03-07-2024

My platform is: ADL-P, 12th Gen Intel(R) Core(TM) i7-1260P.

yuzhang3_intel · ‎03-07-2024

Please get code from below link, build & install.

https://github.com/intel/metrics-discovery

IonutNechita · ‎03-07-2024

Hi @yuzhang3_intel ,

I installed debian package for metrics-discovery project and now i have this:

 /opt/intel/oneapi/vtune/latest/bin64/vtune-self-checker.sh
Intel(R) VTune(TM) Profiler Self Check Utility
Copyright (C) 2009 Intel Corporation. All rights reserved.
Build Number: 627177

HW event-based analysis (counting mode) (Intel driver)
Example of analysis types: Performance Snapshot
    Collection: Ok
    Finalization: Ok...
    Report: Ok

Instrumentation based analysis check
Example of analysis types: Hotspots and Threading with user-mode sampling
    Collection: Ok
    Finalization: Ok...
    Report: Ok

HW event-based analysis check (Intel driver)
Example of analysis types: Hotspots with HW event-based sampling, HPC Performance Characterization, etc.
    Collection: Ok
vtune: Warning: To enable hardware event-based sampling, VTune Profiler has disabled the NMI watchdog timer. The watchdog timer will be re-enabled after collection completes.
    Finalization: Ok...
vtune: Warning: Cannot locate debugging information for the Linux kernel. Source-level analysis will not be possible. Function-level analysis will be limited to kernel symbol tables. See the Enabling Linux Kernel Analysis topic in the product online help for instructions.

    Report: Ok

HW event-based analysis check (Intel driver)
Example of analysis types: Microarchitecture Exploration
    Collection: Ok
vtune: Warning: To enable hardware event-based sampling, VTune Profiler has disabled the NMI watchdog timer. The watchdog timer will be re-enabled after collection completes.
    Finalization: Ok...
vtune: Warning: Cannot locate debugging information for the Linux kernel. Source-level analysis will not be possible. Function-level analysis will be limited to kernel symbol tables. See the Enabling Linux Kernel Analysis topic in the product online help for instructions.

    Report: Ok

HW event-based analysis with uncore events (Intel driver)
Example of analysis types: Memory Access
    Collection: Ok
vtune: Warning: To enable hardware event-based sampling, VTune Profiler has disabled the NMI watchdog timer. The watchdog timer will be re-enabled after collection completes.
    Finalization: Ok...
vtune: Warning: The result contains a lot of raw data. Finalization may take a long time to complete.

vtune: Warning: Cannot locate debugging information for the Linux kernel. Source-level analysis will not be possible. Function-level analysis will be limited to kernel symbol tables. See the Enabling Linux Kernel Analysis topic in the product online help for instructions.

    Report: Ok

HW event-based analysis with stacks (Intel driver)
Example of analysis types: Hotspots with HW event-based sampling and call stacks
    Collection: Ok
vtune: Warning: To enable hardware event-based sampling, VTune Profiler has disabled the NMI watchdog timer. The watchdog timer will be re-enabled after collection completes.
    Finalization: Ok...
vtune: Warning: Cannot locate debugging information for the Linux kernel. Source-level analysis will not be possible. Function-level analysis will be limited to kernel symbol tables. See the Enabling Linux Kernel Analysis topic in the product online help for instructions.

    Report: Ok

HW event-based analysis with context switches (Intel driver)
Example of analysis types: Threading with HW event-based sampling
    Collection: Ok
vtune: Warning: To enable hardware event-based sampling, VTune Profiler has disabled the NMI watchdog timer. The watchdog timer will be re-enabled after collection completes.
    Finalization: Ok...
vtune: Warning: Cannot locate debugging information for the Linux kernel. Source-level analysis will not be possible. Function-level analysis will be limited to kernel symbol tables. See the Enabling Linux Kernel Analysis topic in the product online help for instructions.

    Report: Ok

Checking DPC++ application as prerequisite for GPU analyses: Ok

GPU HW event-based analysis with runtime tracing
Example of analysis types: GPU Compute/Media Hotspots (characterization mode)
    Collection: Ok
vtune: Warning: To enable hardware event-based sampling, VTune Profiler has disabled the NMI watchdog timer. The watchdog timer will be re-enabled after collection completes.
    Finalization: Ok...
vtune: Warning: Cannot locate debugging information for the Linux kernel. Source-level analysis will not be possible. Function-level analysis will be limited to kernel symbol tables. See the Enabling Linux Kernel Analysis topic in the product online help for instructions.

    Report: Ok

GPU software event-based analysis with runtime tracing
Example of analysis types: GPU Compute/Media Hotspots (source analysis mode)
    Collection: Fail
vtune: Error: [Instrumentation Engine]: GTPin:  IGC-GTPin interface version 4112247088 is not supported
vtune: Warning: [Instrumentation Engine]: GTPin:  GTPin didn't find any kernels... Exiting without doing anything.

The check observed a product failure on your system.
Review errors in the output above to fix a problem or contact Intel technical support.

The system is ready for the following analyses:
* Performance Snapshot
* Hotspots and Threading with user-mode sampling
* Hotspots with HW event-based sampling, HPC Performance Characterization, etc.
* Microarchitecture Exploration
* Memory Access
* Hotspots with HW event-based sampling and call stacks
* Threading with HW event-based sampling
* GPU HW event-based analysis with runtime tracing

The following analyses have failed on the system:
* GPU Compute/Media Hotspots (source analysis mode)

Log location: /tmp/vtune-tmp-root/self-checker-2024.03.07_16.32.23/log.txt

yuzhang3_intel · ‎03-07-2024

It looks there is issue related to GPU driver. Did you install GPU driver before? If not, please install driver following steps in the below link.

https://dgpu-docs.intel.com/driver/client/overview.html

IonutNechita · ‎03-07-2024

Hi @yuzhang3_intel ,

What debian package from this page: https://dgpu-docs.intel.com/driver/client/overview.html is necessary to install for have "GPU Compute/Media Hotspots (source analysis mode)"?

yuzhang3_intel · ‎03-07-2024

Follow the steps, 2.1.2 and 2.1.3 first.