Analyzers
Talk to fellow users of Intel Analyzer tools (Intel VTune™ Profiler, Intel Advisor)
4975 Discussions

Vtune seems get wrong result profiling my program

TCJ
Beginner
3,420 Views

when I profiling my program without `-qopenmp`, it works properly. But when I use it, the elapsed Time become extremely low and wrong. However, when I use CUI to profile the same openMP program,  it works properly.

TCJ_0-1626162128853.png

the Analysis Configuration and Collection Log does not offer any warnings.

 

Labels (1)
0 Kudos
18 Replies
AlekhyaV_Intel
Moderator
3,385 Views

Hi,


Thank you for posting in Intel Forums. We need the exact steps to reproduce so that we can debug further. Please provide the details of the OpenMP project you're currently working on. Also please provide us your VTune Profiler version and system information(OS & version)


Regards,

Alekhya


0 Kudos
Bernard
Valued Contributor I
3,364 Views

"Elapsed Time" count seems to be small. You may be advised to increase the workload. The sampling mode  incurres and overhead and for very short program (less than milisecond) the results may be skewed.

What is the average workload of your program?

0 Kudos
TCJ
Beginner
3,325 Views

the duration of the program is about 480ms, but we also tested a 20seconds program. the problem repeat.

0 Kudos
Bernard
Valued Contributor I
3,303 Views

@TCJ 

 

You may try to minimize the overhead by creating a custom analysis for a single [architectural] performance events e.g. CPU_CLK_UNHALTED.THREAD. Set the "sample after value" to 2000003 and enable only counting of user mode counter overflow code triggers. Disable any stack collection option.

 

0 Kudos
AlekhyaV_Intel
Moderator
3,344 Views

Hi,


Could you please give us an update regarding this issue?


Regards,

Alekhya


0 Kudos
TCJ
Beginner
3,326 Views

sorry for being late, the notice email has been put into trash box automatically.

for the openMP project information : 

    we are optimizing it recently, and using vTune to profile. It's a image processing program which execution time has been optimized to 430ms.

    recently, we have used vTune on other project, which occur the same problem while profiling multi thread omp program.(which has a longer executing time about 20seconds )

TCJ_0-1627195762501.png

Vtune version and sys info :

    host: windows 10 (windows 11 is the same) Intel oneAPI vTune Profier 2021.5.0

    remote: ubuntu 18.04 Intel oneAPI vTune Profier 2021.5.0

Also: 

    due to the problem in remote profiling, we use Xmanager transfer remote GUI to local, and it work properly. So this problem seems only happen using remote ssh profile.

   we also notice that the remote ssh profile could not automatically trace the source code of the profiling program (used -g flag  when compiling )

 

 

 

0 Kudos
TCJ
Beginner
3,326 Views

I also tried advanced option duration time  but the problem still.

0 Kudos
Dmitry_P_Intel1
Employee
3,296 Views

Hello,

It looks like the application that is run through VTune remote flow could not load OpenMP runtime. If so - one option is to create a wrapper script on the target that sources needed vars etc. and then invokes the application.  And put this wrapper script as an application to launch for VTune remote collection.

Thanks & Regards, Dmitry

0 Kudos
TCJ
Beginner
3,288 Views

did you mean .sh script, we have used it but that seems doesn't work

0 Kudos
Dmitry_P_Intel1
Employee
3,265 Views

Hello, yes, I meant a kind of .sh script. Could you please redirect the output from the application run inside the script to a file and provide the output if it emitted by the app? It still looks like something was wrong at application startup.

Thanks & Regards, Dmitry

0 Kudos
TCJ
Beginner
3,251 Views

thank, you for your reply.

there are some reason I can't provide the output file, but I checked whether the app start up properly. 
And yes, the application not start up, but vtune launched .

 

0 Kudos
TCJ
Beginner
3,251 Views

and the amplex-python on host doesn't output anything.

It used to display properly if the profile perform correctly

0 Kudos
Dmitry_P_Intel1
Employee
3,246 Views

Acknowledging that VTune remote flow has limitation on propagating stferr from a target app to be displayed on a host. Have you been able to figure out the reason of the app startup issue and get it working through VTune remote flow?

Thanks & Regards, Dmitry

 

0 Kudos
TCJ
Beginner
3,233 Views

not yet, I am complete confuse about this ORZ.

0 Kudos
侯玉山
Novice
3,176 Views

I think,

it's because your program didn't run successfully. You can paste the execution command from VTune GUI to the command line to see whether the output prompt program runs successfully.

0 Kudos
AlekhyaV_Intel
Moderator
3,185 Views

Hi,


Could you please provide us VTune result directory & self-checker logs so that we can debug your issue further.


Regards,

Alekhya


0 Kudos
AlekhyaV_Intel
Moderator
3,122 Views

Hi,


Could you please give us an update regarding this issue? Please send us your VTune Result folder and self checker logs to work on your issue.


Regards,

Alekhya


0 Kudos
AlekhyaV_Intel
Moderator
3,058 Views

Hi,


We have not heard back from you, So we'll close this inquiry now. If you need any further assistance, please post a new question as this thread will no longer be monitored.


Regards,

Alekhya


0 Kudos
Reply