I can't run hardware event-based sampling analysis

[Jin_Tak_L_]

Hello. I've been tried running enable-call-counts knob by specifying the code on pbs script. 

amplxe-cl -collect-with runsa -knob enable-call-count=true -data-limit=0 knob ./raxmlHPC -m GTRCAT -n 23S.E.ALL -s 23S.E/R0/cleaned.alignment.fasta -j -p 6154307

When I run vtune. I keep getting error saying it cannot load data collected.

What am I doing wrong with my code?

[Dmitry_P_Intel1]

Hello,

If you use "-collect-with runsa" you need to point events that you want to collect e.g.:

amplxe-cl -collect-with runsa -knob event-config=CPU_CLK_UNHALTED.THREAD -knob enable-call-count=true -data-limit=0 knob ./raxmlHPC -m GTRCAT -n 23S.E.ALL -s 23S.E/R0/cleaned.alignment.fasta -j -p 6154307

Or if you want hotspot collection with call counts you can use predefined advanced-hotspots:

amplxe-cl -collect advanced-hotspots -knob collection-detail=stack-and-callcount -data-limit=0 knob ./raxmlHPC -m GTRCAT -n 23S.E.ALL -s 23S.E/R0/cleaned.alignment.fasta -j -p 6154307

Please note that call count collection might bring additional overhead.

Thanks & Regards, Dmitry

[Jin_Tak_L_]

Hello dmitry. Thanks for such quick reply. I've run with the second code you've provided but it doesn't seem to work.

I'm profiling on multiple threads with different cpu cores. The only thread seems to be working is the thread with one processor but I can only view the summary tab. The other tabs like bottom up or caller/callee don't display anything or just uploading data forever. 

I cannot even open up the other threads' output saying they haven't collected enough data.

What would be the possible problem with my case?

Is it possible that there are just too many functions calling one another and those calls cause overhead?

Thank you

[Dmitry_P_Intel1]

Hello,

Could you please specify what VTune version do you use?

Also - on summary pane - do you see reasonable Elapsed Time number and relevant other metrics?

Thanks & Regards, Dmitry