Hello,

I am usint VTune 2013 update 3 to profile some java application.
Unfortunately, everytime i do, just when the profiled application is going to finish VTunes reports a Segmentation fault in libabstract.so
This does not happen when i debug one of the sample C++ programs. But it deos happen even if i try to profile just "java -version".
I'm using java version "1.7.0_06"
Java(TM) SE Runtime Environment (build 1.7.0_06-b24)
Java HotSpot(TM) 64-Bit Server VM (build 23.2-b09, mixed mode)

Does anyone knows what could be the problem?

Hi, all

There is a problem with Vtune troubled me a few days as specific as bellows. I will be very grateful if someone can help me to solve this problem.

The tool version is Intel Vtune amplifier xe 2013  update2 evaluation version (noncommercial license), which was download from intel official web site.The purpose is to use Vtune to measure the required bandwidth of my application. My CPU is Intel Core i7 2600, whose architeture is sandy bridge. My  OS is redhat enterprise linux 6.2. The Vtune mode used is GUI.

The problem is: when I choose analysis type, I choose "sandy bridge / ivy bridge" -> "General Exploration" and ticked the "Analyze memory bandwidth". After finished anaylsis, it displayed "Analysis is completed" and "Finalization completed successfully". But in the "Summay", id displayed  "Elapsed time: 0.000s", and the viewpoint button is only one option to choose, which is "Task Time", and not my wanted "bandwidth". as the Vtune Help document says, there would be a "bandwidth" option.

Then I shoosed the "bandwidth" type, get the same result. 

I want to know, if  my evaluation version or my CPU cannot support harware event-based sampling collection, or there is something must be set but I igored.   

Is there someone knows, thanks!

Hi,

I am interested in computing for my application a metric that I would call "fetch ratio": The total amount of data manipulated by the program compared to the total amount of data fetched from main memory (It is crucial that the total amount of data fetched also includes prefetched data).

I'm running my experiments on a Sandy Bridge i5-2410M running Linux x64 3.7

My problem is that I can't find the good events for my needs. 
I ran into this article: http://software.intel.com/en-us/articles/detecting-memory-bandwidth-saturation-in-threaded-applications

But it seems that none of the events mentioned in this article are available on Sandy Bridge. I can't even find any mention of them in the Intel Manuel Volume 3B that list all available events for quite a lot of processors.

In the comments of the article mentionned above, someone asked about bandwidth measurement on Sandy Bridge. Romain Dementiev said that Intel Performance Counter Monitor can do it.

I looked at the code and it doesn't seem to be implemented for Sandy Bridge: http://pastebin.com/qw52cfeb

What event should I use? or at least, what is recommended by Intel for bandwidth measurement on Sandy Bridge? (even system-wide bandwidth measurement would be acceptable, I would just have to keep my machine dead quiet which is not a problem).

Thanks,

Moncef.

Hi,

Could someone tell me if it is possible to use VTune with MPI applications on the Xeon Phi and, if so, explain how to do it.

The suggested way with VTune and MPI is to use the comand-line tool:

http://software.intel.com/sites/products/Whitepaper/Clustertools/amplxe_...

but the command line tool does not run directly on our Xeon Phi, and the Xeon Phi specific information about VTune that I've found e.g.:

http://software.intel.com/en-us/articles/optimization-and-performance-tu...

does not mention MPI.

Apologies if there is already documentation on how to do this.

Many thanks

Rupert

Dear administrator,

The Intel VTune Amplifier XE and VTune Performance Analyzer are really powerful tools to analyze and improve code, the hotspot detector and memory leakage checker help me greatly. I know that they can check problems for my whole program, at most time, however, I wish to check only the hotspot or memory leakage of some subroutine, can Intel VTune Amplifier XE and VTune Performance Analyzer do so? For example,

program test

...

call funA(a1, a2)

!!!!!!!!!!!!!!!!!!!!!!!!!!

$ begin VTune check

call funB(b1, b2)

$ end VTune check

!!!!!!!!!!!!!!!!!!!!!!!!!!

...

end program

Thanks,

Zhanghong Tang

Hi,

After installing vtune_aplifier_xe_2013, I tried to build driver.

cd /opt/intel/vtune_amplifier_xe_2013/powerdk/src

./build_driver

In file included from /opt/intel/vtune_amplifier_xe_2013/powerdk/src/src/apwr_driver.c:109:
include/trace/events/workqueue.h:37: warning: âstruct cpu_workqueue_structâ declared inside parameter list
include/trace/events/workqueue.h:37: warning: its scope is only this definition or declaration, which is probably not what you want
include/trace/events/workqueue.h:37: warning: âstruct cpu_workqueue_structâ declared inside parameter list
include/trace/events/workqueue.h:37: warning: âstruct cpu_workqueue_structâ declared inside parameter list
include/trace/events/workqueue.h:37: warning: âstruct cpu_workqueue_structâ declared inside parameter list
/opt/intel/vtune_amplifier_xe_2013/powerdk/src/src/apwr_driver.c: In function âregister_non_pausable_probesâ:
/opt/intel/vtune_amplifier_xe_2013/powerdk/src/src/apwr_driver.c:4954: error: âprobe_cpu_idleâ undeclared (first use in this function)
/opt/intel/vtune_amplifier_xe_2013/powerdk/src/src/apwr_driver.c:4954: error: (Each undeclared identifier is reported only once
/opt/intel/vtune_amplifier_xe_2013/powerdk/src/src/apwr_driver.c:4954: error: for each function it appears in.)
/opt/intel/vtune_amplifier_xe_2013/powerdk/src/src/apwr_driver.c: In function âunregister_non_pausable_probesâ:
/opt/intel/vtune_amplifier_xe_2013/powerdk/src/src/apwr_driver.c:5046: error: âprobe_cpu_idleâ undeclared (first use in this function)
/opt/intel/vtune_amplifier_xe_2013/powerdk/src/src/apwr_driver.c: In function âpw_alrm_suspend_notifier_callback_iâ:
/opt/intel/vtune_amplifier_xe_2013/powerdk/src/src/apwr_driver.c:6631: warning: ISO C90 forbids mixed declarations and code
make[2]: *** [/opt/intel/vtune_amplifier_xe_2013/powerdk/src/src/apwr_driver.o] Error 1
make[1]: *** [_module_/opt/intel/vtune_amplifier_xe_2013/powerdk/src] Error 2
make[1]: Leaving directory `/usr/src/kernels/2.6.39-200.24.1.el6uek.x86_64'

It complains probe_cpu_idle is not defined.  In apwr_driver.c, it seems probe_cpu_idle is defined:

#if LINUX_VERSION_CODE  < KERNEL_VERSION(2,6,38)
#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,35)
static void probe_power_start(unsigned int type, unsigned int state)
#elif LINUX_VERSION_CODE < KERNEL_VERSION(2,6,36)
static void probe_power_start(void *ignore, unsigned int type, unsigned int state)
#else // 2.6.36 <= version < 2.6.38
static void probe_power_start(void *ignore, unsigned int type, unsigned int state, unsigned int cpu_id)
#endif
{
    // tps_i(type, state);
    DO_PER_CPU_OVERHEAD_FUNC(tps, type, state);
};
#else // version >= 2.6.38
static void probe_cpu_idle(void *ignore, unsigned int state, unsigned int cpu_id)
{
   if (state == PWR_EVENT_EXIT) {
       return;
   }

    // tps_i(type, state);
   DO_PER_CPU_OVERHEAD_FUNC(tps, 0 /*type*/, state);
};
#endif // version
#endif // APWR_RED_HAT

I am using OEL(based on RedHat) kernel 2.6.39-200.24.1.el6uek.x86_64. gcc version 4.4.6 20120305 (Red Hat 4.4.6-4) (GCC)

Any suggestions how I should fix this?

Regards,

Jenny

hi !

plz help me out with this issue.!when i ran a very simple C++ program for hotspot analysis.it succesfully collected ,analysed and finalized the result

still sometimes its shows functions with some CPU values and sometimes show " NO DATA TO SHOW / NO DATA IS AVAILABLE".

please tell me how to  get the data when this "NO DATA IS AVAILABLE" situation arises.

hi !

plz help me out with this issue.!when i ran a very simple C++ program for hotspot analysis.it succesfully collected ,analysed and finalized the result

still sometimes its shows functions with some CPU values and sometimes show " NO DATA TO SHOW / NO DATA IS AVAILABLE".

please tell me how to  get the data when this "NO DATA IS AVAILABLE" situation arises.

Currently i am working with Vtune. I am trying to calculate CPU% utilization give the "CPU seconds used" and the elapsed time.

Is this write :  CPU% = cpu time(secs) / elapsed time. ?

e.g. in 2 core machine

cpu sec =5.6 , elapsed time = 8.4

cpu% = 5.6 / 8.4 .

Also how to map cpu time and utilization in multicore environment.

Thanks,

rupinder

I installed vtune_amplifier xe 2013 update3 in my HP Proliant DL380 G7 server with Xeon X5680 3.33 (2 sockets, each with 6 core CPU)

with RedHat (2.6.18-164.11.1.el5 #1 SMP) kernel

The installation was completed successfully except it warned that "The drivers for Power analysis require a Linux kernel version of 2.6.32 or higher and will be disabled".  Since I don't care about power analysis, as I'm going to turn all CPU to max power/performance model anyway, I ignored that (since my production environment won't use the newer kernel anyway)

However, when I tried run test with "collect stacks" option checked, I got the following:

"Advanced collection of call stacks and thread context switches cannot be enabled.  Make sure the driver is installed and you have permissions to access it"

I'm running the 64bit verison of amplxe-gui using root.  So, permissioning should not be an issue.  I don't know what driver is supposed to be installed and hence I can't check it.  Can anyone shred someone on how to check that?

Thank you very much in advance.

hello everyone !

while doing hotspot analysis ,concurrency,locks and waits  etc . I am getting "failed to finalize the result ,the result you are opening is empty" error ....

while for some other analysis like lightweight hotspot,memory acces its working fine ,giving results.

what may be the plausible reason behind it and how to resolve it ...

please help !

Hi,

I would like to obtain the following basic information using vune

1. No of times a function is called
2. The function call graph (also if possible, the detailed call graph with basic block info)
3. No of times a statement is executed

I cannot find a way to do the same. When I looked on this forum, there are some methods given for older versions. But I cannot find the same for this version. Any help is appreciated.

hello friends !

I am stucked with a serious issue .Due to some development issues .I need to generate reports for all analysis like hotpsot,memory access etc in linux version of vtune ,then i brought these files to windows 7, and trying to analyse and view these results in vtune for windows.

For that I created new project  in vtune for  windows and in the new project copied these files  like r000hs,r001hs,r002lw which were generated in linux. and to see the GUI form of results when I clicked on them I am getting this error "Failed to finalize the result ,cannot finalize the result, Error 0x40000019 (Incompatible database schema) --Database schema is not upgradable ".

Please help to resolve or if there is any other alternative .please do tell. ( generating results in vtune -linux and viewing them in vtune -windows 7)

thanks

I downloaded VTune Amplifier XE, got the registration code, registered it and when I try to install it I got the following message:

In Order to complete the offline activation process, you will need to use a system that is connected to the internet.

1. On the connected system, go to:

https://registration.intel.com/RegCenter/remoteactivation.aspx

2. Enter this produce activation code on the above web page:

XXXX-XXXX-XXXX-XXXX-XXXX-XXXX-XXX

3. Save....

Well, going to the page above and entering the product activation code results in a message that says that for evaluation version I have to go to another page, where, when I go, I can't find any link suggesting anything related to getting the "product unlock code"

Please note that on the target machine I cannot possibly have internet connection so offline activation is the only choice I have.

Please help,

Cristian

Hi,

I am running Amplifier 2013 update2 on SandyBridge.  OS is Linux.

From snb-general-exploration, I can see a function has a lot of OFFCORE_RESPONSE.ANY_DATA_RD.LLC_MISS_ANY.REMOTE_HITM_1.  I want to dig further and see the memory access patten and the source code associate with it.

This question was asked earlier, but I would like to confirm if there is a way to do it now.

http://software.intel.com/en-us/forums/topic/310172
http://software.intel.com/en-us/forums/topic/284531

Regards,

Jenny

Hi,
Recently I have downloaded the Intel package cpp_studio_xe_2013_update1_intel64.tar for 64 bit Linux OS and 64 bit Intel architecture. I am trying to run the tachyon example with vtune-amplifier_xe_2013 for hotspots analysis, but it is showing error “failed to finalize the result”.

Could you please help me how to resolve this issue?

This package is installed in the ubuntu Linux server and exploring this example from windows xp client.

Regards,
Srinivasu

Greetings,

I am pretty sure I have a configuration error, but I can't seem to figure it out. Hopefully someone here can help. I am running the latest of the Cluster Studio XE. I have c++/OpenMP code that I would like to trace out to find the problem areas. It is also worth noting that at the moment I can compile successfully and run other Intel tools (or the few I have tried) from any of the nodes as well as the frontend of the cluster.

On the frontend, I can open the gui, load the project, run, and view the results. If I copy the command line output from the gui, I get:
amplxe-cl -v -collect hotspots -- /home/chris/code/intel/run
Running that on the frontend from the cmd line, gives expected output. I can then open that output in the GUI and view the details.

However, the frontend is a weakling in comparison to the compute nodes. So I tried to run that command on the compute nodes:
$ export INTEL_LICENSE_FILE=28518@frontend01
$ source /software/intel/composer_xe_2013/bin/compilervars.sh intel64
$ source /software/intel/vtune_amplifier_xe/amplxe-vars.sh
$ amplxe-cl -v -collect hotspots -- /home/chris/code/intel/run
Using result path `/home/chris/tmp/r007hs'
Executing actions  8 % Clearing the database                                   
The database has been cleared, elapsed time is 0.623 seconds.
Executing actions 16 % Loading raw data to the database                        
Raw data has been loaded to the database, elapsed time is 0.009 seconds.
Executing actions 50 % Loading raw data to the database                        
Finalizing the result took 0.653 seconds.
Executing actions 50 % done                                                    
Error: Error 0x4000001e (Cannot load raw collector data)

Err.....OK. I went looking for that error and I found a few similar errors, but none that I saw apply to me (that I can tell).

I decided to try again on the login nodes (which have Gnome installed; the nodes are just cmd). It ran just fine in the GUI, but failed on the command line with the same error. I figured that it must be something in the environment variables, but checking `env` on all of the nodes yields similar results. So I decided to install all the same pacakges on the login and compute nodes (cluster is still in early testing so I can do that without angry users:P). Now the GUI fails to run the job on the login node in additon to the cmd version (I haven't figured out how to get better error output yet, but what it does say is just as minimal as the cmd output). On the compute node, I can now open the GUI version, but it too fails like the login node does now. So at the moment, the only way to run the VTune application with any success is on the frontend. This is obviously a problem for me and the cluster...

My end goal here, is to eventually allow the users to run the GUI on the login nodes and submit jobs through the Torque resource manager. However, before I get to that step I need to be able to run the job successfully from the cmd line on the node. I don't know what failed or where as that error is probably the most informative output I have found at the moment.

Can anyone please give me some pointers as to what configuration I may have messed up here?

Thank you!

Hi,
I have downloaded non-commercial Intel C++ Studio XE package to experiment tools supported in it.  To profile code, I am using VTune amplifier update 1, but it is giving error “failed to finalize result”. I want to try with VTune amplifier update 3.
Is this non- commercial downloadable package?, if so could you please provide the link to download vtune amplifier xe 2013 update 3.tar.gz  file.

Regards,
Srinivasu

Hi,

I have compiled the fortran program with different optimization flags. I have one program with default optimization flag -O2 and another compiled program with -fast optimization flag. I was able to open the program compiled with optimization flag -O2. So, when we try to load the program in Vtune, the program also executes ?

However, when I try to load the program compiled with the optimization flag -fast, Vtune throws an error "Failed to finalize the result". The result you are opening is empty. This may be caused by an error during the data collection. What does this actually mean ? I went throught other threads in this forum , but couldn't quite understand what it means.

The collector Messages I got are as follows:

Error: Binary file of the analysis target does not contain symbols required for profiling. See documentation for more details.
Error: Valid dlopen symbol is not found in the static binary of the analysis target.

I am using ubuntu 12.04 system.

Thanks so much. Please reply soon.

Jdbaba

Hi All,

I am trying to open the openmp compiled executable code into vtune. I am using Intel compiler and vtune in linux.

The error that I got while loading the file is "error while loading shared libraries: libiomp5.so : cannot open shared object file: No such file or directory"

However, if I try to load the program compiled without openmp then I can load the program on vtune. Why is it behaving so strange ? If I run the program along, then it will run.Do I need to set any environmental variable to load the openmp compiled program ?

Thank you so much.

Best Regards,

Jdbaba