I have run a memory analysis (the widest scope memory analysis type) via Inpsector XE 2013. It includes analysing of stack acceses, to cach uninitialized arrays/variables.
Attached is a PNG figure with the results.
Note the strange behavior in Polar_HUCM.F90 module, which imply there is uninitialized memory access in line 119. However, note that line 120 is identical in terms of using the same arrays and specifically the same part of the array being controlled by "kb" loop.
We are evluating Inspector XE 2015 and have version:
Update 1 (build 379161)
When working with C++ compiled with the Visual Studio 2013 compiler inspector gives stack traces to memory or gdi leaks
with names like:
whereas we would expect to see C++ class names included:
The Intel® Parallel Studio XE 2016 Beta program is now available. In this beta test, you will have early access to Intel® Parallel Studio XE 2016 products and the opportunity to provide feedback to help make our products better. Registration is easy through the Intel® Parallel Studio XE 2016 Pre-Beta Survey.
This suite of products brings together exciting new technologies along with improvements to Intel’s existing software development tools:
i have parallel 2013. the release notes says Inspector and Advisor are included. how do i start them and is there a user guide/ thx
Is it possible to install an evaluation of Intel Inspector XE 2015 on Linux without root access? I usually just use it on Windows, and I'm just evaluating it on Linux, so I don't want to bother system administrators unless the evaluation proves successful. Can someone please email me a license file? As far as I can tell my purchased Windows license file won't work for a Linux install.
I'm trying to attach my debugger to see a data race in the act, but I can't find the context menu item "Debug This Problem" explained in the documentation "About Investigating Problems Using Interactive Debugging" ( https://software.intel.com/en-us/node/528095 ):
I am running inspector on the OmpScr benchmark. The benchmark allow you to specify a problem size for each of the programs. When I run inspector using small size problem it is able to finish the verification process most of the time but for some of the problem (i.e. c_loopB.badSolution1.par). I get the error:
Error: Internal error. Please contact Intel customer support team.
What does "Use maximum resources" option do on the data race detection analysis?
I found that there is a significant performance differcence between with and without that option. But the analysis results are not quite different.
For instance, for one program it took about 15 seconds without the option while it took about 5 minutes with the option.
Does it make skip analysis for some parts of a program? Some kind of sampling technique?
I tried to use Intel Inspector (from XE 2015) with little success. The application I try to analyze uses boost threads and a boost spirit parser. Thread safety defines have been added (eg BOOST_SPIRIT_THREADSAFE).
The following example displays the problem. In normal Release/Debug builds the program will output :
If Intel Inspector XE 2015 is used with "Thread Error Analysis" the program fails with stack corruption during the execution of boost::call_once.
- Seite 1