From cycle accounting, my program have a high IMISS rate. I'd like to find out the exact location in code where they occurs. From Intel Itanium 2 Processor Reference Manual For software Development and Optimization Table 10-2, EAR event Instruction Cache seems able to provide this information. But I can not find L1I events to do what I want. Please help. Thanks.
I am using gentoo and would like to try the vtune linux version to pack results and view them on windows platform. Is this possible?
I have an evaluation copy of VTA070WSGB00139 which runs under its own gui. I've removed VS.NET 2002 and I've yet to replace it with .NET 2003. A colleague leads me to understand that MS appear to have changed how debug symbols work in VC++ 7.1 (.NET 2003) over how they work in VC++ 6 or 7 (VS.NYET or .NET 2002).
If the debug symbols are different does this have any bearing on VTA070WSGB00139 versus VTA070WSGB00120. What precisely is the difference between VTA070WSGB00139 and VTA070WSGB00120 vis-a-vis CVF6.6, IFC, IVF, and .NET 2002 or 2003.
Are you currently using callgraph activities to analyzer your Linux code?
In the VTUNE 1.1 Overview "Fully linux based solution" was mentioned. Where'e the GUI ?
This thread is an extension of the poll posted this week. If you haven't answered it, bump up a level to the answer the poll. Then, come back if you want to discuss the following:
1) Are you currently using callgraph activities to analyzer your Linux code?
2) If no, why not?
3) If yes, do you prefer the character output that you are currently viewing? Or would you prefer a graphical viewer instead if you could get one?
All comments welcomed, team.
I have inherited a project that has speed and size constraints. It currently has I think too many methods inlined, and to make code smaller, I want to only inline functions that actually increase performance. How can I use Vtune to determine this?
I don't know if you are familiar with a product called Quexal, but it allows you to type in MMX/SSE assembly and it will analyze it for you and compute for each instruction the cycle time and make graphs to maximize parrellism, and speed.
I like how VTune looks as a profiler, but I was wondering if would also be able to do something like that? Or does intel have another product like this?
I have a DirectX application. It's a cloth simulation. The debug and release version run fine. I did this before:
-Set to Release
-Linker->Command Line, Add "/fixed:no"
-Turn "Generate Debug Info" to "Yes(/DEBUG)"
-Turn "Generate Map File" to "Yes(/MAP)"
-Turn "Map Exports" to "Yes(/MAPINFO:EXPORTS)"
It works fine with the a sample DirectX program that comes with the SDK. But with mine, it crashes right at the beginning.