http://blogs.msdn.com/b/profiler/archive/2004/08/04/208286.aspxThis is the second in my series of posts on what the heck all these numbers mean in the profiler. In instrumentation mode, we give 4 varieties of statistics for each function in the reports: - Elapsed Inclusive - Elapsed Exclusive - Applicationen-USTelligent Evolution Platform Developer Build (Build: 5.6.50428.7875)http://blogs.msdn.com/b/profiler/archive/2004/08/04/208286.aspx#3959531Thu, 19 Jul 2007 20:44:00 GMT91d46819-8472-40ad-a661-2c78acb4018c:3959531Visual Studio profiler : Difference between Elapsed and Application Time « Simple thoughts from a simple mind<p>PingBack from <a rel="nofollow" target="_new" href="http://learningdotnet.wordpress.com/2007/07/19/visual-studio-profiler-difference-between-elapsed-and-application-time/">http://learningdotnet.wordpress.com/2007/07/19/visual-studio-profiler-difference-between-elapsed-and-application-time/</a></p> <img src="http://blogs.msdn.com/aggbug.aspx?PostID=3959531" width="1" height="1">http://blogs.msdn.com/b/profiler/archive/2004/08/04/208286.aspx#210022Fri, 06 Aug 2004 18:10:00 GMT91d46819-8472-40ad-a661-2c78acb4018c:210022steve carrollYeah, I know what you mean about the magic. Great feedback. We'll definitely see what we can do about the filtering you discuss going forward.<div style="clear:both;"></div><img src="http://blogs.msdn.com/aggbug.aspx?PostID=210022" width="1" height="1">http://blogs.msdn.com/b/profiler/archive/2004/08/04/208286.aspx#209836Fri, 06 Aug 2004 13:54:00 GMT91d46819-8472-40ad-a661-2c78acb4018c:209836Dmitriy ZaslavskiyWhy not just have a list of function (customizable) time of which not to include in app time? <br>All locking/file/gdi/etc. Let us pick categories. <br>IMHO this will be very usefull. <br> <br>Thanks for being honest. I hate when products do magic. When that happens I cann't use the results. <br> <br>Thanks again<div style="clear:both;"></div><img src="http://blogs.msdn.com/aggbug.aspx?PostID=209836" width="1" height="1">http://blogs.msdn.com/b/profiler/archive/2004/08/04/208286.aspx#209265Thu, 05 Aug 2004 21:13:00 GMT91d46819-8472-40ad-a661-2c78acb4018c:209265steve carrollYeah. The saving grace here is that every call statement will have instrumentation points surrounding it, so hopefully the amount of good application time that you lose due to these false positives will be kept to a minimum. (hopefully ;-)) <br> <br>Like I said above, Application Time is a big approximation. We figured it was best to just be open and honest about that up front rather than have our most sophisticated users scratching their heads. This is definitely an area where we know we have room for improvement, but thanks for the feedback. In practice, this approximation seems to be good enough to find areas of major contention.<div style="clear:both;"></div><img src="http://blogs.msdn.com/aggbug.aspx?PostID=209265" width="1" height="1">http://blogs.msdn.com/b/profiler/archive/2004/08/04/208286.aspx#209257Thu, 05 Aug 2004 21:01:00 GMT91d46819-8472-40ad-a661-2c78acb4018c:209257Dmitriy ZaslavskiyI don't understand: <br>Beetwen Point3 &amp; Point4 there could be a lot app code and all of will be subtracted just because some other thread had a ring0 transition? <br><div style="clear:both;"></div><img src="http://blogs.msdn.com/aggbug.aspx?PostID=209257" width="1" height="1">http://blogs.msdn.com/b/profiler/archive/2004/08/04/208286.aspx#208912Thu, 05 Aug 2004 17:03:00 GMT91d46819-8472-40ad-a661-2c78acb4018c:208912Rob Caron's Blog<div style="clear:both;"></div><img src="http://blogs.msdn.com/aggbug.aspx?PostID=208912" width="1" height="1">