This Bugzilla instance is a read-only archive of historic NetBeans bug reports. To report a bug in NetBeans please follow the project's instructions for reporting issues.
This is similar to what I reported in http://www.netbeans.org/issues/show_bug.cgi?id=77537, but different. In this situation, the snapshot is being taken, but it appears that the reset results profiling point is being ignored. Running NB 6.0 QBuild from 2006-05-31 on JDK5u5, using the sample application and instructions that are attached to http://www.netbeans.org/issues/show_bug.cgi?id=77542, the reset results profiling point seems to not be working. In other words, on the instances when I avoid the crash that is reported in http://www.netbeans.org/issues/show_bug.cgi?id=77542, the end result I see is not correct. The snapshot gets taken at line 63 of PrimeCalculator.java, but when I open that snapshot it shows information from when the program started up. The method I'm interested in (where the reset and snapshot profiling points are located in PrimeCalculator.java) runs on its own thread and I see results from *all* threads. An additional indicator is that in the "Profiling Points" window in the lower-right hand corner of the IDE, the entry for the reset results profiling point does not have anything in its "Result" column. Shouldn't it have some sort of indication that the profiling point was reached and therefore results have been reset?
I cannot reproduce it. It either crashes or results are reset properly. please reopen if it happens with newest daily build and provide step by step instructions. I improved usability so that Reset Results point now shows the time it reset results.
Could you please check if the problem is still valid after Profiling Points reimplementation?
Well, I *guess* it's correct now, although I'm not really sure. Things certainly work *much* better than they did last summer when I opened this IZ. :-) I have attached the latest version of the project, if you want to see what things look like now. I set a Reset Results profiling point at line 45 of PrimeCalculator.java and then a Take Snapshot profiling point at line 81 of PrimeCalculator.java. I then did CPU profiling with "entire application" and "profile all classes" and "use profiling points" selected. I put in 1000000 as the value in the application and clicked the Calculate button three times. I got three snapshots, which is correct. When I open the snapshots, though, all of them include not only the thread on which the code ran, they also include the event handler thread. It doesn't seem like to me the event handler thread should be listed, but perhaps I'm not understanding something here....
Created attachment 40511 [details] sample application
I've found out that for some reason PrimeNumbers.actionPerformed() at line 215 is being called from the EventQueue during computing of the prime numbers by separate thread. In fact these two threads are both running at the same time and that's why you can see them in the snapshot. I'll close this issue as the Reset Results profiling point seems to be processed correctly. Please file a separate issue if you think that the behavior describe above is a bug (but it seems to be behaving this way even without profiling). Fixed by Profiling Points reimplementation.