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.

Bug 134320 - Sometimes profiler shows wrong timings
Summary: Sometimes profiler shows wrong timings
Status: RESOLVED INVALID
Alias: None
Product: profiler
Classification: Unclassified
Component: Base (show other bugs)
Version: 6.x
Hardware: All All
: P3 blocker (vote)
Assignee: Tomas Hurka
URL:
Keywords:
Depends on:
Blocks:
 
Reported: 2008-05-01 10:14 UTC by tkellerer
Modified: 2009-09-18 11:44 UTC (History)
0 users

See Also:
Issue Type: DEFECT
Exception Reporter:


Attachments
Sample snapshot with invalid timings (220.21 KB, application/octet-stream)
2008-05-01 10:15 UTC, tkellerer
Details

Note You need to log in before you can comment on or make changes to this bug.
Description tkellerer 2008-05-01 10:14:09 UTC
Sometimes (unfortunately I cannot reliably reproduce this) the profiler shows ridiculous high execution times, e.g.
510185365 ms which makes the timings not really usable.

Switching between JDK 1.5 and JDK 1.6 does not change this.

Once the Profiler "goes" into this mode, it cannot get it to work properly again unless I start with a fresh userdir.

This has happened with 6.0 also sometimes.
Comment 1 tkellerer 2008-05-01 10:15:14 UTC
Created attachment 60913 [details]
Sample snapshot with invalid timings
Comment 2 Jiri Sedlacek 2008-06-16 16:10:24 UTC
Do you have dynamic CPU frequency switching (PowerNow!/SpeedStep) enabled on your system? The problem typically occurs
when the calibration data are incorrect. You should try to re-calibrate the profiler (Profile | Advanced Commands | Run
Profiler Calibration) and definitely disable dynamic CPU freq. sw. during profiling if enabled.

If the re-calibration doesn't help, could you please provide a sample code which reproduces the problem? Also please
attach the IDE logfile, it contains important data about your system (OS/JDK version etc.).
Comment 3 tkellerer 2008-06-16 16:13:05 UTC
Yes I do have a CPU with dynamic freq. but for the tests I have disabled the switching and the CPUs were running at full
speed. After disabling the dynamic frequency I re-run the calibration. But I still get these strange figures. 
The next time it occurs I will attach the IDE log file.
Comment 4 Tomas Hurka 2008-06-16 16:25:18 UTC
Can you also attach code for that long method? Is it always the same one?
Comment 5 J Bachorik 2008-08-28 10:37:17 UTC
Would it be possible to get the code for reproducible test case? Without that it's rather impossible to find the root
cause. Thanks.
Comment 6 matusdekanek 2009-09-18 11:44:16 UTC
no response for more than a year - closing invalid. If encountered again, please reopen.