Memory usage/ memory leak

[Abhishek Ghuwalewala]

The implementation of your Remover method is already built into the graph. The graph will dispose all its sub-objects such as plots, cursors, axes, and annotations when the graph is disposed, so you should not have to do this manually. It is essential to remove any event handlers when disposing objects. After removing your event handlers there should not be any memory being used by the graph.

[pbrowntwi]

So even if I explicitly call dispose on the graph object I still need to track every event that has been bound during it's lifetime and unsubscribe them ?

This seems like something the graph should do itself (especially if it's events that have been bound in the designer)

[pbrowntwi]

This attachement shows the memory creep being experienced (pressing Creator + Once + Destroy should give a baseline for running multiple times, however if I then press 200 times my memory goes from 26Mb upto 35Mb

and the highlights of my stackdump are


03d2efd4      201        24120 NationalInstruments.UI.Internal.b3
03d2e8b8      201        24120 NationalInstruments.UI.Internal.do
03d2ed5c      201        24924 NationalInstruments.UI.Internal.m
049fd108      201        25728 NationalInstruments.UI.Internal.b2
03c1ccfc      201        26532 NationalInstruments.UI.WindowsForms.Legend
03d2e36c      201        28944 NationalInstruments.UI.Internal.r
7a74d9c8      247        29640 System.ComponentModel.ReflectPropertyDescriptor
03c1742c      201        31356 NationalInstruments.UI.WindowsForms.WaveformGraph
049fcd94      201        32160 NationalInstruments.Restricted.LegendManager
7912c444        1        40016 System.Double[]
03d2d314      402        51456 NationalInstruments.UI.Internal.aq
03d2b160      402        51456 NationalInstruments.UI.Internal.cz
03d2ae20      402        51456 NationalInstruments.UI.Internal.ag
03d24dac      201        57888 NationalInstruments.Restricted.WaveformGraphManager
790fa3e0      932       214284 System.String
79124228      513       306324 System.Object[]
0018abf0       19       483648      Free
791242ec     2867      1069416 System.Collections.Hashtable+bucket[]

The finalizer is commented out but was being called for each instance.

Memory will eventually come back but it takes time (almost as if there's a timer holding onto it) and then it'll drop back to ~32Mb which is still a 6Mb increase (measured using Process Explorer).

Can someone tell me what I'm doing wrong ?

Thanks

Regards Paul

[Abhishek Ghuwalewala]

We have looked at the project you have provided. I am going to file a bug report for the issue. I will let you know when the problem has been investigated and fixed.

[pbrowntwi]

Is there a viewable list of currently know bugs/issues anywhere ? We are nearing shipping and it would be good to know of any other issues that are likely to bite us, annotations were the previous problem to hit us and now this one...

Thanks

Paul

[Abhishek Ghuwalewala]

Yes. The list of known issues at the time of release are documented in the readme. You can find the readme through your Start menu

Start >> All Programs >> National Instruments >> Measurement Studio 8.0.1 for [Visual Studio .NET 2003] | [Visual Studio 2005] >> Readme

[pbrowntwi]

I was more interested in issues that have been found since release.

Paul

[Abhishek Ghuwalewala]

The problem has been investigated and fixed. The fix will be available in the next release of Measurement Studio after 8.1. Thank you very much for reporting the issue and providing us with code to reproduce it.

[pbrown]

What is the timescale for this ? As I said we are about to ship a product and it has much more serious consequences in that than in the test application.

[bilalD]

Hello

We're going to release a patch for this issue. The patch will be available in a week or two.

Sorry for the inconvienience.