[vtkusers] vtkFixedPointVolumeRayCastMapper problem in java [SOLUTION]

[Granseier, Michael (VISUS)]

Hello vtk users,

I have found a solution to the problem I have posted on 17 June 2010.
(vtkusers Digest, Vol 74, Issue 24)

So everyone who has the same error may try this:

To avoid the stack overflow while using the FixedPointVolumeRayCastMapper under java
on an Windows 32-bit machine, set the size of the native method stacks on virtual machine start-up 
manually using the "-Xss" flag.

I have used -Xss1024k and it works now without throwing an exception.

For more information on this issue I refer to chapter "3.5.6 Native Method Stacks"
in http://java.sun.com/docs/books/jvms/second_edition/html/Overview.doc.html
This was very helpful for me to find the solution.

I hope, I could help,

Michael




-----Ursprüngliche Nachricht-----
Von: Granseier, Michael (VISUS) 
Gesendet: Mittwoch, 16. Juni 2010 17:03
An: 'vtkusers at vtk.org'
Betreff: vtkFixedPointVolumeRayCastMapper problem in java

Hello everybody,

after a lot of search in the mailing list I found this old post 
concerning a crash with the vtkFixedPointVolumeRayCastMapper in java.

http://old.nabble.com/vtkFixedPointVolumeRayCastMapper-Problem-in-java-td22659058.html

The last message which deals with this problem was from 7 Jun 2010.
http://old.nabble.com/SimpleRayCast-example-in-Java-tp28799209p28801962.html

Actually I am dealing with exactly the same problem.
Using another volume-mapper is definitely not an option because I have to render large datasets.
The vtkFixedPointVolumeRayCastMapper is the only mapper which is fast enough to 
interact with the volumes on a 32-bit machine.

So I just wanted to ask again, if someone has found a solution for that problem.
Or if someone has reported this as a VTK bug which will be fixed in any future VTK release.

Thanks a lot for any kind of help!


Michael