[vtk-developers] Fwd: Bugs in the MultiThreader.

[Berk Geveci]

Luis,

I am forwarding this to the VTK developers list. Someone should
probably create a bug tracker entry for this. Thanks for the report.

-berk


---------- Forwarded message ----------
From: Luis Ibanez <luis.ibanez at kitware.com>
Date: Tue, Jan 20, 2009 at 9:27 AM
Subject: Bugs in the MultiThreader.
To: Berk Geveci <berk.geveci at kitware.com>
Cc: Bill Hoffman <bill.hoffman at kitware.com>, Brad King
<brad.king at kitware.com>, Hans Johnson <hans-johnson at uiowa.edu>,
Karthik Krishnan <karthik.krishnan at kitware.com>



Berk, Bill, Brad,


During the NAMIC week, Hans Johnson identified a bug in the ITK
Multi-Threader, that we fixed later in the week with Hans' and
Karthik's help.


Since the ITK Multi-Threader was mostly a copy-pasted version
of the VTK Multi-Threader, we review the VTK code as well, and
it seems to have the same bug.

The problem relates to the verification of assumptions regarding

   * Global default number of threads
   * Global maximum number of threads

The ITK fix can be seen at:

http://www.itk.org/cgi-bin/viewcvs.cgi/Code/Common/itkMultiThreader.cxx?root=Insight&r1=1.44&r2=1.45
http://www.itk.org/cgi-bin/viewcvs.cgi/Code/Common/itkMultiThreader.h?root=Insight&r1=1.28&r2=1.29


You may want to review the behavior of the vtkMultiThreader
accordingly.


One way to notice the effect is to use the updated test of
the Multi-Threader in ITK:

http://www.itk.org/cgi-bin/viewcvs.cgi/Testing/Code/Common/itkMultiThreaderTest.cxx?root=Insight&r1=1.2&r2=1.3

where we added the cases that expose the bug.


If you prefer I can simply log a bug entry in MANTIS.

Just let me know,


 Thanks



  Luis