1. 07 Sep, 2017 1 commit
  2. 03 Jan, 2017 1 commit
  3. 28 Oct, 2016 1 commit
  4. 22 Jul, 2016 1 commit
  5. 20 Jul, 2016 1 commit
  6. 31 Mar, 2016 1 commit
  7. 02 Mar, 2016 1 commit
  8. 28 Oct, 2015 1 commit
    • loring's avatar
      depth sort update · eaf68626
      loring authored
      this patch modernizes VTK's depth sorting code. The following improvements
      were made:
      
      * transfrom GetCell to GetCellPoints. Building the cells one by one
        is expensive and we only need points to determine the depth. We
        can also then access the points in place.
      
      * transform qsort to std::sort. Comparisons in std::sort get inlined
        so std::sort is much faster. I also reduced the overhead of swaps
        by using a functor so that only cell id is swapped.
      
      * use templates to deal with point types instead of going through
        virtual GetValue API that converts them to double.
      
      * restructure depth computations so that they can be vectorized by
        the compiler.
      
      * I added a Cxx test to improve test coverage of the various sorting
        modes supported by this class.
      
      In tests the improved class is 2 x faster when sorting by first point
      and cell bounds, and 1.5 x faster when sorting by parametric center.
      The tests made use of a 900M cell iso-surface computed from a cosmology
      simulation, and gcc 4.9.2 with -DNDEBUG -Ofast -march=native -mavx
      
      
      git-svn-id: http://visit.ilight.com/svn/visit/trunk/src@27619 18c085ea-50e0-402c-830e-de6fd14e8384
      eaf68626
  9. 16 Oct, 2015 1 commit
    • loring's avatar
      ordered compositing · fb3550f9
      loring authored
      add alpha blend compositing to visit and use it to implement
      ordered compositing. there is some logic to enable it automatically
      and a gui option to explicitly disable it. it is on by default.
      I also added an option for enabling depth peeling. this is off by
      default because it's very slow. I changed all GetCell methods to
      return long long so that overflow is less of an issue. I fixed a bug
      in the transparency actor that was causing transparent geometry
      to be sorted twice.
      
      
      git-svn-id: http://visit.ilight.com/svn/visit/trunk/src@27514 18c085ea-50e0-402c-830e-de6fd14e8384
      fb3550f9
  10. 05 Oct, 2015 1 commit
    • loring's avatar
      address two issues in visit's osmesa extension manager · 8b890274
      loring authored
      first issue:
      vtk's opengl extension manager (base class for osmesa
      extension manager) had changed in vtk 6.1. we were
      overriding all the base class methods using a copy of
      an out of date version of vtk class code. as a result
      base class was not getting initialized and vtk's
      open gl feature detection, driver, and gl identification
      were not working.
      
      To address this I got rid of the overrides that didn't
      change vtk behavior and instead use the base class
      implementation in those cases. this makes sure the
      base class member variables get initialized and feature
      detection etc works again.
      
      second issue:
      I also let visit enable the blend_function_separate
      extension. in vtk this is the most efficient way
      to ensure correct alpha values during transparent
      rendering and is required for vtk's depth peeling
      algorithm. In a number of plots in visit that need
      a specific blend function I had to set the blend
      function explicitly and save and restore the related
      gl blending state.
      
      note:
      plots that make use of GL_SRC_ALPHA,GL_ONE_MINUS_SRC_ALPHA
      mode for alpha blending don't produce correct alpha
      values. I didn't change that here.
      
      
      git-svn-id: http://visit.ilight.com/svn/visit/trunk/src@27366 18c085ea-50e0-402c-830e-de6fd14e8384
      8b890274
  11. 04 Oct, 2015 1 commit
  12. 02 Oct, 2015 2 commits
  13. 03 Aug, 2015 1 commit
  14. 29 May, 2015 1 commit
  15. 12 Feb, 2015 1 commit
  16. 11 Feb, 2015 1 commit
  17. 08 Jan, 2015 1 commit
  18. 03 Dec, 2014 1 commit
    • brugger's avatar
      · bb97930a
      brugger authored
      I enhanced the Multi-Pass CSG discretization method to only process 
      duplicate boundaries once. This results in better quality discretizations,
      a reduction in the amount of memory used, and an improvement in processing
      speed.
      
      
      
      git-svn-id: http://visit.ilight.com/svn/visit/trunk/src@25150 18c085ea-50e0-402c-830e-de6fd14e8384
      bb97930a
  19. 24 Nov, 2014 1 commit
    • brugger's avatar
      · e2641ec9
      brugger authored
      I modified the multipass csg discretization so that it can be specified
      if all the regions should be discretized at once or individually.
      Currently, it is hardcode to do them individually, but I will add a 
      user control for VisIt 2.9. 
      
      
      
      git-svn-id: http://visit.ilight.com/svn/visit/trunk/src@25093 18c085ea-50e0-402c-830e-de6fd14e8384
      e2641ec9
  20. 21 Nov, 2014 1 commit
    • brugger's avatar
      · a44388e7
      brugger authored
      I modified VisIt so that it no longer crashes discretizing CSG meshes 
      with the multi pass method when the number of boundaries in a region 
      exceeds an internal limit. It now issues a warning message and tries 
      the uniform discretization method. This resolves #2067.
      
      
      
      git-svn-id: http://visit.ilight.com/svn/visit/trunk/src@25083 18c085ea-50e0-402c-830e-de6fd14e8384
      a44388e7
  21. 02 Oct, 2014 1 commit
  22. 26 Sep, 2014 1 commit
  23. 10 Sep, 2014 1 commit
  24. 09 Sep, 2014 1 commit
    • brugger's avatar
      · 46c4096e
      brugger authored
      I corrected a bug with the multipass CSG discretization. It created
      incorrect results in regions where one or more boundaries was referenced
      multiple times and the total number of boundaries in the mesh was greater
      than 512. This resolves #1975.
      
      
      
      git-svn-id: http://visit.ilight.com/svn/visit/trunk/src@24397 18c085ea-50e0-402c-830e-de6fd14e8384
      46c4096e
  25. 10 Aug, 2014 1 commit
  26. 04 Aug, 2014 1 commit
  27. 28 Jul, 2014 1 commit
    • camp's avatar
      visit_vtk/full/vtkRectilinearGridFacelistFilter.C · a1c9da78
      camp authored
      visit_vtk/full/vtkRectilinearGridFacelistFilter.h
        Tried to optimize the code.
        Fixed an issue with Threading,
        there was two static variables in the ConsolidateFacesWithoutGhostZones function.
        I change them to static const members of the class.
      
      avt/Expressions/General/avtPerMaterialValueExpression.C
        Tried to optimize the code.
        Fixed an issue with Threading,
        the function modified a class member varable. I create a local variable.
        Fixed a memory leak if an exception happened.
      
      avt/IVP/avtPICSFilter.C
        Change the names of the parallel algorithms to match the GUI.
      
      avt/Database/Database/avtDatabase.C
        Tried to optimize the code.
        Fixed an issue with Threading,
        the metadata and sil lists are shared by the threads and need to be protected.
      
      common/misc/RuntimeSetting.C
        Added the threads option as a commandline variable to track.
      
      
      
      git-svn-id: http://visit.ilight.com/svn/visit/trunk/src@23895 18c085ea-50e0-402c-830e-de6fd14e8384
      a1c9da78
  28. 26 Jul, 2014 1 commit
  29. 25 Jul, 2014 1 commit
  30. 17 Jul, 2014 1 commit
  31. 16 Jul, 2014 1 commit
    • brugger's avatar
      · 03924acd
      brugger authored
      1) I corrected a cast that was added to remove a warning that was causing 
         the compiler to calculate a normalized normal differently, which was
         causing a few test suite failures. 
      
      2) I modified the setting of the *_RELEASE flags that gets passed to cmake
         when running the regression suite on edge to add back the "-O3" that Mark 
         just removed. He changed it as a result of some bad information I gave him.
      
      
      
      git-svn-id: http://visit.ilight.com/svn/visit/trunk/src@23787 18c085ea-50e0-402c-830e-de6fd14e8384
      03924acd
  32. 10 Jul, 2014 2 commits
  33. 01 Jul, 2014 1 commit
  34. 30 Jun, 2014 1 commit
  35. 29 Jun, 2014 1 commit
  36. 02 May, 2014 1 commit
  37. 25 Apr, 2014 1 commit
    • brugger's avatar
      · 45fa6023
      brugger authored
      I modified the multi-pass discretization of CSG meshes to process each
      domain independently if the number total number of boundary surfaces is
      above the internal limit. This will make if possible to handle larger
      CSG meshes. I also added the class vtkCSGFixedLengthBitField in order
      to make it easier to change the number of bits used by the multi-pass
      discretization. I added a description of the change to the release notes.
      This resolves #1793.
      
      
      
      git-svn-id: http://visit.ilight.com/svn/visit/trunk/src@23226 18c085ea-50e0-402c-830e-de6fd14e8384
      45fa6023
  38. 10 Mar, 2014 1 commit