1. 17 Feb, 2016 1 commit
    • Ken Martin's avatar
      remove glFlush calls · e473e02b
      Ken Martin authored
      These calls should not be needed but pollute the code
      and the profiling results and potentially slow down execution
      e473e02b
  2. 29 Jan, 2016 1 commit
  3. 28 Jan, 2016 2 commits
  4. 20 Nov, 2015 1 commit
  5. 20 Oct, 2015 1 commit
    • Ken Martin's avatar
      A fix for picking combined with depth peeling · 8af6e111
      Ken Martin authored
      A fix for picking combined with depth peeling and some cleanup.
      Picking (not using a hardware picker) was not clearing the
      LastRenderingUsedDepth peelign flag and that was causing issues.
      Switched to that that flag is not sued but instead the information
      objects are used.
      
      (cherry picked from commit 7e23fd65)
      8af6e111
  6. 28 Sep, 2015 1 commit
    • Ken Martin's avatar
      A fix for picking combined with depth peeling · 7e23fd65
      Ken Martin authored
      A fix for picking combined with depth peeling and some cleanup.
      Picking (not using a hardware picker) was not clearing the
      LastRenderingUsedDepth peelign flag and that was causing issues.
      Switched to that that flag is not sued but instead the information
      objects are used.
      7e23fd65
  7. 21 Jul, 2015 1 commit
  8. 26 Jun, 2015 1 commit
  9. 24 Jun, 2015 1 commit
    • Ken Martin's avatar
      Some fixes for PV ColorOpacityTableEditing · 60df241e
      Ken Martin authored
      This patch fixes some issues realed to texture
      mapping with multiple textures. This was showing up
      when depthpeeling was being used at the same time as
      scalar coloring by texturem where the texture was
      changing often.
      60df241e
  10. 01 Jun, 2015 1 commit
    • Ken Martin's avatar
      More cleanups · f1535309
      Ken Martin authored
      vtkgl::CellBO renamed vtkOpenGLHelper
      vtkgl::substitute moved to vtkShaderProgram::Substitute
      Creation of vtkOpenGLIndexBufferObject
      Creation of vtkOpenGLVertexBufferObject
      Move methods that were in vtkglVBOHelper into new classes
      Move IndexCount from CellBO to vtkOpenGLIndexBufferObject
      etc
      f1535309
  11. 13 Dec, 2014 1 commit
    • Ken Martin's avatar
      Fix some release graphics resource issues · 5c1c4b19
      Ken Martin authored
      ParaView IceT does a WindowRemap when using tile display. That
      means everythgin has to work with a new window/context so it
      identifies a bunch of release graphics resource issues. This
      patch fixes a bunch of them. Mostly minor changes the one
      large change is to make ReleaseGraphicsResources in Renderer
      public like it is for every other class. This is the way it
      should be and the original protected method was a mistake from
      way way back.  Making it public save us from doing the
      
      SetRenderWindow(NULL)
      SetRenderWindow(this)
      
      hack that we have been using to free up resources in the
      renderer since we could nto call the protected method.
      
      Change-Id: If31406fa9425417d240311aac110c82409637163
      5c1c4b19
  12. 04 Nov, 2014 1 commit
  13. 31 Oct, 2014 1 commit
    • Ken Martin's avatar
      Add in depth peeling pass and a couple other passes · 925f5066
      Ken Martin authored
      Convert depth peeling render pass to the OpenGL2 backend.
      Also add a new class that encapsulates the basic rendering
      steps that VTK uses.
      
      Modify the renderer to use the depth peeling pass instead of
      having two copies of the code lying around. Removed the depth
      peeling code from Mapper2D as it is no longer needed now that we
      have CopyToFrameBuffer working. That is a lighter weight way to
      get a quad up on the screen. Plus that keeps the depth peeling
      code more localized as it should be. Clean up a lot of depth
      peeling code that was in the renderer and now there are cleaner
      ways of doing it (passing information keys).
      
      Change-Id: I8d0a53bb25920bd30eade49ed12126b9b44be358
      925f5066
  14. 24 Oct, 2014 1 commit
  15. 15 Oct, 2014 1 commit
    • Ken Martin's avatar
      A first cut at a CompositePolyDataMapper2 · c75f3f6f
      Ken Martin authored
      So this is a very different use case from what we have done so far.
      In many ways this is the lots of independent actors case with the
      exception that they are not fully independent. There were a number
      of CPU botlenecks hindering this classes performance:
      
        stl maps and strings
        recomputing shaders/uniforms
        GetInput/SetInput (touching the pipeline is bad)
        Default Culler
      
      There are a couple open issues with this change. I'm
      not thrilled with the CurrentInput approach in PolyDataMapper
      but SetInput and GetInput really are expensive. Second there is
      an issue if a bunch of datasets share a single
      large pointset. Right now we duplicate the pointset VBO for each
      dataset. Will need a VBO cache at some point to allow actors/etc
      to share a pointset or have the VBO building only populate the
      VBO with the points actually used by the cells (or both).
      
      This patch also contains a minor fix to the VAO missing an
      initialization for ForceEmulation.
      
      Change-Id: If2cf9531a26838b0fd8d47dafd1dcff6d4133c4b
      c75f3f6f
  16. 27 Sep, 2014 1 commit
    • Ken Martin's avatar
      Fix otherPrint and TestEmptyInput tests · f5a7a1a8
      Ken Martin authored
      There were a lot of abstract classes defined in Rendering/Core or
      Rendering/Volume that the OpenGL2 backend has not implemented yet.
      This causes a crach whenever such a class is instantiated as is
      done by some of the tests. The fix is to either remove or provide
      implementations for these classes. For the volme code this patch adds
      a VolumeOpenGL2 directory which will be where the new/converted
      volume rendering implementations will go.
      
      The other big piece is the painters. As we are not/can not support that API
      the classes were moved to OpenGL. The excpetion is the PainterDeviceAdapter
      which is part of RenderWindow. When building with OpenGL2 that class is
      treated in RenderWindow as a plain vtkObject that does nothing.
      
      vtkSmartVolumeMapper is not currenty built by OpenGL2 but could be if we
      are willing to add a bunch of ifdefs. The basic approach is there already
      in the class, just need to add a bunch more if we want to have that in the
      short term prior to the new GPU mapper being completed.
      
      This patch includes some other fixes that were required to get the tests to
      pass. Generally related to missing file properties or extra includes that
      were not needed.
      
      Change-Id: Id41db2d90526c109b1acb7442d99d8c2371cc396
      f5a7a1a8
  17. 10 Sep, 2014 1 commit
  18. 08 Sep, 2014 1 commit
  19. 05 Sep, 2014 1 commit
  20. 06 Aug, 2014 2 commits
  21. 16 Jul, 2014 2 commits
  22. 15 Jul, 2014 1 commit
  23. 26 Jun, 2014 1 commit
  24. 24 Jun, 2014 1 commit
  25. 18 Jun, 2014 1 commit
  26. 13 Jun, 2014 4 commits
    • Marcus D. Hanwell's avatar
      Very judicious use of sed, vtkOpenGL2* -> vtkOpenGL* · a6a584b1
      Marcus D. Hanwell authored
      This now gives us a vtkRenderingOpenGL2 that has most of the same
      classes as vtkRenderingOpenGL - the two will clobber each other's
      headers, symbols etc and so can never be built/linked at the same
      time.
      
      Unfortunately a lot of classes SafeDownCast or dynamic_cast to the
      derived OpenGL forms of the classes in vtkRenderingCore, and so
      using a different class name presented significant issues despite
      the fact that these classes are (in an ideal world) implementations
      of the API in vtkRenderingCore for OpenGL 2.1 and above.
      
      Change-Id: Ie8c6234621e550d3fd67da732d2cf4839290d665
      a6a584b1
    • Marcus D. Hanwell's avatar
      Rename vtkOpenGL2* -> vtkOpenGL* · b40b181d
      Marcus D. Hanwell authored
      Using pure renames to make things clearer to git...
      
      Change-Id: Icbf5f782179a5e511aee06e7a1fa494be0a5be26
      b40b181d
    • Marcus D. Hanwell's avatar
      Made vtkRenderingOpenGL2 independent of vtkRenderingOpenGL · 8bc98a02
      Marcus D. Hanwell authored
      This required several classes to be copied over, they may not stay.
      It also required me to port many calls from vtkgl to GLEW, and get
      rid of checks for features in OpenGL 2.1 or less. The extension
      manager has been removed, and its functionality replaced.
      
      Change-Id: I60f8858a315a0e91c175aef6152ab57d26650cb4
      8bc98a02
    • Ken Martin's avatar
      add depth peeling · 28e81660
      Ken Martin authored
      Change-Id: Ic3f725b5c7ef2ed7f7d639acaba5a727bb1768b8
      28e81660
  27. 11 Jun, 2014 1 commit
  28. 10 Jun, 2014 1 commit
    • Ken Martin's avatar
      some bug fixes · 7b424819
      Ken Martin authored
      Change-Id: I834e94a505fdb209dc29793931a9c68b0e676111
      7b424819
  29. 09 Jun, 2014 1 commit
  30. 06 Jun, 2014 1 commit
  31. 03 Jun, 2014 1 commit
  32. 20 May, 2014 2 commits
  33. 19 May, 2014 1 commit