1. 19 Jun, 2017 1 commit
  2. 08 Jun, 2017 1 commit
  3. 03 Apr, 2017 1 commit
  4. 31 Mar, 2017 1 commit
  5. 30 Mar, 2017 1 commit
  6. 27 Mar, 2017 3 commits
  7. 24 Mar, 2017 2 commits
  8. 23 Mar, 2017 1 commit
  9. 22 Mar, 2017 1 commit
    • brugger's avatar
      · 4211a475
      brugger authored
      I moved the SubdivideQuads operator from optional to required so that
      it will always be available to users. This resolves #2789.
      
      
      
      git-svn-id: http://visit.ilight.com/svn/visit/trunk/src@30598 18c085ea-50e0-402c-830e-de6fd14e8384
      4211a475
  10. 03 Mar, 2017 1 commit
  11. 15 Feb, 2017 2 commits
  12. 24 Jan, 2017 3 commits
  13. 23 Jan, 2017 3 commits
  14. 20 Jan, 2017 1 commit
  15. 16 Jan, 2017 1 commit
  16. 13 Jan, 2017 1 commit
  17. 03 Jan, 2017 2 commits
    • bonnell's avatar
      Update copyright date · cdd36204
      bonnell authored
      git-svn-id: http://visit.ilight.com/svn/visit/trunk/src@30108 18c085ea-50e0-402c-830e-de6fd14e8384
      cdd36204
    • gmorris2's avatar
      This change improves support for VisIt's threading (pthread mode). · 0520534c
      gmorris2 authored
      This work is from UO undergraduate students Jeremy Brennan and Alister Maguire.
      
      Changes:
      1) We added a virtual function to avtFilter to indicate whether it is thread safe, and set the default to false.  We also modified the Single-Input-Multiple-Output (SIMO) data tree iterator to only apply threading if the filter in question is thread safe.  This makes debugging a lot easier, and guarantees that only thread-safe filters will be accessed in threading mode.
      2) Fixed race conditions in avtGenericDatabase
      3) Reflect filter is now thread-safe.  This one presented a serious issue, in that reflect was creating arrays that was shared among multiple outputs, and then VTK was choking when iterating on that array in subsequent filters.  (I.e., two data sets shared an array, and then later two threads were iterating on the same array.)  We fixed this by having reflect duplicate data when in threading mode.
      4) The facelist filter is now thread-safe.  The key change was to check for cached external faces in the input file (from Silo).  Having multiple threads attempt to get this data at one time breaks Silo, so we disable retrieval of cached external faces (again only in threaded mode).  Note that the external faces returned by Silo for globe.silo are wrong, so threaded now produces different (and more correct answers).  We hope to pursue improved external facelist calculation in Silo in the future.
      5) The slice filter is now thread-safe.  This involved changing data members to be scoped at the function level.
      6) The interval tree acceleration structure is now thread-safe.  This involved adding a mutex.  (Adding a mutex is a no-op when threading is not enabled.)
      7) The onion peel & index select operators and also the distance to best fit line expression are all also now thread-safe.
      
      
      
      git-svn-id: http://visit.ilight.com/svn/visit/trunk/src@30097 18c085ea-50e0-402c-830e-de6fd14e8384
      0520534c
  18. 22 Dec, 2016 1 commit
  19. 07 Nov, 2016 1 commit
  20. 05 Nov, 2016 1 commit
  21. 24 Oct, 2016 1 commit
    • brugger's avatar
      · 9348f426
      brugger authored
      I modified the polyline to ribbon filter to clean the polydata since
      the vtkRibbonFilter can't handle duplicate points. I also corrected
      several bugs with the integral curver filter with the generation of 
      normals for displaying the curves as ribbons. This resoles #2700.
      
      
      
      git-svn-id: http://visit.ilight.com/svn/visit/trunk/src@29478 18c085ea-50e0-402c-830e-de6fd14e8384
      9348f426
  22. 19 Oct, 2016 1 commit
  23. 26 Sep, 2016 1 commit
  24. 02 Sep, 2016 1 commit
  25. 25 Aug, 2016 1 commit
  26. 18 Aug, 2016 2 commits
  27. 17 Aug, 2016 1 commit
  28. 09 Aug, 2016 1 commit
  29. 07 Jul, 2016 1 commit
  30. 23 Jun, 2016 1 commit