- 20 Dec, 2017 4 commits
-
-
Matt Larsen authored -
b530a5ce Fix issue with Managed Memory for 0 size arrays Acked-by:
Kitware Robot <kwrobot@kitware.com>
Acked-by: Robert Maynard <robert.maynard@kitware.com> Merge-request: !1033
-
a73fc3fd Enable all point gradient worklet tests Acked-by:
Kitware Robot <kwrobot@kitware.com>
Merge-request: !1031 -
1d816534 Local var for the array portal instead of inside the for loop. Acked-by:
Kitware Robot <kwrobot@kitware.com>
Merge-request: !1029
-
- 19 Dec, 2017 6 commits
-
-
Sujin Philip authored
-
Robert Maynard authored
-
c6a4e335 Normalize worklet (both in place, and copy). Acked-by:
Kitware Robot <kwrobot@kitware.com>
Acked-by: -
Dave Pugmire authored -
Dave Pugmire authored
-
186e6869 Add ability to run tests with `mpiexec`. Acked-by:
Kitware Robot <kwrobot@kitware.com>
Acked-by:
-
- 18 Dec, 2017 1 commit
-
-
Utkarsh Ayachit authored
Extends `vtkm_unit_tests` to support running tests with `mpiexec`. Updates `UnitTestMultiBlock` to use mpi, if available.
-
- 15 Dec, 2017 11 commits
-
-
1bb7dde9 Update MultiBlock to use `diy` for block-based operations. e9c7e561 Add `EnvironmentTracker` to track MPI communicator. d52b275e Making vtkm-cont depend on diy. c63f3635 diy: pass operator instance to mpi_op<>::get() 42d5be31 diy: Add support for MPI_COMM_NULL. 6a4edd87 Exclude diy from VTK-m Copyright checks. 9b544471 Add `diy` 1f989b1f exclude `diy` from clang format. ... Acked-by:
Kitware Robot <kwrobot@kitware.com>
Acked-by: -
Utkarsh Ayachit authored
Updating MultiBlock to use `diy` for computing block summaries like ranges, bounds etc. This makes it possible to MultiBlock to work in distributed operations without explicit logic.
-
Utkarsh Ayachit authored
When using MPI, we need a mechanism in VTK-m to set/get the MPI controller. Added `EnvironmentTracker` to do that.
-
Utkarsh Ayachit authored
-
Utkarsh Ayachit authored
This makes it possible to add custom MPI reduction operations without having to add too complex logic to allocating and freeing MPI operation.
-
Utkarsh Ayachit authored
-
Utkarsh Ayachit authored
-
Utkarsh Ayachit authored
-
Utkarsh Ayachit authored
-
Utkarsh Ayachit authored
Adding option to enable MPI. When enabled, we do `find_package(MPI)` in the Base component itself.
-
Utkarsh Ayachit authored
Use `this` and rename `blocks` to `Blocks`.
-
- 14 Dec, 2017 2 commits
-
-
06c5f161 Teach brigand about more intel compiler limitations introduced in 18.X. Acked-by:
Kitware Robot <kwrobot@kitware.com>
Merge-request: !1020 -
Robert Maynard authored
-
- 12 Dec, 2017 3 commits
-
-
7420f72c Simplify the boilerplate functors in vtkm/cont/Algorithms. Acked-by:
Kitware Robot <kwrobot@kitware.com>
Merge-request: !1019 -
Robert Maynard authored
-
f9f205e9 ListCrossProduct now uses a special version for MSVC2013 c02349a8 ListCrossProduct now uses a lazy evaluation implementation 7b1b9e44 Correctly forward rvalue functors when passed to CastAndCall Acked-by:
Kitware Robot <kwrobot@kitware.com>
Merge-request: !1018
-
- 11 Dec, 2017 4 commits
-
-
bf6cb1bd removing state from functors and adding variadic templates 013113e3 Fixing streaming scan doc eaebaea7 adding a wrapper for device adapter algorithm Acked-by:
Kitware Robot <kwrobot@kitware.com>
Merge-request: !1011 -
Matt Larsen authored
-
Robert Maynard authored
The lazy version that was implemented to get the Intel compiler to work doesn't work under MSVC2013. Since MSVC2013 is our least c++11 conforming compiler we add a special MSVS2013 code path, since it will be the first compiler we drop.
-
Robert Maynard authored
The intel compiler could not generate code in a timely manner ( 12+ hours ) when asked to produce a cross product of very long lists. By moving to a lazy evaluation scheme we now have all compilers product a cross product in a reasonable amount of time ( 2-4 seconds ). This resolves Issues: - vtk/vtk-m#190 - vtk/vtk#17196
-
- 08 Dec, 2017 1 commit
-
-
Robert Maynard authored
-
- 07 Dec, 2017 3 commits
-
-
62fa1d56 Fix location of VTKmCheckPyexpander.cmake Acked-by:
Kitware Robot <kwrobot@kitware.com>
Acked-by: -
Kenneth Moreland authored
The location of VTKmCheckPyexpander.cmake was originally set to $ {CMAKE_SOURCE_DIR}/CMake. This is correct with respect to the VTK-m install, but incorrect if VTK-m is being included as a module in another project (like VTK). Change the location to ${VTKm_CMAKE_MODULE_PATH}, which should be correct in every case. -
d51ddda9 Since sizeof returns std::size_t we should iterate using unsigned integers Acked-by:
Kitware Robot <kwrobot@kitware.com>
Merge-request: !1016
-
- 06 Dec, 2017 2 commits
-
-
Robert Maynard authored
-
e28309f0 Remove VTKM_EXEC_CONSTANT Acked-by:
Kitware Robot <kwrobot@kitware.com>
Acked-by:
-
- 05 Dec, 2017 3 commits
-
-
39abcb99 fixing compile errors 655412fc adding support for log plots with the wireframer Acked-by:
Kitware Robot <kwrobot@kitware.com>
Merge-request: !1014 -
Matt Larsen authored
-
Sujin Philip authored
If a global static array is declared with VTKM_EXEC_CONSTANT and the code is compiled by nvcc (for multibackend code) then the array is only accesible on the GPU. If for some reason a worklet fails on the cuda backend and it is re-executed on any of the CPU backends, it will continue to fail. We couldn't find a simple way to declare the array once and have it available on both CPU and GPU. The approach we are using here is to declare the arrays as static inside some "Get" function which is marked as VTKM_EXEC_CONT.
-
