Using GCC 9.2 the pre change vtkm::filter::Threshold would
take 31.84sec to compile and use 1.1GB of memory. After
these changes the filter takes 24.16sec to compile and
uses 885MB of memory.
This file was using the Matrix class but not including Matrix.h. When I
tried to use the MeshQuality filter in an example that did not include
Matrix.h on its own, I got a compile error.
If a worklet doesn't explicitly state an ExecutionSignature, VTK-m
assumes the worklet has no return value, and each ControlSignature
argument is passed to the worklet in the same order.
For example if we had this worklet:
```cxx
struct DotProduct : public vtkm::worklet::WorkletMapField
{
using ControlSignature = void(FieldIn, FieldIn, FieldOut);
using ExecutionSignature = void(_1, _2, _3);
template <typename T, vtkm::IdComponent Size>
VTKM_EXEC void operator()(const vtkm::Vec<T, Size>& v1,
const vtkm::Vec<T, Size>& v2,
T& outValue) const
{
outValue = vtkm::Dot(v1, v2);
}
};
```
It can be simplified to be:
```cxx
struct DotProduct : public vtkm::worklet::WorkletMapField
{
using ControlSignature = void(FieldIn, FieldIn, FieldOut);
template <typename T, vtkm::IdComponent Size>
VTKM_EXEC void operator()(const vtkm::Vec<T, Size>& v1,
const vtkm::Vec<T, Size>& v2,
T& outValue) const
{
outValue = vtkm::Dot(v1, v2);
}
};
973878b8b Improve the performance of the Image and Graph Connectivity algorithms
Acked-by: Kitware Robot <kwrobot@kitware.com>
Acked-by: Li-Ta Lo <ollie@lanl.gov>
Merge-request: !1873
The collection of connectivity algorithms had a couple of inefficiencies.
By moving to using WorkId we can remove a couple of arrays of the same size
as the input domain. In addition by moving to using atomics we can remove
an bool output array with a size equivalent to the input domain and
a call to reduce.
9c41daa2b change input to spacing and radius to make it more intuitive
b991d3c6b merge style differences
7d0ee16fd add the real double radius for scaling
Acked-by: Kitware Robot <kwrobot@kitware.com>
Merge-request: !1825
8520d70e0 Compile most frequently used VTK-m filters into a library
d1d61b9eb vtkm::filter::Filter passes filter policies by value
4ff021b08 Improve VTK-m compilation times by compiling more keys<T> types
Acked-by: Kitware Robot <kwrobot@kitware.com>
Merge-request: !1836
3ef234ef1 Move Oscillator to the sources directory
5c69f1bae Update Wavelet to use WorkletVisitPointsWithCells
62357dea0 Tangle source has better runtime performance
eb3d86654 Add a test explicitly for vtkm::source::Tangle
63f243809 Add vtkm::source::Wavelet as a new source
257dc1f84 Sources have a more consistent API and internal style.
2d455bbef Document what the Tangle source will generate
Acked-by: Kitware Robot <kwrobot@kitware.com>
Merge-request: !1850
f9d02b374 remove file that should not have been checked in
d59917473 rename half variable to avoid shadowing CUDA global
954e25874 fix problem with filename
91f7e763b fix copyright, warnings, cmake for Jacobian-cellmetric branch
e040ea05e formatting suggestions from KMorel
f4edb6a8e remove trailing whitespace
ccedb5aac Jacobian working, unit test working
560aa1f67 add support for outputname
...
Acked-by: Kitware Robot <kwrobot@kitware.com>
Merge-request: !1855
VTK-m now provides the following filters with the default policy
as part of the vtkm_filter library:
- CellAverage
- CleanGrid
- ClipWithField
- ClipWithImplicitFunction
- Contour
- ExternalFaces
- ExtractStructured
- PointAverage
- Threshold
- VectorMagnitude
By building these as a library we hope to provide faster compile
times for consumers of VTK-m when using common configurations.
`vtkm::source::Source` now uses virtual inheritance for
the `Execute` method. This will allow us to enforce that
all vtkm::source have a consistent calling convention.
To make the code inside filters and sources more similar the
base `vtkm::source::Source` class has a `vtkm::cont::Invoker`
as a member variable. This will allow sources to use the
`this->Invoke()` pattern that filters use to launch worklets.
627cdc99b Fix host/device flags.
7148ab342 Fix type conversion warning when `size_t` differs from `vtkm::Id`.
1d9f6b37c Add support for boundary augmentation for MPI optimization
2a1e1dcfd Parallel Contour Tree
Acked-by: Kitware Robot <kwrobot@kitware.com>
Merge-request: !1737
05b679250 Merge branch 'master' into tangle_source
a6c044df9 added changelog
0d818701c put VTKM_SOURCE_EXPORT in .cxx
e0aae7d86 remove EXPORT from .cxx
4d7de67ee use VTKM_SOURCE_EXPORT
cd49136d5 add copyright notice, add installation of header
0c094a568 used the Tangle source
aeb8877a9 add newline at EOF, minor change based review
...
Acked-by: Kitware Robot <kwrobot@kitware.com>
Merge-request: !1804
58a587a7f Merge branch 'master' of https://gitlab.kitware.com/vtk/vtk-m into particleAdvectionFloatType
a83629e95 compile error.
52fe3c403 compile warnings.
5af465efd Merge branch 'master' of https://gitlab.kitware.com/vtk/vtk-m into particleAdvectionFloatType
bd9bc91ac Merge branch 'master' of https://gitlab.kitware.com/vtk/vtk-m into particleAdvectionFloatType
c057f6b83 Merge branch 'particleAdvectionFloatType' of gitlab.kitware.com:dpugmire/vtk-m into particleAdvectionFloatType
9ca4cc2d0 type cleanup.
dff384572 Use vtkm::FloatDefault for particle advection code.
Acked-by: Kitware Robot <kwrobot@kitware.com>
Merge-request: !1827
By using the dual of the cellset we can quickly compute the GhostCells
of structured data using WorkletPointNeighborhood boundary condition
object
Using a 1024x1024x512 test grid we see the following perf:
Master Serial : 5.658144 sec
This MR Serial: 0.519684 sec
Master OpenMP : 0.532256 sec
This MR OpenMP: 0.080604 sec
