ArrayHandles in DAX have a CopyInto function which allows the user to copy an array handle's data into a compatible STL type iterator. Originally this was fairly straight forward to implement since array handles in DAX are templated on the DeviceAdapterTag. In contrast, VTKm array handles use a polymorphic ArrayHandleExecutionManager under the hood allowing a single array handle to interface with multiple devices at runtime. To achieve this virtual functions are used. This makes implementing the CopyInto function difficult since it is templated on the IteratorType and virtual functions cannot be templated.
To work around this, I've implemented a concrete templated CopyInto function in the class derived from ArrayHandleExecutionManagerBase. In the ArrayHandle class, CopyInto dynamically casts the base class into the derived class, then calls the CopyInto function defined in the derived class.
The drawback to this approach is that, should the user define their own class that inherits from ArrayHandleExectionManagerBase, they are not forced to implement the CopyInto function, unlike the other virtual functions.
5e72d3a8 Rename kernels directory to splatkernels to avoid confusion
7a2225cf Add Copyright text
d04e4dfa Remove c++11 constexpr keyword
ed5faf5b Fix for M_PI on windows, use vtkm::Pi()
fe284ffb Add unit test for splat kernel integral.
29001e37 Change GaussianSplatter to KernelSplatter to support other kernels
378cb17e Code cleanup, style, debug, unused vars
65d2980f Fix clang compile error, cleanup debug messages
...
Acked-by: Kitware Robot <kwrobot@kitware.com>
Acked-by: Robert Maynard <robert.maynard@kitware.com>
Merge-request: !193
fd685210 Always install all device headers even when device isn't enabled.
b1663b24 Add an example of using multiple backends from a single translation unit.
fc0ff69d Methods with try/catch need to be host only.
4d635d64 DeviceAdapter Tags now always exist, and contain if the device is valid.
cf32b430 Teach Configure.h to store if TBB and CUDA are enabled.
Acked-by: Kitware Robot <kwrobot@kitware.com>
Acked-by: Kenneth Moreland <kmorel@sandia.gov>
Merge-request: !198
When compiling with cuda and tbb enabled in a single translation unit you
need to make sure all try/catch blocks are marked as host only otherwise
the cuda compiler will error out.
Previously it was really hard to verify if a device adapter was valid. Since
you would have to check for the existence of the tag. Now the tag always
exists, but instead you query the traits of the DeviceAdapter to see if
it is a valid adapter.
This makes compiling with multiple backends alot easier.
Performance testing has shown that the combined lower-bounds + update-connectivity
worklet is marginally, but consistently faster than DeviceAdapterAlgorithm::LowerBounds
followed by an update-connectivity worklet. It also requires less memory.
Add ArrayHandleIndex
Add an ```ArrayHandleIndex``` class that echoes the index. This is really a special version of ```ArrayHandleCounting``` that is of type ```vtkm::Id``` and starts at 0, but it is the most common case and does not need the extra addition.
Also modified ```ArrayHandleCounting``` to support a step, which makes the use a bit more powerful.
See merge request !192
Modify ArrayHandleCounting so that it supports both a starting value and
a step (increment). This adds a multiplication, but the common case that
does not use it is already in a separate class (ArrayHandleIndex).
Template the splatter algorithm over Kernel type and use abstract kernel
interface to fetch the kernel value.
Add Gaussian, Spline3rdOrder kernel classes templated over dimension.
Other kernels can/will be added in future.
Kernel classes are defined such that the integral of the volume is unity
as is the convention in (for example) SPH simulations.
Clean up the implementation to take a uniform grid dataset as input
and splat using the correct world coordinates of the points into
the volume. Kernels are centred around the splat points in world coordinates
(floating point) and not quantized to the nearest voxel.
Skip the generation of all volume points which is unnecessary (memory usage).
Add ability to specify a radius array with one entry per splat point so
that the size of each splat may be controlled arbitrarily.
Change ExecutionWholeArray interface to match expected for ArrayPortal
When ExecutionWholeArray is passed to a worklet, it is expected to
behave like an array portal. However, it was missing the
GetNumberOfValues method and the ValueType typedef. These are now added.
See merge request !189
When ExecutionWholeArray is passed to a worklet, it is expected to
behave like an array portal. However, it was missing the
GetNumberOfValues method and the ValueType typedef. These are now added.