This warning is emitted when the compiler performs an optimization, and wants
you to verify that the presumptions it is making are valid. For this case
we can correctly promise the compiler that the values will not overflow and
the optimizations are correct.
73c05c10 Suppressions of __host__ warnings shouldn't trigger internal compile errors
5100e559 Suppress warnings about calling __host__ functions from CUDA 9.0
282c515e Suppress warnings about calling __host__ functions from CUDA 8.0
5dfdc830 Cuda will also print error/warning pragma values now.
Acked-by: Kitware Robot <kwrobot@kitware.com>
Merge-request: !1150
3211c150 add support to run test with MPI.
Acked-by: Kitware Robot <kwrobot@kitware.com>
Acked-by: Kenneth Moreland <kmorel@sandia.gov>
Merge-request: !1148
faf916f1 Fix condition on which to get the TBB version
Acked-by: Kitware Robot <kwrobot@kitware.com>
Acked-by: Robert Maynard <robert.maynard@kitware.com>
Merge-request: !1140
`vtkm_unit_tests` now supports an MPI option that can be used to add
test that run with MPI. Adding `UnitTestFieldRangeGlobalCompute` to test
global ranges for fields.
a37a3f39 Examples that use glu functions now properly link to OpenGL::GLU
a44ad273 Make sure Rendering example doesn't conflict with rendering testing name
Acked-by: Kitware Robot <kwrobot@kitware.com>
Acked-by: Utkarsh Ayachit <utkarsh.ayachit@kitware.com>
Merge-request: !1147
c6dcbc1f Check for git version only once
f2f8796d Properly install all required cmake files needed by vtk-m
b658a72a Fix broken VTK-m PyExpander functin
Acked-by: Kitware Robot <kwrobot@kitware.com>
Merge-request: !1144
The check to determine the version of VTK-m from git was duplicated in
CMakeLists.txt. Although pointless, it generally was not a big deal
(only an extra check when running CMake). Except for some reason with
the latest changes to the CMake build the second time find_package(Git)
was called on my system the GIT_EXECUTABLE variable got cleared out and
that caused the configure to fail. I have no idea why this happens (and
running CMake again seems to fix the problem), but simply removing the
extraneous find seems fix the problem.
CUDA currently doesn't support building for `compute_` and having compiled
in virtuals ( using separable compilation ). So we need to transition everything
over to `sm_`
Removing another API that need not be on MultiBlock. There's generally
no need for apps to know this. If needed, we can add `...Compute`
function. This removes another API on MultiBlock that could trigger
parallel communication/synchronization.
Removing MultiBlock::GetGlobalRange API to keep things consistent with
DataSet API. Instead, one should use `FieldRangeCompute` or
`FieldRangeGlobalCompute` as appropriate.
8472b64e Suppress more CUDA host function from a host/device function warnings
cd333bf8 Remove unneeded semicolon after closing parentheses on a some methods.
84311a24 Merge branch 'master' into cmake_refactor
83b360f1 vtkm::Lerp now casts integer types to an appropriate scalar type.
0d45c5cf Correct warnings found after disabling all warning when CUDA is enabled.
8808b41f Merge branch 'master' into vtk-m-cmake_refactor
d205f953 Update the documentation on when we need CMake 3.9+
e630ac5a Merge branch 'master' into vtk-m-cmake_refactor
...
Acked-by: Kitware Robot <kwrobot@kitware.com>
Acked-by: Kenneth Moreland <kmorel@sandia.gov>
Merge-request: !1135
3dd66e85 Do not use __assume for icc in gcc compatability mode
Acked-by: Kitware Robot <kwrobot@kitware.com>
Acked-by: Robert Maynard <robert.maynard@kitware.com>
Merge-request: !1139
c1237969 VTK-m ArrayHandle can now take ownership of a user allocated memory location
707970f4 VTK-m StorageBasic is now able to give/take ownership of user allocated memory.
Acked-by: Kitware Robot <kwrobot@kitware.com>
Acked-by: Kenneth Moreland <kmorel@sandia.gov>
Merge-request: !1137
The FindTBB.cmake module gets the TBB version by reading and grepping
one of the header files. Before doing so, it checks to see if the
version variable already exists. The problem was it was checking
TBB_VERSION, which was never created. Worse, other build systems might
set this for us (for example, the ParaView superbuild). Correct this by
making sure both TBB_VERSION_MAJOR and TBB_VERSION_MINOR are set.
When using the GNU header files on a system, the icc compiler emulates
the behavior of the gcc compiler on the system. (See https://
software.intel.com/en-us/node/522750) This appears to mean that icc
features that are not available in gcc could get turned off. In
particular, we found that the __assume feature stopped working.
To get around this problem, do not use __assume when compiling with icc
and __GNUC__ is defined. Instead, use the available gcc features.
Previously memory that was allocated outside of VTK-m was impossible to transfer to
VTK-m as we didn't know how to free it. By extending the ArrayHandle constructors
to support a Storage object that is being moved, we can clearly express that
the ArrayHandle now owns memory it didn't allocate.
Here is an example of how this is done:
```cpp
T* buffer = new T[100];
auto user_free_function = [](void* ptr) { delete[] static_cast<T*>(ptr); };
vtkm::cont::internal::Storage<T, vtkm::cont::StorageTagBasic>
storage(buffer, 100, user_free_function);
vtkm::cont::ArrayHandle<T> arrayHandle(std::move(storage));
```
This fixes the three following issues with StorageBasic.
1. Memory that was allocated by VTK-m and Stolen by the user needed the
proper free function called which is generally StorageBasicAllocator::deallocate.
But that was hard for the user to hold onto. So now we provide a function
pointer to the correct free function.
2. Memory that was allocated outside of VTK-m was impossible to transfer to
VTK-m as we didn't know how to free it. This is now resolved by allowing the
user to specify a free function to be called on release.
3. When the CUDA backend allocates memory for an ArrayHandle that has no
control representation, and the location we are running on supports concurrent
managed access we want to specify that cuda managed memory as also the host memory.
This requires that StorageBasic be able to call an arbitrary new delete function
which is chosen at runtime.
Previously types such as vtkm::Id or vtkm::U/Int32 would be cast to whatever
the weight type was. This is problematic as they should actually be casted to
a double type as the weight type could be a float and therefore the results
This commit removes `vtkm::filter::Result`. All methods that used
`vtkm::filter::Result` simply change to use `vtkm::cont::Dataset` instead.
The utility API on `Result` that was used to add fields to the resulting
dataset is now available via `vtkm::filter::internal::CreateResult`.
