Shuenhoy/vtk-m
1
0
Fork 0
mirror of https://gitlab.kitware.com/vtk/vtk-m synced 2024-09-16 17:22:55 +00:00
Code Issues 2 Actions Packages Projects Releases Wiki Activity
cec1091b0f
vtk-m/benchmarking/CMakeLists.txt

108 lines
4.2 KiB
CMake
Raw Normal View History

Adding micro benchmark suite
2015-07-06 21:44:29 +00:00
##============================================================================
## Copyright (c) Kitware, Inc.
## All rights reserved.
## See LICENSE.txt for details.
conslidate the license statement
2019-04-15 23:24:21 +00:00
##
Adding micro benchmark suite
2015-07-06 21:44:29 +00:00
## This software is distributed WITHOUT ANY WARRANTY; without even
## the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR
## PURPOSE. See the above copyright notice for more information.
##============================================================================
Port benchmarking framework to Google Benchmark.
2019-12-26 19:48:51 +00:00
Add module mechanism This mechanism sets up CMake variables that allow a user to select which modules/libraries to create. Dependencies will be tracked down to ensure that all of a module's dependencies are also enabled. The modules are also arranged into groups. Groups allow you to set the enable flag for a group of modules at once. Thus, if you have several modules that are likely to be used together, you can create a group for them. This can be handy in converting user-friendly CMake options (such as `VTKm_ENABLE_RENDERING`) to the modules that enable that by pointing to the appropriate group.
2022-02-16 12:50:39 +00:00
if(NOT vtkm_module_current STREQUAL "benchmarking")
message(FATAL_ERROR "Benchmarking CMakeLists.txt not loaded by benchmarking module.")
endif()
Add changes for supporting Kokkos/HIP Some of the unit tests for serial and kokkos are disable for hipcc to properly compile. VTKM_MATH_ASSERT and VTKM_TEST_ASSERT fail to compile with HIP in execution environment so they are disabled with building with HIP. Kokkos::finalize is causing error so it is temporarily disabled.
2021-08-06 19:41:07 +00:00
#Find Google Benchmark.Note that benchmark_DIR must be pointed at an
#installation, not a build directory.
Mark the find_package(benchmark) with required Add the REQUIRED flag to the find_package command to find the Google benchmark package. Without that flag, the find_package command will happily pass if the benchmarks are not found. However, if benchmarks is not found, then the CMake configure will then fail because the benchmark::benchmark library is not defined. That is a confusing error to get. Since benchmark is not a typical thing installed on a system, this can be problematic.
2020-01-09 19:28:48 +00:00
find_package(benchmark REQUIRED)
Port benchmarking framework to Google Benchmark.
2019-12-26 19:48:51 +00:00
Add module mechanism This mechanism sets up CMake variables that allow a user to select which modules/libraries to create. Dependencies will be tracked down to ensure that all of a module's dependencies are also enabled. The modules are also arranged into groups. Groups allow you to set the enable flag for a group of modules at once. Thus, if you have several modules that are likely to be used together, you can create a group for them. This can be handy in converting user-friendly CMake options (such as `VTKm_ENABLE_RENDERING`) to the modules that enable that by pointing to the appropriate group.
2022-02-16 12:50:39 +00:00
function(add_benchmark exe_name)
add_executable(${exe_name} ${exe_name}.cxx)
Port benchmarking framework to Google Benchmark.
2019-12-26 19:48:51 +00:00
target_link_libraries(${exe_name} PRIVATE benchmark::benchmark)
Add module mechanism This mechanism sets up CMake variables that allow a user to select which modules/libraries to create. Dependencies will be tracked down to ensure that all of a module's dependencies are also enabled. The modules are also arranged into groups. Groups allow you to set the enable flag for a group of modules at once. Thus, if you have several modules that are likely to be used together, you can create a group for them. This can be handy in converting user-friendly CMake options (such as `VTKm_ENABLE_RENDERING`) to the modules that enable that by pointing to the appropriate group.
2022-02-16 12:50:39 +00:00
vtkm_module_get_property(depends benchmarking DEPENDS)
target_link_libraries(${exe_name} PRIVATE ${depends})
vtkm_module_get_property(optional_depends benchmarking OPTIONAL_DEPENDS)
foreach(dep IN LISTS optional_depends)
if(TARGET ${dep})
target_link_libraries(${exe_name} PRIVATE ${dep})
endif()
endforeach()
VTK-m builds with separate function sections to allow smaller binaries Consumers of VTK-m when enabling of dropping of unused functions will see VTK-m functions dropped. Previously this didn't happen as VTK-m didn't build object files with the correct flags for this. By allowing the linker to remove unused symbols we see a significant saving the file size of VTK-m tests, examples, and benchmarks. An OpenMP build of the tests and benchmarks goes from 168MB to 141MB which is roughly a 16% filesize reduction. Initially I had presumed that these changes would increase link times. But in measurements the total wall time for compilation of VTK-m has stayed about the same ( seeing a decrease of 1.5% ). Presumably the increased computation is offset by the reduction in file writing.
2019-09-09 16:31:01 +00:00
vtkm_add_drop_unused_function_flags(${exe_name})
vtkm_add_target_information(${exe_name})
Merge benchmark executables into a device dependent shared library VTK-m has been updated to replace old per device benchmark executables with a device dependent shared library so that it's able to accept a device adapter at runtime through the "--device=" argument.
2019-02-15 19:47:07 +00:00
set_target_properties(${exe_name} PROPERTIES
RUNTIME_OUTPUT_DIRECTORY ${VTKm_EXECUTABLE_OUTPUT_PATH}
Make VTK-m use consistent symbol visibility for benchmarks
2019-06-18 13:59:43 +00:00
CXX_VISIBILITY_PRESET "hidden"
Update examples and benchmarks to use new VTK-m CMake helper function
2019-07-08 21:10:50 +00:00
CUDA_VISIBILITY_PRESET "hidden"
Make VTK-m use consistent symbol visibility for benchmarks
2019-06-18 13:59:43 +00:00
)
VTK-m builds with separate function sections to allow smaller binaries Consumers of VTK-m when enabling of dropping of unused functions will see VTK-m functions dropped. Previously this didn't happen as VTK-m didn't build object files with the correct flags for this. By allowing the linker to remove unused symbols we see a significant saving the file size of VTK-m tests, examples, and benchmarks. An OpenMP build of the tests and benchmarks goes from 168MB to 141MB which is roughly a 16% filesize reduction. Initially I had presumed that these changes would increase link times. But in measurements the total wall time for compilation of VTK-m has stayed about the same ( seeing a decrease of 1.5% ). Presumably the increased computation is offset by the reduction in file writing.
2019-09-09 16:31:01 +00:00
Add module mechanism This mechanism sets up CMake variables that allow a user to select which modules/libraries to create. Dependencies will be tracked down to ensure that all of a module's dependencies are also enabled. The modules are also arranged into groups. Groups allow you to set the enable flag for a group of modules at once. Thus, if you have several modules that are likely to be used together, you can create a group for them. This can be handy in converting user-friendly CMake options (such as `VTKm_ENABLE_RENDERING`) to the modules that enable that by pointing to the appropriate group.
2022-02-16 12:50:39 +00:00
# At some point, we might not want to compile benchmarks with device compilers.
vtkm_add_target_information(${exe_name} DEVICE_SOURCES ${exe_name}.cxx)
Added benchmarking function for building benchmarks #3 added an explicit approach for adding benchmarks since we are now using the new cmake style
2017-11-08 22:24:02 +00:00
endfunction()
Adding micro benchmark suite
2015-07-06 21:44:29 +00:00
Benchmarks build with the correct backend when cuda is enabled
2017-12-29 19:39:31 +00:00
set(benchmarks
BenchmarkArrayTransfer
Add benchmark for atomic arrays.
2018-08-29 16:03:07 +00:00
BenchmarkAtomicArray
Benchmarks build with the correct backend when cuda is enabled
2017-12-29 19:39:31 +00:00
BenchmarkCopySpeeds
BenchmarkDeviceAdapter
BenchmarkFieldAlgorithms
Add filter benchmarks.
2018-07-13 16:08:28 +00:00
BenchmarkFilters
Add benchmark for 2D explicit grids.
2023-12-05 14:39:08 +00:00
BenchmarkLocators
Benchmark ODE integrators.
2020-08-18 19:47:13 +00:00
BenchmarkODEIntegrators
Benchmarks build with the correct backend when cuda is enabled
2017-12-29 19:39:31 +00:00
BenchmarkTopologyAlgorithms
)
Add module mechanism This mechanism sets up CMake variables that allow a user to select which modules/libraries to create. Dependencies will be tracked down to ensure that all of a module's dependencies are also enabled. The modules are also arranged into groups. Groups allow you to set the enable flag for a group of modules at once. Thus, if you have several modules that are likely to be used together, you can create a group for them. This can be handy in converting user-friendly CMake options (such as `VTKm_ENABLE_RENDERING`) to the modules that enable that by pointing to the appropriate group.
2022-02-16 12:50:39 +00:00
if(TARGET vtkm_rendering)
list(APPEND benchmarks
BenchmarkRayTracing
BenchmarkInSitu
)
endif()
benchmarks: pass unparsed args to Google benchmark - It also adds Google's benchmarch compare.py script - It is installed to the build directory. - It add a wrapper script called compare-benchmarks.py which: - Let you run each of the benchmarks with different devices - It adds a README.md explaining how to run the benchmarks - BenchmarkDeviceAdapter input size range parametrized at compile time Signed-off-by: Vicente Adolfo Bolea Sanchez <vicente.bolea@kitware.com>
2020-03-19 19:15:32 +00:00
set(VTKm_BENCHS_RANGE_LOWER_BOUNDARY 4096 CACHE STRING "Smallest sample for input size bench for BenchmarkDeviceAdapter")
set(VTKm_BENCHS_RANGE_UPPER_BOUNDARY 134217728 CACHE STRING "Biggest sample for input size bench for BenchmarkDeviceAdapter")
mark_as_advanced(VTKm_BENCHS_RANGE_LOWER_BOUNDARY VTKm_BENCHS_RANGE_UPPER_BOUNDARY)
Added benchmarking function for building benchmarks #3 added an explicit approach for adding benchmarks since we are now using the new cmake style
2017-11-08 22:24:02 +00:00
foreach (benchmark ${benchmarks})
Add module mechanism This mechanism sets up CMake variables that allow a user to select which modules/libraries to create. Dependencies will be tracked down to ensure that all of a module's dependencies are also enabled. The modules are also arranged into groups. Groups allow you to set the enable flag for a group of modules at once. Thus, if you have several modules that are likely to be used together, you can create a group for them. This can be handy in converting user-friendly CMake options (such as `VTKm_ENABLE_RENDERING`) to the modules that enable that by pointing to the appropriate group.
2022-02-16 12:50:39 +00:00
add_benchmark(${benchmark})
Added newlines to resolve kitware robot errors I was getting missing newlines errors from the kitware robot so this should fix them
2017-11-17 16:12:16 +00:00
endforeach ()
Benchmarks build with the correct backend when cuda is enabled
2017-12-29 19:39:31 +00:00
Re-enable hip builds There were several tests that were disabled for hip because they either took too long to compile or were failing. We are getting closer to making everything work, so re-enable this part of the build.
2022-05-12 20:00:30 +00:00
target_compile_definitions(BenchmarkDeviceAdapter PUBLIC VTKm_BENCHS_RANGE_LOWER_BOUNDARY=${VTKm_BENCHS_RANGE_LOWER_BOUNDARY})
target_compile_definitions(BenchmarkDeviceAdapter PUBLIC VTKm_BENCHS_RANGE_UPPER_BOUNDARY=${VTKm_BENCHS_RANGE_UPPER_BOUNDARY})
benchmarks: pass unparsed args to Google benchmark - It also adds Google's benchmarch compare.py script - It is installed to the build directory. - It add a wrapper script called compare-benchmarks.py which: - Let you run each of the benchmarks with different devices - It adds a README.md explaining how to run the benchmarks - BenchmarkDeviceAdapter input size range parametrized at compile time Signed-off-by: Vicente Adolfo Bolea Sanchez <vicente.bolea@kitware.com>
2020-03-19 19:15:32 +00:00
perftest: add perf regression test
2022-10-07 21:42:12 +00:00
if(VTKm_ENABLE_PERFORMANCE_TESTING)
include("${VTKm_SOURCE_DIR}/CMake/testing/VTKmPerformanceTest.cmake")
add_benchmark_test(BenchmarkFilters
Split up the filters benchmark tests This should help catch regressions that only affect a single filter. It also allows us to set up a partitioned test that has many small blocks. It also fixes an issue we were seeing on the dashboard where the contour benchmark was apparently running out of resources.
2022-12-04 04:44:20 +00:00
NAME BenchThreshold
REGEX BenchThreshold
)
add_benchmark_test(BenchmarkFilters
NAME BenchTetrahedralize
REGEX BenchTetrahedralize
)
add_benchmark_test(BenchmarkFilters
NAME BenchContour
REGEX
BenchContour/IsStructuredDataSet:1/NIsoVals:12/MergePts:1/GenNormals:0.*/MultiPartitioned:0
BenchContour/IsStructuredDataSet:1/NIsoVals:12/MergePts:0/GenNormals:1/FastNormals:1.*/MultiPartitioned:0
BenchContour/IsStructuredDataSet:0/NIsoVals:12/MergePts:1/GenNormals:0.*/MultiPartitioned:0
BenchContour/IsStructuredDataSet:0/NIsoVals:12/MergePts:0/GenNormals:1/FastNormals:1.*/MultiPartitioned:0
)
add_benchmark_test(BenchmarkFilters
NAME BenchContourPartitioned
ARGS --wavelet-dim=32 --num-partitions=128
REGEX
BenchContour/IsStructuredDataSet:1/NIsoVals:12/MergePts:1/GenNormals:0.*/MultiPartitioned:1
BenchContour/IsStructuredDataSet:1/NIsoVals:12/MergePts:0/GenNormals:1/FastNormals:1.*/MultiPartitioned:1
BenchContour/IsStructuredDataSet:0/NIsoVals:12/MergePts:1/GenNormals:0.*/MultiPartitioned:1
BenchContour/IsStructuredDataSet:0/NIsoVals:12/MergePts:0/GenNormals:1/FastNormals:1.*/MultiPartitioned:1
)
add_benchmark_test(BenchmarkFilters
NAME BenchVertexClustering
REGEX BenchVertexClustering/NumDivs:256
Add performance configuration options Added a name option that allows the same benchmark executable to be used in multiple benchmark tests. This allows the benchmarks to be separated. Also added an option to pass customized arguments to the benchmark executable to overwrite the default values.
2022-12-04 04:13:26 +00:00
)
perftest: add perf regression test
2022-10-07 21:42:12 +00:00
if(TARGET vtkm_rendering)
Split up the filters benchmark tests This should help catch regressions that only affect a single filter. It also allows us to set up a partitioned test that has many small blocks. It also fixes an issue we were seeing on the dashboard where the contour benchmark was apparently running out of resources.
2022-12-04 04:44:20 +00:00
add_benchmark_test(BenchmarkInSitu REGEX "BenchContour")
perftest: add perf regression test
2022-10-07 21:42:12 +00:00
endif()
endif()
Reference in New Issue Copy Permalink