mtapi_cuda_c: first implementation

CMakeLists.txt

@@ -30,6 +30,8 @@ set (EMBB_BASE_VERSION_MAJOR 0)
 set (EMBB_BASE_VERSION_MINOR 3)
 set (EMBB_BASE_VERSION_PATCH 2)
 
+include(FindCUDA)
+
 # Fix compilation for CMake versions >= 3.1
 #
 # New Policy 0054:
@@ -144,6 +146,9 @@ set(EXPECTED_EMBB_TEST_EXECUTABLES "embb_algorithms_cpp_test"
 if(BUILD_OPENCL_PLUGIN STREQUAL ON)
   list(APPEND EXPECTED_EMBB_TEST_EXECUTABLES "embb_mtapi_opencl_c_test")
 endif()
+if(CUDA_FOUND)
+  list(APPEND EXPECTED_EMBB_TEST_EXECUTABLES "embb_mtapi_cuda_c_test")
+endif()
 
 
 ## Copy test execution script to local binaries folder
@@ -180,6 +185,9 @@ add_subdirectory(mtapi_plugins_c/mtapi_network_c)
 if(BUILD_OPENCL_PLUGIN STREQUAL ON)
   add_subdirectory(mtapi_plugins_c/mtapi_opencl_c)
 endif()
+if(CUDA_FOUND)
+  add_subdirectory(mtapi_plugins_c/mtapi_cuda_c)
+endif()
 add_subdirectory(tasks_cpp)
 add_subdirectory(mtapi_cpp)
 add_subdirectory(containers_cpp)

mtapi_plugins_c/mtapi_cuda_c/CMakeLists.txt

+project (project_embb_mtapi_cuda_c)
+
+file(GLOB_RECURSE EMBB_MTAPI_CUDA_C_SOURCES "src/*.c" "src/*.h")
+file(GLOB_RECURSE EMBB_MTAPI_CUDA_C_HEADERS "include/*.h")
+
+file(GLOB_RECURSE EMBB_MTAPI_CUDA_TEST_SOURCES "test/*.cc" "test/*.cu" "test/*.h")
+
+IF(MSVC8 OR MSVC9 OR MSVC10 OR MSVC11)
+FOREACH(src_tmp ${EMBB_MTAPI_CUDA_TEST_SOURCES})
+    SET_PROPERTY(SOURCE ${src_tmp} PROPERTY LANGUAGE CXX)
+ENDFOREACH(src_tmp)
+FOREACH(src_tmp ${EMBB_MTAPI_CUDA_C_SOURCES})
+    SET_PROPERTY(SOURCE ${src_tmp} PROPERTY LANGUAGE CXX)
+ENDFOREACH(src_tmp)
+ENDIF()
+
+IF(CMAKE_COMPILER_IS_GNUCC)
+  set (EMBB_MTAPI_CUDA_C_LIBS dl)
+ENDIF()
+
+# Execute the GroupSources macro
+include(${CMAKE_SOURCE_DIR}/CMakeCommon/GroupSourcesMSVC.cmake)
+GroupSourcesMSVC(include)
+GroupSourcesMSVC(src)
+GroupSourcesMSVC(test)
+
+set (EMBB_MTAPI_CUDA_INCLUDE_DIRS "include" "src" "test")
+include_directories(${EMBB_MTAPI_CUDA_INCLUDE_DIRS}
+                    ${CUDA_TOOLKIT_INCLUDE}
+                    ${CMAKE_CURRENT_SOURCE_DIR}/../../base_c/include
+                    ${CMAKE_CURRENT_BINARY_DIR}/../../base_c/include
+                    ${CMAKE_CURRENT_SOURCE_DIR}/../../mtapi_c/include
+                    ${CMAKE_CURRENT_SOURCE_DIR}/../../mtapi_c/src
+                    )
+
+add_library(embb_mtapi_cuda_c ${EMBB_MTAPI_CUDA_C_SOURCES} ${EMBB_MTAPI_CUDA_C_HEADERS})
+target_link_libraries(embb_mtapi_cuda_c embb_mtapi_c embb_base_c)
+
+if (BUILD_TESTS STREQUAL ON)
+  add_custom_command(
+    DEPENDS "${CMAKE_CURRENT_SOURCE_DIR}/test/embb_mtapi_cuda_test_kernel.cu"
+    OUTPUT "${CMAKE_CURRENT_BINARY_DIR}/embb_mtapi_cuda_test_kernel.ptx"
+    COMMAND ${CUDA_NVCC_EXECUTABLE} -ptx -m 32
+      "${CMAKE_CURRENT_SOURCE_DIR}/test/embb_mtapi_cuda_test_kernel.cu"
+      -o "${CMAKE_CURRENT_BINARY_DIR}/embb_mtapi_cuda_test_kernel.ptx"
+    WORKING_DIRECTORY ${CMAKE_CURRENT_BINARY_DIR}
+  )
+  add_custom_command(
+    DEPENDS "${CMAKE_CURRENT_BINARY_DIR}/embb_mtapi_cuda_test_kernel.ptx"
+    OUTPUT "${CMAKE_CURRENT_BINARY_DIR}/embb_mtapi_cuda_test_kernel.h"
+    COMMAND ${CUDA_TOOLKIT_ROOT_DIR}/bin/bin2c -p 0
+      "${CMAKE_CURRENT_BINARY_DIR}/embb_mtapi_cuda_test_kernel.ptx" >
+      "${CMAKE_CURRENT_BINARY_DIR}/embb_mtapi_cuda_test_kernel.h"
+    WORKING_DIRECTORY ${CMAKE_CURRENT_BINARY_DIR}
+  )
+  include_directories(
+    ${CMAKE_CURRENT_BINARY_DIR}
+    ${CMAKE_CURRENT_BINARY_DIR}/../../partest/include)
+  add_executable (embb_mtapi_cuda_c_test ${EMBB_MTAPI_CUDA_TEST_SOURCES} "${CMAKE_CURRENT_BINARY_DIR}/embb_mtapi_cuda_test_kernel.h")
+  target_link_libraries(embb_mtapi_cuda_c_test embb_mtapi_cuda_c embb_mtapi_c partest embb_base_c ${compiler_libs} ${EMBB_MTAPI_CUDA_C_LIBS} ${CUDA_CUDA_LIBRARY})
+  CopyBin(BIN embb_mtapi_cuda_c_test DEST ${local_install_dir})
+endif()
+
+install(DIRECTORY ${CMAKE_CURRENT_SOURCE_DIR}/include/
+        DESTINATION include FILES_MATCHING PATTERN "*.h")
+install(TARGETS embb_mtapi_cuda_c DESTINATION lib)

mtapi_plugins_c/mtapi_cuda_c/include/embb/mtapi/c/mtapi_cuda.h

+/*
+ * Copyright (c) 2014, Siemens AG. All rights reserved.
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions are met:
+ *
+ * 1. Redistributions of source code must retain the above copyright notice,
+ * this list of conditions and the following disclaimer.
+ *
+ * 2. Redistributions in binary form must reproduce the above copyright notice,
+ * this list of conditions and the following disclaimer in the documentation
+ * and/or other materials provided with the distribution.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
+ * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE
+ * LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
+ * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
+ * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
+ * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
+ * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
+ * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
+ * POSSIBILITY OF SUCH DAMAGE.
+ */
+
+#ifndef EMBB_MTAPI_C_MTAPI_CUDA_H_
+#define EMBB_MTAPI_C_MTAPI_CUDA_H_
+
+
+#include <embb/mtapi/c/mtapi_ext.h>
+
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+
+/**
+ * \defgroup C_MTAPI_CUDA MTAPI CUDA Plugin
+ *
+ * \ingroup C_MTAPI_EXT
+ *
+ * Provides functionality to execute tasks on CUDA devices.
+ */
+
+
+/**
+ * Initializes the MTAPI CUDA environment on a previously initialized MTAPI
+ * node.
+ *
+ * It must be called on all nodes using the MTAPI CUDA plugin.
+ *
+ * Application software using MTAPI network must call
+ * mtapi_cuda_plugin_initialize() once per node. It is an error to call
+ * mtapi_cuda_plugin_initialize() multiple times
+ * from a given node, unless mtapi_cuda_plugin_finalize() is called in
+ * between.
+ *
+ * On success, \c *status is set to \c MTAPI_SUCCESS. On error, \c *status is
+ * set to the appropriate error defined below.
+ * Error code                  | Description
+ * --------------------------- | ----------------------------------------------
+ * \c MTAPI_ERR_UNKNOWN        | MTAPI CUDA couldn't be initialized.
+ *
+ * \see mtapi_cuda_plugin_finalize()
+ *
+ * \notthreadsafe
+ * \ingroup C_MTAPI_CUDA
+ */
+void mtapi_cuda_plugin_initialize(
+  MTAPI_OUT mtapi_status_t* status     /**< [out] Pointer to error code,
+                                            may be \c MTAPI_NULL */
+);
+
+/**
+ * Finalizes the MTAPI CUDA environment on the local MTAPI node.
+ *
+ * It has to be called by each node using MTAPI CUDA. It is an error to call
+ * mtapi_cuda_plugin_finalize() without first calling
+ * mtapi_cuda_plugin_initialize(). An MTAPI node can call
+ * mtapi_cuda_plugin_finalize() once for each call to
+ * mtapi_cuda_plugin_initialize(), but it is an error to call
+ * mtapi_cuda_plugin_finalize() multiple times from a given node
+ * unless mtapi_cuda_plugin_initialize() has been called prior to each
+ * mtapi_cuda_plugin_finalize() call.
+ *
+ * All network tasks that have not completed and that have been started on the
+ * node where mtapi_cuda_plugin_finalize() is called will be canceled
+ * (see mtapi_task_cancel()). mtapi_opencl_plugin_finalize() blocks until all
+ * tasks that have been started on the same node return. Tasks that execute
+ * actions on the node where mtapi_opencl_plugin_finalize() is called, also
+ * block finalization of the MTAPI CUDA system on that node.
+ *
+ * On success, \c *status is set to \c MTAPI_SUCCESS. On error, \c *status is
+ * set to the appropriate error defined below.
+ * Error code                    | Description
+ * ----------------------------- | --------------------------------------------
+ * \c MTAPI_ERR_UNKNOWN          | MTAPI CUDA couldn't be finalized.
+ *
+ * \see mtapi_opencl_plugin_initialize(), mtapi_task_cancel()
+ *
+ * \notthreadsafe
+ * \ingroup C_MTAPI_CUDA
+ */
+void mtapi_cuda_plugin_finalize(
+  MTAPI_OUT mtapi_status_t* status     /**< [out] Pointer to error code,
+                                            may be \c MTAPI_NULL */
+);
+
+/**
+ * This function creates an CUDA action.
+ *
+ * It is called on the node where the user wants to execute an action on an
+ * CUDA device. A CUDA action contains a reference to a local job, the
+ * kernel source to compile and execute on the CUDA device, the name of the
+ * kernel function, a local work size (see CUDA specification for details)
+ * and the size of one element in the result buffer.
+ * After an CUDA action is created, it is referenced by the application using
+ * a node-local handle of type \c mtapi_action_hndl_t, or indirectly through a
+ * node-local job handle of type \c mtapi_job_hndl_t. An CUDA action's
+ * life-cycle begins with mtapi_cuda_action_create(), and ends when
+ * mtapi_action_delete() or mtapi_finalize() is called.
+ *
+ * To create an action, the application must supply the domain-wide job ID of
+ * the job associated with the action. Job IDs must be predefined in the
+ * application and runtime, of type \c mtapi_job_id_t, which is an
+ * implementation-defined type. The job ID is unique in the sense that it is
+ * unique for the job implemented by the action. However several actions may
+ * implement the same job for load balancing purposes.
+ *
+ * If \c node_local_data_size is not zero, \c node_local_data specifies the
+ * start of node local data shared by kernel functions executed on the same
+ * node. \c node_local_data_size can be used by the runtime for cache coherency
+ * operations.
+ *
+ * On success, an action handle is returned and \c *status is set to
+ * \c MTAPI_SUCCESS. On error, \c *status is set to the appropriate error
+ * defined below. In the case where the action already exists, \c status will
+ * be set to \c MTAPI_ERR_ACTION_EXISTS and the handle returned will not be a
+ * valid handle.
+ * <table>
+ *   <tr>
+ *     <th>Error code</th>
+ *     <th>Description</th>
+ *   </tr>
+ *   <tr>
+ *     <td>\c MTAPI_ERR_JOB_INVALID</td>
+ *     <td>The \c job_id is not a valid job ID, i.e., no action was created for
+ *         that ID or the action has been deleted.</td>
+ *   </tr>
+ *   <tr>
+ *     <td>\c MTAPI_ERR_ACTION_EXISTS</td>
+ *     <td>This action is already created.</td>
+ *   </tr>
+ *   <tr>
+ *     <td>\c MTAPI_ERR_ACTION_LIMIT</td>
+ *     <td>Exceeded maximum number of actions allowed.</td>
+ *   </tr>
+ *   <tr>
+ *     <td>\c MTAPI_ERR_NODE_NOTINIT</td>
+ *     <td>The calling node is not initialized.</td>
+ *   </tr>
+ *   <tr>
+ *     <td>\c MTAPI_ERR_UNKNOWN</td>
+ *     <td>The kernel could not be compiled or no CUDA device was
+ *         available.</td>
+ *   </tr>
+ * </table>
+ *
+ * \see mtapi_action_delete(), mtapi_finalize()
+ *
+ * \returns Handle to newly created CUDA action, invalid handle on error
+ * \threadsafe
+ * \ingroup C_MTAPI_CUDA
+ */
+mtapi_action_hndl_t mtapi_cuda_action_create(
+  MTAPI_IN mtapi_job_id_t job_id,      /**< [in] Job id */
+  MTAPI_IN char* kernel_source,        /**< [in] Pointer to kernel source */
+  MTAPI_IN char* kernel_name,          /**< [in] Name of the kernel function */
+  MTAPI_IN mtapi_size_t local_work_size,
+                                       /**< [in] Size of local work group */
+  MTAPI_IN mtapi_size_t element_size,  /**< [in] Size of one element in the
+                                            result buffer */
+  MTAPI_IN void* node_local_data,      /**< [in] Data shared across tasks */
+  MTAPI_IN mtapi_size_t node_local_data_size,
+                                       /**< [in] Size of shared data */
+  MTAPI_OUT mtapi_status_t* status     /**< [out] Pointer to error code,
+                                            may be \c MTAPI_NULL */
+);
+
+
+#ifdef __cplusplus
+}
+#endif
+
+
+#endif // EMBB_MTAPI_C_MTAPI_CUDA_H_

mtapi_plugins_c/mtapi_cuda_c/test/embb_mtapi_cuda_test_kernel.cu

+extern "C" __global__ void test(
+  void* arguments,
+  int arguments_size,
+  void* result_buffer,
+  int result_buffer_size,
+  void* node_local_data,
+  int node_local_data_size) {
+  int ii = blockDim.x * blockIdx.x + threadIdx.x;
+  int elements = arguments_size / sizeof(float) / 2;
+  if (ii >= elements)
+    return;
+  float* a = (float*)arguments;
+  float* b = ((float*)arguments) + elements;
+  float* c = (float*)result_buffer;
+  float* d = (float*)node_local_data;
+  c[ii] = a[ii] + b[ii] + d[0];
+}

mtapi_plugins_c/mtapi_cuda_c/test/embb_mtapi_cuda_test_task.cc

+/*
+ * Copyright (c) 2014, Siemens AG. All rights reserved.
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions are met:
+ *
+ * 1. Redistributions of source code must retain the above copyright notice,
+ * this list of conditions and the following disclaimer.
+ *
+ * 2. Redistributions in binary form must reproduce the above copyright notice,
+ * this list of conditions and the following disclaimer in the documentation
+ * and/or other materials provided with the distribution.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
+ * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE
+ * LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
+ * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
+ * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
+ * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
+ * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
+ * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
+ * POSSIBILITY OF SUCH DAMAGE.
+ */
+
+#include <embb_mtapi_cuda_test_task.h>
+
+#include <embb/mtapi/c/mtapi_cuda.h>
+
+#define MTAPI_CHECK_STATUS(status) \
+PT_ASSERT(MTAPI_SUCCESS == status)
+
+#define CUDA_DOMAIN 1
+#define CUDA_NODE 2
+#define CUDA_JOB 2
+
+// CUDA Kernel Function for element by element vector addition
+#include "embb_mtapi_cuda_test_kernel.h"
+
+TaskTest::TaskTest() {
+  CreateUnit("mtapi cuda task test").Add(&TaskTest::TestBasic, this);
+}
+
+void TaskTest::TestBasic() {
+  mtapi_status_t status;
+  mtapi_job_hndl_t job;
+  mtapi_task_hndl_t task;
+  mtapi_action_hndl_t action;
+
+  const int kElements = 64;
+  float arguments[kElements * 2];
+  float results[kElements];
+
+  for (int ii = 0; ii < kElements; ii++) {
+    arguments[ii] = static_cast<float>(ii);
+    arguments[ii + kElements] = static_cast<float>(ii);
+  }
+
+  mtapi_cuda_plugin_initialize(&status);
+  if (status == MTAPI_ERR_FUNC_NOT_IMPLEMENTED) {
+    // CUDA unavailable
+    return;
+  }
+  MTAPI_CHECK_STATUS(status);
+
+  mtapi_initialize(
+    CUDA_DOMAIN,
+    CUDA_NODE,
+    MTAPI_NULL,
+    MTAPI_NULL,
+    &status);
+  MTAPI_CHECK_STATUS(status);
+
+  float node_local = 1.0f;
+  action = mtapi_cuda_action_create(
+    CUDA_JOB,
+    reinterpret_cast<char const *>(imageBytes), "test", 32, 4,
+    &node_local, sizeof(float),
+    &status);
+  MTAPI_CHECK_STATUS(status);
+
+  status = MTAPI_ERR_UNKNOWN;
+  job = mtapi_job_get(CUDA_JOB, CUDA_DOMAIN, &status);
+  MTAPI_CHECK_STATUS(status);
+
+  task = mtapi_task_start(
+    MTAPI_TASK_ID_NONE,
+    job,
+    arguments, kElements * 2 * sizeof(float),
+    results, kElements*sizeof(float),
+    MTAPI_DEFAULT_TASK_ATTRIBUTES,
+    MTAPI_GROUP_NONE,
+    &status);
+  MTAPI_CHECK_STATUS(status);
+
+  mtapi_task_wait(task, MTAPI_INFINITE, &status);
+  MTAPI_CHECK_STATUS(status);
+
+  for (int ii = 0; ii < kElements; ii++) {
+    PT_EXPECT_EQ(results[ii], ii * 2 + 1);
+  }
+
+  mtapi_action_delete(action, MTAPI_INFINITE, &status);
+  MTAPI_CHECK_STATUS(status);
+
+  mtapi_finalize(&status);
+  MTAPI_CHECK_STATUS(status);
+
+  mtapi_cuda_plugin_finalize(&status);
+  MTAPI_CHECK_STATUS(status);
+}

mtapi_plugins_c/mtapi_cuda_c/test/embb_mtapi_cuda_test_task.h

+/*
+ * Copyright (c) 2014, Siemens AG. All rights reserved.
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions are met:
+ *
+ * 1. Redistributions of source code must retain the above copyright notice,
+ * this list of conditions and the following disclaimer.
+ *
+ * 2. Redistributions in binary form must reproduce the above copyright notice,
+ * this list of conditions and the following disclaimer in the documentation
+ * and/or other materials provided with the distribution.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
+ * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE
+ * LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
+ * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
+ * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
+ * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
+ * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
+ * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
+ * POSSIBILITY OF SUCH DAMAGE.
+ */
+
+#ifndef MTAPI_PLUGINS_C_MTAPI_CUDA_C_TEST_EMBB_MTAPI_CUDA_TEST_TASK_H_
+#define MTAPI_PLUGINS_C_MTAPI_CUDA_C_TEST_EMBB_MTAPI_CUDA_TEST_TASK_H_
+
+#include <partest/partest.h>
+
+class TaskTest : public partest::TestCase {
+ public:
+  TaskTest();
+
+ private:
+  void TestBasic();
+};
+
+#endif // MTAPI_PLUGINS_C_MTAPI_CUDA_C_TEST_EMBB_MTAPI_CUDA_TEST_TASK_H_

mtapi_plugins_c/mtapi_cuda_c/test/main.cc

+/*
+ * Copyright (c) 2014, Siemens AG. All rights reserved.
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions are met:
+ *
+ * 1. Redistributions of source code must retain the above copyright notice,
+ * this list of conditions and the following disclaimer.
+ *
+ * 2. Redistributions in binary form must reproduce the above copyright notice,
+ * this list of conditions and the following disclaimer in the documentation
+ * and/or other materials provided with the distribution.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
+ * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE
+ * LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
+ * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
+ * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
+ * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
+ * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
+ * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
+ * POSSIBILITY OF SUCH DAMAGE.
+ */
+
+#include <partest/partest.h>
+
+#include <embb_mtapi_cuda_test_task.h>
+
+
+PT_MAIN("MTAPI CUDA") {
+  PT_RUN(TaskTest);
+}