/* * Copyright (c) 2017, Intel Corporation * * Permission is hereby granted, free of charge, to any person obtaining a * copy of this software and associated documentation files (the "Software"), * to deal in the Software without restriction, including without limitation * the rights to use, copy, modify, merge, publish, distribute, sublicense, * and/or sell copies of the Software, and to permit persons to whom the * Software is furnished to do so, subject to the following conditions: * * The above copyright notice and this permission notice shall be included * in all copies or substantial portions of the Software. * * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS * OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR * OTHER DEALINGS IN THE SOFTWARE. */ #include "cm_device.h" #include "cm_event_base.h" #include "cm_queue_base_hw.h" #include "cm_timer.h" //! //! \brief Destroys CM Device //! \details Destroys the CM Device. Also destroys surfaces, //! kernels, samplers and the queue that was created //! using this device instance that haven't already been //! explicitly destroyed by calling respective destroy functions. //! \param [in] device //! Reference to the pointer to the CmDevice to be destroyed //! \retval CM_SUCCESS if device successfully destroyed //! \retval CM_FAILURE if failed to destroy device //! extern "C" CM_RT_API int32_t DestroyCmDevice(CmDevice* &device) { INSERT_PROFILER_RECORD(); CmDevice_RT* p = static_cast(device); int32_t result = CmDevice_RT::Destroy(p); if(result == CM_SUCCESS) { device = nullptr; } return result; } //! //! \brief Enqueue a task //! \details Enqueues a task represented by CmTask. //! The kernels in the CmTask object may be run concurrently. //! Each task can have up to CAP_KERNEL_COUNT_PER_TASK kernels. //! Tasks are executed according to the order they //! were enqueued. The next task will not start execution until //! the current task finishes. //! This is a nonblocking call. i.e. it returns immediately //! without waiting for GPU to start or finish execution of the //! task. A CmEvent is generated each time a task is enqueued. //! The CmEvent can be used to check the status or other data //! regarding the task execution. The generated event needs to //! be managed and released by the user. //! To avoid generating an event, the user can set the event //! as CM_NO_EVENT and pass it to this function. //! \param [in] queue //! Pointer to the CmQueue //! \param [in] task //! Pointer to the CmTask //! \param [out] event //! Reference to the pointer for the CmEvent to be created //! \param [in] threadSpace //! Pointer to the thread space //! \retval CM_SUCCESS if task successfully enqueued //! \retval CM_OUT_OF_HOST_MEMORY if out of host memory //! \retval CM_FAILURE otherwise //! EXTERN_C CM_RT_API int32_t CMRT_Enqueue(CmQueue* queue, CmTask* task, CmEvent** event, const CmThreadSpace* threadSpace = nullptr) { INSERT_PROFILER_RECORD(); return queue->Enqueue(task, *event, threadSpace); }