ie_iinfer_request.hpp

Go to the documentation of this file.

 // Copyright (C) 2018-2019 Intel Corporation


 // SPDX-License-Identifier: Apache-2.0


 //


 


 /**


  * @brief a header file for IInferRequest interface


  * @file ie_iinfer_request.hpp


  */


 

 #pragma once


 

 #include "ie_common.h"


 #include <ie_blob.h>


 #include <memory>


 #include <string>


 #include <map>


 #include <details/ie_irelease.hpp>


 

 namespace InferenceEngine {

 


 /**


  * @brief This is an interface of asynchronous infer request


  */


 class IInferRequest : public details::IRelease {

 public:


     /**


      * @enum WaitMode


      * @brief Enumeration to hold wait mode for IInferRequest


      */


     enum WaitMode : int64_t {


         /** Wait until inference result becomes available */


         RESULT_READY = -1,


         /** IInferRequest doesn't block or interrupt current thread and immediately returns inference status */


         STATUS_ONLY = 0,

     };

 

     using Ptr = std::shared_ptr<IInferRequest>;

     using WeakPtr = std::weak_ptr<IInferRequest>;

 


     /**


      * @brief Sets input/output data to infer


      * @note: Memory allocation does not happen


      * @param name Name of input or output blob.


      * @param data Reference to input or output blob. The type of a blob must match the network input precision and size.


      * @param resp Optional: pointer to an already allocated object to contain information in case of failure


      * @return Status code of the operation: OK (0) for success


      */


     virtual StatusCode SetBlob(const char *name, const Blob::Ptr &data, ResponseDesc *resp) noexcept = 0;

 


     /**


      * @brief Gets input/output data for inference


      * @note: Memory allocation does not happen


      * @param name Name of input or output blob.


      * @param data Reference to input or output blob. The type of Blob must match the network input precision and size.


      * @param resp Optional: pointer to an already allocated object to contain information in case of failure


      * @return Status code of the operation: OK (0) for success


      */


     virtual StatusCode GetBlob(const char *name, Blob::Ptr &data, ResponseDesc *resp) noexcept = 0;

 


     /**


      * @brief Infers specified input(s) in synchronous mode


      * @note blocks all methods of IInferRequest while request is ongoing (running or waiting in queue)


      * @param resp Optional: pointer to an already allocated object to contain information in case of failure


      * @return Status code of the operation: OK (0) for success


      */


     virtual StatusCode Infer(ResponseDesc *resp) noexcept = 0;

 


     /**


      * @brief Queries performance measures per layer to get feedback of what is the most time consuming layer


      * @note: not all plugins provide meaningful data


      * @param perfMap Map of layer names to profiling information for that layer


      * @param resp Optional: pointer to an already allocated object to contain information in case of failure


      * @return Status code of the operation: OK (0) for success


      */


     virtual StatusCode GetPerformanceCounts(std::map<std::string, InferenceEngineProfileInfo> &perfMap,

                                             ResponseDesc *resp) const noexcept = 0;

 


     /**


      * @brief Waits for the result to become available. Blocks until specified millis_timeout has elapsed or the result becomes available, whichever comes first.


      * @param millis_timeout Maximum duration in milliseconds to block for


      * @note There are special cases when millis_timeout is equal some value of the WaitMode enum:


      * * STATUS_ONLY - immediately returns inference status (IInferRequest::RequestStatus). It does not block or interrupt current thread


      * * RESULT_READY - waits until inference result becomes available


      * @param resp Optional: a pointer to an already allocated object to contain extra information of a failure (if occurred)


      * @return Enumeration of the resulted action: OK (0) for success


      */


     virtual InferenceEngine::StatusCode Wait(int64_t millis_timeout, ResponseDesc *resp) noexcept = 0;

 


     /**


      * @brief Starts inference of specified input(s) in asynchronous mode


      * @note: It returns immediately. Inference starts also immediately


      * @param resp Optional: a pointer to an already allocated object to contain extra information of a failure (if occurred)


      * @return Enumeration of the resulted action: OK (0) for success


      */


     virtual StatusCode StartAsync(ResponseDesc *resp) noexcept = 0;

 


     /**


      * @brief Completion callback definition as pointer to a function


      * @param context Pointer to request for providing context inside callback


      * @param code Completion result status: OK (0) for success


      */


     typedef void (*CompletionCallback)(InferenceEngine::IInferRequest::Ptr context,

                                        InferenceEngine::StatusCode code);

 


     /**


      * @brief Sets a callback function that will be called on success or failure of asynchronous request


      * @param callback A function to be called


      * @return Enumeration of the resulted action: OK (0) for success


      */


     virtual StatusCode SetCompletionCallback(CompletionCallback callback) noexcept = 0;

 


     /**


      * @brief Gets arbitrary data for the request and stores a pointer to a pointer to the obtained data


      * @param data Pointer to a pointer to the gotten arbitrary data


      * @param resp Optional: a pointer to an already allocated object to contain extra information of a failure (if occurred)


      * @return Enumeration of the resulted action: OK (0) for success


      */


     virtual StatusCode GetUserData(void **data, ResponseDesc *resp) noexcept = 0;

 


     /**


      * @brief Sets arbitrary data for the request


      * @param data Pointer to a pointer to arbitrary data to set


      * @param resp Optional: a pointer to an already allocated object to contain extra information of a failure (if occurred)


      * @return Enumeration of the resulted action: OK (0) for success


      */


     virtual StatusCode SetUserData(void *data, ResponseDesc *resp) noexcept = 0;

 


     /**


     * @brief Sets new batch size when dynamic batching is enabled in executable network that created this request.


     * @param batch_size new batch size to be used by all the following inference calls for this request.


     * @param resp Optional: a pointer to an already allocated object to contain extra information of a failure (if occurred)


     * @return Enumeration of the resulted action: OK (0) for success


     */


     virtual InferenceEngine::StatusCode SetBatch(int batch_size, ResponseDesc *resp) noexcept = 0;

 };

 

 }  // namespace InferenceEngine