Abnormal exit debugging method and device for heterogeneous computing, equipment, medium and product

By detecting device-side exceptions on the host side of the heterogeneous computing system, obtaining on-site information and generating an abnormal exit file, the problems of low efficiency of abnormal exit debugging and large equipment occupation in heterogeneous computing are solved, and rapid and resource-saving exceptions are achieved Exit debugging.

CN118964204APending Publication Date: 2024-11-15SUIYUAN INTELLIGENT TECH (CHENGDU) CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411118933.0
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-15
Publication Date
2024-11-15

AI Technical Summary

Technical Problem

The exception exit debugging method for existing heterogeneous computing is difficult to be effectively carried out in any scenario, especially in a time-consuming program, debugging error locations takes a long time and occupies a large amount of device resources in distributed computing applications.

Method used

When detecting abnormalities on the device side on the host side of the heterogeneous computing system, stop the calculation task, obtain the on-site information of the process and associated device side, generate an exception exit file, and use the file to perform abnormal exit debugging of heterogeneous computing to avoid reloading the program , shorten the debugging time and reduce equipment usage.

Benefits of technology

It realizes abnormal exit debugging in any scenario, avoids reloading the program, shortens debugging time, reduces device usage, and improves debugging efficiency.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118964204A_ABST
    Figure CN118964204A_ABST
Patent Text Reader

Abstract

The invention relates to the technical field of computers, and discloses an abnormal exit debugging method and device for heterogeneous computing, equipment, a medium and a product. The method comprises the steps that when it is detected that an equipment end of the heterogeneous computing system is abnormal, issuing of a computing task is stopped, and a process corresponding to the equipment end and at least one associated equipment end of the process are obtained; establishing an abnormal exit thread corresponding to each associated device end, and obtaining field information of the corresponding associated device end through each abnormal exit thread; and generating an abnormal exit file according to the field information of each associated equipment end, and performing abnormal exit debugging of heterogeneous calculation according to the abnormal exit file. According to the scheme of the embodiment, reloading of the program can be avoided, the time spent on abnormal quit debugging can be shortened, and equipment occupation can be avoided.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of computer technology, and in particular to a method, device, equipment, medium and product for debugging abnormal exit of heterogeneous computing. Background Art

[0002] With the popularization and development of artificial intelligence, the requirements for computing power are getting higher and higher. Heterogeneous computing with high parallel and high-density computing power has become an inevitable choice to support the development of artificial intelligence. The software stack of heterogeneous computing is complex, and the host side and the device side have their own division of labor. Once an exception occurs on the device side, the entire application will exit, and the difficulty of debugging the abnormal exit will be greatly increased.

[0003] At present, the existing abnormal exit debugging method of heterogeneous computing usually controls the failed heterogeneous computing application to be executed by the device side through the hardware debugging register exposed on the control bus, so as to find the error location of the program. However, in the existing technology, only the program that must have the problem can be reproduced, and the program that will fail only during the stress test is almost impossible to reproduce; secondly, the debugging of the error program requires reloading the program and gradually executing the calculation instructions. For the time-consuming program, it takes a long time to debug to the error location; moreover, when debugging the error program, it is necessary to use the host environment with the device side, which will occupy the device during debugging, especially the distributed computing application using multiple devices, which will occupy a lot of device resources. Summary of the invention

[0004] The present invention provides a method, device, equipment, medium and product for abnormal exit debugging of heterogeneous computing, which can realize abnormal exit debugging in any scenario, avoid reloading the program, shorten the time spent on abnormal exit debugging, and avoid equipment occupation.

[0005] According to one aspect of the present invention, a method for debugging abnormal exit of heterogeneous computing is provided, which is applied to a host side of a heterogeneous computing system, and includes:

[0006] When an abnormality is detected on a device end of the heterogeneous computing system, the computing task is stopped from being issued, and a process corresponding to the device end and at least one associated device end of the process are obtained;

[0007] Establishing an abnormal exit thread corresponding to each of the associated device ends, and obtaining the corresponding on-site information of the associated device end through each of the abnormal exit threads;

[0008] An abnormal exit file is generated according to the field information of each associated device end, and abnormal exit debugging of heterogeneous computing is performed according to the abnormal exit file.

[0009] According to another aspect of the present invention, there is provided an abnormal exit debugging device for heterogeneous computing, which is applied to a host end of a heterogeneous computing system, and includes:

[0010] A process acquisition module, configured to stop issuing computing tasks when an abnormality is detected on a device end of the heterogeneous computing system, and acquire a process corresponding to the device end and at least one associated device end of the process;

[0011] A site information acquisition module, used to establish an abnormal exit thread corresponding to each associated device end, and acquire the site information of the corresponding associated device end through each abnormal exit thread;

[0012] The abnormal exit debugging module is used to generate an abnormal exit file according to the field information of each associated device end, and perform abnormal exit debugging of heterogeneous computing according to the abnormal exit file.

[0013] According to another aspect of the present invention, there is provided an electronic device, the electronic device comprising:

[0014] at least one processor, and

[0015] a memory communicatively connected to the at least one processor; wherein,

[0016] The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the abnormal exit debugging method for heterogeneous computing described in any embodiment of the present invention.

[0017] According to another aspect of the present invention, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores a computer program, wherein the computer program is used to enable a processor to implement the abnormal exit debugging method for heterogeneous computing described in any embodiment of the present invention when executed.

[0018] According to another aspect of the present invention, a computer program product is provided, including a computer program, wherein when the computer program is executed by a processor, the abnormal exit debugging method for heterogeneous computing described in any embodiment of the present invention is implemented.

[0019] The technical solution of the embodiment of the present invention, through the host side of the heterogeneous computing system, when an exception is detected in the device side of the heterogeneous computing system, stops the sending of computing tasks, and obtains the process corresponding to the device side, and at least one associated device side of the process; establishes an exception exit thread corresponding to each associated device side, and obtains the corresponding field information of the associated device side through each exception exit thread; generates an exception exit file according to the field information of each associated device side, and performs abnormal exit debugging of heterogeneous computing according to the exception exit file; by concurrently collecting and saving field information of all device sides under the process when it is determined that an exception has occurred in any device side, and generating an exception exit file according to the field information, abnormal exit debugging in any scenario can be achieved, reloading of the program can be avoided, the time spent on abnormal exit debugging can be shortened, and device occupation can be avoided.

[0020] It should be understood that the contents described in this section are not intended to identify the key or important features of the embodiments of the present invention, nor are they intended to limit the scope of the present invention. Other features of the present invention will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0022] Figure 1 This is a flowchart of a heterogeneous computing abnormal exit debugging method provided in accordance with Embodiment 1 of the present invention;

[0023] Figure 2 This is a schematic diagram of a section of an abnormal exit file provided according to the first embodiment of the present invention;

[0024] Figure 3 This is a schematic diagram of a segment portion of an abnormal exit file provided according to the first embodiment of the present invention;

[0025] Figure 4 is a flowchart of another abnormal exit debugging method for heterogeneous computing provided according to the first embodiment of the present invention;

[0026] Figure 5 This is a schematic diagram of the structure of a heterogeneous computing abnormal exit debugging device provided in accordance with Embodiment 2 of the present invention;

[0027] Figure 6 It is a structural schematic diagram of an electronic device for implementing the abnormal exit debugging method for heterogeneous computing according to an embodiment of the present invention. DETAILED DESCRIPTION

[0028] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0029] It should be noted that the terms "first", "second", "target", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0030] Embodiment 1

[0031] Figure 1 A flowchart of a method for debugging abnormal exit of heterogeneous computing is provided for the first embodiment of the present invention. This embodiment is applicable to the case of performing on-site protection and recovery of abnormal exit of a device in a heterogeneous computing scenario. The method can be executed by an abnormal exit debugging device for heterogeneous computing. The method can be applied to the host side of a heterogeneous computing system. Typically, the host side can be an electronic device, such as a computer device or a server. Figure 1 As shown, the method includes:

[0032] S110. When an abnormality is detected on a device side of the heterogeneous computing system, stop issuing computing tasks, and obtain a process corresponding to the device side and at least one associated device side of the process.

[0033] Among them, the heterogeneous computing system may include a host end and multiple device ends. When executing heterogeneous computing tasks, the host end creates a process for heterogeneous computing and sends computing tasks to multiple device ends through the process. Then, after receiving the computing task, each device end can create multiple heterogeneous computing threads to control multiple computing units to jointly execute the computing task, and after the calculation is completed, the calculation results can be fed back to the host end. Finally, the host end integrates the calculation results fed back by each device end to obtain the final calculation result.

[0034] In this embodiment, when performing heterogeneous computing, if a device terminal has an operation exception, the exception information (e.g., exception code, exception description information, etc.) can be reported to the host terminal. When the host terminal detects the exception information reported by the device terminal, it can be known that the device terminal has an exception, and at this time, the computing task can be stopped from being issued, and the process to which the device terminal with the exception belongs and all the device terminals occupied by the process can be found according to the preset mapping relationship between the process and the device terminal, as the associated device terminals.

[0035] S120: Establish an abnormal exit thread corresponding to each associated device end, and obtain the corresponding on-site information of the associated device end through each abnormal exit thread.

[0036] Specifically, the host side can establish a corresponding thread for each associated device side, which is responsible for the abnormal exit of each device side. After the abnormal exit thread is established, each abnormal exit thread can be used to perform on-site storage for the corresponding associated device side. Typically, on-site information of the associated device side can be collected and saved. The on-site information can include basic device information of the device side, thread information of heterogeneous computing threads running in the device side, etc.

[0037] S130. Generate an abnormal exit file according to the on-site information of each associated device end, and perform abnormal exit debugging of heterogeneous computing according to the abnormal exit file.

[0038] The abnormal exit file may be based on the Executable and Linking Format (ELF). Specifically, the current abnormal exit file may be generated according to the standard ELF format with the field information of all associated device ends, typically a dev-core file.

[0039] In this embodiment, since the abnormal exit file conforms to the standard ELF format, the saved dev-core file can be directly used as a device file for debugging, and does not require a host environment with a heterogeneous computing system, and can be directly read using the readelf tool. Specifically, the dev-gdb tool can be used to directly load the dev-core file to read the scene information and reproduce the program error scene, thereby realizing the abnormal exit debugging of heterogeneous computing.

[0040] Optionally, generating an abnormal exit file according to the on-site information of each associated device terminal may include:

[0041] The process basic information corresponding to the process is obtained, and an abnormal exit file is generated according to the process basic information and the field information of each associated device end.

[0042] In this embodiment, while the field information of the device side is saved, the process basic information of the process can also be saved. Specifically, when the host side detects an abnormality on the device side, it can collect the process basic information of the process to which the device side belongs, such as the process number, running information, etc., and can generate the current abnormal exit file according to the standard ELF format with the process basic information and the field information of each associated device side.

[0043] In this embodiment, by saving the field data when an error occurs during program operation, it is possible to solve the problem in the prior art that the problem can only be reproduced stably on the device side, but the problem that the program that will fail only during stress testing cannot be reproduced, and the problem that debugging an error program requires reloading the program, and for a time-consuming program, it takes a long time to debug to the error location; secondly, by saving the field data as an abnormal exit file in a standard ELF format, the tool can directly read the required data from the abnormal exit file without occupying the device, which can solve the problem in the prior art that when debugging an error program, a host environment with heterogeneous computing devices is required, and the device will be occupied during debugging, especially distributed computing applications using multiple devices, which will occupy a large amount of device resources.

[0044] The technical solution of the embodiment of the present invention, through the host side of the heterogeneous computing system, when an exception is detected in the device side of the heterogeneous computing system, stops the sending of computing tasks, and obtains the process corresponding to the device side, and at least one associated device side of the process; establishes an exception exit thread corresponding to each associated device side, and obtains the corresponding field information of the associated device side through each exception exit thread; generates an exception exit file according to the field information of each associated device side, and performs abnormal exit debugging of heterogeneous computing according to the exception exit file; by concurrently collecting and saving field information of all device sides under the process when it is determined that an exception has occurred in any device side, and generating an exception exit file according to the field information, abnormal exit debugging in any scenario can be achieved, reloading of the program can be avoided, the time spent on abnormal exit debugging can be shortened, and device occupation can be avoided.

[0045] In an optional implementation of this embodiment, obtaining the corresponding on-site information of the associated device end through each abnormal exit thread may include:

[0046] Obtaining basic device information corresponding to each of the associated device ends through each of the abnormal exit threads, and obtaining each heterogeneous computing thread corresponding to the process in the corresponding associated device end;

[0047] Modifying the computing unit executing each of the heterogeneous computing threads to a pause state, and obtaining heterogeneous computing thread information corresponding to each of the heterogeneous computing threads;

[0048] The on-site information of each associated device end is acquired according to the device basic information corresponding to each associated device end and the heterogeneous computing thread information corresponding to each heterogeneous computing thread.

[0049] The heterogeneous computing thread information may include register information, stack information and / or storage information at all levels. In this embodiment, the field information of the associated device end may include basic device information and heterogeneous computing thread information.

[0050] Specifically, after establishing the abnormal exit thread corresponding to each associated device end, first, the basic device information corresponding to each associated device end, such as device type, device identification, manufacturer name, etc., can be collected through each abnormal exit thread; at the same time, all heterogeneous computing threads on each associated device end occupied by the process can be obtained through each abnormal exit thread. Then, the computing units of these heterogeneous computing threads can be suspended, and the thread information of each heterogeneous computing thread can be collected, such as register information, stack, storage at all levels, etc. Finally, the basic device information corresponding to each associated device end and the heterogeneous computing thread information corresponding to each heterogeneous computing thread can be saved together as field information. For example, it can be stored in the memory of the host end.

[0051] Correspondingly, in this embodiment, an abnormal exit file in a standard ELF format can be jointly generated based on the device basic information corresponding to each associated device end, the heterogeneous computing thread information corresponding to each heterogeneous computing thread, and the process basic information corresponding to the process.

[0052] Among them, the standard ELF format can occupy a PROGBITS segment for each block of continuous storage, and a PROGBITS segment for each block of continuous public registers. The private information of each heterogeneous computing thread is placed in a NOTE segment in the order of Elf64_Nhdr, name, and specific description (custom format), and the STRTAB segment containing each segment name string is generated to generate the corresponding section. Taking the abnormal exit file generated when an application using 8 heterogeneous computing threads exits as an example, the section part and segment part of the abnormal exit file can be respectively as follows: Figure 2 and 3 shown.

[0053] Optionally, obtaining heterogeneous computing thread information corresponding to each of the heterogeneous computing threads may include:

[0054] Determining whether each of the heterogeneous computing threads has ended normally;

[0055] If it is detected that the current heterogeneous computing thread does not end normally, the current heterogeneous computing thread is determined as an abnormal heterogeneous computing thread, and heterogeneous computing thread information corresponding to each abnormal heterogeneous computing thread is obtained.

[0056] In this embodiment, when collecting heterogeneous computing thread information, the heterogeneous computing threads can be screened first. Specifically, each heterogeneous computing thread is traversed to determine whether the thread ends normally. Typically, if it is detected that the heterogeneous computing thread has been executed, it is determined that the thread ends normally; and if it is detected that the heterogeneous computing thread is executing or has not yet started executing, it is determined that the thread does not end normally. In this way, multiple heterogeneous computing threads that did not end normally can be screened as abnormal heterogeneous computing threads, and the thread information of each abnormal heterogeneous computing thread can be collected.

[0057] Correspondingly, acquiring the on-site information of each associated device end according to the device basic information corresponding to each associated device end and the heterogeneous computing thread information corresponding to each heterogeneous computing thread may include:

[0058] The on-site information of each associated device end is acquired according to the basic device information corresponding to each associated device end and the heterogeneous computing thread information corresponding to each abnormal heterogeneous computing thread.

[0059] Specifically, the basic device information corresponding to the currently associated device end and the heterogeneous computing thread information corresponding to each abnormal heterogeneous computing thread in the currently associated device end can be used together as the field information of the currently associated device end.

[0060] In a specific implementation of this embodiment, the process of the abnormal exit debugging method for heterogeneous computing can be as follows: Figure 4 As shown. First, when an exception is detected in the process, determine whether the Coredump (abnormal exit mechanism) is enabled. If so, perform Coredump on each associated device in parallel; then, save key process information, such as device type number, process number, etc., obtain all heterogeneous computing threads occupied by the process, and suspend the computing units of these threads; further, traverse each heterogeneous computing thread and determine whether the thread ends normally; if it is determined that the thread does not end normally, export and save heterogeneous computing thread information, such as registers, stacks, and storage at all levels; finally, a core file in ELF format can be generated based on the saved key process information and heterogeneous computing thread information. After the core file is generated, the process is controlled to exit and the abnormal scene is cleaned up.

[0061] Embodiment 2

[0062] Figure 5This is a schematic diagram of the structure of a heterogeneous computing abnormal exit debugging device provided in the second embodiment of the present invention. Figure 5 As shown, the device can be applied to the host side of a heterogeneous computing system, including: a process acquisition module 210, a field information acquisition module 220 and an abnormal exit debugging module 230; wherein,

[0063] The process acquisition module 210 is used to stop issuing computing tasks when an abnormality is detected on a device end of the heterogeneous computing system, and to acquire a process corresponding to the device end and at least one associated device end of the process;

[0064] The on-site information acquisition module 220 is used to establish an abnormal exit thread corresponding to each of the associated device ends, and obtain the on-site information of the corresponding associated device end through each of the abnormal exit threads;

[0065] The abnormal exit debugging module 230 is used to generate an abnormal exit file according to the field information of each associated device end, and perform abnormal exit debugging of heterogeneous computing according to the abnormal exit file.

[0066] The technical solution of the embodiment of the present invention, through the host side of the heterogeneous computing system, when an exception is detected in the device side of the heterogeneous computing system, stops the sending of computing tasks, and obtains the process corresponding to the device side, and at least one associated device side of the process; establishes an exception exit thread corresponding to each associated device side, and obtains the corresponding field information of the associated device side through each exception exit thread; generates an exception exit file according to the field information of each associated device side, and performs abnormal exit debugging of heterogeneous computing according to the exception exit file; by concurrently collecting and saving field information of all device sides under the process when it is determined that an exception has occurred in any device side, and generating an exception exit file according to the field information, abnormal exit debugging in any scenario can be achieved, reloading of the program can be avoided, the time spent on abnormal exit debugging can be shortened, and device occupation can be avoided.

[0067] Optionally, the on-site information acquisition module 220 is specifically used to obtain the basic device information corresponding to each of the associated device ends through each of the abnormal exit threads, and to obtain each heterogeneous computing thread corresponding to the process in the corresponding associated device end;

[0068] Modifying the computing unit executing each of the heterogeneous computing threads to a pause state, and obtaining heterogeneous computing thread information corresponding to each of the heterogeneous computing threads;

[0069] The on-site information of each associated device end is acquired according to the device basic information corresponding to each associated device end and the heterogeneous computing thread information corresponding to each heterogeneous computing thread.

[0070] Optionally, the on-site information acquisition module 220 is specifically used to determine whether each of the heterogeneous computing threads has ended normally;

[0071] If it is detected that the current heterogeneous computing thread has not ended normally, the current heterogeneous computing thread is determined as an abnormal heterogeneous computing thread, and heterogeneous computing thread information corresponding to each abnormal heterogeneous computing thread is obtained;

[0072] The on-site information of each associated device end is acquired according to the basic device information corresponding to each associated device end and the heterogeneous computing thread information corresponding to each abnormal heterogeneous computing thread.

[0073] Optionally, the heterogeneous computing thread information includes register information, stack information and / or storage information at each level.

[0074] Optionally, the abnormal exit debugging module 230 is specifically used to obtain the process basic information corresponding to the process, and generate an abnormal exit file according to the process basic information and the field information of each associated device end.

[0075] Optionally, the abnormal exit file is based on the executable linkable format.

[0076] The abnormal exit debugging device for heterogeneous computing provided by the embodiment of the present invention can execute the abnormal exit debugging method for heterogeneous computing provided by any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0077] In the technical solution of the present disclosure, the collection, storage, use, processing, transmission, provision and disclosure of user personal information involved are in compliance with the provisions of relevant laws and regulations and do not violate public order and good morals.

[0078] Embodiment 3

[0079] Figure 6 A schematic diagram of the structure of an electronic device 30 that can be used to implement an embodiment of the present invention is shown. The electronic device 30 is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device 30 can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices (such as helmets, glasses, watches, etc.) and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present invention described and / or required herein.

[0080] like Figure 6As shown, the electronic device 30 includes at least one processor 31, and a memory connected to the at least one processor 31, such as a read-only memory (ROM) 32, a random access memory (RAM) 33, etc., wherein the memory stores a computer program that can be executed by at least one processor, and the processor 31 can perform various appropriate actions and processes according to the computer program stored in the read-only memory 32 or the computer program loaded from the storage unit 38 to the random access memory 33. In the RAM 33, various programs and data required for the operation of the electronic device 30 can also be stored. The processor 31, the ROM 32 and the RAM 33 are connected to each other through a bus 34. The input / output (I / O) interface 35 is also connected to the bus 34.

[0081] A number of components in the electronic device 30 are connected to the I / O interface 35, including: an input unit 36, such as a keyboard, a mouse, etc.; an output unit 37, such as various types of displays, speakers, etc.; a storage unit 38, such as a disk, an optical disk, etc.; and a communication unit 39, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 39 allows the electronic device 30 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0082] The processor 31 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the processor 31 include, but are not limited to, a central processing unit, a graphics processing unit, various special artificial intelligence computing chips, various processors running machine learning model algorithms, digital signal processors, and any appropriate processors, controllers, microcontrollers, etc. The processor 31 executes the various methods and processes described above, such as the abnormal exit debugging method of heterogeneous computing.

[0083] In some embodiments, the abnormal exit debugging method of heterogeneous computing can be implemented as a computer program, which is tangibly contained in a computer-readable storage medium, such as a storage unit 38. In some embodiments, part or all of the computer program can be loaded and / or installed on the electronic device 30 via the ROM 32 and / or the communication unit 39. When the computer program is loaded into the RAM 33 and executed by the processor 31, one or more steps of the abnormal exit debugging method of heterogeneous computing described above can be performed. Alternatively, in other embodiments, the processor 31 can be configured to execute the abnormal exit debugging method of heterogeneous computing in any other appropriate manner (for example, by means of firmware).

[0084] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays, application specific integrated circuits, application specific standard products, systems on a chip, load programmable logic devices, computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0085] Computer programs for implementing the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that when the computer program is executed by the processor, the functions / operations specified in the flow chart and / or block diagram are implemented. The computer program may be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.

[0086] In the context of the present invention, a computer-readable storage medium may be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, device, or equipment. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices, or equipment, or any suitable combination of the foregoing. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory, a read-only memory, an erasable programmable read-only memory, an optical fiber, a portable compact disk read-only memory, an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0087] To provide interaction with a user, the systems and techniques described herein may be implemented on an electronic device 30 having: a display device (e.g., a cathode ray tube or a liquid crystal display) for displaying information to the user; and a keyboard and a pointing device (e.g., a mouse or a trackball) through which the user can provide input to the electronic device 30. Other types of devices may also be used to provide interaction with the user; for example, the feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form (including acoustic input, voice input, or tactile input).

[0088] The systems and techniques described herein may be implemented in a computing system that includes backend components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes frontend components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such backend components, middleware components, or frontend components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: local area networks, wide area networks, blockchain networks, and the Internet.

[0089] A computing system may include clients and servers. Clients and servers are generally remote from each other and typically interact through a communication network. The relationship of client and server is generated by computer programs running on respective computers and having a client-server relationship to each other. The server may be a cloud server.

[0090] This embodiment may also include a computer program product, which includes a computer program. When the computer program is executed by a processor, it implements the abnormal exit debugging method for heterogeneous computing provided by any embodiment of the present invention.

[0091] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps described in the present invention can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solution of the present invention can be achieved, and this document does not limit this.

[0092] The above specific implementations do not constitute a limitation on the protection scope of the present invention. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for debugging abnormal exit of heterogeneous computing. It is characterized in that Applied to the host side of heterogeneous computing systems, including: When an abnormality is detected on a device end of the heterogeneous computing system, the computing task is stopped from being issued, and a process corresponding to the device end and at least one associated device end of the process are obtained; Establishing an abnormal exit thread corresponding to each of the associated device ends, and obtaining the corresponding on-site information of the associated device end through each of the abnormal exit threads; An abnormal exit file is generated according to the field information of each associated device end, and abnormal exit debugging of heterogeneous computing is performed according to the abnormal exit file.

2. The method according to claim 1, It is characterized in that By means of each of the abnormal exit threads, the corresponding on-site information of the associated device end is obtained, including: Obtaining basic device information corresponding to each of the associated device ends through each of the abnormal exit threads, and obtaining each heterogeneous computing thread corresponding to the process in the corresponding associated device end; Modifying the computing unit executing each of the heterogeneous computing threads to a pause state, and obtaining heterogeneous computing thread information corresponding to each of the heterogeneous computing threads; The on-site information of each associated device end is acquired according to the device basic information corresponding to each associated device end and the heterogeneous computing thread information corresponding to each heterogeneous computing thread.

3. The method according to claim 2, It is characterized in that Obtaining heterogeneous computing thread information corresponding to each of the heterogeneous computing threads includes: Determining whether each of the heterogeneous computing threads has ended normally; If it is detected that the current heterogeneous computing thread has not ended normally, the current heterogeneous computing thread is determined as an abnormal heterogeneous computing thread, and heterogeneous computing thread information corresponding to each abnormal heterogeneous computing thread is obtained; According to the basic device information corresponding to each of the associated device ends and the heterogeneous computing thread information corresponding to each of the heterogeneous computing threads, the on-site information of each of the associated device ends is obtained, including: The on-site information of each associated device end is acquired according to the basic device information corresponding to each associated device end and the heterogeneous computing thread information corresponding to each abnormal heterogeneous computing thread.

4. The method according to claim 2 or 3, It is characterized in that Heterogeneous computing thread information includes register information, stack information and / or storage information at all levels.

5. The method according to claim 1, It is characterized in that Generate an abnormal exit file based on the on-site information of each associated device end, including: The process basic information corresponding to the process is obtained, and an abnormal exit file is generated according to the process basic information and the field information of each associated device end.

6. The method according to claim 1 or 5, It is characterized in that Abnormal exit files are based on the executable linkable format.

7. An abnormal exit debugging device for heterogeneous computing, It is characterized in that Applied to the host side of heterogeneous computing systems, including: A process acquisition module, configured to stop issuing computing tasks when an abnormality is detected on a device end of the heterogeneous computing system, and acquire a process corresponding to the device end and at least one associated device end of the process; A site information acquisition module, used to establish an abnormal exit thread corresponding to each associated device end, and acquire the site information of the corresponding associated device end through each abnormal exit thread; The abnormal exit debugging module is used to generate an abnormal exit file according to the field information of each associated device end, and perform abnormal exit debugging of heterogeneous computing according to the abnormal exit file.

8. An electronic device, It is characterized in that The electronic device comprises: at least one processor, and a memory communicatively connected to the at least one processor; wherein, The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the abnormal exit debugging method for heterogeneous computing according to any one of claims 1 to 6.

9. A computer-readable storage medium, It is characterized in that The computer-readable storage medium stores a computer program, and the computer program is used to enable a processor to implement the abnormal exit debugging method for heterogeneous computing according to any one of claims 1 to 6 when executed.

10. A computer program product, It is characterized in that The invention comprises a computer program, which, when executed by a processor, implements the abnormal exit debugging method for heterogeneous computing according to any one of claims 1 to 6.