System and method for dynamically loading PCIE slave device based on FPGA, electronic device and storage medium
By using the AURORA bus and AXI4 module in the FPGA system for information forwarding and parsing, and combining with the firewall to maintain the communication state, the time-consuming and low efficiency problems of FPGA dynamic loading are solved, and efficient CPU-to-FPGA communication is achieved.
Patent Information
- Application Number
- CN202411106048.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-13
- Publication Date
- 2024-12-03
AI Technical Summary
In the prior art, dynamic loading of FPGAs as CPU PCIE slave devices takes time and is inefficient, and restarting the operating system to restore connections will consume a lot of resources, affecting system performance.
Two FPGA systems based on AURORA bus communication are adopted. One FPGA saves the PCIE interface configuration and connects to the CPU. The other FPGA supports dynamic loading of function service modules. Information forwarding and analysis is realized through AXI4FULL and AXI4LITE modules to avoid restarting the operating system and through The firewall maintains communication status.
It realizes efficient communication during dynamic loading of FPGA, avoids loss of PCIE interface configuration, reduces the time-consuming system restart, improves efficiency, and ensures the normal communication status between the CPU and the FPGA.
Smart Images

Figure CN119066026A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of embedded technology, and in particular to a system, method, electronic device and storage medium for dynamically loading a PCIE slave device based on FPGA. Background Art
[0002] FPGA development and application are developing rapidly. With its advantages such as high flexibility, short development cycle, and high parallel computing efficiency, FPGA is being applied to more and more fields, forming an embedded platform together with DSP, CPU, GPU, MCU and other chips.
[0003] When the FPGA in an embedded system is used as a PCIE slave device, it needs to be prepared before the CPU starts so that the CPU can scan the FPGA and work with it when it starts. Due to FPGA resource limitations, the corresponding firmware in the FPGA is sometimes reloaded when it is needed during development. Reloading will clear the original configuration in the FPGA, including the configuration of the PCIE interface, causing the CPU to be unable to connect to the FPGA.
[0004] In the prior art, after the FPGA configuration is reloaded, the PCIE connection between the CPU and the FPGA can be reestablished by restarting the operating system, but restarting the operating system takes minutes, which is inefficient and time-consuming. Another partial reloading method is to retain part of the PCIE configuration and reload other configurations, but this method consumes a lot of FPGA resources, resulting in a significant reduction in the available resources for functional services, and even unable to guarantee normal operation. Summary of the invention
[0005] In view of the above analysis, the embodiments of the present invention aim to provide a system, method, electronic device and storage medium for dynamically loading PCIE slave devices based on FPGA, so as to solve the problem that the existing FPGA dynamic loading as a CPU PCIE slave device is time-consuming and inefficient.
[0006] On the one hand, an embodiment of the present invention provides a system for dynamically loading a PCIE slave device based on an FPGA, the system comprising a first FPGA and a second FPGA based on AURORA bus communication, wherein:
[0007] The first FPGA is used to save the PCIE interface configuration and connect to the CPU through the PCIE interface, and forward information between the CPU and the second FPGA;
[0008] The second FPGA includes a functional service module and supports dynamic loading of configuration of the functional service module.
[0009] The beneficial effects of the above technical solution are as follows: the first FPGA is connected to the CPU through the PCIE interface, the second FPGA runs the functional business module and supports dynamic loading, the first FPGA is connected to the second FPGA through the AURORA bus, and the information is forwarded between the CPU and the second FPGA. This solution not only meets the configuration needs of the second FPGA for dynamically loading the functional business module, but also avoids restarting the operating system by connecting the first FPGA to the CPU through the PCIE interface, so that the PCIE interface configuration will not be cleared. The use of the AURORA bus for communication between the first FPGA and the second FPGA also guarantees the communication rate, solving the problem of long time and low efficiency when the CPU and the FPGA as a PCIE slave device are dynamically loaded.
[0010] Based on the further improvement of the above system, an AXI4FULL module and an AXI4LITE module are provided between the PCIE interface of the first FPGA and the AURORA bus, wherein:
[0011] The AXI4FULL module of the first FPGA is used to parse and output data from the information sent by the PCIE interface or the AURORA bus;
[0012] The AXI4LITE module of the first FPGA is used to parse and output the configuration from the information sent by the PCIE interface or the AURORA bus.
[0013] The beneficial effect of the above further improvement scheme is that the AXI4FULL module and AXI4LITE module set in the first FPGA meet the parsing and transmission rate requirements of data transmitted between the PCIE interface and the AURORA bus.
[0014] Based on the further improvement of the above system, the second FPGA is provided with an AXI4FULL module and an AXI4LITE module between the AURORA bus and the functional service module, wherein:
[0015] The AXI4FULL module of the second FPGA is used to parse and output data from the information sent by the AURORA bus or the functional service module;
[0016] The AXI4LITE module is used to parse and output the configuration from the information sent by the AURORA bus and the functional business module.
[0017] The beneficial effect of the above further improvement scheme is that the AXI4FULL module and AXI4LITE module set in the second FPGA meet the parsing and transmission rate requirements of data transmitted between the AURORA bus and the functional business module.
[0018] Based on the further improvement of the above system, a firewall for generating analog communication signals is provided between the AXI4FULL module, the AXI4LITE module of the first FPGA and the AURORA bus.
[0019] The beneficial effect of the above further improved solution is that when the second FPGA is dynamically loaded, the firewall generates an analog communication signal to maintain communication with the first FPGA, thereby maintaining a normal communication state between the first FPGA and the CPU.
[0020] Based on further improvements of the above system, when the second FPGA is dynamically loaded, the firewall simulates the second FPGA and maintains the communication state with the first FPGA until the second FPGA completes the dynamic loading and then the firewall stops running.
[0021] The beneficial effect of the above further improvement scheme is: ensuring that the communication status of the first FPGA and CPU is normal during the dynamic loading of the second FPGA, avoiding the time-consuming and low-efficiency problem caused by system restart due to abnormal communication status of the first FPGA and CPU.
[0022] Based on further improvements of the above system, the AXI4FULL module and AXI4LITE module of the first FPGA and the second FPGA both include a write address channel, a write data channel, a write response channel, a read address channel, and a read response channel, and each channel of the first FPGA and the second FPGA is set with a valid bit.
[0023] The beneficial effects of the above further improved solution are: data transmission is realized in different channels, and flow control of data transmission is achieved through effective bit setting.
[0024] Based on further improvements of the above system, the read response channel of the first FPGA, the write address channel, the write data channel and the read address channel of the second FPGA are all provided with FIFOs for communication feedback and data flow control.
[0025] The beneficial effect of the above further improved solution is: it is used to solve the problem that when the first FPGA and the second FPGA communicate through the AURORA bus, the data receiving end cannot directly provide status feedback and flow control to the data sending end through the AURORA bus.
[0026] On the other hand, an embodiment of the present invention provides a PCIE slave device communication method supporting dynamic loading, the method specifically comprising:
[0027] The data sent by the first FPGA or the second FPGA as a sending end through the corresponding channel of the AXI4FULL module of the sending end and the configuration information sent by the AXI4LITE module of the sending end are input into the AURORA bus;
[0028] The AURORA bus integrates the data and configuration information into a data packet and sends it to the data receiving end;
[0029] The data packet is parsed by the AXI4FULL module of the receiving end to obtain data, and the AXI4LITE module of the receiving end parses the configuration information and stores them in corresponding FIFOs respectively;
[0030] Determine whether the FIFO capacity corresponding to the AXI4FULL module or the AXI4LITE module of the receiving end reaches a preset threshold, and if so, the sending end stops sending data;
[0031] The FIFOs corresponding to the AXI4FULL module and the AXI4LITE module at the receiving end forward the received data to the channels corresponding to the AXI4FULL module and the AXI4LITE module at the receiving end in sequence.
[0032] The beneficial effects of the above technical solution are as follows: by setting AXI4FULL and AXI4LITE modules in both the first FPGA and the second FPGA, the first FPGA and the second FPGA are connected and communicated through the AURORA bus, so that the information sent by the CPU through PCIE can be parsed and sent to the second FPGA through the corresponding channels of the AXI4FULL module and the AXI4LITE module of the first FPGA. The CPU can also read the complete data from the second FPGA received by the first FPGA through PCIE; and by setting FIFO after the AXI4FULL module and the AXI4LITE module at the receiving end, and controlling the sending end to stop or start sending data by judging whether the FIFO capacity reaches a preset threshold, the problem that the receiving end cannot perform status feedback and flow control with the sending end when the AURORA bus is used as the communication bus between the first FPGA and the second FPGA is solved.
[0033] An embodiment of the present invention also provides an electronic device, which includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, the steps of a PCIE slave device communication method supporting dynamic loading as provided in an embodiment of the present invention are implemented.
[0034] The beneficial effects of the above technical solution are as follows: the electronic device can dynamically load the functional services in the FPGA without restarting the operating system, which solves the problem of long time consumption and low efficiency caused by restarting the operating system when dynamically loading the FPGA in the prior art.
[0035] An embodiment of the present invention further provides a non-transitory computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the steps of the PCIE slave device communication method supporting dynamic loading provided by an embodiment of the present invention are implemented.
[0036] The beneficial effects of the above technical solution are as follows: the computer program stored in the storage medium can realize high-speed data communication between the CPU and the second FPGA through the first FPGA, and when the second FPGA is dynamically loaded, there is no need to restart the operating system, which solves the problem of long time and low efficiency caused by restarting the operating system when dynamically loading the FPGA in the prior art.
[0037] In the present invention, the above-mentioned technical solutions can also be combined with each other to achieve more preferred combination solutions. Other features and advantages of the present invention will be described in the subsequent description, and some advantages can become obvious from the description, or can be understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained through the contents particularly pointed out in the description and the drawings. BRIEF DESCRIPTION OF THE DRAWINGS
[0038] The accompanying drawings are only used for the purpose of illustrating specific embodiments and are not to be considered as limiting the present invention. In the entire drawings, the same reference symbols represent the same components;
[0039] Figure 1 This is a schematic diagram of the system structure of Example 1 of the present invention.
[0040] Figure 2 This is a data flow diagram of the method in Example 2 of the present invention. DETAILED DESCRIPTION
[0041] The preferred embodiments of the present invention are described in detail below in conjunction with the accompanying drawings, wherein the accompanying drawings constitute a part of this application and are used together with the embodiments of the present invention to illustrate the principles of the present invention, but are not used to limit the scope of the present invention.
[0042] Embodiment 1:
[0043] A specific embodiment of the present invention discloses a system for dynamically loading PCIE slave devices based on FPGA, such as Figure 1 shown.
[0044] The system includes a first FPGA and a second FPGA communicating based on an AURORA bus, wherein:
[0045] The first FPGA is used to save the PCIE interface configuration and connect to the CPU through the PCIE interface, and forward information between the CPU and the second FPGA;
[0046] The second FPGA includes a functional service module and supports dynamic loading of configuration of the functional service module.
[0047] Specifically, the first FPGA saves the PCIE interface configuration and connects to the CPU, keeping the CPU able to read the PCIE slave device and communicate; the second FPGA includes a functional business module and supports dynamic loading; the first FPGA and the second FPGA communicate via the AURORA bus.
[0048] The communication process between the CPU and the PCIE slave device is actively initiated by the CPU, and specifically includes two types of processes: reading and writing. Among them, the writing is performed by the CPU sending information to the first FGPA through the PCIE interface. Generally, the sent information includes signaling and / or data, corresponding to the corresponding functional business that the CPU needs to process and execute as the FPGA of the PCIE slave device.
[0049] When the CPU transmits information to the second FPGA through the first FPGA, it is necessary to ensure that the first FPGA and the second FPGA can communicate with each other and meet the communication rate and data integrity index requirements.
[0050] Furthermore, an AXI4FULL module and an AXI4LITE module are provided between the PCIE interface of the first FPGA and the AURORA bus, wherein:
[0051] The AXI4FULL module of the first FPGA is used to parse and output data from the information sent by the PCIE interface or the AURORA bus;
[0052] The AXI4LITE module of the first FPGA is used to parse and output the configuration from the information sent by the PCIE interface or the AURORA bus.
[0053] Furthermore, the second FPGA is provided with an AXI4FULL module and an AXI4LITE module between the AURORA bus and the functional service module, wherein:
[0054] The AXI4FULL module of the second FPGA is used to parse and output data from the information sent by the AURORA bus or the functional service module;
[0055] The AXI4LITE module is used to parse and output the configuration from the information sent by the AURORA bus and the functional business module.
[0056] AXI4 is a bus standard for FPGAs that supports high-speed data transmission and channel separation, and can meet the application needs of different specific scenarios.
[0057] In the embodiment of the present invention, an AXI4FULL module and an AXI4LITE module are provided in both the first FPGA and the second FPGA. The channel separation technology based on the AXI4 bus standard transmits the received information through the AXI4FULL full-rate channel, transmits the configuration through the AXI4LITE low-rate channel, and then sends it out in the form of a complete data packet, thereby ensuring the rate and integrity of information transmission.
[0058] Exemplarily, when the CPU sends data to the first FPGA through the PCIE interface, the AXI4FULL module of the first FPGA decodes and outputs the data from the information transmitted from the PCIE interface, and the AXI4LITE module of the first FPGA decodes and outputs the configuration from the information transmitted from the PCIE interface; the output data and configuration are then integrated into a complete data packet, which is transmitted from the first FPGA to the second FPGA through the AURORA bus. The AXI4FULL module of the second FPGA decodes and outputs the data from the complete data packet transmitted from the AURORA interface, and the AXI4LITE module of the second FPGA decodes and outputs the configuration from the complete data packet transmitted from the AURORA interface; the output data and configuration are synchronously output to the functional service module.
[0059] Furthermore, the AXI4FULL module and AXI4LITE module of the first FPGA and the second FPGA include a write address channel, a write data channel, a write response channel, a read address channel, and a read response channel, and each channel of the first FPGA and the second FPGA is set with a valid bit.
[0060] Specifically, according to the AXI4 protocol definition, the AXI4FULL module and the AXI4LITE module each include 5 channels, corresponding to 3 write operations and 2 read operations respectively; based on the information sent by the CPU to the first FPGA through the PCIE interface or the data packet sent by the second FPGA to the first FPGA through the AURORA bus corresponds to one of the above 5 channels, therefore, the data parsed by the AXI4FULL module of the first FPGA and the configuration information parsed by the AXI4LITE will be synchronously transmitted through their respective corresponding channels, ensuring the correctness and integrity of the information. Similarly, the AXI4FULL module and AXI4LITE module of the second FPGA support the same processing for the data packets sent by the first FPGA through the AURORA bus or the data packets generated by the functional module.
[0061] The first FPGA and the second FPGA communicate via the AURORA bus, which has a set of 512-bit wide data interfaces shared by channels of the AXI4FULL modules and AXI4LITE modules of the first FPGA and the second FPGA, and controls the switching logic and timing of the AURORA bus data interface shared by each channel by setting a valid bit for each channel, thereby achieving maximum efficient use of the AURORA bus data interface.
[0062] When the second FPGA is dynamically loaded, the communication with the first FPGA will be temporarily disconnected, and corresponding processing is required to maintain the communication status of the first FPGA to avoid abnormal communication with the CPU due to abnormality of the first FPGA and the need to restart the system.
[0063] Specifically, a firewall for generating analog communication signals is provided between the AXI4FULL module and the AXI4LITE module of the first FPGA and the AURORA bus.
[0064] When the second FPGA is dynamically loaded, the firewall simulates the second FPGA and maintains the communication state with the first FPGA until the second FPGA completes the dynamic loading and then the firewall stops running.
[0065] Preferably, by detecting the second FPGA, when the second FPGA does not respond after timeout, the firewall is immediately started to generate an analog signal to maintain the communication status with the first FPGA by setting a status code; when the second FPGA completes dynamic loading and is detected, the status code is reset, the firewall stops, and the second FPGA resumes communication with the first FPGA.
[0066] The AURORA bus is a high-speed serial bus protocol used for communication between FPGAs, and is used for protocol conversion and data transmission when the first FPGA and the second FPGA communicate. However, the AURORA bus does not have a "handshake" signal for the data receiving end, which means that when receiving data, the data receiving end and the data sending end cannot achieve a "handshake", which will cause the loss of sent data.
[0067] To solve this problem, the read response channel of the first FPGA, the write address channel, the write data channel and the read address channel of the second FPGA are all provided with FIFOs for communication feedback and data flow control.
[0068] This embodiment uses two FPGAs to replace the solution of using only a single FPGA in the prior art, wherein the first FPGA is connected to the CPU through the PCIE interface as a PCIE slave device, and supports information transmission between the CPU and the second FPGA; the second FPGA supports dynamic loading, includes a functional business module, and communicates with the first FPGA through the AURORA bus; to ensure the communication between the CPU, the first FPGA, and the second FPGA, the first FPGA and the second FPGA are both provided with an AXI4FULL module and an AXI4LITE module for data parsing and protocol conversion, and a firewall for simulating the second FPGA is set between the AXI4FULL and AXI4LITE modules of the first FPGA and the AURORA bus to prevent the first FPGA from having an abnormal state during the dynamic loading of the second FPGA; the AXI4FULL module and the AXI4LITE module of the first FPGA and the second FPGA each include 5 channels, a valid bit is set for each channel, and a FIFO is set in the corresponding data receiving channel to adapt to the communication mechanism of the AURORA bus to achieve communication feedback and data flow control. Compared with the prior art, the technical solution of this embodiment solves the time-consuming and inefficient problem of dynamically loading FPGA causing loss of PCIE configuration and having to restart the operating system, and the information transmission rate and integrity meet the index requirements.
[0069] Embodiment 2:
[0070] The second specific embodiment of the present invention discloses a PCIE slave device communication method supporting dynamic loading. Embodiment 2 includes all the technical features of Embodiment 1, and the method specifically includes:
[0071] The data sent by the first FPGA or the second FPGA as a sending end through the corresponding channel of the AXI4FULL module of the sending end and the configuration information sent by the AXI4LITE module of the sending end are input into the AURORA bus;
[0072] The AURORA bus integrates the data and configuration information into a data packet and sends it to the data receiving end;
[0073] The data packet is parsed by the AXI4FULL module of the receiving end to obtain data, and the AXI4LITE module of the receiving end parses the configuration information and stores them in corresponding FIFOs respectively;
[0074] Determine whether the FIFO capacity corresponding to the AXI4FULL module or the AXI4LITE module of the receiving end reaches a preset threshold, and if so, the sending end stops sending data;
[0075] The FIFOs corresponding to the AXI4FULL module and the AXI4LITE module at the receiving end forward the received data to the channels corresponding to the AXI4FULL module and the AXI4LITE module at the receiving end in sequence.
[0076] Figure 2 A data flow diagram of the method of Example 2 is shown, wherein AXI4 represents AXI4FULL or AXI4LITE.
[0077] In the system disclosed in Example 1, data transparent transmission is adopted to transmit the information of the CPU through the PCIE interface to the functional service module of the second FPGA.
[0078] Specifically, each channel of the AXI4FULL module and the AXI4LITE module set in the first FPGA and the second FPGA connects the AURORA bus of the first FPGA and the second FPGA, forming an information transmission channel between the PCIE interface and the functional modules included in the second FPGA. According to the AXI4 bus protocol, AXI4FULL and AXI4LITE each include 5 channels, and each channel has valid and ready handshake signals; the AURORA bus has only 5 signals, including: tx_data, tx_valid, tx_ready, rx_data, rx_valid, which means that the AURORA bus has a ready handshake signal when sending data, but no handshake signal when receiving data, so the data must be cached immediately, otherwise the data will be lost.
[0079] Correspondingly, FIFOs are set in the read response channel of the first FPGA, the write address channel, the write data channel, and the read address channel of the second FPGA.
[0080] By default, when the valid bits of each channel of the first FPGA and the second FPGA are set to valid, and the signals corresponding to the AURORA bus are valid at the same time, the negative feedback handshake and flow control of the data receiving end can be implemented by the method disclosed in this embodiment.
[0081] The AXI4FULL and AXI4LITE modules of the first FPGA and the second FPGA do not parse the content and address of the transmitted data, and only complete the protocol conversion and transceiver of the information transmitted between the PICE interface and the AURORA bus through the timing of the AXI4FULL and AXI4LITE of the first FPGA parsing the transmitted information. The AXI4FULL and AXI4LITE of the second FPGA parse the timing of the transmitted information to complete the protocol conversion and transceiver of the information transmitted between the AURORA bus and the functional business module, and then realize the information transmission between the first FPGA and the second FPGA through the AURORA bus, thereby realizing the transparent transmission of information from the CPU through the PCIE interface to the functional business module of the second FPGA. Compared with the prior art, it not only solves the problem of long time consumption and low efficiency caused by the restart of the operating system due to dynamic loading of the FPGA as a PCIE slave device, but also ensures that the data transmission rate and integrity meet the index requirements.
[0082] Embodiment 3:
[0083] The third embodiment of the present invention discloses an electronic device, which includes all the technical features of Embodiment 1 and Embodiment 2. The electronic device includes a memory, a processor, and a computer program stored in the memory and executable on the processor. The processor includes the steps of implementing the PCIE slave device communication method supporting dynamic loading as disclosed in Embodiment 2 when the CPU, the first FPGA, and the second FPGA execute the program.
[0084] Embodiment 4:
[0085] The fourth embodiment of the present invention discloses a non-transitory computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the steps of the PCIE slave device communication method supporting dynamic loading disclosed in Embodiment 2 are implemented.
[0086] Those skilled in the art will appreciate that all or part of the processes of the above-mentioned embodiments can be implemented by instructing related hardware through a computer program, and the program can be stored in a computer-readable storage medium, wherein the computer-readable storage medium is a disk, an optical disk, a read-only storage memory, or a random access memory, etc.
[0087] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by any technician familiar with the technical field within the technical scope disclosed by the present invention should be covered within the protection scope of the present invention.
Claims
1. A system for dynamically loading PCIE slave devices based on FPGA, characterized in that: The system includes a first FPGA and a second FPGA communicating based on an AURORA bus, wherein: The first FPGA is used to save the PCIE interface configuration and connect to the CPU through the PCIE interface, and forward information between the CPU and the second FPGA; The second FPGA includes a functional service module and supports dynamic loading of configuration of the functional service module.
2. A system for dynamically loading PCIE slave devices based on FPGA according to claim 1, characterized in that: An AXI4FULLFULL module and an AXI4LITE module are provided between the PCIE interface of the first FPGA and the AURORA bus, wherein: The AXI4FULL module of the first FPGA is used to parse and output data from the information sent by the PCIE interface or the AURORA bus; The AXI4LITE module of the first FPGA is used to parse and output the configuration from the information sent by the PCIE interface or the AURORA bus.
3. The system for dynamically loading PCIE slave devices based on FPGA according to claim 1, characterized in that: The second FPGA is provided with an AXI4FULL module and an AXI4LITE module between the AURORA bus and the functional service module, wherein: The AXI4FULL module of the second FPGA is used to parse and output data from the information sent by the AURORA bus or the functional service module; The AXI4LITE module is used to parse and output the configuration from the information sent by the AURORA bus and the functional business module.
4. A system for dynamically loading PCIE slave devices based on FPGA according to claim 2, characterized in that: A firewall for generating analog communication signals is provided between the AXI4FULL module, the AXI4LITE module of the first FPGA and the AURORA bus.
5. A system for dynamically loading PCIE slave devices based on FPGA according to claim 4, characterized in that: When the second FPGA is dynamically loaded, the firewall simulates the second FPGA and maintains the communication state with the first FPGA until the second FPGA completes the dynamic loading and then the firewall stops running.
6. The system for dynamically loading PCIE slave devices based on FPGA according to claim 5, characterized in that: The AXI4FULL module and AXI4LITE module of the first FPGA and the second FPGA both include a write address channel, a write data channel, a write response channel, a read address channel, and a read response channel, and each channel of the first FPGA and the second FPGA is provided with a valid bit.
7. The system for dynamically loading PCIE slave devices based on FPGA according to claim 6, characterized in that: The read response channel of the first FPGA, the write address channel, the write data channel and the read address channel of the second FPGA are all provided with FIFOs for communication feedback and data flow control.
8. A PCIE slave device communication method supporting dynamic loading of a system for dynamically loading PCIE slave devices based on FPGA according to any one of claims 1 to 7, characterized in that: The method specifically comprises: The data sent by the first FPGA or the second FPGA as a sending end through the corresponding channel of the AXI4FULL module of the sending end and the configuration information sent by the AXI4LITE module of the sending end are input into the AURORA bus; The AURORA bus integrates the data and configuration information into a data packet and sends it to the data receiving end; The data packet is parsed by the AXI4FULL module of the receiving end to obtain data, and the AXI4LITE module of the receiving end parses the configuration information and stores them in corresponding FIFOs respectively; Determine whether the FIFO capacity corresponding to the AXI4FULL module or the AXI4LITE module of the receiving end reaches a preset threshold, and if so, the sending end stops sending data; The FIFOs corresponding to the AXI4FULL module and the AXI4LITE module at the receiving end forward the received data to the channels corresponding to the AXI4FULL module and the AXI4LITE module at the receiving end in sequence.
9. An electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the program, the steps of the PCIE slave device communication method supporting dynamic loading as claimed in claim 8 are implemented.
10. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the PCIE slave device communication method supporting dynamic loading as claimed in claim 8 are implemented.
