WO2022227409A1

WO2022227409A1 - Embedded terminal remote software updating method

Info

Publication number: WO2022227409A1
Application number: PCT/CN2021/122776
Authority: WO
Inventors: 王宜怀; 施连敏; 章晓芳; 姚望舒; 刘强
Original assignee: 苏州大学
Priority date: 2021-04-29
Filing date: 2021-10-09
Publication date: 2022-11-03
Also published as: CN113110860A; CN113110860B

Abstract

The present invention provides an embedded terminal remote software updating method, used for being adapted to software updating requirements in different application scenes, and considering the portability and reusability of a system framework. A dynamic command serves as a main carrier and provides flexibility and adaptive support for a software updating function. The present invention uses a communication mode of serial communication as the preamble study of communication module adaption, and on this basis, summarizes a user program updating method and a BIOS self-updating method of the communication module adaption. For a dynamic command space problem in an implementation process, the present invention provides a solution for a dynamic component library, also has certain development practicability when simplifying the dynamic command, and also provide a perfect solution for a Flash load balancing problem in the dynamic component library.

Description

A kind of embedded terminal remote software update method

technical field

The invention relates to the technical field of software development and debugging, in particular to a remote software update method for an embedded terminal.

Background technique

As the core function of the integrated development environment, software update is the main way for the software written by the developer to run on the embedded terminal. Under the background of remote online development of embedded terminals, the terminal node needs to have the ability to update the code remotely, so that the node can be flexibly configured and upgraded to adapt to changes in the environment.

In the traditional development mode, developers often select a communication method in advance for specific application scenarios, and solidify the corresponding communication module driver components and software update codes in the BIOS program, so as to complete the subsequent User program. Developed and updated. However, such a development model often has the following two disadvantages:

(1) The program is portable and has poor reusability. In this development mode, whenever different communication modules are selected for different scenarios, the previous BIOS programs lose the value of porting and reuse to a large extent. Development will bring meaningless duplication of work, reduce the overall development efficiency and prolong the development cycle.

(2) The upgrade of the communication method is blocked. The current development and evolution of communication modules is extremely fast. Whether it is for the replacement of the communication module itself or the firmware update for the same communication module, there may be demands within the working cycle of the embedded terminal. However, in the traditional development mode, the related components of the communication module have been solidified in the BIOS, which is difficult to meet the replacement and upgrade requirements.

Therefore, it is particularly important to propose an adaptive software update method for communication modules. At the same time, in the development process of the embedded terminal, the update requirements for user programs are often larger.

technical problem

technical solutions

The object of the present invention is achieved through the following technical solutions.

Aiming at the key technology of adaptive program update of communication modules, this application proposes a remote software update method for embedded terminals, so as to adapt to the software update requirements in different application scenarios and take into account the portability and reusability of the system framework itself . As the main carrier, dynamic commands provide flexibility and adaptive support for the software update function.

Specifically, the present invention provides a remote software update method for an embedded terminal, comprising: sending dynamic command data to a user serial port and writing it into a user program; sending a system instruction for program jumping to the user serial port, and the user program receives the instruction Then, jump to the BIOS; the dynamic command function is called by default during the BIOS startup process; in the dynamic command, the program running state is first modified to complete the BIOS residency; secondly, the initialization and interrupt enable of the user serial port are completed; and the BIOS program interrupt vector is set The address of the interrupt service routine corresponding to the user serial port in the table is a dynamic command function; finally, the host computer is informed that the relevant initialization steps have been completed, and the program update starts; the host computer starts to send the user program data frame to the user serial port, and the terminal passes the user serial port. The dynamic command completes receiving, checking and writing; when the dynamic command of the user serial port completes the processing of the last frame of data, the program running state is modified, the BIOS is reset, and the BIOS basic operation process completes the subsequent guidance for the user program.

Further, the outermost frame structure of the user serial port is the framing structure of the communication layer, including frame header, frame trailer, data length, communication layer data and check bit.

Further, the establishment process of described dynamic command function is as follows: revise link file, increase the FLASH space corresponding to dynamic command in MEMORY field and SECTIONS field; Increase the declaration and definition of dynamic command function, function body is empty, and use attribute grammar The dynamic command function is placed in the FLASH space; in the interrupt handler of the communication module receiving data, two frame marks are reserved for receiving the machine code of the dynamic command function and the execution dynamic command function respectively; In the update state judgment function Add dynamic command execution branch; add dynamic command reset step in key reset.

Further, further include:

The BIOS is self-updated: the writing of the dynamic command and the program jump are completed by the system command, the takeover of the communication module is completed with the help of the dynamic command, and then the BIOS self-update process is started, and the machine code of the new version of the BIOS is first received and stored in the current BIOS. Behind the BIOS machine code, after the machine code of the new BIOS is received and verified, the machine code of the old BIOS is replaced and updated by the machine code of the new BIOS sector by sector.

Further, the steps of the self-update of the BIOS are as follows: (1) write a related function to make a dynamic command code; (2) send the dynamic command installation package of the BIOS self-update to the target chip user program, and write the dynamic command after the target chip receives it. (3) Send system commands to the target chip user program to make it jump and reside in the BIOS; (4) BIOS calls dynamic commands to complete the takeover of the user communication module; (5) Starts sending new messages to the target chip The machine code of the BIOS project is stored in the back of the current BIOS storage space after frame-by-frame verification, and the corresponding frame header is given for the last frame of data; (6) After the target chip detects the last frame of data and verifies the storage, Do not exit the interrupt immediately, and enter the data migration work in the interrupt, so that the new BIOS officially replaces the old BIOS, and after the migration is completed, a soft reset is performed to start the new BIOS process.

Further, the dynamic command can be simplified as a dynamic component library, including: a component function list area and a component function code area, wherein the component function list area is used to store the function pointer and necessary parameters of the component, and the component function code area is used to store the component. The specific implementation of each function in the system, the dynamic command function is responsible for the management of the data in these two areas.

Further, the dynamic command can perform query operations on each component and function, and add, modify and delete operations in combination with the component function code area.

Further, the query operation process includes: reading the data in the component function list area by means of a dynamic command and sending it back to the upper computer for analysis by the upper computer.

Further, the operation process of the increase includes: when adding a component, first determine the recording position of the overall information of the current component in the component function list area according to the next available number field in the component function list area, and after the recording is completed. Then modify the next available number field, and determine the write address of the component machine code according to the next writable address field of the function code area, and modify the next writable address and function code area of the function code area after the writing is completed. The used space field value.

Further, the Flash load balancing method in the dynamic component library is as follows:

(1) For the delete operation, do not erase the space originally occupied, but only modify the used space parameters of the function code area in the component function list area; (2) For the modified operation, also do not erase the originally occupied space, Directly look for the next storable space to write, and the absolute change of the function storage space size is recorded in the used space parameter of the function code area. The space occupied by the discarded code data generated in (1) (2) is called Flash fragment; (3) When performing addition operations and addition operations in the process of modification, if there is no continuous storage space that can accommodate components, then according to the used space of the function code area, the first address of the function code area and the end address of the function code area. parameters, calculate whether the remaining continuous free space and the sum of all Flash fragments generated by (1) (2) can accommodate the component, if not, return the corresponding storage space emergency prompt information, if yes, clear all sectors at one time , organize all components into close and continuous storage before and after, and inform the corresponding upper computer that this operation will take a certain amount of time.

beneficial effect

The advantage of the present invention is that: the present invention takes serial communication as a leading research on the self-adaptation of the communication module, and based on this, summarizes the user program update method of the self-adaptation of the communication module. And for the dynamic command space problem in the implementation process, the solution of the dynamic component library is given, which simplifies the dynamic command and has certain development practicability, and also provides a complete solution for the Flash load balancing problem in the dynamic component library. solution.

Description of drawings

Various other advantages and benefits will become apparent to those of ordinary skill in the art upon reading the following detailed description of the preferred embodiments. The drawings are for the purpose of illustrating preferred embodiments only and are not to be considered limiting of the invention. Also, the same components are denoted by the same reference numerals throughout the drawings. In the attached image:

FIG. 1 shows a schematic diagram of a frame structure of a user serial port according to an embodiment of the present invention.

FIG. 2 shows a flow chart of dynamic command execution based on user serial port User program update according to an embodiment of the present invention.

FIG. 3 shows a schematic diagram of functional dependencies of errors according to an embodiment of the present invention.

FIG. 4 shows a schematic diagram of function rewriting for functional dependency problems according to an embodiment of the present invention.

FIG. 5 shows a schematic diagram of the functional dependency of the BIOS self-update dynamic command according to an embodiment of the present invention.

FIG. 6 is a schematic diagram showing the trend of erasing frequency and service life of each sector under the compact principle according to an embodiment of the present invention.

FIG. 7 shows a schematic diagram of Flash defragmentation according to an embodiment of the present invention.

Embodiments of the present invention

Exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be embodied in various forms and should not be limited by the embodiments set forth herein. Rather, these embodiments are provided so that the present disclosure will be more thoroughly understood, and will fully convey the scope of the present disclosure to those skilled in the art.

The invention proposes a remote software update method for an embedded terminal, so as to adapt to the software update requirements in different application scenarios and take into account the portability and reusability of the system framework itself. As the main carrier, dynamic commands provide flexibility and adaptive support for the software update function. The present invention takes the communication mode of serial communication as the leading research on the self-adaptation of the communication module, and based on this, summarizes the user program update and BIOS self-update method of the self-adaptation of the communication module. And for the dynamic command space problem in the implementation process, the solution of the dynamic component library is given, which simplifies the dynamic command and has certain development practicability, and also provides a complete solution for the Flash load balancing problem in the dynamic component library. solution. The present invention selects a relatively fixed BIOS program to be responsible for adapting to different communication modules, and the user program carries the changed communication module drive components. Thus, the basic software architecture under this software update method is determined.

The communication module selected in embedded application development often does not have the ability to communicate directly with the corresponding embedded terminal. The general solution is to use the inherent communication method in the embedded terminal as a bridge to achieve the communication between the two. Data interaction. In addition, the core problem of the adaptive software update of the communication module is that the BIOS realizes the program update by means of the User's communication module. Therefore, as a leading research, a relatively simple communication method should be selected. Serial communication is a common peripheral interface of MCU, and the communication speed and security are guaranteed to a certain extent, and it can be easily converted to RS485/RS232 interface through a simple peripheral circuit to further ensure Communication stability and transmission distance. So choose serial communication as User's communication method to study the realization of User program update method, and prepare for summarizing the general method of communication module adaptive software update.

1 Basic points and design ideas

The user program update based on the user serial port is not a frequent requirement, and considering that the subsequent program update may not be limited to serial communication, the user program update based on the user serial port should be combined with dynamic commands as a temporary function. deal with. In addition, the main object of the program update is still the user program, so the simpler implementation method is still to use the BIOS program as the main place to perform the update, so there is a link where the User program and the BIOS program jump to each other. All data sending and receiving are based on the serial port of the user program, which means that the user program should write dynamic commands into the BIOS dynamic command area, and considering that the sending and receiving of data is still mainly performed by the BIOS program, the BIOS program must Serial port takeover.

1. Basic design points

(1) Program jump. In the program update scenario of the user serial port, a version of the User program must already be running in the target embedded terminal, so the program jump is firstly the jump from the User program to the BIOS program.

The jump from the User program to the BIOS program is still determined by the data received by the user serial port. Therefore, in the analysis of the data received by the user serial port, it is necessary to reserve certain keywords to identify this frame of data as a system command. jump.

After the data of the new version of the user program is received, the step of jumping from the BIOS program to the User program is relatively simple, and the basic User program booting method under the general embedded computer framework (hereinafter referred to as the GEC framework) can be used.

(2) The establishment of dynamic command system in user program. Similarly, the specific function of the dynamic command is received and written by the user serial port, and the execution of the dynamic command is still carried out by the BIOS program. Therefore, like the program jump above, under the system command, it is also necessary to reserve a keyword to use It is used to guide the writing of dynamic commands.

(3) The BIOS program takes over the user serial port. The takeover of the user serial port by the BIOS program not only needs to initialize the user serial port and enable the interrupt, but also needs to redirect the interrupt handler of the user serial port. For functions that do not exist in these BIOS programs, dynamic commands can be used to implement them.

2. Main design ideas

Based on the foregoing, the main design ideas of the User program update based on the user serial port are as follows:

(1) Send dynamic command data to the serial port of the user program and write it in by the User program.

(2) Send the system command of program jump to the serial port of the user program. After the User program receives the instruction, it jumps to the BIOS.

(3) The dynamic command function is called by default during the BIOS startup process. In the dynamic command, firstly modify the running state of the program to complete the residency of the BIOS; secondly, complete the initialization and interrupt enable of the user serial port; and set the interrupt service routine address corresponding to the user serial port in the BIOS program interrupt vector table as the dynamic command function; Finally, inform the host computer that the relevant initialization steps have been completed, and the program update can be started.

(4) The host computer starts to send the User program data frame to the user serial port, and the terminal completes the reception, checksum writing through the interrupt service routine (ie dynamic command) of the user serial port.

(5) When the interrupt service routine (ie dynamic command) of the user serial port completes the processing of the last frame of data, modify the running state of the program, reset the BIOS, and complete the subsequent booting of the User program by the basic running process of the BIOS.

2 Specific implementation methods

After giving the basic design ideas, the next step is to describe the specific implementation of the User program update based on the user serial port, mainly focusing on the frame structure and dynamic command system of the user serial port.

1. Frame structure of user serial port

The outermost frame structure of the user serial port is the framing structure of the communication layer, which may include frame header, frame end, data length, communication layer data and check digit. The frame structure in the communication layer data should reserve contents such as the unique serial number of the device and the software version number to assist in judging the choice of data frames in some broadcast scenarios. The effective data frame structure in the communication layer data is defined by the developer, but in order to complete the software update such as the communication module adaptation, it is necessary to reserve a keyword as the frame identifier of the system command to provide rich pre- Set functions, such as program jump, device serial number query, etc. The frame structure of the user serial port is shown in Figure 1.

2. The establishment of dynamic command system under GEC

The establishment of the dynamic command system under GEC should also be considered from the perspectives of BIOS engineering and User engineering.

(1) The establishment of the dynamic command system in the BIOS project. The software foundation of the dynamic command system is the P _new project framework, so establishing a dynamic command system in the BIOS project means that the BIOS project corresponds to the P _new project, and the modification steps for the BIOS project are mainly as follows.

Step 1: Modify the link file and add the FLASH space corresponding to the dynamic command in the MEMORY field and the SECTIONS field.

Step 2: Add the declaration and definition of the dynamic command function, the function body is empty, and use the attribute syntax to place the dynamic command function in the FLASH space in Step 1.

Step 3: Reserve two frame identifiers in the interrupt handler for receiving data of the communication module for receiving the machine code of the dynamic command function and executing the dynamic command function respectively.

Step 4: Add a dynamic command execution branch to the update state judgment function.

Step 5: Add dynamic command reset step in key reset. The reset here refers to restoring the function of the dynamic command to an empty function, that is, the reset machine code corresponds to "return 0;" of the C language code.

(2) Access to the dynamic command system in the User project. After the dynamic command system is established in the BIOS project, most of the temporary function addition requirements can be met. However, for some special temporary functions, especially those functions that need to be implemented with resources in the User project, the dynamic command system must be connected to the User project. User program update based on user communication module is a typical example.

Different from establishing the dynamic command system in the BIOS project, only the dynamic command system needs to be accessed in the User project. The dynamic command storage space remains unchanged in the BIOS, but only provides a way to write dynamic commands for the User project. , that is, the two projects jointly maintain a dynamic command space.

The User project is the main development site of the developer. The specific communication module is determined by the developer, and the dynamic command system must depend on the transmission and reception of the data of the communication module. Therefore, the writing function of the dynamic command should be placed in the system command. The system command is used as a fixed file of the User project template, so that the access to the dynamic command system in different communication modules can be realized.

3. Development process based on dynamic commands

Next, the development process based on the dynamic command system is given.

(1) The programming of P _new . As discussed above, the BIOS project that has established the dynamic command system is the P _new project. After compiling, H _new is obtained, and it can be burned into the target embedded terminal.

(2) Compilation, verification and extraction of dynamic commands. Firstly, the specific functions of the dynamic commands are written in the copy P _copy of the P _new project, and the preliminary dynamic command syntax rules are checked. After compiling, compare H _new and H _copy to ensure that the machine code except the dynamic command area is completely consistent. The machine code corresponding to the dynamic command area is extracted by the written host computer software, and sent to the target embedded terminal running the P _new project, and finally the command to run the dynamic command is sent to observe the running result of the dynamic command.

(3) Integration of dynamic commands and host computer software. According to the running result, after confirming that the dynamic command is made correctly, the extraction, sending and running of the dynamic command machine code are added to the event of the specified host computer window as required. In this way, the target embedded terminal running the P _new project can run dynamic commands by relying on the host computer.

4. The specific design of dynamic commands

As mentioned above, the specific content of the dynamic command should be divided into two parts. The first part should be executed at the default call of the BIOS program, mainly to complete the initialization of the user serial port, interrupt enable and take over the user serial port interrupt, and reserve the code for subsequent updates; the second part is the user serial port interrupt processing specific content of the program. Among them, the user serial port interrupt service routine function needs to be registered when taking over the user serial port interruption. Here, the dynamic command address can be directly selected. For the two parts of the dynamic command, the dynamic command is divided according to whether the communication module interrupt is triggered or not. To ensure that the two parts do not interfere with each other. The dynamic command execution flow based on user serial port User program update is shown in Figure 2.

Communication module adaptive user program update method

The external communication module selected in the embedded application development often needs to establish the data connection with the help of the inherent communication module in the embedded terminal. In the following text, serial communication will be used as an example to serve as a bridge connecting the embedded terminal and the external communication module. Therefore, the component design of the external communication module in a specific project can be understood as the upper-level component design based on uplink communication.

Through in-depth analysis of the previous user program update method based on user serial port, the main work related to serial communication is summarized and extracted, mainly including: serial communication initialization, interrupt enable, interrupt service routine address designation, data transmission and data reception. Therefore, in the communication module adaptive program update method, the common elements of the communication module are abstracted as: initialization of the communication module, interrupt enable, designation of interrupt service routine address, data transmission and data reception.

As a result, the adaptive User program update design method of the communication module becomes relatively clear, and the related work of the communication module is only carried out in the corresponding interrupt and dynamic commands of the User program, so a detailed design method can be given for both. .

Communication module data reception interruption in User program

For the data reception interruption of the communication module in the User program, it is mainly explained from the data frame.

The data frame of the communication module does not change in structure with the change of the communication module, so the specific frame structure still adopts the user serial port frame structure design method shown in Figure 1.

In the same way, since the adaptive User program update of the communication module is essentially the update of the User program by the BIOS program, it is still necessary to reserve a program jump instruction in the system instruction identifier in the data frame for converting the target terminal's The program running state is pulled back from User to BIOS.

In addition, due to the dependence of the communication module driving component and the need of the program execution process, the system instruction identifier still needs to be connected to the dynamic command system to provide the User program with a writing interface of the dynamic command.

Design of Dynamic Commands

The main functions of the dynamic commands under the communication module adaptation include: initialization of the communication module, interrupt enablement of the communication module, redirection of the communication module interrupt service routine, update process based on the communication module and BIOS interrupt vector table recovery, etc. The main decision algorithm is as follows.

.

Overall process design

The overall process of the adaptive User program update of the communication module is designed as follows:

(1) Send dynamic command data to the communication module and write it in by the User program.

(2) Send the system command of program jump to the communication module. The User program receives the instruction and jumps to the BIOS.

(3) The dynamic command function is called by default during the BIOS startup process.

(4) In the dynamic command, first modify the program running state to complete the BIOS residency; secondly complete the initialization and interrupt enable of the communication module; and set the interrupt service routine address corresponding to the communication module in the BIOS program interrupt vector table It is a dynamic command function; finally inform the host computer that the relevant initialization steps have been executed, and the program update can be started.

(5) The host computer starts to send the User program data frame to the communication module, and the terminal completes the reception, checksum writing through the interrupt service routine (ie dynamic command) of the communication module.

(6) When the interrupt service routine (that is, the dynamic command) of the communication module completes the processing of the last frame of data, the program running state is modified, the BIOS is reset, and the BIOS basic operation process completes the subsequent guidance for the User program.

BIOS self-update method for adaptive communication module

An important prerequisite for ensuring the stability of the embedded terminal software framework is the stability of the BIOS project. When the BIOS is found to have fatal hidden dangers, a BIOS self-update mechanism is urgently needed. The support of dynamic commands enables BIOS self-update to be realized.

main method design

Similar to the update of the User program, the general idea of the adaptive BIOS self-update of the communication module is: complete the writing of the dynamic command and the program jump through the system command, complete the takeover of the communication module with the help of the dynamic command, and then start the BIOS self-update. The process is to first receive the machine code of the new version of BIOS and store it behind the current BIOS machine code. After the machine code of the new BIOS is received and verified, replace the machine code of the old BIOS with the machine code of the new BIOS sector by sector. . The main steps of BIOS self-update are as follows:

(1) Write related functions to make dynamic command codes.

(2) Send the dynamic command installation package of BIOS self-update to the target chip User program, and the target chip writes the dynamic command area after receiving it.

(3) Send a system command to the target chip User program to make it jump and reside in the BIOS.

(4) The BIOS invokes the dynamic command to complete the takeover of the user communication module.

(5) Start sending the machine code of the new BIOS project to the target chip, check it frame by frame and save it in the back of the current BIOS storage space, and give a special frame header for the last frame of data.

(6) After the target chip detects the last frame of data and verifies the storage, it does not exit the interrupt immediately, but enters the data relocation work in the interrupt, so that the new BIOS officially replaces the old BIOS. After the relocation is completed, it performs a soft reset and starts to execute the new BIOS process.

The main decision-making algorithm within the dynamic command is shown below.

.

functional dependency problem

In the process of replacing the old BIOS sector by sector, the key problem encountered is: when using the erasing function of the Flash component in the old BIOS to erase the old BIOS, when the erasing function tries to erase the sector where the erasing function itself is located When erasing, a fatal error will inevitably occur, and other functions of the Flash component in the subsequent old BIOS will not be available, as shown in Figure 3.

In order to avoid the function dependence on the old BIOS, the initial strategy adopted is to rewrite the Flash component functions that will be used in the entire update process in the dynamic command function, and the BIOS self-update area skips the dynamic command function area, as shown in Figure 4 shown.

Since the dynamic command is essentially called in the interrupt, after the dynamic command receives the last frame of data of the new BIOS, the program does not jump out of the dynamic command and returns to the main function of the old BIOS, but continues to stay in the dynamic command area. Perform sector-by-sector replacement of the old and new BIOS, so that there is no problem of jumping back to the main program but the main program has been erased. It is only necessary to perform a software reset after the dynamic command completes the sector-by-sector replacement of the old and new BIOS. Can.

Further analysis of the above problems can be found that the functional dependency problem of the code to be erased only begins with the process of replacing the old BIOS data sector by sector. The function that depends on the problem can be rewritten. The functional dependencies of the BIOS self-update dynamic commands are shown in Figure 5.

Simplification of Dynamic Commands: Dynamic Component Library

A series of program update methods given above all rely on dynamic commands to complete the realization of core functions. However, the original intention of the dynamic command is to achieve the effect of dynamic loading of multiple functions in a limited space through the common use of a small area of space by multiple temporary functions. In the program update scenario mentioned above, since the communication module belongs to the resource of the User program in the software framework, it is necessary to rewrite a communication module in the dynamic command when transmitting the dynamic command of the related functions of the communication module to the BIOS. function copy or expand the communication module function. This will inevitably lead to an excessively large amount of code for dynamic commands, and even in the face of communication modules with complex implementations, there may be a problem of insufficient space.

The Proposition of Dynamic Component Library

In the case of function rewriting and expansion of dynamic commands due to a large number of functional dependencies, the storage space of dynamic commands may not be enough to accommodate huge codes. Dynamic command invocation. In addition, the BIOS project in the GEC software framework has been solidified inside the chip when the developer obtains the embedded terminal, and is maintained by the GEC framework developer. Therefore, the GEC framework also needs to release stable components for developers to use. And there needs to be a way of publishing that doesn't reveal the source code. Therefore, the technology of dynamic component library based on dynamic command is introduced.

Dynamic component library is a component management mode provided by BIOS. In this mode, dynamic commands are responsible for adding, deleting, checking, and modifying components, and combined with the host computer, the corresponding component header files are generated for users to use.

Design Method of Dynamic Component Library

The design of the dynamic component library mainly relies on two areas opened up in the BIOS link file: the component function list area and the component function code area. The component function list area is used to store the function pointer and other necessary parameters of the component, and the component function code area is used for It is used to store the specific implementation of each function in the component. In addition, the dynamic command function is responsible for the management of the data in these two areas. The User project uses the functions in the dynamic component library through the API (Application Programming Interface) interface provided by the BIOS project.

1. Design of the component function list area

The programming form of the component function list area is an array of pointer type (32-bit), which is mainly divided into three areas: component library parameter area, system related area and component area. The main organization form and the function of corresponding parameters are shown in Table 3-1. Show.

.

2. Preliminary design of main management methods

With the index and related parameters in the component function list area, dynamic commands can easily query each component and even each function, and add, modify and delete operations in combination with the component function code area.

(1) Query. The component function list area provides detailed records for all the data in the component library itself, so when there is a query request for the corresponding information, you only need to rely on dynamic commands to read the data in the component function list area and send it back to the host computer. machine analysis.

(2) Increase. When adding a component, first determine the record position of the overall information of the current component in the component function list area according to the "next available number" field in the component function list area, and modify the "next available number" field after the recording is completed . In addition, it is necessary to determine the write address of the component machine code according to the field of "Next Writable Address of Function Code Area", and modify the "Next Writable Address of Function Code Area" and "Function Code Area Has Been Written" accordingly after the writing is completed. Use space" field value.

(3) Delete. When there is a need for component deletion, firstly, according to the "area occupancy number" field of each area in the component function list area, you need to chain query component names in sequence. When the component to be deleted is located, delete the corresponding component function list area and component function code. and modify the fields of "Next Available Number", "Next Writable Address of Function Code Area" and "Used Space of Function Code Area".

(4) Modification. Due to the mechanism that Flash uses sectors as the minimum erasing unit, the modification operation for components can be simply disassembled as first delete and then add. The difference is that the Flash write operation of the two areas is not performed immediately after deletion, but is written together after the addition operation is performed. The specific method has been described in (2)(3).

Thanks to the mechanism of the dynamic command itself, these operations will not affect the current running state of the system, nor will it affect the real-time performance too much, eliminating the cumbersome shutdown and repeated programming of the program.

Flash load balancing

Considering the service life of Flash itself, if each operation ensures the compactness of the component function code area in multiple component addition, deletion and modification operations, these operations will often lead to serious polarization of the component function code area life. This is because In normal scenarios, the actual usage ratio of the component function code area will not be close to 100%. Taking 50% of the average actual usage space as an example, the erasure frequency and corresponding life of each sector in the component function code area will roughly appear as follows: Figure 6. Distribution. And the smaller the average actual use space of the component function code area, and the more the number of component additions, deletions and modifications, the more serious this dichotomy will be.

In response to this problem, this application proposes a Flash load balancing solution:

(1) For the delete operation, the original occupied space is not erased, and only the parameter "used space of function code area" in the component function list area is modified;

(2) For the modified operation, the original occupied space is also not erased, but the next storable space to write is directly searched for. The absolute change in the size of the function storage space is recorded in the "used space in the function code area". a parameter. (1) The space occupied by the discarded code data generated in (2) is hereinafter referred to as Flash fragment.

(3) When adding operations and adding operations during modification, if there is no continuous storage space that can accommodate the component, according to the "used space in the function code area", "the first address of the function code area" and "function code area" The three parameters of "code area tail address", calculate whether the remaining continuous free space and the sum of all Flash fragments generated by (1) (2) can accommodate the component, if not, return the corresponding storage space emergency prompt information, if yes , clean up all sectors at one time, organize all components into close and continuous storage before and after as shown in Figure 7, and inform the corresponding upper computer that this operation will take a certain amount of time.

The core idea of this scheme is that when there is enough storage space, the free storage space in the rear is used first, and while the utilization rate of the tail sector is improved, the flash erase operation is largely avoided until the free storage space is insufficient. Only consider the overall processing of Flash fragments. The detailed algorithm is expressed as follows:

.

The above description is only a preferred embodiment of the present invention, but the protection scope of the present invention is not limited to this. Substitutions should be covered within the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

A method for remote software updating of an embedded terminal, comprising: sending dynamic command data to a user serial port and writing it in by a user program; sending a system instruction for program jumping to the user serial port, and after the user program receives the instruction, Jump to BIOS; the dynamic command function is called by default during the BIOS startup process; in the dynamic command, the program running state is first modified to complete the BIOS residency; secondly, the initialization and interrupt enable of the user serial port are completed; and the BIOS program interrupt vector table is set The address of the interrupt service routine corresponding to the user serial port is a dynamic command function; finally, the host computer is informed that the relevant initialization steps have been completed, and the program update is started; the host computer starts to send the user program data frame to the user serial port, and the terminal passes the user serial port. Dynamic commands Complete receiving, checking and writing; when the dynamic command of the user serial port completes the processing of the last frame of data, modify the program running state, reset the BIOS, and complete the follow-up guidance for the user program by the BIOS basic running process; the dynamic command The creation process of the function is as follows: modify the link file, add the FLASH space corresponding to the dynamic command in the MEMORY field and the SECTIONS field; add the declaration and definition of the dynamic command function, leave the function body empty, and use the attribute syntax to place the dynamic command function in all places. Describe the FLASH space; Reserve two frame marks in the interrupt processing program that the communication module receives data respectively for receiving the machine code of the dynamic command function and executing the dynamic command function; In the update state judgment function, increase the dynamic command execution branch; Added dynamic command reset steps in key reset.
The method for remote software updating of an embedded terminal according to claim 1, wherein the outermost frame structure of the user serial port is a framing structure of a communication layer, comprising a frame header, a frame trailer, a data length, a communication Layer data and check digit.
The method for remote software updating of an embedded terminal according to claim 1, further comprising: the BIOS is self-updated: writing of dynamic commands and program jumping are completed through system instructions, and communication is completed by means of dynamic commands The module takes over, and then starts the BIOS self-update process. First, the machine code of the new BIOS is received and stored in the back of the current BIOS machine code. After the machine code of the new BIOS is received and verified, the old BIOS is updated sector by sector. Replace the machine code of the updated BIOS with the machine code of the new BIOS.
The method for remote software update of an embedded terminal according to claim 3, wherein the steps of the BIOS self-update are as follows: (1) write a related function to make a dynamic command code; (2) send the target chip user program BIOS self-updated dynamic command installation package, the target chip will write the dynamic command area after receiving it; (3) Send system commands to the target chip user program to make it jump and reside in the BIOS; (4) The BIOS calls the dynamic command , complete the takeover of the user communication module; (5) Start to send the machine code of the new BIOS project to the target chip, check frame by frame and save it in the back of the current BIOS storage space, and give the corresponding frame header for the last frame of data ; (6) After the target chip detects the last frame of data and verifies the storage, it does not exit the interrupt immediately, and enters the data relocation work in the interrupt, so that the new BIOS officially replaces the old BIOS. BIOS flow.
The method for remote software updating of an embedded terminal according to claim 1, wherein the dynamic command can be simplified into a dynamic component library, comprising: a component function list area and a component function code area, wherein the component function list area uses In order to store the function pointer and necessary parameters of the component, the component function code area is used to store the specific implementation of each function in the component, and the dynamic command function is responsible for the management of the data in these two areas.
The method for remote software updating of an embedded terminal according to claim 5, wherein the dynamic command can perform query operations on each component and function, and add, modify and delete operations in combination with the component function code area.
The method for remote software updating of an embedded terminal according to claim 6, wherein the query operation process comprises: reading the data in the component function list area by means of dynamic commands and sending it back to the host computer, and the host computer parses it. .
The method for remote software updating of an embedded terminal according to claim 6, wherein the operation process of adding comprises: when adding a component, firstly, it is determined according to the next available number field in the component function list area. The overall information of the current component is in the record position of the component function list area, and after the recording is completed, the next available number field is modified, and the write address of the component machine code is determined according to the next writable address field in the function code area, and when writing After the entry is completed, modify the next writable address of the function code area and the field value of the used space in the function code area accordingly.
A method for remote software updating of an embedded terminal according to claim 5, wherein the Flash load balancing method in the dynamic component library is as follows: (1) For the deletion operation, the space originally occupied is not erased , only modify the used space parameters of the function code area in the component function list area; (2) For the modified operation, also do not erase the originally occupied space, directly find the next storable space to write, the function storage space size The absolute change is recorded in the used space parameter of the function code area, of which (1) the space occupied by the discarded code data generated in (2) is called Flash fragment; (3) When performing addition operations and addition operations in the modification process, If there is no continuous storage space that can accommodate the component, then according to the three parameters of the used space of the function code area, the first address of the function code area and the end address of the function code area, calculate the remaining continuous free space and by (1) (2) Whether the sum of all generated Flash fragments can accommodate the component, if not, return the corresponding storage space emergency prompt information, if yes, clear all sectors at one time, organize all components into close and continuous storage before and after, and inform the corresponding host computer This operation will take some time.