KR20080054592A - Log storing method for embedded system by using fixed memory area - Google Patents

Abstract

A method for storing a log with a fixed memory area in an embedded system is provided to reduce burden of a developer and prevent mistake omitting the log by leaving the log with a simple output function without making a device driver developer consider or use the log separately, and leave the log even if the system is failed by fatal device error by recording the log in real-time. A fixed log area is reserved in a memory to prevent an OS(Operating System) or an application program from using the fixed log area when the OS is started(S510). A pointer of a buffer used for output is set in the fixed log area when contents are output from the application program or a kernel. The output contents are stored in the fixed log area together with time information related to output(S540). A cache of a CPU is flushed when the system supports a watchdog function(S560). The fixed log area includes a log header area storing log information, an application program log area storing an application program log, and a kernel log area storing the log related to the kernel. The log header area includes location information storing the last log and a checksum for checking whether log header information is correct or incorrect.

Description

How to save log using fixed location memory area in embedded system {LOG STORING METHOD FOR EMBEDDED SYSTEM BY USING FIXED MEMORY AREA}

1 illustrates a conventional method of storing logs in a storage medium or an allocated memory.

2 is a diagram illustrating a structure of a log area according to an exemplary embodiment.

3 is a diagram illustrating a log storing method according to an exemplary embodiment.

4 is a diagram illustrating an example of a log output for tracking a cause of a failure, according to an exemplary embodiment;

5 is a flowchart illustrating a log storage using a fixed location memory area according to an exemplary embodiment.

The present invention relates to a fault tracking method in an embedded system, and more particularly, to a method of tracking faults in an embedded system by leaving a log in a fixed position of a memory such as a RAM.

The log is a record of the history information (errors, events, etc.) performed by the software. The log is information that can trace the cause of the failure when a problem occurs in the system. In general, there are two ways to log so that you can track failures in your system. One method is to log on the storage medium and the other method is to log on the allocated RAM area. As shown in FIG. 1, a system for recording a log on a storage medium such as a disk or flash includes an application 110 requesting log storage through an API call and a library (or manager) 120 for storing the log. ), The storage medium 140 is a space in which log contents are actually stored. The storage medium 140 refers to a device capable of storing data. In general, the storage medium 140 refers to a nonvolatile storage medium in which data remains as it is even after power is stored. When the application 110 requests the library or the manager 120 to store the log to store the log, the log library or the manager 120 stores the log in the storage medium 140. Depending on the size, a large amount of logs can be left, and even if a problem occurs and the system is restarted, the contents of the log remain in the storage medium 140 so that the problem can be tracked.

On the other hand, the system that leaves a log in the allocated RAM area, the application 110 requesting the log storage through the API call, the library (or administrator) 120 to store the contents of the log allocated to the memory, the log contents It is composed of a RAM (RAM) 130 which is a space that is actually stored. When the application 110 requests the library or the manager 120 to store the log to store the log, the log library or the manager 120 is allocated the memory 130 as necessary for the log to store the log. As a result, the system load is relatively low compared to leaving a log in the storage medium 140, and the life is semi-permanent and the stability is higher than that of the storage medium 140.

However, the method of leaving a log on a storage medium is to leave a log on a storage medium such as disk or flash, so that the operation is relatively slow and the load is generated on a storage medium that has a high load. . In addition, in order to save the log, an application program must call an API that stores the log separately. Due to the characteristics of the storage medium, its life is fixed, and the storage medium has a high probability of failure, resulting in poor stability.

Logging in the allocated RAM area also causes problems in the system. When the system is restarted, the log disappears and the problem cannot be tracked. The application must call a separate API to save the log. If there is an additional load and the application does not leave a log, there is a problem that the log cannot be viewed.

An object of the present invention is to provide a log storage method using a fixed location memory area in an embedded system to perform a load-free failure tracking.

To this end, in one embodiment, the fixed location log area in memory is reserved at operating system startup to prevent the operating system or application from using it, and the pointer to the buffer used for output when outputting content from the application or kernel is fixed to the fixed location. By providing a log area, a method of storing the output content in a fixed location log area is provided.

In one embodiment, the log area may include a log header area for storing information about the log, an application log area for storing the application log, and a kernel log area for storing logs related to the kernel, The log header area may include location information where the log was last stored and checksum information for checking whether the log header information is correct.

In one embodiment, the log may be stored further including time information related to the output, and the CPU cache may be flushed when the watchdog function is supported.

In one embodiment, the log area may be managed in the form of a fixed size circular buffer to overwrite the newly added buffer contents circularly from the initial position when the buffer is used up. Can be.

Hereinafter, various embodiments of the present invention will be described with reference to the accompanying drawings.

2 is a diagram illustrating a structure of a log area according to an exemplary embodiment. As the system boots, the operating system maps memory for different purposes. Memory may be allocated to the operating system area 210, the log area 240, and the application area 270, respectively, and the log area 240 is a RAM area in a fixed position, even if the system is restarted unless power is cut off. The content is preserved as it is. In this embodiment, RAM refers to a volatile memory that retains stored data only while power is supplied, but is not limited to any suitable storage medium in the art that performs this function. Can be implemented.

The log area 240 may change according to the size of the system-wide memory, a log header area 242 for storing information about the log, an application log area 245 for storing an application log, and an operating system Alternatively, it may be divided into a kernel log area 248 that stores a log of a device driver. The log header area 242 may include a location 243 where the log was last stored, and checksum information 244 for checking whether the log header information is correct. Finally, the buffer pointer is changed to the location where the log was last stored based on the information on the location of the log 243 so that the log can be stored thereafter. A checksum is a sum and error detection method used to check the accuracy of data. It checks all data sent before the last input data or the last transmission data to check whether the data is correctly entered or transmitted. It is sending the total sum separately. The receiving party can check the integrity of the log header information by taking one by one, summing them up, and then comparing them with the final check sum. As described above, using a fixed memory area as a log area 240 and protecting this area is that the contents of the log remain intact unless the power is shut down even if the system fails and restarts. This is because the cause of the failure can be traced afterwards.

3 is a diagram illustrating a log storing method according to an exemplary embodiment. When the application 310 or the kernel 350 outputs arbitrary contents to the console device, the contents to be output to the temporary buffer must be stored. The kernel 350 is the most basic and core function of the operating system, and manages important resources constituting the system such as storage, memory, files, and peripheral devices. In addition, it handles key aspects of computer operations such as time management, processor management, CPU scheduling, I / O control, and system resource allocation. On the other hand, the console device refers to a terminal device that shows the performance of the system, such as a monitor on a PC, a serial device or LAN in an embedded system. In the system execution step, for example, the contents of the application 310 or the kernel 350 are output to the console device through a print output function such as printf 320 or printk 360, and a system module (not shown). ) May change the pointer to the buffer used to the application log area 330 or the kernel log area 370, which is a memory area of a fixed location. Since the memory area of the fixed position is limited in size, the log area may be managed in the form of circular buffers 340 and 380 to be used within a predetermined size. The circular buffers 340 and 380 may have a fixed size when the buffers are used up in the circular buffer by overwriting the newly added buffer contents in a circular form from the initial position. Through this, the log can be saved without the additional load of copying and storing the buffer contents separately, and the contents of all messages output from the application 310 or the kernel 350 remain in the log area, thereby leaving the application 310 to leave the log. ) Or kernel 350 does not need to use a separate API.

Depending on the system, a watchdog device may be included. Watchdog is a device that automatically resets the entire system when the watchdog is not reset by the software for a certain period of time when a hardware or software failure causes an infinite loop to be performed at a specific position and no other work can be done at all. to be. In a system having such a watchdog function, when hardware is reset through the watchdog, the contents stored in the memory may exist only in the CPU cache of the central processing unit instead of the actual physical memory. Therefore, in a system having a watchdog function, after the output is finished, the cache may be flushed to match the contents of the cache and the memory of the CPU.

Meanwhile, when storing logs in the log areas 330 and 370, time information may be stored together. In the failure analysis, the time when a problem occurs is often important, so when the visual information is stored together in this way, the analysis can be performed more easily.

In a Unix-like operating system such as Linux, when an application 310 accesses the wrong address and terminates, the execution contents of the running application program are stored as a file and the file can be analyzed later. If the storage space of such a file does not exist separately, it cannot be stored and analyzed, but according to an exemplary embodiment, since the log is automatically stored in the log area simply by outputting the termination information, an error of the application 310 may occur. Analysis is possible.

4 is a diagram illustrating an example of a log output for tracking a cause of a failure, according to an exemplary embodiment. For example, when an error occurs in an application program or a device driver, the operating system may restart the system or restart the application program after outputting register information in the central processing unit at the time of forced termination and termination. At this time, the output message is stored as a log as described above, so that the cause of the failure can be traced. The log information shown in FIG. 4 is an example of a log output from FTLinux (Fault Tolerant Linux), which is a Linux having improved stability for use in a communication device developed by the applicant, and analyzes the log to track the cause of the failure. can do.

5 is a flowchart illustrating a log storing using a fixed location memory area, according to an exemplary embodiment. When the system starts, the operating system allocates memory for each purpose and reserves the fixed location memory area, that is, the log area 240 (FIG. 2) from being used (S510). When the operating program or the kernel outputs information to the console, for example, a print output function such as printf or printk is called (S520). In step S530, the buffer pointer is changed to the location where the log was last stored so that the log can be stored thereafter. Next, the content output to the console is stored in the log area of the memory such as RAM, that is, the application log area 245 or the kernel log area 248 (S540). It is determined whether the system supports the watchdog function (S550), and if the watchdog function is supported, the CPU cache is flushed to match the contents of the CPU cache (S560).

The above embodiments can also be embodied as computer readable codes on a computer readable recording medium. Computer-readable recording media include any type of recording device that stores data that can be read by a computer system. Examples of computer-readable recording media include ROM, RAM, CD-ROM, magnetic tape, floppy disks, optical data storage devices, and the like, which are also implemented in the form of carrier waves (for example, transmission over the Internet). Include. The computer readable recording medium can also be distributed over network coupled computer systems so that the computer readable code is stored and executed in a distributed fashion. In addition, functional programs, codes, and code segments for implementing the above embodiments can be easily inferred by programmers in the art to which the present invention belongs.

Although the present invention has been described in detail with reference to the preferred embodiments of the present invention, these are merely exemplary, and those skilled in the art to which the present invention pertains have various modifications and equivalents therefrom. It will be appreciated that embodiments are possible. Therefore, the true technical protection scope of the present invention will be defined by the technical spirit of the appended claims.

According to the above embodiments, the application program developer can leave a log with a simple output function without the device driver developer considering or using a separate log, thereby reducing the burden on the developer and preventing a mistake of dropping the log. You don't need to install a separate library or manager. And because the log remains in real time, it can be logged even if the system suddenly fails due to a fatal device fault.

In addition, according to the above embodiments, there is no additional buffer copy to store the log, so that the system load can be reduced, and the log is stored in memory, thereby reducing the possibility of log corruption, even if the system is restarted unless the power is turned off. Will remain intact. Finally, the effect is that you can leave not only the application log but also the kernel log.

Claims (11)

Log storage method using fixed location memory area in embedded system, Reserving a fixed location log area in memory at startup of the operating system to prevent use by the operating system or an application program; When outputting content from an application or kernel, setting a pointer to a buffer used for the output to the fixed location log area; And Storing the output content in the fixed location log area How to include. The method of claim 1, The log area includes a log header area for storing information about the log, an application log area for storing the application log, and a kernel log area for storing logs related to the kernel. The method of claim 2, And the log header area includes checksum information for checking whether or not the log header information is correct. The method of claim 1, The storing step further includes storing the time information associated with the output. The method of claim 1, If the system supports a watchdog function, further comprising flushing a CPU cache. The method of claim 1, The log area is managed in the form of a fixed size circular buffer to overwrite the newly added buffer contents from the initial position when the buffer is used up in the buffer. A log storage system using a fixed location memory area in an embedded system, A memory in which its fixed location log area is reserved at the start of the operating system to prevent an operating system or an application from using it; And A module that, when outputting content from an application or kernel, sets a pointer to the buffer used for the output to the fixed location log area. Log storage system comprising a. The method of claim 7, wherein The log area includes a log header area for storing information about the log, an application log area for storing the application log, and a kernel log area for storing logs related to the kernel. The method of claim 7, wherein The log area is managed in the form of a fixed size circular buffer for overwriting the newly added buffer contents in a circular form when the buffer is used up in the buffer. The method of claim 7, wherein Further comprising a watchdog device, And the fixed location log area after the output is consistent with information in the cache through flushing a CPU cache. A computer readable recording medium storing computer executable instructions for performing a log storage method using a fixed location memory area in an embedded system, the method comprising: Reserving a fixed location log area in memory at startup of the operating system to prevent use by the operating system or an application program; When outputting content from an application or kernel, setting a pointer to a buffer used for the output to the fixed location log area; And Storing the output content in the fixed location log area Computer-readable recording medium comprising a.

Images