JPH1040177A - Program check system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To take a speedy countermeasure if a program is destroyed by preventing the time for abnormality detection when abnormality of the program is detected from increasing with the program size. SOLUTION: A program check processing part 4a calculates the check sum of one program block specified by a check block specification part 3 and stores it in a check sum storage part 5 each time a timer interruption generation part 2 initiates a periodic interruption to store it in a check sum storage part 5a. Then the newly calculated check sum is used for the program block specified by the check block specification part 3 and check sums read out of the check sum storage part 5a is used for all other program blocks to calculate the cumulative value of the check sums of the program blocks. The cumulative value of the calculated check sums is collated with the check sum in the check sum storage part 6 and when they do not match each other, it is decided that the program is abnormal.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a program check method, and more particularly to a program check method in a computer system which is developed on a memory and operates a central processing unit.

[0002]

2. Description of the Related Art A rewritable memory (hereinafter referred to as RA) which is read from a read-only memory (hereinafter referred to as ROM).
M), a computer system using a central processing unit (hereinafter referred to as a CPU) for reading and executing a program developed on a computer is disclosed in Japanese Patent Laid-Open No. 4-542 in order to check whether the RAM is operating normally. As disclosed in Japanese Patent Application Laid-Open No. H10-157, a method of checking whether a program developed on a RAM is broken or not and indirectly checking the RAM by this program check is performed. As a method of this program check, RA is divided into a plurality of program blocks.
A system is used in which a program stored in M is checked one block at a time by a periodic timer interrupt during the operation of the system.

FIG. 4 shows a block diagram of the above-mentioned conventional program check system. FIG. 5 is a block diagram showing a configuration of a computer system having the program check method of FIG. In the conventional program check method shown in FIG. 4, a check target program storage unit 1 that stores a check target program divided into a plurality of program blocks and an interrupt is generated at a cycle that does not significantly impair the processing performance of the system. A timer interrupt generating unit 2, a check block specifying unit 3 for specifying a program block to be checked, a checksum of the specified program block is calculated, and a cumulative value obtained by adding the calculated checksum of each program block is calculated. A program check processing unit 4b for checking a program abnormality by checking against a checksum of the entire program created in advance, and a checksum storage for saving a checksum of each program block calculated by the program check processing unit 4b. Part 5b Dimethyl created, the checksum storage unit 6 for storing a checksum of the program, and a program anomaly correction unit 7 for performing the corresponding processing when the abnormality of the program is detected.

The computer system having the program check method shown in FIG.
The CPU 12b having a built-in timer interrupt generation unit 2, check block designation unit 3, program check processing unit 4b, and program abnormality response processing unit 7, an area for storing a main processing program and a program check program, and a check target program A ROM 13b having an area for storing the calculated checksum, ie, the checksum storage unit 6 in FIG.
An area for storing a main processing program and a program check program, which are programs to be checked, read from the ROM 13b by the CPU 12b, that is, an area for storing the check target program storage unit 1 of FIG. RA having 4 checksum storage units 5b
M17b.

Next, the operation will be described.

First, the main operation of the computer system will be described. FIG. 6 is a flowchart showing a main operation in the computer system having the program check method of FIG. When the power is turned on or the CPU 12b in FIG. 5 is turned on or the reset is executed, the main processing program and the program check program read from the ROM 13b by the CPU 12b are divided into appropriate sizes in advance so as not to compress the main processing of the system. Into blocks and as multiple program blocks,
The CPU 12b stores it in the check target program storage 1 (step 51 in FIG. 5). The CPU 12b is started by the main processing program stored in the check target program storage 1, and executes a predetermined initialization process (step 52). Next, when the value of the interrupt cycle of the timer interrupt generation unit 2 for periodically checking the program is set to an appropriate value by the operator so as not to impair the operation of the system (step 53). CP
The U12b starts the main processing by the main processing program, and sequentially executes the processing repeatedly given until the timer interrupt generation unit 2 in which the interrupt generation time is set times out and an interrupt occurs (step 54).
When the timer interrupt generation unit 2 of the CPU 12b generates a timer interrupt at the set cycle, the CPU 12b temporarily suspends the execution of the main processing and executes the program check processing (Step 55). In the routine of the program check process, the result of the program check performed at that time is determined (step 56). If the result is normal, the program check process assigned at the time of the timer interrupt is terminated, and the main process of step 54 Return to execution. The check of the program block to be checked is not performed for all program blocks at the same time of one timer interrupt, but every time a timer interrupt is performed, the check target program divided into multiple blocks is checked for each block It is checked sequentially. If the result of the program check is abnormal in step 56, a check error is displayed and the program check process ends (step 57). Thereafter, the CPU 12b resets itself (step 58) and restarts. I do.

Next, the operation of the program check program will be described. FIG. 7 is a flowchart showing the operation in the program check method of FIG.

When the timer interrupt shown in FIG. 6 is performed by the timer interrupt generator 2 shown in FIG. 4 (step 7 in FIG. 7).
1) First, the timer interrupt generation unit 2 activates the check block designating unit 3, and the check block designating unit 3 performs an interrupt by the timer from among a plurality of program blocks stored in the check target program storage unit 1. At this time, a program block to be checked is specified and notified to the program check processing unit 4b (step 72). FIG. 7 shows a state in which the i-th program block of the n program blocks is specified by the check block specifying unit 3. The program check processing unit 4b reads the checksum from the checksum storage unit 5b (step 73), calculates the checksum for the program block specified by the check block specifying unit 3, and checks the obtained program block. The sum is added to the checksum read from the checksum storage 5b (step 74) and saved in the checksum storage 5b (step 75). Next, the program check processing unit 4b
It is checked whether or not the above processing has been performed on all the program blocks and the addition of the checksum has been completed (step 76). The above processing has not been performed on all the program blocks, and the above processing has been completed. If there is a program that has not been executed, the process ends as a program check for the program block at that time, and the operation is passed to the main routine. CPU 12b
Restarts the main processing that was interrupted by the end of the program check processing. Further, when an interrupt of the next timer occurs, the operation from step 72 on the program block next to the program block on which the previous program check was performed, and the operation from step 72 are repeated as before.
As described above, in the program check process, each time an interrupt for the program check is performed, each block is checked for each program block in a predetermined order, and a checksum of the block is obtained. Is repeated from the checksum read out from the checksum storage unit 5b, which is the total value of the checksums of the program blocks that have been completed, and stored in the memory 5. After the checksum of the last program block has been added, step 7
When it is determined in step 6 whether addition of all blocks has been completed, the program check processing unit 4b reads out a checksum created in advance and stored in the checksum storage unit 6 (step 77). It is checked whether the checksum matches the total value of the checksums of all the blocks stored in the checksum storage 5b (step 78). As a result of this collation, if they match, the program check processing unit 4b determines that there is no abnormality, clears the checksum (step 79), passes the operation to the main routine, and ends the check processing by the timer interrupt. The operation of the main routine returns to the normal processing at the end of the program check processing by the timer interrupt, but the periodic program check processing by the timer interrupt is continued. If they do not match, the program check processing unit 4b determines that the program is abnormal and activates the program error handling unit 7, and the program error handling unit 7 determines whether the detected program is abnormal. A predetermined response process is performed (step 80).

[0009]

In the above-mentioned conventional program check method, each time a timer interrupt is performed, a plurality of divided program blocks are checked one by one to obtain their respective checksums. After the block check is completed, the checksum of each program block is added to obtain the checksum of the entire program, and then it is compared with the checksum calculated in advance. Unless the program check processing is performed to the end, it is impossible to detect a program abnormality, and there is a disadvantage that the time required to detect the program abnormality increases as the size of the program increases.

[0010] An object of the present invention is to prevent a time required for detecting a program error from increasing due to an increase in the size of a program, and to enable a quick response when a program error is detected. It is to provide a check method.

[0011]

According to the program check method of the present invention, when a program check is performed by using a checksum to detect an abnormality in a program of a CPU, the program is divided into a plurality of program blocks and the RAM is divided into a plurality of program blocks. And a timer that operates periodically
The checksum of one program block is calculated and stored in one interrupt, and the program block to be checked is sequentially changed for each interrupt by the timer. Calculating a checksum, calculating an accumulated value of the checksums of the plurality of program blocks, and comparing the accumulated value with a previously calculated checksum of the program to determine whether there is an abnormality in the program. In the program check method, for each interruption by the timer, the checksum is newly calculated for one block of the program block to be checked at that time, and the checksum previously stored and updated with the newly calculated checksum is updated. And said check Of the plurality of program blocks by adding the updated checksum of the program block and the checksums stored so far of all the remaining program blocks, and calculating the sum of the checksums of the plurality of program blocks. It is configured to collate with a checksum to determine the presence or absence of an abnormality in the program for each interruption by the timer.

Further, the program check method of the present invention provides (A) a rewritable check target program storage means for storing a check target program divided into a plurality of program blocks, and (B) a periodic interrupt is generated. Timer interrupt generating means, (C) first checksum storing means for storing a checksum of the program calculated in advance, (D) activated by the output of the timer interrupt generating means, and each time it is activated, Check block designating means for sequentially designating one program block of a plurality of program blocks stored in the target program storage means; (E) reading the program block designated by the check block designating means, and checking the program block checksum After calculating It calculates the cumulative value of all the checksums of the lock, by matching the checksum stored in the said accumulated value first checksum storage means,
Program check processing means for determining that the program is abnormal when both values do not match; and (F) second checksum storage means for storing a calculated checksum of the program block, wherein the check target In a program check method in a computer system that executes the program expanded on a program storage means, the program check processing means includes a check block designating means each time a periodic interrupt is performed by the timer interrupt generating means. Calculates a checksum for one block of the program block designated by the above and stores the checksum in the second checksum storage means. The second checksum is calculated for the newly calculated checksum and all the remaining program blocks. From storage means The calculated checksum of the program block is added to the calculated checksum to calculate the total value of the checksum of the program block, and is compared with the checksum of the first checksum storage means. It is configured to determine the presence or absence.

[0013]

Next, embodiments of the present invention will be described with reference to the drawings.

FIG. 1 is a block diagram showing an embodiment of a program check method according to the present invention. FIG. 2 shows FIG.
FIG. 2 is a block diagram showing a configuration of a computer system having the program check method.

The program check method according to the present embodiment shown in FIG. 1 includes a check target program storage unit 1 for storing a check target program divided into a plurality of program blocks, and a check program storage unit 1 which does not significantly impair the processing performance of the system. Calculated by a timer interrupt generating unit 2 for generating an interrupt at a periodic interval, a check block specifying unit 3 for specifying a program block to be checked, a program check processing unit 4a for checking the normality of a program, and a program check processing unit 4a. Checksum storage 5 for saving the checksum of each program block
a, a checksum storage unit 6 that stores a checksum of the program created in advance, and a program error handling unit 7 that performs a process when an error of the program is detected.

The computer system shown in FIG. 2 and having the program check method of FIG.
CPU 12a having a built-in timer interrupt generation unit 2, check block designation unit 3, program check processing unit 4a, and program abnormality handling processing unit 7, an area for storing a main processing program and a program check program, and a A ROM 13a having an area for storing a checksum, that is, a checksum storage unit 6 shown in FIG.
a area for storing a main processing program and a program for checking programs, which are programs to be checked, read out from the storage area a, that is, an area for storing the checksums of a plurality of program blocks while having the check target program storage unit 1 of FIG. And a RAM 17a having one checksum storage unit 5a.

Next, the operation will be described.

First, the operation of the program check program will be described. FIG. 3 is a flowchart showing an example of the operation in the program check method of FIG.

When the timer interrupt shown in FIG. 6 is performed by the timer interrupt generation unit 2 in FIG. 1 (step 2 in FIG. 3).
1) First, the timer interrupt generation unit 2 activates the check block designating unit 3, and the check block designating unit 3 performs an interrupt by the timer from among a plurality of program blocks stored in the check target program storage unit 1. At this time, a program block to be checked is specified and notified to the program check processing unit 4a (step 22). FIG. 3 shows a state in which the i-th program block of the n program blocks is specified by the check block specifying unit 3. The program check processing unit 4a calculates a checksum for the program block specified by the check block specifying unit 3 (step 23), and saves the checksum in the checksum storage unit 5a using the obtained checksum of the program block. The corresponding checksum is updated (step 24). Next, the program check processing unit 4a adds all the checksums of the first block to the n-th block saved in the checksum storage unit 5a (step 25). Program check processing unit 4a
Reads out the checksum created in advance and stored in the checksum storage unit 6 (step 26), and checksums the sum of the checksum and the checksum of all n blocks saved in the checksum storage unit 5a. Is checked to see if they match (step 27). As a result of this comparison, if the two match, the program check processing unit 4a determines that there is no abnormality, clears the accumulated value of the checksum of each program block, passes the operation to the main routine, and executes the check processing by the timer interrupt. finish. The operation of the main routine returns to the normal processing at the end of the check processing by the timer interrupt, but the check processing by the timer interrupt that is periodically performed continues. If they do not match, the program check processing unit 4a determines that the program is abnormal and activates the program error handling unit 7, and the program error handling unit 7 determines whether the detected program is abnormal. A predetermined response process is performed (step 28).

The above processing is performed for all n program blocks. When an interrupt for a program check is performed, one of the n program blocks is designated and checked in a predetermined order, only the checksum of the program block is calculated, and the checksum is stored in the checksum storage unit 5a. Of the saved checksums of the n blocks, only the checksum of the program block for which the above checksum has been calculated is updated with the above calculated checksum. The checksums of the other program blocks are not updated and remain as they were. In such a state, the checksums of all the n program blocks saved in the checksum storage unit 5a are added and compared with the checksums created in advance and stored in the checksum storage unit 6. However, as mentioned above, 1
In the first interrupt, only one of the n program blocks has its checksum calculated, and the checksum storage unit 5 corresponding to the data is calculated.
Since only the checksum saved in a is updated, the number of interrupts has not yet reached n when the system has just been started up, that is, the number of interrupts of n program blocks has not been reached. When the checksum has not been calculated for some of the program blocks, the checksum cannot be compared with the checksum of the entire program stored in the checksum storage unit 6, so that the checksum of the n program blocks is not obtained. It is necessary to wait for the collation until the stage where all have been calculated.

In the above description, the interrupt by the timer is described as being performed periodically, but may be performed at an arbitrary determined time.

[0022]

As described above, according to the program check method of the present invention, when a timer interrupt is performed and the checksum of one designated program block is newly calculated, the processing of the timer interrupt at that time is as follows. For the checksums of other program blocks, the total checksum of each program block is calculated using the checksum as before, and the checksum of the entire program is obtained. Since the entire checksum can be compared with the checksum calculated in advance, the time required to detect a program error can be prevented from increasing as the size of the program increases. Allows quick response when anomalies are detected To have the advantage of being able to minimize the various adverse effects occur when abnormality of program is detected.

[Brief description of the drawings]

FIG. 1 is a block diagram showing an embodiment of a program check method according to the present invention.

FIG. 2 is a block diagram showing a configuration of a computer system having the program check method of FIG.

FIG. 3 is a flowchart illustrating an example of an operation in the program check method according to the present embodiment.

FIG. 4 is a block diagram of a conventional program check method.

FIG. 5 is a block diagram showing a configuration of a computer system having the program check method of FIG.

6 is a flowchart showing a main operation in a computer system having the program check method of FIG.

FIG. 7 is a flowchart showing an operation in the program check method of FIG. 4;

[Explanation of symbols]

 1 Check target program storage unit 2 Timer interrupt generation unit 3 Check block designation unit 4a, 4b Program check processing unit 5a, 5b Checksum storage unit 6 Checksum storage unit 7 Program abnormality handling processing unit 12a, 12b CPU 13a, 13b ROM 17a, 17b RAM

Claims (2)

[Claims] When a program check is performed using a checksum to detect an abnormality in a program of a CPU, the program is divided into a plurality of program blocks, stored in a RAM, and a timer that operates periodically.
The checksum of one program block is calculated and stored in one interrupt, and the program block to be checked is sequentially changed for each interrupt by the timer. Calculating a checksum, calculating an accumulated value of the checksums of the plurality of program blocks, and comparing the accumulated value with a previously calculated checksum of the program to determine whether there is an abnormality in the program. In the program check method, for each interruption by the timer, the checksum is newly calculated for one block of the program block to be checked at that time, and the checksum previously stored and updated with the newly calculated checksum is updated. And said check Of the plurality of program blocks by adding the updated checksum of the program block and the checksums stored so far of all the remaining program blocks, and calculating the sum of the checksums of the plurality of program blocks. A program check method, wherein a check is made against a checksum to determine whether or not the program is abnormal for each interruption by the timer. (A) a rewritable check target program storing means for storing a check target program divided into a plurality of program blocks; (B) a timer interrupt generating means for generating a periodic interrupt; A) first checksum storage means for storing a checksum of the program calculated in advance; (D) activated by an output of the timer interrupt generation means, and stored in the check target program storage means each time the program is activated. Check block designating means for sequentially designating one program block of a plurality of program blocks; (E) reading out the program block designated by the check block designating means and calculating the checksum of the program block, Calculate the cumulative value of all checksums of To, and collated with the check sum stored said accumulated value and said first checksum storing means, the program check processing means determines that the program is abnormal when both values do not match,
(F) a second checksum storing means for storing the calculated checksum of the program block, wherein the program check method in a computer system for executing the program developed on the check target program storing means comprises: The program check processing means calculates a checksum for one program block designated by the check block designation means every time a periodic interruption is performed by the timer interrupt generation means, and calculates the second checksum. The checksum stored in the storage means is added to the newly calculated checksum and the checksums read from the second checksum storage means for all the remaining program blocks. Is calculated, Serial first checksum storage means checksums and then matching the program check method, which comprises determining the presence or absence of an abnormality of the program for each interrupt by the timer.