JPH064499A - Processor diagnostic system - Google Patents

Abstract

PURPOSE:To appropriately deal with processor abnormality during a system operation by detecting the processor abnormality even during the system operation in addition to diagnosing respective processors at the time of starting the system and storing diagnostic result, thereby replacing the result with the diagnostic result in an initial state diagnosis at the time of starting a system. CONSTITUTION:This system is provided with an initial state diagnostic means 1 for diagnosing the functions or the like of the own processor by a diagnostic program on the respective processors at the time of starting the system, a diagnosed result storage means 2 for storing the diagnosed result on a local processor by the diagnostic program on a main processor in a diagnosed result storage area 3 provided on the main processor, a processor abnormality detection means 5 at the time of operation by the communication program on the own local processor and the main processor and a diagnosed result replacing means 4 for replacing the initial state diagnosed results in the diagnosed result storage area by the communication program on the main processor when the processor abnormality at the time of the operation is generated.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a processor diagnosis system in a multiprocessor system.

[0002]

2. Description of the Related Art FIG. 3 is a block diagram showing an example of a multiprocessor system to which a conventional processor diagnosis method is applied. Generally, in the conventional processor diagnosis method in a multiprocessor system, as shown in FIG. 3, each local processor 33, 35,
The diagnostic programs 34, 36, and 38 operating on 37 self-diagnose, and the diagnostic program 31 operating on the main processor 30 uses the diagnostic results to detect the main processor 3
0 is stored in the diagnostic result storage area 32.

[0003]

In the conventional processor diagnosis method described above, each local processor self-diagnoses only when the system is started up and the diagnosis result is stored in the main processor. When you refer to the diagnosis result before doing it and prevent the error and the illegal operation, only the initial state of the system is affected.
It has the drawback that no measures are taken against processor abnormalities during operation.

In addition to the initial diagnosis of the processor, the object of the present invention is to notify the main processor to the effect that a processor abnormality is detected on the local processor even during operation, and cannot detect the abnormality on the local processor. For the system, by detecting the processor abnormality of the local processor on the main processor depending on the response of the local processor to the polling from the main processor, the result is replaced with the diagnosis result in the initial diagnosis at system startup. It is to provide a processor diagnosis method in a multiprocessor system capable of coping with a processor abnormality in the inside.

[0005]

SUMMARY OF THE INVENTION A processor diagnostic system of the present invention uses an initial diagnostic means for diagnosing the function of its own processor by a diagnostic program on each processor at system startup and a local diagnostic program on the main processor. In a processor diagnostic method in a multiprocessor system including a diagnostic result storage means for storing an initial diagnostic result on a processor in a diagnostic result storage area provided on a main processor, in a case of operation by a communication program on a local processor itself and a main processor. The processor abnormality detecting means and the diagnostic result replacing means for replacing the initial diagnostic result in the diagnostic result storage area by the communication program on the main processor when a processor abnormality occurs during operation.

[0006]

Embodiments of the present invention will now be described with reference to the drawings. FIG. 1 is a block diagram showing an embodiment of a processor diagnosis system of the present invention. 2 is a block diagram showing an example of a multiprocessor system to which this embodiment is applied. As shown in FIG. 1, this embodiment includes an initial diagnosis means 1, a diagnosis result storage means 2, a diagnosis result storage area 3, a processor abnormality detection means 5 during operation, and a diagnosis result replacement means 4. Has been done.

Next, referring to FIG. 2, the storage and replacement of the diagnosis result by the initial diagnosis of the processor in the multiprocessor system and the detection of the processor abnormality during the operation will be described. First, when the power of the system is turned on, the diagnostic program 11,
15, 18, and 21 operate to diagnose their own processor functions and the like.

The diagnostic result is the diagnostic program 15, 18, 2 on each local processor 14, 17, 20.
1, the diagnostic program 1 on the main processor 10
1 is notified and the diagnostic result indicates that the processor operates normally, the diagnostic program 11 on the main processor 10 sets “1” to the corresponding bit of the diagnostic result storage area 2 on the main processor 10. When it is set and the operation is defective, the corresponding bit of the diagnostic result storage area 12 on the main processor 10 is left as "0".

After that, when a program for communicating between the processors is activated from one processor to another processor, the software instruction refers to the diagnostic result storage area 12 on the main processor 10. At this time,
When the bit corresponding to the processor of the communication destination is "1", the software instruction is executed and an interrupt is generated for the processor. If the bit corresponding to the processor of the communication destination is "0", execution of the software is stopped.

When a processor abnormality is detected on the local processor 14, 17, 20 during system operation, the fact is notified from the communication program 16, 19, 22 on the local processor 14, 17, 20 on the main processor 10. The communication program 13 is notified.

Further, the local processors 14, 17,
If the processor abnormality cannot be detected on the local processor 20, the local processor 14, 17, for the polling from the communication program 13 on the main processor 10,
The communication program 13 on the main processor 10 detects an abnormality in the local processors 14, 17, 20 depending on the presence or absence of a response from the communication programs 16, 19, 22 on 20.

Then, the communication program 13 on the main processor 10 causes the local processors 14, 17, 2 to operate.
When an abnormality of 0 is detected, the communication program 13 on the main processor 10 replaces the corresponding bit in the diagnostic result storage area 12 on the main processor 10 with “0”.

[0013]

As described above, according to the processor diagnosis method of the present invention, in addition to diagnosing each processor when the system is started up and storing the diagnosis result, the system can be operated on the local processor even during system operation. When a processor abnormality is detected, the fact is notified to the main processor, and if it cannot be detected on the local processor, the presence or absence of a response from the local processor to the polling from the main processor indicates whether or not the system was started up. Since the diagnosis result in the initial diagnosis is replaced, by referring to the diagnosis result before performing inter-processor communication, it is possible to appropriately deal with a processor abnormality during system operation.

[Brief description of drawings]

FIG. 1 is a block diagram showing an embodiment of a processor diagnosis system of the present invention.

FIG. 2 is a block diagram showing an example of a multiprocessor system to which this embodiment is applied.

FIG. 3 is a block diagram showing an example of a multiprocessor system to which a conventional processor diagnosis method is applied.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Initial diagnosis means 2 Diagnostic result storage means 3 Diagnostic result storage area 4 Diagnostic result replacement means 5 Processor abnormality detection means during operation 10 Main processor 11 Diagnostic program 12 Diagnostic result storage area 13 Communication program 14 Local processor 15 Diagnostic program 16 Communication program 17 Local Processor 18 Diagnostic Program 19 Communication Program 20 Local Processor 21 Diagnostic Program 22 Communication Program 30 Main Processor 31 Diagnostic Program 32 Diagnostic Result Storage Area 33 Local Processor 34 Diagnostic Program 35 Local Processor 36 Diagnostic Program 37 Local Processor 38 Diagnostic Program

Claims (1)

[Claims] 1. An initial diagnostic means for diagnosing the function of its own processor by a diagnostic program on each processor when the system is started up, and an initial diagnostic result on the local processor by a diagnostic program on the main processor on the main processor. In a processor diagnostic method in a multiprocessor system including a diagnostic result storage unit for storing in a diagnostic result storage area, a processor abnormality detection unit during operation by a communication program on a local processor itself and a main processor, and a processor abnormality during operation And a diagnostic result replacement means for replacing the initial diagnostic result in the diagnostic result storage area by a communication program on the main processor when the processor diagnostic system is used.