JPH09160843A - Method and device for diagnosing hardware - Google Patents

Abstract

PROBLEM TO BE SOLVED: To automatically make a diagnosis without placing any load on a CPU by detecting abnormality occurring in an input device, an input/output controller, or various interfaces when there is no access. SOLUTION: A bus monitor device 4 monitors a control line connected to an I/O 3 by a control line monitor part 4a, and sets status information showing a state wherein there is no access on a system bus 2 for a specific time corresponding to a time-up time set previously in a timer 4c in a status register 4b and issues an interruption to the CPU 1 once detecting the said state. The CPU 1 once receiving the interruption from the bus monitor device 4 issues a diagnostic program to the I/O 3 to be diagnosed. This method does not load the CPU 1 with the detection of the no-access state and can make an efficient diagnosis by making good use of the time when there is no access.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a hardware diagnosing method for diagnosing whether a function or operation of hardware in a computer system is normal, and a hardware diagnosing apparatus used for the method.

[0002]

2. Description of the Related Art FIG. 14 shows a plurality of input / output devices (hereinafter, I
Is a block configuration diagram showing a computer system in which (/ O) is connected to a CPU via a system bus. In the figure, 1 is a CPU, 2 is a system bus, and 3a to 3c.
Is I / O.

Next, when diagnosing each I / O in such a computer system, a diagnostic OS is loaded on the computer that executes the diagnostics, and each I / O is loaded on the loaded OS.
/ O is executed by executing various diagnostic programs as needed. Further, in this case, there is no means for judging the load of the system bus 2 in any of the modules constituting the computer system, and the CPU 1 does not consider the load state of the system bus and does not consider the load state of the system bus. Diagnosing.

FIG. 15 is a block diagram showing the configuration of a computer system obtained by further expanding the computer system shown in FIG. In the figure, parts that are the same as or correspond to those in FIG. 14 are assigned the same reference numerals and explanations thereof are omitted. Reference numerals 4'and 5'are input / output interfaces for the disk device, for example, the input / output interface 4'is arranged on the CPU side, and the input / output interface 5'is arranged on the disk device side.
They are connected in series.

Next, the operation of the hardware diagnosing method performed in this computer system will be described. When diagnosing each I / O in such a computer system, the CPU 1 diagnoses the I / O 3a, 3c and the input / output interface 4 '. In particular, for the I / O interface 4 ', the integrated diagnosis including the I / O interface 5'and the disk device 6 will be performed. Independent diagnosis cannot be performed, and a diagnostic program for diagnosing I / O 3a, a diagnostic program for diagnosing I / O 3c, an I / O interface 4 ', an I / O interface 5', and a disk device 6 are integrally diagnosed. Like the diagnostic program, it is necessary to execute the three diagnostic programs separately by inputting three commands.

Further, as a hardware diagnostic device for diagnosing the hardware of such a computer system, especially a fault occurring in I / O, a "health check of input / output device" disclosed in Japanese Patent Laid-Open No. 5-108509. There is a device. The health check device of this input / output device uses the above-mentioned I / O device when normal access to the I / O is frequently used.
The I / O failure detection operation is stopped, and the I / O failure detection operation is started when the usage frequency is low.

[0007]

Since the conventional hardware diagnosing method and its apparatus are configured as described above, the frequency of access is low when the diagnostic program is executed. The access in the low state is invalid while the diagnostic program is being executed, and there is a problem that the access or service of the CPU 1 does not reach the I / O to be diagnosed.

When a failure occurs in a computer system in which the input / output interfaces are hierarchized, the CPU 1 causes the input / output interface 5'and the disk device 6 via the input / output interface 4 '.
Therefore, there is a problem that it is not possible to discriminate and determine in which layer the fault is due to the integrated diagnosis.

The present invention has been made to solve the above problems, and determines the load state of the system bus by monitoring the system bus, especially the control line, and access is not performed for a predetermined time. The load on the CPU for the hardware in the state is reduced, or the CPU is
An object of the present invention is to provide a hardware diagnostic method and an apparatus therefor, which automatically diagnoses without burdening the user.

Further, the present invention provides a hardware diagnosing method and an apparatus therefor capable of efficiently executing diagnosis by the same command even if devices such as input / output interfaces of different types exist. To aim.

Further, according to the present invention, when the input / output device, the input / output control device, various interfaces, etc. are hierarchized, the return diagnosis is performed for each device so that it is possible to precisely determine in which layer the fault has occurred. It is an object of the present invention to obtain a hardware diagnostic method and an apparatus therefor which make it possible to isolate an I / O in which a failure has occurred before it develops into a failure in the entire computer system.

[0012]

According to a first aspect of the present invention, there is provided a hardware diagnosing method, wherein a control line of a system bus is monitored to detect a state in which no access is made for a predetermined time.
The CPU uses the interrupt signal generated based on the detection result.
Executes a diagnostic program and detects an abnormality occurring in the input / output device, the input / output control device, or various interfaces when there is no access.

In the hardware diagnosing method according to a second aspect of the present invention, the CPU indicates the various data on the system bus at the beginning of diagnosis by the diagnostic program by the notification indicating that there is no access from the state of the system bus. It is provided with a configuration in which it is confirmed that there is no exchange, and after the confirmation, the process according to the started diagnostic program is advanced.

According to a third aspect of the present invention, there is provided a hardware diagnosing method, wherein a CPU indicates various types of information on the system bus by a notification indicating that there is an access detected from the state of the system bus at the beginning of diagnosis by a diagnostic program. The CPU is configured to forcibly end the execution of the started diagnostic program when it is confirmed that data has been transferred.

According to a fourth aspect of the present invention, there is provided a hardware diagnosing method, wherein log data for diagnosing an input / output device, an input / output control device, or various interfaces is saved on the side where an interrupt signal is generated. is there.

In the hardware diagnosing method according to a fifth aspect of the present invention, the access monitoring means provided in addition to the CPU, the input / output device, the input / output control device or various interfaces monitors the control line of the system bus for access. Is detected for a predetermined time, the C
The PU, the input / output device, the input / output control device, or a microprocessor provided other than the various interfaces executes a diagnostic program, and the access monitoring unit monitors the control line of the system bus to access. When the occurrence is detected, the running diagnostic program is forcibly terminated, and the microprocessor diagnoses the input / output device or the input / output control device or various interfaces when there is no access. is there.

According to a sixth aspect of the present invention, there is provided a hardware diagnosing method in which a microprocessor provided other than a CPU, an input / output device, an input / output control device, or various interfaces has a system bus at the beginning of diagnosis by executing a diagnostic program. From the above state, it is confirmed that there is no access again, and after the confirmation, the diagnosis by the diagnosis program for the input / output device, the input / output control device, or the various interfaces is advanced.

According to a seventh aspect of the present invention, there is provided a hardware diagnosing method, wherein a diagnosis result of an input / output device, an input / output control device or various interfaces is used except for the CPU, the input / output device, the input / output control device and the various interfaces. It is provided with a configuration for saving in the memory used by the microprocessor provided in.

According to an eighth aspect of the present invention, there is provided a hardware diagnosing method, wherein an interrupt signal generated and output when a state in which no access is made for a predetermined time is detected by monitoring a control line of a system bus is input. The output device, the input / output control device, or the various interfaces may be the other input / output device other than itself, the input / output control device, or the various types other than itself via the system bus by the microprocessor and the diagnostic program installed in the self. The interface device is provided with a configuration for making a diagnosis.

In the hardware diagnosing method according to a ninth aspect of the present invention, the order of diagnosis is determined in advance as a priority order, and the input / output device, the input / output control device, or the various interfaces are themselves determined according to the priority order. A microprocessor and a diagnostic program installed in the system provide a configuration for diagnosing the input / output device other than itself, the input / output control device, or the various interface devices via the system bus.

According to a tenth aspect of the present invention, in a hardware diagnosing method, an input / output device, an input / output control device, or various interfaces each monitor a control line of the system bus to detect a state in which there is no access for a predetermined time. Then, an interrupt signal is output based on the detection result, and the input / output device or the input / output control device that receives the interrupt signal or the input / output device that outputs the interrupt signal from various interfaces or the Diagnosis is performed by executing a diagnostic program owned by itself for various interfaces, and after the diagnosis is completed, an interrupt signal is output from the side that performed the diagnosis, and this time the input / output device that has received the diagnosis or The input / output control device or the various interfaces are Those having a structure for performing diagnosis by executing the diagnostic program is itself held against that side was the diagnosis that accepts the interrupt signal.

In the hardware diagnosing method according to an eleventh aspect of the present invention, time data defining a predetermined time when detecting a state of no access by monitoring a control line of a system bus is provided to a side to be diagnosed. I / O device or I / O control device or I / O device or I / O with high priority when receiving a diagnosis The control device or the various interfaces are provided with a configuration for receiving a diagnosis prior to the input / output device having a low priority, the input / output control device, or the various interfaces.

According to a twelfth aspect of the present invention, there is provided a hardware diagnosing method, wherein an input / output device, an input / output control device, or various interfaces each monitor a control line of a system bus to detect a state in which no access is made for a predetermined time. An input / output device, an input / output control device, or various interfaces that output the interrupt signal based on the detection result, and the other input / output device or input / output control device that receives the interrupt signal, or various interfaces Other than the I / O device, the I / O control device, or the various interfaces to execute a diagnostic program that the user has to make a diagnosis, and the I / O device or the I / O control device or the I / O control device that is the diagnosis target in the diagnosis is executed. For the various interfaces Only when the disconnection result is "normal", after the diagnosis is completed, the side that has performed the diagnosis outputs an interrupt signal, and the input / output device or the input / output control device or the The input / output device other than the various interfaces, the input / output control device, or the various interfaces accepts the output interrupt signal, so that the input / output device, the input / output control device, or the various interfaces that are not the parties involved in the diagnosis On the other hand, it is provided with a configuration for making a diagnosis by executing a diagnosis program that the user has.

According to a thirteenth aspect of the present invention, in the hardware diagnosing method, the I / O device, the I / O controller, or the various interfaces that have made the diagnosis save the log data at the time of the diagnosis in a predetermined memory area. It is equipped with a configuration.

In the hardware diagnosing method according to the fourteenth aspect of the present invention, the input / output device or the input / output control device or various interfaces which are diagnosed and whose diagnostic result is judged to be "abnormal" indicate that the diagnostic result is "abnormal". Based on the interrupt signal output from the I / O device or input / output control device or various interfaces on the diagnosing side, the log data of the time when it is judged to be “abnormal” is stored in the memory of its own memory. It is provided with a configuration for storing in a predetermined area.

According to a fifteenth aspect of the present invention, in the hardware diagnosing method, the contents of the diagnosis performed on the input / output device, the input / output control device, or the various interfaces, and the input / output device, the input / output control device, or the various interfaces. A command common to the input / output device, the input / output control device, or the various interfaces for performing the diagnosis according to the above contents for each interface is determined in advance, and the CPU outputs the input / output device or the input / output control device. Alternatively, the input / output device, the input / output control device, or the various interfaces may be configured to perform a self-diagnosis according to the content of the predetermined diagnosis by a diagnostic program by the command given to the various interfaces. And .

According to a sixteenth aspect of the present invention, in the hardware diagnosing method, the CPU for executing the diagnostic program, the input / output device, the input / output control device, or various interfaces is configured so that the state of the system bus is started at the beginning of the diagnosis by the diagnostic program. From the above, it is confirmed that various kinds of data are not exchanged on the system bus, and after the confirmation, the processing according to the started diagnostic program is performed.

According to a seventeenth aspect of the present invention, in the hardware diagnosing method, the CPU for executing the diagnostic program, the input / output device, the input / output control device or various interfaces is configured so that the state of the system bus is started at the beginning of the diagnosis by the diagnostic program. From the above, when it is confirmed that various kinds of data are exchanged on the system bus, the execution of the started diagnostic program is forcibly terminated.

In the hardware diagnosing method according to the eighteenth aspect of the present invention, various data is transmitted and received between a control line of the system bus and an input / output device or input / output control device or various interfaces connected hierarchically. The CPU detects a state in which there is no access for a predetermined time by monitoring the state of, and the CPU executes a diagnostic program when the access is not performed for a predetermined time, and the CPU is connected in a hierarchical manner. The output device, the input / output control device, or the various interfaces from the input / output device, the input / output control device, or the various interfaces in a higher layer to the input / output device, the input / output control device, or the various interfaces in a lower layer The diagnosis is performed in the order of Rui is obtained by so provided an arrangement for detecting an abnormality occurring in the input-output control unit or the various interfaces.

A hardware diagnosing method according to a nineteenth aspect of the present invention is an input / output device or an input / output control device or various interfaces connected to a system bus, a hierarchical input / output device, an input / output control device or various interfaces. For the above, the contents of the diagnosis performed for the highest-level I / O device, the I / O control device, or the various interfaces, and the diagnosis according to the above contents for the I / O device, the I / O control device, or the various interfaces, respectively A command common to the input / output device, the input / output control device, or the various interfaces for performing the above is determined in advance, and the command given from the CPU to the input / output device, the input / output control device, or the various interfaces is determined. The input / output device, the input / output control device, or the various interfaces each perform self-diagnosis according to the predetermined diagnostic content by a diagnostic program, and the hierarchically connected input / output device Alternatively, as for the input / output control device or the various interfaces, the input / output device or the input / output control device of the highest layer or the input / output device or the input / output control device of the lower layer from the various interfaces It is equipped with a configuration for performing diagnosis in order to various interfaces.

According to a twentieth aspect of the present invention, there is provided a hardware diagnosing method for an input / output device, an input / output control device, various interfaces, and the hierarchical input / output device, the input / output control device, or various interfaces. As for the uppermost input / output device, the input / output control device, or various interfaces, the contents of the diagnosis are performed, and the input / output device, the input / output control device, or the various interfaces are diagnosed according to the above contents. A common command for predetermining a predetermined time is used, and the microprocessor, which is provided other than the CPU, the input / output device, the input / output control device, and the various interfaces, monitors the control line of the system bus to access for a predetermined time. When a condition without The microprocessor executes a diagnostic program, and when the microprocessor does not access the input / output device, the input / output control device, or the various interfaces, the microprocessor gives the input / output device or the input / output control device or A self-diagnosis for each of the various interfaces is performed according to the content of the predetermined diagnosis, and the hierarchical input / output device, the input / output control device, or the various interfaces are
A self-diagnosis by the diagnostic program is performed in the order of the input / output device or the input / output control device or the various interfaces of the highest layer to the input / output device, the input / output control device or the various interfaces of the lower layer. It has a configuration for performing.

According to a twenty-first aspect of the present invention, there is provided a hardware diagnosing device, which comprises a bus monitoring device for detecting whether or not there is no access for a predetermined time by monitoring a control line of a system bus, and the bus monitoring device. The CPU for accepting an interrupt signal generated and output by the bus monitoring device when a state in which the access is not performed for a predetermined time is detected is used by the CPU to diagnose the input / output device, the input / output control device, or various interfaces. It is provided with a diagnostic program provided on the side, and storage means for storing the data obtained by executing the diagnostic program as log data in the bus monitoring device.

According to a twenty-second aspect of the present invention, there is provided a hardware diagnosing device, which is provided at the beginning of diagnosing by a diagnosing program.
When the bus monitoring device detects that there is no access from the state of the system bus and the notification sent from the bus monitoring device, the CPU confirms that various data is not transmitted and received on the system bus. After the confirmation, the configuration is such that the processing according to the started diagnostic program is advanced.

According to a twenty third aspect of the present invention, there is provided a hardware diagnosing apparatus, which is provided at the beginning of diagnosing by a diagnosing program.
When the CPU confirms by the notification sent from the bus monitoring device that the bus monitoring device detects that there is an access from the state of the system bus, the CPU confirms that various data is transmitted and received on the system bus, It is provided with a configuration for forcibly ending the execution of the started diagnostic program.

According to a twenty-fourth aspect of the present invention, there is provided a hardware diagnosing device, which has a diagnostic program for diagnosing an input / output device, an input / output control device or various interfaces, and a predetermined access by monitoring a control line of a system bus. And a control line monitoring unit for detecting a state where there is no time, and the control line monitoring unit monitors the control line of the system bus to execute the diagnostic program when detecting a state where there is no access for a predetermined time, A bus monitor having a writing output device, the input / output control device, or a microprocessor for diagnosing the various interfaces is provided.

According to the twenty-fifth aspect of the present invention, in the hardware diagnosing device, the microprocessor confirms from the state of the system bus that there is no access again at the beginning of the diagnosis started by executing the diagnostic program. The configuration is provided so as to proceed with the diagnosis later.

A hardware diagnostic apparatus according to a twenty-sixth aspect of the present invention has a configuration for storing the diagnostic result of the input / output device, the input / output control device or various interfaces in the bus monitoring device.

A hardware diagnostic device according to a twenty-seventh aspect of the present invention detects a state in which there is no access for a predetermined time by monitoring the control line of the system bus, and generates and outputs an interrupt signal based on the detection result. Bus monitoring device, and a diagnostic program installed in each of the input / output device, the input / output control device, and the various interfaces for diagnosing other input / output devices, the input / output control device, or the various interfaces other than itself and the interrupt signal. Upon receipt, the diagnostic program is executed, and a microprocessor for diagnosing other input / output devices or input / output control devices or various interfaces other than itself is configured.

A hardware diagnostic device according to a twenty-eighth aspect of the present invention comprises a priority order determining means for predetermining the order of diagnosis among the input / output device, the input / output control device or various interfaces as a priority order. According to the priority order determined by the priority order determining means, a microprocessor mounted on each of the input / output device, the input / output control device, and the various types of interfaces causes a diagnostic program to be transmitted via the system bus. In addition, a configuration for diagnosing the input / output device other than itself, the input / output control device, or the various interface devices is provided.

In the hardware diagnosing apparatus according to the twenty-ninth aspect of the invention, there is no access for a predetermined time by monitoring the control lines of the system bus mounted on the input / output device or the input / output control device or various interfaces. A control line monitoring unit that detects a state, and an interrupt signal that is generated and output based on the detection result of the control line monitoring unit, and that is output from an I / O device other than itself, an I / O control device, or various interfaces. A microprocessor for diagnosing the other input / output device or input / output control device or various interfaces that output the interrupt signal by accepting the interrupt signal, and the other input / output device or input / output control device or various interfaces To be equipped with a diagnostic program for diagnosing Those were.

In the hardware diagnosing device according to the thirtieth aspect of the invention, time data defining a predetermined time when the control line monitoring unit monitors the control line of the system bus to detect a state of no access is provided. Set to a value shortened according to the priority of the diagnosis of the input / output device or the input / output control device or various interfaces to be diagnosed, the predetermined value when detecting the non-access state A CPU for detecting the passage of time
And an input / output device, an input / output control device, or various interfaces connected to each other via a system bus.

[0042]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS One embodiment of the present invention will be described below. Embodiment 1 FIG. FIG. 1 is a block configuration diagram showing a configuration of a hardware diagnostic device for implementing the hardware diagnostic method of the first embodiment. In the figure, 1 is CP
U (storage means), 2 is a system bus, 3 is an input / output device (hereinafter referred to as I / O), 4 is a bus monitoring device for monitoring the load state of the system bus 2,
It is connected to the. Reference numeral 4a denotes a control line monitoring unit that detects a state where the control line connected to the I / O 3 is not accessed for a predetermined time, and 4b indicates that the control line is not accessed for the predetermined time, or is in a state of diagnosis or a state in which diagnosis is completed. The status register 4c for storing status information is a timer in which time-up time information corresponding to the predetermined time for detecting the lapse of the predetermined time is set. The predetermined time has a different value depending on each system.

Next, the operation will be described. The bus monitoring device 4 monitors the control line connected to the I / O 3 by the control line monitoring unit 4a, and accesses on the system bus 2 for a predetermined time corresponding to the time-up time preset in the timer 4c. When it detects a state in which is not generated, the status information indicating the state is set in the status register 4b and an interrupt is issued to the CPU 1. When the CPU 1 receives an interrupt from the bus monitoring device 4, the CPU 1 issues a diagnostic program to the I / O 3 which is a diagnostic target.

In this diagnostic program, a flag indicating that self-diagnosis is in progress is set in a predetermined register at the start of execution so that an error during self-diagnosis is not reported as a system error. To release. Further, when an access is detected on the system bus 2 during the self-diagnosis, the self-diagnosis is interrupted, the set flag indicating that the self-diagnosis is in progress is released, and normal access is given priority. . Further, the bus monitoring device 4 blinks an LED indicating that the self-diagnosis is in progress when the diagnostic program is executed, and when a failure occurs, logs data about the failure that has occurred and L that indicates the occurrence of the failure.
Turn on the ED.

FIG. 2 is a flowchart showing the series of processing operations described above. According to this flowchart, the control line monitoring unit 4a of the bus monitoring device 4 is connected to the CPU 1 and the I / O 3
The control line of the system bus 2 which connects to and is monitored (step ST1), and it is monitored whether a state in which no access occurs on the system bus 2 continues for a predetermined time (step ST2). When the state of no access continues for the predetermined time, status information indicating that there is no access is set in the status register 4b (step S
T3), an interrupt signal is issued to the CPU 1 (step ST4). Upon receiving the interrupt signal issued by the bus monitoring device 4 (step ST5), the CPU 1 sets a flag indicating that diagnosis is in progress in a predetermined register (step ST6), and executes a diagnostic program for diagnosing I / O3. It is activated (step ST7). In this diagnostic program, first, it is determined whether or not another access has occurred on the system bus 2 (step ST8). This judgment is made based on the notification sent from the bus monitoring device 4 when an access occurs on the system bus 2, and when it is judged that an access has occurred, the diagnostic program started in step ST7 is forced. End (step S
T19). Then, the flag which is set in step ST6 and indicates that the diagnosis is in progress is released (step ST
16).

On the other hand, the control line monitoring unit 4a of the bus monitoring device 4 which monitors the control line detects the access that has occurred on the system bus 2 and outputs status information indicating that the access is in progress to the status register 4b. Is set (step ST17), and the log data is written in the memory (step ST18). Although not shown, this memory is a predetermined area of the memory in which the status register 4b and the like are set.

On the other hand, in step ST8, the CPU 1
When it is determined that the system bus 2 is not accessed, the diagnostic program started in step ST7 is executed to determine whether an error occurs in the I / O 3 according to a predetermined procedure (step ST9). . If no error occurs, the diagnostic program is judged to be normal and the diagnostic program is terminated (step ST15), and the step S
The processing from T16 to step ST18 is executed.

When the CPU 1 determines that an error has occurred in the I / O 3 in step ST9, the running diagnostic program is forcibly terminated and this time the CPU
1 issues an interrupt to the bus monitoring device 4 (step S
(T11), the CPU 1 cancels the flag set in step ST6, which indicates that the diagnosis is in progress (step S).
T12).

On the other hand, when the bus monitor 4 detects that the CPU 1 has issued an interrupt, it sets the status information indicating that the diagnosis is completed in the status register 4b (step ST13). The error log data is written to the memory (step ST14).

As described above, according to the first embodiment, by providing the bus monitoring device 4 for monitoring the access that has occurred on the system bus 2, the system can be monitored without burdening the CPU when monitoring the access. Bus 2
The above configuration detects that there is no access for a predetermined period of time and automatically executes the diagnostic program based on the detection result. There is an effect that the failure of / O3 can be diagnosed and detected, and the influence of the failure can be prevented from affecting the entire computer system.

Embodiment 2 FIG. 3 shows the second embodiment.
FIG. 3 is a block configuration diagram showing a configuration of a hardware diagnostic device that implements the hardware diagnostic method of FIG. 3, parts that are the same as or correspond to those in FIG. 1 are assigned the same reference numerals and explanations thereof are omitted. In the figure, 5 and 6 are input / output devices of the same type or different types connected to the system bus 2 (hereinafter,
Reference numeral 5a denotes a microprocessor (priority order determining means) mounted on the I / O 5. 5b is I
This is a ROM mounted on the I / O 5 and stores diagnostic programs for I / O other than itself. 6a
Is a microprocessor (priority order determination means) mounted on the I / O 6. A ROM 6b is mounted on the I / O 6, and stores a diagnostic program for other I / Os other than itself.

Next, the operation will be described. In the hardware diagnosing method of this embodiment, the bus monitoring device 4 monitors the control line, and there is no access to the system bus 2 for a predetermined time corresponding to the time-up time preset in the timer 4c. Sometimes status information indicating that there is no access to the status register 4b is set, and an interrupt is issued to I / O5 and I / O6. Interrupts issued by the bus monitoring device 4 are prioritized between I / O5 and I / O6. In this embodiment, I / O5 has a higher priority than I / O6. Is high. Therefore, when the I / O 5 receives the interrupt issued by the bus monitoring device 4, the I / O 5
Reference numeral 5 executes a diagnostic program stored in the ROM 5b to diagnose I / O6. When the diagnosis of I / O6 is completed, the I / O6 knows that the diagnosis by I / O5 is completed, and conversely, from the I / O6 to the ROM6 for I / O5.
Run using the diagnostic program stored in b.

As in the case of the first embodiment, the I / O which is diagnosing the flag indicating that the self-diagnosis is underway is also performed by this diagnostic program so that the error under self-diagnosis is not reported as a system error. It is set in a predetermined register of the microprocessor when the execution is started, and is released when the self-diagnosis is completed. Further, when an access occurs on the system bus during the self-diagnosis, the bus monitoring device 4 notifies the I / O which is performing the diagnosis, and the I / O which has received the notification executes the diagnostic program. The operation is suspended, the set flag is released, and normal access is given priority. In addition, the bus monitoring device 4 blinks an LED indicating that the diagnostic program is being executed, leaves error log data regarding the occurred fault, and turns on the LED indicating the fault occurrence.

FIG. 4 is a flowchart showing the series of processing operations described above. According to this flowchart, the control line monitoring unit 4a of the bus monitoring device 4 monitors the control line of the system bus 2 connected to the I / Os 5 and 6 (step ST2).
1) It is determined whether or not there is no access for the predetermined time (step ST22), and when it is determined that there is no access for the predetermined time, a status indicating that there is no access. Information is set in the status register 4b (step ST23), and I
Issue an interrupt to / O5,6 (step ST2
4).

The I / O 5 which has detected the interrupt issued from the control line monitoring unit 4a of the bus monitoring device 4 (step ST25) sets a flag indicating that diagnosis is in progress in a predetermined register of the microprocessor 5a. (Step ST26), I / O6 stored in advance in the ROM 5b
The diagnostic program for diagnosing is executed and the diagnosis of I / O6 is started (step ST27). In this diagnostic program, first, the system bus 2 connected to the I / Os 5 and 6 is connected.
It is determined whether or not the access has occurred above (step ST28), and if there is access, the execution of the diagnostic program started in step ST27 is forcibly ended (step ST39). Then, the flag indicating that the diagnosis is in progress, which is set in step ST26, is canceled (step ST40).

On the other hand, the bus monitoring device 4 sets status information indicating that an access has occurred on the system bus 2 in the status register 4b (step ST
41), the log data in the process of forcibly ending the diagnosis is written in the memory (step ST42). Although not shown, this memory is a predetermined area of the memory in which the status register 4b is set.

On the other hand, when it is determined in step ST28 that the control line is not accessed,
The diagnosis of I / O6 started in step ST27 is continued, and it is determined according to the diagnostic program whether or not an error occurs in a predetermined procedure (step ST2).
9). When it is determined that no error occurs, I / O6
Determines that it operates normally and ends the diagnosis (step ST35). Then, the flag indicating that the diagnosis is in progress, which is set in step ST26, is released (step ST36), an interrupt is issued to the I / O6 (step ST37), and the log data for the terminated diagnosis is written in the Write to memory (step ST3
8).

Further, the I / O 6 which has received the interrupt issued from the I / O 5 in step ST37 (step ST25), knows that the diagnosis for itself has been completed, and performs the processing of step ST26 and subsequent steps. O
5 or other I / O diagnosis.

When the I / O 5 determines in step ST29 that an error has occurred, the I / O 5 forcibly terminates the running diagnostic program (step S29).
T30), the I / O 5 issues an interrupt to the bus monitoring device 4 (step ST31), and cancels the flag set in step ST26 indicating that the diagnosis is in progress (step ST32).

The bus monitoring device 4, which has received the interrupt issued from the I / O 5 in step ST31, sets status information indicating that the diagnosis is forcibly completed in the status register 4b (step ST3).
3) Further, the error log data is written to the memory (step ST34).

As described above, according to the second embodiment, the bus monitoring device for detecting the state where there is no access for a predetermined time by monitoring the control line on the system bus 2 and each I / O.
A microprocessor and a ROM storing a diagnostic program for diagnosing other I / Os are mounted on the O, and each I / O automatically executes the diagnosis for the other I / Os when there is no access. Since it is configured like this, the diagnosis between I / O is C
There is an effect that it can be performed without burdening the PU1.

Embodiment 3 FIG. 5 shows the third embodiment.
FIG. 3 is a block configuration diagram showing a configuration of a hardware diagnostic device that implements the hardware diagnostic method of FIG. 5, parts that are the same as or correspond to those in FIG. 1 are assigned the same reference numerals and explanations thereof are omitted. In this embodiment, each I / O is configured to have the function of the bus monitoring device shown in the first and second embodiments. In the figure, 7
And 9 are input / output devices (hereinafter referred to as I / Os), and only two I / Os 7 and 9 are shown for the sake of explanation, but there may be a plurality of other devices. Reference numeral 7a denotes a microprocessor (priority determining means) mounted on the I / O 7, 7b denotes another I / O, in this case, a ROM storing a diagnostic program for diagnosing the I / O 9, and 8 denotes the I / O 7. A control line monitoring unit (access monitoring means) 8a, a status register 8b, which is a provided bus monitoring device and detects a state in which no access is made for a predetermined time by monitoring the control line of the system bus 2 and the predetermined time. And a timer 8c in which time-up time information corresponding to the predetermined time for detecting the elapse of is set.

Reference numeral 9a denotes a microprocessor (priority determining means) mounted on the I / O 9, 9b denotes another I / O, in this case, a ROM storing a diagnostic program for diagnosing the I / O 7, and 10 denotes an I / O. / O9 is a bus monitoring device, which monitors a control line of the system bus 2 and detects a state in which there is no access for a predetermined time, an access monitoring unit 10a, a status register 10b, and the predetermined The timer 10c is provided with time-up time information corresponding to the predetermined time for detecting the passage of time. The time-up time information corresponding to the predetermined time has different values depending on the respective systems, and also differs according to the priority of I / O to be diagnosed. That is, the time-up time information set in the high-priority I / O timer is set shorter than the time-up time information set in the low-priority I / O timer. The higher the I / O, the shorter. As a result, if the I / O 7 has a higher priority than the I / O 9 and the control line is not accessed for more than the time when the timer 8c of the I / O 7 times up, the I / O having the higher priority is
An interrupt is issued from 7 first, and the diagnosis for I / O 9 to I / O 7 that has accepted this interrupt is given priority.

Next, the operation will be described. The I / O 7 and I / O 9 control line monitoring units 8 a and 10 a monitor the control lines of the system bus 2. Then, a time-up that defines a predetermined time set according to the system in advance and a time during which a state in which there is no access to the control line set according to the priority of the I / O for each timer lasts If there is no access to the control line for time, the status register 8b of each I / O 7, 9
Status information indicating that there is no access is set in 10b, and the I / Os 7 and 9 issue interrupts respectively, but the timers 8c and 10c are timed up according to the priority of their own I / O. The time is set in advance, and in this case, the time-up time for prioritizing the diagnosis of I / O7 over the diagnosis of I / O9 is set in each timer 8c, 10c, so that I / O7 precedes I / O9. The interrupt is issued to the I / O7, and the I / O7 that has accepted the interrupt issued by the I / O7 is given priority to diagnose the I / O7.

FIG. 6 is a flow chart showing the series of operations described above, and the operation will be described below with reference to this flow chart. According to this flowchart, each I / O
7 and 9 bus monitoring devices 8 and 10 control line monitoring units 8a and 1
0a monitors the control lines of the system bus 2 connected to the I / Os 7 and 9 (step ST51), and detects its own I / O.
Based on the time-up time set in the O timer, it is judged whether or not there is no access for the predetermined time (step ST52). When the state without access continues for a certain time, the time-up time information for prioritizing diagnosis of I / O7 over I / O9 is set in each of the timers 8c and 10c. I when the time to time up is exceeded
An interrupt is issued from / O7 to I / O9 (step ST53).

The interrupt issued by this I / O7 is accepted by the I / O9 (step ST54),
The I / O 9 sets status information indicating that there is no access in the status register 10b (step ST5
5) Further, a flag indicating that the diagnosis is in progress is set in a predetermined register (step ST56), the diagnosis program previously stored in the ROM 9b is executed, and the I / O 7
The diagnosis is started (step ST57).

In this diagnostic program, I / O7,
It is determined whether or not another access has occurred on the system bus 2 connected to the CPU 9, that is, whether or not the state of no access continues at this point (step ST5).
8) If there is an access, the execution of the diagnostic program started in step ST57 is forcibly ended (step ST65). Then, the status information set in the status register 10b in the step ST55 is reset to the status information indicating the accessed state (step ST66), and the flag set in the step ST56 is released (step ST6).
7) The log data in the process of forcibly ending the diagnosis is written in the memory (step ST68), and the process ends.
Although not shown, the memory in this case is a predetermined area of the memory in which the status register 10b is set.

On the other hand, if it is determined in step ST58 that there is no access, the diagnosis of the I / O7 started in step ST57 is continued and an error occurs in a predetermined procedure according to the diagnostic program. Is determined (step ST59). When it is determined that no error occurs, the I / O 7 is determined to operate normally and the diagnosis ends (step ST6).
0). Then, the status information set in the status register 10b in the step ST55 is reset to the status information indicating that the diagnosis is normally completed (step ST61), and further the step ST56.
The flag indicating that the diagnosis is in progress is set (step ST62), and another I / O, that is, I / O7.
An interrupt is issued to (step ST6
3), The log data regarding the terminated diagnosis is written in the memory (step ST64). In this case, instead of issuing an interrupt to the I / O 7 in step ST63, an interrupt may be issued to another I / O other than the I / O 7.

When it is determined in step ST59 that an error has occurred, the running diagnostic program is forcibly terminated (step ST70), and step S70 is executed.
The flag set in T56 indicating that the diagnosis is in progress is canceled (step ST71), and another I / O, that is, I / O7.
An interrupt is issued to (step ST72).

In the I / O7, status information indicating that an error has occurred is set in the status register 8b based on the interrupt issued by the I / O9 (step ST7).
3) Further, the error log data is written to a predetermined area of a memory (not shown) in which the status register 8b of the bus monitoring device 8 is set (step ST7).
4), end the process.

The I / O 7 is I in step ST63.
When the interrupt issued by the I / O 9 is received (step ST54), the diagnosis result indicating that the own I / O is normal is obtained. Therefore, the status register 8b is not accessed or is normal. The status information indicating that the diagnosis result is obtained is set (step ST55), and the flag indicating that the diagnosis is being performed is set in a predetermined register (step ST56). This time, I / O7 to I The diagnosis for / O9 is started (step ST57), and the processes after step ST58 are executed.

As described above, according to the third embodiment, each I / O is provided with the bus monitoring function for detecting the absence of access for a predetermined time by monitoring the control line of the system bus 2, and further Since the priority order with respect to other I / Os is determined in advance for each I / O and each I / O is configured to diagnose the other I / O, the CPU 1 at the time of performing the diagnosis The burden is reduced, and the priority of each I / O is determined based on the time-up time information set in the timer of each I / O, so it is more flexible as a diagnostic system than when determined by hardware. There is an effect that the degree is improved.

Embodiment 4 FIG. 7 shows the fourth embodiment.
FIG. 3 is a block configuration diagram showing the configuration of the hardware diagnosis method of FIG. 7, parts that are the same as or correspond to those in FIG. 1 are assigned the same reference numerals and explanations thereof are omitted. In this embodiment, a ROM which stores a microprocessor and a diagnostic program in the bus monitoring device shown in the first embodiment
Are provided. In the figure, 12 is a bus monitoring device, 12a
Is a control line monitoring unit (access monitoring means) for monitoring the control line of the system bus 2 and detecting a state in which there is no access for a predetermined time, 12b is a status register for storing status information, and 12c is for the lapse of the predetermined time. A timer in which time-up time information corresponding to the predetermined time for detection is set, 12d is the microprocessor (priority order determining means), and 12e is the ROM storing a diagnostic program. The predetermined time has a different value depending on each system.

Next, the operation will be described. The control line monitoring unit 12a monitors the control line of the system bus 2 connected to the I / O 3, and when it detects that there is no access for the predetermined time, it activates a diagnostic program stored in the ROM 12e, I by the bus monitoring device 12
/ O3 is diagnosed. Also in the diagnostic program of this embodiment, a flag indicating that self-diagnosis is being performed is set in a predetermined register at the start of execution so that an error during self-diagnosis is not reported as a system error, and the self-diagnosis ends. Release it. If some other access occurs on the system bus 2 during the self-diagnosis, the execution of the diagnostic program is suspended, the flag is released, and normal access is given priority. When the diagnostic program is executed, the bus monitoring device 12 also blinks the LED indicating that the self-diagnosis is in progress, and when a failure occurs, an error log is left and the LED indicating the failure occurrence of the I / O3 is turned on. Let

FIG. 8 is a flow chart showing the series of operations described above, and the operation will be described below with reference to this flow chart. According to this flowchart, the control line monitoring unit 12a of the bus monitoring device 12 monitors the control line connected to the I / O 3 (step ST81), and monitors whether or not there is no access for a predetermined time. (Step ST82). When the state of no access continues for the predetermined time, status information indicating that there is no access for the predetermined time is set in the status register 12b (step ST83), and a flag indicating that the diagnosis is in progress is predetermined. To the register (step ST8
4) Start a diagnostic program for diagnosing I / O3 (step ST85). In this diagnostic program, first, it is determined whether or not another access has occurred on the system bus 2 connecting the CPU 1 and the I / O 3 (step S).
T86). When it is determined that an access has occurred, the diagnostic program started in step ST85 is forcibly terminated (step ST92). Then, the flag set in step ST84 indicating that the diagnosis is in progress is canceled (step ST93).

Then, the control line monitoring unit 12a of the bus monitoring device 12 monitoring the control line sends status information indicating that an access has occurred to the status register 12b.
Is set (step ST94), and log data in the process of forcibly ending the diagnosis is written to the memory (step ST95). This memory is not shown,
It indicates a predetermined area of the memory in which the status register 12b and the like are set.

On the other hand, when it is determined in step ST86 that an access has not occurred, the above-mentioned step S
The diagnostic program started at T85 is executed, and it is determined whether or not an error occurs in I / O3 according to a predetermined procedure (step ST87). If no error occurs,
It is judged to be normal and the diagnostic program is terminated (step S
T88), the flag set in the step ST84 is released (step ST89), and status information indicating that the diagnosis is normally completed is displayed in the status register 12
It is set to b (step ST90), log information about the diagnosis normally performed is written to the memory (step ST91), and the process is ended.

When an error occurs in the I / O 3 which is the object of diagnosis in step ST87, the running diagnostic program is forcibly terminated (step ST9
6) The flag set in step ST84 is released (step ST97), and status information indicating that the diagnosis is forcibly ended or an error has occurred in I / O3 is set in the status register 12b (step S
T98), and further write error log information about the diagnosis in which the error has occurred to the memory (step ST9).
9), end the processing.

In this series of operations, the I / O cannot be diagnosed simultaneously from the bus monitoring device 12, so that a priority order for diagnosis is preset for each I / O. Each I / O is diagnosed according to the set priority.

Alternatively, at the time of diagnosis by the bus monitoring device 12, each I / O is selected by the address output on the address bus by the microprocessor 12d, and the diagnosis is sequentially performed on each I / O according to this selection order. .

As described above, according to the fourth embodiment, the bus monitor 12 is equipped with the microprocessor 12d, and the I / O 3 is diagnosed from the bus monitor 12 by using the diagnostic program stored in the ROM 12e. By doing so, there is an effect that each I / O 3 can be diagnosed without imposing any burden on the CPU 1.

Embodiment 5 FIG. 9 shows the fifth embodiment.
3 is a system configuration diagram showing the configuration of a hardware diagnostic device that implements the hardware diagnostic method of FIG. In the hardware diagnostic device of this embodiment, each I / O performs self-diagnosis on a predetermined item according to a command sent from the CPU 19. In the figure, 19 is CP
U and 20 are system buses 21 to 24 are system buses 2
Input / output devices (hereinafter referred to as I / O) of different types connected to 0.

In this hardware diagnosing device, for example, as shown in FIG. 10, command numbers are set in advance corresponding to common test items of each I / O, and the user assigns command numbers to the CPU 19 according to the test items to be executed.
Enables diagnosis for all different I / O by simply inputting from. Therefore, it is detected that each I / O has not been accessed on the system bus 20 and the I / O has not been exchanged other than data for self-diagnosis for a predetermined time. A control line monitoring unit and a timer, a command number for performing self-diagnosis, a ROM storing a diagnostic program to be executed in accordance with the identified command number, and a diagnostic result for its own I / O are saved. And a status register for setting. Also, the command number for the CPU 19 to read the status from each I / O is determined in advance for each I / O.

Therefore, the CPU 19 outputs the command shown in FIG. 10 to the system bus 20 on condition that the control line monitoring units of all I / Os detect the state in which no access occurs on the system bus 20. Then each I / O is a CPU
The self-diagnosis according to the command sent from the CPU 19 is executed, and when this self-diagnosis is completed, each I / O issues an interrupt to notify the CPU 19. At this time, each I
/ O stores in the status register whether the diagnosis result is “OK” or “NG” as a fixed-length status, so the command number “3” is sent from the CPU 19 to each I / O.
By sending it to O, the CPU 19 can read the diagnosis result from each I / O.

In preparation for the case where each I / O simultaneously issues an interrupt to the CPU 19 in this series of operations, when there is an I / O that simultaneously issued an interrupt, the interrupt from which I / O is prioritized. The priority order for determining whether or not is set in advance. Further, when the CPU 19 reads the diagnostic result stored in the status register of each I / O, it is read in accordance with the preset priority order.

As described above, according to the hardware diagnosing method of the fifth embodiment, by using the command number predetermined for each test item, all I / O of different types can be performed by one command. On the other hand, the self-diagnosis of the item corresponding to the command is performed in each I / O, and the diagnosis result can be easily read from each I / O.

Embodiment 6 FIG. FIG. 11 is a block diagram for explaining the hardware diagnostic method of the sixth embodiment. In the figure, 31 is a CPU, 32 is a system bus, 33 is an interface on the computer system side for connecting the disk device 35 to the CPU 31, and 34 is an interface on the disk device 35 side. Conventionally, in a system configured in this way, the CPU 31 cannot diagnose each of the interface 34 and the disk device 35, and the interface 33, the interface 34, and the disk device 35 are diagnosed as one piece of hardware. There is. Therefore, when a failure occurs in any of the interface 33, the interface 34, and the disk device 35, it is difficult to identify where the failure has occurred.

In this embodiment, first, the interface 3 based on the loopback test of only the interface 33 is performed.
3 is diagnosed (the route of (1)), and when a diagnostic result indicating that the interface 33 is normal is obtained, the interface 33 and the interface 33 are integrated by a return test in which the interface 33 and the interface 34 are integrated. Diagnosis of 34 (route (2)) is performed. As a result, when the diagnostic result indicating that the interfaces 33 and 34 are normal is obtained, the interface 33, the interface 34, and the disk are discriminated by the return test in which the interface 33, the interface 34, and the disk device 35 are integrated. A diagnosis (path (3)) for the device 35 is performed.

By performing such diagnosis, for example, when an abnormality occurs in the interface 34,
The diagnostic result of the loopback test by the route (1) is “normal”, the diagnostic result of the loopback test by the route (2) is “abnormal”, and the interface 34 is abnormal when the loopback test by the route (2) is performed. Will be revealed.

In order to automatically perform the diagnosis by such a loopback test during normal access, the interfaces and devices connected to the CPU 31 by these interfaces, in this case the disk device 35, are used. However, these interfaces and devices need to be equipped with a microprocessor or functions or diagnostic programs to replace it. Also, the CPU 31
Of the bus monitoring device described in the first to fifth embodiments, which detects that data is not exchanged by a normal access between the disk device 35 and the disk device 35. 35 need to be provided.

With this configuration, it is possible to make a diagnosis for each of the interface 33, the interface 34, and the disk device 35 connected to the system bus 32. There is an effect that it is possible to easily identify whether or not the failure has occurred.

In the fifth embodiment described above, the CPU
Although the device connected to 31 is a disk device, it may be another device, for example, a storage device such as a printer or a CD-ROM, or any device connected to the CPU 31 through an interface.

Embodiment 7 FIG. FIG. 12 is a block diagram for explaining the hardware diagnostic method of the seventh embodiment. In the figure, 41 is a CPU, 42 is a system bus, and 43 is an interface on the computer system side for connecting a disk device 45 to the CPU 41. Control signals and status signals necessary for controlling the operation of the disk device 45 are sent to the CPU 41 and the disk. It includes an I / O port for inputting / outputting from / to the device 45. Reference numeral 44 denotes an interface on the side of the disk device 45, which, like the interface 43, sends control signals and status signals necessary for controlling the operation of the disk device 45 to the CPU 41.
And an I / O port for inputting / outputting data from / to the disk device 45. Reference numeral 45 is a disk device including a transmission / reception interface for transmitting / receiving data read from or written to the disk to / from the CPU 41. The interfaces 43 and 44 and the disk device 45 have a function of transmitting / receiving data necessary for diagnosis to / from the CPU 41 and performing self-diagnosis, and stored therein a microprocessor and a self-diagnosis program for this purpose. The CPU 41 includes a ROM, a control line monitoring unit for detecting a state in which data other than the self-diagnosis data is not transferred by normal access for a predetermined time, a status register, a timer, and the like.
There are three layers when viewed from the side.

Reference numeral 47 is an input / output device (hereinafter referred to as I / O) connected to the system bus 42, 48 is a memory interface, and 49 is a memory. These I / O 47, memory interface 48, and memory 49 also perform self-diagnosis. It has a function to perform, a microprocessor for this purpose, a ROM storing a self-diagnosis program, and control for detecting a state in which data other than self-diagnosis data is not transferred for a predetermined time by normal access. It is equipped with a line monitoring unit, a timer, and a status register that stores the self-diagnosis results.

Further, as described in the fifth embodiment, the interfaces 43 and 44, the disk device 45,
I / O 47, memory interface 48, memory 49
The items of self-diagnosis performed by executing the above-mentioned diagnostic program are determined in advance in accordance with the commands issued by the CPU, and the self-diagnosis results are stored in the respective status registers.

Next, the operation will be described. The interfaces 43, 44 connected to the system bus 42,
The CPU 41 is provided on condition that the control line monitoring unit of each of the devices such as the disk device 45, the I / O 47, the memory interface 48, and the memory 49 detects a state in which no access occurs on the system bus 42 for a predetermined time. Outputs a predetermined command, the interface 43, I / O
47 and the memory interface 48 execute self-diagnosis according to the command sent from the CPU 41, and when this self-diagnosis ends, each device issues an interrupt and notifies the CPU 41. At this time, each device stores in the status register whether the diagnosis result is “OK” or “NG” as a fixed-length status.
By sending a diagnostic result reading command from 1 to each device, the CPU 41 can read the diagnostic result from each device. The diagnostic routes performed in this case are shown as route (1), route (2), and route (5).

In this series of operations, each device simultaneously performs C
In case of issuing an interrupt to the PU 41, a priority order for deciding which device gives priority to an interrupt when there is a device simultaneously issuing the interrupt is preset, and the status of each device is set. Even when the CPU 41 reads the diagnosis result stored in the register, the diagnosis result is read in accordance with the preset priority order.

If all the self-diagnosis results thus obtained are "normal", the CPU 41 further diagnoses the interface 44 and the memory 49 this time. In this diagnosis, the interface 43 and the interface 44 are integrally captured (diagram of (3)), and the memory interface 48 and the memory 49 are integrally captured.
(Diagram (6)).

As a result, when the diagnosis result that the interface 44 and the memory 49 are normal is obtained, the disk device 45 is further diagnosed. This diagnosis is a diagnosis (path (4)) in which the interface 43, the interface 44, and the disk device 45 are integrated.

As described above, by sequentially diagnosing the respective hierarchies of the hierarchized system from the upper level, there is an effect that it is possible to surely identify the hierarchy of the device in which the failure occurred.

Eighth Embodiment FIG. 13 is a block diagram showing a configuration and a diagnostic processing operation of a hardware diagnostic device that implements the hardware diagnostic method of the eighth embodiment. In FIG. 13, parts that are the same as or correspond to those in FIG. 12 are given the same reference numerals and description thereof is omitted. In the figure,
Reference numeral 51 is a bus monitoring device, 51a is a control line monitoring unit that monitors the control line of the system bus 42 and detects a state in which there is no access for a predetermined time, 51b is a status register for storing status information, and 51c is a status register for the predetermined time. A timer in which time-up time information corresponding to the predetermined time for detecting progress is set, 51d is a microprocessor (priority order determination means) mounted in the bus monitoring device 51,
Reference numeral 51e is a ROM that stores a diagnostic program. The predetermined time has a different value depending on each system.

Next, the operation will be described. In the hardware diagnostic device according to the eighth embodiment, the system bus 42
The bus monitoring device 51 connected to the control line monitoring unit 51a and the status register 51 are connected to the bus monitoring device 51 as in the fourth embodiment.
b, timer 51c, microprocessor 51d, ROM
51e is provided, and the interfaces 43 and 44, the disk device 45, the I / O 47, and the memory interface 4 are provided without burdening the CPU 41 from the bus monitoring device 51.
8. Each diagnosis of the memory 49 is performed. However, since the interfaces 43 and 44, the disk device 45, the I / O 47, the memory interface 48, and the memory 49 are not equipped with the microprocessor and the diagnostic program as in the seventh embodiment, the I / O 47 and the hierarchization are performed. Interface 43, interface 44, disk device 45, and similarly hierarchical memory interface 48
And the memory 49 need to be diagnosed differently,
Therefore, the microprocessor 51d of the bus monitoring device 51
Is an interface 43, an interface 44, a disk device 45, an I / O 47, and a memory interface 4.
8 and the memory 49 are respectively identified, and the diagnostic program for each identified device is executed to make a diagnosis.

Therefore, as described in the fourth embodiment, the interface 43, the interface 44,
The disk device 45, the I / O 47, the memory interface 48, and the memory 49 have respective diagnostic priorities set in advance, and the respective devices are diagnosed according to the set priorities.

Alternatively, at the time of diagnosis by the bus monitoring device 51, the uppermost I / O 47, the interface 43 and the memory interface 48 connected to the system bus 42 are selected by the address output on the address bus by the microprocessor 51d. Thus, according to this selection order, the I / O 47, the interface 4
3, the diagnosis is sequentially performed on the hierarchy with the memory interface 48 as the highest level.

The control line monitoring section 51a monitors the control line of the system bus 42, and when it detects that there is no access for the predetermined time, it activates the diagnostic program stored in the ROM 51e, and the bus monitoring unit 51a. Interface 43, 44, disk device 45, I /
Diagnosis of each device such as the O47, the memory interface 48, and the memory 49 is performed. In this case, the bus monitoring device 51
In order to prevent an error during self-diagnosis from being reported as a system error, a flag indicating that self-diagnosis is in progress is set in a predetermined register at the start of execution, and is cleared when the self-diagnosis ends. If any access other than the access by executing the diagnostic program occurs on the system bus 42 during the self-diagnosis, the execution of the diagnostic program is interrupted, the flag is canceled, and the normal access is prioritized. When the diagnostic program is executed, the bus monitoring device 51 also blinks an LED indicating that self-diagnosis is being performed, and when a failure occurs, an error log is left and the interfaces 43 and 44, the disk device 45, and the I / O device are connected. O4
The LED indicating that a failure has occurred in devices such as 7, the memory interface 48, and the memory 49 is turned on.

Further, since the devices such as the interface 43, the interface 44, the disk device 45, the memory interface 48, the memory 49, etc. are hierarchized, as described in the sixth and seventh embodiments, the higher hierarchy. It is necessary to sequentially diagnose from the above devices, but the command to instruct self-diagnosis to the above devices is the bus monitoring device 5.
The CPU 41 performs the processing operation of the entire system when it is output from the first microprocessor 51d and the diagnosis result is determined to be "abnormal".

As described above, in a system in which a plurality of devices such as different types of interfaces are hierarchized, when there is no normal access to the system bus, diagnosis for each hierarchy is performed from the bus monitoring device 51. By automatically performing and diagnosing each device constituting each hierarchy in order from the upper level, it is possible to identify and specify which of a plurality of hierarchized devices has a failure.

[0108]

As described above, according to the first aspect of the present invention, the CPU diagnoses by the interrupt signal generated by monitoring the control line of the system bus and detecting the state of no access for a predetermined time. Since the program is executed and the I / O device, the I / O controller, or various interfaces are diagnosed when there is no access, the diagnosis is not executed prior to the access. Original normal operations in the CPU and the input / output device, the input / output control device, the various interfaces, etc. are performed as usual, and the detection of the non-access state does not burden the CPU, and the time during the non-access time is reduced. There is an effect that effective diagnosis can be performed efficiently.

According to the second aspect of the present invention, at the beginning of the diagnosis started by executing the diagnostic program, the CPU notifies the system by the notification indicating that no access is detected from the state of the system bus. Since it is configured to confirm that various data is not sent and received on the bus and to proceed with the diagnosis according to the diagnostic program started after the confirmation, the CPU does not detect the non-access state. The access is obtained by the notification, does not impose a burden on the CPU during execution of the diagnostic program, and does not interfere with the normal normal operation of the CPU and the input / output device or the input / output control device or various interfaces. There is an effect that an efficient diagnosis can be performed by effectively utilizing the dead time.

According to the third aspect of the invention, the CP is notified by the notification indicating that there is an access detected from the state of the system bus at the beginning of the diagnosis start by the execution of the diagnosis program.
When it is confirmed that U has sent and received various data on the system bus, the CPU is configured to forcibly terminate the execution of the started diagnostic program. Therefore, the CPU detects the presence of the access. The CP that is obtained by the notification instead of being executed and is being executed by the diagnostic program
It is not a burden on U, and the execution of the diagnostic program is forcibly terminated when the access is detected. Therefore, the CPU and the input / output device, the input / output control device, or the normal normal operation in various interfaces are executed. There is an effect that an efficient diagnosis can be performed by effectively utilizing the time when there is no access without hindering the operation.

According to the invention described in claim 4, since the log data at the time of diagnosing the input / output device or the input / output control device or various interfaces is configured to be saved on the side where the interrupt signal is generated, There is an effect that a log for the diagnosis can be surely left when an abnormality is detected or when the diagnosis is forcibly terminated halfway.

According to the fifth aspect of the present invention, the access monitoring means provided in addition to the input / output device or the input / output control device or various interfaces connected to the CPU via the CPU and the system bus controls the system bus. When a state in which no access is made for a predetermined time is detected by monitoring the line, the microprocessor provided in other than the input / output device, the input / output control device, or the various interfaces, as in the access monitoring means, executes a diagnostic program. Since the input / output device, the input / output control device, or the various interfaces are configured to be executed when the access is not performed, the input / output device, the input / output control device, or the various interfaces are diagnosed in the above-described manner. CPU
This eliminates the need for the intervention, and has the effect of eliminating the burden on the CPU when executing the diagnosis.

According to the sixth aspect of the present invention, the microprocessor provided other than the CPU, the input / output device or the input / output control device connected to the CPU via the system bus, or various interfaces starts the diagnosis by the diagnostic program. Make sure there is no access at the beginning of
Since the diagnosis for the input / output device, the input / output control device, or the various interfaces is advanced after the confirmation, even if the access occurs immediately after the start of the diagnosis by the diagnostic program, the diagnosis immediately starts. The execution of the diagnostic program does not impose a burden on the CPU, and the access is prioritized over the execution of the diagnostic program. There is an effect that it is possible to perform efficient diagnosis by effectively utilizing the time when there is no access without hindering the original normal operation of the output control device or the various interfaces.

According to the invention described in claim 7, the diagnostic result of the input / output device or the input / output control device or various interfaces is sent to the CP via the CPU and the system bus.
Since the memory is configured to be stored in the memory on the microprocessor side provided other than the input / output device or the input / output control device connected to U or various interfaces, the CPU executes the diagnosis and the storage of the diagnosis result. There is no need to use the CP when executing the diagnosis.
This has the effect of eliminating the burden on U.

According to the present invention, the system bus control line is monitored to detect a state in which no access is made for a predetermined time, and the system is generated by an interrupt signal generated and output based on the detection result. The input / output device, the input / output control device, or the various interfaces connected to the bus are controlled by a microprocessor and a diagnostic program installed in each of the input / output devices other than itself or the input / output control via the system bus. Since the device or the various interface devices is configured to perform the diagnosis, the I / O device, the input / output control device, or the various interface devices can access the access line by monitoring the control line of the system bus. Generated and output based on the detection result of the timeless condition On the basis of the embedded signal, another I / O device, an I / O control device, or various interface devices will be diagnosed by the diagnostic program owned by themselves, and the burden on the CPU can be eliminated when the diagnostic is executed. There is an effect that can be done.

According to the invention described in claim 9, the order of diagnosis is determined in advance as a priority order, and the input / output device, the input / output control device, or the various interfaces are mounted on themselves according to the priority order. Since the microprocessor and the diagnostic program are configured to diagnose the other input / output device other than itself, the input / output control device, or the various interface devices via the system bus, the input / output device or the input / output device The output control device or the various interface devices will diagnose the other input / output device, the input / output control device, or the various interface devices according to the priority according to the diagnostic program owned by itself, and execute the diagnostic program. At the time of CP There is an effect that it is possible to eliminate the burden.

According to the tenth aspect of the present invention, the input / output device, the input / output control device, or the various interfaces each monitor the control line of the system bus to detect a state in which there is no access for a predetermined time. Other input / output device or input / output control device or various interfaces that have received the interrupt signal are output to the input / output device, the input / output control device, or the various interfaces that output the interrupt signal. On the other hand, the diagnosis is carried out by executing the diagnostic program possessed by itself, and after the diagnosis is completed, the side performing the diagnosis outputs an interrupt signal so that the side receiving the interrupt signal diagnoses itself. Since it is configured as described above, the input / output device, the input / output control device, or the various types of Each other it is possible to make a diagnosis by the diagnosis program by itself held between interfaces, there is an effect that can be eliminated burden of CPU during the diagnostic execution.

According to the eleventh aspect of the present invention, the time data defining the predetermined time when detecting the state of no access by monitoring the control line of the system bus is used as the priority of the diagnosis. Predetermine to a short value according to the height, and when receiving a diagnosis, the I / O device or I / O controller or various interfaces with high priority should be changed to the I / O device, I / O controller or various interfaces with low priority. Since the system is configured to receive the diagnosis first, the priority of the input / output device or the input / output control device or various interfaces to be diagnosed can be freely changed according to the setting method of the time data, and the system flexibility There is an effect that can improve the sex.

According to the twelfth aspect of the present invention, the input / output device, the input / output control device, or the various interfaces each monitor the control line of the system bus to detect a state in which there is no access for a predetermined time, and the detection result is detected. Other input / output device or input / output control device or various interfaces that have received the interrupt signal, the input / output device or input / output control device or various interfaces other than the input / output device that output the interrupt signal. The output device, the input / output control device, or the various interfaces are diagnosed by executing their own diagnostic programs, and the input / output device, the input / output control device, or the various interfaces that are diagnosed in the diagnosis. The diagnostic result for "normal Only when the diagnosis is completed, the input / output device, the input / output control device, or the various interfaces that have performed the diagnosis output an interrupt signal, and the diagnosed input / output device or the input / output control is executed. I / O device other than the device or the various interfaces, or the I / O control device, or the I / O device or the I / O control device that is not a party to the side receiving the diagnosis, by the I / O control device or the various interfaces accepting the output interrupt signal Since the diagnostics are configured to be executed by executing the diagnostic program owned by itself to the various interfaces, the input / output device, the input / output control device, or the various interfaces are input by the diagnostic program owned by the user. Output device or above It is possible to diagnose between output controller or said various interfaces, there is an effect that can be eliminated burden of CPU during the diagnostic execution.

According to the thirteenth aspect of the present invention, the input / output device, the input / output control device, or various interfaces that have made the diagnosis are configured to store the log data when the diagnosis is made in a predetermined memory area. Therefore, there is an effect that the log data can be saved without burdening the CPU.

According to the fourteenth aspect of the present invention, the input / output device, the input / output control device, or the various interfaces which are the objects of diagnosis for which the diagnosis result is judged as "abnormal" judges the diagnosis result as "abnormal". Save the log data when it is judged as "abnormal" based on the interrupt signal output from the I / O device or I / O control device or various interfaces on the diagnosis side in a predetermined area of the memory of its own Therefore, the log data when it is judged as "abnormal" is the "abnormal"
It is saved in the input / output device, the input / output control device, or various interfaces that have been determined to be effective, which has the effect of facilitating the management of diagnostic results.

According to the fifteenth aspect of the present invention, the contents of the diagnosis performed for the input / output device, the input / output control device, or the various interfaces, and the diagnosis contents for the input / output device, the input / output control device, or the various interfaces A command common to the input / output device, the input / output control device, or the various interfaces for performing diagnosis according to the above contents is determined in advance, and the input / output device or the input / output device is input from the CPU connected via the system bus. According to the command given to the output control device or the various interfaces, the input / output device, the input / output control device, or the various interfaces, respectively, perform self-diagnosis according to the contents of the predetermined diagnosis by a diagnostic program. Configured Since, there is an effect of efficiently direct the self-diagnosis in accordance with the contents in a single command to all the input and output device or the input-output control unit or the various interfaces.

According to the sixteenth aspect of the present invention, the input / output device, the input / output control device, or the various interfaces are provided on the system bus by normal access from the state of the system bus at the beginning of diagnosis by the diagnostic program. Since it is configured to confirm that no data has been exchanged and to proceed with the diagnosis according to the diagnostic program started after the confirmation, the diagnostic program is started in the state where various data is not exchanged on the system bus. Therefore, there is an effect that it is possible to eliminate the influence of the execution of the diagnostic program on the transfer of various data on the system bus by the normal access that occurs immediately after the start of the diagnostic program.

According to the seventeenth aspect of the present invention, the input / output device, the input / output control device, or various interfaces for executing the diagnostic program are connected to the system bus from the state of the system bus at the beginning of the diagnosis by the diagnostic program. When it is confirmed that the various types of data have been exchanged, the execution of the started diagnostic program is forcibly terminated. Therefore, immediately after the diagnosis is started, various types of data on the system bus are transferred by normal access. When there is a transfer, immediately stop the diagnosis that started,
The normal access can be prioritized, and the adverse effect of the execution of the diagnosis on the transfer of various data on the system bus due to the normal access can be eliminated.

According to the eighteenth aspect of the present invention, the status of transmission / reception of various data between the control lines of the system bus and the hierarchically connected input / output devices or input / output control devices or various interfaces is monitored. By doing so, the CPU detects a state where there is no access for a predetermined time, and when there is no access, the CPU executes a diagnostic program, and the hierarchically connected input / output devices or input / output control devices or various types are connected. Since the interface is configured to perform diagnosis in the order from the upper hierarchy to the lower hierarchy, it is possible to diagnose each input / output device or input / output control device or various interfaces connected hierarchically. effective.

According to the nineteenth aspect of the present invention, the input / output device or the input / output control device or various interfaces directly connected to the system bus, or the uppermost input / output device or the input / output control when hierarchized. The contents of diagnosis performed on the device or various interfaces, and the input / output device or the input / output controller for performing diagnosis on the input / output device, the input / output control device, or the various interfaces, respectively. Alternatively, a command common to the various interfaces is determined in advance, and the command given from the CPU causes the input / output device, the input / output control device, or the various interfaces to comply with the predetermined diagnostic content. Self-diagnosis program Since the self-diagnosis is performed sequentially from the uppermost layer to the lower layer when the hierarchical structure is adopted, the input / output device or the previous device connected to the system bus directly or in the hierarchical structure is used. There is an effect that the entry output control device or each of the various interfaces can be diagnosed with the content according to the command given from the CPU.

According to the twentieth aspect of the invention, the input / output device or the input / output control device or various interfaces directly connected to the system bus, or the highest level input / output device or the input / output in the case of hierarchization The contents of the diagnosis to be performed on the control device or various interfaces, and the common command for performing the diagnosis according to the contents on the input / output device, the input / output control device, or the various interfaces are determined in advance, A state in which there is no access for a predetermined time by monitoring a control line of the system bus on the side of the CPU connected to the system bus, the input / output device, the input / output control device, and a microprocessor provided other than the various interfaces. When the microprocessor detects a The input / output device, the input / output control device, or the various interfaces according to the command given by the microprocessor to the input / output device, the input / output control device, or the various interfaces when there is no access. The self-diagnosis is performed according to the content of the predetermined diagnosis, and in the case where the layers are hierarchized, the self-diagnosis is performed by the diagnostic program from the uppermost layer to the lower layers in sequence. So the CP
Content of diagnosis of each of the input / output device, the input / output control device, or the various interfaces connected directly or hierarchically to the system bus without going through U according to a command given from the microprocessor. There is an effect that can be done in.

According to the twenty-first aspect of the present invention, there is provided a bus monitoring device for detecting whether or not there is no access for a predetermined time by monitoring the control line of the system bus, and a bus monitoring device for not detecting the access for a predetermined time. A diagnostic program provided on the CPU side for allowing the CPU that has received the interrupt signal generated by the bus monitoring device to perform a diagnosis on the input / output device, the input / output control device, or various interfaces, and the diagnostic program. Since the bus monitoring device is configured to store the data obtained at the time of execution as log data in the bus monitoring device, the CPU does not detect the absence of the access but the bus monitoring device. Does not impose a burden on the CPU, and the CPU can know from the notification that there is no access. There is an effect that can reduce the load.

According to the twenty-second aspect of the invention, the CPU sends the system at the beginning of the diagnosis by the diagnostic program by the notification sent from the bus monitoring device when it detects that there is no access from the state of the system bus. Since it is configured to confirm that various kinds of data are not transmitted / received by the normal access on the bus, and the processing according to the started diagnostic program is advanced after the confirmation, the CPU detects the absence of the access. Instead of the above, the bus monitoring device performs the notification, the CPU can know from the notification that there is no access, does not impose a burden on the CPU during execution of the diagnostic program, and the execution of the diagnostic program causes the access The diagnostic program is executed by the CPU and the input / output device or Controller or without hindering the original normal operation at various interfaces, there is an effect that it is possible to perform an efficient diagnosis and effective use of the no access time.

According to the twenty-third aspect of the invention, when the bus monitor detects that there is an access from the state of the system bus, the CPU notifies the diagnosis start by the diagnostic program by the notification sent from the bus monitor. First, it is confirmed that various kinds of data are exchanged on the system bus and the execution of the started diagnostic program is forcibly ended. Therefore, the CPU detects the presence of the access. Instead of the above, the bus monitoring device does not burden the CPU during execution of the diagnostic program, and the access is performed prior to the execution of the diagnostic program and the diagnostic program is forcibly terminated. To prevent the original normal operation of the CPU and the input / output device, the input / output control device, or various interfaces No Rukoto, there is an effect that it is possible to perform an efficient diagnosis and effective use of the no access time.

According to the twenty-fourth aspect of the present invention, a diagnostic program for diagnosing the input / output device, the input / output control device, or various interfaces, and the control line of the system bus are monitored so that access is not made for a predetermined time. A control line monitoring unit for detecting a state, and the control line monitoring unit executes the diagnostic program when the control line monitoring unit detects a state in which no access is made for a predetermined time by monitoring the control line of the system bus. Alternatively, since the bus monitoring device having the input / output control device or the microprocessor for diagnosing the various interfaces is provided, the CPU for diagnosing the input / output device or the input / output control device or the various interfaces
The intervention of the
It has the effect of eliminating the burden on U.

According to the twenty-fifth aspect of the invention, in the diagnostic program executed when the control line monitoring unit monitors the control line of the system bus to detect a state in which there is no access for a predetermined time, the microprocessor is Immediately after the start of the diagnosis by the diagnosis program, since it is configured to confirm that there is no access at the beginning of the started diagnosis based on the detection result of the control line monitoring unit and advance the diagnosis after the confirmation. Even if the access occurs, the started diagnosis can be stopped immediately, the execution of the diagnosis does not burden the CPU, and the access is performed prior to the execution of the diagnostic program. , Without hindering the normal normal operation of the CPU and the input / output device, the input / output control device, or various interfaces There is an effect that it is possible to perform an efficient diagnosis and effective use of the no access time.

According to the twenty-sixth aspect of the invention, since the microcomputer provided in the bus monitoring device is configured to save the diagnostic results of the input / output device, the input / output control device, or various interfaces in the bus monitoring device, There is no need for the CPU to execute the output device, the input / output control device, or the various interface diagnostics and the storage of the diagnostic result, and it is possible to eliminate the burden on the CPU when executing the diagnostics.

According to the twenty-seventh aspect of the present invention, a bus monitoring device for detecting a state in which there is no access for a predetermined time by monitoring the control line of the system bus, and generating and outputting an interrupt signal based on the detection result. And an input / output device, an input / output control device, or various interfaces, respectively, and by receiving a diagnostic program for diagnosing other input / output devices, input / output control devices, or various interfaces other than itself and the interrupt signal, The input / output device, the input / output control device, and the various types are configured to include the input / output device other than the self, the input / output control device, and the microprocessor for diagnosing various interfaces. The bus monitoring device is the interface device. Based on the interrupt signal generated and output when the access is detected for a predetermined period of time by monitoring the control line of the stem bus, the other input / output devices or input / output control devices or various interface devices Diagnosis is made by the diagnosis program owned by itself, and there is an effect that the load of the CPU connected to the system bus can be eliminated when the diagnosis is executed.

According to the twenty-eighth aspect of the present invention, there is provided a priority order determining means for predetermining the order of diagnosis among the input / output devices, the input / output control device, or various interfaces as priorities. According to the priority determined by the priority determining means, the input / output device, the input / output control device, or the microprocessors respectively mounted on the various interfaces are
The input / output device is configured so as to diagnose the other input / output device other than itself, the input / output control device, or the various interface devices via a system bus by a diagnostic program installed in itself. Alternatively, the input / output control device or the various interface devices are assigned to the other input / output devices, the input / output control device or the various interface devices by the priority determining means according to the diagnostic program stored in the priority determination means. According to the CP when executing the diagnosis
This has the effect of eliminating the burden on U.

According to the twenty-ninth aspect of the present invention, a control line monitoring unit that detects a state in which there is no access for a predetermined time by monitoring the control line of the system bus, and a detection result of the control line monitoring unit are used. The other input / output device or the input / output control that outputs the interrupt signal by generating and outputting the interrupt signal and receiving the interrupt signal output from the other input / output device or input / output control device or various interfaces other than itself A microprocessor for diagnosing a device or various interfaces, and a diagnostic program for diagnosing the other I / O device, the I / O control device, or the various interfaces, the I / O device, the I / O control device, or the various Since it was configured to be mounted on each interface, Or said output control device or the various interfaces it is possible to perform mutually diagnosed by the diagnostic program is itself held, there is an effect that can be eliminated CPU load of when the diagnosis execution.

According to the thirtieth aspect of the present invention, the time data defining the predetermined time when the control line monitoring unit monitors the control line of the system bus to detect the state of no access is the object of diagnosis. A timer that is set to a short value according to the high priority of the side receiving the diagnosis of the input / output device or input / output control device or various interfaces
Since the input / output device, the input / output control device, or the various interfaces connected to the CPU via the system bus are provided respectively, the input / output device, the input / output control device, or the input / output control device to be diagnosed according to the setting method of the time data or The priority of various interfaces can be freely changed, and there is an effect that a system with improved flexibility can be constructed.

[Brief description of the drawings]

FIG. 1 is a block configuration diagram showing a configuration of a hardware diagnostic device for implementing a hardware diagnostic method according to a first embodiment of the present invention.

FIG. 2 is a flowchart showing an operation of the hardware diagnostic device for implementing the hardware diagnostic method according to the first embodiment of the present invention.

FIG. 3 is a block configuration diagram showing a configuration of a hardware diagnostic device for implementing a hardware diagnostic method according to a second embodiment of the present invention.

FIG. 4 is a flowchart showing an operation of the hardware diagnostic device for implementing the hardware diagnostic method according to the second embodiment of the present invention.

FIG. 5 is a block configuration diagram showing a configuration of a hardware diagnostic device for realizing a hardware diagnostic method according to a third embodiment of the present invention.

FIG. 6 is a flowchart showing the operation of the hardware diagnostic device for implementing the hardware diagnostic method according to the third embodiment of the present invention.

FIG. 7 is a block configuration diagram showing a configuration of a hardware diagnostic device for realizing a hardware diagnostic method according to a fourth embodiment of the present invention.

FIG. 8 is a flowchart showing an operation of the hardware diagnostic device for realizing the hardware diagnostic method according to the fourth embodiment of the present invention.

FIG. 9 is a block configuration diagram showing a configuration of a hardware diagnostic device for implementing a hardware diagnostic method according to a fifth embodiment of the present invention.

FIG. 10 is an explanatory diagram showing the relationship between commands and diagnostic items in the hardware diagnostic device for implementing the hardware diagnostic method according to the fifth embodiment of the present invention.

FIG. 11 is a block diagram illustrating a hardware diagnostic method according to a sixth embodiment of the present invention.

FIG. 12 is a block diagram for explaining a hardware diagnosis method according to a seventh embodiment of the present invention.

FIG. 13 is a block diagram showing a configuration and a diagnostic processing operation in a hardware diagnostic device for implementing a hardware diagnostic method according to an eighth embodiment of the present invention.

FIG. 14 is a block diagram showing a computer system to which a conventional hardware diagnosis method is applied, in which a plurality of input / output devices are connected to a CPU via a system bus.

FIG. 15 is a block configuration diagram showing a configuration of a computer system obtained by further expanding the computer system shown in FIG. [Brief Description of Drawings] 1, 19, 31, 41 CPU, 2, 20, 32, 42
System bus, 3,5-7,9,21-24,47
Input / output device, 4, 8, 10, 12, 51 Bus monitoring device, 5a, 6a, 7a, 9a, 12d, 51d Microprocessor (priority determining means), 8a, 10a, 1
2a Control line monitoring unit (access monitoring means), 8c, 10
c Timer, 33, 34, 43, 44 interface, 49 memory.

Claims (30)

[Claims] 1. A hardware diagnostic method for detecting an abnormality occurring in an input / output device, an input / output control device, or various interfaces connected to a CPU via a system bus, monitoring a control line of the system bus. The CPU detects a state in which no access is made for a predetermined time, and uses the interrupt signal generated based on the detection result to detect the CPU.
Executing a diagnostic program to detect an abnormality occurring in the input / output device, the input / output control device, or the various interfaces. 2. The CPU confirms by the notification indicating that no access is detected from the state of the system bus, at the beginning of diagnosis by the diagnostic program, confirms that various data is not exchanged on the system bus. The hardware diagnosis method according to claim 1, further comprising the step of advancing processing according to the diagnosis program started after the confirmation. 3. When the CPU confirms that various kinds of data are transmitted and received on the system bus by a notification indicating that there is an access detected from the state of the system bus at the beginning of diagnosis by the diagnostic program, The hardware diagnostic method according to claim 2, wherein the execution of the started diagnostic program is forcibly terminated. 4. Log data for diagnosing an input / output device, an input / output control device, or various interfaces,
The hardware diagnostic method according to claim 1, wherein the hardware diagnostic method stores the interrupt signal on a generation side. 5. A hardware diagnostic method for detecting an abnormality occurring in an input / output device, an input / output control device, or various interfaces connected to a CPU via a system bus, comprising: the CPU, the input / output device, and the input / output device. The CPU, the input / output device, and the input / output device when the access control unit provided in addition to the output control device and the various interfaces detects a state in which there is no access for a predetermined time by monitoring the control line of the system bus. While the control device and the microprocessor provided other than the various interfaces execute a diagnostic program, the access monitoring means monitors the control line of the system bus to detect that an access has occurred, and the diagnosis being executed. The program is forcibly terminated and the I / O device or the I / O device A hardware diagnostic method characterized by detecting an abnormality occurring in a force control device or the various interfaces. 6. The microprocessor confirms that there is no access again from the state of the system bus at the beginning of the diagnosis started by executing the diagnostic program, and after the confirmation, diagnoses the input / output device or the input / output control device or various interfaces. The method for diagnosing hardware according to claim 5, wherein the method advances. 7. A diagnostic result of an input / output device, an input / output control device or various interfaces is stored in a memory on a microprocessor side.
Described hardware diagnostic method. 8. A hardware diagnostic method for detecting an abnormality occurring in an input / output device, an input / output control device, or various interfaces connected to a CPU via a system bus, monitoring a control line of the system bus. In the above, the I / O device, the I / O controller, or the various interfaces are detected by the interrupt signal generated on the basis of the detection result of no access for a predetermined time. A hardware diagnostic method, wherein a diagnostic program diagnoses the other input / output devices other than itself, the input / output control device, or the various interface devices via the system bus. 9. The order of performing diagnosis is determined in advance as a priority order, and the input / output device, the input / output control device, or various interfaces are controlled by the microprocessor and the diagnostic program installed in the system according to the priority order. 9. The hardware diagnosing method according to claim 8, wherein diagnosis is performed on the other input / output device, the input / output control device, or the various interface devices other than itself via a bus. 10. A hardware diagnostic method for detecting an abnormality occurring in an input / output device, an input / output control device, or various interfaces connected to a CPU via a system bus, the input / output device or the input / output control device. Alternatively, each of the various interfaces detects a state in which there is no access for a predetermined time by monitoring the control line of the system bus, outputs an interrupt signal based on the detection result, and inputs and outputs other interrupt signals. The device, the input / output control device, or the various interfaces makes a diagnosis by executing a diagnostic program owned by itself to the input / output device, the input / output control device, or the various interfaces that output the interrupt signal, and the diagnosis is performed. After the end, the input / output device that has performed the diagnosis Alternatively, the input / output control device or the various interfaces outputs an interrupt signal,
The input / output device or the input / output control device or the various interfaces diagnosed as “normal” that is the target of the diagnosis outputs the interrupt signal by accepting the output interrupt signal, or A hardware diagnostic method, wherein diagnosis is performed by executing a diagnostic program owned by the input / output control device or the various interfaces. 11. An input / output device, an input / output control device, or various interfaces on the side to be diagnosed, which provides time data defining a predetermined time for detecting a state of no access by monitoring a control line of a system bus. I / O device with high priority or I / O controller or various interfaces with low priority when receiving a diagnosis. 11. The hardware diagnosis method according to claim 10, wherein the diagnosis is performed prior to the input / output control device or various interfaces. 12. A hardware diagnostic method for detecting an abnormality occurring in an input / output device, an input / output control device, or various interfaces connected to a CPU via a system bus, comprising: the input / output device or the input / output control device. Alternatively, each of the various interfaces detects a state of no access for a predetermined time by monitoring the control line of the system bus, outputs an interrupt signal based on the detection result, and another input / output device that receives the interrupt signal Alternatively, the input / output control device or the various interfaces have their own diagnostics for the input / output device, the input / output control device, or the input / output device other than the various interfaces, the input / output control device, or the various interfaces that output the interrupt signal. Run the program Only when the diagnosis result for the input / output device, the input / output control device, or the various interfaces that is the target of diagnosis in the diagnosis is “normal”, after completion of the diagnosis,
The input / output device, the input / output control device, or the various interfaces that have performed the diagnosis output an interrupt signal, and the input / output device, the input / output control device, or the various interfaces other than the “normal” are diagnosed. When the input / output device, the input / output control device, or the various interfaces accept the output interrupt signal, the input / output device, the input / output control device, or the various interfaces that are not parties to the side that has received the diagnosis are themselves A hardware diagnosing method, characterized in that diagnosis is performed by executing a diagnostic program stored therein. 13. The input / output device, the input / output control device, or various interfaces which have made a diagnosis store log data at the time of making the diagnosis in a predetermined memory area. 13. The hardware diagnostic method according to any one of 12. 14. The input / output device, the input / output control device, or various interfaces which are the objects of diagnosis for which the diagnosis result is determined to be “abnormal” are input to the side which has performed the diagnosis for which the diagnosis result is determined to be “abnormal”. The log data when the "abnormal" is determined based on an interrupt signal output from the output device, the input / output control device, or various interfaces is stored in a predetermined area of a memory included in itself. The hardware diagnostic method according to any one of claims 10 to 13. 15. A hardware diagnostic method for detecting an abnormality occurring in an input / output device, an input / output control device, or various interfaces connected to a CPU via a system bus, comprising: the input / output device or the input / output control device. Alternatively, the contents of diagnosis performed on each of the various interfaces and the input / output device or the input / output control device for performing diagnosis according to the contents on the input / output device, the input / output control device, or the various interfaces, respectively. Alternatively, a command common to the various interfaces is previously determined,
According to the command given from the CPU to the input / output device, the input / output control device, or the various interfaces, the input / output device, the input / output control device, or the various interfaces respectively follow the predetermined diagnostic content. A hardware diagnostic method, characterized in that self-diagnosis is performed by a diagnostic program. 16. An input / output device, an input / output control device, or various interfaces for executing a diagnostic program must not exchange various data on the system bus from the state of the system bus at the beginning of diagnosis by the diagnostic program. 16. The hardware diagnostic method according to claim 8, further comprising the step of confirming, and advancing the processing according to the started diagnostic program. 17. An input / output device, an input / output control device, or various interfaces that execute a diagnostic program must transfer various data on the system bus from the state of the system bus at the beginning of diagnosis by the diagnostic program. 17. The hardware diagnostic method according to claim 16, further comprising: forcibly ending the execution of the started diagnostic program when is confirmed. 18. The CPU through the system bus to the C
In a hardware diagnostic method for detecting an abnormality occurring in an input / output device, an input / output control device, or various interfaces that are hierarchically connected to a PU, the control line of the system bus and the hierarchically connected The CPU detects that there is no access for a predetermined time by monitoring the state of transmission and reception of various data between the input / output device, the input / output control device, and the various interfaces, and the access is performed for a predetermined time. When there is no state, the CPU executes the diagnostic program, and the input / output device or the input / output control device or various interfaces connected in the hierarchical form are connected to the input / output device or the input / output control device in the higher hierarchy. Alternatively, there is the input / output device in a lower layer from the various interfaces. Diagnoses in the order of the input / output control device or the various interfaces, and detects an abnormality occurring in the hierarchical input / output device, the input / output control device, or the various interfaces. Hardware diagnostic method. 19. An input / output device, an input / output control device, or various interfaces including an input / output device, an input / output control device, or various interfaces, which are connected to the CPU via a system bus and are hierarchical to the CPU. In the hardware diagnostic method for detecting an abnormal state, the highest level for the input / output device, the input / output control device, or the various interfaces, and the hierarchical input / output device, the input / output control device, or the various interfaces. Of the input / output device, the input / output control device, or the various interfaces, and the input for performing the diagnosis according to the contents for the input / output device, the input / output control device, or the various interfaces, respectively. There is an output device Alternatively, a command common to the input / output control device or the various interfaces is determined in advance, and the input / output device or the input / output device is controlled by the command given from the CPU to the input / output device or the input / output control device or the various interfaces. The input / output control device or each of the various interfaces performs self-diagnosis according to the content of the predetermined diagnosis by a diagnostic program, and the input / output device or the input / output control device connected in a hierarchical manner or With respect to the various interfaces, diagnosis is performed in the order from the input / output device or the input / output control device or the various interfaces in the highest hierarchy to the input / output device, the input / output control device or the various interfaces in the lower hierarchy. There, the hardware diagnostic method characterized by detecting an abnormality has occurred in the input device or the output control device or the various interfaces. 20. An input / output device, an input / output control device, or various interfaces including an input / output device, an input / output control device, or various interfaces which are connected to the CPU via a system bus and which are layered with respect to the CPU. In the hardware diagnosis method for detecting an abnormality, the input / output device, the input / output control device, or the various interfaces, and the hierarchical input / output device, the input / output control device, or the various interfaces are The diagnostic contents to be performed on the highest-level I / O device, I / O control device, or various interfaces, and the I / O device, the I / O control device, or various interfaces, and the highest-level I / O device or input. Output control device A common command for performing diagnosis according to the above contents is determined in advance for each of the interfaces or various interfaces, and a micro-computer provided in addition to the CPU, the input / output device, the input / output control device, and the various interfaces. By monitoring the control line of the system bus on the processor side, the microprocessor executes a diagnostic program when it detects that there is no access for a predetermined time, and when there is no access, the I / O device or the I / O By the command given by the microprocessor to the control device or the various interfaces, self-diagnosis for the input / output device, the input / output control device, or the various interfaces is performed according to the content of the predetermined diagnosis, and the hierarchy is provided. Connected For the input / output device, the input / output control device, or the various interfaces, the input / output device or the input / output control device or the various interfaces of the highest layer is in the lower layer of the input / output device or the previous interface. A hardware characterized in that a self-diagnosis is performed by the diagnostic program in the order of the entry output control device or the various interfaces, and an abnormality occurring in the input / output device, the input / output control device or the various interfaces is detected. Wear diagnostic method. 21. A hardware diagnostic device for detecting an abnormality occurring in an input / output device, an input / output control device, or various interfaces connected to a CPU via a system bus, monitoring the control line of the system bus. And a bus monitoring device that detects whether there is no access for a predetermined time, and an interrupt signal generated and output by the bus monitoring device when the bus monitoring device detects a state where there is no access for a predetermined time. The received CP
A diagnostic program provided on the CPU side for U to diagnose the input / output device, the input / output control device, or the various interfaces, and data obtained by executing the diagnostic program as log data on the bus. A hardware diagnostic device comprising: a storage unit that stores the data in a monitoring device. 22. At the beginning of the diagnosis by the diagnostic program, the CPU causes the bus monitoring device to send a notification on the system bus when the bus monitoring device detects that there is no access from the state of the system bus. 22. The hardware diagnostic device according to claim 21, wherein it is confirmed that various data has not been transmitted and received, and the process according to the diagnostic program started after the confirmation is performed. 23. When the CPU detects that there is access from the state of the system bus at the beginning of the diagnosis by the diagnostic program, the CPU sends a notification on the system bus by the notification sent from the bus monitor. 23. The hardware diagnostic apparatus according to claim 22, wherein the execution of the started diagnostic program is forcibly terminated when it is confirmed that various kinds of data have been exchanged. 24. In a hardware diagnostic device for detecting an abnormality occurring in an input / output device, an input / output control device, or various interfaces connected to a CPU via a system bus, the input / output device is connected to the system bus, A diagnostic program for diagnosing the device, the input / output control device, or the various interfaces, a control line monitoring unit that detects a state in which there is no access for a predetermined time by monitoring the control line of the system bus, and the control line A microprocessor that executes a diagnostic program when a monitoring unit monitors a control line of the system bus and detects a state in which there is no access for a predetermined time, and diagnoses the input / output device or the input / output control device or various interfaces. A bus monitoring device having Hardware diagnostic device. 25. The microprocessor confirms that there is no access again from the state of the system bus at the beginning of the diagnosis started by executing the diagnosis program, and advances the diagnosis after the confirmation. 24. The hardware diagnostic device according to item 24. 26. The hardware diagnosis device according to claim 25, wherein the diagnosis result of the input / output device, the input / output control device, or various interfaces is stored in the bus monitoring device. 27. A hardware diagnostic device for detecting an abnormality occurring in an input / output device, an input / output control device, or various interfaces connected to a CPU via a system bus, monitoring the control line of the system bus. With a bus monitoring device that detects a state in which there is no access for a predetermined time, and generates and outputs an interrupt signal based on the detection result, and is mounted on each of the input / output device, the input / output control device, or the various interfaces,
An input / output device other than oneself or an input / output control device or a diagnostic program for diagnosing various interfaces and the diagnosis program is executed by receiving the interrupt signal, and another input / output device other than oneself or the input / output control A hardware diagnostic device, comprising: a device or a microprocessor for diagnosing various interfaces. 28. A priority order determining means for determining in advance a priority order for performing diagnosis among the input / output device, the input / output control device or various interfaces, and the priority order determining means determines the priority order. According to the priority order, the microprocessor installed in each of the input / output device, the input / output control device, or the various interfaces is connected to the other input / output other than itself via the system bus by the diagnostic program installed in itself. 28. The hardware diagnostic device according to claim 27, which diagnoses the device, the input / output control device, or the various interface devices. 29. A hardware diagnostic device for detecting an abnormality occurring in an input / output device or an input / output control device or various interfaces connected to a CPU via a system bus, wherein the input / output device or the input / output control device or A control line monitoring unit, which is mounted on each interface, detects a state in which there is no access for a predetermined time by monitoring the control line of the system bus, and generates an interrupt signal based on the detection result of the control line monitoring unit. And outputs the interrupt signal by receiving an interrupt signal output from another input / output device or input / output control device or various interfaces other than itself. A microprocessor that diagnoses various interfaces, And a diagnostic program for diagnosing the other input / output device, the input / output control device, or the various interfaces. 30. The input / output device or the input / output to be diagnosed is time data defining a predetermined time when the control line monitoring unit monitors the control line of the system bus to detect a state of no access. A timer, which is set to a short value according to the high priority of the side receiving the diagnosis of the control device or various interfaces, for detecting the passage of the predetermined time when detecting the state of no access,
30. The hardware diagnostic device according to claim 29, comprising an input / output device, an input / output control device, or various interfaces connected to U via a system bus.