JPH03130835A - Restarting device for information processor - Google Patents

Abstract

PURPOSE:To quickly restart an information processor by checking the presence or absence of an error after reading successively the data out of a main storage to accurately decide a faulty area and loading the data again only to the main storage address that includes a fault. CONSTITUTION:When a main storage has a fault, a diagnostic read means 101 reads successively the data to be stored in a main storage 103. An error detection means 104 detects the presence or absence of an error of the read data. Then a reloading part 107 of a restart control means 109 sets again the prescribed data to only the corresponding address of the storage 103 where an error is detected out of the read data via the means 104. In such a constitution, the area of a fault can be accurately decided in the storage 103 and then the data are loaded again only in the main storage address including the fault. Thus an information processor can be restarted at a high speed.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to an apparatus for restarting an information processing apparatus, and more particularly to an apparatus for restarting an information processing apparatus when a main memory failure occurs.

[Conventional technology]

Normally, when a failure is detected in an information processing device, this failure is reported to a diagnostic processing device, and the diagnostic processing device executes restart processing of the information processing device. That is, when the diagnostic processing device receives the above report,
The operation of the information processing equipment is stopped, the software visible information in the information processing equipment at the time of a failure is extracted through a path called a scan canvas, and after the information processing equipment is cleared and the failure is removed, the extracted software visible information is extracted. Reset the information in the information processing device. As a result, the information processing device is set to the same operating state as it was at the time of the failure from the software perspective. If the fault in the information processing device is fixed, the same fault will be detected again and the device will become inoperable, but if it is an intermittent fault, it will be able to operate normally after being restarted.

Additionally, if a failure occurs when accessing the main memory from an information processing device, there is a high possibility that the main memory data has been destroyed. As shown, the main memory was restored by reloading data from the external storage device to all recoverable areas of the main memory area.

[Problem to be solved by the invention]

As mentioned above, conventionally, when a main memory access failure occurs, main memory is recovered by reloading data to all recoverable areas of the main memory area. It took a long time to load, making it difficult to restart quickly.

The present invention has been made in view of these circumstances,
The purpose of this is to accurately locate the fault in main memory and reload data only to the main memory address where the fault exists, thereby enabling a fast restart of the information processing equipment. The goal is to provide equipment.

[Means to solve the problem]

In order to achieve the above object, the present invention provides an apparatus for restarting an information processing apparatus when a failure occurs in the main memory. A reference-F means for sequentially reading data;
Error detection means for detecting errors in the data sequentially read by the diagnostic read means, and re-setting of predetermined data only for the corresponding address of the main memory where the error detection means detected an error in the read data. and a restart control means including a reload section to perform the restart.

[Effect]

In the restart device for an information processing device according to the present invention, when a failure occurs in the main memory, the diagnostic read means sequentially reads data stored in the main memory, and the error detection means detects an error in the read data. Detect the presence or absence of. The reload section of the restart control means resets predetermined data only to the corresponding address of the main memory where an error in the read data is detected by the error detection means.

[Example] Next, an example of the present invention will be described in detail with reference to the drawings.

FIG. 1 is a block diagram showing an example of a computer system to which the present invention can be applied, and shows a system including a diagnostic processing device 101, an information processing device 102, a main memory 103, and an error detection circuit 104.

In the same figure, the error detection circuit 104 is connected to the main memory 103.
This is a circuit that detects whether or not there is an error in the data (READ DATA) read from the data processing device 102. When an error is detected, a report to that effect (PTYER) is sent to the information processing device 102 via the X error report m1lo. and output to the diagnostic processing device 101.

Software executed by the information processing device 102 is stored in the main memory 103, and the information processing device 102 stores an address (ME) in the address register 105 for accessing the main memory.
By setting M ADR3), the main memory 103
can be accessed. That is, data (WRIT
E DATA) can be written to the main memory 103, and data (READ DATA) can be written from the main memory 103.
) can be read. This read data is applied to the information processing device 102 and also to the error detection circuit 104 and diagnostic processing device 101 described above. Note that the main memory 103 is
The data read from the diagnostic processing bag! In step 101, the log information is stored in an external storage device (not shown) as log information. The information processing device 102 also includes a fault detector #IC (not shown), and when a fault is detected by this mechanism, a fault report (CHECK) is sent to the diagnostic processing device 101 via the fault report line 111. . Here, when a failure occurs in the main memory 103, an error in data (READ DATA) read from the main memory 103 is detected by the error detection circuit 1.
04, an error report is made to the information processing device 102, and the information processing device 102 reports to the diagnostic processing device 101 via the fault report line 111 that a fault has occurred in the main memory 103.

The diagnostic processing device 101 is a device that restarts the information processing device 102, and has a means for extracting and restoring software visible information of the information processing device 102 using a scan path 113, and a device for restarting the information processing device 102. In addition to means for sending data (DATA) to and controlling it, the following means are provided.

- Diagnosis read means (not shown) Address register 10 for main memory access in information processing device 102 in order to extract data in main memory 103
This is a means for reading data stored in the main memory 103 by setting an address (ADR3) to the main memory 103 via the diagnostic read address setting path +12. Note that the main memory 1 set in the address register 105 in the diagnostic read
A copy of address 03 (ADR3) is held in the diagnostic read address register 106. Also, during this diagnostic read, the error detection circuit 104 detects errors in the read data.

- Restarting control means 109 This is means for performing main control related to restarting the information processing device 102. If the error detection circuit 104 detects an error in the read data during diagnostic reading of an address in the main memory 103, the restart control means 109 temporarily suspends the diagnostic reading and activates the reloading unit 107. The reload unit 107 reloads the address held in the diagnostic read address register 106, that is, the address of the main memory 103 where the error was detected, with recovery data from an external storage device (not shown).

Note that the load destination address is an address held in the diagnostic read address register 106.

When the reloading section 107 finishes reloading the data, the restart control means 109 resumes diagnostic reading for the next address.

Next, the operation of this embodiment configured as described above will be explained.

When some kind of failure is detected in the information processing device 102,
This is reported to the diagnostic processing device 101 via the fault report line 111.The diagnostic processing bag 2101 receives this report and
A process of extracting software visible information in the information processing device 102 using the scan path 113, a process of determining the possibility of restarting while reading data in the main memory 103 using a diagnostic read, and restarting is possible. The process of clearing the information processing device 102 when it is determined that the information processing device 1
02, reloading the main memory data, and then making the information processing device 102 operational.

FIG. 2 shows that among the above-mentioned processes performed by the restart control means 109, the information processing device 102 is
This shows a part of the processing that is executed when an access error to the main memory 103 is reported from. As shown in the figure,
When a main memory access error occurs, the data is read out one by one from the main memory 103 by diagnostic read (S1), and the error detection circuit 104 checks whether an error has been detected in the read data (S2). If not, the process proceeds to step S4, and if an error is detected, the reload unit 107 reloads the predetermined data to the main memory address pointed to by the address of the diagnostic read address register 106 that holds a copy of the read address ( S3), proceed to process S4. In process S4, it is determined whether diagnostic reading has been performed for all addresses in the main memory 103, and if it has not been completed, the diagnostic read address is counted up. S5), and the process returns to process S1 for the next Perform a diagnostic read of the address, and if all addresses have been completed, continue the remaining startup processing necessary. Check the normality of the data each time you extract individual main memory data with a diagnostic read, such as 0 or more. By doing so, the fault location in the main memory 103 is located, and reloading is performed only for the fault location.

〔Effect of the invention〕

As explained above, in the restart device for an information processing device according to the present invention, when a fault occurs in the main memory, the location of the fault can be accurately identified by sequentially reading data in the main memory and checking for errors. By identifying the fault and reloading data only to the main memory address where the fault exists, it is possible to quickly restart the system.

[Brief explanation of the drawing]

FIG. 1 is a block diagram showing an example of a computer system to which the present invention is applied, FIG. 2 is a flowchart showing part of the processing of the restart control means 109, and FIG. 3 is a flowchart showing a part of the processing of the restart control means 109. This is a flowchart. In the figure, 101...Diagnostic processing device 102...Information processing device 103...Main memory 104...Error detection circuit 105...Address register 10 for main memory access
6...Diagnostic read address register 107...Reload unit 109...Restart control means 110...Error report line 111...Failure report line 112...Diagnostic read address setting bus 13...・Scan path

Claims (1)

[Scope of Claims] In an apparatus for restarting an information processing apparatus when a failure occurs in the main memory, the diagnostic read means sequentially reads data stored in the main memory when a failure occurs in the main memory. and an error detection means for detecting an error in the data sequentially read by the diagnostic read means, and a predetermined data only for the corresponding address of the main memory where an error in the read data is detected by the error detection means. 1. A restarting device for an information processing device, comprising a restarting control means including a reloading section for resetting.