JPS63316242A - Diagnosing system - Google Patents

Abstract

PURPOSE:To improve a system throughput and to shorten a diagnosing executing time by executing autonomously the diagnosis with an input output control device in accordance with the contents of a diagnosis conditions list set beforehand. CONSTITUTION:A diagnosed activated input/output control device 3 reads a 'diagnosis conditions list' to show a diagnosis object, the number of the times of the execution, execution time, a diagnosis classification, etc., from a memory 20a of a main memory device 20, then analyzes the contents and executes autonomously a designation test. Next, when the diagnosis is completed, the diagnosis result is analyzed, a 'diagnosis result list' is prepared, and stored into a memory 20b of a main memory device 20. Thus, the role of a host software is only to give a single command for the preparations of the diagnosis conditions list and the diagnosis activation, the processing burden and the over- head are reduced, a system throughput is improved and the diagnosis execution time is shortened.

Description

[Detailed Description of the Invention] [Summary] In a diagnostic method in a data processing system, a diagnostic condition list is provided in the main memory in advance, and the input/output control device executes the above list by a diagnostic activation command from the central processing unit. By autonomously executing the diagnosis according to the contents of the data processing system, the processing load on the upper software in the data processing system is reduced, and the throughput of the system is improved.

[Industrial application field]

The present invention relates to a diagnostic method used in a data processing system such as a central processing system of an exchange.

[Conventional technology]

As shown in Fig. 3, a general data processing system operates under the control of a central processing unit (CPU) 1, a main memory (MM) 2, and a plurality of input/output control units (I Oco-n) 3. It is composed of input/output devices (10o~, ) 4, etc., and these are connected via a processor bus (PB) 5. The above 10C3 is a microprocessor (MP
U) 3a, peripheral control circuit (PHCTL) 3bS 103
ROM3 with built-in firmware to control
c, RAM 3d as work memory, processor bus interface circuit (PB-INF) 3e, I O
It includes an interface circuit (To-INF) 3f, a register group for upper-level software interface (CTL-REG) 3g, and the like. The main registers in CTLREG3i are the command register (CMR), device status register (DSR), and memory address register (M
AR) and a byte count register (BCR).

In addition, IO2 is, for example, a magnetic disk unit (DK
U) , (key tape unit (MT4J)
, personal computer (PC), etc.

The conventional diagnostic method for a data processing system with the above configuration is that upper software (software for executing the original processing of the system; for example, in the case of a switch, software for switching processing) performs detailed diagnosis of 10C3 and 104. It was designed to isolate equipment failures based on the results of each diagnostic phase. An outline of the process is shown in FIG.

In the figure, first, in step a1, diagnosis execution is started online. This is C in l0C3 shown in Figure 3.
The command register CMH in TLREG3g is set to
This is done by I (upper software) writing the command code CMC and subcommand code SUBCMC. The command code CMC is DIAG (=D
iagnosis (diagnosis) is used, and the subcommand code SUBCMC is 1 from (01)s to (FF)H.
A hexadecimal representation is used, each indicating a diagnostic location.

Next, in step at, the subcommand code SUBCM
Analyze C. For example, (01)o, (02)111 (
03) +1 is RAM3d and ROM in l0Ca respectively
3 Specify the test for c1PHCTL3 b, (04)
)l specifies a data transfer test between l0C3 and r04, and (05), . . . -(FF)H specifies other tests. Then, in step a3, diagnosis is executed only for each designated test obtained through the above analysis.

When the diagnosis is completed (step a4), the completion status of the corresponding test is stored in the device status register DS in step as.
Set to R. CMP ends here (Complete
on). After that, in step ab, the higher CPU
Executes an interrupt to I.

[Problem that the invention seeks to solve]

In the conventional diagnosis method described above, the upper software (online)
Since it is started and diagnosed by giving multiple commands via the command register CMR as necessary,
There was a problem in that the processing load and overhead on the host software increased, resulting in a decrease in system throughput. Additionally, this has also led to the problem that it takes a long time to diagnose and analyze the diagnosis results.

SUMMARY OF THE INVENTION In view of the above-mentioned problems, it is an object of the present invention to provide a diagnostic method that improves system throughput and reduces diagnostic execution time.

[Means for solving problems]

In the diagnosis method of the present invention, a "diagnosis condition list" indicating conditions at the time of diagnosis execution is provided in advance on the main memory (MM),
Based on the diagnostic startup command from the central processing unit (CPU),
The input/output control device (IOC) reads the above list, autonomously executes diagnosis according to the contents, and stores the analyzed contents of the diagnosis results as a "diagnosis result list" in the main memory (MM).
This feature is characterized in that it can be written to.

[For production]

In the diagnosis method of the present invention, as described above, the input/output control device autonomously executes diagnosis according to the contents of the "diagnosis condition list." As a result, the role of the higher-level software in diagnosis is reduced to preparing a list of diagnostic conditions and providing only one command for activating the diagnosis. Therefore, the processing load and overhead on the host software are reduced, which makes it possible to improve throughput and
Diagnostic execution time is also reduced.

〔Example〕

Embodiments of the present invention will be described below with reference to the drawings.

FIG. 1 is a system configuration diagram of a data processing system that also shows a processing outline of an embodiment of the present invention. In the same figure, a central processing unit (CPU) 1, an input/output control unit (I
The hardware configurations of the OC) 3 and the input/output device (TO) 4 are the same as the conventional device shown in FIG. However, the main storage device 20, unlike the main storage device 2 of FIG. 3, has memory areas 20a and 20b for storing a "diagnosis condition list" and a "diagnosis result list" which will be described in detail later. As a feature of this embodiment, the l0C3 is made intelligent to a certain extent, and it is possible to autonomously execute diagnosis using the above diagnosis condition list (diagnosis execution order ■-■-■-■-
■).

FIG. 2 specifically shows the diagnostic processing in this embodiment.

First, a "diagnosis condition list" that indicates the conditions for diagnosis execution (for example, diagnosis target, number of executions, execution time, diagnosis type, etc.)
is stored in the area 20a of the RAM 20. This setting is performed by the upper software. As a diagnostic condition list, for example,
As shown in Figure 2, the target of each test for diagnosis (
For example, RAM3d, RO in IOC3 shown in FIG.
M3c, PHCTL3b, etc.) for each test content (
For example, the above-mentioned number of executions, time, etc.) can be cited.

When executing a diagnosis, first, in step b1, diagnosis execution is started online. This is done by the CPU1 (upper software) writing the command code CMC to the command register CMR in l0Ca. However, at this time, unlike the conventional method shown in FIG. 4, there is no need to write the subcommand code SUBCMC to the command register CMR.

The command code CMC for starting diagnosis includes D T
AG (=Diagnosis) is used.

The ICO 3 that has been activated for diagnosis reads the above-mentioned "diagnosis condition list" from the memory area 20a of the main memory (MM) 20 in step b2. l0C3 then analyzes the contents of the diagnostic condition list in step b (for example, checks whether the execution flag for each test is rlJ (=execution) or rOJ (=execution skip)), and then performs step b4 accordingly.
Specified test (for example, ROM3c test in ICO3, data transfer test between 10C3 and 104, etc.)
be executed autonomously. The execution flag “1” and “0” are
Decide in advance online.

10C3, after completing the above diagnosis (Step 7 b)
, in step b, the diagnosis results are analyzed to create a "diagnosis result list", which is stored in the main memory (MM) in step b.
20 memory areas 20b. For example, as shown in Fig. 2, this diagnostic result list shows the results and pass/fail determination flags (rlJ = bad, "0" = good) for each test executed, and also shows temporary failures (retries). If there are any faults that have been remedied, the data showing the faults will be listed. l0C3 continues step b,
After setting the device status register DSR in the diagnostic completed state, the contents are transferred to the upper level (CPUI) in step b9.
) to send an interrupt notification. From the above-mentioned "diagnosis result list", the host software can know whether or not there is a fault in the device to be diagnosed (IOC3, 104), and if there is a fault, it can know the detailed contents thereof.

According to this embodiment, since the 10C3 activated for diagnosis by the higher-level software autonomously executes the diagnosis according to the contents of the "diagnosis condition list," the role of the higher-level software at this time is to prepare the diagnostic condition list. It only requires giving a single command (DIA, G) for diagnostic activation. Therefore, the processing load and overhead of the upper-level software are reduced, thereby improving system throughput. Also,
At the same time, the diagnostic execution time is shortened, and the number of man-hours for creating the diagnostic program can also be reduced accordingly.

Note that the above-mentioned diagnostic processing is executed separately on each l0Cz (i-0 to n) that receives the diagnostic activation from the CPUI, but if multiple IOCs and IOs are of the same type, these You can share one diagnostic condition list.

Further, in addition to the contents shown in FIG. 2, the "diagnosis condition list" can easily include various conditions corresponding to each device to be diagnosed, thereby making it possible to perform a very wide variety of diagnoses. Even in this case, by setting the execution flag appropriately for each test in advance, the higher-level software can start the diagnosis with just one command.

Furthermore, the diagnostic method of the present invention can be applied to various data processing systems in addition to the central processing system of an exchange.

〔Effect of the invention〕

According to the diagnostic method of the present invention, since the processing load and overhead of the upper-level software are reduced, the job usage time for other processing of the upper-level software is increased, and therefore, an improvement in system throughput is realized.

Further, by shortening the diagnostic program steps, the diagnostic execution time is shortened, and the number of man-hours for creating the diagnostic program is accordingly reduced.

Furthermore, the reliability of the system is improved by reducing the number of times the CPU accesses the input/output control device that is subject to diagnosis due to a failure or the like.

[Brief explanation of the drawing]

Fig. 1 is a system configuration diagram of a data processing system that also shows a processing overview of an embodiment of the present invention, Fig. 2 is a flowchart showing diagnostic processing in the embodiment, and Fig. 3 is a general diagram of the data processing system. FIG. 4 is a flowchart showing conventional diagnostic processing. 1... Central processing unit (CPU), 3... Input/output control device (IOC), 4... Input/output device (10), 20... Main memory (MM). Patent applicant: Fujitsu Ltd.
Figure 3

Claims (1)

[Claims] 1) A central processing unit (1), a main storage device (20), at least one input/output control device (3), and an input/output device that operates under the control of the input/output control device. In a data processing system comprising a device (4), the input/output control device (3) and the input/output device (4)
When executing the diagnosis, a diagnosis condition list indicating the conditions when executing the diagnosis is provided in advance on the main storage device (20),
In response to a diagnostic activation command from the central processing unit (1), the input/output control device (3) reads the diagnostic condition list from the main storage device (20) and autonomously performs the diagnosis according to the contents of the diagnostic condition list. A diagnostic method characterized in that the analysis content of the diagnostic results is written into the main storage device (20) as a diagnostic result list. 2) The input/output control device (3) includes at least a microprocessor (3a), a ROM (3c) containing firmware for controlling the input/output device (4), and a RAM (3c) as a work memory. 3d). The diagnostic method according to claim 1, characterized by comprising: