JPH0375844A - Automatic analysis system for fault - Google Patents

Abstract

PURPOSE:To analyze a dump when a panic occurs without a manual operation by retrieving a code analysis table and an address symbol corresponding table in a system editing and outputting the result. CONSTITUTION:The system consists of a data input means 1, the code analysis table 2, a code analysis means 3, the address symbol corresponding table 4, an address symbol conversion means 5, a data editing means 6 and a data output means 7. When the panic occurs, a panic code and a physical address where the fault occurs are taken out from a dump area, and an explanation sentence corresponding to the panic code, a measure and the name of a symbol corresponding to the physical address are retrieved. Then, edited data is outputted from the system. Thus, the need of the manual operation using a manual and a corresponding table is eliminated.

Description

DETAILED DESCRIPTION OF THE INVENTION C. Industrial Field of Application The present invention provides a method for detecting a fault in an operating system that stores code representing the state of the fault and memory contents in a dump area when a fault that makes it impossible to continue processing is detected. Regarding analysis method.

[Prior Art] Conventionally, this type of failure analysis has been carried out using manuals and the like that include explanations of codes (hereinafter referred to as panic codes) that represent the failure state of failures that make it impossible to continue processing (hereinafter referred to as panics). This was done manually using a printed address-symbol correspondence table.

If the operating system panics, it outputs a message and terminates the process. In order to investigate the cause of the panic, the memory image at that time (
It's called a dump. ) and examine the contents of the data.

[Problems to be Solved by the Invention] In the conventional failure analysis method described above, output is output to this dump terminal or list, the panic code and physical address where the failure has occurred are found, and the address and symbol name are The problem was that the cause had to be determined by manually checking the correspondence table and panic code manual, which was time-consuming and complicated.

[Means for analyzing problems]

The automatic fault analysis method of the present invention includes data input means for extracting a code representing a fault state and a physical address that caused the fault from a dump area, and explanatory text corresponding to each of the codes representing the fault state. a code analysis table storing countermeasures; a code analysis means for searching the code analysis table with a code representing the state of the fault extracted by the data input means and extracting a corresponding explanatory text and countermeasure; and a physical address. an address-symbol correspondence table storing symbol names corresponding to the address-symbol correspondence table; and address-symbol conversion means for searching the address-symbol correspondence table with the physical address taken out by the data input means and extracting the corresponding symbol name; data editing for editing the code, explanatory text, countermeasure, physical address, and symbol name extracted by the data input means, the code analysis means, and the address/symbol conversion means and representing the state of the fault; and a data output means for inputting the data edited by the data editing means and outputting it to a terminal or list.

[Function] When a panic occurs, the panic code and the physical address that caused the failure are retrieved from the dump area, the explanatory text corresponding to the panic code, the countermeasure, and the symbol name corresponding to the physical address are searched and the edited data is saved. Since it is output from the system, there is no need for manual work using manuals or correspondence tables.

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

FIG. 1 is a block diagram of an embodiment of the automatic fault analysis method of the present invention, FIG. 2 is a flowchart of the processing of the code analysis means 3 of FIG. 1, and FIG. 2 is a flowchart of processing.

The data input means 1 takes out the dump tape (not shown) at the time of the panic from the dump area (not shown), and also takes out the panic code and the physical address where the failure occurred from the dump data.

The code analysis table 2 stores panic codes, explanations and countermeasures corresponding to each panic code. The code analysis means 3 searches the code analysis table 2 for the panic code extracted by the data input means 1, and extracts the explanatory text and countermeasures corresponding to the panic code. The address/symbol correspondence table 4 stores physical addresses and symbol names corresponding to each physical address. The address/symbol conversion means 5 converts the address/symbol using the physical address taken out by the data input means 1.
The symbol correspondence table 4 is searched and the symbol name corresponding to the physical address is extracted. The data editing means 6 stores the panic code and physical address taken out by the data input means 1, the explanatory text and countermeasure taken out by the code analysis means 3, and the symbol name taken out by the address/symbol conversion means 5. Input and edit for output. Data output means 7 The data edited for output by the data editing means 6 is input and output to a terminal or a list (not shown).

Next, the operation of this embodiment will be explained.

When a panic occurs, the data input means 1 first takes out dump data from the dump area, and then takes out the panic code and the physical address that caused the failure from the dump data.

Next, as shown in FIG. 2, the code analysis means 3 first reads one record of the code analysis table 2 (step 11).
), check whether it is the same as the panic code retrieved by data input means 1 (Step 12). If it is different, return to the successive processing of Step 11, and if it is the same, write the corresponding explanation and countermeasure for each. Storage area (not shown)
(Step 13). At the same time, the address/symbol conversion means 5 reads the address/symbol correspondence table 4 into one record 1 (step 2), as shown in FIG.
1), check whether it is the same as the physical address retrieved by the data input means 1 (step 22); if it is different, return to the reading process of step 21; if it is the same, the corresponding symbol name is stored in the storage area (not shown). ) (step 23). Subsequently, the data editing means 6 extracts the panic code and its explanation as described above,
Workaround and edit the physical address and its symbolic name. The data edited by the data editing means 6 is transferred to the data output means 7
to the terminal or list.

[Effects of the Invention] As explained above, the present invention searches the code analysis table and the address/symbol correspondence table within the system, edits the results, and outputs them, thereby eliminating the need for any human intervention when a panic occurs. Since it is possible to analyze dumps of the following types, there is an effect that failure analysis can be performed easily and in a short time.

[Brief explanation of drawings]

FIG. 1 is a block diagram of an embodiment of the automatic fault analysis method of the present invention, FIG. 2 is a flowchart of the processing of the code analysis means 3 of FIG. 1, and FIG. 2 is a flowchart of processing. 1...Data input means, 2...Code analysis table, 3...Code analysis means, 4...Address/symbol correspondence table, 5.
...Address/symbol conversion means, 6...
- Data editing means, 7...Data output means, 11-1321-23...Steps.

Claims (1)

[Claims] In an operating system that stores a code representing the state of the fault and the contents of memory in a dump area when a fault that makes it impossible to continue processing is detected, the code representing the state of the fault and the contents of the memory are stored from the dump area. a data input means for extracting the physical address that caused the fault; a code analysis table storing explanations and countermeasures corresponding to each of the codes representing the state of the fault; and the fault extracted by the data input means. a code analysis means for searching the code analysis table with a code representing the state of the code and extracting a corresponding explanatory text and countermeasure; an address/symbol correspondence table storing physical addresses and corresponding symbol names; and the data input. address/symbol converting means for searching an address/symbol correspondence table with the physical address extracted by the means and extracting a corresponding symbol name; data editing means for editing a code, explanation, countermeasure, physical address, and symbol name representing the state of the failure for output; and data for inputting the data edited by the data editing means and outputting it to a terminal or list. An automatic failure analysis method having an output means.