JPS5942343B2 - Trace information logging method - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a trace information logging method, and particularly to a trace information logging method that allows trace information to be directly read by a program.

As is well known, in a data processing system where multiple devices are connected via a common bus, if an unrecoverable error occurs in the system and the system goes down, trace information of the common bus is left as log information. The log information is then used to analyze the cause of the error.

In the conventional technology, signals on a common bus are traced to a address area by a tracing means in the hardware configuration, and when an error occurs, the signal is written to an external recording medium by a logout mechanism in the hardware configuration. However, providing a hardware-based logout mechanism increases costs, and in the event of a system failure, the logout mechanism itself may not function perfectly, resulting in a situation where trace information is not recorded on the external storage medium. is expected to occur. In order to write trace information to an external storage medium without increasing hardware, it has been proposed to write trace information in a trace area onto an external storage medium by executing a program.
In this proposed method, the program for logging is executed in the same state after the error occurs, so
It is also expected that an error will occur in the program processing for this logging, and that trace information will not be recorded on the external storage medium. The present invention takes the above points into consideration and aims to enable reliable logging of trace information using a program.

Therefore, the trace information logging method of the present invention is
A common bus, a plurality of devices connected to the common bus, and a trace means for storing signals on the common bus in a trace area, and when an error occurs, it is recorded in the trace area. A trace information logging method in a computer that writes trace information to an external storage medium by executing a program, which freezes the function of the tracing means upon the occurrence of an error,
Prevent the trace information recorded in the above trace area from being changed, then re-initial program load the above program, and execute the program to update the trace information recorded in the above trace area. It is characterized in that it is written to the external storage medium.

DESCRIPTION OF THE PREFERRED EMBODIMENTS The present invention will be described below with reference to the drawings. FIG. 1 shows a configuration example of a data processing system to which the present invention is applied, and FIG. 2 shows a flowchart illustrating an embodiment of the present invention.

In FIG. 1, 1 is a common bus, 2 is an external storage medium such as a magnetic disk device, 3 is a storage device, 4 is a main processor, and 5 is a trace in which trace information of the common bus 1, that is, address information and data information is written.・Represents each area.

The processing operations will be sequentially explained below with reference to the flowchart shown in FIG. (1) When an error occurs during data processing, the operation of writing trace information to the trace area 5 is stopped.

In other words, the trace function is frozen. (To) When initial program loading (PL) is performed again, it is first determined whether or not trace information at the time of error occurrence is recorded in the trace area 5. In addition, IP
Since the registers and memory are initialized by L, the program executed after 1PL again will not be affected by the above error. (3) If the judgment result is "Yes", the above trace area 5 The trace information at the time of error occurrence is read from the CPU 2 and the information is recorded on the external storage medium 2.

(4) Then, it instructs to start writing the trace information to the trace area 5, and starts executing the job.

(3) On the other hand, if the above judgment result is "NO", an instruction is immediately given to start writing processing of trace information to the above trace area 5, and execution of the job is started.

Here, the processing of above I2 (3) and (4) and the said (3)
The processing is performed by a program. As explained above, the present invention allows trace information to be read directly by a program, and log information can be re-recorded after a system failure.
It is now possible to do this programmatically during PL. In other words, in addition to the simple trace function, (1) start of trace, (2)
) report the presence or absence of trace information in the event of an error, and (3) read trace information according to instructions from the program, freeze the trace function when an error occurs, and have a function that will not be reset except when the power is turned off. I made it.
Therefore, it is possible to reliably record trace information at the time of an error.

[Brief explanation of drawings]

FIG. 1 shows a configuration example of a data processing system to which the present invention is applied, and FIG. 2 shows a flowchart for explaining an embodiment of the present invention.

Claims (1)

[Claims] 1 A common bus, a plurality of devices connected to the common bus, and a trace means for storing signals on the common bus in a trace area, and when an error occurs, the signal is recorded in the trace area. This is a trace information logging method for computers in which trace information is written to an external storage medium by executing a program, and the function of the tracing means is frozen when an error occurs, and the trace information is recorded in the trace area. Make sure that the trace information in the program is not changed, and then re-initial program load the above program.
A trace information logging method, characterized in that trace information recorded in the trace area is written to the external storage medium by executing the program.