JP2000010793A - OS macro issue PC determination method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To distinctively inform a user of a micro issue place where abnormality occurs and to facilitate an analysis of the abnormality by storing in a stack 'disp' indicating the relative position from the current stack pointer to the stack address where PC when an OS macro is issued in a task is stored. SOLUTION: A microlibrary 20 calculates the distance from the current stack pointer to the address of an area 46 where the PC in the stack 10 as a microlibrary right before a trap instruction is executed (22). The value is stored as disp in the address that the current SP indicates (21) and then the trap instruction is executed (22). Consequently, when an OS macro 30 detects an error of a requested process (31), disp stored in the area 44 of the address that current SP indicates plus 4 addresses is taken out and the area 46 at a distance of a relative value indicated by the disp from there is read out to obtain a desired PC.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a computer system in which a PC is loaded on a stack when a subroutine call and an exception occur, and an OS macro for requesting the OS from a task, which is a unit of work of a user program, to the OS. The present invention relates to a determination method for knowing a PC of a task that has issued an OS macro when issued.

[0002]

2. Description of the Related Art Generally, in a computer system, a user program operates in units of work called tasks.
The task advances its processing while using an OS function called an OS macro. To link to the OS macro, a subroutine called a macro library is prepared, and each task can link to the OS macro, that is, request the OS to perform processing by calling the macro library. Here, the OS macro checks whether there is an error in the requested processing, and if there is an error, notifies the user by outputting a call PC of the OS macro to a message so that the error location can be analyzed quickly and easily.

In a conventional computer system, the PC
Will be described with reference to FIG. Here the stack 40
An area is used to store a return PC when a subroutine is called from the main program, an area for transferring parameters, a work area used in the program, and the like. The used area of the stack area is constantly changing, and the extent to which the stack area is used is managed by the stack pointer SP.

In a computer system, tasks operate at a user level of 1 and an OS at a kernel level of 2.
In order to link the task 10 to the OS macro, first, the macro library 20 prepared as a subroutine is called subroutine calls 11, 12, and 13. The macro library 20 is composed of user level 1 to kernel level 2
Executes the special instruction trap22 to move to OS macro 3 and generates an exception (60) to enable OS macro 3
Link to 0. Subroutine call PC at this time
Also, the exception generation PC is stacked on the stack 40.

That is, OS macro issuance 11 or 12
Or P when the macro library is called in step 13
C (PC1, PC2, or PC3) is stacked on the stack (46), and a pointer SP (stack pointer) (41) indicating how much the stack is used is (42-
Refers to 1).

Next, PC4 when the trap instruction is executed
Is stored in the next stack area 43 used as a work by the macro library, and SP indicates (42-2).
When the OS macro 30 detects an error in the requested processing (31), the PC (PC1 or PC2) for knowing the macro issuer 11 or 12 or 13 originally.
Or PC3) should be output 33 in the message, but the SP has already progressed to (42-2), and the OS macro 30 knows only SP (42-2), and how it is loaded on the actual stack It is not known whether it is in the state of 43, 45 or 46.

Therefore, PC1 or P
There is no way to obtain the address of the stack area 46 that remembers C2 or PC3, and there is no choice but to output the trap execution PC (PC4) of the macro library to the message. In this case, when the task 10 issues the same OS macro 11, 12, 13, even if an error is detected in the OS macro, the PC from which a message is output is always the macro library PC (PC4), and the user It is difficult to know whether the abnormal OS macro issuance is 11, 12, or 13.

[0008]

In the OS macro issuing PC determination method according to the conventional method described above, when the OS macro detects an error in the requested processing, the return PC output as a message points to the macro library. However, when a user issues a plurality of identical OS macros in a task, there is a problem that the user cannot know the location of the abnormal macro issuance.

As described above, according to the present invention, even when a plurality of identical OS macros are issued from a task via a macro library, the original OS macro issuance location is output in the message, thereby causing an abnormality to the user. It is an object of the present invention to clearly notify a macro issuance location and facilitate anomaly analysis.

[0010]

As a mechanism for knowing where the target PCs 11, 12, and 13 are stored in the stack in the OS macro 30, immediately before linking from the macro library to the OS macro by a trap. The disp indicating the relative position from the current stack pointer to the stack address where the PC at the time of issuing the OS macro in the task is stored in the stack, so that the OS macro issuing PC can be traced on the stack. I do.

[0011]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below with reference to FIG.

The operation is the same as before until the task 10 calls the macro library 20. In this embodiment, immediately before executing 22 the trap instruction, the macro library 20 determines from the current stack pointer up to the address of the area 46 where the PC (PC1, PC2, or PC3) of the task that is the macro library call source is stored. Is calculated, and the value is stored as a disp in the address (42-2) indicated by the current SP (21), and then trap
Execute the instruction (22).

Thus, when the OS macro 30 detects an error in the requested processing (31), the OS macro 30 extracts the disp stored in the area 44 of the address +4 address indicated by the current SP (42-3), and By reading out the area 46 separated from the relative value indicated by disp therefrom, P
C1 or PC2 or PC3 can be obtained.

Thus, it becomes possible to output the PC (PC1, PC2, or PC3) of the task from which the OS macro is issued as a message. Therefore, task 10
However, even when the same macro issuance 11, 12, and 13 are performed, if an error is detected in the OS macro, the PC that outputs a message becomes the PC of the task that is the macro issuance source, and the user issues the abnormal macro issuance. The location can be clearly known, and the program abnormality can be easily and quickly analyzed.

[0015]

As described above, according to the present invention, it is possible to obtain the PC of the macro issuing task in the OS macro from the disp information stored in the stack. By outputting the macro issuing PC to the message, the user can easily know the issuing PC, and the analysis of the program abnormality can be easily and quickly performed.

[Brief description of the drawings]

FIG. 1 is an explanatory diagram of a PC determination method using disp information.

FIG. 2 is an explanatory diagram of a PC determination method in a conventional computer system.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 ... User level, 2 ... Kernel level, 10 ... Task, 11-13 ... OS macro issuance, 20 ... Macro library, 21 ... Disp information storage, 22 ... Trap, 30 ... Macro, 31 ... Abnormality determination, 32 ... Error message output,
40: stack, 41: stack pointer, 42-1
42-3: Stack pointed by stack pointer, 43 ... tr
ap execution PC, 44 disp information, 45 macro library work area, 46 macro library execution PC, 47
... Stack pointed to by disp.

 ──────────────────────────────────────────────────続 き Continuing on the front page (72) Inventor Keiji Kuwahara 5-2-1 Omikacho, Hitachi City, Ibaraki Prefecture Inside the Omika Plant, Hitachi, Ltd. (72) Inventor Junichi Kobayashi 5-chome, Omikacho, Hitachi City, Ibaraki Prefecture No. 1 Hitachi Process Computer Engineering Co., Ltd. F-term (reference) 5B098 BA05 BA06 GA02 GA04 JJ06

Claims (1)

[Claims] 1. A macro library which is a computer on which a PC (program counter) at the time of occurrence of a subroutine call and an exception is stacked and which is linked from a task which is a unit of work of a user program to an OS function called an OS macro. In a computer system that has a subroutine called, the macro library generates an exception, and a disp indicating where the stack area where the PC of the task that called the macro library is stored immediately before linking to the OS is stored. , The OS macro can obtain the PC of the user task that issued the macro.
S macro PC determination method.