JPH02304635A - Program runaway detecting system - Google Patents

Abstract

PURPOSE:To detect the runaway of a program in an early stage and to suppress the damage caused by the runaway to the minimum by providing a means for checking the validity of a return address extending from a program of a call destination to a program of a call origin. CONSTITUTION:At the time of return address copy processing of a program 2 of call destination, a return address brought to save to a stack 3 is fetched once and copied to a return address copy area 4 of a memory. Subsequently, the return address is reloaded to the stack 3 again. At the time of return address validity check processing 2c, the return address fetched from the stack 3 is compared with the return address copied in the return address area 4. As a result, in the case of coincidence, it is decided to be valid and returned to a call origin program 1, and in the case of discrepancy, it is decided to be an error of the return address and a runaway state of the program is recognized. Also, a program stop and an error processing by an error processing routine are executed.

Description

DETAILED DESCRIPTION OF THE INVENTION [Summary] The present invention relates to a method for detecting program runaway when performing linked processing between a main routine and a subroutine, or between subroutines, etc. in a computer system.

The purpose of this invention is to enable early detection of program runaway and to minimize damage caused by program runaway by providing a means for checking the validity of the return address from the called program to the calling program.

In a computer system that saves and manages return addresses, etc. on a stack when calling another program from a running program, and performs coordination processing between multiple programs, the called program reads the return address from the stack. The return address is copied to a predetermined area of the memory, then the original processing is executed, and after the original processing is finished, the return address is read from the stack again, and this is copied to the predetermined area of the memory. It is configured to compare it with the address and determine that the program has runaway if there is a mismatch.

[Industrial application field]

The present invention relates to a method for detecting program runaway when performing linked processing between a main routine and a subroutine, or between subroutines, etc. in a computer system.

[Conventional technology]

Figure 4 (al) shows an example of program linkage processing.
In the illustrated example, main routine A calls subroutine B during processing, and subroutine B further calls subroutine C during its processing.

When calling these subroutines, return information is saved at the time of the call. The return information includes a return address, register information, etc. When the called subroutine finishes processing, the previously saved return information is retrieved, and based on that, the caller's main routine or subroutine resumes processing.

When subroutines are called multiple times, as in the linked process shown in FIG.

Restoration needs to be controlled quickly. Conventionally, the management of this return information has been carried out under the control of the O8 using a stack and stack pointer as shown in FIG.

In FIG. 4Cb), ■ is the state of the stack while main routine A is being executed, and is empty.

■ is the state of the stack when main routine A calls subroutine B, "A" indicates the saved return information of main routine A, and ■ is the state of the stack when subroutine B calls subroutine C. where "B" indicates return information to the saved subroutine B. ■ is the state of the stack when returning from subroutine C to subroutine B, and “B” is returned from the stack.
is extracted and used. And ■ is subroutine B
This is the state of the stack when returning from main routine A to main routine A, and "A" is taken out from the stack and used.

[Problem to be solved by the invention]

In conventional computer systems, during program linkage processing, a program error may cause the program to run out of control, causing the stack pointer to shift or the stack value to be destroyed. In this state, a return instruction is executed from the called program. As a result, the called program incorrectly returns to the calling program, causing damage such as memory contents being destroyed due to runaway programs, and making it difficult to discover bugs.

The present invention enables early detection of program runaway and prevents damage caused by program runaway by providing means for checking the validity of the return address from the called program to the calling program in program linkage processing. The aim is to keep it to a minimum.

[Means to solve the problem]

In the present invention, when a program is called, the called program copies the return address saved on the stack into memory, and when processing ends, copies the return address retrieved from the stack into memory. The return address of the stack is compared with a certain address if the two match, and the return address of the stack is determined to be valid, and if the two do not match, it is recognized as a runaway.

FIG. 1 is a diagram showing the principle of the present invention.

In fig.

1 is a calling program such as a main routine or subroutine.

2 is a called program such as a subroutine.

2a, 2b, and 2c are the constituent parts of the called program 2, where 2a is the return address copying process placed at the beginning of the program, 2b is the original process, and 2C is the return address validating process placed just before the return instruction. This is a gender check process.

3 is a stack for saving the return address.

4 is a return address copy area of the memory.

In the return address copying process of the called program 2, the return address saved in the stack 3 is once taken out, copied to the return address copy area 4 of the memory, and the return address is loaded back onto the stack 3.

In the return address validity check process, stack 3
The return address taken out from the stack is compared with the return address copied in the return address copy area 4 to determine the validity of the return address in the stack. If they match, it is assumed to be valid and a return to the calling program 1 is executed.

If they do not match, it is considered a return address error and the program is running out of control. This causes the program to stop and the error handling routine to perform error handling.

[Effect]

During processing by a program call, if the program goes out of control due to some program bug and the stack pointer shifts and the restack contents are destroyed, the return address will change.0 The present invention copies this when the called program starts processing. This can be easily detected by checking the validity using the returned address.

When multiple program calls are made.

Each called program performs processing to check the validity of this return address.

〔Example〕

FIG. 2 is a configuration diagram of a computer system according to an embodiment of the present invention, and only elements related to the present invention are shown in a simplified manner.

In the figure, 5 is the CPU, 6 is the register [R1], 7 is the stack pointer, and 8 is the memory with the stack area (1
), 9 is a memory (2) with a data storage area, and 10 is a return address copy area [M1].

FIG. 3 shows the flow of a linked program of the main routine 11 and subroutine 12 according to one embodiment of the present invention. These are executed under the computer system shown in FIG. The operation will be explained below.

The main routine 11 is composed of processing A, processing by subroutine calls, and processing B.

When a subroutine call is issued, No. 3 (not shown) saves the return address to the stack area of memory [1] pointed to by the stack pointer 7, and starts the subroutine 12. The subroutine has 2
Execute the following steps ■ or ■.

(2) The subroutine 12 first executes the instruction POP R1. This instruction pumps up the top return address from the stack and stores it in register (R1).

00 Next, the subroutine 12 executes the instruction PUSHR1. This instruction pushes the return address of register [R1] onto the stack and backs it up. At this time, the previous return address remains in register [R1].

00 Next, the return address of register [R1] is stored in the return address copy area (Ml).

(2) Subroutine 12 executes the original process C of 1.

(2) After completing processing C, execute command POP R1.

Pump up the return address from the stack and store it in register [R1].

(2) Next, execute the instruction PUSHR1 and load the return address back onto the stack.

(2) Compare the contents of the register (R1) and the contents of the return address copy area [M1].

Comparing (2) and (2), if R1=M1, a return instruction (not shown) is executed, and return control is performed by O8 using the return address of the stack.

Comparing (2) and (2), if R1≠M1, the program is stopped.

When returning to the main routine 11, processing B is executed from the return address.

〔Effect of the invention〕

According to the present invention, bugs in a program can be discovered early in cooperative processing of a plurality of programs.
Additionally, programs containing bugs can be easily identified, making bug analysis and investigation more efficient. Another advantage is that malfunctions caused by bugs can be prevented from expanding.

[Brief explanation of drawings]

FIG. 1 is a principle diagram of the present invention, FIG. 2 is a configuration diagram of a computer system according to an embodiment of the present invention, and FIG. 3 is a flow diagram of a program according to an embodiment of the present invention. FIG. 4 is an explanatory diagram of conventional program linkage processing. 1 in Figure 1: Caller program 2: Callee program 2a: Return address copy process 2b: Original process 2C: Return address validity check process 3 stack 4: Return address copy area Patent applicant: Bfn Corporation Manager Patent attorney Fumihiro Hase (2 others) This issue includes Sun and Moon Hara Koko - Figure 1 *! ffa month 1 prong fang 11: I run name + 1ji 9 (system futakeman Y cum 2 figures

Claims (1)

[Claims] In a computer system that saves and manages return information such as return addresses on a stack when calling another program from a running program, and performs linked processing between multiple programs, the called program returns from the stack. After reading the address and copying it to a predetermined area of the memory, the original processing is executed, and after the original processing is finished, the return address is read again from the stack and is copied to the predetermined area of the memory. A program runaway detection method that compares with the return address and determines that the program has runaway if they do not match.