Method for supervising running stack and static data area overlapping

Abstract

The invention relates to a method for checking the superposition between operation stack and static data area, wherein in linkage device, it adjusts the distribution of operation stack and static data area in data memory, then detects it in software simulator when operating, adds the relation code between judgment stack point value and minimum address size of data memory in the software simulator; if the stack point value is lower than the minimum address, there is descending overflow of operation stack, and there is superposition error between them. The invention has the advantages that: (1), the simulator only needs to add the relation code between judgment stack point value and base address size of data memory, to automatically information mistake when there is descending overflow in operation stack; (2), it can avoid memory consumption; (3), the generated executable document can be loaded on hardware.

Description

technical field [0001] The invention relates to the detection of program runtime errors in an embedded system, in particular to a method for detecting the overlapping of a running stack and a static data area. Background technique [0002] Due to price and application reasons, the instruction memory and data memory in the embedded system are relatively small, so when the program is executed, there may be overlapping of different types of data (such as static data area, heap, and running stack) in the data memory. Condition. The static data section (static data sections) includes each data section in the ELF executable file. The loading addresses of these data areas are usually specified in linker scripts. In a simple embedded system, the running program only loads the instruction and data parts in the ELF executable file, and discards the Section Header Table (Section Header Table) that stores the size information of each section in the ELF file. However, most of such sim...

AI Technical Summary

Problems solved by technology

Although this method is much more accurate than the previous method, it still has relatively large disadvantages: first, it requires testers to make comparisons, and also requires testers to know the size of the static data area, so this method is not easy to use ; Disadvantage 2, adding instructions needs to occupy instruction memory, which is not suitable for programs with high pressure on instruction memory; Supported in , the executable file generated at this time cannot run in the actual hardware environment

Images