Kernel Memory Error Detection for Selected Processes in Real Time
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for detecting memory errors in operating system kernels are inefficient and degrade kernel performance due to the need for source code modifications, which are not applicable to closed-source systems, and perform tests on all processes, causing unnecessary performance degradation.
Innovation Solution
A method for real-time memory error detection in operating system kernels that selectively allocates memory areas with red zones and shadow memory for specific processes, using interrupt handlers to minimize performance impact by only testing object processes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If memory error detection is applied to all processes in the kernel, then memory error detection coverage is improved, but kernel performance deteriorates
Solution Approach 1:
The patent segments the kernel memory space into detection object memory areas and normal memory areas. By dividing the memory detection scope, the system applies error detection mechanisms only to specific segments (detection object memory areas) rather than the entire kernel memory, thereby reducing the performance overhead while maintaining detection coverage for critical regions.
Solution Approach 2:
The patent implements local quality by applying different memory detection policies to different processes. Detection object processes receive enhanced monitoring with red zones and shadow memory, while other processes use normal memory allocation. This localized approach ensures high reliability for critical processes without degrading overall kernel performance.
2Reliability
If source code modification is applied to implement memory error detection, then detection capability is improved, but adaptability to closed-source systems deteriorates
Solution Approach 1:
The patent introduces an intermediary mechanism that operates at the memory management level rather than requiring source code modifications. By using kernel-level memory allocation hooks and shadow memory structures, the system achieves memory error detection functionality without directly modifying application source code, enabling deployment on both open-source and closed-source systems.
3Measurement precision
If red zone and shadow memory are allocated for all memory areas, then memory error detection accuracy is improved, but memory overhead increases
Solution Approach 1:
The patent segments the memory detection resources and allocates red zones and shadow memory only to detection object memory areas rather than all memory areas. This selective allocation maintains high detection accuracy for critical memory regions while significantly reducing the overall memory overhead by excluding normal memory areas from the detection mechanism.
Data Source
AI summary
The present embodiments provide a method and an apparatus for detecting an error of a kernel memory in real time, wherein the method and the apparatus select only an object process in real time without a source code and check a memory error of an operating system kernel only for the object process, thereby minimizing an impact on kernel performance.