"Researchers Develop New Approach That Protects 90 Percent Of Stack Memory Data"

An international Penn State-led team has developed a new and more reliable approach to defending vulnerable data on the stack, which is a memory region responsible for storing computer program data for processes. This vulnerable data could include return addresses and other objects that can be exploited by malicious actors through memory errors to gain access to more data. Despite extensive research into defenses to protect stack objects from memory error exploitation, much stack data remains vulnerable, according to project lead Trent Jaeger, professor of computer science and engineering at the Penn State School of Electrical Engineering and Computer Science. Memory errors are classified into three types: spatial, temporal, and type. Spatial errors allow access to memory outside of the object's allotted space. Temporal errors allow access to memory before or after it was assigned, and type errors enable access by assuming a format other than the actual format of an object. In each case, an adversary may access objects other than those intended by the programmer when programming objects on the stack to access specific data. Recent stack defense methods are said to provide an incomplete view of security by failing to account for memory errors comprehensively and limiting the set of objects that can be protected. Therefore, the team has presented the DATAGUARD system, which improves security by performing a more comprehensive and accurate safety analysis that proves a greater number of stack objects are safe from memory errors while ensuring that no unsafe stack objects are mistakenly classified as safe. DATAGUARD uses static analysis and symbolic execution to validate stack objects free from spatial, type, and temporal memory errors. Jaeger explains that this process involves analyzing the safety of items that point to the objects and generating safety constraints for the objects' safety parameters before validating an object's safe or unsafe status. During tests, DATAGUARD identified and removed 6.3 percent of objects that the Safe Stack technique misclassified as safe, and proved that 65 percent of objects labeled as unsafe by Safe Stack were actually safe. DATAGUARD demonstrates that a more comprehensive and accurate but conservative analysis increases the scope of data protection to over 90 percent of stack objects on average, while also reducing overhead, or the extra run time the system uses to protect safe objects. This article continues to discuss memory errors and the new data security approach developed to protect against such errors while using less system power. 

PSU reports "Researchers Develop New Approach That Protects 90 Percent Of Stack Memory Data"