A method and system for dynamically upgrading software based on intermediate auxiliary functions

Abstract

The invention discloses an intermediate auxiliary function-based method for dynamically updating software. The method comprises the following steps: generating an intermediate auxiliary function by analyzing to-be-updated functions in integrated new / old version of code source files; respectively updating to-be-updated functions located at a function call stack, to-be-updated functions not located at the function call stack and to-be-updated static variables by utilizing the intermediate auxiliary function and a binary rewriting mechanism. Correspondingly, the invention also discloses a corresponding system. According to the method and the system disclosed by the invention, system loss and deadlock problems brought by inserting a plurality of check points are effectively solved; while updating security is ensured, real-time dynamic updating can be performed on a program, and operation complexity of a user and system overheads are obviously reduced; meanwhile, rollback can be finished when any exception occurs during updating to ensure reliability of an updating process. The method and the system disclosed by the invention are suitable for application services which need high-reliability continuous demands, and are capable of reducing loss caused by system restart due to version updating.

Description

technical field [0001] The invention belongs to the security field in cloud software reliability research, and more specifically relates to a method and system for dynamically upgrading software based on intermediate auxiliary functions. Background technique [0002] The software needs to be continuously upgraded and updated to fix bugs and add functions during the use process. However, traditional software updates need to close the software first, and then restart the software after the update is completed. In this way, on the one hand, the software service will be down for a period of time, and on the other hand, the software service will lose the state before the update. For software services that require 7 / 24 hours, this undoubtedly reduces the availability of services. But if the update is not adopted, it will undoubtedly reduce the security of the software and the reliability of the service. To solve this problem, the software dynamic upgrade method can complete the ...

AI Technical Summary

Problems solved by technology

[0007] Some dynamic software upgrade methods (polus) can only update functions that are not in the function call stack, and cannot update functions that are being executed in real time. If some functions that cannot return in a short time, such as the main loop function and scheduling function, will not be able to update to the next version

Some dynamic upgrade methods (upstare) require specific compiler support, so programs compiled with general compilers cannot be dynamically updated

Some dynamic software upgrade methods (Kitsune) take the process as the update unit. During the update process, all the code data of the new version needs to be stored, resulting in high overhead, but the actual update often only has a few changes.

The security update point of some dynamic update methods can only be applied to single-threaded programs, and multi-threaded programs cannot reach the security update point at the same time, so dynamic updates cannot be performed on multi-threaded programs

Images