Practical dynamic software updating for c
The challenge for making DSU practical is that DSU should be flexible, and yet safe, efficient, and easy to use.In this paper, we present Ginseng, a DSU implementation which can compile C programs specially to support dynamic updates.Software updates are a fact of life which means updating to add new some features and fix some bugs.Current software update methodology is stop/restart. However, many system would prefer no-line upgrades in lien of having to stop and restart the system every time it must be patched.DSU is appealing compared to other approaches for on-line upgrades because it is quite general and requires no redundant hardware.The challenge is in making DSU flexible, and yet safe and easy to use.Cambridge Epigenetics Club BCS East Anglia International Relations & History Working Group Neurodevelopment disorders of genetic origin – what can we learn?
Performance experiments show that all patches could be applied in less than 5 ms, and that the overhead on application throughput due to updating support ranged from 0 to at most 32%.
The majority of updates were easy to generate, though more effort was required for signficant functional changes.
Compiling for updateability had virtually no impact on server performance, but would be more significant for compute-bound programs. Hicks is an assistant professor in the Computer Science department at the University of Maryland, College Park. dissertation, he designed and implemented a general, language-based approach for dynamically upgrading running software.
His research bridges the areas of “systems” and programming languages, in that he frequently applies or develops language-based technology to solve problems that occur in systems software. This dissertation won the 2002 ACM SIGPLAN Doctoral Dissertation Award.
With his students he has been significantly extending this work to make it more flexible and practical, to apply to real C programs.