Title
Memory Footprint Reduction with Quasi-Static Shared Libraries in MMU-less Embedded Systems
Abstract
Despite a rapid decrease in the price of solid state memory devices, system memory is still a very precious resource in embedded systems. The use of shared libraries is known to be effective in significantly reducing memory usage. Unfortunately, many resource-constrained embedded systems lack MMU, making it extremely difficult to support this technique. To address this problem, we propose a novel shared library scheme called the quasi-static shared library. In quasi-static shared libraries, global symbols are bound to pseudo-addresses at linking time and the actual physical addresses are bound at loading time. This scheme is made possible by emulating MMU's memory mapping feature with a Data Section Base Register (DSBR) and a Data Section Base Table (DSBT). Quasi-static shared libraries do not require symbol tables which take up time and space at runtime. We have implemented the proposed scheme in a commercial ADSL (Asymmetric Digital Subscriber Line) home network gateway and conducted a series of experiments measuring its memory usage and performance overhead. The result is drastic: a 35% reduction in flash memory usage and a 10% reduction in RAM usage. These results were achieved with only a negligible performance penalty of less than 4%. Even though this scheme was applied to uClinux-based embedded systems, it can be used for any MMU-less real-time operating system.
Year
DOI
Venue
2006
10.1109/RTAS.2006.26
IEEE Real-Time Technology and Applications Symposium
Keywords
Field
DocType
library scheme,memory mapping feature,solid state memory device,memory usage,shared library,system memory,quasi-static shared libraries,memory footprint reduction,proposed scheme,quasi-static shared library,flash memory usage,ram usage,registers,home automation,embedded system,asymmetric digital subscriber line,real time operating system,dsl
Interleaved memory,Uniform memory access,Shared memory,Computer science,Distributed memory,Real-time computing,Data diffusion machine,Memory management,Memory map,Distributed shared memory,Operating system,Embedded system
Conference
ISBN
Citations 
PageRank 
0-7695-2516-4
2
0.36
References 
Authors
1
3
Name
Order
Citations
PageRank
Jaesoo Lee1223.91
Jiyong Park24011.02
Seongsoo Hong377177.57