Reducing Program Image Size by Extracting Frozen Code and Data

Constraints on the memory size of embedded systems require reducing the image size of executing programs. Common techniques include code compression and reduced instruction sets. We propose a novel technique that eliminates large portions of the executable image without compromising execution time (due to decompression) or code generation (due to reduced instruction sets). Frozen code and data portions are identified using profiling techniques and removed from the loadable image. They are replaced with interrupts that load them in the unlikely case that they are accessed. The executable is sustained in a runnable mode.

Analysis of the frozen portions reveals that most are error and uncommon input handlers. Only a minority of the code (less that 1%) which was identified as frozen during a training run, is also accessed with production datasets.

Applying this technique results in a maximal code reduction of 90% (average 62%) in the image size of the SPECINT CPU 2000 benchmarks and 40% (average 20%) in the data image size. For the SPECFP CPU 2000 maximial code reduction reached up to 90% reduction (average 80%) and static data reductuion reached up to 90% (average 50%).