Abstract
The decompression of selected code fragments during program execution using a software-based technique is discussed. The profiling information from the original, uncompressed program is used to choose code fragments that are infrequently executed. This limit the effect of time overhead involved in dynamic decompression on the execution speed of the entire program. It is expected that the better compression of the infrequently executed code can contribute to a significant improvement in overall size reduction, but the increased decompression effort does not lead to a significant runtime penalty. The benefits of using data compression techniques to dynamically decompressed infrequently executed codes are also discussed.
| Original language | English (US) |
|---|---|
| Pages (from-to) | 54-60 |
| Number of pages | 7 |
| Journal | Communications of the ACM |
| Volume | 46 |
| Issue number | 8 |
| DOIs | |
| State | Published - Aug 2003 |
ASJC Scopus subject areas
- General Computer Science