Title :
SYZYGY - a framework for scalable cross-module IPO
Author :
Moon, Sungdo ; Li, Xinliang D. ; Hundt, Robert ; Chakrabarti, Dhruva R. ; Lozano, Luis A. ; Srinivasan, Uma ; Liu, Shin-Ming
Author_Institution :
Java, Compilers, & Tools Lab., Hewlett-Packard Co., Cupertino, CA, USA
Abstract :
Performing analysis across module boundaries for an entire program is important for exploiting several runtime performance opportunities. However, due to scalability problems in existing full-program analysis frameworks, such performance opportunities are only realized by paying tremendous compile-time costs. Alternative solutions, such as partial compilations or user assertions, are complicated or unsafe and as a result, not many commercial applications are compiled today with cross-module optimizations. We present SYZYGY, a practical framework for performing efficient, scalable, interprocedural optimizations. The framework is implemented in the HP-UX Itanium® compilers and we have successfully compiled many very large applications consisting of millions of lines of code. We achieved performance improvements of up to 40% over optimization level two and compilation time improvements in the order of 100% and more compared to a previous approach.
Keywords :
optimising compilers; program control structures; program diagnostics; storage management; HP-UX Itanium® compilers; SYZYGY framework; cross-module optimizations; interprocedural optimizations; partial compilations; program analysis; user assertions; Costs; Java; Memory management; Moon; Optimization methods; Performance analysis; Program processors; Runtime; Scalability; Usability;
Conference_Titel :
Code Generation and Optimization, 2004. CGO 2004. International Symposium on
Print_ISBN :
0-7695-2102-9
DOI :
10.1109/CGO.2004.1281664