Title :
Annotating user-defined abstractions for optimization
Author :
Quinlan, Dan ; Schordan, Markus ; Vuduc, Richard ; Yi, Qing
Author_Institution :
Center for Appl. Sci. Comput., Lawrence Livermore Nat. Lab., CA
Abstract :
Although conventional compilers implement a wide range of optimization techniques, they frequently miss opportunities to optimize the use of abstractions, largely because they are not designed to recognize and use the relevant semantic information about such abstractions. In this position paper, we propose a set of annotations to help communicate high-level semantic information about abstractions to the compiler, thereby enabling the large body of traditional compiler optimizations to be applied to the use of those abstractions. Our annotations explicitly describe properties of abstractions that are needed to guarantee the applicability and profitability of a broad variety of such optimizations, including memoization, reordering, data layout transformations, and inlining and specialization
Keywords :
computational linguistics; data flow analysis; optimising compilers; compiler optimization; conventional compiler; data layout transformation; memorization; relevant semantic information; reordering; user-defined abstraction; Computer languages; Computer science; Containers; Costs; Design optimization; Laboratories; Optimizing compilers; Profitability; Scientific computing; US Department of Energy;
Conference_Titel :
Parallel and Distributed Processing Symposium, 2006. IPDPS 2006. 20th International
Conference_Location :
Rhodes Island
Print_ISBN :
1-4244-0054-6
DOI :
10.1109/IPDPS.2006.1639722