Imperfectly - Nested Loops Yonghong

نویسندگان

  • Yonghong Song
  • Zhiyuan Li
چکیده

This paper presents an integrated compiler framework for tiling a class of nontrivial imperfectly-nested loops such that cache locality is improved. We develop a new memory cost model to analyze data reuse in terms of both the cache and the TLB, based on which we compute the tile size with or without array duplication. We determine whether to duplicate arrays for tiling by comparing the respective exploited reuse factors. We identify compatible loops in order to improve the prootability of tiling. The preliminary results with several benchmark programs show that the transformed programs run faster by 9% to 282%.

برای دانلود رایگان متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

ثبت نام

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

منابع مشابه

Further Results for Improving Loop Interchange in Non-Adjacent and Imperfectly Nested Loops

Abstract Loop interchange is a powerful restructuring technique for supporting vectorization and parallelization. In this paper, we propose a technique which is better to determine whether loops can be interchanged between two non-adjacent loops. We also present a method for determining whether we can directly make loop interchange on an imperfectly nested loop. Some experimental results are al...

متن کامل

A Scheme for Detecting the Termination of a Parallel Loop Nest

One central problem in the execution of parallel nested loops with non-aane bounds is the precise scanning (i.e., enumeration) of the points in their iteration space and the detection of their termination. Scanning schemes have been proposed for both shared-memory and distributed-memory implementations. However, these schemes work only for perfectly nested while loops. We propose a scheme which...

متن کامل

Statement-Level Communication-Free Hyperplane Partitioning Techniques for Parallelizing Compilers on Multicomputers

This paper addresses the problems of communication free partitions of statement-iterations of nested loops and data accessed by these statement-iterations. Communication-free hyperplane partitions of disjoint subsets of data and statement-iterations are considered. This approach is more possible than existing methods in nding the data and program distribution patterns that can cause the process...

متن کامل

Affine-by-Statement Transformations of Imperfectly Nested Loops

A majority of loop restructuring techniques developed so far assume that loops are perfectly nested. The unimodular approach unifies three individual transformations – loop interchange, skewing and reversal – but is still limited to perfect loop nests. This paper outlines a framework that enables the use of unimodular transformations to restructure imperfect loop nests. The concepts previously ...

متن کامل

ذخیره در منابع من


  با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید

عنوان ژورنال:

دوره   شماره 

صفحات  -

تاریخ انتشار 1999