M. Hind, M. Burke, P. Carini, and S. Midki. An Empirical Study of Precise Interprocedural Array Analysis. Scienti c Programming, 3(3):255-271, 1994.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....machines. We just summarize the dynamic results here; the full results appear in [50] Most previous evaluations of interprocedural parallelization systems provided static measurements of the number of additional loops parallelized as a result of interpro cedural dependence analysis [56, 59, 77, 114]. The SUIF results were compared with the most recent of these empirical studies, which examines the SPEC89 and PERFECT benchmark suites using the FIDA system [59] When considering only those loops containing calls for this set of 16 programs, the SUIF system is able to parallelize greater than ....

.... static measurements of the number of additional loops parallelized as a result of interpro cedural dependence analysis [56, 59, 77, 114] The SUIF results were compared with the most recent of these empirical studies, which examines the SPEC89 and PERFECT benchmark suites using the FIDA system [59]. When considering only those loops containing calls for this set of 16 programs, the SUIF system is able to parallelize greater than five times more of these loops (a comparison with FIDA is presented in detail in [53] Static loop counts, however, are not good indicators of whether ....

[Article contains additional citation context not shown here]

M. Hind, M. Burke, P. Carini, and S. Midkiff. An empirical study of precise interprocedural array analysis. Scientific Programming, 3(3):255 271, 1994.

.... efficient interprocedural analysis for shared memory machines or compiling for distributed memory machines, but usually not in combination [17] Others have focused on efficiently solving classic data flow problems interprocedurally [11, 32] or precise interprocedural analyses of array accesses [14, 15, 22, 23, 24, 27]. Several recent distributed memory compilation systems employ integer polyhedra instead of RSDs for greater flexibility [3, 4] Few other distributed memory compilation systems have discussed interprocedural issues, especially interprocedural optimization. The CM Fortran compiler utilizes ....

M. Hind, M. Burke, P. Carini, and S. Midkiff. An empirical study of precise interprocedural array analysis. Scientific Programming, 3(3):255--271, 1994.

.... success on linear algebra libraries [2, 4, 6, 7] but the results on larger programs have been much less promising [4] We have compared our results with the Fida system (Full Interprocedural Data Flow Analysis) an interprocedural system that performs precise flow insensitive array analysis [3]. The Fida system was the first to measure how interprocedural analysis on full applications (from the Perfect and Spec89 benchmark suites) affects the number of parallel loops that the system can recognize. In comparing how many loops containing procedure calls are parallelized using the two ....

M. Hind, M. Burke, P. Carini, and S. Midkiff. An empirical study of precise interprocedural array analysis. Scientific Programming, 3(3):255--271, 1994.

....accesses in a region of code; a data dependence analysis is then applied to a small number of summaries. The summaries also provide the representation used by the array data flow analysis, described below. There have been many different designs of summaries that trade off efficiency and precision [13, 14, 18, 24]. We represent a summary of a set of array accesses by a list of systems of linear inequalities: the array indices are equated to affine expressions of outer loop indices and loop invariant values, constrained further by inequalities derived from the loop bounds. The representation of data ....

....arrays (intersect Write and ExposedRead) ffl Recognize array reductions and record reduction operator type Figure 4: Phases of Interprocedural Parallelization Analysis. recently, the Fida system was developed at IBM to obtain more precise array sections through partial inlining of array accesses [14] (see Section 6) Irigoin et al. have developed the PIPS system, an interprocedural analysis system that is part of an environment for parallel programming [16] More recently, PIPS has been extended to incorporate interprocedural array privatization [15, 6] PIPS is most similar to our work, but ....

[Article contains additional citation context not shown here]

M. Hind, M. Burke, P. Carini, and S. Midkiff. An empirical study of precise interprocedural array analysis. Scientific Programming, 3(3):255--271, 1994.

....that the exactness criterion for IF constructs could be deduced from trivial and intuitive definitions of R and R derived from the exact semantics R. choice of the representation are still alive: Time and space complexity versus accuracy is the main issue, but usefulness is widely acknowledged [31, 4, 24, 21, 15]. The need for under approximations appeared only recently [32, 26, 21] mainly for locality analysis, to allow advanced program transformations such as array privatization. But no study of the underlying semantical analysis framework had been made, thus missing its inherent problems due to the ....

Michael Hind, Michael Burke, Paul Carini, and Sam Midkiff. An empirical study of precise interprocedural array analysis. Scientific Programming, 3(3):255--271, May 1994.

....distinct from those that are referenced or modified in other invocations. These techniques were shown to be effective in parallelizing linear algebra libraries. More recently, the Fida system was developed at IBM to obtain more precise array sections through partial inlining of array accesses [10] (see Section 7) Irigoin et al. developed an interprocedural analysis system, called PIPS, that is part of an environment for parallel programming [12] More recently, PIPS has been extended to incorporate interprocedural array privatization [11, 5] PIPS is most similar to our work, but lacks ....

....full results are presented elsewhere [7] but we present a few highlights in this section. We have compared the results of our interpocedural analysis with the Fida system (Full Interprocedural Data Flow Analysis) an interprocedural system that performs precise flow insensitive array analysis [10] (see Section 2) The Fida system was the first to measure how interprocedural analysis on full applications (from the Perfect and Spec89 benchmark suites) affects the number of parallel loops that the system can automatically recognize. We compare how many loops containing procedure calls are ....

M. Hind, M. Burke, P. Carini, and S. Midkiff. An empirical study of precise interprocedural array analysis. Scientific Programming, 3(3):255--271, 1994.

....distinct from those that are referenced or modified in other invocations. These techniques were shown to be effective in parallelizing linear algebra libraries. More recently, the Fida system was developed at IBM to obtain more precise array sections through partial inlining of array accesses [14] (see 1. Flow insensitive pass: ffl Find modified and referenced variables ffl Find scalar reductions 2. Bottom up pass: scalar analysis ffl Find privatizable scalars ffl Summarize symbolic behaviors (side effects) 3. Top down pass: scalar analysis ffl Apply calling context to symbolic ....

....empirical evaluation of the results of the parallelization analysis on a large collection of benchmark programs. Previous evaluations of interprocedural parallelization systems have provided static measurements of the number of additional loops parallelized as a result of interprocedural analysis [13, 14, 18, 24]. We have compared our results with the most recent of these empirical studies, which examines the Spec89 and Perfect benchmark suites [14] When considering only those loops containing calls for this set of 16 programs, the SUIF system is able to parallelize greater than five times more of these ....

[Article contains additional citation context not shown here]

M. Hind, M. Burke, P. Carini, and S. Midkiff. An empirical study of precise interprocedural array analysis. Scientific Programming, 3(3):255--271, 1994.

No context found.

M. Hind, M. Burke, P. Carini, and S. Midki. An Empirical Study of Precise Interprocedural Array Analysis. Scienti c Programming, 3(3):255-271, 1994.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC