Real-World Loops are Easy to Predict: A Case Study - Inria - Institut national de recherche en sciences et technologies du numérique Accéder directement au contenu
Communication Dans Un Congrès Année : 2014

Real-World Loops are Easy to Predict: A Case Study

Résumé

In this paper we study the relevance of fast and simple solutions to compute approximations of the number of iterations of loops (loop trip count) of imperative real-world programs. The context of this work is the use of these approximations in compiler optimizations: most of the time, the optimizations yield greater benefits for large trip counts, and are either innocuous or detrimental for small ones. In this particular work, we argue that, although predicting exactly the trip count of a loop is undecidable, most of the time, there is no need to use computationally expensive state-of-the-art methods to compute (an approximation of) it. We support our position with an actual case study. We show that a fast predictor can be used to speedup the JavaScript JIT compiler of Firefox - one of the most well-engineered runtime environments in use today. We have accurately predicted over 85% of all the interval loops found in typical JavaScript benchmarks, and in millions of lines of C code. Furthermore, we have been able to speedup several JavaScript programs by over 5%, reaching 24% of improvement in one benchmark.
Fichier principal
Vignette du fichier
wst_realloops_author_version.pdf (515.5 Ko) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)
Loading...

Dates et versions

hal-01006208 , version 1 (09-03-2015)

Licence

Paternité - Pas d'utilisation commerciale

Identifiants

  • HAL Id : hal-01006208 , version 1

Citer

Raphael Ernani Rodrigues, Péricles Alves, Fernando Pereira, Laure Gonnord. Real-World Loops are Easy to Predict: A Case Study. Workshop on Software Termination (WST'14), Jul 2014, Vienne, Austria. ⟨hal-01006208⟩
227 Consultations
481 Téléchargements

Partager

Gmail Facebook X LinkedIn More