Skip to Main content Skip to Navigation
Conference papers

Code Bones: Fast and Flexible Code Generation for Dynamic and Speculative Polyhedral Optimization

Juan Manuel Martinez Caamaño 1, 2, 3 Willy Wolff 1, 2, 3 Philippe Clauss 1, 2, 3
1 CAMUS - Compilation pour les Architectures MUlti-coeurS
Inria Nancy - Grand Est, ICube - Laboratoire des sciences de l'ingénieur, de l'informatique et de l'imagerie
Abstract : In this paper, we present a new runtime code generation technique for speculative loop optimization and parallelization, that allows to generate on-the-fly codes resulting from any polyhedral optimizing transformation of loop nests, such as tiling, skewing, fission, fusion or interchange, without introducing a penalizing time overhead. The proposed strategy is based on the generation of code bones at compile-time, which are parametrized code snippets either dedicated to speculation management or to computations of the original target program. These code bones are then instantiated and assembled at runtime to constitute the speculatively-optimized code, as soon as an optimizing polyhedral transformation has been determined. Their granularity threshold is sufficient to apply any polyhedral transformation, while still enabling fast runtime code generation. This strategy has been implemented in the speculative loop parallelizing framework Apollo.
Document type :
Conference papers
Complete list of metadata

Cited literature [18 references]  Display  Hide  Download
Contributor : Philippe Clauss Connect in order to contact the contributor
Submitted on : Friday, October 7, 2016 - 12:43:55 PM
Last modification on : Thursday, December 2, 2021 - 3:17:29 AM
Long-term archiving on: : Friday, February 3, 2017 - 9:22:11 PM


Files produced by the author(s)



Juan Manuel Martinez Caamaño, Willy Wolff, Philippe Clauss. Code Bones: Fast and Flexible Code Generation for Dynamic and Speculative Polyhedral Optimization. Euro-Par 2016, Aug 2016, Grenoble, France. pp.12, ⟨10.1007/978-3-319-43659-3_17⟩. ⟨hal-01377656⟩



Record views


Files downloads