The Polyhedral Model of Nonlinear Loops

Sukumaran-Rajam, Aravind; Clauss, Philippe

doi:10.1145/2838734

Cited by 13 publications

(22 citation statements)

References 29 publications

(27 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…If it is greater than 0.9, future memory accesses are predicted to happen inside a tube, otherwise inside a range. This correlation criterion is derived from experimental evaluation . The tube consists of the regression hyperplane, a tube width and a predicted alignment.…”

Section: When Polyhedra Meet Speculationmentioning

confidence: 99%

“…While executing, the prediction model is continually verified by comparing the actual reached values against their linear predictions. At the end of the chunk, if no misprediction was detected, APOLLO performs a new backup 6 for the next chunk and executes this chunk using the same optimized version 7 . If a misprediction is detected, memory is restored to cancel the execution of the current chunk 8 .…”

Section: Of 16mentioning

confidence: 99%

“…This correlation criterion is derived from experimental evaluation. 6,11 The tube consists of the regression hyperplane, a tube width and a predicted alignment. The tube width is the maximum observed deviation from the regression hyperplane, rounded to the next bigger multiple of the word size.…”

Section: Memory Accessesmentioning

confidence: 99%

“…Traditional TLS systems implement only a straightforward loop parallelization strategy consisting of slicing the target loop into consecutive parallel threads. To combine the best of both the polyhedral model and the TLS approach, new TLS frameworks implementing a speculative and dynamic adaptation of the polyhedral model have been proposed . In these frameworks, parallelizing and optimizing transformations are performed for loops exhibiting a polyhedral‐compliant behavior at runtime.…”

Section: Introductionmentioning

confidence: 99%

“…The resulting assembled code is then further optimized and compiled using the LLVM just‐in‐time (JIT) compiler. Our contribution has been implemented in the speculative parallelization framework APOLLO . We show on a set of benchmark programs that this code generation technique enables significant parallel speedups on loops that cannot be handled at compile‐time.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Full runtime polyhedral optimizing loop transformations with the generation, instantiation, and scheduling of code‐bones

Caamaño

Selva

Clauss

et al. 2017

Concurrency and Computation

Self Cite

View full text Add to dashboard Cite

Summary In this paper, we present a new runtime code generation technique for speculative loop optimization and parallelization. The main benefit of this technique, compared to previous approaches, is to enable advanced optimizing loop transformations at runtime with an acceptable time overhead. The loop transformations that may be applied are those handled by the polyhedral model. The proposed code generation 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 approach has been implemented in the speculative loop parallelizing framework APOLLO.

show abstract

Section: When Polyhedra Meet Speculationmentioning

confidence: 99%

Section: Of 16mentioning

confidence: 99%

Section: Memory Accessesmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Full runtime polyhedral optimizing loop transformations with the generation, instantiation, and scheduling of code‐bones

Caamaño

Selva

Clauss

et al. 2017

Concurrency and Computation

Self Cite

View full text Add to dashboard Cite

show abstract

Rec2Poly: Converting Recursions to Polyhedral Optimized Loops Using an Inspector-Executor Strategy

Kobeissi

Ketterlin

Clauss

2020

Lecture Notes in Computer Science

Self Cite

View full text Add to dashboard Cite

In this paper, we propose Rec2Poly, a framework which detects automatically if recursive programs may be transformed into affine loops that are compliant with the polyhedral model. If successful, the replacing loops can then take advantage of advanced loop optimizing and parallelizing transformations as tiling or skewing. Rec2Poly is made of two main phases: an offline profiling phase and an inspector-executor phase. In the profiling phase, the original recursive program, which has been instrumented, is run. Whenever possible, the trace of collected information is used to build equivalent affine loops from the runtime behavior. Then, an inspector-executor program is automatically generated, where the inspector is made of a light version of the original recursive program, whose aim is reduced to the generation and verification of the information which is essential to ensure the correctness of the equivalent affine loop program. The collected information is mainly related to the touched memory addresses and the control flow of the socalled "impacting" basic blocks of instructions. Moreover, in order to exhibit the lowest possible time-overhead, the inspector is implemented as a parallel process where several memory buffers of information are verified simultaneously. Finally, the executor is made of the equivalent affine loops that have been optimized and parallelized.

show abstract