A view of coroutines

Generators offer an elegant way to express iterators. However, performance has always been their Achilles heel and has prevented widespread adoption. We present techniques to efficiently implement and optimize generators.We have implemented our optimizations in ZipPy, a modern, light-weight AST interpreter based Python 3 implementation targeting the Java virtual machine. Our implementation builds on a framework that optimizes AST interpreters using just-in-time compilation. In such a system, it is crucial that AST optimizations do not prevent subsequent optimizations. Our system was carefully designed to avoid this problem. We report an average speedup of 3.58× for generatorbound programs. As a result, using generators no longer has downsides and programmers are free to enjoy their upsides.

show abstract

“…A generator is a more restricted variation of a coroutine [12,16]. It encompasses two control abstractions: suspend and resume.…”

Section: Generators In Pythonmentioning

confidence: 99%

Accelerating iterators in optimizing AST interpreters

Zhang

Larsen

Brunthaler

et al. 2014

Proceedings of the 2014 ACM International Conference on Object Oriented Programming Systems Languages &Amp; Applications

View full text Add to dashboard Cite

show abstract

“…However, recall that the semantics of the Chapel iterators, differing from normal functions, require that once program execution reaches the last statement in the loop body, control resumes inside the iterator body on the statement immediately following the yield statement for each iterator. Zippered iteration would be implemented naturally using coroutines [8], which allow for execution to begin anywhere inside of a function, unlike functions in most current languages. However, without coroutines, zipper-product iteration may still be implemented using techniques we describe in Section 4.…”

Section: Invoking Multiple Iteratorsmentioning

confidence: 99%

“…A coroutine [8] is a routine that yields or produces values for another routine to consume. Unlike functions in most modern languages, coroutines have multiple points of entry.…”

Section: Related Workmentioning

confidence: 99%

Iterators in Chapel

Joyner¹,

Chamberlain²,

Deitz³

2006

Proceedings 20th IEEE International Parallel &Amp; Distributed Processing Symposium

View full text Add to dashboard Cite

A long-held tenet of software engineering is that algorithms and data structures should be specified orthogonally in order to minimize the impact that changes to one will have on the other. Unfortunately, this principle is often not well-supported in scientific and parallel codes due to the lack of abstractions for factoring iteration away from computation in traditional scientific languages. The result is a fragile situation in which complex loop nests are used to express parallelism and maximize performance, yet must be maintained individually as the algorithm and data structures evolve. In this paper, we introduce the iterator concept in the Chapel parallel programming language, designed to address this problem and provide a means for factoring iteration away from computation. The paper illustrates iterators using several examples, compares our approach with those taken in other languages, and describes our implementation in the Chapel compiler.

show abstract

“…When the application halts (at SUSPEND [3]), the CURRENT execution environment will be the LOCKED environment (environment SIX).…”

Section: An F~mentioning

confidence: 99%

“…UNAME÷GET USER NAME ~EXTERNAL PROC [2] DIO~I 0 'T~EMO~ 'PAYROLL ABORT UNAME' [3] 'INTERRUPTtOEMOD 'PAYROL~ TERMINATE' [4] [4] ADD TO ERROR LOG NAME,vOENUM,PWD RLOG B Figure 6: Fragment of example secure environment.…”

Section: The Implementer Can Then Determine What the Error Was Withomentioning

confidence: 99%

Secure application environments in APL\3000

Marcum

1979

SIGAPL APL Quote Quad

View full text Add to dashboard Cite

Various mechanisms have been proposed for implementing secure applications (that is, a function or package of functions over which the author can maintain total control during execution, and whose possibly sensitive information can be protected). This paper describes another set of tools, a “Spaghetti Stack” with “Exception Handling,” for implementing secure applications. (A Spaghetti Stack is a means for capturing a State Indicator execution point, retaining the state, and executing from it later.)

show abstract

A view of coroutines

Abstract: The coroutine mechanism is explained as a simplified implementation of a special case in parallel processing.

Cited by 17 publications

References 1 publication

Accelerating iterators in optimizing AST interpreters

Accelerating iterators in optimizing AST interpreters

Iterators in Chapel

Secure application environments in APL\3000

Contact Info

Product

Resources

About