Allocation removal by partial evaluation in a tracing JIT

Bolz, Carl Friedrich; Cuni, Antonio; FijaBkowski, Maciej; Leuschel, Michaël; Pedroni, Samuele; Rigo, Armin

doi:10.1145/1929501.1929508

Cited by 42 publications

(28 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The PyPy benchmarks are used to measure the performance of PyPy and are composed of a series of microbenchmarks and larger programs. 5 The benchmarks were taken from the PyPy benchmarks repository using revision ff7b35837d0f. 6 The benchmarks were run on a version of PyPy based on revision 0b77afaafdd0 and patched to collect additional data about guards in the machine code backends.…”

Section: Discussionmentioning

confidence: 99%

See 1 more Smart Citation

The efficient handling of guards in the design of RPython's tracing JIT

Schneider

Bolz

2012

Proceedings of the Sixth ACM Workshop on Virtual Machines and Intermediate Languages

Self Cite

View full text Add to dashboard Cite

Tracing just-in-time (JIT) compilers record linear control flow paths, inserting operations called guards at points of possible divergence. These operations occur frequently in generated traces and therefore it is important to design and implement them carefully to find the right trade-off between deoptimization, memory overhead, and (partly) execution speed. In this paper, we perform an empirical analysis of runtime properties of guards. This is used to guide the design of guards in the RPython tracing JIT.

show abstract

Section: Discussionmentioning

confidence: 99%

“…The other important point of interaction between resume data and the optimizer is RPython's allocation removal optimization [5]. This optimization discovers allocations in the trace that create objects that do not survive long.…”

Section: Interaction With Optimizationmentioning

confidence: 99%

The efficient handling of guards in the design of RPython's tracing JIT

Schneider

Bolz

2012

Proceedings of the Sixth ACM Workshop on Virtual Machines and Intermediate Languages

Self Cite

View full text Add to dashboard Cite

show abstract

“…In Jit compilation partial evaluation has the potential to both reduce compilation time and increase code quality. This has been demonstrated, for example, in the PyPy Python Jit compiler [6].…”

Section: Related Workmentioning

confidence: 99%

Early partial evaluation in a JIT-compiled, retargetable instruction set simulator generated from a high-level architecture description

Wagstaff

Gould

Franke

et al. 2013

Proceedings of the 50th Annual Design Automation Conference

View full text Add to dashboard Cite

Modern processor design tools integrate in their workflows generators for instruction set simulators (Iss) from architecture descriptions. Whilst these generated simulators are useful for design evaluation and software development, they suffer from poor performance. We present an ultra-fast Jitcompiled Iss generated from an ArchC description. We also introduce a novel partial evaluation optimisation, which further improves Jit compilation time and code quality. This results in a simulation rate of 510Mips for an Arm target across 45 Eembc and Spec benchmarks. On average, our Iss is 1.7 times faster than Simit-Arm, one of the fastest Iss generated from an architecture description.

show abstract

“…Since both integers and floating-point numbers are boxed, unlike in typical Scheme implementations, many of these allocations must be eliminated to obtain high performance. Fortunately, RPython's optimizer is able to see and remove allocations that do not live beyond the scope of a trace (Bolz et al 2011). …”

Section: Allocation Removalmentioning

confidence: 99%

Pycket: a tracing JIT for a functional language

Bauman

Bolz

Hirschfeld³

et al. 2015

SIGPLAN Not.

Self Cite

View full text Add to dashboard Cite

We present Pycket, a high-performance tracing JIT compiler for Racket. Pycket supports a wide variety of the sophisticated features in Racket such as contracts, continuations, classes, structures, dynamic binding, and more. On average, over a standard suite of benchmarks, Pycket outperforms existing compilers, both Racket's JIT and other highly-optimizing Scheme compilers. Further, Pycket provides much better performance for proxies than existing systems, dramatically reducing the overhead of contracts and gradual typing. We validate this claim with performance evaluation on multiple existing benchmark suites.The Pycket implementation is of independent interest as an application of the RPython meta-tracing framework (originally created for PyPy), which automatically generates tracing JIT compilers from interpreters. Prior work on meta-tracing focuses on bytecode interpreters, whereas Pycket is a high-level interpreter based on the CEK abstract machine and operates directly on abstract syntax trees. Pycket supports proper tail calls and first-class continuations. In the setting of a functional language, where recursion and higher-order functions are more prevalent than explicit loops, the most significant performance challenge for a tracing JIT is identifying which control flows constitute a loop-we discuss two strategies for identifying loops and measure their impact.

show abstract

Allocation removal by partial evaluation in a tracing JIT

Cited by 42 publications

References 22 publications

The efficient handling of guards in the design of RPython's tracing JIT

The efficient handling of guards in the design of RPython's tracing JIT

Early partial evaluation in a JIT-compiled, retargetable instruction set simulator generated from a high-level architecture description

Pycket: a tracing JIT for a functional language

Contact Info

Product

Resources

About