Configuring lowest latency kernel

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F68407700%3A21230%2F15%3A00233902" target="_blank" >RIV/68407700:21230/15:00233902 - isvavai.cz</a>

Výsledek na webu

—

DOI - Digital Object Identifier

—

Jazyk výsledku

angličtina

Název v původním jazyce

Configuring lowest latency kernel

Popis výsledku v původním jazyce

Linux kernel performance is a key parameter for many applications and tuning the kernel for optimal performance can be a challenging task. One reason is that there are many knobs one can use to tweak the performance. One such knob is compile-time con?guration of the Linux kernel and in this paper, we are interested in how this con?guration in?uences system response-time (latency), which is an important parameter for real-time applications. We developed a tool that uses a SAT solver to systematically generate valid kernel con?gurations. For each con?guration (typically, we haveseveral thousands con?gurations) the kernel is compiled, booted and a set of benchmarks is executed. The benchmark results are processed and we try to calculate the contribution of individual con?guration options to the benchmark results. The benchmark used in this paper is the well-known cyclictest utility and we investigate the e?ect of 54 con?guration options to its results. This is still work in progress and t

Název v anglickém jazyce

Configuring lowest latency kernel

Popis výsledku anglicky

Linux kernel performance is a key parameter for many applications and tuning the kernel for optimal performance can be a challenging task. One reason is that there are many knobs one can use to tweak the performance. One such knob is compile-time con?guration of the Linux kernel and in this paper, we are interested in how this con?guration in?uences system response-time (latency), which is an important parameter for real-time applications. We developed a tool that uses a SAT solver to systematically generate valid kernel con?gurations. For each con?guration (typically, we haveseveral thousands con?gurations) the kernel is compiled, booted and a set of benchmarks is executed. The benchmark results are processed and we try to calculate the contribution of individual con?guration options to the benchmark results. The benchmark used in this paper is the well-known cyclictest utility and we investigate the e?ect of 54 con?guration options to its results. This is still work in progress and t

Druh

O - Ostatní výsledky

CEP obor

JC - Počítačový hardware a software

OECD FORD obor

—

Projekt

<a href="/cs/project/EE2.3.30.0034" target="_blank" >EE2.3.30.0034: Podpora zkvalitnění týmů výzkumu a vývoje a rozvoj intersektorální mobility na ČVUT v Praze</a><br>

Návaznosti

P - Projekt vyzkumu a vyvoje financovany z verejnych zdroju (s odkazem do CEP)

Rok uplatnění

2015

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů