Logo des Repositoriums

Auflistung nach Autor:in "Spinczyk, Olaf"

1 - 10 von 12
  • Textdokument
    Aspektorientierung und Programmfamilien im Betriebssystembau
    (Ausgezeichnete Informatikdissertationen 2002, 2003) Spinczyk, Olaf
    In der hier zusammengefassten Dissertation geht es um die Analyse, den Entwurf und die Implementierung von maßgeschneiderter Systemsoftware, wie sie beispielsweise im Bereich kleinster eingebetteter Systeme benötigt wird, um Hardware-Anforderungen zu reduzieren und damit Kosten zu sparen. Dabei wurde der Gedanke der Betriebssystemfamilie aufgegriffen und mit dem modernen Ansatz der aspektorientierten Programmierung (AOP) kombiniert. Die so entstandenen Methoden, Werkzeuge und Fallstudien zeigen neue Perspektiven für diese Klasse von Betriebssystemen auf und lassen sich auch auf andere Bereiche sinnvoll übertragen. Die vollständige Fassung der Arbeit ist elektronisch über die Universitätbibliothek Magdeburg (http://diglib.uni-magdeburg.de/) verfügbar .
  • Konferenzbeitrag
    Continuous performance analysis of fault-tolerant virtual machines
    (INFORMATIK 2012, 2012) Jablkowski, Boguslaw; Spinczyk, Olaf
    Virtual machine technology has been successfully applied for the construction of fault-tolerant computing systems. For example, vmware Fault Tolerance and Xen Remus support transparent failover of VMs running on different physical machines in a local area network. However, high availability alone is in many application domains not sufficient. Especially in the context of Cyber-Physical Systems, which interact with the physical environment, realtime constraints have to be fulfilled in order to avoid damage. Therefore, we are working on the combination of VM-based fault tolerance with a performance analysis technique, namely the modular performance analysis with real-time calculus. Such enhanced system would at any time be aware of its own performance and could use this information for smarter reconfiguration decisions in case of faults. This paper will sketch the underlying model, the envisioned system architecture, and discuss beneficial application scenarios.
  • Konferenzbeitrag
    Dynamic aspect weaver family for family-based adaptable systems
    (NODe 2005 – GSEM 2005, 2005) Gilani, Wasif; Spinczyk, Olaf
  • Konferenzbeitrag
    Exploiting Modular Language Extensions in Legacy C Code: An Automotive Case Study
    (Modellierung 2018, 2018) Grosche, Andreas; Igel, Burkhard; Spinczyk, Olaf
    Model-driven software development using language workbenches like JetBrains MPS provide many advantages compared to traditional software development. Base languages can be incrementally extended to increase the abstractness up to domain-specific languages (DSLs). Changes can be performed more efficiently in problem-oriented language extensions or DSLs, than in a base language. In addition, formal analysis can be performed on abstract models. To benefit from the model-driven approach, non-model-based legacy code has to be reusable and transformable to language extensions and DSLs. For the development of embedded systems, mbeddr provides a C99-like base language and extensions for MPS, such as mathematical symbols and state machines. This paper presents a case study that shows how many legacy C code fragments of three automotive series projects could be replaced by mbeddr language extensions. Furthermore, a proof of concept shows the feasibility of fraction and foreach loop refactorings. This work is a first approach for future language extension refactorings.
  • Konferenzbeitrag
    Fail*: towards a versatile fault-injection experiment framework
    (ARCS 2012 Workshops, 2012) Schirmeier, Horst; Hoffmann, Martin; Kapitza, Rüdiger; Lohmann, Daniel; Spinczyk, Olaf
    Many years of research on dependable, fault-tolerant software systems yielded many tool implementations for vulnerability analysis and experimental validation of resilience measures. We identify two disjoint classes of fault-injection (FI) experiment tools in the field, and argue that both are plagued by inherent deficiencies, such as insufficient target state access, little or no means to switch to another target system, and non-reusable experiment code. In this article, we present a novel design approach for a FI infrastructure that aims at combining the strengths of both classes. Our FAIL* experiment framework provides carefully-chosen abstractions simplifying both the implementation of different simulator/hardware target backends and the reuse of experiment code, while retaining the ability for deep target-state access for specialized FI experiments. An exemplary report on first experiences with a prototype implementation based on existing X86 and ARM simulators demonstrates the tool's versatility.
  • Textdokument
    First Things First: A Discussion of Modelling Approaches for Disruptive Memory Technologies
    (Tagungsband des FG-BS Herbsttreffens 2021, 2021) Müller, Michael; Kessener, Daniel; Spinczyk, Olaf
    Disruptive memory technologies (DMT) are dramatically changing the architecture of modern computer systems and affect important design decisions throughout the whole software stack. For their success it is crucial that developers of system software and applications find ways to fully exploit the potential of the novel hardware. Having appropriate DMT hardware models is the key to successful optimization in the world of system software and application development. Therefore, this paper introduces four relevant new DMTs and corresponding performance analyzes as well as modelling approaches. We conclude with the insight that there is a lack of system-wide models that are practically applicable by system software designers for proper optimization and, thus, an important domain for future research.
  • Konferenzbeitrag
    LavA: model-driven development of configurable MPSoC hardware structures for robots
    (INFORMATIK 2011 – Informatik schafft Communities, 2011) Meier, Matthias; Spinczyk, Olaf
    Deploying multicore or multiprocessor hardware for robotics applications is highly beneficial. Parallel hardware structures can be utilized to improve the performance, real-time characteristics, or fault tolerance. Special accelerator components can boost the performance and energy efficiency even more. However, the optimal hardware design is application-specific. This is a dilemma especially for modular general purpose robots, because the application scenario is unknown at design time. Therefore, more and more robots are being equipped with configurable hardware such as FPGAs. In this paper we describe the LavA framework, which facilitates the development of application-specific MPSoC hardware structures. Our prototype can interact with Lego Mindstorms NXT sensors and actuators. A DSL is used to describe the hardware structure. Syntactic and semantic checks are performed on the high-level hardware model and a resource model quickly provides an estimate of the required FPGA resources. The hardware synthesis itself is fully automated and requires no special know how. Optionally, the framework can even statically analyze the C/C++ application code. Based on the hardware access patterns found in the code, a suitable hardware description is derived automatically.
  • Zeitschriftenartikel
    mxkernel: A Novel System Software Stack for Data Processing on Modern Hardware
    (Datenbank-Spektrum: Vol. 20, No. 3, 2020) Mühlig, Jan; Müller, Michael; Spinczyk, Olaf; Teubner, Jens
    Emerging hardware platforms are characterized by large degrees of parallelism, complex memory hierarchies, and increasing hardware heterogeneity. Their theoretical peak data processing performance can only be unleashed if the different pieces of systems software collaborate much more closely and if their traditional dependencies and interfaces are redesigned. We have developed the key concepts and a prototype implementation of a novel system software stack named mxkernel . For MxKernel, efficient large scale data processing capabilities are a primary design goal. To achieve this, heterogeneity and parallelism become first-class citizens and deep memory hierarchies are considered from the very beginning. Instead of a classical “thread” model, mxkernel provides a simpler control flow abstraction: mxtasks model closed units of work, for which mxkernel will guarantee the required execution semantics, such exclusive access to a specific object in memory. They can be a very elegant abstraction also for heterogeneity and resource sharing. Furthermore, mxtasks are annotated with metadata, such as code variants (to support heterogeneity), memory access behavior (to improve cache efficiency and support memory hierarchies), or dependencies between mxtasks (to improve scheduling and avoid synchronization cost). With precisely the required metadata available, mxkernel can provide a lightweight, yet highly efficient form of resource management, even across applications, operating system, and database. Based on the mxkernel prototype we present preliminary results from this ambitious undertaking. We argue that threads are an ill-suited control flow abstraction for our modern computer architectures and that a task-based execution model is to be favored.
  • Textdokument
    Performance is not Boolean: Supporting Scalar Configuration Variables in NFP Models
    (Tagungsband des FG-BS Frühjahrstreffens 2022, 2022) Friesel, Daniel; Spinczyk, Olaf
    Non-functional properties (NFPs) such as memory requirements, timing, or energy consumption are important characteristics of embedded software systems and software product lines (SPLs) in general. Both during system design and at runtime, the goal is to optimize resource utilization (and, thus, NFPs) by appropriate system configuration or orchestration. NFP models, learned from benchmarks of various SPL configurations, allow for the prediction of these properties, thus enabling NFP-aware software configuration and runtime decisions. However, many existing approaches for automated learning of NFP models limit their scope to boolean variables. We argue that this is no longer sufficient: NFP models must accommodate scalar variables to achieve suitable accuracy when faced with today’s highly configurable software systems and variable workloads. To this end, we evaluate four regression tree-based NFP modeling approaches on eight use cases, and examine model complexity and model accuracy. We find that models with support for scalar variables achieve up to three times lower mean model error when predicting configurations that were not part of the training set. At the same time, the complexity of scalar and boolean-only models is nearly the same; only benchmarking becomes more time-intensive due to the need to explore scalar variables. We conclude that scalar-enabled models provide increased accuracy almost free of charge, and recommend using them when generating NFP models for embedded systems and workloads with scalar configuration variables.
  • Konferenzbeitrag
    Protecting the dynamic dispatch in C++ by dependability aspects
    (INFORMATIK 2012, 2012) Borchert, Christoph; Schirmeier, Horst; Spinczyk, Olaf
    Computer systems, especially devices with highly-miniaturized feature sizes, are unreliable. Data memory is susceptible to a number of physical effects that cause faults, which can be observed as spontaneous bit flips. Although in many application scenarios corrupt data is harmless (“almost” correct result often suffices), control-flow transitions are very sensitive to faults. Indirect jumps, such as the dynamic dispatch of virtual functions in C++, often crash the system in case of a single bit flip. This paper describes a suitable software-based fault-tolerance mechanism, which can be applied to arbitrary C++ software by source-to-source compilation. The overall cost for this mechanism is below 10 % for both runtime and memory overhead. Our evaluation results show that this approach eliminates 67.1 % of all irregular program terminations in a case study using an embedded weather-station software, whose entire data memory is corrupted by single-bit flips.