Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                
Skip to main content
In this paper we present MCoreSim, an open-source simulation framework for massively parallel and many-core computing systems based on OMNeT++. The simulator supports tile-based architectures with distributed memory and mesh-based... more
In this paper we present MCoreSim, an open-source simulation framework for massively parallel and many-core computing systems based on OMNeT++. The simulator supports tile-based architectures with distributed memory and mesh-based interconnects. Its primary purpose is to allow for investigations on the impact of the heterogeneous in-chip communication latencies, as arising due to the network-on-a-chip structure of future and emerging many-core processors, on the performance of the hosted applications. We plan to use MCoreSim to study the variety of possible choices in realizing a suitable software stack for these systems, especially in terms of the choices at the kernel design level.
ABSTRACT short paper / presentation
Research Interests:
In this paper we present an arbitrator implementation for enabling multiprocessor synchronization and resource sharing primitives with guaranteed maximum latency in real-time operating systems (RTOS). The arbitrator has been developed for... more
In this paper we present an arbitrator implementation for enabling multiprocessor synchronization and resource sharing primitives with guaranteed maximum latency in real-time operating systems (RTOS). The arbitrator has been developed for use in conjunction with the NIOS II softcore in the Altera family of system on programmable chips.
We propose an adaptive scheduling technique to schedule highly dynamic multimedia tasks on a CPU. We use a combination of two techniques: the first one is a feedback mechanism to track the resource requirements of the tasks based on... more
We propose an adaptive scheduling technique to schedule highly dynamic multimedia tasks on a CPU. We use a combination of two techniques: the first one is a feedback mechanism to track the resource requirements of the tasks based on “local” observations. The second one is a mechanism that operates with a “global” visibility, reclaiming unused bandwidth. The combination proves very effective: resource reclaiming increases the robustness of the feedback, while the identification of the correct bandwidth made by the feedback increases the effectiveness of the reclamation. We offer both theoretical results and an extensive experimental validation of the approach.
Hardware platforms for real-time embedded systems are evolving towards heterogeneous architectures comprising different types of processing cores and dedicated hardware accelerators, which can be implemented on silicon or dynamically... more
Hardware platforms for real-time embedded systems are evolving towards heterogeneous architectures comprising different types of processing cores and dedicated hardware accelerators, which can be implemented on silicon or dynamically deployed on FPGA fabric. Such accelerators typically access a shared memory to exchange a significant amount of data with other processing elements. Existing COTS solutions focus on maximizing the overall throughput of the system, rather than guaranteeing the timing constraints of individual hardware accelerators. This paper presents the AXI budgeting unit (ABU), a hardware-based solution to implement a bandwidth reservation mechanism on top of the AMBA AXI standard infrastructure for hardware accelerators deployed on FPGAs. An accurate and tractable model, as well as the corresponding analysis, are also proposed to bound the response time of hardware accelerators in the presence of ABUs, in order to verify whether they can complete before their deadlin...
The problem of reducing energy consumption is becoming very important in the design of embedded real-time systems. Many of these systems, in fact, are powered by rechargeable batteries, and the goal is to extend, as much as it is... more
The problem of reducing energy consumption is becoming very important in the design of embedded real-time systems. Many of these systems, in fact, are powered by rechargeable batteries, and the goal is to extend, as much as it is possible, the autonomy of the system. To reduce energy consumption, one possible approach is to selectively slow down the processor frequency. In this paper we propose a modification of the Linux kernel to schedule aperiodic tasks in a soft realtime environment. The proposed solution consists in a new scheduling strategy based on the Resource Reservation Framework [9], which introduces very little modification to the Linux API. Our scheduler is based on Algorithm GRUB (Greedy Reclamation of Unused Bandwidth), presented by Baruah and Lipari [5]. After presenting the algorithm, we describe its implementation on a Intrinsyc CerfCube 250, which uses a Intel PXA250 processor. We show with an example of multimedia application that, by using our approach, we save ...
Real-time systems are traditionally classified into hard real-time and soft real-time: in the first category we have safety critical real-time systems where missing a deadline can have catastrophic consequences, whereas in the second... more
Real-time systems are traditionally classified into hard real-time and soft real-time: in the first category we have safety critical real-time systems where missing a deadline can have catastrophic consequences, whereas in the second class we find systems or which we need to optimise the Quality of service provided to the user. However, the frontier between these two classes is thinner than one may think, and many systems that were considered as hard real-time in the past should now be reconsidered under a different light. In this paper we shall first recall the fundamental notion of time-predictability and criticality, in order to understand where the real-time deadlines that we use in our theoretical models come from. We shall then introduce the model of a soft real-time system and present one popular method for scheduling hard and soft real-time tasks, the resource reservation framework. Finally, we shall show how resource reservation techniques can be successfully applied to the...
In this paper, we address the problem of schedulingbased on their deadlines. Bounded blocking on exclusiveresources is achieved by means of a dynamicresource access protocol which also prevents deadlocksand chained blocking. A tunable... more
In this paper, we address the problem of schedulingbased on their deadlines. Bounded blocking on exclusiveresources is achieved by means of a dynamicresource access protocol which also prevents deadlocksand chained blocking. A tunable servicing technique isused to improve aperiodic responsiveness in the presenceof resource constraints. The schedulability analysisis also extended to the case in which aperiodic deadlinesvary at runtime. The results achieved in this papercan also be used for developing adaptive real- ...
The constant bandwidth server (CBS) framework can be implemented on a preemptive uniprocessor platform to make full use of the computing capacity of the platform; when implemented upon a preemptive multiprocessor platform, however, it is... more
The constant bandwidth server (CBS) framework can be implemented on a preemptive uniprocessor platform to make full use of the computing capacity of the platform; when implemented upon a preemptive multiprocessor platform, however, it is known that the schedulable utilization is strictly less than the capacity of the platform. The issue of using the excess processing capacity is addressed here,
Resource reservations have proven an effective technique to support hard and soft real-time appli-cations in open systems, and some implementations for Linux have already been proposed in the past. However, such implementations generally... more
Resource reservations have proven an effective technique to support hard and soft real-time appli-cations in open systems, and some implementations for Linux have already been proposed in the past. However, such implementations generally focus on providing guarantees to real-time applications, dis-regarding the performance of non real-time activities. In this paper, the problems encountered using a reservation-based scheduling algorithm in Linux (the Constant Bandwidth Server) are described, showing why the original algorithm is not suitable for scheduling non real-time activities. Then, the properties required for properly scheduling non real-time tasks are described, and a novel algorithm (called HGRUB) is analysed, showing how it effectively addresses the presented issues. The performance of HGRUB are then evaluated (by using our implementation in Linux) and compared with the performance of traditional reservation systems.
Research Interests:
Hard real-time systems are very predictable, but not sufficiently flexible to adapt to dynamic situations. They are built under pessimistic assumptions to cope with worst-case scenarios, so they often waste resources. Soft real-time... more
Hard real-time systems are very predictable, but not sufficiently flexible to adapt to dynamic situations. They are built under pessimistic assumptions to cope with worst-case scenarios, so they often waste resources. Soft real-time systems are built to reduce resource consumption, tolerate overloads and adapt to system changes. They are also more suited to novel applications of real-time technology, such as multimedia systems, monitoring apparatuses, telecommunication networks, mobile robotics, virtual reality, and interactive ...
ABSTRACT
... Mangesh Chitnis, Claudio Salvadori, Matteo Petracca, Giuseppe Lipari ReTiS Lab Scuola Superiore Sant'Anna, Pisa, Italy name.surname@sssup.it ... This information can help in estimating the traffic flow entering in (or... more
... Mangesh Chitnis, Claudio Salvadori, Matteo Petracca, Giuseppe Lipari ReTiS Lab Scuola Superiore Sant'Anna, Pisa, Italy name.surname@sssup.it ... This information can help in estimating the traffic flow entering in (or exit-ing from) a city area, and take appropriate actions to pre ...
Research Interests:
Research Interests:
Increasing interest for real-time support in general purpose operating systems has driven a lot of development efforts inside the Linux kernel community. Thus, a Linux based system may be a suitable platform to run heterogeneous real-time... more
Increasing interest for real-time support in general purpose operating systems has driven a lot of development efforts inside the Linux kernel community. Thus, a Linux based system may be a suitable platform to run heterogeneous real-time and non real-time, periodic and ...
Luis Almeida, University of Aveiro, Portugal James Anderson, University of North Carolina, USA Sanjoy Baruah, University of North Carolina, USA Lucia Lo Bello, University of Catania, Italy Azer Bestavros, Boston University, USA Riccardo... more
Luis Almeida, University of Aveiro, Portugal James Anderson, University of North Carolina, USA Sanjoy Baruah, University of North Carolina, USA Lucia Lo Bello, University of Catania, Italy Azer Bestavros, Boston University, USA Riccardo Bettati, Texas A&M University, USA Giorgio Buttazzo, University of Pavia, Italy Marco Caccamo, University of Illinois at Urbana Champaign, USA Albert MK Cheng, University of Houston, USA Gerhard Fohler, Malardalen University, Sweden Steve Goddard, University of Nebraska-Lincoln, USA Hans Hansson, ...

And 111 more