Real-time scheduling of parallel tasks in the Linux Kernel

This paper proposes a global multiprocessor scheduling algorithm for the Linux kernel that combines the global EDF scheduler with a priority-aware work-stealing load balancing scheme, enabling parallel real-time tasks to be executed on more than one processor at a given time instant. We state that some priority inversion may actually be acceptable, provided it helps reduce contention, communication, synchronisation and coordination between parallel threads, while still guaranteeing the expected system’s predictability. Experimental results demonstrate the low scheduling overhead of the proposed approach comparatively to an existing real-time deadline-oriented scheduling class for the Linux kernel.

Supporting real-time parallel task models with work-stealing

Dynamic parallel scheduling using work-stealing has gained popularity in academia and industry for its good performance, ease of implementation and theoretical bounds on space and time. Cores treat their own double-ended queues (deques) as a stack, pushing and popping threads from the bottom, but treat the deque of another randomly selected busy core as a queue, stealing threads only from the top, whenever they are idle. However, this standard approach cannot be directly applied to real-time systems, where the importance of parallelising tasks is increasing due to the limitations of multiprocessor scheduling theory regarding parallelism. Using one deque per core is obviously a source of priority inversion since high priority tasks may eventually be enqueued after lower priority tasks, possibly leading to deadline misses as in this case the lower priority tasks are the candidates when a stealing operation occurs. Our proposal is to replace the single non-priority deque of work-stealing with ordered per-processor priority deques of ready threads. The scheduling algorithm starts with a single deque per-core, but unlike traditional work-stealing, the total number of deques in the system may now exceed the number of processors. Instead of stealing randomly, cores steal from the highest priority deque.

Task assignment algorithms for two-type heterogeneous multiprocessors

Consider the problem of assigning real-time tasks on a heterogeneous multiprocessor platform comprising two different types of processors — such a platform is referred to as two-type platform. We present two linearithmic timecomplexity algorithms, SA and SA-P, each providing the follow- ing guarantee. For a given two-type platform and a given task set, if there exists a feasible task-to-processor-type assignment such that tasks can be scheduled to meet deadlines by allowing them to migrate only between processors of the same type, then (i) using SA, it is guaranteed to find such a feasible task-to- processor-type assignment where the same restriction on task migration applies but given a platform in which processors are 1+α/2 times faster and (ii) SA-P succeeds in finding 2 a feasible task-to-processor assignment where tasks are not allowed to migrate between processors but given a platform in which processors are 1+α/times faster, where 0<α≤1. The parameter α is a property of the task set — it is the maximum utilization of any task which is less than or equal to 1.

Online intra-task device scheduling for hard real-time systems

A large part of power dissipation in a system is generated by I/O devices. Increasingly these devices provide power saving mechanisms, inter alia to enhance battery life. While I/O device scheduling has been studied in the past for realtime systems, the use of energy resources by these scheduling algorithms may be improved. These approaches are crafted considering a very large overhead of device transitions. Technology enhancements have allowed the hardware vendors to reduce the device transition overhead and energy consumption. We propose an intra-task device scheduling algorithm for real time systems that allows to shut-down devices while ensuring system schedulability. Our results show an energy gain of up to 90% when compared to the techniques proposed in the state-of-the-art.

Extending fixed task-priority schedulability by interference limitation

While the earliest deadline first algorithm is known to be optimal as a uniprocessor scheduling policy, the implementation comes at a cost in terms of complexity. Fixed taskpriority algorithms on the other hand have lower complexity but higher likelihood of task sets being declared unschedulable, when compared to earliest deadline first (EDF). Various attempts have been undertaken to increase the chances of proving a task set schedulable with similar low complexity. In some cases, this was achieved by modifying applications to limit preemptions, at the cost of flexibility. In this work, we explore several variants of a concept to limit interference by locking down the ready queue at certain instances. The aim is to increase the prospects of schedulability of a given task system, without compromising on complexity or flexibility, when compared to the regular fixed task-priority algorithm. As a final contribution, a new preemption threshold assignment algorithm is provided which is less complex and more straightforward than the previous method available in the literature.

A framework for offloading real-time applications in a distributed environment

This work focuses on highly dynamic distributed systems with Quality of Service (QoS) constraints (most importantly real-time constraints). To that purpose, real-time applications may benefit from code offloading techniques, so that parts of the application can be offloaded and executed, as services, by neighbour nodes, which are willing to cooperate in such computations. These applications explicitly state their QoS requirements, which are translated into resource requirements, in order to evaluate the feasibility of accepting other applications in the system.

A conjecture about provably good task assignment on heterogeneous multiprocessor platforms but with a stronger adversary

Consider the problem of scheduling a set of implicit-deadline sporadic tasks to meet all deadlines on a heterogeneous multiprocessor platform. We use an algorithm proposed in [1] (we refer to it as LP-EE) from state-of-the-art for assigning tasks to heterogeneous multiprocessor platform and (re-)prove its performance guarantee but for a stronger adversary.We conjecture that if a task set can be scheduled to meet deadlines on a heterogeneous multiprocessor platform by an optimal task assignment scheme that allows task migrations then LP-EE meets deadlines as well with no migrations if given processors twice as fast. We illustrate this with an example.

Provably good task assignment on heterogeneous multiprocessor platforms for a restricted case but with a stronger adversary

Consider the problem of scheduling a set of implicit-deadline sporadic tasks to meet all deadlines on a heterogeneous multiprocessor platform. We consider a restricted case where the maximum utilization of any task on any processor in the system is no greater than one. We use an algorithm proposed in [1] (we refer to it as LP-EE) from state-of-the-art for assigning tasks to heterogeneous multiprocessor platform and (re-)prove its performance guarantee for this restricted case but for a stronger adversary. We show that if a task set can be scheduled to meet deadlines on a heterogeneous multiprocessor platform by an optimal task assignment scheme that allows task migrations then LP-EE meets deadlines as well with no migrations if given processors twice as fast.

Pattern operators for grid

The definition and programming of distributed applications has become a major research issue due to the increasing availability of (large scale) distributed platforms and the requirements posed by the economical globalization. However, such a task requires a huge effort due to the complexity of the distributed environments: large amount of users may communicate and share information across different authority domains; moreover, the “execution environment” or “computations” are dynamic since the number of users and the computational infrastructure change in time. Grid environments, in particular, promise to be an answer to deal with such complexity, by providing high performance execution support to large amount of users, and resource sharing across different organizations. Nevertheless, programming in Grid environments is still a difficult task. There is a lack of high level programming paradigms and support tools that may guide the application developer and allow reusability of state-of-the-art solutions. Specifically, the main goal of the work presented in this thesis is to contribute to the simplification of the development cycle of applications for Grid environments by bringing structure and flexibility to three stages of that cycle through a commonmodel. The stages are: the design phase, the execution phase, and the reconfiguration phase. The common model is based on the manipulation of patterns through pattern operators, and the division of both patterns and operators into two categories, namely structural and behavioural. Moreover, both structural and behavioural patterns are first class entities at each of the aforesaid stages. At the design phase, patterns can be manipulated like other first class entities such as components. This allows a more structured way to build applications by reusing and composing state-of-the-art patterns. At the execution phase, patterns are units of execution control: it is possible, for example, to start or stop and to resume the execution of a pattern as a single entity. At the reconfiguration phase, patterns can also be manipulated as single entities with the additional advantage that it is possible to perform a structural reconfiguration while keeping some of the behavioural constraints, and vice-versa. For example, it is possible to replace a behavioural pattern, which was applied to some structural pattern, with another behavioural pattern. In this thesis, besides the proposal of the methodology for distributed application development, as sketched above, a definition of a relevant set of pattern operators was made. The methodology and the expressivity of the pattern operators were assessed through the development of several representative distributed applications. To support this validation, a prototype was designed and implemented, encompassing some relevant patterns and a significant part of the patterns operators defined. This prototype was based in the Triana environment; Triana supports the development and deployment of distributed applications in the Grid through a dataflow-based programming model. Additionally, this thesis also presents the analysis of a mapping of some operators for execution control onto the Distributed Resource Management Application API (DRMAA). This assessment confirmed the suitability of the proposed model, as well as the generality and flexibility of the defined pattern operators

On the implementation of real-time slotbased task-splitting scheduling algorithms for multiprocessor systems

In this paper we discuss challenges and design principles of an implementation of slot-based tasksplitting algorithms into the Linux 2.6.34 version. We show that this kernel version is provided with the required features for implementing such scheduling algorithms. We show that the real behavior of the scheduling algorithm is very close to the theoretical. We run and discuss experiments on 4-core and 24-core machines.

Implementing slot-based task-splitting multiprocessor scheduling

Consider the problem of scheduling a set of sporadic tasks on a multiprocessor system to meet deadlines using a task-splitting scheduling algorithm. Task-splitting (also called semi-partitioning) scheduling algorithms assign most tasks to just one processor but a few tasks are assigned to two or more processors, and they are dispatched in a way that ensures that a task never executes on two or more processors simultaneously. A particular type of task-splitting algorithms, called slot-based task-splitting dispatching, is of particular interest because of its ability to schedule tasks with high processor utilizations. Unfortunately, no slot-based task-splitting algorithm has been implemented in a real operating system so far. In this paper we discuss and propose some modifications to the slot-based task-splitting algorithm driven by implementation concerns, and we report the first implementation of this family of algorithms in a real operating system running Linux kernel version 2.6.34. We have also conducted an extensive range of experiments on a 4-core multicore desktop PC running task-sets with utilizations of up to 88%. The results show that the behavior of our implementation is in line with the theoretical framework behind it.

Practical aspects of slot-based task- splitting dispatching in its schedulability analysis

Consider the problem of scheduling a set of sporadic tasks on a multiprocessor system to meet deadlines using a tasksplitting scheduling algorithm. Task-splitting (also called semipartitioning) scheduling algorithms assign most tasks to just one processor but a few tasks are assigned to two or more processors, and they are dispatched in a way that ensures that a task never executes on two or more processors simultaneously. A certain type of task-splitting algorithms, called slot-based task-splitting, is of particular interest because of its ability to schedule tasks at high processor utilizations. We present a new schedulability analysis for slot-based task-splitting scheduling algorithms that takes the overhead into account and also a new task assignment algorithm.

A 12*(1+|R|/(4m))-speed algorithm for scheduling constrained-deadline sporadic real-time tasks on a multiprocessor comprising m processors where a task may request one of |R| sequentially-reusable shared resources

We present a 12*(1+|R|/(4m))-speed algorithm for scheduling constrained-deadline sporadic real-time tasks on a multiprocessor comprising m processors where a task may request one of |R| sequentially-reusable shared resources.

Implementing multicore real-time scheduling algorithms based on task splitting using ada 2012

Multiprocessors, particularly in the form of multicores, are becoming standard building blocks for executing reliable software. But their use for applications with hard real-time requirements is non-trivial. Well-known realtime scheduling algorithms in the uniprocessor context (Rate-Monotonic [1] or Earliest-Deadline-First [1]) do not perform well on multiprocessors. For this reason the scientific community in the area of real-time systems has produced new algorithms specifically for multiprocessors. In the meanwhile, a proposal [2] exists for extending the Ada language with new basic constructs which can be used for implementing new algorithms for real-time scheduling; the family of task splitting algorithms is one of them which was emphasized in the proposal [2]. Consequently, assessing whether existing task splitting multiprocessor scheduling algorithms can be implemented with these constructs is paramount. In this paper we present a list of state-of-art task-splitting multiprocessor scheduling algorithms and, for each of them, we present detailed Ada code that uses the new constructs.

Optimising and adapting the QoS of a dynamic set of inter-dependent tasks

Due to the growing complexity and adaptability requirements of real-time systems, which often exhibit unrestricted Quality of Service (QoS) inter-dependencies among supported services and user-imposed quality constraints, it is increasingly difficult to optimise the level of service of a dynamic task set within an useful and bounded time. This is even more difficult when intending to benefit from the full potential of an open distributed cooperating environment, where service characteristics are not known beforehand and tasks may be inter-dependent. This paper focuses on optimising a dynamic local set of inter-dependent tasks that can be executed at varying levels of QoS to achieve an efficient resource usage that is constantly adapted to the specific constraints of devices and users, nature of executing tasks and dynamically changing system conditions. Extensive simulations demonstrate that the proposed anytime algorithms are able to quickly find a good initial solution and effectively optimise the rate at which the quality of the current solution improves as the algorithms are given more time to run, with a minimum overhead when compared against their traditional versions.