BUNDLEP: Prioritizing Conflict Free Regions in Multi-threaded Programs to Improve Cache Reuse

Abstract

In "BUNDLE: Real-Time Multi-Threaded Scheduling to Reduce Cache Contention", Tessler and Fisher propose a scheduling mechanism and combined worst-case execution time calculation method that treats the instruction cache as a beneficial resource shared between threads. Object analysis produces a worst-case execution time bound and separates code segments into regions. Threads are dynamically placed in bundles associated with regions at run time by the BUNDLE scheduling algorithm where they benefit from shared cache values.In the evaluation of the previous work, tasks were created with a predetermined worst-case execution time path through the control flow graph. Apriori knowledge of the worst-case path is an impractical restriction on any analysis. At the time, the only other solution available was an all-paths search of the graph, which is an equally impractical approach due to its complexity.The primary focus of this work is to build upon BUNDLE, expanding its applicability beyond a proof of concept. We present a complete a worst-case execution time calculation method that includes thread level context switch costs, operating on real programs, with representative architecture parameters, and compare our results to those produced by Heptane's state of the art method. To these ends, we propose a modification to the BUNDLE scheduling algorithm called BUNDLEP. Bundles are assigned priorities that enforce an ordered flow of threads through the control flow graph -avoiding the need for multiple all-paths searches through the graph. In many cases, our evaluation shows a run-time and analytical benefit for BUNLDEP compared to serialized thread execution and state of the art WCET analysis.

Bringing Inter-Thread Cache Benefits to Federated Scheduling -- Extended Results & Technical Report

“…A positive perspective of caches is taken by BUNDLE [54][55][56] for multi-threaded tasks running on a single processor. This positive perspective is reflected in the WCET of a task which includes the inter-thread cache benefit: the speed-up one thread experiences by another placing values in the cache.…”

“…Tasks in the BUNDLE [54][55][56] model are represented by a tuple τ i = (p i , d i , c i (η i ), η i , o i ). A task has the familiar minimum inter-arrival time p i and relative deadline d i .…”

“…To produce the WCETO for a task, the control flow graph of the executable object is divided into conflict free regions: sub-graphs of the control flow graph where no two instructions map to the same cache block. Conflict free regions serve as input to the BUNDLE scheduling algorithm, which maximizes the number of threads executing over each region [55] in turn maximizing the inter-thread cache benefit.…”

“…Thus, for any η a < η b < η c the point (η b , c(η b )) lies above the line defined by (η a , c(η a )) and (η c , c(η c )), therefore c(η) is concave. Multi-threaded programs executed by BUNDLEP [55] illustrate the discrete concave growth described by the analysis.…”

“…In the setting of uniprocessor multi-threaded task systems, the BUNDLE-based approaches [54][55][56] (referred to as BUNDLE throughout the rest of this work) treat cache memory positively, creating a benefit to schedulability. BUNDLE restricts the execution of the multiple threads of task on a single processor in a cache cognizant manner.…”