A True O(1) Parallel Deadlock Detection Algorithm for Single-Unit Resource Systems and Its Hardware Implementation

Abstract

Due to rapid technology advance, Multiprocessor System-on-Chips (MPSoCs) are likely to become commodity computing platforms for embedded applications. In the future, it is possible that an MPSoC is equipped with a large number of processing elements as well as on-chip resources. The management of these faces many challenges, among which deadlock is one of the most crucial issues. This paper presents a novel hardware-oriented deadlock detection algorithm suitable for current and future MPSoCs. Unlike previously published methods whose runtime complexities are often affected by the number of processing elements and resources in the system, the proposed algorithm leverages specialized hardware to guarantee O(1) overall runtime complexity. Such complexity is achieved by: 1) classifying resource allocation events; 2) for each type of events, using hardware to perform a set of specific detection and/or preparation operations that only takes constant runtime; and 3) updating necessary information for multiple resources in parallel in hardware. We implement the algorithm in Verilog HDL and demonstrate through simulation that each algorithm invocation takes at most four clock cycles.

GPU-OSDDA: a bit-vector GPU-based deadlock detection algorithm for single-unit resource systems

“…Later, Xiao and Lee developed a new approach to classifying resource events in a single-unit resource system. This development led to a new algorithm known as O(1) Single-Unit Deadlock Detection Algorithm (OSDDA) [9]. By utilizing the new classification of resource events, deadlock preparation was able to be completed in O(1) time.…”

“…As a result, OSDDA was able to achieve an overall run-time complexity of O(1) in hardware. The basis of our algorithm is rooted in the methodology of OSDDA [9]. The core operation of OSDDA is based upon a classification of re-source events in the system which is further discussed in Section 2.4.…”

“…Thus, a cycle in the RAG is a necessary and sufficient condition for deadlock [9]. (2) A process requests one resource at a time (see Definition 3).…”

“…It is able to achieve this by performing parallel computations on a RAG based on the classification of resource events in the system. The three types of resource events in OSDDA are: granted resource requests, blocked resource requests, and resource release [9]. The resource events and deadlock detection capability of OSDDA are briefly discussed in Sections 2.4.1 and 2.4.2.…”

“…As a result, researchers have developed hardware-based algorithms [7][8][9][10][11] that expanded upon the findings of these software algorithms. Hardware algorithms led to very fast and deterministic results but lacked the ability to handle an increasing amount of processes and resources (due to size constraints and hardware complexity), as would be seen in a real world system.…”