SHIM: a deterministic model for heterogeneous embedded systems

Abstract

Typical embedded hardware/software systems are implemented using a combination of C and an HDL such as Verilog. While each is well-behaved in isolation, combining the two gives a nondeterministic model whose ultimate behavior must be validated through expensive (cycle-accurate) simulation.We propose an alternative for describing such systems. Our SHIM (software/hardware integration medium) model, effectively Kahn networks with rendezvous communication, provides deterministic concurrency. We present the Tiny-SHIM language for such systems and its semantics, demonstrate how to implement it in hardware and software, and discuss how it can be used to model a real-world system.By providing a powerful, deterministic formalism for expressing systems, designing systems and verifying their correctness will become easier.

A Capability Calculus for Concurrency and Determinism

“…5 In our example, the first barrier in each process matches the other, and the second barrier in each process matches the other. Synchronizations that are not syntactically identifiable are often considered as a sign of potential bugs [1].…”

Deterministic Message Passing for Distributed Parallel Computing

“…We also show the information of control messages compared with normal application messages (Column [5][6][7][8]. Note that DMPI incurs little overhead for applications that do not trigger the two deterministic mechanisms.…”

“…They achieve determinism by using a restricted message-passing model. For example, the SHIM language [5], [22], [23] leverages FIFO pipes to pass messages between processes, and prevents the using of asynchronous messages and wildcard receiving operations. However, these programming models are usually domain-specific.…”

“…For example, Kahn's network is a deterministic message-passing model which leverages FIFO pipes to make communication. This message-passing model requires that all receiving operations should wait for a specified sending message [4], [5]. However, these models are too strict, thus they are easy to cause performance slowdown or introduce deadlocks if programs are not carefully designed.…”

Representation of synchronous, asynchronous, and polychronous components by clocked guarded actions

“…using several clocks, data flow languages like CAL [25] are based on data-driven paradigms [50][51][52], and others like SHIM [24] or most multi-threaded languages are based on asynchronous threads with a rendezvous-style communication [38,39].…”