We consider the problem of transforming a given sequential implementation of a data structure into a wait-free concurrent implementation. Given the code for different operations of an object that is designed to work under the assumption that only a single process accesses it, we want to construct an implementation that works correctly in a concurrent environment where it may be accessed by many processes.
We assume a shared memory model with atomic registers, that is, registers that support either a read or a write in one atomic step. It is well known that using atomic registers it is impossible to construct concurrent implementations of even very simple objects such as test-and-set bits. However, we show that the knowledge about relative speeds of processes {\em can} be used for such implementations. We assume that there is a known upper bound on the time taken by the slowest process to execute a statement involving an access to the shared memory. This timing assumption is very powerful and enables us to construct wait-free implementations of data structures such as queues, stacks and synchronization primitives such as test-and-set, compare-and-swap, fetch-and-add, etc.
Our transformation works only when the given sequential implementation is bounded, that is, there is a known upper bound on the number of steps required to complete any of the operations it supports. Our transformation ensures that in the absence of contention there is only a small overhead in the cost of executing the concurrent operations over the sequential ones, namely, only a constant number of accesses to the shared memory.
Proceedings of the Fifth IEEE Symposium on Parallel and Distributed Processing (SPDP 1993), pp. 470-477.