What it means for a concurrent program to satisfy a specification: why no one has specified priority
L Lamport - Proceedings of the 12th ACM SIGACT-SIGPLAN …, 1985 - dl.acm.org
L Lamport
Proceedings of the 12th ACM SIGACT-SIGPLAN symposium on Principles of …, 1985•dl.acm.orgThe formal correspondence between an implementation and its specification is examined. It
is shown that existing specifications that claim to describe priority are either vacuous or else
too restrictive to be implemented in some reasonable situations. This is illustrated with a
precisely formulated problem of specifying a first-come-first-served mutual exclusion
algorithm, which it is claimed cannot be solved by existing methods.
is shown that existing specifications that claim to describe priority are either vacuous or else
too restrictive to be implemented in some reasonable situations. This is illustrated with a
precisely formulated problem of specifying a first-come-first-served mutual exclusion
algorithm, which it is claimed cannot be solved by existing methods.
The formal correspondence between an implementation and its specification is examined. It is shown that existing specifications that claim to describe priority are either vacuous or else too restrictive to be implemented in some reasonable situations. This is illustrated with a precisely formulated problem of specifying a first-come-first-served mutual exclusion algorithm, which it is claimed cannot be solved by existing methods.