arXiv subjects

Brijesh Dongol

Publications and source records attributed to Brijesh Dongol.

34 records · Page 2Linked to original sources

arXiv2020

Defining and Verifying Durable Opacity: Correctness for Persistent Software Transactional Memory

Non-volatile memory (NVM), aka persistent memory, is a new paradigm for memory that preserves its contents even after power loss. The expected ubiquity of NVM has stimulated interest in the design of novel concepts ensuring correctness of concurrent programming abstractions in the face of persistency. So far, this has lead to the design of a number of persistent concurrent data structures, built to satisfy an associated notion of correctness: durable linearizability. In this paper, we transfer the principle of durable concurrent correctness to the area of software transactional memory (STM). Software transactional memory algorithms allow for concurrent access to shared state. Like linearizability for concurrent data structures, opacity is the established notion of correctness for STMs. First, we provide a novel definition of durable opacity extending opacity to handle crashes and recovery in the context of NVM. Second, we develop a durably opaque version of an existing STM algorithm, namely the Transactional Mutex Lock (TML). Third, we design a proof technique for durable opacity based on refinement between TML and an operational characterisation of durable opacity by adapting the TMS2 specification. Finally, we apply this proof technique to show that the durable version of TML is indeed durably opaque. The correctness proof is mechanized within Isabelle.

cs.DC↗

arXiv2020

Integrating Owicki-Gries for C11-Style Memory Models into Isabelle/HOL

Weak memory presents a new challenge for program verification and has resulted in the development of a variety of specialised logics. For C11-style memory models, our previous work has shown that it is possible to extend Hoare logic and Owicki-Gries reasoning to verify correctness of weak memory programs. The technique introduces a set of high-level assertions over C11 states together with a set of basic Hoare-style axioms over atomic weak memory statements (e.g., reads/writes), but retains all other standard proof obligations for compound statements. This paper takes this line of work further by showing Nipkow and Nieto's encoding of Owicki-Gries in the Isabelle theorem prover can be extended to handle C11-style weak memory models in a straightforward manner. We exemplify our techniques over several litmus tests from the literature and a non-trivial example: Peterson's algorithm adapted for C11. For the examples we consider, the proof outlines can be automatically discharged using the existing Isabelle tactics developed by Nipkow and Nieto. The benefit here is that programs can be written using a familiar pseudocode syntax with assertions embedded directly into the program.

cs.PL↗

arXiv2018

Verifying C11 Programs Operationally

This paper develops an operational semantics for a release-acquire fragment of the C11 memory model with relaxed accesses. We show that the semantics is both sound and complete with respect to the axiomatic model. The semantics relies on a per-thread notion of observability, which allows one to reason about a weak memory C11 program in program order. On top of this, we develop a proof calculus for invariant-based reasoning, which we use to verify the release-acquire version of Peterson's mutual exclusion algorithm.

cs.PL↗

arXiv2018

Proceedings 18th Refinement Workshop

Refinement is one of the cornerstones of a formal approach to software engineering. Refinement is the process of developing a more detailed design or implementation from an abstract specification through a sequence of mathematically-based steps that maintain correctness with respect to the original specification. Work on the foundations of languages such as Z, B, VDM and CSP have led to their widespread use in certain industrial sectors, e.g., those with security or safety critical concerns. In addition to precise specification, formal methods also allow the possibility of precise and verifiable development, as captured by the concept of refinement. The 18th Refinement Workshop was held as part of FLoC 2018 at Oxford, UK.

cs.LO↗

arXiv2018

Causal Linearizability: Compositionality for Partially Ordered Executions

In the interleaving model of concurrency, where events are totally ordered, linearizability is compositional: the composition of two linearizable objects is guaranteed to be linearizable. However, linearizability is not compositional when events are only partially ordered, as in many weak-memory models that describe multicore memory systems. In this paper, we present causal linearizability, a correctness condition for concurrent objects implemented in weak-memory models. We abstract from the details of specific memory models by defining our condition using Lamport's execution structures. We apply our condition to the C11 memory model, providing a correctness condition for C11 objects. We develop a proof method for verifying objects implemented in C11 and related models. Our method is an adaptation of simulation-based methods, but in contrast to other such methods, it does not require that the implementation totally order its events. We also show that causal linearizability reduces to linearizability in the totally ordered case.

cs.LO↗

arXiv2016

Reducing Opacity to Linearizability: A Sound and Complete Method

Transactional memory is a mechanism that manages thread synchronisation on behalf of a programmer so that blocks of code execute with an illusion of atomicity. The main safety criterion for transactional memory is opacity, which defines conditions for serialising concurrent transactions. Proving opacity is complicated because it allows concurrent transactions to observe distinct memory states, while TM implementations are typically based on one single shared store. This paper presents a sound and complete method, based on coarse-grained abstraction, for reducing proofs of opacity to the relatively simpler correctness condition: linearizability. We use our methods to verify TML and NORec from the literature and show our techniques extend to relaxed memory models by showing that both are opaque under TSO without requiring additional fences. Our methods also elucidate TM designs at higher level of abstraction; as an application, we develop a variation of NORec with fast-path reads transactions. All our proofs have been mechanised, either in the Isabelle theorem prover or the PAT model checker.

cs.LO↗

arXiv2016

Towards linking correctness conditions for concurrent objects and contextual trace refinement

Correctness conditions for concurrent objects describe how atomicity of an abstract sequential object may be decomposed. Many different concurrent objects and proof methods for them have been developed. However, arguments about correctness are conducted with respect to an object in isolation. This is in contrast to real-world practice, where concurrent objects are often implemented as part of a programming language library (e.g., java.util.concurrent) and are instantiated within a client program. A natural question to ask, then is: How does a correctness condition for a concurrent object ensure correctness of a client program that uses the concurrent object? This paper presents the main issues that surround this question and provides some answers by linking different correctness conditions with a form of trace refinement.

cs.LO↗

arXiv2016

Contextual trace refinement for concurrent objects: Safety and progress

Correctness of concurrent objects is defined in terms of safety properties such as linearizability, sequential consistency, and quiescent consistency, and progress properties such as wait-, lock-, and obstruction-freedom. These properties, however, only refer to the behaviours of the object in isolation, which does not tell us what guarantees these correctness conditions on concurrent objects provide to their client programs. This paper investigates the links between safety and progress properties of concurrent objects and a form of trace refinement for client programs, called contextual trace refinement. In particular, we show that linearizability together with a minimal notion of progress are sufficient properties of concurrent objects to ensure contextual trace refinement, but sequential consistency and quiescent consistency are both too weak. Our reasoning is carried out in the action systems framework with procedure calls, which we extend to cope with non-atomic operations.

cs.DC↗

arXiv2015

Decidability and Complexity for Quiescent Consistency and its Variations

Quiescent consistency is a notion of correctness for a concurrent object that gives meaning to the object's behaviours in quiescent states, i.e., states in which none of the object's operations are being executed. Correctness of an implementation object is defined in terms of a corresponding abstract specification. This gives rise to two important verification questions: membership (checking whether a behaviour of the implementation is allowed by the specification) and correctness (checking whether all behaviours of the implementation are allowed by the specification). In this paper, we show that the membership problem for quiescent consistency is NP-complete and that the correctness problem is decidable, but coNP-hard and in EXPSPACE. For both problems, we consider restricted versions of quiescent consistency by assuming an upper limit on the number of events between two quiescent points. Here, we show that the membership problem is in PTIME, whereas correctness is in PSPACE. Quiescent consistency does not guarantee sequential consistency, i.e., it allows operation calls by the same process to be reordered when mapping to an abstract specification. Therefore, we also consider quiescent sequential consistency, which strengthens quiescent consistency with an additional sequential consistency condition. We show that the unrestricted versions of membership and correctness are NP-complete and undecidable, respectively. When by placing a limit on the number of events between two quiescent points, membership is in PTIME, while correctness is in PSPACE. Finally, we consider a version of quiescent sequential consistency that places an upper limit on the number of processes for every run of the implementation, and show that the membership problem for quiescent sequential consistency with this restriction is in PTIME.

cs.LO↗

arXiv2015

Verifying linearizability: A comparative survey

Linearizability has become the key correctness criterion for concurrent data structures, ensuring that histories of the concurrent object under consideration are consistent, where consistency is judged with respect to a sequential history of a corresponding abstract data structure. Linearizability allows any order of concurrent (i.e., overlapping) calls to operations to be picked, but requires the real-time order of non-overlapping to be preserved. Over the years numerous techniques for verifying linearizability have been developed, using a variety of formal foundations such as refinement, shape analysis, reduction, etc. However, as the underlying framework, nomenclature and terminology for each method differs, it has become difficult for practitioners to judge the differences between each approach, and hence, judge the methodology most appropriate for the data structure at hand. We compare the major of methods used to verify linearizability, describe the main contribution of each method, and compare their advantages and limitations.

cs.LO↗

arXiv2014

Principles for Verification Tools: Separation Logic

A principled approach to the design of program verification and con- struction tools is applied to separation logic. The control flow is modelled by power series with convolution as separating conjunction. A generic construction lifts resource monoids to assertion and predicate transformer quantales. The data flow is captured by concrete store/heap models. These are linked to the separation algebra by soundness proofs. Verification conditions and transformation laws are derived by equational reasoning within the predicate transformer quantale. This separation of concerns makes an implementation in the Isabelle/HOL proof as- sistant simple and highly automatic. The resulting tool is correct by construction; it is explained on the classical linked list reversal example.

cs.LO↗

arXiv2014

Convolution, Separation and Concurrency

A notion of convolution is presented in the context of formal power series together with lifting constructions characterising algebras of such series, which usually are quantales. A number of examples underpin the universality of these constructions, the most prominent ones being separation logics, where convolution is separating conjunction in an assertion quantale; interval logics, where convolution is the chop operation; and stream interval functions, where convolution is used for analysing the trajectories of dynamical or real-time systems. A Hoare logic is constructed in a generic fashion on the power series quantale, which applies to each of these examples. In many cases, commutative notions of convolution have natural interpretations as concurrency operations.

cs.LO↗

arXiv2013

Simplifying proofs of linearisability using layers of abstraction

Linearisability has become the standard correctness criterion for concurrent data structures, ensuring that every history of invocations and responses of concurrent operations has a matching sequential history. Existing proofs of linearisability require one to identify so-called linearisation points within the operations under consideration, which are atomic statements whose execution causes the effect of an operation to be felt. However, identification of linearisation points is a non-trivial task, requiring a high degree of expertise. For sophisticated algorithms such as Heller et al's lazy set, it even is possible for an operation to be linearised by the concurrent execution of a statement outside the operation being verified. This paper proposes an alternative method for verifying linearisability that does not require identification of linearisation points. Instead, using an interval-based logic, we show that every behaviour of each concrete operation over any interval is a possible behaviour of a corresponding abstraction that executes with coarse-grained atomicity. This approach is applied to Heller et al's lazy set to show that verification of linearisability is possible without having to consider linearisation points within the program code.

cs.LO↗

arXiv2013

Data refinement for true concurrency

The majority of modern systems exhibit sophisticated concurrent behaviour, where several system components modify and observe the system state with fine-grained atomicity. Many systems (e.g., multi-core processors, real-time controllers) also exhibit truly concurrent behaviour, where multiple events can occur simultaneously. This paper presents data refinement defined in terms of an interval-based framework, which includes high-level operators that capture non-deterministic expression evaluation. By modifying the type of an interval, our theory may be specialised to cover data refinement of both discrete and continuous systems. We present an interval-based encoding of forward simulation, then prove that our forward simulation rule is sound with respect to our data refinement definition. A number of rules for decomposing forward simulation proofs over both sequential and parallel composition are developed.

cs.LO↗

arXiv2012

Proving linearisability via coarse-grained abstraction

Linearisability has become the standard safety criterion for concurrent data structures ensuring that the effect of a concrete operation takes place after the execution some atomic statement (often referred to as the linearisation point). Identification of linearisation points is a non-trivial task and it is even possible for an operation to be linearised by the execution of other concurrent operations. This paper presents a method for verifying linearisability that does not require identification of linearisation points in the concrete code. Instead, we show that the concrete program is a refinement of some coarse-grained abstraction. The linearisation points in the abstraction are straightforward to identify and the linearisability proof itself is simpler due to the coarse granularity of its atomic statements. The concrete fine-grained program is a refinement of the coarse-grained program, and hence is also linearisable because every behaviour of the concrete program is a possible behaviour its abstraction.

cs.LO↗

arXiv2006

Extending the theory of Owicki and Gries with a logic of progress

This paper describes a logic of progress for concurrent programs. The logic is based on that of UNITY, molded to fit a sequential programming model. Integration of the two is achieved by using auxiliary variables in a systematic way that incorporates program counters into the program text. The rules for progress in UNITY are then modified to suit this new system. This modification is however subtle enough to allow the theory of Owicki and Gries to be used without change.

cs.LO↗