Concurrent software verification with states, events, and deadlocks

We present a framework for model checking concurrent software systems which incorporates both states and events. Contrary to other state/event approaches, our work also integrates two powerful verification techniques, counterexample-guided abstraction refinement and compositional reasoning. Our specification language is a state/event extension of linear temporal logic, and allows us to express many properties of software in a concise and intuitive manner. We show how standard automata-theoretic LTL model checking algorithms can be ported to our framework at no extra cost, enabling us to directly benefit from the large body of research on efficient LTL verification. We also present an algorithm to detect deadlocks in concurrent message-passing programs. Deadlock- freedom is not only an important and desirable property in its own right, but is also a prerequisite for the soundness of our model checking algorithm. Even though deadlock is inherently non-compositional and is not preserved by classical abstractions, our iterative algorithm employs both (non-standard) abstractions and compositional reasoning to alleviate the state-space explosion problem. The resulting framework differs in key respects from other instances of the counterexample-guided abstraction refinement paradigm found in the literature. We have implemented this work in the magic verification tool for concurrent C programs and performed tests on a broad set of benchmarks. Our experiments show that this new approach not only eases the writing of specifications, but also yields important gains both in space and in time during verification. In certain cases, we even encountered specifications that could not be verified using traditional pure event-based or state-based approaches, but became tractable within our state/event framework. We also recorded substantial reductions in time and memory consumption when performing deadlock-freedom checks with our new abstractions. Finally, we report two bugs (including a deadlock) in the source code of Micro-C/OS versions 2.0 and 2.7, which we discovered during our experiments. This research was sponsored by the National Science Foundation (NSF) under grants no. CCR-9803774 and CCR-0121547, the Office of Naval Research (ONR) and the Naval Research Laboratory (NRL) under contract no. N00014-01-1-0796, the Army Research Office (ARO) under contract no. DAAD19-01-1-0485, and was conducted as part of the Predictable Assembly from Certifiable Components (PACC) project at the Software Engineering Institute (SEI). This article combines and builds upon the papers (CCO⁺04) and (CCOS04). Received December 2004 Revised July 2005 Accepted July 2005 by Eerke A. Boiten, John Derrick, Graeme Smith and Ian Hayes     

Testing and evaluating virus detectors for handheld devices

The widespread use of personal digital assistants and smartphones gives securing these devices a high priority. Yet little attention has been placed on protecting handheld devices against viruses. Currently available antivirus software for handhelds is few in number. At this stage, the opportunity exists for the evaluation and improvement of current solutions. By pinpointing weaknesses in the current antivirus software, improvements can be made to properly protect these devices from a future tidal wave of viruses. This research evaluates four currently available antivirus solutions for handheld devices. A formal model of virus transformation that provides transformation traceability is presented. Two sets of ten tests each were administered; nine tests from each set involved the modification of source code of two known viruses for handheld devices. The testing techniques used are well established in PC testing; thus the focus of this research is solely on handheld devices. Statistical analysis of the test results show high false negative production rates for the antivirus software and an overall false negative production rate of 47.5% with a 95% confidence interval between 36.6% and 58.4%. This high rate shows that current solutions poorly identify modified versions of a virus. The virus is left undetected and capable of spreading, infecting and causing damage.     

Effectiveness of single and coastal superposphates applied either in autumn or spring

Phosphorus (P) is known to leach laterally in water flowing during winter over pasture growing on flat soils that are shallow sands over slowly permeable lateritic ironstone gravel or clay soils in the high rainfall (> 800 mm annual average) areas of south-western Australia. The climate is Mediterranean, with hot dry December to March and cool wet April to November growing season, with excess water flowing over the surface from June to early-August. Fertilizer P is presently applied at about mid-March, near the start of the growing season. Single superphosphate has been applied for many years, which has a good residual value, and so the soils are no longer acutely P deficient. Consequently, a better method may be to apply the fertilizer at mid-August, after waterlogging and P leaching have usually receded, and radiation and temperature are rising, so pasture growth is increasing. The field experiment reported here was on a shallow sand over lateritic ironstone gravel where lateral leaching of P occurs. The experiment compared from 1990 to 1994 the effectiveness of single superphosphate (SSP), the fertilizer used at present, and coastal superphosphate (CSP), a partially acidulated rock phosphate containing about half the total P and one third the water-soluble P initially present in SSP. The fertilizers were applied annually either at mid-March or at mid-August. SSP applied at mid-March was the most effective treatment studied in the years when pasture plants had emerged before fertilizer was applied at mid-March. This is attributed to pasture plants being able to take up P from SSP applied at mid-March before leaching of P occurred, so that relative to SSP applied at mid-March, the other P fertilizer treatments (CSP applied at mid-march and mid-August, SSP applied at mid-August) were about equally or less effective. However, in years when the growing season had yet to start before fertilizer was applied at mid-March, then relative to SSP applied at mid-March, the other fertilizer treatments were equally or more effective. This is attributed to extensive leaching of P from SSP applied at mid-March, so that due to P losses from SSP applied at mid-March, the other treatments were equally or more effective. It is therefore concluded that profitable pasture production with reduced leaching is achieved by applying SSP at mid-March if soils are moist and pasture plants are growing at that time. However, if the soils are dry and no pasture plants are growing at mid-March, then CSP should be applied at mid-August.     

A method of limited replication for the efficient in vivo delivery of adenovirus to cancer cells

Replication-deficient viral vectors are currently being used in gene transfer strategies to treat cancer cells. Unfortunately, viruses are limited in their ability to diffuse through tissue. This makes it virtually impossible to infect the majority of tumor cells in vivo and results in inadequate gene transfer. This problem can be addressed by allowing limited viral replication. Limited viral replication facilitates greater penetration of virions into tissue and can improve gene transfer. We have developed a strategy of limited viral replication using AdRSVlaclys, a chemically modified E1-deleted adenovirus, to codeliver an exogenous plasmid encoding the adenovirus E1 region. This system allows one round of viral replication. We examined the effect of this limited adenovirus replication in vitro and in vivo. In culture, codelivery of virus and pE1 resulted in a large increase in infected cells when compared with control cells exposed to virus and pUC19. In experiments on nude mice bearing HeLa ascites tumors, intraperitoneal injection of AdRSVlaclys/pE1 resulted in a significantly higher percentage of infected HeLa cells as compared with the PBS controls (p < 0.05) or the AdRSVlaclys/pUC19 controls (p < 0.01). These data demonstrate that the transcomplementation of replication-deficient adenovirus with exogenous E1 DNA leads to limited replication, and this controlled replication enhances gene transfer efficiency of adenovirus in vivo.     

Immunoreactivity of neural crest-derived cells in thymic tissue developing under the rat kidney capsule

In order to study the functional development of a thymus in an experimental model, small pieces of adult rat thymic tissue were cultured for 9 days and implanted under the kidney capsule of littermates. The tissues were examined with a panel of antibodies raised against thymic and neural factors and neural crest cells at intervals from 5 to 13 days. At 5 days post-implantation, there were groups of L1+ cells within the implants that reacted with antibodies raised against neural and neural crest cell markers. L1+ cells were highly mitotic, rounded cells measuring 8.7 +/- 0.6 micrometer in diameter. Double immunostaining with different combinations of antibodies showed that 94% of the L1+ cells were also TH+, and many were HNK-1/NCAM+, PGP 9.5+, NGF+, chromogranin A+, VIP+, S100+, CGRP+, GAD+, and A2B5+. A few were also pan-cytokeratin+. These results indicate that these cells are derived from neural crest derived cells and belong to the neuroepithelial line of development. The L1+ cells were most numerous before nerves appeared (about Day 9) and reduced in number and extent as the thymus differentiated. The neural crest cells occasionally had long cytoplasmic extensions, but it was not possible to decide if they formed the nerves that appeared in the implants. Adult thymuses also contained a population of L1+ and HNK-1/NCAM+ cells, mainly in the subcapsular cortex, the septa, and the medulla. These cells could be a source of neural crest cells able to repopulate the implant. The adult thymus may always contain a reservoir of cells potentially capable of producing neuropeptides and transmitter factors required for thymic growth and regeneration.     

New preprocessing technique for digital facsimile transmission

A new preprocessing technique which removes unwanted notches and random noise generated by electronic scanners is presented. Using a 3 element by 3 element mask, the central picture element is altered from black to white, or vice versa, if the surrounding elements correspond to a predetermined pattern. A significant improvement in both subjective image quality and coding efficiency is observed.     

Regulator for controlling liquid helium bath near 4.2 K

A simple regulator for stabilizing the temperature of a liquid (4)He bath near 4.2 K is described. The device controls the pressure of the (4)He gas, and achieves a stability of +/-50 muK for the hold-time of the cryostat.     

International coffee standardisation and legislation

The author describes the particular technical features of coffee standards used commercially and/or legislatively. Existing national standards on green coffee and their harmonisation of terminology, sampling and test method are described and discussed. Comparisons are made of the various legislative standards of different countries with respect to instant coffee and their harmonisation within the EEC member countries. Roasted coffee is considered to a lesser extent as it is not an important element in the world trade for coffee. The role of various international organisations in standardisation is described. It is emphasised that the flavour of coffee is not amenable to international standardisation.     

Thin Film Adhesion: Effect of Glow Discharge on Substrate

Glow discharge treatment of soda lime glass and silica substrates prior to depositing copper films in oil-free ultra high vacuum as well as in conventional vacuum has been found to increase film adhesion suggesting that gross contamination removal is not the only effect of importance in this method of improving film adhesion. There is evidence from Auger electron emission spectroscopy for sorption of gas from the discharge into the substrate surface layers. Desorption of gases from the substrate also appears to take place. A difference in electron emission from a variety of insulating surfaces following glow discharge was observed in a scanning electron microscope, the sign of the change depending on the nature of the discharge gas. No sputtered material from system components could be detected on the substrate and there was no detectable micro-roughening nor net surface electric charge.     

Polarization mapping: a method to improve sum frequency generation spectral analysis

A "polarization mapping" method has been applied to improve the fitting quality of sum frequency generation (SFG) vibrational spectra of complicated systems and validate the data analysis of simple SFG spectra. Using such a method, two-dimensional SFG spectra can be constructed, from which more reliable spectral information can be obtained from a surface/interface. Model calculations as well as experiments have been employed to illustrate the power of the polarization mapping method for spectral analysis of SFG spectra. By using a deuterated polystyrene surface and interfacial protein molecules as examples, we demonstrate that this method is especially important for complicated molecules, such as polymers and proteins.