Many large-scale multithreaded programs incur deadlock bugs. Existing deadlock warning detection techniques only report warning scenarios, which may or may not be real deadlocks. Each warning should be further verified on whether it may manifest into a real deadlock. For this purpose, a number of active randomized testing schedulers have been developed to trigger them, and yet pervious experiments show that their deadlock confirmation probability can be low. This paper presents ASN, a novel barrier-based randomized scheduler that triggers real deadlocks with high probabilities. We exploit the insights that in a confirmation run, the threads involved in a real deadlock should properly acquire one or more sets of locks prior to deadlocking. ASN automatically identifies three interesting sets of such positions. It guides the threads participating in a given warning to stay at these position sets in turn. When all the threads are staying at the last position set, ASN checks whether any deadlock that matches with the given warning has been triggered. We have evaluated ASN on 15 deadlock bugs in a suite of real-world multithreaded programs. The results show that ASN either confirms more deadlocks from the benchmark suite or triggers the same deadlocks with significantly higher probabilities than existing schedulers.

An external web service may evolve without prior notification. In the course of the regression testing of a workflow-based web service, existing test case prioritization techniques may only verify the latest service composition using the not-yet-executed test cases, overlooking high-priority test cases that have already been applied to the service composition before the evolution. In this paper, we propose Preemptive Regression Testing (PRT), an adaptive testing approach to addressing this challenge. Whenever a change in the coverage of any service artifact is detected, PRT recursively preempts the current session of regression test and creates a sub-session of the current test session to assure such lately identified changes in coverage by adjusting the execution priority of the test cases in the test suite. Then, the sub-session will resume the execution from the suspended position. PRT terminates only when each test case in the test suite has been executed at least once without any preemption activated in between any test case executions. The experimental result confirms that testing workflow-based web service in the face of such changes is very challenging; and one of the PRT-enriched techniques shows its potential to overcome the challenge.

Location-based services (LBS) are widely deployed. When the implementation of an LBS-enabled service has evolved, regression testing can be employed to assure the previously established behaviors not having been adversely affected. Proper test case prioritization helps reveal service anomalies efficiently so that fixes can be scheduled earlier to minimize the nuisance to service consumers. A key observation is that locations captured in the inputs and the expected outputs of test cases are physically correlated by the LBS-enabled service, and these services heuristically use estimated and imprecise locations for their computations, making these services tend to treat locations in close proximity homogenously. This paper exploits this observation. It proposes a suite of metrics and initializes them to demonstrate input-guided techniques and point-of-interest (POI) aware test case prioritization techniques, differing by whether the location information in the expected outputs of test cases is used. It reports a case study on a stateful LBS-enabled service. The case study shows that the POI-aware techniques can be more effective and more stable than the baseline, which reorders test cases randomly, and the input-guided techniques. We also find that one of the POI-aware techniques, cdist, is either the most effective or the second most effective technique among all the studied techniques in our evaluated aspects, although no technique excels in all studied SOA fault classes.

This paper presents the aim of TeamWork, our ongoing effort to develop a comprehensive dynamic deadlock confirmation tool for multithreaded programs. It also presents a refined object abstraction algorithm that refines the existing stack hash abstraction.

A workflow-based web service may use ultra-late binding to invoke external web services to concretize its implementation at run time. Nonetheless, such external services or the availability of recently used external services may evolve without prior notification, dynamically triggering the workflow-based service to bind to new replacement external services to continue the current execution. Any integration mismatch may cause a failure. In this paper, we propose Preemptive Regression Testing (PRT), a novel testing approach that addresses this adaptive issue. Whenever such a late-change on the service under regression test is detected, PRT preempts the currently executed regression test suite, searches for additional test cases as fixes, runs these fixes, and then resumes the execution of the regression test suite from the preemption point.

Dynamic race detectors can explore multiple thread schedules of a multithreaded program over the same input to detect data races. Although existing sampling-based precise race detectors reduce overheads effectively so that lightweight precise race detection can be performed in testing or post-deployment environments, they are ineffective in detecting races if the sampling rates are low. This paper presents CARISMA to address this problem. CARISMA exploits the insight that along an execution trace, a program may potentially handle many accesses to the memory locations created at the same site for similar purposes. Iterating over multiple execution trials of the same input, CARISMA estimates and distributes the sampling budgets among such location creation sites, and probabilistically collects a fraction of all accesses to the memory locations associated with such sites for subsequent race detection. Our experiment shows that, compared with PACER on the same platform and at the same sampling rate (such as 1%), CARISMA is significantly more effective.

Dynamic service compositions pose new verification and validation challenges such as uncertainty in service membership. Moreover, applying an entire test suite to loosely coupled services one after another in the same composition can be too rigid and restrictive. In this paper, we investigate the impact of service selection on service-centric testing techniques. Specifically, we propose to incorporate service selection in executing a test suite and develop a suite of metrics and test case prioritization techniques for the testing of location-aware services. A case study shows that a test case prioritization technique that incorporates service selection can outperform their traditional counterpart - the impact of service selection is noticeable on software engineering techniques in general and on test case prioritization techniques in particular. Further-more, we find that points-of-interest-aware techniques can be significantly more effective than input-guided techniques in terms of the number of invocations required to expose the first failure of a service composition.