Luk Levrouw

A New Trace And Replay System For Shared Memory Programs Based On Lamport Clocks

One of the key pwblema iin debugging parallel pwgrama is that the behavior of a parallel program in response to a jbed input may be indeterminate. As a consequence of this non-repeatability, cyclic monitoring techniques for error isollation are not guaranteed to work. To tackle this problem, diferent trace and replay mechanwma have been proposed to support the debugging taak. In this paper we introduce a novel trace and replag mechanism for shared-memoty programa, baaed on Lamport cdocka. We compare the new scheme with Leblanc’s ‘Instant Replay’ and Netzer ’s ‘optimal tracing’. Our approach produces (much) smaller traces than both ezirting approaches. Moreover, in comparison with Ntstzer’s approach, our approach induces much less ovlerhead during the recording phase.

Interrupt replay: a debugging method for parallel programs with interrupts

Abstract The behaviour of programs for multiprocessors may be indeterminate, due to processor timing variations. This poses a problem for cyclic debugging, since a bug may disappear from one execution to another. Replay is an elegant solution to this problem, in which ‘sufficient’ information is recorded in a log. This information is then used to control subsequent executions of the same program so that repeatability is guaranteed. Interrupts are another source of non-determinism, even in sequential programs. This paper presents an extension of the well-known Instant Replay method, termed Interrupt Replay, for replaying programs in the presence of interrupts. The correctness of Interrupt Replay is based on the assumption that there are no interrupt races: an interrupt service routine must not access data that is also accessed by the foreground process whenever the interrupt is enabled. If such races are present then replay may fail to produce deterministic results. This assumption is similar to the basic assumption of Instant Replay that shared variables are properly protected by mutual exclusion. Also as in Instant Replay, it is assumed that the behaviour of the environment (input data, external interrupts) is replayed by some other tracing mechanism.

Space efficient data race detection for parallel programs with series-parallel task graphs

A data race or access anomaly is a bug in parallel programs, occurring when two parallel processes access the same shared resource in an unsynchronised fashion, and at least one access modifies the resource. The effects of a data race can be non-deterministic and can give the program debugger a very hard time. Race detection represents a class of methods which automatically detect the presence of data races in a parallel program. This paper deals with on-the-fly methods, in which race detection is performed at run-time, during the execution of the program. We present a novel method for on-the-fly race detection, which reduces the storage requirements by several orders of magnitude over previous methods. Moreover, the method is language independent and can handle the problems associated with pointer variables (aliasing). The programs it deals with are restricted to those that have a series-parallel task graph. This includes fork-join and spawning types of parallelism.<<ETX>>

Minimizing the Log Size for Execution Replay of Shared-Memory Programs

Execution replay is an essential part of parallel debugging. Since explicitly parallel programs can be non-repeatable, trace-and-replay mechanisms have been proposed to support the debugging task.

On the implementation of a replay mechanism

Parallel programs can be nondeterministic: consecutive runs with the same input can result in different executions. Therefore we cannot use cyclic debugging techniques. In order to be able to use those techniques we need a tool that traces information about an execution so it can be replayed for debugging. Because the recording interferes with the program, possibly perturbating the execution, we must limit the amount of information and keep the algorithm simple. This paper presents an implementation of the ROLT replay mechanism for a multithreaded operating system (Solaris).

An efficient record-replay mechanism for shared memory programs

One of the main difficulties in debugging parallel programs is that parallel programs execute non-deterministically. As a consequence, cyclic debugging techniques for error isolation are not guaranteed to work. For that reason, different record-replay mechanisms have been proposed. Ordering based mechanisms log the relative order in which process interactions take place in order to obtain reproducibility. In this paper, existing ordering based approaches for shared memory programs are compared, and a novel approach, situated somewhere in between the existing approaches, is presented. To obtain repeatability, we record the relative ordering of the processor-level synchronization events. The proposed mechanism can also be used in support of performance debugging.<<ETX>>

Steps toward triaural perception

By virtue of their low cost and simplicity, ultrasonic sensors are widely used in rangefinding applications. However, when a more detailed map of the environment is needed, as is the case in mobile robotics, the simple-minded use of a single ultrasonic sensor is often insufficient. In this paper we propose a measurement system composed of three ultrasonic sensors, one transmitting and three receiving, to overcome these problems. By combining the information from the three sensors we can accurately determine the position, both in distance and bearing, of all objects in the field of view taking one single snapshot of the scene. Within the limits derived in this paper, it is even possible to discriminate between different types of reflectors. To get these results we have to find the reflections belonging to the same object in the signals received by the three sensors, a problem analogous to the correspondence problem in stereo- vision. We present a maximum likelihood approach as a solution to this problem. A realistic scene will be used to compare our sensor with the traditional ultrasonic ranging sensor.