System for debugging shared memory multiprocessor computers.

摘要

System for debugging application programs being executed in a parallel processor computer system. A debugging program running in one processor identifies data events occurring on a shared memory bus as significant breakpoints to identify data errors. A bus monitor is programmed to detect the data events as they occur in the shared memory bus. A trace memory is provided to record the detected data events along with a time stamp. The trace memory may be read and the program analyzed for the determination of data errors which may have occurred. <IMAGE>