| 摘要 |
A computer system, program product and method that monitor the threads executing within a region of a computer program during debugging. This region may be a plurality of nonadjacent sections of code, each with defined entry and exit addresses identified by control points. Some or all threads may be halted depending on a predetermined criteria related to threads of interest or the number of threads executing in the region. Of special interest is monitoring for a thread count so that timing errors may be analyzed for when some plurality of threads simultaneously execute within the region. Moreover, in the illustrative embodiment, control points are implemented for thread monitoring in a manner similar to a break point, utilizing a break point table to determine whether a system exception is due to a break point or to a thread monitor control point. If the latter, program execution continues after storing the thread identifier in a record for the thread monitor control point in the break point table.
|
