You have created your application and resolved the build errors. Now it's time to find and correct those logic errors that keep your application or stored procedures from running correctly. You can do this with the mpC Workshop's integrated mpC debugger. The mpC debugger allows you to test your mpC programs. You can manage and set breakpoints and watchpoints, as well as view values of variables.
The mpC programming language provides facilities to manage a kind of resource named computing space. It is a set of virtual processors of different performances interconnected with links of different bandwidth. mpC Programmer can manage computing space by allocating and discarding regions of computing space named networks. Managing computing space in mpC is similar to managing memory in C. At run-time virtual processors correspond to processes of parallel program. When debugging a parallel program, you can view information about all available virtual processors in the program. Using Parallel Debugger, you can start and stop individual groups of virtual processors (see Cursors overview).
The mpC Parallel Debugger provides information about the nets and subnets, which exist in the debugged program. So you donít have to group virtual processors by hand to achieve the grouping of virtual processors that is logically correct for the debugged program. The mpC Parallel Debugger groups processes automatically.
The mpC Parallel Debugger has client-server architecture. Such architecture allows the debugger to support remote debugging. Since client and server communicate with each other over TCP/IP protocol, you can debug programs running on remote machines in New York from your workplace in London.
This chapter explains how to debug mpC parallel programs using the mpC parallel debugger. This chapter discusses: