Control section dependency is determined by the requirements of a control section for <references to> or <access to> a given <routine> of <entry point> in another control section. A control section is dependent upon any control section from which it receives control or that processes its data. For example, if control section C receives control from control section B, C is dependent upon B. That is, both control sections must be in storage before execution can continue beyond a given point in the program.

Assume that a program contains seven control sections, CSA through CSG, and exceeds the amount of storage available for its execution. Before the program is rewritten, it is examined to see if it could be placed into an overlay structure. Figure 1 shows the groups of dependent control sections in the program (the arrows indicate dependencies).

Figure 1. Control section dependencies

Each dependent group is also a path. That is, if control section CSG is executed, CSB and CSA must also be in storage. Because CSA and CSB are in each path, they must be in the root segment. Control section CSC is in two groups and therefore is a common segment in two different paths.

A better way to show the relationship between segments is with a tree structure. A tree graphically shows how segments can use virtual storage at different times. It does not imply the order of execution, although the root segment is the first to receive control. Figure 2 shows the tree structure for the dependent groups shown in Figure 1. The structure has five segments and is contained in one region.

Figure 2. Single-region overlay tree structure