Structure of the module map

A program module map consists of a small header followed by fixed length records. These records represent segments (G), classes (C), sections (S), entry points or LD records (E) and parts (P). The fixed length records are followed by a string area containing null-terminated symbols. At the end of the module map is the module map locator. The locator is a 24 byte area that contains an eyecatcher (IEWBMMP), which may be the first or second doubleword, followed by a doubleword containing the offset to the beginning of the module map from the beginning of the segment (or zero if the map is in a NOLOAD class).

Note: The IEWBMMP macro is shipped in the SYS1.MACLIB member IEWBMMP, which contains the mapping description for assembler programs in Accessing program object class information.

All sections and classes (that is, all elements) and parts resident in initial load and deferred load classes have entries in the map. Segment numbers are stored in all entries.

Unnamed sections are included in the map, but unnamed entry points and parts are omitted. Unnamed entities are those whose named are printed as $PRIV****** in the binder sysprint map. C++ function or variable names are the complete mangled name. Entry records include amode, xplink/nonxplink and code/data attributes. The table contains offsets only. There are no adcons. This choice enables the table to be stored in a NOLOAD class, if desired.

The entries are ordered by segment and then offset within segment.

Entries are organized in a hierarchy, segment/class/section/entry or segment/class/section/part. For each class, there is a class entry followed by entries for all sections that contribute to that class. These section entries correspond to ED ESD entries in the module. Each section entry is followed by entries for the entry points (corresponding to LD ESD records) and parts (corresponding to PD ESD records) in that section and class.