SLAU132Y September 2004 – June 2021
Similarly to code sections described in the previous section, data can either be placed in a single section or multiple sections. The benefit of multiple data sections is that the linker may omit unused data structures from the executable. This option causes aggregate data—arrays, structs, and unions—to be placed in separate subsections of the data section.
If this option is not used, the default is "on". If this option is used but neither "on" nor "off" is specified, an error message is provided.
If the SET_DATA_SECTION pragma is used, the --gen_data_subsections=on option is ignored. User-defined section placement takes precedence over automatic generation of subsections.