Supported data types passed by value (direct) between C and COBOL
Table 1 identifies the data types that can be passed by value (direct) as parameters between C and COBOL applications.
| C | COBOL (by value) |
|---|---|
| char | PIC X, PIC A |
| signed short int | PIC S9(4) USAGE IS BINARY |
| unsigned short int | PIC 9(4) USAGE IS BINARY |
| signed int, signed long int | PIC S9(9) USAGE IS BINARY |
| unsigned int | PIC 9(9) USAGE IS BINARY, LENGTH OF |
| unsigned long int | PIC 9(9) USAGE IS BINARY |
| float | COMP-1 |
| double | COMP-2 |
| pointer to... | POINTER, ADDRESS OF |
If the COBOL program receives int parameters from the C calling function that might have a value that is larger than that declared as the maximum size by the COBOL picture clause, the COBOL program must either be compiled with the TRUNC(BIN) compiler option or each binary data item that receives int parameters from C must be declared as USAGE IS COMP-5. Taking these actions will guarantee that truncation of high-order digits does not occur. For more information about the TRUNC(BIN) compiler option or about using COMP-5 data items, see the appropriate version of the programming guide in the Enterprise COBOL for z/OS library.