TCP/IP API Support › Socket Macro Interface for Assembler Programs › Notes

Notes

• The syntax does not show Assembler column conventions (label starts in column 1; statement in column 10; continuation line in column 16; continuation character in column 72).
• On return from the #SOCKET call, R15 is always 0, except in cases of a parameter-list error where the RETCODE field cannot be found; in this case R15 is set to -1.
• The parameter values assigned to the three return code parameters (RETCODE, ERRNO and RSNCODE) and to all the function-specific-parameters can be specified in data field notation or in register notation.

  In data field notation, the program specifies the name of a variable field containing the parameter value.

  In register notation, the program specifies a register containing the address of the variable field containing the parameter value (not the value itself). General registers 2 to 15 can be used in this notation; the register reference must be enclosed in parentheses.

• Some parameters also accept a value in the form of an absolute expression. Where applicable, this is mentioned under the corresponding parameter's description.
• Some parameters from the #SOCKET macro are optional. There are two ways to address an optional parameter:
  • Omit the parameter on the #SOCKET macro call.
  • Assign a null value to the parameter. For example, HOSTNAME=NULL.

  Both ways are equivalent.

• The #SOCKET macro uses the following registers when building its parameter list:
  • R0—A work register to build the parameter list.
  • R1—Address the parameter list.
  • R14 and R15—The branch and link register for the call sequence to socket services.
• #SOCKET TCPIPDEF generates DSECTs and EQUates needed to write a TCP/IP program.
• #SOCKET ERRNOS generates all EQUates for CA IDMS specific errno values.