Message Formats

This section describes the SNA Device Interface Specification interface in terms of message formats. These are presented in a language-independent notation that is described below.

The messages used between the node and the SNALinks are shown in the following table.

Details of the message format notation and key assumptions about the contents of the message formats are as follows:

• "Reserved" indicates that the field must be set to zero (for a numeric field) or all nulls (for names) by the sender of the message.
• "Undefined" indicates that the value of the field is indeterminate. The field is not set by the sender and should not be examined by the receiver of the message.
• Fields that occupy two bytes—the srci field in all messages, and fields such as opresid in Open(LINK) Request —are represented with the arithmetically most significant byte in the lowest byte address, irrespective of the normal byte order used by the processor on which the software executes. That is, the 2-byte value 0x1234 has the byte 0x12 in the lowest byte address. The exception to this is the startd and endd fields in all elements, which are always stored in the processor's normal byte order.
• Messages are composed of buffers, consisting of a buffer header and zero or more buffer elements; see Messages for more information on buffer formats.
• The startd field in each element gives the offset of the first byte of data in the element after the trpad field. Its value will either be 1 (data starts in the byte after the trpad field), 10 (nine bytes of padding are included between the trpad field and the start of the data), or 13 (12 bytes of padding are included between the trpad field and the start of the data). Any extra bytes are used by the local node for additional header information. This avoids having to copy data into a new buffer when adding this information.
• Because startd indicates the index into dataru starting from 1, not 0, the first byte of valid data will always be at dataru[startd–1].
• All fields within dataru are of type unsigned character (UCHAR), except where the notes indicate otherwise.