Fields

A field corresponds to a certain number of bits of a machine instruction. The first field of an instruction stores the opcode of the instruction. The remaining fields include the operands of the instruction. There can also be 0-bit fields, which just correspond to punctuation in the assembly language format of the instruction.

Parameters:

• name
  The names of the fields have a form that is a little different than symbols in the assembly language. A field name can be either any single non-whitespace character or a mixture of two or more letters, numbers, and hyphen (-) characters.
• type
  The type is either required, optional, or ignored.
  • A required field is one that must appear as an operand in the instruction. If the field has positive length, then the assembly code must include an operand for that field consisting of a literal or a symbol whose value will fit in the field. If the field has length 0, then the name of the field must be included as an operand in the instruction in assembly language programs and is used just as punctuation. Therefore, a field of length 0 must have a name that forms a token in assembly language.
  • If a field is optional, then an operand need not be specified for that field in assembly language, in which case the field's default value is used for the operand instead. If the field has length 0 and is optional, then it may be omitted. In each machine instruction, at most one field with length > 0 can be optional.
  • An ignored field is one that is never included as an operand in assembly language. Instead, the field's default value is always used for the field. Fields of length 0 may not have this type.
• numBits
  The length or number of bits in this field. It must be a base-10 integer that is 0 or greater.
• relativity
  The relativity of a field is either absolute, pcRelativePreIncr, or pcRelativePostIncr. This relativity allows for two pc-relative addressing modes.
  • A field with absolute relativity is one whose operand's value is placed directly in the field when assembled.
  • If a field has pcRelativePreIncr relativity, then the value of the operand in the assembly program for the field is subtracted from the address of the instruction containing the field. The assembler puts the difference as the actual value of the operand in the field in the assembled instruction.
  • If a field has pcRelativePostIncr relativity, then the value of the operand in the assembly program for the field is subtracted from the address of the instruction after the one containing the field. The assembler puts the difference as the actual value of the operand in the field in the assembled instruction.
• defaultValue
  The default value must be a base-10 integer that fits in the field. It is used as the operand for fields of type ignored or for fields of type optional when operands for those fields are not included in the assembly program.
• signed
  If a field is signed, it can legally hold any value from -2^n-1 to 2ⁿ-1 where n is the number of bits in the field. If the field is unsigned, the values must be in the range 0 to 2ⁿ-1.
• values
  If a field has absolute relativity and a positive number of bits, then you can optionally create a set of named values for the field. If you create such a set, then the names for these values become the only legal values for the field. Furthermore, in this case, you must use one of the names of those values as the operand in the field in each assembly language program. If you do not specify a set of values, then any value that fits in the field is acceptable.

Examples:

Figure 23. The dialog for editing the fields of machine instructions.

Consider the fields shown in Figure 23 above. Assume that the "reg" field has a set of three values associated with it: A=2, B=3, and C=4. Also, assume that the punctuation characters '[' and ']' are individual character tokens (see the discussion of tokens for information on setting the token usage of punctuation characters). Now consider two 16-bit machine instructions with names "foo" and "bar" with opcodes 0 and 1 and with the following respective formats:

    op reg reg un7
    op delta[reg]

    foo B C
    bar -2[A]

    00000 011 100 0000101
    00001 11111110 010

Note that the delta value of -2 is converted to binary using 8-bit 2's complement notation.

If the two punctuation characters '[' and ']' are not individual tokens and instead form parts of symbols, then you need to separate them with whitespace from neighboring characters in instruction calls to avoid ambiguity or incorrect parsing. For example, you would need to write

    foo B C
    bar -2 [ A ]

Notes:

• Fields of length 0 are just punctuation and their name must appear as the operand in assembly language. Such fields must be of type required or optional. All other properties of such fields are unused.
• To preserve backward compatibility of this version of CPU Sim with previous versions that used field lengths only instead of these new fields, CPU Sim looks for fields when loading a machine previously saved to a file. If no such fields are included in the file and instead field lengths are included, then CPU Sim will add new fields with names "1" through"16" and "-16" through"-1" to the machine when it is loaded. In this way, the old field lengths become actual fields instead.