Page 74 - DCAP507_SYSTEM_SOFTWARE

System Software

Notes 3. MOT_TABLE Search the Machine-Op Table (MOT) for a match with the operation of
the current source card.
4. SYM_TABLE Store a label and its associated value into the Symbol Table (ST). If the
symbol is already in the table, return error indication (multiply defined
symbol).

5. LIT_TABLE Store a literal into the Literal Table (LT); do not store the same literal
twice.
6. WRITE1 Write a copy of the assembly source record on a storage device for use
by pass 2.
7. DLENGTH Scan operand field of space consuming pseudo-ops to determine the
amount of storage required.
8. EVAL Evaluate an arithmetic expression consisting of constants and symbols
(e.g. 6, ALPHA, BETA+14*GAMMA).

9. STGET Search the Symbol Table (ST) for the entry corresponding to a specific
symbol (used by SYM_TABLE, and EVAL).
10. LIT_LC Assign storage locations to each literal in the literal table (may use
DLENGTH).

Pass 2

1. READ2 Read the next assembly source record from the file copy.
2. POT_TABLE Same as in pass 1
3. MOT_TABLE Same as in pass 1 (evaluate expressions).

4. EVAL Same as in pass 1 (evaluate expressions).
5. OBJECT Transform produced instruction to object record format; write the
record when it is filled with data.

6. PRINT Transform relative location and produced code to character format;
print the line along with copy of the source card.
7. DATAGEN Process the fields of the data-generating pseudo-op to produce object
code (uses EVAL).
8. DLENGTH Similar as in pass 1.
9. LIT_GEN Produce code for literals (uses DATAGEN).

Each of these functions should separately go via the whole design process (problem statement,
data basics, algorithm, modularity, etc.). These functions can be executed as disconnected external
subroutines, as internal subroutines, or as sections of the pass 1 and pass 2 programs. In any case,
the aptitude to tract functions separately makes it much simpler to design the structure of the
assembler and each of its parts. So, instead of observing the assembler as a single program (of
1,000 to 10,000 source statements), we examine it as a coordinated collection of routines each of
relatively minor size and complexity. We will not attempt to scrutinize all of these functional
routines in detail since they are quite uncomplicated. There are two particular observations of
interest:
1. Many of the routines involve the scanning or evaluation of fields (e.g., DLENGTH, EVAL,
DATA_GEN);

68 LOVELY PROFESSIONAL UNIVERSITY