ZSM, a CP/M Z80 assembler.

ZSM is an assembler for the Z80 microprocessor. It uses standard

Zilog/Mostek mnemonics, supports conditional assembly and
produces Intel HEX file suitable for loading with the CP/M LOAD
utility. ZSM is a disk based assembler, i.e. source code is read
from disk, the resulting object code is written to disk and an
optional print file can be written to disk. Alternatively the
print file can be redirected to the console.
1. Using the assembler.
----------------------The assembler is initiated by typing:
ZSM filename.pqr
This takes the file 'filename.ZSM' and assembles it according to
the parameters 'p','q' and 'r' and the options chosen. The
parameters must be selected as follows:
p

- the disk drive containing the source

'filename.ZSM'. A to P or @ for default drive.

code,

q - the disk drive where the object code, 'filename.HEX' is

to be written. A to P or @ for default drive, Z if not
required.
r

- the disk drive where the listing is to be written,

'filename.PRN'. A to O or @ for default drive, P to
the list device, X to the console, Y to the console
with errors to the pinter, Z to suppress listing.

For example assembling the file TEST.ZSM:ZSM TEST.AAB

This expects the file TEST.Z80 on drive A, will write the object
code file, TEST.HEX to drive A and the listing file, TEST.PRN to
drive B.

2. Assembler syntax
------------------2.1 ASSEMBLER MNEMONICS
The assembler recognises all standard Z80 mnemonics as defined in
the Z80 CPU Technical Manual. The following mnemonics are
recognised in both forms shown.
ADC
ADD
ADD
ADD
ADD
ADD

A,s
A,n
A,r
a,(HL)
a,(IX+d)
a,(IY+d)

ADC
ADD
ADD
ADD
ADD
ADD

s
n
r
(HL)
(IX+d)
(IY+d)

2.2 LABELS
Labels may be up to 11 characters long. The first character of
the label must be alphabetic (i.e. A-Z) or '_' or '$'. A colon,
':' following the label is optional.
2.3 CONSTANTS
ZSM allows binary, octal, decimal, hexadecimal and ASCII
constants according to the following conventions:
Binary - number formed from binary digits (0,1) and terminated by
the character 'B'. Range 0000000000000000B to 1111111111111111B.
Example:

LD HL,10001111101101B

Octal - number formed from octal digits (0-7) and terminated by

the character 'Q'. Range 000000Q to 177777Q.
Example:

LD HL,23670Q

Decimal - number formed from decimal digits (0-9) and EITHER left
unterminated or terminated by the character 'D'.
Range: 0 to 65535.
Example:

LD HL,32145

Hexadecimal - number formed from hexadecimal digits (0-9 and A-F)

and terminated by the character 'H'. A hex number beginning with
a letter must be preceded by a 0 to distinguish it from a label
or register name. Range 0000H to FFFFH.
Example:

LD HL,0A3FH

ASCII - number formed from ASCII value of characters enclosed in

single quotes. Range ' ' to '~' or 20H to 7EH including all
alphanumerics and punctuation.
Examples:

LD HL,'Wk'
LD A,'*'

The $ sign used as a constant represents the current value of the

program counter.

2.4 EXPRESSION EVALUATION

Operands in instructions and pseudo-ops may consist of arithmetic
expressions which are evaluated at assembly time and the
calculated value used as part of the object code. The following
operators may be used to form expressions:
+
*
/
.MOD.
< or .GT.
.GE.

Addition
Subtraction (or unary Negation)
Multiplication
Division
Modulus
Greater Than. X.GT.Y true if X greater than Y.
Greater than or Equal. X.GE.Y true if X
greater than or equal to Y.
Less Than. X.LT.Y true if X less than Y.
Less than or Equal. X.LE.Y true if X less
than or equal to Y.
Equal. X.EQ.Y true if X equals Y.
Not Equal. X.NE.Y true if X not equal to Y.
Shift Left n bits.
Shift Right n bits.
Logical NOT (unary)
Logical AND
Logical OR
Logical Exclusive OR
Low byte of expression (.MOD. 256)
High byte of expression (.SHR. 8)

> or .LT.
.LE.
= or .EQ.
.NE.
.SHL. n
.SHR. n
.NOT.
.AND.
.OR.
.XOR.
.LOW.
.HIGH.

Expressions are evaluated left to right with no precedence. No

parentheses can be used in expressions so some care is required
to ensure that expressions are correctly formed. Unary
expressions .NOT., .LOW. and .HIGH. must be at the beginning of
an expression and only act on the next parameter in the
expression. For example:
DB

.LOW. 1234H + 50

is evaluated as (.LOW. 1234H) + 50, not .LOW.(1234H + 50).

2.5 PSEUDO-OPCODES
DB or DEFB (define byte)
The DB pseudo-op tells the assembler to reserve a byte or string
of bytes with specific values. The bytes may be defined as
constants, expressions as described earlier or as already defines
symbols.
<label>: DB

<item or list of items>

For example:
DB

3,-5,34+LABEL,'A'

DS or DEFS (define storage)

This tell the assembler to reserve a specified number of bytes
for storage. Note that the data stored in the bytes is not
defined.
DW or DEFW (define word)
This is exactly the same as the DB Pseudo-op except that the
data is stored in words (2 bytes each) rather than bytes.
END (end of assembly)
This pseudo-op is used to terminate assembly of a block of source
code.
ENDIF (end of IF definition)
Used to terminate a block of conditionally assembled code. See IF
Pseudo-op for further details.
EQU (equate)
The EQU pseudo-op assigns a value to a label or symbol. The
format is:
label:

EQU <item>

Label is the name of the symbol and item is any of: a constant,
an address, a label or an expression. Once the value of the
symbol is assigned it is in effect for the entire source program.
FORM (form feed)
Forces a new page in the listing.
IF (begin conditional assembly)
The IF, ELSE and ENDIF pseudo-ops define a sequence of assembly
language statements which are to be included or excluded during
the assembly. The form is:
IF <expression>
statement #1
statement #2
.....
statement #n
[ELSE
.....
.....
]
ENDIF
Conditionals may be nested up to 8 levels deep. Values used in
the expression must be defined before the IF statement. When the
assembler finds an IF pseudo-op the expression is evaluated. If
the expression evaluates to a logical TRUE (non-zero) then the
statements following, up to the ENDIF are included in the
assembly. The optional ELSE pseudo-op allows alternate code to be
generated when the opposite condition exists. An ELSE is always
related to the most recent active IF.

IF

<expr1> -----------+
|
IF <expr2> ------+
|
|
|
ELSE ---------------+
|
|
|
ENDIF --------------+
|
|
IF <expr3> -------+
|
|
|
ENDIF --------------+
|
|
ENDIF -------------------+
LIST (use following options in listing)
This is used to control what parts of the assembly listing get
sent to the listing file. The word LIST is followed by a string
of option word separated by commas.
LIST OFF
Suppresses the listing until a LIST ON command.
LIST ON (default)
Turns on the listing file after a LIST OFF statement.
LIST COND (default)
Forces generation and printing of all blocks of code which
are part of conditional assemblies whether they are
assembled or not.
LIST NOCOND
Suppresses listing of IF, ELSE and ENDIF statements and any
conditionally assembled code which is NOT included in the
assembly, i.e. only the code actually assembled is shown in
the listing.
LIST SYMBOL (default)
Generate a sorted symbol table at the end of the listing
file.
LIST NOSYMBOL
Suppress printing of a sorted symbol table at the end of the
listing file.
LIST TABS (default)
Tab characters (ASCII 9) are to be used in the listing file.
LIST NOTABS
Do not use tabs in the listing file.

2.6 COMMENTS
Remarks or comments are free format including any printable ASCII
character as long as the comment is preceded by a semicolon, ';'.
The comment may follow an opcode, operand or label or may exist
on a line by itself.
3. Error Messages
ZSM generates a number of error messages while assembling to
inform you of its progress. They fall into two categories: errors
found while setting up files for assembly operation and those
encountered while assembling the file. The first type of error
message will be one of the following:
Source (.ZSM) file not found.
Unable to create/open object (.HEX) file.
Unable to create/open output (.PRN) file.
The first message is generated when the filename specified in the
command line cannot be found on the selected disk. The other
messages occur while the output (HEX and PRN) files are being set
up, usually the disk or directory is full. These errors are sent
only to the console and the assembly is aborted.
The following errors can occur during the assembly process. The
offending source line is displayed on the console, followed by
the error message. If a listing file has been selected the error
message appears in the listing on the line following the error.
Errors are displayed even if the listing has been disabled using
the 'LIST OFF' pseudo-op.
"Too many IF statements"
Only 8 levels of nested IF statements are allowed
"ENDIF without matching IF statement"
IF and ENDIF should always be paired to indicate the start
and end of a section of conditional code.
"ELSE without matching IF statement"
The ELSE pseudo-op should only appear within a section of
conditional code between IF and ENDIF.
"Relative jump range error"
The address specified in a relative jump instruction is out
of the range -128 to +127.
"Expression error"
This error is caused by a badly formed expression. Note that
brackets are not allowed in expressions.
"Illegal opcode"
The assembler cannot recognise the instruction.
"Multiple definition"
The symbol has already been defined.

"Syntax error"
This error message covers a multitude of sins. It is usually
caused by misspelling or omission of a delimiter such as a
comma or semicolon.
"Undefined symbol"
The symbol used in the instruction or expression has not
been defined.
"Value error"
This error appears when the value of a symbol or expression
exceeds the range allowable. For example an index register
displacement value must be in the range +127 to -128.
"Symbol Table Overflow"
"Divide by Zero"
The expression being evaluated contains a division by zero.