140/317
6 - STMicroelectronics Programming Tools
Thus the assembler relieves the programmer from any calculation. Any change in the program
will be automatically accounted for when the source text is re-assembled.
6.1.4 Declaring constants
A constant, in assembler, may be of one of two kinds: the constant data, and the symbol def-
inition. In both cases, they are numeric (or string) values that are defined in the source text,
and remain unchanged for the whole life of the program.
6.1.4.1 Constant data
Constant data are similar to variables, in that they take up some bytes in memory to hold data.
The difference is that the data are defined in the source text, and cannot be changed while the
program is executing. For so, they are located in read-only memory (ROM). They are ac-
cessed in exactly the same way as variables. The value of these memory locations are de-
fined in the source text. Thus, a special pseudo-op is available that both reserves memory and
sets it to a user-defined value.
Syntax: The constant data are define using one of the pseudo-ops
DC.B, DC.W, DC.L, BYTE,
WORD, LONG and STRING. The DC pseudo-ops work like the DS pseudo-ops above, but in addi-
tion, they set the memory to the value of the operand field. Example:
.PowerOf2 DC.B 1,2,4,8,16,32,64,128 ; powers of 2
The following instruction reads one byte from the table according to the value of the index X.
The input value being loaded first into
X,thevalueinA after the execution is 2 raised to the
power
X (1 for X=0, 2 for X=1, 4 for X=2, etc.):
LD A, PowerOf2,(X)
The pseudo-ops BYTE, WORD and LONG are similar to DC.B, DC.W and DC.L, respectively, but
with an important difference. When word and long data are stored in memory, it is important to
take care of the order of the bytes that make up a word or a long value. For example, the hex-
adecimal number
1234h can be stored in bytes of increasing addresses either as the se-
quence
12h, 34h, or as the sequence 34h, 12h. The same applies to long values that are
stored in four bytes. Either method has its advantages and drawbacks, and in fact the market
is divided into the proponents of one method or the other. The Motorola style is to put the most
significant byte first, that is
34h, 12h. The Intel style is the opposite, i.e. 12h, 34h. The ST7 fol-
lows the Motorola style in its instruction syntax: when extended addressing mode is used, the
first byte of the address is the most significant one. The
JP instruction, using the long indexed
mode, reads the destination address in memory starting with the most significant byte. If you
To Next Page
Loading...
