Filter
Top Products
Calling Non-BASIC-80 Subroutines
E-2
When you have determined the optimum starting address for your subroutine, you
can
LOCATE
it
there with this command. LOCATE converts the relocatable object
code to absolute object code, according to the starting address given. An example
of
giving the starting address for your subroutine code
is
shown below.
LOCATE
:F1
:PLMSUB.OBJ
TO
:F1
:PLMSUB.LD
CODE
(OEECOH)
This example converts the relocatable object code in
:Fl
:PLMSUB.OBJ to absolute
object code in the
output
file
:Fl
:PLMSUB.LD, and makes all addresses in the
subroutine relative
to
OEECOH. Refer to the ISIS-II User's Guide for further details
of using LOCATE.
After
LOCATE has converted your relocatable code to absolute object code, you
can load it into memory with the
ISIS-II Monitor DEBUG command. When you
enter
DEBUG followed
by
the filename
of
the load module, the subroutine
is
loaded
into memory at the address specified in the load module. The starting address
displays
on
your terminal. You then enter GS, which returns you to ISIS-II with the
subroutine loaded
at
the specified address. From here.you can invoke
BASIC-SO
and
call the subroutine.
NOTE
You must give the MEMTOP option to reserve memory each time
BASIC-SO
is
invoked.
Calling Subroutines
After a subroutine
is
loaded into memory, you can call it from BASIC-80. First,
invoke
BASIC-SO as you normally do, and give the MEMTOP option as previously
specified.
The first step in calling a subroutine
is
defining its address with the DEFUSR func-
tion. Up to
25
subroutines can be addressed in this way, with an integer in the range
0-24. The starting address
of
the subroutine
is
given in hexadecimal:
DEFUSR5=OEECOH
Once BASIC-SO knows where USR5
is
located, you can call it. When you call it, you
must supply any needed variables.
Since the protocol for passing parameters follows
PL/M
conventions, you can only directly pass 2-byte integer variables or 2-byte
addresses.
If
you specify a floating-point variable
or
a string variable you must use
the
V
ARPTR
function to pass the address
of
the desired variable.
For
example, to
pass the addresses
of
two floating-point numbers (Kl and
K2):
120
A = USR5 (VARPTR(K1), (VARPTR(K2»
BASIC-SO
goes to the address where
Kl
is
stored (VARPTR(Kl) and the address
where
K2
is
stored (VARPTR(K2». Once it has found these two values, it passes
them to the subroutine.
If
BASIC-SO
encounters new variables after executing line
120
above, the memory
locations
of
Kl
and
K2
can change, causing errors.
Be
sure that all variables are
defined before using the VARPTR instruction.
In the example shown, once the parameters are passed, the subroutine executes.
Because
of
PL/M-80
calling conventions BASIC-80 expects the returned result to be
a two-byte integer in the
HL
register-pair, and assigns this 16-bit value to A.
BASIC-80