
C165UTAH
System Programming
Data Sheet
477
2001-02-23
Internal System Stack
A system stack is provided to store return vectors, segment pointers, and processor
status for procedures and interrupt routines. A system register, SP, points to the top of
the stack. This pointer is decremented when data is pushed onto the stack, and
incremented when data is popped.
The internal system stack can also be used to temporarily store data or pass it between
subroutines or tasks. Instructions are provided to push or pop registers on/from the
system stack. However, in most cases the register banking scheme provides the best
performance for passing data between multiple tasks.
Note:
The system stack allows the storage of words only. Bytes must either be
converted to words or the respective other byte must be disregarded.
Register SP can only be loaded with even byte addresses (The LSB of SP is
always '0').
Detection of stack overflow/underflow is supported by two registers, STKOV (Stack
Overflow Pointer) and STKUN (Stack Underflow Pointer). Specific system traps (Stack
Overflow trap, Stack Underflow trap) will be entered whenever the SP reaches either
boundary specified in these registers.
The contents of the stack pointer are compared to the contents of the overflow register,
whenever the SP is DECREMENTED either by a CALL, PUSH or SUB instruction. An
overflow trap will be entered, when the SP value is less than the value in the stack
overflow register.
The contents of the stack pointer are compared to the contents of the underflow register,
whenever the SP is INCREMENTED either by a RET, POP or ADD instruction. An
underflow trap will be entered, when the SP value is greater than the value in the stack
underflow register.
Note:
When a value is MOVED into the stack pointer, NO check against the overflow/
underflow registers is performed.
In many cases the user will place a software reset instruction (SRST) into the stack
underflow and overflow trap service routines. This is an easy approach, which does not
require special programming. However, this approach assumes that the defined internal
stack is sufficient for the current software and that exceeding its upper or lower boundary
represents a fatal error.
It is also possible to use the stack underflow and stack overflow traps to cache portions
of a larger external stack. Only the portion of the system stack currently being used is
placed into the internal memory, thus allowing a greater portion of the internal RAM to
be used for program, data or register banking. This approach assumes no error but
requires a set of control routines (see below).