Page 418 - Programming Microcontrollers in C

Delays Revisited 403

inituart(38400);
Enable_Fast_Interrupts();
vector(handler,FAST_AUTOVECTOR);
while(count++<300)
{
if((semaphore=attach_semaphore())==-1)
{
puts(“semaphore attachment error\n”);
exit(0);
}

delay(1000,semaphore);
wait_for_semaphore(semaphore);
printd(count);
putchar(‘\r’);
}
}
Listing 8-4: Delay Test Routine

The delay test routine first initializes the on-board UART so that
signals can be sent to a terminal and then enables the fast interrupts
and places the handler address in the FAST_AUTOVECTOR location
as needed. The main test attaches a semaphore, executes a delay of
1000 milliseconds, waits for the delay to expire and prints a value to
the screen. This sequence is placed in a loop that is executed 300 times.
Note that it is important that attach_semaphore() be checked
for a –1 return. If not, you could attempt to attach too many semaphores,
and your code would not catch the error. In the above case, the program
was exited with the exit() function. It is not usually necessary to
abort the program when there is no semaphore when one is needed.
You can merely wait for a semaphore with any value between 0 and 9.
When that semaphore is released, you can proceed with a request to
attach a semaphore and it should then succeed.
The attach semaphore is in the calling program in this case. The
semaphore number is sent to the delay program where it is in turn
passed to the pit interrupt service routine. When the interrupt service
routine, after the proper delay, is executed, the specified semaphore
is released. Therefore, in the calling program, the wait for semaphore
routine is executed until the semaphore is released. If needed, the
semaphore status can be polled synchronously by the program to
determine when the delay time has expired. In other words, the