Take this pseudo code,
Semaphore S <- 0
non-critical section
wait(S)
critical section
signal(S)
Does this solution to the critical section problem support mutual exclusion only?
I know that there is no freedom from deadlock since the critical section is never reached, however would that also mean that it does not support mutual exclusion.
Mutual Exclusion means that only 1 thread(or process) can enter the Critical Section. So by definition the code does not violate mutual exclusion.
The important thing to note here is that all threads, would be in busy-waiting state, and there is no progress at all, but actually there is no deadlock, because one of the 4 must occur conditions for deadlock is CIRCULAR WAIT. But its not that threads are waiting for each other in this case.
