The solution for n-thread mutual exclusion presented by Eisenberg and McGuire65 guarantees that any single process will enter its critical section within n–1 tries in the worst case. Should one hope for better performance with n processes?

What will be an ideal response?

In the worst case, all n processes will contend to enter their critical sections at the same time. Since the bound has to apply to all n processes, and only one process can be in its critical section at a time, it would be impossible to guarantee that all processes will enter their critical sections in less than n ? 1 tries in the worst case.