Suppose
a multiprocessor system has 8 equal capability processors. The OS supports multi-threading
by providing up to 8 kernel threads per application program and uses
many-to-many model to map user threads to kernel threads. Suppose a
multithreaded application with 12 threads is currently running on the system.
Assuming there is no other process on the system, and hence no context switch
overhead, do you think 100% processor utilization is achievable throughout the
life of this process? It is also assumed that not all kernel threads can block
at the same time. Explain.
Field of study:
No answers yet