Ch. 7 Process Synchronization (1/2)

7.I Background F Producer - Consumer process :  Compiler, Assembler, Loader, · · · · · · F Bounded buffer :  Assume that there is a fixed buffer size.  Consumer must wait if the buffer is empty.  Producer must wait if the buffer is full F Unbounded buffer In = out > empty Next free position Full : in + 1 = out first full position

7.I Background F Producer while(1){ /* produce an item */ while (counter == BUFFER_SIZE) ; buffer[in] = nextProduced; in = (in + 1) % BUFFER_SIZE counter++; } F Consumer while(1){ while (counter == 0) ; nextConsumed = buffer[out]; out = (out + 1) % BUFFER_SIZE counter--; /* consume the item */ }

7.I Background F Concurrent processing : base of multi-programmed O.S.  Concurrent access of shared data raise inconsistency  synchronization(process) needed example ) T 0 : producer execute register 1 := counter T 1 : producer execute register 1 := register T 2 : consumer execute register 2 := counter T 3 : consumer execute register 2 := register T 4 : consumer execute counter := register 1 T 5 : producer execute counter := register 2 In case of T 4 and T 5, ‘count’ value will be inconsistent

7.2 Critical Section Problem F Critical Section : Cooperating protocol  Critical Section must satisfy the following three requirement  Mutual Exclusion  Process P i is executing in Critical Section, other can’t be executing in Critical Section  Progress  If no process is executing in its critical section and some processes wish to enter their critical section, then  Only those process that are not executing in the Remainder can participate in the decision on which will enter its critical section  Bounded waiting  There must exist a bound on the number of times that other processes are allowed to enter their critical section after a process has made a request to enter the Critical Section and that request is granted  Critical-section problem  design a protocol that the processes can use to cooperate  For each process  assume that executing at a nonzero speed  no assumption concerning the relative speed

7.2 Critical Section Problem  Generate Structure of Concurrent process  General structure of a typical process P i do { Common variable declaration; parbegin; p 0 ; p 1 ; parend; } while do { entry section Critical Section exit section Remainder section } while(1);

7.2 Critical Section Problem F Two process solution  restrict only two processes at a time  Algorithm 1 do { while ( turn != i ); critical section turn = j; remainder section } while (1)

7.2 Critical Section Problem  Algorithm 1  ensure that only one process at a time can be execute in Critical Section  does not satisfy the progress requirement  let process share a common integer variable turn  turn initialized to 0 or 1  if turn == i then process P i is allowed to execute in Critical Section  Problem  if turn == 0 and P 1 is ready to enter in C.S, P 1 can’t enter even though P 0 may be in remainder section

7.2 Critical Section Problem  The Structure of process P i in algorithm 1 do { while (turn != i ) ; Critical Section turn = j ; Remainder Section } while (1);

7.2 Critical Section Problem  Algorithm 2  replace variable turn (in Algorithm 1) with following array boolean flag[2];  elements of the array are initialized to false  flag[i] is true : P i is ready to enter the Critical Section  P i first sets flag[i] to be true, P i checks to verify that P j is not also ready to enter the Critical Section  if P j were ready, then P i wait until P j exit the Critical Section  mutual-exclusion requirement is satisfied  does not satisfy progress requirement

7.2 Critical Section Problem  The structure of process P i in algorithm 2 do { flag[i] = true; while ( flag[j] ) ; Critical Section flag[i] = false ; Remainder Section } while (1);

7.2 Critical Section Problem  Problem  consider the following execution sequence T 0 : P 0 sets flag[0] = true T 1 : P 1 sets flag[1] = true  means that P 0 and P 1 want to enter the C.S at the same time  P 0 and P 1 are looping forever in their respective while statements

7.2 Critical Section Problem  Algorithm 3  combining the key ideas of algorithm 1 and 2  All three requirements are satisfied  The process share next two variable boolean flag[2]; int turn;  initially flag[0] = flag[1] = false  if both process try to enter at the same time, turn will be set to both i and j at roughly at the same time  eventual value of turn decides which of the two processes is allowed to enter its critical section first

7.2 Critical Section Problem  The structure of process P i in algorithm 3 do { flag[i] = true; turn = j; while (flag[j] and turn == j) ; Critical Section flag[i] = false ; Remainder Section } while (1) ;

7.2 Critical Section Problem F Multiple-Process Solutions  solving the critical section problem for n processes  known as the bakery algorithm  each customer receives a number  The customer with the lowest number is served next  can’t guarantee that two process do not receive the same number  if P i and P j receive the same number and if i < j then, P i served first  Common data structures boolean choosing [n]; int number [n];  initialized to false and 0  Define the following notation (a, b) < (c, d) if a < c or if a = c and b < d. max(a 0, …, a n-1 ) is a number; k, such that k  a i for i = 0, …, n-1

7.2 Critical Section Problem  The structure of process P i in the bakery algorithm do { choosing[i] = true; number[i] = max(number[0], number[1], …, number[n-1]) +1; choosing[i] = false; for ( j = 0 ; j < n-1 ; j++ ) { while ( choosing[j] ) ; while ( number[j]  0) && (number[j], j) < (number[i], i) ; } Critical Section number[i] = 0 ; Remainder Section } while (1);

7.3 Synchronization Hardware F Critical Section problem  could be solved  disabling interrupts on a multiprocessor  The message passing delay entry into each critical section  system efficiency decrease F Test-and-Set  this instruction is executed atomically = uninterruptible unit  if two Test-and -Set instructions are executed simultaneously, they will be executed sequentially in some arbitrary order  can implement mutual exclusion by declaring a Boolean variable lock ( initialized to false )

7.3 Synchronization Hardware  Definition of TestAndSet instruction  Mutual-exclusion implementation with TestAndSet boolean TestAndSet(boolean &target) { boolean rv = target; target = true; return rv; } do { while ( TestAndSet ( lock ) ) ; Critical Section lock = false; Remainder Section } while (1)

7.3 Synchronization Hardware F Swap  swaps the contents of two words, atomically  can implement mutual exclusion by declaring global variable lock ( initialized to false )  does not satisfy the bounded-waiting requirement  Definition of Swap instruction void Swap(boolean &a, boolean &b ) { booleantemp = a; a = b; b = temp; }

7.3 Synchronization Hardware  Mutual-exclusion implementation with the Swap do { key = true; while (key == true) Swap(lock, key); Critical Section lock = false; Remainder Section } while (1);

7.3 Synchronization Hardware F An algorithm that uses the Test-and-Set  satisfied all the critical section requirements  common data structure boolean waiting[n] boolean lock  this data structure are initialized to false

7.3 Synchronization Hardware  Bounded-waiting mutual exclusion with Test-and-Set do { waiting[i] = true; key = true; while (waiting[i] && key) key = TestAndSet(lock); waiting[i]= false; Critical Section j = (i+1) % n; while(( j != i) && (!waiting[j])) j = (j+1) % n; if (j == i) lock = false else waiting[j] = false; Remainder Section until false

7.4 Semaphores  The solutions in Section 7.3 are not easy to generalize to more complex problems  Synchronization tool (Semaphore)  semaphore S : integer variable  standard atomic operations : P(for wait, to test), V(for signal, to increment)  Modification to S in operations must be executed indivisibly  The testing of S(S  0) and modification(S:=S-1) must be executed without interruption wait(S) : while S  0 do no-op; S = S - 1 signal(S) : S = S + 1

7.4 Semaphores F Usage  use semaphore to deal with the n-process critical-section problem  The n processes share a semaphore, mutex initialized to 1  Ex) two concurrently running processes  P 1 with a statement S 1, P 2 with a statement S 2  S 2 be executed only after S 1 has completed  P 1 and P 2 share a semaphore synch, initialized to 0 S 1 ; signal(synch); wait(synch); S 2 ; P1P1 P2P2

7.4 Semaphores F Implementation  The main disadvantage of semaphore is busy waiting  while a process in critical section, any other process must loop continuously in the entry code  Busy waiting wastes CPU cycles  This type of semaphore is called a spinlock  The advantage of a spinlock  No context switch is required when a process must wait on a lock  Mutual-exclusion implementation with semaphores repeat wait(mutex); signal(mutex); critical section remainder section until false;

7.4 Semaphores F Implementation  To overcome the busy waiting, modify the wait and signal  wait operation  Rather than busy waiting, process can block itself  place a process into a waiting queue  the state of process is switched to the waiting state  control is transferred to the CPU scheduler  scheduler selects another process to execute  signal operation  blocked process is restarted by wakeup operation when some other process executes a signal operation  wakeup change process from waiting state to ready state  process is placed in the ready queue

7.4 Semaphores type struct { int value; struct process *L; } semaphore; void wait (semaphore S) { S.value--; if (S.value <0) { add this process to S.L; block(); } void signal (semaphore S) { S.value++; if (S.value <=0) { remove a process P from S.L; wakeup (P); }

7.4 Semaphores F Deadlocks and Starvation  Deadlock : two or more processes are waiting indefinitely for an event that can be caused by only one of the waiting processes  Ex) two processes P 0, P 1, two semaphore S, Q (value 1)  when P 0 executes wait(Q), it must wait until P 1 executes signal(Q)  when P 1 executes wait(S), it must wait until P 0 executes signal(S)  Starvation(indefinite blocking)  processes wait indefinitely within the semaphore P 0 wait(S) wait(Q) signal(S); signal(Q); P 1 wait(Q) wait(S) signal(Q); signal(S);

7.4 Semaphores F Binary Semaphores  Counting semaphore : semaphore described in the previous sections  Binary semaphore : a semaphore with an integer value that can range only between 0 and 1  implement a counting semaphore using binary semaphore  S : a counting semaphore binary-semaphore S1, S2; int C;

7.4 Semaphores F Binary Semaphores  Initially S1 = 1, S2 = 0  C is set to the initial value of the counting semaphore S  wait, signal operations on the counting semaphore S can be implemented as follows: wait(S1); C --; if (C  0) { signal(S1); wait(S2); } signal(S1); wait(S1); C ++; if (C  0) signal(S2); else signal(S1); wait operationsignal operation