1. Ch4. Processes 2015-04-15.

2. 4.1 Process Concept (1) new waiting terminate readyrunning admitted interrupt exit scheduler dispatch I/O or event completion I/O or event wait Diagram of process state Process control block pointer process state process number program counter register Memory limits List of open files

3. 4.1 Process Concept (2) Diagram showing CPU switch from process to process. process P 0 process P 1 Operating system Interrupt or system call Save state into PCB 0 reload state from PCB 1 Save state into PCB 1 reload state from PCB 0 executing idle executing idle Interrupt or system call

4. 4.2 Process Scheduling (1) F CPU utilization and higher throughput  Two processes, A and B, for execution  Process execution with multiprogramming Idle; input start Process A Idle; input Idle; input stop Idle; input start Process B Idle; input Idle; input stop Idle; input start Process P 0 Idle; input Idle; input stop Idle; input start Process P 1 Idle; input Idle; input stop

5. 4.2 Process Scheduling (2) F Process scheduling Job SubmitHOLD Ready Queue I/O Request time slice expired Fork a child Wait for an interrupt CPU I/OI/O Queue Child executes Interrupt occurs Child terminate High level scheduling Intermediate level scheduling Low-level scheduling

6. 4.2 Process Scheduling (3) F Scheduling Algorithm  High Level Scheduling by Spooler  Intermediate Level Scheduling = job scheduling = Long-term scheduler  Low Level Scheduling = process scheduling =short-term scheduler ( CPU )

7. Addition of medium-term scheduling to the queuing diagram partially executed swapped-out processes swap inswap out Ready Queue CPUEND I/O waiting Queue I/O 4.2 Process Scheduling (4)

8. 4.3 Operation on Process (1) F O.S. mechanism Process Creation Process Termination F Process Graph : Graph for the relation of several processes Example) P1P1 P3P3 P2P2 P4P4 P5P5 P6P6 P i -> P j : node P i is parent of P j P j is child of P i -> : edge means that “ P i create P j ”

9. 4.3 Operation on Process (2) F Process Creation  Execution ① Parent runs concurrently with children ② Parent waits until some or all of its children have terminated  Sharing Parent shares all common resources with child F Process Termination  Parent terminate the execution of children if  child use allocated resource excessively  task that assigned to child is over  cascading termination : if parent terminated, children terminated

10. F Process creation (parent process  childern) Fork S 1 ; Fork L ; S 2 ; L : S 3 ; Join Fork L 1 ; S 1 ; ¦ Go To L 2 ; ¦ L 1 : S 2 ; L 2 : Join Count ; S1 S2S3 Fork S1S2 S3 Join 4.3 Operation on Process (3)

11. ex) Concurrent read(a) ; read(b) ; c := a+b ; write(c) ; Count := 2 Fork L 1 ; read(a) ; Go to L 2 ; L 1 : read(b) ; L 2 : Join count ; c := a+b ; write(c) ; s1s1 s2s2 s3s3 s4s4 s6s6 s5s5 s7s7 S 1 ; count := 3 ; Fork L 1 ; S 2 ; S 4 ; Fork L 2 ; S 5 ; Go To L 3 ; L 2 : S 6 ; Go To L 3 ; L 1 : S 3 ; L 3 : Join count ; S 7 ; 4.3 Operation on Process (4)

12. 4.4 ~ 4.5. Cooperating Process / Threads F Cooperating Process: Process that shares data with other processes F Task >> Thread a basic unit of CPU utilization =Light weight process a group of thread =heavy weight process * T raditional process : a task with one thread. UNIX Kernel : Single Tasking. MACH Kernel : multithread. F Producer & Consumer Processes ex) Assembly code Object Modules F Bounded Buffer / Unbounded - Buffer Producer. CompilerAssembler Loader

13. 4.4 Cooperating Process F Concurrent processes  Independent processes : cannot affect the other processes  Cooperating processes : share data with other processes F Why cooperating?  Information sharing  Computation speedup  Modularity  convenience

14. 4.4 Cooperating Process F Represent program that copy file f  g with concurrent Stmt fork/join and parbegin/parend statement Var f, g : file of T; r, s : T; count : integer; begin reset(f); read(f,r); while not eof(f) do begin count:=2 s:=r fork L1; write(g,s); go to L2; L1:read(f,r); L2:join count; end write(g,r); end Var f, g : file of T; r, s : T; count : integer; begin reset(f); read(f,r); while not eof(f) do begin s:=r parbegin write(g,s); read(f,r); parend end write(g,r); end

15. 4.4 Cooperating process Ex) parbegin read(a); read(b); parend c:=a+b; write(c); Problem) for i=1,5 x(i)=x(i-1)+y(i-1) y(i)=x(i-1)  y(i-1) s(i)=x(i)+y(i) if s(i)>100 then a=a+1 else b=b+1 end  execute P 1 (i) and P 2 (i-1) concurrently P 1 (i) P 2 (i) P1P1 P2P2

16. PQ Send MSG Receive MSG 4.6 InterProcess Communication F Communication of Cooperating process  Shared Memory  IPC(method of Message communication) F IPC operation : Send, Receive F Consideration Issues  How many links  buffer space  Message size  unidirectional / bidirectional

17. 4.6 Interprocess Communication F Communication link F Naming  Direct Communication  Indirect Communication F direct Communication  send(P,message) : Send a message to process P  receive(Q,message) : Receive a message to process Q ex) Producer-Consumer Problem  The producer process Repeat ……. Produce an item in nextp ……. Until false;

18. 4.6 Interprocess Communication  The consumer process Repeat receive(procedure,nextc); ……. Consume the item in nextc ……. Until false;  send(P.MSG) : send a MSG to process P  receive(id,MSG) : Receive a MSG from any process

19. 4.6 Interprocess Communication F Indirect communication : by mailbox(ports)  Send(A,MSG) : send a MSG to mailbox A  Receive(A,MSG) : Receive a MSG from mailbox A F Buffering  by Queue Capacity for link  Zero Capacity : maximum length Zero  MSG system = Rendezvous

20. 4.6 Interprocess Communication  Bounded Capacity : finite length if full, sender wait  Unbounded Capacity : Sender is never delayed ex) P -- (MSG) -->Q P Q Send(Q,MSG) Receive(P,MSG) Receive(Q,MSG) Send(P,”ACK”) Asynchrouously