Processes

Outline
 Process Concept
 Process Scheduling
 Operations on Processes
 Interprocess Communication
 IPC in Shared-Memory Systems
 IPC in Message-Passing Systems
 Process Concept
 An operating system executes a variety of programs that run as a process.
 Process – a program in execution; process execution must progress in sequential
fashion. No parallel execution of instructions of a single process
 Multiple parts
 The program code, also called text section
 Current activity including program counter, processor registers
 Stack containing temporary data
 Function parameters, return addresses, local variables
 Data section containing global variables
 Heap containing memory dynamically allocated during run time
 Process Concept (Cont.)
 Program is passive entity stored on disk (executable file); process is active
 Program becomes process when an executable file is loaded into memory
 Execution of program started via GUI mouse clicks, command line entry of its
name, etc.
 One program can be several processes
 Consider multiple users executing the same program
Process in Memory
Memory Layout of a C Program
 Process State
 As a process executes, it changes state
 New: The process is being created
 Running: Instructions are being executed
 Waiting: The process is waiting for some event to occur
 Ready: The process is waiting to be assigned to a processor
 Terminated: The process has finished execution
Diagram of Process State
 Process Control Block (PCB)
Information associated with each process(also called task control block)

 Process state – running, waiting, etc.

 Program counter – location of instruction to next execute
 CPU registers – contents of all process-centric registers
 CPU scheduling information- priorities, scheduling queue
pointers
 Memory-management information – memory allocated to
the process
 Accounting information – CPU used, clock time elapsed
since start, time limits
 I/O status information – I/O devices allocated to process,
list of open files
 Threads
 So far, process has a single thread of execution
 Consider having multiple program counters per process
 Multiple locations can execute at once
 Multiple threads of control -> threads

 Must then have storage for thread details, multiple program counters in
PCB
 Explore in detail in Chapter 4
 Process Scheduling
 Process scheduler selects among available processes for next
execution on CPU core
 Goal -- Maximize CPU use, quickly switch processes onto CPU
core
 Maintains scheduling queues of processes
 Ready queue – set of all processes residing in main memory, ready
and waiting to execute
 Wait queues – set of processes waiting for an event (i.e., I/O)
 Processes migrate among the various queues
Ready and Wait Queues
Representation of Process Scheduling
CPU Switch From Process to Process
A context switch occurs when the CPU switches from
one process to another.
Context Switch
 When CPU switches to another process, the system must save the
state of the old process and load the saved state for the new
process via a context switch
 Context of a process represented in the PCB
 Context-switch time is pure overhead; the system does no useful
work while switching
 The more complex the OS and the PCB  the longer the context
switch
 Time dependent on hardware support
 Some hardware provides multiple sets of registers per CPU 
multiple contexts loaded at once
Operations on Processes
 System must provide mechanisms for:
 Process creation
 Process termination
Process Creation
 Parent process create children processes, which, in turn create
other processes, forming a tree of processes
 Generally, process identified and managed via a process identifier
(pid)
 Resource sharing options
 Parent and children share all resources
 Children share subset of parent’s resources
 Parent and child share no resources
 Execution options
 Parent and children execute concurrently
 Parent waits until children terminate
 Process Creation (Cont.)
 Address space
 Child duplicate of parent
 Child has a program loaded into it
 UNIX examples
 fork() system call creates new process
 exec() system call used after a fork() to replace the process’ memory
space with a new program
 Parent process calls wait()waiting for the child to terminate
A Tree of Processes in Linux
C Program Forking Separate Process
Process Termination
 Process executes last statement and then asks the operating system to
delete it using the exit() system call.
 Returns status data from child to parent (via wait())
 Process’ resources are deallocated by operating system
 Parent may terminate the execution of children processes using the
abort() system call. Some reasons for doing so:
 Child has exceeded allocated resources
 Task assigned to child is no longer required
 The parent is exiting, and the operating systems does not allow a child to
continue if its parent terminates
 Process Termination

 Some operating systems do not allow child to exists if its parent has terminated. If a
process terminates, then all its children must also be terminated.
 cascading termination. All children, grandchildren, etc., are terminated.
 The termination is initiated by the operating system.
 The parent process may wait for termination of a child process by using the
wait()system call. The call returns status information and the pid of the terminated
process
pid = wait(&status);
 A zombie process is a process that has completed
execution (exited) but still has an entry in the process table
because its parent process hasn't yet "claimed" its exit status.
 If parent terminated without invoking wait(), process is an
orphan
 Interprocess Communication
 Processes within a system may be independent or cooperating
 Cooperating process can affect or be affected by other processes, including
sharing data
 Reasons for cooperating processes:
 Information sharing
 Computation speedup
 Modularity
 Convenience
 Cooperating processes need interprocess communication (IPC)
 Two models of IPC
 Shared memory
 Message passing
Communications Models
(a) Shared memory. (b) Message passing.
Producer-Consumer Problem
 Paradigm for cooperating processes:
 producer process produces information that is consumed by a consumer
process
 Two variations:
 unbounded-buffer places no practical limit on the size of the buffer:
 Producer never waits
 Consumer waits if there is no buffer to consume

 bounded-buffer assumes that there is a fixed buffer size

 Producer must wait if all buffers are full
 Consumer waits if there is no buffer to consume
 IPC – Shared Memory

 An area of memory shared among the processes that wish to

communicate
 The communication is under the control of the users processes not the
operating system.
 Major issues is to provide mechanism that will allow the user processes
to synchronize their actions when they access shared memory.
 Bounded-Buffer – Shared-Memory Solution

 Shared data
#define BUFFER_SIZE 10

typedef struct {

. . .

} item;

item buffer[BUFFER_SIZE];

int in = 0;

int out = 0;

 The buffer is empty when in = =out; the buffer is full when ((in + 1) %
BUFFER SIZE) = = out.
 Producer Process – Shared Memory

item next_consumed;
item next_produced;

while (true) {
while (true) {
while (in == out)
/* produce an item in next produced */
; /* do nothing */
while (((in + 1) % BUFFER_SIZE) == out) next_consumed = buffer[out];
; /* do nothing */ out = (out + 1) % BUFFER_SIZE;
buffer[in] = next_produced;
in = (in + 1) % BUFFER_SIZE; /* consume the item in next
consumed */
}
}
 IPC – Message Passing
 Processes communicate with each other without resorting to
shared variables

 IPC facility provides two operations:

 send(message)
 receive(message)

 The message size is either fixed or variable

 Message Passing (Cont.)
 If processes P and Q wish to communicate, they need to:
 Establish a communication link between them
 Exchange messages via send/receive
 Implementation issues:
 How are links established?
 Can a link be associated with more than two processes?
 How many links can there be between every pair of communicating
processes?
 What is the capacity of a link?
 Is the size of a message that the link can accommodate fixed or variable?
 Is a link unidirectional or bi-directional?
Implementation of Communication Link

 Physical:
 Shared memory
 Hardware bus
 Network
 Logical:
 Direct or indirect
 Synchronous or asynchronous
 Automatic or explicit buffering
Direct Communication
 Processes must name each other explicitly:
 send (P, message) – send a message to process P
 receive(Q, message) – receive a message from process Q
 Properties of communication link
 Links are established automatically
 A link is associated with exactly one pair of communicating processes
 Between each pair there exists exactly one link
 The link may be unidirectional, but is usually bi-directional
Indirect Communication
 Messages are directed and received from mailboxes (also referred to as
ports)
 Each mailbox has a unique id
 Processes can communicate only if they share a mailbox
 Properties of communication link
 Link established only if processes share a common mailbox
 A link may be associated with many processes
 Each pair of processes may share several communication links
 Link may be unidirectional or bi-directional
 Indirect Communication (Cont.)
 Operations
 Create a new mailbox (port)
 Send and receive messages through mailbox
 Delete a mailbox
 Primitives are defined as:
 send(A, message) – send a message to mailbox A
 receive(A, message) – receive a message from mailbox A
 Indirect Communication (Cont.)
 Mailbox sharing
 P1, P2, and P3 share mailbox A
 P1, sends; P2 and P3 receive
 Who gets the message?
 Solutions
 Allow a link to be associated with at most two processes
 Allow only one process at a time to execute a receive operation
 Allow the system to select arbitrarily the receiver. Sender is notified
who the receiver was.
Synchronization
Message passing may be either blocking or non-blocking

 Blocking is considered synchronous

 Blocking send -- the sender is blocked until the message is received
 Blocking receive -- the receiver is blocked until a message is available
 Non-blocking is considered asynchronous
 Non-blocking send -- the sender sends the message and continue
 Non-blocking receive -- the receiver receives:
 A valid message, or
 Null message
 Producer-Consumer: Message Passing

 Producer
message next_produced;
while (true) {
/* produce an item in next_produced
*/

send(next_produced);
}

 Consumer
message next_consumed;
while (true) {
receive(next_consumed)

/* consume the item in next_consumed

*/
}
Buffering
 Queue of messages attached to the link.
 Implemented in one of three ways
1. Zero capacity – no messages are queued on a link.
Sender must wait for receiver (rendezvous)
2. Bounded capacity – finite length of n messages
Sender must wait if link full
3. Unbounded capacity – infinite length
Sender never waits
Pipes
 Acts as a conduit allowing two processes to communicate
 Issues:
 Is communication unidirectional or bidirectional?
 In the case of two-way communication, is it half or full-duplex?
 Must there exist a relationship (i.e., parent-child) between the communicating
processes?
 Can the pipes be used over a network?
 Ordinary pipes – cannot be accessed from outside the process that created
it. Typically, a parent process creates a pipe and uses it to communicate with
a child process that it created.
 Named pipes – can be accessed without a parent-child relationship.
Ordinary Pipes
 Ordinary Pipes allow communication in standard producer-consumer style
 Producer writes to one end (the write-end of the pipe)
 Consumer reads from the other end (the read-end of the pipe)
 Ordinary pipes are therefore unidirectional
 Require parent-child relationship between communicating processes

 Windows calls these anonymous pipes

Named Pipes
 Named Pipes are more powerful than ordinary pipes
 Communication is bidirectional
 No parent-child relationship is necessary between the communicating
processes
 Several processes can use the named pipe for communication
 Provided on both UNIX and Windows systems
Communications in Client-Server Systems

 Sockets
 Remote Procedure Calls
Sockets
 Used for communication in client-server environment
 A socket is defined as an endpoint for communication
 A pair of process communication over a network employ a pair of sockets ,
on for each process.
 A Socket is identified by and IP address concatenated by a port number
 The socket 161.25.19.8:1625 refers to port 1625 on host 161.25.19.8
 Server waits for incoming client requests by listening to a specified port.
Once a request is received the server accepts a connection from the client
sockets t o complete the connection.
 Servers implementing specific services (such as telnet, FTP, http) listen to
well-known ports
 All ports below 1024 are well known, used for standard services
 Socket Communication

• When a client process

initiate a request for
connection , it is assigned a
port by host computer
• This port is some arbitrary
number greater than 1024 • The packets travelling
between the hosts are
delivered to the appropriate
process based on the
destination port number
Remote Procedure Calls
 RPC is a protocol that one program can use to request a service from a
program located in a network in another computer on a network with out
having to understand the network’s details
 Remote procedure call (RPC) abstracts procedure calls between processes on
networked systems
 It is similar in many respects to the IPC Mechanism
 However , because we are dealing with an environment in which the processes
are executing on separate systems we must use a message based communication
scheme to provide remoter service.
 In contrast to the IPC facility, the message exchanged in RPC communication are
well structured and are thus no longer just packets of data
 Each message is addressed on RPC daemon listening to a port on the remote
system and each contains an identifier of the function to execute and the
parameters to pass to that function.
 The Function is then executes as requested and any output is sent back to the
requester in a separate message.
Remote Procedure Calls…
 The semantics of RPCs allows a client to invoke a procedure on remote host
as it would invoke a procedure locally.
 the RPC system hides the details that allows communication to take place by
providing a stub on the client side
 Typically a separate stub exist for each separated remote procedure
 When a client invokes a remote procedure, the RPC system calls the appropriate
stub, passing it the parameters provided to the remote procedure, This stub
locates the port on the server and marshals the parameters.
 Parameters marshalling involves packaging the parameters into a form that can be
transmitted over a network.
 The stub then transmit the message to the server using message passing
 A similar stub on the server side receive the message and invoke the related
procedure on the server side.
 If necessary, return values are passed back to the client using the same technique
Issues in RPC and How they are resolved

Issue How they are resolved

Difference in data representation RPC system defines a machine independent data. One such
on the client side and server side representation is called external data representation (XDR)
On client side parameter marshaling involves converting
the machine dependent data into XDR before they are sent
to the server.
ON the server side the XDR data are un-marshaled to
converted to machine dependent representation for the
server
Locally procedure fails only on The Operating system must ensure that messages are acted
extreme cases, RPC system can on exactly once rather than ata most once . Most local
fails or be duplicated and procedures have the exactly once functionality, but is more
executed more than once, as a difficult to implement.
result of common network error.
Execution of RPC