InterProcess Communication

Inter-Process Communication
 A process has access to the memory which constitutes its
own address space.
 When a child process is created, the only way to
communicate between a parent and a child process is:
 The variables are replicas
 The parent receives the exit status of the child
 So far, we’ve discussed communication mechanisms only
during process creation/termination
 Processes may need to communicate during their life time.

Operating Systems -
Cooperating Processes
 Independent process cannot affect or be affected by the
execution of another process.
 Cooperating process can affect or be affected by the execution
of another process
 Advantages of process cooperation
 Information sharing
 Computation speed-up:
 make use of multiple processing elements
 Modularity
 Convenience:
 editing, printing, compiling in parallel
 Dangers of process cooperation
 Data corruption, deadlocks, increased complexity
 Requires processes to synchronize their processing

Operating Systems -
Purposes for IPC
 IPC allows processes to communicate and synchronize their
actions without sharing the same address space
 Data Transfer
 Sharing Data
 Event notification
 Resource Sharing and Synchronization

Operating Systems -
IPC Mechanisms
 Mechanisms used for communication and synchronization
 Message Passing
 message passing interfaces, mailboxes and message

queues
 sockets, pipes

 Shared Memory: Non-message passing systems

Operating Systems -
Message Passing
 In a Message system there are no shared variables.
 IPC facility provides two operations for fixed or variable
sized message:
 send(message)
 receive(message)
 If processes P and Q wish to communicate, they need to:
 establish a communication link
 exchange messages via send and receive
 Implementation of communication link
 physical (e.g., memory, network etc)
 logical (e.g., syntax and semantics, abstractions)

Operating Systems -
Implementation Questions
 How are links established?
 Can a link be associated with more than two processes?
 How are links made known to 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?

Operating Systems -
Message Passing Systems
 Exchange messages over a communication link
 Methods for implementing the communication link and
primitives (send/receive):
 Direct or Indirect communications (Naming)
 Symmetric or Asymmetric communications (blocking
versus non-blocking)
 Buffering
 Send-by-copy or send-by-reference
 fixed or variable sized messages

Operating Systems -
Direct Communication – Internet and
Sockets
 Processes must name each other explicitly:
 Symmetric Addressing
 send (P, message) – send to process P
 receive(Q, message) – receive from Q
 Asymmetric Addressing
 send (P, message) – send to process P
 receive(id, message) – rx from any; system sets id = sender
 Properties of communication link
 Links established automatically between pairs
 processes must know each others ID
 Exactly one link per pair of communicating processes
 Disadvantage: a process must know the name or ID of the
process(es) it wishes to communicate with

Operating Systems -
Indirect Communication
 Messages are sent to or received from mailboxes (also
referred to as ports).
 Each mailbox has a unique id.
 Processes can communicate only if they share a mailbox.
 Primitives:
 send(A, message) – send a message to mailbox A
 receive(A, message) – receive a message from mailbox A
 Properties of communication link
 Link established only if processes share a common mailbox
 A link may be associated with more than 2 processes.
 Each pair of processes may share several communication
links.

Operating Systems -
Indirect Communication-Ownership
 process owns (i.e. mailbox is implemented in user space):
 only the owner may receive messages through this mailbox.
 Other processes may only send.
 When process terminates any “owned” mailboxes are
destroyed.
 kernel owns
 then mechanisms provided to create, delete, send and
receive through mailboxes.
 Process that creates mailbox owns it (and so may receive
through it)
 but may transfer ownership to another process.

Operating Systems -
Indirect Communication
 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.
 OR allow only one process at a time to execute a receive
operation.
 OR allow the system to select arbitrarily the receiver. Sender
is notified who the receiver was.

Operating Systems -
Synchronization
 Message passing may be either blocking or non-blocking.
 blocking send:
 sender blocked until message received by mailbox or process
 nonblocking send:
 sender resumes operation immediately after sending
 blocking receive:
 receiver blocks until a message is available
 nonblocking receive:
 receiver returns immediately with either a valid or null message.

Operating Systems -
Buffering
 All messaging system require framework to temporarily
buffer messages.
 These queues are implemented in one of three ways:
 Zero capacity
 No messages may be queued within the link, requires
sender to block until receiver retrieves message.
 Bounded capacity
 Link has finite number of message buffers. If no buffers are
available then sender must block until one is freed up.
 Unbounded capacity
 Link has unlimited buffer space, consequently send never
needs to block.

Operating Systems -