Concurrency

BAB XI

Concurrency

Concurrency in software execution can occur at four different levels: instruction level (executing two or more machine instructions simultaneously), statement level (executing two or more high-level language statements simultaneously), unit level (executing two or more subprogram units simultaneously), and program level (executing two or more programs simultaneously). Because no language design issues are involved with them, instruction-level and program-level concurrency are not discussed in this chapter. Concurrency at both the subprogram and the statement levels is discussed, with most of the focus on the subprogram level.

There are two distinct categories of concurrent unit control. The most natural category of concurrency is that in which, assuming that more than one processor is available, several program units from the same program literally execute simultaneously. This is physical concurrency. A slight relaxation of this concept of concurrency allows the programmer and the application software to assume that there are multiple processors providing actual concurrency, when in fact the actual execution of programs is taking place in interleaved fashion on a single processor. This is logical concurrency. From the programmer’s and language designer’s points of view, logical concurrency is the same as physical concurrency. It is the language implementor’s task, using the capabilities of the underlying operating system, to map the logical concurrency to the host hardware. Both logical and physical concurrency allow the concept of concurrency to be used as a program design methodology.

Concurrency can occur at four levels:

·         Machine instruction level

·         High-level language statement level

·         Unit level

·         Program level

Because there are no language issues in instruction- and program-level concurrency, they are not addressed here.

Introduction to Subprogram-Level Concurrency

·         A task or process or thread is a program unit that can be in concurrent execution with other program units

·         Tasks differ from ordinary subprograms in that:

o   A task may be implicitly started

o   When a program unit starts the execution of a task, it is not necessarily suspended

o   When a task’s execution is completed, control may not return to the caller

·         Tasks usually work together

·         Two general categories of task

o   Heavyweight tasks execute in their own address space

o   Lightweight tasks all run in the same address space – more efficient

o   A task is disjoint if it does not communicate with or affect the execution of any other task in the program in any way

·         Task synchronization

o   A mechanism that controls the order in which tasks execute

o   Two kinds of synchronization

§  Cooperation synchronization

§  Competition synchronization

o   Task communication is necessary for synchronization, provided by:

§  Shared nonlocal variables

§  Parameters

§  Message passing

·         Kinds of synchronization

o   Cooperation: Task A must wait for task B to complete some specific activity before task A can continue its execution, e.g., the producer-consumer problem

o   Competition: Two or more tasks must use some resource that cannot be simultaneously used, e.g., a shared counter.

Scheduler

o   Providing synchronization requires a mechanism for delaying task execution

o   Task execution control is maintained by a program called the scheduler, which maps task execution onto available processors

·         Task Execution States

o   New - created but not yet started

o   Ready - ready to run but not currently running (no available processor)

o   Running

o   Blocked - has been running, but cannot now continue (usually waiting for some event to occur)

o   Dead - no longer active in any sense

·         Liveness and Deadlock

o   Liveness is a characteristic that a program unit may or may not have

§  In sequential code, it means the unit will

    eventually complete its execution

o   In a concurrent environment, a task can easily lose its liveness

o   If all tasks in a concurrent environment lose their liveness, it is called deadlock

·         Design issues for concurrency

o   Competition and cooperation synchronization*

o   Controlling task scheduling

o   How can an application influence task scheduling

o   How and when tasks start and end execution

o   How and when are tasks created (The most important issue)

·         Methods of providing synchronization

o   Semaphores

o   Monitors

o   Message Passing

Semaphores

·         Dijkstra – 1965

o   In an effort to provide competition synchronization through mutually exclusive access to shared data structures, Edsger Dijkstra devised semaphores in 1965 (Dijkstra, 1968b). Semaphores can also be used to provide cooperation synchronization.

·         A semaphore is a data structure consisting of a counter and a queue for storing task descriptors

o   A task descriptor is a data structure that stores all of the relevant information about the execution state of the task

·         Semaphores can be used to implement guards on the code that accesses shared data structures

·         Semaphores have only two operations, wait and release (originally called P and V by Dijkstra)

·         Semaphores can be used to provide both competition and cooperation synchronization

·         Cooperation Synchronization with Semaphores

o   Example: A shared buffer

o   The buffer is implemented as an ADT with the operations DEPOSIT and FETCH as the only ways to access the buffer

o   Use two semaphores for cooperation: emptyspots and fullspots

o   The semaphore counters are used to store the numbers of empty spots and full spots in the buffer

o   DEPOSIT must first check emptyspots to see if there is room in the buffer

o   If there is room, the counter of emptyspots is decremented and the value is inserted

o   If there is no room, the caller is stored in the queue of emptyspots

o   When DEPOSIT is finished, it must increment the counter of fullspots

o   FETCH must first check fullspots to see if there is a value

§  If there is a full spot, the counter of fullspots is decremented and the value is removed

§  If there are no values in the buffer, the caller must be placed in the queue of fullspots

§  When FETCH is finished, it increments the counter of emptyspots

o   The operations of FETCH and DEPOSIT on the semaphores are accomplished through two semaphore operations named wait and release

·         Semaphores: Wait and Release Operations

wait(aSemaphore)

if aSemaphore’s counter > 0 then

   decrement aSemaphore’s counter

else

   put the caller in aSemaphore’s queue

   attempt to transfer control to a ready task

     -- if the task ready queue is empty,

     -- deadlock occurs

end

release(aSemaphore)

if aSemaphore’s queue is empty then

   increment aSemaphore’s counter

else

   put the calling task in the task ready queue

   transfer control to a task from aSemaphore’s queue

end

·         Producer and consumer tasks

semaphore fullspots, emptyspots;

fullstops.count = 0;

emptyspots.count = BUFLEN;

task producer;

                loop

                -- produce VALUE –-

                wait (emptyspots); {wait for space}

                DEPOSIT(VALUE);

                release(fullspots); {increase filled}

                end loop;

end producer;

task consumer;

                loop

                wait (fullspots);{wait till not empty}}

                FETCH(VALUE);

                release(emptyspots); {increase empty}

                -- consume VALUE –-

                end loop;

end consumer;

·         Competition Synchronization with Semaphores

o   A third semaphore, named access, is used to control access (competition synchronization)

§  The counter of access will only have the values 0 and 1

§  Such a semaphore is called a binary semaphore

o   Note that wait and release must be atomic!

·         Evaluation of Semaphores

o   Misuse of semaphores can cause failures in cooperation synchronization, e.g., the buffer will overflow if the wait of fullspots is left out

o   Misuse of semaphores can cause failures in competition synchronization, e.g., the program will deadlock if the release of access is left out

Monitors

·         Ada, Java, C#

·         The idea: encapsulate the shared data and its operations to restrict access

·         A monitor is an abstract data type for shared data

·         Competition Synchronization

o   Shared data is resident in the monitor (rather than in the client units)

o   All access resident in the monitor

§  Monitor implementation guarantee synchronized access by allowing only one access at a time

§  Calls to monitor procedures are implicitly queued if the monitor is busy at the time of the call

Message Passing

·         Message passing is a general model for concurrency

o   It can model both semaphores and monitors

o   It is not just for competition synchronization

·         Central idea: task communication is like seeing a doctor--most of the time she waits for you or you wait for her, but when you are both ready, you get together, or rendezvous

·         Rendezvous Time Lines