Threads and Concurrency

In this Topic, we will cover the following:

• Overview
• Multicore Programming
• Multithreading Models
• Thread Libraries
• Implicit Threading
• Operating System Examples

• Identify the basic components of a thread, and contrast threads and processes
• Describe the benefits and challenges of designing multithreaded applications
• Describe how the Windows and Linux operating systems represent threads
• Describe multithreaded applications using the Pthreads, Java, and Windows threading APIs

• Most modern applications are multithreaded
• Threads run within application
• Multiple tasks with the application can be implemented by separate threads
  • Update display
  • Fetch data
  • Spell checking
  • Answer a network request
• Can simplify code, increase efficiency
• Kernels are generally multithreaded

• Heavyweight process: Traditional processes that can have only one thread
• Multi-threaded processes: A process that includes multiple threads that share the resources allocated to the process
• lightweight process: another name for A thread, is a basic unit of CPU utilization; it is comprised of:
  • A thread ID
  • program counter
  • register set
  • stack space
• A thread shares with other threads its:
  • code section
  • data section
  • O/S resources

• Threads are important for any application with multiple tasks that can be run with separate threads of control.
• A web server could produce a thread for each client
  • Serve clients with multiple threads concurrently.
  • less overhead to use multiple threads than to use multiple processes.
• A Word processor may have separate threads for:
  • Reading user input
  • Display graphics
  • Performing Spell & grammar check

• Benefits of threads are:
  • Responsiveness:
    • Threads allow a program to continue running even if part of it is blocked/ performing a lengthy operation. e.g: a multithreaded web browser can allow user interaction in 1 thread while loading an image in another thread.
  • Resource Sharing:
    • threads share memory and resources of the process they belong to.
  • Economy:
    • Allocating memory and resources to a process is costly- threads share resources of the process .
    • Threads are faster to create and to switch between them than processes .(i.e. in Solaris, creating P is 30 times slower than creating thread).
  • Scalability- Utilization of Multiprocessor architectures:
    • Multiple threads can run in parallel on different CPUs - A single threaded process can run only on 1 CPU no matter how many are available.
• Thread management could be done at either
  • User level
  • Kernel level

• Multicore or multiprocessor systems putting pressure on programmers, challenges include:
  • Dividing activities- area of the application that can be divided into separate tasks.
  • Balance- ensure that tasks perform equal work or equal value
  • Data splitting- data accessed by the tasks must be divided to run on separate cores
  • Data dependency- 1 task depend on data from another, the execution synchronized
  • Testing and debugging- When a program is running in parallel on multiple cores, many different execution paths are possible. Testing and debugging such concurrent programs is inherently more difficult than testing and debugging single-threaded applications

• A thread is managed by an application (by user-level threads library) not in the kernel
• Scheduled by the thread library
• Fast to create and manage: all thread creation and scheduling are done in user space without the need for kernel intervention
• Any user-level thread performing a blocking system call will cause the entire process to block if the kernel is single threaded.
• Examples of three primary thread libraries - POSIX Pthreads - Java threads - Win32 threads

• All thread management is done by the kernel itself
• Generally slower to create and manage than user threads
• If a thread is performing a blocking system, the kernel can schedule another thread in an application to be executed
• Kernel can schedule threads on different processors
• Kernel does creation, scheduling and management of threads.
• Examples – virtually all general purpose operating systems, including:
  • Windows
  • Linux
  • Mac OS X
  • iOS
  • Android

What is the relationship between the user & kernel threads:

• Many-to-One
• One-to-One
• Many-to-Many