1. Operating System Concepts

2. Lecture Contents Motivation for threads Thread Concept
Similarity and Difference between Process and Thread
Advantages of threads
Implementation of Multithreading
Multithreading Models
Thread Library Concept
Multithreading Issues
Case Study
*Please follow Silberchatz, Galvin and Gagne (Chapter 4), 7th edition

3. Issues with Processes
So far we have discussed Process, Creation of new process concept and its execution; However there are some issues with processes.
Processes are EXPENSIVE ! (E.g. resources that are to be allocated with every process)
Hardware resources (memory), kernel resources (PCB)
Link establishment for IPC (unrelated processes/related processes)
Concurrency
Motivation ? .. A mechanism that provides an alternative of a process (so that a new process is not needed)

4. THREAD
Thread- A piece of code in a process which runs concurrently with execution of other piece of codes within the same process (same address space)
No need to create a new process
Resources that thread need to update/allocate privately:
Thread ID
CPU Context (Program Counter (PC), register set)
Stack
Priority
Kernel maintained external variable errno, (in order to tell the programmer that an error has occurred)

5. Threads within a same process share…
Multiple and concurrent threads within a process share:
Code
Data
Open File descriptor table
PCB (Process Control Block)
For a thread creation, only need to allocate CPU context and Stack !!
Threads are in the same process address space

6. Single and Multithreaded Processes

7. How processes are similar to threads or different ?
A thread can be in a state, similar to process (new, ready, running, waiting, terminate)
Threads can also create other threads.
Difference
Processes execute only within their own address space, whereas threads of same process execute within address space of one major process
Automatic protection mechanism for processes but what about threads… ?
Synchronization issues ?

8. Advantages of thread
Responsiveness - allow a program to continue running even if part of it is blocked (e.g. In a browser, multiple threads are working simultaneously to increase responsiveness)
Resource sharing - threads share the memory and resources of the process (allows an application to have several threads of activity within the same address space)
Economy - it is more economical to create and context-switch threads (because they share resources of the process to which they belong)
In Solaris a process is thirty times slower to create and five times slower to context-switch
Scheduling a thread is easy (just save the CPU context)
Scalability (utilizing multiprocessor architectures)

9. Some other things..
The main reason for having threads is that in many applications, multiple activities are going on at once.
Some of these may block from time to time. By decomposing such an application into multiple sequential threads that run in parallel, the programming model becomes simpler.

10. How multithreading is implemented ?
There are two techniques for implementing multithreading
User-Level threads
Thread creation and management is done by the application using a thread library
Any application can be programmed to be multithreaded by using a thread library which is a package of application level functions.
The kernel is not aware of existence of threads
Kernel Level threads
Thread management and implementation is done by operating system
Kernel provides a way for user processes to create multiple threads

11. Difference between the two techniques…
Advantages of User level threads
User-level threads can be created in a system that doesn’t provide multithreading facility
Management and scheduling of thread is done by threading library in user space, so user level threads don’t invoke kernel for doing these tasks
User level threads can run on any OS, no changes are required to do on underlying kernel (Portability)
Disadvantages
Using User Level threading strategy, a multithreaded application cannot take advantage of multiprocessor system. (different threads of a same process cannot run on different CPU’s because OS considers a process a single threaded program)
If a thread is blocked, all the threads within a process gets blocked.

12. Advantages of Kernel level threads
Since kernel threads use the kernel scheduler, different kernel threads can run on different CPUs (taking advantage of multiprocessor system)
If a thread within a process gets blocked (e.g. waiting for an I/O), the scheduler can schedule another thread
Disadvantages
Frequent invoking of kernel for management of threads
More expensive to create than user threads

15. Examples User Threads Kernel Threads
Thread management done by user-level threads library
Three primary thread libraries:
POSIX Pthreads
Win32 threads
Java threads
Kernel Threads
Supported by the Kernel
Examples
Windows XP/2000
Solaris
Linux
POSIX: an acronym for Portable Operating System Interface, is a family of standards specified by the IEEE Computer Society for maintaining compatibility between operating systems. POSIX defines the application programming interface (API), along with command line shells and utility interfaces, for software compatibility with variants of Unix and other operating systems.
This standard defines a standard operating system interface and environment, including a command interpreter (or "shell"), and common utility programs to support applications portability at the source code level. 
Pthreads: Pthreads refers to the POSIX standard (IEEE c) defining an API for thread creation and synchronization. This is a specification for thread behavior, not an implementation.
Threads have been standardized by the IEEE Technical Committee on Operating Systems, POSIX (Portable Operating System Interface). This standard allows users to write portable multi-thread programs.

16. Multithreading Models
Based on the number of kernel threads allocated correspondingly to the user level threads, we have three multithreading models
Many-to-one Model
One-to-one Model
Many-to-Many Model

17. Many-to-One Model
Many user-level threads mapped to single kernel thread
Thread management is done by the thread library in user space, so it is efficient.
Drawback - the entire process will block if a thread makes a blocking system call.
Examples:
Solaris Green Threads
GNU Portable Threads
Kernel is scheduling a single process

18. One-to-One Model Each user-level thread maps to kernel thread
True concurrency - compared to many-to-one model since another thread will be able to run when a thread makes a blocking system call.
Allows multiple threads to run in parallel on multiprocessors.
Drawback:
Creating a user thread requires creating a corresponding kernel thread. (Overhead)
Examples
Windows NT/XP/2000
Linux
Solaris 9 and later

19. Many-to-Many Model (Hybrid Model)
Multiple user level threads multiplexed over a smaller or equal number of kernel threads
Allows the operating system to create a sufficient number of kernel threads
Solaris 2
Windows NT/2000 with the ThreadFiber package
Which of three models solaris model is built?

20. Thread Libraries
Thread library provides programmer with API for creating and managing threads
Two primary ways of implementing
Library entirely in user space
Kernel-level library supported by the OS -> invoking a function in the library results with a system call

21. Pthreads
A POSIX standard (IEEE c) API for thread creation and synchronization
API specifies behavior of the thread library, implementation is up to development of the library (i.e. the OS designers)
This standard has been defined to make the writing of PORTABLE threaded program possible
This standard defines over 60 function calls
May be provided either as user-level or kernel-level.
Common in UNIX operating systems (Solaris, Linux, Mac OS X)

22. Threading Issues Thread cancellation of target thread
Asynchronous or deferred
Signal handling
Thread pools
Thread-specific data
Scheduler activations

23. Thread Cancellation Terminating a thread before it has finished
Example: multiple threads concurrently searching through a database and one thread returns the result
Stopping a webpage from loading any further
Two general approaches:
Asynchronous cancellation terminates the target thread immediately – troublesome if it’s in middle of updating some data
Deferred cancellation allows the target thread to periodically check if it should be cancelled – checking a flag.
Can only be cancelled at defined points called cancellation points

24. Signal Handling
Signals are used in UNIX systems to notify a process that a particular event has occurred
Example: illegal memory access, division by zero
A signal handler is used to process signals
Signal is generated by particular event
Signal is delivered to a process
Signal is handled
In a multithreaded process, where should a signal be delivered?
Options:
Deliver the signal to the thread to which the signal applies
Deliver the signal to every thread in the process
Deliver the signal to certain threads in the process
Assign a specific thread to receive all signals for the process

25. Thread Pools Reason for the need of thread pool:
Time consumed to create a new thread and destroy the previous ones – leads to slow performance
Unlimited threads could exhaust system resources, such as CPU time or memory
Solution:
Thread Pools
Create a number of threads at process startup and place them in a pool where they await work
Advantages:
Usually slightly faster to service a request with an existing thread than create a new thread
Also allows an application to specify the number of threads it requires

27. Thread Specific Data Scheduler activation
Allows each thread to have its own copy of specific data
Example, in a transaction-processing system, service each transaction in a separate thread. Furthermore, each transaction might be assigned a unique identifier. To associate each thread with its unique identifier, use thread-specific data.
Scheduler activation
User-Level Threads: If a single user-level thread blocks, the operating system blocks the entire multithreaded process. Another limitation to a many-to-one thread mapping is that a process's threads cannot simultaneously execute on multiple processors. Scheduler activations attempt to address these limitations to user-level threads. A scheduler activation is a kernel thread that can notify a user-level threading library of events (e.g., a thread has blocked or a processor is available). This type of kernel thread is called a "scheduler activation,“ because the user-level threading library can perform thread-scheduling operations when "activated" by an event notification, sometimes called an upcall.

28. Case Study
Windows XP
Linux

29. Windows XP

30. Windows XP Threads

31. Linux Threads Linux refers to them as tasks rather than threads
Thread creation is done through clone() system call
clone() accepts arguments that specify which resources to share with the child process