Download Threads - Wikispaces

Chapter 4: Threads
Operating System Concepts – 9th Edition
Silberschatz, Galvin and Gagne ©2013
Chapter 4: Threads
n
Overview
n
Multicore Programming
n
Multithreading Models
n
Thread Libraries
n
Implicit Threading
n
Threading Issues
n
Operating System Examples
Operating System Concepts – 9th Edition
4.2
Silberschatz, Galvin and Gagne ©2013
Objectives
n
To introduce the notion of a thread—a fundamental unit of CPU
utilization that forms the basis of multithreaded computer
systems
n
To discuss the APIs for the Pthreads, Windows, and Java
thread libraries
n
To explore several strategies that provide implicit threading
n
To examine issues related to multithreaded programming
n
To cover operating system support for threads in Windows and
Linux
Operating System Concepts – 9th Edition
4.3
Silberschatz, Galvin and Gagne ©2013
Motivation
n
Most modern applications are multithreaded
n
Threads run within application – a process always has one or
more threads
n
Multiple tasks within the application can be implemented by
separate threads
n
l
Update display
l
Fetch data
l
Spell checking
l
Answer a network request
Process creation is heavy-weight while thread creation is
light-weight
l
Threads are often referred to as light-weight processes
n
Threading can greatly simplify code and increase efficiency
n
Kernels are generally multithreaded
Operating System Concepts – 9th Edition
4.4
Silberschatz, Galvin and Gagne ©2013
Single and Multithreaded Processes
Operating System Concepts – 9th Edition
4.5
Silberschatz, Galvin and Gagne ©2013
Multithreaded Server Architecture
Operating System Concepts – 9th Edition
4.6
Silberschatz, Galvin and Gagne ©2013
Benefits
n
Benefits of threading:
l
Responsiveness – threading may allow one thread to
continue execution if another thread is blocked. This is
especially important for user interfaces.
l
Resource Sharing – threads share the resources of process
(heap, globals, etc.) making communication easier than with
shared memory or message passing.
l
Economy – Creating a thread is cheaper (faster) than
creating a process. Switching between threads is lower
overhead than context switching between processes
l
Scalability – threaded processes can take advantage of
multiprocessor architectures
Operating System Concepts – 9th Edition
4.7
Silberschatz, Galvin and Gagne ©2013
Multicore Programming
n
Multicore or multiprocessor systems are putting pressure on
programmers, challenges include:
l
Dividing activities
l
Balance
l
Data splitting
l
Data dependency
l
Testing and debugging
n
Parallelism implies a system can perform more than one task
simultaneously
n
Concurrency supports more than one task making progress
l
A single single processor / core can use a scheduler to provide
concurrency (multiprogramming / timesharing) without parallelism
Operating System Concepts – 9th Edition
4.8
Silberschatz, Galvin and Gagne ©2013
Concurrency vs. Parallelism
n
Concurrent execution on single-core system:
n
Parallelism on a multi-core system:
Operating System Concepts – 9th Edition
4.9
Silberschatz, Galvin and Gagne ©2013
Multicore Programming (Cont.)
n
n
Types of parallelism
l
Data parallelism – distributes subsets of the same data
across multiple cores, same operation on each
l
Task parallelism – distributing threads across cores, each
thread performing unique operation
As # of threads grows, so does architectural support for threading
l
CPUs have multiple cores

Each core may be able to store the context of several
threads concurrently to speed switching between threads

These concurrently loaded threads are known as
hardware threads
–
Consider Oracle SPARC T4 with 8 cores and 8
hardware threads per core (64 hardware threads)
–
Consider Intel i7 with 4 cores and 2 threads per core
(8 hardware threads) (Hyper Threading)
Operating System Concepts – 9th Edition
4.10
Silberschatz, Galvin and Gagne ©2013
Amdahl’s Law
n
Amdahl’s Law identifies the performance gains from adding additional
cores to an application that has both serial and parallel components
n
S is serial portion
n
N processing cores
n
That is, if application is 75% parallel / 25% serial, moving from 1 to 2
cores results in speedup of 1.6 times
n
As N approaches infinity, speedup approaches 1 / S
The serial portion of an application has disproportionate effect on
performance gained by adding additional cores
n
But does the law take into account contemporary multicore systems?
Operating System Concepts – 9th Edition
4.11
Silberschatz, Galvin and Gagne ©2013
User Threads and Kernel Threads
n
There are two different types of threads:
n
User threads - management done by user-level threads library
l
n
Three primary thread libraries:

POSIX Pthreads

Windows threads

Java threads
Kernel threads - Supported by the Kernel
l
Examples – virtually all general purpose operating systems, including:

Windows

Solaris

Linux

Tru64 UNIX

Mac OS X
Operating System Concepts – 9th Edition
4.12
Silberschatz, Galvin and Gagne ©2013
Multithreading Models
n
Kernel threads are maintained by the operating system and are used to
allocate processor time.
l
The short-term scheduler is actually switch between kernel threads.
n
User threads are managed without kernel support.
n
For the user threads to run there must be a relationship between kernel
threads and user threads.
l
Many-to-One: Many user threads are attached to one kernel thread
l
One-to-One: Every user thread is attached to exactly one kernel thread
l
Many-to-Many: User threads are not attached to specific kernel threads
but multiplexed between them
Operating System Concepts – 9th Edition
4.13
Silberschatz, Galvin and Gagne ©2013
Many-to-One
n
Many user-level threads mapped to
single kernel thread
n
One thread blocking causes all to block
n
Multiple threads may not run in parallel
on muticore system because only one
may be in kernel at a time
n
Few systems currently use this model
n
Examples:
l
Solaris Green Threads
l
GNU Portable Threads
Operating System Concepts – 9th Edition
4.14
Silberschatz, Galvin and Gagne ©2013
One-to-One
n
Each user-level thread maps to kernel thread
n
Creating a user-level thread creates a kernel thread
n
More concurrency than many-to-one – one thread
blocking does not cause others to block
n
Number of threads per process sometimes
restricted due to overhead – the system may be
limited on the number of kernel threads that can be
supported.
n
Examples
l
Windows
l
Linux
l
Solaris 9 and later
Operating System Concepts – 9th Edition
4.15
Silberschatz, Galvin and Gagne ©2013
Many-to-Many Model
n
Allows many user level threads to be
mapped to many kernel threads
n
Allows concurrency without imposing
a limit on the number of user threads
n
Examples:
l
Solaris prior to version 9
l
Windows with the ThreadFiber
package
Operating System Concepts – 9th Edition
4.16
Silberschatz, Galvin and Gagne ©2013
Two-level Model
n
Similar to M:M, except that it allows a user thread to be
bound to kernel thread
n
Examples:
l
IRIX
l
HP-UX
l
Tru64 UNIX
l
Solaris 8 and earlier
Operating System Concepts – 9th Edition
4.17
Silberschatz, Galvin and Gagne ©2013
Thread Libraries
n
A Thread library provides the programmer with an API for
creating and managing threads
n
There are two primary ways this is implemented:
l
Library entirely in user space
l
Kernel-level library supported by the OS
Operating System Concepts – 9th Edition
4.18
Silberschatz, Galvin and Gagne ©2013
Pthreads
n
Pthreads is a POSIX standard (IEEE 1003.1c) API for thread
creation and synchronization
l
May be provided either as user-level or kernel-level
l
Specification, not implementation – different systems
may implement the threads in different ways, but the API is
consistent
l
API specifies behavior of the thread library, implementation
is up to development of the library
l
Common in UNIX operating systems (Solaris, Linux, Mac
OS X)
Operating System Concepts – 9th Edition
4.19
Silberschatz, Galvin and Gagne ©2013
Pthreads Example
/* Note: In Linux link with option -pthread */
#include <stdio.h>
#include <pthread.h>
int sum; /* this data is shared by the thread(s) */
void *runner(void *param); /* threads call this function */
int main(int argc, char *argv[])
{
pthread_t tid; /* the thread identifier */
pthread_attr_t attr; /* set of thread attributes */
if (argc != 2) {
fprintf(stderr, "usage: a.out <integer value>\n");
return -1;
}
if (atoi(argv[1]) < 0) {
fprintf(stderr, "%d must be >= 0\n", atoi(argv[1]));
return -1;
}
Operating System Concepts – 9th Edition
4.20
Silberschatz, Galvin and Gagne ©2013
Pthreads Example
pthread_attr_init(&attr); /*get the default attributes */
/* create the thread */
pthread_create(&tid, &attr, runner, argv[1]);
pthread_join(tid, NULL); /*wait for the thread to exit */
printf("sum = %d\n", sum);
return 0;
}
/* The thread will begin control in this function */
void *runner(void *param)
{
int i, upper = atoi(param);
sum = 0;
for(i = 1; i <= upper; i++)
sum += i;
pthread_exit(0);
}
Operating System Concepts – 9th Edition
4.21
Silberschatz, Galvin and Gagne ©2013
Pthreads Code for Joining 10 Threads
n
When programming for multicore systems, a programmer may
need to create multiple threads.
l
It may be helpful to keep track of the threads using an array.
l
The following code shows how to join on an array of 10
threads using a for loop.
Operating System Concepts – 9th Edition
4.22
Silberschatz, Galvin and Gagne ©2013
Windows Threads
n
Creating threads in Windows is similar in concept.
l
Global data is shared between threads.
l
When a thread is created it begins in a programmer defined
function.

A point to this function is passed as a parameter when
creating the thread.

This function can also be passed parameters
l
The thread is created using CreateThread().
l
The parent waits for the thread using
WaitForSingleObject().
l
To wait for multiple threads Windows has the
WaitForMultipleObjects() function.
Operating System Concepts – 9th Edition
4.23
Silberschatz, Galvin and Gagne ©2013
Windows Multithreaded C Program
Operating System Concepts – 9th Edition
4.24
Silberschatz, Galvin and Gagne ©2013
Windows Multithreaded C Program (Cont.)
Operating System Concepts – 9th Edition
4.25
Silberschatz, Galvin and Gagne ©2013
Java Threads
n
Java threads are managed by the JVM
n
Typically implemented using the threads model provided by
underlying OS
n
Java threads may be created by:
l
Extending Thread class
l
Implementing the Runnable interface
Operating System Concepts – 9th Edition
4.26
Silberschatz, Galvin and Gagne ©2013
Java Multithreaded Program
Operating System Concepts – 9th Edition
4.27
Silberschatz, Galvin and Gagne ©2013
Java Multithreaded Program (Cont.)
Operating System Concepts – 9th Edition
4.28
Silberschatz, Galvin and Gagne ©2013
Implicit Threading
n
As the number of threads used increases, it can be difficult for
the programmer to keep track.
n
One way to address this is to transfer the creation of threads to
compilers and runtime libraries.
n
Threads and created and joined automatically by library
functions and the compiler.
n
We will explore three methods:
n
l
Thread Pools
l
OpenMP
l
Grand Central Dispatch
Other methods include Microsoft Threading Building Blocks
(TBB), java.util.concurrent package
Operating System Concepts – 9th Edition
4.29
Silberschatz, Galvin and Gagne ©2013
Thread Pools
n
Create a number of threads in a pool where they await work
n
Advantages:
l
It is usually slightly faster to service a request with an
existing thread than create a new thread
l
It allows the number of threads in the application(s) to be
bound to the size of the pool
l
Separating the task to be performed from the mechanics of
creating task allows us to use different strategies for
running task

n
i.e. Tasks could be scheduled to run periodically
Windows API supports thread pools:
Operating System Concepts – 9th Edition
4.30
Silberschatz, Galvin and Gagne ©2013
OpenMP
n
Set of compiler directives and an
API for C, C++, FORTRAN
n
Provides support for parallel
programming in shared-memory
environments
n
Identifies parallel regions –
blocks of code that can run in
parallel
#pragma omp parallel
Create as many threads as there are
cores
#pragma omp parallel for
for(i=0;i<N;i++) {
c[i] = a[i] + b[i];
}
Run for loop in parallel
Operating System Concepts – 9th Edition
4.31
Silberschatz, Galvin and Gagne ©2013
Grand Central Dispatch
n
Apple technology for Mac OS X and iOS operating systems
n
Extensions to C, C++ languages, API, and run-time library
n
Allows identification of parallel sections
n
Manages most of the details of threading
n
Block is in “^{ }” - ˆ{ printf("I am a block"); }
n
Blocks placed in dispatch queue
l
Assigned to available thread in thread pool when removed
from queue
Operating System Concepts – 9th Edition
4.32
Silberschatz, Galvin and Gagne ©2013
Grand Central Dispatch
n
Two types of dispatch queues:
l
serial – blocks removed in FIFO order, queue is per process,
called main queue

l
concurrent – removed in FIFO order but several may be
removed at a time

n
Programmers can create additional serial queues within
program
Three system wide queues with priorities low, default, high
The following code submits a block to the concurrent queue with
default priority.
Operating System Concepts – 9th Edition
4.33
Silberschatz, Galvin and Gagne ©2013
Threading Issues
n
There are several issues that need to be considered in
implementing a multithreaded environment.
l
The semantics of fork() and exec() system calls must change
l
Process signals be handling

l
Synchronous and asynchronous
Thread cancellation of target thread

Asynchronous or deferred
l
Thread-local storage
l
Scheduler activations
Operating System Concepts – 9th Edition
4.34
Silberschatz, Galvin and Gagne ©2013
Semantics of fork() and exec()
n
n
In a multithreaded environment, does fork()duplicate only
the calling thread or all threads?
l
Some UNIXes have two versions of fork, one
duplicates all threads and the other duplicates only the
one that invoked fork.
l
If exec will be called immediately after fork, then it is
wasteful to create multiple threads.
exec() usually works as normal – replace the running
process including all threads
Operating System Concepts – 9th Edition
4.35
Silberschatz, Galvin and Gagne ©2013
Signal Handling
n
Signals are used in UNIX systems to notify a process that a
particular event has occurred.
n
A signal handler is used to process signals
n
1.
Signal is generated by particular event
2.
Signal is delivered to a process
3.
Signal is handled by one of two signal handlers:
1.
default
2.
user-defined
Every signal has default handler that the kernel runs when
handling signal
l
User-defined signal handler can override the default
l
For a single-threaded process, the signal is delivered to
process
Operating System Concepts – 9th Edition
4.36
Silberschatz, Galvin and Gagne ©2013
Signal Handling (Cont.)
n
Where should a signal be delivered for a multi-threaded
process?
n
Possible answers:
Deliver the signal to the thread to which the signal
applies
l
n
l
Deliver the signal to every thread in the process
l
Deliver the signal to certain threads in the process
l
Assign a specific thread to receive all signals for the
process
Different types of signals may be delivered using different
methods.
l
Synchronous signals need to be delivered to the thread
generating the signal.
l
Asynchronous signals may be delivered in different ways

Some should be delivered to all threads (e.g.
<control><C> process termination).
Operating System Concepts – 9th Edition
4.37
Silberschatz, Galvin and Gagne ©2013
Thread Cancellation
n
It is sometimes necessary to terminate a thread before it has
finished
n
The thread to be canceled is the target thread
n
Two general approaches:
n
l
Asynchronous cancellation terminates the target thread
immediately
l
Deferred cancellation allows the target thread to periodically
check if it should be cancelled
The following is Pthread code to create and cancel a thread:
Operating System Concepts – 9th Edition
4.38
Silberschatz, Galvin and Gagne ©2013
Thread Cancellation (Cont.)
Invoking thread cancellation requests cancellation, but actual
cancellation depends on thread state
If thread has cancellation disabled, cancellation remains pending
until thread enables it
The default type is deferred
Cancellation only occurs when the thread reaches a
cancellation point

i.e. the thread might call pthread_testcancel()to
indicate a cancellation point

If the thread has been cancelled, then a cleanup handler
is invoked
On Linux systems, thread cancellation is handled through signals.
Operating System Concepts – 9th Edition
4.39
Silberschatz, Galvin and Gagne ©2013
Thread-Local Storage
Thread-local storage (TLS) allows each thread to have its
own copy of data
Useful when you do not have control over the thread creation
process (i.e., when using a thread pool)
Different from local variables
Local variables visible only during single function
invocation
TLS visible across function invocations within the thread
Similar to static data
TLS is unique to each thread
Operating System Concepts – 9th Edition
4.40
Silberschatz, Galvin and Gagne ©2013
Scheduler Activations
Both M:M and Two-level models require
communication to maintain the appropriate
number of kernel threads allocated to the
application
Many systems use an intermediate data
structure between user and kernel threads –
lightweight process (LWP)
Appears to be a virtual processor on which
process can schedule the user thread to
run
Each LWP is attached to a kernel thread
How many LWPs to create?
Scheduler activations provide upcalls - a
communication mechanism from the kernel to
the upcall handler in the thread library
This communication allows an application to
maintain the correct number kernel threads
Operating System Concepts – 9th Edition
4.41
Silberschatz, Galvin and Gagne ©2013
Operating System Examples
Windows Threads
Linux Threads
Operating System Concepts – 9th Edition
4.42
Silberschatz, Galvin and Gagne ©2013
Windows Threads
Windows implements the Windows API – primary API for Win
98, Win NT, Win 2000, Win XP, and Win 7
Implements the one-to-one mapping
Each thread contains
A thread id
Register set representing state of processor
Separate user and kernel stacks for when thread runs in
user mode or kernel mode
Private data storage area used by run-time libraries and
dynamic link libraries (DLLs)
The register set, stacks, and private storage area are known as
the context of the thread
Operating System Concepts – 9th Edition
4.43
Silberschatz, Galvin and Gagne ©2013
Windows Threads (Cont.)
The primary data structures of a thread include:
ETHREAD (executive thread block) – includes pointer to
process to which thread belongs and to KTHREAD, in
kernel space
KTHREAD (kernel thread block) – scheduling and
synchronization info, kernel-mode stack, pointer to TEB, in
kernel space
TEB (thread environment block) – thread id, user-mode
stack, thread-local storage, in user space
Operating System Concepts – 9th Edition
4.44
Silberschatz, Galvin and Gagne ©2013
Windows Threads Data Structures
Operating System Concepts – 9th Edition
4.45
Silberschatz, Galvin and Gagne ©2013
Linux Threads
Linux refers to threads as tasks
Thread creation is done through clone() system call
clone() allows a child task to share the address space of the
parent task (process)
Flags control behavior
struct task_struct points to process data structures
(shared or unique)
Operating System Concepts – 9th Edition
4.46
Silberschatz, Galvin and Gagne ©2013
End of Chapter 4
Operating System Concepts – 9th Edition
Silberschatz, Galvin and Gagne ©2013
Homework
Chapter 4, page 191
1, 3, 4, 6, 7, 8, 11, 13, 15, 17, 18, 21, 26
Operating System Concepts – 9th Edition
4.48
Silberschatz, Galvin and Gagne ©2013