Op Sys Quiz 2 - Threads

Front 1

Why use threads?

Back 1

1) Multiple tasks of an application can be implemented by threads

2) Process creation is heavy-weight, thread creation is light-weight

3) threads can simplify code, increase efficiency

Front 2

What is a thread?

Back 2

An independent stream of instructions that can be scheduled to run as such by the kernel

Front 3

Threads exist _____ the process and ______ its resources

Back 3

Threads exist within the process and share its resources

Front 4

Each thread has its own:

Back 4

essential resources (per-thread resources)

stack
registers
program counter
thread-specific data
*independent flow of control

Front 5

threads are individually scheduled by ______

Back 5

the kernel

Front 6

each thread can be in any of the scheduling states (T/F)

Back 6

true

Front 7

Benefits of Threads

Back 7

*Responsiveness
- multithreading an interactive application allows a program to continue running even if part of it is blocked or performing a lengthy operation
*Resource Sharing
- sharing resources may result in efficient communication and high degree of cooperation
*Economy
- thread is more light weight than process
*Scalability
- better utilization of multiprocessor architectures

Front 8

Threads may be provided either at the ______ or by the ______

Back 8

user level (user threads) or by the kernel (kernel threads)

Front 9

user threads are supported ____ the kernel, ______ kernel support

Back 9

above the kernel, without kernel support

Front 10

kernel threads are supported and managed directly by ____

Back 10

the kernel

Front 11

Multi-threading models

Back 11

ways to map user threads to kernel threads

Many to One
One to One
Many to Many

Front 12

Many to One

Back 12

Many user-level threads are mapped to one kernel thread

thread management is done by the thread library in user space

entire process will block if a thread makes a blocking system call

multiple threads are unable to run in parallel in multi-processors

Front 13

One to One

Back 13

One user-level thread maps to one kernel thread

allows other threads to run when a thread blocks

multiple threads can run in parallel on multiprocessors

creating a user thread requires creating a corresponding kernel thread (overhead)

limits the number of threads

Front 14

Many to Many

Back 14

Many user level threads are mapped to many kernel threads

solves shortcomings of 1:1 and m:1

create as many threads as necessary

corresponding threads can run in parallel on a multiprocessor

Front 15

Two-level model

Back 15

similar to many to many

allows a user thread to be BOUND to a kernel thread

Front 16

Thread Libraries

Back 16

provides programmer with API for creating and managing threads

Front 17

2 ways to implement thread libraries

Back 17

1) library entirely in user space with no kernel support

2) kernel level library supported by the OS

Front 18

Pthreads

Back 18

A POSIX standard API for thread creation and synchronization

specification for thread behavior

implementation is up to developer of the library

Front 19

Threading Issues

Back 19

1) Semantics of fork and exec system calls
2) thread cancellation of target thread
3) signal handling
4) thread pools
5) thread specific data
6) scheduler activities

Front 20

Semantics of fork and exec

Back 20

fork duplicates the whole single-threaded process

exec typically replaces the entire process, multithreaded or not

Front 21

thread cancellation

Back 21

termination a (target) thread before it has finished

2 approaches

Front 22

2 approaches to thread cancellation

Back 22

* asynchronous cancellation
- terminates the target thread immediately

* deferred cancellation
- allows the target thread to periodically check if it should be cancelled

Front 23

Signal handling

Back 23

signals are used in UNIX systems to notify a process that a particular event has occurred

* signal is generated by the occurrence of a particular event
* signal is delivered to a process
* once delivered, the signal must be handled

Front 24

Synchronous signals (exceptions) are delivered to:

Back 24

the same thread causing the signal

Front 25

Asynchronous signals (I/O) can be delivered to:

Back 25

* the thread to which the signal applies
* every thread in the process
* certain threads in the process
* specific thread to receive all signals for the process

Front 26

One thread per request model has _______ problems

Back 26

scalability (overhead to create a thread before each request, it may exhaust the resource)

Front 27

Thread pools

Back 27

Solve the scalability problems of the one thread per request model

* create a number of threads in a pool, where the sit and wait for work
* when received a request, they server wakens a thread and passes it to the request
* thread returns to the pool and completing the job

Front 28

Advantage of thread pools

Back 28

* Servicing a request with an existing thread is usually faster than waiting to create a new thread

* thread pool limits the number of threads that exist at any one point

Front 29

Thread specific data

Back 29

data is shared in multithreaded programs

thread specific data allows each thread to have its own copy of data

Front 30

Light weight process (LWP)

Back 30

an intermediate data structure between the user and kernel thread in many to many and two level models

to the user thread library, appears as virtual processors on which to schedule user threads

Front 31

each LWP is attached to a _________

Back 31

kernel thread

Front 32

Upcall

Back 32

scheduler activation notifies the thread library via upcalls

Front 33

Linux Thread Flag:
CLONE_FS

Back 33

file-system information is shared

Front 34

Linux Thread Flag:
CLONE_VM

Back 34

the same memory space is shared

Front 35

Linux Thread Flag:
CLONE_SIGHAND

Back 35

signal handlers are shared

Front 36

Linux Thread Flag:
CLONE_FILES

Back 36

the set of open files is shared