Threads

1. Threads

2. Processes versus Threads

3. Why Threads? • Processes do not share resources very well • Why? • Process context switching cost is very high • Why? • Thread: a light-weighted process • A sequence of execution

4. Threads: Lightweight Processes execution Environment (resource)

5. Real Life Example? • Process • “system programming” course • Different from “internet engineering” • Thread • homework, Reading, Self-assessment quiz • Each is a different “execution” • But all share • Content • Textbook • Personnel (TAs, instructors) • Affect each other

6. Review: threads vs. processes (created with fork)

7. Thread components • A thread has its own program counter and stack, but shares a number of resources with its process and other threads of the process: • address space: code and global variables • open files • semaphores • signals • timers • process ID • Thread specific resource: • Thread ID • Program counter • Register set • Stack space • Signal mask (later…)

8. Thread vs. Process • Each thread execute separately • Threads in the same process share resources • No protection among threads!!

9. Storage for Threads

10. Thread Model : Context Switch • Extensive sharing makes CPU switching among peer threads and creation of threads inexpensive compared to processes • Thread context switch still requires • Register set switch • But no memory management related work!!! • Why need to switch from one thread (process) to another? • Some thread (process) may block for I/O • Many threads (processes) share limited #CPUs

11. Thread State • Threads states are • Ready • Blocked • Running • Terminated • Why these states? • Threads share CPU • On single processor machine only one thread can run at a time • Threads can block waiting for a system call to be completed

12. Creating a Thread • When a new thread is created it runs concurrently with the creating process. • When creating a thread you indicate which function the thread should execute.

13. Normal function call

14. Threaded function call Difference from the previous one?

15. Benefits of Threads • Responsiveness • Multi-threading allows applications to run even if part of it is blocked • Resource sharing • Sharing of memory, files and other resources of the process to which the threads belong • Economy • Much more costly and time consuming to create and manage processes than threads • Utilization of multiprocessor architectures • Each thread can run in parallel on a different processor

16. Thread Creation vs. Process Creation Time in seconds for 50000 fork or thread creations

17. Threaded Function Call Detail • A function that is used as a thread must have a special format. • It takes a single parameter and returns a single parameter. • Can point to a structure, so in effect, the function can use any number of parameters. • http://www.llnl.gov/computing/tutorials/pthreads/

18. An Example The thread function casts and unpacks the first argument: void * myWorkerFunction(void *arg) {           int fd1 = ((int *)arg) [0]; int fd2 = ((int *)arg) [1]; }

19. Pthread Operations

20. pthread_* return values • Unlike most POSIX functions, they do not set errno but the value returned when an error occurs has the value that errno would have. • None of the POSIX thread functions ever return the error EINTR.

21. Example program • #include <phtread.h> • #include <thread.h> • #include <stdio.h> • void *threadex(void *); • int main() { • pthread_t tid; /* stores the new thread ID */ • pthread_create(&tid, NULL, threadex, NULL); /*create a new thread*/ • pthread_join(tid, NULL); /*main thread waits for other thread to terminate */ • return 0; /* main thread exits */ • } • void *threadex(void *arg) /*thread routine*/ { • int i; • for (i=0; i<5; i++) • fprintf(stderr, `Hello, world! \n''); • return NULL; }

22. Thread Usage: word processor • What if it is single-threaded?

23. Thread Usage: Web Server

24. Web Server • Rough outline of code for previous slide • (a) Dispatcher thread • (b) Worker thread

25. Designing Threaded Programs • Thread candidates? • Discrete, independent tasks which can execute concurrently • E.g. if routine1 and routine2 can be interchanged, interleaved and/or overlapped in real time, they are candidates for threading

26. Tasks Suitable for threading • Block for potentially long waits • Use many CPU cycles • Must respond to asynchronous events • Are of lesser or greater importance than other tasks • Are able to be performed in parallel with other tasks

27. Common Multi-thread Software Architectures • Manager/worker • a single thread, the manager assigns work to other threads, the workers. • Typically, the manager handles all input and parcels out work to the other tasks • Pipeline: • a task is broken into a series of suboperations, each of which is handled in series, but concurrently, by a different thread. • An automobile assembly line best describes this model • Peer • similar to the manager/worker model, but after the main thread creates other threads, it participates in the work

28. A Challenge: Making Single-Threaded Code Multithreaded • Conflicts between threads over the use of a global variable

29. A solution: Private Global Variables

30. Thread Packages • Kernel thread packages • Implemented and supported at kernel level • User-level thread packages • Implemented at user level

31. User-level Thread

32. User Level Threads • User-level threads without direct O/S support • Threads are invisible to the kernel • Simpler kernel implementation • Can only use one processor at a time • Implementation dependent • Some threads can block other threads • (or) Requires a special library of system calls to prevent blocking

33. User-level Threads • Advantages • Fast Context Switching: • User level threads are implemented using user level thread libraries, hence no call to OS and no interrupts to kernel • One key difference with processes: • when a thread is finished running for the moment, it can call thread_yield. • This instruction • (a) saves the thread information in the thread table itself, and • (b) calls the thread scheduler to pick another thread to run. • The procedure that saves the local thread state and the scheduler are local procedures, hence no trap to kernel, no context switch, no memory switch, and this makes the thread scheduling very fast. • Customized Scheduling

34. Kernel-level Threads • Kernel can schedule threads in addition to processes. • Multiple threads of a process can run simultaneously on multiple CPUs. • Synchronization more efficient than for processes (but less than for user-level threads). • Kernel-level threads can make blocking I/O calls without blocking other threads of same process

35. Kernel-Level Thread

36. Trade-offs (review)? • Kernel thread packages • Each thread can make blocking I/O calls • Can run concurrently on multiple processors • Threads in User-level • Fast context switch • Customized scheduling

37. Implementing Threads in User Space (old Linux) • A user-level threads package

38. Hybrid Implementations (Solaris) • Multiplexing user-level threads onto kernel- level threads

39. What’s POSIX Got To Do With It? • Each OS had it’s own thread library and style • That made writing multithreaded programs difficult because: • you had to learn a new API with each new OS • you had to modify your code with each port to a new OS • POSIX (IEEE 1003.1c-1995) provided a standard known as Pthreads • Unix International (UI) threads (Solaris threads) are available on Solaris (which also supports POSIX threads)

40. Pthreads--- POSIX Threads • It is a standard API • Supported by most vendors • General concepts applicable to other thread APIs (java threads, NT threads,etc). • Low level functions

41. Creating a thread with pthread • A thread is created with int pthread_create( pthread_t *restrict thread, const pthread_attr_t *restrict attr, void *(*start_routine)(void *), void *restrict arg); • The creating process (or thread) must provide a location for storage of the thread id. • The third parameter is just the name of the function for the thread to run. • The last parameter is a pointer to the arguments.

42. Restrict Keyword • One of the new features in the recently approved C standard C99 • This qualifier can be applied to a data pointer to indicate that • During the scope of that pointer declaration, all data accessed through it will be accessed only through that pointer but not through any other pointer. • It enables the compiler to perform certain optimizations based on the premise that a given object cannot be changed through another pointer • http://www.cellperformance.com/mike_acton/2006/05/demystifying_the_restrict_keyw.html

43. The Thread ID • pthread_t pthread_self(void) • Each thread has an id of type pthread_t. • On most systems this is just an integer (like a process ID) • But it does not have to be • A thread can get its ID with pthread_self • Compare two threads • int pthread_equal(thread_t t1, pthread_t t2)

44. Exiting and Cancellation • Question: • If a thread calls exit(), what about other threads in the same process? • A process can terminate when: • it calls exit directly • one of its threads calls exit • it returns from main() • it receives a termination signal • In any of these cases, all threads of the process terminate.

45. Exiting • When a thread is done, it can return from its first function (the one used by pthread_create) or it can call pthread_exit void pthread_exit(void *value_ptr);

46. Cancel that thread! • One thread can request that another exit with pthread_cancel • int pthread_cancel(pthread_t thread); • The pthread_cancel returns after making the request. • A successful return does not mean that the target thread has terminated or even that it eventually will terminate as a result of the request

47. Thread Attributes • Create an attribute object (initialize it with default properties) • Modify the properties of the attribute object • Create a thread using the attribute object • The object can be changed or reused without affecting the thread • The attribute object affects the thread only at the time of thread creation

48. Attribute Initialization and Deletion Initialize or destroy an attribute with: int pthread_attr_destroy(pthread_attr_t *attr); int pthread_attr_init(pthread_attr_t *attr);

49. Example: Create a detached thread int e, fd; pthread_attr_t tattr; pthread_t tid; if(e = pthread_attr_init(&tattr)) fprintf(stderr, "Failed to create attribute object: %s\n", strerror(e)); else if(e = pthread_attr_setdetachstate(&tattr, PTHREAD_CREATE_DETACHED)) fprintf(stderr, "Failed to set attribute state to detached: %s\n", strerror(e)); else if(e = pthread_create(&tid, &tattr, data, &fd)) fprintf(stderr, "Failed to create thread: %s\n", strerror(e));

50. The thread stack You can set a location and size for the thread stack. int pthread_attr_setstack( pthread_attr_t *attr, void *stackaddr, size_t stacksize ) (there’s a getstack function too) Some systems allow you to set a guard for the stack so that an overflow into the guard area can generate a SIGSEGV signal. int pthread_attr_getguardsize(const pthread_attr_t *restrict attr, size_t *restrict guardsize); int pthread_attr_setguardsize(pthread_attr_t *attr, size_t guardsize);