1 / 30

Inter Process Communication & Timers

Inter Process Communication & Timers. Time.h (page R:Ch9 pp302-320). #include <time.h> time_t time(time_t *calptr); Epoch: 00:00 (midnight), Jan 1, 1970 GMT Day is 86,400 seconds time_t is usually a long If the long is 32 bits, time overflows in 2038 Extensions POSIX:XSI microseconds

cybil
Download Presentation

Inter Process Communication & Timers

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Inter Process Communication & Timers

  2. Time.h (page R:Ch9 pp302-320) #include <time.h> time_t time(time_t *calptr); Epoch: 00:00 (midnight), Jan 1, 1970 GMT Day is 86,400 seconds time_t is usually a long If the long is 32 bits, time overflows in 2038 Extensions POSIX:XSI microseconds POSIX:TMR nanoseconds

  3. Timing a function #include <stdio.h> #include <time.h> void function_to_time(void); int main(void) { time_t tstart; tstart = time(NULL); function_to_time(); printf(“function_to_time took %f seconds of elapsed time\n”, difftime(time(NULL), tstart)); return(0); }

  4. POSIX XSI struct timeval time_t tv_sec; /* seconds since the Epoch*/ time_t tv_usec /* and microsoeconds*/ #include <sys/time.h> int gettimeofday(struct timeval *restrict tp, void *restrict tzp); tzp is null, historical

  5. P307 Measure running time using gettimeofday #include <stdio.h> #include <sys/time.h> #define MILLION 1000000L void function_to_time(void); int main(void) { long timedif; struct timeval tpend; struct timeval tpstart; if (gettimeofday(&tpstart, NULL)) { fprintf(stderr, “Failed to get start time\n”); return 1; }

  6. Measure running time using gettimeofday function_to_time(); /* timed code goes here */ if (gettimeofday(&tpend, NULL)) { fprintf(stderr, “Failed to get end time\n”); return 1; } timedif = MILLION*(tpend.tv_sec - tpstart.tv_sec) + tpend.tv_usec – tpstart.tv_usec; printf(“The function_to_time took %ld microseconds\n”, timedif); return 0; }

  7. Sleep & nanosleep #include <unistd.h> unsigned sleep(unsigned seconds); sleep interacts with SIGALRM #include <time.h> Int nanosleep(const struct timespec *rqtp, struct timespec *rmtp); Resolution of CLOCK_REALTIME is of order 10ms Does not affect SIGALRM

  8. Producer Consumer • Problem occurs in system & application programming • e.g. Web Server dispatches incoming web requests to a waiting process(es) • e.g. GUI events from keyboard,& mouse are queued by O/S and consumed by applications.

  9. insertPtr removePtr Producer-Consumer Problem • Producers insert items • Consumers remove items • Shared bounded buffer • Efficient implementation is a circular buffer with an insert and a removal pointer.

  10. Challenge • Prevent buffer overflow • Producer inserts too many items and the buffer overflows • Prevent buffer underflow • Producer inserts an item • Consumer removes an item • Consumer removes another item • Proper synchronization • Mutual Exclusion • Progress • Bounded wait • i.e. Prevent deadlock

  11. 2 Counting Semaphores and a mutex • Counting semaphore to count # items in buffer • Counting semaphore to count # free slots • Mutex to protect accesses to shared buffer & pointers.

  12. Assembling the solution • sem_wait(slots), sem_post(slots) • sem_wait(items), sem_post(items) • mutex_lock(m) mutex_unlock(m) • insertptr=(insertptr+1) % N • removalptr=(removalptr+1) % N • Initialize semaphore slots to size of buffer • Initialize semaphore items to zero.

  13. Pseudocode getItem() • sem_wait(items) • mutex_lock(mutex) • result=buffer[ removePtr ]; • removePtr=(removePtr +1 ) % N • mutex_unlock(mutex) • sem_post(slots)

  14. Pseudocode putItem(data) • sem_wait(slots) • mutex_lock(mutex) • buffer[ insertPtr]= data; • insertPtr=(insertPtr + 1 ) % N • mutex_unlock(mutex) • sem_post(items)

  15. putItem(data) { mutex_lock(mutex) sem_wait(slots) buffer[ insertPtr]= … insertPtr=… sem_post(items) mutex_unlock(mutex) } getItem() { mutex_lock(mutex) sem_wait(items) result=buffer[ removePtr ]; removePtr=… sem_post(slots) mutex_unlock(mutex)} Analysis#1 What's the precise problem?

  16. putItem(data) { mutex_lock(mutex) #2 sem_wait(slots) buffer[ insertPtr]= … insertPtr=… sem_post(items) mutex_unlock(mutex) } getItem() { mutex_lock(mutex) sem_wait(items) BLOCKS #1 result=buffer[ removePtr ]; removePtr=… sem_post(slots) mutex_unlock(mutex)} Deadlock e.g Consumer waits for producer to insert a new item but Producer is waiting for Consumer to release mutex

  17. putItem(data) { sem_wait(slots) mutex_lock(mutex) buffer[ insertPtr]= … insertPtr=… sem_post(items) mutex_unlock(mutex) } getItem() { sem_wait(items) sem_post(slots) mutex_lock(mutex) result=buffer[ removePtr ]; removePtr=… mutex_unlock(mutex) } Analysis#2

  18. putItem(data) { sem_wait(slots) mutex_lock(mutex) buffer[ insertPtr]= … insertPtr=… sem_post(items) mutex_unlock(mutex) } getItem() { sem_wait(items) sem_post(slots) mutex_lock(mutex) result=buffer[ removePtr ]; removePtr=… mutex_unlock(mutex) } Buffer overflow when reader removes item from a full buffer: Producer inserts item too early

  19. First Reader-Writer Problem • readers: read data • writers: write data • Rule: • Multiple readers can read the data simultaneously • Only one writer can write the data at any time • A reader and a writer cannot in critical section together. • Locking table: whether any two can be in the critical section simultaneously

  20. First Readers-Writers Problem(Readers have priority) • Let processes reading do so concurrently • Let processes writing do so one at a time • Introduce semaphores • Semaphore mutex = 1; • Semaphore wrt = 1; • int readc = 0; /* reader counter */

  21. Writer while (TRUE) { Think_up_data(); /*noncritical section*/ lock(&wrt); do writing unlock(&wrt); } Reader while (TRUE) { lock(&mutex); readcount =readcount+1; if readcount == 1 then lock(&wrt); unlock(&mutex); do reading lock(&mutex) readcount=readcount-1; if readcount == 0 then unlock(&wrt); unlock(&mutex); Use read data }

  22. Topics: Hardware/OS Overview • Chapter 1.1-1.7 (Stallings) • Chapter 2.1-2.2 (Stallings) • Chapter 1 (R&R) • Keywords Need to know • Processors – registers • Interrupts and Interrupt Handling • Polling and Programmed I/O • Basic Memory principles • Kernel mode, user mode • Multiprogramming, uni-programming • Time sharing • Buffer Overflow and security

  23. Topics: Processes • Chapter 3.1-3.4 (Stallings) • Chapter 2 and 3 (R&R) • Keywords need to know • What is process? • What is the difference between process and program? • What is the program image layout? • Understand argument arrays • What does it mean to have a thread-safe function? • What is the difference between static and dynamic variables? • What are the major process states? • What is the difference between dispatcher and scheduler? • What is PCB? • What happens when process switches from running to ready state? • What is process chain, process fan?

  24. Topics: Threads • Chapter 12.1-12.5 (R&R) • Chapter 4.1, 4.5 (Stallings) • Keywords need to know • What is the difference between processes and threads? • What is the difference between user-level threads and kernel-level threads? • Detaching and joining threads • What happens if you if you call exit(1) in a thread? • What is a graceful way to exit a thread without causing process termination?

  25. Topics: Concurrency (Mutual Exclusion) • Chapter 14.1-14.3 (R&R) • Chapter 5.1-5.3 (Stallings – and don’t forget Appendix A about the Software Solutions) • Keywords need to know • What are the four conditions to provide appropriate synchronization and enter critical region? • What is the difference between counting semaphore and mutex? • What do sem_wait and sem_post do? • How can counting semaphores be implemented using binary semaphores? • How can test_and_set be used for synchronization? • How can you make a function atomic? • Consider increment (i++) and decrement function (i--). How do you ensure that race condition does not occur on the shared variable ‘i’ when two processes use them at the same time?

  26. Topics: Thread Synchronization • Chapter 13.1-13.2 (R&R) • Keywords to know • What are mutex locks? • How do you initialize mutex locks? • When would you use mutex instead of counting semaphore? • When would you use counting semaphore instead of mutex? • Are mutex functions interrupted by signals?

  27. Topics: Scheduling • Chapter 9.1-9.2 Scheduling (Stallings) • Keywords need to know • Scheduling policies • FCFC, SJF, Round Robin, Priority Scheduling • Preemptive vs. Non-Preemptive Scheduling • Queues in Process management – what is ready queue? How are process states related to process management queues? • What is average waiting time? • What is the difference between process waiting time and turn-around time?

  28. Topics: Signals • Chapter 8.1-8.5 (R&R • Keywords need to know • Signal basic concepts – generating signals, blocked, pending signals, delivered signals, ignored signals, … • What is signal mask and what are the operations to modify signal mask? • What is signal handler? • What is the role of sigaction? • How do you wait for signals?

  29. Topics: Timers • Chapter 9.1-9.3 (R&R) • Keywords need to know • Understand what the various time functions are for • Gettimeofday • Understand the different clock resolutions • Sleep function

  30. Topics: Classical Sync Problems • Chapter 5.3 and 6.6 (Stallings ) • Keywords need to know • What is the producer-consumer problem? • What are the various semaphores in the producer/consumer solution for? • What is the dining philosopher problem? • What is the danger of a simple solution for dining philosopher problem?

More Related