Scribd
Upload
38 views

Systems Programming - Lab9

This document describes two problems for a lab assignment on concurrency control. Problem 1 involves using threads to calculate the square roots of integers in parallel. Problem 2 compares the runtime of a serial matrix summation algorithm to a parallel algorithm using threads. Students are instructed to write C programs to solve the problems, capture output, and measure runtimes for different input sizes.

Uploaded by

AAKASH VAGHELA
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
38 views

Systems Programming - Lab9

This document describes two problems for a lab assignment on concurrency control. Problem 1 involves using threads to calculate the square roots of integers in parallel. Problem 2 compares the runtime of a serial matrix summation algorithm to a parallel algorithm using threads. Students are instructed to write C programs to solve the problems, capture output, and measure runtimes for different input sizes.

Uploaded by

AAKASH VAGHELA
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 3

Systems Programming – Lab Manual

______________________________________________________________________________

Lab 9 – Concurrency Control

Write C program for each of the problem below. After executing each of the C
program, you will need to capture the output in text file format.
C program file will be named as StudentID_Lab9_x.c
Text file with output captured will be names as StudentID_Lab9_x.txt

Note: None of the below problems require thread synchronization mechanisms.


These assignments will teach how to divide the computation work among
different threads so that it can be done with improved run time.

Problem 1: Write a C program StudentID_Lab9_1.c using POSIX thread to compute


square root of integers from 0 to 99 and return an array of double sroot[100] with
computed value of square roots:
• Main thread will create 10 threads and stores threadID in an array. Main
thread must wait till all the threads are completed
• Each of the threads will calculate square root of 10 integers by calling
sqr_root function depending on which thread it is
o E.g. thread 0 (first thread) will calculate square root of 0 to 9, thread 1
(2nd thread) will calculate square root of 10 to 19 and so on.
o Once square root is of an integer is calculated it will stored in an array
at an offset pointed by the input value e.g. if thread 1 has finished
calculating square root of 15 then it will place calculated value at
sroot[15].
• While main thread is waiting for all the thread to finish, main thread should
display a short message to the user and then display the results of the
computation when they are ready.

Problem 2: Compare serial versus parallel algorithm runtime for matrix summation.
Parallel Algorithm:
Write a C program StudentID_Lab9_2.c using POSIX thread for summation of matrix
mat[ROWS][COLS] by a number of threads provided by user.
• Main thread accepts from user input number threads to use (num_threads).
num_threads must be less than ROWS
• Main thread create number of threads as given in num_thread.
• Each thread will:
o Each thread will be assigned specified number of rows of a matrix to
calculate the sum. i.e. number of rows assigned to each thread will be
ROWS/num_threads. if ROWS value is not multiple of num_threads
then you need to handle that situation by assigning additional rows to
some threads. e.g. ROWS = 10 and num_thread=4 then 2 rows each
will be first allocated to all 4 threads but then we are left with 2 more
rows which should be assigned to 2 out of 4 threads.
o Local calculated sum value must be returned back from thread to main
thread
• Main thread will access the sum value returned from each thread and display
the final total sum of the entire matrix.

Serial Algorithm:
Write the serial algorithm in StudentID_Lab9_2.c for summation of matrix. For
large matrices (e.g. 500x500 etc) compare the time difference between parallel
versus serial algorithm.

Get the runtime information for both Serial and Parallel version by adding the
following code in your programs:
You can measure the time it took to run the program using following code:
/* this function returns the time of day in microsecond */
uint64_t get_gtod_clock_time ()
{
struct timeval tv;
if (gettimeofday (&tv, NULL) == 0)
return (uint64_t) (tv.tv_sec * 1000000 + tv.tv_usec);
else
return 0;
}

/* At the beginning of main function declare variables */


uint64_t start_time_value, end_time_value, time_diff;
/* Get the start time */
start_time_value = get_gtod_clock_time();

-------- your code -------------------------

/* Get the end time */


end_time_value = get_gtod_clock_time();

/* Time difference */
time_diff = end_time_value - start_time_value;
/* display the difference */
printf("%"PRIu64"\n", time_diff);

Compare Runtime:
Note the different in output for time_diff value between Serial and Parallel version
an algorithm for different values of num_threads and ROWS/COLS

You might also like