Assignment 2: sum of square roots

Assignment 2: 
In this project, you are asked to write a program which uses three threads to concurrently
compute the sum of square roots from 1 to n, where n is a multiple of 3 and is specified in the
command line.
The program gets the value n from the command line argument (using argc / argv) and converts
it to an integer. Next, it creates two threads using pthread_create(). Then, the main thread
computes the sum of square roots from 1 to n/3, while the two child threads compute the sum of
the square roots from n/3+1 to 2n/3 and 2n/3+1 to n, respectively, at the same time as main
thread's calculation. All three partial sums should be added together to a shared global variable
by each thread at the end of execution. The main thread calls pthread_join() to wait for the
termination of the two child threads and then prints the result.
Below is an example of running your thread program:
spirit> thr 9999
sum of square roots: 666616.459197
You need to think about the location in program execution to call thread creation / mutex locking
etc carefully. Be sure that three threads do calculation concurrently, no excessive mutex locking
which creates unnecessary overhead, and is inefficient. Use mutex when necessary to avoid race
conditions which cause your program to produce incorrect answers randomly. Programs which
fail to do so will receive partial credit only even if the result is (sometimes) correct. You can
assume the input is always valid and won't cause overflows if you use standard data types (such
as int and doubles). You don't need to handle errors.
If you're not sure where to start, a guideline for the overall structure is provided below
(pseudocode). You don't have to use this structure if your own code can meet all requirements
above.
int start_points[4]
double grand_sum
mutex grand_sum_lock
function do_partial(start_point) {
 set sum to sum of square roots from start_point[0] to start_point[1]
 add sum to grand_sum, protected by grand_sum_lock
}
function main() {
 read n from argv and calculate start_points at 1, n/3+1, 2n/3+1, n+1
 start thread_1 using do_partial with start_points+1 as arg
 start thread_2 using do_partial with start_points+2 as arg
 call do_partial with start_points+0 as arg in main thread
 join thread_1 and thread_2
 print grand_sum as the final output
}