Example#pragma omp parallelint id $=$ omp$\_$get$\_$ thread$\_$num $()$ ;Alid$]=$ big$\_$call $($id$)$; $//$some large function#pragma omp barrier $//$Tells all threads to wait at this pointBlid$]=$ big$\_$calc$2($id$,$ A$)//$barrier is used so that A is finished$//$updating before using.In the case of mutual exclusion, a section of code is defined such that only one thread can access it at a time. This section of code is called a critical section. If a thread reaches the critical section but a $2$nd thread is still executing it$,$ the first thread has to wait to execute the critical section. The following statement is used to define that block of code:#pragma omp criticalExample:float res;#pragma omp parallelfloat B; int i$,$ id$,$ nthrds;B $=$ big$\_$job $($i$)$;#pragma omp critical $//$critical sectionres $+=$ consume $($B$)$; $//$only one thread will add to res at a timeAlternatively, #pragma omp atomic may be used instead of #pragma omp critical. Atomic works the same as critical but uses hardware that may speed up specific operations. This should only be used to do quick updates in memory $($e$.$g$.$ x$++,--$x$).$ When in doubt, use critical because is the more general version.ExerciseMake copies and modify your parallelized $"$pi$.$c$"$ program $3$ times. First use omp barrier and record the processing time for $1,2,3,$ and $4$ threads. Repeat this for omp critical and omp atomic. Compare their processing times. Which is the best method to use?Hint: When using critical and atomic update sum without array to avoid false sharingDashboard$68$CalendarOTo Do