Barrier synchronization among a group of process works as follows: For each process in the group, after a process enters the code segment of the barrier, it will not leave the code segment until every other process in the group has entered its respective barrier code segment. Write an MPI C code using MPI_Send()'s and MPI_Recv()'s to implement barrier synchronization among P processes (for general) so that the number of steps of (possibly parallel) send-receive operations is no more than 2 log_2 P.

Need this problem in C Language please, with steps and comments