Lab01: Global Broadcast

Objective
To distribute simple data to each process. Topics include scanf, MPI_Bcast, MPI_Send, MPI_Recv,
and decisions involving the rank of a process.

Background
Copies of variables are created for each distinct process.

Assignment

Create a file named share.c that reads in a single integer value from standard
input and broadcasts that value to all nodes. The input will only take place
on node zero, the machine you are sitting at; use the rank of the process to
check for this.
Version 1: Use MPI_Send and MPI_Recv. Send the integer from the root to each other process:
```
            P0
	  / | \    
	P1  P2 P3   etc.
 
```
Version 2: Use MPI_Bcast instead of MPI_Send and MPI_Recv.
MPI_Bcast works as a send from root node and a receive from the other nodes.
All processes need to see the MPI_Bcast.
```
            P0
	  / | \    
	P1  P2 P3   etc.
 
```
Version 3: Use MPI_Send and MPI_Recv, but send the integer from the process to process in a ring:
```
            P0 --> P1 --> P2 --> P3
 
```
For sample programs see
- MPI_Send and MPI_Recv: Lab01Starter.c and Lab01Starter.cpp
- MPI_Bcast: Lab01StarterB.c and Lab01StarterB.cpp
Sample Output
Enter a value: 37
Process 0 got value 37.
Process 13 got value 37.
Process 8 got value 37.
Process 1 got value 37.
Process 2 got value 37.
Process 4 got value 37.
Process 10 got value 37.
Process 6 got value 37.
Process 9 got value 37.
Process 5 got value 37.
Process 12 got value 37.
Process 7 got value 37.
Process 11 got value 37.
Process 3 got value 37.