Thanks for your reply. 

I have checked the tags, master and worker tags match. 

The deadlock happens in the case of 2 tasks scheduled on one processor. 

If there is only one task on one processor, there is no deadlock.

It works well.

The master is resopnsible for scheduling tasks to workers, which need to run 

the assigned tasks and feedback results to master. 

if I assign one task to each worker, it works well. 

But, when I increase the # of task to 2 on worker node, it is deadlock. 

The master only schedules 2 tasks to one worker in order to simplify the analysis

for the poential deadlock. 

The worker can receive the 2 tasks and run them, but the master cannot get the 

results from worker.

the main idea: 

counter=0;

totalTaskNum =2;

while (counter < totalTaskNum )

{

TaskPackage myTaskPackage(world); 

world.isend(node1, downStreamTaskTag, myTaskPackage); 

recvReqs[counter] = world.irecv(node1, upStreamtaskTag, taskResultPackage[counter]);

counter++;

}

world.wait_all(recvReqs, recvReqs+(totalTaskNum));

worker (node 1): 

while(1)

{

TaskPackage workerTaskPackage(world); 

world.recv(node0,downStreamTaskTag,  workerTaskPackage ); 

do it local work;

world.isend(node0, upStreamTaskTag, workerTaskPackage); 

if (no new task)

break;

}

My code has many classes, I am trying to find out how to cut out the main part from it. 

Any help is appreciated. 

thanks

Jack