[Cbe-oss-dev] A question for the MARS tutorial example 7.2

[XU,HONGSHENG]

 I am a graduate student working with my advisor to promote the 
use of Linux on PS3 for
optimization purpose. Now I am reading the help file for the MARS 
projects, however I encounter a problem in understanding the 
tutorial example 7.2.
    In the host program, on line 42-43, there are two waits for 
the sub-task 2_0 and 2_1 which are initiated by task 1, but in the 
task 1 program, there are the same waits for the sub-task 2_0 and
2_1 on line 18-19 . As my understanding, the host program waits 
for task 1 to finalize, and task 1 waits for sub-task 2_0 and 2_1 
to finalize. So it seems that host program does not need to wait
for sub-task 2_0 and 2_1 to finalize, since it is guaranteed 
through the codes of task 1. I am wondering whether there is any 
specific purpose for doing so here or it is just redundant.


Thank you