matrix transpose example (#9)

* get index right for block and phase, but still have bugs

* try to modify transpose loop

* 4*4 matrix it works

* cleanup code and needs standard input interface

* add barrier after data init

* make index right

* add options for user

* change dataflow to async for local transpose to reduce cost

* the program works as Thomas's example for matrix transposition

* add iterations, testing sections and comments

* finalize comments, variables naming

* get index right for block and phase, but still have bugs

* try to modify transpose loop

* 4*4 matrix it works

* cleanup code and needs standard input interface

* add barrier after data init

* make index right

* add options for user

* change dataflow to async for local transpose to reduce cost

* the program works as Thomas's example for matrix transposition

* add iterations, testing sections and comments

* finalize comments, variables naming

* modify command line options