Just some comments on the tutorial from: http://developer.amd.com/tools-and-sdks/heterogeneous-computing/amd-accelerated-parallel-processing-...
1. It would be very useful to have source code download links
2. the 'gcc' compile command will not work, as the example uses C++ header files - you will need to use g++
3. In line with 1. above, copying and pasting the code yields pretty unreadable code.
4. Compiling on linux, you may need to use '#define CL_USE_DEPRECATED_OPENCL_1_1_APIS' at the top of the main cpp file (I had to do this on Debian 7 64-bit, using Nvidias opencl implementation and header files)
Also, thanks for the tutorial!
I have attached the full source code, as well as a bash script to compile on linux.