Adventures in OpenCL: experimenting with #includes on Apple OpenCL Implementation
http://enja.org

If you build and run this program, you will get the expected results (two arrays are summed as a linear combination).
If you modify inc.cl after building and running, the modification will not be detected by the Apple OpenCL compiler.
Once you modify a.cl after modifying inc.cl the changes will be picked up... but only for a.cl (you must also modify b.cl in some way).

I even found that changing the modification back will result in the previous behavior.

None of this happens on NVIDIA stack on Ubuntu, i.e. it detects changes in #included files...


Build:
make a build directory and build the tutorial
mkdir build
cd build
cmake ..
make

execute the example program
./inc.x


