Inter procedural optimization

These patches enable inter procedural optimization for the GCC and Intel compilers. They may also work for the LLVM/clang.

Further response to #104 (closed).

Matthieu, I tried these and didn't see any improvements. Can you also check please.

Thanks !

The improvement is not big but it might help. It can't harm for sure.

I have tried a "dry run" in which all the tasks directly return without doing anything (note that they don't get optimised out). This essentially measures the overhead due to the scheduler and the construction of the tasks. The ipo did reduce the time of such a run and this is where the ipo is expected to do its job since the functions are scattered over multiple compilation units (engine.o, scheduler.o and runner.o mainly).

Status changed to merged

mentioned in commit d3f5f664