Delete inline optimization due to lcov issue:
Coverage test report of EFL API can't check called or not inline function.
Uh ? We do call alignof in a lot of place which may be part of the critical path in some case. It is also a very small function which mostly does math and should inline quite nicely. Now removing this just for a coverage report issue isn't a good enough reason for me. Please add benchmark and make sure that there is no side effect of removing that inline (Also it is unlikely that by itself work as there is an inline directive also in eina_inline_mempool.x).