Declare platform-independent byteswapping functions QINLINE

[smcv]

In addition to enabling inlining for these very simple functions,
this suppresses warnings from gcc 6:

…/shared/qcommon/q_platform.h:302:13: warning: 'void CopyShortSwap(void*, const void*)' defined but not used [-Wunused-function]

---

The travis-ci tests on my branch demonstrate this building successfully on g++ and clang, but it would be good if someone could try this on MSVC before merging (hopefully the Appveyor builds do that for us automatically).

static QINLINE provokes warnings from clang (because QINLINE already contains static for clang), but that isn't a regression: there's a lot of that already. As a follow-up change, it would be good if the OpenJK maintainers could choose one of these options:

• define QINLINE to inline or __inline (whichever is supported) on all compilers, the same as ID_INLINE in ioquake3; make sure that every C-style function that is declared QINLINE is static QINLINE (this resembles what is done in ioquake3)
• define inline to __inline if the compiler doesn't support the inline keyword, and replace all QINLINE or static QINLINE with static inline (this resembles what is commonly done in Unix-centric open source projects like GLib)
• define QINLINE to static inline or static __inline (whichever is supported) on all compilers; make sure that every C-style function that is declared QINLINE is not static QINLINE

My vote would be the first one, since OpenJK seems to be 90% of the way to it already.

[xycaleth]

Looks good. Would you mind submitting another PR to implement your first option? It think that sounds like the best way to go here.

[smcv]

Would you mind submitting another PR to implement your first option?

Sure, will try to get to that soon.