[PATCH] Define __NO_INLINE__ if not inlining (take 2)

[Neil Booth]

Kaveh R. Ghazi wrote:-

> No wait, where did that come from?  We're not supposed to alter the
> behavior of the compiler via the use of warning flags (which is what
> -pedantic is essentially.)
> 
> If you do this, then someone will inevitably write:
> 
> #ifdef __PEDANTIC__
> do something
> #else
> do something else
> #endif
> 
> which specifically breaks this contract WRT warning flags.

Hmm.  See PR 2677.  I admit I find your argument more persuasive;
I prefer to be conservative.  I don't know if Zack has changed his
mind in the light of your point.

Neil.