Re: [PATCH] log2.h: Macro-ize is_power_of_2() for use in BUILD_BUG_ON

[Roland Dreier <rdreier@cisco.com>]

 > We've had recurring struggles with various versions of gcc screwing up
 > constructs of this form and trying to emit the non-constant code when
 > the arg was clearly a compile-time constant.  One episode which comes
 > to mind was when we made changes to kmalloc().
 > 
 > Of course, that might not bite us in this case - it would need a lot of
 > coverage testing to find out.

Actually, after thinking about this a bit more I don't think this is
much of a risk in this case.  The bad case is where gcc gets the wrong
answer for __builtin_constant_p(), and in that case, we're left with the
status quo ante, ie a call to the original inline function implementation.

This will break a call to BUILD_BUG_ON() but that's visible at compile
time.  It's hard to think of a silent failure case that actually hurts
anything -- the macro and inline implementation are identical, except
that the macro implementation will work when gcc really needs a compile
time constant.

 - R.