Steve Woodford comments;

> on ARM, a PC-relative load just to find 
> the address of the variable. This is all inlined code, so even small 
> functions like this contrived example:

[ ... inline designation ends up with larger number of insns ... ]

I've been concerned that recent trend of excessive use of inline code
by *.h.  It does not necessarily contribute the code compactness, the
ease of maintenance, future growth and HW adaptation.  I would
toss the vote for "do-not-make-spl-inline" discipline since "it's _the_
place of possible improvements by funcy tricks and insns."

> Using the above two suggestions in tandem, the spl(9) implementation will 
> be much smaller, the code will scale to multiple hardware masks much 
> more easily, and you may find some more opportunities to factor out 
> common/shared code.

That's what I expect common/re-targetable example for future growth.

Toru Nishimura/ALKYL Technology