🚀 go-pugleaf

Nick Maclaren (nmm1@cus.cam.ac.uk) writes:
> The example that really baffles me is '(++i,j)+(j,++i)', because the
> standard assumes an ordering of operations without defining what that
> ordering is. ...

Right; the ordering is unspecified -- except that the left-hand ++i must
be evaluated before the left-hand j with a sequence point intervening,
and the right-hand j must be evaluated before the right-hand ++i with
a sequence point intervening.  Therefore the sequence

	1. right-hand j
	2. left-hand ++i
	3. right-hand ++i
	4. left-hand j

is possible.  (The ++ can be broken up into subsidiary operations, but it
isn't necessary for this analysis.)  The comma operators force sequence
points between steps 1 and 3, and between 2 and 4, but neither of these
sequence points has to be between 2 and 3.  Therefore i may be modified
twice without a sequence point intervening - whether this happens or not
is unspecified, since it depends on the evaluation order.  Therefore it
is unspecified whether or not the expression causes undefined behavior.
Therefore the expression does cause undefined behavior.$

See?  It's not so hard.

$ - This follows from the rule about implicitly undefined behavior that is
bundled in with the definition if "undefined behavior" in 3.16/1.6.
-- 
Mark Brader, msb@sq.com             "Logic is logic.  That's all I say."
SoftQuad Inc., Toronto                      -- Oliver Wendell Holmes

My text in this article is in the public domain.