Re: order-only dependency and transitivity

[Paul D. Smith]

%% Ross Boylan <address@hidden> writes:

  >> >> $(SOURCES): m.touch
  >> >> 
  >> >> m.touch: m.web
  >> >>        ftangle -F m.web
  >> >>        @touch $@

  rb> "Every non-.PHONY rule must update a file with the exact name of its 
target."

  rb> $(SOURCES) is the target in the line
  rb> $(SOURCES): m.touch
  rb> yet the $(SOURCES) are not updated by the rule.  Further, at least
  rb> some members of $(SOURCES) are likely not to be updated by actions of
  rb> other rules (the one for m.touch) in the makefile (specifically,
  rb> ftangle -F).

  rb> But perhaps since this rule only expresses a dependency, and there
  rb> are no other rules with $(SOURCES) targets, we're in a different
  rb> ballpark.

Correct; this is just a dependency definition.  A rule would have a
command associated with it.

  rb> Though I didn't have that rule specifically in mind, I did think
  rb> I'd get in some trouble if some of $(SOURCES) were permanently
  rb> older than m.touch.  Apparently not...

Make doesn't know about "permanently"; it has no concept of state.  All
it knows is that some files are older, and some are newer; in a
dependency definition like the one above where multiple files appear on
the left-hand side each one is considered individually.

  rb> Ah.  I didn't know such caching was going on.  Does the "behind
  rb> the back" ness derive from the fact that the files do not appear
  rb> in the command line for ftangle, or in the fact that the relevant
  rb> rule doesn't list them as targets?

The latter.  Make doesn't know anything about command lines: it doesn't
parse them in any way except expanding variables [*].  It just passes
them to the shell.

  rb> It might be useful at least to document it.  I didn't notice any
  rb> mention of it in the make info pages, and there is no index entry
  rb> for cache or caching.

The theory is that the cache is an implementation detail which is
invisible to the user.  Unfortunately in some situations that's not true
(enough).

  rb> In my particular case it seems there is a solution that doesn't
  rb> require disabling the cache.

Yes.

  rb> A final general comment on the documentation: I couldn't tell from
  rb> it exactly when out-of-dateness was determined.  For example, it
  rb> could be done once before the makefile is processed; it could be
  rb> done just at the time a rule is checked; it could be done some
  rb> other way.  It would be very helpful to say explicitly how this
  rb> happens.  I suppose the caching makes that a little tricky.

There is really only one possible answer for this, though:
out-of-date-ness for a given target is evaluated after all its
prequisites have been determined to be up-to-date, to decide whether the
target needs to be updated.  I suppose this could be made more explicit
in the manual.

-- 
-------------------------------------------------------------------------------
 Paul D. Smith <address@hidden>          Find some GNU make tips at:
 http://www.gnu.org                      http://make.paulandlesley.org
 "Please remain calm...I may be mad, but I am a professional." --Mad Scientist