Re: do phony deps take "precedence" over regular deps?

[Robert P. J. Day]

On Mon, 7 Feb 2005, John Graham-Cumming wrote:

> On Mon, 2005-02-07 at 09:54, Robert P. J. Day wrote:
> > $(Program): build_msg (OBJECTS) $(BUILTINS_DEP) ...
> >     ... regular commands here ...
> >
> > .PHONY: build_msg
> > build_msg:
> >     @printf ... building $(Program) ...
> >
> > the text reads:
> >
> >   "Because the printf is in a phony target, the message is printed
> > immediately before any prerequisites are updated."
> >
> > really?
>
> No.  Just try an experiment:
>
>     .PHONY: all
>     all: real-target phony-target other-real-target
>
>     real-target: ; @echo $@
>     other-real-target: ; @echo $@
>
>     .PHONY: phony-target
>     phony-target: ; @echo $@
>
> And you'll see the output is
>
>     real-target
>     phony-target
>     other-real-target
>
> Try permuting the prerequisites of all and you see that Make
> traverses them left to right.  PHONY only affects whether the
> prerequisite will be updated or not.

i suspected as much, but you can see how the way that sentence is
phrased in the book would lead someone to a noticeably different
conclusion, right?

certainly, if you were doing a parallel make, make would have the
freedom to process the dependencies any way it wanted, no?

rday