[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Documentation and other things

From: Paul D. Smith
Subject: Re: Documentation and other things
Date: Wed, 14 Feb 2001 14:26:12 -0500

%% Magnus Fromreide <address@hidden> writes:

  mf> In the info maual section Phony Target it is recommended that
  mf> submakes are started using a pattern where each subdir is used as
  mf> a target in order to make paralelization easier but this makes it
  mf> very hard to communicate to the submake what the target should be.

You can use $(MAKECMDGOALS).

  mf> In the section Recursive use of `make' no reference is made to
  mf> that discussion and the only case that is treated is the one where
  mf> there is only a single subdirectory.

  mf> Which method should be preferred here?

I don't see where they conflict?

  mf> Could the documentation be clarified?

Undoubtedly.  I'll look at it.

  mf> On a completly different point, is there any way to make GNU make
  mf> turn off the automatic rebuilding of include files


  mf> Additionally it would be nice if there was a shorter version of
  mf> --no-print-directory -- the obvious fix would seem to be that not
  mf> turn it on for recursive makes, if there is a need for it then the
  mf> short option -w turns it on. By the current behaviour there is no
  mf> way to specify that it should be turned off save for that long
  mf> name.

I guess I don't see why this is such a big problem--you don't need it on
the command line, since the top-level make doesn't print that message
anyway, and you can add it to MAKEFLAGS in the Makefile itself so what
difference does it make how long the option is?

  mf> Oh yes, if there is no file to be included but make knows how to
  mf> build it then that isn't strictly an error and thus it shouldn't
  mf> be reported but it still is reported,

That's because the time/place when make knows there is no file to be
included and the time/place that make knows it can rebuild it are _far_
removed from each other.

  mf> will the new read.c version fix this as well?

I doubt it.  The read.c change is quite trivial (don't print a message
but instead set some kind of flag); it's the change in the other places
that would be needed to print the message once make discovers that it
really _can't_ build the file which are the hard parts.

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

reply via email to

[Prev in Thread] Current Thread [Next in Thread]