[U-Boot] Nested Makefiles

Wolfgang Denk wd at denx.de
Tue Jun 21 00:23:23 CEST 2011


Dear Mike Frysinger,

In message <BANLkTim5nXA8fF=O24gHEGL+XyBTQLMJUA at mail.gmail.com> you wrote:
> On Thu, Jun 16, 2011 at 18:03, Wolfgang Denk <wd at denx.de> wrote:
> > I think it is fundamentally wrong to implement such a feature (let's
> > call it "terse make output") in the Makefiles of many projects, using
> > a lot of trickery and magic. =A0If a specific verbosity of make output
> > is needed, this should most naturally be implemented as a feature in
> > make itself - we already have make options like "-d" (for very verbose
> > output) or "-s" (for silent mode), so why not add a new verbosity
> > level which produces exactly the short output format you want?
> >
> >
> > So yes, patches are welcome, but these should go directly to the make
> > mailing lists / patch system, see
> > http://savannah.gnu.org/mail/?group=3Dmake
> 
> not to kick sand just for fun, but this is an example of you getting
> final veto power.  this has been requested by many people, and ive
> seen very few people against it (off the top of my head, i can only
> recall you, but i havent looked at previous threads to be sure).
> 
> i'm not saying your logic is without merit, just that in all
> practicality, i dont think it's going to happen the way you desire.

We should probably split this thread and adapt the Subject, as we are
discussing two completely independend topics here:

One is about "Nested Makefiles" (or recursive Makefiles), which the
subject mentions.  I have to admit that I don't have a clear opinion
about this yet, so I lean back and wait what the results of this
discussion might be.  When we see any code we have probably a better
base to discuss this.

The other topic is actually terse make output, and here indeed I
insist that the problem should be solved at the base, like I always
insisted that GCC / libgcc issues should not be worked around again
and again and in every software package that uses GCC but rather at
the base of the problem, in GCC itself.

Any other approach might just appear to need less efforts, but you
have to multiply this with the number of projects that do the same,
and the silly waste of efforts for multiple similar implementations.
In total, a clean solution at the correct place is not only less
effort, but the only mentally clean solution.

At least that's my 0.02 <insert your favorite unit of currency here>.

YMMV...

Best regards,

Wolfgang Denk

-- 
DENX Software Engineering GmbH,     MD: Wolfgang Denk & Detlev Zundel
HRB 165235 Munich, Office: Kirchenstr.5, D-82194 Groebenzell, Germany
Phone: (+49)-8142-66989-10 Fax: (+49)-8142-66989-80 Email: wd at denx.de
A dog always bit deepest on the veterinary hand.
                                    - Terry Pratchett, _Wyrd Sisters_


More information about the U-Boot mailing list