[Top][All Lists]

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

Re: Automake and makedepend

From: Ralf Wildenhues
Subject: Re: Automake and makedepend
Date: Sat, 14 May 2005 13:53:00 +0200
User-agent: Mutt/1.5.9i

Hi Stan,

Please try to answer below the text you refer to, thanks.

* Stan Guillory wrote on Fri, May 13, 2005 at 06:14:38PM CEST:
> Ralf Wildenhues <address@hidden> 5/13/2005 10:41:11 AM >>>
> > * Stan Guillory wrote on Thu, May 12, 2005 at 07:56:17PM CEST:
> > > 
> > > 1) Should the dependency tracking mechanism in automake work by
> > > default with Sun Forte C++ 5.4?
> > 
> > I only have access to a different compiler version.  What does yours
> > do after this series of commands (in a junk directory):
> > ?  Over here, the second one triggers the error
> > | CC: Suffix mismatch between -o and produced file (should produce .i)
> > 
> > which prevents the dashXmstdout depmode to work.

Never mind.  See below.

> > However, on this system
> >   depmode=makedepend source=a/ object=d/a.o libtool=no \
> >   sh -x path/to/depcomp CC -Ib -I. -c a/ -o d/a.o
> > 
> > works.  So, the question is whether `depcomp' needs to cut out any of
> > your $CPPFLAGS or $CXXFLAGS which could make makedepend barf.
> On the first invocation, I get the following:
> etc, i.e. a list of what I think I need. The errno was 0 after
> executing that sequence of commands.
> I got the same error you did when I executed the second CC invocation.
> What version of the compiler are you using, btw?

5.3, I think.  

Call me stupid, I just forgot how the dependency mechanism was chosen
again:  Basically, there are two classes of mechanisms: ones that are
fast (don't require separate invocation of either compiler or other
tool), and all others, considered slow.  You have to add
to your configure options in order to enable the slow ones.
All nicely documented, too (see configure --help).

Now my tests show that dashXmstdout is selected, and that should work
with your compiler as well.  Could you confirm that, please?


reply via email to

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