autoconf tenderfoot

Brian Bruns camber at ais.org
Sun Jun 2 14:27:53 EDT 2002


On Sat, 1 Jun 2002, James K. Lowden wrote:

> All, 
> 
> What I know about autoconf ends approximately at its "f".  So I'm probably
> doing something wrong.  What I want to know is, Is it sufficient for me to
> run "./configure", or should I run autoconf/automake/oughtathink first to
> generate a configure script?  
> 
> On my NetBSD system, I'm using the CVS version to run down some things,
> spurious (I hope) warnings and such.  I'm looking at this:
> 
> In file included from ../../include/tds.h:35,
>                  from server.c:21:
> /usr/include/netinet/in.h:170: warning: `INADDR_NONE' redefined
> ../../include/config.h:13: warning: this is the location of the previous
> definition

config.h.in is created by "autoheader" and config.h by the configure 
script.  What should happen is that configure checks whether INADDR_NONE 
exists already and optionally includes it in config.h.  So most likely 
there is something wrong with the check for INADDR_NONE not detecting its 
existance. The check for it is in acinclude.m4 as AC_HAVE_INADDR_NONE.

So, now the trick is to figure out why the macro isn't working...

Yours seems to define this in netinet/in.h but there is no check for that 
in the configure.in  So, my reccomendation would be to add

AC_CHECK_HEADERS(netinet/in.h)

just above AC_HAVE_INADDR_NONE (the test for INADDR_NONE has a 
conditional include of in.h) and run ./autogen.sh to regenerate everything 
and give it a go.  We might have to conditionally include it where 
INADDR_NONE is used too.

Brian

> which, bless his heart, is pretty comprehensible.  But what to do? Reading
> between the lines of Brian's similar question:
> 
> > is MAXPATHLEN always in sys/param.h if not already defined?  If so, we 
> > could add AC_CHECK_HEADERS(sys/param.h) to configure.in and then add an 
> > #ifdef HAVE_SYS_PARAM to config.c to include it.
> 
> I looked at configure.in, which led me to autoconfincognita.  
> 
> I'm not sure if I'm supposed to have to add something to some .in file, or
> whether the tools are supposed to infer the right answer and aren't
> (possibly because of some misconfiguration in my system).  
> 
> I'm not afraid to read, but if someone could guide me to the right
> reading, I'd be very much obliged.  
> 
> TIA.
> 
> --jkl
> 




More information about the FreeTDS mailing list