CPP problem

[Peter Eisentraut]

Can someone comment on this?  The user had CPP defined in the environment
and pointed to a less-than-optimal version.  A number of CPP related
autoconf tests gave losing results, but for compilation gcc was used,
whose preprocessor has different properties and so things went wrong.

It looks like a bit of a design problem, but I'm not sure whose.

We use Autoconf 2.13.

-- 
Peter Eisentraut   address@hidden   http://funkturm.homeip.net/~peter

---------- Forwarded message ----------
Date: Tue, 17 Jul 2001 18:07:38 +0200
From: Kovács Péter <address@hidden>
To: Peter Eisentraut <address@hidden>
Cc: address@hidden
Subject: RE: [PORTS] Unable to build on Solaris 2.6

In my .profile I had CPP set to /usr/ccs/lib/cpp and autoconf probably used
this instead of gcc -E. On the other hand, at build time gcc is probably not
looking at the CPP environment variable for a preprocessor, but simply uses
its own. After all I would not say that this is a bug in autoconf, but being
a bit more idiot-proof :) would certainly not hurt.

Peter

-----Original Message-----
From: Peter Eisentraut [mailto:address@hidden
Sent: Tuesday, July 17, 2001 5:21 PM
To: Kovács Péter
Cc: address@hidden
Subject: RE: [PORTS] Unable to build on Solaris 2.6


Kovács Péter writes:

> Apparently, the configure script did not correctly recognize the
stringizing
> capibality of the preprocessor. (Should I report this as a bug in
> autoconf???)

Hmm.  I've never seen this test give a wrong result.  Can you check *why*
it does that?

-- 
Peter Eisentraut   address@hidden   http://funkturm.homeip.net/~peter