[Bacula-users] Re: configure script OS/Distribution detection

[Alan B. <aj...@ms...>]

On Fri, 6 Feb 2004, Kern Sibbald wrote:

> The OS testing code in autoconf/aclocal.m4 was written by Riccardo
> Facchetti for apcupsd. It used to work quite fine for all those
> machines, but I have not touched it in a *very* long time.  

Hrm, I appear to be chasing a red herring.

The configure script writes everything important to confdefs.h, but after
it's run, that file disappears (no debugging output...)

To compound matters, there's no actual result written to the screen or the 
config.out file from the OS tests.

This patch will fix that. There are probably better ways of achieving it 
however and I can't see where confdefs.h is being deleted, so I've just 
added a cat line near the end to dump the file to screen. You may wish to 
snip the 3rd chunk accordingly

$ diff -u configure.orig configure

--- configure.orig      2004-02-06 10:39:04.000000000 +0000
+++ configure   2004-02-06 10:53:58.000000000 +0000
@@ -4131,8 +4131,8 @@
 fi
 fi

-echo "$as_me:$LINENO: result: \" \"" >&5
-echo "${ECHO_T}\" \"" >&6
+echo "$as_me:$LINENO: result: \"`tail -1 confdefs.h | cut -f2 -d_`\"" >&5
+echo "${ECHO_T}\"`tail -1 confdefs.h | cut -f2 -d_`\"" >&6


 # -----------------------------------------------------------
@@ -4206,8 +4206,8 @@
 else
         DISTNAME=unknown
 fi
-echo "$as_me:$LINENO: result: \" \"" >&5
-echo "${ECHO_T}\" \"" >&6
+echo "$as_me:$LINENO: result: \"$DISTNAME\"" >&5
+echo "${ECHO_T}\"$DISTNAME\"" >&6


 # -----------------------------------------------------------
@@ -19190,6 +19190,7 @@
 # If some macros were called several times there might be several times
 # the same #defines, which is useless.  Nevertheless, we may not want to
 # sort them, since we want the *last* AC-DEFINE to be honored.
+cat confdefs.h
 uniq confdefs.h | sed -n -f confdef2sed.sed >conftest.defines
 sed 's/ac_d/ac_u/g' conftest.defines >conftest.undefs
 rm -f confdef2sed.sed