The gtk+ ./configure script displays:
[...]
checking X11/extensions/XShm.h... yes
checking for Xinerama support on XFree86... checking for XineramaQueryExtension in -lXinerama... no
no
checking for XRRUpdateConfiguration in -lXrandr... no
[...]

You can see a "no" alone on its own line. This is because we have two checks imbricated: checking for XineramaQueryExtension is inside checking for Xinerama.

With the patch I propose we have:
[...]
checking X11/extensions/XShm.h... yes
checking for XineramaQueryExtension in -lXinerama... no
checking for Xinerama support on XFree86... no
checking for XRRUpdateConfiguration in -lXrandr... no
[...]

The patch is very simple and just move the call to AC_MSG_CHECKING() a but later in the script.

Regards,

--- gtk+/configure.in	2006-05-01 22:26:42.000000000 +0000
+++ gtk+.new/configure.in	2006-05-09 15:26:31.000000000 +0000
@@ -1247,14 +1247,14 @@
   case "$host" in
       *-*-solaris*)
           # Check for solaris
-	  AC_MSG_CHECKING(for Xinerama support on Solaris)
-	  
 	  have_solaris_xinerama=false
 	  AC_CHECK_FUNC(XineramaGetInfo,
 	      [AC_CHECK_HEADER(X11/extensions/xinerama.h,
   		  [have_solaris_xinerama=true], :,
 		  [#include <X11/Xlib.h>])])
 		
+	  AC_MSG_CHECKING(for Xinerama support on Solaris)
+	  
           if $have_solaris_xinerama ; then
             AC_DEFINE(HAVE_SOLARIS_XINERAMA)
 	    AC_DEFINE(HAVE_XINERAMA) 
@@ -1265,8 +1265,6 @@
           ;;
       *)
 	  # Check for XFree
-          AC_MSG_CHECKING(for Xinerama support on XFree86)
-	  
           have_xfree_xinerama=false
           if $PKG_CONFIG --exists xinerama ; then
              have_xfree_xinerama=true
@@ -1279,6 +1277,8 @@
                       [#include <X11/Xlib.h>])])
           fi
 
+          AC_MSG_CHECKING(for Xinerama support on XFree86)
+	  
           if $have_xfree_xinerama ; then
             AC_DEFINE(HAVE_XFREE_XINERAMA)
 	    AC_DEFINE(HAVE_XINERAMA)