trouble with gsl on AIX

[Daniel Rohe <rohe at physik dot RWTH-Aachen dot DE>]

I've just tried to compile gsl-0.9 on AIX 4.3.2. 
while the configure script (--disable-shared) seems to run without any
serious 
complaints (the only one being a message like "shared: not found"), 
trying to compile returns the following: (!?)



	cd .  && CONFIG_FILES= CONFIG_HEADERS=config.h  /bin/sh ./config.status
creating config.h
config.h is unchanged
	make  all-recursive
Making all in gsl
	rm -f gsl*.h
	HEADERLIST="../gsl*.h ../*/gsl*.h";  for h in $HEADERLIST; do 
BASENAME=`basename $h`;  ln -s $h $BASENAME;  done
Target "all" is up to date.
Making all in utils
	/bin/sh ../libtool --mode=compile cc -DHAVE_CONFIG_H -I. -I. -I..    
-g -c placeholder.c
cc -DHAVE_CONFIG_H -I. -I. -I.. -g -c placeholder.c
echo timestamp > placeholder.lo
	/bin/sh ../libtool --mode=link cc  -g  -o libutils.la   placeholder.lo 
-lm
mkdir .libs
rm -fr .libs/libutils.la .libs/libutils.* .libs/libutils.*
ar cru .libs/libutils.a  placeholder.o
ranlib .libs/libutils.a
creating libutils.la
(cd .libs && rm -f libutils.la && ln -s ../libutils.la libutils.la)
Target "all" is up to date.
Making all in sys
	/bin/sh ../libtool --mode=compile cc -DHAVE_CONFIG_H -I. -I. -I..
-I..    -g -c minmax.c
cc -DHAVE_CONFIG_H -I. -I. -I.. -I.. -g -c minmax.c
echo timestamp > minmax.lo
	/bin/sh ../libtool --mode=compile cc -DHAVE_CONFIG_H -I. -I. -I..
-I..    -g -c prec.c
cc -DHAVE_CONFIG_H -I. -I. -I.. -I.. -g -c prec.c
echo timestamp > prec.lo
	/bin/sh ../libtool --mode=compile cc -DHAVE_CONFIG_H -I. -I. -I..
-I..    -g -c hypot.c
cc -DHAVE_CONFIG_H -I. -I. -I.. -I.. -g -c hypot.c
echo timestamp > hypot.lo
	/bin/sh ../libtool --mode=compile cc -DHAVE_CONFIG_H -I. -I. -I..
-I..    -g -c log1p.c
cc -DHAVE_CONFIG_H -I. -I. -I.. -I.. -g -c log1p.c
echo timestamp > log1p.lo
	/bin/sh ../libtool --mode=compile cc -DHAVE_CONFIG_H -I. -I. -I..
-I..    -g -c expm1.c
cc -DHAVE_CONFIG_H -I. -I. -I.. -I.. -g -c expm1.c
echo timestamp > expm1.lo
	/bin/sh ../libtool --mode=compile cc -DHAVE_CONFIG_H -I. -I. -I..
-I..    -g -c coerce.c
cc -DHAVE_CONFIG_H -I. -I. -I.. -I.. -g -c coerce.c
echo timestamp > coerce.lo
	/bin/sh ../libtool --mode=compile cc -DHAVE_CONFIG_H -I. -I. -I..
-I..    -g -c invhyp.c
cc -DHAVE_CONFIG_H -I. -I. -I.. -I.. -g -c invhyp.c
echo timestamp > invhyp.lo
	/bin/sh ../libtool --mode=compile cc -DHAVE_CONFIG_H -I. -I. -I..
-I..    -g -c pow_int.c
cc -DHAVE_CONFIG_H -I. -I. -I.. -I.. -g -c pow_int.c
echo timestamp > pow_int.lo
	/bin/sh ../libtool --mode=compile cc -DHAVE_CONFIG_H -I. -I. -I..
-I..    -g -c infnan.c
cc -DHAVE_CONFIG_H -I. -I. -I.. -I.. -g -c infnan.c
echo timestamp > infnan.lo
	/bin/sh ../libtool --mode=compile cc -DHAVE_CONFIG_H -I. -I. -I..
-I..    -g -c fdiv.c
cc -DHAVE_CONFIG_H -I. -I. -I.. -I.. -g -c fdiv.c
echo timestamp > fdiv.lo
	/bin/sh ../libtool --mode=link cc  -g  -o libgslsys.la   minmax.lo
prec.lo hypot.lo log1p.lo expm1.lo  coerce.lo invhyp.lo pow_int.lo
infnan.lo fdiv.lo  -lm 
mkdir .libs
rm -fr .libs/libgslsys.la .libs/libgslsys.* .libs/libgslsys.*
ar cru .libs/libgslsys.a  minmax.o prec.o hypot.o log1p.o expm1.o
coerce.o invhyp.o pow_int.o infnan.o fdiv.o
ranlib .libs/libgslsys.a
creating libgslsys.la
(cd .libs && rm -f libgslsys.la && ln -s ../libgslsys.la libgslsys.la)
Target "all" is up to date.
Making all in test
	/bin/sh ../libtool --mode=compile cc -DHAVE_CONFIG_H -I. -I. -I..    
-g -c results.c
cc -DHAVE_CONFIG_H -I. -I. -I.. -g -c results.c
"results.c", line 70.7: 1506-045 (S) Undeclared identifier va_alist.
"results.c", line 122.7: 1506-045 (S) Undeclared identifier va_alist.
"results.c", line 181.7: 1506-045 (S) Undeclared identifier va_alist.
"results.c", line 241.7: 1506-045 (S) Undeclared identifier va_alist.
"results.c", line 296.7: 1506-045 (S) Undeclared identifier va_alist.
"results.c", line 345.7: 1506-045 (S) Undeclared identifier va_alist.
make: 1254-004 The error code from the last command is 1.


Stop.
make: 1254-004 The error code from the last command is 1.


Stop.
make: 1254-004 The error code from the last command is 2.



maybe someone has an idea what this means ?!


Daniel Rohe