isnanf syntax error

[Eric Blake]

Found this while cross-compiling:

checking where to find the exponent in a 'double'... checking whether byte
ordering is bigendian... no
word 1 bit 20
checking whether isnan(float) can be used without linking with libm... yes
checking whether isnan(float) works... guessing yes
../configure: line 14918: test: too many arguments

This patch does not address the first bug (namely, that the check for
exponent inside double and bigendian are nested, which is a no-no for
output purposes); but does address the second.

2007-05-29  Eric Blake  <address@hidden>

        * m4/isnanf.m4 (gl_FUNC_ISNANF_NO_LIBM): Avoid syntax error on
        cross compiles.

Index: m4/isnanf.m4
===================================================================
RCS file: /sources/gnulib/gnulib/m4/isnanf.m4,v
retrieving revision 1.3
diff -u -p -r1.3 isnanf.m4
--- m4/isnanf.m4        20 May 2007 00:25:31 -0000      1.3
+++ m4/isnanf.m4        29 May 2007 12:48:50 -0000
@@ -1,4 +1,4 @@
-# isnanf.m4 serial 2
+# isnanf.m4 serial 3
 dnl Copyright (C) 2007 Free Software Foundation, Inc.
 dnl This file is free software; the Free Software Foundation
 dnl gives unlimited permission to copy and/or distribute it,
@@ -13,7 +13,7 @@ AC_DEFUN([gl_FUNC_ISNANF_NO_LIBM],
     gl_ISNANF_WORKS
   fi
   if test $gl_cv_func_isnanf_no_libm = yes \
-     && test $gl_cv_func_isnanf_works = yes; then
+     && test "$gl_cv_func_isnanf_works" = yes; then
     AC_DEFINE([HAVE_ISNANF_IN_LIBC], 1,
       [Define if the isnan(float) function is available in libc.])
   else

-- 
Don't work too hard, make some time for fun as well!

Eric Blake             address@hidden