--- a/ChangeLog
+++ b/ChangeLog
@@ -1,5 +1,8 @@
 2007-03-21  Bruno Haible  <bruno@clisp.org>
 
+	* modules/frexp-tests: New file.
+	* tests/test-frexp.c: New file.
+
 	* modules/frexp: New file.
 	* lib/frexp.c: New file.
 	* m4/frexp.m4: New file.

new file mode 100644
--- /dev/null
+++ b/modules/frexp-tests
@@ -0,0 +1,15 @@
+Files:
+tests/test-frexp.c
+
+Depends-on:
+
+configure.ac:
+AC_SUBST([FREXP_LIBM])
+
+Makefile.am:
+TESTS += test-frexp
+check_PROGRAMS += test-frexp
+test_frexp_LDADD = @FREXP_LIBM@ $(LDADD)
+
+License:
+LGPL

new file mode 100644
--- /dev/null
+++ b/tests/test-frexp.c
@@ -0,0 +1,185 @@
+/* Test of splitting a double into fraction and mantissa.
+   Copyright (C) 2007 Free Software Foundation, Inc.
+
+   This program is free software; you can redistribute it and/or modify
+   it under the terms of the GNU General Public License as published by
+   the Free Software Foundation; either version 2, or (at your option)
+   any later version.
+
+   This program is distributed in the hope that it will be useful,
+   but WITHOUT ANY WARRANTY; without even the implied warranty of
+   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+   GNU General Public License for more details.
+
+   You should have received a copy of the GNU General Public License
+   along with this program; if not, write to the Free Software Foundation,
+   Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.  */
+
+/* Written by Bruno Haible <bruno@clisp.org>, 2007.  */
+
+#include <config.h>
+
+#include <math.h>
+
+#include <float.h>
+#include <stdlib.h>
+
+#define ASSERT(expr) if (!(expr)) abort ();
+
+static double
+my_ldexp (double x, int d)
+{
+  for (; d > 0; d--)
+    x *= 2.0;
+  for (; d < 0; d++)
+    x *= 0.5;
+  return x;
+}
+
+int
+main ()
+{
+  int i;
+  /* The use of 'volatile' guarantees that excess precision bits are dropped
+     when dealing with denormalized numbers.  It is necessary on x86 systems
+     where double-floats are not IEEE compliant by default, to avoid that the
+     results become platform and compiler option dependent.  'volatile' is a
+     portable alternative to gcc's -ffloat-store option.  */
+  volatile double x;
+
+  { /* NaN.  */
+    int exp = -9999;
+    double mantissa;
+    x = 0.0 / 0.0;
+    mantissa = frexp (x, &exp);
+    ASSERT (mantissa != mantissa);
+  }
+
+  { /* Positive infinity.  */
+    int exp = -9999;
+    double mantissa;
+    x = 1.0 / 0.0;
+    mantissa = frexp (x, &exp);
+    ASSERT (mantissa == x);
+  }
+
+  { /* Negative infinity.  */
+    int exp = -9999;
+    double mantissa;
+    x = -1.0 / 0.0;
+    mantissa = frexp (x, &exp);
+    ASSERT (mantissa == x);
+  }
+
+  { /* Positive zero.  */
+    int exp = -9999;
+    double mantissa;
+    x = 0.0;
+    mantissa = frexp (x, &exp);
+    ASSERT (exp == 0);
+    ASSERT (mantissa == x);
+  }
+
+  { /* Negative zero.  */
+    int exp = -9999;
+    double mantissa;
+    x = -0.0;
+    mantissa = frexp (x, &exp);
+    ASSERT (exp == 0);
+    ASSERT (mantissa == x);
+  }
+
+  for (i = 1, x = 1.0; i <= DBL_MAX_EXP; i++, x *= 2.0)
+    {
+      int exp = -9999;
+      double mantissa = frexp (x, &exp);
+      ASSERT (exp == i);
+      ASSERT (mantissa == 0.5);
+    }
+  for (i = 1, x = 1.0; i >= DBL_MIN_EXP; i--, x *= 0.5)
+    {
+      int exp = -9999;
+      double mantissa = frexp (x, &exp);
+      ASSERT (exp == i);
+      ASSERT (mantissa == 0.5);
+    }
+  for (; i >= DBL_MIN_EXP - 100 && x > 0.0; i--, x *= 0.5)
+    {
+      int exp = -9999;
+      double mantissa = frexp (x, &exp);
+      ASSERT (exp == i);
+      ASSERT (mantissa == 0.5);
+    }
+
+  for (i = 1, x = -1.0; i <= DBL_MAX_EXP; i++, x *= 2.0)
+    {
+      int exp = -9999;
+      double mantissa = frexp (x, &exp);
+      ASSERT (exp == i);
+      ASSERT (mantissa == -0.5);
+    }
+  for (i = 1, x = -1.0; i >= DBL_MIN_EXP; i--, x *= 0.5)
+    {
+      int exp = -9999;
+      double mantissa = frexp (x, &exp);
+      ASSERT (exp == i);
+      ASSERT (mantissa == -0.5);
+    }
+  for (; i >= DBL_MIN_EXP - 100 && x < 0.0; i--, x *= 0.5)
+    {
+      int exp = -9999;
+      double mantissa = frexp (x, &exp);
+      ASSERT (exp == i);
+      ASSERT (mantissa == -0.5);
+    }
+
+  for (i = 1, x = 1.01; i <= DBL_MAX_EXP; i++, x *= 2.0)
+    {
+      int exp = -9999;
+      double mantissa = frexp (x, &exp);
+      ASSERT (exp == i);
+      ASSERT (mantissa == 0.505);
+    }
+  for (i = 1, x = 1.01; i >= DBL_MIN_EXP; i--, x *= 0.5)
+    {
+      int exp = -9999;
+      double mantissa = frexp (x, &exp);
+      ASSERT (exp == i);
+      ASSERT (mantissa == 0.505);
+    }
+  for (; i >= DBL_MIN_EXP - 100 && x > 0.0; i--, x *= 0.5)
+    {
+      int exp = -9999;
+      double mantissa = frexp (x, &exp);
+      ASSERT (exp == i);
+      ASSERT (mantissa >= 0.5);
+      ASSERT (mantissa < 1.0);
+      ASSERT (mantissa == my_ldexp (x, - exp));
+    }
+
+  for (i = 1, x = 1.73205; i <= DBL_MAX_EXP; i++, x *= 2.0)
+    {
+      int exp = -9999;
+      double mantissa = frexp (x, &exp);
+      ASSERT (exp == i);
+      ASSERT (mantissa == 0.866025);
+    }
+  for (i = 1, x = 1.73205; i >= DBL_MIN_EXP; i--, x *= 0.5)
+    {
+      int exp = -9999;
+      double mantissa = frexp (x, &exp);
+      ASSERT (exp == i);
+      ASSERT (mantissa == 0.866025);
+    }
+  for (; i >= DBL_MIN_EXP - 100 && x > 0.0; i--, x *= 0.5)
+    {
+      int exp = -9999;
+      double mantissa = frexp (x, &exp);
+      ASSERT (exp == i || exp == i + 1);
+      ASSERT (mantissa >= 0.5);
+      ASSERT (mantissa < 1.0);
+      ASSERT (mantissa == my_ldexp (x, - exp));
+    }
+
+  return 0;
+}