floatformat.c (floatformat_to_double): Add a comment about a potential source of...

	* floatformat.c (floatformat_to_double): Add a comment about a
	potential source of warnings when compiling this file.

From-SVN: r101914

diff --git a/libiberty/ChangeLog b/libiberty/ChangeLog
index bb11111..e2cbeb4 100644 (file)
--- a/libiberty/ChangeLog
+++ b/libiberty/ChangeLog
@@ -1,5 +1,10 @@
 2005-07-12  Ben Elliston  <bje@au.ibm.com>
 
+       * floatformat.c (floatformat_to_double): Add a comment about a
+       potential source of warnings when compiling this file.
+
+2005-07-12  Ben Elliston  <bje@au.ibm.com>
+
        * pexecute.c (pexecute): Cast string litrals to char *.
        * regex.c (re_comp): Cast a call to gettext() to char *.
 

diff --git a/libiberty/floatformat.c b/libiberty/floatformat.c
index 41000f4..e466f53 100644 (file)
--- a/libiberty/floatformat.c
+++ b/libiberty/floatformat.c
@@ -306,6 +306,13 @@ floatformat_to_double (const struct floatformat *fmt,
          mant_bits_left -= mant_bits;
        }
 
+      /* On certain systems (such as GNU/Linux), the use of the
+        INFINITY macro below may generate a warning that can not be
+        silenced due to a bug in GCC (PR preprocessor/11931).  The
+        preprocessor fails to recognise the __extension__ keyword in
+        conjunction with the GNU/C99 extension for hexadecimal
+        floating point constants and will issue a warning when
+        compiling with -pedantic.  */
       if (nan)
        dto = NAN;
       else