diff --git a/ChangeLog b/ChangeLog
index 3b7f98869d..9983dca092 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,3 +1,9 @@
+2012-06-18  H.J. Lu  <hongjiu.lu@intel.com>
+
+	[BZ #14117]
+	* sysdeps/x86_64/fpu_control.h (_FPU_GETCW): Add __volatile__.
+	(_FPU_SETCW): Likewise.
+
 2012-06-15  H.J. Lu  <hongjiu.lu@intel.com>
 
 	[BZ #14117]
diff --git a/sysdeps/x86_64/fpu_control.h b/sysdeps/x86_64/fpu_control.h
index 61bc30e973..605302c09b 100644
--- a/sysdeps/x86_64/fpu_control.h
+++ b/sysdeps/x86_64/fpu_control.h
@@ -1,5 +1,5 @@
-/* FPU control word bits.  x86-64 version.
-   Copyright (C) 1993,1995,1996,1997,1998,2000,2001,2002 Free Software Foundation, Inc.
+/* FPU control word bits.  x86 version.
+   Copyright (C) 1993-2012 Free Software Foundation, Inc.
    This file is part of the GNU C Library.
    Contributed by Olaf Flebbe.
 
@@ -90,9 +90,18 @@
 /* Type of the control word.  */
 typedef unsigned int fpu_control_t __attribute__ ((__mode__ (__HI__)));
 
-/* Macros for accessing the hardware control word.  */
-#define _FPU_GETCW(cw) __asm__ ("fnstcw %0" : "=m" (*&cw))
-#define _FPU_SETCW(cw) __asm__ ("fldcw %0" : : "m" (*&cw))
+/* Macros for accessing the hardware control word.  "*&" is used to
+   work around a bug in older versions of GCC.  __volatile__ is used
+   to support combination of writing the control register and reading
+   it back.  Without __volatile__, the old value may be used for reading
+   back under compiler optimization.
+
+   Note that the use of these macros is not sufficient anymore with
+   recent hardware nor on x86-64.  Some floating point operations are
+   executed in the SSE/SSE2 engines which have their own control and
+   status register.  */
+#define _FPU_GETCW(cw) __asm__ __volatile__ ("fnstcw %0" : "=m" (*&cw))
+#define _FPU_SETCW(cw) __asm__ __volatile__ ("fldcw %0" : : "m" (*&cw))
 
 /* Default control word set at startup.  */
 extern fpu_control_t __fpu_control;