diff --git a/Documentation/NuttShell.html b/Documentation/NuttShell.html
index ad204f5dc9..78a5651074 100644
--- a/Documentation/NuttShell.html
+++ b/Documentation/NuttShell.html
@@ -8,7 +8,7 @@
   <tr align="center" bgcolor="#e4e4e4">
     <td>
       <h1><big><font color="#3c34ec"><i>NuttShell (NSH)</i></font></big></h1>
-      <p>Last Updated: August 28, 2012</p>
+      <p>Last Updated: October 20, 2012</p>
     </td>
   </tr>
 </table>
diff --git a/Documentation/NuttxPortingGuide.html b/Documentation/NuttxPortingGuide.html
index a16032db57..e43ca8a2fa 100644
--- a/Documentation/NuttxPortingGuide.html
+++ b/Documentation/NuttxPortingGuide.html
@@ -12,7 +12,7 @@
       <h1><big><font color="#3c34ec">
         <i>NuttX RTOS Porting Guide</i>
       </font></big></h1>
-      <p>Last Updated: August 28, 2012</p>
+      <p>Last Updated: October 20, 2012</p>
     </td>
   </tr>
 </table>
@@ -4449,12 +4449,12 @@ build
   If <code>CONFIG_ARCH_MEMCPY</code> is <b>not</b> selected, then you make also select Daniel
   Vik's optimized implementation of <code>memcpy()</code>:
 </p>
-  <ul><li>
-    <code>CONFIG_MEMCPY_VIK</code>:
-    Select this option to use the optimized <code>memcpy()</code> function by Daniel Vik.
-    See licensing information in the top-level <code>COPYING</code> file.
-    Default: <code>n</code>.
-  </li></ul>
+<ul><li>
+  <code>CONFIG_MEMCPY_VIK</code>:
+  Select this option to use the optimized <code>memcpy()</code> function by Daniel Vik.
+  See licensing information in the top-level <code>COPYING</code> file.
+  Default: <code>n</code>.
+</li></ul>
 
 <p>
   And if <code>CONFIG_MEMCPY_VIK</code>, the following tuning options are available:
@@ -4474,6 +4474,15 @@ build
     Compiles memcpy for 64 bit architectures
 </li></ul>
 
+<p><li>
+  If <code>CONFIG_ARCH_MEMSET</code> is <b>not</b> selected, then the following option is also available:
+</p>
+<ul><li>
+  <code>CONFIG_MEMSET_OPTSPEED</code>:
+  Select this option to use a version of <code>memset()</code> optimized for speed.
+  Default: <code>memset()</code> is optimized for size.
+</li></ul>
+
 <li>
   <p>
     The architecture may provide custom versions of certain standard header files: