<?xml version="1.0" encoding="UTF-8"?>
<!-- Generated on Tue, 24 Nov 2009 21:48:17 -0800 -->
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
  <channel>
    <atom:link href="http://software.intel.com/en-us/articles/fast-floating-point-to-integer-conversions/feed/" rel="self" type="application/rss+xml" />
    <title>Intel Software Network Comments feed</title>
    <link>http://software.intel.com/en-us/articles/fast-floating-point-to-integer-conversions/feed/</link>
    <description></description>
    <language>en-us</language>
    <item>
      <title>By George Geczy</title>
      <description><![CDATA[ 
I believe that the reference to the _controlfp command in this article has the parameters shown reversed, they should be:
_controlfp( _RC_CHOP, _MCW_RC )

Also, /Qifist is 'deprecated' in the latest version of Visual Studio, but /fp:fast does not provide the same results and /Qifist is still required to achieve maximum performance.
 ]]></description>
      <link>http://software.intel.com/en-us/articles/fast-floating-point-to-integer-conversions/#comment-250</link>
      <pubDate>Fri, 12 Sep 2008 12:39:15 -0700</pubDate>
      <guid isPermaLink="true">http://software.intel.com/en-us/articles/fast-floating-point-to-integer-conversions/#comment-250</guid>
    </item>
    <item>
      <title>By Mike Stoner</title>
      <description><![CDATA[ 
George, thanks for the comment.  You are right, the operands are reversed but we didn't detect the error since the mask and the new value map to the same number(0x300).  Note that this paper was written in 2002 when MSVC 6.0 was still widely used.  Now that most developers are compiling with some level of SSE support we don't see very many codes that convert from the x87 stack.
 ]]></description>
      <link>http://software.intel.com/en-us/articles/fast-floating-point-to-integer-conversions/#comment-252</link>
      <pubDate>Mon, 15 Sep 2008 18:15:14 -0700</pubDate>
      <guid isPermaLink="true">http://software.intel.com/en-us/articles/fast-floating-point-to-integer-conversions/#comment-252</guid>
    </item>
    <item>
      <title>By MJW</title>
      <description><![CDATA[ Perhaps this matter should be a hint to chip designers to find out what programmers need before designing processors.  Were the Pentium designers unaware that the most commonly used programming languages convert floats to ints by truncation?  And if they were aware, why for heaven&#39;s sake didn&#39;t they include an instruction to do this regardless of the current rounding mode?  I remember all the headaches this issue caused me. ]]></description>
      <link>http://software.intel.com/en-us/articles/fast-floating-point-to-integer-conversions/#comment-9431</link>
      <pubDate>Thu, 04 Dec 2008 01:14:00 -0800</pubDate>
      <guid isPermaLink="true">http://software.intel.com/en-us/articles/fast-floating-point-to-integer-conversions/#comment-9431</guid>
    </item>
  </channel></rss>