sdk/1.0/docs/man/integerFunctions.xml

<?xml version="1.0" encoding="UTF-8"?>
<!DOCTYPE book PUBLIC "-//OASIS//DTD DocBook MathML Module V1.1b1//EN"
              "http://www.oasis-open.org/docbook/xml/mathml/1.1CR1/dbmathml.dtd">
<refentry>
	<refentryinfo>
		<keywordset>
			<keyword>Integer Built-In Functions</keyword>
		</keywordset>
	</refentryinfo>
    <refmeta>
        <refentrytitle>Integer Built-In Functions</refentrytitle>
		<refmiscinfo>
            <copyright>
                <year>2007-2009</year>
                <holder>The Khronos Group Inc. 
 Permission is hereby granted, free of charge, to any person obtaining a
copy of this software and/or associated documentation files (the
"Materials"), to deal in the Materials without restriction, including
without limitation the rights to use, copy, modify, merge, publish,
distribute, sublicense, and/or sell copies of the Materials, and to
permit persons to whom the Materials are furnished to do so, subject to
the condition that this copyright notice and permission notice shall be included
in all copies or substantial portions of the Materials.</holder>
            </copyright>
        </refmiscinfo>
        <manvolnum>2</manvolnum>
    </refmeta>
  	<refnamediv>
		<refname>Integer Built-In Functions</refname> 
		<refpurpose>Click an item in the table below for details about that function.</refpurpose> 
	</refnamediv>
	 
<refsect1 id="springboard"><title></title>
	<informaltable frame="all"><tgroup cols="2" align="left" colsep="1" rowsep="1">
	<colspec colnum="1" colwidth="2*" colname="col1" />
	<colspec colnum="2" colwidth="8*" colname="col2" />
	<tbody>
		<row>
			<entry><citerefentry>
				<refentrytitle>abs</refentrytitle>
			</citerefentry></entry>
			<entry>|<varname>x</varname>|</entry>
		</row>
		<row>
			<entry><citerefentry href="abs">
				<refentrytitle>abs_diff</refentrytitle>
			</citerefentry></entry>
			<entry>|<varname>x</varname>-<varname>y</varname>| without modulo overflow</entry>
		</row>
		<row>			
			<entry><citerefentry>
				<refentrytitle>add_sat</refentrytitle>
			</citerefentry></entry>
			<entry><varname>x</varname>+<varname>y</varname> and saturate result</entry>
		</row>
		<row>				
			<entry><citerefentry>
				<refentrytitle>hadd</refentrytitle>
			</citerefentry></entry>
			<entry>(<varname>x</varname>+<varname>y</varname>) &gt;&gt; 1 without modulo overflow </entry>
		</row>
		<row>
			<entry><citerefentry href="hadd">
				<refentrytitle>rhadd</refentrytitle>
			</citerefentry></entry>
			<entry>(<varname>x</varname>+<varname>y</varname>+1) &gt;&gt; 1. The intermediate sum
does not modulo overflow. </entry>
		</row>
		<row>				
			<entry><citerefentry>
				<refentrytitle>clz</refentrytitle>
			</citerefentry></entry>
			<entry>Number of leading 0-bits in <varname>x</varname></entry>
		</row>
		<row>			
			<entry><citerefentry>
				<refentrytitle>mad_hi</refentrytitle>
			</citerefentry></entry>
			<entry><function>mul_hi</function>(<varname>a</varname>,<varname>b</varname>)+<varname>c</varname></entry>
		</row>
		<row>			
			<entry><citerefentry>
				<refentrytitle>mad24</refentrytitle>
			</citerefentry></entry>
			<entry>(Fast integer function.) Multiply 24-bit integer then add the 32-bit result to 32-bit integer</entry>
		</row>		
		<row>			
			<entry><citerefentry>
				<refentrytitle>mad_sat</refentrytitle>
			</citerefentry></entry>
			<entry><varname>a</varname>*<varname>b</varname>+<varname>c</varname> and saturate the result</entry>
		</row>
		<row>			
			<entry><citerefentry href="integerMax">
				<refentrytitle>max</refentrytitle>
			</citerefentry></entry>
			<entry>The greater of <varname>x</varname> or <varname>y</varname></entry>
		</row>
		<row>			
			<entry><citerefentry href="integerMax">
				<refentrytitle>min</refentrytitle>
			</citerefentry></entry>
			<entry>The lessor of <varname>x</varname> or <varname>y</varname></entry>
		</row>
		<row>			
			<entry><citerefentry>
				<refentrytitle>mul_hi</refentrytitle>
			</citerefentry></entry>
			<entry>High half of the product of <varname>x</varname> and <varname>y</varname></entry>
		</row>
		<row>			
			<entry><citerefentry>
				<refentrytitle>mul24</refentrytitle>
			</citerefentry></entry>
			<entry>(Fast integer function.) Multiply 24-bit integer values <varname>a</varname> and <varname>b</varname></entry>
		</row>		
		<row>			
			<entry><citerefentry>
				<refentrytitle>rotate</refentrytitle>
			</citerefentry></entry>
			<entry>result[<varname>indx</varname>]=<varname>v</varname>[<varname>indx</varname>]&lt;&lt;<varname>i</varname>[<varname>indx</varname>]</entry>
		</row>
		<row>			
			<entry><citerefentry>
				<refentrytitle>sub_sat</refentrytitle>
			</citerefentry></entry>
			<entry><varname>x</varname> - <varname>y</varname> and saturate the result</entry>
		</row>
		<row>			
			<entry><citerefentry>
				<refentrytitle>upsample</refentrytitle>
			</citerefentry></entry>
			<entry><varname>result</varname>[i] = ((gentype)<varname>hi</varname>[i] &lt;&lt; 8|16|32) | <varname>lo</varname>[i]</entry>
		</row>
	</tbody>	
		
	</tgroup></informaltable>	
    </refsect1>
	
    <refsect1 id="description"><title>Description</title>
        <para>
			Built-in integer functions can be used to take scalar or vector arguments. We use
the generic type name <type>gentype</type> to indicate that the function can take <type>char</type>, <type>char{2|4|8|16}</type>, <type>uchar</type>, <type>uchar{2|4|8|16}</type>, <type>short</type>, <type>short{2|4|8|16}</type>, <type>ushort</type>, <type>ushort{2|4|8|16}</type>, <type>int</type>, <type>int{2|4|8|16}</type>, <type>uint</type>,
<type>uint{2|4|8|16}</type>, <type>long</type>, <type>long{2|4|8|16}</type>, <type>ulong</type>, or <type>ulong{2|4|8|16}</type> as
the type for the arguments. We use the generic type name <type>ugentype</type> to refer to unsigned
versions of <type>gentype</type>. For example, if <type>gentype</type> is <type>char4</type>, <type>ugentype</type> is <constant>uchar4</constant>.</para>
 <para>For any specific use of a function, the actual type has to be the same for all arguments and the return type unless otherwise specified.
        </para>
    </refsect1>
	
    <refsect1 id="specification"><title>Specification</title>
        <para>
            <imageobject>
                <imagedata fileref="pdficon_small1.gif" format="gif" />
            </imageobject>
                                                 
            <olink uri="integerFunctions">OpenCL Specification</olink>
        </para>
    </refsect1> 	
<!-- ================================ ALSO SEE  
    <refsect1 id="seealso"><title>Also see</title>
        <para>
            <citerefentry><refentrytitle>clEnqueueNDRangeKernel</refentrytitle></citerefentry>,
            <citerefentry><refentrytitle>clEnqueueTask</refentrytitle></citerefentry>
        </para>
    </refsect1>-->	

<!-- ============================== COPYRIGHT -->
<!-- Content included from copyright.inc.xsl -->

    <refsect3 id="Copyright"><title></title>
		<imageobject>
                <imagedata fileref="KhronosLogo.jpg" format="jpg" />
        </imageobject>
		<para />
    </refsect3>

</refentry>