Opentk/Source/Bind/Specifications/Docs/glDispatchComputeIndirect.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 id="glDispatchComputeIndirect">
    <refentryinfo>
        <copyright>
            <year>2013</year>
            <holder>Khronos Group</holder>
        </copyright>
    </refentryinfo>
    <refmeta>
        <refentrytitle>glDispatchComputeIndirect</refentrytitle>
        <manvolnum>3G</manvolnum>
    </refmeta>
    <refnamediv>
        <refname>glDispatchComputeIndirect</refname>
        <refpurpose>launch one or more compute work groups using parameters stored in a buffer</refpurpose>
    </refnamediv>
    <refsynopsisdiv><title>C Specification</title>
        <funcsynopsis>
            <funcprototype>
                <funcdef>void <function>glDispatchComputeIndirect</function></funcdef>
                <paramdef>GLintptr <parameter>indirect</parameter></paramdef>
            </funcprototype>
        </funcsynopsis>
    </refsynopsisdiv>
    <refsect1 id="parameters"><title>Parameters</title>
        <variablelist>
        <varlistentry>
            <term><parameter>indirect</parameter></term>
            <listitem>
                <para>
                    The offset into the buffer object currently bound to the
                    <constant>GL_DISPATCH_INDIRECT_BUFFER</constant> buffer target at
                    which the dispatch parameters are stored.
                </para>
            </listitem>
        </varlistentry>
        </variablelist>
    </refsect1>
    <refsect1 id="description"><title>Description</title>
        <para>
            <function>glDispatchComputeIndirect</function> launches one or more compute
            work groups using parameters stored in the buffer object currently bound
            to the <constant>GL_DISPATCH_INDIRECT_BUFFER</constant> target.
            Each work group is processed by the active program object for the compute
            shader stage. While the individual shader invocations within a work group are
            executed as a unit, work groups are executed completely independently and in
            unspecified order. <parameter>indirect</parameter> contains the offset
            into the data store of the buffer object bound to the <constant>GL_DISPATCH_INDIRECT_BUFFER</constant>
            target at which the parameters are stored.
        </para>
        <para>
            The parameters addressed by <parameter>indirect</parameter> are packed a structure,
            which takes the form (in C):
            <programlisting><![CDATA[    typedef  struct {
        uint  num_groups_x;
        uint  num_groups_y;
        uint  num_groups_z;
    } DispatchIndirectCommand;]]></programlisting>
        </para>
        <para>
            A call to <function>glDispatchComputeIndirect</function> is equivalent, assuming no
            errors are generated, to:
            <programlisting><![CDATA[    cmd = (const DispatchIndirectCommand  *)indirect;
    glDispatchComputeIndirect(cmd->num_groups_x, cmd->num_groups_y, cmd->num_groups_z);]]></programlisting>
        </para>
        <para>
            Unlike <citerefentry><refentrytitle>glDispatchCompute</refentrytitle></citerefentry>,
            no error is generated if any of the <code>num_groups_x</code>,
            <code>num_groups_y</code> or <code>num_groups_z</code> members
            of the <code>DispatchIndirectCommand</code> is larger than the
            value of <constant>GL_MAX_COMPUTE_WORK_GROUP_COUNT</constant> for the
            corresponding dimension. In such circumstances, behavior is undefined and
            may lead to application termination.
        </para>
    </refsect1>
    <refsect1 id="errors"><title>Errors</title>
        <para>
            <constant>GL_INVALID_OPERATION</constant> is generated if there is no active program
            for the compute shader stage.
        </para>
        <para>
            <constant>GL_INVALID_VALUE</constant> is generated if <parameter>indirect</parameter> is
            less than zero or not a multiple of four.
        </para>
        <para>
            <constant>GL_INVALID_OPERATION</constant> is generated if no buffer is bound to the
            <constant>GL_DISPATCH_INDIRECT_BUFFER</constant> target or if the command would
            source data beyond the end of the buffer object's data store.
        </para>
    </refsect1>
    <refsect1 id="associatedgets"><title>Associated Gets</title>
        <para>
            <citerefentry><refentrytitle>glGet</refentrytitle></citerefentry> with argument <constant>GL_MAX_COMPUTE_WORK_GROUP_COUNT</constant>
        </para>
    </refsect1>
    <refsect1 id="seealso"><title>See Also</title>
        <para>
            <citerefentry><refentrytitle>glDispatchCompute</refentrytitle></citerefentry>.
        </para>
    </refsect1>
    <refsect1 id="Copyright"><title>Copyright</title>
        <para>
            Copyright <trademark class="copyright"></trademark> 2013 Khronos Group.
            This material may be distributed subject to the terms and conditions set forth in
            the Open Publication License, v 1.0, 8 June 1999.
            <ulink url="http://opencontent.org/openpub/">http://opencontent.org/openpub/</ulink>.
        </para>
    </refsect1>
</refentry>