--- /dev/null
+<refentry id="vidioc-g-enc-index">
+ <refmeta>
+ <refentrytitle>ioctl VIDIOC_G_ENC_INDEX</refentrytitle>
+ &manvol;
+ </refmeta>
+
+ <refnamediv>
+ <refname>VIDIOC_G_ENC_INDEX</refname>
+ <refpurpose>Get meta data about a compressed video stream</refpurpose>
+ </refnamediv>
+
+ <refsynopsisdiv>
+ <funcsynopsis>
+ <funcprototype>
+ <funcdef>int <function>ioctl</function></funcdef>
+ <paramdef>int <parameter>fd</parameter></paramdef>
+ <paramdef>int <parameter>request</parameter></paramdef>
+ <paramdef>struct v4l2_enc_idx *<parameter>argp</parameter></paramdef>
+ </funcprototype>
+ </funcsynopsis>
+ </refsynopsisdiv>
+
+ <refsect1>
+ <title>Arguments</title>
+
+ <variablelist>
+ <varlistentry>
+ <term><parameter>fd</parameter></term>
+ <listitem>
+ <para>&fd;</para>
+ </listitem>
+ </varlistentry>
+ <varlistentry>
+ <term><parameter>request</parameter></term>
+ <listitem>
+ <para>VIDIOC_G_ENC_INDEX</para>
+ </listitem>
+ </varlistentry>
+ <varlistentry>
+ <term><parameter>argp</parameter></term>
+ <listitem>
+ <para></para>
+ </listitem>
+ </varlistentry>
+ </variablelist>
+ </refsect1>
+
+ <refsect1>
+ <title>Description</title>
+
+ <para>The <constant>VIDIOC_G_ENC_INDEX</constant> ioctl provides
+meta data about a compressed video stream the same or another
+application currently reads from the driver, which is useful for
+random access into the stream without decoding it.</para>
+
+ <para>To read the data applications must call
+<constant>VIDIOC_G_ENC_INDEX</constant> with a pointer to a
+&v4l2-enc-idx;. On success the driver fills the
+<structfield>entry</structfield> array, stores the number of elements
+written in the <structfield>entries</structfield> field, and
+initializes the <structfield>entries_cap</structfield> field.</para>
+
+ <para>Each element of the <structfield>entry</structfield> array
+contains meta data about one picture. A
+<constant>VIDIOC_G_ENC_INDEX</constant> call reads up to
+<constant>V4L2_ENC_IDX_ENTRIES</constant> entries from a driver
+buffer, which can hold up to <structfield>entries_cap</structfield>
+entries. This number can be lower or higher than
+<constant>V4L2_ENC_IDX_ENTRIES</constant>, but not zero. When the
+application fails to read the meta data in time the oldest entries
+will be lost. When the buffer is empty or no capturing/encoding is in
+progress, <structfield>entries</structfield> will be zero.</para>
+
+ <para>Currently this ioctl is only defined for MPEG-2 program
+streams and video elementary streams.</para>
+
+ <table pgwide="1" frame="none" id="v4l2-enc-idx">
+ <title>struct <structname>v4l2_enc_idx</structname></title>
+ <tgroup cols="3">
+ &cs-str;
+ <tbody valign="top">
+ <row>
+ <entry>__u32</entry>
+ <entry><structfield>entries</structfield></entry>
+ <entry>The number of entries the driver stored in the
+<structfield>entry</structfield> array.</entry>
+ </row>
+ <row>
+ <entry>__u32</entry>
+ <entry><structfield>entries_cap</structfield></entry>
+ <entry>The number of entries the driver can
+buffer. Must be greater than zero.</entry>
+ </row>
+ <row>
+ <entry>__u32</entry>
+ <entry><structfield>reserved</structfield>[4]</entry>
+ <entry spanname="hspan">Reserved for future extensions.
+Drivers must set the array to zero.</entry>
+ </row>
+ <row>
+ <entry>&v4l2-enc-idx-entry;</entry>
+ <entry><structfield>entry</structfield>[<constant>V4L2_ENC_IDX_ENTRIES</constant>]</entry>
+ <entry>Meta data about a compressed video stream. Each
+element of the array corresponds to one picture, sorted in ascending
+order by their <structfield>offset</structfield>.</entry>
+ </row>
+ </tbody>
+ </tgroup>
+ </table>
+
+ <table pgwide="1" frame="none" id="v4l2-enc-idx-entry">
+ <title>struct <structname>v4l2_enc_idx_entry</structname></title>
+ <tgroup cols="3">
+ &cs-str;
+ <tbody valign="top">
+ <row>
+ <entry>__u64</entry>
+ <entry><structfield>offset</structfield></entry>
+ <entry>The offset in bytes from the beginning of the
+compressed video stream to the beginning of this picture, that is a
+<wordasword>PES packet header</wordasword> as defined in <xref
+ linkend="mpeg2part1" /> or a <wordasword>picture
+header</wordasword> as defined in <xref linkend="mpeg2part2" />. When
+the encoder is stopped, the driver resets the offset to zero.</entry>
+ </row>
+ <row>
+ <entry>__u64</entry>
+ <entry><structfield>pts</structfield></entry>
+ <entry>The 33 bit <wordasword>Presentation Time
+Stamp</wordasword> of this picture as defined in <xref
+ linkend="mpeg2part1" />.</entry>
+ </row>
+ <row>
+ <entry>__u32</entry>
+ <entry><structfield>length</structfield></entry>
+ <entry>The length of this picture in bytes.</entry>
+ </row>
+ <row>
+ <entry>__u32</entry>
+ <entry><structfield>flags</structfield></entry>
+ <entry>Flags containing the coding type of this picture, see <xref
+ linkend="enc-idx-flags" />.</entry>
+ </row>
+ <row>
+ <entry>__u32</entry>
+ <entry><structfield>reserved</structfield>[2]</entry>
+ <entry>Reserved for future extensions.
+Drivers must set the array to zero.</entry>
+ </row>
+ </tbody>
+ </tgroup>
+ </table>
+
+ <table pgwide="1" frame="none" id="enc-idx-flags">
+ <title>Index Entry Flags</title>
+ <tgroup cols="3">
+ &cs-def;
+ <tbody valign="top">
+ <row>
+ <entry><constant>V4L2_ENC_IDX_FRAME_I</constant></entry>
+ <entry>0x00</entry>
+ <entry>This is an Intra-coded picture.</entry>
+ </row>
+ <row>
+ <entry><constant>V4L2_ENC_IDX_FRAME_P</constant></entry>
+ <entry>0x01</entry>
+ <entry>This is a Predictive-coded picture.</entry>
+ </row>
+ <row>
+ <entry><constant>V4L2_ENC_IDX_FRAME_B</constant></entry>
+ <entry>0x02</entry>
+ <entry>This is a Bidirectionally predictive-coded
+picture.</entry>
+ </row>
+ <row>
+ <entry><constant>V4L2_ENC_IDX_FRAME_MASK</constant></entry>
+ <entry>0x0F</entry>
+ <entry><wordasword>AND</wordasword> the flags field with
+this mask to obtain the picture coding type.</entry>
+ </row>
+ </tbody>
+ </tgroup>
+ </table>
+ </refsect1>
+
+ <refsect1>
+ &return-value;
+ </refsect1>
+</refentry>
Code Review / kvmfornfv.git / blobdiff