blob: fa6f059fc4c9037997b0d5e38955bafbee71f3f9 [file] [log] [blame]
<?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" [
<!ENTITY clCreate_memflagsInc SYSTEM "clCreate_memflagsInc.xml">
]>
<refentry>
<refentryinfo>
<keywordset>
<keyword>clGetSupportedImageFormats</keyword>
</keywordset>
</refentryinfo>
<refmeta>
<refentrytitle>
clGetSupportedImageFormats
</refentrytitle>
<refmiscinfo>
<copyright>
<year>2007-2013</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>3</manvolnum>
</refmeta>
<!-- ================================ SYNOPSIS -->
<refnamediv id="clGetSupportedImageFormats">
<refname>
clGetSupportedImageFormats
</refname>
<refpurpose>
Get the list of image formats supported by an OpenCL implementation.
</refpurpose>
</refnamediv>
<refsynopsisdiv xmlns:xlink="http://www.w3.org/1999/xlink">
<title></title>
<funcsynopsis>
<funcprototype>
<funcdef>
<link xlink:href="scalarDataTypes.html">cl_int</link>
<function>clGetSupportedImageFormats</function>
</funcdef>
<paramdef><link xlink:href="abstractDataTypes.html">cl_context</link><parameter>context</parameter></paramdef>
<paramdef><link xlink:href="enums.html#cl_mem_flags">cl_mem_flags</link><parameter>flags</parameter></paramdef>
<paramdef><link xlink:href="enums.html#cl_mem_object_type">cl_mem_object_type</link><parameter>image_type</parameter></paramdef>
<paramdef><link xlink:href="scalarDataTypes.html">cl_uint</link><parameter>num_entries</parameter></paramdef>
<paramdef><link xlink:href="cl_image_format.html">cl_image_format</link><parameter>*image_formats</parameter></paramdef>
<paramdef><link xlink:href="scalarDataTypes.html">cl_uint</link><parameter>*num_image_formats</parameter></paramdef>
</funcprototype>
</funcsynopsis>
</refsynopsisdiv>
<!-- ================================ PARAMETERS -->
<refsect1 id="parameters">
<title>Parameters</title>
<variablelist>
<varlistentry>
<term> <varname> context </varname> </term>
<listitem>
<para>
A valid OpenCL context on which the image object(s) will be created.
</para>
</listitem>
</varlistentry>
<varlistentry>
<term> <varname> flags </varname> </term>
<listitem>
<para>
A bit-field that is used to specify allocation and usage information about
the image memory object being queried and is described in the table below.
To get a list of supported image
formats that can be read from or written to by a
kernel, <varname>flags</varname> must be set to
<constant>CL_MEM_READ_WRITE</constant>
(get a list of images that can be read from and written
to by different kernel instances when correctly ordered
by event dependencies),
<constant>CL_MEM_READ_ONLY</constant> (list of images
that can be read from by a kernel) or
<constant>CL_MEM_WRITE_ONLY</constant> (list of images
that can be written to by a kernel). To get a list of
supported image formats that can be both read from and
written to by a kernel, <varname>flags</varname> must be set
to <constant>CL_MEM_KERNEL_READ_AND_WRITE</constant>.
Please see section 5.3.2.2 for clarification.
</para>
<!-- Table 5.3 -->
&clCreate_memflagsInc;
</listitem>
</varlistentry>
<varlistentry>
<term> <varname> image_type </varname> </term>
<listitem>
<para>
Describes the image type and must be either
<constant>CL_MEM_OBJECT_IMAGE1D</constant>,
<constant>CL_MEM_OBJECT_IMAGE1D_BUFFER</constant>,
<constant>CL_MEM_OBJECT_IMAGE2D</constant>,
<constant>CL_MEM_OBJECT_IMAGE3D</constant>,
<constant>CL_MEM_OBJECT_IMAGE1D_ARRAY</constant> or
<constant>CL_MEM_OBJECT_IMAGE2D_ARRAY</constant>.
</para>
</listitem>
</varlistentry>
<varlistentry>
<term> <varname> num_entries </varname> </term>
<listitem>
<para>
Specifies the number of entries that can be returned in the memory location
given by <varname>image_formats</varname>.
</para>
</listitem>
</varlistentry>
<varlistentry>
<term> <varname> image_formats </varname> </term>
<listitem>
<para>
A pointer to a memory location where the list of
supported image formats are returned. Each entry describes a
<citerefentry><refentrytitle>cl_image_format</refentrytitle></citerefentry>
structure supported by the OpenCL implementation. If
<varname>image_formats</varname> is NULL, it is ignored.
</para>
</listitem>
</varlistentry>
<varlistentry>
<term> <varname> num_image_formats </varname> </term>
<listitem>
<para>
The actual number of supported image formats for a
specific <varname>context</varname> and values specified by
<varname>flags</varname>. If <varname>num_image_formats</varname> is NULL,
it is ignored.
</para>
</listitem>
</varlistentry>
</variablelist>
</refsect1>
<!-- ================================ NOTES -->
<refsect1 id="notes"><title>Notes</title>
<para>
<function>clGetSupportedImageFormats</function> can be used to get the list of image
formats supported by an OpenCL implementation when the following information about
an image memory object is specified:
</para>
<itemizedlist mark="disc">
<listitem>
Context
</listitem>
<listitem>
Image type - 1D, 2D, or 3D image, 1D image buffer, 1D or 2D image array.
</listitem>
<listitem>
Image object allocation information
</listitem>
</itemizedlist>
<para>
<function>clGetSupportedImageFormats</function> returns a union of image formats
supported by all devices in the context.
</para>
<para>
If <constant>CL_DEVICE_IMAGE_SUPPORT</constant> specified in table 4.3
(see <citerefentry><refentrytitle>clGetDeviceInfo</refentrytitle></citerefentry>)
is <constant>CL_TRUE</constant>, the values assigned to
<constant>CL_DEVICE_MAX_READ_IMAGE_ARGS</constant>, <constant>CL_DEVICE_MAX_WRITE_IMAGE_ARGS</constant>,
<constant>CL_DEVICE_IMAGE2D_MAX_WIDTH</constant>, <constant>CL_DEVICE_IMAGE2D_MAX_HEIGHT</constant>,
<constant>CL_DEVICE_IMAGE3D_MAX_WIDTH</constant>, <constant>CL_DEVICE_IMAGE3D_MAX_HEIGHT</constant>,
<constant>CL_DEVICE_IMAGE3D_MAX_DEPTH</constant> and <constant>CL_DEVICE_MAX_SAMPLERS</constant> by the
implementation must be greater than or equal to the minimum values specified in table 4.3
(see <citerefentry><refentrytitle>clGetDeviceInfo</refentrytitle></citerefentry>).
</para>
<bridgehead>Minimum List of Supported Image Formats</bridgehead>
<para>
For 1D, 1D image from buffer, 2D, 3D image objects, 1D and 2D image array objects, the
mandated minimum list of image formats that must be supported by all devices (that can be read
from or written to by a kernel but not both) that support images is described
in the table below (Table 5.8a):
</para>
<!-- table 5.8a -->
<informaltable frame="all">
<tgroup cols="4" align="left" colsep="1" rowsep="1">
<colspec colname="col1" colnum="1" />
<colspec colname="col2" colnum="2" />
<colspec colname="col3" colnum="3" />
<colspec colname="col4" colnum="4" />
<thead>
<row>
<entry><literallayout>num_channels</literallayout></entry>
<entry><literallayout>channel_order</literallayout></entry>
<entry><literallayout>channel_data_type</literallayout></entry>
<entry><literallayout>read/write</literallayout></entry>
</row>
</thead>
<tbody>
<row>
<entry><literallayout>1</literallayout></entry>
<entry><literallayout>CL_R</literallayout></entry>
<entry><literallayout>CL_UNORM_INT8
CL_UNORM_INT16
CL_SNORM_INT8
CL_SNORM_INT16
CL_SIGNED_INT8
CL_SIGNED_INT16
CL_SIGNED_INT32
CL_UNSIGNED_INT8
CL_UNSIGNED_INT16
CL_UNSIGNED_INT32
CL_HALF_FLOAT
CL_FLOAT
</literallayout>
</entry>
<entry><literallayout>read+write</literallayout></entry>
</row>
<row>
<entry><literallayout>1</literallayout></entry>
<entry><literallayout>CL_DEPTH</literallayout></entry>
<entry><literallayout>CL_UNORM_INT16
CL_FLOAT
</literallayout>
</entry>
<entry><literallayout>read+write</literallayout>
<constant>CL_DEPTH</constant> channel order is supported only for 2D image and 2D image array objects.
</entry>
</row>
<row>
<entry><literallayout>2</literallayout></entry>
<entry><literallayout>CL_RG</literallayout></entry>
<entry><literallayout>CL_UNORM_INT8
CL_UNORM_INT16
CL_SNORM_INT8
CL_SNORM_INT16
CL_SIGNED_INT8
CL_SIGNED_INT16
CL_SIGNED_INT32
CL_UNSIGNED_INT8
CL_UNSIGNED_INT16
CL_UNSIGNED_INT32
CL_HALF_FLOAT
CL_FLOAT
</literallayout>
</entry>
<entry><literallayout>read+write</literallayout></entry>
</row>
<row>
<entry><literallayout>4</literallayout></entry>
<entry><literallayout>CL_RGBA</literallayout></entry>
<entry><literallayout>CL_UNORM_INT8
CL_UNORM_INT16
CL_SNORM_INT8
CL_SNORM_INT16
CL_SIGNED_INT8
CL_SIGNED_INT16
CL_SIGNED_INT32
CL_UNSIGNED_INT8
CL_UNSIGNED_INT16
CL_UNSIGNED_INT32
CL_HALF_FLOAT
CL_FLOAT
</literallayout>
</entry>
<entry><literallayout>read+write</literallayout></entry>
</row>
<row>
<entry><literallayout>4</literallayout></entry>
<entry><literallayout>CL_BGRA</literallayout></entry>
<entry><literallayout>CL_UNORM_INT8</literallayout></entry>
<entry><literallayout>read+write</literallayout></entry>
</row>
<row>
<entry><literallayout>4</literallayout></entry>
<entry><literallayout>CL_sRGBA</literallayout></entry>
<entry><literallayout>CL_UNORM_INT8</literallayout></entry>
<entry>
read+write if the
<citerefentry><refentrytitle>cl_khr_srgb_image_writes</refentrytitle></citerefentry>
extension is supported, else read only.
sRGB channel order support is not required for 1D image buffers.
Writes to images with sRGB channel orders
requires device support of the
<citerefentry><refentrytitle>cl_khr_srgb_image_writes</refentrytitle></citerefentry>
extension.
</entry>
</row>
<row>
<entry><literallayout>1</literallayout></entry>
<entry><literallayout>CL_DEPTH_STENCIL</literallayout></entry>
<entry><literallayout>CL_UNORM_INT24
CL_FLOAT
</literallayout>
</entry>
<entry><literallayout>read only (applies if the
<citerefentry><refentrytitle>cl_khr_gl_depth_images</refentrytitle></citerefentry>
extension is enabled)</literallayout></entry>
</row>
</tbody>
</tgroup>
</informaltable>
<para>
For 1D, 1D image from buffer, 2D, 3D image objects, 1D and 2D image array objects, the
mandated minimum list of image formats that must be supported by all devices (that can be read
from and written to by a kernel) that support images is described
in the table below (Table 5.8b):
</para>
<!-- table 5.8b -->
<informaltable frame="all">
<tgroup cols="3" align="left" colsep="1" rowsep="1">
<colspec colname="col1" colnum="1" />
<colspec colname="col2" colnum="2" />
<colspec colname="col3" colnum="3" />
<thead>
<row>
<entry><literallayout>num_channels</literallayout></entry>
<entry><literallayout>channel_order</literallayout></entry>
<entry><literallayout>channel_data_type</literallayout></entry>
</row>
</thead>
<tbody>
<row>
<entry><literallayout>1</literallayout></entry>
<entry><literallayout>CL_R</literallayout></entry>
<entry><literallayout>CL_UNORM_INT8
CL_SIGNED_INT8
CL_SIGNED_INT16
CL_SIGNED_INT32
CL_UNSIGNED_INT8
CL_UNSIGNED_INT16
CL_UNSIGNED_INT32
CL_HALF_FLOAT
CL_FLOAT
</literallayout>
</entry>
</row>
<row>
<entry><literallayout>4</literallayout></entry>
<entry><literallayout>CL_RGBA</literallayout></entry>
<entry><literallayout>CL_UNORM_INT8
CL_UNORM_INT16
CL_SIGNED_INT8
CL_SIGNED_INT16
CL_SIGNED_INT32
CL_UNSIGNED_INT8
CL_UNSIGNED_INT16
CL_UNSIGNED_INT32
CL_HALF_FLOAT
CL_FLOAT
</literallayout>
</entry>
</row>
<row>
<entry><literallayout>1</literallayout></entry>
<entry><literallayout>CL_DEPTH_STENCIL (applies if the
<citerefentry><refentrytitle>cl_khr_gl_depth_images</refentrytitle></citerefentry>
extension is enabled)</literallayout></entry>
<entry><literallayout>CL_UNORM_INT24
CL_FLOAT
</literallayout>
</entry>
</row>
</tbody>
</tgroup>
</informaltable>
<bridgehead> Image format mapping to OpenCL C image access qualifiers</bridgehead>
<para>
Image arguments to kernels may have the <constant>read_only</constant>, <constant>write_only</constant> or <constant>read_write</constant>
qualifier. Not all image formats supported by the device and platform are valid to be passed to all
of these access qualifiers. For each access qualifier, only images whose format is in the list of
formats returned by clGetSupportedImageFormats with the given flag arguments in the table below are
permitted. It is not valid to pass an image supporting writing as both a read_only image and a
write_only image parameter, or to a read_write image parameter and any other image parameter.
</para>
<!-- table 5.9 -->
<informaltable frame="all">
<tgroup cols="2" align="left" colsep="1" rowsep="1">
<colspec colname="col1" colnum="1" />
<colspec colname="col2" colnum="2" />
<thead>
<row>
<entry><literallayout>Access Qualifier</literallayout></entry>
<entry><literallayout>cl_mem_flags</literallayout></entry>
</row>
</thead>
<tbody>
<row>
<entry><constant>read_only</constant></entry>
<entry>
<constant>CL_MEM_READ_ONLY</constant>,
<constant>CL_MEM_READ_WRITE</constant>,
<constant>CL_MEM_KERNEL_READ_AND_WRITE</constant>
</entry>
</row>
<row>
<entry><constant>write_only</constant></entry>
<entry>
<constant>CL_MEM_WRITE_ONLY</constant>,
<constant>CL_MEM_READ_WRITE</constant>,
<constant>CL_MEM_KERNEL_READ_AND_WRITE</constant>
</entry>
</row>
<row>
<entry><constant>read_write</constant></entry>
<entry>
<constant>CL_MEM_KERNEL_READ_AND_WRITE</constant>
</entry>
</row>
</tbody>
</tgroup>
</informaltable>
<bridgehead>Additional notes if the <citerefentry><refentrytitle>cl_khr_gl_depth_images</refentrytitle></citerefentry>
extension is enabled:</bridgehead>
<para>
For the image format given by channel order of
<constant>CL_DEPTH_STENCIL</constant> and channel data type of
<constant>CL_UNORM_INT24</constant>, the depth is
stored as an unsigned normalized 24-bit value.
</para>
<para>
For the image format given by channel order of
<constant>CL DEPTH_STENCIL</constant> and channel data type of
<constant>CL_FLOAT</constant>, each pixel is two 32-bit values.
The depth is stored as a single precision floating
point value followed by the stencil which is stored as a 8-bit integer value.
</para>
<para>
The stencil value cannot be read or written using the
<function>read_imagef</function> and
<function>write_imagef</function>
built-in functions in an OpenCL kernel.
</para>
<para>
Depth image objects with an image channel order =
<constant>CL_DEPTH_STENCIL</constant> cannot be used as
arguments to
<citerefentry><refentrytitle>clEnqueueReadImage</refentrytitle></citerefentry>,
<citerefentry><refentrytitle>clEnqueueWriteImage</refentrytitle></citerefentry>,
<citerefentry><refentrytitle>clEnqueueCopyImage</refentrytitle></citerefentry>,
<citerefentry><refentrytitle>clEnqueueCopyImageToBuffer</refentrytitle></citerefentry>,
<citerefentry><refentrytitle>clEnqueueCopyBufferToImage</refentrytitle></citerefentry>,
<citerefentry><refentrytitle>clEnqueueMapImage</refentrytitle></citerefentry> and
<citerefentry><refentrytitle>clEnqueueFillImage</refentrytitle></citerefentry>
and will return a <errorname>CL_INVALID_OPERATION</errorname> error.
</para>
</refsect1>
<!-- ================================ ERRORS -->
<refsect1 id="errors"><title>Errors</title>
<para>
Returns <errorname>CL_SUCCESS</errorname> if the function is executed successfully.
Otherwise, it returns one of the following errors:
</para>
<itemizedlist mark="disc">
<listitem>
<errorname>CL_INVALID_CONTEXT</errorname> if <varname>context</varname> is not a valid context.
</listitem>
<listitem>
<errorname>CL_INVALID_VALUE</errorname> if <varname>flags</varname> or
<varname>image_type</varname> are not valid, or if <varname>num_entries</varname>
is 0 and <varname>image_formats</varname> is not NULL.
</listitem>
<listitem>
<errorname>CL_OUT_OF_RESOURCES</errorname> if there is a failure to allocate
resources required by the OpenCL implementation on the device.
</listitem>
<listitem>
<errorname>CL_OUT_OF_HOST_MEMORY</errorname> if there is a failure to allocate
resources required by the OpenCL implementation on the host.
</listitem>
</itemizedlist>
</refsect1>
<!-- ================================ EXAMPLE -->
<!-- DO NOT DELETE IN CASE AN EXAMPLE IS ADDED IN THE FUTURE -->
<!--
<refsect2 id="example1">
<title>
Example
</title>
<informaltable frame="none">
<tgroup cols="1" align="left" colsep="0" rowsep="0">
<colspec colname="col1" colnum="1" />
<tbody>
<row>
<entry>
Example goes here - it will be set in "code" type with white space preserved.
</entry>
</row>
</tbody>
</tgroup>
</informaltable>
</refsect2>
-->
<!-- ================================ SPECIFICATION -->
<!-- Set the "uri" attribute in the <olink /> element to the "named destination" for the PDF page
-->
<refsect1 id="specification"><title>Specification</title>
<para>
<imageobject>
<imagedata fileref="pdficon_small1.gif" format="gif" />
</imageobject>
<olink uri="clGetSupportedImageFormats">OpenCL Specification</olink>
</para>
</refsect1>
<!-- ================================ ALSO SEE -->
<refsect1 id="seealso"><title>Also see</title>
<para>
<citerefentry><refentrytitle>cl_image_format</refentrytitle></citerefentry>,
<citerefentry><refentrytitle>clGetDeviceInfo</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>
<!-- 26-Oct-2015, API rev. 19 -->
</refentry>