Name

eglChooseConfig — return a list of EGL frame buffer configurations that match specified attributes

C Specification

EGLBoolean eglChooseConfig( EGLDisplay display,
  EGLint const * attrib_list,
  EGLConfig * configs,
  EGLint config_size,
  EGLint * num_config);
 

Parameters

display

Specifies the EGL display connection.

attrib_list

Specifies attributes required to match by configs.

configs

Returns an array of frame buffer configurations.

config_size

Specifies the size of the array of frame buffer configurations.

num_config

Returns the number of frame buffer configurations returned.

Description

eglChooseConfig returns in configs a list of all EGL frame buffer configurations that match the attributes specified in attrib_list. The returned EGLConfigs can be used in any EGL function that requires an EGL frame buffer configuration.

If configs is not NULL, up to config_size configs will be returned in the array pointed to by configs. The number of configs actually returned will be returned in *num_config.

If configs is NULL, no configs will be returned in configs. Instead, the total number of configs matching attrib_list will be returned in *num_config. In this case config_size is ignored. This form of eglChooseConfig is used to determine the number of matching frame buffer configurations, followed by allocating an array of EGLConfig to pass into another call to eglChooseConfig with all other parameters unchanged.

All attributes in attrib_list, including boolean attributes, are immediately followed by the corresponding desired value. The list is terminated with EGL_NONE. If an attribute is not specified in attrib_list then the default value (see below) is used (and the attribute is said to be specified implicitly). For example, if EGL_DEPTH_SIZE is not specified then it is assumed to be zero. For some attributes, the default is EGL_DONT_CARE meaning that any value is OK for this attribute, so the attribute will not be checked.

Attributes are matched in an attribute-specific manner. Some of the attributes, such as EGL_LEVEL, must match the specified value exactly. Others, such as, EGL_RED_SIZE must meet or exceed the specified minimum values. If more than one EGL frame buffer configuration matching all attributes is found, then a list of configurations, sorted according to the ``best'' match criteria, is returned. The match criteria for each attribute and the exact sorting order is defined below.

For the bitmask attributes EGL_CONFORMANT, EGL_RENDERABLE_TYPE, and EGL_SURFACE_TYPE, only the nonzero bits of the mask are considered when matching. Any bits that are zero in the specified bitmask attribute value may be either zero or one in the resulting config's attribute value.

Attributes which may appear in attrib_list, and their descriptions and allowed values, are:

EGL_ALPHA_MASK_SIZE

Must be followed by a nonnegative integer that indicates the desired alpha mask buffer size, in bits. The smallest alpha mask buffers of at least the specified size are preferred. The default value is zero.

The alpha mask buffer is used only by OpenGL and OpenGL ES client APIs.

EGL_ALPHA_SIZE

Must be followed by a nonnegative integer that indicates the desired size of the alpha component of the color buffer, in bits. If this value is zero, color buffers with the smallest alpha component size are preferred. Otherwise, color buffers with the largest alpha component of at least the specified size are preferred. The default value is zero.

EGL_BIND_TO_TEXTURE_RGB

Must be followed by EGL_DONT_CARE, EGL_TRUE, or EGL_FALSE. If EGL_TRUE is specified, then only frame buffer configurations that support binding of color buffers to an OpenGL ES RGB texture will be considered. Currently only frame buffer configurations that support pbuffers allow this. The default value is EGL_DONT_CARE.

EGL_BIND_TO_TEXTURE_RGBA

Must be followed by one of EGL_DONT_CARE, EGL_TRUE, or EGL_FALSE. If EGL_TRUE is specified, then only frame buffer configurations that support binding of color buffers to an OpenGL ES RGBA texture will be considered. Currently only frame buffer configurations that support pbuffers allow this. The default value is EGL_DONT_CARE.

EGL_BLUE_SIZE

Must be followed by a nonnegative integer that indicates the desired size of the blue component of the color buffer, in bits. If this value is zero, color buffers with the smallest blue component size are preferred. Otherwise, color buffers with the largest blue component of at least the specified size are preferred. The default value is zero.

EGL_BUFFER_SIZE

Must be followed by a nonnegative integer that indicates the desired color buffer size, in bits. The smallest color buffers of at least the specified size are preferred. The default value is zero.

The color buffer size is the sum of EGL_RED_SIZE, EGL_GREEN_SIZE, EGL_BLUE_SIZE, and EGL_ALPHA_SIZE, and does not include any padding bits which may be present in the pixel format. It is usually preferable to specify desired sizes for these color components individually.

EGL_COLOR_BUFFER_TYPE

Must be followed by one of EGL_RGB_BUFFER or EGL_LUMINANCE_BUFFER.

EGL_RGB_BUFFER indicates an RGB color buffer; in this case, attributes EGL_RED_SIZE, EGL_GREEN_SIZE and EGL_BLUE_SIZE must be non-zero, and EGL_LUMINANCE_SIZE must be zero.

EGL_LUMINANCE_BUFFER indicates a luminance color buffer. In this case EGL_RED_SIZE, EGL_GREEN_SIZE, EGL_BLUE_SIZE must be zero, and EGL_LUMINANCE_SIZE must be non-zero.

For both RGB and luminance color buffers, EGL_ALPHA_SIZE may be zero or non-zero.

EGL_CONFIG_CAVEAT

Must be followed by EGL_DONT_CARE, EGL_NONE, EGL_SLOW_CONFIG, or EGL_NON_CONFORMANT_CONFIG.

If EGL_DONT_CARE is specified, then configs are not matched for this attribute. The default value is EGL_DONT_CARE.

If EGL_NONE is specified, then configs are matched for this attribute, but only configs with no caveats (neither EGL_SLOW_CONFIG or EGL_NON_CONFORMANT_CONFIG) will be considered.

If EGL_SLOW_CONFIG is specified, then only slow configs configurations will be considered. The meaning of``slow'' is implementation-dependent, but typically indicates a non-hardware-accelerated (software) implementation.

If EGL_NON_CONFORMANT_CONFIG is specified, then only configs supporting non-conformant OpenGL ES contexts will be considered.

If the EGL version is 1.3 or later, caveat EGL_NON_CONFORMANT_CONFIG is obsolete, since the same information can be specified via the EGL_CONFORMANT attribute on a per-client-API basis, not just for OpenGL ES.

EGL_CONFIG_ID

Must be followed by a valid integer ID that indicates the desired EGL frame buffer configuration. When a EGL_CONFIG_ID is specified, all other attributes are ignored. The default value is EGL_DONT_CARE.

The meaning of config IDs is implementation-dependent. They are used only to uniquely identify different frame buffer configurations.

EGL_CONFORMANT

Must be followed by a bitmask indicating which types of client API contexts created with respect to the frame buffer configuration config must pass the required conformance tests for that API. Mask bits include:

EGL_OPENGL_BIT

Config supports creating OpenGL contexts.

EGL_OPENGL_ES_BIT

Config supports creating OpenGL ES 1.0 and/or 1.1 contexts.

EGL_OPENGL_ES2_BIT

Config supports creating OpenGL ES 2.0 contexts.

EGL_OPENVG_BIT

Config supports creating OpenVG contexts.

For example, if the bitmask is set to EGL_OPENGL_ES_BIT, only frame buffer configurations that support creating conformant OpenGL ES contexts will match. The default value is zero.

Most EGLConfigs should be conformant for all supported client APIs, and it is rarely desirable to select a nonconformant config. Conformance requirements limit the number of non-conformant configs that an implementation can define.

EGL_DEPTH_SIZE

Must be followed by a nonnegative integer that indicates the desired depth buffer size, in bits. The smallest depth buffers of at least the specified size is preferred. If the desired size is zero, frame buffer configurations with no depth buffer are preferred. The default value is zero.

The depth buffer is used only by OpenGL and OpenGL ES client APIs.

EGL_GREEN_SIZE

Must be followed by a nonnegative integer that indicates the desired size of the green component of the color buffer, in bits. If this value is zero, color buffers with the smallest green component size are preferred. Otherwise, color buffers with the largest green component of at least the specified size are preferred. The default value is zero.

EGL_LEVEL

Must be followed by an integer buffer level specification. This specification is honored exactly. Buffer level zero corresponds to the default frame buffer of the display. Buffer level one is the first overlay frame buffer, level two the second overlay frame buffer, and so on. Negative buffer levels correspond to underlay frame buffers. The default value is zero.

Most imlementations do not support overlay or underlay planes (buffer levels other than zero).

EGL_LUMINANCE_SIZE

Must be followed by a nonnegative integer that indicates the desired size of the luminance component of the color buffer, in bits. If this value is zero, color buffers with the smallest luminance component size are preferred. Otherwise, color buffers with the largest luminance component of at least the specified size are preferred. The default value is zero.

EGL_MATCH_NATIVE_PIXMAP

Must be followed by the handle of a valid native pixmap, cast to EGLint, or EGL_NONE. If the value is not EGL_NONE, only configs which support creating pixmap surfaces with this pixmap using eglCreatePixmapSurface will match this attribute. If the value is EGL_NONE, then configs are not matched for this attribute. The default value is EGL_NONE.

EGL_MATCH_NATIVE_PIXMAP was introduced due to the difficulty of determining an EGLConfig compatibile with a native pixmap using only color component sizes.

EGL_NATIVE_RENDERABLE

Must be followed by EGL_DONT_CARE, EGL_TRUE, or EGL_FALSE. If EGL_TRUE is specified, then only frame buffer configurations that allow native rendering into the surface will be considered. The default value is EGL_DONT_CARE.

EGL_MAX_SWAP_INTERVAL

Must be followed by a integer that indicates the maximum value that can be passed to eglSwapInterval. The default value is EGL_DONT_CARE.

EGL_MIN_SWAP_INTERVAL

Must be followed by a integer that indicates the minimum value that can be passed to eglSwapInterval. The default value is EGL_DONT_CARE.

EGL_RED_SIZE

Must be followed by a nonnegative integer that indicates the desired size of the red component of the color buffer, in bits. If this value is zero, color buffers with the smallest red component size are preferred. Otherwise, color buffers with the largest red component of at least the specified size are preferred. The default value is zero.

EGL_SAMPLE_BUFFERS

Must be followed by the minimum acceptable number of multisample buffers. Configurations with the smallest number of multisample buffers that meet or exceed this minimum number are preferred. Currently operation with more than one multisample buffer is undefined, so only values of zero or one will produce a match. The default value is zero.

EGL_SAMPLES

Must be followed by the minimum number of samples required in multisample buffers. Configurations with the smallest number of samples that meet or exceed the specified minimum number are preferred. Note that it is possible for color samples in the multisample buffer to have fewer bits than colors in the main color buffers. However, multisampled colors maintain at least as much color resolution in aggregate as the main color buffers.

EGL_STENCIL_SIZE

Must be followed by a nonnegative integer that indicates the desired stencil buffer size, in bits. The smallest stencil buffers of at least the specified size are preferred. If the desired size is zero, frame buffer configurations with no stencil buffer are preferred. The default value is zero.

The stencil buffer is used only by OpenGL and OpenGL ES client APIs.

EGL_RENDERABLE_TYPE

Must be followed by a bitmask indicating which types of client API contexts the frame buffer configuration must support creating with eglCreateContext). Mask bits are the same as for attribute EGL_CONFORMANT. The default value is EGL_OPENGL_ES_BIT.

EGL_SURFACE_TYPE

Must be followed by a bitmask indicating which EGL surface types and capabilities the frame buffer configuration must support. Mask bits include:

EGL_MULTISAMPLE_RESOLVE_BOX_BIT

Config allows specifying box filtered multisample resolve behavior with eglSurfaceAttrib.

EGL_PBUFFER_BIT

Config supports creating pixel buffer surfaces.

EGL_PIXMAP_BIT

Config supports creating pixmap surfaces.

EGL_SWAP_BEHAVIOR_PRESERVED_BIT

Config allows setting swap behavior for color buffers with eglSurfaceAttrib.

EGL_VG_ALPHA_FORMAT_PRE_BIT

Config allows specifying OpenVG rendering with premultiplied alpha values at surface creation time (see eglCreatePbufferSurface, eglCreatePixmapSurface, and eglCreateWindowSurface).

EGL_VG_COLORSPACE_LINEAR_BIT

Config allows specifying OpenVG rendering in a linear colorspace at surface creation time (see eglCreatePbufferSurface, eglCreatePixmapSurface, and eglCreateWindowSurface).

EGL_WINDOW_BIT

Config supports creating window surfaces.

For example, if the bitmask is set to EGL_WINDOW_BIT | EGL_PIXMAP_BIT, only frame buffer configurations that support both windows and pixmaps will be considered. The default value is EGL_WINDOW_BIT.

EGL_TRANSPARENT_TYPE

Must be followed by one of EGL_NONE or EGL_TRANSPARENT_RGB. If EGL_NONE is specified, then only opaque frame buffer configurations will be considered. If EGL_TRANSPARENT_RGB is specified, then only transparent frame buffer configurations will be considered. The default value is EGL_NONE.

Most implementations support only opaque frame buffer configurations.

EGL_TRANSPARENT_RED_VALUE

Must be followed by an integer value indicating the transparent red value. The value must be between zero and the maximum color buffer value for red. Only frame buffer configurations that use the specified transparent red value will be considered. The default value is EGL_DONT_CARE.

This attribute is ignored unless EGL_TRANSPARENT_TYPE is included in attrib_list and specified as EGL_TRANSPARENT_RGB.

EGL_TRANSPARENT_GREEN_VALUE

Must be followed by an integer value indicating the transparent green value. The value must be between zero and the maximum color buffer value for green. Only frame buffer configurations that use the specified transparent green value will be considered. The default value is EGL_DONT_CARE.

This attribute is ignored unless EGL_TRANSPARENT_TYPE is included in attrib_list and specified as EGL_TRANSPARENT_RGB.

EGL_TRANSPARENT_BLUE_VALUE

Must be followed by an integer value indicating the transparent blue value. The value must be between zero and the maximum color buffer value for blue. Only frame buffer configurations that use the specified transparent blue value will be considered. The default value is EGL_DONT_CARE.

This attribute is ignored unless EGL_TRANSPARENT_TYPE is included in attrib_list and specified as EGL_TRANSPARENT_RGB.

When more than one EGL frame buffer configuration matches the specified attributes, a list of matching configurations is returned. The list is sorted according to the following precedence rules, which are applied in ascending order (i.e., configurations that are considered equal by a lower numbered rule are sorted by the higher numbered rule):

  1. Special: by EGL_CONFIG_CAVEAT, where the precedence is EGL_NONE, EGL_SLOW_CONFIG, and EGL_NON_CONFORMANT_CONFIG.

  2. Special: by EGL_COLOR_BUFFER_TYPE, where the precedence is EGL_RGB_BUFFER, EGL_LUMINANCE_BUFFER.

  3. Special: by larger total number of color bits (for an RGB color buffer, this is the sum of EGL_RED_SIZE, EGL_GREEN_SIZE, EGL_BLUE_SIZE, and EGL_ALPHA_SIZE; for a luminance color buffer, the sum of EGL_LUMINANCE_SIZE and EGL_ALPHA_SIZE). If the requested number of bits in attrib_list is 0 or EGL_DONT_CARE for a particular color component, then the number of bits for that component is not considered.

    This sort rule places configs with deeper color buffers before configs with shallower color buffers, which may be counterintuitive.

  4. Smaller EGL_BUFFER_SIZE.

  5. Smaller EGL_SAMPLE_BUFFERS.

  6. Smaller EGL_SAMPLES.

  7. Smaller EGL_DEPTH_SIZE.

  8. Smaller EGL_STENCIL_SIZE.

  9. Smaller EGL_ALPHA_MASK_SIZE.

  10. Special: EGL_NATIVE_VISUAL_TYPE (the actual sort order is implementation-defined, depending on the meaning of native visual types).

  11. Smaller EGL_CONFIG_ID (this is always the last sorting rule, and guarantees a unique ordering).

EGLConfigs are not sorted with respect to the attributes EGL_BIND_TO_TEXTURE_RGB, EGL_BIND_TO_TEXTURE_RGBA, EGL_CONFORMANT, EGL_LEVEL, EGL_NATIVE_RENDERABLE, EGL_MAX_SWAP_INTERVAL, EGL_MIN_SWAP_INTERVAL, EGL_RENDERABLE_TYPE, EGL_SURFACE_TYPE, EGL_TRANSPARENT_TYPE, EGL_TRANSPARENT_RED_VALUE, EGL_TRANSPARENT_GREEN_VALUE, and EGL_TRANSPARENT_BLUE_VALUE.

Examples

The following example specifies a frame buffer configuration in the normal frame buffer (not an overlay or underlay). The returned frame buffer configuration supports a color buffer with at least 4 bits each of red, green and blue, and possibly no alpha bits. The code shown in the example may or may not have a depth buffer, or a stencil buffer.

EGLint const attrib_list[] = {
        EGL_RED_SIZE, 4,
        EGL_GREEN_SIZE, 4,
        EGL_BLUE_SIZE, 4,
        EGL_NONE
};

Notes

EGL_RENDERABLE_TYPE bit EGL_OPENGL_BIT, and EGL_SURFACE_TYPE bits EGL_MULTISAMPLE_RESOLVE_BOX_BIT and EGL_SWAP_BEHAVIOR_PRESERVED_BIT are supported only if the EGL version is 1.4 or greater.

EGL_CONFORMANT, EGL_MATCH_NATIVE_PIXMAP, EGL_RENDERABLE_TYPE bit EGL_OPENGL_ES2_BIT, and EGL_SURFACE_TYPE bits EGL_VG_ALPHA_FORMAT_PRE_BIT and EGL_VG_COLORSPACE_LINEAR_BIT are supported only if the EGL version is 1.3 or greater.

EGL_ALPHA_MASK_SIZE, EGL_COLOR_BUFFER_TYPE, EGL_LUMINANCE_SIZE, EGL_RENDERABLE_TYPE, and EGL_RENDERABLE_TYPE bits EGL_OPENGL_ES_BIT and EGL_OPENVG_BIT are supported only if the EGL version is 1.2 or greater.

If OpenGL or OpenGL ES rendering is supported for a luminance color buffer, it is treated as RGB rendering with the value of GL_RED_BITS equal to EGL_LUMINANCE_SIZE and the values of GL_GREEN_BITS and GL_BLUE_BITS equal to zero. The red component of fragments is written to the luminance channel of the color buffer while the green and blue components are discarded.

eglGetConfigs and eglGetConfigAttrib can be used to implement selection algorithms other than the generic one implemented by eglChooseConfig. Call eglGetConfigs to retrieve all the frame buffer configurations, or alternatively, all the frame buffer configurations with a particular set of attributes. Next call eglGetConfigAttrib to retrieve additional attributes for the frame buffer configurations and then select between them.

EGL implementors are strongly discouraged, but not proscribed, from changing the selection algorithm used by eglChooseConfig. Therefore, selections may change from release to release of the client-side library.

Errors

EGL_FALSE is returned on failure, EGL_TRUE otherwise. configs and num_config are not modified when EGL_FALSE is returned.

EGL_BAD_DISPLAY is generated if display is not an EGL display connection.

EGL_BAD_ATTRIBUTE is generated if attribute_list contains an invalid frame buffer configuration attribute or an attribute value that is unrecognized or out of range.

EGL_NOT_INITIALIZED is generated if display has not been initialized.

EGL_BAD_PARAMETER is generated if num_config is NULL.