Pipeline

Pipeline — Functions for creating and manipulating the GPU pipeline

Synopsis

                    CoglPipeline;
CoglPipeline *      cogl_pipeline_new                   (CoglContext *context);
CoglPipeline *      cogl_pipeline_copy                  (CoglPipeline *source);
CoglBool            cogl_is_pipeline                    (void *object);
void                cogl_pipeline_set_color             (CoglPipeline *pipeline,
                                                         const CoglColor *color);
void                cogl_pipeline_set_color4ub          (CoglPipeline *pipeline,
                                                         uint8_t red,
                                                         uint8_t green,
                                                         uint8_t blue,
                                                         uint8_t alpha);
void                cogl_pipeline_set_color4f           (CoglPipeline *pipeline,
                                                         float red,
                                                         float green,
                                                         float blue,
                                                         float alpha);
void                cogl_pipeline_get_color             (CoglPipeline *pipeline,
                                                         CoglColor *color);
enum                CoglPipelineAlphaFunc;
void                cogl_pipeline_set_alpha_test_function
                                                        (CoglPipeline *pipeline,
                                                         CoglPipelineAlphaFunc alpha_func,
                                                         float alpha_reference);
#define             COGL_BLEND_STRING_ERROR
enum                CoglBlendStringError;
CoglBool            cogl_pipeline_set_blend             (CoglPipeline *pipeline,
                                                         const char *blend_string,
                                                         CoglError **error);
void                cogl_pipeline_set_blend_constant    (CoglPipeline *pipeline,
                                                         const CoglColor *constant_color);
void                cogl_pipeline_set_point_size        (CoglPipeline *pipeline,
                                                         float point_size);
float               cogl_pipeline_get_point_size        (CoglPipeline *pipeline);
CoglBool            cogl_pipeline_set_per_vertex_point_size
                                                        (CoglPipeline *pipeline,
                                                         CoglBool enable,
                                                         CoglError **error);
CoglBool            cogl_pipeline_get_per_vertex_point_size
                                                        (CoglPipeline *pipeline);
CoglColorMask       cogl_pipeline_get_color_mask        (CoglPipeline *pipeline);
void                cogl_pipeline_set_color_mask        (CoglPipeline *pipeline,
                                                         CoglColorMask color_mask);
CoglBool            cogl_pipeline_set_depth_state       (CoglPipeline *pipeline,
                                                         const CoglDepthState *state,
                                                         CoglError **error);
void                cogl_pipeline_get_depth_state       (CoglPipeline *pipeline,
                                                         CoglDepthState *state_out);
enum                CoglPipelineCullFaceMode;
void                cogl_pipeline_set_cull_face_mode    (CoglPipeline *pipeline,
                                                         CoglPipelineCullFaceMode cull_face_mode);
enum                CoglWinding;
void                cogl_pipeline_set_front_face_winding
                                                        (CoglPipeline *pipeline,
                                                         CoglWinding front_winding);
void                cogl_pipeline_set_layer_texture     (CoglPipeline *pipeline,
                                                         int layer_index,
                                                         CoglTexture *texture);
void                cogl_pipeline_set_layer_null_texture
                                                        (CoglPipeline *pipeline,
                                                         int layer_index,
                                                         CoglTextureType texture_type);
CoglTexture *       cogl_pipeline_get_layer_texture     (CoglPipeline *pipeline,
                                                         int layer_index);
enum                CoglPipelineFilter;
void                cogl_pipeline_set_layer_filters     (CoglPipeline *pipeline,
                                                         int layer_index,
                                                         CoglPipelineFilter min_filter,
                                                         CoglPipelineFilter mag_filter);
CoglPipelineFilter  cogl_pipeline_get_layer_min_filter  (CoglPipeline *pipeline,
                                                         int layer_index);
CoglPipelineFilter  cogl_pipeline_get_layer_mag_filter  (CoglPipeline *pipeline,
                                                         int layer_index);
enum                CoglPipelineWrapMode;
void                cogl_pipeline_set_layer_wrap_mode   (CoglPipeline *pipeline,
                                                         int layer_index,
                                                         CoglPipelineWrapMode mode);
void                cogl_pipeline_set_layer_wrap_mode_s (CoglPipeline *pipeline,
                                                         int layer_index,
                                                         CoglPipelineWrapMode mode);
void                cogl_pipeline_set_layer_wrap_mode_t (CoglPipeline *pipeline,
                                                         int layer_index,
                                                         CoglPipelineWrapMode mode);
void                cogl_pipeline_set_layer_wrap_mode_p (CoglPipeline *pipeline,
                                                         int layer_index,
                                                         CoglPipelineWrapMode mode);
CoglBool            cogl_pipeline_set_layer_combine     (CoglPipeline *pipeline,
                                                         int layer_index,
                                                         const char *blend_string,
                                                         CoglError **error);
void                cogl_pipeline_set_layer_combine_constant
                                                        (CoglPipeline *pipeline,
                                                         int layer_index,
                                                         const CoglColor *constant);
CoglBool            cogl_pipeline_set_layer_point_sprite_coords_enabled
                                                        (CoglPipeline *pipeline,
                                                         int layer_index,
                                                         CoglBool enable,
                                                         CoglError **error);
CoglBool            cogl_pipeline_get_layer_point_sprite_coords_enabled
                                                        (CoglPipeline *pipeline,
                                                         int layer_index);
void                cogl_pipeline_remove_layer          (CoglPipeline *pipeline,
                                                         int layer_index);
int                 cogl_pipeline_get_n_layers          (CoglPipeline *pipeline);
CoglBool            (*CoglPipelineLayerCallback)        (CoglPipeline *pipeline,
                                                         int layer_index,
                                                         void *user_data);
void                cogl_pipeline_foreach_layer         (CoglPipeline *pipeline,
                                                         CoglPipelineLayerCallback callback,
                                                         void *user_data);
int                 cogl_pipeline_get_uniform_location  (CoglPipeline *pipeline,
                                                         const char *uniform_name);
void                cogl_pipeline_set_uniform_1f        (CoglPipeline *pipeline,
                                                         int uniform_location,
                                                         float value);
void                cogl_pipeline_set_uniform_1i        (CoglPipeline *pipeline,
                                                         int uniform_location,
                                                         int value);
void                cogl_pipeline_set_uniform_float     (CoglPipeline *pipeline,
                                                         int uniform_location,
                                                         int n_components,
                                                         int count,
                                                         const float *value);
void                cogl_pipeline_set_uniform_int       (CoglPipeline *pipeline,
                                                         int uniform_location,
                                                         int n_components,
                                                         int count,
                                                         const int *value);
void                cogl_pipeline_set_uniform_matrix    (CoglPipeline *pipeline,
                                                         int uniform_location,
                                                         int dimensions,
                                                         int count,
                                                         CoglBool transpose,
                                                         const float *value);
void                cogl_pipeline_add_snippet           (CoglPipeline *pipeline,
                                                         CoglSnippet *snippet);
void                cogl_pipeline_add_layer_snippet     (CoglPipeline *pipeline,
                                                         int layer,
                                                         CoglSnippet *snippet);

Description

Cogl allows creating and manipulating objects representing the full configuration of the GPU pipeline. In simplified terms the GPU pipeline takes primitive geometry as the input, it first performs vertex processing, allowing you to deform your geometry, then rasterizes that (turning it from pure geometry into fragments) then performs fragment processing including depth testing and texture mapping. Finally it blends the result with the framebuffer.

Details

CoglPipeline

typedef struct _CoglPipeline CoglPipeline;

cogl_pipeline_new ()

CoglPipeline *      cogl_pipeline_new                   (CoglContext *context);

Allocates and initializes a default simple pipeline that will color a primitive white.

context :

a CoglContext

Returns :

a pointer to a new CoglPipeline

Since 2.0

Stability Level: Unstable


cogl_pipeline_copy ()

CoglPipeline *      cogl_pipeline_copy                  (CoglPipeline *source);

Creates a new pipeline with the configuration copied from the source pipeline.

We would strongly advise developers to always aim to use cogl_pipeline_copy() instead of cogl_pipeline_new() whenever there will be any similarity between two pipelines. Copying a pipeline helps Cogl keep track of a pipelines ancestry which we may use to help minimize GPU state changes.

source :

a CoglPipeline object to copy

Returns :

a pointer to the newly allocated CoglPipeline

Since 2.0

Stability Level: Unstable


cogl_is_pipeline ()

CoglBool            cogl_is_pipeline                    (void *object);

Gets whether the given object references an existing pipeline object.

object :

A CoglObject

Returns :

TRUE if the object references a CoglPipeline, FALSE otherwise

Since 2.0

Stability Level: Unstable


cogl_pipeline_set_color ()

void                cogl_pipeline_set_color             (CoglPipeline *pipeline,
                                                         const CoglColor *color);

Sets the basic color of the pipeline, used when no lighting is enabled.

Note that if you don't add any layers to the pipeline then the color will be blended unmodified with the destination; the default blend expects premultiplied colors: for example, use (0.5, 0.0, 0.0, 0.5) for semi-transparent red. See cogl_color_premultiply().

The default value is (1.0, 1.0, 1.0, 1.0)

pipeline :

A CoglPipeline object

color :

The components of the color

Since 2.0

Stability Level: Unstable


cogl_pipeline_set_color4ub ()

void                cogl_pipeline_set_color4ub          (CoglPipeline *pipeline,
                                                         uint8_t red,
                                                         uint8_t green,
                                                         uint8_t blue,
                                                         uint8_t alpha);

Sets the basic color of the pipeline, used when no lighting is enabled.

The default value is (0xff, 0xff, 0xff, 0xff)

pipeline :

A CoglPipeline object

red :

The red component

green :

The green component

blue :

The blue component

alpha :

The alpha component

Since 2.0

Stability Level: Unstable


cogl_pipeline_set_color4f ()

void                cogl_pipeline_set_color4f           (CoglPipeline *pipeline,
                                                         float red,
                                                         float green,
                                                         float blue,
                                                         float alpha);

Sets the basic color of the pipeline, used when no lighting is enabled.

The default value is (1.0, 1.0, 1.0, 1.0)

pipeline :

A CoglPipeline object

red :

The red component

green :

The green component

blue :

The blue component

alpha :

The alpha component

Since 2.0

Stability Level: Unstable


cogl_pipeline_get_color ()

void                cogl_pipeline_get_color             (CoglPipeline *pipeline,
                                                         CoglColor *color);

Retrieves the current pipeline color.

pipeline :

A CoglPipeline object

color :

The location to store the color. [out]

Since 2.0

Stability Level: Unstable


enum CoglPipelineAlphaFunc

typedef enum {
  COGL_PIPELINE_ALPHA_FUNC_NEVER    = 0x0200,
  COGL_PIPELINE_ALPHA_FUNC_LESS	    = 0x0201,
  COGL_PIPELINE_ALPHA_FUNC_EQUAL    = 0x0202,
  COGL_PIPELINE_ALPHA_FUNC_LEQUAL   = 0x0203,
  COGL_PIPELINE_ALPHA_FUNC_GREATER  = 0x0204,
  COGL_PIPELINE_ALPHA_FUNC_NOTEQUAL = 0x0205,
  COGL_PIPELINE_ALPHA_FUNC_GEQUAL   = 0x0206,
  COGL_PIPELINE_ALPHA_FUNC_ALWAYS   = 0x0207
} CoglPipelineAlphaFunc;

Alpha testing happens before blending primitives with the framebuffer and gives an opportunity to discard fragments based on a comparison with the incoming alpha value and a reference alpha value. The CoglPipelineAlphaFunc determines how the comparison is done.

COGL_PIPELINE_ALPHA_FUNC_NEVER

Never let the fragment through.

COGL_PIPELINE_ALPHA_FUNC_LESS

Let the fragment through if the incoming alpha value is less than the reference alpha value

COGL_PIPELINE_ALPHA_FUNC_EQUAL

Let the fragment through if the incoming alpha value equals the reference alpha value

COGL_PIPELINE_ALPHA_FUNC_LEQUAL

Let the fragment through if the incoming alpha value is less than or equal to the reference alpha value

COGL_PIPELINE_ALPHA_FUNC_GREATER

Let the fragment through if the incoming alpha value is greater than the reference alpha value

COGL_PIPELINE_ALPHA_FUNC_NOTEQUAL

Let the fragment through if the incoming alpha value does not equal the reference alpha value

COGL_PIPELINE_ALPHA_FUNC_GEQUAL

Let the fragment through if the incoming alpha value is greater than or equal to the reference alpha value.

COGL_PIPELINE_ALPHA_FUNC_ALWAYS

Always let the fragment through.

cogl_pipeline_set_alpha_test_function ()

void                cogl_pipeline_set_alpha_test_function
                                                        (CoglPipeline *pipeline,
                                                         CoglPipelineAlphaFunc alpha_func,
                                                         float alpha_reference);

Before a primitive is blended with the framebuffer, it goes through an alpha test stage which lets you discard fragments based on the current alpha value. This function lets you change the function used to evaluate the alpha channel, and thus determine which fragments are discarded and which continue on to the blending stage.

The default is COGL_PIPELINE_ALPHA_FUNC_ALWAYS

pipeline :

A CoglPipeline object

alpha_func :

A CoglPipelineAlphaFunc constant

alpha_reference :

A reference point that the chosen alpha function uses to compare incoming fragments to.

Since 2.0

Stability Level: Unstable


COGL_BLEND_STRING_ERROR

#define COGL_BLEND_STRING_ERROR (cogl_blend_string_error_domain ())

CoglError domain for blend string parser errors

Since 1.0


enum CoglBlendStringError

typedef enum {
 /*< prefix=COGL_BLEND_STRING_ERROR >*/
  COGL_BLEND_STRING_ERROR_PARSE_ERROR,
  COGL_BLEND_STRING_ERROR_ARGUMENT_PARSE_ERROR,
  COGL_BLEND_STRING_ERROR_INVALID_ERROR,
  COGL_BLEND_STRING_ERROR_GPU_UNSUPPORTED_ERROR
} CoglBlendStringError;

Error enumeration for the blend strings parser

COGL_BLEND_STRING_ERROR_PARSE_ERROR

Generic parse error

COGL_BLEND_STRING_ERROR_ARGUMENT_PARSE_ERROR

Argument parse error

COGL_BLEND_STRING_ERROR_INVALID_ERROR

Internal parser error

COGL_BLEND_STRING_ERROR_GPU_UNSUPPORTED_ERROR

Blend string not supported by the GPU

Since 1.0


cogl_pipeline_set_blend ()

CoglBool            cogl_pipeline_set_blend             (CoglPipeline *pipeline,
                                                         const char *blend_string,
                                                         CoglError **error);

If not already familiar; please refer here for an overview of what blend strings are, and their syntax.

Blending occurs after the alpha test function, and combines fragments with the framebuffer.

Currently the only blend function Cogl exposes is ADD(). So any valid blend statements will be of the form:

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18










<channel-mask>=ADD(SRC_COLOR*(<factor>), DST_COLOR*(<factor>))

This is the list of source-names usable as blend factors:

The source names can be used according to the color-source and factor syntax, so for example "(1-SRC_COLOR[A])" would be a valid factor, as would "(CONSTANT[RGB])"

These can also be used as factors:

  • 0: (0, 0, 0, 0)
  • 1: (1, 1, 1, 1)
  • SRC_ALPHA_SATURATE_FACTOR: (f,f,f,1) where f = MIN(SRC_COLOR[A],1-DST_COLOR[A])

Note

Remember; all color components are normalized to the range [0, 1] before computing the result of blending.

Example 1. Blend Strings/1

Blend a non-premultiplied source over a destination with premultiplied alpha:

"RGB = ADD(SRC_COLOR*(SRC_COLOR[A]), DST_COLOR*(1-SRC_COLOR[A]))"
"A   = ADD(SRC_COLOR, DST_COLOR*(1-SRC_COLOR[A]))"
  


Example 2. Blend Strings/2

Blend a premultiplied source over a destination with premultiplied alpha

"RGBA = ADD(SRC_COLOR, DST_COLOR*(1-SRC_COLOR[A]))"
  


The default blend string is:

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19










RGBA = ADD (SRC_COLOR, DST_COLOR*(1-SRC_COLOR[A]))

That gives normal alpha-blending when the calculated color for the pipeline is in premultiplied form.

pipeline :

A CoglPipeline object

blend_string :

A Cogl blend string describing the desired blend function.

error :

return location for a CoglError that may report lack of driver support if you give separate blend string statements for the alpha channel and RGB channels since some drivers, or backends such as GLES 1.1, don't support this feature. May be NULL, in which case a warning will be printed out using GLib's logging facilities if an error is encountered.

Returns :

TRUE if the blend string was successfully parsed, and the described blending is supported by the underlying driver/hardware. If there was an error, FALSE is returned and error is set accordingly (if present).

Since 2.0

Stability Level: Unstable


cogl_pipeline_set_blend_constant ()

void                cogl_pipeline_set_blend_constant    (CoglPipeline *pipeline,
                                                         const CoglColor *constant_color);

When blending is setup to reference a CONSTANT blend factor then blending will depend on the constant set with this function.

pipeline :

A CoglPipeline object

constant_color :

The constant color you want

Since 2.0

Stability Level: Unstable


cogl_pipeline_set_point_size ()

void                cogl_pipeline_set_point_size        (CoglPipeline *pipeline,
                                                         float point_size);

Changes the size of points drawn when COGL_VERTICES_MODE_POINTS is used with the attribute buffer API. Note that typically the GPU will only support a limited minimum and maximum range of point sizes. If the chosen point size is outside that range then the nearest value within that range will be used instead. The size of a point is in screen space so it will be the same regardless of any transformations.

If the point size is set to 0.0 then drawing points with the pipeline will have undefined results. This is the default value so if an application wants to draw points it must make sure to use a pipeline that has an explicit point size set on it.

pipeline :

a CoglPipeline pointer

point_size :

the new point size.

Since 2.0

Stability Level: Unstable


cogl_pipeline_get_point_size ()

float               cogl_pipeline_get_point_size        (CoglPipeline *pipeline);

Get the size of points drawn when COGL_VERTICES_MODE_POINTS is used with the vertex buffer API.

pipeline :

a CoglPipeline pointer

Returns :

the point size of the pipeline.

Since 2.0

Stability Level: Unstable


cogl_pipeline_set_per_vertex_point_size ()

CoglBool            cogl_pipeline_set_per_vertex_point_size
                                                        (CoglPipeline *pipeline,
                                                         CoglBool enable,
                                                         CoglError **error);

Sets whether to use a per-vertex point size or to use the value set by cogl_pipeline_set_point_size(). If per-vertex point size is enabled then the point size can be set for an individual point either by drawing with a CoglAttribute with the name ‘cogl_point_size_in’ or by writing to the GLSL builtin ‘cogl_point_size_out’ from a vertex shader snippet.

If per-vertex point size is enabled and this attribute is not used and cogl_point_size_out is not written to then the results are undefined.

Note that enabling this will only work if the COGL_FEATURE_ID_PER_VERTEX_POINT_SIZE feature is available. If this is not available then the function will return FALSE and set a CoglError.

pipeline :

a CoglPipeline pointer

enable :

whether to enable per-vertex point size

error :

a location to store a CoglError if the change failed

Returns :

TRUE if the change suceeded or FALSE otherwise

Since 2.0

Stability Level: Unstable


cogl_pipeline_get_per_vertex_point_size ()

CoglBool            cogl_pipeline_get_per_vertex_point_size
                                                        (CoglPipeline *pipeline);

pipeline :

a CoglPipeline pointer

Returns :

TRUE if the pipeline has per-vertex point size enabled or FALSE otherwise. The per-vertex point size can be enabled with cogl_pipeline_set_per_vertex_point_size().

Since 2.0

Stability Level: Unstable


cogl_pipeline_get_color_mask ()

CoglColorMask       cogl_pipeline_get_color_mask        (CoglPipeline *pipeline);

Gets the current CoglColorMask of which channels would be written to the current framebuffer. Each bit set in the mask means that the corresponding color would be written.

pipeline :

a CoglPipeline object.

Returns :

A CoglColorMask

Since 1.8

Stability Level: Unstable


cogl_pipeline_set_color_mask ()

void                cogl_pipeline_set_color_mask        (CoglPipeline *pipeline,
                                                         CoglColorMask color_mask);

Defines a bit mask of which color channels should be written to the current framebuffer. If a bit is set in color_mask that means that color will be written.

pipeline :

a CoglPipeline object.

color_mask :

A CoglColorMask of which color channels to write to the current framebuffer.

Since 1.8

Stability Level: Unstable


cogl_pipeline_set_depth_state ()

CoglBool            cogl_pipeline_set_depth_state       (CoglPipeline *pipeline,
                                                         const CoglDepthState *state,
                                                         CoglError **error);

This commits all the depth state configured in state struct to the given pipeline. The configuration values are copied into the pipeline so there is no requirement to keep the CoglDepthState struct around if you don't need it any more.

Note: Since some platforms do not support the depth range feature it is possible for this function to fail and report an error.

pipeline :

A CoglPipeline object

state :

A CoglDepthState struct

error :

A CoglError to report failures to setup the given state.

Returns :

TRUE if the GPU supports all the given state else FALSE and returns an error.

Since 2.0

Stability Level: Unstable


cogl_pipeline_get_depth_state ()

void                cogl_pipeline_get_depth_state       (CoglPipeline *pipeline,
                                                         CoglDepthState *state_out);

Retrieves the current depth state configuration for the given pipeline as previously set using cogl_pipeline_set_depth_state().

pipeline :

A CoglPipeline object

state_out :

A destination CoglDepthState struct. [out]

Since 2.0

Stability Level: Unstable


enum CoglPipelineCullFaceMode

typedef enum {
  COGL_PIPELINE_CULL_FACE_MODE_NONE,
  COGL_PIPELINE_CULL_FACE_MODE_FRONT,
  COGL_PIPELINE_CULL_FACE_MODE_BACK,
  COGL_PIPELINE_CULL_FACE_MODE_BOTH
} CoglPipelineCullFaceMode;

Specifies which faces should be culled. This can be set on a pipeline using cogl_pipeline_set_cull_face_mode().

COGL_PIPELINE_CULL_FACE_MODE_NONE

Neither face will be culled. This is the default.

COGL_PIPELINE_CULL_FACE_MODE_FRONT

Front faces will be culled.

COGL_PIPELINE_CULL_FACE_MODE_BACK

Back faces will be culled.

COGL_PIPELINE_CULL_FACE_MODE_BOTH

All faces will be culled.

cogl_pipeline_set_cull_face_mode ()

void                cogl_pipeline_set_cull_face_mode    (CoglPipeline *pipeline,
                                                         CoglPipelineCullFaceMode cull_face_mode);

Sets which faces will be culled when drawing. Face culling can be used to increase efficiency by avoiding drawing faces that would get overridden. For example, if a model has gaps so that it is impossible to see the inside then faces which are facing away from the screen will never be seen so there is no point in drawing them. This can be acheived by setting the cull face mode to COGL_PIPELINE_CULL_FACE_MODE_BACK.

Face culling relies on the primitives being drawn with a specific order to represent which faces are facing inside and outside the model. This order can be specified by calling cogl_pipeline_set_front_face_winding().

Status: Unstable

pipeline :

A CoglPipeline

cull_face_mode :

The new mode to set

Since 2.0


enum CoglWinding

typedef enum {
  COGL_WINDING_CLOCKWISE,
  COGL_WINDING_COUNTER_CLOCKWISE
} CoglWinding;

Enum used to represent the two directions of rotation. This can be used to set the front face for culling by calling cogl_pipeline_set_front_face_winding().

COGL_WINDING_CLOCKWISE

Vertices are in a clockwise order

COGL_WINDING_COUNTER_CLOCKWISE

Vertices are in a counter-clockwise order

cogl_pipeline_set_front_face_winding ()

void                cogl_pipeline_set_front_face_winding
                                                        (CoglPipeline *pipeline,
                                                         CoglWinding front_winding);

The order of the vertices within a primitive specifies whether it is considered to be front or back facing. This function specifies which order is considered to be the front faces. COGL_WINDING_COUNTER_CLOCKWISE sets the front faces to primitives with vertices in a counter-clockwise order and COGL_WINDING_CLOCKWISE sets them to be clockwise. The default is COGL_WINDING_COUNTER_CLOCKWISE.

Status: Unstable

pipeline :

a CoglPipeline

front_winding :

the winding order

Since 2.0


cogl_pipeline_set_layer_texture ()

void                cogl_pipeline_set_layer_texture     (CoglPipeline *pipeline,
                                                         int layer_index,
                                                         CoglTexture *texture);

cogl_pipeline_set_layer_null_texture ()

void                cogl_pipeline_set_layer_null_texture
                                                        (CoglPipeline *pipeline,
                                                         int layer_index,
                                                         CoglTextureType texture_type);

Sets the texture for this layer to be the default texture for the given type. This is equivalent to calling cogl_pipeline_set_layer_texture() with NULL for the texture argument except that you can also specify the type of default texture to use. The default texture is a 1x1 pixel white texture.

This function is mostly useful if you want to create a base pipeline that you want to create multiple copies from using cogl_pipeline_copy(). In that case this function can be used to specify the texture type so that any pipeline copies can share the internal texture type state for efficiency.

pipeline :

A CoglPipeline

layer_index :

The layer number to modify

texture_type :

The type of the default texture to use

Since 1.10

Stability Level: Unstable


cogl_pipeline_get_layer_texture ()

CoglTexture *       cogl_pipeline_get_layer_texture     (CoglPipeline *pipeline,
                                                         int layer_index);

pipeline :

A CoglPipeline object

layer_index :

the index of the layer

Returns :

the texture that was set for the given layer of the pipeline or NULL if no texture was set.

Since 1.10

Stability Level: Unstable


enum CoglPipelineFilter

typedef enum {
  COGL_PIPELINE_FILTER_NEAREST = 0x2600,
  COGL_PIPELINE_FILTER_LINEAR = 0x2601,
  COGL_PIPELINE_FILTER_NEAREST_MIPMAP_NEAREST = 0x2700,
  COGL_PIPELINE_FILTER_LINEAR_MIPMAP_NEAREST = 0x2701,
  COGL_PIPELINE_FILTER_NEAREST_MIPMAP_LINEAR = 0x2702,
  COGL_PIPELINE_FILTER_LINEAR_MIPMAP_LINEAR = 0x2703
} CoglPipelineFilter;

Texture filtering is used whenever the current pixel maps either to more than one texture element (texel) or less than one. These filter enums correspond to different strategies used to come up with a pixel color, by possibly referring to multiple neighbouring texels and taking a weighted average or simply using the nearest texel.

COGL_PIPELINE_FILTER_NEAREST

Measuring in manhatten distance from the, current pixel center, use the nearest texture texel

COGL_PIPELINE_FILTER_LINEAR

Use the weighted average of the 4 texels nearest the current pixel center

COGL_PIPELINE_FILTER_NEAREST_MIPMAP_NEAREST

Select the mimap level whose texel size most closely matches the current pixel, and use the COGL_PIPELINE_FILTER_NEAREST criterion

COGL_PIPELINE_FILTER_LINEAR_MIPMAP_NEAREST

Select the mimap level whose texel size most closely matches the current pixel, and use the COGL_PIPELINE_FILTER_LINEAR criterion

COGL_PIPELINE_FILTER_NEAREST_MIPMAP_LINEAR

Select the two mimap levels whose texel size most closely matches the current pixel, use the COGL_PIPELINE_FILTER_NEAREST criterion on each one and take their weighted average

COGL_PIPELINE_FILTER_LINEAR_MIPMAP_LINEAR

Select the two mimap levels whose texel size most closely matches the current pixel, use the COGL_PIPELINE_FILTER_LINEAR criterion on each one and take their weighted average

cogl_pipeline_set_layer_filters ()

void                cogl_pipeline_set_layer_filters     (CoglPipeline *pipeline,
                                                         int layer_index,
                                                         CoglPipelineFilter min_filter,
                                                         CoglPipelineFilter mag_filter);

Changes the decimation and interpolation filters used when a texture is drawn at other scales than 100%.

Note

It is an error to pass anything other than COGL_PIPELINE_FILTER_NEAREST or COGL_PIPELINE_FILTER_LINEAR as magnification filters since magnification doesn't ever need to reference values stored in the mipmap chain.

pipeline :

A CoglPipeline object

layer_index :

the layer number to change.

min_filter :

the filter used when scaling a texture down.

mag_filter :

the filter used when magnifying a texture.

Since 1.10

Stability Level: Unstable


cogl_pipeline_get_layer_min_filter ()

CoglPipelineFilter  cogl_pipeline_get_layer_min_filter  (CoglPipeline *pipeline,
                                                         int layer_index);

Retrieves the currently set minification CoglPipelineFilter set on the specified layer. The miniifcation filter determines how the layer should be sampled when down-scaled.

The default filter is COGL_PIPELINE_FILTER_LINEAR but this can be changed using cogl_pipeline_set_layer_filters().

pipeline :

A CoglPipeline object

layer_index :

the layer number to change.

Returns :

The minification CoglPipelineFilter for the specified layer.

Since 1.10

Stability Level: Unstable


cogl_pipeline_get_layer_mag_filter ()

CoglPipelineFilter  cogl_pipeline_get_layer_mag_filter  (CoglPipeline *pipeline,
                                                         int layer_index);

Retrieves the currently set magnification CoglPipelineFilter set on the specified layer. The magnification filter determines how the layer should be sampled when up-scaled.

The default filter is COGL_PIPELINE_FILTER_LINEAR but this can be changed using cogl_pipeline_set_layer_filters().

pipeline :

A CoglPipeline object

layer_index :

the layer number to change.

Returns :

The magnification CoglPipelineFilter for the specified layer.

Since 1.10

Stability Level: Unstable


enum CoglPipelineWrapMode

typedef enum {
  COGL_PIPELINE_WRAP_MODE_REPEAT = 0x2901,
  COGL_PIPELINE_WRAP_MODE_MIRRORED_REPEAT = 0x8370,
  COGL_PIPELINE_WRAP_MODE_CLAMP_TO_EDGE = 0x812F,
  COGL_PIPELINE_WRAP_MODE_AUTOMATIC = 0x0207 /* GL_ALWAYS */
} CoglPipelineWrapMode;

The wrap mode specifies what happens when texture coordinates outside the range 0→1 are used. Note that if the filter mode is anything but COGL_PIPELINE_FILTER_NEAREST then texels outside the range 0→1 might be used even when the coordinate is exactly 0 or 1 because OpenGL will try to sample neighbouring pixels. For example if you are trying to render the full texture then you may get artifacts around the edges when the pixels from the other side are merged in if the wrap mode is set to repeat.

COGL_PIPELINE_WRAP_MODE_REPEAT

The texture will be repeated. This is useful for example to draw a tiled background.

COGL_PIPELINE_WRAP_MODE_MIRRORED_REPEAT

COGL_PIPELINE_WRAP_MODE_CLAMP_TO_EDGE

The coordinates outside the range 0→1 will sample copies of the edge pixels of the texture. This is useful to avoid artifacts if only one copy of the texture is being rendered.

COGL_PIPELINE_WRAP_MODE_AUTOMATIC

Cogl will try to automatically decide which of the above two to use. For cogl_framebuffer_draw_rectangle(), it will use repeat mode if any of the texture coordinates are outside the range 0→1, otherwise it will use clamp to edge. For cogl_framebuffer_draw_attributes() or cogl_primitive_draw() it will use repeat mode except for layers that have point sprite coordinate generation enabled. This is the default value.

Since 2.0


cogl_pipeline_set_layer_wrap_mode ()

void                cogl_pipeline_set_layer_wrap_mode   (CoglPipeline *pipeline,
                                                         int layer_index,
                                                         CoglPipelineWrapMode mode);

Sets the wrap mode for all three coordinates of texture lookups on this layer. This is equivalent to calling cogl_pipeline_set_layer_wrap_mode_s(), cogl_pipeline_set_layer_wrap_mode_t() and cogl_pipeline_set_layer_wrap_mode_p() separately.

pipeline :

A CoglPipeline object

layer_index :

the layer number to change.

mode :

the new wrap mode

Since 2.0

Stability Level: Unstable


cogl_pipeline_set_layer_wrap_mode_s ()

void                cogl_pipeline_set_layer_wrap_mode_s (CoglPipeline *pipeline,
                                                         int layer_index,
                                                         CoglPipelineWrapMode mode);

Sets the wrap mode for the 's' coordinate of texture lookups on this layer.

pipeline :

A CoglPipeline object

layer_index :

the layer number to change.

mode :

the new wrap mode

Since 2.0

Stability Level: Unstable


cogl_pipeline_set_layer_wrap_mode_t ()

void                cogl_pipeline_set_layer_wrap_mode_t (CoglPipeline *pipeline,
                                                         int layer_index,
                                                         CoglPipelineWrapMode mode);

Sets the wrap mode for the 't' coordinate of texture lookups on this layer.

pipeline :

A CoglPipeline object

layer_index :

the layer number to change.

mode :

the new wrap mode

Since 2.0

Stability Level: Unstable


cogl_pipeline_set_layer_wrap_mode_p ()

void                cogl_pipeline_set_layer_wrap_mode_p (CoglPipeline *pipeline,
                                                         int layer_index,
                                                         CoglPipelineWrapMode mode);

Sets the wrap mode for the 'p' coordinate of texture lookups on this layer. 'p' is the third coordinate.

pipeline :

A CoglPipeline object

layer_index :

the layer number to change.

mode :

the new wrap mode

Since 2.0

Stability Level: Unstable


cogl_pipeline_set_layer_combine ()

CoglBool            cogl_pipeline_set_layer_combine     (CoglPipeline *pipeline,
                                                         int layer_index,
                                                         const char *blend_string,
                                                         CoglError **error);

If not already familiar; you can refer here for an overview of what blend strings are and there syntax.

These are all the functions available for texture combining:

  • REPLACE(arg0) = arg0
  • MODULATE(arg0, arg1) = arg0 x arg1
  • ADD(arg0, arg1) = arg0 + arg1
  • ADD_SIGNED(arg0, arg1) = arg0 + arg1 - 0.5
  • INTERPOLATE(arg0, arg1, arg2) = arg0 x arg2 + arg1 x (1 - arg2)
  • SUBTRACT(arg0, arg1) = arg0 - arg1
  •  DOT3_RGB(arg0, arg1) = 4 x ((arg0[R] - 0.5)) * (arg1[R] - 0.5) +
                                 (arg0[G] - 0.5)) * (arg1[G] - 0.5) +
                                 (arg0[B] - 0.5)) * (arg1[B] - 0.5))
        
  •  DOT3_RGBA(arg0, arg1) = 4 x ((arg0[R] - 0.5)) * (arg1[R] - 0.5) +
                                  (arg0[G] - 0.5)) * (arg1[G] - 0.5) +
                                  (arg0[B] - 0.5)) * (arg1[B] - 0.5))
        

Refer to the color-source syntax for describing the arguments. The valid source names for texture combining are:

TEXTURE

Use the color from the current texture layer

TEXTURE_0, TEXTURE_1, etc

Use the color from the specified texture layer

CONSTANT

Use the color from the constant given with cogl_pipeline_set_layer_combine_constant()

PRIMARY

Use the color of the pipeline as set with cogl_pipeline_set_color()

PREVIOUS

Either use the texture color from the previous layer, or if this is layer 0, use the color of the pipeline as set with cogl_pipeline_set_color()

Layer Combine Examples

This is effectively what the default blending is:

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18










RGBA = MODULATE (PREVIOUS, TEXTURE)

This could be used to cross-fade between two images, using the alpha component of a constant as the interpolator. The constant color is given by calling cogl_pipeline_set_layer_combine_constant().

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18










RGBA = INTERPOLATE (PREVIOUS, TEXTURE, CONSTANT[A])

Note

You can't give a multiplication factor for arguments as you can with blending.

pipeline :

A CoglPipeline object

layer_index :

Specifies the layer you want define a combine function for

blend_string :

A Cogl blend string describing the desired texture combine function.

error :

A CoglError that may report parse errors or lack of GPU/driver support. May be NULL, in which case a warning will be printed out if an error is encountered.

Returns :

TRUE if the blend string was successfully parsed, and the described texture combining is supported by the underlying driver and or hardware. On failure, FALSE is returned and error is set

Since 2.0

Stability Level: Unstable


cogl_pipeline_set_layer_combine_constant ()

void                cogl_pipeline_set_layer_combine_constant
                                                        (CoglPipeline *pipeline,
                                                         int layer_index,
                                                         const CoglColor *constant);

When you are using the 'CONSTANT' color source in a layer combine description then you can use this function to define its value.

pipeline :

A CoglPipeline object

layer_index :

Specifies the layer you want to specify a constant used for texture combining

constant :

The constant color you want

Since 2.0

Stability Level: Unstable


cogl_pipeline_set_layer_point_sprite_coords_enabled ()

CoglBool            cogl_pipeline_set_layer_point_sprite_coords_enabled
                                                        (CoglPipeline *pipeline,
                                                         int layer_index,
                                                         CoglBool enable,
                                                         CoglError **error);

When rendering points, if enable is TRUE then the texture coordinates for this layer will be replaced with coordinates that vary from 0.0 to 1.0 across the primitive. The top left of the point will have the coordinates 0.0,0.0 and the bottom right will have 1.0,1.0. If enable is FALSE then the coordinates will be fixed for the entire point.

This function will only work if COGL_FEATURE_ID_POINT_SPRITE is available. If the feature is not available then the function will return FALSE and set error.

pipeline :

A CoglPipeline object

layer_index :

the layer number to change.

enable :

whether to enable point sprite coord generation.

error :

A return location for a CoglError, or NULL to ignore errors.

Returns :

TRUE if the function succeeds, FALSE otherwise.

Since 2.0

Stability Level: Unstable


cogl_pipeline_get_layer_point_sprite_coords_enabled ()

CoglBool            cogl_pipeline_get_layer_point_sprite_coords_enabled
                                                        (CoglPipeline *pipeline,
                                                         int layer_index);

Gets whether point sprite coordinate generation is enabled for this texture layer.

pipeline :

A CoglPipeline object

layer_index :

the layer number to check.

Returns :

whether the texture coordinates will be replaced with point sprite coordinates.

Since 2.0

Stability Level: Unstable


cogl_pipeline_remove_layer ()

void                cogl_pipeline_remove_layer          (CoglPipeline *pipeline,
                                                         int layer_index);

This function removes a layer from your pipeline

pipeline :

A CoglPipeline object

layer_index :

Specifies the layer you want to remove

Since 1.10

Stability Level: Unstable


cogl_pipeline_get_n_layers ()

int                 cogl_pipeline_get_n_layers          (CoglPipeline *pipeline);

Retrieves the number of layers defined for the given pipeline

pipeline :

A CoglPipeline object

Returns :

the number of layers

Since 2.0

Stability Level: Unstable


CoglPipelineLayerCallback ()

CoglBool            (*CoglPipelineLayerCallback)        (CoglPipeline *pipeline,
                                                         int layer_index,
                                                         void *user_data);

The callback prototype used with cogl_pipeline_foreach_layer() for iterating all the layers of a pipeline.

pipeline :

The CoglPipeline whos layers are being iterated

layer_index :

The current layer index

user_data :

The private data passed to cogl_pipeline_foreach_layer()

Since 2.0

Stability Level: Unstable


cogl_pipeline_foreach_layer ()

void                cogl_pipeline_foreach_layer         (CoglPipeline *pipeline,
                                                         CoglPipelineLayerCallback callback,
                                                         void *user_data);

Iterates all the layer indices of the given pipeline.

pipeline :

A CoglPipeline object

callback :

A CoglPipelineLayerCallback to be called for each layer index

user_data :

Private data that will be passed to the callback

Since 2.0

Stability Level: Unstable


cogl_pipeline_get_uniform_location ()

int                 cogl_pipeline_get_uniform_location  (CoglPipeline *pipeline,
                                                         const char *uniform_name);

This is used to get an integer representing the uniform with the name uniform_name. The integer can be passed to functions such as cogl_pipeline_set_uniform_1f() to set the value of a uniform.

This function will always return a valid integer. Ie, unlike OpenGL, it does not return -1 if the uniform is not available in this pipeline so it can not be used to test whether uniforms are present. It is not necessary to set the program on the pipeline before calling this function.

pipeline :

A CoglPipeline object

uniform_name :

The name of a uniform

Returns :

A integer representing the location of the given uniform.

Since 2.0

Stability Level: Unstable


cogl_pipeline_set_uniform_1f ()

void                cogl_pipeline_set_uniform_1f        (CoglPipeline *pipeline,
                                                         int uniform_location,
                                                         float value);

Sets a new value for the uniform at uniform_location. If this pipeline has a user program attached and is later used as a source for drawing, the given value will be assigned to the uniform which can be accessed from the shader's source. The value for uniform_location should be retrieved from the string name of the uniform by calling cogl_pipeline_get_uniform_location().

This function should be used to set uniforms that are of type float. It can also be used to set a single member of a float array uniform.

pipeline :

A CoglPipeline object

uniform_location :

The uniform's location identifier

value :

The new value for the uniform

Since 2.0

Stability Level: Unstable


cogl_pipeline_set_uniform_1i ()

void                cogl_pipeline_set_uniform_1i        (CoglPipeline *pipeline,
                                                         int uniform_location,
                                                         int value);

Sets a new value for the uniform at uniform_location. If this pipeline has a user program attached and is later used as a source for drawing, the given value will be assigned to the uniform which can be accessed from the shader's source. The value for uniform_location should be retrieved from the string name of the uniform by calling cogl_pipeline_get_uniform_location().

This function should be used to set uniforms that are of type int. It can also be used to set a single member of a int array uniform or a sampler uniform.

pipeline :

A CoglPipeline object

uniform_location :

The uniform's location identifier

value :

The new value for the uniform

Since 2.0

Stability Level: Unstable


cogl_pipeline_set_uniform_float ()

void                cogl_pipeline_set_uniform_float     (CoglPipeline *pipeline,
                                                         int uniform_location,
                                                         int n_components,
                                                         int count,
                                                         const float *value);

Sets new values for the uniform at uniform_location. If this pipeline has a user program attached and is later used as a source for drawing, the given values will be assigned to the uniform which can be accessed from the shader's source. The value for uniform_location should be retrieved from the string name of the uniform by calling cogl_pipeline_get_uniform_location().

This function can be used to set any floating point type uniform, including float arrays and float vectors. For example, to set a single vec4 uniform you would use 4 for n_components and 1 for count. To set an array of 8 float values, you could use 1 for n_components and 8 for count.

pipeline :

A CoglPipeline object

uniform_location :

The uniform's location identifier

n_components :

The number of components in the corresponding uniform's type

count :

The number of values to set

value :

Pointer to the new values to set

Since 2.0

Stability Level: Unstable


cogl_pipeline_set_uniform_int ()

void                cogl_pipeline_set_uniform_int       (CoglPipeline *pipeline,
                                                         int uniform_location,
                                                         int n_components,
                                                         int count,
                                                         const int *value);

Sets new values for the uniform at uniform_location. If this pipeline has a user program attached and is later used as a source for drawing, the given values will be assigned to the uniform which can be accessed from the shader's source. The value for uniform_location should be retrieved from the string name of the uniform by calling cogl_pipeline_get_uniform_location().

This function can be used to set any integer type uniform, including int arrays and int vectors. For example, to set a single ivec4 uniform you would use 4 for n_components and 1 for count. To set an array of 8 int values, you could use 1 for n_components and 8 for count.

pipeline :

A CoglPipeline object

uniform_location :

The uniform's location identifier

n_components :

The number of components in the corresponding uniform's type

count :

The number of values to set

value :

Pointer to the new values to set

Since 2.0

Stability Level: Unstable


cogl_pipeline_set_uniform_matrix ()

void                cogl_pipeline_set_uniform_matrix    (CoglPipeline *pipeline,
                                                         int uniform_location,
                                                         int dimensions,
                                                         int count,
                                                         CoglBool transpose,
                                                         const float *value);

Sets new values for the uniform at uniform_location. If this pipeline has a user program attached and is later used as a source for drawing, the given values will be assigned to the uniform which can be accessed from the shader's source. The value for uniform_location should be retrieved from the string name of the uniform by calling cogl_pipeline_get_uniform_location().

This function can be used to set any matrix type uniform, including matrix arrays. For example, to set a single mat4 uniform you would use 4 for dimensions and 1 for count. To set an array of 8 mat3 values, you could use 3 for dimensions and 8 for count.

If transpose is FALSE then the matrix is expected to be in column-major order or if it is TRUE then the matrix is in row-major order. You can pass a CoglMatrix by calling by passing the result of cogl_matrix_get_array() in value and setting transpose to FALSE.

pipeline :

A CoglPipeline object

uniform_location :

The uniform's location identifier

dimensions :

The size of the matrix

count :

The number of values to set

transpose :

Whether to transpose the matrix

value :

Pointer to the new values to set

Since 2.0

Stability Level: Unstable


cogl_pipeline_add_snippet ()

void                cogl_pipeline_add_snippet           (CoglPipeline *pipeline,
                                                         CoglSnippet *snippet);

Adds a shader snippet to pipeline. The snippet will wrap around or replace some part of the pipeline as defined by the hook point in snippet. Note that some hook points are specific to a layer and must be added with cogl_pipeline_add_layer_snippet() instead.

pipeline :

A CoglPipeline

snippet :

The CoglSnippet to add to the vertex processing hook

Since 1.10

Stability Level: Unstable


cogl_pipeline_add_layer_snippet ()

void                cogl_pipeline_add_layer_snippet     (CoglPipeline *pipeline,
                                                         int layer,
                                                         CoglSnippet *snippet);

Adds a shader snippet that will hook on to the given layer of the pipeline. The exact part of the pipeline that the snippet wraps around depends on the hook that is given to cogl_snippet_new(). Note that some hooks can't be used with a layer and need to be added with cogl_pipeline_add_snippet() instead.

pipeline :

A CoglPipeline

layer :

The layer to hook the snippet to

snippet :

A CoglSnippet

Since 1.10

Stability Level: Unstable