GraphicsStateGuardian

class GraphicsStateGuardian

Bases: GraphicsStateGuardianBase

Encapsulates all the communication with a particular instance of a given rendering backend. Tries to guarantee that redundant state-change requests are not issued (hence “state guardian”).

There will be one of these objects for each different graphics context active in the system.

Inheritance diagram

Inheritance diagram of GraphicsStateGuardian

enum ShaderModel
enumerator SM_00 = 0
enumerator SM_11 = 1
enumerator SM_20 = 2
enumerator SM_2X = 3
enumerator SM_30 = 4
enumerator SM_40 = 5
enumerator SM_50 = 6
enumerator SM_51 = 7
virtual bool begin_scene(void)
void clear_flash_texture(void)
virtual void end_scene(void)
TextureStage *get_alpha_scale_texture_stage(void)

Returns the TextureStage that will be used to apply an alpha scale, if get_alpha_scale_via_texture() returns true.

bool get_alpha_scale_via_texture(void) const
bool get_alpha_scale_via_texture(TextureAttrib const *tex_attrib) const

Returns true if this particular GSG can implement (or would prefer to implement) an alpha scale via an additional Texture layer, or false if we need to actually munge the alpha.

This variant of get_alpha_scale_via_texture() answers the question of whether the GSG can implement an alpha scale via an additional Texture layer, considering the current TextureAttrib that will be in effect. This considers whether there is at least one additional texture slot available on the GSG.

static TypeHandle get_class_type(void)
bool get_color_scale_via_lighting(void) const

Returns true if this particular GSG can implement (or would prefer to implement) set color and/or color scale using materials and/or ambient lights, or false if we need to actually munge the color.

CoordinateSystem get_coordinate_system(void) const

Returns the coordinate system in effect on this particular gsg. Normally, this will be the default coordinate system, but it might be set differently at runtime.

bool get_copy_texture_inverted(void) const

Returns true if this particular GSG has the property that any framebuffer- to-texture copy results in a texture that is upside-down and backwards from Panda’s usual convention; that is, it copies into a texture from the bottom up instead of from the top down.

If this is true, then on offscreen GraphicsBuffer created for the purposes of rendering into a texture should be created with the invert flag set true, to compensate. Panda will do this automatically if you create an offscreen buffer using GraphicsOutput::make_texture_buffer().

virtual std::string get_driver_renderer(void)
virtual int get_driver_shader_version_major(void)
virtual int get_driver_shader_version_minor(void)
virtual std::string get_driver_vendor(void)
virtual std::string get_driver_version(void)
virtual int get_driver_version_major(void)
virtual int get_driver_version_minor(void)
GraphicsEngine *get_engine(void) const
Texture *get_flash_texture(void) const
PN_stdfloat get_gamma(void) const
virtual CoordinateSystem get_internal_coordinate_system(void) const
Loader *get_loader(void) const

Returns the Loader object that will be used by this GSG to load textures when necessary, if get_incomplete_render() is true.

int get_max_2d_texture_array_layers(void) const

//z axis

//z axis

/**
  • Returns the largest possible number of pages, or -1 if there is no

  • particular limit. Returns 0 if 2-d texture arrays not supported.

  • The value returned may not be meaningful until after the graphics context

  • has been fully created (e.g. the window has been opened).

*/

int get_max_3d_texture_dimension(void) const

Returns the largest possible texture size in any one dimension for a 3-d texture, or -1 if there is no particular limit. Returns 0 if 3-d textures are not supported.

The value returned may not be meaningful until after the graphics context has been fully created (e.g. the window has been opened).

int get_max_buffer_texture_size(void) const

Returns the largest possible buffer texture size, or -1 if there is no particular limit. Returns 0 if cube map textures are not supported.

The value returned may not be meaningful until after the graphics context has been fully created (e.g. the window has been opened).

int get_max_clip_planes(void) const

Returns the maximum number of simultaneous clip planes that may be applied to geometry, or -1 if there is no particular limit.

The value returned may not be meaningful until after the graphics context has been fully created (e.g. the window has been opened).

int get_max_color_targets(void) const

Returns the maximum number of simultaneous color textures that may be attached for render-to-texture, as supported by this particular GSG. If you exceed this number, the lowest-priority render targets will not be applied. Use RenderTarget::set_priority() to adjust the relative importance of the different render targets.

The value returned may not be meaningful until after the graphics context has been fully created (e.g. the window has been opened).

int get_max_cube_map_dimension(void) const

//z axis

//z axis

/**
  • Returns the largest possible texture size in any one dimension for a cube

  • map texture, or -1 if there is no particular limit. Returns 0 if cube map

  • textures are not supported.

  • The value returned may not be meaningful until after the graphics context

  • has been fully created (e.g. the window has been opened).

*/

int get_max_lights(void) const

Returns the maximum number of simultaneous lights that may be rendered on geometry, or -1 if there is no particular limit.

The value returned may not be meaningful until after the graphics context has been fully created (e.g. the window has been opened).

int get_max_texture_stages(void) const

Returns the maximum number of simultaneous textures that may be applied to geometry with multitexturing, as supported by this particular GSG. If you exceed this number, the lowest-priority texture stages will not be applied. Use TextureStage::set_priority() to adjust the relative importance of the different texture stages.

The value returned may not be meaningful until after the graphics context has been fully created (e.g. the window has been opened).

int get_max_vertex_transform_indices(void) const

Returns the maximum number of transforms there may be in a single TransformTable for this graphics hardware. If this number is 0 (but get_max_transforms() is nonzero), then the graphics hardware (or API) doesn’t support indexed transforms, but can support direct transform references.

The value returned may not be meaningful until after the graphics context has been fully created (e.g. the window has been opened).

int get_max_vertex_transforms(void) const

Returns the maximum number of transform matrices that may be simultaneously used to transform any one vertex by the graphics hardware. If this number is 0, then the hardware (or the graphics backend) doesn’t support soft- skinned vertices (in which case Panda will animate the vertices in software).

The value returned may not be meaningful until after the graphics context has been fully created (e.g. the window has been opened).

int get_maximum_simultaneous_render_targets(void) const

Deprecated. Use get_max_color_targets() instead, which returns the exact same value.

GraphicsPipe *get_pipe(void) const

Returns the graphics pipe on which this GSG was created.

virtual PreparedGraphicsObjects *get_prepared_objects(void)
PyObject *get_prepared_textures(void) const
bool get_runtime_color_scale(void) const

Returns true if this particular GSG can implement (or would prefer to implement) set color and/or color scale directly, without requiring any munging of vertices or tricks with lighting.

virtual SceneSetup *get_scene(void) const final
ShaderGenerator *get_shader_generator(void) const

Returns the ShaderGenerator object that will be used by this GSG to generate shaders when necessary.

ShaderModel get_shader_model(void) const

Returns the ShaderModel

bool get_supports_2d_texture_array(void) const

Returns true if this GSG can render 2-d textures array.

bool get_supports_3d_texture(void) const

Returns true if this GSG can render 3-d (volumetric) textures.

bool get_supports_basic_shaders(void) const

Returns true if this particular GSG supports arbfp1+arbvp1 or above.

bool get_supports_buffer_texture(void) const

Returns true if this GSG can render buffer textures.

virtual bool get_supports_cg_profile(std::string const &name) const
bool get_supports_compressed_texture(void) const

Returns true if this GSG can compress textures as it loads them into texture memory, and/or accept pre-compressed textures for storing.

bool get_supports_compute_shaders(void) const

Returns true if this particular GSG supports compute shaders.

bool get_supports_cube_map(void) const

Returns true if this GSG can render cube map textures.

bool get_supports_cube_map_array(void) const

Returns true if this GSG can render cube map arrays.

bool get_supports_depth_stencil(void) const

Returns true if this particular GSG supports textures whose format is F_depth_stencil. This only returns true if the GSG supports the full packed depth-stencil functionality.

bool get_supports_depth_texture(void) const

Returns true if this particular GSG supports textures whose format is F_depth_stencil. This returns true if the GSG supports GL_DEPTH_COMPONENT textures, which are considered a limited but still valid case of F_depth_stencil.

bool get_supports_dual_source_blending(void) const

Returns true if dual source (incoming1_color and incoming1_alpha) blend operands are supported by this GSG.

bool get_supports_generate_mipmap(void) const

Returns true if this particular GSG can generate mipmaps for a texture automatically, or if they must be generated in software. If this is true, then mipmaps can safely be enabled for rendered textures (e.g. using the MultitexReducer).

bool get_supports_geometry_instancing(void) const

Returns true if this particular GSG supports hardware geometry instancing: the ability to render multiple copies of a model. In OpenGL, this is done using the EXT_draw_instanced extension.

bool get_supports_geometry_shaders(void) const

Returns true if this particular GSG supports geometry shaders.

bool get_supports_glsl(void) const

Returns true if this particular GSG supports GLSL shaders.

bool get_supports_indirect_draw(void) const

Returns true if this particular GSG supports draw calls for which the information comes from a buffer.

bool get_supports_luminance_texture(void) const

Returns true if this particular GSG supports luminance textures.

bool get_supports_occlusion_query(void) const

Returns true if this GSG supports an occlusion query. If this is true, then begin_occlusion_query() and end_occlusion_query() may be called to bracket a sequence of draw_triangles() (or whatever) calls to measure pixels that pass the depth test.

bool get_supports_sampler_objects(void) const

Returns true if this particular GSG supports the use of sampler objects to record texture sampling parameters separately from the texture objects. This doesn’t really affect functionality, but if this is false, it may mean that using the same texture with different SamplerState objects will result in reduced performance.

bool get_supports_stencil(void) const

Returns true if this particular GSG supports stencil buffers at all.

bool get_supports_tessellation_shaders(void) const

Returns true if this particular GSG supports tesselation shaders.

bool get_supports_tex_non_pow2(void) const

Returns true if this GSG can handle non power of two sized textures.

bool get_supports_texture_combine(void) const

Returns true if this particular GSG can use the TextureStage::M_combine mode, which includes all of the texture blend modes specified by set_combine_rgb() and/or set_combine_alpha(). If this is false, you must limit yourself to using the simpler blend modes.

bool get_supports_texture_dot3(void) const

Returns true if this GSG can use the TextureStage::CM_dot3_rgb or CM_dot3_rgba combine modes.

bool get_supports_texture_saved_result(void) const

Returns true if this GSG can use the TextureStage::CS_last_saved_result source, which allows you to save the result of a TextureStage and re-use it for multiple inputs.

bool get_supports_timer_query(void) const

Returns true if this GSG supports a timer query.

bool get_supports_two_sided_stencil(void) const

Returns true if this particular GSG supports two sided stencil: different stencil settings for the front and back side of the same polygon.

Texture::QualityLevel get_texture_quality_override(void) const

Returns the global quality_level override specified by set_texture_quality_override.

This is mainly useful for the tinydisplay software renderer. See Texture::set_quality_level().

GraphicsThreadingModel const &get_threading_model(void) const

Returns the threading model that was used to create this GSG.

bool get_timer_queries_active(void) const

Returns true if timer queries are currently enabled on this GSG.

virtual bool has_extension(std::string const &extension) const
bool is_active(void) const

Returns the active flag associated with the GraphicsStateGuardian.

bool is_hardware(void) const

Returns true if this GSG appears to be hardware-accelerated, or false if it is known to be software only.

bool is_valid(void) const

Returns true if the GSG has been correctly initialized within a graphics context, false if there has been some problem or it hasn’t been initialized yet.

bool needs_reset(void) const

Returns true if the gsg is marked as needing a reset.

void release_all(void)

Releases all prepared objects.

int release_all_geoms(void)

Frees the resources for all geoms associated with this GSG.

int release_all_index_buffers(void)

Frees the resources for all index buffers associated with this GSG.

int release_all_samplers(void)

Frees the resources for all samplers associated with this GSG.

int release_all_shader_buffers(void)

Frees the resources for all index buffers associated with this GSG.

int release_all_textures(void)

Frees the resources for all textures associated with this GSG.

int release_all_vertex_buffers(void)

Frees the resources for all vertex buffers associated with this GSG.

virtual void restore_gamma(void)
void set_active(bool active)

Sets the active flag associated with the GraphicsStateGuardian. If the GraphicsStateGuardian is marked inactive, nothing is rendered. This is not normally turned off unless there is a problem with the rendering detected at a low level.

void set_coordinate_system(CoordinateSystem cs)
void set_flash_texture(Texture *tex)
virtual bool set_gamma(PN_stdfloat gamma)
void set_incomplete_render(bool incomplete_render)

Sets the incomplete_render flag. When this is true, the frame will be rendered even if some of the geometry or textures in the scene are not available (e.g. they have been temporarily paged out). When this is false, the frame will be held up while this data is reloaded.

Setting this true allows for a smoother frame rate, but occasionally parts of the frame will be invisible or missing (they will generally come in within a second or two). Setting this false guarantees that every frame will be complete, but may cause more chugs as things are loaded up at runtime.

You may want to set this false during loading screens, to guarantee that all of your assets are available by the time you take the loading screen down.

This flag may also be set individually on each DisplayRegion. It will be considered true for a given DisplayRegion only if it is true on both the GSG and on the DisplayRegion.

void set_loader(Loader *loader)

Sets the Loader object that will be used by this GSG to load textures when necessary, if get_incomplete_render() is true.

bool set_scene(SceneSetup *scene_setup)
void set_shader_generator(ShaderGenerator *shader_generator)

Sets the ShaderGenerator object that will be used by this GSG to generate shaders when necessary.

void set_shader_model(GraphicsStateGuardian::ShaderModel shader_model)

Sets the ShaderModel. This will override the auto- detected shader model during GSG reset. Useful for testing lower-end shaders.

void set_texture_quality_override(Texture::QualityLevel quality_level)

Specifies the global quality_level to be imposed for all Textures rendered by this GSG. This overrides the value set on individual textures via Texture::set_quality_level(). Set this to Texture::QL_default in order to allow the individual texture quality levels to be respected.

This is mainly useful for the tinydisplay software renderer. See Texture::set_quality_level().

void traverse_prepared_textures(GraphicsStateGuardian::TextureCallback *func, void *callback_arg)