Name ARB_shader_viewport_layer_array Name Strings GL_ARB_shader_viewport_layer_array Contact Graham Sellers, AMD (graham.sellers 'at' amd.com) Contributors Graham Sellers Notice Copyright (c) 2015 The Khronos Group Inc. Copyright terms at http://www.khronos.org/registry/speccopyright.html Specification Update Policy Khronos-approved extension specifications are updated in response to issues and bugs prioritized by the Khronos OpenGL Working Group. For extensions which have been promoted to a core Specification, fixes will first appear in the latest version of that core Specification, and will eventually be backported to the extension document. This policy is described in more detail at https://www.khronos.org/registry/OpenGL/docs/update_policy.php Status Complete. Approved by the ARB on June 26, 2015. Ratified by the Khronos Board of Promoters on August 7, 2015. Version Last Modified Date: February 12, 2019 Revision: 2 Number ARB Extension #185 Dependencies OpenGL 4.1 is required. The extension is written against the OpenGL 4.5 Specification, Core Profile, February 2, 2015 and the OpenGL Shading Language Specification, version 4.50.5. Overview The gl_ViewportIndex and gl_Layer built-in variables were introduced by the in OpenGL 4.1. These variables are available in un-extended OpenGL only to the geometry shader. When written in the geometry shader, they cause geometry to be directed to one of an array of several independent viewport rectangles or framebuffer attachment layers, respectively. In order to use any viewport or attachment layer other than zero, a geometry shader must be present. Geometry shaders introduce processing overhead and potential performance issues. The AMD_vertex_shader_layer and AMD_vertex_shader_viewport_index extensions allowed the gl_Layer and gl_ViewportIndex outputs to be written directly from the vertex shader with no geometry shader present. This extension effectively merges the AMD_vertex_shader_layer and AMD_vertex_shader_viewport_index extensions together and extends them further to allow both outputs to be written from tessellation evaluation shaders. New Procedures and Functions None. New Tokens None. Additions to Chapter 9 of the OpenGL 4.5 (Core) Specification (Framebuffers and Framebuffer Objects) Modify section 9.8, "Layered Framebuffers" as follows: Remove the bullet list explaining when layer number zero is used and replace the language with: The layer number of a fragment is zero if no vertex processing shader (vertex, tessellation evaluation or geometry) statically assigns a value to the built-in output gl_Layer. Replace the following paragraph with: Otherwise, the layer for each point, line, or triangle is taken from the gl_Layer output of the last active vertex processing stage to one of the vertices of the primitive. The vertex used is implementation-dependent. To get defined results, all vertices of each primitive should set the same value for gl_Layer. The layer number written by a vertex procesing shader has no effect if the framebuffer is not layered. Additions to Chapter 11 of the OpenGL 4.5 (Core) Specification (Programmable Vertex Processing) Add the following to Subsection 11.1.3.10, "Shader Outputs" The built-in output gl_Layer is used in layered rendering, and is discussed further in the next section. The built-in output gl_ViewportIndex is used to direct rendering to one of several viewports and is discussed further in the next section. Move Subsection 11.3.4.6, "Layer and Viewport Selection" into a new section 11.4, "Layer and Viewport Selection", renumber surrounding sections appropriately. Replace all instances of "Geometry shaders" in Section 11.4 with "Vertex, tessellation and geometry shaders". Additions to Chapter 13 of the OpenGL 4.5 (Core) Specification (Fixed-Function Vertex Post-Processing) Modify section 13.6.1 "Controlling the Viewport" as follows: Replace the paragraph beginning "Multiple viewports are available ..." with: Multiple viewports are available and are numbered zero through the value of MAX_VIEWPORTS minus one. If an active vertex processing shader is active and writes to gl_ViewportIndex, the viewport transformation uses the viewport corresponding to the value assigned to gl_ViewportIndex taken from an implementation-dependent primitive vertex. If the value of the viewport index is outside the range zero to the value of MAX_VIEWPORTS minus one, the results of the viewport transformation are undefined. If the active vertex, tessellation or geometry shaders (if present) do not write to gl_ViewportIndex, the viewport numbered zero is used by the viewport transformation. Modifications to the OpenGL Shading Language Specification, Version 4.50 Including the following line in a shader can be used to control the language features described in this extension: #extension GL_ARB_shader_viewport_layer_array : where is as specified in section 3.3. New preprocessor #defines are added to the OpenGL Shading Language: #define GL_ARB_shader_viewport_layer_array 1 Modify Section 7.1 (Built-In Language Variables), p. 122 Add to the list of vertex shader built-ins: out gl_PerVertex { ... int gl_ViewportIndex; int gl_Layer; }; Add to the list of tessellation evaluation shader built-ins: out gl_PerVertex { ... int gl_ViewportIndex; int gl_Layer; }; Modify the language introducing "gl_Layer" on p.127 as follows: The variable gl_Layer is available as an output variable in the vertex, tessellation evaluation, and geometry (VTG) languages and an input variable in the fragment language. In the VTG languages, it is used to select a specific layer (or face and layer of a cube map) of a multi-layer framebuffer attachment. The actual layer used will come from one of the vertices in the primitive being shaded. Which vertex the layer comes from is discussed in section 11.4 "Layer and Viewport Selection" of the OpenGL Specification. It might be undefined, so it is best to write the same layer value for all vertices of a primitive. If a shader statically assigns a value to gl_Layer, layered rendering mode is enabled. See section 11.3.4.5 "Geometry Shader Outputs" and section 9.4.9 "Layered Framebuffers" of the OpenGL Graphics System Specification for more information. If a shader statically assigns a value to gl_Layer, and there is an execution path through the shader that does not set gl_Layer, then the value of gl_Layer is undefined for executions of the shader that take that path. ... The input variable gl_Layer in the fragment language will have the same value that was written to the output variable gl_Layer in the VTG languages. If the final VTG stage does not dynamically assign a value to gl_Layer, the value of gl_Layer in the fragment stage will be undefined. If the final VTG stage makes no static assignment to gl_Layer, the input gl_Value in the fragment stage will be zero. Otherwise, the fragment stage will read the same value written by the final VTG stage, even if that value is out of range. If a fragment shader contains a static access to gl_Layer, it will count against the implementation defined limit for the maximum number of inputs to the fragment stage. Modify the language introducing "gl_ViewportIndex" on p.128 as follows: The variable gl_ViewportIndex is available as an output variable in the VTG languages and an input variable in the fragment language. In the geometry language, it provides the index of the viewport to which the next primitive emitted from the geometry shader should be drawn. In the vertex and tessellation evaluation languages, it provides the index of the viewport associated with the vertex being shaded. Primitives will undergo viewport transformation and scissor testing using the viewport transformation and scissor rectangle selected by the value of gl_ViewportIndex. The viewport index used will come from one of the vertices in the primitive being shaded. However, which vertex the viewport index comes from is implementation-dependent, so it is best to use the same viewport index for all vertices of the primitive. If the final VTG stage does not assign a value to gl_ViewportIndex, viewport transform and scissor rectangle zero will be used. If a shader statically assigns a value to gl_ViewportIndex and there is a path through the shader that does not assign a value to gl_ViewportIndex, the value of gl_ViewportIndex is undefined for executions of the shader that take that path. See section 11.4 "Layer and Viewport Selection" of the OpenGL Graphics System Specification for more information. The input variable gl_ViewportIndex in the fragment stage will have the same value that was written to the output variable gl_ViewportIndex in the final VTG stage. If the final VTG stage does not dynamically assign to gl_ViewportIndex, the value of gl_ViewportIndex in the fragment shader will be undefined. If the final VTG stage makes no static assignment to gl_ViewportIndex, the fragment stage will read zero. Otherwise, the fragment stage will read the same value written by the final VTG stage, even if that value is out of range. If a fragment shader contains a static access to gl_ViewportIndex, it will count against the implementation defined limit for the maximum number of inputs to the fragment stage. Additions to the AGL/GLX/WGL Specifications None. GLX Protocol None. Errors None. New State None. New Implementation Dependent State None. Issues 1) Do we want to allow writes to these built-in variables from tessellation shaders? RESOLVED: Yes, this is allowed in tessellation evaluation shaders. Note that it is not possible to write either variable in the tessellation control shader. 2) What happens if gl_ViewportIndex or gl_Layer is written in the vertex shader and a geometry shader is present? RESOLVED: The value written by the last vertex processing stage is used. If the last vertex processing stage (vertex, tessellation evaluation or geometry) does not statically assign to gl_ViewportIndex or gl_Layer, index or layer zero is assumed. 3) Are provoking vertex semantics honored for the purposes of writes to gl_ViewportIndex in the VS? RESOLVED: Yes, they are. Query VIEWPORT_INDEX_PROVOKING_VERTEX to determine the 'provokingness' of gl_ViewportIndex. In general, though, it's best practice to ensure that all vertices of a single primitive (including strips, fans and loops) have the same value for gl_ViewportIndex. 4) Do we want gl_ViewportIndexIn or gl_LayerIn inputs to shaders to allow automatic passing of these variables along the pipe? DISCUSSION: Rather not. This is just emualating things that the application could and should do itself. Pushing built-ins down the pipe that aren't consuming dedicated hardware resources just complicates things like counting rules and ultimately ends up pretty inefficient. 5) Are the gl_ViewportIndex and gl_Layer builtins part of the "out gl_PerVertex" block or not? RESOLVED: The initial version of this extension didn't specify and many implementations treated them as loose variables (or didn't care). However, the spec language makes it quite clear that there is an instance of these variables and a shader is expected to write the same value for each vertex. In practise, unless an application is trying to redeclare these variables it isn't really possible to tell where they are declared. Revision History Rev. Date Author Changes ---- -------- -------- ----------------------------------------- 1 03/19/2015 gsellers Initial version merging AMD_vertex_shader_viewport_index and AMD_vertex_shader_layer 2 02/12/2019 dkoch opengl/api/44: clarify that builtins should be part of the "out gl_PerVertex" block.