| /hardware/google/gfxstream/codegen/vulkan/vulkan-docs-next/appendices/ |
| D | VK_EXT_pipeline_library_group_handles.adoc | 24 When using pipeline libraries in ray tracing pipelines, a library might get
25 linked into different pipelines in an incremental way.
32 record buffers and the linkage of ray tracing pipelines.
41 bitwise identical between parent and child pipelines.
|
| D | VK_EXT_dynamic_rendering_unused_attachments.adoc | 26 and bound pipelines within those render pass instances to have an unused
28 It also allows pipelines to use different formats in a render pass as long
|
| D | VK_EXT_graphics_pipeline_library.adoc | 30 graphics pipelines, with the intent of allowing faster pipeline loading for
31 applications reusing the same shaders or state in multiple pipelines.
|
| D | VK_NVX_multiview_per_view_attributes.adoc | 28 <<pipelines-graphics-subsets-pre-rasterization,pre-rasterization shader
35 <<pipelines-graphics-subsets-pre-rasterization,pre-rasterization shader
44 A subpass can declare via a second creation flag whether all pipelines
|
| D | VK_NV_device_generated_commands_compute.adoc | 24 pipelines, setting push constants and launching compute dispatches.
|
| D | VK_EXT_pipeline_protected_access.adoc | 22 access to protected memory will be limited to the specific pipelines that
|
| D | VK_KHR_pipeline_library.adoc | 20 pipelines.
|
| D | VK_EXT_pipeline_creation_cache_control.adoc | 44 It is strongly advised that applications create pipelines sufficiently ahead
48 Proper usage of pipeline caches and/or derivative pipelines help mitigate
84 pipelines before returning.
|
| D | VK_FUCHSIA_external_memory.adoc | 26 processing workload for more complex memory manipulation pipelines.
|
| D | VK_EXT_load_store_op_none.adoc | 29 pipelines have been created.
|
| D | VK_EXT_pipeline_properties.adoc | 21 Vulkan SC requires offline compilation of pipelines.
32 Once the pipelines have been compiled by the offline pipeline cache
|
| D | vulkanscdeviations.adoc | 29 | <<pipelines,Pipelines>> a| include::{chapters}/pipelines.adoc[tag=scaddition]
52 | <<pipelines,Pipelines>> a| include::{chapters}/pipelines.adoc[tag=scdeviation]
81 | <<pipelines,Pipelines>> a| include::{chapters}/pipelines.adoc[tag=scremoved]
|
| D | VK_EXT_extended_dynamic_state3.adoc | 28 state to help applications further reduce the number of monolithic pipelines
|
| /hardware/google/gfxstream/codegen/vulkan/vulkan-docs-next/chapters/ |
| D | pipelines.adoc | 5 [[pipelines]] anchor
8 The following <<pipelines-block-diagram,figure>> shows a block diagram of
9 the Vulkan pipelines.
29 The first stage of the <<pipelines-graphics,graphics pipeline>>
87 The <<pipelines-compute,compute pipeline>> is a separate pipeline from the
93 organize the various operations of the pipelines.
97 [[pipelines-block-diagram]]
114 Any pipeline object state that is specified as <<pipelines-dynamic-state,
128 and graphics pipelines are each represented by sname:VkPipeline handles:
134 [[pipelines-compute]]
[all …]
|
| D | VK_AMD_pipeline_compiler_control.adoc | 5 // This section is included inside the Pipelines chapter (pipelines.adoc)
7 [[pipelines-compiler-control]]
|
| /hardware/google/gfxstream/codegen/vulkan/vulkan-docs-next/proposals/ |
| D | VK_EXT_pipeline_protected_access.adoc | 11 means for applications to distinguish between pipelines that do and do not
24 pipelines while the rest of the pipelines do not.
90 In practice, pipelines that actually access protected memory are scarce and
112 When this feature is enabled, pipelines can be flagged as not accessing
114 Such pipelines are not allowed to be bound to protected command buffers.
132 The `pipelineProtectedAccess` feature allows pipelines to be restricted to or
139 === RESOLVED: Should the `pipelineProtectedAccess` feature allow pipelines to opt into protected ac…
|
| D | VK_EXT_graphics_pipeline_library.adoc | 11 of pipelines, improving the performance of pipeline compilation for
18 The original promise of monolithic pipelines in Vulkan was to enable
33 telling them to reduce the number of pipelines.
88 choices about how and when these pipelines are compiled.
99 For link:{refpage}VK_KHR_ray_tracing_pipeline.html[VK_KHR_ray_tracing_pipeline], pipelines
103 applications to create partial pipelines (pipeline libraries) containing
106 Ray pipelines were relatively straightforward as only shaders are linked,
109 Graphics pipelines by comparison contain a lot of static state that needs
146 linking pipelines without `VK_PIPELINE_CREATE_LINK_TIME_OPTIMIZATION_BIT_EXT`
148 applications can link pipelines on demand while recording commands.
[all …]
|
| D | VK_EXT_pipeline_library_group_handles.adoc | 14 and handles are guaranteed to remain bitwise identical when linked into other pipelines.
20 …le model where pipeline libraries are compiled in isolation, and later, pipelines can be linked ra…
41 This is problematic since fully linked pipelines must include a ray generation shader stage. Forcin…
44 The assumption and premise of this extension is that drivers already compile pipelines into binary …
69 Note that invariance is not relevant for linked pipelines that are created without `VK_PIPELINE_CRE…
70 These pipelines cannot be linked into any other pipelines, so the question of invariance does not e…
120 The downside of this approach is that we cannot easily remove pipelines once linked.
|
| D | VK_EXT_dynamic_rendering_unused_attachments.adoc | 11 pipelines can be used with render pass instances, and may subsequently require
16 The `apiext:VK_KHR_dynamic_rendering` has several restrictions on what pipelines can be
75 extension that lifts these restrictions and allows pipelines to be bound to render
83 feature which allows apps to bind pipelines to render pass instances where attachment
|
| D | VK_EXT_shader_module_identifier.adoc | 29 Optimizing the translation might not be enough when faced with tens of thousand pipelines being com…
92 === Soft guarantees of successfully compiling pipelines
112 …is will easily contribute to pipeline cache bloat, since the exact same pipelines might end up in …
139 This allows applications to speculatively create link-time optimized pipelines from identifiers onl…
140 ray-tracing pipelines from libraries.
150 as there is no good excuse why an implementation should not be able to satisfy those pipelines.
|
| /hardware/google/gfxstream/codegen/vulkan/vulkan-docs-next/chapters/VK_NV_device_generated_commands/ |
| D | generatedcommands.adoc | 14 Groups>> for graphics pipelines.
18 * Retrieve device addresses of compute pipelines by
|
| /hardware/google/camera/devices/EmulatedCamera/hwl/ |
| D | EmulatedRequestProcessor.cpp | 77 const std::vector<EmulatedPipeline>& pipelines, in ProcessPipelineRequests() argument
86 if (request.pipeline_id >= pipelines.size()) { in ProcessPipelineRequests()
103 &request, pipelines, dynamic_stream_id_map, use_default_physical_camera); in ProcessPipelineRequests()
112 pipelines[request.pipeline_id].streams, request.pipeline_id, in ProcessPipelineRequests()
113 pipelines[request.pipeline_id].cb, /*override_width*/ 0, in ProcessPipelineRequests()
121 pipelines[request.pipeline_id].streams, request.pipeline_id, in ProcessPipelineRequests()
122 pipelines[request.pipeline_id].cb, request.input_width, in ProcessPipelineRequests()
143 .callback = pipelines[request.pipeline_id].cb, in ProcessPipelineRequests()
|
| D | EmulatedLogicalRequestState.cpp | 299 HwlPipelineRequest* request, const std::vector<EmulatedPipeline>& pipelines, in UpdateRequestForDynamicStreams() argument
308 if (pipeline_id >= pipelines.size()) { in UpdateRequestForDynamicStreams()
336 const auto& current_pipeline = pipelines[pipeline_id]; in UpdateRequestForDynamicStreams()
|
| /hardware/google/gfxstream/include/OpenGLESDispatch/ |
| D | gles31_only_functions.h | 14 X(void, glGenProgramPipelines, (GLsizei n, GLuint * pipelines), (n, pipelines)) \
15 X(void, glDeleteProgramPipelines, (GLsizei n, const GLuint * pipelines), (n, pipelines)) \
|
| /hardware/google/gfxstream/host/gl/OpenGLESDispatch/ |
| D | gles31_only_dispatch_logging_wrappers.impl | 19 void glGenProgramPipelines_dispatchLoggingWrapper(GLsizei n, GLuint * pipelines) {
20 DISPATCH_DEBUG_LOG("glGenProgramPipelines(n:%d, pipelines:%p)", n, pipelines);
21 GLDispatch::glGenProgramPipelines_underlying(n, pipelines);
24 void glDeleteProgramPipelines_dispatchLoggingWrapper(GLsizei n, const GLuint * pipelines) {
25 DISPATCH_DEBUG_LOG("glDeleteProgramPipelines(n:%d, pipelines:%p)", n, pipelines);
26 GLDispatch::glDeleteProgramPipelines_underlying(n, pipelines);
|