Go to the source code of this file.
|
| enum | eGPULoadOp { GPU_LOADACTION_CLEAR = 0
, GPU_LOADACTION_LOAD
, GPU_LOADACTION_DONT_CARE
} |
| |
| enum | eGPUStoreOp { GPU_STOREACTION_STORE = 0
, GPU_STOREACTION_DONT_CARE
} |
| |
| enum | GPUAttachmentState { GPU_ATTACHMENT_IGNORE = 0
, GPU_ATTACHMENT_WRITE
, GPU_ATTACHMENT_READ
} |
| |
| enum | eGPUFrontFace { GPU_CLOCKWISE
, GPU_COUNTERCLOCKWISE
} |
| |
| enum class | blender::gpu::shader::Type {
blender::gpu::shader::float_t = 0
, blender::gpu::shader::float2_t
, blender::gpu::shader::float3_t
, blender::gpu::shader::float4_t
,
blender::gpu::shader::float3x3_t
, blender::gpu::shader::float4x4_t
, blender::gpu::shader::uint_t
, blender::gpu::shader::uint2_t
,
blender::gpu::shader::uint3_t
, blender::gpu::shader::uint4_t
, blender::gpu::shader::int_t
, blender::gpu::shader::int2_t
,
blender::gpu::shader::int3_t
, blender::gpu::shader::int4_t
, blender::gpu::shader::bool_t
, blender::gpu::shader::float3_10_10_10_2_t
,
blender::gpu::shader::uchar_t
, blender::gpu::shader::uchar2_t
, blender::gpu::shader::uchar3_t
, blender::gpu::shader::uchar4_t
,
blender::gpu::shader::char_t
, blender::gpu::shader::char2_t
, blender::gpu::shader::char3_t
, blender::gpu::shader::char4_t
,
blender::gpu::shader::ushort_t
, blender::gpu::shader::ushort2_t
, blender::gpu::shader::ushort3_t
, blender::gpu::shader::ushort4_t
,
blender::gpu::shader::short_t
, blender::gpu::shader::short2_t
, blender::gpu::shader::short3_t
, blender::gpu::shader::short4_t
} |
| |
◆ eGPUFrontFace
◆ eGPULoadOp
Describes the load operation of a frame-buffer attachment at the start of a render pass.
| Enumerator |
|---|
| GPU_LOADACTION_CLEAR | Clear the frame-buffer attachment using the clear value.
|
| GPU_LOADACTION_LOAD | Load the value from the attached texture. Cannot be used with memoryless attachments. Slower than GPU_LOADACTION_CLEAR or GPU_LOADACTION_DONT_CARE.
|
| GPU_LOADACTION_DONT_CARE | Do not care about the content of the attachment when the render pass starts. Useful if only the values being written are important. Faster than GPU_LOADACTION_CLEAR.
|
Definition at line 17 of file GPU_common_types.hh.
◆ eGPUStoreOp
Describes the store operation of a frame-buffer attachment at the end of a render pass.
| Enumerator |
|---|
| GPU_STOREACTION_STORE | Do not care about the content of the attachment when the render pass ends. Useful if only the values being written are important. Cannot be used with memoryless attachments.
|
| GPU_STOREACTION_DONT_CARE | The result of the rendering for this attachment will be discarded. No writes to the texture memory will be done which makes it faster than GPU_STOREACTION_STORE. IMPORTANT: The actual values of the attachment is to be considered undefined. Only to be used on transient attachment that are only used within the boundaries of a render pass (ex.: Unneeded depth buffer result).
|
Definition at line 39 of file GPU_common_types.hh.
◆ GPUAttachmentState
Describes the state of a frame-buffer attachment during a sub-pass.
NOTE: Until this is correctly implemented in all backend, reading and writing from the same attachment will not work. Although there is no case where it would currently be useful.
| Enumerator |
|---|
| GPU_ATTACHMENT_IGNORE | Attachment will not be written during rendering.
|
| GPU_ATTACHMENT_WRITE | Attachment will be written during render sub-pass. This also works with blending.
|
| GPU_ATTACHMENT_READ | Attachment is used as input in the fragment shader. Incompatible with depth on Metal.
|
Definition at line 63 of file GPU_common_types.hh.