2022-02-11 09:07:11 +11:00
|
|
|
/* SPDX-License-Identifier: GPL-2.0-or-later
|
|
|
|
|
* Copyright 2021 Blender Foundation. All rights reserved. */
|
OpenSubDiv: add support for an OpenGL evaluator
This evaluator is used in order to evaluate subdivision at render time, allowing for
faster renders of meshes with a subdivision surface modifier placed at the last
position in the modifier list.
When evaluating the subsurf modifier, we detect whether we can delegate evaluation
to the draw code. If so, the subdivision is first evaluated on the GPU using our own
custom evaluator (only the coarse data needs to be initially sent to the GPU), then,
buffers for the final `MeshBufferCache` are filled on the GPU using a set of
compute shaders. However, some buffers are still filled on the CPU side, if doing so
on the GPU is impractical (e.g. the line adjacency buffer used for x-ray, whose
logic is hardly GPU compatible).
This is done at the mesh buffer extraction level so that the result can be readily used
in the various OpenGL engines, without having to write custom geometry or tesselation
shaders.
We use our own subdivision evaluation shaders, instead of OpenSubDiv's vanilla one, in
order to control the data layout, and interpolation. For example, we store vertex colors
as compressed 16-bit integers, while OpenSubDiv's default evaluator only work for float
types.
In order to still access the modified geometry on the CPU side, for use in modifiers
or transform operators, a dedicated wrapper type is added `MESH_WRAPPER_TYPE_SUBD`.
Subdivision will be lazily evaluated via `BKE_object_get_evaluated_mesh` which will
create such a wrapper if possible. If the final subdivision surface is not needed on
the CPU side, `BKE_object_get_evaluated_mesh_no_subsurf` should be used.
Enabling or disabling GPU subdivision can be done through the user preferences (under
Viewport -> Subdivision).
See patch description for benchmarks.
Reviewed By: campbellbarton, jbakker, fclem, brecht, #eevee_viewport
Differential Revision: https://developer.blender.org/D12406
2021-12-27 16:34:47 +01:00
|
|
|
|
|
|
|
|
#include "BKE_subdiv_modifier.h"
|
|
|
|
|
|
|
|
|
|
#include "MEM_guardedalloc.h"
|
|
|
|
|
|
|
|
|
|
#include "DNA_mesh_types.h"
|
|
|
|
|
#include "DNA_modifier_types.h"
|
|
|
|
|
#include "DNA_object_types.h"
|
|
|
|
|
#include "DNA_scene_types.h"
|
|
|
|
|
#include "DNA_userdef_types.h"
|
|
|
|
|
|
|
|
|
|
#include "BKE_modifier.h"
|
|
|
|
|
#include "BKE_subdiv.h"
|
|
|
|
|
|
|
|
|
|
#include "GPU_capabilities.h"
|
|
|
|
|
#include "GPU_context.h"
|
|
|
|
|
|
|
|
|
|
#include "opensubdiv_capi.h"
|
|
|
|
|
|
2022-06-13 15:21:05 +02:00
|
|
|
bool BKE_subsurf_modifier_runtime_init(SubsurfModifierData *smd, const bool use_render_params)
|
OpenSubDiv: add support for an OpenGL evaluator
This evaluator is used in order to evaluate subdivision at render time, allowing for
faster renders of meshes with a subdivision surface modifier placed at the last
position in the modifier list.
When evaluating the subsurf modifier, we detect whether we can delegate evaluation
to the draw code. If so, the subdivision is first evaluated on the GPU using our own
custom evaluator (only the coarse data needs to be initially sent to the GPU), then,
buffers for the final `MeshBufferCache` are filled on the GPU using a set of
compute shaders. However, some buffers are still filled on the CPU side, if doing so
on the GPU is impractical (e.g. the line adjacency buffer used for x-ray, whose
logic is hardly GPU compatible).
This is done at the mesh buffer extraction level so that the result can be readily used
in the various OpenGL engines, without having to write custom geometry or tesselation
shaders.
We use our own subdivision evaluation shaders, instead of OpenSubDiv's vanilla one, in
order to control the data layout, and interpolation. For example, we store vertex colors
as compressed 16-bit integers, while OpenSubDiv's default evaluator only work for float
types.
In order to still access the modified geometry on the CPU side, for use in modifiers
or transform operators, a dedicated wrapper type is added `MESH_WRAPPER_TYPE_SUBD`.
Subdivision will be lazily evaluated via `BKE_object_get_evaluated_mesh` which will
create such a wrapper if possible. If the final subdivision surface is not needed on
the CPU side, `BKE_object_get_evaluated_mesh_no_subsurf` should be used.
Enabling or disabling GPU subdivision can be done through the user preferences (under
Viewport -> Subdivision).
See patch description for benchmarks.
Reviewed By: campbellbarton, jbakker, fclem, brecht, #eevee_viewport
Differential Revision: https://developer.blender.org/D12406
2021-12-27 16:34:47 +01:00
|
|
|
{
|
|
|
|
|
const int requested_levels = (use_render_params) ? smd->renderLevels : smd->levels;
|
|
|
|
|
|
2022-06-13 15:21:05 +02:00
|
|
|
SubdivSettings settings;
|
|
|
|
|
settings.is_simple = (smd->subdivType == SUBSURF_TYPE_SIMPLE);
|
|
|
|
|
settings.is_adaptive = !(smd->flags & eSubsurfModifierFlag_UseRecursiveSubdivision);
|
|
|
|
|
settings.level = settings.is_simple ? 1 :
|
|
|
|
|
(settings.is_adaptive ? smd->quality : requested_levels);
|
|
|
|
|
settings.use_creases = (smd->flags & eSubsurfModifierFlag_UseCrease);
|
|
|
|
|
settings.vtx_boundary_interpolation = BKE_subdiv_vtx_boundary_interpolation_from_subsurf(
|
OpenSubDiv: add support for an OpenGL evaluator
This evaluator is used in order to evaluate subdivision at render time, allowing for
faster renders of meshes with a subdivision surface modifier placed at the last
position in the modifier list.
When evaluating the subsurf modifier, we detect whether we can delegate evaluation
to the draw code. If so, the subdivision is first evaluated on the GPU using our own
custom evaluator (only the coarse data needs to be initially sent to the GPU), then,
buffers for the final `MeshBufferCache` are filled on the GPU using a set of
compute shaders. However, some buffers are still filled on the CPU side, if doing so
on the GPU is impractical (e.g. the line adjacency buffer used for x-ray, whose
logic is hardly GPU compatible).
This is done at the mesh buffer extraction level so that the result can be readily used
in the various OpenGL engines, without having to write custom geometry or tesselation
shaders.
We use our own subdivision evaluation shaders, instead of OpenSubDiv's vanilla one, in
order to control the data layout, and interpolation. For example, we store vertex colors
as compressed 16-bit integers, while OpenSubDiv's default evaluator only work for float
types.
In order to still access the modified geometry on the CPU side, for use in modifiers
or transform operators, a dedicated wrapper type is added `MESH_WRAPPER_TYPE_SUBD`.
Subdivision will be lazily evaluated via `BKE_object_get_evaluated_mesh` which will
create such a wrapper if possible. If the final subdivision surface is not needed on
the CPU side, `BKE_object_get_evaluated_mesh_no_subsurf` should be used.
Enabling or disabling GPU subdivision can be done through the user preferences (under
Viewport -> Subdivision).
See patch description for benchmarks.
Reviewed By: campbellbarton, jbakker, fclem, brecht, #eevee_viewport
Differential Revision: https://developer.blender.org/D12406
2021-12-27 16:34:47 +01:00
|
|
|
smd->boundary_smooth);
|
2022-06-13 15:21:05 +02:00
|
|
|
settings.fvar_linear_interpolation = BKE_subdiv_fvar_interpolation_from_uv_smooth(
|
OpenSubDiv: add support for an OpenGL evaluator
This evaluator is used in order to evaluate subdivision at render time, allowing for
faster renders of meshes with a subdivision surface modifier placed at the last
position in the modifier list.
When evaluating the subsurf modifier, we detect whether we can delegate evaluation
to the draw code. If so, the subdivision is first evaluated on the GPU using our own
custom evaluator (only the coarse data needs to be initially sent to the GPU), then,
buffers for the final `MeshBufferCache` are filled on the GPU using a set of
compute shaders. However, some buffers are still filled on the CPU side, if doing so
on the GPU is impractical (e.g. the line adjacency buffer used for x-ray, whose
logic is hardly GPU compatible).
This is done at the mesh buffer extraction level so that the result can be readily used
in the various OpenGL engines, without having to write custom geometry or tesselation
shaders.
We use our own subdivision evaluation shaders, instead of OpenSubDiv's vanilla one, in
order to control the data layout, and interpolation. For example, we store vertex colors
as compressed 16-bit integers, while OpenSubDiv's default evaluator only work for float
types.
In order to still access the modified geometry on the CPU side, for use in modifiers
or transform operators, a dedicated wrapper type is added `MESH_WRAPPER_TYPE_SUBD`.
Subdivision will be lazily evaluated via `BKE_object_get_evaluated_mesh` which will
create such a wrapper if possible. If the final subdivision surface is not needed on
the CPU side, `BKE_object_get_evaluated_mesh_no_subsurf` should be used.
Enabling or disabling GPU subdivision can be done through the user preferences (under
Viewport -> Subdivision).
See patch description for benchmarks.
Reviewed By: campbellbarton, jbakker, fclem, brecht, #eevee_viewport
Differential Revision: https://developer.blender.org/D12406
2021-12-27 16:34:47 +01:00
|
|
|
smd->uv_smooth);
|
2022-06-13 15:21:05 +02:00
|
|
|
|
|
|
|
|
SubsurfRuntimeData *runtime_data = (SubsurfRuntimeData *)smd->modifier.runtime;
|
|
|
|
|
if (settings.level == 0) {
|
|
|
|
|
/* Modifier is effectively disabled, but still update settings if runtime data
|
|
|
|
|
* was already allocated. */
|
|
|
|
|
if (runtime_data) {
|
|
|
|
|
runtime_data->settings = settings;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
return false;
|
|
|
|
|
}
|
2022-06-22 18:58:25 -05:00
|
|
|
|
|
|
|
|
/* Allocate runtime data if it did not exist yet. */
|
|
|
|
|
if (runtime_data == NULL) {
|
|
|
|
|
runtime_data = MEM_callocN(sizeof(*runtime_data), "subsurf runtime");
|
|
|
|
|
smd->modifier.runtime = runtime_data;
|
2022-06-13 15:21:05 +02:00
|
|
|
}
|
2022-06-22 18:58:25 -05:00
|
|
|
runtime_data->settings = settings;
|
|
|
|
|
return true;
|
OpenSubDiv: add support for an OpenGL evaluator
This evaluator is used in order to evaluate subdivision at render time, allowing for
faster renders of meshes with a subdivision surface modifier placed at the last
position in the modifier list.
When evaluating the subsurf modifier, we detect whether we can delegate evaluation
to the draw code. If so, the subdivision is first evaluated on the GPU using our own
custom evaluator (only the coarse data needs to be initially sent to the GPU), then,
buffers for the final `MeshBufferCache` are filled on the GPU using a set of
compute shaders. However, some buffers are still filled on the CPU side, if doing so
on the GPU is impractical (e.g. the line adjacency buffer used for x-ray, whose
logic is hardly GPU compatible).
This is done at the mesh buffer extraction level so that the result can be readily used
in the various OpenGL engines, without having to write custom geometry or tesselation
shaders.
We use our own subdivision evaluation shaders, instead of OpenSubDiv's vanilla one, in
order to control the data layout, and interpolation. For example, we store vertex colors
as compressed 16-bit integers, while OpenSubDiv's default evaluator only work for float
types.
In order to still access the modified geometry on the CPU side, for use in modifiers
or transform operators, a dedicated wrapper type is added `MESH_WRAPPER_TYPE_SUBD`.
Subdivision will be lazily evaluated via `BKE_object_get_evaluated_mesh` which will
create such a wrapper if possible. If the final subdivision surface is not needed on
the CPU side, `BKE_object_get_evaluated_mesh_no_subsurf` should be used.
Enabling or disabling GPU subdivision can be done through the user preferences (under
Viewport -> Subdivision).
See patch description for benchmarks.
Reviewed By: campbellbarton, jbakker, fclem, brecht, #eevee_viewport
Differential Revision: https://developer.blender.org/D12406
2021-12-27 16:34:47 +01:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
static ModifierData *modifier_get_last_enabled_for_mode(const Scene *scene,
|
|
|
|
|
const Object *ob,
|
|
|
|
|
int required_mode)
|
|
|
|
|
{
|
|
|
|
|
ModifierData *md = ob->modifiers.last;
|
|
|
|
|
|
|
|
|
|
while (md) {
|
|
|
|
|
if (BKE_modifier_is_enabled(scene, md, required_mode)) {
|
|
|
|
|
break;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
md = md->prev;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
return md;
|
|
|
|
|
}
|
|
|
|
|
|
2022-02-25 03:53:49 +01:00
|
|
|
bool BKE_subsurf_modifier_use_custom_loop_normals(const SubsurfModifierData *smd, const Mesh *mesh)
|
OpenSubDiv: add support for an OpenGL evaluator
This evaluator is used in order to evaluate subdivision at render time, allowing for
faster renders of meshes with a subdivision surface modifier placed at the last
position in the modifier list.
When evaluating the subsurf modifier, we detect whether we can delegate evaluation
to the draw code. If so, the subdivision is first evaluated on the GPU using our own
custom evaluator (only the coarse data needs to be initially sent to the GPU), then,
buffers for the final `MeshBufferCache` are filled on the GPU using a set of
compute shaders. However, some buffers are still filled on the CPU side, if doing so
on the GPU is impractical (e.g. the line adjacency buffer used for x-ray, whose
logic is hardly GPU compatible).
This is done at the mesh buffer extraction level so that the result can be readily used
in the various OpenGL engines, without having to write custom geometry or tesselation
shaders.
We use our own subdivision evaluation shaders, instead of OpenSubDiv's vanilla one, in
order to control the data layout, and interpolation. For example, we store vertex colors
as compressed 16-bit integers, while OpenSubDiv's default evaluator only work for float
types.
In order to still access the modified geometry on the CPU side, for use in modifiers
or transform operators, a dedicated wrapper type is added `MESH_WRAPPER_TYPE_SUBD`.
Subdivision will be lazily evaluated via `BKE_object_get_evaluated_mesh` which will
create such a wrapper if possible. If the final subdivision surface is not needed on
the CPU side, `BKE_object_get_evaluated_mesh_no_subsurf` should be used.
Enabling or disabling GPU subdivision can be done through the user preferences (under
Viewport -> Subdivision).
See patch description for benchmarks.
Reviewed By: campbellbarton, jbakker, fclem, brecht, #eevee_viewport
Differential Revision: https://developer.blender.org/D12406
2021-12-27 16:34:47 +01:00
|
|
|
{
|
2022-02-25 03:53:49 +01:00
|
|
|
return (smd->flags & eSubsurfModifierFlag_UseCustomNormals) && (mesh->flag & ME_AUTOSMOOTH) &&
|
|
|
|
|
CustomData_has_layer(&mesh->ldata, CD_CUSTOMLOOPNORMAL);
|
|
|
|
|
}
|
OpenSubDiv: add support for an OpenGL evaluator
This evaluator is used in order to evaluate subdivision at render time, allowing for
faster renders of meshes with a subdivision surface modifier placed at the last
position in the modifier list.
When evaluating the subsurf modifier, we detect whether we can delegate evaluation
to the draw code. If so, the subdivision is first evaluated on the GPU using our own
custom evaluator (only the coarse data needs to be initially sent to the GPU), then,
buffers for the final `MeshBufferCache` are filled on the GPU using a set of
compute shaders. However, some buffers are still filled on the CPU side, if doing so
on the GPU is impractical (e.g. the line adjacency buffer used for x-ray, whose
logic is hardly GPU compatible).
This is done at the mesh buffer extraction level so that the result can be readily used
in the various OpenGL engines, without having to write custom geometry or tesselation
shaders.
We use our own subdivision evaluation shaders, instead of OpenSubDiv's vanilla one, in
order to control the data layout, and interpolation. For example, we store vertex colors
as compressed 16-bit integers, while OpenSubDiv's default evaluator only work for float
types.
In order to still access the modified geometry on the CPU side, for use in modifiers
or transform operators, a dedicated wrapper type is added `MESH_WRAPPER_TYPE_SUBD`.
Subdivision will be lazily evaluated via `BKE_object_get_evaluated_mesh` which will
create such a wrapper if possible. If the final subdivision surface is not needed on
the CPU side, `BKE_object_get_evaluated_mesh_no_subsurf` should be used.
Enabling or disabling GPU subdivision can be done through the user preferences (under
Viewport -> Subdivision).
See patch description for benchmarks.
Reviewed By: campbellbarton, jbakker, fclem, brecht, #eevee_viewport
Differential Revision: https://developer.blender.org/D12406
2021-12-27 16:34:47 +01:00
|
|
|
|
2022-02-25 22:01:17 +01:00
|
|
|
static bool subsurf_modifier_use_autosmooth_or_split_normals(const SubsurfModifierData *smd,
|
|
|
|
|
const Mesh *mesh)
|
2022-02-25 03:53:49 +01:00
|
|
|
{
|
|
|
|
|
return (mesh->flag & ME_AUTOSMOOTH) || BKE_subsurf_modifier_use_custom_loop_normals(smd, mesh);
|
|
|
|
|
}
|
OpenSubDiv: add support for an OpenGL evaluator
This evaluator is used in order to evaluate subdivision at render time, allowing for
faster renders of meshes with a subdivision surface modifier placed at the last
position in the modifier list.
When evaluating the subsurf modifier, we detect whether we can delegate evaluation
to the draw code. If so, the subdivision is first evaluated on the GPU using our own
custom evaluator (only the coarse data needs to be initially sent to the GPU), then,
buffers for the final `MeshBufferCache` are filled on the GPU using a set of
compute shaders. However, some buffers are still filled on the CPU side, if doing so
on the GPU is impractical (e.g. the line adjacency buffer used for x-ray, whose
logic is hardly GPU compatible).
This is done at the mesh buffer extraction level so that the result can be readily used
in the various OpenGL engines, without having to write custom geometry or tesselation
shaders.
We use our own subdivision evaluation shaders, instead of OpenSubDiv's vanilla one, in
order to control the data layout, and interpolation. For example, we store vertex colors
as compressed 16-bit integers, while OpenSubDiv's default evaluator only work for float
types.
In order to still access the modified geometry on the CPU side, for use in modifiers
or transform operators, a dedicated wrapper type is added `MESH_WRAPPER_TYPE_SUBD`.
Subdivision will be lazily evaluated via `BKE_object_get_evaluated_mesh` which will
create such a wrapper if possible. If the final subdivision surface is not needed on
the CPU side, `BKE_object_get_evaluated_mesh_no_subsurf` should be used.
Enabling or disabling GPU subdivision can be done through the user preferences (under
Viewport -> Subdivision).
See patch description for benchmarks.
Reviewed By: campbellbarton, jbakker, fclem, brecht, #eevee_viewport
Differential Revision: https://developer.blender.org/D12406
2021-12-27 16:34:47 +01:00
|
|
|
|
2022-02-25 03:53:49 +01:00
|
|
|
static bool is_subdivision_evaluation_possible_on_gpu(void)
|
|
|
|
|
{
|
OpenSubDiv: add support for an OpenGL evaluator
This evaluator is used in order to evaluate subdivision at render time, allowing for
faster renders of meshes with a subdivision surface modifier placed at the last
position in the modifier list.
When evaluating the subsurf modifier, we detect whether we can delegate evaluation
to the draw code. If so, the subdivision is first evaluated on the GPU using our own
custom evaluator (only the coarse data needs to be initially sent to the GPU), then,
buffers for the final `MeshBufferCache` are filled on the GPU using a set of
compute shaders. However, some buffers are still filled on the CPU side, if doing so
on the GPU is impractical (e.g. the line adjacency buffer used for x-ray, whose
logic is hardly GPU compatible).
This is done at the mesh buffer extraction level so that the result can be readily used
in the various OpenGL engines, without having to write custom geometry or tesselation
shaders.
We use our own subdivision evaluation shaders, instead of OpenSubDiv's vanilla one, in
order to control the data layout, and interpolation. For example, we store vertex colors
as compressed 16-bit integers, while OpenSubDiv's default evaluator only work for float
types.
In order to still access the modified geometry on the CPU side, for use in modifiers
or transform operators, a dedicated wrapper type is added `MESH_WRAPPER_TYPE_SUBD`.
Subdivision will be lazily evaluated via `BKE_object_get_evaluated_mesh` which will
create such a wrapper if possible. If the final subdivision surface is not needed on
the CPU side, `BKE_object_get_evaluated_mesh_no_subsurf` should be used.
Enabling or disabling GPU subdivision can be done through the user preferences (under
Viewport -> Subdivision).
See patch description for benchmarks.
Reviewed By: campbellbarton, jbakker, fclem, brecht, #eevee_viewport
Differential Revision: https://developer.blender.org/D12406
2021-12-27 16:34:47 +01:00
|
|
|
/* Only OpenGL is supported for OpenSubdiv evaluation for now. */
|
|
|
|
|
if (GPU_backend_get_type() != GPU_BACKEND_OPENGL) {
|
|
|
|
|
return false;
|
|
|
|
|
}
|
|
|
|
|
|
2022-01-24 18:46:30 +01:00
|
|
|
if (!(GPU_compute_shader_support() && GPU_shader_storage_buffer_objects_support())) {
|
OpenSubDiv: add support for an OpenGL evaluator
This evaluator is used in order to evaluate subdivision at render time, allowing for
faster renders of meshes with a subdivision surface modifier placed at the last
position in the modifier list.
When evaluating the subsurf modifier, we detect whether we can delegate evaluation
to the draw code. If so, the subdivision is first evaluated on the GPU using our own
custom evaluator (only the coarse data needs to be initially sent to the GPU), then,
buffers for the final `MeshBufferCache` are filled on the GPU using a set of
compute shaders. However, some buffers are still filled on the CPU side, if doing so
on the GPU is impractical (e.g. the line adjacency buffer used for x-ray, whose
logic is hardly GPU compatible).
This is done at the mesh buffer extraction level so that the result can be readily used
in the various OpenGL engines, without having to write custom geometry or tesselation
shaders.
We use our own subdivision evaluation shaders, instead of OpenSubDiv's vanilla one, in
order to control the data layout, and interpolation. For example, we store vertex colors
as compressed 16-bit integers, while OpenSubDiv's default evaluator only work for float
types.
In order to still access the modified geometry on the CPU side, for use in modifiers
or transform operators, a dedicated wrapper type is added `MESH_WRAPPER_TYPE_SUBD`.
Subdivision will be lazily evaluated via `BKE_object_get_evaluated_mesh` which will
create such a wrapper if possible. If the final subdivision surface is not needed on
the CPU side, `BKE_object_get_evaluated_mesh_no_subsurf` should be used.
Enabling or disabling GPU subdivision can be done through the user preferences (under
Viewport -> Subdivision).
See patch description for benchmarks.
Reviewed By: campbellbarton, jbakker, fclem, brecht, #eevee_viewport
Differential Revision: https://developer.blender.org/D12406
2021-12-27 16:34:47 +01:00
|
|
|
return false;
|
|
|
|
|
}
|
|
|
|
|
|
2022-05-11 15:19:10 +02:00
|
|
|
if (GPU_max_compute_shader_storage_blocks() < MAX_GPU_SUBDIV_SSBOS) {
|
2022-03-15 16:07:32 +01:00
|
|
|
return false;
|
|
|
|
|
}
|
|
|
|
|
|
OpenSubDiv: add support for an OpenGL evaluator
This evaluator is used in order to evaluate subdivision at render time, allowing for
faster renders of meshes with a subdivision surface modifier placed at the last
position in the modifier list.
When evaluating the subsurf modifier, we detect whether we can delegate evaluation
to the draw code. If so, the subdivision is first evaluated on the GPU using our own
custom evaluator (only the coarse data needs to be initially sent to the GPU), then,
buffers for the final `MeshBufferCache` are filled on the GPU using a set of
compute shaders. However, some buffers are still filled on the CPU side, if doing so
on the GPU is impractical (e.g. the line adjacency buffer used for x-ray, whose
logic is hardly GPU compatible).
This is done at the mesh buffer extraction level so that the result can be readily used
in the various OpenGL engines, without having to write custom geometry or tesselation
shaders.
We use our own subdivision evaluation shaders, instead of OpenSubDiv's vanilla one, in
order to control the data layout, and interpolation. For example, we store vertex colors
as compressed 16-bit integers, while OpenSubDiv's default evaluator only work for float
types.
In order to still access the modified geometry on the CPU side, for use in modifiers
or transform operators, a dedicated wrapper type is added `MESH_WRAPPER_TYPE_SUBD`.
Subdivision will be lazily evaluated via `BKE_object_get_evaluated_mesh` which will
create such a wrapper if possible. If the final subdivision surface is not needed on
the CPU side, `BKE_object_get_evaluated_mesh_no_subsurf` should be used.
Enabling or disabling GPU subdivision can be done through the user preferences (under
Viewport -> Subdivision).
See patch description for benchmarks.
Reviewed By: campbellbarton, jbakker, fclem, brecht, #eevee_viewport
Differential Revision: https://developer.blender.org/D12406
2021-12-27 16:34:47 +01:00
|
|
|
const int available_evaluators = openSubdiv_getAvailableEvaluators();
|
|
|
|
|
if ((available_evaluators & OPENSUBDIV_EVALUATOR_GLSL_COMPUTE) == 0) {
|
|
|
|
|
return false;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
return true;
|
|
|
|
|
}
|
|
|
|
|
|
2022-02-25 03:53:49 +01:00
|
|
|
bool BKE_subsurf_modifier_force_disable_gpu_evaluation_for_mesh(const SubsurfModifierData *smd,
|
|
|
|
|
const Mesh *mesh)
|
|
|
|
|
{
|
|
|
|
|
if ((U.gpu_flag & USER_GPU_FLAG_SUBDIVISION_EVALUATION) == 0) {
|
2022-03-01 09:09:24 +11:00
|
|
|
/* GPU subdivision is explicitly disabled, so we don't force it. */
|
2022-02-25 03:53:49 +01:00
|
|
|
return false;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
if (!is_subdivision_evaluation_possible_on_gpu()) {
|
|
|
|
|
/* The GPU type is not compatible with the subdivision. */
|
|
|
|
|
return false;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
return subsurf_modifier_use_autosmooth_or_split_normals(smd, mesh);
|
|
|
|
|
}
|
|
|
|
|
|
2022-05-17 15:31:37 +02:00
|
|
|
bool BKE_subsurf_modifier_can_do_gpu_subdiv(const Scene *scene,
|
|
|
|
|
const Object *ob,
|
|
|
|
|
const Mesh *mesh,
|
|
|
|
|
const SubsurfModifierData *smd,
|
|
|
|
|
int required_mode)
|
2022-02-25 03:53:49 +01:00
|
|
|
{
|
|
|
|
|
if ((U.gpu_flag & USER_GPU_FLAG_SUBDIVISION_EVALUATION) == 0) {
|
|
|
|
|
return false;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* Deactivate GPU subdivision if autosmooth or custom split normals are used as those are
|
|
|
|
|
* complicated to support on GPU, and should really be separate workflows. */
|
|
|
|
|
if (subsurf_modifier_use_autosmooth_or_split_normals(smd, mesh)) {
|
|
|
|
|
return false;
|
|
|
|
|
}
|
|
|
|
|
|
2022-05-17 15:31:37 +02:00
|
|
|
ModifierData *md = modifier_get_last_enabled_for_mode(scene, ob, required_mode);
|
|
|
|
|
if (md != (const ModifierData *)smd) {
|
|
|
|
|
return false;
|
2022-02-25 03:53:49 +01:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
return is_subdivision_evaluation_possible_on_gpu();
|
|
|
|
|
}
|
|
|
|
|
|
2022-05-17 15:31:37 +02:00
|
|
|
bool BKE_subsurf_modifier_has_gpu_subdiv(const Mesh *mesh)
|
OpenSubDiv: add support for an OpenGL evaluator
This evaluator is used in order to evaluate subdivision at render time, allowing for
faster renders of meshes with a subdivision surface modifier placed at the last
position in the modifier list.
When evaluating the subsurf modifier, we detect whether we can delegate evaluation
to the draw code. If so, the subdivision is first evaluated on the GPU using our own
custom evaluator (only the coarse data needs to be initially sent to the GPU), then,
buffers for the final `MeshBufferCache` are filled on the GPU using a set of
compute shaders. However, some buffers are still filled on the CPU side, if doing so
on the GPU is impractical (e.g. the line adjacency buffer used for x-ray, whose
logic is hardly GPU compatible).
This is done at the mesh buffer extraction level so that the result can be readily used
in the various OpenGL engines, without having to write custom geometry or tesselation
shaders.
We use our own subdivision evaluation shaders, instead of OpenSubDiv's vanilla one, in
order to control the data layout, and interpolation. For example, we store vertex colors
as compressed 16-bit integers, while OpenSubDiv's default evaluator only work for float
types.
In order to still access the modified geometry on the CPU side, for use in modifiers
or transform operators, a dedicated wrapper type is added `MESH_WRAPPER_TYPE_SUBD`.
Subdivision will be lazily evaluated via `BKE_object_get_evaluated_mesh` which will
create such a wrapper if possible. If the final subdivision surface is not needed on
the CPU side, `BKE_object_get_evaluated_mesh_no_subsurf` should be used.
Enabling or disabling GPU subdivision can be done through the user preferences (under
Viewport -> Subdivision).
See patch description for benchmarks.
Reviewed By: campbellbarton, jbakker, fclem, brecht, #eevee_viewport
Differential Revision: https://developer.blender.org/D12406
2021-12-27 16:34:47 +01:00
|
|
|
{
|
2022-06-13 15:21:05 +02:00
|
|
|
SubsurfRuntimeData *runtime_data = mesh->runtime.subsurf_runtime_data;
|
|
|
|
|
return runtime_data && runtime_data->has_gpu_subdiv;
|
OpenSubDiv: add support for an OpenGL evaluator
This evaluator is used in order to evaluate subdivision at render time, allowing for
faster renders of meshes with a subdivision surface modifier placed at the last
position in the modifier list.
When evaluating the subsurf modifier, we detect whether we can delegate evaluation
to the draw code. If so, the subdivision is first evaluated on the GPU using our own
custom evaluator (only the coarse data needs to be initially sent to the GPU), then,
buffers for the final `MeshBufferCache` are filled on the GPU using a set of
compute shaders. However, some buffers are still filled on the CPU side, if doing so
on the GPU is impractical (e.g. the line adjacency buffer used for x-ray, whose
logic is hardly GPU compatible).
This is done at the mesh buffer extraction level so that the result can be readily used
in the various OpenGL engines, without having to write custom geometry or tesselation
shaders.
We use our own subdivision evaluation shaders, instead of OpenSubDiv's vanilla one, in
order to control the data layout, and interpolation. For example, we store vertex colors
as compressed 16-bit integers, while OpenSubDiv's default evaluator only work for float
types.
In order to still access the modified geometry on the CPU side, for use in modifiers
or transform operators, a dedicated wrapper type is added `MESH_WRAPPER_TYPE_SUBD`.
Subdivision will be lazily evaluated via `BKE_object_get_evaluated_mesh` which will
create such a wrapper if possible. If the final subdivision surface is not needed on
the CPU side, `BKE_object_get_evaluated_mesh_no_subsurf` should be used.
Enabling or disabling GPU subdivision can be done through the user preferences (under
Viewport -> Subdivision).
See patch description for benchmarks.
Reviewed By: campbellbarton, jbakker, fclem, brecht, #eevee_viewport
Differential Revision: https://developer.blender.org/D12406
2021-12-27 16:34:47 +01:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
void (*BKE_subsurf_modifier_free_gpu_cache_cb)(Subdiv *subdiv) = NULL;
|
|
|
|
|
|
2022-06-13 15:21:05 +02:00
|
|
|
Subdiv *BKE_subsurf_modifier_subdiv_descriptor_ensure(SubsurfRuntimeData *runtime_data,
|
OpenSubDiv: add support for an OpenGL evaluator
This evaluator is used in order to evaluate subdivision at render time, allowing for
faster renders of meshes with a subdivision surface modifier placed at the last
position in the modifier list.
When evaluating the subsurf modifier, we detect whether we can delegate evaluation
to the draw code. If so, the subdivision is first evaluated on the GPU using our own
custom evaluator (only the coarse data needs to be initially sent to the GPU), then,
buffers for the final `MeshBufferCache` are filled on the GPU using a set of
compute shaders. However, some buffers are still filled on the CPU side, if doing so
on the GPU is impractical (e.g. the line adjacency buffer used for x-ray, whose
logic is hardly GPU compatible).
This is done at the mesh buffer extraction level so that the result can be readily used
in the various OpenGL engines, without having to write custom geometry or tesselation
shaders.
We use our own subdivision evaluation shaders, instead of OpenSubDiv's vanilla one, in
order to control the data layout, and interpolation. For example, we store vertex colors
as compressed 16-bit integers, while OpenSubDiv's default evaluator only work for float
types.
In order to still access the modified geometry on the CPU side, for use in modifiers
or transform operators, a dedicated wrapper type is added `MESH_WRAPPER_TYPE_SUBD`.
Subdivision will be lazily evaluated via `BKE_object_get_evaluated_mesh` which will
create such a wrapper if possible. If the final subdivision surface is not needed on
the CPU side, `BKE_object_get_evaluated_mesh_no_subsurf` should be used.
Enabling or disabling GPU subdivision can be done through the user preferences (under
Viewport -> Subdivision).
See patch description for benchmarks.
Reviewed By: campbellbarton, jbakker, fclem, brecht, #eevee_viewport
Differential Revision: https://developer.blender.org/D12406
2021-12-27 16:34:47 +01:00
|
|
|
const Mesh *mesh,
|
|
|
|
|
const bool for_draw_code)
|
|
|
|
|
{
|
|
|
|
|
if (runtime_data->subdiv && runtime_data->set_by_draw_code != for_draw_code) {
|
|
|
|
|
BKE_subdiv_free(runtime_data->subdiv);
|
|
|
|
|
runtime_data->subdiv = NULL;
|
|
|
|
|
}
|
2022-06-13 15:21:05 +02:00
|
|
|
Subdiv *subdiv = BKE_subdiv_update_from_mesh(
|
|
|
|
|
runtime_data->subdiv, &runtime_data->settings, mesh);
|
OpenSubDiv: add support for an OpenGL evaluator
This evaluator is used in order to evaluate subdivision at render time, allowing for
faster renders of meshes with a subdivision surface modifier placed at the last
position in the modifier list.
When evaluating the subsurf modifier, we detect whether we can delegate evaluation
to the draw code. If so, the subdivision is first evaluated on the GPU using our own
custom evaluator (only the coarse data needs to be initially sent to the GPU), then,
buffers for the final `MeshBufferCache` are filled on the GPU using a set of
compute shaders. However, some buffers are still filled on the CPU side, if doing so
on the GPU is impractical (e.g. the line adjacency buffer used for x-ray, whose
logic is hardly GPU compatible).
This is done at the mesh buffer extraction level so that the result can be readily used
in the various OpenGL engines, without having to write custom geometry or tesselation
shaders.
We use our own subdivision evaluation shaders, instead of OpenSubDiv's vanilla one, in
order to control the data layout, and interpolation. For example, we store vertex colors
as compressed 16-bit integers, while OpenSubDiv's default evaluator only work for float
types.
In order to still access the modified geometry on the CPU side, for use in modifiers
or transform operators, a dedicated wrapper type is added `MESH_WRAPPER_TYPE_SUBD`.
Subdivision will be lazily evaluated via `BKE_object_get_evaluated_mesh` which will
create such a wrapper if possible. If the final subdivision surface is not needed on
the CPU side, `BKE_object_get_evaluated_mesh_no_subsurf` should be used.
Enabling or disabling GPU subdivision can be done through the user preferences (under
Viewport -> Subdivision).
See patch description for benchmarks.
Reviewed By: campbellbarton, jbakker, fclem, brecht, #eevee_viewport
Differential Revision: https://developer.blender.org/D12406
2021-12-27 16:34:47 +01:00
|
|
|
runtime_data->subdiv = subdiv;
|
|
|
|
|
runtime_data->set_by_draw_code = for_draw_code;
|
|
|
|
|
return subdiv;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
int BKE_subsurf_modifier_eval_required_mode(bool is_final_render, bool is_edit_mode)
|
|
|
|
|
{
|
|
|
|
|
if (is_final_render) {
|
|
|
|
|
return eModifierMode_Render;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
return eModifierMode_Realtime | (is_edit_mode ? eModifierMode_Editmode : 0);
|
|
|
|
|
}
|