griefith/test
1
0
Fork 0
Code Issues Pull Requests Actions 1 Packages Projects Releases Wiki Activity
Files
f32d7dd0c8aeee53d5bca4e6ff2f2e96468cd062
test/intern/cycles/kernel/geom/shader_data.h

Ignoring revisions in .git-blame-ignore-revs. Click here to bypass and see the normal blame view.

449 lines
13 KiB
C
Raw Normal View History

Cycles: use SPDX license headers * Replace license text in headers with SPDX identifiers. * Remove specific license info from outdated readme.txt, instead leave details to the source files. * Add list of SPDX license identifiers used, and corresponding license texts. * Update copyright dates while we're at it. Ref D14069, T95597
2022-02-11 13:53:21 +01:00
/* SPDX-License-Identifier: Apache-2.0
* Copyright 2011-2022 Blender Foundation */
Cycles: merge of cycles-x branch, a major update to the renderer This includes much improved GPU rendering performance, viewport interactivity, new shadow catcher, revamped sampling settings, subsurface scattering anisotropy, new GPU volume sampling, improved PMJ sampling pattern, and more. Some features have also been removed or changed, breaking backwards compatibility. Including the removal of the OpenCL backend, for which alternatives are under development. Release notes and code docs: https://wiki.blender.org/wiki/Reference/Release_Notes/3.0/Cycles https://wiki.blender.org/wiki/Source/Render/Cycles Credits: * Sergey Sharybin * Brecht Van Lommel * Patrick Mours (OptiX backend) * Christophe Hery (subsurface scattering anisotropy) * William Leeson (PMJ sampling pattern) * Alaska (various fixes and tweaks) * Thomas Dinges (various fixes) For the full commit history, see the cycles-x branch. This squashes together all the changes since intermediate changes would often fail building or tests. Ref T87839, T87837, T87836 Fixes T90734, T89353, T80267, T80267, T77185, T69800
2021-09-20 17:59:20 +02:00
/* Functions to initialize ShaderData given.
*
* Could be from an incoming ray, intersection or sampled position. */
#pragma once
CCL_NAMESPACE_BEGIN
/* ShaderData setup from incoming ray */
#ifdef __OBJECT_MOTION__
Cycles: replace integrator state argument macros * Rename struct KernelGlobals to struct KernelGlobalsCPU * Add KernelGlobals, IntegratorState and ConstIntegratorState typedefs that every device can define in its own way. * Remove INTEGRATOR_STATE_ARGS and INTEGRATOR_STATE_PASS macros and replace with these new typedefs. * Add explicit state argument to INTEGRATOR_STATE and similar macros In preparation for decoupling main and shadow paths. Differential Revision: https://developer.blender.org/D12888
2021-10-17 16:10:10 +02:00
ccl_device void shader_setup_object_transforms(KernelGlobals kg,
Cycles: Kernel address space changes for MSL This is the first of a sequence of changes to support compiling Cycles kernels as MSL (Metal Shading Language) in preparation for a Metal GPU device implementation. MSL requires that all pointer types be declared with explicit address space attributes (device, thread, etc...). There is already precedent for this with Cycles' address space macros (ccl_global, ccl_private, etc...), therefore the first step of MSL-enablement is to apply these consistently. Line-for-line this represents the largest change required to enable MSL. Applying this change first will simplify future patches as well as offering the emergent benefit of enhanced descriptiveness. The vast majority of deltas in this patch fall into one of two cases: - Ensuring ccl_private is specified for thread-local pointer types - Ensuring ccl_global is specified for device-wide pointer types Additionally, the ccl_addr_space qualifier can be removed. Prior to Cycles X, ccl_addr_space was used as a context-dependent address space qualifier, but now it is either redundant (e.g. in struct typedefs), or can be replaced by ccl_global in the case of pointer types. Associated function variants (e.g. lcg_step_float_addrspace) are also redundant. In cases where address space qualifiers are chained with "const", this patch places the address space qualifier first. The rationale for this is that the choice of address space is likely to have the greater impact on runtime performance and overall architecture. The final part of this patch is the addition of a metal/compat.h header. This is partially complete and will be extended in future patches, paving the way for the full Metal implementation. Ref T92212 Reviewed By: brecht Maniphest Tasks: T92212 Differential Revision: https://developer.blender.org/D12864
2021-10-14 13:53:40 +01:00
ccl_private ShaderData *ccl_restrict sd,
Cycles: merge of cycles-x branch, a major update to the renderer This includes much improved GPU rendering performance, viewport interactivity, new shadow catcher, revamped sampling settings, subsurface scattering anisotropy, new GPU volume sampling, improved PMJ sampling pattern, and more. Some features have also been removed or changed, breaking backwards compatibility. Including the removal of the OpenCL backend, for which alternatives are under development. Release notes and code docs: https://wiki.blender.org/wiki/Reference/Release_Notes/3.0/Cycles https://wiki.blender.org/wiki/Source/Render/Cycles Credits: * Sergey Sharybin * Brecht Van Lommel * Patrick Mours (OptiX backend) * Christophe Hery (subsurface scattering anisotropy) * William Leeson (PMJ sampling pattern) * Alaska (various fixes and tweaks) * Thomas Dinges (various fixes) For the full commit history, see the cycles-x branch. This squashes together all the changes since intermediate changes would often fail building or tests. Ref T87839, T87837, T87836 Fixes T90734, T89353, T80267, T80267, T77185, T69800
2021-09-20 17:59:20 +02:00
float time)
{
if (sd->object_flag & SD_OBJECT_MOTION) {
sd->ob_tfm_motion = object_fetch_transform_motion(kg, sd->object, time);
sd->ob_itfm_motion = transform_quick_inverse(sd->ob_tfm_motion);
}
}
#endif
/* TODO: break this up if it helps reduce register pressure to load data from
Cleanup: spelling in comments
2021-10-06 14:44:27 +11:00
* global memory as we write it to shader-data. */
Cycles: replace integrator state argument macros * Rename struct KernelGlobals to struct KernelGlobalsCPU * Add KernelGlobals, IntegratorState and ConstIntegratorState typedefs that every device can define in its own way. * Remove INTEGRATOR_STATE_ARGS and INTEGRATOR_STATE_PASS macros and replace with these new typedefs. * Add explicit state argument to INTEGRATOR_STATE and similar macros In preparation for decoupling main and shadow paths. Differential Revision: https://developer.blender.org/D12888
2021-10-17 16:10:10 +02:00
ccl_device_inline void shader_setup_from_ray(KernelGlobals kg,
Cycles: Kernel address space changes for MSL This is the first of a sequence of changes to support compiling Cycles kernels as MSL (Metal Shading Language) in preparation for a Metal GPU device implementation. MSL requires that all pointer types be declared with explicit address space attributes (device, thread, etc...). There is already precedent for this with Cycles' address space macros (ccl_global, ccl_private, etc...), therefore the first step of MSL-enablement is to apply these consistently. Line-for-line this represents the largest change required to enable MSL. Applying this change first will simplify future patches as well as offering the emergent benefit of enhanced descriptiveness. The vast majority of deltas in this patch fall into one of two cases: - Ensuring ccl_private is specified for thread-local pointer types - Ensuring ccl_global is specified for device-wide pointer types Additionally, the ccl_addr_space qualifier can be removed. Prior to Cycles X, ccl_addr_space was used as a context-dependent address space qualifier, but now it is either redundant (e.g. in struct typedefs), or can be replaced by ccl_global in the case of pointer types. Associated function variants (e.g. lcg_step_float_addrspace) are also redundant. In cases where address space qualifiers are chained with "const", this patch places the address space qualifier first. The rationale for this is that the choice of address space is likely to have the greater impact on runtime performance and overall architecture. The final part of this patch is the addition of a metal/compat.h header. This is partially complete and will be extended in future patches, paving the way for the full Metal implementation. Ref T92212 Reviewed By: brecht Maniphest Tasks: T92212 Differential Revision: https://developer.blender.org/D12864
2021-10-14 13:53:40 +01:00
ccl_private ShaderData *ccl_restrict sd,
ccl_private const Ray *ccl_restrict ray,
ccl_private const Intersection *ccl_restrict isect)
Cycles: merge of cycles-x branch, a major update to the renderer This includes much improved GPU rendering performance, viewport interactivity, new shadow catcher, revamped sampling settings, subsurface scattering anisotropy, new GPU volume sampling, improved PMJ sampling pattern, and more. Some features have also been removed or changed, breaking backwards compatibility. Including the removal of the OpenCL backend, for which alternatives are under development. Release notes and code docs: https://wiki.blender.org/wiki/Reference/Release_Notes/3.0/Cycles https://wiki.blender.org/wiki/Source/Render/Cycles Credits: * Sergey Sharybin * Brecht Van Lommel * Patrick Mours (OptiX backend) * Christophe Hery (subsurface scattering anisotropy) * William Leeson (PMJ sampling pattern) * Alaska (various fixes and tweaks) * Thomas Dinges (various fixes) For the full commit history, see the cycles-x branch. This squashes together all the changes since intermediate changes would often fail building or tests. Ref T87839, T87837, T87836 Fixes T90734, T89353, T80267, T80267, T77185, T69800
2021-09-20 17:59:20 +02:00
{
/* Read intersection data into shader globals.
*
* TODO: this is redundant, could potentially remove some of this from
* ShaderData but would need to ensure that it also works for shadow
* shader evaluation. */
sd->u = isect->u;
sd->v = isect->v;
sd->ray_length = isect->t;
sd->type = isect->type;
Cycles: fully decouple triangle and curve primitive storage from BVH2 Previously the storage here was optimized to avoid indirections in BVH2 traversal. This helps improve performance a bit, but makes performance and memory usage of Embree and OptiX BVHs a bit worse also. It also adds code complexity in other parts of the code. Now decouple triangle and curve primitive storage from BVH2. * Reduced peak memory usage on all devices * Bit better performance for OptiX and Embree * Bit worse performance for CUDA * Simplified code: ** Intersection.prim/object now matches ShaderData.prim/object ** No more offset manipulation for mesh displacement before a BVH is built ** Remove primitive packing code and flags for Embree and OptiX ** Curve segments are now stored in a KernelCurve struct * Also happens to fix a bug in baking with incorrect prim/object Fixes T91968, T91770, T91902 Differential Revision: https://developer.blender.org/D12766
2021-02-28 23:23:24 +01:00
sd->object = isect->object;
Cleanup: renaming and consistency for kernel data * Rename "texture" to "data array". This has not used textures for a long time, there are just global memory arrays now. (On old CUDA GPUs there was a cache for textures but not global memory, so we used to put all data in textures.) * For CUDA and HIP, put globals in KernelParams struct like other devices. * Drop __ prefix for data array names, no possibility for naming conflict now that these are in a struct.
2022-06-17 17:16:37 +02:00
sd->object_flag = kernel_data_fetch(object_flag, sd->object);
Cycles: fully decouple triangle and curve primitive storage from BVH2 Previously the storage here was optimized to avoid indirections in BVH2 traversal. This helps improve performance a bit, but makes performance and memory usage of Embree and OptiX BVHs a bit worse also. It also adds code complexity in other parts of the code. Now decouple triangle and curve primitive storage from BVH2. * Reduced peak memory usage on all devices * Bit better performance for OptiX and Embree * Bit worse performance for CUDA * Simplified code: ** Intersection.prim/object now matches ShaderData.prim/object ** No more offset manipulation for mesh displacement before a BVH is built ** Remove primitive packing code and flags for Embree and OptiX ** Curve segments are now stored in a KernelCurve struct * Also happens to fix a bug in baking with incorrect prim/object Fixes T91968, T91770, T91902 Differential Revision: https://developer.blender.org/D12766
2021-02-28 23:23:24 +01:00
sd->prim = isect->prim;
Cycles: merge of cycles-x branch, a major update to the renderer This includes much improved GPU rendering performance, viewport interactivity, new shadow catcher, revamped sampling settings, subsurface scattering anisotropy, new GPU volume sampling, improved PMJ sampling pattern, and more. Some features have also been removed or changed, breaking backwards compatibility. Including the removal of the OpenCL backend, for which alternatives are under development. Release notes and code docs: https://wiki.blender.org/wiki/Reference/Release_Notes/3.0/Cycles https://wiki.blender.org/wiki/Source/Render/Cycles Credits: * Sergey Sharybin * Brecht Van Lommel * Patrick Mours (OptiX backend) * Christophe Hery (subsurface scattering anisotropy) * William Leeson (PMJ sampling pattern) * Alaska (various fixes and tweaks) * Thomas Dinges (various fixes) For the full commit history, see the cycles-x branch. This squashes together all the changes since intermediate changes would often fail building or tests. Ref T87839, T87837, T87836 Fixes T90734, T89353, T80267, T80267, T77185, T69800
2021-09-20 17:59:20 +02:00
sd->lamp = LAMP_NONE;
sd->flag = 0;
/* Read matrices and time. */
sd->time = ray->time;
#ifdef __OBJECT_MOTION__
shader_setup_object_transforms(kg, sd, ray->time);
#endif
/* Read ray data into shader globals. */
sd->I = -ray->D;
#ifdef __HAIR__
Fix Cycles OptiX crash with 3D curves after point cloud changes Includes refactoring to reduce the number of bits taken by primitive types, so they more easily fit in the OptiX limit.
2021-12-20 02:52:56 +01:00
if (sd->type & PRIMITIVE_CURVE) {
Cycles: merge of cycles-x branch, a major update to the renderer This includes much improved GPU rendering performance, viewport interactivity, new shadow catcher, revamped sampling settings, subsurface scattering anisotropy, new GPU volume sampling, improved PMJ sampling pattern, and more. Some features have also been removed or changed, breaking backwards compatibility. Including the removal of the OpenCL backend, for which alternatives are under development. Release notes and code docs: https://wiki.blender.org/wiki/Reference/Release_Notes/3.0/Cycles https://wiki.blender.org/wiki/Source/Render/Cycles Credits: * Sergey Sharybin * Brecht Van Lommel * Patrick Mours (OptiX backend) * Christophe Hery (subsurface scattering anisotropy) * William Leeson (PMJ sampling pattern) * Alaska (various fixes and tweaks) * Thomas Dinges (various fixes) For the full commit history, see the cycles-x branch. This squashes together all the changes since intermediate changes would often fail building or tests. Ref T87839, T87837, T87836 Fixes T90734, T89353, T80267, T80267, T77185, T69800
2021-09-20 17:59:20 +02:00
/* curve */
curve_shader_setup(kg, sd, ray->P, ray->D, isect->t, isect->object, isect->prim);
}
else
Cycles: pointcloud rendering This add support for rendering of the point cloud object in Blender, as a native geometry type in Cycles that is more memory and time efficient than instancing sphere meshes. This can be useful for rendering sand, water splashes, particles, motion graphics, etc. Points are currently always rendered as spheres, with backface culling. More shapes are likely to be added later, but this is the most important one and can be customized with shaders. For CPU rendering the Embree primitive is used, for GPU there is our own intersection code. Motion blur is suppored. Volumes inside points are not currently supported. Implemented with help from: * Kévin Dietrich: Alembic procedural integration * Patrick Mourse: OptiX integration * Josh Whelchel: update for cycles-x changes Ref T92573 Differential Revision: https://developer.blender.org/D9887
2021-12-01 17:30:46 +01:00
#endif
#ifdef __POINTCLOUD__
Fix Cycles OptiX crash with 3D curves after point cloud changes Includes refactoring to reduce the number of bits taken by primitive types, so they more easily fit in the OptiX limit.
2021-12-20 02:52:56 +01:00
if (sd->type & PRIMITIVE_POINT) {
Cycles: pointcloud rendering This add support for rendering of the point cloud object in Blender, as a native geometry type in Cycles that is more memory and time efficient than instancing sphere meshes. This can be useful for rendering sand, water splashes, particles, motion graphics, etc. Points are currently always rendered as spheres, with backface culling. More shapes are likely to be added later, but this is the most important one and can be customized with shaders. For CPU rendering the Embree primitive is used, for GPU there is our own intersection code. Motion blur is suppored. Volumes inside points are not currently supported. Implemented with help from: * Kévin Dietrich: Alembic procedural integration * Patrick Mourse: OptiX integration * Josh Whelchel: update for cycles-x changes Ref T92573 Differential Revision: https://developer.blender.org/D9887
2021-12-01 17:30:46 +01:00
/* point */
point_shader_setup(kg, sd, isect, ray);
}
else
Cycles: merge of cycles-x branch, a major update to the renderer This includes much improved GPU rendering performance, viewport interactivity, new shadow catcher, revamped sampling settings, subsurface scattering anisotropy, new GPU volume sampling, improved PMJ sampling pattern, and more. Some features have also been removed or changed, breaking backwards compatibility. Including the removal of the OpenCL backend, for which alternatives are under development. Release notes and code docs: https://wiki.blender.org/wiki/Reference/Release_Notes/3.0/Cycles https://wiki.blender.org/wiki/Source/Render/Cycles Credits: * Sergey Sharybin * Brecht Van Lommel * Patrick Mours (OptiX backend) * Christophe Hery (subsurface scattering anisotropy) * William Leeson (PMJ sampling pattern) * Alaska (various fixes and tweaks) * Thomas Dinges (various fixes) For the full commit history, see the cycles-x branch. This squashes together all the changes since intermediate changes would often fail building or tests. Ref T87839, T87837, T87836 Fixes T90734, T89353, T80267, T80267, T77185, T69800
2021-09-20 17:59:20 +02:00
#endif
Fix T94251: Cycles wrong pointcloud normal for instanced objects Refactor code a bit also so we need to do fewer matrix transforms for shader data setup of points and curves.
2021-12-20 06:00:38 +01:00
{
if (sd->type == PRIMITIVE_TRIANGLE) {
/* static triangle */
float3 Ng = triangle_normal(kg, sd);
Cleanup: renaming and consistency for kernel data * Rename "texture" to "data array". This has not used textures for a long time, there are just global memory arrays now. (On old CUDA GPUs there was a cache for textures but not global memory, so we used to put all data in textures.) * For CUDA and HIP, put globals in KernelParams struct like other devices. * Drop __ prefix for data array names, no possibility for naming conflict now that these are in a struct.
2022-06-17 17:16:37 +02:00
sd->shader = kernel_data_fetch(tri_shader, sd->prim);
Cycles: merge of cycles-x branch, a major update to the renderer This includes much improved GPU rendering performance, viewport interactivity, new shadow catcher, revamped sampling settings, subsurface scattering anisotropy, new GPU volume sampling, improved PMJ sampling pattern, and more. Some features have also been removed or changed, breaking backwards compatibility. Including the removal of the OpenCL backend, for which alternatives are under development. Release notes and code docs: https://wiki.blender.org/wiki/Reference/Release_Notes/3.0/Cycles https://wiki.blender.org/wiki/Source/Render/Cycles Credits: * Sergey Sharybin * Brecht Van Lommel * Patrick Mours (OptiX backend) * Christophe Hery (subsurface scattering anisotropy) * William Leeson (PMJ sampling pattern) * Alaska (various fixes and tweaks) * Thomas Dinges (various fixes) For the full commit history, see the cycles-x branch. This squashes together all the changes since intermediate changes would often fail building or tests. Ref T87839, T87837, T87836 Fixes T90734, T89353, T80267, T80267, T77185, T69800
2021-09-20 17:59:20 +02:00
Fix T94251: Cycles wrong pointcloud normal for instanced objects Refactor code a bit also so we need to do fewer matrix transforms for shader data setup of points and curves.
2021-12-20 06:00:38 +01:00
/* vectors */
Cycles: compute triangle location from barycentric instead of re-intersecting This is a bit more efficient than what we did before. Ref D12954
2022-01-13 17:12:03 +01:00
sd->P = triangle_point_from_uv(kg, sd, isect->object, isect->prim, isect->u, isect->v);
Fix T94251: Cycles wrong pointcloud normal for instanced objects Refactor code a bit also so we need to do fewer matrix transforms for shader data setup of points and curves.
2021-12-20 06:00:38 +01:00
sd->Ng = Ng;
sd->N = Ng;
Cycles: merge of cycles-x branch, a major update to the renderer This includes much improved GPU rendering performance, viewport interactivity, new shadow catcher, revamped sampling settings, subsurface scattering anisotropy, new GPU volume sampling, improved PMJ sampling pattern, and more. Some features have also been removed or changed, breaking backwards compatibility. Including the removal of the OpenCL backend, for which alternatives are under development. Release notes and code docs: https://wiki.blender.org/wiki/Reference/Release_Notes/3.0/Cycles https://wiki.blender.org/wiki/Source/Render/Cycles Credits: * Sergey Sharybin * Brecht Van Lommel * Patrick Mours (OptiX backend) * Christophe Hery (subsurface scattering anisotropy) * William Leeson (PMJ sampling pattern) * Alaska (various fixes and tweaks) * Thomas Dinges (various fixes) For the full commit history, see the cycles-x branch. This squashes together all the changes since intermediate changes would often fail building or tests. Ref T87839, T87837, T87836 Fixes T90734, T89353, T80267, T80267, T77185, T69800
2021-09-20 17:59:20 +02:00
Fix T94251: Cycles wrong pointcloud normal for instanced objects Refactor code a bit also so we need to do fewer matrix transforms for shader data setup of points and curves.
2021-12-20 06:00:38 +01:00
/* smooth normal */
if (sd->shader & SHADER_SMOOTH_NORMAL)
sd->N = triangle_smooth_normal(kg, Ng, sd->prim, sd->u, sd->v);
Cycles: merge of cycles-x branch, a major update to the renderer This includes much improved GPU rendering performance, viewport interactivity, new shadow catcher, revamped sampling settings, subsurface scattering anisotropy, new GPU volume sampling, improved PMJ sampling pattern, and more. Some features have also been removed or changed, breaking backwards compatibility. Including the removal of the OpenCL backend, for which alternatives are under development. Release notes and code docs: https://wiki.blender.org/wiki/Reference/Release_Notes/3.0/Cycles https://wiki.blender.org/wiki/Source/Render/Cycles Credits: * Sergey Sharybin * Brecht Van Lommel * Patrick Mours (OptiX backend) * Christophe Hery (subsurface scattering anisotropy) * William Leeson (PMJ sampling pattern) * Alaska (various fixes and tweaks) * Thomas Dinges (various fixes) For the full commit history, see the cycles-x branch. This squashes together all the changes since intermediate changes would often fail building or tests. Ref T87839, T87837, T87836 Fixes T90734, T89353, T80267, T80267, T77185, T69800
2021-09-20 17:59:20 +02:00
#ifdef __DPDU__
Fix T94251: Cycles wrong pointcloud normal for instanced objects Refactor code a bit also so we need to do fewer matrix transforms for shader data setup of points and curves.
2021-12-20 06:00:38 +01:00
/* dPdu/dPdv */
triangle_dPdudv(kg, sd->prim, &sd->dPdu, &sd->dPdv);
Cycles: merge of cycles-x branch, a major update to the renderer This includes much improved GPU rendering performance, viewport interactivity, new shadow catcher, revamped sampling settings, subsurface scattering anisotropy, new GPU volume sampling, improved PMJ sampling pattern, and more. Some features have also been removed or changed, breaking backwards compatibility. Including the removal of the OpenCL backend, for which alternatives are under development. Release notes and code docs: https://wiki.blender.org/wiki/Reference/Release_Notes/3.0/Cycles https://wiki.blender.org/wiki/Source/Render/Cycles Credits: * Sergey Sharybin * Brecht Van Lommel * Patrick Mours (OptiX backend) * Christophe Hery (subsurface scattering anisotropy) * William Leeson (PMJ sampling pattern) * Alaska (various fixes and tweaks) * Thomas Dinges (various fixes) For the full commit history, see the cycles-x branch. This squashes together all the changes since intermediate changes would often fail building or tests. Ref T87839, T87837, T87836 Fixes T90734, T89353, T80267, T80267, T77185, T69800
2021-09-20 17:59:20 +02:00
#endif
Fix T94251: Cycles wrong pointcloud normal for instanced objects Refactor code a bit also so we need to do fewer matrix transforms for shader data setup of points and curves.
2021-12-20 06:00:38 +01:00
}
else {
/* motion triangle */
motion_triangle_shader_setup(
kg, sd, ray->P, ray->D, isect->t, isect->object, isect->prim, false);
}
if (!(sd->object_flag & SD_OBJECT_TRANSFORM_APPLIED)) {
/* instance transform */
object_normal_transform_auto(kg, sd, &sd->N);
object_normal_transform_auto(kg, sd, &sd->Ng);
Cycles: merge of cycles-x branch, a major update to the renderer This includes much improved GPU rendering performance, viewport interactivity, new shadow catcher, revamped sampling settings, subsurface scattering anisotropy, new GPU volume sampling, improved PMJ sampling pattern, and more. Some features have also been removed or changed, breaking backwards compatibility. Including the removal of the OpenCL backend, for which alternatives are under development. Release notes and code docs: https://wiki.blender.org/wiki/Reference/Release_Notes/3.0/Cycles https://wiki.blender.org/wiki/Source/Render/Cycles Credits: * Sergey Sharybin * Brecht Van Lommel * Patrick Mours (OptiX backend) * Christophe Hery (subsurface scattering anisotropy) * William Leeson (PMJ sampling pattern) * Alaska (various fixes and tweaks) * Thomas Dinges (various fixes) For the full commit history, see the cycles-x branch. This squashes together all the changes since intermediate changes would often fail building or tests. Ref T87839, T87837, T87836 Fixes T90734, T89353, T80267, T80267, T77185, T69800
2021-09-20 17:59:20 +02:00
#ifdef __DPDU__
Fix T94251: Cycles wrong pointcloud normal for instanced objects Refactor code a bit also so we need to do fewer matrix transforms for shader data setup of points and curves.
2021-12-20 06:00:38 +01:00
object_dir_transform_auto(kg, sd, &sd->dPdu);
object_dir_transform_auto(kg, sd, &sd->dPdv);
Cycles: merge of cycles-x branch, a major update to the renderer This includes much improved GPU rendering performance, viewport interactivity, new shadow catcher, revamped sampling settings, subsurface scattering anisotropy, new GPU volume sampling, improved PMJ sampling pattern, and more. Some features have also been removed or changed, breaking backwards compatibility. Including the removal of the OpenCL backend, for which alternatives are under development. Release notes and code docs: https://wiki.blender.org/wiki/Reference/Release_Notes/3.0/Cycles https://wiki.blender.org/wiki/Source/Render/Cycles Credits: * Sergey Sharybin * Brecht Van Lommel * Patrick Mours (OptiX backend) * Christophe Hery (subsurface scattering anisotropy) * William Leeson (PMJ sampling pattern) * Alaska (various fixes and tweaks) * Thomas Dinges (various fixes) For the full commit history, see the cycles-x branch. This squashes together all the changes since intermediate changes would often fail building or tests. Ref T87839, T87837, T87836 Fixes T90734, T89353, T80267, T80267, T77185, T69800
2021-09-20 17:59:20 +02:00
#endif
Fix T94251: Cycles wrong pointcloud normal for instanced objects Refactor code a bit also so we need to do fewer matrix transforms for shader data setup of points and curves.
2021-12-20 06:00:38 +01:00
}
Cycles: merge of cycles-x branch, a major update to the renderer This includes much improved GPU rendering performance, viewport interactivity, new shadow catcher, revamped sampling settings, subsurface scattering anisotropy, new GPU volume sampling, improved PMJ sampling pattern, and more. Some features have also been removed or changed, breaking backwards compatibility. Including the removal of the OpenCL backend, for which alternatives are under development. Release notes and code docs: https://wiki.blender.org/wiki/Reference/Release_Notes/3.0/Cycles https://wiki.blender.org/wiki/Source/Render/Cycles Credits: * Sergey Sharybin * Brecht Van Lommel * Patrick Mours (OptiX backend) * Christophe Hery (subsurface scattering anisotropy) * William Leeson (PMJ sampling pattern) * Alaska (various fixes and tweaks) * Thomas Dinges (various fixes) For the full commit history, see the cycles-x branch. This squashes together all the changes since intermediate changes would often fail building or tests. Ref T87839, T87837, T87836 Fixes T90734, T89353, T80267, T80267, T77185, T69800
2021-09-20 17:59:20 +02:00
}
Cleanup: renaming and consistency for kernel data * Rename "texture" to "data array". This has not used textures for a long time, there are just global memory arrays now. (On old CUDA GPUs there was a cache for textures but not global memory, so we used to put all data in textures.) * For CUDA and HIP, put globals in KernelParams struct like other devices. * Drop __ prefix for data array names, no possibility for naming conflict now that these are in a struct.
2022-06-17 17:16:37 +02:00
sd->flag = kernel_data_fetch(shaders, (sd->shader & SHADER_MASK)).flags;
Fix T94251: Cycles wrong pointcloud normal for instanced objects Refactor code a bit also so we need to do fewer matrix transforms for shader data setup of points and curves.
2021-12-20 06:00:38 +01:00
Cycles: merge of cycles-x branch, a major update to the renderer This includes much improved GPU rendering performance, viewport interactivity, new shadow catcher, revamped sampling settings, subsurface scattering anisotropy, new GPU volume sampling, improved PMJ sampling pattern, and more. Some features have also been removed or changed, breaking backwards compatibility. Including the removal of the OpenCL backend, for which alternatives are under development. Release notes and code docs: https://wiki.blender.org/wiki/Reference/Release_Notes/3.0/Cycles https://wiki.blender.org/wiki/Source/Render/Cycles Credits: * Sergey Sharybin * Brecht Van Lommel * Patrick Mours (OptiX backend) * Christophe Hery (subsurface scattering anisotropy) * William Leeson (PMJ sampling pattern) * Alaska (various fixes and tweaks) * Thomas Dinges (various fixes) For the full commit history, see the cycles-x branch. This squashes together all the changes since intermediate changes would often fail building or tests. Ref T87839, T87837, T87836 Fixes T90734, T89353, T80267, T80267, T77185, T69800
2021-09-20 17:59:20 +02:00
/* backfacing test */
bool backfacing = (dot(sd->Ng, sd->I) < 0.0f);
if (backfacing) {
sd->flag |= SD_BACKFACING;
sd->Ng = -sd->Ng;
sd->N = -sd->N;
#ifdef __DPDU__
sd->dPdu = -sd->dPdu;
sd->dPdv = -sd->dPdv;
#endif
}
#ifdef __RAY_DIFFERENTIALS__
/* differentials */
differential_transfer_compact(&sd->dP, ray->dP, ray->D, ray->dD, sd->Ng, sd->ray_length);
differential_incoming_compact(&sd->dI, ray->D, ray->dD);
differential_dudv(&sd->du, &sd->dv, sd->dPdu, sd->dPdv, sd->dP, sd->Ng);
#endif
}
/* ShaderData setup from position sampled on mesh */
Cycles: replace integrator state argument macros * Rename struct KernelGlobals to struct KernelGlobalsCPU * Add KernelGlobals, IntegratorState and ConstIntegratorState typedefs that every device can define in its own way. * Remove INTEGRATOR_STATE_ARGS and INTEGRATOR_STATE_PASS macros and replace with these new typedefs. * Add explicit state argument to INTEGRATOR_STATE and similar macros In preparation for decoupling main and shadow paths. Differential Revision: https://developer.blender.org/D12888
2021-10-17 16:10:10 +02:00
ccl_device_inline void shader_setup_from_sample(KernelGlobals kg,
Cycles: Kernel address space changes for MSL This is the first of a sequence of changes to support compiling Cycles kernels as MSL (Metal Shading Language) in preparation for a Metal GPU device implementation. MSL requires that all pointer types be declared with explicit address space attributes (device, thread, etc...). There is already precedent for this with Cycles' address space macros (ccl_global, ccl_private, etc...), therefore the first step of MSL-enablement is to apply these consistently. Line-for-line this represents the largest change required to enable MSL. Applying this change first will simplify future patches as well as offering the emergent benefit of enhanced descriptiveness. The vast majority of deltas in this patch fall into one of two cases: - Ensuring ccl_private is specified for thread-local pointer types - Ensuring ccl_global is specified for device-wide pointer types Additionally, the ccl_addr_space qualifier can be removed. Prior to Cycles X, ccl_addr_space was used as a context-dependent address space qualifier, but now it is either redundant (e.g. in struct typedefs), or can be replaced by ccl_global in the case of pointer types. Associated function variants (e.g. lcg_step_float_addrspace) are also redundant. In cases where address space qualifiers are chained with "const", this patch places the address space qualifier first. The rationale for this is that the choice of address space is likely to have the greater impact on runtime performance and overall architecture. The final part of this patch is the addition of a metal/compat.h header. This is partially complete and will be extended in future patches, paving the way for the full Metal implementation. Ref T92212 Reviewed By: brecht Maniphest Tasks: T92212 Differential Revision: https://developer.blender.org/D12864
2021-10-14 13:53:40 +01:00
ccl_private ShaderData *ccl_restrict sd,
Cycles: merge of cycles-x branch, a major update to the renderer This includes much improved GPU rendering performance, viewport interactivity, new shadow catcher, revamped sampling settings, subsurface scattering anisotropy, new GPU volume sampling, improved PMJ sampling pattern, and more. Some features have also been removed or changed, breaking backwards compatibility. Including the removal of the OpenCL backend, for which alternatives are under development. Release notes and code docs: https://wiki.blender.org/wiki/Reference/Release_Notes/3.0/Cycles https://wiki.blender.org/wiki/Source/Render/Cycles Credits: * Sergey Sharybin * Brecht Van Lommel * Patrick Mours (OptiX backend) * Christophe Hery (subsurface scattering anisotropy) * William Leeson (PMJ sampling pattern) * Alaska (various fixes and tweaks) * Thomas Dinges (various fixes) For the full commit history, see the cycles-x branch. This squashes together all the changes since intermediate changes would often fail building or tests. Ref T87839, T87837, T87836 Fixes T90734, T89353, T80267, T80267, T77185, T69800
2021-09-20 17:59:20 +02:00
const float3 P,
const float3 Ng,
const float3 I,
int shader,
int object,
int prim,
float u,
float v,
float t,
float time,
bool object_space,
int lamp)
{
/* vectors */
sd->P = P;
sd->N = Ng;
sd->Ng = Ng;
sd->I = I;
sd->shader = shader;
if (prim != PRIM_NONE)
sd->type = PRIMITIVE_TRIANGLE;
else if (lamp != LAMP_NONE)
sd->type = PRIMITIVE_LAMP;
else
sd->type = PRIMITIVE_NONE;
/* primitive */
sd->object = object;
sd->lamp = LAMP_NONE;
/* Currently no access to bvh prim index for strand sd->prim. */
sd->prim = prim;
sd->u = u;
sd->v = v;
sd->time = time;
sd->ray_length = t;
Cleanup: renaming and consistency for kernel data * Rename "texture" to "data array". This has not used textures for a long time, there are just global memory arrays now. (On old CUDA GPUs there was a cache for textures but not global memory, so we used to put all data in textures.) * For CUDA and HIP, put globals in KernelParams struct like other devices. * Drop __ prefix for data array names, no possibility for naming conflict now that these are in a struct.
2022-06-17 17:16:37 +02:00
sd->flag = kernel_data_fetch(shaders, (sd->shader & SHADER_MASK)).flags;
Cycles: merge of cycles-x branch, a major update to the renderer This includes much improved GPU rendering performance, viewport interactivity, new shadow catcher, revamped sampling settings, subsurface scattering anisotropy, new GPU volume sampling, improved PMJ sampling pattern, and more. Some features have also been removed or changed, breaking backwards compatibility. Including the removal of the OpenCL backend, for which alternatives are under development. Release notes and code docs: https://wiki.blender.org/wiki/Reference/Release_Notes/3.0/Cycles https://wiki.blender.org/wiki/Source/Render/Cycles Credits: * Sergey Sharybin * Brecht Van Lommel * Patrick Mours (OptiX backend) * Christophe Hery (subsurface scattering anisotropy) * William Leeson (PMJ sampling pattern) * Alaska (various fixes and tweaks) * Thomas Dinges (various fixes) For the full commit history, see the cycles-x branch. This squashes together all the changes since intermediate changes would often fail building or tests. Ref T87839, T87837, T87836 Fixes T90734, T89353, T80267, T80267, T77185, T69800
2021-09-20 17:59:20 +02:00
sd->object_flag = 0;
if (sd->object != OBJECT_NONE) {
Cleanup: renaming and consistency for kernel data * Rename "texture" to "data array". This has not used textures for a long time, there are just global memory arrays now. (On old CUDA GPUs there was a cache for textures but not global memory, so we used to put all data in textures.) * For CUDA and HIP, put globals in KernelParams struct like other devices. * Drop __ prefix for data array names, no possibility for naming conflict now that these are in a struct.
2022-06-17 17:16:37 +02:00
sd->object_flag |= kernel_data_fetch(object_flag, sd->object);
Cycles: merge of cycles-x branch, a major update to the renderer This includes much improved GPU rendering performance, viewport interactivity, new shadow catcher, revamped sampling settings, subsurface scattering anisotropy, new GPU volume sampling, improved PMJ sampling pattern, and more. Some features have also been removed or changed, breaking backwards compatibility. Including the removal of the OpenCL backend, for which alternatives are under development. Release notes and code docs: https://wiki.blender.org/wiki/Reference/Release_Notes/3.0/Cycles https://wiki.blender.org/wiki/Source/Render/Cycles Credits: * Sergey Sharybin * Brecht Van Lommel * Patrick Mours (OptiX backend) * Christophe Hery (subsurface scattering anisotropy) * William Leeson (PMJ sampling pattern) * Alaska (various fixes and tweaks) * Thomas Dinges (various fixes) For the full commit history, see the cycles-x branch. This squashes together all the changes since intermediate changes would often fail building or tests. Ref T87839, T87837, T87836 Fixes T90734, T89353, T80267, T80267, T77185, T69800
2021-09-20 17:59:20 +02:00
#ifdef __OBJECT_MOTION__
shader_setup_object_transforms(kg, sd, time);
#endif
Fix Cycles assert in light sampling There is no object transform on lights.
2022-01-28 14:16:06 +01:00
/* transform into world space */
if (object_space) {
object_position_transform_auto(kg, sd, &sd->P);
object_normal_transform_auto(kg, sd, &sd->Ng);
sd->N = sd->Ng;
object_dir_transform_auto(kg, sd, &sd->I);
}
Cycles: merge of cycles-x branch, a major update to the renderer This includes much improved GPU rendering performance, viewport interactivity, new shadow catcher, revamped sampling settings, subsurface scattering anisotropy, new GPU volume sampling, improved PMJ sampling pattern, and more. Some features have also been removed or changed, breaking backwards compatibility. Including the removal of the OpenCL backend, for which alternatives are under development. Release notes and code docs: https://wiki.blender.org/wiki/Reference/Release_Notes/3.0/Cycles https://wiki.blender.org/wiki/Source/Render/Cycles Credits: * Sergey Sharybin * Brecht Van Lommel * Patrick Mours (OptiX backend) * Christophe Hery (subsurface scattering anisotropy) * William Leeson (PMJ sampling pattern) * Alaska (various fixes and tweaks) * Thomas Dinges (various fixes) For the full commit history, see the cycles-x branch. This squashes together all the changes since intermediate changes would often fail building or tests. Ref T87839, T87837, T87836 Fixes T90734, T89353, T80267, T80267, T77185, T69800
2021-09-20 17:59:20 +02:00
Fix Cycles assert in light sampling There is no object transform on lights.
2022-01-28 14:16:06 +01:00
if (sd->type == PRIMITIVE_TRIANGLE) {
/* smooth normal */
if (sd->shader & SHADER_SMOOTH_NORMAL) {
sd->N = triangle_smooth_normal(kg, Ng, sd->prim, sd->u, sd->v);
Cycles: merge of cycles-x branch, a major update to the renderer This includes much improved GPU rendering performance, viewport interactivity, new shadow catcher, revamped sampling settings, subsurface scattering anisotropy, new GPU volume sampling, improved PMJ sampling pattern, and more. Some features have also been removed or changed, breaking backwards compatibility. Including the removal of the OpenCL backend, for which alternatives are under development. Release notes and code docs: https://wiki.blender.org/wiki/Reference/Release_Notes/3.0/Cycles https://wiki.blender.org/wiki/Source/Render/Cycles Credits: * Sergey Sharybin * Brecht Van Lommel * Patrick Mours (OptiX backend) * Christophe Hery (subsurface scattering anisotropy) * William Leeson (PMJ sampling pattern) * Alaska (various fixes and tweaks) * Thomas Dinges (various fixes) For the full commit history, see the cycles-x branch. This squashes together all the changes since intermediate changes would often fail building or tests. Ref T87839, T87837, T87836 Fixes T90734, T89353, T80267, T80267, T77185, T69800
2021-09-20 17:59:20 +02:00
Fix Cycles assert in light sampling There is no object transform on lights.
2022-01-28 14:16:06 +01:00
if (!(sd->object_flag & SD_OBJECT_TRANSFORM_APPLIED)) {
object_normal_transform_auto(kg, sd, &sd->N);
}
Cycles: merge of cycles-x branch, a major update to the renderer This includes much improved GPU rendering performance, viewport interactivity, new shadow catcher, revamped sampling settings, subsurface scattering anisotropy, new GPU volume sampling, improved PMJ sampling pattern, and more. Some features have also been removed or changed, breaking backwards compatibility. Including the removal of the OpenCL backend, for which alternatives are under development. Release notes and code docs: https://wiki.blender.org/wiki/Reference/Release_Notes/3.0/Cycles https://wiki.blender.org/wiki/Source/Render/Cycles Credits: * Sergey Sharybin * Brecht Van Lommel * Patrick Mours (OptiX backend) * Christophe Hery (subsurface scattering anisotropy) * William Leeson (PMJ sampling pattern) * Alaska (various fixes and tweaks) * Thomas Dinges (various fixes) For the full commit history, see the cycles-x branch. This squashes together all the changes since intermediate changes would often fail building or tests. Ref T87839, T87837, T87836 Fixes T90734, T89353, T80267, T80267, T77185, T69800
2021-09-20 17:59:20 +02:00
}
Fix Cycles assert in light sampling There is no object transform on lights.
2022-01-28 14:16:06 +01:00
/* dPdu/dPdv */
Cycles: merge of cycles-x branch, a major update to the renderer This includes much improved GPU rendering performance, viewport interactivity, new shadow catcher, revamped sampling settings, subsurface scattering anisotropy, new GPU volume sampling, improved PMJ sampling pattern, and more. Some features have also been removed or changed, breaking backwards compatibility. Including the removal of the OpenCL backend, for which alternatives are under development. Release notes and code docs: https://wiki.blender.org/wiki/Reference/Release_Notes/3.0/Cycles https://wiki.blender.org/wiki/Source/Render/Cycles Credits: * Sergey Sharybin * Brecht Van Lommel * Patrick Mours (OptiX backend) * Christophe Hery (subsurface scattering anisotropy) * William Leeson (PMJ sampling pattern) * Alaska (various fixes and tweaks) * Thomas Dinges (various fixes) For the full commit history, see the cycles-x branch. This squashes together all the changes since intermediate changes would often fail building or tests. Ref T87839, T87837, T87836 Fixes T90734, T89353, T80267, T80267, T77185, T69800
2021-09-20 17:59:20 +02:00
#ifdef __DPDU__
Fix Cycles assert in light sampling There is no object transform on lights.
2022-01-28 14:16:06 +01:00
triangle_dPdudv(kg, sd->prim, &sd->dPdu, &sd->dPdv);
Cycles: merge of cycles-x branch, a major update to the renderer This includes much improved GPU rendering performance, viewport interactivity, new shadow catcher, revamped sampling settings, subsurface scattering anisotropy, new GPU volume sampling, improved PMJ sampling pattern, and more. Some features have also been removed or changed, breaking backwards compatibility. Including the removal of the OpenCL backend, for which alternatives are under development. Release notes and code docs: https://wiki.blender.org/wiki/Reference/Release_Notes/3.0/Cycles https://wiki.blender.org/wiki/Source/Render/Cycles Credits: * Sergey Sharybin * Brecht Van Lommel * Patrick Mours (OptiX backend) * Christophe Hery (subsurface scattering anisotropy) * William Leeson (PMJ sampling pattern) * Alaska (various fixes and tweaks) * Thomas Dinges (various fixes) For the full commit history, see the cycles-x branch. This squashes together all the changes since intermediate changes would often fail building or tests. Ref T87839, T87837, T87836 Fixes T90734, T89353, T80267, T80267, T77185, T69800
2021-09-20 17:59:20 +02:00
Fix Cycles assert in light sampling There is no object transform on lights.
2022-01-28 14:16:06 +01:00
if (!(sd->object_flag & SD_OBJECT_TRANSFORM_APPLIED)) {
object_dir_transform_auto(kg, sd, &sd->dPdu);
object_dir_transform_auto(kg, sd, &sd->dPdv);
}
#endif
Cycles: merge of cycles-x branch, a major update to the renderer This includes much improved GPU rendering performance, viewport interactivity, new shadow catcher, revamped sampling settings, subsurface scattering anisotropy, new GPU volume sampling, improved PMJ sampling pattern, and more. Some features have also been removed or changed, breaking backwards compatibility. Including the removal of the OpenCL backend, for which alternatives are under development. Release notes and code docs: https://wiki.blender.org/wiki/Reference/Release_Notes/3.0/Cycles https://wiki.blender.org/wiki/Source/Render/Cycles Credits: * Sergey Sharybin * Brecht Van Lommel * Patrick Mours (OptiX backend) * Christophe Hery (subsurface scattering anisotropy) * William Leeson (PMJ sampling pattern) * Alaska (various fixes and tweaks) * Thomas Dinges (various fixes) For the full commit history, see the cycles-x branch. This squashes together all the changes since intermediate changes would often fail building or tests. Ref T87839, T87837, T87836 Fixes T90734, T89353, T80267, T80267, T77185, T69800
2021-09-20 17:59:20 +02:00
}
Fix Cycles assert in light sampling There is no object transform on lights.
2022-01-28 14:16:06 +01:00
else {
#ifdef __DPDU__
sd->dPdu = zero_float3();
sd->dPdv = zero_float3();
Cycles: merge of cycles-x branch, a major update to the renderer This includes much improved GPU rendering performance, viewport interactivity, new shadow catcher, revamped sampling settings, subsurface scattering anisotropy, new GPU volume sampling, improved PMJ sampling pattern, and more. Some features have also been removed or changed, breaking backwards compatibility. Including the removal of the OpenCL backend, for which alternatives are under development. Release notes and code docs: https://wiki.blender.org/wiki/Reference/Release_Notes/3.0/Cycles https://wiki.blender.org/wiki/Source/Render/Cycles Credits: * Sergey Sharybin * Brecht Van Lommel * Patrick Mours (OptiX backend) * Christophe Hery (subsurface scattering anisotropy) * William Leeson (PMJ sampling pattern) * Alaska (various fixes and tweaks) * Thomas Dinges (various fixes) For the full commit history, see the cycles-x branch. This squashes together all the changes since intermediate changes would often fail building or tests. Ref T87839, T87837, T87836 Fixes T90734, T89353, T80267, T80267, T77185, T69800
2021-09-20 17:59:20 +02:00
#endif
Fix Cycles assert in light sampling There is no object transform on lights.
2022-01-28 14:16:06 +01:00
}
Cycles: merge of cycles-x branch, a major update to the renderer This includes much improved GPU rendering performance, viewport interactivity, new shadow catcher, revamped sampling settings, subsurface scattering anisotropy, new GPU volume sampling, improved PMJ sampling pattern, and more. Some features have also been removed or changed, breaking backwards compatibility. Including the removal of the OpenCL backend, for which alternatives are under development. Release notes and code docs: https://wiki.blender.org/wiki/Reference/Release_Notes/3.0/Cycles https://wiki.blender.org/wiki/Source/Render/Cycles Credits: * Sergey Sharybin * Brecht Van Lommel * Patrick Mours (OptiX backend) * Christophe Hery (subsurface scattering anisotropy) * William Leeson (PMJ sampling pattern) * Alaska (various fixes and tweaks) * Thomas Dinges (various fixes) For the full commit history, see the cycles-x branch. This squashes together all the changes since intermediate changes would often fail building or tests. Ref T87839, T87837, T87836 Fixes T90734, T89353, T80267, T80267, T77185, T69800
2021-09-20 17:59:20 +02:00
}
else {
Fix Cycles assert in light sampling There is no object transform on lights.
2022-01-28 14:16:06 +01:00
if (lamp != LAMP_NONE) {
sd->lamp = lamp;
}
Cycles: merge of cycles-x branch, a major update to the renderer This includes much improved GPU rendering performance, viewport interactivity, new shadow catcher, revamped sampling settings, subsurface scattering anisotropy, new GPU volume sampling, improved PMJ sampling pattern, and more. Some features have also been removed or changed, breaking backwards compatibility. Including the removal of the OpenCL backend, for which alternatives are under development. Release notes and code docs: https://wiki.blender.org/wiki/Reference/Release_Notes/3.0/Cycles https://wiki.blender.org/wiki/Source/Render/Cycles Credits: * Sergey Sharybin * Brecht Van Lommel * Patrick Mours (OptiX backend) * Christophe Hery (subsurface scattering anisotropy) * William Leeson (PMJ sampling pattern) * Alaska (various fixes and tweaks) * Thomas Dinges (various fixes) For the full commit history, see the cycles-x branch. This squashes together all the changes since intermediate changes would often fail building or tests. Ref T87839, T87837, T87836 Fixes T90734, T89353, T80267, T80267, T77185, T69800
2021-09-20 17:59:20 +02:00
#ifdef __DPDU__
sd->dPdu = zero_float3();
sd->dPdv = zero_float3();
#endif
}
/* backfacing test */
if (sd->prim != PRIM_NONE) {
bool backfacing = (dot(sd->Ng, sd->I) < 0.0f);
if (backfacing) {
sd->flag |= SD_BACKFACING;
sd->Ng = -sd->Ng;
sd->N = -sd->N;
#ifdef __DPDU__
sd->dPdu = -sd->dPdu;
sd->dPdv = -sd->dPdv;
#endif
}
}
#ifdef __RAY_DIFFERENTIALS__
/* no ray differentials here yet */
sd->dP = differential3_zero();
sd->dI = differential3_zero();
sd->du = differential_zero();
sd->dv = differential_zero();
#endif
}
/* ShaderData setup for displacement */
Cycles: replace integrator state argument macros * Rename struct KernelGlobals to struct KernelGlobalsCPU * Add KernelGlobals, IntegratorState and ConstIntegratorState typedefs that every device can define in its own way. * Remove INTEGRATOR_STATE_ARGS and INTEGRATOR_STATE_PASS macros and replace with these new typedefs. * Add explicit state argument to INTEGRATOR_STATE and similar macros In preparation for decoupling main and shadow paths. Differential Revision: https://developer.blender.org/D12888
2021-10-17 16:10:10 +02:00
ccl_device void shader_setup_from_displace(KernelGlobals kg,
Cycles: Kernel address space changes for MSL This is the first of a sequence of changes to support compiling Cycles kernels as MSL (Metal Shading Language) in preparation for a Metal GPU device implementation. MSL requires that all pointer types be declared with explicit address space attributes (device, thread, etc...). There is already precedent for this with Cycles' address space macros (ccl_global, ccl_private, etc...), therefore the first step of MSL-enablement is to apply these consistently. Line-for-line this represents the largest change required to enable MSL. Applying this change first will simplify future patches as well as offering the emergent benefit of enhanced descriptiveness. The vast majority of deltas in this patch fall into one of two cases: - Ensuring ccl_private is specified for thread-local pointer types - Ensuring ccl_global is specified for device-wide pointer types Additionally, the ccl_addr_space qualifier can be removed. Prior to Cycles X, ccl_addr_space was used as a context-dependent address space qualifier, but now it is either redundant (e.g. in struct typedefs), or can be replaced by ccl_global in the case of pointer types. Associated function variants (e.g. lcg_step_float_addrspace) are also redundant. In cases where address space qualifiers are chained with "const", this patch places the address space qualifier first. The rationale for this is that the choice of address space is likely to have the greater impact on runtime performance and overall architecture. The final part of this patch is the addition of a metal/compat.h header. This is partially complete and will be extended in future patches, paving the way for the full Metal implementation. Ref T92212 Reviewed By: brecht Maniphest Tasks: T92212 Differential Revision: https://developer.blender.org/D12864
2021-10-14 13:53:40 +01:00
ccl_private ShaderData *ccl_restrict sd,
Cycles: merge of cycles-x branch, a major update to the renderer This includes much improved GPU rendering performance, viewport interactivity, new shadow catcher, revamped sampling settings, subsurface scattering anisotropy, new GPU volume sampling, improved PMJ sampling pattern, and more. Some features have also been removed or changed, breaking backwards compatibility. Including the removal of the OpenCL backend, for which alternatives are under development. Release notes and code docs: https://wiki.blender.org/wiki/Reference/Release_Notes/3.0/Cycles https://wiki.blender.org/wiki/Source/Render/Cycles Credits: * Sergey Sharybin * Brecht Van Lommel * Patrick Mours (OptiX backend) * Christophe Hery (subsurface scattering anisotropy) * William Leeson (PMJ sampling pattern) * Alaska (various fixes and tweaks) * Thomas Dinges (various fixes) For the full commit history, see the cycles-x branch. This squashes together all the changes since intermediate changes would often fail building or tests. Ref T87839, T87837, T87836 Fixes T90734, T89353, T80267, T80267, T77185, T69800
2021-09-20 17:59:20 +02:00
int object,
int prim,
float u,
float v)
{
float3 P, Ng, I = zero_float3();
int shader;
triangle_point_normal(kg, object, prim, u, v, &P, &Ng, &shader);
/* force smooth shading for displacement */
shader |= SHADER_SMOOTH_NORMAL;
Cleanup: renaming and consistency for kernel data * Rename "texture" to "data array". This has not used textures for a long time, there are just global memory arrays now. (On old CUDA GPUs there was a cache for textures but not global memory, so we used to put all data in textures.) * For CUDA and HIP, put globals in KernelParams struct like other devices. * Drop __ prefix for data array names, no possibility for naming conflict now that these are in a struct.
2022-06-17 17:16:37 +02:00
shader_setup_from_sample(kg,
sd,
P,
Ng,
I,
shader,
object,
prim,
u,
v,
0.0f,
0.5f,
!(kernel_data_fetch(object_flag, object) & SD_OBJECT_TRANSFORM_APPLIED),
LAMP_NONE);
Cycles: merge of cycles-x branch, a major update to the renderer This includes much improved GPU rendering performance, viewport interactivity, new shadow catcher, revamped sampling settings, subsurface scattering anisotropy, new GPU volume sampling, improved PMJ sampling pattern, and more. Some features have also been removed or changed, breaking backwards compatibility. Including the removal of the OpenCL backend, for which alternatives are under development. Release notes and code docs: https://wiki.blender.org/wiki/Reference/Release_Notes/3.0/Cycles https://wiki.blender.org/wiki/Source/Render/Cycles Credits: * Sergey Sharybin * Brecht Van Lommel * Patrick Mours (OptiX backend) * Christophe Hery (subsurface scattering anisotropy) * William Leeson (PMJ sampling pattern) * Alaska (various fixes and tweaks) * Thomas Dinges (various fixes) For the full commit history, see the cycles-x branch. This squashes together all the changes since intermediate changes would often fail building or tests. Ref T87839, T87837, T87836 Fixes T90734, T89353, T80267, T80267, T77185, T69800
2021-09-20 17:59:20 +02:00
}
Cycles: bake transparent shadows for hair These transparent shadows can be expansive to evaluate. Especially on the GPU they can lead to poor occupancy when only some pixels require many kernel launches to trace and evaluate many layers of transparency. Baked transparency allows tracing a single ray in many cases by accumulating the throughput directly in the intersection program without recording hits or evaluating shaders. Transparency is baked at curve vertices and interpolated, for most shaders this will look practically the same as actual shader evaluation. Fixes T91428, performance regression with spring demo file due to transparent hair, and makes it render significantly faster than Blender 2.93. Differential Revision: https://developer.blender.org/D12880
2021-09-20 16:16:11 +02:00
/* ShaderData setup for point on curve. */
ccl_device void shader_setup_from_curve(KernelGlobals kg,
ccl_private ShaderData *ccl_restrict sd,
int object,
int prim,
int segment,
float u)
{
/* Primitive */
sd->type = PRIMITIVE_PACK_SEGMENT(PRIMITIVE_CURVE_THICK, segment);
sd->lamp = LAMP_NONE;
sd->prim = prim;
sd->u = u;
sd->v = 0.0f;
sd->time = 0.5f;
sd->ray_length = 0.0f;
/* Shader */
Cleanup: renaming and consistency for kernel data * Rename "texture" to "data array". This has not used textures for a long time, there are just global memory arrays now. (On old CUDA GPUs there was a cache for textures but not global memory, so we used to put all data in textures.) * For CUDA and HIP, put globals in KernelParams struct like other devices. * Drop __ prefix for data array names, no possibility for naming conflict now that these are in a struct.
2022-06-17 17:16:37 +02:00
sd->shader = kernel_data_fetch(curves, prim).shader_id;
sd->flag = kernel_data_fetch(shaders, (sd->shader & SHADER_MASK)).flags;
Cycles: bake transparent shadows for hair These transparent shadows can be expansive to evaluate. Especially on the GPU they can lead to poor occupancy when only some pixels require many kernel launches to trace and evaluate many layers of transparency. Baked transparency allows tracing a single ray in many cases by accumulating the throughput directly in the intersection program without recording hits or evaluating shaders. Transparency is baked at curve vertices and interpolated, for most shaders this will look practically the same as actual shader evaluation. Fixes T91428, performance regression with spring demo file due to transparent hair, and makes it render significantly faster than Blender 2.93. Differential Revision: https://developer.blender.org/D12880
2021-09-20 16:16:11 +02:00
/* Object */
sd->object = object;
Cleanup: renaming and consistency for kernel data * Rename "texture" to "data array". This has not used textures for a long time, there are just global memory arrays now. (On old CUDA GPUs there was a cache for textures but not global memory, so we used to put all data in textures.) * For CUDA and HIP, put globals in KernelParams struct like other devices. * Drop __ prefix for data array names, no possibility for naming conflict now that these are in a struct.
2022-06-17 17:16:37 +02:00
sd->object_flag = kernel_data_fetch(object_flag, sd->object);
Cycles: bake transparent shadows for hair These transparent shadows can be expansive to evaluate. Especially on the GPU they can lead to poor occupancy when only some pixels require many kernel launches to trace and evaluate many layers of transparency. Baked transparency allows tracing a single ray in many cases by accumulating the throughput directly in the intersection program without recording hits or evaluating shaders. Transparency is baked at curve vertices and interpolated, for most shaders this will look practically the same as actual shader evaluation. Fixes T91428, performance regression with spring demo file due to transparent hair, and makes it render significantly faster than Blender 2.93. Differential Revision: https://developer.blender.org/D12880
2021-09-20 16:16:11 +02:00
#ifdef __OBJECT_MOTION__
shader_setup_object_transforms(kg, sd, sd->time);
#endif
/* Get control points. */
Cleanup: renaming and consistency for kernel data * Rename "texture" to "data array". This has not used textures for a long time, there are just global memory arrays now. (On old CUDA GPUs there was a cache for textures but not global memory, so we used to put all data in textures.) * For CUDA and HIP, put globals in KernelParams struct like other devices. * Drop __ prefix for data array names, no possibility for naming conflict now that these are in a struct.
2022-06-17 17:16:37 +02:00
KernelCurve kcurve = kernel_data_fetch(curves, prim);
Cycles: bake transparent shadows for hair These transparent shadows can be expansive to evaluate. Especially on the GPU they can lead to poor occupancy when only some pixels require many kernel launches to trace and evaluate many layers of transparency. Baked transparency allows tracing a single ray in many cases by accumulating the throughput directly in the intersection program without recording hits or evaluating shaders. Transparency is baked at curve vertices and interpolated, for most shaders this will look practically the same as actual shader evaluation. Fixes T91428, performance regression with spring demo file due to transparent hair, and makes it render significantly faster than Blender 2.93. Differential Revision: https://developer.blender.org/D12880
2021-09-20 16:16:11 +02:00
int k0 = kcurve.first_key + PRIMITIVE_UNPACK_SEGMENT(sd->type);
int k1 = k0 + 1;
int ka = max(k0 - 1, kcurve.first_key);
int kb = min(k1 + 1, kcurve.first_key + kcurve.num_keys - 1);
float4 P_curve[4];
Cleanup: renaming and consistency for kernel data * Rename "texture" to "data array". This has not used textures for a long time, there are just global memory arrays now. (On old CUDA GPUs there was a cache for textures but not global memory, so we used to put all data in textures.) * For CUDA and HIP, put globals in KernelParams struct like other devices. * Drop __ prefix for data array names, no possibility for naming conflict now that these are in a struct.
2022-06-17 17:16:37 +02:00
P_curve[0] = kernel_data_fetch(curve_keys, ka);
P_curve[1] = kernel_data_fetch(curve_keys, k0);
P_curve[2] = kernel_data_fetch(curve_keys, k1);
P_curve[3] = kernel_data_fetch(curve_keys, kb);
Cycles: bake transparent shadows for hair These transparent shadows can be expansive to evaluate. Especially on the GPU they can lead to poor occupancy when only some pixels require many kernel launches to trace and evaluate many layers of transparency. Baked transparency allows tracing a single ray in many cases by accumulating the throughput directly in the intersection program without recording hits or evaluating shaders. Transparency is baked at curve vertices and interpolated, for most shaders this will look practically the same as actual shader evaluation. Fixes T91428, performance regression with spring demo file due to transparent hair, and makes it render significantly faster than Blender 2.93. Differential Revision: https://developer.blender.org/D12880
2021-09-20 16:16:11 +02:00
/* Interpolate position and tangent. */
sd->P = float4_to_float3(catmull_rom_basis_derivative(P_curve, sd->u));
#ifdef __DPDU__
sd->dPdu = float4_to_float3(catmull_rom_basis_derivative(P_curve, sd->u));
#endif
/* Transform into world space */
if (!(sd->object_flag & SD_OBJECT_TRANSFORM_APPLIED)) {
object_position_transform_auto(kg, sd, &sd->P);
#ifdef __DPDU__
object_dir_transform_auto(kg, sd, &sd->dPdu);
#endif
}
/* No view direction, normals or bitangent. */
sd->I = zero_float3();
sd->N = zero_float3();
sd->Ng = zero_float3();
#ifdef __DPDU__
sd->dPdv = zero_float3();
#endif
/* No ray differentials currently. */
#ifdef __RAY_DIFFERENTIALS__
sd->dP = differential3_zero();
sd->dI = differential3_zero();
sd->du = differential_zero();
sd->dv = differential_zero();
#endif
}
Cycles: merge of cycles-x branch, a major update to the renderer This includes much improved GPU rendering performance, viewport interactivity, new shadow catcher, revamped sampling settings, subsurface scattering anisotropy, new GPU volume sampling, improved PMJ sampling pattern, and more. Some features have also been removed or changed, breaking backwards compatibility. Including the removal of the OpenCL backend, for which alternatives are under development. Release notes and code docs: https://wiki.blender.org/wiki/Reference/Release_Notes/3.0/Cycles https://wiki.blender.org/wiki/Source/Render/Cycles Credits: * Sergey Sharybin * Brecht Van Lommel * Patrick Mours (OptiX backend) * Christophe Hery (subsurface scattering anisotropy) * William Leeson (PMJ sampling pattern) * Alaska (various fixes and tweaks) * Thomas Dinges (various fixes) For the full commit history, see the cycles-x branch. This squashes together all the changes since intermediate changes would often fail building or tests. Ref T87839, T87837, T87836 Fixes T90734, T89353, T80267, T80267, T77185, T69800
2021-09-20 17:59:20 +02:00
/* ShaderData setup from ray into background */
Cycles: replace integrator state argument macros * Rename struct KernelGlobals to struct KernelGlobalsCPU * Add KernelGlobals, IntegratorState and ConstIntegratorState typedefs that every device can define in its own way. * Remove INTEGRATOR_STATE_ARGS and INTEGRATOR_STATE_PASS macros and replace with these new typedefs. * Add explicit state argument to INTEGRATOR_STATE and similar macros In preparation for decoupling main and shadow paths. Differential Revision: https://developer.blender.org/D12888
2021-10-17 16:10:10 +02:00
ccl_device_inline void shader_setup_from_background(KernelGlobals kg,
Cycles: Kernel address space changes for MSL This is the first of a sequence of changes to support compiling Cycles kernels as MSL (Metal Shading Language) in preparation for a Metal GPU device implementation. MSL requires that all pointer types be declared with explicit address space attributes (device, thread, etc...). There is already precedent for this with Cycles' address space macros (ccl_global, ccl_private, etc...), therefore the first step of MSL-enablement is to apply these consistently. Line-for-line this represents the largest change required to enable MSL. Applying this change first will simplify future patches as well as offering the emergent benefit of enhanced descriptiveness. The vast majority of deltas in this patch fall into one of two cases: - Ensuring ccl_private is specified for thread-local pointer types - Ensuring ccl_global is specified for device-wide pointer types Additionally, the ccl_addr_space qualifier can be removed. Prior to Cycles X, ccl_addr_space was used as a context-dependent address space qualifier, but now it is either redundant (e.g. in struct typedefs), or can be replaced by ccl_global in the case of pointer types. Associated function variants (e.g. lcg_step_float_addrspace) are also redundant. In cases where address space qualifiers are chained with "const", this patch places the address space qualifier first. The rationale for this is that the choice of address space is likely to have the greater impact on runtime performance and overall architecture. The final part of this patch is the addition of a metal/compat.h header. This is partially complete and will be extended in future patches, paving the way for the full Metal implementation. Ref T92212 Reviewed By: brecht Maniphest Tasks: T92212 Differential Revision: https://developer.blender.org/D12864
2021-10-14 13:53:40 +01:00
ccl_private ShaderData *ccl_restrict sd,
Cycles: merge of cycles-x branch, a major update to the renderer This includes much improved GPU rendering performance, viewport interactivity, new shadow catcher, revamped sampling settings, subsurface scattering anisotropy, new GPU volume sampling, improved PMJ sampling pattern, and more. Some features have also been removed or changed, breaking backwards compatibility. Including the removal of the OpenCL backend, for which alternatives are under development. Release notes and code docs: https://wiki.blender.org/wiki/Reference/Release_Notes/3.0/Cycles https://wiki.blender.org/wiki/Source/Render/Cycles Credits: * Sergey Sharybin * Brecht Van Lommel * Patrick Mours (OptiX backend) * Christophe Hery (subsurface scattering anisotropy) * William Leeson (PMJ sampling pattern) * Alaska (various fixes and tweaks) * Thomas Dinges (various fixes) For the full commit history, see the cycles-x branch. This squashes together all the changes since intermediate changes would often fail building or tests. Ref T87839, T87837, T87836 Fixes T90734, T89353, T80267, T80267, T77185, T69800
2021-09-20 17:59:20 +02:00
const float3 ray_P,
const float3 ray_D,
const float ray_time)
{
/* for NDC coordinates */
sd->ray_P = ray_P;
/* vectors */
sd->P = ray_D;
sd->N = -ray_D;
sd->Ng = -ray_D;
sd->I = -ray_D;
sd->shader = kernel_data.background.surface_shader;
Cleanup: renaming and consistency for kernel data * Rename "texture" to "data array". This has not used textures for a long time, there are just global memory arrays now. (On old CUDA GPUs there was a cache for textures but not global memory, so we used to put all data in textures.) * For CUDA and HIP, put globals in KernelParams struct like other devices. * Drop __ prefix for data array names, no possibility for naming conflict now that these are in a struct.
2022-06-17 17:16:37 +02:00
sd->flag = kernel_data_fetch(shaders, (sd->shader & SHADER_MASK)).flags;
Cycles: merge of cycles-x branch, a major update to the renderer This includes much improved GPU rendering performance, viewport interactivity, new shadow catcher, revamped sampling settings, subsurface scattering anisotropy, new GPU volume sampling, improved PMJ sampling pattern, and more. Some features have also been removed or changed, breaking backwards compatibility. Including the removal of the OpenCL backend, for which alternatives are under development. Release notes and code docs: https://wiki.blender.org/wiki/Reference/Release_Notes/3.0/Cycles https://wiki.blender.org/wiki/Source/Render/Cycles Credits: * Sergey Sharybin * Brecht Van Lommel * Patrick Mours (OptiX backend) * Christophe Hery (subsurface scattering anisotropy) * William Leeson (PMJ sampling pattern) * Alaska (various fixes and tweaks) * Thomas Dinges (various fixes) For the full commit history, see the cycles-x branch. This squashes together all the changes since intermediate changes would often fail building or tests. Ref T87839, T87837, T87836 Fixes T90734, T89353, T80267, T80267, T77185, T69800
2021-09-20 17:59:20 +02:00
sd->object_flag = 0;
sd->time = ray_time;
sd->ray_length = 0.0f;
sd->object = OBJECT_NONE;
sd->lamp = LAMP_NONE;
sd->prim = PRIM_NONE;
sd->u = 0.0f;
sd->v = 0.0f;
#ifdef __DPDU__
/* dPdu/dPdv */
sd->dPdu = zero_float3();
sd->dPdv = zero_float3();
#endif
#ifdef __RAY_DIFFERENTIALS__
/* differentials */
sd->dP = differential3_zero(); /* TODO: ray->dP */
differential_incoming(&sd->dI, sd->dP);
sd->du = differential_zero();
sd->dv = differential_zero();
#endif
}
/* ShaderData setup from point inside volume */
#ifdef __VOLUME__
Cycles: replace integrator state argument macros * Rename struct KernelGlobals to struct KernelGlobalsCPU * Add KernelGlobals, IntegratorState and ConstIntegratorState typedefs that every device can define in its own way. * Remove INTEGRATOR_STATE_ARGS and INTEGRATOR_STATE_PASS macros and replace with these new typedefs. * Add explicit state argument to INTEGRATOR_STATE and similar macros In preparation for decoupling main and shadow paths. Differential Revision: https://developer.blender.org/D12888
2021-10-17 16:10:10 +02:00
ccl_device_inline void shader_setup_from_volume(KernelGlobals kg,
Cycles: Kernel address space changes for MSL This is the first of a sequence of changes to support compiling Cycles kernels as MSL (Metal Shading Language) in preparation for a Metal GPU device implementation. MSL requires that all pointer types be declared with explicit address space attributes (device, thread, etc...). There is already precedent for this with Cycles' address space macros (ccl_global, ccl_private, etc...), therefore the first step of MSL-enablement is to apply these consistently. Line-for-line this represents the largest change required to enable MSL. Applying this change first will simplify future patches as well as offering the emergent benefit of enhanced descriptiveness. The vast majority of deltas in this patch fall into one of two cases: - Ensuring ccl_private is specified for thread-local pointer types - Ensuring ccl_global is specified for device-wide pointer types Additionally, the ccl_addr_space qualifier can be removed. Prior to Cycles X, ccl_addr_space was used as a context-dependent address space qualifier, but now it is either redundant (e.g. in struct typedefs), or can be replaced by ccl_global in the case of pointer types. Associated function variants (e.g. lcg_step_float_addrspace) are also redundant. In cases where address space qualifiers are chained with "const", this patch places the address space qualifier first. The rationale for this is that the choice of address space is likely to have the greater impact on runtime performance and overall architecture. The final part of this patch is the addition of a metal/compat.h header. This is partially complete and will be extended in future patches, paving the way for the full Metal implementation. Ref T92212 Reviewed By: brecht Maniphest Tasks: T92212 Differential Revision: https://developer.blender.org/D12864
2021-10-14 13:53:40 +01:00
ccl_private ShaderData *ccl_restrict sd,
ccl_private const Ray *ccl_restrict ray)
Cycles: merge of cycles-x branch, a major update to the renderer This includes much improved GPU rendering performance, viewport interactivity, new shadow catcher, revamped sampling settings, subsurface scattering anisotropy, new GPU volume sampling, improved PMJ sampling pattern, and more. Some features have also been removed or changed, breaking backwards compatibility. Including the removal of the OpenCL backend, for which alternatives are under development. Release notes and code docs: https://wiki.blender.org/wiki/Reference/Release_Notes/3.0/Cycles https://wiki.blender.org/wiki/Source/Render/Cycles Credits: * Sergey Sharybin * Brecht Van Lommel * Patrick Mours (OptiX backend) * Christophe Hery (subsurface scattering anisotropy) * William Leeson (PMJ sampling pattern) * Alaska (various fixes and tweaks) * Thomas Dinges (various fixes) For the full commit history, see the cycles-x branch. This squashes together all the changes since intermediate changes would often fail building or tests. Ref T87839, T87837, T87836 Fixes T90734, T89353, T80267, T80267, T77185, T69800
2021-09-20 17:59:20 +02:00
{
/* vectors */
sd->P = ray->P;
sd->N = -ray->D;
sd->Ng = -ray->D;
sd->I = -ray->D;
sd->shader = SHADER_NONE;
sd->flag = 0;
sd->object_flag = 0;
sd->time = ray->time;
sd->ray_length = 0.0f; /* todo: can we set this to some useful value? */
sd->object = OBJECT_NONE; /* todo: fill this for texture coordinates */
sd->lamp = LAMP_NONE;
sd->prim = PRIM_NONE;
sd->type = PRIMITIVE_VOLUME;
sd->u = 0.0f;
sd->v = 0.0f;
# ifdef __DPDU__
/* dPdu/dPdv */
sd->dPdu = zero_float3();
sd->dPdv = zero_float3();
# endif
# ifdef __RAY_DIFFERENTIALS__
/* differentials */
sd->dP = differential3_zero(); /* TODO ray->dD */
differential_incoming(&sd->dI, sd->dP);
sd->du = differential_zero();
sd->dv = differential_zero();
# endif
/* for NDC coordinates */
sd->ray_P = ray->P;
sd->ray_dP = ray->dP;
}
#endif /* __VOLUME__ */
CCL_NAMESPACE_END
Reference in New Issue Copy Permalink