This repository has been archived on 2023-10-09. You can view files and clone it. You cannot open issues or pull requests or push a commit.
Files
blender-archive/intern/cycles/kernel/kernel_camera.h
Michael Jones (Apple) a0f269f682 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 16:14:43 +01:00

522 lines
17 KiB
C++

/*
* Copyright 2011-2013 Blender Foundation
*
* Licensed under the Apache License, Version 2.0 (the "License");
* you may not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
#pragma once
#include "kernel_differential.h"
#include "kernel_lookup_table.h"
#include "kernel_montecarlo.h"
#include "kernel_projection.h"
CCL_NAMESPACE_BEGIN
/* Perspective Camera */
ccl_device float2 camera_sample_aperture(ccl_constant KernelCamera *cam, float u, float v)
{
float blades = cam->blades;
float2 bokeh;
if (blades == 0.0f) {
/* sample disk */
bokeh = concentric_sample_disk(u, v);
}
else {
/* sample polygon */
float rotation = cam->bladesrotation;
bokeh = regular_polygon_sample(blades, rotation, u, v);
}
/* anamorphic lens bokeh */
bokeh.x *= cam->inv_aperture_ratio;
return bokeh;
}
ccl_device void camera_sample_perspective(ccl_global const KernelGlobals *ccl_restrict kg,
float raster_x,
float raster_y,
float lens_u,
float lens_v,
ccl_private Ray *ray)
{
/* create ray form raster position */
ProjectionTransform rastertocamera = kernel_data.cam.rastertocamera;
float3 raster = make_float3(raster_x, raster_y, 0.0f);
float3 Pcamera = transform_perspective(&rastertocamera, raster);
#ifdef __CAMERA_MOTION__
if (kernel_data.cam.have_perspective_motion) {
/* TODO(sergey): Currently we interpolate projected coordinate which
* gives nice looking result and which is simple, but is in fact a bit
* different comparing to constructing projective matrix from an
* interpolated field of view.
*/
if (ray->time < 0.5f) {
ProjectionTransform rastertocamera_pre = kernel_data.cam.perspective_pre;
float3 Pcamera_pre = transform_perspective(&rastertocamera_pre, raster);
Pcamera = interp(Pcamera_pre, Pcamera, ray->time * 2.0f);
}
else {
ProjectionTransform rastertocamera_post = kernel_data.cam.perspective_post;
float3 Pcamera_post = transform_perspective(&rastertocamera_post, raster);
Pcamera = interp(Pcamera, Pcamera_post, (ray->time - 0.5f) * 2.0f);
}
}
#endif
float3 P = zero_float3();
float3 D = Pcamera;
/* modify ray for depth of field */
float aperturesize = kernel_data.cam.aperturesize;
if (aperturesize > 0.0f) {
/* sample point on aperture */
float2 lensuv = camera_sample_aperture(&kernel_data.cam, lens_u, lens_v) * aperturesize;
/* compute point on plane of focus */
float ft = kernel_data.cam.focaldistance / D.z;
float3 Pfocus = D * ft;
/* update ray for effect of lens */
P = make_float3(lensuv.x, lensuv.y, 0.0f);
D = normalize(Pfocus - P);
}
/* transform ray from camera to world */
Transform cameratoworld = kernel_data.cam.cameratoworld;
#ifdef __CAMERA_MOTION__
if (kernel_data.cam.num_motion_steps) {
transform_motion_array_interpolate(&cameratoworld,
kernel_tex_array(__camera_motion),
kernel_data.cam.num_motion_steps,
ray->time);
}
#endif
P = transform_point(&cameratoworld, P);
D = normalize(transform_direction(&cameratoworld, D));
bool use_stereo = kernel_data.cam.interocular_offset != 0.0f;
if (!use_stereo) {
/* No stereo */
ray->P = P;
ray->D = D;
#ifdef __RAY_DIFFERENTIALS__
float3 Dcenter = transform_direction(&cameratoworld, Pcamera);
float3 Dcenter_normalized = normalize(Dcenter);
/* TODO: can this be optimized to give compact differentials directly? */
ray->dP = differential_zero_compact();
differential3 dD;
dD.dx = normalize(Dcenter + float4_to_float3(kernel_data.cam.dx)) - Dcenter_normalized;
dD.dy = normalize(Dcenter + float4_to_float3(kernel_data.cam.dy)) - Dcenter_normalized;
ray->dD = differential_make_compact(dD);
#endif
}
else {
/* Spherical stereo */
spherical_stereo_transform(&kernel_data.cam, &P, &D);
ray->P = P;
ray->D = D;
#ifdef __RAY_DIFFERENTIALS__
/* Ray differentials, computed from scratch using the raster coordinates
* because we don't want to be affected by depth of field. We compute
* ray origin and direction for the center and two neighboring pixels
* and simply take their differences. */
float3 Pnostereo = transform_point(&cameratoworld, zero_float3());
float3 Pcenter = Pnostereo;
float3 Dcenter = Pcamera;
Dcenter = normalize(transform_direction(&cameratoworld, Dcenter));
spherical_stereo_transform(&kernel_data.cam, &Pcenter, &Dcenter);
float3 Px = Pnostereo;
float3 Dx = transform_perspective(&rastertocamera,
make_float3(raster_x + 1.0f, raster_y, 0.0f));
Dx = normalize(transform_direction(&cameratoworld, Dx));
spherical_stereo_transform(&kernel_data.cam, &Px, &Dx);
differential3 dP, dD;
dP.dx = Px - Pcenter;
dD.dx = Dx - Dcenter;
float3 Py = Pnostereo;
float3 Dy = transform_perspective(&rastertocamera,
make_float3(raster_x, raster_y + 1.0f, 0.0f));
Dy = normalize(transform_direction(&cameratoworld, Dy));
spherical_stereo_transform(&kernel_data.cam, &Py, &Dy);
dP.dy = Py - Pcenter;
dD.dy = Dy - Dcenter;
ray->dD = differential_make_compact(dD);
ray->dP = differential_make_compact(dP);
#endif
}
#ifdef __CAMERA_CLIPPING__
/* clipping */
float z_inv = 1.0f / normalize(Pcamera).z;
float nearclip = kernel_data.cam.nearclip * z_inv;
ray->P += nearclip * ray->D;
ray->dP += nearclip * ray->dD;
ray->t = kernel_data.cam.cliplength * z_inv;
#else
ray->t = FLT_MAX;
#endif
}
/* Orthographic Camera */
ccl_device void camera_sample_orthographic(ccl_global const KernelGlobals *ccl_restrict kg,
float raster_x,
float raster_y,
float lens_u,
float lens_v,
ccl_private Ray *ray)
{
/* create ray form raster position */
ProjectionTransform rastertocamera = kernel_data.cam.rastertocamera;
float3 Pcamera = transform_perspective(&rastertocamera, make_float3(raster_x, raster_y, 0.0f));
float3 P;
float3 D = make_float3(0.0f, 0.0f, 1.0f);
/* modify ray for depth of field */
float aperturesize = kernel_data.cam.aperturesize;
if (aperturesize > 0.0f) {
/* sample point on aperture */
float2 lensuv = camera_sample_aperture(&kernel_data.cam, lens_u, lens_v) * aperturesize;
/* compute point on plane of focus */
float3 Pfocus = D * kernel_data.cam.focaldistance;
/* update ray for effect of lens */
float3 lensuvw = make_float3(lensuv.x, lensuv.y, 0.0f);
P = Pcamera + lensuvw;
D = normalize(Pfocus - lensuvw);
}
else {
P = Pcamera;
}
/* transform ray from camera to world */
Transform cameratoworld = kernel_data.cam.cameratoworld;
#ifdef __CAMERA_MOTION__
if (kernel_data.cam.num_motion_steps) {
transform_motion_array_interpolate(&cameratoworld,
kernel_tex_array(__camera_motion),
kernel_data.cam.num_motion_steps,
ray->time);
}
#endif
ray->P = transform_point(&cameratoworld, P);
ray->D = normalize(transform_direction(&cameratoworld, D));
#ifdef __RAY_DIFFERENTIALS__
/* ray differential */
differential3 dP;
dP.dx = float4_to_float3(kernel_data.cam.dx);
dP.dy = float4_to_float3(kernel_data.cam.dx);
ray->dP = differential_make_compact(dP);
ray->dD = differential_zero_compact();
#endif
#ifdef __CAMERA_CLIPPING__
/* clipping */
ray->t = kernel_data.cam.cliplength;
#else
ray->t = FLT_MAX;
#endif
}
/* Panorama Camera */
ccl_device_inline void camera_sample_panorama(ccl_constant KernelCamera *cam,
#ifdef __CAMERA_MOTION__
ccl_global const DecomposedTransform *cam_motion,
#endif
float raster_x,
float raster_y,
float lens_u,
float lens_v,
ccl_private Ray *ray)
{
ProjectionTransform rastertocamera = cam->rastertocamera;
float3 Pcamera = transform_perspective(&rastertocamera, make_float3(raster_x, raster_y, 0.0f));
/* create ray form raster position */
float3 P = zero_float3();
float3 D = panorama_to_direction(cam, Pcamera.x, Pcamera.y);
/* indicates ray should not receive any light, outside of the lens */
if (is_zero(D)) {
ray->t = 0.0f;
return;
}
/* modify ray for depth of field */
float aperturesize = cam->aperturesize;
if (aperturesize > 0.0f) {
/* sample point on aperture */
float2 lensuv = camera_sample_aperture(cam, lens_u, lens_v) * aperturesize;
/* compute point on plane of focus */
float3 Dfocus = normalize(D);
float3 Pfocus = Dfocus * cam->focaldistance;
/* calculate orthonormal coordinates perpendicular to Dfocus */
float3 U, V;
U = normalize(make_float3(1.0f, 0.0f, 0.0f) - Dfocus.x * Dfocus);
V = normalize(cross(Dfocus, U));
/* update ray for effect of lens */
P = U * lensuv.x + V * lensuv.y;
D = normalize(Pfocus - P);
}
/* transform ray from camera to world */
Transform cameratoworld = cam->cameratoworld;
#ifdef __CAMERA_MOTION__
if (cam->num_motion_steps) {
transform_motion_array_interpolate(
&cameratoworld, cam_motion, cam->num_motion_steps, ray->time);
}
#endif
P = transform_point(&cameratoworld, P);
D = normalize(transform_direction(&cameratoworld, D));
/* Stereo transform */
bool use_stereo = cam->interocular_offset != 0.0f;
if (use_stereo) {
spherical_stereo_transform(cam, &P, &D);
}
ray->P = P;
ray->D = D;
#ifdef __RAY_DIFFERENTIALS__
/* Ray differentials, computed from scratch using the raster coordinates
* because we don't want to be affected by depth of field. We compute
* ray origin and direction for the center and two neighboring pixels
* and simply take their differences. */
float3 Pcenter = Pcamera;
float3 Dcenter = panorama_to_direction(cam, Pcenter.x, Pcenter.y);
Pcenter = transform_point(&cameratoworld, Pcenter);
Dcenter = normalize(transform_direction(&cameratoworld, Dcenter));
if (use_stereo) {
spherical_stereo_transform(cam, &Pcenter, &Dcenter);
}
float3 Px = transform_perspective(&rastertocamera, make_float3(raster_x + 1.0f, raster_y, 0.0f));
float3 Dx = panorama_to_direction(cam, Px.x, Px.y);
Px = transform_point(&cameratoworld, Px);
Dx = normalize(transform_direction(&cameratoworld, Dx));
if (use_stereo) {
spherical_stereo_transform(cam, &Px, &Dx);
}
differential3 dP, dD;
dP.dx = Px - Pcenter;
dD.dx = Dx - Dcenter;
float3 Py = transform_perspective(&rastertocamera, make_float3(raster_x, raster_y + 1.0f, 0.0f));
float3 Dy = panorama_to_direction(cam, Py.x, Py.y);
Py = transform_point(&cameratoworld, Py);
Dy = normalize(transform_direction(&cameratoworld, Dy));
if (use_stereo) {
spherical_stereo_transform(cam, &Py, &Dy);
}
dP.dy = Py - Pcenter;
dD.dy = Dy - Dcenter;
ray->dD = differential_make_compact(dD);
ray->dP = differential_make_compact(dP);
#endif
#ifdef __CAMERA_CLIPPING__
/* clipping */
float nearclip = cam->nearclip;
ray->P += nearclip * ray->D;
ray->dP += nearclip * ray->dD;
ray->t = cam->cliplength;
#else
ray->t = FLT_MAX;
#endif
}
/* Common */
ccl_device_inline void camera_sample(ccl_global const KernelGlobals *ccl_restrict kg,
int x,
int y,
float filter_u,
float filter_v,
float lens_u,
float lens_v,
float time,
ccl_private Ray *ray)
{
/* pixel filter */
int filter_table_offset = kernel_data.film.filter_table_offset;
float raster_x = x + lookup_table_read(kg, filter_u, filter_table_offset, FILTER_TABLE_SIZE);
float raster_y = y + lookup_table_read(kg, filter_v, filter_table_offset, FILTER_TABLE_SIZE);
#ifdef __CAMERA_MOTION__
/* motion blur */
if (kernel_data.cam.shuttertime == -1.0f) {
ray->time = 0.5f;
}
else {
/* TODO(sergey): Such lookup is unneeded when there's rolling shutter
* effect in use but rolling shutter duration is set to 0.0.
*/
const int shutter_table_offset = kernel_data.cam.shutter_table_offset;
ray->time = lookup_table_read(kg, time, shutter_table_offset, SHUTTER_TABLE_SIZE);
/* TODO(sergey): Currently single rolling shutter effect type only
* where scan-lines are acquired from top to bottom and whole scan-line
* is acquired at once (no delay in acquisition happens between pixels
* of single scan-line).
*
* Might want to support more models in the future.
*/
if (kernel_data.cam.rolling_shutter_type) {
/* Time corresponding to a fully rolling shutter only effect:
* top of the frame is time 0.0, bottom of the frame is time 1.0.
*/
const float time = 1.0f - (float)y / kernel_data.cam.height;
const float duration = kernel_data.cam.rolling_shutter_duration;
if (duration != 0.0f) {
/* This isn't fully physical correct, but lets us to have simple
* controls in the interface. The idea here is basically sort of
* linear interpolation between how much rolling shutter effect
* exist on the frame and how much of it is a motion blur effect.
*/
ray->time = (ray->time - 0.5f) * duration;
ray->time += (time - 0.5f) * (1.0f - duration) + 0.5f;
}
else {
ray->time = time;
}
}
}
#endif
/* sample */
if (kernel_data.cam.type == CAMERA_PERSPECTIVE) {
camera_sample_perspective(kg, raster_x, raster_y, lens_u, lens_v, ray);
}
else if (kernel_data.cam.type == CAMERA_ORTHOGRAPHIC) {
camera_sample_orthographic(kg, raster_x, raster_y, lens_u, lens_v, ray);
}
else {
#ifdef __CAMERA_MOTION__
ccl_global const DecomposedTransform *cam_motion = kernel_tex_array(__camera_motion);
camera_sample_panorama(&kernel_data.cam, cam_motion, raster_x, raster_y, lens_u, lens_v, ray);
#else
camera_sample_panorama(&kernel_data.cam, raster_x, raster_y, lens_u, lens_v, ray);
#endif
}
}
/* Utilities */
ccl_device_inline float3 camera_position(ccl_global const KernelGlobals *kg)
{
Transform cameratoworld = kernel_data.cam.cameratoworld;
return make_float3(cameratoworld.x.w, cameratoworld.y.w, cameratoworld.z.w);
}
ccl_device_inline float camera_distance(ccl_global const KernelGlobals *kg, float3 P)
{
Transform cameratoworld = kernel_data.cam.cameratoworld;
float3 camP = make_float3(cameratoworld.x.w, cameratoworld.y.w, cameratoworld.z.w);
if (kernel_data.cam.type == CAMERA_ORTHOGRAPHIC) {
float3 camD = make_float3(cameratoworld.x.z, cameratoworld.y.z, cameratoworld.z.z);
return fabsf(dot((P - camP), camD));
}
else {
return len(P - camP);
}
}
ccl_device_inline float camera_z_depth(ccl_global const KernelGlobals *kg, float3 P)
{
if (kernel_data.cam.type != CAMERA_PANORAMA) {
Transform worldtocamera = kernel_data.cam.worldtocamera;
return transform_point(&worldtocamera, P).z;
}
else {
Transform cameratoworld = kernel_data.cam.cameratoworld;
float3 camP = make_float3(cameratoworld.x.w, cameratoworld.y.w, cameratoworld.z.w);
return len(P - camP);
}
}
ccl_device_inline float3 camera_direction_from_point(ccl_global const KernelGlobals *kg, float3 P)
{
Transform cameratoworld = kernel_data.cam.cameratoworld;
if (kernel_data.cam.type == CAMERA_ORTHOGRAPHIC) {
float3 camD = make_float3(cameratoworld.x.z, cameratoworld.y.z, cameratoworld.z.z);
return -camD;
}
else {
float3 camP = make_float3(cameratoworld.x.w, cameratoworld.y.w, cameratoworld.z.w);
return normalize(camP - P);
}
}
ccl_device_inline float3 camera_world_to_ndc(ccl_global const KernelGlobals *kg,
ccl_private ShaderData *sd,
float3 P)
{
if (kernel_data.cam.type != CAMERA_PANORAMA) {
/* perspective / ortho */
if (sd->object == PRIM_NONE && kernel_data.cam.type == CAMERA_PERSPECTIVE)
P += camera_position(kg);
ProjectionTransform tfm = kernel_data.cam.worldtondc;
return transform_perspective(&tfm, P);
}
else {
/* panorama */
Transform tfm = kernel_data.cam.worldtocamera;
if (sd->object != OBJECT_NONE)
P = normalize(transform_point(&tfm, P));
else
P = normalize(transform_direction(&tfm, P));
float2 uv = direction_to_panorama(&kernel_data.cam, P);
return make_float3(uv.x, uv.y, 0.0f);
}
}
CCL_NAMESPACE_END