RetroArch/gfx/drivers_shader/shader_gl_core.cpp
2020-08-16 02:29:54 +02:00

2507 lines
77 KiB
C++

/* RetroArch - A frontend for libretro.
* Copyright (C) 2019 - Hans-Kristian Arntzen
*
* RetroArch is free software: you can redistribute it and/or modify it under the terms
* of the GNU General Public License as published by the Free Software Found-
* ation, either version 3 of the License, or (at your option) any later version.
*
* RetroArch is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY;
* without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR
* PURPOSE. See the GNU General Public License for more details.
*
* You should have received a copy of the GNU General Public License along with RetroArch.
* If not, see <http://www.gnu.org/licenses/>.
*/
#include "shader_gl_core.h"
#include "glslang_util.h"
#include "glslang_util_cxx.h"
#include <vector>
#include <memory>
#include <functional>
#include <utility>
#include <math.h>
#include <string.h>
#include <compat/strl.h>
#include <formats/image.h>
#include <retro_miscellaneous.h>
#include "slang_reflection.h"
#include "slang_reflection.hpp"
#include "spirv_glsl.hpp"
#include "../common/gl_core_common.h"
#include "../../verbosity.h"
#include "../../msg_hash.h"
using namespace std;
GLuint gl_core_cross_compile_program(
const uint32_t *vertex, size_t vertex_size,
const uint32_t *fragment, size_t fragment_size,
gl_core_buffer_locations *loc, bool flatten)
{
GLuint program = 0;
try
{
spirv_cross::ShaderResources vertex_resources;
spirv_cross::ShaderResources fragment_resources;
spirv_cross::CompilerGLSL vertex_compiler(vertex, vertex_size / 4);
spirv_cross::CompilerGLSL fragment_compiler(fragment, fragment_size / 4);
spirv_cross::CompilerGLSL::Options opts;
#ifdef HAVE_OPENGLES3
opts.es = true;
#else
opts.es = false;
#endif
opts.version = gl_core_get_cross_compiler_target_version();
opts.fragment.default_float_precision = spirv_cross::CompilerGLSL::Options::Precision::Highp;
opts.fragment.default_int_precision = spirv_cross::CompilerGLSL::Options::Precision::Highp;
opts.enable_420pack_extension = false;
vertex_compiler.set_common_options(opts);
fragment_compiler.set_common_options(opts);
vertex_resources = vertex_compiler.get_shader_resources();
fragment_resources = fragment_compiler.get_shader_resources();
for (auto &res : vertex_resources.stage_inputs)
{
uint32_t location = vertex_compiler.get_decoration(res.id, spv::DecorationLocation);
vertex_compiler.set_name(res.id, string("RARCH_ATTRIBUTE_") + to_string(location));
vertex_compiler.unset_decoration(res.id, spv::DecorationLocation);
}
for (auto &res : vertex_resources.stage_outputs)
{
uint32_t location = vertex_compiler.get_decoration(res.id, spv::DecorationLocation);
vertex_compiler.set_name(res.id, string("RARCH_VARYING_") + to_string(location));
vertex_compiler.unset_decoration(res.id, spv::DecorationLocation);
}
for (auto &res : fragment_resources.stage_inputs)
{
uint32_t location = fragment_compiler.get_decoration(res.id, spv::DecorationLocation);
fragment_compiler.set_name(res.id, string("RARCH_VARYING_") + to_string(location));
fragment_compiler.unset_decoration(res.id, spv::DecorationLocation);
}
if (vertex_resources.push_constant_buffers.size() > 1)
{
RARCH_ERR("[GLCore]: Cannot have more than one push constant buffer.\n");
return 0;
}
for (auto &res : vertex_resources.push_constant_buffers)
{
vertex_compiler.set_name(res.id, "RARCH_PUSH_VERTEX_INSTANCE");
vertex_compiler.set_name(res.base_type_id, "RARCH_PUSH_VERTEX");
}
if (vertex_resources.uniform_buffers.size() > 1)
{
RARCH_ERR("[GLCore]: Cannot have more than one uniform buffer.\n");
return 0;
}
for (auto &res : vertex_resources.uniform_buffers)
{
if (flatten)
vertex_compiler.flatten_buffer_block(res.id);
vertex_compiler.set_name(res.id, "RARCH_UBO_VERTEX_INSTANCE");
vertex_compiler.set_name(res.base_type_id, "RARCH_UBO_VERTEX");
vertex_compiler.unset_decoration(res.id, spv::DecorationDescriptorSet);
vertex_compiler.unset_decoration(res.id, spv::DecorationBinding);
}
if (fragment_resources.push_constant_buffers.size() > 1)
{
RARCH_ERR("[GLCore]: Cannot have more than one push constant block.\n");
return 0;
}
for (auto &res : fragment_resources.push_constant_buffers)
{
fragment_compiler.set_name(res.id, "RARCH_PUSH_FRAGMENT_INSTANCE");
fragment_compiler.set_name(res.base_type_id, "RARCH_PUSH_FRAGMENT");
}
if (fragment_resources.uniform_buffers.size() > 1)
{
RARCH_ERR("[GLCore]: Cannot have more than one uniform buffer.\n");
return 0;
}
for (auto &res : fragment_resources.uniform_buffers)
{
if (flatten)
fragment_compiler.flatten_buffer_block(res.id);
fragment_compiler.set_name(res.id, "RARCH_UBO_FRAGMENT_INSTANCE");
fragment_compiler.set_name(res.base_type_id, "RARCH_UBO_FRAGMENT");
fragment_compiler.unset_decoration(res.id, spv::DecorationDescriptorSet);
fragment_compiler.unset_decoration(res.id, spv::DecorationBinding);
}
std::vector<uint32_t> texture_binding_fixups;
for (auto &res : fragment_resources.sampled_images)
{
uint32_t binding = fragment_compiler.get_decoration(res.id, spv::DecorationBinding);
fragment_compiler.set_name(res.id, string("RARCH_TEXTURE_") + to_string(binding));
fragment_compiler.unset_decoration(res.id, spv::DecorationDescriptorSet);
fragment_compiler.unset_decoration(res.id, spv::DecorationBinding);
texture_binding_fixups.push_back(binding);
}
auto vertex_source = vertex_compiler.compile();
auto fragment_source = fragment_compiler.compile();
GLuint vertex_shader = gl_core_compile_shader(GL_VERTEX_SHADER, vertex_source.c_str());
GLuint fragment_shader = gl_core_compile_shader(GL_FRAGMENT_SHADER, fragment_source.c_str());
#if 0
RARCH_LOG("[GLCore]: Vertex shader:\n========\n%s\n=======\n", vertex_source.c_str());
RARCH_LOG("[GLCore]: Fragment shader:\n========\n%s\n=======\n", fragment_source.c_str());
#endif
if (!vertex_shader || !fragment_shader)
{
RARCH_ERR("[GLCore]: One or more shaders failed to compile.\n");
if (vertex_shader)
glDeleteShader(vertex_shader);
if (fragment_shader)
glDeleteShader(fragment_shader);
return 0;
}
program = glCreateProgram();
glAttachShader(program, vertex_shader);
glAttachShader(program, fragment_shader);
for (auto &res : vertex_resources.stage_inputs)
{
uint32_t location = vertex_compiler.get_decoration(res.id, spv::DecorationLocation);
glBindAttribLocation(program, location, (string("RARCH_ATTRIBUTE_") + to_string(location)).c_str());
}
glLinkProgram(program);
glDeleteShader(vertex_shader);
glDeleteShader(fragment_shader);
GLint status;
glGetProgramiv(program, GL_LINK_STATUS, &status);
if (!status)
{
GLint length;
glGetProgramiv(program, GL_INFO_LOG_LENGTH, &length);
if (length > 0)
{
char *info_log = (char*)malloc(length);
if (info_log)
{
glGetProgramInfoLog(program, length, &length, info_log);
RARCH_ERR("[GLCore]: Failed to link program: %s\n", info_log);
free(info_log);
glDeleteProgram(program);
return 0;
}
}
}
glUseProgram(program);
if (loc)
{
loc->flat_ubo_fragment = -1;
loc->flat_ubo_vertex = -1;
loc->flat_push_vertex = -1;
loc->flat_push_fragment = -1;
loc->buffer_index_ubo_vertex = GL_INVALID_INDEX;
loc->buffer_index_ubo_fragment = GL_INVALID_INDEX;
if (flatten)
{
loc->flat_ubo_vertex = glGetUniformLocation(program, "RARCH_UBO_VERTEX");
loc->flat_ubo_fragment = glGetUniformLocation(program, "RARCH_UBO_FRAGMENT");
loc->flat_push_vertex = glGetUniformLocation(program, "RARCH_PUSH_VERTEX");
loc->flat_push_fragment = glGetUniformLocation(program, "RARCH_PUSH_FRAGMENT");
}
else
{
loc->buffer_index_ubo_vertex = glGetUniformBlockIndex(program, "RARCH_UBO_VERTEX");
loc->buffer_index_ubo_fragment = glGetUniformBlockIndex(program, "RARCH_UBO_FRAGMENT");
}
}
/* Force proper bindings for textures. */
for (auto &binding : texture_binding_fixups)
{
GLint location = glGetUniformLocation(program, (string("RARCH_TEXTURE_") + to_string(binding)).c_str());
if (location >= 0)
glUniform1i(location, binding);
}
glUseProgram(0);
}
catch (const exception &e)
{
RARCH_ERR("[GLCore]: Failed to cross compile program: %s\n", e.what());
if (program != 0)
glDeleteProgram(program);
return 0;
}
return program;
}
namespace gl_core_shader
{
static const uint32_t opaque_vert[] =
#include "../drivers/vulkan_shaders/opaque.vert.inc"
;
static const uint32_t opaque_frag[] =
#include "../drivers/vulkan_shaders/opaque.frag.inc"
;
struct Texture
{
gl_core_filter_chain_texture texture;
glslang_filter_chain_filter filter;
glslang_filter_chain_filter mip_filter;
glslang_filter_chain_address address;
};
static GLenum address_to_gl(glslang_filter_chain_address type)
{
switch (type)
{
#ifdef HAVE_OPENGLES3
case GLSLANG_FILTER_CHAIN_ADDRESS_CLAMP_TO_BORDER:
#if 0
RARCH_WARN("[GLCore]: No CLAMP_TO_BORDER in GLES3. Falling back to edge clamp.\n");
#endif
return GL_CLAMP_TO_EDGE;
#else
case GLSLANG_FILTER_CHAIN_ADDRESS_CLAMP_TO_BORDER:
return GL_CLAMP_TO_BORDER;
#endif
case GLSLANG_FILTER_CHAIN_ADDRESS_REPEAT:
return GL_REPEAT;
case GLSLANG_FILTER_CHAIN_ADDRESS_MIRRORED_REPEAT:
return GL_MIRRORED_REPEAT;
case GLSLANG_FILTER_CHAIN_ADDRESS_CLAMP_TO_EDGE:
default:
break;
}
return GL_CLAMP_TO_EDGE;
}
static GLenum convert_filter_to_mag_gl(glslang_filter_chain_filter filter)
{
switch (filter)
{
case GLSLANG_FILTER_CHAIN_LINEAR:
return GL_LINEAR;
case GLSLANG_FILTER_CHAIN_NEAREST:
default:
break;
}
return GL_NEAREST;
}
static GLenum convert_filter_to_min_gl(glslang_filter_chain_filter filter, glslang_filter_chain_filter mipfilter)
{
if ( (filter == GLSLANG_FILTER_CHAIN_LINEAR)
&& (mipfilter == GLSLANG_FILTER_CHAIN_LINEAR)
)
return GL_LINEAR_MIPMAP_LINEAR;
else if (filter == GLSLANG_FILTER_CHAIN_LINEAR)
return GL_LINEAR_MIPMAP_NEAREST;
else if (mipfilter == GLSLANG_FILTER_CHAIN_LINEAR)
return GL_NEAREST_MIPMAP_LINEAR;
return GL_NEAREST_MIPMAP_NEAREST;
}
static GLenum convert_glslang_format(glslang_format fmt)
{
#undef FMT
#define FMT(x, r) case SLANG_FORMAT_##x: return GL_##r
switch (fmt)
{
FMT(R8_UNORM, R8);
FMT(R8_SINT, R8I);
FMT(R8_UINT, R8UI);
FMT(R8G8_UNORM, RG8);
FMT(R8G8_SINT, RG8I);
FMT(R8G8_UINT, RG8UI);
FMT(R8G8B8A8_UNORM, RGBA8);
FMT(R8G8B8A8_SINT, RGBA8I);
FMT(R8G8B8A8_UINT, RGBA8UI);
FMT(R8G8B8A8_SRGB, SRGB8_ALPHA8);
FMT(A2B10G10R10_UNORM_PACK32, RGB10_A2);
FMT(A2B10G10R10_UINT_PACK32, RGB10_A2UI);
FMT(R16_UINT, R16UI);
FMT(R16_SINT, R16I);
FMT(R16_SFLOAT, R16F);
FMT(R16G16_UINT, RG16UI);
FMT(R16G16_SINT, RG16I);
FMT(R16G16_SFLOAT, RG16F);
FMT(R16G16B16A16_UINT, RGBA16UI);
FMT(R16G16B16A16_SINT, RGBA16I);
FMT(R16G16B16A16_SFLOAT, RGBA16F);
FMT(R32_UINT, R32UI);
FMT(R32_SINT, R32I);
FMT(R32_SFLOAT, R32F);
FMT(R32G32_UINT, RG32UI);
FMT(R32G32_SINT, RG32I);
FMT(R32G32_SFLOAT, RG32F);
FMT(R32G32B32A32_UINT, RGBA32UI);
FMT(R32G32B32A32_SINT, RGBA32I);
FMT(R32G32B32A32_SFLOAT, RGBA32F);
default:
return 0;
}
}
class StaticTexture
{
public:
StaticTexture(string id,
GLuint image,
unsigned width, unsigned height,
bool linear,
bool mipmap,
glslang_filter_chain_address address);
~StaticTexture();
StaticTexture(StaticTexture&&) = delete;
void operator=(StaticTexture&&) = delete;
void set_id(string name) { id = move(name); }
const string &get_id() const { return id; }
const Texture &get_texture() const { return texture; }
private:
string id;
GLuint image;
Texture texture;
};
StaticTexture::StaticTexture(string id_, GLuint image_,
unsigned width, unsigned height, bool linear, bool mipmap,
glslang_filter_chain_address address)
: id(std::move(id_)), image(image_)
{
GLenum gl_address = address_to_gl(address);
texture.filter = GLSLANG_FILTER_CHAIN_NEAREST;
texture.mip_filter = GLSLANG_FILTER_CHAIN_NEAREST;
texture.address = address;
texture.texture.width = width;
texture.texture.height = height;
texture.texture.format = 0;
texture.texture.image = image;
if (linear)
texture.filter = GLSLANG_FILTER_CHAIN_LINEAR;
if (mipmap && linear)
texture.mip_filter = GLSLANG_FILTER_CHAIN_LINEAR;
glBindTexture(GL_TEXTURE_2D, image);
glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_WRAP_S, gl_address);
glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_WRAP_T, gl_address);
glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_MAG_FILTER, linear ? GL_LINEAR : GL_NEAREST);
if (linear && mipmap)
glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_MIN_FILTER, GL_LINEAR_MIPMAP_LINEAR);
else if (linear)
glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_MIN_FILTER, GL_LINEAR);
else
glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_MIN_FILTER, GL_NEAREST);
glBindTexture(GL_TEXTURE_2D, 0);
}
StaticTexture::~StaticTexture()
{
if (image != 0)
glDeleteTextures(1, &image);
}
struct CommonResources
{
CommonResources();
~CommonResources();
vector<Texture> original_history;
vector<Texture> framebuffer_feedback;
vector<Texture> pass_outputs;
vector<unique_ptr<StaticTexture>> luts;
unordered_map<string, slang_texture_semantic_map> texture_semantic_map;
unordered_map<string, slang_texture_semantic_map> texture_semantic_uniform_map;
unique_ptr<video_shader> shader_preset;
GLuint quad_program = 0;
GLuint quad_vbo = 0;
gl_core_buffer_locations quad_loc = {};
};
CommonResources::CommonResources()
{
static float quad_data[] = {
0.0f, 0.0f, 0.0f, 0.0f,
1.0f, 0.0f, 1.0f, 0.0f,
0.0f, 1.0f, 0.0f, 1.0f,
1.0f, 1.0f, 1.0f, 1.0f,
};
glGenBuffers(1, &quad_vbo);
glBindBuffer(GL_ARRAY_BUFFER, quad_vbo);
glBufferData(GL_ARRAY_BUFFER, sizeof(quad_data), quad_data, GL_STATIC_DRAW);
glBindBuffer(GL_ARRAY_BUFFER, 0);
quad_program = gl_core_cross_compile_program(
opaque_vert, sizeof(opaque_vert),
opaque_frag, sizeof(opaque_frag), &quad_loc, true);
}
CommonResources::~CommonResources()
{
if (quad_program != 0)
glDeleteProgram(quad_program);
if (quad_vbo != 0)
glDeleteBuffers(1, &quad_vbo);
}
class Framebuffer
{
public:
Framebuffer(GLenum format, unsigned max_levels);
~Framebuffer();
Framebuffer(Framebuffer&&) = delete;
void operator=(Framebuffer&&) = delete;
void set_size(const Size2D &size, GLenum format = 0);
const Size2D &get_size() const { return size; }
GLenum get_format() const { return format; }
GLuint get_image() const { return image; }
GLuint get_framebuffer() const { return framebuffer; }
bool is_complete() const { return complete; }
unsigned get_levels() const { return levels; }
private:
GLuint image = 0;
Size2D size;
GLenum format;
unsigned max_levels;
unsigned levels = 0;
GLuint framebuffer = 0;
void init();
bool complete = false;
};
Framebuffer::Framebuffer(GLenum format_, unsigned max_levels_)
: size({1, 1}), format(format_), max_levels(max_levels_)
{
glGenFramebuffers(1, &framebuffer);
/* Need to bind to create */
glBindFramebuffer(GL_FRAMEBUFFER, framebuffer);
glBindFramebuffer(GL_FRAMEBUFFER, 0);
if (format == 0)
format = GL_RGBA8;
}
void Framebuffer::set_size(const Size2D &size_, GLenum format_)
{
size = size_;
if (format_ != 0)
format = format_;
init();
}
void Framebuffer::init()
{
GLenum status;
glBindFramebuffer(GL_FRAMEBUFFER, framebuffer);
if (image != 0)
{
glFramebufferTexture2D(GL_FRAMEBUFFER,
GL_COLOR_ATTACHMENT0, GL_TEXTURE_2D, 0, 0);
glDeleteTextures(1, &image);
}
glGenTextures(1, &image);
glBindTexture(GL_TEXTURE_2D, image);
if (size.width == 0)
size.width = 1;
if (size.height == 0)
size.height = 1;
levels = glslang_num_miplevels(size.width, size.height);
if (max_levels < levels)
levels = max_levels;
if (levels == 0)
levels = 1;
glTexStorage2D(GL_TEXTURE_2D, levels,
format,
size.width, size.height);
glFramebufferTexture2D(GL_FRAMEBUFFER,
GL_COLOR_ATTACHMENT0, GL_TEXTURE_2D, image, 0);
status = glCheckFramebufferStatus(GL_FRAMEBUFFER);
complete = true;
if (status != GL_FRAMEBUFFER_COMPLETE)
{
complete = false;
switch (status)
{
case GL_FRAMEBUFFER_INCOMPLETE_ATTACHMENT:
RARCH_ERR("[GLCore]: Incomplete attachment.\n");
break;
case GL_FRAMEBUFFER_INCOMPLETE_MISSING_ATTACHMENT:
RARCH_ERR("[GLCore]: Incomplete, missing attachment.\n");
break;
case GL_FRAMEBUFFER_UNSUPPORTED:
{
unsigned levels;
RARCH_ERR("[GLCore]: Unsupported FBO, falling back to RGBA8.\n");
glFramebufferTexture2D(GL_FRAMEBUFFER, GL_COLOR_ATTACHMENT0, GL_TEXTURE_2D, 0, 0);
glDeleteTextures(1, &image);
glGenTextures(1, &image);
glBindTexture(GL_TEXTURE_2D, image);
levels = glslang_num_miplevels(size.width, size.height);
if (max_levels < levels)
levels = max_levels;
glTexStorage2D(GL_TEXTURE_2D, levels,
GL_RGBA8,
size.width, size.height);
glFramebufferTexture2D(GL_FRAMEBUFFER, GL_COLOR_ATTACHMENT0, GL_TEXTURE_2D, image, 0);
complete = glCheckFramebufferStatus(GL_FRAMEBUFFER) == GL_FRAMEBUFFER_COMPLETE;
}
break;
}
}
glBindFramebuffer(GL_FRAMEBUFFER, 0);
glBindTexture(GL_TEXTURE_2D, 0);
}
Framebuffer::~Framebuffer()
{
if (framebuffer != 0)
glDeleteFramebuffers(1, &framebuffer);
if (image != 0)
glDeleteTextures(1, &image);
}
class UBORing
{
public:
~UBORing();
std::vector<GLuint> buffers;
unsigned buffer_index = 0;
};
UBORing::~UBORing()
{
glDeleteBuffers(buffers.size(), buffers.data());
}
class Pass
{
public:
explicit Pass(bool final_pass) :
final_pass(final_pass)
{}
~Pass();
Pass(Pass&&) = delete;
void operator=(Pass&&) = delete;
const Framebuffer &get_framebuffer() const
{
return *framebuffer;
}
Framebuffer *get_feedback_framebuffer()
{
return framebuffer_feedback.get();
}
void set_pass_info(const gl_core_filter_chain_pass_info &info);
void set_shader(GLenum stage,
const uint32_t *spirv,
size_t spirv_words);
bool build();
bool init_feedback();
void build_commands(
const Texture &original,
const Texture &source,
const gl_core_viewport &vp,
const float *mvp);
void set_frame_count(uint64_t count)
{
frame_count = count;
}
void set_frame_count_period(unsigned period)
{
frame_count_period = period;
}
void set_frame_direction(int32_t direction)
{
frame_direction = direction;
}
void set_name(const char *name)
{
pass_name = name;
}
const string &get_name() const
{
return pass_name;
}
glslang_filter_chain_filter get_source_filter() const
{
return pass_info.source_filter;
}
glslang_filter_chain_filter get_mip_filter() const
{
return pass_info.mip_filter;
}
glslang_filter_chain_address get_address_mode() const
{
return pass_info.address;
}
void set_common_resources(CommonResources *common)
{
this->common = common;
}
const slang_reflection &get_reflection() const
{
return reflection;
}
void set_pass_number(unsigned pass)
{
pass_number = pass;
}
void add_parameter(unsigned parameter_index, const std::string &id);
void end_frame();
void allocate_buffers();
private:
bool final_pass;
Size2D get_output_size(const Size2D &original_size,
const Size2D &max_source) const;
GLuint pipeline = 0;
CommonResources *common = nullptr;
Size2D current_framebuffer_size = {};
gl_core_viewport current_viewport;
gl_core_filter_chain_pass_info pass_info;
vector<uint32_t> vertex_shader;
vector<uint32_t> fragment_shader;
unique_ptr<Framebuffer> framebuffer;
unique_ptr<Framebuffer> framebuffer_feedback;
bool init_pipeline();
void set_semantic_texture(slang_texture_semantic semantic,
const Texture &texture);
slang_reflection reflection;
std::vector<uint8_t> uniforms;
void build_semantics(uint8_t *buffer,
const float *mvp,
const Texture &original, const Texture &source);
void build_semantic_vec4(uint8_t *data, slang_semantic semantic,
unsigned width, unsigned height);
void build_semantic_uint(uint8_t *data,
slang_semantic semantic, uint32_t value);
void build_semantic_int(uint8_t *data,
slang_semantic semantic, int32_t value);
void build_semantic_parameter(uint8_t *data, unsigned index, float value);
void build_semantic_texture_vec4(uint8_t *data,
slang_texture_semantic semantic,
unsigned width, unsigned height);
void build_semantic_texture_array_vec4(uint8_t *data,
slang_texture_semantic semantic, unsigned index,
unsigned width, unsigned height);
void build_semantic_texture(uint8_t *buffer,
slang_texture_semantic semantic, const Texture &texture);
void build_semantic_texture_array(uint8_t *buffer,
slang_texture_semantic semantic,
unsigned index, const Texture &texture);
uint64_t frame_count = 0;
unsigned frame_count_period = 0;
int32_t frame_direction = 1;
unsigned pass_number = 0;
size_t ubo_offset = 0;
string pass_name;
struct Parameter
{
string id;
unsigned index;
unsigned semantic_index;
};
vector<Parameter> parameters;
vector<Parameter> filtered_parameters;
vector<uint8_t> push_constant_buffer;
gl_core_buffer_locations locations = {};
UBORing ubo_ring;
void reflect_parameter(const std::string &name, slang_semantic_meta &meta);
void reflect_parameter(const std::string &name, slang_texture_semantic_meta &meta);
void reflect_parameter_array(const std::string &name, std::vector<slang_texture_semantic_meta> &meta);
};
bool Pass::build()
{
unordered_map<string, slang_semantic_map> semantic_map;
unsigned i;
unsigned j = 0;
framebuffer.reset();
framebuffer_feedback.reset();
if (!final_pass)
framebuffer = unique_ptr<Framebuffer>(
new Framebuffer(pass_info.rt_format, pass_info.max_levels));
for (i = 0; i < parameters.size(); i++)
{
if (!slang_set_unique_map(semantic_map, parameters[i].id,
slang_semantic_map{ SLANG_SEMANTIC_FLOAT_PARAMETER, j }))
return false;
j++;
}
reflection = slang_reflection{};
reflection.pass_number = pass_number;
reflection.texture_semantic_map = &common->texture_semantic_map;
reflection.texture_semantic_uniform_map = &common->texture_semantic_uniform_map;
reflection.semantic_map = &semantic_map;
if (!slang_reflect_spirv(vertex_shader, fragment_shader, &reflection))
return false;
/* Filter out parameters which we will never use anyways. */
filtered_parameters.clear();
for (i = 0; i < reflection.semantic_float_parameters.size(); i++)
{
if (reflection.semantic_float_parameters[i].uniform ||
reflection.semantic_float_parameters[i].push_constant)
filtered_parameters.push_back(parameters[i]);
}
if (!init_pipeline())
return false;
return true;
}
void Pass::reflect_parameter(const std::string &name, slang_semantic_meta &meta)
{
if (meta.uniform)
{
int vert = glGetUniformLocation(pipeline, (std::string("RARCH_UBO_VERTEX_INSTANCE.") + name).c_str());
int frag = glGetUniformLocation(pipeline, (std::string("RARCH_UBO_FRAGMENT_INSTANCE.") + name).c_str());
if (vert >= 0)
meta.location.ubo_vertex = vert;
if (frag >= 0)
meta.location.ubo_fragment = frag;
}
if (meta.push_constant)
{
int vert = glGetUniformLocation(pipeline, (std::string("RARCH_PUSH_VERTEX_INSTANCE.") + name).c_str());
int frag = glGetUniformLocation(pipeline, (std::string("RARCH_PUSH_FRAGMENT_INSTANCE.") + name).c_str());
if (vert >= 0)
meta.location.push_vertex = vert;
if (frag >= 0)
meta.location.push_fragment = frag;
}
}
void Pass::reflect_parameter(const std::string &name, slang_texture_semantic_meta &meta)
{
if (meta.uniform)
{
int vert = glGetUniformLocation(pipeline, (std::string("RARCH_UBO_VERTEX_INSTANCE.") + name).c_str());
int frag = glGetUniformLocation(pipeline, (std::string("RARCH_UBO_FRAGMENT_INSTANCE.") + name).c_str());
if (vert >= 0)
meta.location.ubo_vertex = vert;
if (frag >= 0)
meta.location.ubo_fragment = frag;
}
if (meta.push_constant)
{
int vert = glGetUniformLocation(pipeline, (std::string("RARCH_PUSH_VERTEX_INSTANCE.") + name).c_str());
int frag = glGetUniformLocation(pipeline, (std::string("RARCH_PUSH_FRAGMENT_INSTANCE.") + name).c_str());
if (vert >= 0)
meta.location.push_vertex = vert;
if (frag >= 0)
meta.location.push_fragment = frag;
}
}
void Pass::reflect_parameter_array(const std::string &name, std::vector<slang_texture_semantic_meta> &meta)
{
size_t i;
for (i = 0; i < meta.size(); i++)
{
std::string n = name + std::to_string(i);
slang_texture_semantic_meta *m = (slang_texture_semantic_meta*)&meta[i];
if (m->uniform)
{
int vert = glGetUniformLocation(pipeline,
(std::string("RARCH_UBO_VERTEX_INSTANCE.") + n).c_str());
int frag = glGetUniformLocation(pipeline,
(std::string("RARCH_UBO_FRAGMENT_INSTANCE.") + n).c_str());
if (vert >= 0)
m->location.ubo_vertex = vert;
if (frag >= 0)
m->location.ubo_fragment = frag;
}
if (m->push_constant)
{
int vert = glGetUniformLocation(pipeline,
(std::string("RARCH_PUSH_VERTEX_INSTANCE.") + n).c_str());
int frag = glGetUniformLocation(pipeline,
(std::string("RARCH_PUSH_FRAGMENT_INSTANCE.") + n).c_str());
if (vert >= 0)
m->location.push_vertex = vert;
if (frag >= 0)
m->location.push_fragment = frag;
}
}
}
bool Pass::init_pipeline()
{
pipeline = gl_core_cross_compile_program(
vertex_shader.data(), vertex_shader.size() * sizeof(uint32_t),
fragment_shader.data(), fragment_shader.size() * sizeof(uint32_t),
&locations, false);
if (!pipeline)
return false;
uniforms.resize(reflection.ubo_size);
if (reflection.ubo_size)
{
unsigned i;
size_t size = reflection.ubo_size;
unsigned count = 16;
ubo_ring.buffers.resize(count);
glGenBuffers(count, ubo_ring.buffers.data());
for (i = 0; i < ubo_ring.buffers.size(); i++)
{
glBindBuffer(GL_UNIFORM_BUFFER, ubo_ring.buffers[i]);
glBufferData(GL_UNIFORM_BUFFER, size, NULL, GL_STREAM_DRAW);
}
glBindBuffer(GL_UNIFORM_BUFFER, 0);
}
push_constant_buffer.resize(reflection.push_constant_size);
reflect_parameter("MVP", reflection.semantics[SLANG_SEMANTIC_MVP]);
reflect_parameter("OutputSize", reflection.semantics[SLANG_SEMANTIC_OUTPUT]);
reflect_parameter("FinalViewportSize", reflection.semantics[SLANG_SEMANTIC_FINAL_VIEWPORT]);
reflect_parameter("FrameCount", reflection.semantics[SLANG_SEMANTIC_FRAME_COUNT]);
reflect_parameter("FrameDirection", reflection.semantics[SLANG_SEMANTIC_FRAME_DIRECTION]);
reflect_parameter("OriginalSize", reflection.semantic_textures[SLANG_TEXTURE_SEMANTIC_ORIGINAL][0]);
reflect_parameter("SourceSize", reflection.semantic_textures[SLANG_TEXTURE_SEMANTIC_SOURCE][0]);
reflect_parameter_array("OriginalHistorySize", reflection.semantic_textures[SLANG_TEXTURE_SEMANTIC_ORIGINAL_HISTORY]);
reflect_parameter_array("PassOutputSize", reflection.semantic_textures[SLANG_TEXTURE_SEMANTIC_PASS_OUTPUT]);
reflect_parameter_array("PassFeedbackSize", reflection.semantic_textures[SLANG_TEXTURE_SEMANTIC_PASS_FEEDBACK]);
reflect_parameter_array("UserSize", reflection.semantic_textures[SLANG_TEXTURE_SEMANTIC_USER]);
for (auto &m : common->texture_semantic_uniform_map)
{
auto &array = reflection.semantic_textures[m.second.semantic];
if (m.second.index < array.size())
reflect_parameter(m.first, array[m.second.index]);
}
for (auto &m : filtered_parameters)
if (m.semantic_index < reflection.semantic_float_parameters.size())
reflect_parameter(m.id, reflection.semantic_float_parameters[m.semantic_index]);
return true;
}
void Pass::set_pass_info(const gl_core_filter_chain_pass_info &info)
{
pass_info = info;
}
Size2D Pass::get_output_size(const Size2D &original,
const Size2D &source) const
{
float width = 0.0f;
float height = 0.0f;
switch (pass_info.scale_type_x)
{
case GLSLANG_FILTER_CHAIN_SCALE_ORIGINAL:
width = float(original.width) * pass_info.scale_x;
break;
case GLSLANG_FILTER_CHAIN_SCALE_SOURCE:
width = float(source.width) * pass_info.scale_x;
break;
case GLSLANG_FILTER_CHAIN_SCALE_VIEWPORT:
width = current_viewport.width * pass_info.scale_x;
break;
case GLSLANG_FILTER_CHAIN_SCALE_ABSOLUTE:
width = pass_info.scale_x;
break;
default:
break;
}
switch (pass_info.scale_type_y)
{
case GLSLANG_FILTER_CHAIN_SCALE_ORIGINAL:
height = float(original.height) * pass_info.scale_y;
break;
case GLSLANG_FILTER_CHAIN_SCALE_SOURCE:
height = float(source.height) * pass_info.scale_y;
break;
case GLSLANG_FILTER_CHAIN_SCALE_VIEWPORT:
height = current_viewport.height * pass_info.scale_y;
break;
case GLSLANG_FILTER_CHAIN_SCALE_ABSOLUTE:
height = pass_info.scale_y;
break;
default:
break;
}
return { unsigned(roundf(width)), unsigned(roundf(height)) };
}
void Pass::end_frame()
{
swap(framebuffer, framebuffer_feedback);
}
void Pass::build_semantic_vec4(uint8_t *data, slang_semantic semantic,
unsigned width, unsigned height)
{
slang_semantic_meta *refl = (slang_semantic_meta*)
&reflection.semantics[semantic];
if (data && refl->uniform)
{
if (refl->location.ubo_vertex >= 0 || refl->location.ubo_fragment >= 0)
{
float v4[4];
glslang_build_vec4(v4, width, height);
if (refl->location.ubo_vertex >= 0)
glUniform4fv(refl->location.ubo_vertex, 1, v4);
if (refl->location.ubo_fragment >= 0)
glUniform4fv(refl->location.ubo_fragment, 1, v4);
}
else
glslang_build_vec4(
reinterpret_cast<float *>(data + refl->ubo_offset),
width,
height);
}
if (refl->push_constant)
{
if ( refl->location.push_vertex >= 0 ||
refl->location.push_fragment >= 0)
{
float v4[4];
glslang_build_vec4(v4, width, height);
if (refl->location.push_vertex >= 0)
glUniform4fv(refl->location.push_vertex, 1, v4);
if (refl->location.push_fragment >= 0)
glUniform4fv(refl->location.push_fragment, 1, v4);
}
else
glslang_build_vec4(
reinterpret_cast<float *>
(push_constant_buffer.data() + refl->push_constant_offset),
width,
height);
}
}
void Pass::build_semantic_parameter(uint8_t *data, unsigned index, float value)
{
slang_semantic_meta *refl = (slang_semantic_meta*)
&reflection.semantic_float_parameters[index];
/* We will have filtered out stale parameters. */
if (data && refl->uniform)
{
if (refl->location.ubo_vertex >= 0 || refl->location.ubo_fragment >= 0)
{
if (refl->location.ubo_vertex >= 0)
glUniform1f(refl->location.ubo_vertex, value);
if (refl->location.ubo_fragment >= 0)
glUniform1f(refl->location.ubo_fragment, value);
}
else
*reinterpret_cast<float *>(data + refl->ubo_offset) = value;
}
if (refl->push_constant)
{
if (refl->location.push_vertex >= 0 || refl->location.push_fragment >= 0)
{
if (refl->location.push_vertex >= 0)
glUniform1f(refl->location.push_vertex, value);
if (refl->location.push_fragment >= 0)
glUniform1f(refl->location.push_fragment, value);
}
else
*reinterpret_cast<float *>(push_constant_buffer.data() + refl->push_constant_offset) = value;
}
}
void Pass::build_semantic_uint(uint8_t *data, slang_semantic semantic,
uint32_t value)
{
auto &refl = reflection.semantics[semantic];
if (data && refl.uniform)
{
if (refl.location.ubo_vertex >= 0 || refl.location.ubo_fragment >= 0)
{
if (refl.location.ubo_vertex >= 0)
glUniform1ui(refl.location.ubo_vertex, value);
if (refl.location.ubo_fragment >= 0)
glUniform1ui(refl.location.ubo_fragment, value);
}
else
*reinterpret_cast<uint32_t *>(data + reflection.semantics[semantic].ubo_offset) = value;
}
if (refl.push_constant)
{
if (refl.location.push_vertex >= 0 || refl.location.push_fragment >= 0)
{
if (refl.location.push_vertex >= 0)
glUniform1ui(refl.location.push_vertex, value);
if (refl.location.push_fragment >= 0)
glUniform1ui(refl.location.push_fragment, value);
}
else
*reinterpret_cast<uint32_t *>(push_constant_buffer.data() + refl.push_constant_offset) = value;
}
}
void Pass::build_semantic_int(uint8_t *data, slang_semantic semantic,
int32_t value)
{
auto &refl = reflection.semantics[semantic];
if (data && refl.uniform)
{
if (refl.location.ubo_vertex >= 0 || refl.location.ubo_fragment >= 0)
{
if (refl.location.ubo_vertex >= 0)
glUniform1i(refl.location.ubo_vertex, value);
if (refl.location.ubo_fragment >= 0)
glUniform1i(refl.location.ubo_fragment, value);
}
else
*reinterpret_cast<int32_t *>(data + reflection.semantics[semantic].ubo_offset) = value;
}
if (refl.push_constant)
{
if (refl.location.push_vertex >= 0 || refl.location.push_fragment >= 0)
{
if (refl.location.push_vertex >= 0)
glUniform1i(refl.location.push_vertex, value);
if (refl.location.push_fragment >= 0)
glUniform1i(refl.location.push_fragment, value);
}
else
*reinterpret_cast<int32_t *>(push_constant_buffer.data() + refl.push_constant_offset) = value;
}
}
void Pass::build_semantic_texture(uint8_t *buffer,
slang_texture_semantic semantic, const Texture &texture)
{
build_semantic_texture_vec4(buffer, semantic,
texture.texture.width, texture.texture.height);
set_semantic_texture(semantic, texture);
}
void Pass::build_semantic_texture_array_vec4(uint8_t *data, slang_texture_semantic semantic,
unsigned index, unsigned width, unsigned height)
{
auto &refl = reflection.semantic_textures[semantic];
if (index >= refl.size())
return;
if (data && refl[index].uniform)
{
if (refl[index].location.ubo_vertex >= 0 || refl[index].location.ubo_fragment >= 0)
{
float v4[4];
glslang_build_vec4(v4, width, height);
if (refl[index].location.ubo_vertex >= 0)
glUniform4fv(refl[index].location.ubo_vertex, 1, v4);
if (refl[index].location.ubo_fragment >= 0)
glUniform4fv(refl[index].location.ubo_fragment, 1, v4);
}
else
glslang_build_vec4(
reinterpret_cast<float *>(data + refl[index].ubo_offset),
width,
height);
}
if (refl[index].push_constant)
{
if (refl[index].location.push_vertex >= 0 || refl[index].location.push_fragment >= 0)
{
float v4[4];
glslang_build_vec4(v4, width, height);
if (refl[index].location.push_vertex >= 0)
glUniform4fv(refl[index].location.push_vertex, 1, v4);
if (refl[index].location.push_fragment >= 0)
glUniform4fv(refl[index].location.push_fragment, 1, v4);
}
else
glslang_build_vec4(
reinterpret_cast<float *>(push_constant_buffer.data() + refl[index].push_constant_offset),
width,
height);
}
}
void Pass::build_semantic_texture_vec4(uint8_t *data, slang_texture_semantic semantic,
unsigned width, unsigned height)
{
build_semantic_texture_array_vec4(data, semantic, 0, width, height);
}
bool Pass::init_feedback()
{
if (final_pass)
return false;
framebuffer_feedback = unique_ptr<Framebuffer>(
new Framebuffer(pass_info.rt_format, pass_info.max_levels));
return true;
}
Pass::~Pass()
{
if (pipeline != 0)
glDeleteProgram(pipeline);
}
void Pass::set_shader(GLenum stage,
const uint32_t *spirv,
size_t spirv_words)
{
switch (stage)
{
case GL_VERTEX_SHADER:
vertex_shader.clear();
vertex_shader.insert(end(vertex_shader),
spirv, spirv + spirv_words);
break;
case GL_FRAGMENT_SHADER:
fragment_shader.clear();
fragment_shader.insert(end(fragment_shader),
spirv, spirv + spirv_words);
break;
default:
break;
}
}
void Pass::add_parameter(unsigned index, const std::string &id)
{
parameters.push_back({ id, index, unsigned(parameters.size()) });
}
void Pass::set_semantic_texture(slang_texture_semantic semantic,
const Texture &texture)
{
if (reflection.semantic_textures[semantic][0].texture)
{
unsigned binding = reflection.semantic_textures[semantic][0].binding;
glActiveTexture(GL_TEXTURE0 + binding);
glBindTexture(GL_TEXTURE_2D, texture.texture.image);
glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_MAG_FILTER, convert_filter_to_mag_gl(texture.filter));
glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_MIN_FILTER, convert_filter_to_min_gl(texture.filter, texture.mip_filter));
glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_WRAP_S, address_to_gl(texture.address));
glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_WRAP_T, address_to_gl(texture.address));
}
}
void Pass::build_semantic_texture_array(uint8_t *buffer,
slang_texture_semantic semantic, unsigned index, const Texture &texture)
{
build_semantic_texture_array_vec4(buffer, semantic, index,
texture.texture.width, texture.texture.height);
if (index < reflection.semantic_textures[semantic].size() &&
reflection.semantic_textures[semantic][index].texture)
{
unsigned binding = reflection.semantic_textures[semantic][index].binding;
glActiveTexture(GL_TEXTURE0 + binding);
glBindTexture(GL_TEXTURE_2D, texture.texture.image);
glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_MAG_FILTER, convert_filter_to_mag_gl(texture.filter));
glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_MIN_FILTER, convert_filter_to_min_gl(texture.filter, texture.mip_filter));
glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_WRAP_S, address_to_gl(texture.address));
glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_WRAP_T, address_to_gl(texture.address));
}
}
void Pass::build_semantics(uint8_t *buffer,
const float *mvp, const Texture &original, const Texture &source)
{
unsigned i;
/* MVP */
if (buffer && reflection.semantics[SLANG_SEMANTIC_MVP].uniform)
{
size_t offset = reflection.semantics[
SLANG_SEMANTIC_MVP].ubo_offset;
if (mvp)
memcpy(buffer + offset,
mvp, sizeof(float) * 16);
else
gl_core_build_default_matrix(reinterpret_cast<float *>(
buffer + offset));
}
if (reflection.semantics[SLANG_SEMANTIC_MVP].push_constant)
{
size_t offset = reflection.semantics[
SLANG_SEMANTIC_MVP].push_constant_offset;
if (mvp)
memcpy(push_constant_buffer.data() + offset,
mvp, sizeof(float) * 16);
else
gl_core_build_default_matrix(reinterpret_cast<float *>(
push_constant_buffer.data() + offset));
}
/* Output information */
build_semantic_vec4(buffer, SLANG_SEMANTIC_OUTPUT,
current_framebuffer_size.width,
current_framebuffer_size.height);
build_semantic_vec4(buffer, SLANG_SEMANTIC_FINAL_VIEWPORT,
unsigned(current_viewport.width),
unsigned(current_viewport.height));
build_semantic_uint(buffer, SLANG_SEMANTIC_FRAME_COUNT,
frame_count_period
? uint32_t(frame_count % frame_count_period)
: uint32_t(frame_count));
build_semantic_int(buffer, SLANG_SEMANTIC_FRAME_DIRECTION,
frame_direction);
/* Standard inputs */
build_semantic_texture(buffer, SLANG_TEXTURE_SEMANTIC_ORIGINAL, original);
build_semantic_texture(buffer, SLANG_TEXTURE_SEMANTIC_SOURCE, source);
/* ORIGINAL_HISTORY[0] is an alias of ORIGINAL. */
build_semantic_texture_array(buffer,
SLANG_TEXTURE_SEMANTIC_ORIGINAL_HISTORY, 0, original);
/* Parameters. */
for (i = 0; i < filtered_parameters.size(); i++)
build_semantic_parameter(buffer,
filtered_parameters[i].semantic_index,
common->shader_preset->parameters[
filtered_parameters[i].index].current);
/* Previous inputs. */
for (i = 0; i < common->original_history.size(); i++)
build_semantic_texture_array(buffer,
SLANG_TEXTURE_SEMANTIC_ORIGINAL_HISTORY, i + 1,
common->original_history[i]);
/* Previous passes. */
for (i = 0; i < common->pass_outputs.size(); i++)
build_semantic_texture_array(buffer,
SLANG_TEXTURE_SEMANTIC_PASS_OUTPUT, i,
common->pass_outputs[i]);
/* Feedback FBOs. */
for (i = 0; i < common->framebuffer_feedback.size(); i++)
build_semantic_texture_array(buffer,
SLANG_TEXTURE_SEMANTIC_PASS_FEEDBACK, i,
common->framebuffer_feedback[i]);
/* LUTs. */
for (i = 0; i < common->luts.size(); i++)
build_semantic_texture_array(buffer,
SLANG_TEXTURE_SEMANTIC_USER, i,
common->luts[i]->get_texture());
}
void Pass::build_commands(
const Texture &original,
const Texture &source,
const gl_core_viewport &vp,
const float *mvp)
{
current_viewport = vp;
Size2D size = get_output_size(
{ original.texture.width, original.texture.height },
{ source.texture.width, source.texture.height });
if (framebuffer &&
(size.width != framebuffer->get_size().width ||
size.height != framebuffer->get_size().height))
framebuffer->set_size(size);
current_framebuffer_size = size;
glUseProgram(pipeline);
build_semantics(uniforms.data(), mvp, original, source);
if (locations.flat_ubo_vertex >= 0)
glUniform4fv(locations.flat_ubo_vertex,
GLsizei((reflection.ubo_size + 15) / 16),
reinterpret_cast<const float *>(uniforms.data()));
if (locations.flat_ubo_fragment >= 0)
glUniform4fv(locations.flat_ubo_fragment,
GLsizei((reflection.ubo_size + 15) / 16),
reinterpret_cast<const float *>(uniforms.data()));
if (locations.flat_push_vertex >= 0)
glUniform4fv(locations.flat_push_vertex,
GLsizei((reflection.push_constant_size + 15) / 16),
reinterpret_cast<const float *>(push_constant_buffer.data()));
if (locations.flat_push_fragment >= 0)
glUniform4fv(locations.flat_push_fragment,
GLsizei((reflection.push_constant_size + 15) / 16),
reinterpret_cast<const float *>(push_constant_buffer.data()));
if (!( locations.buffer_index_ubo_vertex == GL_INVALID_INDEX
&& locations.buffer_index_ubo_fragment == GL_INVALID_INDEX))
{
/* UBO Ring - update and bind */
unsigned vertex_binding = locations.buffer_index_ubo_vertex;
unsigned fragment_binding = locations.buffer_index_ubo_fragment;
const void *data = uniforms.data();
size_t size = reflection.ubo_size;
GLuint id = ubo_ring.buffers[ubo_ring.buffer_index];
glBindBuffer(GL_UNIFORM_BUFFER, id);
glBufferSubData(GL_UNIFORM_BUFFER, 0, size, data);
glBindBuffer(GL_UNIFORM_BUFFER, 0);
if (vertex_binding != GL_INVALID_INDEX)
glBindBufferBase(GL_UNIFORM_BUFFER, vertex_binding, id);
if (fragment_binding != GL_INVALID_INDEX)
glBindBufferBase(GL_UNIFORM_BUFFER, fragment_binding, id);
ubo_ring.buffer_index++;
if (ubo_ring.buffer_index >= ubo_ring.buffers.size())
ubo_ring.buffer_index = 0;
}
/* The final pass is always executed inside
* another render pass since the frontend will
* want to overlay various things on top for
* the passes that end up on-screen. */
if (!final_pass && framebuffer->is_complete())
{
glBindFramebuffer(GL_FRAMEBUFFER, framebuffer->get_framebuffer());
glColorMask(GL_TRUE, GL_TRUE, GL_TRUE, GL_TRUE);
glClearColor(0.0f, 0.0f, 0.0f, 0.0f);
glClear(GL_COLOR_BUFFER_BIT);
}
if (final_pass)
glViewport(current_viewport.x, current_viewport.y,
current_viewport.width, current_viewport.height);
else
glViewport(0, 0, size.width, size.height);
#if !defined(HAVE_OPENGLES)
if (framebuffer && framebuffer->get_format() == GL_SRGB8_ALPHA8)
glEnable(GL_FRAMEBUFFER_SRGB);
else
glDisable(GL_FRAMEBUFFER_SRGB);
#endif
/* Draw quad */
glDisable(GL_CULL_FACE);
glDisable(GL_BLEND);
glDisable(GL_DEPTH_TEST);
glEnableVertexAttribArray(0);
glEnableVertexAttribArray(1);
glBindBuffer(GL_ARRAY_BUFFER, common->quad_vbo);
glVertexAttribPointer(0, 2, GL_FLOAT, GL_FALSE, 4 * sizeof(float),
reinterpret_cast<void *>(uintptr_t(0)));
glVertexAttribPointer(1, 2, GL_FLOAT, GL_FALSE, 4 * sizeof(float),
reinterpret_cast<void *>(uintptr_t(2 * sizeof(float))));
glDrawArrays(GL_TRIANGLE_STRIP, 0, 4);
glBindBuffer(GL_ARRAY_BUFFER, 0);
glDisableVertexAttribArray(0);
glDisableVertexAttribArray(1);
#if !defined(HAVE_OPENGLES)
glDisable(GL_FRAMEBUFFER_SRGB);
#endif
glBindFramebuffer(GL_FRAMEBUFFER, 0);
if (!final_pass)
if (framebuffer->get_levels() > 1)
{
glBindFramebuffer(GL_FRAMEBUFFER, 0);
glBindTexture(GL_TEXTURE_2D, framebuffer->get_image());
glGenerateMipmap(GL_TEXTURE_2D);
glBindTexture(GL_TEXTURE_2D, 0);
}
}
}
struct gl_core_filter_chain
{
public:
gl_core_filter_chain(unsigned num_passes) { set_num_passes(num_passes); }
inline void set_shader_preset(unique_ptr<video_shader> shader)
{
common.shader_preset = move(shader);
}
inline video_shader *get_shader_preset()
{
return common.shader_preset.get();
}
void set_pass_info(unsigned pass,
const gl_core_filter_chain_pass_info &info);
void set_shader(unsigned pass, GLenum stage,
const uint32_t *spirv, size_t spirv_words);
bool init();
void set_input_texture(const gl_core_filter_chain_texture &texture);
void build_offscreen_passes(const gl_core_viewport &vp);
void build_viewport_pass(const gl_core_viewport &vp, const float *mvp);
void end_frame();
void set_frame_count(uint64_t count);
void set_frame_count_period(unsigned pass, unsigned period);
void set_frame_direction(int32_t direction);
void set_pass_name(unsigned pass, const char *name);
void add_static_texture(unique_ptr<gl_core_shader::StaticTexture> texture);
void add_parameter(unsigned pass, unsigned parameter_index, const std::string &id);
void set_num_passes(unsigned passes);
private:
vector<unique_ptr<gl_core_shader::Pass>> passes;
vector<gl_core_filter_chain_pass_info> pass_info;
vector<vector<function<void ()>>> deferred_calls;
std::unique_ptr<gl_core_shader::Framebuffer> copy_framebuffer;
gl_core_shader::CommonResources common;
gl_core_filter_chain_texture input_texture = {};
bool init_history();
bool init_feedback();
bool init_alias();
vector<unique_ptr<gl_core_shader::Framebuffer>> original_history;
bool require_clear = false;
void clear_history_and_feedback();
void update_feedback_info();
void update_history_info();
};
void gl_core_filter_chain::update_history_info()
{
unsigned i;
for (i = 0; i < original_history.size(); i++)
{
gl_core_shader::Texture *source = (gl_core_shader::Texture*)
&common.original_history[i];
if (!source)
continue;
source->texture.image = original_history[i]->get_image();
source->texture.width = original_history[i]->get_size().width;
source->texture.height = original_history[i]->get_size().height;
source->filter = passes.front()->get_source_filter();
source->mip_filter = passes.front()->get_mip_filter();
source->address = passes.front()->get_address_mode();
}
}
void gl_core_filter_chain::update_feedback_info()
{
unsigned i;
for (i = 0; i < passes.size() - 1; i++)
{
gl_core_shader::Framebuffer *fb = passes[i]->get_feedback_framebuffer();
if (!fb)
continue;
gl_core_shader::Texture *source = (gl_core_shader::Texture*)
&common.framebuffer_feedback[i];
if (!source)
continue;
source->texture.image = fb->get_image();
source->texture.width = fb->get_size().width;
source->texture.height = fb->get_size().height;
source->filter = passes[i]->get_source_filter();
source->mip_filter = passes[i]->get_mip_filter();
source->address = passes[i]->get_address_mode();
}
}
void gl_core_filter_chain::build_offscreen_passes(const gl_core_viewport &vp)
{
unsigned i;
/* First frame, make sure our history and feedback textures
* are in a clean state. */
if (require_clear)
{
clear_history_and_feedback();
require_clear = false;
}
update_history_info();
if (!common.framebuffer_feedback.empty())
update_feedback_info();
const gl_core_shader::Texture original = {
input_texture,
passes.front()->get_source_filter(),
passes.front()->get_mip_filter(),
passes.front()->get_address_mode(),
};
gl_core_shader::Texture source = original;
for (i = 0; i < passes.size() - 1; i++)
{
passes[i]->build_commands(original, source, vp, nullptr);
const gl_core_shader::Framebuffer &fb = passes[i]->get_framebuffer();
source.texture.image = fb.get_image();
source.texture.width = fb.get_size().width;
source.texture.height = fb.get_size().height;
source.filter = passes[i + 1]->get_source_filter();
source.mip_filter = passes[i + 1]->get_mip_filter();
source.address = passes[i + 1]->get_address_mode();
common.pass_outputs[i] = source;
}
}
void gl_core_filter_chain::end_frame()
{
/* If we need to keep old frames, copy it after fragment is complete.
* TODO: We can improve pipelining by figuring out which
* pass is the last that reads from
* the history and dispatch the copy earlier. */
if (!original_history.empty())
{
/* Update history */
unique_ptr<gl_core_shader::Framebuffer> tmp;
unique_ptr<gl_core_shader::Framebuffer> &back = original_history.back();
swap(back, tmp);
if (input_texture.width != tmp->get_size().width ||
input_texture.height != tmp->get_size().height ||
(input_texture.format != 0
&& input_texture.format != tmp->get_format()))
tmp->set_size({ input_texture.width, input_texture.height }, input_texture.format);
if (tmp->is_complete())
gl_core_framebuffer_copy(
tmp->get_framebuffer(),
common.quad_program,
common.quad_vbo,
common.quad_loc.flat_ubo_vertex,
tmp->get_size(),
input_texture.image);
/* Should ring buffer, but we don't have *that* many passes. */
move_backward(begin(original_history), end(original_history) - 1, end(original_history));
swap(original_history.front(), tmp);
}
}
void gl_core_filter_chain::build_viewport_pass(
const gl_core_viewport &vp, const float *mvp)
{
unsigned i;
/* First frame, make sure our history and
* feedback textures are in a clean state. */
if (require_clear)
{
clear_history_and_feedback();
require_clear = false;
}
gl_core_shader::Texture source;
const gl_core_shader::Texture original = {
input_texture,
passes.front()->get_source_filter(),
passes.front()->get_mip_filter(),
passes.front()->get_address_mode(),
};
if (passes.size() == 1)
{
source = {
input_texture,
passes.back()->get_source_filter(),
passes.back()->get_mip_filter(),
passes.back()->get_address_mode(),
};
}
else
{
const gl_core_shader::Framebuffer &fb = passes[passes.size() - 2]
->get_framebuffer();
source.texture.image = fb.get_image();
source.texture.width = fb.get_size().width;
source.texture.height = fb.get_size().height;
source.filter = passes.back()->get_source_filter();
source.mip_filter = passes.back()->get_mip_filter();
source.address = passes.back()->get_address_mode();
}
passes.back()->build_commands(original, source, vp, mvp);
/* For feedback FBOs, swap current and previous. */
for (i = 0; i < passes.size(); i++)
{
gl_core_shader::Framebuffer *fb = passes[i]->get_feedback_framebuffer();
if (fb)
passes[i]->end_frame();
}
}
bool gl_core_filter_chain::init_history()
{
unsigned i;
size_t required_images = 0;
original_history.clear();
common.original_history.clear();
for (i = 0; i < passes.size(); i++)
required_images =
max(required_images,
passes[i]->get_reflection().semantic_textures[
SLANG_TEXTURE_SEMANTIC_ORIGINAL_HISTORY].size());
if (required_images < 2)
{
RARCH_LOG("[GLCore]: Not using frame history.\n");
return true;
}
/* We don't need to store array element #0,
* since it's aliased with the actual original. */
required_images--;
original_history.reserve(required_images);
common.original_history.resize(required_images);
for (i = 0; i < required_images; i++)
original_history.emplace_back(new gl_core_shader::Framebuffer(0, 1));
RARCH_LOG("[GLCore]: Using history of %u frames.\n", unsigned(required_images));
/* On first frame, we need to clear the textures to
* a known state, but we need
* a command buffer for that, so just defer to first frame.
*/
require_clear = true;
return true;
}
bool gl_core_filter_chain::init_feedback()
{
unsigned i;
bool use_feedbacks = false;
common.framebuffer_feedback.clear();
/* Final pass cannot have feedback. */
for (i = 0; i < passes.size() - 1; i++)
{
bool use_feedback = false;
for (auto &pass : passes)
{
auto &r = pass->get_reflection();
auto &feedbacks = r.semantic_textures[SLANG_TEXTURE_SEMANTIC_PASS_FEEDBACK];
if (i < feedbacks.size() && feedbacks[i].texture)
{
use_feedback = true;
use_feedbacks = true;
break;
}
}
if (use_feedback && !passes[i]->init_feedback())
return false;
if (use_feedback)
RARCH_LOG("[GLCore]: Using framebuffer feedback for pass #%u.\n", i);
}
if (!use_feedbacks)
{
RARCH_LOG("[GLCore]: Not using framebuffer feedback.\n");
return true;
}
common.framebuffer_feedback.resize(passes.size() - 1);
require_clear = true;
return true;
}
bool gl_core_filter_chain::init_alias()
{
unsigned i, j;
common.texture_semantic_map.clear();
common.texture_semantic_uniform_map.clear();
for (i = 0; i < passes.size(); i++)
{
const string name = passes[i]->get_name();
if (name.empty())
continue;
j = &passes[i] - passes.data();
if (!slang_set_unique_map(common.texture_semantic_map, name,
slang_texture_semantic_map{ SLANG_TEXTURE_SEMANTIC_PASS_OUTPUT, j }))
return false;
if (!slang_set_unique_map(common.texture_semantic_uniform_map,
name + "Size",
slang_texture_semantic_map{ SLANG_TEXTURE_SEMANTIC_PASS_OUTPUT, j }))
return false;
if (!slang_set_unique_map(common.texture_semantic_map,
name + "Feedback",
slang_texture_semantic_map{ SLANG_TEXTURE_SEMANTIC_PASS_FEEDBACK, j }))
return false;
if (!slang_set_unique_map(common.texture_semantic_uniform_map,
name + "FeedbackSize",
slang_texture_semantic_map{ SLANG_TEXTURE_SEMANTIC_PASS_FEEDBACK, j }))
return false;
}
for (i = 0; i < common.luts.size(); i++)
{
j = &common.luts[i] - common.luts.data();
if (!slang_set_unique_map(common.texture_semantic_map,
common.luts[i]->get_id(),
slang_texture_semantic_map{ SLANG_TEXTURE_SEMANTIC_USER, j }))
return false;
if (!slang_set_unique_map(common.texture_semantic_uniform_map,
common.luts[i]->get_id() + "Size",
slang_texture_semantic_map{ SLANG_TEXTURE_SEMANTIC_USER, j }))
return false;
}
return true;
}
void gl_core_filter_chain::set_pass_info(unsigned pass, const gl_core_filter_chain_pass_info &info)
{
if (pass >= pass_info.size())
pass_info.resize(pass + 1);
pass_info[pass] = info;
}
void gl_core_filter_chain::set_num_passes(unsigned num_passes)
{
unsigned i;
pass_info.resize(num_passes);
passes.reserve(num_passes);
for (i = 0; i < num_passes; i++)
{
passes.emplace_back(new gl_core_shader::Pass(i + 1 == num_passes));
passes.back()->set_common_resources(&common);
passes.back()->set_pass_number(i);
}
}
void gl_core_filter_chain::set_shader(unsigned pass, GLenum stage, const uint32_t *spirv, size_t spirv_words)
{
passes[pass]->set_shader(stage, spirv, spirv_words);
}
void gl_core_filter_chain::add_parameter(unsigned pass,
unsigned index, const std::string &id)
{
passes[pass]->add_parameter(index, id);
}
bool gl_core_filter_chain::init()
{
unsigned i;
if (!init_alias())
return false;
for (i = 0; i < passes.size(); i++)
{
RARCH_LOG("[slang]: Building pass #%u (%s)\n", i,
passes[i]->get_name().empty() ?
msg_hash_to_str(MENU_ENUM_LABEL_VALUE_NOT_AVAILABLE) :
passes[i]->get_name().c_str());
passes[i]->set_pass_info(pass_info[i]);
if (!passes[i]->build())
return false;
}
require_clear = false;
if (!init_history())
return false;
if (!init_feedback())
return false;
common.pass_outputs.resize(passes.size());
return true;
}
void gl_core_filter_chain::clear_history_and_feedback()
{
unsigned i;
for (i = 0; i < original_history.size(); i++)
{
if (original_history[i]->is_complete())
gl_core_framebuffer_clear(original_history[i]->get_framebuffer());
}
for (i = 0; i < passes.size(); i++)
{
gl_core_shader::Framebuffer *fb = passes[i]->get_feedback_framebuffer();
if (fb && fb->is_complete())
gl_core_framebuffer_clear(fb->get_framebuffer());
}
}
void gl_core_filter_chain::set_input_texture(
const gl_core_filter_chain_texture &texture)
{
input_texture = texture;
/* Need a copy to remove padding.
* GL HW render interface in libretro is kinda garbage now ... */
if (input_texture.padded_width != input_texture.width ||
input_texture.padded_height != input_texture.height)
{
if (!copy_framebuffer)
copy_framebuffer.reset(new gl_core_shader::Framebuffer(texture.format, 1));
if (input_texture.width != copy_framebuffer->get_size().width ||
input_texture.height != copy_framebuffer->get_size().height ||
(input_texture.format != 0 &&
input_texture.format != copy_framebuffer->get_format()))
copy_framebuffer->set_size({ input_texture.width, input_texture.height }, input_texture.format);
if (copy_framebuffer->is_complete())
gl_core_framebuffer_copy_partial(
copy_framebuffer->get_framebuffer(),
common.quad_program,
common.quad_loc.flat_ubo_vertex,
copy_framebuffer->get_size(),
input_texture.image,
float(input_texture.width)
/ input_texture.padded_width,
float(input_texture.height)
/ input_texture.padded_height);
input_texture.image = copy_framebuffer->get_image();
}
}
void gl_core_filter_chain::add_static_texture(unique_ptr<gl_core_shader::StaticTexture> texture)
{
common.luts.push_back(move(texture));
}
void gl_core_filter_chain::set_frame_count(uint64_t count)
{
unsigned i;
for (i = 0; i < passes.size(); i++)
passes[i]->set_frame_count(count);
}
void gl_core_filter_chain::set_frame_count_period(unsigned pass, unsigned period)
{
passes[pass]->set_frame_count_period(period);
}
void gl_core_filter_chain::set_frame_direction(int32_t direction)
{
unsigned i;
for (i = 0; i < passes.size(); i++)
passes[i]->set_frame_direction(direction);
}
void gl_core_filter_chain::set_pass_name(unsigned pass, const char *name)
{
passes[pass]->set_name(name);
}
static unique_ptr<gl_core_shader::StaticTexture> gl_core_filter_chain_load_lut(
gl_core_filter_chain *chain,
const video_shader_lut *shader)
{
texture_image image;
GLuint tex = 0;
image.width = 0;
image.height = 0;
image.pixels = NULL;
image.supports_rgba = true;
if (!image_texture_load(&image, shader->path))
return {};
unsigned levels = shader->mipmap ? glslang_num_miplevels(image.width, image.height) : 1;
glGenTextures(1, &tex);
glBindTexture(GL_TEXTURE_2D, tex);
glTexStorage2D(GL_TEXTURE_2D, levels,
GL_RGBA8, image.width, image.height);
glPixelStorei(GL_UNPACK_ROW_LENGTH, 0);
glPixelStorei(GL_UNPACK_ALIGNMENT, 4);
glBindBuffer(GL_PIXEL_UNPACK_BUFFER, 0);
glTexSubImage2D(GL_TEXTURE_2D, 0, 0, 0,
image.width, image.height,
GL_RGBA, GL_UNSIGNED_BYTE, image.pixels);
if (levels > 1)
glGenerateMipmap(GL_TEXTURE_2D);
glBindTexture(GL_TEXTURE_2D, 0);
if (image.pixels)
image_texture_free(&image);
return unique_ptr<gl_core_shader::StaticTexture>(new gl_core_shader::StaticTexture(shader->id,
tex, image.width, image.height,
shader->filter != RARCH_FILTER_NEAREST,
levels > 1,
rarch_wrap_to_address(shader->wrap)));
}
static bool gl_core_filter_chain_load_luts(
gl_core_filter_chain *chain,
video_shader *shader)
{
unsigned i;
for (i = 0; i < shader->luts; i++)
{
unique_ptr<gl_core_shader::StaticTexture> image = gl_core_filter_chain_load_lut(chain, &shader->lut[i]);
if (!image)
{
RARCH_ERR("[GLCore]: Failed to load LUT \"%s\".\n", shader->lut[i].path);
return false;
}
chain->add_static_texture(move(image));
}
return true;
}
gl_core_filter_chain_t *gl_core_filter_chain_create_default(
glslang_filter_chain_filter filter)
{
struct gl_core_filter_chain_pass_info pass_info;
unique_ptr<gl_core_filter_chain> chain{ new gl_core_filter_chain(1) };
if (!chain)
return nullptr;
pass_info.scale_type_x = GLSLANG_FILTER_CHAIN_SCALE_VIEWPORT;
pass_info.scale_type_y = GLSLANG_FILTER_CHAIN_SCALE_VIEWPORT;
pass_info.scale_x = 1.0f;
pass_info.scale_y = 1.0f;
pass_info.rt_format = 0;
pass_info.source_filter = filter;
pass_info.mip_filter = GLSLANG_FILTER_CHAIN_NEAREST;
pass_info.address = GLSLANG_FILTER_CHAIN_ADDRESS_CLAMP_TO_EDGE;
pass_info.max_levels = 0;
chain->set_pass_info(0, pass_info);
chain->set_shader(0, GL_VERTEX_SHADER,
gl_core_shader::opaque_vert,
sizeof(gl_core_shader::opaque_vert) / sizeof(uint32_t));
chain->set_shader(0, GL_FRAGMENT_SHADER,
gl_core_shader::opaque_frag,
sizeof(gl_core_shader::opaque_frag) / sizeof(uint32_t));
if (!chain->init())
return nullptr;
return chain.release();
}
gl_core_filter_chain_t *gl_core_filter_chain_create_from_preset(
const char *path, glslang_filter_chain_filter filter)
{
unsigned i;
config_file_t *conf = NULL;
unique_ptr<video_shader> shader{ new video_shader() };
if (!shader)
return nullptr;
if (!(conf = video_shader_read_preset(path)))
return nullptr;
if (!video_shader_read_conf_preset(conf, shader.get()))
{
config_file_free(conf);
return nullptr;
}
bool last_pass_is_fbo = shader->pass[shader->passes - 1].fbo.valid;
unique_ptr<gl_core_filter_chain> chain{ new gl_core_filter_chain(shader->passes + (last_pass_is_fbo ? 1 : 0)) };
if (!chain)
goto error;
if (shader->luts && !gl_core_filter_chain_load_luts(chain.get(), shader.get()))
goto error;
shader->num_parameters = 0;
for (i = 0; i < shader->passes; i++)
{
glslang_output output;
struct gl_core_filter_chain_pass_info pass_info;
const video_shader_pass *pass = &shader->pass[i];
const video_shader_pass *next_pass =
i + 1 < shader->passes ? &shader->pass[i + 1] : nullptr;
pass_info.scale_type_x = GLSLANG_FILTER_CHAIN_SCALE_ORIGINAL;
pass_info.scale_type_y = GLSLANG_FILTER_CHAIN_SCALE_ORIGINAL;
pass_info.scale_x = 0.0f;
pass_info.scale_y = 0.0f;
pass_info.rt_format = 0;
pass_info.source_filter = GLSLANG_FILTER_CHAIN_LINEAR;
pass_info.mip_filter = GLSLANG_FILTER_CHAIN_LINEAR;
pass_info.address = GLSLANG_FILTER_CHAIN_ADDRESS_REPEAT;
pass_info.max_levels = 0;
if (!glslang_compile_shader(pass->source.path, &output))
{
RARCH_ERR("Failed to compile shader: \"%s\".\n",
pass->source.path);
goto error;
}
for (auto &meta_param : output.meta.parameters)
{
if (shader->num_parameters >= GFX_MAX_PARAMETERS)
{
RARCH_ERR("[GLCore]: Exceeded maximum number of parameters.\n");
goto error;
}
auto itr = find_if(shader->parameters, shader->parameters + shader->num_parameters,
[&](const video_shader_parameter &param)
{
return meta_param.id == param.id;
});
if (itr != shader->parameters + shader->num_parameters)
{
/* Allow duplicate #pragma parameter, but
* only if they are exactly the same. */
if (meta_param.desc != itr->desc ||
meta_param.initial != itr->initial ||
meta_param.minimum != itr->minimum ||
meta_param.maximum != itr->maximum ||
meta_param.step != itr->step)
{
RARCH_ERR("[GLCore]: Duplicate parameters found for \"%s\", but arguments do not match.\n",
itr->id);
goto error;
}
chain->add_parameter(i, itr - shader->parameters, meta_param.id);
}
else
{
video_shader_parameter *param = &shader->parameters[shader->num_parameters];
strlcpy(param->id, meta_param.id.c_str(), sizeof(param->id));
strlcpy(param->desc, meta_param.desc.c_str(), sizeof(param->desc));
param->current = meta_param.initial;
param->initial = meta_param.initial;
param->minimum = meta_param.minimum;
param->maximum = meta_param.maximum;
param->step = meta_param.step;
chain->add_parameter(i, shader->num_parameters, meta_param.id);
shader->num_parameters++;
}
}
chain->set_shader(i,
GL_VERTEX_SHADER,
output.vertex.data(),
output.vertex.size());
chain->set_shader(i,
GL_FRAGMENT_SHADER,
output.fragment.data(),
output.fragment.size());
chain->set_frame_count_period(i, pass->frame_count_mod);
if (!output.meta.name.empty())
chain->set_pass_name(i, output.meta.name.c_str());
/* Preset overrides. */
if (*pass->alias)
chain->set_pass_name(i, pass->alias);
if (pass->filter == RARCH_FILTER_UNSPEC)
pass_info.source_filter = filter;
else
{
pass_info.source_filter =
pass->filter == RARCH_FILTER_LINEAR
? GLSLANG_FILTER_CHAIN_LINEAR
: GLSLANG_FILTER_CHAIN_NEAREST;
}
pass_info.address = rarch_wrap_to_address(pass->wrap);
pass_info.max_levels = 1;
/* TODO: Expose max_levels in slangp.
* CGP format is a bit awkward in that it uses mipmap_input,
* so we much check if next pass needs the mipmapping.
*/
if (next_pass && next_pass->mipmap)
pass_info.max_levels = ~0u;
pass_info.mip_filter = pass->filter != RARCH_FILTER_NEAREST && pass_info.max_levels > 1
? GLSLANG_FILTER_CHAIN_LINEAR
: GLSLANG_FILTER_CHAIN_NEAREST;
bool explicit_format = output.meta.rt_format != SLANG_FORMAT_UNKNOWN;
/* Set a reasonable default. */
if (output.meta.rt_format == SLANG_FORMAT_UNKNOWN)
output.meta.rt_format = SLANG_FORMAT_R8G8B8A8_UNORM;
if (!pass->fbo.valid)
{
pass_info.scale_type_x = i + 1 == shader->passes
? GLSLANG_FILTER_CHAIN_SCALE_VIEWPORT
: GLSLANG_FILTER_CHAIN_SCALE_SOURCE;
pass_info.scale_type_y = i + 1 == shader->passes
? GLSLANG_FILTER_CHAIN_SCALE_VIEWPORT
: GLSLANG_FILTER_CHAIN_SCALE_SOURCE;
pass_info.scale_x = 1.0f;
pass_info.scale_y = 1.0f;
if (i + 1 == shader->passes)
{
pass_info.rt_format = 0;
if (explicit_format)
RARCH_WARN("[slang]: Using explicit format for last pass in chain,"
" but it is not rendered to framebuffer, using swapchain format instead.\n");
}
else
{
pass_info.rt_format = gl_core_shader::convert_glslang_format(output.meta.rt_format);
RARCH_LOG("[slang]: Using render target format %s for pass output #%u.\n",
glslang_format_to_string(output.meta.rt_format), i);
}
}
else
{
/* Preset overrides shader.
* Kinda ugly ... */
if (pass->fbo.srgb_fbo)
output.meta.rt_format = SLANG_FORMAT_R8G8B8A8_SRGB;
else if (pass->fbo.fp_fbo)
output.meta.rt_format = SLANG_FORMAT_R16G16B16A16_SFLOAT;
pass_info.rt_format = gl_core_shader::convert_glslang_format(output.meta.rt_format);
RARCH_LOG("[slang]: Using render target format %s for pass output #%u.\n",
glslang_format_to_string(output.meta.rt_format), i);
switch (pass->fbo.type_x)
{
case RARCH_SCALE_INPUT:
pass_info.scale_x = pass->fbo.scale_x;
pass_info.scale_type_x = GLSLANG_FILTER_CHAIN_SCALE_SOURCE;
break;
case RARCH_SCALE_ABSOLUTE:
pass_info.scale_x = float(pass->fbo.abs_x);
pass_info.scale_type_x = GLSLANG_FILTER_CHAIN_SCALE_ABSOLUTE;
break;
case RARCH_SCALE_VIEWPORT:
pass_info.scale_x = pass->fbo.scale_x;
pass_info.scale_type_x = GLSLANG_FILTER_CHAIN_SCALE_VIEWPORT;
break;
}
switch (pass->fbo.type_y)
{
case RARCH_SCALE_INPUT:
pass_info.scale_y = pass->fbo.scale_y;
pass_info.scale_type_y = GLSLANG_FILTER_CHAIN_SCALE_SOURCE;
break;
case RARCH_SCALE_ABSOLUTE:
pass_info.scale_y = float(pass->fbo.abs_y);
pass_info.scale_type_y = GLSLANG_FILTER_CHAIN_SCALE_ABSOLUTE;
break;
case RARCH_SCALE_VIEWPORT:
pass_info.scale_y = pass->fbo.scale_y;
pass_info.scale_type_y = GLSLANG_FILTER_CHAIN_SCALE_VIEWPORT;
break;
}
}
chain->set_pass_info(i, pass_info);
}
if (last_pass_is_fbo)
{
struct gl_core_filter_chain_pass_info pass_info;
pass_info.scale_type_x = GLSLANG_FILTER_CHAIN_SCALE_VIEWPORT;
pass_info.scale_type_y = GLSLANG_FILTER_CHAIN_SCALE_VIEWPORT;
pass_info.scale_x = 1.0f;
pass_info.scale_y = 1.0f;
pass_info.rt_format = 0;
pass_info.source_filter = filter;
pass_info.mip_filter = GLSLANG_FILTER_CHAIN_NEAREST;
pass_info.address = GLSLANG_FILTER_CHAIN_ADDRESS_CLAMP_TO_EDGE;
pass_info.max_levels = 0;
chain->set_pass_info(shader->passes, pass_info);
chain->set_shader(shader->passes,
GL_VERTEX_SHADER,
gl_core_shader::opaque_vert,
sizeof(gl_core_shader::opaque_vert) / sizeof(uint32_t));
chain->set_shader(shader->passes,
GL_FRAGMENT_SHADER,
gl_core_shader::opaque_frag,
sizeof(gl_core_shader::opaque_frag) / sizeof(uint32_t));
}
if (!video_shader_resolve_current_parameters(conf, shader.get()))
goto error;
chain->set_shader_preset(move(shader));
if (!chain->init())
goto error;
config_file_free(conf);
return chain.release();
error:
config_file_free(conf);
return nullptr;
}
struct video_shader *gl_core_filter_chain_get_preset(
gl_core_filter_chain_t *chain)
{
return chain->get_shader_preset();
}
void gl_core_filter_chain_free(
gl_core_filter_chain_t *chain)
{
delete chain;
}
void gl_core_filter_chain_set_shader(
gl_core_filter_chain_t *chain,
unsigned pass,
GLenum shader_stage,
const uint32_t *spirv,
size_t spirv_words)
{
chain->set_shader(pass, shader_stage, spirv, spirv_words);
}
void gl_core_filter_chain_set_pass_info(
gl_core_filter_chain_t *chain,
unsigned pass,
const struct gl_core_filter_chain_pass_info *info)
{
chain->set_pass_info(pass, *info);
}
bool gl_core_filter_chain_init(gl_core_filter_chain_t *chain)
{
return chain->init();
}
void gl_core_filter_chain_set_input_texture(
gl_core_filter_chain_t *chain,
const struct gl_core_filter_chain_texture *texture)
{
chain->set_input_texture(*texture);
}
void gl_core_filter_chain_set_frame_count(
gl_core_filter_chain_t *chain,
uint64_t count)
{
chain->set_frame_count(count);
}
void gl_core_filter_chain_set_frame_direction(
gl_core_filter_chain_t *chain,
int32_t direction)
{
chain->set_frame_direction(direction);
}
void gl_core_filter_chain_set_frame_count_period(
gl_core_filter_chain_t *chain,
unsigned pass,
unsigned period)
{
chain->set_frame_count_period(pass, period);
}
void gl_core_filter_chain_set_pass_name(
gl_core_filter_chain_t *chain,
unsigned pass,
const char *name)
{
chain->set_pass_name(pass, name);
}
void gl_core_filter_chain_build_offscreen_passes(
gl_core_filter_chain_t *chain,
const gl_core_viewport *vp)
{
chain->build_offscreen_passes(*vp);
}
void gl_core_filter_chain_build_viewport_pass(
gl_core_filter_chain_t *chain,
const gl_core_viewport *vp, const float *mvp)
{
chain->build_viewport_pass(*vp, mvp);
}
void gl_core_filter_chain_end_frame(gl_core_filter_chain_t *chain)
{
chain->end_frame();
}