video_core/host_shaders: Add CMake integration for string shaders

This commit is contained in:
ReinUsesLisp
2022-12-22 17:52:27 +02:00
committed by GPUCode
parent b8583f9af3
commit f7cb308243
37 changed files with 670 additions and 494 deletions

View File

@ -1,3 +1,5 @@
add_subdirectory(host_shaders)
add_library(video_core STATIC
command_processor.cpp
command_processor.h
@ -157,36 +159,6 @@ add_library(video_core STATIC
video_core.h
)
set(SHADER_FILES
renderer_opengl/depth_to_color.frag
renderer_opengl/depth_to_color.vert
renderer_opengl/ds_to_color.frag
renderer_opengl/texture_filters/anime4k/refine.frag
renderer_opengl/texture_filters/anime4k/x_gradient.frag
renderer_opengl/texture_filters/anime4k/y_gradient.frag
renderer_opengl/texture_filters/bicubic/bicubic.frag
renderer_opengl/texture_filters/nearest_neighbor/nearest_neighbor.frag
renderer_opengl/texture_filters/scale_force/scale_force.frag
renderer_opengl/texture_filters/tex_coord.vert
renderer_opengl/texture_filters/xbrz/xbrz_freescale.frag
renderer_opengl/texture_filters/xbrz/xbrz_freescale.vert
)
include(${CMAKE_CURRENT_SOURCE_DIR}/generate_shaders.cmake)
foreach(shader_file ${SHADER_FILES})
get_filename_component(shader_file_name ${shader_file} NAME)
GetShaderHeaderFile(${shader_file_name})
list(APPEND SHADER_HEADERS ${shader_header_file})
endforeach()
add_custom_target(shaders
BYPRODUCTS ${SHADER_HEADERS}
COMMAND "${CMAKE_COMMAND}" -P ${CMAKE_CURRENT_SOURCE_DIR}/generate_shaders.cmake
SOURCES ${SHADER_FILES}
)
add_dependencies(video_core shaders)
target_include_directories(video_core PRIVATE ${CMAKE_CURRENT_BINARY_DIR})
if(ARCHITECTURE_x86_64)
@ -202,16 +174,19 @@ endif()
create_target_directory_groups(video_core)
# Ignore nullability warnings generated from VMA
if (NOT MSVC)
# Ignore nullability warnings generated from VMA
target_compile_options(vma INTERFACE -Wno-unused-variable -Wno-nullability-completeness)
endif()
target_link_libraries(video_core PUBLIC common core)
target_link_libraries(video_core PRIVATE nihstro-headers Boost::serialization glm::glm)
target_link_libraries(video_core PRIVATE vulkan-headers vma sirit SPIRV glslang glad)
set_target_properties(video_core PROPERTIES INTERPROCEDURAL_OPTIMIZATION ${ENABLE_LTO})
add_dependencies(video_core host_shaders)
target_include_directories(video_core PRIVATE ${HOST_SHADERS_INCLUDE})
target_link_libraries(video_core PRIVATE vulkan-headers vma sirit SPIRV glslang glad)
target_link_libraries(video_core PRIVATE nihstro-headers Boost::serialization glm::glm)
if (ARCHITECTURE_x86_64)
target_link_libraries(video_core PUBLIC xbyak)
endif()

View File

@ -0,0 +1,100 @@
# Copyright 2022 Citra Emulator Project
# Licensed under GPLv2 or any later version
# Refer to the license.txt file included.
set(SHADER_FILES
texture_filtering/bicubic.frag
texture_filtering/nearest_neighbor.frag
texture_filtering/refine.frag
texture_filtering/scale_force.frag
texture_filtering/tex_coord.vert
texture_filtering/xbrz_freescale.frag
texture_filtering/xbrz_freescale.vert
texture_filtering/x_gradient.frag
texture_filtering/y_gradient.frag
opengl_present.frag
opengl_present.vert
opengl_present_anaglyph.frag
opengl_present_interlaced.frag
vulkan_present.frag
vulkan_present.vert
vulkan_present_anaglyph.frag
vulkan_present_interlaced.frag
)
find_program(GLSLANGVALIDATOR "glslangValidator")
if ("${GLSLANGVALIDATOR}" STREQUAL "GLSLANGVALIDATOR-NOTFOUND")
message(FATAL_ERROR "Required program `glslangValidator` not found.")
endif()
set(MACROS "-Dgl_VertexID=gl_VertexIndex")
set(QUIET_FLAG "--quiet")
set(SHADER_INCLUDE ${CMAKE_CURRENT_BINARY_DIR}/include)
set(SHADER_DIR ${SHADER_INCLUDE}/video_core/host_shaders)
set(HOST_SHADERS_INCLUDE ${SHADER_INCLUDE} PARENT_SCOPE)
set(INPUT_FILE ${CMAKE_CURRENT_SOURCE_DIR}/source_shader.h.in)
set(HEADER_GENERATOR ${CMAKE_CURRENT_SOURCE_DIR}/StringShaderHeader.cmake)
# Check if `--quiet` is available on host's glslangValidator version
# glslangValidator prints to STDERR iff an unrecognized flag is passed to it
execute_process(
COMMAND
${GLSLANGVALIDATOR} ${QUIET_FLAG}
ERROR_VARIABLE
GLSLANG_ERROR
# STDOUT variable defined to silence unnecessary output during CMake configuration
OUTPUT_VARIABLE
GLSLANG_OUTPUT
)
if (NOT GLSLANG_ERROR STREQUAL "")
message(WARNING "Refusing to use unavailable flag `${QUIET_FLAG}` on `${GLSLANGVALIDATOR}`")
set(QUIET_FLAG "")
endif()
foreach(FILENAME IN ITEMS ${SHADER_FILES})
string(REPLACE "." "_" SHADER_NAME ${FILENAME})
set(SOURCE_FILE ${CMAKE_CURRENT_SOURCE_DIR}/${FILENAME})
# Skip generating source headers on Vulkan exclusive files
if (NOT ${FILENAME} MATCHES "vulkan.*")
set(SOURCE_HEADER_FILE ${SHADER_DIR}/${SHADER_NAME}.h)
add_custom_command(
OUTPUT
${SOURCE_HEADER_FILE}
COMMAND
${CMAKE_COMMAND} -P ${HEADER_GENERATOR} ${SOURCE_FILE} ${SOURCE_HEADER_FILE} ${INPUT_FILE}
MAIN_DEPENDENCY
${SOURCE_FILE}
DEPENDS
${INPUT_FILE}
# HEADER_GENERATOR should be included here but msbuild seems to assume it's always modified
)
set(SHADER_HEADERS ${SHADER_HEADERS} ${SOURCE_HEADER_FILE})
endif()
# Skip compiling to SPIR-V OpenGL exclusive files
if (NOT ${FILENAME} MATCHES "opengl.*")
string(TOUPPER ${SHADER_NAME}_SPV SPIRV_VARIABLE_NAME)
set(SPIRV_HEADER_FILE ${SHADER_DIR}/${SHADER_NAME}_spv.h)
add_custom_command(
OUTPUT
${SPIRV_HEADER_FILE}
COMMAND
${GLSLANGVALIDATOR} --target-env vulkan1.1 --glsl-version 450 ${QUIET_FLAG} ${MACROS} --variable-name ${SPIRV_VARIABLE_NAME} -o ${SPIRV_HEADER_FILE} ${SOURCE_FILE}
MAIN_DEPENDENCY
${SOURCE_FILE}
)
set(SHADER_HEADERS ${SHADER_HEADERS} ${SPIRV_HEADER_FILE})
endif()
endforeach()
set(SHADER_SOURCES ${SHADER_FILES})
list(APPEND SHADER_SOURCES ${GLSL_INCLUDES})
add_custom_target(host_shaders
DEPENDS
${SHADER_HEADERS}
SOURCES
${SHADER_SOURCES}
)

View File

@ -0,0 +1,36 @@
# SPDX-FileCopyrightText: 2020 yuzu Emulator Project
# SPDX-License-Identifier: GPL-2.0-or-later
set(SOURCE_FILE ${CMAKE_ARGV3})
set(HEADER_FILE ${CMAKE_ARGV4})
set(INPUT_FILE ${CMAKE_ARGV5})
get_filename_component(CONTENTS_NAME ${SOURCE_FILE} NAME)
string(REPLACE "." "_" CONTENTS_NAME ${CONTENTS_NAME})
string(TOUPPER ${CONTENTS_NAME} CONTENTS_NAME)
FILE(READ ${SOURCE_FILE} line_contents)
# Replace double quotes with single quotes,
# as double quotes will be used to wrap the lines
STRING(REGEX REPLACE "\"" "'" line_contents "${line_contents}")
# CMake separates list elements with semicolons, but semicolons
# are used extensively in the shader code.
# Replace with a temporary marker, to be reverted later.
STRING(REGEX REPLACE ";" "{{SEMICOLON}}" line_contents "${line_contents}")
# Make every line an individual element in the CMake list.
STRING(REGEX REPLACE "\n" ";" line_contents "${line_contents}")
# Build the shader string, wrapping each line in double quotes.
foreach(line IN LISTS line_contents)
string(CONCAT CONTENTS "${CONTENTS}" \"${line}\\n\"\n)
endforeach()
# Revert the original semicolons in the source.
STRING(REGEX REPLACE "{{SEMICOLON}}" ";" CONTENTS "${CONTENTS}")
get_filename_component(OUTPUT_DIR ${HEADER_FILE} DIRECTORY)
make_directory(${OUTPUT_DIR})
configure_file(${INPUT_FILE} ${HEADER_FILE} @ONLY)

View File

@ -0,0 +1,18 @@
// Copyright 2022 Citra Emulator Project
// Licensed under GPLv2 or any later version
// Refer to the license.txt file included.
//? #version 430 core
layout(location = 0) in vec2 frag_tex_coord;
layout(location = 0) out vec4 color;
layout(binding = 0) uniform sampler2D color_texture;
uniform vec4 i_resolution;
uniform vec4 o_resolution;
uniform int layer;
void main() {
color = texture(color_texture, frag_tex_coord);
}

View File

@ -0,0 +1,23 @@
// Copyright 2022 Citra Emulator Project
// Licensed under GPLv2 or any later version
// Refer to the license.txt file included.
//? #version 430 core
layout(location = 0) in vec2 vert_position;
layout(location = 1) in vec2 vert_tex_coord;
layout(location = 0) out vec2 frag_tex_coord;
// This is a truncated 3x3 matrix for 2D transformations:
// The upper-left 2x2 submatrix performs scaling/rotation/mirroring.
// The third column performs translation.
// The third row could be used for projection, which we don't need in 2D. It hence is assumed to
// implicitly be [0, 0, 1]
uniform mat3x2 modelview_matrix;
void main() {
// Multiply input position by the rotscale part of the matrix and then manually translate by
// the last column. This is equivalent to using a full 3x3 matrix and expanding the vector
// to `vec3(vert_position.xy, 1.0)`
gl_Position = vec4(mat2(modelview_matrix) * vert_position + modelview_matrix[2], 0.0, 1.0);
frag_tex_coord = vert_tex_coord;
}

View File

@ -0,0 +1,32 @@
// Copyright 2022 Citra Emulator Project
// Licensed under GPLv2 or any later version
// Refer to the license.txt file included.
//? #version 430 core
// Anaglyph Red-Cyan shader based on Dubois algorithm
// Constants taken from the paper:
// "Conversion of a Stereo Pair to Anaglyph with
// the Least-Squares Projection Method"
// Eric Dubois, March 2009
const mat3 l = mat3( 0.437, 0.449, 0.164,
-0.062,-0.062,-0.024,
-0.048,-0.050,-0.017);
const mat3 r = mat3(-0.011,-0.032,-0.007,
0.377, 0.761, 0.009,
-0.026,-0.093, 1.234);
layout(location = 0) in vec2 frag_tex_coord;
layout(location = 0) out vec4 color;
layout(binding = 0) uniform sampler2D color_texture;
layout(binding = 1) uniform sampler2D color_texture_r;
uniform vec4 resolution;
uniform int layer;
void main() {
vec4 color_tex_l = texture(color_texture, frag_tex_coord);
vec4 color_tex_r = texture(color_texture_r, frag_tex_coord);
color = vec4(color_tex_l.rgb*l+color_tex_r.rgb*r, color_tex_l.a);
}

View File

@ -0,0 +1,22 @@
// Copyright 2022 Citra Emulator Project
// Licensed under GPLv2 or any later version
// Refer to the license.txt file included.
//? #version 430 core
layout(location = 0) in vec2 frag_tex_coord;
layout(location = 0) out vec4 color;
layout(binding = 0) uniform sampler2D color_texture;
layout(binding = 1) uniform sampler2D color_texture_r;
uniform vec4 o_resolution;
uniform int reverse_interlaced;
void main() {
float screen_row = o_resolution.x * frag_tex_coord.x;
if (int(screen_row) % 2 == reverse_interlaced)
color = texture(color_texture, frag_tex_coord);
else
color = texture(color_texture_r, frag_tex_coord);
}

View File

@ -0,0 +1,15 @@
// Copyright 2022 Citra Emulator Project
// Licensed under GPLv2 or any later version
// Refer to the license.txt file included.
#pragma once
#include <string_view>
namespace HostShaders {
constexpr std::string_view @CONTENTS_NAME@ = {
@CONTENTS@
};
} // namespace HostShaders

View File

@ -1,11 +1,14 @@
// Copyright 2022 Citra Emulator Project
// Licensed under GPLv2 or any later version
// Refer to the license.txt file included.
//? #version 330
precision mediump float;
in vec2 tex_coord;
layout(location = 0) in vec2 tex_coord;
layout(location = 0) out vec4 frag_color;
out vec4 frag_color;
uniform sampler2D input_texture;
layout(binding = 0) uniform sampler2D input_texture;
// from http://www.java-gaming.org/index.php?topic=35123.0
vec4 cubic(float v) {
@ -18,9 +21,9 @@ vec4 cubic(float v) {
return vec4(x, y, z, w) * (1.0 / 6.0);
}
vec4 textureBicubic(sampler2D sampler, vec2 texCoords) {
vec4 textureBicubic(sampler2D tex_sampler, vec2 texCoords) {
vec2 texSize = vec2(textureSize(sampler, 0));
vec2 texSize = vec2(textureSize(tex_sampler, 0));
vec2 invTexSize = 1.0 / texSize;
texCoords = texCoords * texSize - 0.5;
@ -38,10 +41,10 @@ vec4 textureBicubic(sampler2D sampler, vec2 texCoords) {
offset *= invTexSize.xxyy;
vec4 sample0 = texture(sampler, offset.xz);
vec4 sample1 = texture(sampler, offset.yz);
vec4 sample2 = texture(sampler, offset.xw);
vec4 sample3 = texture(sampler, offset.yw);
vec4 sample0 = texture(tex_sampler, offset.xz);
vec4 sample1 = texture(tex_sampler, offset.yz);
vec4 sample2 = texture(tex_sampler, offset.xw);
vec4 sample3 = texture(tex_sampler, offset.yw);
float sx = s.x / (s.x + s.y);
float sy = s.z / (s.z + s.w);

View File

@ -0,0 +1,15 @@
// Copyright 2022 Citra Emulator Project
// Licensed under GPLv2 or any later version
// Refer to the license.txt file included.
//? #version 430 core
precision mediump float;
layout(location = 0) in vec2 tex_coord;
layout(location = 0) out vec4 frag_color;
layout(binding = 0) uniform sampler2D input_texture;
void main() {
frag_color = texture(input_texture, tex_coord);
}

View File

@ -1,12 +1,15 @@
//? #version 330
// Copyright 2022 Citra Emulator Project
// Licensed under GPLv2 or any later version
// Refer to the license.txt file included.
//? #version 430 core
precision mediump float;
in vec2 tex_coord;
layout(location = 0) in vec2 tex_coord;
layout(location = 0) out vec4 frag_color;
out vec4 frag_color;
uniform sampler2D HOOKED;
uniform sampler2D LUMAD;
layout(binding = 0) uniform sampler2D HOOKED;
layout(binding = 1) uniform sampler2D LUMAD;
const float LINE_DETECT_THRESHOLD = 0.4;
const float STRENGTH = 0.6;

View File

@ -1,5 +1,3 @@
//? #version 320 es
// from https://github.com/BreadFish64/ScaleFish/tree/master/scale_force
// MIT License
@ -24,13 +22,18 @@
// OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
// SOFTWARE.
// Copyright 2022 Citra Emulator Project
// Licensed under GPLv2 or any later version
// Refer to the license.txt file included.
//? #version 320 es
precision mediump float;
in vec2 tex_coord;
layout(location = 0) in vec2 tex_coord;
layout(location = 0) out vec4 frag_color;
out vec4 frag_color;
uniform sampler2D input_texture;
layout(binding = 0) uniform sampler2D input_texture;
vec2 tex_size;
vec2 inv_tex_size;

View File

@ -1,5 +1,9 @@
//? #version 330
out vec2 tex_coord;
// Copyright 2022 Citra Emulator Project
// Licensed under GPLv2 or any later version
// Refer to the license.txt file included.
//? #version 430 core
layout(location = 0) out vec2 tex_coord;
const vec2 vertices[4] =
vec2[4](vec2(-1.0, -1.0), vec2(1.0, -1.0), vec2(-1.0, 1.0), vec2(1.0, 1.0));

View File

@ -1,11 +1,14 @@
//? #version 330
// Copyright 2022 Citra Emulator Project
// Licensed under GPLv2 or any later version
// Refer to the license.txt file included.
//? #version 430 core
precision mediump float;
in vec2 tex_coord;
layout(location = 0) in vec2 tex_coord;
layout(location = 0) out vec2 frag_color;
out vec2 frag_color;
uniform sampler2D tex_input;
layout(binding = 0) uniform sampler2D tex_input;
const vec3 K = vec3(0.2627, 0.6780, 0.0593);
// TODO: improve handling of alpha channel

View File

@ -1,14 +1,25 @@
//? #version 330
// Copyright 2022 Citra Emulator Project
// Licensed under GPLv2 or any later version
// Refer to the license.txt file included.
//? #version 430 core
precision mediump float;
in vec2 tex_coord;
in vec2 source_size;
in vec2 output_size;
layout(location = 0) in vec2 tex_coord;
layout(location = 1) in vec2 source_size;
layout(location = 2) in vec2 output_size;
out vec4 frag_color;
layout(location = 0) out vec4 frag_color;
uniform sampler2D tex;
layout(binding = 0) uniform sampler2D tex;
#ifdef VULKAN
layout(push_constant, std140) uniform XbrzInfo {
lowp float scale;
};
#else
uniform lowp float scale;
#endif
const int BLEND_NONE = 0;
const int BLEND_NORMAL = 1;

View File

@ -0,0 +1,28 @@
// Copyright 2022 Citra Emulator Project
// Licensed under GPLv2 or any later version
// Refer to the license.txt file included.
//? #version 430 core
layout(location = 0) out vec2 tex_coord;
layout(location = 1) out vec2 source_size;
layout(location = 2) out vec2 output_size;
layout(binding = 0) uniform sampler2D tex;
#ifdef VULKAN
layout(push_constant, std140) uniform XbrzInfo {
lowp float scale;
};
#else
uniform lowp float scale;
#endif
const vec2 vertices[4] =
vec2[4](vec2(-1.0, -1.0), vec2(1.0, -1.0), vec2(-1.0, 1.0), vec2(1.0, 1.0));
void main() {
gl_Position = vec4(vertices[gl_VertexID], 0.0, 1.0);
tex_coord = (vertices[gl_VertexID] + 1.0) / 2.0;
source_size = vec2(textureSize(tex, 0));
output_size = source_size * scale;
}

View File

@ -1,11 +1,14 @@
//? #version 330
// Copyright 2022 Citra Emulator Project
// Licensed under GPLv2 or any later version
// Refer to the license.txt file included.
//? #version 430 core
precision mediump float;
in vec2 tex_coord;
layout(location = 0) in vec2 tex_coord;
layout(location = 0) out float frag_color;
out float frag_color;
uniform sampler2D tex_input;
layout(binding = 0) uniform sampler2D tex_input;
void main() {
vec2 t = textureLodOffset(tex_input, tex_coord, 0.0, ivec2(0, 1)).xy;

View File

@ -0,0 +1,26 @@
// Copyright 2022 Citra Emulator Project
// Licensed under GPLv2 or any later version
// Refer to the license.txt file included.
#version 450 core
#extension GL_ARB_separate_shader_objects : enable
layout (location = 0) in vec2 frag_tex_coord;
layout (location = 0) out vec4 color;
layout (push_constant, std140) uniform DrawInfo {
mat4 modelview_matrix;
vec4 i_resolution;
vec4 o_resolution;
int screen_id_l;
int screen_id_r;
int layer;
int reverse_interlaced;
};
layout (set = 0, binding = 0) uniform texture2D screen_textures[3];
layout (set = 0, binding = 1) uniform sampler screen_sampler;
void main() {
color = texture(sampler2D(screen_textures[screen_id_l], screen_sampler), frag_tex_coord);
}

View File

@ -0,0 +1,25 @@
// Copyright 2022 Citra Emulator Project
// Licensed under GPLv2 or any later version
// Refer to the license.txt file included.
#version 450 core
#extension GL_ARB_separate_shader_objects : enable
layout (location = 0) in vec2 vert_position;
layout (location = 1) in vec2 vert_tex_coord;
layout (location = 0) out vec2 frag_tex_coord;
layout (push_constant, std140) uniform DrawInfo {
mat4 modelview_matrix;
vec4 i_resolution;
vec4 o_resolution;
int screen_id_l;
int screen_id_r;
int layer;
};
void main() {
vec4 position = vec4(vert_position, 0.0, 1.0) * modelview_matrix;
gl_Position = vec4(position.x, -position.y, 0.0, 1.0);
frag_tex_coord = vert_tex_coord;
}

View File

@ -0,0 +1,40 @@
// Copyright 2022 Citra Emulator Project
// Licensed under GPLv2 or any later version
// Refer to the license.txt file included.
#version 450 core
#extension GL_ARB_separate_shader_objects : enable
layout (location = 0) in vec2 frag_tex_coord;
layout (location = 0) out vec4 color;
// Anaglyph Red-Cyan shader based on Dubois algorithm
// Constants taken from the paper:
// "Conversion of a Stereo Pair to Anaglyph with
// the Least-Squares Projection Method"
// Eric Dubois, March 2009
const mat3 l = mat3( 0.437, 0.449, 0.164,
-0.062,-0.062,-0.024,
-0.048,-0.050,-0.017);
const mat3 r = mat3(-0.011,-0.032,-0.007,
0.377, 0.761, 0.009,
-0.026,-0.093, 1.234);
layout (push_constant, std140) uniform DrawInfo {
mat4 modelview_matrix;
vec4 i_resolution;
vec4 o_resolution;
int screen_id_l;
int screen_id_r;
int layer;
int reverse_interlaced;
};
layout (set = 0, binding = 0) uniform texture2D screen_textures[3];
layout (set = 0, binding = 1) uniform sampler screen_sampler;
void main() {
vec4 color_tex_l = texture(sampler2D(screen_textures[screen_id_l], screen_sampler), frag_tex_coord);
vec4 color_tex_r = texture(sampler2D(screen_textures[screen_id_r], screen_sampler), frag_tex_coord);
color = vec4(color_tex_l.rgb*l+color_tex_r.rgb*r, color_tex_l.a);
}

View File

@ -0,0 +1,30 @@
// Copyright 2022 Citra Emulator Project
// Licensed under GPLv2 or any later version
// Refer to the license.txt file included.
#version 450 core
#extension GL_ARB_separate_shader_objects : enable
layout (location = 0) in vec2 frag_tex_coord;
layout (location = 0) out vec4 color;
layout (push_constant, std140) uniform DrawInfo {
mat4 modelview_matrix;
vec4 i_resolution;
vec4 o_resolution;
int screen_id_l;
int screen_id_r;
int layer;
int reverse_interlaced;
};
layout (set = 0, binding = 0) uniform texture2D screen_textures[3];
layout (set = 0, binding = 1) uniform sampler screen_sampler;
void main() {
float screen_row = o_resolution.x * frag_tex_coord.x;
if (int(screen_row) % 2 == reverse_interlaced)
color = texture(sampler2D(screen_textures[screen_id_l], screen_sampler), frag_tex_coord);
else
color = texture(sampler2D(screen_textures[screen_id_r], screen_sampler), frag_tex_coord);
}

View File

@ -117,7 +117,7 @@ void OGLSampler::Release() {
handle = 0;
}
void OGLShader::Create(const char* source, GLenum type) {
void OGLShader::Create(std::string_view source, GLenum type) {
if (handle != 0)
return;
if (source == nullptr)
@ -144,7 +144,7 @@ void OGLProgram::Create(bool separable_program, const std::vector<GLuint>& shade
handle = LoadProgram(separable_program, shaders);
}
void OGLProgram::Create(const char* vert_shader, const char* frag_shader) {
void OGLProgram::Create(std::string_view vert_shader, std::string_view frag_shader) {
OGLShader vert, frag;
vert.Create(vert_shader, GL_VERTEX_SHADER);
frag.Create(frag_shader, GL_FRAGMENT_SHADER);

View File

@ -109,7 +109,7 @@ public:
return *this;
}
void Create(const char* source, GLenum type);
void Create(std::string_view source, GLenum type);
void Release();
@ -136,10 +136,10 @@ public:
void Create(bool separable_program, const std::vector<GLuint>& shaders);
/// Creates a new program from given shader soruce code
void Create(const char* vert_shader, const char* frag_shader);
void Create(std::string_view vert_shader, std::string_view frag_shader);
/// Creates a new compute shader program
void Create(const std::string_view compute_shader);
void Create(std::string_view compute_shader);
/// Deletes the internal OpenGL resource
void Release();

View File

@ -13,7 +13,7 @@
namespace OpenGL {
GLuint LoadShader(const char* source, GLenum type) {
GLuint LoadShader(std::string_view source, GLenum type) {
const std::string version = GLES ? R"(
#version 320 es
#define CITRA_GLES
@ -46,7 +46,7 @@ GLuint LoadShader(const char* source, GLenum type) {
UNREACHABLE();
}
std::array<const char*, 2> src_arr{version.data(), source};
std::array<const char*, 2> src_arr{version.data(), source.data()};
GLuint shader_id = glCreateShader(type);
glShaderSource(shader_id, static_cast<GLsizei>(src_arr.size()), src_arr.data(), nullptr);
LOG_DEBUG(Render_OpenGL, "Compiling {} shader...", debug_type);
@ -71,7 +71,7 @@ GLuint LoadShader(const char* source, GLenum type) {
return shader_id;
}
GLuint LoadProgram(bool separable_program, const std::vector<GLuint>& shaders) {
GLuint LoadProgram(bool separable_program, std::span<const GLuint> shaders) {
// Link the program
LOG_DEBUG(Render_OpenGL, "Linking program...");

View File

@ -4,7 +4,8 @@
#pragma once
#include <vector>
#include <span>
#include <string_view>
#include <glad/glad.h>
namespace OpenGL {
@ -29,7 +30,7 @@ precision mediump uimage2D;
* @param source String of the GLSL shader program
* @param type Type of the shader (GL_VERTEX_SHADER, GL_GEOMETRY_SHADER or GL_FRAGMENT_SHADER)
*/
GLuint LoadShader(const char* source, GLenum type);
GLuint LoadShader(std::string_view source, GLenum type);
/**
* Utility function to create and link an OpenGL GLSL shader program
@ -37,6 +38,6 @@ GLuint LoadShader(const char* source, GLenum type);
* @param shaders ID of shaders to attach to the program
* @returns Handle of the newly created OpenGL program object
*/
GLuint LoadProgram(bool separable_program, const std::vector<GLuint>& shaders);
GLuint LoadProgram(bool separable_program, std::span<const GLuint> shaders);
} // namespace OpenGL

View File

@ -23,6 +23,14 @@
#include "video_core/renderer_opengl/renderer_opengl.h"
#include "video_core/video_core.h"
#include "video_core/host_shaders/opengl_present_vert.h"
#include "video_core/host_shaders/opengl_present_frag.h"
#include "video_core/host_shaders/opengl_present_anaglyph_frag.h"
#include "video_core/host_shaders/opengl_present_interlaced_frag.h"
MICROPROFILE_DEFINE(OpenGL_RenderFrame, "OpenGL", "Render Frame", MP_RGB(128, 128, 64));
MICROPROFILE_DEFINE(OpenGL_WaitPresent, "OpenGL", "Wait For Present", MP_RGB(128, 128, 128));
namespace OpenGL {
// If the size of this is too small, it ends up creating a soft cap on FPS as the renderer will have
@ -220,93 +228,6 @@ public:
}
};
static const char vertex_shader[] = R"(
in vec2 vert_position;
in vec2 vert_tex_coord;
out vec2 frag_tex_coord;
// This is a truncated 3x3 matrix for 2D transformations:
// The upper-left 2x2 submatrix performs scaling/rotation/mirroring.
// The third column performs translation.
// The third row could be used for projection, which we don't need in 2D. It hence is assumed to
// implicitly be [0, 0, 1]
uniform mat3x2 modelview_matrix;
void main() {
// Multiply input position by the rotscale part of the matrix and then manually translate by
// the last column. This is equivalent to using a full 3x3 matrix and expanding the vector
// to `vec3(vert_position.xy, 1.0)`
gl_Position = vec4(mat2(modelview_matrix) * vert_position + modelview_matrix[2], 0.0, 1.0);
frag_tex_coord = vert_tex_coord;
}
)";
static const char fragment_shader[] = R"(
in vec2 frag_tex_coord;
layout(location = 0) out vec4 color;
uniform vec4 i_resolution;
uniform vec4 o_resolution;
uniform int layer;
uniform sampler2D color_texture;
void main() {
color = texture(color_texture, frag_tex_coord);
}
)";
static const char fragment_shader_anaglyph[] = R"(
// Anaglyph Red-Cyan shader based on Dubois algorithm
// Constants taken from the paper:
// "Conversion of a Stereo Pair to Anaglyph with
// the Least-Squares Projection Method"
// Eric Dubois, March 2009
const mat3 l = mat3( 0.437, 0.449, 0.164,
-0.062,-0.062,-0.024,
-0.048,-0.050,-0.017);
const mat3 r = mat3(-0.011,-0.032,-0.007,
0.377, 0.761, 0.009,
-0.026,-0.093, 1.234);
in vec2 frag_tex_coord;
out vec4 color;
uniform vec4 resolution;
uniform int layer;
uniform sampler2D color_texture;
uniform sampler2D color_texture_r;
void main() {
vec4 color_tex_l = texture(color_texture, frag_tex_coord);
vec4 color_tex_r = texture(color_texture_r, frag_tex_coord);
color = vec4(color_tex_l.rgb*l+color_tex_r.rgb*r, color_tex_l.a);
}
)";
static const char fragment_shader_interlaced[] = R"(
in vec2 frag_tex_coord;
out vec4 color;
uniform vec4 o_resolution;
uniform sampler2D color_texture;
uniform sampler2D color_texture_r;
uniform int reverse_interlaced;
void main() {
float screen_row = o_resolution.x * frag_tex_coord.x;
if (int(screen_row) % 2 == reverse_interlaced)
color = texture(color_texture, frag_tex_coord);
else
color = texture(color_texture_r, frag_tex_coord);
}
)";
/**
* Vertex structure that the drawn screen rectangles are composed of.
*/
@ -391,9 +312,6 @@ VideoCore::RasterizerInterface* RendererOpenGL::Rasterizer() {
/// Shutdown the renderer
void RendererOpenGL::ShutDown() {}
MICROPROFILE_DEFINE(OpenGL_RenderFrame, "OpenGL", "Render Frame", MP_RGB(128, 128, 64));
MICROPROFILE_DEFINE(OpenGL_WaitPresent, "OpenGL", "Wait For Present", MP_RGB(128, 128, 128));
/// Swap buffers (render frame)
void RendererOpenGL::SwapBuffers() {
// Maintain the rasterizer's state as a priority
@ -710,13 +628,13 @@ void RendererOpenGL::ReloadShader() {
if (Settings::values.render_3d.GetValue() == Settings::StereoRenderOption::Anaglyph) {
if (Settings::values.pp_shader_name.GetValue() == "dubois (builtin)") {
shader_data += fragment_shader_anaglyph;
shader_data += HostShaders::OPENGL_PRESENT_ANAGLYPH_FRAG;
} else {
std::string shader_text = OpenGL::GetPostProcessingShaderCode(
true, Settings::values.pp_shader_name.GetValue());
if (shader_text.empty()) {
// Should probably provide some information that the shader couldn't load
shader_data += fragment_shader_anaglyph;
shader_data += HostShaders::OPENGL_PRESENT_ANAGLYPH_FRAG;
} else {
shader_data += shader_text;
}
@ -725,32 +643,32 @@ void RendererOpenGL::ReloadShader() {
Settings::values.render_3d.GetValue() ==
Settings::StereoRenderOption::ReverseInterlaced) {
if (Settings::values.pp_shader_name.GetValue() == "horizontal (builtin)") {
shader_data += fragment_shader_interlaced;
shader_data += HostShaders::OPENGL_PRESENT_INTERLACED_FRAG;
} else {
std::string shader_text = OpenGL::GetPostProcessingShaderCode(
false, Settings::values.pp_shader_name.GetValue());
if (shader_text.empty()) {
// Should probably provide some information that the shader couldn't load
shader_data += fragment_shader_interlaced;
shader_data += HostShaders::OPENGL_PRESENT_INTERLACED_FRAG;
} else {
shader_data += shader_text;
}
}
} else {
if (Settings::values.pp_shader_name.GetValue() == "none (builtin)") {
shader_data += fragment_shader;
shader_data += HostShaders::OPENGL_PRESENT_FRAG;
} else {
std::string shader_text = OpenGL::GetPostProcessingShaderCode(
false, Settings::values.pp_shader_name.GetValue());
if (shader_text.empty()) {
// Should probably provide some information that the shader couldn't load
shader_data += fragment_shader;
shader_data += HostShaders::OPENGL_PRESENT_FRAG;
} else {
shader_data += shader_text;
}
}
}
shader.Create(vertex_shader, shader_data.c_str());
shader.Create(HostShaders::OPENGL_PRESENT_VERT, shader_data.c_str());
state.draw.shader_program = shader.handle;
state.Apply();
uniform_modelview_matrix = glGetUniformLocation(shader.handle, "modelview_matrix");

View File

@ -32,10 +32,10 @@
#include "video_core/renderer_opengl/texture_filters/anime4k/anime4k_ultrafast.h"
#include "shaders/refine.frag"
#include "shaders/tex_coord.vert"
#include "shaders/x_gradient.frag"
#include "shaders/y_gradient.frag"
#include "video_core/host_shaders/texture_filtering/refine_frag.h"
#include "video_core/host_shaders/texture_filtering/tex_coord_vert.h"
#include "video_core/host_shaders/texture_filtering/x_gradient_frag.h"
#include "video_core/host_shaders/texture_filtering/y_gradient_frag.h"
namespace OpenGL {
@ -56,9 +56,9 @@ Anime4kUltrafast::Anime4kUltrafast(u16 scale_factor) : TextureFilterBase(scale_f
}
state.draw.vertex_array = vao.handle;
gradient_x_program.Create(tex_coord_vert.data(), x_gradient_frag.data());
gradient_y_program.Create(tex_coord_vert.data(), y_gradient_frag.data());
refine_program.Create(tex_coord_vert.data(), refine_frag.data());
gradient_x_program.Create(HostShaders::TEX_COORD_VERT, HostShaders::X_GRADIENT_FRAG);
gradient_y_program.Create(HostShaders::TEX_COORD_VERT, HostShaders::Y_GRADIENT_FRAG);
refine_program.Create(HostShaders::TEX_COORD_VERT, HostShaders::REFINE_FRAG);
state.draw.shader_program = gradient_y_program.handle;
state.Apply();

View File

@ -4,13 +4,13 @@
#include "video_core/renderer_opengl/texture_filters/bicubic/bicubic.h"
#include "shaders/bicubic.frag"
#include "shaders/tex_coord.vert"
#include "video_core/host_shaders/texture_filtering/bicubic_frag.h"
#include "video_core/host_shaders/texture_filtering/tex_coord_vert.h"
namespace OpenGL {
Bicubic::Bicubic(u16 scale_factor) : TextureFilterBase(scale_factor) {
program.Create(tex_coord_vert.data(), bicubic_frag.data());
program.Create(HostShaders::TEX_COORD_VERT, HostShaders::BICUBIC_FRAG);
vao.Create();
src_sampler.Create();

View File

@ -4,13 +4,13 @@
#include "video_core/renderer_opengl/texture_filters/nearest_neighbor/nearest_neighbor.h"
#include "shaders/nearest_neighbor.frag"
#include "shaders/tex_coord.vert"
#include "video_core/host_shaders/texture_filtering/nearest_neighbor_frag.h"
#include "video_core/host_shaders/texture_filtering/tex_coord_vert.h"
namespace OpenGL {
NearestNeighbor::NearestNeighbor(u16 scale_factor) : TextureFilterBase(scale_factor) {
program.Create(tex_coord_vert.data(), nearest_neighbor_frag.data());
program.Create(HostShaders::TEX_COORD_VERT, HostShaders::NEAREST_NEIGHBOR_FRAG);
vao.Create();
src_sampler.Create();

View File

@ -1,12 +0,0 @@
//? #version 330
precision mediump float;
in vec2 tex_coord;
out vec4 frag_color;
uniform sampler2D input_texture;
void main() {
frag_color = texture(input_texture, tex_coord);
}

View File

@ -4,13 +4,13 @@
#include "video_core/renderer_opengl/texture_filters/scale_force/scale_force.h"
#include "shaders/scale_force.frag"
#include "shaders/tex_coord.vert"
#include "video_core/host_shaders/texture_filtering/scale_force_frag.h"
#include "video_core/host_shaders/texture_filtering/tex_coord_vert.h"
namespace OpenGL {
ScaleForce::ScaleForce(u16 scale_factor) : TextureFilterBase(scale_factor) {
program.Create(tex_coord_vert.data(), scale_force_frag.data());
program.Create(HostShaders::TEX_COORD_VERT, HostShaders::SCALE_FORCE_FRAG);
vao.Create();
src_sampler.Create();

View File

@ -42,8 +42,8 @@
#include "video_core/renderer_opengl/texture_filters/xbrz/xbrz_freescale.h"
#include "shaders/xbrz_freescale.frag"
#include "shaders/xbrz_freescale.vert"
#include "video_core/host_shaders/texture_filtering/xbrz_freescale_frag.h"
#include "video_core/host_shaders/texture_filtering/xbrz_freescale_vert.h"
namespace OpenGL {
@ -51,7 +51,7 @@ XbrzFreescale::XbrzFreescale(u16 scale_factor) : TextureFilterBase(scale_factor)
const OpenGLState cur_state = OpenGLState::GetCurState();
program.Create(xbrz_freescale_vert.data(), xbrz_freescale_frag.data());
program.Create(HostShaders::XBRZ_FREESCALE_VERT, HostShaders::XBRZ_FREESCALE_FRAG);
vao.Create();
src_sampler.Create();

View File

@ -1,17 +0,0 @@
//? #version 330
out vec2 tex_coord;
out vec2 source_size;
out vec2 output_size;
uniform sampler2D tex;
uniform lowp float scale;
const vec2 vertices[4] =
vec2[4](vec2(-1.0, -1.0), vec2(1.0, -1.0), vec2(-1.0, 1.0), vec2(1.0, 1.0));
void main() {
gl_Position = vec4(vertices[gl_VertexID], 0.0, 1.0);
tex_coord = (vertices[gl_VertexID] + 1.0) / 2.0;
source_size = vec2(textureSize(tex, 0));
output_size = source_size * scale;
}

View File

@ -19,127 +19,16 @@
#include "video_core/renderer_vulkan/vk_shader_util.h"
#include "video_core/video_core.h"
#include "video_core/host_shaders/vulkan_present_vert_spv.h"
#include "video_core/host_shaders/vulkan_present_frag_spv.h"
#include "video_core/host_shaders/vulkan_present_anaglyph_frag_spv.h"
#include "video_core/host_shaders/vulkan_present_interlaced_frag_spv.h"
namespace Vulkan {
constexpr std::string_view vertex_shader = R"(
#version 450 core
#extension GL_ARB_separate_shader_objects : enable
layout (location = 0) in vec2 vert_position;
layout (location = 1) in vec2 vert_tex_coord;
layout (location = 0) out vec2 frag_tex_coord;
// This is a truncated 3x3 matrix for 2D transformations:
// The upper-left 2x2 submatrix performs scaling/rotation/mirroring.
// The third column performs translation.
// The third row could be used for projection, which we don't need in 2D. It hence is assumed to
// implicitly be [0, 0, 1]
layout (push_constant, std140) uniform DrawInfo {
mat4 modelview_matrix;
vec4 i_resolution;
vec4 o_resolution;
int screen_id_l;
int screen_id_r;
int layer;
};
void main() {
vec4 position = vec4(vert_position, 0.0, 1.0) * modelview_matrix;
gl_Position = vec4(position.x, -position.y, 0.0, 1.0);
frag_tex_coord = vert_tex_coord;
}
)";
constexpr std::string_view fragment_shader = R"(
#version 450 core
#extension GL_ARB_separate_shader_objects : enable
layout (location = 0) in vec2 frag_tex_coord;
layout (location = 0) out vec4 color;
layout (push_constant, std140) uniform DrawInfo {
mat4 modelview_matrix;
vec4 i_resolution;
vec4 o_resolution;
int screen_id_l;
int screen_id_r;
int layer;
int reverse_interlaced;
};
layout (set = 0, binding = 0) uniform texture2D screen_textures[3];
layout (set = 0, binding = 1) uniform sampler screen_sampler;
void main() {
color = texture(sampler2D(screen_textures[screen_id_l], screen_sampler), frag_tex_coord);
}
)";
constexpr std::string_view fragment_shader_anaglyph = R"(
#version 450 core
#extension GL_ARB_separate_shader_objects : enable
layout (location = 0) in vec2 frag_tex_coord;
layout (location = 0) out vec4 color;
// Anaglyph Red-Cyan shader based on Dubois algorithm
// Constants taken from the paper:
// "Conversion of a Stereo Pair to Anaglyph with
// the Least-Squares Projection Method"
// Eric Dubois, March 2009
const mat3 l = mat3( 0.437, 0.449, 0.164,
-0.062,-0.062,-0.024,
-0.048,-0.050,-0.017);
const mat3 r = mat3(-0.011,-0.032,-0.007,
0.377, 0.761, 0.009,
-0.026,-0.093, 1.234);
layout (push_constant, std140) uniform DrawInfo {
mat4 modelview_matrix;
vec4 i_resolution;
vec4 o_resolution;
int screen_id_l;
int screen_id_r;
int layer;
int reverse_interlaced;
};
layout (set = 0, binding = 0) uniform texture2D screen_textures[3];
layout (set = 0, binding = 1) uniform sampler screen_sampler;
void main() {
vec4 color_tex_l = texture(sampler2D(screen_textures[screen_id_l], screen_sampler), frag_tex_coord);
vec4 color_tex_r = texture(sampler2D(screen_textures[screen_id_r], screen_sampler), frag_tex_coord);
color = vec4(color_tex_l.rgb*l+color_tex_r.rgb*r, color_tex_l.a);
}
)";
constexpr std::string_view fragment_shader_interlaced = R"(
#version 450 core
#extension GL_ARB_separate_shader_objects : enable
layout (location = 0) in vec2 frag_tex_coord;
layout (location = 0) out vec4 color;
layout (push_constant, std140) uniform DrawInfo {
mat4 modelview_matrix;
vec4 i_resolution;
vec4 o_resolution;
int screen_id_l;
int screen_id_r;
int layer;
int reverse_interlaced;
};
layout (set = 0, binding = 0) uniform texture2D screen_textures[3];
layout (set = 0, binding = 1) uniform sampler screen_sampler;
void main() {
float screen_row = o_resolution.x * frag_tex_coord.x;
if (int(screen_row) % 2 == reverse_interlaced)
color = texture(sampler2D(screen_textures[screen_id_l], screen_sampler), frag_tex_coord);
else
color = texture(sampler2D(screen_textures[screen_id_r], screen_sampler), frag_tex_coord);
}
)";
/// Vertex structure that the drawn screen rectangles are composed of.
/**
* Vertex structure that the drawn screen rectangles are composed of.
*/
struct ScreenRectVertex {
ScreenRectVertex() = default;
ScreenRectVertex(float x, float y, float u, float v)
@ -151,33 +40,13 @@ struct ScreenRectVertex {
constexpr u32 VERTEX_BUFFER_SIZE = sizeof(ScreenRectVertex) * 8192;
/**
* Defines a 1:1 pixel ortographic projection matrix with (0,0) on the top-left
* corner and (width, height) on the lower-bottom.
*
* The projection part of the matrix is trivial, hence these operations are represented
* by a 3x2 matrix.
*
* @param flipped Whether the frame should be flipped upside down.
*/
static std::array<float, 3 * 2> MakeOrthographicMatrix(float width, float height, bool flipped) {
std::array<float, 3 * 2> matrix; // Laid out in column-major order
// Last matrix row is implicitly assumed to be [0, 0, 1].
if (flipped) {
// clang-format off
matrix[0] = 2.f / width; matrix[2] = 0.f; matrix[4] = -1.f;
matrix[1] = 0.f; matrix[3] = 2.f / height; matrix[5] = -1.f;
// clang-format on
} else {
// clang-format off
matrix[0] = 2.f / width; matrix[2] = 0.f; matrix[4] = -1.f;
matrix[1] = 0.f; matrix[3] = -2.f / height; matrix[5] = 1.f;
// clang-format on
}
return matrix;
constexpr std::array<f32, 4 * 4> MakeOrthographicMatrix(f32 width, f32 height) {
// clang-format off
return { 2.f / width, 0.f, 0.f, 0.f,
0.f, 2.f / height, 0.f, 0.f,
0.f, 0.f, 1.f, 0.f,
-1.f, -1.f, 0.f, 1.f};
// clang-format on
}
namespace {
@ -418,14 +287,10 @@ void RendererVulkan::LoadFBToScreenInfo(const GPU::Regs::FramebufferConfig& fram
void RendererVulkan::CompileShaders() {
vk::Device device = instance.GetDevice();
present_vertex_shader =
Compile(vertex_shader, vk::ShaderStageFlagBits::eVertex, device, ShaderOptimization::Debug);
present_shaders[0] = Compile(fragment_shader, vk::ShaderStageFlagBits::eFragment, device,
ShaderOptimization::Debug);
present_shaders[1] = Compile(fragment_shader_anaglyph, vk::ShaderStageFlagBits::eFragment,
device, ShaderOptimization::Debug);
present_shaders[2] = Compile(fragment_shader_interlaced, vk::ShaderStageFlagBits::eFragment,
device, ShaderOptimization::Debug);
present_vertex_shader = CompileSPV(VULKAN_PRESENT_VERT_SPV, device);
present_shaders[0] = CompileSPV(VULKAN_PRESENT_FRAG_SPV, device);
present_shaders[1] = CompileSPV(VULKAN_PRESENT_ANAGLYPH_FRAG_SPV, device);
present_shaders[2] = CompileSPV(VULKAN_PRESENT_INTERLACED_FRAG_SPV, device);
auto properties = instance.GetPhysicalDevice().getProperties();
for (std::size_t i = 0; i < present_samplers.size(); i++) {
@ -441,7 +306,8 @@ void RendererVulkan::CompileShaders() {
.compareEnable = false,
.compareOp = vk::CompareOp::eAlways,
.borderColor = vk::BorderColor::eIntOpaqueBlack,
.unnormalizedCoordinates = false};
.unnormalizedCoordinates = false,
};
present_samplers[i] = device.createSampler(sampler_info);
}

View File

@ -12,111 +12,113 @@
namespace Vulkan {
constexpr TBuiltInResource DefaultTBuiltInResource = {.maxLights = 32,
.maxClipPlanes = 6,
.maxTextureUnits = 32,
.maxTextureCoords = 32,
.maxVertexAttribs = 64,
.maxVertexUniformComponents = 4096,
.maxVaryingFloats = 64,
.maxVertexTextureImageUnits = 32,
.maxCombinedTextureImageUnits = 80,
.maxTextureImageUnits = 32,
.maxFragmentUniformComponents = 4096,
.maxDrawBuffers = 32,
.maxVertexUniformVectors = 128,
.maxVaryingVectors = 8,
.maxFragmentUniformVectors = 16,
.maxVertexOutputVectors = 16,
.maxFragmentInputVectors = 15,
.minProgramTexelOffset = -8,
.maxProgramTexelOffset = 7,
.maxClipDistances = 8,
.maxComputeWorkGroupCountX = 65535,
.maxComputeWorkGroupCountY = 65535,
.maxComputeWorkGroupCountZ = 65535,
.maxComputeWorkGroupSizeX = 1024,
.maxComputeWorkGroupSizeY = 1024,
.maxComputeWorkGroupSizeZ = 64,
.maxComputeUniformComponents = 1024,
.maxComputeTextureImageUnits = 16,
.maxComputeImageUniforms = 8,
.maxComputeAtomicCounters = 8,
.maxComputeAtomicCounterBuffers = 1,
.maxVaryingComponents = 60,
.maxVertexOutputComponents = 64,
.maxGeometryInputComponents = 64,
.maxGeometryOutputComponents = 128,
.maxFragmentInputComponents = 128,
.maxImageUnits = 8,
.maxCombinedImageUnitsAndFragmentOutputs = 8,
.maxCombinedShaderOutputResources = 8,
.maxImageSamples = 0,
.maxVertexImageUniforms = 0,
.maxTessControlImageUniforms = 0,
.maxTessEvaluationImageUniforms = 0,
.maxGeometryImageUniforms = 0,
.maxFragmentImageUniforms = 8,
.maxCombinedImageUniforms = 8,
.maxGeometryTextureImageUnits = 16,
.maxGeometryOutputVertices = 256,
.maxGeometryTotalOutputComponents = 1024,
.maxGeometryUniformComponents = 1024,
.maxGeometryVaryingComponents = 64,
.maxTessControlInputComponents = 128,
.maxTessControlOutputComponents = 128,
.maxTessControlTextureImageUnits = 16,
.maxTessControlUniformComponents = 1024,
.maxTessControlTotalOutputComponents = 4096,
.maxTessEvaluationInputComponents = 128,
.maxTessEvaluationOutputComponents = 128,
.maxTessEvaluationTextureImageUnits = 16,
.maxTessEvaluationUniformComponents = 1024,
.maxTessPatchComponents = 120,
.maxPatchVertices = 32,
.maxTessGenLevel = 64,
.maxViewports = 16,
.maxVertexAtomicCounters = 0,
.maxTessControlAtomicCounters = 0,
.maxTessEvaluationAtomicCounters = 0,
.maxGeometryAtomicCounters = 0,
.maxFragmentAtomicCounters = 8,
.maxCombinedAtomicCounters = 8,
.maxAtomicCounterBindings = 1,
.maxVertexAtomicCounterBuffers = 0,
.maxTessControlAtomicCounterBuffers = 0,
.maxTessEvaluationAtomicCounterBuffers = 0,
.maxGeometryAtomicCounterBuffers = 0,
.maxFragmentAtomicCounterBuffers = 1,
.maxCombinedAtomicCounterBuffers = 1,
.maxAtomicCounterBufferSize = 16384,
.maxTransformFeedbackBuffers = 4,
.maxTransformFeedbackInterleavedComponents =
64,
.maxCullDistances = 8,
.maxCombinedClipAndCullDistances = 8,
.maxSamples = 4,
.maxMeshOutputVerticesNV = 256,
.maxMeshOutputPrimitivesNV = 512,
.maxMeshWorkGroupSizeX_NV = 32,
.maxMeshWorkGroupSizeY_NV = 1,
.maxMeshWorkGroupSizeZ_NV = 1,
.maxTaskWorkGroupSizeX_NV = 32,
.maxTaskWorkGroupSizeY_NV = 1,
.maxTaskWorkGroupSizeZ_NV = 1,
.maxMeshViewCountNV = 4,
.maxDualSourceDrawBuffersEXT = 1,
.limits = TLimits{
.nonInductiveForLoops = 1,
.whileLoops = 1,
.doWhileLoops = 1,
.generalUniformIndexing = 1,
.generalAttributeMatrixVectorIndexing = 1,
.generalVaryingIndexing = 1,
.generalSamplerIndexing = 1,
.generalVariableIndexing = 1,
.generalConstantMatrixVectorIndexing = 1,
}};
constexpr TBuiltInResource DefaultTBuiltInResource = {
.maxLights = 32,
.maxClipPlanes = 6,
.maxTextureUnits = 32,
.maxTextureCoords = 32,
.maxVertexAttribs = 64,
.maxVertexUniformComponents = 4096,
.maxVaryingFloats = 64,
.maxVertexTextureImageUnits = 32,
.maxCombinedTextureImageUnits = 80,
.maxTextureImageUnits = 32,
.maxFragmentUniformComponents = 4096,
.maxDrawBuffers = 32,
.maxVertexUniformVectors = 128,
.maxVaryingVectors = 8,
.maxFragmentUniformVectors = 16,
.maxVertexOutputVectors = 16,
.maxFragmentInputVectors = 15,
.minProgramTexelOffset = -8,
.maxProgramTexelOffset = 7,
.maxClipDistances = 8,
.maxComputeWorkGroupCountX = 65535,
.maxComputeWorkGroupCountY = 65535,
.maxComputeWorkGroupCountZ = 65535,
.maxComputeWorkGroupSizeX = 1024,
.maxComputeWorkGroupSizeY = 1024,
.maxComputeWorkGroupSizeZ = 64,
.maxComputeUniformComponents = 1024,
.maxComputeTextureImageUnits = 16,
.maxComputeImageUniforms = 8,
.maxComputeAtomicCounters = 8,
.maxComputeAtomicCounterBuffers = 1,
.maxVaryingComponents = 60,
.maxVertexOutputComponents = 64,
.maxGeometryInputComponents = 64,
.maxGeometryOutputComponents = 128,
.maxFragmentInputComponents = 128,
.maxImageUnits = 8,
.maxCombinedImageUnitsAndFragmentOutputs = 8,
.maxCombinedShaderOutputResources = 8,
.maxImageSamples = 0,
.maxVertexImageUniforms = 0,
.maxTessControlImageUniforms = 0,
.maxTessEvaluationImageUniforms = 0,
.maxGeometryImageUniforms = 0,
.maxFragmentImageUniforms = 8,
.maxCombinedImageUniforms = 8,
.maxGeometryTextureImageUnits = 16,
.maxGeometryOutputVertices = 256,
.maxGeometryTotalOutputComponents = 1024,
.maxGeometryUniformComponents = 1024,
.maxGeometryVaryingComponents = 64,
.maxTessControlInputComponents = 128,
.maxTessControlOutputComponents = 128,
.maxTessControlTextureImageUnits = 16,
.maxTessControlUniformComponents = 1024,
.maxTessControlTotalOutputComponents = 4096,
.maxTessEvaluationInputComponents = 128,
.maxTessEvaluationOutputComponents = 128,
.maxTessEvaluationTextureImageUnits = 16,
.maxTessEvaluationUniformComponents = 1024,
.maxTessPatchComponents = 120,
.maxPatchVertices = 32,
.maxTessGenLevel = 64,
.maxViewports = 16,
.maxVertexAtomicCounters = 0,
.maxTessControlAtomicCounters = 0,
.maxTessEvaluationAtomicCounters = 0,
.maxGeometryAtomicCounters = 0,
.maxFragmentAtomicCounters = 8,
.maxCombinedAtomicCounters = 8,
.maxAtomicCounterBindings = 1,
.maxVertexAtomicCounterBuffers = 0,
.maxTessControlAtomicCounterBuffers = 0,
.maxTessEvaluationAtomicCounterBuffers = 0,
.maxGeometryAtomicCounterBuffers = 0,
.maxFragmentAtomicCounterBuffers = 1,
.maxCombinedAtomicCounterBuffers = 1,
.maxAtomicCounterBufferSize = 16384,
.maxTransformFeedbackBuffers = 4,
.maxTransformFeedbackInterleavedComponents =
64,
.maxCullDistances = 8,
.maxCombinedClipAndCullDistances = 8,
.maxSamples = 4,
.maxMeshOutputVerticesNV = 256,
.maxMeshOutputPrimitivesNV = 512,
.maxMeshWorkGroupSizeX_NV = 32,
.maxMeshWorkGroupSizeY_NV = 1,
.maxMeshWorkGroupSizeZ_NV = 1,
.maxTaskWorkGroupSizeX_NV = 32,
.maxTaskWorkGroupSizeY_NV = 1,
.maxTaskWorkGroupSizeZ_NV = 1,
.maxMeshViewCountNV = 4,
.maxDualSourceDrawBuffersEXT = 1,
.limits = TLimits{
.nonInductiveForLoops = 1,
.whileLoops = 1,
.doWhileLoops = 1,
.generalUniformIndexing = 1,
.generalAttributeMatrixVectorIndexing = 1,
.generalVaryingIndexing = 1,
.generalSamplerIndexing = 1,
.generalVariableIndexing = 1,
.generalConstantMatrixVectorIndexing = 1,
},
};
EShLanguage ToEshShaderStage(vk::ShaderStageFlagBits stage) {
switch (stage) {
@ -222,10 +224,13 @@ vk::ShaderModule Compile(std::string_view code, vk::ShaderStageFlagBits stage, v
MICROPROFILE_DEFINE(Vulkan_SPVCompilation, "Vulkan", "SPIR-V Shader Compilation",
MP_RGB(100, 255, 52));
vk::ShaderModule CompileSPV(std::vector<u32> code, vk::Device device) {
vk::ShaderModule CompileSPV(std::span<const u32> code, vk::Device device) {
MICROPROFILE_SCOPE(Vulkan_SPVCompilation);
const vk::ShaderModuleCreateInfo shader_info = {.codeSize = code.size() * sizeof(u32),
.pCode = code.data()};
const vk::ShaderModuleCreateInfo shader_info = {
.codeSize = code.size() * sizeof(u32),
.pCode = code.data(),
};
try {
return device.createShaderModule(shader_info);
} catch (vk::SystemError& err) {

View File

@ -13,6 +13,6 @@ enum class ShaderOptimization { High = 0, Debug = 1 };
vk::ShaderModule Compile(std::string_view code, vk::ShaderStageFlagBits stage, vk::Device device,
ShaderOptimization level);
vk::ShaderModule CompileSPV(std::vector<u32> code, vk::Device device);
vk::ShaderModule CompileSPV(std::span<const u32> code, vk::Device device);
} // namespace Vulkan

View File

@ -13,7 +13,7 @@ struct ShaderRegs;
namespace Pica::Shader {
class ShaderSetup;
struct ShaderSetup;
enum class UniformBindings : u32 { Common, VS, GS };