cry/Hyprland
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions
Hyprland/src/render/Shader.cpp

209 lines
6.6 KiB
C++
Raw Normal View History

added CPU uniform location lookup optimization
2022-06-30 12:39:10 +02:00
#include "Shader.hpp"
opengl: optimize shaders and reduce unneeded drawcalls (#10364) * opengl: remove unnecessery glflush calls glflushing forces the driver to break batching and issue commands prematurely and prevents optimisations like command reordering and merging. many glFunctions already internally glflushes and eglsync creation still has a glflush at end render. so lets reduce the overhead of these calls. * opengl: reduce glUseProgram calls apitrace shows cases where the same program gets called multiple times, add a helper function that keeps track of current program and only call it once on same program. reduces slight overhead. * opengl: use more efficient vertex array object use a more modern vertex array object approach with the shaders, makes it a onetime setup on shader creation instead of once per drawcall, also should make the driver not have to revalidate the vertex format on each call.
2025-05-11 18:36:20 +02:00
#include "render/OpenGL.hpp"
added CPU uniform location lookup optimization
2022-06-30 12:39:10 +02:00
renderer: reduce a lot of glcalls and cache various states (#10757) * opengl: cache viewport state according to nvidia docs calling glViewPort unnecessarily on the same already set viewport is wasteful and can cause state changes when not needed. cache it in a struct and only call it when the viewport is actually changing. * opengl: cache glenable/gldisable state avoid making multiple glenable/gldisable calls on already set caps, can cause state changes and incur driver overhead. * opengl: cache glscissor box only call glscissor if the box actually has changed, try to avoid state changes. * opengl: cache gluniform calls cache the gluniform calls, the uniform values are cached in driver per program only the drawcalls setting the uniform yet again with the same value on same location is causing more overhead then caching it ourself and just no oping on it if no changes. * shader: rewrite handling of uniforms and state this is way faster as we don't need to mess with maps (hashing, etc) and instead can just use an array * opengl: stuff and 300 shaders * opengl: typo * opengl: get the uniform locations properly now that the legacy shaders are gone get the uniformlocations for SKIP_CM etc, so they can be properly set and used depending on if cm_enabled is set to false or true, before it was falling back to a legacy shader that didnt even have those uniforms. * opengl: check epsilon on float and remove extra glcall seems an extra unset glcall was added, remove it. and check the float epsilon on the glfloat. * opengl: remove instanced shader draw remove the instanced boolean from the vertex shader, might be neglible differences, needs more benchmark/work to see if its even worth it. * texture: cache texture paramaters parameters where occasionally set twice or more on same texture, short version wrap it and cache it. and move gpu churn to cpu churn. add a bind/unbind to texture aswell. * texture: use fast std::array caching cache the texparameter values in fast array lookups and incase we dont want it cached, apply it anyways. * shader: fix typo and hdr typo actually use Matrix4x2fv in the 4x2fv cache function, and send the proper float array for hdr. * texture: make caching not linear lookup make caching of texture params not linear. * minor style changes * opengl: revert drawarrays revert the mostly code style reduce loc change of drawarrays, and focus on the caching. its a if else case going wrong here breaking blur/contrast amongst others drawing. --------- Co-authored-by: Vaxry <vaxry@vaxry.net>
2025-06-25 12:42:32 +02:00
#define EPSILON(x, y) (std::abs((x) - (y)) < 1e-5f)
static bool compareFloat(auto a, auto b) {
if (a.size() != b.size())
return false;
for (size_t i = 0; i < a.size(); ++i)
if (std::fabs(a[i] - b[i]) > 1e-5f)
return false;
return true;
}
SShader::SShader() {
uniformLocations.fill(-1);
}
refactor: cshader class to sshader struct (#10324)
2025-05-08 00:07:35 +02:00
SShader::~SShader() {
Added screen shaders
2022-12-01 13:36:07 +00:00
destroy();
}
opengl: optimize shaders and reduce unneeded drawcalls (#10364) * opengl: remove unnecessery glflush calls glflushing forces the driver to break batching and issue commands prematurely and prevents optimisations like command reordering and merging. many glFunctions already internally glflushes and eglsync creation still has a glflush at end render. so lets reduce the overhead of these calls. * opengl: reduce glUseProgram calls apitrace shows cases where the same program gets called multiple times, add a helper function that keeps track of current program and only call it once on same program. reduces slight overhead. * opengl: use more efficient vertex array object use a more modern vertex array object approach with the shaders, makes it a onetime setup on shader creation instead of once per drawcall, also should make the driver not have to revalidate the vertex format on each call.
2025-05-11 18:36:20 +02:00
void SShader::createVao() {
renderer: reduce a lot of glcalls and cache various states (#10757) * opengl: cache viewport state according to nvidia docs calling glViewPort unnecessarily on the same already set viewport is wasteful and can cause state changes when not needed. cache it in a struct and only call it when the viewport is actually changing. * opengl: cache glenable/gldisable state avoid making multiple glenable/gldisable calls on already set caps, can cause state changes and incur driver overhead. * opengl: cache glscissor box only call glscissor if the box actually has changed, try to avoid state changes. * opengl: cache gluniform calls cache the gluniform calls, the uniform values are cached in driver per program only the drawcalls setting the uniform yet again with the same value on same location is causing more overhead then caching it ourself and just no oping on it if no changes. * shader: rewrite handling of uniforms and state this is way faster as we don't need to mess with maps (hashing, etc) and instead can just use an array * opengl: stuff and 300 shaders * opengl: typo * opengl: get the uniform locations properly now that the legacy shaders are gone get the uniformlocations for SKIP_CM etc, so they can be properly set and used depending on if cm_enabled is set to false or true, before it was falling back to a legacy shader that didnt even have those uniforms. * opengl: check epsilon on float and remove extra glcall seems an extra unset glcall was added, remove it. and check the float epsilon on the glfloat. * opengl: remove instanced shader draw remove the instanced boolean from the vertex shader, might be neglible differences, needs more benchmark/work to see if its even worth it. * texture: cache texture paramaters parameters where occasionally set twice or more on same texture, short version wrap it and cache it. and move gpu churn to cpu churn. add a bind/unbind to texture aswell. * texture: use fast std::array caching cache the texparameter values in fast array lookups and incase we dont want it cached, apply it anyways. * shader: fix typo and hdr typo actually use Matrix4x2fv in the 4x2fv cache function, and send the proper float array for hdr. * texture: make caching not linear lookup make caching of texture params not linear. * minor style changes * opengl: revert drawarrays revert the mostly code style reduce loc change of drawarrays, and focus on the caching. its a if else case going wrong here breaking blur/contrast amongst others drawing. --------- Co-authored-by: Vaxry <vaxry@vaxry.net>
2025-06-25 12:42:32 +02:00
GLuint shaderVao = 0, shaderVbo = 0, shaderVboUv = 0;
opengl: optimize shaders and reduce unneeded drawcalls (#10364) * opengl: remove unnecessery glflush calls glflushing forces the driver to break batching and issue commands prematurely and prevents optimisations like command reordering and merging. many glFunctions already internally glflushes and eglsync creation still has a glflush at end render. so lets reduce the overhead of these calls. * opengl: reduce glUseProgram calls apitrace shows cases where the same program gets called multiple times, add a helper function that keeps track of current program and only call it once on same program. reduces slight overhead. * opengl: use more efficient vertex array object use a more modern vertex array object approach with the shaders, makes it a onetime setup on shader creation instead of once per drawcall, also should make the driver not have to revalidate the vertex format on each call.
2025-05-11 18:36:20 +02:00
glGenVertexArrays(1, &shaderVao);
glBindVertexArray(shaderVao);
renderer: reduce a lot of glcalls and cache various states (#10757) * opengl: cache viewport state according to nvidia docs calling glViewPort unnecessarily on the same already set viewport is wasteful and can cause state changes when not needed. cache it in a struct and only call it when the viewport is actually changing. * opengl: cache glenable/gldisable state avoid making multiple glenable/gldisable calls on already set caps, can cause state changes and incur driver overhead. * opengl: cache glscissor box only call glscissor if the box actually has changed, try to avoid state changes. * opengl: cache gluniform calls cache the gluniform calls, the uniform values are cached in driver per program only the drawcalls setting the uniform yet again with the same value on same location is causing more overhead then caching it ourself and just no oping on it if no changes. * shader: rewrite handling of uniforms and state this is way faster as we don't need to mess with maps (hashing, etc) and instead can just use an array * opengl: stuff and 300 shaders * opengl: typo * opengl: get the uniform locations properly now that the legacy shaders are gone get the uniformlocations for SKIP_CM etc, so they can be properly set and used depending on if cm_enabled is set to false or true, before it was falling back to a legacy shader that didnt even have those uniforms. * opengl: check epsilon on float and remove extra glcall seems an extra unset glcall was added, remove it. and check the float epsilon on the glfloat. * opengl: remove instanced shader draw remove the instanced boolean from the vertex shader, might be neglible differences, needs more benchmark/work to see if its even worth it. * texture: cache texture paramaters parameters where occasionally set twice or more on same texture, short version wrap it and cache it. and move gpu churn to cpu churn. add a bind/unbind to texture aswell. * texture: use fast std::array caching cache the texparameter values in fast array lookups and incase we dont want it cached, apply it anyways. * shader: fix typo and hdr typo actually use Matrix4x2fv in the 4x2fv cache function, and send the proper float array for hdr. * texture: make caching not linear lookup make caching of texture params not linear. * minor style changes * opengl: revert drawarrays revert the mostly code style reduce loc change of drawarrays, and focus on the caching. its a if else case going wrong here breaking blur/contrast amongst others drawing. --------- Co-authored-by: Vaxry <vaxry@vaxry.net>
2025-06-25 12:42:32 +02:00
if (uniformLocations[SHADER_POS_ATTRIB] != -1) {
glGenBuffers(1, &shaderVbo);
glBindBuffer(GL_ARRAY_BUFFER, shaderVbo);
opengl: optimize shaders and reduce unneeded drawcalls (#10364) * opengl: remove unnecessery glflush calls glflushing forces the driver to break batching and issue commands prematurely and prevents optimisations like command reordering and merging. many glFunctions already internally glflushes and eglsync creation still has a glflush at end render. so lets reduce the overhead of these calls. * opengl: reduce glUseProgram calls apitrace shows cases where the same program gets called multiple times, add a helper function that keeps track of current program and only call it once on same program. reduces slight overhead. * opengl: use more efficient vertex array object use a more modern vertex array object approach with the shaders, makes it a onetime setup on shader creation instead of once per drawcall, also should make the driver not have to revalidate the vertex format on each call.
2025-05-11 18:36:20 +02:00
glBufferData(GL_ARRAY_BUFFER, sizeof(fullVerts), fullVerts, GL_STATIC_DRAW);
renderer: reduce a lot of glcalls and cache various states (#10757) * opengl: cache viewport state according to nvidia docs calling glViewPort unnecessarily on the same already set viewport is wasteful and can cause state changes when not needed. cache it in a struct and only call it when the viewport is actually changing. * opengl: cache glenable/gldisable state avoid making multiple glenable/gldisable calls on already set caps, can cause state changes and incur driver overhead. * opengl: cache glscissor box only call glscissor if the box actually has changed, try to avoid state changes. * opengl: cache gluniform calls cache the gluniform calls, the uniform values are cached in driver per program only the drawcalls setting the uniform yet again with the same value on same location is causing more overhead then caching it ourself and just no oping on it if no changes. * shader: rewrite handling of uniforms and state this is way faster as we don't need to mess with maps (hashing, etc) and instead can just use an array * opengl: stuff and 300 shaders * opengl: typo * opengl: get the uniform locations properly now that the legacy shaders are gone get the uniformlocations for SKIP_CM etc, so they can be properly set and used depending on if cm_enabled is set to false or true, before it was falling back to a legacy shader that didnt even have those uniforms. * opengl: check epsilon on float and remove extra glcall seems an extra unset glcall was added, remove it. and check the float epsilon on the glfloat. * opengl: remove instanced shader draw remove the instanced boolean from the vertex shader, might be neglible differences, needs more benchmark/work to see if its even worth it. * texture: cache texture paramaters parameters where occasionally set twice or more on same texture, short version wrap it and cache it. and move gpu churn to cpu churn. add a bind/unbind to texture aswell. * texture: use fast std::array caching cache the texparameter values in fast array lookups and incase we dont want it cached, apply it anyways. * shader: fix typo and hdr typo actually use Matrix4x2fv in the 4x2fv cache function, and send the proper float array for hdr. * texture: make caching not linear lookup make caching of texture params not linear. * minor style changes * opengl: revert drawarrays revert the mostly code style reduce loc change of drawarrays, and focus on the caching. its a if else case going wrong here breaking blur/contrast amongst others drawing. --------- Co-authored-by: Vaxry <vaxry@vaxry.net>
2025-06-25 12:42:32 +02:00
glEnableVertexAttribArray(uniformLocations[SHADER_POS_ATTRIB]);
glVertexAttribPointer(uniformLocations[SHADER_POS_ATTRIB], 2, GL_FLOAT, GL_FALSE, 0, nullptr);
opengl: optimize shaders and reduce unneeded drawcalls (#10364) * opengl: remove unnecessery glflush calls glflushing forces the driver to break batching and issue commands prematurely and prevents optimisations like command reordering and merging. many glFunctions already internally glflushes and eglsync creation still has a glflush at end render. so lets reduce the overhead of these calls. * opengl: reduce glUseProgram calls apitrace shows cases where the same program gets called multiple times, add a helper function that keeps track of current program and only call it once on same program. reduces slight overhead. * opengl: use more efficient vertex array object use a more modern vertex array object approach with the shaders, makes it a onetime setup on shader creation instead of once per drawcall, also should make the driver not have to revalidate the vertex format on each call.
2025-05-11 18:36:20 +02:00
}
// UV VBO (dynamic, may be updated per frame)
renderer: reduce a lot of glcalls and cache various states (#10757) * opengl: cache viewport state according to nvidia docs calling glViewPort unnecessarily on the same already set viewport is wasteful and can cause state changes when not needed. cache it in a struct and only call it when the viewport is actually changing. * opengl: cache glenable/gldisable state avoid making multiple glenable/gldisable calls on already set caps, can cause state changes and incur driver overhead. * opengl: cache glscissor box only call glscissor if the box actually has changed, try to avoid state changes. * opengl: cache gluniform calls cache the gluniform calls, the uniform values are cached in driver per program only the drawcalls setting the uniform yet again with the same value on same location is causing more overhead then caching it ourself and just no oping on it if no changes. * shader: rewrite handling of uniforms and state this is way faster as we don't need to mess with maps (hashing, etc) and instead can just use an array * opengl: stuff and 300 shaders * opengl: typo * opengl: get the uniform locations properly now that the legacy shaders are gone get the uniformlocations for SKIP_CM etc, so they can be properly set and used depending on if cm_enabled is set to false or true, before it was falling back to a legacy shader that didnt even have those uniforms. * opengl: check epsilon on float and remove extra glcall seems an extra unset glcall was added, remove it. and check the float epsilon on the glfloat. * opengl: remove instanced shader draw remove the instanced boolean from the vertex shader, might be neglible differences, needs more benchmark/work to see if its even worth it. * texture: cache texture paramaters parameters where occasionally set twice or more on same texture, short version wrap it and cache it. and move gpu churn to cpu churn. add a bind/unbind to texture aswell. * texture: use fast std::array caching cache the texparameter values in fast array lookups and incase we dont want it cached, apply it anyways. * shader: fix typo and hdr typo actually use Matrix4x2fv in the 4x2fv cache function, and send the proper float array for hdr. * texture: make caching not linear lookup make caching of texture params not linear. * minor style changes * opengl: revert drawarrays revert the mostly code style reduce loc change of drawarrays, and focus on the caching. its a if else case going wrong here breaking blur/contrast amongst others drawing. --------- Co-authored-by: Vaxry <vaxry@vaxry.net>
2025-06-25 12:42:32 +02:00
if (uniformLocations[SHADER_TEX_ATTRIB] != -1) {
opengl: optimize shaders and reduce unneeded drawcalls (#10364) * opengl: remove unnecessery glflush calls glflushing forces the driver to break batching and issue commands prematurely and prevents optimisations like command reordering and merging. many glFunctions already internally glflushes and eglsync creation still has a glflush at end render. so lets reduce the overhead of these calls. * opengl: reduce glUseProgram calls apitrace shows cases where the same program gets called multiple times, add a helper function that keeps track of current program and only call it once on same program. reduces slight overhead. * opengl: use more efficient vertex array object use a more modern vertex array object approach with the shaders, makes it a onetime setup on shader creation instead of once per drawcall, also should make the driver not have to revalidate the vertex format on each call.
2025-05-11 18:36:20 +02:00
glGenBuffers(1, &shaderVboUv);
glBindBuffer(GL_ARRAY_BUFFER, shaderVboUv);
glBufferData(GL_ARRAY_BUFFER, sizeof(fullVerts), fullVerts, GL_DYNAMIC_DRAW); // Initial dummy UVs
renderer: reduce a lot of glcalls and cache various states (#10757) * opengl: cache viewport state according to nvidia docs calling glViewPort unnecessarily on the same already set viewport is wasteful and can cause state changes when not needed. cache it in a struct and only call it when the viewport is actually changing. * opengl: cache glenable/gldisable state avoid making multiple glenable/gldisable calls on already set caps, can cause state changes and incur driver overhead. * opengl: cache glscissor box only call glscissor if the box actually has changed, try to avoid state changes. * opengl: cache gluniform calls cache the gluniform calls, the uniform values are cached in driver per program only the drawcalls setting the uniform yet again with the same value on same location is causing more overhead then caching it ourself and just no oping on it if no changes. * shader: rewrite handling of uniforms and state this is way faster as we don't need to mess with maps (hashing, etc) and instead can just use an array * opengl: stuff and 300 shaders * opengl: typo * opengl: get the uniform locations properly now that the legacy shaders are gone get the uniformlocations for SKIP_CM etc, so they can be properly set and used depending on if cm_enabled is set to false or true, before it was falling back to a legacy shader that didnt even have those uniforms. * opengl: check epsilon on float and remove extra glcall seems an extra unset glcall was added, remove it. and check the float epsilon on the glfloat. * opengl: remove instanced shader draw remove the instanced boolean from the vertex shader, might be neglible differences, needs more benchmark/work to see if its even worth it. * texture: cache texture paramaters parameters where occasionally set twice or more on same texture, short version wrap it and cache it. and move gpu churn to cpu churn. add a bind/unbind to texture aswell. * texture: use fast std::array caching cache the texparameter values in fast array lookups and incase we dont want it cached, apply it anyways. * shader: fix typo and hdr typo actually use Matrix4x2fv in the 4x2fv cache function, and send the proper float array for hdr. * texture: make caching not linear lookup make caching of texture params not linear. * minor style changes * opengl: revert drawarrays revert the mostly code style reduce loc change of drawarrays, and focus on the caching. its a if else case going wrong here breaking blur/contrast amongst others drawing. --------- Co-authored-by: Vaxry <vaxry@vaxry.net>
2025-06-25 12:42:32 +02:00
glEnableVertexAttribArray(uniformLocations[SHADER_TEX_ATTRIB]);
glVertexAttribPointer(uniformLocations[SHADER_TEX_ATTRIB], 2, GL_FLOAT, GL_FALSE, 0, nullptr);
opengl: optimize shaders and reduce unneeded drawcalls (#10364) * opengl: remove unnecessery glflush calls glflushing forces the driver to break batching and issue commands prematurely and prevents optimisations like command reordering and merging. many glFunctions already internally glflushes and eglsync creation still has a glflush at end render. so lets reduce the overhead of these calls. * opengl: reduce glUseProgram calls apitrace shows cases where the same program gets called multiple times, add a helper function that keeps track of current program and only call it once on same program. reduces slight overhead. * opengl: use more efficient vertex array object use a more modern vertex array object approach with the shaders, makes it a onetime setup on shader creation instead of once per drawcall, also should make the driver not have to revalidate the vertex format on each call.
2025-05-11 18:36:20 +02:00
}
glBindVertexArray(0);
glBindBuffer(GL_ARRAY_BUFFER, 0);
renderer: reduce a lot of glcalls and cache various states (#10757) * opengl: cache viewport state according to nvidia docs calling glViewPort unnecessarily on the same already set viewport is wasteful and can cause state changes when not needed. cache it in a struct and only call it when the viewport is actually changing. * opengl: cache glenable/gldisable state avoid making multiple glenable/gldisable calls on already set caps, can cause state changes and incur driver overhead. * opengl: cache glscissor box only call glscissor if the box actually has changed, try to avoid state changes. * opengl: cache gluniform calls cache the gluniform calls, the uniform values are cached in driver per program only the drawcalls setting the uniform yet again with the same value on same location is causing more overhead then caching it ourself and just no oping on it if no changes. * shader: rewrite handling of uniforms and state this is way faster as we don't need to mess with maps (hashing, etc) and instead can just use an array * opengl: stuff and 300 shaders * opengl: typo * opengl: get the uniform locations properly now that the legacy shaders are gone get the uniformlocations for SKIP_CM etc, so they can be properly set and used depending on if cm_enabled is set to false or true, before it was falling back to a legacy shader that didnt even have those uniforms. * opengl: check epsilon on float and remove extra glcall seems an extra unset glcall was added, remove it. and check the float epsilon on the glfloat. * opengl: remove instanced shader draw remove the instanced boolean from the vertex shader, might be neglible differences, needs more benchmark/work to see if its even worth it. * texture: cache texture paramaters parameters where occasionally set twice or more on same texture, short version wrap it and cache it. and move gpu churn to cpu churn. add a bind/unbind to texture aswell. * texture: use fast std::array caching cache the texparameter values in fast array lookups and incase we dont want it cached, apply it anyways. * shader: fix typo and hdr typo actually use Matrix4x2fv in the 4x2fv cache function, and send the proper float array for hdr. * texture: make caching not linear lookup make caching of texture params not linear. * minor style changes * opengl: revert drawarrays revert the mostly code style reduce loc change of drawarrays, and focus on the caching. its a if else case going wrong here breaking blur/contrast amongst others drawing. --------- Co-authored-by: Vaxry <vaxry@vaxry.net>
2025-06-25 12:42:32 +02:00
uniformLocations[SHADER_SHADER_VAO] = shaderVao;
uniformLocations[SHADER_SHADER_VBO_POS] = shaderVbo;
uniformLocations[SHADER_SHADER_VBO_UV] = shaderVboUv;
RASSERT(uniformLocations[SHADER_SHADER_VAO] >= 0, "SHADER_SHADER_VAO could not be created");
RASSERT(uniformLocations[SHADER_SHADER_VBO_POS] >= 0, "SHADER_SHADER_VBO_POS could not be created");
RASSERT(uniformLocations[SHADER_SHADER_VBO_UV] >= 0, "SHADER_SHADER_VBO_UV could not be created");
}
void SShader::setUniformInt(eShaderUniform location, GLint v0) {
if (uniformLocations.at(location) == -1)
return;
auto& cached = uniformStatus.at(location);
if (cached.index() != 0 && std::get<GLint>(cached) == v0)
return;
cached = v0;
glUniform1i(uniformLocations[location], v0);
}
void SShader::setUniformFloat(eShaderUniform location, GLfloat v0) {
if (uniformLocations.at(location) == -1)
return;
auto& cached = uniformStatus.at(location);
if (cached.index() != 0) {
auto val = std::get<GLfloat>(cached);
if (EPSILON(val, v0))
return;
}
cached = v0;
glUniform1f(uniformLocations[location], v0);
}
void SShader::setUniformFloat2(eShaderUniform location, GLfloat v0, GLfloat v1) {
if (uniformLocations.at(location) == -1)
return;
auto& cached = uniformStatus.at(location);
if (cached.index() != 0) {
auto val = std::get<std::array<GLfloat, 2>>(cached);
if (EPSILON(val[0], v0) && EPSILON(val[1], v1))
return;
}
cached = std::array<GLfloat, 2>{v0, v1};
glUniform2f(uniformLocations[location], v0, v1);
}
void SShader::setUniformFloat3(eShaderUniform location, GLfloat v0, GLfloat v1, GLfloat v2) {
if (uniformLocations.at(location) == -1)
return;
auto& cached = uniformStatus.at(location);
if (cached.index() != 0) {
auto val = std::get<std::array<GLfloat, 3>>(cached);
if (EPSILON(val[0], v0) && EPSILON(val[1], v1) && EPSILON(val[2], v2))
return;
}
cached = std::array<GLfloat, 3>{v0, v1, v2};
glUniform3f(uniformLocations[location], v0, v1, v2);
}
void SShader::setUniformFloat4(eShaderUniform location, GLfloat v0, GLfloat v1, GLfloat v2, GLfloat v3) {
if (uniformLocations.at(location) == -1)
return;
auto& cached = uniformStatus.at(location);
if (cached.index() != 0) {
auto val = std::get<std::array<GLfloat, 4>>(cached);
if (EPSILON(val[0], v0) && EPSILON(val[1], v1) && EPSILON(val[2], v2) && EPSILON(val[3], v3))
return;
}
cached = std::array<GLfloat, 4>{v0, v1, v2, v3};
glUniform4f(uniformLocations[location], v0, v1, v2, v3);
}
void SShader::setUniformMatrix3fv(eShaderUniform location, GLsizei count, GLboolean transpose, std::array<GLfloat, 9> value) {
if (uniformLocations.at(location) == -1)
return;
auto& cached = uniformStatus.at(location);
if (cached.index() != 0) {
auto val = std::get<SUniformMatrix3Data>(cached);
if (val.count == count && val.transpose == transpose && compareFloat(val.value, value))
return;
}
cached = SUniformMatrix3Data{.count = count, .transpose = transpose, .value = value};
glUniformMatrix3fv(uniformLocations[location], count, transpose, value.data());
}
void SShader::setUniformMatrix4x2fv(eShaderUniform location, GLsizei count, GLboolean transpose, std::array<GLfloat, 8> value) {
if (uniformLocations.at(location) == -1)
return;
auto& cached = uniformStatus.at(location);
if (cached.index() != 0) {
auto val = std::get<SUniformMatrix4Data>(cached);
if (val.count == count && val.transpose == transpose && compareFloat(val.value, value))
return;
}
cached = SUniformMatrix4Data{.count = count, .transpose = transpose, .value = value};
glUniformMatrix4x2fv(uniformLocations[location], count, transpose, value.data());
}
void SShader::setUniform4fv(eShaderUniform location, GLsizei count, std::vector<float> value) {
if (uniformLocations.at(location) == -1)
return;
auto& cached = uniformStatus.at(location);
if (cached.index() != 0) {
auto val = std::get<SUniform4Data>(cached);
if (val.count == count && compareFloat(val.value, value))
return;
}
cached = SUniform4Data{.count = count, .value = value};
glUniform4fv(uniformLocations[location], count, value.data());
opengl: optimize shaders and reduce unneeded drawcalls (#10364) * opengl: remove unnecessery glflush calls glflushing forces the driver to break batching and issue commands prematurely and prevents optimisations like command reordering and merging. many glFunctions already internally glflushes and eglsync creation still has a glflush at end render. so lets reduce the overhead of these calls. * opengl: reduce glUseProgram calls apitrace shows cases where the same program gets called multiple times, add a helper function that keeps track of current program and only call it once on same program. reduces slight overhead. * opengl: use more efficient vertex array object use a more modern vertex array object approach with the shaders, makes it a onetime setup on shader creation instead of once per drawcall, also should make the driver not have to revalidate the vertex format on each call.
2025-05-11 18:36:20 +02:00
}
refactor: cshader class to sshader struct (#10324)
2025-05-08 00:07:35 +02:00
void SShader::destroy() {
renderer: reduce a lot of glcalls and cache various states (#10757) * opengl: cache viewport state according to nvidia docs calling glViewPort unnecessarily on the same already set viewport is wasteful and can cause state changes when not needed. cache it in a struct and only call it when the viewport is actually changing. * opengl: cache glenable/gldisable state avoid making multiple glenable/gldisable calls on already set caps, can cause state changes and incur driver overhead. * opengl: cache glscissor box only call glscissor if the box actually has changed, try to avoid state changes. * opengl: cache gluniform calls cache the gluniform calls, the uniform values are cached in driver per program only the drawcalls setting the uniform yet again with the same value on same location is causing more overhead then caching it ourself and just no oping on it if no changes. * shader: rewrite handling of uniforms and state this is way faster as we don't need to mess with maps (hashing, etc) and instead can just use an array * opengl: stuff and 300 shaders * opengl: typo * opengl: get the uniform locations properly now that the legacy shaders are gone get the uniformlocations for SKIP_CM etc, so they can be properly set and used depending on if cm_enabled is set to false or true, before it was falling back to a legacy shader that didnt even have those uniforms. * opengl: check epsilon on float and remove extra glcall seems an extra unset glcall was added, remove it. and check the float epsilon on the glfloat. * opengl: remove instanced shader draw remove the instanced boolean from the vertex shader, might be neglible differences, needs more benchmark/work to see if its even worth it. * texture: cache texture paramaters parameters where occasionally set twice or more on same texture, short version wrap it and cache it. and move gpu churn to cpu churn. add a bind/unbind to texture aswell. * texture: use fast std::array caching cache the texparameter values in fast array lookups and incase we dont want it cached, apply it anyways. * shader: fix typo and hdr typo actually use Matrix4x2fv in the 4x2fv cache function, and send the proper float array for hdr. * texture: make caching not linear lookup make caching of texture params not linear. * minor style changes * opengl: revert drawarrays revert the mostly code style reduce loc change of drawarrays, and focus on the caching. its a if else case going wrong here breaking blur/contrast amongst others drawing. --------- Co-authored-by: Vaxry <vaxry@vaxry.net>
2025-06-25 12:42:32 +02:00
uniformStatus.fill(std::monostate());
renderer: various fixes towards improving gpu calls robustness (#9188) ensure framebuffer textures are detached and deleted, avoid leaving framebuffers bound when not needed * render: avoid calling glDeleteProgram on no program its safe to do so but it adds a bunch of unnecessery lines in apitrace when tracing. if guard it and return early. * opengl: ensure texture and buffers are unbound ensure bound buffers are unbound after use, also detach textures from framebuffer before deleting it otherwise it will become dangling and essentially leak.
2025-02-08 01:46:26 +01:00
if (program == 0)
return;
renderer: reduce a lot of glcalls and cache various states (#10757) * opengl: cache viewport state according to nvidia docs calling glViewPort unnecessarily on the same already set viewport is wasteful and can cause state changes when not needed. cache it in a struct and only call it when the viewport is actually changing. * opengl: cache glenable/gldisable state avoid making multiple glenable/gldisable calls on already set caps, can cause state changes and incur driver overhead. * opengl: cache glscissor box only call glscissor if the box actually has changed, try to avoid state changes. * opengl: cache gluniform calls cache the gluniform calls, the uniform values are cached in driver per program only the drawcalls setting the uniform yet again with the same value on same location is causing more overhead then caching it ourself and just no oping on it if no changes. * shader: rewrite handling of uniforms and state this is way faster as we don't need to mess with maps (hashing, etc) and instead can just use an array * opengl: stuff and 300 shaders * opengl: typo * opengl: get the uniform locations properly now that the legacy shaders are gone get the uniformlocations for SKIP_CM etc, so they can be properly set and used depending on if cm_enabled is set to false or true, before it was falling back to a legacy shader that didnt even have those uniforms. * opengl: check epsilon on float and remove extra glcall seems an extra unset glcall was added, remove it. and check the float epsilon on the glfloat. * opengl: remove instanced shader draw remove the instanced boolean from the vertex shader, might be neglible differences, needs more benchmark/work to see if its even worth it. * texture: cache texture paramaters parameters where occasionally set twice or more on same texture, short version wrap it and cache it. and move gpu churn to cpu churn. add a bind/unbind to texture aswell. * texture: use fast std::array caching cache the texparameter values in fast array lookups and incase we dont want it cached, apply it anyways. * shader: fix typo and hdr typo actually use Matrix4x2fv in the 4x2fv cache function, and send the proper float array for hdr. * texture: make caching not linear lookup make caching of texture params not linear. * minor style changes * opengl: revert drawarrays revert the mostly code style reduce loc change of drawarrays, and focus on the caching. its a if else case going wrong here breaking blur/contrast amongst others drawing. --------- Co-authored-by: Vaxry <vaxry@vaxry.net>
2025-06-25 12:42:32 +02:00
GLuint shaderVao, shaderVbo, shaderVboUv;
shaderVao = uniformLocations[SHADER_SHADER_VAO] == -1 ? 0 : uniformLocations[SHADER_SHADER_VAO];
shaderVbo = uniformLocations[SHADER_SHADER_VBO_POS] == -1 ? 0 : uniformLocations[SHADER_SHADER_VBO_POS];
shaderVboUv = uniformLocations[SHADER_SHADER_VBO_UV] == -1 ? 0 : uniformLocations[SHADER_SHADER_VBO_UV];
opengl: optimize shaders and reduce unneeded drawcalls (#10364) * opengl: remove unnecessery glflush calls glflushing forces the driver to break batching and issue commands prematurely and prevents optimisations like command reordering and merging. many glFunctions already internally glflushes and eglsync creation still has a glflush at end render. so lets reduce the overhead of these calls. * opengl: reduce glUseProgram calls apitrace shows cases where the same program gets called multiple times, add a helper function that keeps track of current program and only call it once on same program. reduces slight overhead. * opengl: use more efficient vertex array object use a more modern vertex array object approach with the shaders, makes it a onetime setup on shader creation instead of once per drawcall, also should make the driver not have to revalidate the vertex format on each call.
2025-05-11 18:36:20 +02:00
if (shaderVao)
glDeleteVertexArrays(1, &shaderVao);
renderer: reduce a lot of glcalls and cache various states (#10757) * opengl: cache viewport state according to nvidia docs calling glViewPort unnecessarily on the same already set viewport is wasteful and can cause state changes when not needed. cache it in a struct and only call it when the viewport is actually changing. * opengl: cache glenable/gldisable state avoid making multiple glenable/gldisable calls on already set caps, can cause state changes and incur driver overhead. * opengl: cache glscissor box only call glscissor if the box actually has changed, try to avoid state changes. * opengl: cache gluniform calls cache the gluniform calls, the uniform values are cached in driver per program only the drawcalls setting the uniform yet again with the same value on same location is causing more overhead then caching it ourself and just no oping on it if no changes. * shader: rewrite handling of uniforms and state this is way faster as we don't need to mess with maps (hashing, etc) and instead can just use an array * opengl: stuff and 300 shaders * opengl: typo * opengl: get the uniform locations properly now that the legacy shaders are gone get the uniformlocations for SKIP_CM etc, so they can be properly set and used depending on if cm_enabled is set to false or true, before it was falling back to a legacy shader that didnt even have those uniforms. * opengl: check epsilon on float and remove extra glcall seems an extra unset glcall was added, remove it. and check the float epsilon on the glfloat. * opengl: remove instanced shader draw remove the instanced boolean from the vertex shader, might be neglible differences, needs more benchmark/work to see if its even worth it. * texture: cache texture paramaters parameters where occasionally set twice or more on same texture, short version wrap it and cache it. and move gpu churn to cpu churn. add a bind/unbind to texture aswell. * texture: use fast std::array caching cache the texparameter values in fast array lookups and incase we dont want it cached, apply it anyways. * shader: fix typo and hdr typo actually use Matrix4x2fv in the 4x2fv cache function, and send the proper float array for hdr. * texture: make caching not linear lookup make caching of texture params not linear. * minor style changes * opengl: revert drawarrays revert the mostly code style reduce loc change of drawarrays, and focus on the caching. its a if else case going wrong here breaking blur/contrast amongst others drawing. --------- Co-authored-by: Vaxry <vaxry@vaxry.net>
2025-06-25 12:42:32 +02:00
if (shaderVbo)
glDeleteBuffers(1, &shaderVbo);
fix shader destroy id unset
2022-12-03 14:45:05 +00:00
opengl: optimize shaders and reduce unneeded drawcalls (#10364) * opengl: remove unnecessery glflush calls glflushing forces the driver to break batching and issue commands prematurely and prevents optimisations like command reordering and merging. many glFunctions already internally glflushes and eglsync creation still has a glflush at end render. so lets reduce the overhead of these calls. * opengl: reduce glUseProgram calls apitrace shows cases where the same program gets called multiple times, add a helper function that keeps track of current program and only call it once on same program. reduces slight overhead. * opengl: use more efficient vertex array object use a more modern vertex array object approach with the shaders, makes it a onetime setup on shader creation instead of once per drawcall, also should make the driver not have to revalidate the vertex format on each call.
2025-05-11 18:36:20 +02:00
if (shaderVboUv)
glDeleteBuffers(1, &shaderVboUv);
glDeleteProgram(program);
fix shader destroy id unset
2022-12-03 14:45:05 +00:00
program = 0;
renderer: various fixes towards improving gpu calls robustness (#9188) ensure framebuffer textures are detached and deleted, avoid leaving framebuffers bound when not needed * render: avoid calling glDeleteProgram on no program its safe to do so but it adds a bunch of unnecessery lines in apitrace when tracing. if guard it and return early. * opengl: ensure texture and buffers are unbound ensure bound buffers are unbound after use, also detach textures from framebuffer before deleting it otherwise it will become dangling and essentially leak.
2025-02-08 01:46:26 +01:00
}
Reference in a new issue Copy permalink