diff options
| author | ReinUsesLisp <reinuseslisp@airmail.cc> | 2019-11-02 04:08:31 -0300 |
|---|---|---|
| committer | ReinUsesLisp <reinuseslisp@airmail.cc> | 2019-11-02 05:05:34 -0300 |
| commit | 76ca2a5f82f4df64cb839af42c93acb6705411ae (patch) | |
| tree | 22ef46bef8f32d8b2aa2f3928a96b30b3f69e213 /src/video_core/renderer_opengl/gl_device.cpp | |
| parent | 11e39da02bec92fe4a332bfb737323ccb8087626 (diff) | |
gl_rasterizer: Upload constant buffers with glNamedBufferSubData
Nvidia's OpenGL driver maps gl(Named)BufferSubData with some requirements
to a fast. This path has an extra memcpy but updates the buffer without
orphaning or waiting for previous calls. It can be seen as a better
model for "push constants" that can upload a whole UBO instead of 256
bytes.
This path has some requirements established here:
http://on-demand.gputechconf.com/gtc/2014/presentations/S4379-opengl-44-scene-rendering-techniques.pdf#page=24
Instead of using the stream buffer, this commits moves constant buffers
uploads to calls of glNamedBufferSubData and from my testing it brings a
performance improvement. This is disabled when the vendor is not Nvidia
since it brings performance regressions.
Diffstat (limited to 'src/video_core/renderer_opengl/gl_device.cpp')
| -rw-r--r-- | src/video_core/renderer_opengl/gl_device.cpp | 4 |
1 files changed, 4 insertions, 0 deletions
diff --git a/src/video_core/renderer_opengl/gl_device.cpp b/src/video_core/renderer_opengl/gl_device.cpp index 64de7e425..c65b24c69 100644 --- a/src/video_core/renderer_opengl/gl_device.cpp +++ b/src/video_core/renderer_opengl/gl_device.cpp @@ -51,8 +51,11 @@ bool HasExtension(const std::vector<std::string_view>& images, std::string_view } // Anonymous namespace Device::Device() { + const std::string_view vendor = reinterpret_cast<const char*>(glGetString(GL_VENDOR)); const std::vector extensions = GetExtensions(); + const bool is_nvidia = vendor == "NVIDIA Corporation"; + uniform_buffer_alignment = GetInteger<std::size_t>(GL_UNIFORM_BUFFER_OFFSET_ALIGNMENT); shader_storage_alignment = GetInteger<std::size_t>(GL_SHADER_STORAGE_BUFFER_OFFSET_ALIGNMENT); max_vertex_attributes = GetInteger<u32>(GL_MAX_VERTEX_ATTRIBS); @@ -64,6 +67,7 @@ Device::Device() { has_variable_aoffi = TestVariableAoffi(); has_component_indexing_bug = TestComponentIndexingBug(); has_precise_bug = TestPreciseBug(); + has_fast_buffer_sub_data = is_nvidia; LOG_INFO(Render_OpenGL, "Renderer_VariableAOFFI: {}", has_variable_aoffi); LOG_INFO(Render_OpenGL, "Renderer_ComponentIndexingBug: {}", has_component_indexing_bug); |