aboutsummaryrefslogtreecommitdiff
path: root/src/video_core/host_shaders
diff options
context:
space:
mode:
authorReinUsesLisp <reinuseslisp@airmail.cc>2021-01-16 20:48:58 -0300
committerReinUsesLisp <reinuseslisp@airmail.cc>2021-02-13 02:17:22 -0300
commit82c2601555b59a94d7160f2fd686cb63d32dd423 (patch)
treecd0ecd865945452fa589b572de614fc487f2f96a /src/video_core/host_shaders
parenta39d9c5194a5fa230ee987ebfc73476f2011d6fa (diff)
video_core: Reimplement the buffer cache
Reimplement the buffer cache using cached bindings and page level granularity for modification tracking. This also drops the usage of shared pointers and virtual functions from the cache. - Bindings are cached, allowing to skip work when the game changes few bits between draws. - OpenGL Assembly shaders no longer copy when a region has been modified from the GPU to emulate constant buffers, instead GL_EXT_memory_object is used to alias sub-buffers within the same allocation. - OpenGL Assembly shaders stream constant buffer data using glProgramBufferParametersIuivNV, from NV_parameter_buffer_object. In theory this should save one hash table resolve inside the driver compared to glBufferSubData. - A new OpenGL stream buffer is implemented based on fences for drivers that are not Nvidia's proprietary, due to their low performance on partial glBufferSubData calls synchronized with 3D rendering (that some games use a lot). - Most optimizations are shared between APIs now, allowing Vulkan to cache more bindings than before, skipping unnecesarry work. This commit adds the necessary infrastructure to use Vulkan object from OpenGL. Overall, it improves performance and fixes some bugs present on the old cache. There are still some edge cases hit by some games that harm performance on some vendors, this are planned to be fixed in later commits.
Diffstat (limited to 'src/video_core/host_shaders')
-rw-r--r--src/video_core/host_shaders/CMakeLists.txt1
-rw-r--r--src/video_core/host_shaders/vulkan_quad_array.comp28
-rw-r--r--src/video_core/host_shaders/vulkan_uint8.comp9
3 files changed, 8 insertions, 30 deletions
diff --git a/src/video_core/host_shaders/CMakeLists.txt b/src/video_core/host_shaders/CMakeLists.txt
index 28f2b8614..970120acc 100644
--- a/src/video_core/host_shaders/CMakeLists.txt
+++ b/src/video_core/host_shaders/CMakeLists.txt
@@ -12,7 +12,6 @@ set(SHADER_FILES
vulkan_blit_depth_stencil.frag
vulkan_present.frag
vulkan_present.vert
- vulkan_quad_array.comp
vulkan_quad_indexed.comp
vulkan_uint8.comp
)
diff --git a/src/video_core/host_shaders/vulkan_quad_array.comp b/src/video_core/host_shaders/vulkan_quad_array.comp
deleted file mode 100644
index 212f4e998..000000000
--- a/src/video_core/host_shaders/vulkan_quad_array.comp
+++ /dev/null
@@ -1,28 +0,0 @@
-// Copyright 2019 yuzu Emulator Project
-// Licensed under GPLv2 or any later version
-// Refer to the license.txt file included.
-
-#version 460 core
-
-layout (local_size_x = 1024) in;
-
-layout (std430, set = 0, binding = 0) buffer OutputBuffer {
- uint output_indexes[];
-};
-
-layout (push_constant) uniform PushConstants {
- uint first;
-};
-
-void main() {
- uint primitive = gl_GlobalInvocationID.x;
- if (primitive * 6 >= output_indexes.length()) {
- return;
- }
-
- const uint quad_map[6] = uint[](0, 1, 2, 0, 2, 3);
- for (uint vertex = 0; vertex < 6; ++vertex) {
- uint index = first + primitive * 4 + quad_map[vertex];
- output_indexes[primitive * 6 + vertex] = index;
- }
-}
diff --git a/src/video_core/host_shaders/vulkan_uint8.comp b/src/video_core/host_shaders/vulkan_uint8.comp
index ad74d7af9..872291670 100644
--- a/src/video_core/host_shaders/vulkan_uint8.comp
+++ b/src/video_core/host_shaders/vulkan_uint8.comp
@@ -16,9 +16,16 @@ layout (std430, set = 0, binding = 1) writeonly buffer OutputBuffer {
uint16_t output_indexes[];
};
+uint AssembleIndex(uint id) {
+ // Most primitive restart indices are 0xFF
+ // Hardcode this to 0xFF for now
+ uint index = uint(input_indexes[id]);
+ return index == 0xFF ? 0xFFFF : index;
+}
+
void main() {
uint id = gl_GlobalInvocationID.x;
if (id < input_indexes.length()) {
- output_indexes[id] = uint16_t(input_indexes[id]);
+ output_indexes[id] = uint16_t(AssembleIndex(id));
}
}