ShuriZma
/
duckstation
mirror of https://github.com/stenzek/duckstation
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

GPU/HW: Reduce number of shaders/pipelines in default config 

~3x speedup in shader compile time on a cold cache.
This commit is contained in:
Stenzek 2024-09-19 21:02:09 +10:00
parent a5cd3ee7a8
commit 0538b95d9b
No known key found for this signature in database
2 changed files with 56 additions and 22 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

76
src/core/gpu_hw.cpp
View File

@ -156,6 +156,11 @@ public:
}
~ShaderCompileProgressTracker() = default;
double GetElapsedMilliseconds() const
{
return Common::Timer::ConvertValueToMilliseconds(Common::Timer::GetCurrentValue() - m_start_time);
}
void Increment(u32 progress = 1)
{
m_progress += progress;
@ -170,10 +175,10 @@ public:
private:
std::string m_title;
u64 m_min_time;
u64 m_update_interval;
u64 m_start_time;
u64 m_last_update_time;
Common::Timer::Value m_min_time;
Common::Timer::Value m_update_interval;
Common::Timer::Value m_start_time;
Common::Timer::Value m_last_update_time;
u32 m_progress;
u32 m_total;
};
@ -386,6 +391,8 @@ void GPU_HW::RestoreDeviceContext()
void GPU_HW::UpdateSettings(const Settings& old_settings)
{
const bool prev_force_progressive_scan = m_force_progressive_scan;
GPU::UpdateSettings(old_settings);
const GPUDevice::Features features = g_gpu_device->GetFeatures();
@ -398,7 +405,7 @@ void GPU_HW::UpdateSettings(const Settings& old_settings)
m_pgxp_depth_buffer != g_settings.UsingPGXPDepthBuffer());
const bool shaders_changed =
(m_resolution_scale != resolution_scale || m_multisamples != multisamples ||
m_true_color != g_settings.gpu_true_color ||
m_true_color != g_settings.gpu_true_color || prev_force_progressive_scan != m_force_progressive_scan ||
(multisamples > 0 && g_settings.gpu_per_sample_shading != old_settings.gpu_per_sample_shading) ||
(resolution_scale > 1 && g_settings.gpu_scaled_dithering != old_settings.gpu_scaled_dithering) ||
(resolution_scale > 1 && g_settings.gpu_texture_filter == GPUTextureFilter::Nearest &&
@ -411,8 +418,8 @@ void GPU_HW::UpdateSettings(const Settings& old_settings)
g_settings.gpu_downsample_scale != old_settings.gpu_downsample_scale))) ||
(features.geometry_shaders && g_settings.gpu_wireframe_mode != old_settings.gpu_wireframe_mode) ||
m_pgxp_depth_buffer != g_settings.UsingPGXPDepthBuffer() ||
(features.noperspective_interpolation &&
ShouldDisableColorPerspective() != old_settings.gpu_pgxp_color_correction) ||
(features.noperspective_interpolation && g_settings.gpu_pgxp_enable &&
g_settings.gpu_pgxp_color_correction != old_settings.gpu_pgxp_color_correction) ||
m_allow_sprite_mode !=
ShouldAllowSpriteMode(m_resolution_scale, g_settings.gpu_texture_filter, g_settings.gpu_sprite_texture_filter));
@ -928,6 +935,7 @@ bool GPU_HW::CompilePipelines(Error* error)
const bool per_sample_shading = g_settings.gpu_per_sample_shading && features.per_sample_shading;
const bool force_round_texcoords = (m_resolution_scale > 1 && m_texture_filtering == GPUTextureFilter::Nearest &&
g_settings.gpu_force_round_texcoords);
const bool true_color = g_settings.gpu_true_color;
// Determine when to use shader blending.
// FBFetch is free, we need it for filtering without DSB, or when accurate blending is forced.
@ -974,22 +982,29 @@ bool GPU_HW::CompilePipelines(Error* error)
const u32 active_texture_modes =
m_allow_sprite_mode ? NUM_TEXTURE_MODES :
(NUM_TEXTURE_MODES - (NUM_TEXTURE_MODES - static_cast<u32>(BatchTextureMode::SpriteStart)));
const u32 total_pipelines =
(m_allow_sprite_mode ? 5 : 3) + // vertex shaders
(active_texture_modes * 5 * 9 * 2 * 2 * 2 * (1 + BoolToUInt32(needs_rov_depth))) + // fragment shaders
((m_pgxp_depth_buffer ? 2 : 1) * 5 * 5 * active_texture_modes * 2 * 2 * 2) + // batch pipelines
((m_wireframe_mode != GPUWireframeMode::Disabled) ? 1 : 0) + // wireframe
1 + // fullscreen quad VS
(2 * 2) + // vram fill
(1 + BoolToUInt32(m_write_mask_as_depth)) + // vram copy
(1 + BoolToUInt32(m_write_mask_as_depth)) + // vram write
1 + // vram write replacement
(m_write_mask_as_depth ? 1 : 0) + // mask -> depth
1 + // vram read
2 + // extract/display
((m_downsample_mode != GPUDownsampleMode::Disabled) ? 1 : 0); // downsample
const u32 total_vertex_shaders = (m_allow_sprite_mode ? 5 : 3);
const u32 total_fragment_shaders =
(active_texture_modes * 5 * 9 * 2 * (1 + BoolToUInt32(!true_color)) *
(1 + BoolToUInt32(!m_force_progressive_scan)) * (1 + BoolToUInt32(needs_rov_depth)));
const u32 total_items =
total_vertex_shaders + total_fragment_shaders +
((m_pgxp_depth_buffer ? 2 : 1) * 5 * 5 * active_texture_modes * 2 * (1 + BoolToUInt32(!true_color)) *
(1 + BoolToUInt32(!m_force_progressive_scan))) + // batch pipelines
((m_wireframe_mode != GPUWireframeMode::Disabled) ? 1 : 0) + // wireframe
1 + // fullscreen quad VS
(2 * 2) + // vram fill
(1 + BoolToUInt32(m_write_mask_as_depth)) + // vram copy
(1 + BoolToUInt32(m_write_mask_as_depth)) + // vram write
1 + // vram write replacement
(m_write_mask_as_depth ? 1 : 0) + // mask -> depth
1 + // vram read
2 + // extract/display
((m_downsample_mode != GPUDownsampleMode::Disabled) ? 1 : 0); // downsample
ShaderCompileProgressTracker progress("Compiling Pipelines", total_pipelines);
INFO_LOG("Compiling {} vertex shaders, {} fragment shaders, and {} pipelines.", total_vertex_shaders,
total_fragment_shaders, total_items);
ShaderCompileProgressTracker progress("Compiling Pipelines", total_items);
// vertex shaders - [textured/palette/sprite]
// fragment shaders - [depth_test][render_mode][transparency_mode][texture_mode][check_mask][dithering][interlacing]
@ -1062,8 +1077,16 @@ bool GPU_HW::CompilePipelines(Error* error)
for (u8 dithering = 0; dithering < 2; dithering++)
{
// Never going to draw with dithering on in true color.
if (dithering && true_color)
continue;
for (u8 interlacing = 0; interlacing < 2; interlacing++)
{
// Never going to draw with line skipping in force progressive.
if (interlacing && m_force_progressive_scan)
continue;
const bool sprite = (static_cast<BatchTextureMode>(texture_mode) >= BatchTextureMode::SpriteStart);
const bool uv_limits = ShouldClampUVs(sprite ? m_sprite_texture_filtering : m_texture_filtering);
const BatchTextureMode shader_texmode = static_cast<BatchTextureMode>(
@ -1150,8 +1173,16 @@ bool GPU_HW::CompilePipelines(Error* error)
{
for (u8 dithering = 0; dithering < 2; dithering++)
{
// Never going to draw with dithering on in true color.
if (dithering && true_color)
continue;
for (u8 interlacing = 0; interlacing < 2; interlacing++)
{
// Never going to draw with line skipping in force progressive.
if (interlacing && m_force_progressive_scan)
continue;
for (u8 check_mask = 0; check_mask < 2; check_mask++)
{
const bool textured = (static_cast<BatchTextureMode>(texture_mode) != BatchTextureMode::Disabled);
@ -1605,6 +1636,7 @@ bool GPU_HW::CompilePipelines(Error* error)
#undef UPDATE_PROGRESS
INFO_LOG("Pipeline creation took {:.2f} ms.", progress.GetElapsedMilliseconds());
return true;
}

2
src/core/gpu_hw_shadergen.cpp
View File

@ -703,6 +703,8 @@ std::string GPU_HW_ShaderGen::GenerateBatchFragmentShader(
GPUTextureFilter texture_filtering, bool uv_limits, bool force_round_texcoords, bool dithering, bool interlacing,
bool check_mask, bool use_rov, bool use_rov_depth, bool rov_depth_test)
{
DebugAssert(!m_true_color || !dithering); // Should not be doing dithering+true color.
// TODO: don't write depth for shader blend
DebugAssert(transparency == GPUTransparencyMode::Disabled || render_mode == GPU_HW::BatchRenderMode::ShaderBlend);
DebugAssert(!rov_depth_test || (use_rov && use_rov_depth));