Compare commits

..

1 Commits

Author SHA1 Message Date
lizzie 3de6cfbe4a [shader_recompiler] use reusable stable_vector<> instead of object pools
Signed-off-by: lizzie <lizzie@eden-emu.dev>
2026-09-01 10:11:21 +00:00
62 changed files with 1122 additions and 1215 deletions
-3
View File
@@ -63,6 +63,3 @@ artifacts
/install*
vulkansdk*.exe
*.tar.zst
# Build-time generated external sources
externals/generated
+3 -3
View File
@@ -214,6 +214,8 @@ if(MSVC)
endif()
# TODO(crueter): Cleanup, each dep that has a bundled option should allow to choose between bundled, external, system
cmake_dependent_option(YUZU_USE_BUNDLED_SDL3 "Download bundled SDL3 build" "${MSVC}" "NOT ANDROID" OFF)
option(ENABLE_CUBEB "Enables the cubeb audio backend" ON)
set(EXT_DEFAULT OFF)
@@ -221,8 +223,6 @@ if (MSVC OR ANDROID)
set(EXT_DEFAULT ON)
endif()
option(YUZU_USE_BUNDLED_SDL3 "Download bundled SDL3 build" ${EXT_DEFAULT})
# ffmpeg
option(YUZU_USE_BUNDLED_FFMPEG "Download bundled FFmpeg" ${EXT_DEFAULT})
cmake_dependent_option(YUZU_USE_EXTERNAL_FFMPEG "Build FFmpeg from external source" "${SOLARIS}" "NOT WIN32 AND NOT ANDROID" OFF)
@@ -238,7 +238,7 @@ option(YUZU_USE_BUNDLED_SIRIT "Download bundled sirit" ${BUNDLED_SIRIT_DEFAULT})
# FreeBSD 15+ has libusb, versions below should disable it
cmake_dependent_option(ENABLE_LIBUSB "Enable the use of LibUSB" ON "WIN32 OR LINUX OR FREEBSD OR APPLE" OFF)
cmake_dependent_option(ENABLE_OPENGL "Enable OpenGL" ON "NOT (WIN32 AND ARCHITECTURE_arm64) AND NOT APPLE AND NOT ANDROID" OFF)
cmake_dependent_option(ENABLE_OPENGL "Enable OpenGL" ON "NOT (WIN32 AND ARCHITECTURE_arm64) AND NOT APPLE" OFF)
mark_as_advanced(FORCE ENABLE_OPENGL)
option(ENABLE_WEB_SERVICE "Enable web services (telemetry, etc.)" ON)
+6
View File
@@ -186,6 +186,12 @@
"repo": "eden-emulator/oaknut",
"version": "v2.0.3"
},
"oboe": {
"bundled": true,
"hash": "ce4011afe7345370d4ead3b891cd69a5ef224b129535783586c0ca75051d303ed446e6c7f10bde8da31fff58d6e307f1732a3ffd03b249f9ef1fd48fd4132715",
"repo": "google/oboe",
"version": "1.10.0"
},
"openssl": {
"hash": "29002ce50cb95a4f4f1d0e9d3f684401fbd4eac34203dc2eef3b6334af5d44aa46bf788b63a6f5c139c383eafb7269ae87a58a9a3ad5912903b9773e545ccc0a",
"min_version": "3.0.0",
+1 -1
View File
@@ -76,7 +76,7 @@ The following options are desktop only.
- `ENABLE_LIBUSB` (ON) Enable the use of the libusb input backend (HIGHLY RECOMMENDED)
- `ENABLE_OPENGL` (ON) Enable the OpenGL graphics backend
- Unavailable on Windows/ARM64 and on Android
- Unavailable on Windows/ARM64
- You probably shouldn't turn this off.
### Qt
+43 -37
View File
@@ -136,50 +136,49 @@ if(ENABLE_CUBEB)
endif()
endif()
if (NOT YUZU_USE_BUNDLED_SDL3)
if (NOT WIN32)
# Yuzu itself needs: Atomic Audio Events Joystick Haptic Sensor Threads Timers
# Since 2.0.18 Atomic+Threads required for HIDAPI/libusb (see https://github.com/libsdl-org/SDL/issues/5095)
# Yuzu-cmd also needs: Video (depends on Loadso/Dlopen)
# CPUinfo also required for SDL Audio, at least until 2.28.0 (see https://github.com/libsdl-org/SDL/issues/7809)
set(SDL_UNUSED_SUBSYSTEMS
File Filesystem
Locale Power Render)
foreach(_SUB ${SDL_UNUSED_SUBSYSTEMS})
string(TOUPPER ${_SUB} _OPT)
set(SDL_${_OPT} OFF)
endforeach()
if (NOT ANDROID)
if (NOT YUZU_USE_BUNDLED_SDL3)
if (NOT WIN32)
# Yuzu itself needs: Atomic Audio Events Joystick Haptic Sensor Threads Timers
# Since 2.0.18 Atomic+Threads required for HIDAPI/libusb (see https://github.com/libsdl-org/SDL/issues/5095)
# Yuzu-cmd also needs: Video (depends on Loadso/Dlopen)
# CPUinfo also required for SDL Audio, at least until 2.28.0 (see https://github.com/libsdl-org/SDL/issues/7809)
set(SDL_UNUSED_SUBSYSTEMS
File Filesystem
Locale Power Render)
foreach(_SUB ${SDL_UNUSED_SUBSYSTEMS})
string(TOUPPER ${_SUB} _OPT)
set(SDL_${_OPT} OFF)
endforeach()
set(HIDAPI ON)
endif()
set(HIDAPI ON)
endif()
if (APPLE)
set(SDL_FILE ON)
endif()
if (APPLE)
set(SDL_FILE ON)
endif()
AddJsonPackage(sdl3)
else()
message(STATUS "Using bundled SDL3")
AddJsonPackage(sdl3-ci)
# copy sdl java sources for consumption by Gradle
if (ANDROID)
file(COPY ${SDL3_SOURCE_DIR}/java DESTINATION ${CMAKE_CURRENT_SOURCE_DIR}/generated/sdl)
endif()
endif()
# Normalize SDL3 link target across package variants.
# Some SDL3 packages export only SDL3::SDL3-shared or SDL3::SDL3-static.
if (NOT TARGET SDL3::SDL3)
if (TARGET SDL3::SDL3-shared)
add_library(SDL3::SDL3 ALIAS SDL3::SDL3-shared)
elseif (TARGET SDL3::SDL3-static)
add_library(SDL3::SDL3 ALIAS SDL3::SDL3-static)
AddJsonPackage(sdl3)
else()
message(FATAL_ERROR "SDL3 package found, but no usable SDL3 target was exported")
message(STATUS "Using bundled SDL3")
AddJsonPackage(sdl3-ci)
endif()
# Normalize SDL3 link target across package variants.
# Some SDL3 packages export only SDL3::SDL3-shared or SDL3::SDL3-static.
if (NOT TARGET SDL3::SDL3)
if (TARGET SDL3::SDL3-shared)
add_library(SDL3::SDL3 ALIAS SDL3::SDL3-shared)
elseif (TARGET SDL3::SDL3-static)
add_library(SDL3::SDL3 ALIAS SDL3::SDL3-static)
else()
message(FATAL_ERROR "SDL3 package found, but no usable SDL3 target was exported")
endif()
endif()
endif()
set(BUILD_SHARED_LIBS OFF)
# SPIRV Headers
AddJsonPackage(spirv-headers)
@@ -386,6 +385,13 @@ if (YUZU_CRASH_DUMPS AND NOT TARGET libbreakpad_client)
endif()
endif()
# oboe
if (ANDROID)
AddJsonPackage(oboe)
add_library(oboe::oboe ALIAS oboe)
endif()
if (APPLE)
# moltenvk
if (NOT YUZU_USE_BUNDLED_MOLTENVK)
-1
View File
@@ -11,7 +11,6 @@
#include <limits>
#include <span>
#include <array>
#include <algorithm>
#include <time.h>
namespace Tz {
-7
View File
@@ -272,13 +272,6 @@ android {
resValue("string", "app_name_suffixed", "$currentName$suffix")
resValue("string", "app_name", "Eden$suffix")
}
sourceSets {
named("main") {
java.srcDir("${edenDir}/externals/generated/sdl/java")
kotlin.srcDir("${edenDir}/externals/generated/sdl/java")
}
}
}
idea {
-3
View File
@@ -9,9 +9,6 @@
-keep class org.ini4j.spi.IniBuilder
-keep class org.ini4j.spi.IniFormatter
-keep class org.libsdl.app.** { *; }
-keepclassmembers class org.libsdl.app.** { *; }
# Suppress warnings for R8
-dontwarn org.bouncycastle.jsse.BCSSLParameters
-dontwarn org.bouncycastle.jsse.BCSSLSocket
@@ -16,7 +16,6 @@ import android.widget.TextView
import androidx.annotation.Keep
import androidx.core.net.toUri
import com.google.android.material.dialog.MaterialAlertDialogBuilder
import org.libsdl.app.SDL
import java.lang.ref.WeakReference
import org.yuzu.yuzu_emu.activities.EmulationActivity
import org.yuzu.yuzu_emu.fragments.CoreErrorDialogFragment
@@ -53,8 +52,6 @@ object NativeLibrary {
init {
try {
System.loadLibrary("yuzu-android")
SDL.setupJNI()
initJvm()
} catch (ex: UnsatisfiedLinkError) {
error("[NativeLibrary] $ex")
}
@@ -370,7 +367,6 @@ object NativeLibrary {
NetPlayManager.clearChat()
}
external fun initJvm()
external fun initMultiplayer()
@Keep
+1 -1
View File
@@ -22,7 +22,7 @@ add_library(yuzu-android SHARED
set_property(TARGET yuzu-android PROPERTY IMPORTED_LOCATION ${FFmpeg_LIBRARY_DIR})
target_link_libraries(yuzu-android PRIVATE audio_core common core input_common frontend_common video_core)
target_link_libraries(yuzu-android PRIVATE android camera2ndk EGL glad jnigraphics log GLESv2)
target_link_libraries(yuzu-android PRIVATE android camera2ndk EGL glad jnigraphics log)
if (ARCHITECTURE_arm64)
target_link_libraries(yuzu-android PRIVATE adrenotools)
endif()
-9
View File
@@ -1730,15 +1730,6 @@ jint Java_org_yuzu_yuzu_1emu_NativeLibrary_loadAmiibo(JNIEnv* env, jobject jobj,
return static_cast<jint>(info);
}
extern "C" JNIEXPORT void JNICALL
Java_org_yuzu_yuzu_1emu_NativeLibrary_initJvm(JNIEnv *env, jclass clazz) {
JavaVM *vm;
if (env->GetJavaVM(&vm) != JNI_OK)
return;
Common::Android::Initialize(vm, env);
}
JNIEXPORT void JNICALL
Java_org_yuzu_yuzu_1emu_NativeLibrary_initMultiplayer(
JNIEnv* env, [[maybe_unused]] jobject obj) {
@@ -800,6 +800,7 @@
<string name="theme_mode_dark">تاریک</string>
<!-- Audio output engines -->
<string name="oboe">oboe</string>
<string name="cubeb">cubeb</string>
<!-- Anisotropic filtering options -->
@@ -465,13 +465,13 @@
<string-array name="outputEngineEntries">
<item>@string/auto</item>
<item>@string/sdl3</item>
<item>@string/oboe</item>
<item>@string/cubeb</item>
<item>@string/string_null</item>
</string-array>
<integer-array name="outputEngineValues">
<item>0</item>
<item>2</item>
<item>4</item>
<item>1</item>
<item>3</item>
</integer-array>
@@ -1240,7 +1240,7 @@
<string name="theme_mode_dark">Dark</string>
<!-- Audio output engines -->
<string name="sdl3" translatable="false">SDL3</string>
<string name="oboe" translatable="false">oboe</string>
<string name="cubeb" translatable="false">cubeb</string>
<!-- Anisotropic filtering options -->
+14 -5
View File
@@ -238,11 +238,20 @@ if (ENABLE_CUBEB)
target_compile_definitions(audio_core PRIVATE HAVE_CUBEB=1)
endif()
target_sources(audio_core PRIVATE
sink/sdl3_sink.cpp
sink/sdl3_sink.h)
if(ANDROID)
target_sources(audio_core PRIVATE
sink/oboe_sink.cpp
sink/oboe_sink.h)
target_link_libraries(audio_core PRIVATE SDL3::SDL3)
target_compile_definitions(audio_core PRIVATE HAVE_SDL3)
target_link_libraries(audio_core PRIVATE oboe)
target_compile_definitions(audio_core PUBLIC HAVE_OBOE)
else()
target_sources(audio_core PRIVATE
sink/sdl3_sink.cpp
sink/sdl3_sink.h)
target_link_libraries(audio_core PRIVATE SDL3::SDL3)
target_compile_definitions(audio_core PRIVATE HAVE_SDL3)
endif()
create_target_directory_groups(audio_core)
+230
View File
@@ -0,0 +1,230 @@
// SPDX-FileCopyrightText: Copyright 2026 Eden Emulator Project
// SPDX-License-Identifier: GPL-3.0-or-later
// SPDX-FileCopyrightText: Copyright 2023 yuzu Emulator Project
// SPDX-License-Identifier: GPL-2.0-or-later
#include <span>
#include <vector>
#include <oboe/Oboe.h>
#include "audio_core/common/common.h"
#include "audio_core/sink/oboe_sink.h"
#include "audio_core/sink/sink_stream.h"
#include "common/logging.h"
#include "common/scope_exit.h"
#include "core/core.h"
namespace AudioCore::Sink {
class OboeSinkStream final : public SinkStream,
public oboe::AudioStreamDataCallback,
public oboe::AudioStreamErrorCallback {
public:
explicit OboeSinkStream(Core::System& system_, StreamType type_, const std::string& name_,
u32 system_channels_)
: SinkStream(system_, type_) {
name = name_;
system_channels = system_channels_;
this->OpenStream();
}
~OboeSinkStream() override {
LOG_INFO(Audio_Sink, "Destroyed Oboe stream");
}
void Finalize() override {
this->Stop();
m_stream.reset();
}
void Start(bool resume = false) override {
if (!m_stream || !paused) {
return;
}
paused = false;
if (m_stream->start() != oboe::Result::OK) {
LOG_CRITICAL(Audio_Sink, "Error starting Oboe stream");
}
}
void Stop() override {
if (!m_stream || paused) {
return;
}
this->SignalPause();
if (m_stream->stop() != oboe::Result::OK) {
LOG_CRITICAL(Audio_Sink, "Error stopping Oboe stream");
}
}
public:
static s32 QueryChannelCount(oboe::Direction direction) {
std::shared_ptr<oboe::AudioStream> temp_stream;
oboe::AudioStreamBuilder builder;
const auto result = ConfigureBuilder(builder, direction)->openStream(temp_stream);
if (result == oboe::Result::OK) {
return temp_stream->getChannelCount() >= 6 ? 6 : 2;
}
LOG_ERROR(Audio_Sink, "Failed to open {} stream. Using default channel count 2",
direction == oboe::Direction::Output ? "output" : "input");
return 2;
}
protected:
oboe::DataCallbackResult onAudioReady(oboe::AudioStream*, void* audio_data,
s32 num_buffer_frames) override {
const size_t num_channels = this->GetDeviceChannels();
const size_t frame_size = num_channels;
const size_t num_frames = static_cast<size_t>(num_buffer_frames);
if (type == StreamType::In) {
std::span<const s16> input_buffer{reinterpret_cast<const s16*>(audio_data),
num_frames * frame_size};
this->ProcessAudioIn(input_buffer, num_frames);
} else {
std::span<s16> output_buffer{reinterpret_cast<s16*>(audio_data),
num_frames * frame_size};
this->ProcessAudioOutAndRender(output_buffer, num_frames);
}
return oboe::DataCallbackResult::Continue;
}
void onErrorAfterClose(oboe::AudioStream*, oboe::Result) override {
LOG_INFO(Audio_Sink, "Audio stream closed, reinitializing");
if (this->OpenStream()) {
m_stream->start();
}
}
private:
static oboe::AudioStreamBuilder* ConfigureBuilder(oboe::AudioStreamBuilder& builder,
oboe::Direction direction) {
// TODO: investigate callback delay issues when using AAudio
return builder.setPerformanceMode(oboe::PerformanceMode::LowLatency)
->setAudioApi(oboe::AudioApi::OpenSLES)
->setDirection(direction)
->setSampleRate(TargetSampleRate)
->setSampleRateConversionQuality(oboe::SampleRateConversionQuality::High)
->setFormat(oboe::AudioFormat::I16)
->setFormatConversionAllowed(true)
->setUsage(oboe::Usage::Game)
->setBufferCapacityInFrames(TargetSampleCount * 2);
}
bool OpenStream() {
const auto direction = [&]() {
switch (type) {
case StreamType::In:
return oboe::Direction::Input;
case StreamType::Out:
case StreamType::Render:
return oboe::Direction::Output;
default:
ASSERT(false);
return oboe::Direction::Output;
}
}();
const auto expected_channels = QueryChannelCount(direction);
const auto expected_mask = [&]() {
switch (expected_channels) {
case 1:
return oboe::ChannelMask::Mono;
case 2:
return oboe::ChannelMask::Stereo;
case 6:
return oboe::ChannelMask::CM5Point1;
default:
ASSERT(false);
return oboe::ChannelMask::Unspecified;
}
}();
oboe::AudioStreamBuilder builder;
const auto result = ConfigureBuilder(builder, direction)
->setChannelCount(expected_channels)
->setChannelMask(expected_mask)
->setChannelConversionAllowed(true)
->setDataCallback(this)
->setErrorCallback(this)
->openStream(m_stream);
ASSERT(result == oboe::Result::OK);
return result == oboe::Result::OK && this->SetStreamProperties();
}
bool SetStreamProperties() {
ASSERT(m_stream);
m_stream->setBufferSizeInFrames(TargetSampleCount * 2);
device_channels = m_stream->getChannelCount();
const auto sample_rate = m_stream->getSampleRate();
const auto buffer_capacity = m_stream->getBufferCapacityInFrames();
const auto stream_backend =
m_stream->getAudioApi() == oboe::AudioApi::AAudio ? "AAudio" : "OpenSLES";
LOG_INFO(Audio_Sink, "Opened Oboe {} stream with {} channels sample rate {} capacity {}",
stream_backend, device_channels, sample_rate, buffer_capacity);
return true;
}
std::shared_ptr<oboe::AudioStream> m_stream{};
};
OboeSink::OboeSink() {
// TODO: This is not generally knowable
// The channel count is distinct based on direction and can change
device_channels = OboeSinkStream::QueryChannelCount(oboe::Direction::Output);
}
OboeSink::~OboeSink() = default;
SinkStream* OboeSink::AcquireSinkStream(Core::System& system, u32 system_channels,
const std::string& name, StreamType type) {
SinkStreamPtr& stream = sink_streams.emplace_back(
std::make_unique<OboeSinkStream>(system, type, name, system_channels));
return stream.get();
}
void OboeSink::CloseStream(SinkStream* to_remove) {
sink_streams.remove_if([&](auto& stream) { return stream.get() == to_remove; });
}
void OboeSink::CloseStreams() {
sink_streams.clear();
}
f32 OboeSink::GetDeviceVolume() const {
if (sink_streams.empty()) {
return 1.0f;
}
return sink_streams.front()->GetDeviceVolume();
}
void OboeSink::SetDeviceVolume(f32 volume) {
for (auto& stream : sink_streams) {
stream->SetDeviceVolume(volume);
}
}
void OboeSink::SetSystemVolume(f32 volume) {
for (auto& stream : sink_streams) {
stream->SetSystemVolume(volume);
}
}
} // namespace AudioCore::Sink
+75
View File
@@ -0,0 +1,75 @@
// SPDX-FileCopyrightText: Copyright 2023 yuzu Emulator Project
// SPDX-License-Identifier: GPL-2.0-or-later
#pragma once
#include <list>
#include <string>
#include "audio_core/sink/sink.h"
namespace Core {
class System;
}
namespace AudioCore::Sink {
class SinkStream;
class OboeSink final : public Sink {
public:
explicit OboeSink();
~OboeSink() override;
/**
* Create a new sink stream.
*
* @param system - Core system.
* @param system_channels - Number of channels the audio system expects.
* May differ from the device's channel count.
* @param name - Name of this stream.
* @param type - Type of this stream, render/in/out.
*
* @return A pointer to the created SinkStream
*/
SinkStream* AcquireSinkStream(Core::System& system, u32 system_channels,
const std::string& name, StreamType type) override;
/**
* Close a given stream.
*
* @param stream - The stream to close.
*/
void CloseStream(SinkStream* stream) override;
/**
* Close all streams.
*/
void CloseStreams() override;
/**
* Get the device volume. Set from calls to the IAudioDevice service.
*
* @return Volume of the device.
*/
f32 GetDeviceVolume() const override;
/**
* Set the device volume. Set from calls to the IAudioDevice service.
*
* @param volume - New volume of the device.
*/
void SetDeviceVolume(f32 volume) override;
/**
* Set the system volume. Comes from the audio system using this stream.
*
* @param volume - New volume of the system.
*/
void SetSystemVolume(f32 volume) override;
private:
/// List of streams managed by this sink
std::list<SinkStreamPtr> sink_streams{};
};
} // namespace AudioCore::Sink
+19
View File
@@ -10,6 +10,9 @@
#include <vector>
#include "audio_core/sink/sink_details.h"
#ifdef HAVE_OBOE
#include "audio_core/sink/oboe_sink.h"
#endif
#ifdef HAVE_CUBEB
#include "audio_core/sink/cubeb_sink.h"
#endif
@@ -40,8 +43,24 @@ struct SinkDetails {
/// SuitableFn is_suitable; // REVERTED FOR LatencyFn latency ABOVE - DIABLO 3 FIX
};
// NOTE TO PROBABLY FIX LATER FOR ANDROID - the return value "0u" for the first HAVE_OBOE
// section below was just copied from the null section so there's a somewhat valid value
// being returned, since the previous "true" value probably isn't compatible with the
// previous EA-3833 code. (HAVE_OBOE was introduced in a later release.) Eventually need
// to change "0u" for something else directly from the oboe_sink.cpp functions.
// sink_details is ordered in terms of desirability, with the best choice at the top.
constexpr SinkDetails sink_details[] = {
#ifdef HAVE_OBOE
SinkDetails{
Settings::AudioEngine::Oboe,
[](std::string_view device_id) -> std::unique_ptr<Sink> {
return std::make_unique<OboeSink>();
},
[](bool capture) { return std::vector<std::string>{"Default"}; },
[]() { return 0u; },
},
#endif
#ifdef HAVE_CUBEB
SinkDetails{
Settings::AudioEngine::Cubeb,
+4 -4
View File
@@ -1,7 +1,7 @@
// SPDX-FileCopyrightText: Copyright 2023 yuzu Emulator Project
// SPDX-License-Identifier: GPL-2.0-or-later
// SPDX-FileCopyrightText: Copyright 2026 Eden Emulator Project
// SPDX-FileCopyrightText: 2025 Eden Emulator Project
// SPDX-License-Identifier: GPL-3.0-or-later
#include "android_common.h"
@@ -51,7 +51,7 @@ jstring ToJString(JNIEnv* env, std::u16string_view str) {
}
double GetJDouble(JNIEnv* env, jobject jdouble) {
return env->CallDoubleMethod(jdouble, GetDoubleValueMethod());
return env->GetDoubleField(jdouble, GetDoubleValueField());
}
jobject ToJDouble(JNIEnv* env, double value) {
@@ -59,7 +59,7 @@ jobject ToJDouble(JNIEnv* env, double value) {
}
s32 GetJInteger(JNIEnv* env, jobject jinteger) {
return env->CallIntMethod(jinteger, GetIntegerValueMethod());
return env->GetIntField(jinteger, GetIntegerValueField());
}
jobject ToJInteger(JNIEnv* env, s32 value) {
@@ -67,7 +67,7 @@ jobject ToJInteger(JNIEnv* env, s32 value) {
}
bool GetJBoolean(JNIEnv* env, jobject jboolean) {
return env->CallBooleanMethod(jboolean, GetBooleanValueMethod());
return env->GetBooleanField(jboolean, GetBooleanValueField());
}
jobject ToJBoolean(JNIEnv* env, bool value) {
+192 -188
View File
@@ -63,15 +63,15 @@ static jfieldID s_patch_title_id_field;
static jclass s_double_class;
static jmethodID s_double_constructor;
static jmethodID s_double_value_method;
static jfieldID s_double_value_field;
static jclass s_integer_class;
static jmethodID s_integer_constructor;
static jmethodID s_integer_value_method;
static jfieldID s_integer_value_field;
static jclass s_boolean_class;
static jmethodID s_boolean_constructor;
static jmethodID s_boolean_value_method;
static jfieldID s_boolean_value_field;
static jclass s_player_input_class;
static jmethodID s_player_input_constructor;
@@ -103,6 +103,7 @@ static jmethodID s_clear_chat;
static constexpr jint JNI_VERSION = JNI_VERSION_1_6;
namespace Common::Android {
JNIEnv *GetEnvForThread() {
thread_local static struct OwnedEnv {
OwnedEnv() {
@@ -298,8 +299,8 @@ namespace Common::Android {
return s_double_constructor;
}
jmethodID GetDoubleValueMethod() {
return s_double_value_method;
jfieldID GetDoubleValueField() {
return s_double_value_field;
}
jclass GetIntegerClass() {
@@ -310,8 +311,8 @@ namespace Common::Android {
return s_integer_constructor;
}
jmethodID GetIntegerValueMethod() {
return s_integer_value_method;
jfieldID GetIntegerValueField() {
return s_integer_value_field;
}
jclass GetBooleanClass() {
@@ -322,8 +323,8 @@ namespace Common::Android {
return s_boolean_constructor;
}
jmethodID GetBooleanValueMethod() {
return s_boolean_value_method;
jfieldID GetBooleanValueField() {
return s_boolean_value_field;
}
jclass GetPlayerInputClass() {
@@ -428,6 +429,188 @@ namespace Common::Android {
jint InitFFmpegOnLoad(JavaVM *vm);
jint JNI_OnLoad(JavaVM *vm, void *reserved) {
s_java_vm = vm;
InitFFmpegOnLoad(vm);
JNIEnv *env;
if (vm->GetEnv(reinterpret_cast<void **>(&env), JNI_VERSION) != JNI_OK)
return JNI_ERR;
// Initialize Java classes
const jclass native_library_class = env->FindClass("org/yuzu/yuzu_emu/NativeLibrary");
s_native_library_class = reinterpret_cast<jclass>(env->NewGlobalRef(native_library_class));
s_disk_cache_progress_class = reinterpret_cast<jclass>(env->NewGlobalRef(
env->FindClass("org/yuzu/yuzu_emu/disk_shader_cache/DiskShaderCacheProgress")));
s_load_callback_stage_class = reinterpret_cast<jclass>(env->NewGlobalRef(env->FindClass(
"org/yuzu/yuzu_emu/disk_shader_cache/DiskShaderCacheProgress$LoadCallbackStage")));
const jclass game_dir_class = env->FindClass("org/yuzu/yuzu_emu/model/GameDir");
s_game_dir_class = reinterpret_cast<jclass>(env->NewGlobalRef(game_dir_class));
s_game_dir_constructor = env->GetMethodID(game_dir_class, "<init>",
"(Ljava/lang/String;Z)V");
env->DeleteLocalRef(game_dir_class);
// Initialize methods
s_exit_emulation_activity =
env->GetStaticMethodID(s_native_library_class, "exitEmulationActivity", "(I)V");
s_disk_cache_load_progress =
env->GetStaticMethodID(s_disk_cache_progress_class, "loadProgress", "(III)V");
s_copy_to_storage = env->GetStaticMethodID(s_native_library_class, "copyFileToStorage",
"(Ljava/lang/String;Ljava/lang/String;)Z");
s_file_exists = env->GetStaticMethodID(s_native_library_class, "exists",
"(Ljava/lang/String;)Z");
s_file_extension = env->GetStaticMethodID(s_native_library_class, "getFileExtension",
"(Ljava/lang/String;)Ljava/lang/String;");
s_on_emulation_started =
env->GetStaticMethodID(s_native_library_class, "onEmulationStarted", "()V");
s_on_emulation_stopped =
env->GetStaticMethodID(s_native_library_class, "onEmulationStopped", "(I)V");
s_on_program_changed =
env->GetStaticMethodID(s_native_library_class, "onProgramChanged", "(I)V");
const jclass game_class = env->FindClass("org/yuzu/yuzu_emu/model/Game");
s_game_class = reinterpret_cast<jclass>(env->NewGlobalRef(game_class));
s_game_constructor = env->GetMethodID(game_class, "<init>",
"(Ljava/lang/String;Ljava/lang/String;Ljava/lang/"
"String;Ljava/lang/String;Ljava/lang/String;Z)V");
s_game_title_field = env->GetFieldID(game_class, "title", "Ljava/lang/String;");
s_game_path_field = env->GetFieldID(game_class, "path", "Ljava/lang/String;");
s_game_program_id_field = env->GetFieldID(game_class, "programId", "Ljava/lang/String;");
s_game_developer_field = env->GetFieldID(game_class, "developer", "Ljava/lang/String;");
s_game_version_field = env->GetFieldID(game_class, "version", "Ljava/lang/String;");
s_game_is_homebrew_field = env->GetFieldID(game_class, "isHomebrew", "Z");
env->DeleteLocalRef(game_class);
const jclass string_class = env->FindClass("java/lang/String");
s_string_class = reinterpret_cast<jclass>(env->NewGlobalRef(string_class));
env->DeleteLocalRef(string_class);
const jclass pair_class = env->FindClass("kotlin/Pair");
s_pair_class = reinterpret_cast<jclass>(env->NewGlobalRef(pair_class));
s_pair_constructor =
env->GetMethodID(pair_class, "<init>", "(Ljava/lang/Object;Ljava/lang/Object;)V");
s_pair_first_field = env->GetFieldID(pair_class, "first", "Ljava/lang/Object;");
s_pair_second_field = env->GetFieldID(pair_class, "second", "Ljava/lang/Object;");
env->DeleteLocalRef(pair_class);
const jclass overlay_control_data_class =
env->FindClass("org/yuzu/yuzu_emu/overlay/model/OverlayControlData");
s_overlay_control_data_class =
reinterpret_cast<jclass>(env->NewGlobalRef(overlay_control_data_class));
s_overlay_control_data_constructor =
env->GetMethodID(overlay_control_data_class, "<init>",
"(Ljava/lang/String;ZLkotlin/Pair;Lkotlin/Pair;Lkotlin/Pair;F)V");
s_overlay_control_data_id_field =
env->GetFieldID(overlay_control_data_class, "id", "Ljava/lang/String;");
s_overlay_control_data_enabled_field =
env->GetFieldID(overlay_control_data_class, "enabled", "Z");
s_overlay_control_data_landscape_position_field =
env->GetFieldID(overlay_control_data_class, "landscapePosition", "Lkotlin/Pair;");
s_overlay_control_data_portrait_position_field =
env->GetFieldID(overlay_control_data_class, "portraitPosition", "Lkotlin/Pair;");
s_overlay_control_data_foldable_position_field =
env->GetFieldID(overlay_control_data_class, "foldablePosition", "Lkotlin/Pair;");
s_overlay_control_data_individual_scale_field =
env->GetFieldID(overlay_control_data_class, "individualScale", "F");
env->DeleteLocalRef(overlay_control_data_class);
const jclass patch_class = env->FindClass("org/yuzu/yuzu_emu/model/Patch");
s_patch_class = reinterpret_cast<jclass>(env->NewGlobalRef(patch_class));
s_patch_constructor = env->GetMethodID(
patch_class, "<init>",
"(ZLjava/lang/String;Ljava/lang/String;ILjava/lang/String;Ljava/lang/String;JI)V");
s_patch_enabled_field = env->GetFieldID(patch_class, "enabled", "Z");
s_patch_name_field = env->GetFieldID(patch_class, "name", "Ljava/lang/String;");
s_patch_version_field = env->GetFieldID(patch_class, "version", "Ljava/lang/String;");
s_patch_type_field = env->GetFieldID(patch_class, "type", "I");
s_patch_program_id_field = env->GetFieldID(patch_class, "programId", "Ljava/lang/String;");
s_patch_title_id_field = env->GetFieldID(patch_class, "titleId", "Ljava/lang/String;");
env->DeleteLocalRef(patch_class);
const jclass double_class = env->FindClass("java/lang/Double");
s_double_class = reinterpret_cast<jclass>(env->NewGlobalRef(double_class));
s_double_constructor = env->GetMethodID(double_class, "<init>", "(D)V");
s_double_value_field = env->GetFieldID(double_class, "value", "D");
env->DeleteLocalRef(double_class);
const jclass int_class = env->FindClass("java/lang/Integer");
s_integer_class = reinterpret_cast<jclass>(env->NewGlobalRef(int_class));
s_integer_constructor = env->GetMethodID(int_class, "<init>", "(I)V");
s_integer_value_field = env->GetFieldID(int_class, "value", "I");
env->DeleteLocalRef(int_class);
const jclass boolean_class = env->FindClass("java/lang/Boolean");
s_boolean_class = reinterpret_cast<jclass>(env->NewGlobalRef(boolean_class));
s_boolean_constructor = env->GetMethodID(boolean_class, "<init>", "(Z)V");
s_boolean_value_field = env->GetFieldID(boolean_class, "value", "Z");
env->DeleteLocalRef(boolean_class);
const jclass player_input_class =
env->FindClass("org/yuzu/yuzu_emu/features/input/model/PlayerInput");
s_player_input_class = reinterpret_cast<jclass>(env->NewGlobalRef(player_input_class));
s_player_input_constructor = env->GetMethodID(
player_input_class, "<init>",
"(Z[Ljava/lang/String;[Ljava/lang/String;[Ljava/lang/String;ZIJJJJLjava/lang/String;Z)V");
s_player_input_connected_field = env->GetFieldID(player_input_class, "connected", "Z");
s_player_input_buttons_field =
env->GetFieldID(player_input_class, "buttons", "[Ljava/lang/String;");
s_player_input_analogs_field =
env->GetFieldID(player_input_class, "analogs", "[Ljava/lang/String;");
s_player_input_motions_field =
env->GetFieldID(player_input_class, "motions", "[Ljava/lang/String;");
s_player_input_vibration_enabled_field =
env->GetFieldID(player_input_class, "vibrationEnabled", "Z");
s_player_input_vibration_strength_field =
env->GetFieldID(player_input_class, "vibrationStrength", "I");
s_player_input_body_color_left_field =
env->GetFieldID(player_input_class, "bodyColorLeft", "J");
s_player_input_body_color_right_field =
env->GetFieldID(player_input_class, "bodyColorRight", "J");
s_player_input_button_color_left_field =
env->GetFieldID(player_input_class, "buttonColorLeft", "J");
s_player_input_button_color_right_field =
env->GetFieldID(player_input_class, "buttonColorRight", "J");
s_player_input_profile_name_field =
env->GetFieldID(player_input_class, "profileName", "Ljava/lang/String;");
s_player_input_use_system_vibrator_field =
env->GetFieldID(player_input_class, "useSystemVibrator", "Z");
env->DeleteLocalRef(player_input_class);
const jclass yuzu_input_device_interface =
env->FindClass("org/yuzu/yuzu_emu/features/input/YuzuInputDevice");
s_yuzu_input_device_interface =
reinterpret_cast<jclass>(env->NewGlobalRef(yuzu_input_device_interface));
s_yuzu_input_device_get_name =
env->GetMethodID(yuzu_input_device_interface, "getName", "()Ljava/lang/String;");
s_yuzu_input_device_get_guid =
env->GetMethodID(yuzu_input_device_interface, "getGUID", "()Ljava/lang/String;");
s_yuzu_input_device_get_port = env->GetMethodID(yuzu_input_device_interface, "getPort",
"()I");
s_yuzu_input_device_get_supports_vibration =
env->GetMethodID(yuzu_input_device_interface, "getSupportsVibration", "()Z");
s_yuzu_input_device_vibrate = env->GetMethodID(yuzu_input_device_interface, "vibrate",
"(F)V");
s_yuzu_input_device_get_axes =
env->GetMethodID(yuzu_input_device_interface, "getAxes", "()[Ljava/lang/Integer;");
s_yuzu_input_device_has_keys =
env->GetMethodID(yuzu_input_device_interface, "hasKeys", "([I)[Z");
env->DeleteLocalRef(yuzu_input_device_interface);
s_add_netplay_message = env->GetStaticMethodID(s_native_library_class, "addNetPlayMessage",
"(ILjava/lang/String;)V");
s_clear_chat = env->GetStaticMethodID(s_native_library_class, "clearChat", "()V");
// Initialize Android Storage
Common::FS::Android::RegisterCallbacks(env, s_native_library_class);
// Initialize applets
Common::Android::SoftwareKeyboard::InitJNI(env);
Common::Android::WebBrowser::InitJNI(env);
return JNI_VERSION;
}
void JNI_OnUnload(JavaVM *vm, void *reserved) {
JNIEnv *env;
if (vm->GetEnv(reinterpret_cast<void **>(&env), JNI_VERSION) != JNI_OK) {
@@ -462,183 +645,4 @@ namespace Common::Android {
}
#endif
void Initialize(JavaVM* vm, JNIEnv *env) {
s_java_vm = vm;
InitFFmpegOnLoad(vm);
if (env->ExceptionCheck()) {
env->ExceptionClear();
}
// Initialize Java classes
const jclass native_library_class = env->FindClass("org/yuzu/yuzu_emu/NativeLibrary");
s_native_library_class = reinterpret_cast<jclass>(env->NewGlobalRef(native_library_class));
s_disk_cache_progress_class = reinterpret_cast<jclass>(env->NewGlobalRef(
env->FindClass("org/yuzu/yuzu_emu/disk_shader_cache/DiskShaderCacheProgress")));
s_load_callback_stage_class = reinterpret_cast<jclass>(env->NewGlobalRef(env->FindClass(
"org/yuzu/yuzu_emu/disk_shader_cache/DiskShaderCacheProgress$LoadCallbackStage")));
const jclass game_dir_class = env->FindClass("org/yuzu/yuzu_emu/model/GameDir");
s_game_dir_class = reinterpret_cast<jclass>(env->NewGlobalRef(game_dir_class));
s_game_dir_constructor = env->GetMethodID(game_dir_class, "<init>",
"(Ljava/lang/String;Z)V");
env->DeleteLocalRef(game_dir_class);
// Initialize methods
s_exit_emulation_activity =
env->GetStaticMethodID(s_native_library_class, "exitEmulationActivity", "(I)V");
s_disk_cache_load_progress =
env->GetStaticMethodID(s_disk_cache_progress_class, "loadProgress", "(III)V");
s_copy_to_storage = env->GetStaticMethodID(s_native_library_class, "copyFileToStorage",
"(Ljava/lang/String;Ljava/lang/String;)Z");
s_file_exists = env->GetStaticMethodID(s_native_library_class, "exists",
"(Ljava/lang/String;)Z");
s_file_extension = env->GetStaticMethodID(s_native_library_class, "getFileExtension",
"(Ljava/lang/String;)Ljava/lang/String;");
s_on_emulation_started =
env->GetStaticMethodID(s_native_library_class, "onEmulationStarted", "()V");
s_on_emulation_stopped =
env->GetStaticMethodID(s_native_library_class, "onEmulationStopped", "(I)V");
s_on_program_changed =
env->GetStaticMethodID(s_native_library_class, "onProgramChanged", "(I)V");
const jclass game_class = env->FindClass("org/yuzu/yuzu_emu/model/Game");
s_game_class = reinterpret_cast<jclass>(env->NewGlobalRef(game_class));
s_game_constructor = env->GetMethodID(game_class, "<init>",
"(Ljava/lang/String;Ljava/lang/String;Ljava/lang/"
"String;Ljava/lang/String;Ljava/lang/String;Z)V");
s_game_title_field = env->GetFieldID(game_class, "title", "Ljava/lang/String;");
s_game_path_field = env->GetFieldID(game_class, "path", "Ljava/lang/String;");
s_game_program_id_field = env->GetFieldID(game_class, "programId", "Ljava/lang/String;");
s_game_developer_field = env->GetFieldID(game_class, "developer", "Ljava/lang/String;");
s_game_version_field = env->GetFieldID(game_class, "version", "Ljava/lang/String;");
s_game_is_homebrew_field = env->GetFieldID(game_class, "isHomebrew", "Z");
env->DeleteLocalRef(game_class);
const jclass string_class = env->FindClass("java/lang/String");
s_string_class = reinterpret_cast<jclass>(env->NewGlobalRef(string_class));
env->DeleteLocalRef(string_class);
const jclass pair_class = env->FindClass("kotlin/Pair");
s_pair_class = reinterpret_cast<jclass>(env->NewGlobalRef(pair_class));
s_pair_constructor =
env->GetMethodID(pair_class, "<init>", "(Ljava/lang/Object;Ljava/lang/Object;)V");
s_pair_first_field = env->GetFieldID(pair_class, "first", "Ljava/lang/Object;");
s_pair_second_field = env->GetFieldID(pair_class, "second", "Ljava/lang/Object;");
env->DeleteLocalRef(pair_class);
const jclass overlay_control_data_class =
env->FindClass("org/yuzu/yuzu_emu/overlay/model/OverlayControlData");
s_overlay_control_data_class =
reinterpret_cast<jclass>(env->NewGlobalRef(overlay_control_data_class));
s_overlay_control_data_constructor =
env->GetMethodID(overlay_control_data_class, "<init>",
"(Ljava/lang/String;ZLkotlin/Pair;Lkotlin/Pair;Lkotlin/Pair;F)V");
s_overlay_control_data_id_field =
env->GetFieldID(overlay_control_data_class, "id", "Ljava/lang/String;");
s_overlay_control_data_enabled_field =
env->GetFieldID(overlay_control_data_class, "enabled", "Z");
s_overlay_control_data_landscape_position_field =
env->GetFieldID(overlay_control_data_class, "landscapePosition", "Lkotlin/Pair;");
s_overlay_control_data_portrait_position_field =
env->GetFieldID(overlay_control_data_class, "portraitPosition", "Lkotlin/Pair;");
s_overlay_control_data_foldable_position_field =
env->GetFieldID(overlay_control_data_class, "foldablePosition", "Lkotlin/Pair;");
s_overlay_control_data_individual_scale_field =
env->GetFieldID(overlay_control_data_class, "individualScale", "F");
env->DeleteLocalRef(overlay_control_data_class);
const jclass patch_class = env->FindClass("org/yuzu/yuzu_emu/model/Patch");
s_patch_class = reinterpret_cast<jclass>(env->NewGlobalRef(patch_class));
s_patch_constructor = env->GetMethodID(
patch_class, "<init>",
"(ZLjava/lang/String;Ljava/lang/String;ILjava/lang/String;Ljava/lang/String;JI)V");
s_patch_enabled_field = env->GetFieldID(patch_class, "enabled", "Z");
s_patch_name_field = env->GetFieldID(patch_class, "name", "Ljava/lang/String;");
s_patch_version_field = env->GetFieldID(patch_class, "version", "Ljava/lang/String;");
s_patch_type_field = env->GetFieldID(patch_class, "type", "I");
s_patch_program_id_field = env->GetFieldID(patch_class, "programId", "Ljava/lang/String;");
s_patch_title_id_field = env->GetFieldID(patch_class, "titleId", "Ljava/lang/String;");
env->DeleteLocalRef(patch_class);
const jclass double_class = env->FindClass("java/lang/Double");
s_double_class = reinterpret_cast<jclass>(env->NewGlobalRef(double_class));
s_double_constructor = env->GetMethodID(double_class, "<init>", "(D)V");
s_double_value_method = env->GetMethodID(double_class, "doubleValue", "()D");
env->DeleteLocalRef(double_class);
const jclass int_class = env->FindClass("java/lang/Integer");
s_integer_class = reinterpret_cast<jclass>(env->NewGlobalRef(int_class));
s_integer_constructor = env->GetMethodID(int_class, "<init>", "(I)V");
s_integer_value_method = env->GetMethodID(int_class, "intValue", "()I");
env->DeleteLocalRef(int_class);
const jclass boolean_class = env->FindClass("java/lang/Boolean");
s_boolean_class = reinterpret_cast<jclass>(env->NewGlobalRef(boolean_class));
s_boolean_constructor = env->GetMethodID(boolean_class, "<init>", "(Z)V");
s_boolean_value_method = env->GetMethodID(boolean_class, "booleanValue", "()Z");
env->DeleteLocalRef(boolean_class);
const jclass player_input_class =
env->FindClass("org/yuzu/yuzu_emu/features/input/model/PlayerInput");
s_player_input_class = reinterpret_cast<jclass>(env->NewGlobalRef(player_input_class));
s_player_input_constructor = env->GetMethodID(
player_input_class, "<init>",
"(Z[Ljava/lang/String;[Ljava/lang/String;[Ljava/lang/String;ZIJJJJLjava/lang/String;Z)V");
s_player_input_connected_field = env->GetFieldID(player_input_class, "connected", "Z");
s_player_input_buttons_field =
env->GetFieldID(player_input_class, "buttons", "[Ljava/lang/String;");
s_player_input_analogs_field =
env->GetFieldID(player_input_class, "analogs", "[Ljava/lang/String;");
s_player_input_motions_field =
env->GetFieldID(player_input_class, "motions", "[Ljava/lang/String;");
s_player_input_vibration_enabled_field =
env->GetFieldID(player_input_class, "vibrationEnabled", "Z");
s_player_input_vibration_strength_field =
env->GetFieldID(player_input_class, "vibrationStrength", "I");
s_player_input_body_color_left_field =
env->GetFieldID(player_input_class, "bodyColorLeft", "J");
s_player_input_body_color_right_field =
env->GetFieldID(player_input_class, "bodyColorRight", "J");
s_player_input_button_color_left_field =
env->GetFieldID(player_input_class, "buttonColorLeft", "J");
s_player_input_button_color_right_field =
env->GetFieldID(player_input_class, "buttonColorRight", "J");
s_player_input_profile_name_field =
env->GetFieldID(player_input_class, "profileName", "Ljava/lang/String;");
s_player_input_use_system_vibrator_field =
env->GetFieldID(player_input_class, "useSystemVibrator", "Z");
env->DeleteLocalRef(player_input_class);
const jclass yuzu_input_device_interface =
env->FindClass("org/yuzu/yuzu_emu/features/input/YuzuInputDevice");
s_yuzu_input_device_interface =
reinterpret_cast<jclass>(env->NewGlobalRef(yuzu_input_device_interface));
s_yuzu_input_device_get_name =
env->GetMethodID(yuzu_input_device_interface, "getName", "()Ljava/lang/String;");
s_yuzu_input_device_get_guid =
env->GetMethodID(yuzu_input_device_interface, "getGUID", "()Ljava/lang/String;");
s_yuzu_input_device_get_port = env->GetMethodID(yuzu_input_device_interface, "getPort",
"()I");
s_yuzu_input_device_get_supports_vibration =
env->GetMethodID(yuzu_input_device_interface, "getSupportsVibration", "()Z");
s_yuzu_input_device_vibrate = env->GetMethodID(yuzu_input_device_interface, "vibrate",
"(F)V");
s_yuzu_input_device_get_axes =
env->GetMethodID(yuzu_input_device_interface, "getAxes", "()[Ljava/lang/Integer;");
s_yuzu_input_device_has_keys =
env->GetMethodID(yuzu_input_device_interface, "hasKeys", "([I)[Z");
env->DeleteLocalRef(yuzu_input_device_interface);
s_add_netplay_message = env->GetStaticMethodID(s_native_library_class, "addNetPlayMessage",
"(ILjava/lang/String;)V");
s_clear_chat = env->GetStaticMethodID(s_native_library_class, "clearChat", "()V");
// Initialize Android Storage
Common::FS::Android::RegisterCallbacks(env, s_native_library_class);
// Initialize applets
Common::Android::SoftwareKeyboard::InitJNI(env);
Common::Android::WebBrowser::InitJNI(env);
}
} // namespace Common::Android
+3 -4
View File
@@ -11,7 +11,6 @@
namespace Common::Android {
void Initialize(JavaVM* vm, JNIEnv *env);
JNIEnv* GetEnvForThread();
/**
@@ -81,15 +80,15 @@ jfieldID GetPatchTitleIdField();
jclass GetDoubleClass();
jmethodID GetDoubleConstructor();
jmethodID GetDoubleValueMethod();
jfieldID GetDoubleValueField();
jclass GetIntegerClass();
jmethodID GetIntegerConstructor();
jmethodID GetIntegerValueMethod();
jfieldID GetIntegerValueField();
jclass GetBooleanClass();
jmethodID GetBooleanConstructor();
jmethodID GetBooleanValueMethod();
jfieldID GetBooleanValueField();
jclass GetPlayerInputClass();
jmethodID GetPlayerInputConstructor();
-1
View File
@@ -5,7 +5,6 @@
// SPDX-License-Identifier: GPL-2.0-or-later
#include <string>
#include <cstdlib>
#include <string_view>
#ifdef _WIN32
#include <llvm/Demangle/Demangle.h>
+3 -3
View File
@@ -92,14 +92,14 @@ struct EnumMetadata {
// AudioEngine must be specified discretely due to having existing but slightly different
// canonicalizations
// TODO (lat9nq): Remove explicit definition of AudioEngine/sink_id
enum class AudioEngine : u32 { Auto, Cubeb, Sdl3, Null, };
enum class AudioEngine : u32 { Auto, Cubeb, Sdl3, Null, Oboe, };
template<>
inline std::vector<std::pair<std::string_view, AudioEngine>> EnumMetadata<AudioEngine>::Canonicalizations() {
return {
{"auto", AudioEngine::Auto},
{"cubeb", AudioEngine::Cubeb},
{"sdl3", AudioEngine::Sdl3},
{"null", AudioEngine::Null},
{"null", AudioEngine::Null}, {"oboe", AudioEngine::Oboe},
};
}
/// @brief This is just a sufficiently large number that is more than the number of other enums declared here
@@ -113,7 +113,7 @@ inline AudioEngine EnumMetadata<AudioEngine>::GetFirst() {
}
template<>
inline AudioEngine EnumMetadata<AudioEngine>::GetLast() {
return AudioEngine::Null;
return AudioEngine::Oboe;
}
ENUM(AudioMode, Mono, Stereo, Surround);
+1 -1
View File
@@ -358,7 +358,7 @@ private:
ConnectionState(boost::asio::ip::tcp::socket&& client_socket_, async_pipe signal_pipe_, Kernel::KernelCore& kernel)
: client_socket{std::move(client_socket_)}
, signal_pipe{std::move(signal_pipe_)}
, signal_pipe{signal_pipe_}
, active_thread{kernel, nullptr}
{}
-3
View File
@@ -1021,9 +1021,6 @@ Result KProcess::Run(KernelCore& kernel, s32 priority, size_t stack_size) {
// Suspend for debug, if we should.
if (kernel.System().DebuggerEnabled()) {
LOG_INFO(Debug_GDBStub,
"GDB stub enabled; suspending guest process until a debugger continues execution on port {}",
Settings::values.gdbstub_port.GetValue());
main_thread->RequestSuspend(kernel, SuspendType::Debug);
}
+1 -1
View File
@@ -114,7 +114,7 @@ std::set<s32> Android::GetDeviceAxes(JNIEnv* env, jobject& j_device) const {
std::set<s32> axes;
for (int i = 0; i < env->GetArrayLength(j_axes); ++i) {
jobject axis = env->GetObjectArrayElement(j_axes, i);
axes.insert(env->CallIntMethod(axis, Common::Android::GetIntegerValueMethod()));
axes.insert(env->GetIntField(axis, Common::Android::GetIntegerValueField()));
}
return axes;
}
+1 -2
View File
@@ -234,13 +234,12 @@ add_library(shader_recompiler STATIC
ir_opt/texture_pass.cpp
ir_opt/vendor_workaround_pass.cpp
ir_opt/verification_pass.cpp
object_pool.h
profile.h
program_header.h
runtime_info.h
shader_info.h
varying_state.h
shader_pool.h
)
target_link_libraries(shader_recompiler PUBLIC common fmt::fmt sirit::sirit)
@@ -1,6 +1,3 @@
// SPDX-FileCopyrightText: Copyright 2026 Eden Emulator Project
// SPDX-License-Identifier: GPL-3.0-or-later
// SPDX-FileCopyrightText: Copyright 2021 yuzu Emulator Project
// SPDX-License-Identifier: GPL-2.0-or-later
@@ -32,7 +29,7 @@ struct FuncTraits<ReturnType_ (*)(Args...)> {
};
template <auto func, typename... Args>
[[maybe_unused]] void SetDefinition(EmitContext& ctx, IR::Inst* inst, Args... args) {
void SetDefinition(EmitContext& ctx, IR::Inst* inst, Args... args) {
inst->SetDefinition<Id>(func(ctx, std::forward<Args>(args)...));
}
@@ -492,9 +492,6 @@ void SetupCapabilities(const Profile& profile, const Info& info, EmitContext& ct
if (ctx.uses_nonuniform_storage_texel_buffer) {
ctx.AddCapability(spv::Capability::StorageTexelBufferArrayNonUniformIndexing);
}
if (ctx.uses_nonuniform_storage_buffer) {
ctx.AddCapability(spv::Capability::StorageBufferArrayNonUniformIndexing);
}
}
}
@@ -4,6 +4,8 @@
// SPDX-FileCopyrightText: Copyright 2021 yuzu Emulator Project
// SPDX-License-Identifier: GPL-2.0-or-later
#include <bit>
#include "shader_recompiler/backend/spirv/emit_spirv.h"
#include "shader_recompiler/backend/spirv/emit_spirv_instructions.h"
#include "shader_recompiler/backend/spirv/spirv_emit_context.h"
@@ -21,13 +23,29 @@ Id SharedPointer(EmitContext& ctx, Id offset, u32 index_offset = 0) {
: ctx.OpAccessChain(ctx.shared_u32, ctx.shared_memory_u32, index);
}
Id StorageIndex(EmitContext& ctx, const IR::Value& offset, size_t element_size) {
if (offset.IsImmediate()) {
const u32 imm_offset{static_cast<u32>(offset.U32() / element_size)};
return ctx.Const(imm_offset);
}
const u32 shift{static_cast<u32>(std::countr_zero(element_size))};
const Id index{ctx.Def(offset)};
if (shift == 0) {
return index;
}
const Id shift_id{ctx.Const(shift)};
return ctx.OpShiftRightLogical(ctx.U32[1], index, shift_id);
}
Id StoragePointer(EmitContext& ctx, const StorageTypeDefinition& type_def,
Id StorageDefinitions::*member_ptr, const IR::Value& binding,
const IR::Value& offset, size_t element_size) {
if (!binding.IsImmediate()) {
throw NotImplementedException("Dynamic storage buffer indexing");
}
return ctx.StoragePointer(binding.U32(), ctx.Def(offset), type_def, static_cast<u32>(element_size), member_ptr);
const Id ssbo{ctx.ssbos[binding.U32()].*member_ptr};
const Id index{StorageIndex(ctx, offset, element_size)};
return ctx.OpAccessChain(type_def.element, ssbo, ctx.u32_zero_value, index);
}
std::pair<Id, Id> AtomicArgs(EmitContext& ctx) {
@@ -198,14 +216,16 @@ Id EmitStorageAtomicUMax32(EmitContext& ctx, const IR::Value& binding, const IR:
Id EmitStorageAtomicInc32(EmitContext& ctx, const IR::Value& binding, const IR::Value& offset,
Id value) {
const Id pointer{StoragePointer(ctx, ctx.storage_types.U32, &StorageDefinitions::U32, binding, offset, sizeof(u32))};
return ctx.OpFunctionCall(ctx.U32[1], ctx.increment_cas_ssbo, pointer, value);
const Id ssbo{ctx.ssbos[binding.U32()].U32};
const Id base_index{StorageIndex(ctx, offset, sizeof(u32))};
return ctx.OpFunctionCall(ctx.U32[1], ctx.increment_cas_ssbo, base_index, value, ssbo);
}
Id EmitStorageAtomicDec32(EmitContext& ctx, const IR::Value& binding, const IR::Value& offset,
Id value) {
const Id pointer{StoragePointer(ctx, ctx.storage_types.U32, &StorageDefinitions::U32, binding, offset, sizeof(u32))};
return ctx.OpFunctionCall(ctx.U32[1], ctx.decrement_cas_ssbo, pointer, value);
const Id ssbo{ctx.ssbos[binding.U32()].U32};
const Id base_index{StorageIndex(ctx, offset, sizeof(u32))};
return ctx.OpFunctionCall(ctx.U32[1], ctx.decrement_cas_ssbo, base_index, value, ssbo);
}
Id EmitStorageAtomicAnd32(EmitContext& ctx, const IR::Value& binding, const IR::Value& offset,
@@ -344,49 +364,56 @@ Id EmitStorageAtomicExchange32x2(EmitContext& ctx, const IR::Value& binding,
Id EmitStorageAtomicAddF32(EmitContext& ctx, const IR::Value& binding, const IR::Value& offset,
Id value) {
const Id pointer{StoragePointer(ctx, ctx.storage_types.U32, &StorageDefinitions::U32, binding, offset, sizeof(u32))};
return ctx.OpFunctionCall(ctx.F32[1], ctx.f32_add_cas, pointer, value);
const Id ssbo{ctx.ssbos[binding.U32()].U32};
const Id base_index{StorageIndex(ctx, offset, sizeof(u32))};
return ctx.OpFunctionCall(ctx.F32[1], ctx.f32_add_cas, base_index, value, ssbo);
}
Id EmitStorageAtomicAddF16x2(EmitContext& ctx, const IR::Value& binding, const IR::Value& offset,
Id value) {
const Id pointer{StoragePointer(ctx, ctx.storage_types.U32, &StorageDefinitions::U32, binding, offset, sizeof(u32))};
const Id result{ctx.OpFunctionCall(ctx.F16[2], ctx.f16x2_add_cas, pointer, value)};
const Id ssbo{ctx.ssbos[binding.U32()].U32};
const Id base_index{StorageIndex(ctx, offset, sizeof(u32))};
const Id result{ctx.OpFunctionCall(ctx.F16[2], ctx.f16x2_add_cas, base_index, value, ssbo)};
return ctx.OpBitcast(ctx.U32[1], result);
}
Id EmitStorageAtomicAddF32x2(EmitContext& ctx, const IR::Value& binding, const IR::Value& offset,
Id value) {
const Id pointer{StoragePointer(ctx, ctx.storage_types.U32, &StorageDefinitions::U32, binding, offset, sizeof(u32))};
const Id result{ctx.OpFunctionCall(ctx.F32[2], ctx.f32x2_add_cas, pointer, value)};
const Id ssbo{ctx.ssbos[binding.U32()].U32};
const Id base_index{StorageIndex(ctx, offset, sizeof(u32))};
const Id result{ctx.OpFunctionCall(ctx.F32[2], ctx.f32x2_add_cas, base_index, value, ssbo)};
return ctx.OpPackHalf2x16(ctx.U32[1], result);
}
Id EmitStorageAtomicMinF16x2(EmitContext& ctx, const IR::Value& binding, const IR::Value& offset,
Id value) {
const Id pointer{StoragePointer(ctx, ctx.storage_types.U32, &StorageDefinitions::U32, binding, offset, sizeof(u32))};
const Id result{ctx.OpFunctionCall(ctx.F16[2], ctx.f16x2_min_cas, pointer, value)};
const Id ssbo{ctx.ssbos[binding.U32()].U32};
const Id base_index{StorageIndex(ctx, offset, sizeof(u32))};
const Id result{ctx.OpFunctionCall(ctx.F16[2], ctx.f16x2_min_cas, base_index, value, ssbo)};
return ctx.OpBitcast(ctx.U32[1], result);
}
Id EmitStorageAtomicMinF32x2(EmitContext& ctx, const IR::Value& binding, const IR::Value& offset,
Id value) {
const Id pointer{StoragePointer(ctx, ctx.storage_types.U32, &StorageDefinitions::U32, binding, offset, sizeof(u32))};
const Id result{ctx.OpFunctionCall(ctx.F32[2], ctx.f32x2_min_cas, pointer, value)};
const Id ssbo{ctx.ssbos[binding.U32()].U32};
const Id base_index{StorageIndex(ctx, offset, sizeof(u32))};
const Id result{ctx.OpFunctionCall(ctx.F32[2], ctx.f32x2_min_cas, base_index, value, ssbo)};
return ctx.OpPackHalf2x16(ctx.U32[1], result);
}
Id EmitStorageAtomicMaxF16x2(EmitContext& ctx, const IR::Value& binding, const IR::Value& offset,
Id value) {
const Id pointer{StoragePointer(ctx, ctx.storage_types.U32, &StorageDefinitions::U32, binding, offset, sizeof(u32))};
const Id result{ctx.OpFunctionCall(ctx.F16[2], ctx.f16x2_max_cas, pointer, value)};
const Id ssbo{ctx.ssbos[binding.U32()].U32};
const Id base_index{StorageIndex(ctx, offset, sizeof(u32))};
const Id result{ctx.OpFunctionCall(ctx.F16[2], ctx.f16x2_max_cas, base_index, value, ssbo)};
return ctx.OpBitcast(ctx.U32[1], result);
}
Id EmitStorageAtomicMaxF32x2(EmitContext& ctx, const IR::Value& binding, const IR::Value& offset,
Id value) {
const Id pointer{StoragePointer(ctx, ctx.storage_types.U32, &StorageDefinitions::U32, binding, offset, sizeof(u32))};
const Id result{ctx.OpFunctionCall(ctx.F32[2], ctx.f32x2_max_cas, pointer, value)};
const Id ssbo{ctx.ssbos[binding.U32()].U32};
const Id base_index{StorageIndex(ctx, offset, sizeof(u32))};
const Id result{ctx.OpFunctionCall(ctx.F32[2], ctx.f32x2_max_cas, base_index, value, ssbo)};
return ctx.OpPackHalf2x16(ctx.U32[1], result);
}
@@ -4,33 +4,46 @@
// SPDX-FileCopyrightText: Copyright 2021 yuzu Emulator Project
// SPDX-License-Identifier: GPL-2.0-or-later
#include <bit>
#include "shader_recompiler/backend/spirv/emit_spirv.h"
#include "shader_recompiler/backend/spirv/emit_spirv_instructions.h"
#include "shader_recompiler/backend/spirv/spirv_emit_context.h"
namespace Shader::Backend::SPIRV {
namespace {
Id StorageByteOffset(EmitContext& ctx, const IR::Value& offset, size_t element_size, u32 index_offset) {
Id byte_offset{ctx.Def(offset)};
if (index_offset != 0) {
byte_offset = ctx.OpIAdd(ctx.U32[1], byte_offset, ctx.Const(static_cast<u32>(index_offset * element_size)));
Id StorageIndex(EmitContext& ctx, const IR::Value& offset, size_t element_size,
u32 index_offset = 0) {
if (offset.IsImmediate()) {
const u32 imm_offset{static_cast<u32>(offset.U32() / element_size) + index_offset};
return ctx.Const(imm_offset);
}
return byte_offset;
const u32 shift{static_cast<u32>(std::countr_zero(element_size))};
Id index{ctx.Def(offset)};
if (shift != 0) {
const Id shift_id{ctx.Const(shift)};
index = ctx.OpShiftRightLogical(ctx.U32[1], index, shift_id);
}
if (index_offset != 0) {
index = ctx.OpIAdd(ctx.U32[1], index, ctx.Const(index_offset));
}
return index;
}
Id StoragePointer(EmitContext& ctx, const IR::Value& binding, const IR::Value& offset,
const StorageTypeDefinition& type_def, size_t element_size,
Id StorageDefinitions::* member_ptr, u32 index_offset = 0) {
Id StorageDefinitions::*member_ptr, u32 index_offset = 0) {
if (!binding.IsImmediate()) {
throw NotImplementedException("Dynamic storage buffer indexing");
}
const Id byte_offset{StorageByteOffset(ctx, offset, element_size, index_offset)};
return ctx.StoragePointer(binding.U32(), byte_offset, type_def, static_cast<u32>(element_size), member_ptr);
const Id ssbo{ctx.ssbos[binding.U32()].*member_ptr};
const Id index{StorageIndex(ctx, offset, element_size, index_offset)};
return ctx.OpAccessChain(type_def.element, ssbo, ctx.u32_zero_value, index);
}
Id LoadStorage(EmitContext& ctx, const IR::Value& binding, const IR::Value& offset, Id result_type,
const StorageTypeDefinition& type_def, size_t element_size,
Id StorageDefinitions::* member_ptr, u32 index_offset = 0) {
Id StorageDefinitions::*member_ptr, u32 index_offset = 0) {
const Id pointer{
StoragePointer(ctx, binding, offset, type_def, element_size, member_ptr, index_offset)};
return ctx.OpLoad(result_type, pointer);
@@ -44,7 +57,7 @@ Id LoadStorage32(EmitContext& ctx, const IR::Value& binding, const IR::Value& of
void WriteStorage(EmitContext& ctx, const IR::Value& binding, const IR::Value& offset, Id value,
const StorageTypeDefinition& type_def, size_t element_size,
Id StorageDefinitions::* member_ptr, u32 index_offset = 0) {
Id StorageDefinitions::*member_ptr, u32 index_offset = 0) {
const Id pointer{
StoragePointer(ctx, binding, offset, type_def, element_size, member_ptr, index_offset)};
ctx.OpStore(pointer, value);
@@ -299,7 +299,7 @@ void DefineConstBuffers(EmitContext& ctx, const Info& info, Id UniformDefinition
}
void DefineSsbos(EmitContext& ctx, StorageTypeDefinition& type_def,
Id StorageDefinitions::* member_type, const Info& info, u32 binding, Id type,
Id StorageDefinitions::*member_type, const Info& info, u32 binding, Id type,
u32 stride) {
const Id array_type{ctx.TypeRuntimeArray(type)};
ctx.Decorate(array_type, spv::Decoration::ArrayStride, stride);
@@ -309,27 +309,23 @@ void DefineSsbos(EmitContext& ctx, StorageTypeDefinition& type_def,
ctx.MemberDecorate(struct_type, 0, spv::Decoration::Offset, 0U);
const Id struct_pointer{ctx.TypePointer(spv::StorageClass::StorageBuffer, struct_type)};
type_def.array = struct_pointer;
type_def.element = ctx.TypePointer(spv::StorageClass::StorageBuffer, type);
u32 index{};
for (const StorageBufferDescriptor& desc : info.storage_buffers_descriptors) {
const Id variable_type{[&] {
if (desc.count == 1) {
return struct_pointer;
}
const Id descriptor_array{ctx.TypeArray(struct_type, ctx.Const(desc.count))};
return ctx.TypePointer(spv::StorageClass::StorageBuffer, descriptor_array);
}()};
const Id id{ctx.AddGlobalVariable(variable_type, spv::StorageClass::StorageBuffer)};
const Id id{ctx.AddGlobalVariable(struct_pointer, spv::StorageClass::StorageBuffer)};
ctx.Decorate(id, spv::Decoration::Binding, binding);
ctx.Decorate(id, spv::Decoration::DescriptorSet, 0U);
ctx.Name(id, fmt::format("ssbo{}", index));
if (ctx.profile.supported_spirv >= 0x00010400) {
ctx.interfaces.push_back(id);
}
ctx.ssbos[index].*member_type = id;
++index;
++binding;
for (size_t i = 0; i < desc.count; ++i) {
ctx.ssbos[index + i].*member_type = id;
}
index += desc.count;
binding += desc.count;
}
}
@@ -372,7 +368,8 @@ Id CasFunction(EmitContext& ctx, Operation operation, Id value_type) {
return func;
}
Id CasLoop(EmitContext& ctx, Operation operation, Id element_pointer, Id value_type, Id memory_type, spv::Scope scope) {
Id CasLoop(EmitContext& ctx, Operation operation, Id array_pointer, Id element_pointer,
Id value_type, Id memory_type, spv::Scope scope) {
const bool is_shared{scope == spv::Scope::Workgroup};
const bool is_struct{!is_shared || ctx.uses_explicit_workgroup_layout};
const Id cas_func{CasFunction(ctx, operation, value_type)};
@@ -382,12 +379,14 @@ Id CasLoop(EmitContext& ctx, Operation operation, Id element_pointer, Id value_t
const Id loop_header{ctx.OpLabel()};
const Id continue_block{ctx.OpLabel()};
const Id merge_block{ctx.OpLabel()};
const Id func_type{is_shared ? ctx.TypeFunction(value_type, ctx.U32[1], value_type)
: ctx.TypeFunction(value_type, element_pointer, value_type)};
const Id func_type{is_shared
? ctx.TypeFunction(value_type, ctx.U32[1], value_type)
: ctx.TypeFunction(value_type, ctx.U32[1], value_type, array_pointer)};
const Id func{ctx.OpFunction(value_type, spv::FunctionControlMask::MaskNone, func_type)};
const Id address{ctx.OpFunctionParameter(is_shared ? ctx.U32[1] : element_pointer)};
const Id index{ctx.OpFunctionParameter(ctx.U32[1])};
const Id op_b{ctx.OpFunctionParameter(value_type)};
const Id base{is_shared ? ctx.shared_memory_u32 : ctx.OpFunctionParameter(array_pointer)};
ctx.AddLabel();
ctx.OpBranch(loop_header);
ctx.AddLabel(loop_header);
@@ -396,13 +395,8 @@ Id CasLoop(EmitContext& ctx, Operation operation, Id element_pointer, Id value_t
ctx.OpBranch(continue_block);
ctx.AddLabel(continue_block);
const Id word_pointer{[&] {
if (!is_shared) {
return address;
}
return is_struct ? ctx.OpAccessChain(element_pointer, ctx.shared_memory_u32, zero, address)
: ctx.OpAccessChain(element_pointer, ctx.shared_memory_u32, address);
}()};
const Id word_pointer{is_struct ? ctx.OpAccessChain(element_pointer, base, zero, index)
: ctx.OpAccessChain(element_pointer, base, index)};
if (value_type.value == ctx.F32[2].value) {
const Id u32_value{ctx.OpLoad(ctx.U32[1], word_pointer)};
const Id value{ctx.OpUnpackHalf2x16(ctx.F32[2], u32_value)};
@@ -486,7 +480,6 @@ EmitContext::EmitContext(const Profile& profile_, const RuntimeInfo& runtime_inf
DefineSharedMemoryFunctions(program);
DefineConstantBuffers(program.info, uniform_binding);
DefineConstantBufferIndirectFunctions(program.info);
DefineStorageBufferMappings(program.info, storage_binding);
DefineStorageBuffers(program.info, storage_binding);
DefineTextureBuffers(program.info, texture_binding);
DefineImageBuffers(program.info, image_binding);
@@ -539,36 +532,6 @@ Id EmitContext::BitOffset16(const IR::Value& offset) {
return OpBitwiseAnd(U32[1], OpShiftLeftLogical(U32[1], Def(offset), Const(3u)), Const(16u));
}
Id EmitContext::StoragePointer(u32 binding, Id byte_offset, const StorageTypeDefinition& type_def,
u32 element_size, Id StorageDefinitions::* member_ptr) {
const Id ssbo{ssbos[binding].*member_ptr};
const u32 segment_count{storage_buffer_mapping_counts[binding]};
if (segment_count <= 1) {
const Id index{
element_size == 1
? byte_offset
: OpShiftRightLogical(U32[1], byte_offset, Const(static_cast<u32>(std::countr_zero(element_size))))};
return OpAccessChain(type_def.element, ssbo, u32_zero_value, index);
}
const Id mapped{OpFunctionCall(U32[2], storage_buffer_map_func,
Const(storage_buffer_mapping_bases[binding]),
Const(segment_count), byte_offset)};
const Id segment{OpCompositeExtract(U32[1], mapped, 0U)};
const Id local_offset{OpCompositeExtract(U32[1], mapped, 1U)};
const Id index{
element_size == 1
? local_offset
: OpShiftRightLogical(U32[1], local_offset, Const(static_cast<u32>(std::countr_zero(element_size))))};
Decorate(segment, spv::Decoration::NonUniform);
non_uniform_ids.insert(segment.value);
const Id pointer{OpAccessChain(type_def.element, ssbo, segment, u32_zero_value, index)};
Decorate(pointer, spv::Decoration::NonUniform);
non_uniform_ids.insert(pointer.value);
uses_nonuniform_storage_buffer = true;
return pointer;
}
void EmitContext::DefineCommonTypes(const Info& info) {
void_id = TypeVoid();
@@ -733,10 +696,12 @@ void EmitContext::DefineSharedMemory(const IR::Program& program) {
void EmitContext::DefineSharedMemoryFunctions(const IR::Program& program) {
if (program.info.uses_shared_increment) {
increment_cas_shared = CasLoop(*this, Operation::Increment, shared_u32, U32[1], U32[1], spv::Scope::Workgroup);
increment_cas_shared = CasLoop(*this, Operation::Increment, shared_memory_u32_type,
shared_u32, U32[1], U32[1], spv::Scope::Workgroup);
}
if (program.info.uses_shared_decrement) {
decrement_cas_shared = CasLoop(*this, Operation::Decrement, shared_u32, U32[1], U32[1], spv::Scope::Workgroup);
decrement_cas_shared = CasLoop(*this, Operation::Decrement, shared_memory_u32_type,
shared_u32, U32[1], U32[1], spv::Scope::Workgroup);
}
}
@@ -980,7 +945,8 @@ void EmitContext::DefineGlobalMemoryFunctions(const Info& info) {
}
using DefPtr = Id StorageDefinitions::*;
const Id zero{u32_zero_value};
const auto define_body{[&](DefPtr ssbo_member, Id addr, const StorageTypeDefinition& type_def, u32 shift, auto&& callback) {
const auto define_body{[&](DefPtr ssbo_member, Id addr, Id element_pointer, u32 shift,
auto&& callback) {
AddLabel();
const size_t num_buffers{info.storage_buffers_descriptors.size()};
for (size_t index = 0; index < num_buffers; ++index) {
@@ -1007,28 +973,30 @@ void EmitContext::DefineGlobalMemoryFunctions(const Info& info) {
OpSelectionMerge(else_label, spv::SelectionControlMask::MaskNone);
OpBranchConditional(cond, then_label, else_label);
AddLabel(then_label);
const Id ssbo_id{ssbos[index].*ssbo_member};
const Id ssbo_offset{OpUConvert(U32[1], OpISub(U64, addr, ssbo_addr))};
const Id ssbo_pointer{StoragePointer(static_cast<u32>(index), ssbo_offset, type_def, 1U << shift, ssbo_member)};
const Id ssbo_index{OpShiftRightLogical(U32[1], ssbo_offset, Const(shift))};
const Id ssbo_pointer{OpAccessChain(element_pointer, ssbo_id, zero, ssbo_index)};
callback(ssbo_pointer);
AddLabel(else_label);
}
}};
const auto define_load{
[&](DefPtr ssbo_member, const StorageTypeDefinition& type_def, Id type, u32 shift) {
const Id function_type{TypeFunction(type, U64)};
const Id func_id{OpFunction(type, spv::FunctionControlMask::MaskNone, function_type)};
const Id addr{OpFunctionParameter(U64)};
define_body(ssbo_member, addr, type_def, shift, [&](Id ssbo_pointer) { OpReturnValue(OpLoad(type, ssbo_pointer)); });
OpReturnValue(ConstantNull(type));
OpFunctionEnd();
return func_id;
}};
const auto define_write{[&](DefPtr ssbo_member, const StorageTypeDefinition& type_def, Id type, u32 shift) {
const auto define_load{[&](DefPtr ssbo_member, Id element_pointer, Id type, u32 shift) {
const Id function_type{TypeFunction(type, U64)};
const Id func_id{OpFunction(type, spv::FunctionControlMask::MaskNone, function_type)};
const Id addr{OpFunctionParameter(U64)};
define_body(ssbo_member, addr, element_pointer, shift,
[&](Id ssbo_pointer) { OpReturnValue(OpLoad(type, ssbo_pointer)); });
OpReturnValue(ConstantNull(type));
OpFunctionEnd();
return func_id;
}};
const auto define_write{[&](DefPtr ssbo_member, Id element_pointer, Id type, u32 shift) {
const Id function_type{TypeFunction(void_id, U64, type)};
const Id func_id{OpFunction(void_id, spv::FunctionControlMask::MaskNone, function_type)};
const Id addr{OpFunctionParameter(U64)};
const Id data{OpFunctionParameter(type)};
define_body(ssbo_member, addr, type_def, shift, [&](Id ssbo_pointer) {
define_body(ssbo_member, addr, element_pointer, shift, [&](Id ssbo_pointer) {
OpStore(ssbo_pointer, data);
OpReturn();
});
@@ -1038,9 +1006,10 @@ void EmitContext::DefineGlobalMemoryFunctions(const Info& info) {
}};
const auto define{
[&](DefPtr ssbo_member, const StorageTypeDefinition& type_def, Id type, size_t size) {
const Id element_type{type_def.element};
const u32 shift{static_cast<u32>(std::countr_zero(size))};
const Id load_func{define_load(ssbo_member, type_def, type, shift)};
const Id write_func{define_write(ssbo_member, type_def, type, shift)};
const Id load_func{define_load(ssbo_member, element_type, type, shift)};
const Id write_func{define_write(ssbo_member, element_type, type, shift)};
return std::make_pair(load_func, write_func);
}};
std::tie(load_global_func_u32, write_global_func_u32) =
@@ -1259,79 +1228,6 @@ void EmitContext::DefineConstantBufferIndirectFunctions(const Info& info) {
}
}
void EmitContext::DefineStorageBufferMappings(const Info& info, u32& binding) {
if (!UsesStorageBufferMappings(info)) {
return;
}
ASSERT(profile.support_storage_buffer_array_nonuniform_indexing);
AddExtension("SPV_KHR_storage_buffer_storage_class");
const u32 num_entries{NumDescriptors(info.storage_buffers_descriptors)};
const Id array_type{TypeArray(U32[1], Const(num_entries))};
Decorate(array_type, spv::Decoration::ArrayStride, sizeof(u32));
const Id struct_type{TypeStruct(array_type)};
Decorate(struct_type, spv::Decoration::Block);
MemberName(struct_type, 0, "segment_sizes");
MemberDecorate(struct_type, 0, spv::Decoration::Offset, 0U);
const Id pointer_type{TypePointer(spv::StorageClass::StorageBuffer, struct_type)};
storage_buffer_mapping_u32 = TypePointer(spv::StorageClass::StorageBuffer, U32[1]);
storage_buffer_mapping = AddGlobalVariable(pointer_type, spv::StorageClass::StorageBuffer);
Decorate(storage_buffer_mapping, spv::Decoration::Binding, binding++);
Decorate(storage_buffer_mapping, spv::Decoration::DescriptorSet, 0U);
Name(storage_buffer_mapping, "storage_buffer_mapping");
//Starting with version 1.4... (https://registry.khronos.org/SPIR-V/specs/unified1/SPIRV.html)
if (profile.supported_spirv >= 0x00010400) {
interfaces.push_back(storage_buffer_mapping);
}
u32 mapping_base{};
for (u32 index = 0; index < info.storage_buffers_descriptors.size(); ++index) {
const StorageBufferDescriptor& desc = info.storage_buffers_descriptors[index];
storage_buffer_mapping_bases[index] = mapping_base;
storage_buffer_mapping_counts[index] = desc.count;
mapping_base += desc.count;
}
const Id function_type{TypeFunction(U32[2], U32[1], U32[1], U32[1])};
storage_buffer_map_func = OpFunction(U32[2], spv::FunctionControlMask::MaskNone, function_type);
const Id base{OpFunctionParameter(U32[1])};
const Id count{OpFunctionParameter(U32[1])};
const Id byte_offset{OpFunctionParameter(U32[1])};
const Id index_pointer_type{TypePointer(spv::StorageClass::Function, U32[1])};
const Id loop_header{OpLabel()};
const Id continue_block{OpLabel()};
const Id merge_block{OpLabel()};
AddLabel();
const Id segment_var{AddLocalVariable(index_pointer_type, spv::StorageClass::Function)};
const Id offset_var{AddLocalVariable(index_pointer_type, spv::StorageClass::Function)};
OpStore(segment_var, u32_zero_value);
OpStore(offset_var, byte_offset);
OpBranch(loop_header);
AddLabel(loop_header);
const Id segment{OpLoad(U32[1], segment_var)};
const Id local_offset{OpLoad(U32[1], offset_var)};
const Id mapping_index{OpIAdd(U32[1], base, segment)};
const Id size_pointer{OpAccessChain(storage_buffer_mapping_u32, storage_buffer_mapping, u32_zero_value, mapping_index)};
const Id segment_size{OpLoad(U32[1], size_pointer)};
const Id fits{OpULessThan(U1, local_offset, segment_size)};
const Id last_segment{OpISub(U32[1], count, Const(1U))};
const Id is_last{OpIEqual(U1, segment, last_segment)};
const Id found{OpLogicalOr(U1, fits, is_last)};
OpLoopMerge(merge_block, continue_block, spv::LoopControlMask::MaskNone);
OpBranchConditional(found, merge_block, continue_block);
AddLabel(continue_block);
OpStore(offset_var, OpISub(U32[1], local_offset, segment_size));
OpStore(segment_var, OpIAdd(U32[1], segment, Const(1U)));
OpBranch(loop_header);
AddLabel(merge_block);
OpReturnValue(OpCompositeConstruct(U32[2], segment, local_offset));
OpFunctionEnd();
Name(storage_buffer_map_func, "map_storage_buffer");
}
void EmitContext::DefineStorageBuffers(const Info& info, u32& binding) {
if (info.storage_buffers_descriptors.empty()) {
return;
@@ -1375,7 +1271,9 @@ void EmitContext::DefineStorageBuffers(const Info& info, u32& binding) {
DefineSsbos(*this, storage_types.U32x4, &StorageDefinitions::U32x4, info, binding, U32[4],
sizeof(u32[4]));
}
binding += static_cast<u32>(info.storage_buffers_descriptors.size());
for (const StorageBufferDescriptor& desc : info.storage_buffers_descriptors) {
binding += desc.count;
}
const bool needs_function{
info.uses_global_increment || info.uses_global_decrement || info.uses_atomic_f32_add ||
info.uses_atomic_f16x2_add || info.uses_atomic_f16x2_min || info.uses_atomic_f16x2_max ||
@@ -1384,31 +1282,40 @@ void EmitContext::DefineStorageBuffers(const Info& info, u32& binding) {
AddCapability(spv::Capability::VariablePointersStorageBuffer);
}
if (info.uses_global_increment) {
increment_cas_ssbo = CasLoop(*this, Operation::Increment, storage_types.U32.element, U32[1], U32[1], spv::Scope::Device);
increment_cas_ssbo = CasLoop(*this, Operation::Increment, storage_types.U32.array,
storage_types.U32.element, U32[1], U32[1], spv::Scope::Device);
}
if (info.uses_global_decrement) {
decrement_cas_ssbo = CasLoop(*this, Operation::Decrement, storage_types.U32.element, U32[1], U32[1], spv::Scope::Device);
decrement_cas_ssbo = CasLoop(*this, Operation::Decrement, storage_types.U32.array,
storage_types.U32.element, U32[1], U32[1], spv::Scope::Device);
}
if (info.uses_atomic_f32_add) {
f32_add_cas = CasLoop(*this, Operation::FPAdd, storage_types.U32.element, F32[1], U32[1], spv::Scope::Device);
f32_add_cas = CasLoop(*this, Operation::FPAdd, storage_types.U32.array,
storage_types.U32.element, F32[1], U32[1], spv::Scope::Device);
}
if (info.uses_atomic_f16x2_add) {
f16x2_add_cas = CasLoop(*this, Operation::FPAdd, storage_types.U32.element, F16[2], F16[2], spv::Scope::Device);
f16x2_add_cas = CasLoop(*this, Operation::FPAdd, storage_types.U32.array,
storage_types.U32.element, F16[2], F16[2], spv::Scope::Device);
}
if (info.uses_atomic_f16x2_min) {
f16x2_min_cas = CasLoop(*this, Operation::FPMin, storage_types.U32.element, F16[2], F16[2], spv::Scope::Device);
f16x2_min_cas = CasLoop(*this, Operation::FPMin, storage_types.U32.array,
storage_types.U32.element, F16[2], F16[2], spv::Scope::Device);
}
if (info.uses_atomic_f16x2_max) {
f16x2_max_cas = CasLoop(*this, Operation::FPMax, storage_types.U32.element, F16[2], F16[2], spv::Scope::Device);
f16x2_max_cas = CasLoop(*this, Operation::FPMax, storage_types.U32.array,
storage_types.U32.element, F16[2], F16[2], spv::Scope::Device);
}
if (info.uses_atomic_f32x2_add) {
f32x2_add_cas = CasLoop(*this, Operation::FPAdd, storage_types.U32.element, F32[2], F32[2], spv::Scope::Device);
f32x2_add_cas = CasLoop(*this, Operation::FPAdd, storage_types.U32.array,
storage_types.U32.element, F32[2], F32[2], spv::Scope::Device);
}
if (info.uses_atomic_f32x2_min) {
f32x2_min_cas = CasLoop(*this, Operation::FPMin, storage_types.U32.element, F32[2], F32[2], spv::Scope::Device);
f32x2_min_cas = CasLoop(*this, Operation::FPMin, storage_types.U32.array,
storage_types.U32.element, F32[2], F32[2], spv::Scope::Device);
}
if (info.uses_atomic_f32x2_max) {
f32x2_max_cas = CasLoop(*this, Operation::FPMax, storage_types.U32.element, F32[2], F32[2], spv::Scope::Device);
f32x2_max_cas = CasLoop(*this, Operation::FPMax, storage_types.U32.array,
storage_types.U32.element, F32[2], F32[2], spv::Scope::Device);
}
}
@@ -114,6 +114,7 @@ struct UniformDefinitions {
};
struct StorageTypeDefinition {
Id array{};
Id element{};
};
@@ -172,8 +173,6 @@ public:
[[nodiscard]] Id BitOffset8(const IR::Value& offset);
[[nodiscard]] Id BitOffset16(const IR::Value& offset);
[[nodiscard]] Id StoragePointer(u32 binding, Id byte_offset, const StorageTypeDefinition& type_def, u32 element_size, Id StorageDefinitions::*member_ptr);
Id Const(u32 value) {
return Constant(U32[1], value);
}
@@ -258,11 +257,6 @@ public:
std::array<UniformDefinitions, Info::MAX_CBUFS> cbufs{};
std::array<StorageDefinitions, Info::MAX_SSBOS> ssbos{};
std::array<u32, Info::MAX_SSBOS> storage_buffer_mapping_bases{};
std::array<u32, Info::MAX_SSBOS> storage_buffer_mapping_counts{};
Id storage_buffer_mapping{};
Id storage_buffer_mapping_u32{};
Id storage_buffer_map_func{};
std::vector<TextureBufferDefinition> texture_buffers;
std::vector<ImageBufferDefinition> image_buffers;
std::vector<TextureDefinition> textures;
@@ -382,7 +376,6 @@ public:
bool uses_nonuniform_storage_image{};
bool uses_nonuniform_uniform_texel_buffer{};
bool uses_nonuniform_storage_texel_buffer{};
bool uses_nonuniform_storage_buffer{};
private:
void DefineCommonTypes(const Info& info);
@@ -393,7 +386,6 @@ private:
void DefineSharedMemoryFunctions(const IR::Program& program);
void DefineConstantBuffers(const Info& info, u32& binding);
void DefineConstantBufferIndirectFunctions(const Info& info);
void DefineStorageBufferMappings(const Info& info, u32& binding);
void DefineStorageBuffers(const Info& info, u32& binding);
void DefineTextureBuffers(const Info& info, u32& binding);
void DefineImageBuffers(const Info& info, u32& binding);
@@ -14,7 +14,7 @@
namespace Shader::IR {
Block::Block(ObjectPool<Inst>& inst_pool_) : inst_pool{&inst_pool_} {}
Block::Block(boost::container::stable_vector<Inst>& inst_pool_) : inst_pool{&inst_pool_} {}
Block::~Block() = default;
@@ -23,13 +23,12 @@ void Block::AppendNewInst(Opcode op, std::initializer_list<Value> args) {
}
Block::iterator Block::PrependNewInst(iterator insertion_point, const Inst& base_inst) {
Inst* const inst{inst_pool->Create(base_inst)};
Inst* const inst{&inst_pool->emplace_back(base_inst)};
return instructions.insert(insertion_point, *inst);
}
Block::iterator Block::PrependNewInst(iterator insertion_point, Opcode op,
std::initializer_list<Value> args, u32 flags) {
Inst* const inst{inst_pool->Create(op, flags)};
Block::iterator Block::PrependNewInst(iterator insertion_point, Opcode op, std::initializer_list<Value> args, u32 flags) {
Inst* const inst{&inst_pool->emplace_back(op, flags)};
const auto result_it{instructions.insert(insertion_point, *inst)};
if (inst->NumArgs() != args.size()) {
@@ -53,12 +52,10 @@ void Block::AddBranch(Block* block) {
block->imm_predecessors.push_back(this);
}
static std::string BlockToIndex(const std::map<const Block*, size_t>& block_to_index,
Block* block) {
if (const auto it{block_to_index.find(block)}; it != block_to_index.end()) {
static std::string BlockToIndex(const std::map<const Block*, size_t>& block_to_index, Block* block) {
if (const auto it{block_to_index.find(block)}; it != block_to_index.end())
return fmt::format("{{Block ${}}}", it->second);
}
return fmt::format("$<unknown block {:016x}>", reinterpret_cast<u64>(block));
return fmt::format("$<unknown block {:016x}>", u64(block));
}
static size_t InstIndex(std::map<const Inst*, size_t>& inst_to_index, size_t& inst_index,
@@ -1,4 +1,4 @@
// SPDX-FileCopyrightText: Copyright 2025 Eden Emulator Project
// SPDX-FileCopyrightText: Copyright 2026 Eden Emulator Project
// SPDX-License-Identifier: GPL-3.0-or-later
// SPDX-FileCopyrightText: Copyright 2021 yuzu Emulator Project
@@ -13,11 +13,11 @@
#include <bit>
#include <numeric>
#include <boost/intrusive/list.hpp>
#include <boost/container/stable_vector.hpp>
#include "common/common_types.h"
#include "shader_recompiler/frontend/ir/condition.h"
#include "shader_recompiler/frontend/ir/value.h"
#include "shader_recompiler/object_pool.h"
namespace Shader::IR {
@@ -30,7 +30,7 @@ public:
using reverse_iterator = InstructionList::reverse_iterator;
using const_reverse_iterator = InstructionList::const_reverse_iterator;
explicit Block(ObjectPool<Inst>& inst_pool_);
explicit Block(boost::container::stable_vector<Inst>& inst_pool_);
~Block();
Block(const Block&) = delete;
@@ -170,7 +170,7 @@ public:
private:
/// Memory pool for instruction list
ObjectPool<Inst>* inst_pool;
boost::container::stable_vector<Inst>* inst_pool;
/// List of instructions in this block
InstructionList instructions;
@@ -1,4 +1,4 @@
// SPDX-FileCopyrightText: Copyright 2025 Eden Emulator Project
// SPDX-FileCopyrightText: Copyright 2026 Eden Emulator Project
// SPDX-License-Identifier: GPL-3.0-or-later
// SPDX-FileCopyrightText: Copyright 2021 yuzu Emulator Project
@@ -174,11 +174,10 @@ bool Block::Contains(Location pc) const noexcept {
return pc >= begin && pc < end;
}
Function::Function(ObjectPool<Block>& block_pool, Location start_address)
: entrypoint{start_address} {
Function::Function(boost::container::stable_vector<Block>& block_pool, Location start_address) : entrypoint{start_address} {
Label& label{labels.emplace_back()};
label.address = start_address;
label.block = block_pool.Create(Block{});
label.block = &block_pool.emplace_back(Block{});
label.block->begin = start_address;
label.block->end = start_address;
label.block->end_class = EndClass::Branch;
@@ -187,12 +186,13 @@ Function::Function(ObjectPool<Block>& block_pool, Location start_address)
label.block->branch_false = nullptr;
}
CFG::CFG(Environment& env_, ObjectPool<Block>& block_pool_, Location start_address,
bool exits_to_dispatcher_)
: env{env_}, block_pool{block_pool_}, program_start{start_address}, exits_to_dispatcher{
exits_to_dispatcher_} {
CFG::CFG(Environment& env_, boost::container::stable_vector<Block>& block_pool_, Location start_address, bool exits_to_dispatcher_)
: env{env_}
, block_pool{block_pool_}
, program_start{start_address}
, exits_to_dispatcher{exits_to_dispatcher_} {
if (exits_to_dispatcher) {
dispatch_block = block_pool.Create(Block{});
dispatch_block = &block_pool.emplace_back(Block{});
dispatch_block->begin = {};
dispatch_block->end = {};
dispatch_block->end_class = EndClass::Exit;
@@ -371,7 +371,7 @@ void CFG::AnalyzeCondInst(Block* block, FunctionId function_id, Location pc,
return;
}
// Create a virtual block and a conditional block
Block* const conditional_block{block_pool.Create()};
Block* const conditional_block{&block_pool.emplace_back()};
Block virtual_block{};
virtual_block.begin = block->begin.Virtual();
virtual_block.end = block->begin.Virtual();
@@ -546,7 +546,7 @@ Block* CFG::AddLabel(Block* block, Stack stack, Location pc, FunctionId function
if (label_it != function.labels.end()) {
return label_it->block;
}
Block* const new_block{block_pool.Create()};
Block* const new_block{&block_pool.emplace_back()};
new_block->begin = pc;
new_block->end = pc;
new_block->end_class = EndClass::Branch;
@@ -1,3 +1,6 @@
// SPDX-FileCopyrightText: Copyright 2026 Eden Emulator Project
// SPDX-License-Identifier: GPL-3.0-or-later
// SPDX-FileCopyrightText: Copyright 2021 yuzu Emulator Project
// SPDX-License-Identifier: GPL-2.0-or-later
@@ -9,6 +12,7 @@
#include <vector>
#include <boost/container/small_vector.hpp>
#include <boost/container/stable_vector.hpp>
#include <boost/intrusive/set.hpp>
#include "shader_recompiler/environment.h"
@@ -17,7 +21,6 @@
#include "shader_recompiler/frontend/maxwell/instruction.h"
#include "shader_recompiler/frontend/maxwell/location.h"
#include "shader_recompiler/frontend/maxwell/opcodes.h"
#include "shader_recompiler/object_pool.h"
namespace Shader::Maxwell::Flow {
@@ -96,7 +99,7 @@ struct Label {
};
struct Function {
explicit Function(ObjectPool<Block>& block_pool, Location start_address);
explicit Function(boost::container::stable_vector<Block>& block_pool, Location start_address);
Location entrypoint;
boost::container::small_vector<Label, 16> labels;
@@ -110,8 +113,7 @@ class CFG {
};
public:
explicit CFG(Environment& env, ObjectPool<Block>& block_pool, Location start_address,
bool exits_to_dispatcher = false);
explicit CFG(Environment& env, boost::container::stable_vector<Block>& block_pool, Location start_address, bool exits_to_dispatcher = false);
CFG& operator=(const CFG&) = delete;
CFG(const CFG&) = delete;
@@ -138,27 +140,20 @@ private:
/// Inspect already visited blocks.
/// Return true when the block has already been visited
bool InspectVisitedBlocks(FunctionId function_id, const Label& label);
AnalysisState AnalyzeInst(Block* block, FunctionId function_id, Location pc);
void AnalyzeCondInst(Block* block, FunctionId function_id, Location pc, EndClass insn_end_class,
IR::Condition cond);
void AnalyzeCondInst(Block* block, FunctionId function_id, Location pc, EndClass insn_end_class, IR::Condition cond);
/// Return true when the branch instruction is confirmed to be a branch
bool AnalyzeBranch(Block* block, FunctionId function_id, Location pc, Instruction inst,
Opcode opcode);
void AnalyzeBRA(Block* block, FunctionId function_id, Location pc, Instruction inst,
bool is_absolute);
AnalysisState AnalyzeBRX(Block* block, Location pc, Instruction inst, bool is_absolute,
FunctionId function_id);
bool AnalyzeBranch(Block* block, FunctionId function_id, Location pc, Instruction inst, Opcode opcode);
void AnalyzeBRA(Block* block, FunctionId function_id, Location pc, Instruction inst, bool is_absolute);
AnalysisState AnalyzeBRX(Block* block, Location pc, Instruction inst, bool is_absolute, FunctionId function_id);
AnalysisState AnalyzeEXIT(Block* block, FunctionId function_id, Location pc, Instruction inst);
/// Return the branch target block id
Block* AddLabel(Block* block, Stack stack, Location pc, FunctionId function_id);
Environment& env;
ObjectPool<Block>& block_pool;
boost::container::stable_vector<Block>& block_pool;
boost::container::small_vector<Function, 1> functions;
Location program_start;
bool exits_to_dispatcher{};
@@ -13,143 +13,19 @@
#include <fmt/ranges.h>
#include <boost/intrusive/list.hpp>
#include <ranges>
#include "shader_recompiler/shader_pool.h"
#include "shader_recompiler/environment.h"
#include "shader_recompiler/frontend/ir/basic_block.h"
#include "shader_recompiler/frontend/ir/ir_emitter.h"
#include "shader_recompiler/frontend/maxwell/structured_control_flow.h"
#include "shader_recompiler/frontend/maxwell/translate/translate.h"
#include "shader_recompiler/frontend/maxwell/translate_program.h"
#include "shader_recompiler/frontend/maxwell/control_flow.h"
#include "shader_recompiler/host_translate_info.h"
#include "shader_recompiler/object_pool.h"
namespace Shader::Maxwell {
namespace {
struct Statement;
// Use normal_link because we are not guaranteed to destroy the tree in order
using ListBaseHook =
boost::intrusive::list_base_hook<boost::intrusive::link_mode<boost::intrusive::normal_link>>;
using Tree = boost::intrusive::list<Statement,
// Allow using Statement without a definition
boost::intrusive::base_hook<ListBaseHook>,
// Avoid linear complexity on splice, size is never called
boost::intrusive::constant_time_size<false>>;
using Node = Tree::iterator;
enum class StatementType {
Code,
Goto,
Label,
If,
Loop,
Break,
Return,
Kill,
Unreachable,
Function,
Identity,
Not,
Or,
SetVariable,
SetIndirectBranchVariable,
Variable,
IndirectBranchCond,
};
bool HasChildren(StatementType type) {
switch (type) {
case StatementType::If:
case StatementType::Loop:
case StatementType::Function:
return true;
default:
return false;
}
}
struct Goto {};
struct Label {};
struct If {};
struct Loop {};
struct Break {};
struct Return {};
struct Kill {};
struct Unreachable {};
struct FunctionTag {};
struct Identity {};
struct Not {};
struct Or {};
struct SetVariable {};
struct SetIndirectBranchVariable {};
struct Variable {};
struct IndirectBranchCond {};
#ifdef _MSC_VER
#pragma warning(push)
#pragma warning(disable : 26495) // Always initialize a member variable, expected in Statement
#endif
struct Statement : ListBaseHook {
Statement(const Flow::Block* block_, Statement* up_)
: block{block_}, up{up_}, type{StatementType::Code} {}
Statement(Goto, Statement* cond_, Node label_, Statement* up_)
: label{label_}, cond{cond_}, up{up_}, type{StatementType::Goto} {}
Statement(Label, u32 id_, Statement* up_) : id{id_}, up{up_}, type{StatementType::Label} {}
Statement(If, Statement* cond_, Tree&& children_, Statement* up_)
: children{std::move(children_)}, cond{cond_}, up{up_}, type{StatementType::If} {}
Statement(Loop, Statement* cond_, Tree&& children_, Statement* up_)
: children{std::move(children_)}, cond{cond_}, up{up_}, type{StatementType::Loop} {}
Statement(Break, Statement* cond_, Statement* up_)
: cond{cond_}, up{up_}, type{StatementType::Break} {}
Statement(Return, Statement* up_) : up{up_}, type{StatementType::Return} {}
Statement(Kill, Statement* up_) : up{up_}, type{StatementType::Kill} {}
Statement(Unreachable, Statement* up_) : up{up_}, type{StatementType::Unreachable} {}
Statement(FunctionTag) : children{}, type{StatementType::Function} {}
Statement(Identity, IR::Condition cond_, Statement* up_)
: guest_cond{cond_}, up{up_}, type{StatementType::Identity} {}
Statement(Not, Statement* op_, Statement* up_) : op{op_}, up{up_}, type{StatementType::Not} {}
Statement(Or, Statement* op_a_, Statement* op_b_, Statement* up_)
: op_a{op_a_}, op_b{op_b_}, up{up_}, type{StatementType::Or} {}
Statement(SetVariable, u32 id_, Statement* op_, Statement* up_)
: op{op_}, id{id_}, up{up_}, type{StatementType::SetVariable} {}
Statement(SetIndirectBranchVariable, IR::Reg branch_reg_, s32 branch_offset_, Statement* up_)
: branch_offset{branch_offset_},
branch_reg{branch_reg_}, up{up_}, type{StatementType::SetIndirectBranchVariable} {}
Statement(Variable, u32 id_, Statement* up_)
: id{id_}, up{up_}, type{StatementType::Variable} {}
Statement(IndirectBranchCond, u32 location_, Statement* up_)
: location{location_}, up{up_}, type{StatementType::IndirectBranchCond} {}
~Statement() {
if (HasChildren(type)) {
std::destroy_at(&children);
}
}
union {
const Flow::Block* block;
Node label;
Tree children;
IR::Condition guest_cond;
Statement* op;
Statement* op_a;
u32 location;
s32 branch_offset;
};
union {
Statement* cond;
Statement* op_b;
u32 id;
IR::Reg branch_reg;
};
Statement* up{};
StatementType type;
};
#ifdef _MSC_VER
#pragma warning(pop)
#endif
std::string DumpExpr(const Statement* stmt) {
switch (stmt->type) {
@@ -314,7 +190,9 @@ bool NeedsLift(Node goto_stmt, Node label_stmt) noexcept {
class GotoPass {
public:
explicit GotoPass(Flow::CFG& cfg, ObjectPool<Statement>& stmt_pool) : pool{stmt_pool} {
explicit GotoPass(Flow::CFG& cfg, ShaderPools& pools_)
: pools{pools_}
{
std::vector gotos{BuildTree(cfg)};
const auto end{gotos.rend()};
for (auto goto_stmt = gotos.rbegin(); goto_stmt != end; ++goto_stmt) {
@@ -385,16 +263,14 @@ private:
return gotos;
}
void BuildTree(Flow::CFG& cfg, Flow::Function& function, u32& label_id,
std::vector<Node>& gotos, Node function_insert_point,
std::optional<Node> return_label) {
Statement* const false_stmt{pool.Create(Identity{}, IR::Condition{false}, &root_stmt)};
void BuildTree(Flow::CFG& cfg, Flow::Function& function, u32& label_id, std::vector<Node>& gotos, Node function_insert_point, std::optional<Node> return_label) {
Statement* const false_stmt{&pools.stmt.emplace_back(Identity{}, IR::Condition{false}, &root_stmt)};
Tree& root{root_stmt.children};
::Common::unordered_map<Flow::Block*, Node> local_labels;
local_labels.reserve(function.blocks.size());
for (Flow::Block& block : function.blocks) {
Statement* const label{pool.Create(Label{}, label_id, &root_stmt)};
Statement* const label{&pools.stmt.emplace_back(Label{}, label_id, &root_stmt)};
const Node label_it{root.insert(function_insert_point, *label)};
local_labels.emplace(&block, label_it);
++label_id;
@@ -406,46 +282,39 @@ private:
// Reset goto variables before the first block and after its respective label
const auto make_reset_variable{[&]() -> Statement& {
return *pool.Create(SetVariable{}, label->id, false_stmt, &root_stmt);
return pools.stmt.emplace_back(SetVariable{}, label->id, false_stmt, &root_stmt);
}};
root.push_front(make_reset_variable());
root.insert(ip, make_reset_variable());
root.insert(ip, *pool.Create(&block, &root_stmt));
root.insert(ip, pools.stmt.emplace_back(&block, &root_stmt));
switch (block.end_class) {
case Flow::EndClass::Branch: {
Statement* const always_cond{
pool.Create(Identity{}, IR::Condition{true}, &root_stmt)};
Statement* const always_cond{&pools.stmt.emplace_back(Identity{}, IR::Condition{true}, &root_stmt)};
if (block.cond == IR::Condition{true}) {
const Node true_label{local_labels.at(block.branch_true)};
gotos.push_back(
root.insert(ip, *pool.Create(Goto{}, always_cond, true_label, &root_stmt)));
gotos.push_back(root.insert(ip, pools.stmt.emplace_back(Goto{}, always_cond, true_label, &root_stmt)));
} else if (block.cond == IR::Condition{false}) {
const Node false_label{local_labels.at(block.branch_false)};
gotos.push_back(root.insert(
ip, *pool.Create(Goto{}, always_cond, false_label, &root_stmt)));
gotos.push_back(root.insert(ip, pools.stmt.emplace_back(Goto{}, always_cond, false_label, &root_stmt)));
} else {
const Node true_label{local_labels.at(block.branch_true)};
const Node false_label{local_labels.at(block.branch_false)};
Statement* const true_cond{pool.Create(Identity{}, block.cond, &root_stmt)};
gotos.push_back(
root.insert(ip, *pool.Create(Goto{}, true_cond, true_label, &root_stmt)));
gotos.push_back(root.insert(
ip, *pool.Create(Goto{}, always_cond, false_label, &root_stmt)));
Statement* const true_cond{&pools.stmt.emplace_back(Identity{}, block.cond, &root_stmt)};
gotos.push_back(root.insert(ip, pools.stmt.emplace_back(Goto{}, true_cond, true_label, &root_stmt)));
gotos.push_back(root.insert(ip, pools.stmt.emplace_back(Goto{}, always_cond, false_label, &root_stmt)));
}
break;
}
case Flow::EndClass::IndirectBranch:
root.insert(ip, *pool.Create(SetIndirectBranchVariable{}, block.branch_reg,
block.branch_offset, &root_stmt));
root.insert(ip, pools.stmt.emplace_back(SetIndirectBranchVariable{}, block.branch_reg, block.branch_offset, &root_stmt));
for (const Flow::IndirectBranch& indirect : block.indirect_branches) {
const Node indirect_label{local_labels.at(indirect.block)};
Statement* cond{
pool.Create(IndirectBranchCond{}, indirect.address, &root_stmt)};
Statement* goto_stmt{pool.Create(Goto{}, cond, indirect_label, &root_stmt)};
Statement* cond{&pools.stmt.emplace_back(IndirectBranchCond{}, indirect.address, &root_stmt)};
Statement* goto_stmt{&pools.stmt.emplace_back(Goto{}, cond, indirect_label, &root_stmt)};
gotos.push_back(root.insert(ip, *goto_stmt));
}
root.insert(ip, *pool.Create(Unreachable{}, &root_stmt));
root.insert(ip, pools.stmt.emplace_back(Unreachable{}, &root_stmt));
break;
case Flow::EndClass::Call: {
Flow::Function& call{cfg.Functions()[block.function_call]};
@@ -454,16 +323,16 @@ private:
break;
}
case Flow::EndClass::Exit:
root.insert(ip, *pool.Create(Return{}, &root_stmt));
root.insert(ip, pools.stmt.emplace_back(Return{}, &root_stmt));
break;
case Flow::EndClass::Return: {
Statement* const always_cond{pool.Create(Identity{}, block.cond, &root_stmt)};
auto goto_stmt{pool.Create(Goto{}, always_cond, return_label.value(), &root_stmt)};
Statement* const always_cond{&pools.stmt.emplace_back(Identity{}, block.cond, &root_stmt)};
auto goto_stmt{&pools.stmt.emplace_back(Goto{}, always_cond, return_label.value(), &root_stmt)};
gotos.push_back(root.insert(ip, *goto_stmt));
break;
}
case Flow::EndClass::Kill:
root.insert(ip, *pool.Create(Kill{}, &root_stmt));
root.insert(ip, pools.stmt.emplace_back(Kill{}, &root_stmt));
break;
}
}
@@ -479,8 +348,8 @@ private:
Tree& body{goto_stmt->up->children};
Tree if_body;
if_body.splice(if_body.begin(), body, std::next(goto_stmt), label_stmt);
Statement* const cond{pool.Create(Not{}, goto_stmt->cond, &root_stmt)};
Statement* const if_stmt{pool.Create(If{}, cond, std::move(if_body), goto_stmt->up)};
Statement* const cond{&pools.stmt.emplace_back(Not{}, goto_stmt->cond, &root_stmt)};
Statement* const if_stmt{&pools.stmt.emplace_back(If{}, cond, std::move(if_body), goto_stmt->up)};
UpdateTreeUp(if_stmt);
body.insert(goto_stmt, *if_stmt);
body.erase(goto_stmt);
@@ -491,7 +360,7 @@ private:
Tree loop_body;
loop_body.splice(loop_body.begin(), body, label_stmt, goto_stmt);
Statement* const cond{goto_stmt->cond};
Statement* const loop{pool.Create(Loop{}, cond, std::move(loop_body), goto_stmt->up)};
Statement* const loop{&pools.stmt.emplace_back(Loop{}, cond, std::move(loop_body), goto_stmt->up)};
UpdateTreeUp(loop);
body.insert(goto_stmt, *loop);
body.erase(goto_stmt);
@@ -516,15 +385,15 @@ private:
const u32 label_id{label->id};
Statement* const goto_cond{goto_stmt->cond};
Statement* const set_var{pool.Create(SetVariable{}, label_id, goto_cond, parent)};
Statement* const set_var{&pools.stmt.emplace_back(SetVariable{}, label_id, goto_cond, parent)};
body.insert(goto_stmt, *set_var);
Tree if_body;
if_body.splice(if_body.begin(), body, std::next(goto_stmt), label_nested_stmt);
Statement* const variable{pool.Create(Variable{}, label_id, &root_stmt)};
Statement* const neg_var{pool.Create(Not{}, variable, &root_stmt)};
Statement* const variable{&pools.stmt.emplace_back(Variable{}, label_id, &root_stmt)};
Statement* const neg_var{&pools.stmt.emplace_back(Not{}, variable, &root_stmt)};
if (!if_body.empty()) {
Statement* const if_stmt{pool.Create(If{}, neg_var, std::move(if_body), parent)};
Statement* const if_stmt{&pools.stmt.emplace_back(If{}, neg_var, std::move(if_body), parent)};
UpdateTreeUp(if_stmt);
body.insert(goto_stmt, *if_stmt);
}
@@ -533,8 +402,7 @@ private:
switch (label_nested_stmt->type) {
case StatementType::If:
// Update nested if condition
label_nested_stmt->cond =
pool.Create(Or{}, variable, label_nested_stmt->cond, &root_stmt);
label_nested_stmt->cond = &pools.stmt.emplace_back(Or{}, variable, label_nested_stmt->cond, &root_stmt);
break;
case StatementType::Loop:
break;
@@ -542,7 +410,7 @@ private:
throw LogicError("Invalid inward movement");
}
Tree& nested_tree{label_nested_stmt->children};
Statement* const new_goto{pool.Create(Goto{}, variable, label, &*label_nested_stmt)};
Statement* const new_goto{&pools.stmt.emplace_back(Goto{}, variable, label, &*label_nested_stmt)};
return nested_tree.insert(nested_tree.begin(), *new_goto);
}
@@ -556,16 +424,16 @@ private:
Tree loop_body;
loop_body.splice(loop_body.begin(), body, label_nested_stmt, goto_stmt);
SanitizeNoBreaks(loop_body);
Statement* const variable{pool.Create(Variable{}, label_id, &root_stmt)};
Statement* const loop_stmt{pool.Create(Loop{}, variable, std::move(loop_body), parent)};
Statement* const variable{&pools.stmt.emplace_back(Variable{}, label_id, &root_stmt)};
Statement* const loop_stmt{&pools.stmt.emplace_back(Loop{}, variable, std::move(loop_body), parent)};
UpdateTreeUp(loop_stmt);
body.insert(goto_stmt, *loop_stmt);
Statement* const new_goto{pool.Create(Goto{}, variable, label, loop_stmt)};
Statement* const new_goto{&pools.stmt.emplace_back(Goto{}, variable, label, loop_stmt)};
loop_stmt->children.push_front(*new_goto);
const Node new_goto_node{loop_stmt->children.begin()};
Statement* const set_var{pool.Create(SetVariable{}, label_id, goto_stmt->cond, loop_stmt)};
Statement* const set_var{&pools.stmt.emplace_back(SetVariable{}, label_id, goto_stmt->cond, loop_stmt)};
loop_stmt->children.push_back(*set_var);
body.erase(goto_stmt);
@@ -577,22 +445,22 @@ private:
Tree& body{parent->children};
const u32 label_id{goto_stmt->label->id};
Statement* const goto_cond{goto_stmt->cond};
Statement* const set_goto_var{pool.Create(SetVariable{}, label_id, goto_cond, &*parent)};
Statement* const set_goto_var{&pools.stmt.emplace_back(SetVariable{}, label_id, goto_cond, &*parent)};
body.insert(goto_stmt, *set_goto_var);
Tree if_body;
if_body.splice(if_body.begin(), body, std::next(goto_stmt), body.end());
if_body.pop_front();
Statement* const cond{pool.Create(Variable{}, label_id, &root_stmt)};
Statement* const neg_cond{pool.Create(Not{}, cond, &root_stmt)};
Statement* const if_stmt{pool.Create(If{}, neg_cond, std::move(if_body), &*parent)};
Statement* const cond{&pools.stmt.emplace_back(Variable{}, label_id, &root_stmt)};
Statement* const neg_cond{&pools.stmt.emplace_back(Not{}, cond, &root_stmt)};
Statement* const if_stmt{&pools.stmt.emplace_back(If{}, neg_cond, std::move(if_body), &*parent)};
UpdateTreeUp(if_stmt);
body.insert(goto_stmt, *if_stmt);
body.erase(goto_stmt);
Statement* const new_cond{pool.Create(Variable{}, label_id, &root_stmt)};
Statement* const new_goto{pool.Create(Goto{}, new_cond, goto_stmt->label, parent->up)};
Statement* const new_cond{&pools.stmt.emplace_back(Variable{}, label_id, &root_stmt)};
Statement* const new_goto{&pools.stmt.emplace_back(Goto{}, new_cond, goto_stmt->label, parent->up)};
Tree& parent_tree{parent->up->children};
return parent_tree.insert(std::next(parent), *new_goto);
}
@@ -602,21 +470,21 @@ private:
Tree& body{parent->children};
const u32 label_id{goto_stmt->label->id};
Statement* const goto_cond{goto_stmt->cond};
Statement* const set_goto_var{pool.Create(SetVariable{}, label_id, goto_cond, parent)};
Statement* const cond{pool.Create(Variable{}, label_id, &root_stmt)};
Statement* const break_stmt{pool.Create(Break{}, cond, parent)};
Statement* const set_goto_var{&pools.stmt.emplace_back(SetVariable{}, label_id, goto_cond, parent)};
Statement* const cond{&pools.stmt.emplace_back(Variable{}, label_id, &root_stmt)};
Statement* const break_stmt{&pools.stmt.emplace_back(Break{}, cond, parent)};
body.insert(goto_stmt, *set_goto_var);
body.insert(goto_stmt, *break_stmt);
body.erase(goto_stmt);
const Node loop{Tree::s_iterator_to(*goto_stmt->up)};
Statement* const new_goto_cond{pool.Create(Variable{}, label_id, &root_stmt)};
Statement* const new_goto{pool.Create(Goto{}, new_goto_cond, goto_stmt->label, loop->up)};
Statement* const new_goto_cond{&pools.stmt.emplace_back(Variable{}, label_id, &root_stmt)};
Statement* const new_goto{&pools.stmt.emplace_back(Goto{}, new_goto_cond, goto_stmt->label, loop->up)};
Tree& parent_tree{loop->up->children};
return parent_tree.insert(std::next(loop), *new_goto);
}
ObjectPool<Statement>& pool;
ShaderPools& pools;
Statement root_stmt{FunctionTag{}};
};
@@ -652,11 +520,11 @@ private:
class TranslatePass {
public:
TranslatePass(ObjectPool<IR::Inst>& inst_pool_, ObjectPool<IR::Block>& block_pool_,
ObjectPool<Statement>& stmt_pool_, Environment& env_, Statement& root_stmt,
IR::AbstractSyntaxList& syntax_list_, const HostTranslateInfo& host_info)
: stmt_pool{stmt_pool_}, inst_pool{inst_pool_}, block_pool{block_pool_}, env{env_},
syntax_list{syntax_list_} {
TranslatePass(ShaderPools& pools_, Environment& env_, Statement& root_stmt, IR::AbstractSyntaxList& syntax_list_, const HostTranslateInfo& host_info)
: pools{pools_}
, env{env_}
, syntax_list{syntax_list_}
{
Visit(root_stmt, nullptr, nullptr);
IR::Block& first_block{*syntax_list.front().data.block};
@@ -674,7 +542,7 @@ private:
if (current_block) {
return;
}
current_block = block_pool.Create(inst_pool);
current_block = &pools.block.emplace_back(pools.inst);
auto& node{syntax_list.emplace_back()};
node.type = IR::AbstractSyntaxNode::Type::Block;
node.data.block = current_block;
@@ -740,7 +608,7 @@ private:
break;
}
case StatementType::Loop: {
IR::Block* const loop_header_block{block_pool.Create(inst_pool)};
IR::Block* const loop_header_block{&pools.block.emplace_back(pools.inst)};
if (current_block) {
current_block->AddBranch(loop_header_block);
}
@@ -748,7 +616,7 @@ private:
header_node.type = IR::AbstractSyntaxNode::Type::Block;
header_node.data.block = loop_header_block;
IR::Block* const continue_block{block_pool.Create(inst_pool)};
IR::Block* const continue_block{&pools.block.emplace_back(pools.inst)};
IR::Block* const merge_block{MergeBlock(parent, stmt)};
const size_t loop_node_index{syntax_list.size()};
@@ -814,7 +682,7 @@ private:
}
case StatementType::Return: {
ensure_block();
IR::Block* return_block{block_pool.Create(inst_pool)};
IR::Block* return_block{&pools.block.emplace_back(pools.inst)};
IR::IREmitter{*return_block}.Epilogue();
current_block->AddBranch(return_block);
@@ -862,10 +730,10 @@ private:
Statement* merge_stmt{TryFindForwardBlock(stmt)};
if (!merge_stmt) {
// Create a merge block we can visit later
merge_stmt = stmt_pool.Create(&dummy_flow_block, &parent);
merge_stmt = &pools.stmt.emplace_back(&dummy_flow_block, &parent);
parent.children.insert(std::next(Tree::s_iterator_to(stmt)), *merge_stmt);
}
return block_pool.Create(inst_pool);
return &pools.block.emplace_back(pools.inst);
}
void DemoteCombinationPass() {
@@ -973,9 +841,7 @@ private:
asl.insert(next_it_2, demote_if_node);
}
ObjectPool<Statement>& stmt_pool;
ObjectPool<IR::Inst>& inst_pool;
ObjectPool<IR::Block>& block_pool;
ShaderPools& pools;
Environment& env;
IR::AbstractSyntaxList& syntax_list;
bool uses_demote_to_helper{};
@@ -983,15 +849,12 @@ private:
};
} // Anonymous namespace
IR::AbstractSyntaxList BuildASL(ObjectPool<IR::Inst>& inst_pool, ObjectPool<IR::Block>& block_pool,
Environment& env, Flow::CFG& cfg,
const HostTranslateInfo& host_info) {
ObjectPool<Statement> stmt_pool{64};
GotoPass goto_pass{cfg, stmt_pool};
IR::AbstractSyntaxList BuildASL(ShaderPools& pools, Environment& env, Flow::CFG& cfg, const HostTranslateInfo& host_info) {
GotoPass goto_pass{cfg, pools};
Statement& root{goto_pass.RootStatement()};
IR::AbstractSyntaxList syntax_list;
TranslatePass{inst_pool, block_pool, stmt_pool, env, root, syntax_list, host_info};
stmt_pool.ReleaseContents();
TranslatePass pass{pools, env, root, syntax_list, host_info};
pools.stmt.clear();
return syntax_list;
}
@@ -1,3 +1,6 @@
// SPDX-FileCopyrightText: Copyright 2026 Eden Emulator Project
// SPDX-License-Identifier: GPL-3.0-or-later
// SPDX-FileCopyrightText: Copyright 2021 yuzu Emulator Project
// SPDX-License-Identifier: GPL-2.0-or-later
@@ -8,15 +11,13 @@
#include "shader_recompiler/frontend/ir/basic_block.h"
#include "shader_recompiler/frontend/ir/value.h"
#include "shader_recompiler/frontend/maxwell/control_flow.h"
#include "shader_recompiler/object_pool.h"
namespace Shader {
struct HostTranslateInfo;
namespace Maxwell {
[[nodiscard]] IR::AbstractSyntaxList BuildASL(ObjectPool<IR::Inst>& inst_pool,
ObjectPool<IR::Block>& block_pool, Environment& env,
Flow::CFG& cfg, const HostTranslateInfo& host_info);
struct ShaderPools;
[[nodiscard]] IR::AbstractSyntaxList BuildASL(ShaderPools& pools, Environment& env, Flow::CFG& cfg, const HostTranslateInfo& host_info);
} // namespace Maxwell
} // namespace Shader
@@ -10,6 +10,7 @@
#include <queue>
#include "common/settings.h"
#include "shader_recompiler/shader_pool.h"
#include "shader_recompiler/exception.h"
#include "shader_recompiler/frontend/ir/basic_block.h"
#include "shader_recompiler/frontend/ir/ir_emitter.h"
@@ -132,14 +133,6 @@ void AddNVNStorageBuffers(IR::Program& program) {
}
}
void ConfigureStorageBufferMappings(IR::Program& program, const HostTranslateInfo& host_info) {
// https://docs.vulkan.org/guide/latest/descriptor_arrays.html
// will be needed: descriptor array elements represent physical spans of one guest virtual buffer.
for (StorageBufferDescriptor& desc : program.info.storage_buffers_descriptors) {
desc.count = host_info.storage_buffer_segment_count;
}
}
using IR::IsLegacyAttribute; //rescoped to attribute.h to make it visible in load_store_attribute.cpp IPA
std::map<IR::Attribute, IR::Attribute> GenerateLegacyToGenericMappings(
@@ -242,13 +235,12 @@ void LowerGeometryPassthrough(const IR::Program& program, const HostTranslateInf
} // Anonymous namespace
IR::Program TranslateProgram(ObjectPool<IR::Inst>& inst_pool, ObjectPool<IR::Block>& block_pool,
Environment& env, Flow::CFG& cfg, const HostTranslateInfo& host_info) {
IR::Program TranslateProgram(ShaderPools& pools, Environment& env, Flow::CFG& cfg, const HostTranslateInfo& host_info) {
HostTranslateInfo normalized_host_info{host_info};
normalized_host_info.ApplyDescriptorLimitPolicy();
IR::Program program;
program.syntax_list = BuildASL(inst_pool, block_pool, env, cfg, normalized_host_info);
program.syntax_list = BuildASL(pools, env, cfg, host_info);
program.blocks = GenerateBlocks(program.syntax_list);
program.post_order_blocks = PostOrder(program.syntax_list.front());
program.stage = env.ShaderStage();
@@ -307,7 +299,6 @@ IR::Program TranslateProgram(ObjectPool<IR::Inst>& inst_pool, ObjectPool<IR::Blo
Optimization::PositionPass(env, program);
Optimization::GlobalMemoryToStorageBufferPass(program, normalized_host_info);
ConfigureStorageBufferMappings(program, normalized_host_info);
Optimization::TexturePass(env, program, normalized_host_info);
if (Settings::values.resolution_info.active || Settings::values.rescale_hack.GetValue()) {
@@ -323,7 +314,6 @@ IR::Program TranslateProgram(ObjectPool<IR::Inst>& inst_pool, ObjectPool<IR::Blo
CollectInterpolationInfo(env, program);
AddNVNStorageBuffers(program);
ConfigureStorageBufferMappings(program, normalized_host_info);
return program;
}
@@ -420,11 +410,7 @@ void ConvertLegacyToGeneric(IR::Program& program, const Shader::RuntimeInfo& run
}
}
IR::Program GenerateGeometryPassthrough(ObjectPool<IR::Inst>& inst_pool,
ObjectPool<IR::Block>& block_pool,
const HostTranslateInfo& host_info,
IR::Program& source_program,
Shader::OutputTopology output_topology) {
IR::Program GenerateGeometryPassthrough(ShaderPools& pools, const HostTranslateInfo& host_info, IR::Program& source_program, Shader::OutputTopology output_topology) {
IR::Program program;
program.stage = Stage::Geometry;
program.output_topology = output_topology;
@@ -436,16 +422,15 @@ IR::Program GenerateGeometryPassthrough(ObjectPool<IR::Inst>& inst_pool,
program.info.stores.Set(IR::Attribute::Layer, true);
program.info.stores.Set(source_program.info.emulated_layer, false);
IR::Block* current_block = block_pool.Create(inst_pool);
IR::Block* current_block = &pools.block.emplace_back(pools.inst);
auto& node{program.syntax_list.emplace_back()};
node.type = IR::AbstractSyntaxNode::Type::Block;
node.data.block = current_block;
IR::IREmitter ir{*current_block};
EmitGeometryPassthrough(ir, program, program.info.stores, true,
source_program.info.emulated_layer);
EmitGeometryPassthrough(ir, program, program.info.stores, true, source_program.info.emulated_layer);
IR::Block* return_block{block_pool.Create(inst_pool)};
IR::Block* return_block{&pools.block.emplace_back(pools.inst)};
IR::IREmitter{*return_block}.Epilogue();
current_block->AddBranch(return_block);
@@ -1,3 +1,6 @@
// SPDX-FileCopyrightText: Copyright 2026 Eden Emulator Project
// SPDX-License-Identifier: GPL-3.0-or-later
// SPDX-FileCopyrightText: Copyright 2021 yuzu Emulator Project
// SPDX-License-Identifier: GPL-2.0-or-later
@@ -7,7 +10,7 @@
#include "shader_recompiler/frontend/ir/basic_block.h"
#include "shader_recompiler/frontend/ir/program.h"
#include "shader_recompiler/frontend/maxwell/control_flow.h"
#include "shader_recompiler/object_pool.h"
#include "shader_recompiler/frontend/maxwell/structured_control_flow.h"
#include "shader_recompiler/runtime_info.h"
namespace Shader {
@@ -16,22 +19,17 @@ struct HostTranslateInfo;
namespace Shader::Maxwell {
[[nodiscard]] IR::Program TranslateProgram(ObjectPool<IR::Inst>& inst_pool,
ObjectPool<IR::Block>& block_pool, Environment& env,
Flow::CFG& cfg, const HostTranslateInfo& host_info);
struct ShaderPools;
[[nodiscard]] IR::Program MergeDualVertexPrograms(IR::Program& vertex_a, IR::Program& vertex_b,
Environment& env_vertex_b);
[[nodiscard]] IR::Program TranslateProgram(ShaderPools& pools, Environment& env, Flow::CFG& cfg, const HostTranslateInfo& host_info);
[[nodiscard]] IR::Program MergeDualVertexPrograms(IR::Program& vertex_a, IR::Program& vertex_b, Environment& env_vertex_b);
void ConvertLegacyToGeneric(IR::Program& program, const RuntimeInfo& runtime_info);
// Maxwell v1 and older Nvidia cards don't support setting gl_Layer from non-geometry stages.
// This creates a workaround by setting the layer as a generic output and creating a
// passthrough geometry shader that reads the generic and sets the layer.
[[nodiscard]] IR::Program GenerateGeometryPassthrough(ObjectPool<IR::Inst>& inst_pool,
ObjectPool<IR::Block>& block_pool,
const HostTranslateInfo& host_info,
IR::Program& source_program,
Shader::OutputTopology output_topology);
[[nodiscard]] IR::Program GenerateGeometryPassthrough(ShaderPools& pools, const HostTranslateInfo& host_info, IR::Program& source_program, Shader::OutputTopology output_topology);
} // namespace Shader::Maxwell
@@ -6,7 +6,6 @@
#pragma once
#include <algorithm>
#include "common/common_types.h"
namespace Shader {
@@ -20,7 +19,6 @@ struct HostTranslateInfo {
u64 min_ssbo_alignment{}; ///< Minimum alignment supported by the device for SSBOs
u32 max_per_stage_descriptor_sampled_images{}; ///< maximum sampled descriptors per stage
u32 max_per_stage_descriptor_storage_buffers{}; ///< maximum storage descriptors per stage
u32 max_per_stage_resources{}; ///< maximum resources per stage
u32 max_descriptor_set_samplers{};
u32 max_descriptor_set_uniform_buffers{};
@@ -40,14 +38,12 @@ struct HostTranslateInfo {
///< passthrough shaders
bool support_conditional_barrier{}; ///< True when the device supports barriers in conditional
///< control flow
u32 storage_buffer_segment_count{1}; ///< Physical ranges available to each guest SSBO
void ApplyDescriptorLimitPolicy() noexcept {
if (min_ssbo_alignment == 0) {
min_ssbo_alignment = 1;
}
ApplyDescriptorLimitFallback(max_per_stage_descriptor_sampled_images);
ApplyDescriptorLimitFallback(max_per_stage_descriptor_storage_buffers);
ApplyDescriptorLimitFallback(max_per_stage_resources);
ApplyDescriptorLimitFallback(max_descriptor_set_samplers);
ApplyDescriptorLimitFallback(max_descriptor_set_uniform_buffers);
@@ -57,7 +53,6 @@ struct HostTranslateInfo {
ApplyDescriptorLimitFallback(max_descriptor_set_sampled_images);
ApplyDescriptorLimitFallback(max_descriptor_set_storage_images);
ApplyDescriptorLimitFallback(max_descriptor_set_input_attachements);
storage_buffer_segment_count = (std::max)(storage_buffer_segment_count, 1U);
}
private:
-106
View File
@@ -1,106 +0,0 @@
// SPDX-FileCopyrightText: Copyright 2021 yuzu Emulator Project
// SPDX-License-Identifier: GPL-2.0-or-later
#pragma once
#include <memory>
#include <type_traits>
#include <utility>
namespace Shader {
template <typename T>
requires std::is_destructible_v<T>
class ObjectPool {
public:
explicit ObjectPool(size_t chunk_size = 8192) : new_chunk_size{chunk_size} {
node = &chunks.emplace_back(new_chunk_size);
}
template <typename... Args>
requires std::is_constructible_v<T, Args...>
[[nodiscard]] T* Create(Args&&... args) {
return std::construct_at(Memory(), std::forward<Args>(args)...);
}
void ReleaseContents() {
if (chunks.empty()) {
return;
}
Chunk& root{chunks.front()};
if (root.used_objects == root.num_objects) {
// Root chunk has been filled, squash allocations into it
const size_t total_objects{root.num_objects + new_chunk_size * (chunks.size() - 1)};
chunks.clear();
chunks.emplace_back(total_objects);
} else {
root.Release();
chunks.resize(1);
}
chunks.shrink_to_fit();
node = &chunks.front();
}
private:
struct NonTrivialDummy {
NonTrivialDummy() noexcept {}
};
union Storage {
Storage() noexcept {}
~Storage() noexcept {}
NonTrivialDummy dummy{};
T object;
};
struct Chunk {
explicit Chunk() = default;
explicit Chunk(size_t size)
: num_objects{size}, storage{std::make_unique<Storage[]>(size)} {}
Chunk& operator=(Chunk&& rhs) noexcept {
Release();
used_objects = std::exchange(rhs.used_objects, 0);
num_objects = std::exchange(rhs.num_objects, 0);
storage = std::move(rhs.storage);
return *this;
}
Chunk(Chunk&& rhs) noexcept
: used_objects{std::exchange(rhs.used_objects, 0)},
num_objects{std::exchange(rhs.num_objects, 0)}, storage{std::move(rhs.storage)} {}
~Chunk() {
Release();
}
void Release() {
std::destroy_n(storage.get(), used_objects);
used_objects = 0;
}
size_t used_objects{};
size_t num_objects{};
std::unique_ptr<Storage[]> storage;
};
[[nodiscard]] T* Memory() {
Chunk* const chunk{FreeChunk()};
return &chunk->storage[chunk->used_objects++].object;
}
[[nodiscard]] Chunk* FreeChunk() {
if (node->used_objects != node->num_objects) {
return node;
}
node = &chunks.emplace_back(new_chunk_size);
return node;
}
Chunk* node{};
std::vector<Chunk> chunks;
size_t new_chunk_size{};
};
} // namespace Shader
-1
View File
@@ -64,7 +64,6 @@ struct Profile {
bool support_storage_image_array_nonuniform_indexing{};
bool support_uniform_texel_buffer_array_nonuniform_indexing{};
bool support_storage_texel_buffer_array_nonuniform_indexing{};
bool support_storage_buffer_array_nonuniform_indexing{};
bool warp_size_potentially_larger_than_guest{};
-5
View File
@@ -6,7 +6,6 @@
#pragma once
#include <algorithm>
#include <array>
#include <bitset>
#include <map>
@@ -341,10 +340,6 @@ struct Info {
ImageDescriptors image_descriptors;
};
[[nodiscard]] inline bool UsesStorageBufferMappings(const Info& info) noexcept {
return std::ranges::any_of(info.storage_buffers_descriptors, [](const auto& desc) { return desc.count > 1; });
}
template <typename Descriptors>
u32 NumDescriptors(const Descriptors& descriptors) {
u32 num{};
+134
View File
@@ -0,0 +1,134 @@
// SPDX-FileCopyrightText: Copyright 2026 Eden Emulator Project
// SPDX-License-Identifier: GPL-3.0-or-later
#pragma once
#include <boost/container/stable_vector.hpp>
#include <boost/intrusive/list.hpp>
#include "shader_recompiler/frontend/maxwell/control_flow.h"
namespace Shader::Maxwell {
struct Statement;
// Use normal_link because we are not guaranteed to destroy the tree in order
using ListBaseHook = boost::intrusive::list_base_hook<boost::intrusive::link_mode<boost::intrusive::normal_link>>;
using Tree = boost::intrusive::list<Statement,
// Allow using Statement without a definition
boost::intrusive::base_hook<ListBaseHook>,
// Avoid linear complexity on splice, size is never called
boost::intrusive::constant_time_size<false>>;
using Node = Tree::iterator;
enum class StatementType {
Code,
Goto,
Label,
If,
Loop,
Break,
Return,
Kill,
Unreachable,
Function,
Identity,
Not,
Or,
SetVariable,
SetIndirectBranchVariable,
Variable,
IndirectBranchCond,
};
[[nodiscard]] inline bool HasChildren(StatementType type) {
switch (type) {
case StatementType::If:
case StatementType::Loop:
case StatementType::Function:
return true;
default:
return false;
}
}
struct Goto {};
struct Label {};
struct If {};
struct Loop {};
struct Break {};
struct Return {};
struct Kill {};
struct Unreachable {};
struct FunctionTag {};
struct Identity {};
struct Not {};
struct Or {};
struct SetVariable {};
struct SetIndirectBranchVariable {};
struct Variable {};
struct IndirectBranchCond {};
#ifdef _MSC_VER
#pragma warning(push)
#pragma warning(disable : 26495) // Always initialize a member variable, expected in Statement
#endif
struct Statement : ListBaseHook {
Statement(const Flow::Block* block_, Statement* up_) : block{block_}, up{up_}, type{StatementType::Code} {}
Statement(Goto, Statement* cond_, Node label_, Statement* up_) : label{label_}, cond{cond_}, up{up_}, type{StatementType::Goto} {}
Statement(Label, u32 id_, Statement* up_) : id{id_}, up{up_}, type{StatementType::Label} {}
Statement(If, Statement* cond_, Tree&& children_, Statement* up_) : children{std::move(children_)}, cond{cond_}, up{up_}, type{StatementType::If} {}
Statement(Loop, Statement* cond_, Tree&& children_, Statement* up_) : children{std::move(children_)}, cond{cond_}, up{up_}, type{StatementType::Loop} {}
Statement(Break, Statement* cond_, Statement* up_) : cond{cond_}, up{up_}, type{StatementType::Break} {}
Statement(Return, Statement* up_) : up{up_}, type{StatementType::Return} {}
Statement(Kill, Statement* up_) : up{up_}, type{StatementType::Kill} {}
Statement(Unreachable, Statement* up_) : up{up_}, type{StatementType::Unreachable} {}
Statement(FunctionTag) : children{}, type{StatementType::Function} {}
Statement(Identity, IR::Condition cond_, Statement* up_) : guest_cond{cond_}, up{up_}, type{StatementType::Identity} {}
Statement(Not, Statement* op_, Statement* up_) : op{op_}, up{up_}, type{StatementType::Not} {}
Statement(Or, Statement* op_a_, Statement* op_b_, Statement* up_) : op_a{op_a_}, op_b{op_b_}, up{up_}, type{StatementType::Or} {}
Statement(SetVariable, u32 id_, Statement* op_, Statement* up_) : op{op_}, id{id_}, up{up_}, type{StatementType::SetVariable} {}
Statement(SetIndirectBranchVariable, IR::Reg branch_reg_, s32 branch_offset_, Statement* up_) : branch_offset{branch_offset_}, branch_reg{branch_reg_}, up{up_}, type{StatementType::SetIndirectBranchVariable} {}
Statement(Variable, u32 id_, Statement* up_) : id{id_}, up{up_}, type{StatementType::Variable} {}
Statement(IndirectBranchCond, u32 location_, Statement* up_) : location{location_}, up{up_}, type{StatementType::IndirectBranchCond} {}
~Statement() {
if (HasChildren(type)) {
std::destroy_at(&children);
}
}
union {
const Flow::Block* block;
Node label;
Tree children;
IR::Condition guest_cond;
Statement* op;
Statement* op_a;
u32 location;
s32 branch_offset;
};
union {
Statement* cond;
Statement* op_b;
u32 id;
IR::Reg branch_reg;
};
Statement* up{};
StatementType type;
};
#ifdef _MSC_VER
#pragma warning(pop)
#endif
struct ShaderPools {
void clear() {
flow_block.clear();
block.clear();
inst.clear();
stmt.clear();
}
boost::container::stable_vector<Shader::IR::Inst> inst{};
boost::container::stable_vector<Shader::IR::Block> block{};
boost::container::stable_vector<Shader::Maxwell::Flow::Block> flow_block{};
boost::container::stable_vector<Shader::Maxwell::Statement> stmt;
};
}
+64 -158
View File
@@ -7,8 +7,6 @@
#pragma once
#include <algorithm>
#include <cstring>
#include <limits>
#include <memory>
#include <numeric>
@@ -425,7 +423,7 @@ void BufferCache<P>::UnbindGraphicsStorageBuffers(size_t stage) {
template <class P>
bool BufferCache<P>::BindGraphicsStorageBuffer(size_t stage, size_t ssbo_index, u32 cbuf_index,
u32 cbuf_offset, bool is_written, u32 descriptor_count) {
u32 cbuf_offset, bool is_written) {
const bool already_enabled =
((channel_state->enabled_storage_buffers[stage] >> ssbo_index) & 1U) != 0;
if constexpr (requires { runtime.ShouldLimitDynamicStorageBuffers(); }) {
@@ -450,8 +448,8 @@ bool BufferCache<P>::BindGraphicsStorageBuffer(size_t stage, size_t ssbo_index,
const auto& cbufs = maxwell3d->state.shader_stages[stage];
const GPUVAddr ssbo_addr = cbufs.const_buffers[cbuf_index].address + cbuf_offset;
channel_state->storage_buffers[stage][ssbo_index] =
StorageBufferBinding(ssbo_addr, cbuf_index, is_written, descriptor_count);
return channel_state->storage_buffers[stage][ssbo_index].gpu_addr != 0;
StorageBufferBinding(ssbo_addr, cbuf_index, is_written);
return (channel_state->storage_buffers[stage][ssbo_index].buffer_id != NULL_BUFFER_ID);
}
template <class P>
@@ -489,7 +487,7 @@ void BufferCache<P>::UnbindComputeStorageBuffers() {
template <class P>
void BufferCache<P>::BindComputeStorageBuffer(size_t ssbo_index, u32 cbuf_index, u32 cbuf_offset,
bool is_written, u32 descriptor_count) {
bool is_written) {
if (ssbo_index >= channel_state->compute_storage_buffers.size()) [[unlikely]] {
LOG_ERROR(HW_GPU, "Storage buffer index {} exceeds maximum storage buffer count",
ssbo_index);
@@ -526,7 +524,7 @@ void BufferCache<P>::BindComputeStorageBuffer(size_t ssbo_index, u32 cbuf_index,
const auto& cbufs = launch_desc.const_buffer_config;
const GPUVAddr ssbo_addr = cbufs[cbuf_index].Address() + cbuf_offset;
channel_state->compute_storage_buffers[ssbo_index] =
StorageBufferBinding(ssbo_addr, cbuf_index, is_written, descriptor_count);
StorageBufferBinding(ssbo_addr, cbuf_index, is_written);
}
template <class P>
@@ -1002,52 +1000,27 @@ void BufferCache<P>::BindHostGraphicsUniformBuffer(size_t stage, u32 index, u32
template <class P>
void BufferCache<P>::BindHostGraphicsStorageBuffers(size_t stage) {
boost::container::small_vector<u32, NUM_STORAGE_BUFFERS> segment_sizes;
bool uses_mapping{};
ForEachEnabledBit(channel_state->enabled_storage_buffers[stage], [&](u32 index) {
const StorageBufferBindingInfo& binding = channel_state->storage_buffers[stage][index];
uses_mapping |= binding.descriptor_count > 1;
for (u32 segment = 0; segment < binding.descriptor_count; ++segment) {
segment_sizes.push_back(segment < binding.segments.size() ? binding.segments[segment].size : 0);
}
});
if (uses_mapping) {
const u32 mapping_size = static_cast<u32>(segment_sizes.size() * sizeof(u32));
if constexpr (!IS_OPENGL) {
const std::span<u8> mapped = runtime.BindMappedStorageBuffer(mapping_size);
std::memcpy(mapped.data(), segment_sizes.data(), mapping_size);
}
}
u32 binding_index = 0;
ForEachEnabledBit(channel_state->enabled_storage_buffers[stage], [&](u32 index) {
const StorageBufferBindingInfo& storage = channel_state->storage_buffers[stage][index];
const Binding& binding = channel_state->storage_buffers[stage][index];
Buffer& buffer = slot_buffers[binding.buffer_id];
TouchBuffer(buffer, binding.buffer_id);
const u32 size = binding.size;
SynchronizeBuffer(buffer, binding.device_addr, size);
const u32 offset = buffer.Offset(binding.device_addr);
buffer.MarkUsage(offset, size);
const bool is_written = ((channel_state->written_storage_buffers[stage] >> index) & 1) != 0;
for (u32 segment = 0; segment < storage.descriptor_count; ++segment) {
Buffer* buffer = &slot_buffers[NULL_BUFFER_ID];
u32 offset{};
u32 size{IS_OPENGL ? 0U : static_cast<u32>(sizeof(u32))};
const bool is_actual_segment = segment < storage.segments.size();
// shall be safe enough if the segment is not actual, use the last available segment or nullptr if none exist.
const Binding* binding = is_actual_segment ? &storage.segments[segment] : storage.segments.empty() ? nullptr : &storage.segments.back();
if (binding) {
buffer = &slot_buffers[binding->buffer_id];
size = binding->size;
offset = buffer->Offset(binding->device_addr);
if (is_actual_segment) {
TouchBuffer(*buffer, binding->buffer_id);
SynchronizeBuffer(*buffer, binding->device_addr, size);
buffer->MarkUsage(offset, size);
if (is_written) {
MarkWrittenBuffer(binding->buffer_id, binding->device_addr, size);
}
}
}
if constexpr (NEEDS_BIND_STORAGE_INDEX) {
runtime.BindStorageBuffer(stage, binding_index++, *buffer, offset, size, is_written);
} else {
runtime.BindStorageBuffer(*buffer, offset, size, is_written);
}
if (is_written) {
MarkWrittenBuffer(binding.buffer_id, binding.device_addr, size);
}
if constexpr (NEEDS_BIND_STORAGE_INDEX) {
runtime.BindStorageBuffer(stage, binding_index, buffer, offset, size, is_written);
++binding_index;
} else {
runtime.BindStorageBuffer(buffer, offset, size, is_written);
}
});
}
@@ -1163,53 +1136,28 @@ void BufferCache<P>::BindHostComputeUniformBuffers() {
template <class P>
void BufferCache<P>::BindHostComputeStorageBuffers() {
boost::container::small_vector<u32, NUM_STORAGE_BUFFERS> segment_sizes;
bool uses_mapping{};
ForEachEnabledBit(channel_state->enabled_compute_storage_buffers, [&](u32 index) {
const StorageBufferBindingInfo& binding = channel_state->compute_storage_buffers[index];
uses_mapping |= binding.descriptor_count > 1;
for (u32 segment = 0; segment < binding.descriptor_count; ++segment) {
segment_sizes.push_back(segment < binding.segments.size() ? binding.segments[segment].size : 0);
}
});
if (uses_mapping) {
const u32 mapping_size = static_cast<u32>(segment_sizes.size() * sizeof(u32));
if constexpr (!IS_OPENGL) {
const std::span<u8> mapped = runtime.BindMappedStorageBuffer(mapping_size);
std::memcpy(mapped.data(), segment_sizes.data(), mapping_size);
}
}
u32 binding_index = 0;
ForEachEnabledBit(channel_state->enabled_compute_storage_buffers, [&](u32 index) {
const StorageBufferBindingInfo& storage = channel_state->compute_storage_buffers[index];
const Binding& binding = channel_state->compute_storage_buffers[index];
Buffer& buffer = slot_buffers[binding.buffer_id];
TouchBuffer(buffer, binding.buffer_id);
const u32 size = binding.size;
SynchronizeBuffer(buffer, binding.device_addr, size);
const u32 offset = buffer.Offset(binding.device_addr);
buffer.MarkUsage(offset, size);
const bool is_written =
((channel_state->written_compute_storage_buffers >> index) & 1) != 0;
for (u32 segment = 0; segment < storage.descriptor_count; ++segment) {
Buffer* buffer = &slot_buffers[NULL_BUFFER_ID];
u32 offset{};
u32 size{IS_OPENGL ? 0U : static_cast<u32>(sizeof(u32))};
const bool is_actual_segment = segment < storage.segments.size();
//same fallback logic
const Binding* binding = is_actual_segment ? &storage.segments[segment] : storage.segments.empty() ? nullptr : &storage.segments.back();
if (binding) {
buffer = &slot_buffers[binding->buffer_id];
size = binding->size;
offset = buffer->Offset(binding->device_addr);
if (is_actual_segment) {
TouchBuffer(*buffer, binding->buffer_id);
SynchronizeBuffer(*buffer, binding->device_addr, size);
buffer->MarkUsage(offset, size);
if (is_written) {
MarkWrittenBuffer(binding->buffer_id, binding->device_addr, size);
}
}
}
if constexpr (NEEDS_BIND_STORAGE_INDEX) {
runtime.BindComputeStorageBuffer(binding_index++, *buffer, offset, size, is_written);
} else {
runtime.BindStorageBuffer(*buffer, offset, size, is_written);
}
if (is_written) {
MarkWrittenBuffer(binding.buffer_id, binding.device_addr, size);
}
if constexpr (NEEDS_BIND_STORAGE_INDEX) {
runtime.BindComputeStorageBuffer(binding_index, buffer, offset, size, is_written);
++binding_index;
} else {
runtime.BindStorageBuffer(buffer, offset, size, is_written);
}
});
}
@@ -1402,7 +1350,10 @@ void BufferCache<P>::UpdateUniformBuffers(size_t stage) {
template <class P>
void BufferCache<P>::UpdateStorageBuffers(size_t stage) {
ForEachEnabledBit(channel_state->enabled_storage_buffers[stage], [&](u32 index) {
UpdateStorageBuffer(channel_state->storage_buffers[stage][index]);
// Resolve buffer
Binding& binding = channel_state->storage_buffers[stage][index];
const BufferId buffer_id = FindBuffer(binding.device_addr, binding.size);
binding.buffer_id = buffer_id;
});
}
@@ -1463,54 +1414,12 @@ void BufferCache<P>::UpdateComputeUniformBuffers() {
template <class P>
void BufferCache<P>::UpdateComputeStorageBuffers() {
ForEachEnabledBit(channel_state->enabled_compute_storage_buffers, [&](u32 index) {
UpdateStorageBuffer(channel_state->compute_storage_buffers[index]);
// Resolve buffer
Binding& binding = channel_state->compute_storage_buffers[index];
binding.buffer_id = FindBuffer(binding.device_addr, binding.size);
});
}
template <class P>
void BufferCache<P>::UpdateStorageBuffer(StorageBufferBindingInfo& binding) {
binding.segments.clear();
if (binding.gpu_addr == 0 || binding.size == 0) { return;}
if (binding.descriptor_count == 1) {
// for safety gotta preserve the legacy path on possible hosts without storage-buffer descriptor indexing.
const std::optional<DAddr> device_addr = gpu_memory->GpuToCpuAddress(binding.gpu_addr);
if (device_addr) {
binding.segments.push_back(Binding{
.device_addr = *device_addr,
.size = binding.size,
.buffer_id = FindBuffer(*device_addr, binding.size),
});
}
return;
}
const auto ranges = gpu_memory->GetSubmappedRange(binding.gpu_addr, binding.size);
const size_t mapped_size =
std::accumulate(ranges.begin(), ranges.end(), size_t{}, [](size_t total, const auto& range) { return total + range.second; });
if (mapped_size != binding.size) {
LOG_ERROR(HW_GPU, "Storage buffer range {:#x}+{:#x} is not fully mapped", binding.gpu_addr, binding.size);
return;
}
if (ranges.size() > binding.descriptor_count) {
LOG_ERROR(HW_GPU, "Storage buffer range {:#x}+{:#x} has {} physical segments, exceeding host capacity {}",
binding.gpu_addr, binding.size, ranges.size(), binding.descriptor_count);
return;
}
for (const auto& [gpu_addr, size] : ranges) {
const std::optional<DAddr> device_addr = gpu_memory->GpuToCpuAddress(gpu_addr);
if (!device_addr || size > (std::numeric_limits<u32>::max)()) {
binding.segments.clear();
return;
}
const u32 segment_size = static_cast<u32>(size);
binding.segments.push_back(Binding{
.device_addr = *device_addr,
.size = segment_size,
.buffer_id = FindBuffer(*device_addr, segment_size),
});
}
}
template <class P>
void BufferCache<P>::UpdateComputeTextureBuffers() {
ForEachEnabledBit(channel_state->enabled_compute_texture_buffers, [&](u32 index) {
@@ -1932,11 +1841,6 @@ void BufferCache<P>::DeleteBuffer(BufferId buffer_id, bool do_not_mark) {
const auto replace = [scalar_replace](std::span<Binding> bindings) {
std::ranges::for_each(bindings, scalar_replace);
};
const auto storage_replace = [scalar_replace](std::span<StorageBufferBindingInfo> bindings) {
for (StorageBufferBindingInfo& binding : bindings) {
std::ranges::for_each(binding.segments, scalar_replace);
}
};
if (channel_state->index_buffer.buffer_id == buffer_id) {
channel_state->index_buffer.buffer_id = BufferId{};
@@ -1952,10 +1856,10 @@ void BufferCache<P>::DeleteBuffer(BufferId buffer_id, bool do_not_mark) {
}
}
std::ranges::for_each(channel_state->uniform_buffers, replace);
std::ranges::for_each(channel_state->storage_buffers, storage_replace);
std::ranges::for_each(channel_state->storage_buffers, replace);
replace(channel_state->transform_feedback_buffers);
replace(channel_state->compute_uniform_buffers);
storage_replace(channel_state->compute_storage_buffers);
replace(channel_state->compute_storage_buffers);
// Mark the whole buffer as CPU written to stop tracking CPU writes
if (!do_not_mark) {
@@ -1992,14 +1896,12 @@ void BufferCache<P>::DeleteBuffer(BufferId buffer_id, bool do_not_mark) {
}
template <class P>
StorageBufferBindingInfo BufferCache<P>::StorageBufferBinding(GPUVAddr ssbo_addr, u32 cbuf_index,
bool is_written, u32 descriptor_count) const {
// time to get rid of these null bindings
ASSERT(descriptor_count > 0); // shant happen
Binding BufferCache<P>::StorageBufferBinding(GPUVAddr ssbo_addr, u32 cbuf_index,
bool is_written) const {
const GPUVAddr gpu_addr = gpu_memory->Read<u64>(ssbo_addr);
if (gpu_addr == 0) {
return {.descriptor_count = descriptor_count};
return NULL_BINDING;
}
const auto size = [&]() {
@@ -2021,17 +1923,21 @@ StorageBufferBindingInfo BufferCache<P>::StorageBufferBinding(GPUVAddr ssbo_addr
const GPUVAddr aligned_gpu_addr = Common::AlignDown(gpu_addr, alignment);
const u32 aligned_size = static_cast<u32>(gpu_addr - aligned_gpu_addr) + size;
if (!gpu_memory->GpuToCpuAddress(aligned_gpu_addr) || size == 0) {
const std::optional<DAddr> aligned_device_addr = gpu_memory->GpuToCpuAddress(aligned_gpu_addr);
if (!aligned_device_addr || size == 0) {
LOG_DEBUG(HW_GPU, "Failed to find storage buffer for cbuf index {}", cbuf_index);
return {.descriptor_count = descriptor_count};
return NULL_BINDING;
}
const std::optional<DAddr> device_addr = gpu_memory->GpuToCpuAddress(gpu_addr);
ASSERT_MSG(device_addr, "Unaligned storage buffer address not found for cbuf index {}",
cbuf_index);
// The end address used for size calculation does not need to be aligned
const GPUVAddr gpu_end = Common::AlignUp(gpu_addr + size, Core::DEVICE_PAGESIZE);
const DAddr cpu_end = Common::AlignUp(*device_addr + size, Core::DEVICE_PAGESIZE);
const StorageBufferBindingInfo binding{
.gpu_addr = aligned_gpu_addr,
.size = is_written ? aligned_size : static_cast<u32>(gpu_end - aligned_gpu_addr),
.descriptor_count = descriptor_count,
const Binding binding{
.device_addr = *aligned_device_addr,
.size = is_written ? aligned_size : static_cast<u32>(cpu_end - *aligned_device_addr),
.buffer_id = BufferId{},
};
return binding;
}
@@ -89,15 +89,6 @@ struct TextureBufferBinding : Binding {
PixelFormat format;
};
struct StorageBufferBindingInfo {
// another good one: guest SSBO is a virtual interval and may span discontiguous device-memory ranges.
// exact case of missing character frames (high sample lane)
GPUVAddr gpu_addr{};
u32 size{};
u32 descriptor_count{1};
boost::container::small_vector<Binding, 1> segments;
};
static constexpr Binding NULL_BINDING{
.device_addr = 0,
.size = 0,
@@ -124,15 +115,14 @@ public:
Binding index_buffer;
std::array<Binding, NUM_VERTEX_BUFFERS> vertex_buffers;
std::array<std::array<Binding, NUM_GRAPHICS_UNIFORM_BUFFERS>, NUM_STAGES> uniform_buffers;
std::array<std::array<StorageBufferBindingInfo, NUM_STORAGE_BUFFERS>, NUM_STAGES>
storage_buffers;
std::array<std::array<Binding, NUM_STORAGE_BUFFERS>, NUM_STAGES> storage_buffers;
std::array<std::array<TextureBufferBinding, NUM_TEXTURE_BUFFERS>, NUM_STAGES> texture_buffers;
std::array<Binding, NUM_TRANSFORM_FEEDBACK_BUFFERS> transform_feedback_buffers;
Binding count_buffer_binding;
Binding indirect_buffer_binding;
std::array<Binding, NUM_COMPUTE_UNIFORM_BUFFERS> compute_uniform_buffers;
std::array<StorageBufferBindingInfo, NUM_STORAGE_BUFFERS> compute_storage_buffers;
std::array<Binding, NUM_STORAGE_BUFFERS> compute_storage_buffers;
std::array<TextureBufferBinding, NUM_TEXTURE_BUFFERS> compute_texture_buffers;
std::array<u32, NUM_STAGES> enabled_uniform_buffer_masks{};
@@ -259,7 +249,7 @@ public:
void UnbindGraphicsStorageBuffers(size_t stage);
bool BindGraphicsStorageBuffer(size_t stage, size_t ssbo_index, u32 cbuf_index, u32 cbuf_offset,
bool is_written, u32 descriptor_count = 1);
bool is_written);
void UnbindGraphicsTextureBuffers(size_t stage);
@@ -269,7 +259,7 @@ public:
void UnbindComputeStorageBuffers();
void BindComputeStorageBuffer(size_t ssbo_index, u32 cbuf_index, u32 cbuf_offset,
bool is_written, u32 descriptor_count = 1);
bool is_written);
void UnbindComputeTextureBuffers();
@@ -410,8 +400,6 @@ private:
void UpdateStorageBuffers(size_t stage);
void UpdateStorageBuffer(StorageBufferBindingInfo& binding);
void UpdateTextureBuffers(size_t stage);
void UpdateTransformFeedbackBuffers();
@@ -461,9 +449,8 @@ private:
void DeleteBuffer(BufferId buffer_id, bool do_not_mark = false);
[[nodiscard]] StorageBufferBindingInfo StorageBufferBinding(GPUVAddr ssbo_addr, u32 cbuf_index,
bool is_written,
u32 descriptor_count) const;
[[nodiscard]] Binding StorageBufferBinding(GPUVAddr ssbo_addr, u32 cbuf_index,
bool is_written) const;
[[nodiscard]] TextureBufferBinding GetTextureBufferBinding(GPUVAddr gpu_addr, u32 size,
PixelFormat format);
@@ -43,7 +43,6 @@ using Shader::Backend::GLASM::EmitGLASM;
using Shader::Backend::GLSL::EmitGLSL;
using Shader::Backend::SPIRV::EmitSPIRV;
using Shader::Maxwell::ConvertLegacyToGeneric;
using Shader::Maxwell::GenerateGeometryPassthrough;
using Shader::Maxwell::MergeDualVertexPrograms;
using Shader::Maxwell::TranslateProgram;
using VideoCommon::ComputeEnvironment;
@@ -317,7 +316,7 @@ void ShaderCache::LoadDiskResources(u64 title_id, std::stop_token stop_loading,
ComputePipelineKey key;
file.read(reinterpret_cast<char*>(&key), sizeof(key));
queue_work([this, key, env_ = std::move(env), &state, &callback](Context* ctx) mutable {
ctx->pools.ReleaseContents();
ctx->pools.clear();
auto pipeline{CreateComputePipeline(ctx->pools, key, env_, true)};
std::scoped_lock lock{state.mutex};
if (pipeline) {
@@ -338,7 +337,7 @@ void ShaderCache::LoadDiskResources(u64 title_id, std::stop_token stop_loading,
for (auto& env : envs_) {
env_ptrs.push_back(&env);
}
ctx->pools.ReleaseContents();
ctx->pools.clear();
auto pipeline{CreateGraphicsPipeline(ctx->pools, key, MakeSpan(env_ptrs), false, true)};
std::scoped_lock lock{state.mutex};
if (pipeline) {
@@ -448,9 +447,8 @@ std::unique_ptr<GraphicsPipeline> ShaderCache::CreateGraphicsPipeline() {
GraphicsEnvironments environments;
GetGraphicsEnvironments(environments, graphics_key.unique_hashes);
main_pools.ReleaseContents();
auto pipeline{CreateGraphicsPipeline(main_pools, graphics_key, environments.Span(),
use_asynchronous_shaders)};
main_pools.clear();
auto pipeline{CreateGraphicsPipeline(main_pools, graphics_key, environments.Span(), use_asynchronous_shaders)};
if (!pipeline || shader_cache_filename.empty()) {
return pipeline;
}
@@ -464,10 +462,7 @@ std::unique_ptr<GraphicsPipeline> ShaderCache::CreateGraphicsPipeline() {
return pipeline;
}
std::unique_ptr<GraphicsPipeline> ShaderCache::CreateGraphicsPipeline(
ShaderContext::ShaderPools& pools, const GraphicsPipelineKey& key,
std::span<Shader::Environment* const> envs, bool use_shader_workers,
bool force_context_flush) try {
std::unique_ptr<GraphicsPipeline> ShaderCache::CreateGraphicsPipeline(Shader::Maxwell::ShaderPools& pools, const GraphicsPipelineKey& key, std::span<Shader::Environment* const> envs, bool use_shader_workers, bool force_context_flush) try {
auto hash = key.Hash();
LOG_INFO(Render_OpenGL, "{:#016x}", hash);
size_t env_index{};
@@ -480,12 +475,10 @@ std::unique_ptr<GraphicsPipeline> ShaderCache::CreateGraphicsPipeline(
Shader::IR::Program* layer_source_program{};
for (size_t index = 0; index < Maxwell::MaxShaderProgram; ++index) {
const bool is_emulated_stage = layer_source_program != nullptr
&& index == u32(Maxwell::ShaderType::Geometry);
const bool is_emulated_stage = layer_source_program != nullptr && index == u32(Maxwell::ShaderType::Geometry);
if (key.unique_hashes[index] == 0 && is_emulated_stage) {
auto topology = MaxwellToOutputTopology(key.gs_input_topology);
programs[index] = GenerateGeometryPassthrough(pools.inst, pools.block, host_info,
*layer_source_program, topology);
programs[index] = Shader::Maxwell::GenerateGeometryPassthrough(pools, host_info, *layer_source_program, topology);
continue;
}
if (key.unique_hashes[index] == 0) {
@@ -503,13 +496,13 @@ std::unique_ptr<GraphicsPipeline> ShaderCache::CreateGraphicsPipeline(
if (!uses_vertex_a || index != 1) {
// Normal path
programs[index] = TranslateProgram(pools.inst, pools.block, env, cfg, host_info);
programs[index] = TranslateProgram(pools, env, cfg, host_info);
total_storage_buffers += Shader::NumDescriptors(programs[index].info.storage_buffers_descriptors);
} else {
// VertexB path when VertexA is present.
auto& program_va{programs[0]};
auto program_vb{TranslateProgram(pools.inst, pools.block, env, cfg, host_info)};
auto program_vb{TranslateProgram(pools, env, cfg, host_info)};
total_storage_buffers += Shader::NumDescriptors(program_vb.info.storage_buffers_descriptors);
programs[index] = MergeDualVertexPrograms(program_va, program_vb, env);
}
@@ -576,7 +569,7 @@ std::unique_ptr<ComputePipeline> ShaderCache::CreateComputePipeline(
ComputeEnvironment env{*kepler_compute, *gpu_memory, program_base, qmd.program_start};
env.SetCachedSize(shader->size_bytes);
main_pools.ReleaseContents();
main_pools.clear();
auto pipeline{CreateComputePipeline(main_pools, key, env)};
if (!pipeline || shader_cache_filename.empty()) {
return pipeline;
@@ -586,9 +579,7 @@ std::unique_ptr<ComputePipeline> ShaderCache::CreateComputePipeline(
return pipeline;
}
std::unique_ptr<ComputePipeline> ShaderCache::CreateComputePipeline(
ShaderContext::ShaderPools& pools, const ComputePipelineKey& key, Shader::Environment& env,
bool force_context_flush) try {
std::unique_ptr<ComputePipeline> ShaderCache::CreateComputePipeline(Shader::Maxwell::ShaderPools& pools, const ComputePipelineKey& key, Shader::Environment& env, bool force_context_flush) try {
auto hash = key.Hash();
LOG_INFO(Render_OpenGL, "{:#016x}", hash);
@@ -598,7 +589,7 @@ std::unique_ptr<ComputePipeline> ShaderCache::CreateComputePipeline(
env.Dump(hash, key.unique_hash);
}
auto program{TranslateProgram(pools.inst, pools.block, env, cfg, host_info)};
auto program{TranslateProgram(pools, env, cfg, host_info)};
const u32 num_storage_buffers{Shader::NumDescriptors(program.info.storage_buffers_descriptors)};
Shader::RuntimeInfo info;
info.glasm_use_storage_buffers = num_storage_buffers <= device.GetMaxGLASMStorageBufferBlocks();
@@ -12,6 +12,7 @@
#include "common/common_types.h"
#include "common/thread_worker.h"
#include "shader_recompiler/host_translate_info.h"
#include "shader_recompiler/shader_pool.h"
#include "shader_recompiler/profile.h"
#include "video_core/renderer_opengl/gl_compute_pipeline.h"
#include "video_core/renderer_opengl/gl_graphics_pipeline.h"
@@ -51,20 +52,9 @@ private:
[[nodiscard]] GraphicsPipeline* BuiltPipeline(GraphicsPipeline* pipeline) const noexcept;
std::unique_ptr<GraphicsPipeline> CreateGraphicsPipeline();
std::unique_ptr<GraphicsPipeline> CreateGraphicsPipeline(
ShaderContext::ShaderPools& pools, const GraphicsPipelineKey& key,
std::span<Shader::Environment* const> envs, bool use_shader_workers,
bool force_context_flush = false);
std::unique_ptr<ComputePipeline> CreateComputePipeline(const ComputePipelineKey& key,
const VideoCommon::ShaderInfo* shader);
std::unique_ptr<ComputePipeline> CreateComputePipeline(ShaderContext::ShaderPools& pools,
const ComputePipelineKey& key,
Shader::Environment& env,
bool force_context_flush = false);
std::unique_ptr<GraphicsPipeline> CreateGraphicsPipeline(Shader::Maxwell::ShaderPools& pools, const GraphicsPipelineKey& key, std::span<Shader::Environment* const> envs, bool use_shader_workers, bool force_context_flush = false);
std::unique_ptr<ComputePipeline> CreateComputePipeline(const ComputePipelineKey& key, const VideoCommon::ShaderInfo* shader);
std::unique_ptr<ComputePipeline> CreateComputePipeline(Shader::Maxwell::ShaderPools& pools, const ComputePipelineKey& key, Shader::Environment& env, bool force_context_flush = false);
std::unique_ptr<ShaderWorker> CreateWorkers() const;
Core::Frontend::EmuWindow& emu_window;
@@ -80,7 +70,7 @@ private:
GraphicsPipelineKey graphics_key{};
GraphicsPipeline* current_pipeline{};
ShaderContext::ShaderPools main_pools;
Shader::Maxwell::ShaderPools main_pools;
::Common::unordered_map<GraphicsPipelineKey, std::unique_ptr<GraphicsPipeline>> graphics_cache;
::Common::unordered_map<ComputePipelineKey, std::unique_ptr<ComputePipeline>> compute_cache;
@@ -1,33 +1,26 @@
// SPDX-FileCopyrightText: Copyright 2026 Eden Emulator Project
// SPDX-License-Identifier: GPL-3.0-or-later
// SPDX-FileCopyrightText: Copyright 2021 yuzu Emulator Project
// SPDX-License-Identifier: GPL-2.0-or-later
#pragma once
#include <boost/container/stable_vector.hpp>
#include "core/frontend/emu_window.h"
#include "core/frontend/graphics_context.h"
#include "shader_recompiler/frontend/ir/basic_block.h"
#include "shader_recompiler/frontend/maxwell/control_flow.h"
#include "shader_recompiler/frontend/maxwell/structured_control_flow.h"
#include "shader_recompiler/frontend/maxwell/translate_program.h"
namespace OpenGL::ShaderContext {
struct ShaderPools {
void ReleaseContents() {
flow_block.ReleaseContents();
block.ReleaseContents();
inst.ReleaseContents();
}
Shader::ObjectPool<Shader::IR::Inst> inst{8192};
Shader::ObjectPool<Shader::IR::Block> block{32};
Shader::ObjectPool<Shader::Maxwell::Flow::Block> flow_block{32};
};
struct Context {
explicit Context(Core::Frontend::EmuWindow& emu_window)
: gl_context{emu_window.CreateSharedContext()}, scoped{*gl_context} {}
explicit Context(Core::Frontend::EmuWindow& emu_window) : gl_context{emu_window.CreateSharedContext()}, scoped{*gl_context} {}
std::unique_ptr<Core::Frontend::GraphicsContext> gl_context;
Core::Frontend::GraphicsContext::Scoped scoped;
ShaderPools pools;
Shader::Maxwell::ShaderPools pools;
};
} // namespace OpenGL::ShaderContext
@@ -6,7 +6,6 @@
#pragma once
#include <array>
#include <cstddef>
#include <optional>
@@ -137,7 +136,6 @@ inline void WriteDescriptorBuffer(const Device& device, const DescriptorBufferLa
[[nodiscard]] inline u32 NumDescriptorEntries(const Shader::Info& info) {
return Shader::NumDescriptors(info.constant_buffer_descriptors) +
static_cast<u32>(Shader::UsesStorageBufferMappings(info)) +
Shader::NumDescriptors(info.storage_buffers_descriptors) +
Shader::NumDescriptors(info.texture_buffer_descriptors) +
Shader::NumDescriptors(info.image_buffer_descriptors) +
@@ -270,12 +268,6 @@ public:
is_compute |= (stage & VK_SHADER_STAGE_COMPUTE_BIT) != 0;
Add(VK_DESCRIPTOR_TYPE_UNIFORM_BUFFER, stage, info.constant_buffer_descriptors);
// for extra implicit storage-buffer binding required by mapped-storage-buffer support
if (Shader::UsesStorageBufferMappings(info)) {
struct Descriptor { u32 count; };
const std::array descriptors{Descriptor{1}};
Add(VK_DESCRIPTOR_TYPE_STORAGE_BUFFER, stage, descriptors);
}
Add(VK_DESCRIPTOR_TYPE_STORAGE_BUFFER, stage, info.storage_buffers_descriptors);
Add(VK_DESCRIPTOR_TYPE_UNIFORM_TEXEL_BUFFER, stage, info.texture_buffer_descriptors);
Add(VK_DESCRIPTOR_TYPE_STORAGE_TEXEL_BUFFER, stage, info.image_buffer_descriptors);
@@ -695,7 +695,7 @@ vk::Buffer BufferCacheRuntime::CreateNullBuffer() {
.flags = 0,
.size = 4,
.usage = VK_BUFFER_USAGE_VERTEX_BUFFER_BIT | VK_BUFFER_USAGE_INDEX_BUFFER_BIT |
VK_BUFFER_USAGE_TRANSFER_DST_BIT | VK_BUFFER_USAGE_INDIRECT_BUFFER_BIT | VK_BUFFER_USAGE_STORAGE_BUFFER_BIT,
VK_BUFFER_USAGE_TRANSFER_DST_BIT | VK_BUFFER_USAGE_INDIRECT_BUFFER_BIT,
.sharingMode = VK_SHARING_MODE_EXCLUSIVE,
.queueFamilyIndexCount = 0,
.pQueueFamilyIndices = nullptr,
@@ -155,13 +155,6 @@ public:
return ref.mapped_span;
}
// compute/graphics new binding
std::span<u8> BindMappedStorageBuffer(u32 size) {
const StagingBufferRef ref = staging_pool.Request(size, MemoryUsage::Upload);
guest_descriptor_queue.AddBuffer(ref.buffer, ref.device_address, static_cast<u32>(ref.offset), size);
return ref.mapped_span;
}
void BindUniformBuffer(const Buffer& buffer, u32 offset, u32 size) {
BindBuffer(buffer, offset, size);
}
@@ -157,7 +157,9 @@ bool ComputePipeline::Configure(Tegra::Engines::KeplerCompute& kepler_compute,
buffer_cache.UnbindComputeStorageBuffers();
size_t ssbo_index{};
for (const auto& desc : info.storage_buffers_descriptors) {
buffer_cache.BindComputeStorageBuffer(ssbo_index, desc.cbuf_index, desc.cbuf_offset, desc.is_written, desc.count);
ASSERT(desc.count == 1);
buffer_cache.BindComputeStorageBuffer(ssbo_index, desc.cbuf_index, desc.cbuf_offset,
desc.is_written);
++ssbo_index;
}
@@ -47,7 +47,7 @@ static DescriptorBankInfo MakeBankInfo(std::span<const Shader::Info> infos) {
DescriptorBankInfo bank;
for (const Shader::Info& info : infos) {
bank.uniform_buffers += Accumulate(info.constant_buffer_descriptors);
bank.storage_buffers += Accumulate(info.storage_buffers_descriptors) + static_cast<u32>(Shader::UsesStorageBufferMappings(info));
bank.storage_buffers += Accumulate(info.storage_buffers_descriptors);
bank.texture_buffers += Accumulate(info.texture_buffer_descriptors);
bank.image_buffers += Accumulate(info.image_buffer_descriptors);
bank.textures += Accumulate(info.texture_descriptors);
@@ -367,8 +367,9 @@ bool GraphicsPipeline::ConfigureImpl(bool is_indexed) {
if constexpr (Spec::has_storage_buffers) {
size_t ssbo_index{};
for (const auto& desc : info.storage_buffers_descriptors) {
ASSERT(desc.count == 1);
buffer_cache.BindGraphicsStorageBuffer(stage, ssbo_index, desc.cbuf_index,
desc.cbuf_offset, desc.is_written, desc.count);
desc.cbuf_offset, desc.is_written);
++ssbo_index;
}
}
@@ -52,22 +52,8 @@
namespace Vulkan {
namespace {
using Shader::Backend::SPIRV::EmitSPIRV;
using Shader::Maxwell::ConvertLegacyToGeneric;
using Shader::Maxwell::GenerateGeometryPassthrough;
using Shader::Maxwell::MergeDualVertexPrograms;
using Shader::Maxwell::TranslateProgram;
using VideoCommon::ComputeEnvironment;
using VideoCommon::FileEnvironment;
using VideoCommon::GenericEnvironment;
using VideoCommon::GraphicsEnvironment;
// SPIR-V descriptor arrays require a fixed pipeline-layout count.
// Exploration ceiling; buffer-cache telemetry records the actual physical-range demand.
// Keep this modest because every mapped SSBO binds the full fixed array on each update.
constexpr u32 MAX_MAPPED_STORAGE_BUFFER_DESCRIPTORS = 32;
constexpr u32 CACHE_VERSION = 19;
constexpr u32 CACHE_VERSION = 18;
constexpr size_t VULKAN_CACHE_FLUSH_PIPELINES = 128;
constexpr size_t VULKAN_CACHE_FLUSH_MIN_SECONDS = 30;
constexpr std::array<char, 8> VULKAN_CACHE_MAGIC_NUMBER{'y', 'u', 'z', 'u', 'v', 'k', 'c', 'h'};
@@ -437,8 +423,6 @@ PipelineCache::PipelineCache(Tegra::MaxwellDeviceMemoryManager& device_memory_,
device.IsUniformTexelBufferArrayNonUniformIndexingSupported(),
.support_storage_texel_buffer_array_nonuniform_indexing =
device.IsStorageTexelBufferArrayNonUniformIndexingSupported(),
.support_storage_buffer_array_nonuniform_indexing =
device.IsStorageBufferArrayNonUniformIndexingSupported(),
.warp_size_potentially_larger_than_guest = device.IsWarpSizePotentiallyBiggerThanGuest(),
@@ -467,7 +451,6 @@ PipelineCache::PipelineCache(Tegra::MaxwellDeviceMemoryManager& device_memory_,
host_info = Shader::HostTranslateInfo{
.min_ssbo_alignment = device.GetStorageBufferAlignment(),
.max_per_stage_descriptor_sampled_images = device.GetMaxPerStageDescriptorSampledImages(),
.max_per_stage_descriptor_storage_buffers = device.GetMaxPerStageDescriptorStorageBuffers(),
.max_per_stage_resources = device.GetMaxPerStageResources(),
.max_descriptor_set_samplers = device.GetMaxDescriptorSetSamplers(),
.max_descriptor_set_uniform_buffers = device.GetMaxDescriptorSetUniformBuffers(),
@@ -487,20 +470,6 @@ PipelineCache::PipelineCache(Tegra::MaxwellDeviceMemoryManager& device_memory_,
.support_viewport_index_layer = device.IsExtShaderViewportIndexLayerSupported(),
.support_geometry_shader_passthrough = device.IsNvGeometryShaderPassthroughSupported(),
.support_conditional_barrier = device.SupportsConditionalBarriers(),
.storage_buffer_segment_count = [&] {
if (!device.IsStorageBufferArrayNonUniformIndexingSupported()) {
return 1U;
}
constexpr u32 MaxGraphicsStages = static_cast<u32>(Maxwell::MaxShaderStage);
const auto reserve = [](u32 limit, u32 count) {
return limit > count ? limit - count : 0U;
};
const u32 per_stage = reserve(device.GetMaxPerStageDescriptorStorageBuffers(), 1) / static_cast<u32>(Shader::Info::MAX_SSBOS);
const u32 per_set = reserve(device.GetMaxDescriptorSetStorageBuffers(), MaxGraphicsStages) / (static_cast<u32>(Shader::Info::MAX_SSBOS) * MaxGraphicsStages);
const u32 resources = reserve(device.GetMaxPerStageResources(), 1) / (static_cast<u32>(Shader::Info::MAX_SSBOS) * 2);
// ensure at least one storage buffer segment is available per stage. max is still arbitrary
return (std::max)(1U, (std::min)({MAX_MAPPED_STORAGE_BUFFER_DESCRIPTORS, per_stage, per_set, resources}));
}(),
};
host_info.ApplyDescriptorLimitPolicy();
@@ -649,12 +618,12 @@ void PipelineCache::LoadDiskResources(u64 title_id, std::stop_token stop_loading
if (device.IsKhrPipelineExecutablePropertiesEnabled()) {
state.statistics = std::make_unique<PipelineStatistics>(device);
}
const auto load_compute{[&](std::ifstream& file, FileEnvironment env) {
const auto load_compute{[&](std::ifstream& file, VideoCommon::FileEnvironment env) {
ComputePipelineCacheKey key;
file.read(reinterpret_cast<char*>(&key), sizeof(key));
workers.QueueWork([this, key, env_ = std::move(env), &state, &callback]() mutable {
ShaderPools pools;
Shader::Maxwell::ShaderPools pools;
auto pipeline{CreateComputePipeline(pools, key, env_, state.statistics.get(), false)};
std::scoped_lock lock{state.mutex};
if (pipeline) {
@@ -667,7 +636,7 @@ void PipelineCache::LoadDiskResources(u64 title_id, std::stop_token stop_loading
});
++state.total;
}};
const auto load_graphics{[&](std::ifstream& file, std::vector<FileEnvironment> envs) {
const auto load_graphics{[&](std::ifstream& file, std::vector<VideoCommon::FileEnvironment> envs) {
GraphicsPipelineCacheKey key;
file.read(reinterpret_cast<char*>(&key), sizeof(key));
@@ -700,7 +669,7 @@ void PipelineCache::LoadDiskResources(u64 title_id, std::stop_token stop_loading
}
workers.QueueWork([this, key, envs_ = std::move(envs), &state, &callback]() mutable {
ShaderPools pools;
Shader::Maxwell::ShaderPools pools;
boost::container::static_vector<Shader::Environment*, 5> env_ptrs;
for (auto& env : envs_) {
env_ptrs.push_back(&env);
@@ -807,10 +776,7 @@ GraphicsPipeline* PipelineCache::BuiltPipeline(GraphicsPipeline* pipeline) const
return nullptr;
}
std::unique_ptr<GraphicsPipeline> PipelineCache::CreateGraphicsPipeline(
ShaderPools& pools, const GraphicsPipelineCacheKey& key,
std::span<Shader::Environment* const> envs, PipelineStatistics* statistics,
bool build_in_parallel) try {
std::unique_ptr<GraphicsPipeline> PipelineCache::CreateGraphicsPipeline(Shader::Maxwell::ShaderPools& pools, const GraphicsPipelineCacheKey& key, std::span<Shader::Environment* const> envs, PipelineStatistics* statistics, bool build_in_parallel) try {
auto hash = key.Hash();
LOG_DEBUG(Render_Vulkan, "{:#016x}", hash);
size_t env_index{0};
@@ -822,12 +788,10 @@ std::unique_ptr<GraphicsPipeline> PipelineCache::CreateGraphicsPipeline(
Shader::IR::Program* layer_source_program{};
for (size_t index = 0; index < Maxwell::MaxShaderProgram; ++index) {
const bool is_emulated_stage = layer_source_program != nullptr &&
index == static_cast<u32>(Maxwell::ShaderType::Geometry);
const bool is_emulated_stage = layer_source_program != nullptr && index == u32(Maxwell::ShaderType::Geometry);
if (key.unique_hashes[index] == 0 && is_emulated_stage) {
auto topology = MaxwellToOutputTopology(key.state.topology);
programs[index] = GenerateGeometryPassthrough(pools.inst, pools.block, host_info,
*layer_source_program, topology);
programs[index] = Shader::Maxwell::GenerateGeometryPassthrough(pools, host_info, *layer_source_program, topology);
continue;
}
if (key.unique_hashes[index] == 0) {
@@ -836,16 +800,16 @@ std::unique_ptr<GraphicsPipeline> PipelineCache::CreateGraphicsPipeline(
Shader::Environment& env{*envs[env_index]};
++env_index;
const u32 cfg_offset{static_cast<u32>(env.StartAddress() + sizeof(Shader::ProgramHeader))};
const u32 cfg_offset{u32(env.StartAddress() + sizeof(Shader::ProgramHeader))};
Shader::Maxwell::Flow::CFG cfg(env, pools.flow_block, cfg_offset, index == 0);
if (!uses_vertex_a || index != 1) {
// Normal path
programs[index] = TranslateProgram(pools.inst, pools.block, env, cfg, host_info);
programs[index] = Shader::Maxwell::TranslateProgram(pools, env, cfg, host_info);
} else {
// VertexB path when VertexA is present.
auto& program_va{programs[0]};
auto program_vb{TranslateProgram(pools.inst, pools.block, env, cfg, host_info)};
programs[index] = MergeDualVertexPrograms(program_va, program_vb, env);
auto program_vb{Shader::Maxwell::TranslateProgram(pools, env, cfg, host_info)};
programs[index] = Shader::Maxwell::MergeDualVertexPrograms(program_va, program_vb, env);
}
if (Settings::values.dump_guest_shaders) {
@@ -875,8 +839,8 @@ std::unique_ptr<GraphicsPipeline> PipelineCache::CreateGraphicsPipeline(
infos[stage_index] = &program.info;
const auto runtime_info{MakeRuntimeInfo(programs, key, program, previous_stage, device)};
ConvertLegacyToGeneric(program, runtime_info);
const std::vector<u32> code{EmitSPIRV(profile, runtime_info, program, binding)};
Shader::Maxwell::ConvertLegacyToGeneric(program, runtime_info);
const std::vector<u32> code{Shader::Backend::SPIRV::EmitSPIRV(profile, runtime_info, program, binding)};
device.SaveShader(code);
modules[stage_index] = BuildShader(device, code);
@@ -932,14 +896,14 @@ std::unique_ptr<GraphicsPipeline> PipelineCache::CreateGraphicsPipeline() {
GraphicsEnvironments environments;
GetGraphicsEnvironments(environments, graphics_key.unique_hashes);
main_pools.ReleaseContents();
main_pools.clear();
auto pipeline{
CreateGraphicsPipeline(main_pools, graphics_key, environments.Span(), nullptr, true)};
if (!pipeline || pipeline_cache_filename.empty()) {
return pipeline;
}
serialization_thread.QueueWork([this, key = graphics_key, envs = std::move(environments.envs)] {
boost::container::static_vector<const GenericEnvironment*, Maxwell::MaxShaderProgram>
boost::container::static_vector<const VideoCommon::GenericEnvironment*, Maxwell::MaxShaderProgram>
env_ptrs;
for (size_t index = 0; index < Maxwell::MaxShaderProgram; ++index) {
if (key.unique_hashes[index] != 0) {
@@ -956,25 +920,23 @@ std::unique_ptr<ComputePipeline> PipelineCache::CreateComputePipeline(
const ComputePipelineCacheKey& key, const ShaderInfo* shader) {
const GPUVAddr program_base{kepler_compute->regs.code_loc.Address()};
const auto& qmd{kepler_compute->launch_description};
ComputeEnvironment env{*kepler_compute, *gpu_memory, program_base, qmd.program_start};
VideoCommon::ComputeEnvironment env{*kepler_compute, *gpu_memory, program_base, qmd.program_start};
env.SetCachedSize(shader->size_bytes);
main_pools.ReleaseContents();
main_pools.clear();
auto pipeline{CreateComputePipeline(main_pools, key, env, nullptr, true)};
if (!pipeline || pipeline_cache_filename.empty()) {
return pipeline;
}
serialization_thread.QueueWork([this, key, env_ = std::move(env)] {
SerializePipeline(key, std::array<const GenericEnvironment*, 1>{&env_},
SerializePipeline(key, std::array<const VideoCommon::GenericEnvironment*, 1>{&env_},
pipeline_cache_filename, CACHE_VERSION);
});
QueueVulkanPipelineCacheFlush();
return pipeline;
}
std::unique_ptr<ComputePipeline> PipelineCache::CreateComputePipeline(
ShaderPools& pools, const ComputePipelineCacheKey& key, Shader::Environment& env,
PipelineStatistics* statistics, bool build_in_parallel) try {
std::unique_ptr<ComputePipeline> PipelineCache::CreateComputePipeline(Shader::Maxwell::ShaderPools& pools, const ComputePipelineCacheKey& key, Shader::Environment& env, PipelineStatistics* statistics, bool build_in_parallel) try {
auto hash = key.Hash();
if (device.HasBrokenCompute()) {
LOG_ERROR(Render_Vulkan, "Skipping {:#016x}", hash);
@@ -990,11 +952,9 @@ std::unique_ptr<ComputePipeline> PipelineCache::CreateComputePipeline(
env.Dump(hash, key.unique_hash);
}
auto program{TranslateProgram(pools.inst, pools.block, env, cfg, host_info)};
auto program{Shader::Maxwell::TranslateProgram(pools, env, cfg, host_info)};
const VkDriverIdKHR driver_id = device.GetDriverID();
const bool needs_shared_mem_clamp =
driver_id == VK_DRIVER_ID_QUALCOMM_PROPRIETARY ||
driver_id == VK_DRIVER_ID_ARM_PROPRIETARY;
const bool needs_shared_mem_clamp = driver_id == VK_DRIVER_ID_QUALCOMM_PROPRIETARY || driver_id == VK_DRIVER_ID_ARM_PROPRIETARY;
const u32 max_shared_memory = device.GetMaxComputeSharedMemorySize();
if (needs_shared_mem_clamp && program.shared_memory_size > max_shared_memory) {
LOG_WARNING(Render_Vulkan,
@@ -1004,7 +964,7 @@ std::unique_ptr<ComputePipeline> PipelineCache::CreateComputePipeline(
max_shared_memory / 1024);
program.shared_memory_size = max_shared_memory;
}
const std::vector<u32> code{EmitSPIRV(profile, program)};
const std::vector<u32> code{Shader::Backend::SPIRV::EmitSPIRV(profile, program)};
device.SaveShader(code);
vk::ShaderModule spv_module{BuildShader(device, code)};
@@ -15,15 +15,18 @@
#include <type_traits>
#include "common/container/unordered_map.h"
#include <vector>
#include <boost/container/stable_vector.hpp>
#include "common/common_types.h"
#include "common/thread_worker.h"
#include "shader_recompiler/frontend/ir/basic_block.h"
#include "shader_recompiler/frontend/ir/value.h"
#include "shader_recompiler/frontend/maxwell/control_flow.h"
#include "shader_recompiler/frontend/maxwell/structured_control_flow.h"
#include "shader_recompiler/frontend/maxwell/translate_program.h"
#include "shader_recompiler/host_translate_info.h"
#include "shader_recompiler/object_pool.h"
#include "shader_recompiler/profile.h"
#include "shader_recompiler/shader_pool.h"
#include "video_core/engines/maxwell_3d.h"
#include "video_core/host1x/gpu_device_memory_manager.h"
#include "video_core/renderer_vulkan/fixed_pipeline_state.h"
@@ -90,18 +93,6 @@ class Scheduler;
using VideoCommon::ShaderInfo;
struct ShaderPools {
void ReleaseContents() {
flow_block.ReleaseContents();
block.ReleaseContents();
inst.ReleaseContents();
}
Shader::ObjectPool<Shader::IR::Inst> inst{8192};
Shader::ObjectPool<Shader::IR::Block> block{32};
Shader::ObjectPool<Shader::Maxwell::Flow::Block> flow_block{32};
};
class PipelineCache : public VideoCommon::ShaderCache {
public:
explicit PipelineCache(Tegra::MaxwellDeviceMemoryManager& device_memory_, const Device& device,
@@ -127,14 +118,14 @@ private:
std::unique_ptr<GraphicsPipeline> CreateGraphicsPipeline();
std::unique_ptr<GraphicsPipeline> CreateGraphicsPipeline(
ShaderPools& pools, const GraphicsPipelineCacheKey& key,
Shader::Maxwell::ShaderPools& pools, const GraphicsPipelineCacheKey& key,
std::span<Shader::Environment* const> envs, PipelineStatistics* statistics,
bool build_in_parallel);
std::unique_ptr<ComputePipeline> CreateComputePipeline(const ComputePipelineCacheKey& key,
const ShaderInfo* shader);
std::unique_ptr<ComputePipeline> CreateComputePipeline(ShaderPools& pools,
std::unique_ptr<ComputePipeline> CreateComputePipeline(Shader::Maxwell::ShaderPools& pools,
const ComputePipelineCacheKey& key,
Shader::Environment& env,
PipelineStatistics* statistics,
@@ -166,7 +157,7 @@ private:
::Common::unordered_map<ComputePipelineCacheKey, std::unique_ptr<ComputePipeline>> compute_cache;
::Common::unordered_map<GraphicsPipelineCacheKey, std::unique_ptr<GraphicsPipeline>> graphics_cache;
ShaderPools main_pools;
Shader::Maxwell::ShaderPools main_pools;
Shader::Profile profile;
Shader::HostTranslateInfo host_info;
+1 -2
View File
@@ -11,7 +11,6 @@
#include "common/assert.h"
#include "shader_recompiler/frontend/maxwell/control_flow.h"
#include "shader_recompiler/object_pool.h"
#include "video_core/control/channel_state.h"
#include "video_core/dirty_flags.h"
#include "video_core/engines/kepler_compute.h"
@@ -236,7 +235,7 @@ const ShaderInfo* ShaderCache::MakeShaderInfo(GenericEnvironment& env, VAddr cpu
} else {
// Slow path, not really hit on commercial games
// Build a control flow graph to get the real shader size
Shader::ObjectPool<Shader::Maxwell::Flow::Block> flow_block;
boost::container::stable_vector<Shader::Maxwell::Flow::Block> flow_block;
Shader::Maxwell::Flow::CFG cfg{env, flow_block, env.StartAddress()};
info->unique_hash = env.CalculateHash();
info->size_bytes = env.ReadSizeBytes();
@@ -708,6 +708,7 @@ Device::Device(VkInstance instance_, vk::PhysicalDevice physical_, VkSurfaceKHR
descriptor_indexing.shaderUniformTexelBufferArrayDynamicIndexing = false;
descriptor_indexing.shaderStorageTexelBufferArrayDynamicIndexing = false;
descriptor_indexing.shaderUniformBufferArrayNonUniformIndexing = false;
descriptor_indexing.shaderStorageBufferArrayNonUniformIndexing = false;
descriptor_indexing.shaderInputAttachmentArrayNonUniformIndexing = false;
descriptor_indexing.descriptorBindingUniformBufferUpdateAfterBind = false;
descriptor_indexing.descriptorBindingSampledImageUpdateAfterBind = false;
@@ -360,7 +360,6 @@ public:
#define FN_MAX_LIMIT_LIST \
FN_MAX_LIMIT_ELEM(ComputeSharedMemorySize) \
FN_MAX_LIMIT_ELEM(PerStageDescriptorSampledImages) \
FN_MAX_LIMIT_ELEM(PerStageDescriptorStorageBuffers) \
FN_MAX_LIMIT_ELEM(PerStageResources) \
FN_MAX_LIMIT_ELEM(DescriptorSetSamplers) \
FN_MAX_LIMIT_ELEM(DescriptorSetUniformBuffers) \
@@ -416,10 +415,6 @@ FN_MAX_LIMIT_LIST
return features.descriptor_indexing.shaderStorageTexelBufferArrayNonUniformIndexing;
}
bool IsStorageBufferArrayNonUniformIndexingSupported() const {
return features.descriptor_indexing.shaderStorageBufferArrayNonUniformIndexing;
}
/// Returns true if the device supports float64 natively.
bool IsFloat64Supported() const {
return features.features.shaderFloat64;