Rasterizer cache refactor v2 (#6479)

* rasterizer_cache: Switch to template

* Eliminates all opengl references in the rasterizer cache headers
  thus completing the backend abstraction

* rasterizer_cache: Switch to page table

* Surface storage isn't particularly interval sensitive so we can use a page table to make it faster

* rasterizer_cache: Move sampler management out of rasterizer cache

* rasterizer_cache: Remove shared_ptr usage

* Switches to yuzu's slot vector for improved memory locality.

* rasterizer_cache: Rework reinterpretation lookup

* citra_qt: Per game texture filter

* rasterizer_cache: Log additional settings

* gl_texture_runtime: Resolve shadow map comment

* rasterizer_cache: Don't use float for viewport

* gl_texture_runtime: Fix custom allocation recycling

* rasterizer_cache: Minor cleanups

* Cleanup texture cubes when all the faces have been unregistered from the cache

* custom_tex_manager: Allow multiple hash mappings per texture

* code: Move slot vector to common

* rasterizer_cache: Prevent texture cube crashes

* rasterizer_cache: Improve mipmap validation

* CanSubRect now works properly when validating multi-level surfaces, for example Dark Moon validates a 4 level surface from a 3 level one and it works

* gl_blit_handler: Unbind sampler on reinterpretation

src/common/CMakeLists.txt

@@ -100,6 +100,7 @@ add_library(citra_common STATIC
     scope_exit.h
     settings.cpp
     settings.h
+    slot_vector.h
     serialization/atomic.h
     serialization/boost_discrete_interval.hpp
     serialization/boost_flat_set.h

src/common/settings.cpp

@@ -163,6 +163,8 @@ void LogSettings() {
     log_setting("Layout_LargeScreenProportion", values.large_screen_proportion.GetValue());
     log_setting("Utility_DumpTextures", values.dump_textures.GetValue());
     log_setting("Utility_CustomTextures", values.custom_textures.GetValue());
+    log_setting("Utility_PreloadTextures", values.preload_textures.GetValue());
+    log_setting("Utility_AsyncCustomLoading", values.async_custom_loading.GetValue());
     log_setting("Utility_UseDiskShaderCache", values.use_disk_shader_cache.GetValue());
     log_setting("Audio_Emulation", GetAudioEmulationName(values.audio_emulation.GetValue()));
     log_setting("Audio_OutputType", values.output_type.GetValue());

src/common/slot_vector.h

@@ -0,0 +1,154 @@
+// Copyright 2020 yuzu Emulator Project
+// Licensed under GPLv2 or any later version
+// Refer to the license.txt file included.
+
+#pragma once
+
+#include <bit>
+#include <compare>
+#include <numeric>
+#include <type_traits>
+#include <utility>
+#include <vector>
+#include "common/assert.h"
+#include "common/common_types.h"
+
+namespace Common {
+
+struct SlotId {
+    static constexpr u32 INVALID_INDEX = std::numeric_limits<u32>::max();
+
+    constexpr auto operator<=>(const SlotId&) const noexcept = default;
+
+    constexpr explicit operator bool() const noexcept {
+        return index != INVALID_INDEX;
+    }
+
+    u32 index = INVALID_INDEX;
+};
+
+template <class T>
+class SlotVector {
+public:
+    ~SlotVector() noexcept {
+        size_t index = 0;
+        for (u64 bits : stored_bitset) {
+            for (size_t bit = 0; bits; ++bit, bits >>= 1) {
+                if ((bits & 1) != 0) {
+                    values[index + bit].object.~T();
+                }
+            }
+            index += 64;
+        }
+        delete[] values;
+    }
+
+    [[nodiscard]] T& operator[](SlotId id) noexcept {
+        ValidateIndex(id);
+        return values[id.index].object;
+    }
+
+    [[nodiscard]] const T& operator[](SlotId id) const noexcept {
+        ValidateIndex(id);
+        return values[id.index].object;
+    }
+
+    template <typename... Args>
+    [[nodiscard]] SlotId insert(Args&&... args) noexcept {
+        const u32 index = FreeValueIndex();
+        new (&values[index].object) T(std::forward<Args>(args)...);
+        SetStorageBit(index);
+
+        return SlotId{index};
+    }
+
+    void erase(SlotId id) noexcept {
+        values[id.index].object.~T();
+        free_list.push_back(id.index);
+        ResetStorageBit(id.index);
+    }
+
+private:
+    struct NonTrivialDummy {
+        NonTrivialDummy() noexcept {}
+    };
+
+    union Entry {
+        Entry() noexcept : dummy{} {}
+        ~Entry() noexcept {}
+
+        NonTrivialDummy dummy;
+        T object;
+    };
+
+    void SetStorageBit(u32 index) noexcept {
+        stored_bitset[index / 64] |= u64(1) << (index % 64);
+    }
+
+    void ResetStorageBit(u32 index) noexcept {
+        stored_bitset[index / 64] &= ~(u64(1) << (index % 64));
+    }
+
+    bool ReadStorageBit(u32 index) noexcept {
+        return ((stored_bitset[index / 64] >> (index % 64)) & 1) != 0;
+    }
+
+    void ValidateIndex(SlotId id) const noexcept {
+        DEBUG_ASSERT(id);
+        DEBUG_ASSERT(id.index / 64 < stored_bitset.size());
+        DEBUG_ASSERT(((stored_bitset[id.index / 64] >> (id.index % 64)) & 1) != 0);
+    }
+
+    [[nodiscard]] u32 FreeValueIndex() noexcept {
+        if (free_list.empty()) {
+            Reserve(values_capacity ? (values_capacity << 1) : 1);
+        }
+
+        const u32 free_index = free_list.back();
+        free_list.pop_back();
+        return free_index;
+    }
+
+    void Reserve(size_t new_capacity) noexcept {
+        Entry* const new_values = new Entry[new_capacity];
+        size_t index = 0;
+        for (u64 bits : stored_bitset) {
+            for (size_t bit = 0; bits; ++bit, bits >>= 1) {
+                const size_t i = index + bit;
+                if ((bits & 1) == 0) {
+                    continue;
+                }
+                T& old_value = values[i].object;
+                new (&new_values[i].object) T(std::move(old_value));
+                old_value.~T();
+            }
+            index += 64;
+        }
+
+        stored_bitset.resize((new_capacity + 63) / 64);
+
+        const size_t old_free_size = free_list.size();
+        free_list.resize(old_free_size + (new_capacity - values_capacity));
+        std::iota(free_list.begin() + old_free_size, free_list.end(),
+                  static_cast<u32>(values_capacity));
+
+        delete[] values;
+        values = new_values;
+        values_capacity = new_capacity;
+    }
+
+    Entry* values = nullptr;
+    size_t values_capacity = 0;
+
+    std::vector<u64> stored_bitset;
+    std::vector<u32> free_list;
+};
+
+} // namespace Common
+
+template <>
+struct std::hash<Common::SlotId> {
+    size_t operator()(const Common::SlotId& id) const noexcept {
+        return std::hash<u32>{}(id.index);
+    }
+};