11 files changed, 107 insertions, 99 deletions

diff --git a/externals/dynarmic b/externals/dynarmic
-Subproject a1cbea7948372989218a4e6159a95998d65876a
+Subproject befe547d5631024a70d81d2ccee808bbfcb3854
diff --git a/src/audio_core/renderer/command/resample/upsample.cpp b/src/audio_core/renderer/command/resample/upsample.cpp
index 6c3ff31f7..5f7db12ca 100644
--- a/src/audio_core/renderer/command/resample/upsample.cpp
+++ b/src/audio_core/renderer/command/resample/upsample.cpp
@@ -20,25 +20,25 @@ static void SrcProcessFrame(std::span<s32> output, std::span<const s32> input,
                             const u32 target_sample_count, const u32 source_sample_count,
                             UpsamplerState* state) {
     constexpr u32 WindowSize = 10;
-    constexpr std::array<Common::FixedPoint<24, 8>, WindowSize> SincWindow1{
-        51.93359375f, -18.80078125f, 9.73046875f, -5.33203125f, 2.84375f,
-        -1.41015625f, 0.62109375f,   -0.2265625f, 0.0625f,      -0.00390625f,
+    constexpr std::array<Common::FixedPoint<17, 15>, WindowSize> WindowedSinc1{
+        0.95376587f,   -0.12872314f, 0.060028076f,  -0.032470703f, 0.017669678f,
+        -0.009124756f, 0.004272461f, -0.001739502f, 0.000579834f,  -0.000091552734f,
     };
-    constexpr std::array<Common::FixedPoint<24, 8>, WindowSize> SincWindow2{
-        105.35546875f, -24.52734375f, 11.9609375f,  -6.515625f, 3.52734375f,
-        -1.796875f,    0.828125f,     -0.32421875f, 0.1015625f, -0.015625f,
+    constexpr std::array<Common::FixedPoint<17, 15>, WindowSize> WindowedSinc2{
+        0.8230896f,    -0.19161987f,  0.093444824f,  -0.05090332f,   0.027557373f,
+        -0.014038086f, 0.0064697266f, -0.002532959f, 0.00079345703f, -0.00012207031f,
     };
-    constexpr std::array<Common::FixedPoint<24, 8>, WindowSize> SincWindow3{
-        122.08203125f, -16.47656250f, 7.68359375f,  -4.15625000f, 2.26171875f,
-        -1.16796875f,  0.54687500f,   -0.22265625f, 0.07421875f,  -0.01171875f,
+    constexpr std::array<Common::FixedPoint<17, 15>, WindowSize> WindowedSinc3{
+        0.6298828f,    -0.19274902f, 0.09725952f,    -0.05319214f,  0.028625488f,
+        -0.014373779f, 0.006500244f, -0.0024719238f, 0.0007324219f, -0.000091552734f,
     };
-    constexpr std::array<Common::FixedPoint<24, 8>, WindowSize> SincWindow4{
-        23.73437500f, -9.62109375f, 5.07812500f,  -2.78125000f, 1.46875000f,
-        -0.71484375f, 0.30859375f,  -0.10546875f, 0.02734375f,  0.00000000f,
+    constexpr std::array<Common::FixedPoint<17, 15>, WindowSize> WindowedSinc4{
+        0.4057312f,    -0.1468811f,  0.07601929f,    -0.041656494f,  0.022216797f,
+        -0.011016846f, 0.004852295f, -0.0017700195f, 0.00048828125f, -0.000030517578f,
     };
-    constexpr std::array<Common::FixedPoint<24, 8>, WindowSize> SincWindow5{
-        80.62500000f, -24.67187500f, 12.44921875f, -6.80859375f, 3.66406250f,
-        -1.83984375f, 0.83203125f,   -0.31640625f, 0.09375000f,  -0.01171875f,
+    constexpr std::array<Common::FixedPoint<17, 15>, WindowSize> WindowedSinc5{
+        0.1854248f,    -0.075164795f, 0.03967285f,    -0.021728516f,  0.011474609f,
+        -0.005584717f, 0.0024108887f, -0.0008239746f, 0.00021362305f, 0.0f,
     };
 
     if (!state->initialized) {
@@ -91,52 +91,31 @@ static void SrcProcessFrame(std::span<s32> output, std::span<const s32> input,
             static_cast<u16>((state->history_output_index + 1) % UpsamplerState::HistorySize);
     };
 
-    auto calculate_sample = [&state](std::span<const Common::FixedPoint<24, 8>> coeffs1,
-                                     std::span<const Common::FixedPoint<24, 8>> coeffs2) -> s32 {
+    auto calculate_sample = [&state](std::span<const Common::FixedPoint<17, 15>> coeffs1,
+                                     std::span<const Common::FixedPoint<17, 15>> coeffs2) -> s32 {
         auto output_index{state->history_output_index};
-        auto start_pos{output_index - state->history_start_index + 1U};
-        auto end_pos{10U};
+        u64 result{0};
 
-        if (start_pos < 10) {
-            end_pos = start_pos;
-        }
-
-        u64 prev_contrib{0};
-        u32 coeff_index{0};
-        for (; coeff_index < end_pos; coeff_index++, output_index--) {
-            prev_contrib += static_cast<u64>(state->history[output_index].to_raw()) *
-                            coeffs1[coeff_index].to_raw();
-        }
+        for (u32 coeff_index = 0; coeff_index < 10; coeff_index++) {
+            result += static_cast<u64>(state->history[output_index].to_raw()) *
+                      coeffs1[coeff_index].to_raw();
 
-        auto end_index{state->history_end_index};
-        for (; start_pos < 9; start_pos++, coeff_index++, end_index--) {
-            prev_contrib += static_cast<u64>(state->history[end_index].to_raw()) *
-                            coeffs1[coeff_index].to_raw();
+            output_index = output_index == state->history_start_index ? state->history_end_index
+                                                                      : output_index - 1;
         }
 
         output_index =
             static_cast<u16>((state->history_output_index + 1) % UpsamplerState::HistorySize);
-        start_pos = state->history_end_index - output_index + 1U;
-        end_pos = 10U;
 
-        if (start_pos < 10) {
-            end_pos = start_pos;
-        }
-
-        u64 next_contrib{0};
-        coeff_index = 0;
-        for (; coeff_index < end_pos; coeff_index++, output_index++) {
-            next_contrib += static_cast<u64>(state->history[output_index].to_raw()) *
-                            coeffs2[coeff_index].to_raw();
-        }
+        for (u32 coeff_index = 0; coeff_index < 10; coeff_index++) {
+            result += static_cast<u64>(state->history[output_index].to_raw()) *
+                      coeffs2[coeff_index].to_raw();
 
-        auto start_index{state->history_start_index};
-        for (; start_pos < 9; start_pos++, start_index++, coeff_index++) {
-            next_contrib += static_cast<u64>(state->history[start_index].to_raw()) *
-                            coeffs2[coeff_index].to_raw();
+            output_index = output_index == state->history_end_index ? state->history_start_index
+                                                                    : output_index + 1;
         }
 
-        return static_cast<s32>(((prev_contrib >> 15) + (next_contrib >> 15)) >> 8);
+        return static_cast<s32>(result >> (8 + 15));
     };
 
     switch (state->ratio.to_int_floor()) {
@@ -150,23 +129,23 @@ static void SrcProcessFrame(std::span<s32> output, std::span<const s32> input,
                 break;
 
             case 1:
-                output[write_index] = calculate_sample(SincWindow3, SincWindow4);
+                output[write_index] = calculate_sample(WindowedSinc1, WindowedSinc5);
                 break;
 
             case 2:
-                output[write_index] = calculate_sample(SincWindow2, SincWindow1);
+                output[write_index] = calculate_sample(WindowedSinc2, WindowedSinc4);
                 break;
 
             case 3:
-                output[write_index] = calculate_sample(SincWindow5, SincWindow5);
+                output[write_index] = calculate_sample(WindowedSinc3, WindowedSinc3);
                 break;
 
             case 4:
-                output[write_index] = calculate_sample(SincWindow1, SincWindow2);
+                output[write_index] = calculate_sample(WindowedSinc4, WindowedSinc2);
                 break;
 
             case 5:
-                output[write_index] = calculate_sample(SincWindow4, SincWindow3);
+                output[write_index] = calculate_sample(WindowedSinc5, WindowedSinc1);
                 break;
             }
             state->sample_index = static_cast<u8>((state->sample_index + 1) % 6);
@@ -183,11 +162,11 @@ static void SrcProcessFrame(std::span<s32> output, std::span<const s32> input,
                 break;
 
             case 1:
-                output[write_index] = calculate_sample(SincWindow2, SincWindow1);
+                output[write_index] = calculate_sample(WindowedSinc2, WindowedSinc4);
                 break;
 
             case 2:
-                output[write_index] = calculate_sample(SincWindow1, SincWindow2);
+                output[write_index] = calculate_sample(WindowedSinc4, WindowedSinc2);
                 break;
             }
             state->sample_index = static_cast<u8>((state->sample_index + 1) % 3);
@@ -204,12 +183,12 @@ static void SrcProcessFrame(std::span<s32> output, std::span<const s32> input,
                 break;
 
             case 1:
-                output[write_index] = calculate_sample(SincWindow1, SincWindow2);
+                output[write_index] = calculate_sample(WindowedSinc4, WindowedSinc2);
                 break;
 
             case 2:
                 increment();
-                output[write_index] = calculate_sample(SincWindow2, SincWindow1);
+                output[write_index] = calculate_sample(WindowedSinc2, WindowedSinc4);
                 break;
             }
             state->sample_index = static_cast<u8>((state->sample_index + 1) % 3);
diff --git a/src/common/settings.cpp b/src/common/settings.cpp
index 1638b79f5..b1a2aa8b2 100644
--- a/src/common/settings.cpp
+++ b/src/common/settings.cpp
@@ -129,6 +129,10 @@ void UpdateRescalingInfo() {
         info.up_scale = 1;
         info.down_shift = 0;
         break;
+    case ResolutionSetup::Res3_2X:
+        info.up_scale = 3;
+        info.down_shift = 1;
+        break;
     case ResolutionSetup::Res2X:
         info.up_scale = 2;
         info.down_shift = 0;
@@ -149,6 +153,14 @@ void UpdateRescalingInfo() {
         info.up_scale = 6;
         info.down_shift = 0;
         break;
+    case ResolutionSetup::Res7X:
+        info.up_scale = 7;
+        info.down_shift = 0;
+        break;
+    case ResolutionSetup::Res8X:
+        info.up_scale = 8;
+        info.down_shift = 0;
+        break;
     default:
         ASSERT(false);
         info.up_scale = 1;
diff --git a/src/common/settings.h b/src/common/settings.h
index a457e3f23..80b2eeabc 100644
--- a/src/common/settings.h
+++ b/src/common/settings.h
@@ -56,11 +56,14 @@ enum class ResolutionSetup : u32 {
     Res1_2X = 0,
     Res3_4X = 1,
     Res1X = 2,
-    Res2X = 3,
-    Res3X = 4,
-    Res4X = 5,
-    Res5X = 6,
-    Res6X = 7,
+    Res3_2X = 3,
+    Res2X = 4,
+    Res3X = 5,
+    Res4X = 6,
+    Res5X = 7,
+    Res6X = 8,
+    Res7X = 9,
+    Res8X = 10,
 };
 
 enum class ScalingFilter : u32 {
diff --git a/src/core/core_timing.cpp b/src/core/core_timing.cpp
index 0e7b5f943..6bac6722f 100644
--- a/src/core/core_timing.cpp
+++ b/src/core/core_timing.cpp
@@ -142,16 +142,24 @@ void CoreTiming::ScheduleLoopingEvent(std::chrono::nanoseconds start_time,
 }
 
 void CoreTiming::UnscheduleEvent(const std::shared_ptr<EventType>& event_type,
-                                 std::uintptr_t user_data) {
-    std::scoped_lock scope{basic_lock};
-    const auto itr = std::remove_if(event_queue.begin(), event_queue.end(), [&](const Event& e) {
-        return e.type.lock().get() == event_type.get() && e.user_data == user_data;
-    });
-
-    // Removing random items breaks the invariant so we have to re-establish it.
-    if (itr != event_queue.end()) {
-        event_queue.erase(itr, event_queue.end());
-        std::make_heap(event_queue.begin(), event_queue.end(), std::greater<>());
+                                 std::uintptr_t user_data, bool wait) {
+    {
+        std::scoped_lock lk{basic_lock};
+        const auto itr =
+            std::remove_if(event_queue.begin(), event_queue.end(), [&](const Event& e) {
+                return e.type.lock().get() == event_type.get() && e.user_data == user_data;
+            });
+
+        // Removing random items breaks the invariant so we have to re-establish it.
+        if (itr != event_queue.end()) {
+            event_queue.erase(itr, event_queue.end());
+            std::make_heap(event_queue.begin(), event_queue.end(), std::greater<>());
+        }
+    }
+
+    // Force any in-progress events to finish
+    if (wait) {
+        std::scoped_lock lk{advance_lock};
     }
 }
 
@@ -190,20 +198,6 @@ u64 CoreTiming::GetClockTicks() const {
     return CpuCyclesToClockCycles(ticks);
 }
 
-void CoreTiming::RemoveEvent(const std::shared_ptr<EventType>& event_type) {
-    std::scoped_lock lock{basic_lock};
-
-    const auto itr = std::remove_if(event_queue.begin(), event_queue.end(), [&](const Event& e) {
-        return e.type.lock().get() == event_type.get();
-    });
-
-    // Removing random items breaks the invariant so we have to re-establish it.
-    if (itr != event_queue.end()) {
-        event_queue.erase(itr, event_queue.end());
-        std::make_heap(event_queue.begin(), event_queue.end(), std::greater<>());
-    }
-}
-
 std::optional<s64> CoreTiming::Advance() {
     std::scoped_lock lock{advance_lock, basic_lock};
     global_timer = GetGlobalTimeNs().count();
diff --git a/src/core/core_timing.h b/src/core/core_timing.h
index b5925193c..da366637b 100644
--- a/src/core/core_timing.h
+++ b/src/core/core_timing.h
@@ -98,10 +98,13 @@ public:
                               const std::shared_ptr<EventType>& event_type,
                               std::uintptr_t user_data = 0, bool absolute_time = false);
 
-    void UnscheduleEvent(const std::shared_ptr<EventType>& event_type, std::uintptr_t user_data);
+    void UnscheduleEvent(const std::shared_ptr<EventType>& event_type, std::uintptr_t user_data,
+                         bool wait = true);
 
-    /// We only permit one event of each type in the queue at a time.
-    void RemoveEvent(const std::shared_ptr<EventType>& event_type);
+    void UnscheduleEventWithoutWait(const std::shared_ptr<EventType>& event_type,
+                                    std::uintptr_t user_data) {
+        UnscheduleEvent(event_type, user_data, false);
+    }
 
     void AddTicks(u64 ticks_to_add);
 
diff --git a/src/core/hle/kernel/k_hardware_timer.cpp b/src/core/hle/kernel/k_hardware_timer.cpp
index 6bba79ea0..4dcd53821 100644
--- a/src/core/hle/kernel/k_hardware_timer.cpp
+++ b/src/core/hle/kernel/k_hardware_timer.cpp
@@ -18,7 +18,8 @@ void KHardwareTimer::Initialize() {
 }
 
 void KHardwareTimer::Finalize() {
-    this->DisableInterrupt();
+    m_kernel.System().CoreTiming().UnscheduleEvent(m_event_type, reinterpret_cast<uintptr_t>(this));
+    m_wakeup_time = std::numeric_limits<s64>::max();
     m_event_type.reset();
 }
 
@@ -59,7 +60,8 @@ void KHardwareTimer::EnableInterrupt(s64 wakeup_time) {
 }
 
 void KHardwareTimer::DisableInterrupt() {
-    m_kernel.System().CoreTiming().UnscheduleEvent(m_event_type, reinterpret_cast<uintptr_t>(this));
+    m_kernel.System().CoreTiming().UnscheduleEventWithoutWait(m_event_type,
+                                                              reinterpret_cast<uintptr_t>(this));
     m_wakeup_time = std::numeric_limits<s64>::max();
 }
 
diff --git a/src/core/hle/service/nvflinger/nvflinger.cpp b/src/core/hle/service/nvflinger/nvflinger.cpp
index d1cbadde4..f4416f5b2 100644
--- a/src/core/hle/service/nvflinger/nvflinger.cpp
+++ b/src/core/hle/service/nvflinger/nvflinger.cpp
@@ -312,8 +312,6 @@ void NVFlinger::Compose() {
 }
 
 s64 NVFlinger::GetNextTicks() const {
-    static constexpr s64 max_hertz = 120LL;
-
     const auto& settings = Settings::values;
     auto speed_scale = 1.f;
     if (settings.use_multi_core.GetValue()) {
@@ -327,9 +325,11 @@ s64 NVFlinger::GetNextTicks() const {
         }
     }
 
-    const auto next_ticks = ((1000000000 * (1LL << swap_interval)) / max_hertz);
+    // As an extension, treat nonpositive swap interval as framerate multiplier.
+    const f32 effective_fps = swap_interval <= 0 ? 120.f * static_cast<f32>(1 - swap_interval)
+                                                 : 60.f / static_cast<f32>(swap_interval);
 
-    return static_cast<s64>(speed_scale * static_cast<float>(next_ticks));
+    return static_cast<s64>(speed_scale * (1000000000.f / effective_fps));
 }
 
 } // namespace Service::NVFlinger
diff --git a/src/core/hle/service/nvflinger/nvflinger.h b/src/core/hle/service/nvflinger/nvflinger.h
index 9b22397db..3828cf272 100644
--- a/src/core/hle/service/nvflinger/nvflinger.h
+++ b/src/core/hle/service/nvflinger/nvflinger.h
@@ -133,7 +133,7 @@ private:
     /// layers.
     u32 next_buffer_queue_id = 1;
 
-    u32 swap_interval = 1;
+    s32 swap_interval = 1;
 
     /// Event that handles screen composition.
     std::shared_ptr<Core::Timing::EventType> multi_composition_event;
diff --git a/src/video_core/engines/maxwell_3d.cpp b/src/video_core/engines/maxwell_3d.cpp
index 97f547789..68ceda519 100644
--- a/src/video_core/engines/maxwell_3d.cpp
+++ b/src/video_core/engines/maxwell_3d.cpp
@@ -467,7 +467,7 @@ void Maxwell3D::ProcessMacroBind(u32 data) {
 }
 
 void Maxwell3D::ProcessFirmwareCall4() {
-    LOG_WARNING(HW_GPU, "(STUBBED) called");
+    LOG_DEBUG(HW_GPU, "(STUBBED) called");
 
     // Firmware call 4 is a blob that changes some registers depending on its parameters.
     // These registers don't affect emulation and so are stubbed by setting 0xd00 to 1.
diff --git a/src/yuzu/configuration/configure_graphics.ui b/src/yuzu/configuration/configure_graphics.ui
index aa02cc63c..bb9910a53 100644
--- a/src/yuzu/configuration/configure_graphics.ui
+++ b/src/yuzu/configuration/configure_graphics.ui
@@ -366,6 +366,11 @@
              </item>
              <item>
               <property name="text">
+                  <string>1.5X (1080p/1620p) [EXPERIMENTAL]</string>
+              </property>
+             </item>
+             <item>
+              <property name="text">
                <string>2X (1440p/2160p)</string>
               </property>
              </item>
@@ -389,6 +394,16 @@
                <string>6X (4320p/6480p)</string>
               </property>
              </item>
+             <item>
+              <property name="text">
+               <string>7X (5040p/7560p)</string>
+              </property>
+             </item>
+             <item>
+              <property name="text">
+               <string>8X (5760p/8640p)</string>
+              </property>
+             </item>
             </widget>
            </item>
           </layout>