LRU-cache NCZBLOCK blocks for faster update boots

Restores performance but need to check TOTK item icons and Jamboree graphic issues.

- [x] I have read and followed the [Contribution Guidelines](https://git.eden-emu.dev/eden-emu/eden/src/branch/master/CONTRIBUTING.md#code-contributions).
- [x] I have read and followed the [AI Policy](https://git.eden-emu.dev/eden-emu/eden/src/branch/master/docs/policies/AI.md)
- [x] I have read and followed the [Coding Guidelines](https://git.eden-emu.dev/eden-emu/eden/src/branch/master/docs/policies/Coding.md) to the best of my ability.

-------------------

Reviewed-on: https://git.eden-emu.dev/eden-emu/eden/pulls/4512
Reviewed-by: CamilleLaVey <camillelavey99@gmail.com>
Reviewed-by: Shinmegumi <shinmegumi@eden-emu.dev>

.gitignore

@@ -7,6 +7,7 @@
 # Build directory
 /[Bb]uild*/
 doc-build/
+out/
 AppDir/
 uruntime
 

cpmfile.json

@@ -351,13 +351,13 @@
     },
     "zstd": {
         "find_args": "MODULE",
-        "hash": "cc5ad4b119a9c2ea57f0b71eeff01113bb506e0d17000159c5409cb8236d22e38c52d5e9e97e7947a4bf1b2dfc44b6c503ab2d9aedbd59458435c6a2849cb029",
+        "hash": "2b006c3c110c38bd2f5896b42d36a38931c8eae261f30552f9a9fd04458d218d8f1a1425f117fd12af545ed77042e4b2d1a1a9b6286c4d0b9d74ef79f20a8f08",
         "min_version": "1.5",
         "options": [
             "ZSTD_BUILD_SHARED OFF"
         ],
         "repo": "facebook/zstd",
         "source_subdir": "build/cmake",
-        "version": "b8d6101fba"
+        "version": "33618c8"
     }
 }

src/android/app/src/main/java/org/yuzu/yuzu_emu/features/settings/model/BooleanSetting.kt

@@ -16,7 +16,6 @@ enum class BooleanSetting(override val key: String) : AbstractBooleanSetting {
     RENDERER_USE_SPEED_LIMIT("use_speed_limit"),
     USE_CUSTOM_CPU_TICKS("use_custom_cpu_ticks"),
     SKIP_CPU_INNER_INVALIDATION("skip_cpu_inner_invalidation"),
-    NCE_INVALIDATION_GPU_READBACK("nce_invalidation_gpu_readback"),
     FIX_BLOOM_EFFECTS("fix_bloom_effects"),
     EMULATE_BGR565("emulate_bgr565"),
     RESCALE_HACK("rescale_hack"),

src/android/app/src/main/java/org/yuzu/yuzu_emu/features/settings/model/view/SettingsItem.kt

@@ -859,13 +859,6 @@ abstract class SettingsItem(
                     descriptionId = R.string.skip_cpu_inner_invalidation_description
                 )
             )
-            put(
-                SwitchSetting(
-                    BooleanSetting.NCE_INVALIDATION_GPU_READBACK,
-                    titleId = R.string.nce_invalidation_gpu_readback,
-                    descriptionId = R.string.nce_invalidation_gpu_readback_description
-                )
-            )
             put(
                 SwitchSetting(
                     BooleanSetting.FIX_BLOOM_EFFECTS,

src/android/app/src/main/java/org/yuzu/yuzu_emu/features/settings/ui/SettingsFragmentPresenter.kt

@@ -554,7 +554,6 @@ class SettingsFragmentPresenter(
             add(HeaderSetting(R.string.hacks))
 
             add(BooleanSetting.SKIP_CPU_INNER_INVALIDATION.key)
-            add(BooleanSetting.NCE_INVALIDATION_GPU_READBACK.key)
             add(BooleanSetting.FIX_BLOOM_EFFECTS.key)
             add(BooleanSetting.EMULATE_BGR565.key)
             add(BooleanSetting.RENDERER_ASYNCHRONOUS_SHADERS.key)

src/android/app/src/main/java/org/yuzu/yuzu_emu/model/Game.kt

@@ -100,7 +100,7 @@ class Game(
 
     companion object {
         val extensions: Set<String> = HashSet(
-            listOf("xci", "nsp", "nca", "nro")
+            listOf("xci", "nsp", "nca", "nro", "nsz", "xcz")
         )
     }
 }

src/android/app/src/main/java/org/yuzu/yuzu_emu/utils/GameHelper.kt

@@ -109,7 +109,7 @@ object GameHelper {
 
     // File extensions considered as external content, buuut should
     // be done better imo.
-    private val externalContentExtensions = setOf("nsp", "xci")
+    private val externalContentExtensions = setOf("nsp", "xci", "nsz", "xcz")
 
     private fun scanContentContainersRecursive(
         files: Array<MinimalDocumentFile>,

src/android/app/src/main/jni/game_metadata.cpp

@@ -76,7 +76,9 @@ jboolean Java_org_yuzu_yuzu_1emu_utils_GameMetadata_getIsValid(JNIEnv* env, jobj
             auto const file_type = loader->GetFileType();
             if (file_type == Loader::FileType::Unknown || file_type == Loader::FileType::Error)
                 return false;
-            if ((file_type == Loader::FileType::NSP || file_type == Loader::FileType::XCI) && !Loader::IsBootableGameContainer(file, file_type))
+            if ((file_type == Loader::FileType::NSP || file_type == Loader::FileType::XCI ||
+                 file_type == Loader::FileType::NSZ || file_type == Loader::FileType::XCZ) &&
+                !Loader::IsBootableGameContainer(file, file_type))
                 return false;
             u64 program_id = 0;
             return loader->ReadProgramId(program_id) == Loader::ResultStatus::Success;

src/android/app/src/main/res/values-ar/strings.xml

@@ -587,8 +587,6 @@
     <string name="fast_gpu_time_description">يجعل اللعبة تعتقد أن عمل وحدة معالجة الرسومات ينتهي بشكل أسرع مما هو عليه في الواقع، لذلك تتوقف عن خفض الدقة ومسافة العرض لتتناسب مع سرعات جهاز Switch.</string>
     <string name="skip_cpu_inner_invalidation">تخطي إبطال صلاحية وحدة المعالجة المركزية الداخلية</string>
     <string name="skip_cpu_inner_invalidation_description">يتخطى بعض عمليات إبطال ذاكرة التخزين المؤقتة من جانب وحدة المعالجة المركزية أثناء تحديثات الذاكرة، مما يقلل من استخدام وحدة المعالجة المركزية ويحسن أداءها. قد يتسبب ذلك في حدوث أعطال أو تعطل في بعض الألعاب.</string>
-    <string name="nce_invalidation_gpu_readback">إعادة قراءة وحدة معالجة الرسومات قبل إبطال صلاحية NCE</string>
-    <string name="nce_invalidation_gpu_readback_description">إصلاح مؤقت. يعالج مشكلة تعطل لعبة Ender Magnolia (وغيرها من ألعاب UE). قد يؤدي إلى انخفاض في الأداء. يجري العمل على تحسينه.</string>
     <string name="fix_bloom_effects">إصلاح تأثيرات التوهج</string>
     <string name="fix_bloom_effects_description">يقلل من ضبابية التوهج في LA/EOW  (Adreno A6XX - A7XX/ Turnip)، ويزيل التوهج في Burnout. تحذير: قد يسبب تشوهات رسومية في ألعاب أخرى.</string>
     <string name="emulate_bgr565">محاكاة BGR565</string>

src/android/app/src/main/res/values-zh-rTW/strings.xml

@@ -577,8 +577,6 @@
     <string name="fast_gpu_time_description">讓遊戲誤以為 GPU 工作能更快完成，避免遊戲為了配合 Switch 的時脈而降低解析度與渲染距離</string>
     <string name="skip_cpu_inner_invalidation">跳過 CPU 內部失效處理</string>
     <string name="skip_cpu_inner_invalidation_description">在記憶體更新期間跳過某些 CPU 端快取的失效處理，減少 CPU 使用率並提高其性能。可能會導致某些遊戲出現故障或當機</string>
-    <string name="nce_invalidation_gpu_readback">在 NCE 無效化前進行 GPU 回讀</string>
-    <string name="nce_invalidation_gpu_readback_description">（暫時性功能）修復 \"終結者瑪格諾莉亞：霧中綻放\" 及其它使用虛幻引擎之遊戲中的當機問題，可能會降低效能。此功能正在改善中</string>
     <string name="fix_bloom_effects">修正光暈特效</string>
     <string name="fix_bloom_effects_description">在 \"薩爾達傳說：織夢島\" 和 \"薩爾達傳說：智慧的再現\" 中減少光暈模糊的問題並移除 \"橫衝直撞：狂飆樂園重製版\" 中的光暈效果（適用於 Adreno A6XX - A7XX/ Turnip）。警告：可能會在其它遊戲中導致圖形異常</string>
     <string name="emulate_bgr565">模擬 BGR565</string>

src/android/app/src/main/res/values/strings.xml

@@ -593,8 +593,6 @@
     <string name="fast_gpu_time_description">Makes the game believe GPU work finishes faster than it does, so it stops lowering resolution and render distance to fit the Switch\'s clocks.</string>
     <string name="skip_cpu_inner_invalidation">Skip CPU Inner Invalidation</string>
     <string name="skip_cpu_inner_invalidation_description">Skips certain CPU-side cache invalidations during memory updates, reducing CPU usage and improving it\'s performance. This may cause glitches or crashes on some games.</string>
-    <string name="nce_invalidation_gpu_readback">GPU Readback Before NCE Invalidation</string>
-    <string name="nce_invalidation_gpu_readback_description">Temporary. Fix Ender Magnolia (and other UE titles) crashes. May reduce performance. Improvement in progress.</string>
     <string name="fix_bloom_effects">Fix Bloom Effects</string>
     <string name="fix_bloom_effects_description">Reduces bloom blur in LA/EOW (Adreno A6XX - A7XX/ Turnip), removes bloom in Burnout. Warning: may cause graphical artifacts in other games.</string>
     <string name="emulate_bgr565">Emulate BGR565</string>

src/common/settings.h

@@ -639,8 +639,6 @@ struct Values {
                                                         Specialization::Default,
                                                         true,
                                                         true};
-    SwitchableSetting<bool> nce_invalidation_gpu_readback{
-        linkage, false, "nce_invalidation_gpu_readback", Category::RendererHacks};
     SwitchableSetting<bool> async_presentation{linkage,
 #ifdef __ANDROID__
                                                true,

src/common/zstd_compression.cpp

@@ -1,3 +1,6 @@
+// SPDX-FileCopyrightText: Copyright 2026 Eden Emulator Project
+// SPDX-License-Identifier: GPL-3.0-or-later
+
 // SPDX-FileCopyrightText: Copyright 2019 yuzu Emulator Project
 // SPDX-License-Identifier: GPL-2.0-or-later
 
@@ -46,4 +49,10 @@ std::vector<u8> DecompressDataZSTD(std::span<const u8> compressed) {
     return decompressed;
 }
 
+std::size_t DecompressDataZSTD(std::span<const u8> compressed, std::span<u8> output) {
+    const std::size_t result =
+        ZSTD_decompress(output.data(), output.size(), compressed.data(), compressed.size());
+    return ZSTD_isError(result) ? 0 : result;
+}
+
 } // namespace Common::Compression

src/common/zstd_compression.h

@@ -1,3 +1,6 @@
+// SPDX-FileCopyrightText: Copyright 2026 Eden Emulator Project
+// SPDX-License-Identifier: GPL-3.0-or-later
+
 // SPDX-FileCopyrightText: Copyright 2019 yuzu Emulator Project
 // SPDX-License-Identifier: GPL-2.0-or-later
 
@@ -42,4 +45,15 @@ namespace Common::Compression {
  */
 [[nodiscard]] std::vector<u8> DecompressDataZSTD(std::span<const u8> compressed);
 
+/**
+ * Decompresses a single Zstandard frame into a caller-provided buffer of known size.
+ *
+ * @param compressed the compressed source memory region.
+ * @param output     the destination buffer. Its size is the maximum number of bytes to write.
+ *
+ * @return the number of bytes written, or 0 if the frame is invalid or doesn't fit. An empty
+ *         frame also returns 0.
+ */
+[[nodiscard]] std::size_t DecompressDataZSTD(std::span<const u8> compressed, std::span<u8> output);
+
 } // namespace Common::Compression

src/core/CMakeLists.txt

@@ -33,6 +33,8 @@ add_library(core STATIC
     crypto/encryption_layer.h
     crypto/key_manager.cpp
     crypto/key_manager.h
+    crypto/ncz_decompressor.cpp
+    crypto/ncz_decompressor.h
     crypto/partition_data_manager.cpp
     crypto/partition_data_manager.h
     crypto/xts_encryption_layer.cpp

src/core/arm/nce/arm_nce.cpp

@@ -369,8 +369,8 @@ void ArmNce::SetContext(const Kernel::Svc::ThreadContext& ctx) {
 }
 
 void ArmNce::SignalInterrupt(Kernel::KThread* thread) {
-    // Add break loop condition.
-    m_guest_ctx.esr_el1.fetch_or(static_cast<u64>(HaltReason::BreakLoop));
+    constexpr u64 break_loop = static_cast<u64>(HaltReason::BreakLoop);
+    if ((m_guest_ctx.esr_el1.fetch_or(break_loop) & break_loop) != 0) return;
 
     auto* params = &thread->GetNativeExecutionParameters();
     LockThreadParameters(params);

src/core/core.cpp

@@ -4,7 +4,6 @@
 #include <array>
 #include <atomic>
 #include <memory>
-#include <unordered_map>
 #include <utility>
 
 #include "game_settings.h"
@@ -249,30 +248,12 @@ struct System::Impl {
         }
     }
 
-    void NotifyNVDECChannelOpen(u64 process_id) {
-        std::scoped_lock lock{nvdec_active_mutex};
-        ++nvdec_active_channels[process_id];
-    }
-
-    void NotifyNVDECChannelClose(u64 process_id) {
-        std::scoped_lock lock{nvdec_active_mutex};
-        const auto it = nvdec_active_channels.find(process_id);
-        if (it == nvdec_active_channels.end()) {
-            return;
-        }
-        if (--it->second == 0) {
-            nvdec_active_channels.erase(it);
-        }
+    void SetNVDECActive(bool is_nvdec_active) {
+        nvdec_active = is_nvdec_active;
     }
 
     bool GetNVDECActive() {
-        std::scoped_lock lock{nvdec_active_mutex};
-        return !nvdec_active_channels.empty();
-    }
-
-    bool IsNVDECActiveForProcess(u64 process_id) {
-        std::scoped_lock lock{nvdec_active_mutex};
-        return nvdec_active_channels.contains(process_id);
+        return nvdec_active;
     }
 
     void InitializeDebugger(System& system, u16 port) {
@@ -528,8 +509,6 @@ struct System::Impl {
 
     mutable std::mutex suspend_guard;
     std::mutex general_channel_mutex;
-    std::mutex nvdec_active_mutex;
-    std::unordered_map<u64, u32> nvdec_active_channels;
     std::atomic_bool is_paused{};
     std::atomic_bool is_shutting_down{};
     std::atomic_bool is_powered_on{};
@@ -537,6 +516,7 @@ struct System::Impl {
     bool extended_memory_layout : 1 = false;
     bool exit_locked : 1 = false;
     bool exit_requested : 1 = false;
+    bool nvdec_active : 1 = false;
 
     void EnsureGeneralChannelInitialized(System& system) {
         if (!general_channel_event) {
@@ -600,22 +580,14 @@ void System::UnstallApplication() {
     impl->UnstallApplication();
 }
 
-void System::NotifyNVDECChannelOpen(u64 process_id) {
-    impl->NotifyNVDECChannelOpen(process_id);
-}
-
-void System::NotifyNVDECChannelClose(u64 process_id) {
-    impl->NotifyNVDECChannelClose(process_id);
+void System::SetNVDECActive(bool is_nvdec_active) {
+    impl->SetNVDECActive(is_nvdec_active);
 }
 
 bool System::GetNVDECActive() {
     return impl->GetNVDECActive();
 }
 
-bool System::IsNVDECActiveForProcess(u64 process_id) {
-    return impl->IsNVDECActiveForProcess(process_id);
-}
-
 void System::InitializeDebugger() {
     impl->InitializeDebugger(*this, Settings::values.gdbstub_port.GetValue());
 }

src/core/core.h

@@ -191,10 +191,8 @@ public:
     std::unique_lock<std::mutex> StallApplication();
     void UnstallApplication();
 
-    void NotifyNVDECChannelOpen(u64 process_id);
-    void NotifyNVDECChannelClose(u64 process_id);
+    void SetNVDECActive(bool is_nvdec_active);
     [[nodiscard]] bool GetNVDECActive();
-    [[nodiscard]] bool IsNVDECActiveForProcess(u64 process_id);
 
     /**
      * Initialize the debugger.

src/core/crypto/ncz_decompressor.cpp

@@ -0,0 +1,457 @@
+// SPDX-FileCopyrightText: Copyright 2026 Eden Emulator Project
+// SPDX-License-Identifier: GPL-3.0-or-later
+
+#include <algorithm>
+#include <array>
+#include <cstring>
+#include <memory>
+#include <mutex>
+#include <optional>
+#include <span>
+#include <list>
+#include <string>
+#include <vector>
+
+#include <zstd.h>
+
+#include "common/common_types.h"
+#include "common/logging.h"
+#include "common/zstd_compression.h"
+#include "core/crypto/aes_util.h"
+#include "core/crypto/key_manager.h"
+#include "core/crypto/ncz_decompressor.h"
+#include "core/file_sys/vfs/vfs.h"
+
+namespace Core::Crypto {
+
+namespace {
+
+constexpr u64 HeaderSize = 0x4000;
+constexpr std::size_t SectionWire = 0x40;
+constexpr std::size_t BlockFixed = 0x18;
+constexpr std::size_t Chunk = 0x100000;
+// Upper bound on NCZ section entries (64 bytes each, so the table is at most 4 MiB). A base NCA
+// has a handful of sections, but nsz stores an update's patch (BKTR) data one entry per
+// subsection, which is thousands of them.
+constexpr u64 MaxSectionCount = 0x10000;
+constexpr std::array<u8, 8> SectionsMagic{'N', 'C', 'Z', 'S', 'E', 'C', 'T', 'N'};
+constexpr std::array<u8, 8> BlockMagic{'N', 'C', 'Z', 'B', 'L', 'O', 'C', 'K'};
+
+struct Section {
+    s64 offset, size;
+    u64 crypto_type;
+    Key128 key, counter;
+    bool Ctr() const { return crypto_type == 3 || crypto_type == 4; }
+};
+
+struct BlockHeader {
+    u32 exponent;
+    u64 decompressed_size;
+    std::vector<u32> sizes;
+};
+
+u64 Le(const u8* p, std::size_t n) {
+    u64 v = 0;
+    for (std::size_t i = n; i-- > 0;) v = (v << 8) | p[i];
+    return v;
+}
+
+// Mirrors nsz AESCTR.seek(): upper 8 bytes = stored counter, lower 8 = absolute NCA offset / 16.
+void SetIv(Key128& iv, const Key128& base, u64 off) {
+    iv = base;
+    u64 blk = off >> 4;
+    for (std::size_t i = 0; i < 8; ++i) { iv[15 - i] = static_cast<u8>(blk & 0xFF); blk >>= 8; }
+}
+
+// Parse header + section table + optional block header. Returns payload_start.
+std::optional<std::pair<std::vector<Section>, std::optional<BlockHeader>>> ParseHeader(
+    const FileSys::VirtualFile& f, u64& payload_start) {
+    const u64 fsize = f->GetSize();
+    if (fsize < HeaderSize + 16) return std::nullopt;
+
+    const auto head = f->ReadBytes(16, HeaderSize);
+    if (head.size() != 16 || !std::equal(head.begin(), head.begin() + 8, SectionsMagic.begin()))
+        return std::nullopt;
+    const u64 count = Le(head.data() + 8, 8);
+    if (count == 0 || count > MaxSectionCount || count * SectionWire > fsize - HeaderSize - 16)
+        return std::nullopt;
+
+    const auto table = f->ReadBytes(count * SectionWire, HeaderSize + 16);
+    if (table.size() != count * SectionWire) return std::nullopt;
+
+    std::vector<Section> sections;
+    sections.reserve(count + 1);
+    for (u64 i = 0; i < count; ++i) {
+        const u8* e = table.data() + i * SectionWire;
+        Section s{};
+        s.offset = static_cast<s64>(Le(e, 8));
+        s.size = static_cast<s64>(Le(e + 8, 8));
+        s.crypto_type = Le(e + 16, 8);
+        std::memcpy(s.key.data(), e + 32, 16);
+        std::memcpy(s.counter.data(), e + 48, 16);
+        sections.push_back(s);
+    }
+
+    const u64 block_start = HeaderSize + 16 + count * SectionWire;
+    const auto fixed = f->ReadBytes(BlockFixed, block_start);
+    std::optional<BlockHeader> bh;
+    if (fixed.size() == BlockFixed &&
+        std::equal(fixed.begin(), fixed.begin() + 8, BlockMagic.begin())) {
+        BlockHeader b{};
+        b.exponent = fixed[11];
+        const s32 n = static_cast<s32>(Le(fixed.data() + 12, 4));
+        if (n <= 0) return std::nullopt;
+        b.decompressed_size = Le(fixed.data() + 16, 8);
+        const auto sizes = f->ReadBytes(static_cast<u64>(n) * 4, block_start + BlockFixed);
+        if (sizes.size() != static_cast<u64>(n) * 4) return std::nullopt;
+        b.sizes.resize(n);
+        for (s32 i = 0; i < n; ++i) b.sizes[i] = static_cast<u32>(Le(sizes.data() + i * 4, 4));
+        payload_start = block_start + BlockFixed + static_cast<u64>(n) * 4;
+        bh = std::move(b);
+    } else {
+        payload_start = block_start;
+    }
+
+    if (!sections.empty() && sections[0].offset > static_cast<s64>(HeaderSize)) {
+        Section fake{};
+        fake.offset = static_cast<s64>(HeaderSize);
+        fake.size = sections[0].offset - static_cast<s64>(HeaderSize);
+        sections.insert(sections.begin(), fake);
+    }
+    return std::make_pair(std::move(sections), std::move(bh));
+}
+
+// Sequential reader over independently-compressed blocks. Keeps a small LRU cache of recently
+// decoded blocks so random reads (e.g. update NCA with thousands of sections) don't re-decompress
+// the same block over and over.
+class BlockReader {
+public:
+    static constexpr std::size_t BlockCacheSize = 8;
+
+    BlockReader(FileSys::VirtualFile f, const BlockHeader& bh, u64 payload_start)
+        : source{std::move(f)}, block_size{1ULL << bh.exponent}, sizes{bh.sizes},
+          decompressed_size{bh.decompressed_size} {
+        if (bh.exponent < 14 || bh.exponent > 32 || sizes.empty())
+            throw std::runtime_error("bad NCZBLOCK header");
+        u64 off = payload_start;
+        offsets.reserve(sizes.size());
+        for (u32 s : sizes) { offsets.push_back(off); off += s; }
+    }
+
+    // Fills `out` from `position`; returns bytes written.
+    std::size_t ReadAt(u64 position, std::span<u8> out) {
+        std::size_t done = 0;
+        while (done < out.size() && position < decompressed_size) {
+            const u64 id = position / block_size;
+            if (id >= sizes.size()) break;
+            const std::vector<u8>* block = GetBlock(id);
+            if (block == nullptr) break;
+            const u64 in = position - id * block_size;
+            if (in >= block->size()) break;
+            const std::size_t n = static_cast<std::size_t>(
+                std::min<u64>(out.size() - done, block->size() - in));
+            std::memcpy(out.data() + done, block->data() + in, n);
+            done += n;
+            position += n;
+        }
+        return done;
+    }
+
+private:
+    const std::vector<u8>* GetBlock(u64 id) {
+        // LRU hit: move to front.
+        for (auto it = cache.begin(); it != cache.end(); ++it) {
+            if (it->first == id) {
+                cache.splice(cache.begin(), cache, it);
+                return &cache.front().second;
+            }
+        }
+
+        // Miss: decode the block.
+        std::vector<u8> buffer;
+        if (cache.size() >= BlockCacheSize) {
+            buffer = std::move(cache.back().second);
+            cache.pop_back();
+        }
+        if (!DecodeBlock(id, buffer)) {
+            return nullptr;
+        }
+        cache.emplace_front(id, std::move(buffer));
+        return &cache.front().second;
+    }
+
+    bool DecodeBlock(u64 id, std::vector<u8>& buffer) {
+        u64 want = block_size;
+        if (id + 1 == sizes.size()) {
+            const u64 rem = decompressed_size % block_size;
+            if (rem > 0) want = rem; // avoid the 0-size bug from the original nsz
+        }
+        try { buffer.resize(static_cast<std::size_t>(want)); }
+        catch (const std::bad_alloc&) { return false; }
+        const u32 csize = sizes[id];
+        if (csize < want) {
+            const auto comp = source->ReadBytes(csize, offsets[id]);
+            if (comp.size() != csize ||
+                Common::Compression::DecompressDataZSTD(comp, buffer) != want) return false;
+        } else if (source->Read(buffer.data(), buffer.size(), offsets[id]) != buffer.size()) {
+            return false;
+        }
+        return true;
+    }
+
+    FileSys::VirtualFile source;
+    u64 block_size;
+    std::vector<u64> offsets;
+    std::vector<u32> sizes;
+    u64 decompressed_size;
+    std::list<std::pair<u64, std::vector<u8>>> cache;  // LRU: front = most recent
+};
+
+class BlockNczFile final : public FileSys::VfsFile {
+public:
+    BlockNczFile(std::string name_, std::vector<u8> hdr_, std::vector<Section> sections_,
+                 std::unique_ptr<BlockReader> body_)
+        : name{std::move(name_)}, hdr{std::move(hdr_)}, sections{std::move(sections_)},
+          body{std::move(body_)} {
+        nca_size = HeaderSize;
+        for (const auto& s : sections)
+            nca_size = std::max<u64>(nca_size, static_cast<u64>(s.offset + s.size));
+        ciphers.reserve(sections.size());
+        for (const auto& s : sections)
+            ciphers.push_back(s.Ctr() ? std::make_unique<AESCipher<Key128>>(s.key, Mode::CTR)
+                                      : nullptr);
+    }
+
+    std::string GetName() const override { return name; }
+    std::size_t GetSize() const override { return static_cast<std::size_t>(nca_size); }
+    bool Resize(std::size_t) override { return false; }
+    FileSys::VirtualDir GetContainingDirectory() const override { return nullptr; }
+    bool IsWritable() const override { return false; }
+    bool IsReadable() const override { return true; }
+
+    std::size_t Read(u8* data, std::size_t len, std::size_t off) const override {
+        if (off >= nca_size) return 0;
+        len = static_cast<std::size_t>(std::min<u64>(len, nca_size - off));
+        const u64 end = off + len;
+        u64 pos = off;
+        if (pos < HeaderSize) {
+            const u64 n = std::min<u64>(end, HeaderSize) - pos;
+            std::memcpy(data, hdr.data() + pos, static_cast<std::size_t>(n));
+            pos += n;
+        }
+        std::scoped_lock lock{mutex};
+        for (std::size_t i = 0; i < sections.size() && pos < end; ++i) {
+            const auto& s = sections[i];
+            const u64 begin = std::max<u64>(static_cast<u64>(s.offset), HeaderSize);
+            const u64 stop = std::max<u64>(static_cast<u64>(s.offset + s.size), begin);
+            if (stop <= pos) continue;
+            if (pos < begin) {
+                const u64 g = std::min(end, begin);
+                std::memset(data + (pos - off), 0, static_cast<std::size_t>(g - pos));
+                pos = g;
+                if (pos >= end) break;
+            }
+            const u64 pe = std::min(end, stop);
+            const std::span<u8> piece(data + (pos - off), static_cast<std::size_t>(pe - pos));
+            if (body->ReadAt(pos - HeaderSize, piece) != piece.size())
+                return static_cast<std::size_t>(pos - off);
+            if (ciphers[i]) {
+                Key128 iv{};
+                SetIv(iv, s.counter, pos);
+                ciphers[i]->SetIV(iv);
+                ciphers[i]->Transcode(piece.data(), piece.size(), piece.data(), Op::Encrypt);
+            }
+            pos = pe;
+        }
+        return len;
+    }
+
+    std::size_t Write(const u8*, std::size_t, std::size_t) override { return 0; }
+    bool Rename(std::string_view) override { return false; }
+
+private:
+    const std::string name;
+    const std::vector<u8> hdr;
+    const std::vector<Section> sections;
+    const std::unique_ptr<BlockReader> body;
+    std::vector<std::unique_ptr<AESCipher<Key128>>> ciphers;
+    mutable std::mutex mutex;
+    u64 nca_size{};
+};
+
+// Solid NCZ: one zstd stream over the whole body. Decoded incrementally into memory, up to the
+// requested offset, so sequential install reads are O(n) instead of O(n^2).
+class SolidNczFile final : public FileSys::VfsFile {
+public:
+    SolidNczFile(std::string name_, FileSys::VirtualFile source_,
+                 std::vector<Section> sections_, u64 payload_start_)
+        : name{std::move(name_)}, source{std::move(source_)}, sections{std::move(sections_)},
+          payload_start{payload_start_} {
+        const auto raw_header = source->ReadBytes(HeaderSize, 0);
+        data.assign(raw_header.begin(), raw_header.end());
+        frontier = HeaderSize;
+    }
+
+    ~SolidNczFile() override {
+        if (stream != nullptr) ZSTD_freeDStream(stream);
+    }
+
+    std::string GetName() const override { return name; }
+    std::size_t GetSize() const override {
+        u64 size = HeaderSize;
+        for (const auto& s : sections)
+            size = std::max<u64>(size, static_cast<u64>(s.offset + s.size));
+        return static_cast<std::size_t>(size);
+    }
+    bool Resize(std::size_t) override { return false; }
+    FileSys::VirtualDir GetContainingDirectory() const override { return nullptr; }
+    bool IsWritable() const override { return false; }
+    bool IsReadable() const override { return true; }
+
+    std::size_t Read(u8* out, std::size_t len, std::size_t off) const override {
+        std::scoped_lock lock{mutex};
+        const u64 nca_size = GetSize();
+        if (off >= nca_size) return 0;
+        len = static_cast<std::size_t>(std::min<u64>(len, nca_size - off));
+        const u64 needed = off + len;
+        if (needed > frontier && !DecodeTo(needed)) return 0;
+        std::memcpy(out, data.data() + off, len);
+        return len;
+    }
+
+    std::size_t Write(const u8*, std::size_t, std::size_t) override { return 0; }
+    bool Rename(std::string_view) override { return false; }
+
+private:
+    bool DecodeTo(u64 needed) const {
+        if (stream == nullptr) {
+            stream = ZSTD_createDStream();
+            if (stream == nullptr) return false;
+            ZSTD_DCtx_setParameter(stream, ZSTD_d_windowLogMax, 31);
+            read_pos = payload_start;
+        }
+
+        const u64 file_size = source->GetSize();
+        std::vector<u8> out_buf(Chunk);
+
+        std::size_t index = 0;
+        for (; index < sections.size(); ++index) {
+            const auto& s = sections[index];
+            const u64 begin = std::max<u64>(static_cast<u64>(s.offset), HeaderSize);
+            const u64 end = std::max<u64>(static_cast<u64>(s.offset) + static_cast<u64>(s.size),
+                                          begin);
+            if (end > frontier) break;
+        }
+
+        for (; index < sections.size() && frontier < needed; ++index) {
+            const auto& s = sections[index];
+            const u64 begin = std::max<u64>(static_cast<u64>(s.offset), HeaderSize);
+            const u64 end = std::max<u64>(static_cast<u64>(s.offset) + static_cast<u64>(s.size),
+                                          begin);
+            if (end <= frontier) continue;
+            u64 pos = std::max<u64>(begin, frontier);
+
+            std::optional<AESCipher<Key128>> cipher;
+            if (s.Ctr()) cipher.emplace(s.key, Mode::CTR);
+
+            while (pos < end && frontier < needed) {
+                const u64 want = std::min<u64>(end - pos, out_buf.size());
+                std::size_t produced = 0;
+                while (produced < want) {
+                    if (in.pos >= in.size) {
+                        const u64 remaining = file_size - read_pos;
+                        const std::size_t chunk = static_cast<std::size_t>(
+                            std::min<u64>(Chunk, remaining));
+                        if (chunk == 0) { LOG_ERROR(Crypto, "Solid NCZ: truncated"); return false; }
+                        const auto bytes = source->ReadBytes(chunk, read_pos);
+                        if (bytes.empty()) { LOG_ERROR(Crypto, "Solid NCZ: truncated"); return false; }
+                        read_pos += bytes.size();
+                        in_buf.assign(bytes.begin(), bytes.end());
+                        in.src = in_buf.data();
+                        in.size = in_buf.size();
+                        in.pos = 0;
+                    }
+                    ZSTD_outBuffer outb{out_buf.data() + produced, want - produced, 0};
+                    const std::size_t res = ZSTD_decompressStream(stream, &outb, &in);
+                    if (ZSTD_isError(res)) {
+                        LOG_ERROR(Crypto, "Solid NCZ: zstd error");
+                        return false;
+                    }
+                    produced += outb.pos;
+                }
+                if (cipher) {
+                    Key128 iv{};
+                    SetIv(iv, s.counter, pos);
+                    cipher->SetIV(iv);
+                    cipher->Transcode(out_buf.data(), produced, out_buf.data(), Op::Encrypt);
+                }
+                data.insert(data.end(), out_buf.begin(), out_buf.begin() + produced);
+                pos += produced;
+                frontier = pos;
+            }
+        }
+        return frontier >= needed;
+    }
+
+    const std::string name;
+    const FileSys::VirtualFile source;
+    const std::vector<Section> sections;
+    const u64 payload_start;
+    mutable std::mutex mutex;
+    mutable std::vector<u8> data;
+    mutable ZSTD_DStream* stream = nullptr;
+    mutable u64 read_pos = 0;
+    mutable u64 frontier = 0;
+    mutable std::vector<u8> in_buf;
+    mutable ZSTD_inBuffer in{nullptr, 0, 0};
+};
+
+// Shared opener: block path always; solid only when allow_solid is true.
+FileSys::VirtualFile OpenNcz(const FileSys::VirtualFile& f, std::string_view name,
+                             bool allow_solid) {
+    if (!f) return nullptr;
+    u64 ps = 0;
+    auto parsed = ParseHeader(f, ps);
+    if (!parsed) return nullptr;
+    auto& [sections, block] = *parsed;
+
+    if (block) {
+        auto hdr = f->ReadBytes(HeaderSize, 0);
+        if (hdr.size() != HeaderSize) return nullptr;
+        std::unique_ptr<BlockReader> body;
+        try { body = std::make_unique<BlockReader>(f, *block, ps); }
+        catch (const std::exception& e) {
+            LOG_ERROR(Crypto, "Failed to open NCZBLOCK: {}", e.what());
+            return nullptr;
+        }
+        return std::make_shared<BlockNczFile>(std::string{name}, std::move(hdr),
+                                              std::move(sections), std::move(body));
+    }
+
+    if (allow_solid) {
+        return std::make_shared<SolidNczFile>(std::string{name}, f, std::move(sections), ps);
+    }
+
+    LOG_DEBUG(Crypto, "Solid NCZ {} skipped (only NCZBLOCK is supported for play)", name);
+    return nullptr;
+}
+
+} // Anonymous namespace
+
+bool IsBlockCompressedNCZ(const FileSys::VirtualFile& f) {
+    if (!f) return false;
+    u64 ps = 0;
+    const auto p = ParseHeader(f, ps);
+    return p && p->second.has_value();
+}
+
+FileSys::VirtualFile DecompressNCZ(const FileSys::VirtualFile& f, std::string_view name) {
+    return OpenNcz(f, name, /*allow_solid=*/false);
+}
+
+FileSys::VirtualFile DecompressNCZForInstall(const FileSys::VirtualFile& f,
+                                             std::string_view name) {
+    return OpenNcz(f, name, /*allow_solid=*/true);
+}
+
+} // namespace Core::Crypto

src/core/crypto/ncz_decompressor.h

@@ -0,0 +1,35 @@
+// SPDX-FileCopyrightText: Copyright 2026 Eden Emulator Project
+// SPDX-License-Identifier: GPL-3.0-or-later
+
+#pragma once
+
+#include <string_view>
+
+#include "core/file_sys/vfs/vfs_types.h"
+
+namespace Core::Crypto {
+
+/**
+ * Opens an NCZBLOCK-compressed NCA (nsz -B / .xcz block format) as a VirtualFile backed by
+ * on-demand block decoding. Nothing is written to disk; each read decodes only the blocks it
+ * covers and re-encrypts CTR sections. Solid NCZ is not supported.
+ *
+ * @param ncz_file    The raw NCZ payload (e.g. "<id>.ncz" entry from an NSZ/XCZ PFS0).
+ * @param output_name Name to give the returned VirtualFile (e.g. "<id>.nca").
+ *
+ * @return a VirtualFile for the NCA, or nullptr if the NCZ is malformed or not NCZBLOCK.
+ */
+[[nodiscard]] FileSys::VirtualFile DecompressNCZ(const FileSys::VirtualFile& ncz_file,
+                                                 std::string_view output_name);
+
+/**
+ * Returns true if the given file is an NCZ payload in block-compressed (NCZBLOCK) form,
+ * i.e. one that DecompressNCZ() can open. Solid NCZ returns false.
+ */
+[[nodiscard]] bool IsBlockCompressedNCZ(const FileSys::VirtualFile& ncz_file);
+
+// Install to NAND: same as DecompressNCZ, but solid NCZ is decoded on demand too.
+[[nodiscard]] FileSys::VirtualFile DecompressNCZForInstall(
+    const FileSys::VirtualFile& ncz_file, std::string_view output_name);
+
+} // namespace Core::Crypto

src/core/file_sys/content_archive.cpp

@@ -45,6 +45,7 @@ NCA::NCA(VirtualFile file_, const NCA* base_nca, bool allow_missing_base)
     // Ensure we have the proper key area keys to continue.
     const u8 master_key_id = MasterKeyIdForKeyGeneration(reader->GetKeyGeneration());
     if (!keys.HasKey(Core::Crypto::S128KeyType::KeyArea, master_key_id, reader->GetKeyIndex())) {
+        LOG_ERROR(Loader, "Cannot decrypt {}: keys for gen 0x{:02X}+ required.", file->GetName(), reader->GetKeyGeneration());
         status = Loader::ResultStatus::ErrorMissingKeyAreaKey;
         return;
     }

src/core/file_sys/registered_cache.cpp

@@ -16,6 +16,7 @@
 #include "common/random.h"
 #include "common/string_util.h"
 #include "core/crypto/key_manager.h"
+#include "core/crypto/ncz_decompressor.h"
 #include "core/file_sys/card_image.h"
 #include "core/file_sys/common_funcs.h"
 #include "core/file_sys/content_archive.h"
@@ -158,6 +159,25 @@ static std::shared_ptr<NSP> OpenContainerAsNsp(const VirtualFile& file, Loader::
     return nullptr;
 }
 
+// Checks that every .ncz entry in the container is NCZBLOCK. Solid NCZ has no random access,
+// and the whole point of this build is to reject it at scan time so such files never appear
+// in the game list. Containers without any .ncz (plain NSP/XCI) pass through.
+static bool ContainerIsFullyBlockCompressed(const std::shared_ptr<NSP>& nsp) {
+    if (!nsp) {
+        return false;
+    }
+    for (const auto& entry : nsp->GetFiles()) {
+        const auto name = entry->GetName();
+        if (name.size() < 4 || name.substr(name.size() - 4) != ".ncz") {
+            continue;
+        }
+        if (!Core::Crypto::IsBlockCompressedNCZ(entry)) {
+            return false; // solid NCZ -> reject the whole container
+        }
+    }
+    return true; // no .ncz at all, or all block-compressed
+}
+
 template <typename Callback>
 bool ForEachContainerEntry(const std::shared_ptr<NSP>& nsp, bool only_content,
                            std::optional<u64> base_program_id, Callback&& on_entry) {
@@ -811,7 +831,14 @@ std::vector<ContentProviderEntry> RegisteredCache::ListEntriesFilter(
 }
 
 static std::shared_ptr<NCA> GetNCAFromNSPForID(const NSP& nsp, const NcaID& id) {
-    auto file = nsp.GetFile(fmt::format("{}.nca", Common::HexToString(id, false)));
+    const auto id_string = Common::HexToString(id, false);
+    auto file = nsp.GetFile(fmt::format("{}.nca", id_string));
+    if (file == nullptr) {
+        if (auto ncz_file = nsp.GetFile(fmt::format("{}.ncz", id_string))) {
+            file = Core::Crypto::DecompressNCZForInstall(
+                ncz_file, fmt::format("{}.nca", id_string));
+        }
+    }
     if (file == nullptr) {
         return nullptr;
     }
@@ -1273,6 +1300,12 @@ bool ManualContentProvider::AddEntriesFromContainer(VirtualFile file, bool only_
         return false;
     }
 
+    // Reject solid NCZ (.nsz without -B): only block-compressed NCZ is supported.
+    if (!ContainerIsFullyBlockCompressed(nsp)) {
+        LOG_DEBUG(Service_FS, "Skipping solid-compressed container: {}", file->GetName());
+        return false;
+    }
+
     return ForEachContainerEntry(
         nsp, only_content, base_program_id,
         [this](TitleType title_type, ContentRecordType content_type, u64 title_id,
@@ -1415,8 +1448,9 @@ void ExternalContentProvider::ScanDirectory(const VirtualDir& dir) {
         }
 
         const auto extension = Common::ToLower(filename.substr(dot_pos + 1));
-
-        if (extension == "nsp") {
+        // .xcz is an XCI container, but its updates only work through the NSP/PFS0 path
+        // (DecompressNCZ), not ProcessXCI. So route .xcz here on purpose.
+        if (extension == "nsp" || extension == "nsz" || extension == "xcz") {
             ProcessNSP(file);
         } else if (extension == "xci") {
             ProcessXCI(file);
@@ -1434,6 +1468,12 @@ void ExternalContentProvider::ProcessNSP(const VirtualFile& file) {
         return;
     }
 
+    // Reject solid NCZ (.nsz without -B): Eden only supports block-compressed NCZ.
+    if (!ContainerIsFullyBlockCompressed(nsp)) {
+        LOG_DEBUG(Service_FS, "Skipping solid-compressed container: {}", file->GetName());
+        return;
+    }
+
     LOG_DEBUG(Service_FS, "Processing NSP file: {}", file->GetName());
     AddExternalEntriesFromContainer(nsp, entries, versions, multi_version_entries);
 }

src/core/file_sys/submission_package.cpp

@@ -12,6 +12,7 @@
 #include "common/hex_util.h"
 #include "common/logging.h"
 #include "core/crypto/key_manager.h"
+#include "core/crypto/ncz_decompressor.h"
 #include "core/file_sys/common_funcs.h"
 #include "core/file_sys/content_archive.h"
 #include "core/file_sys/nca_metadata.h"
@@ -246,6 +247,17 @@ void NSP::ReadNCAs(const std::vector<VirtualFile>& files) {
                 const auto id_string = Common::HexToString(rec.nca_id, false);
                 auto next_file = pfs->GetFile(fmt::format("{}.nca", id_string));
 
+                if (next_file == nullptr) {
+                    auto ncz_file = pfs->GetFile(fmt::format("{}.ncz", id_string));
+                    if (ncz_file != nullptr) {
+                        next_file = Core::Crypto::DecompressNCZ(ncz_file, fmt::format("{}.nca", id_string));
+                        if (next_file == nullptr) {
+                            LOG_DEBUG(Service_FS, "Skipping solid or unsupported NCZ content {}.ncz", id_string);
+                            continue;
+                        }
+                    }
+                }
+
                 if (next_file == nullptr) {
                     if (rec.type != ContentRecordType::DeltaFragment) {
                         LOG_WARNING(Service_FS,

src/core/hle/service/filesystem/fsp/fs_i_storage.cpp

@@ -4,6 +4,10 @@
 // SPDX-FileCopyrightText: Copyright 2023 yuzu Emulator Project
 // SPDX-License-Identifier: GPL-2.0-or-later
 
+#include <algorithm>
+#include <chrono>
+#include <thread>
+
 #include "core/file_sys/errors.h"
 #include "core/hle/service/cmif_serialization.h"
 #include "core/hle/service/filesystem/fsp/fs_i_storage.h"
@@ -31,6 +35,14 @@ Result IStorage::Read(
     R_UNLESS(length >= 0, FileSys::ResultInvalidSize);
     R_UNLESS(offset >= 0, FileSys::ResultInvalidOffset);
 
+    const auto now = std::chrono::steady_clock::now();
+    if (now < next_read_tick) {
+        std::this_thread::sleep_for(next_read_tick - now);
+    }
+    next_read_tick = std::chrono::steady_clock::now() +
+                     std::chrono::nanoseconds{std::min(length, MaxChargedReadSize) *
+                                              1'000'000'000 / ReadBytesPerSecond};
+
     // Read the data from the Storage backend
     backend->Read(out_bytes.data(), length, offset);
 

src/core/hle/service/filesystem/fsp/fs_i_storage.h

@@ -6,6 +6,8 @@
 
 #pragma once
 
+#include <chrono>
+
 #include "core/file_sys/vfs/vfs.h"
 #include "core/hle/service/cmif_types.h"
 #include "core/hle/service/filesystem/filesystem.h"
@@ -20,6 +22,10 @@ public:
 private:
     FileSys::VirtualFile backend;
 
+    static constexpr s64 ReadBytesPerSecond = 40'000'000;
+    static constexpr s64 MaxChargedReadSize = 64 * 1024;
+    std::chrono::steady_clock::time_point next_read_tick{};
+
     Result Read(
         OutBuffer<BufferAttr_HipcMapAlias | BufferAttr_HipcMapTransferAllowsNonSecure> out_bytes,
         s64 offset, s64 length);

src/core/hle/service/nvdrv/devices/nvhost_nvdec.cpp

@@ -8,7 +8,6 @@
 #include "common/assert.h"
 #include "common/logging.h"
 #include "core/core.h"
-#include "core/hle/kernel/k_process.h"
 #include "core/hle/service/nvdrv/core/container.h"
 #include "core/hle/service/nvdrv/devices/ioctl_serialization.h"
 #include "core/hle/service/nvdrv/devices/nvhost_nvdec.h"
@@ -72,23 +71,17 @@ NvResult nvhost_nvdec::Ioctl3(DeviceFD fd, Ioctl command, std::span<const u8> in
 
 void nvhost_nvdec::OnOpen(NvCore::SessionId session_id, DeviceFD fd) {
     LOG_INFO(Service_NVDRV, "NVDEC video stream started");
+    system.SetNVDECActive(true);
     sessions[fd] = session_id;
-    if (const auto* session = core.GetSession(session_id);
-        session != nullptr && session->process != nullptr) {
-        system.NotifyNVDECChannelOpen(session->process->GetId());
-    }
     host1x.StartDevice(fd, Tegra::Host1x::ChannelType::NvDec, channel_syncpoint);
 }
 
 void nvhost_nvdec::OnClose(DeviceFD fd) {
     LOG_INFO(Service_NVDRV, "NVDEC video stream ended");
     host1x.StopDevice(fd, Tegra::Host1x::ChannelType::NvDec);
+    system.SetNVDECActive(false);
     auto it = sessions.find(fd);
     if (it != sessions.end()) {
-        if (const auto* session = core.GetSession(it->second);
-            session != nullptr && session->process != nullptr) {
-            system.NotifyNVDECChannelClose(session->process->GetId());
-        }
         sessions.erase(it);
     }
 }

src/core/hle/service/service.cpp

@@ -4,16 +4,12 @@
 // SPDX-FileCopyrightText: Copyright 2018 yuzu Emulator Project
 // SPDX-License-Identifier: GPL-2.0-or-later
 
-#include <chrono>
 #include <fmt/ranges.h>
-#include <string_view>
-#include <thread>
 #include "common/assert.h"
 #include "common/logging.h"
 #include "common/settings.h"
 #include "core/core.h"
 #include "core/hle/ipc.h"
-#include "core/hle/kernel/k_process.h"
 #include "core/hle/kernel/kernel.h"
 #include "core/hle/service/ipc_helpers.h"
 #include "core/hle/service/service.h"
@@ -37,7 +33,6 @@ ServiceFrameworkBase::ServiceFrameworkBase(Core::System& system_, const char* se
     : SessionRequestHandler(system_.Kernel(), service_name_)
     , system{system_}
     , service_name{service_name_}
-    , is_i_storage{std::string_view{service_name_} == "IStorage"}
     , handler_invoker{handler_invoker_}
     , max_sessions{max_sessions_}
 {}
@@ -82,22 +77,13 @@ void ServiceFrameworkBase::ReportUnimplementedFunction(HLERequestContext& ctx,
 }
 
 void ServiceFrameworkBase::InvokeRequest(HLERequestContext& ctx) {
-    const auto command = ctx.GetCommand();
-    auto it = handlers.find(command);
-    const bool is_cmd_read = command == 0;
+    auto it = handlers.find(ctx.GetCommand());
     FunctionInfoBase const* info = it == handlers.end() ? nullptr : &it->second;
     if (info == nullptr || info->handler_callback == nullptr)
         return ReportUnimplementedFunction(ctx, info);
 
     LOG_TRACE(Service, "{}", MakeFunctionString(info->name, GetServiceName(), ctx.CommandBuffer()));
     handler_invoker(this, info->handler_callback, ctx);
-
-    if (is_i_storage && is_cmd_read) {
-        const auto* const process = ctx.GetThread().GetOwnerProcess();
-        if (process != nullptr && system.IsNVDECActiveForProcess(process->GetId())) {
-            std::this_thread::sleep_for(std::chrono::microseconds{600});
-        }
-    }
 }
 
 void ServiceFrameworkBase::InvokeRequestTipc(HLERequestContext& ctx) {

src/core/hle/service/service.h

@@ -107,8 +107,6 @@ protected:
     Core::System& system;
     /// Identifier string used to connect to the service.
     const char* service_name;
-    /// Whether this is the IStorage service.
-    const bool is_i_storage;
     /// Function used to safely up-cast pointers to the derived class before invoking a handler.
     InvokerFn* handler_invoker;
     /// Maximum number of concurrent sessions that this service can handle.

src/core/loader/loader.cpp

@@ -82,23 +82,27 @@ std::shared_ptr<FileSys::NSP> OpenContainerAsNsp(FileSys::VirtualFile file, File
         return nullptr;
     }
 
-    if (type == FileType::NSP) {
+    if (type == FileType::NSP || type == FileType::NSZ) {
         auto nsp = std::make_shared<FileSys::NSP>(file, program_id, program_index);
         return nsp->GetStatus() == ResultStatus::Success ? nsp : nullptr;
     }
 
-    if (type == FileType::XCI) {
+    if (type == FileType::XCI || type == FileType::XCZ) {
         FileSys::XCI xci{file, program_id, program_index};
-        if (xci.GetStatus() != ResultStatus::Success) {
-            return nullptr;
+        if (xci.GetStatus() == ResultStatus::Success) {
+            auto secure_nsp = xci.GetSecurePartitionNSP();
+            if (secure_nsp != nullptr && secure_nsp->GetStatus() == ResultStatus::Success) {
+                return secure_nsp;
+            }
         }
 
-        auto secure_nsp = xci.GetSecurePartitionNSP();
-        if (secure_nsp == nullptr || secure_nsp->GetStatus() != ResultStatus::Success) {
-            return nullptr;
+        // PFS0-shaped .xcz (NSZ-style container) does not parse as XCI; fall back to NSP.
+        if (type == FileType::XCZ) {
+            auto nsp = std::make_shared<FileSys::NSP>(file, program_id, program_index);
+            return nsp->GetStatus() == ResultStatus::Success ? nsp : nullptr;
         }
 
-        return secure_nsp;
+        return nullptr;
     }
 
     return nullptr;
@@ -144,7 +148,8 @@ FileType IdentifyFile(FileSys::VirtualFile file) {
 }
 
 bool IsContainerType(FileType type) {
-    return type == FileType::NSP || type == FileType::XCI;
+    return type == FileType::NSP || type == FileType::NSZ || type == FileType::XCI ||
+           type == FileType::XCZ;
 }
 
 bool IsBootableGameContainer(FileSys::VirtualFile file, FileType type, u64 program_id,
@@ -180,8 +185,12 @@ FileType GuessFromFilename(const std::string& name) {
         return FileType::NCA;
     else if (extension == "xci")
         return FileType::XCI;
+    else if (extension == "xcz")
+        return FileType::XCZ;
     else if (extension == "nsp")
         return FileType::NSP;
+    else if (extension == "nsz")
+        return FileType::NSZ;
     else if (extension == "kip")
         return FileType::KIP;
     return FileType::Unknown;
@@ -197,10 +206,14 @@ std::string GetFileTypeString(FileType type) {
         return "NCA";
     case FileType::XCI:
         return "XCI";
+    case FileType::XCZ:
+        return "XCZ";
     case FileType::NAX:
         return "NAX";
     case FileType::NSP:
         return "NSP";
+    case FileType::NSZ:
+        return "NSZ";
     case FileType::KIP:
         return "KIP";
     case FileType::DeconstructedRomDirectory:
@@ -326,12 +339,29 @@ static std::unique_ptr<AppLoader> GetFileLoader(Core::System& system, FileSys::V
                                                system.GetContentProvider(), program_id,
                                                program_index);
 
+    // NCZ-compressed XCI container. XCI-shaped .xcz goes to the XCI loader, while
+    // PFS0-shaped (NSZ-style) .xcz containers fall back to the NSP loader.
+    case FileType::XCZ: {
+        FileSys::XCI probe{file, program_id, program_index};
+        if (probe.GetStatus() == ResultStatus::Success &&
+            probe.GetSecurePartitionNSP() != nullptr) {
+            return std::make_unique<AppLoader_XCI>(std::move(file),
+                                                   system.GetFileSystemController(),
+                                                   system.GetContentProvider(), program_id,
+                                                   program_index);
+        }
+        return std::make_unique<AppLoader_NSP>(std::move(file), system.GetFileSystemController(),
+                                               system.GetContentProvider(), program_id,
+                                               program_index);
+    }
+
     // NX NAX (NintendoAesXts) file format.
     case FileType::NAX:
         return std::make_unique<AppLoader_NAX>(std::move(file));
 
     // NX NSP (Nintendo Submission Package) file format
     case FileType::NSP:
+    case FileType::NSZ:
         return std::make_unique<AppLoader_NSP>(std::move(file), system.GetFileSystemController(),
                                                system.GetContentProvider(), program_id,
                                                program_index);

src/core/loader/loader.h

@@ -42,7 +42,9 @@ enum class FileType {
     NRO,
     NCA,
     NSP,
+    NSZ,
     XCI,
+    XCZ,
     NAX,
     KIP,
     DeconstructedRomDirectory,

src/core/loader/nsp.cpp

@@ -56,6 +56,17 @@ AppLoader_NSP::AppLoader_NSP(FileSys::VirtualFile file_,
 AppLoader_NSP::~AppLoader_NSP() = default;
 
 FileType AppLoader_NSP::IdentifyType(const FileSys::VirtualFile& nsp_file) {
+    // .nsz is NSP-compatible; return NSZ so the game list shows the right type.
+    if (nsp_file->GetName().ends_with(".nsz")) {
+        return FileType::NSZ;
+    }
+
+    // PFS0-shaped .xcz containers are handled by this loader too; report XCZ so the game
+    // list does not show them as NSP. XCI-shaped .xcz is identified by AppLoader_XCI.
+    if (nsp_file->GetName().ends_with(".xcz")) {
+        return FileType::XCZ;
+    }
+
     const FileSys::NSP nsp(nsp_file);
 
     if (nsp.GetStatus() != ResultStatus::Success) {

src/core/loader/xci.cpp

@@ -54,7 +54,8 @@ FileType AppLoader_XCI::IdentifyType(const FileSys::VirtualFile& xci_file) {
     // Identify XCI as a valid container even when it does not include a bootable Program NCA.
     // Bootability is handled by AppLoader_XCI::Load().
     if (xci.GetSecurePartitionNSP() != nullptr) {
-        return FileType::XCI;
+        return xci_file->GetName().ends_with(".xcz") ? FileType::XCZ
+                                                     : FileType::XCI;
     }
 
     return FileType::Error;

src/frontend_common/content_manager.h

@@ -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: 2024 yuzu Emulator Project
@@ -163,7 +163,9 @@ inline InstallResult InstallNSP(Core::System& system, FileSys::VfsFilesystem& vf
 
     std::shared_ptr<FileSys::NSP> nsp;
     FileSys::VirtualFile file = vfs.OpenFile(filename, FileSys::OpenMode::Read);
-    if (boost::to_lower_copy(file->GetName()).ends_with("nsp")) {
+    const auto lower_name = boost::to_lower_copy(file->GetName());
+
+    if (lower_name.ends_with("nsp") || lower_name.ends_with("nsz")) {
         nsp = std::make_shared<FileSys::NSP>(file);
         if (nsp->IsExtractedType()) {
             return InstallResult::Failure;

src/qt_common/config/shared_translation.cpp

@@ -194,10 +194,6 @@ std::unique_ptr<TranslationMap> InitializeTranslations(QObject* parent) {
     INSERT(Settings, skip_cpu_inner_invalidation, tr("Skip CPU Inner Invalidation"),
            tr("Skips certain cache invalidations during memory updates, reducing CPU usage and "
               "improving latency. This may cause soft-crashes."));
-    INSERT(Settings, nce_invalidation_gpu_readback,
-           tr("GPU Readback Before NCE Invalidation"),
-           tr("Flushes GPU buffer data back to guest memory before NCE invalidates the whole "
-              "guest page, preserving it while the CPU store proceeds."));
     INSERT(Settings, vsync_mode, tr("VSync Mode:"),
            tr("FIFO (VSync) does not drop frames or exhibit tearing but is limited by the screen "
               "refresh rate.\nFIFO Relaxed allows tearing as it recovers from a slow down.\n"

src/qt_common/game_list/worker.cpp

@@ -360,7 +360,8 @@ void GameListWorker::ScanFileSystem(ScanTarget target, const std::string& dir_pa
             }
 
             if (target == ScanTarget::PopulateGameList &&
-                (file_type == Loader::FileType::XCI || file_type == Loader::FileType::NSP) &&
+                (file_type == Loader::FileType::XCI || file_type == Loader::FileType::XCZ ||
+                 file_type == Loader::FileType::NSP || file_type == Loader::FileType::NSZ) &&
                 !Loader::IsBootableGameContainer(file, file_type)) {
                 return true;
             }
@@ -375,7 +376,9 @@ void GameListWorker::ScanFileSystem(ScanTarget target, const std::string& dir_pa
                                        program_id, file);
                 } else if (Settings::values.ext_content_from_game_dirs.GetValue() &&
                            (file_type == Loader::FileType::XCI ||
-                            file_type == Loader::FileType::NSP)) {
+                            file_type == Loader::FileType::XCZ ||
+                            file_type == Loader::FileType::NSP ||
+                            file_type == Loader::FileType::NSZ)) {
                     void(provider->AddEntriesFromContainer(file));
                 }
             } else {
@@ -402,7 +405,8 @@ void GameListWorker::ScanFileSystem(ScanTarget target, const std::string& dir_pa
                 };
 
                 if (res2 == Loader::ResultStatus::Success && program_ids.size() > 1 &&
-                    (file_type == Loader::FileType::XCI || file_type == Loader::FileType::NSP)) {
+                    (file_type == Loader::FileType::XCI || file_type == Loader::FileType::XCZ ||
+                     file_type == Loader::FileType::NSP || file_type == Loader::FileType::NSZ)) {
                     for (const auto id : program_ids) {
                         // dravee suggested this, only viable way to
                         // not show sub-games in qlaunch for now.

src/qt_common/qt_common.cpp

@@ -57,7 +57,8 @@ std::unique_ptr<EmuThread> emu_thread = nullptr;
 
 const QStringList supported_file_extensions = {QStringLiteral("nro"), QStringLiteral("nso"),
                                                QStringLiteral("nca"), QStringLiteral("xci"),
-                                               QStringLiteral("nsp"), QStringLiteral("kip")};
+                                               QStringLiteral("xcz"), QStringLiteral("nsp"),
+                                               QStringLiteral("nsz"), QStringLiteral("kip")};
 
 Core::Frontend::WindowSystemType GetWindowSystemType() {
     // Determine WSI type based on Qt platform.

src/shader_recompiler/environment.h

@@ -40,10 +40,9 @@ struct HandleKey {
     u32 sec_index, sec_offset, sec_shift_left;
     u32 count;
     bool has_secondary;
-    u32 mask, sec_mask;
     constexpr bool operator==(const HandleKey& o) const noexcept {
-        return std::tie(index, offset, shift_left, sec_index, sec_offset, sec_shift_left, count, has_secondary, mask, sec_mask)
-            == std::tie(o.index, o.offset, o.shift_left, o.sec_index, o.sec_offset, o.sec_shift_left, o.count, o.has_secondary, o.mask, o.sec_mask);
+        return std::tie(index, offset, shift_left, sec_index, sec_offset, sec_shift_left, count, has_secondary)
+            == std::tie(o.index, o.offset, o.shift_left, o.sec_index, o.sec_offset, o.sec_shift_left, o.count, o.has_secondary);
     }
 };
 struct HandleKeyHash {
@@ -53,7 +52,6 @@ struct HandleKeyHash {
         h ^= (size_t(k.sec_index) << 33) ^ (size_t(k.sec_offset) << 2);
         h ^= (size_t(k.sec_shift_left) << 3);
         h ^= (size_t(k.count) << 7);
-        h ^= (size_t(k.mask) << 11) ^ (size_t(k.sec_mask) << 13);
         return h ^ (k.has_secondary ? 0x9e3779b97f4a7c15ULL : 0ULL);
     }
 };
@@ -68,14 +66,6 @@ struct ConstBufferAddr {
     IR::U32 dynamic_offset;
     u32 count;
     bool has_secondary;
-    /// The dynamic offset is a constant buffer value shifted left, the same for the whole draw
-    bool has_index_cbuf;
-    u32 index_cbuf_index;
-    u32 index_cbuf_offset;
-    u32 index_shift;
-    /// Applied after the shifts
-    u32 mask = 0xffffffff;
-    u32 secondary_mask = 0xffffffff;
 };
 
 class Environment {

src/shader_recompiler/frontend/ir/breadth_first_search.h

@@ -1,8 +1,12 @@
+// 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 <functional>
 #include <optional>
 #include <type_traits>
 #include <queue>
@@ -13,9 +17,10 @@
 
 namespace Shader::IR {
 
-template <typename Pred>
-auto BreadthFirstSearch(const Value& value, Pred&& pred)
+template <typename Pred, typename Resolve = std::identity>
+auto BreadthFirstSearch(Value value, Pred&& pred, Resolve&& resolve = {})
     -> std::invoke_result_t<Pred, const Inst*> {
+    value = resolve(value);
     if (value.IsImmediate()) {
         // Nothing to do with immediates
         return std::nullopt;
@@ -36,7 +41,7 @@ auto BreadthFirstSearch(const Value& value, Pred&& pred)
         }
         // Visit the right most arguments first
         for (size_t arg = inst->NumArgs(); arg--;) {
-            const Value arg_value{inst->Arg(arg)};
+            const Value arg_value{resolve(inst->Arg(arg))};
             if (arg_value.IsImmediate()) {
                 continue;
             }

src/shader_recompiler/frontend/maxwell/translate/impl/floating_point_conversion_integer.cpp

@@ -1,4 +1,4 @@
-// SPDX-FileCopyrightText: Copyright 2026 Eden Emulator Project
+// SPDX-FileCopyrightText: Copyright 2025 Eden Emulator Project
 // SPDX-License-Identifier: GPL-3.0-or-later
 
 // SPDX-FileCopyrightText: Copyright 2021 yuzu Emulator Project
@@ -41,9 +41,9 @@ union F2I {
     BitField<39, 2, Rounding> rounding;
     BitField<41, 1, u64> half;
     BitField<44, 1, u64> ftz;
-    BitField<45, 1, u64> neg;
+    BitField<45, 1, u64> abs;
     BitField<47, 1, u64> cc;
-    BitField<49, 1, u64> abs;
+    BitField<49, 1, u64> neg;
 };
 
 size_t BitSize(DestFormat dest_format) {

src/shader_recompiler/ir_opt/texture_pass.cpp

@@ -31,6 +31,60 @@ struct TextureInst {
 
 using TextureInstVector = boost::container::small_vector<TextureInst, 24>;
 
+const IR::Block* ConditionalHeader(const IR::Block* body) {
+    const auto predecessors{body->ImmPredecessors()};
+    const auto successors{body->ImmSuccessors()};
+    if (predecessors.size() != 1 || successors.size() != 1) return nullptr;
+    const IR::Block* const header{predecessors[0]};
+    const auto targets{header->ImmSuccessors()};
+    if (targets.size() != 2 || targets[0] != body || targets[1] != successors[0] ||
+        targets[1]->ImmPredecessors().size() != 2 || header->empty() ||
+        header->back().GetOpcode() != IR::Opcode::ConditionRef) {
+        return nullptr;
+    }
+    return header;
+}
+
+std::pair<IR::Value, bool> Condition(IR::Value value) {
+    bool negated{};
+    while (const IR::Inst* inst = value.TryInstRecursive()) {
+        if (inst->GetOpcode() == IR::Opcode::LogicalNot) negated = !negated;
+        else if (inst->GetOpcode() != IR::Opcode::ConditionRef) break;
+        value = inst->Arg(0);
+    }
+    return {value.Resolve(), negated};
+}
+
+IR::Value ResolveConditionalPhi(IR::Value value, const IR::Block* block) {
+    if (!value.Resolve().IsPhi() || !(block = ConditionalHeader(block))) return value;
+    const auto [condition, negated]{Condition(block->back().Arg(0))};
+    boost::container::small_vector<const IR::Block*, 16> visited;
+    while (value.Resolve().IsPhi() && std::ranges::find(visited, block) == visited.end()) {
+        visited.push_back(block);
+        const IR::Inst* const phi{value.InstRecursive()};
+        const auto predecessors{block->ImmPredecessors()};
+        if (predecessors.size() == 1) {
+            block = predecessors[0];
+            continue;
+        }
+        if (predecessors.size() != 2 || phi->NumArgs() != 2) break;
+        const IR::Block* body{predecessors[0]};
+        const IR::Block* header{ConditionalHeader(body)};
+        if (header != predecessors[1]) {
+            body = predecessors[1];
+            header = ConditionalHeader(body);
+            if (header != predecessors[0]) break;
+        }
+        const auto [other_condition, other_negated]{Condition(header->back().Arg(0))};
+        if (condition == other_condition && (phi->PhiBlock(0) == header || phi->PhiBlock(0) == body)) {
+            const IR::Block* const incoming{negated == other_negated ? body : header};
+            value = phi->Arg(phi->PhiBlock(0) == incoming ? 0 : 1);
+        }
+        block = header;
+    }
+    return value;
+}
+
 constexpr u32 DESCRIPTOR_SIZE = 8;
 constexpr u32 DESCRIPTOR_SIZE_SHIFT = u32(std::countr_zero(DESCRIPTOR_SIZE));
 constexpr u32 DESCRIPTOR_MAX_COUNT = 1024;
@@ -43,8 +97,7 @@ u32 DynamicDescriptorSizeShift(const IR::U32& dynamic_offset) {
     if (!shift.IsImmediate())
         return DESCRIPTOR_SIZE_SHIFT;
     const u32 size_shift = shift.U32();
-    // Separate texture and sampler handles are 4 bytes apart
-    return size_shift >= 2 && size_shift < 31 ? size_shift : DESCRIPTOR_SIZE_SHIFT;
+    return size_shift >= DESCRIPTOR_SIZE_SHIFT && size_shift < 31 ? size_shift : DESCRIPTOR_SIZE_SHIFT;
 }
 
 u32 DynamicDescriptorCount(u32 base_offset, u32 size_shift, u32 max_descriptors) {
@@ -60,37 +113,6 @@ u32 DynamicDescriptorCount(u32 base_offset, u32 size_shift, u32 max_descriptors)
     return std::min(descriptor_limit, available_count);
 }
 
-struct UniformIndex {
-    u32 cbuf_index;
-    u32 cbuf_offset;
-    u32 shift;
-};
-
-/// Matches a dynamic handle offset that is a constant buffer value, optionally shifted left
-std::optional<UniformIndex> TrackUniformIndex(const IR::U32& dynamic_offset) {
-    if (dynamic_offset.IsImmediate()) {
-        return std::nullopt;
-    }
-    const IR::Inst* inst = dynamic_offset.InstRecursive();
-    u32 shift = 0;
-    if (inst->GetOpcode() == IR::Opcode::ShiftLeftLogical32) {
-        if (!inst->Arg(1).IsImmediate() || inst->Arg(0).IsImmediate()) {
-            return std::nullopt;
-        }
-        shift = inst->Arg(1).U32();
-        inst = inst->Arg(0).InstRecursive();
-    }
-    if (inst->GetOpcode() != IR::Opcode::GetCbufU32 || !inst->Arg(0).IsImmediate() ||
-        !inst->Arg(1).IsImmediate() || shift >= 32) {
-        return std::nullopt;
-    }
-    return UniformIndex{
-        .cbuf_index = inst->Arg(0).U32(),
-        .cbuf_offset = inst->Arg(1).U32(),
-        .shift = shift,
-    };
-}
-
 u32 SaturatingSub(u32 lhs, u32 rhs) {
     return lhs > rhs ? lhs - rhs : 0;
 }
@@ -305,15 +327,14 @@ static inline u32 GetTextureHandleCached(Environment& env, const ConstBufferAddr
     // If has secondary, then it will be cbuf.secondary_{index|offset}, else its 0.
     // So we can just hand it out the raw variable without using sec_idx or sec_off
     // because comparing 0 against 0 will yield true.
-    const HandleKey hk{cbuf.index, cbuf.offset, cbuf.shift_left, cbuf.secondary_index, cbuf.secondary_offset, cbuf.secondary_shift_left, cbuf.count, cbuf.has_secondary, cbuf.mask, cbuf.secondary_mask};
+    const HandleKey hk{cbuf.index, cbuf.offset, cbuf.shift_left, cbuf.secondary_index, cbuf.secondary_offset, cbuf.secondary_shift_left, cbuf.count, cbuf.has_secondary};
     if (auto it = env.handle_cache.find(hk); it != env.handle_cache.end())
         return it->second;
-    u32 handle = (ReadCbufCached(env, cbuf.index, cbuf.offset) << cbuf.shift_left) & cbuf.mask;
-    if (cbuf.has_secondary) {
-        handle |= (ReadCbufCached(env, cbuf.secondary_index, cbuf.secondary_offset)
-                   << cbuf.secondary_shift_left) &
-                  cbuf.secondary_mask;
-    }
+    const u32 sec_idx = cbuf.has_secondary ? cbuf.secondary_index  : cbuf.index;
+    const u32 sec_off = cbuf.has_secondary ? cbuf.secondary_offset : cbuf.offset;
+    const u32 lhs = ReadCbufCached(env, cbuf.index, cbuf.offset) << cbuf.shift_left;
+    const u32 rhs = ReadCbufCached(env, sec_idx, sec_off) << cbuf.secondary_shift_left;
+    const u32 handle = lhs | rhs;
     env.handle_cache.emplace(hk, handle);
     return handle;
 }
@@ -332,23 +353,23 @@ static inline bool IsTexturePixelFormatIntegerCached(Environment& env,
 }
 
 
-std::optional<ConstBufferAddr> Track(const IR::Value& value, Environment& env, const HostTranslateInfo& host_info);
-static inline std::optional<ConstBufferAddr> TrackCached(const IR::Value& v, Environment& env, const HostTranslateInfo& host_info) {
-    if (const IR::Inst* key = v.InstRecursive()) {
+std::optional<ConstBufferAddr> Track(const IR::Value& value, Environment& env, const HostTranslateInfo& host_info, const IR::Block* block);
+static inline std::optional<ConstBufferAddr> TrackCached(const IR::Value& v, Environment& env, const HostTranslateInfo& host_info, const IR::Block* block) {
+    if (const IR::Inst* key = v.TryInstRecursive()) {
         if (auto it = env.track_cache.find(key); it != env.track_cache.end()) return it->second;
-        auto found = Track(v, env, host_info);
+        auto found = Track(v, env, host_info, block);
         if (found) env.track_cache.emplace(key, *found);
         return found;
     }
-    return Track(v, env, host_info);
+    return Track(v, env, host_info, block);
 }
 
-std::optional<ConstBufferAddr> TryGetConstBuffer(const IR::Inst* inst, Environment& env, const HostTranslateInfo& host_info);
+std::optional<ConstBufferAddr> TryGetConstBuffer(const IR::Inst* inst, Environment& env, const HostTranslateInfo& host_info, const IR::Block* block);
 
-std::optional<ConstBufferAddr> Track(const IR::Value& value, Environment& env, const HostTranslateInfo& host_info) {
-    return IR::BreadthFirstSearch(value, [&env, &host_info](const IR::Inst* inst) {
-        return TryGetConstBuffer(inst, env, host_info);
-    });
+std::optional<ConstBufferAddr> Track(const IR::Value& value, Environment& env, const HostTranslateInfo& host_info, const IR::Block* block) {
+    return IR::BreadthFirstSearch(value, [&env, &host_info, block](const IR::Inst* inst) {
+        return TryGetConstBuffer(inst, env, host_info, block);
+    }, [&](IR::Value operand) { return ResolveConditionalPhi(operand, block); });
 }
 
 std::optional<u32> TryGetConstant(IR::Value& value, Environment& env) {
@@ -372,35 +393,23 @@ std::optional<u32> TryGetConstant(IR::Value& value, Environment& env) {
     return ReadCbufCached(env, index_number, offset_number);
 }
 
-std::optional<ConstBufferAddr> TryGetConstBuffer(const IR::Inst* inst, Environment& env, const HostTranslateInfo& host_info) {
+std::optional<ConstBufferAddr> TryGetConstBuffer(const IR::Inst* inst, Environment& env, const HostTranslateInfo& host_info, const IR::Block* block) {
     switch (inst->GetOpcode()) {
     default:
         return std::nullopt;
     case IR::Opcode::BitwiseOr32: {
-        std::optional lhs{TrackCached(inst->Arg(0), env, host_info)};
-        std::optional rhs{TrackCached(inst->Arg(1), env, host_info)};
+        std::optional lhs{TrackCached(inst->Arg(0), env, host_info, block)};
+        std::optional rhs{TrackCached(inst->Arg(1), env, host_info, block)};
         if (!lhs || !rhs) {
             return std::nullopt;
         }
         if (lhs->has_secondary || rhs->has_secondary) {
             return std::nullopt;
         }
-        if (lhs->count > 1 && rhs->count > 1) {
+        if (lhs->count > 1 || rhs->count > 1) {
             return std::nullopt;
         }
-        if (rhs->has_index_cbuf || lhs->has_index_cbuf) {
-            if (rhs->has_index_cbuf && lhs->has_index_cbuf) {
-                return std::nullopt;
-            }
-            if (rhs->has_index_cbuf) {
-                std::swap(lhs, rhs);
-            }
-        } else if (rhs->count > 1) {
-            // An indexed handle array combined with a single handle (e.g. texture indices with
-            // a shared sampler), only the primary handle is indexed
-            std::swap(lhs, rhs);
-        } else if (lhs->count == 1 &&
-                   (lhs->shift_left > 0 || lhs->index > rhs->index || lhs->offset > rhs->offset)) {
+        if (lhs->shift_left > 0 || lhs->index > rhs->index || lhs->offset > rhs->offset) {
             std::swap(lhs, rhs);
         }
         return ConstBufferAddr{
@@ -410,15 +419,9 @@ std::optional<ConstBufferAddr> TryGetConstBuffer(const IR::Inst* inst, Environme
             .secondary_index = rhs->index,
             .secondary_offset = rhs->offset,
             .secondary_shift_left = rhs->shift_left,
-            .dynamic_offset = lhs->dynamic_offset,
-            .count = lhs->count,
+            .dynamic_offset = {},
+            .count = 1,
             .has_secondary = true,
-            .has_index_cbuf = lhs->has_index_cbuf,
-            .index_cbuf_index = lhs->index_cbuf_index,
-            .index_cbuf_offset = lhs->index_cbuf_offset,
-            .index_shift = lhs->index_shift,
-            .mask = lhs->mask,
-            .secondary_mask = rhs->mask,
         };
     }
     case IR::Opcode::ShiftLeftLogical32: {
@@ -426,11 +429,9 @@ std::optional<ConstBufferAddr> TryGetConstBuffer(const IR::Inst* inst, Environme
         if (!shift.IsImmediate()) {
             return std::nullopt;
         }
-        std::optional lhs{TrackCached(inst->Arg(0), env, host_info)};
+        std::optional lhs{TrackCached(inst->Arg(0), env, host_info, block)};
         if (lhs) {
             lhs->shift_left = shift.U32();
-            // A mask applied before the shift moves with the value
-            lhs->mask <<= shift.U32();
         }
         return lhs;
     }
@@ -456,9 +457,9 @@ std::optional<ConstBufferAddr> TryGetConstBuffer(const IR::Inst* inst, Environme
                 return std::nullopt;
             } while (false);
         }
-        std::optional lhs{TrackCached(op1, env, host_info)};
+        std::optional lhs{TrackCached(op1, env, host_info, block)};
         if (lhs) {
-            lhs->mask &= op2.U32();
+            lhs->shift_left = static_cast<u32>(std::countr_zero(op2.U32()));
         }
         return lhs;
         break;
@@ -488,12 +489,12 @@ std::optional<ConstBufferAddr> TryGetConstBuffer(const IR::Inst* inst, Environme
         };
     }
     IR::Inst* const offset_inst{offset.InstRecursive()};
+    if (offset_inst->GetOpcode() != IR::Opcode::IAdd32) {
+        return std::nullopt;
+    }
     u32 base_offset{};
     IR::U32 dynamic_offset;
-    if (offset_inst->GetOpcode() != IR::Opcode::IAdd32) {
-        // The handle array starts at the beginning of the constant buffer
-        dynamic_offset = IR::U32{offset};
-    } else if (offset_inst->Arg(0).IsImmediate()) {
+    if (offset_inst->Arg(0).IsImmediate()) {
         base_offset = offset_inst->Arg(0).U32();
         dynamic_offset = IR::U32{offset_inst->Arg(1)};
     } else if (offset_inst->Arg(1).IsImmediate()) {
@@ -503,7 +504,6 @@ std::optional<ConstBufferAddr> TryGetConstBuffer(const IR::Inst* inst, Environme
         return std::nullopt;
     }
     auto const size_shift = DynamicDescriptorSizeShift(dynamic_offset);
-    const std::optional<UniformIndex> uniform_index{TrackUniformIndex(dynamic_offset)};
     auto const sampled_limit = (std::max)(1U, (std::min)(host_info.max_per_stage_descriptor_sampled_images,
                                                          host_info.max_descriptor_set_sampled_images));
     auto const resource_limit = (std::max)(1U, host_info.max_per_stage_resources);
@@ -517,17 +517,13 @@ std::optional<ConstBufferAddr> TryGetConstBuffer(const IR::Inst* inst, Environme
         .dynamic_offset = dynamic_offset,
         .count = DynamicDescriptorCount(base_offset, size_shift, (std::min)({DESCRIPTOR_MAX_COUNT, sampled_limit, resource_limit})),
         .has_secondary = false,
-        .has_index_cbuf = uniform_index.has_value(),
-        .index_cbuf_index = uniform_index ? uniform_index->cbuf_index : 0,
-        .index_cbuf_offset = uniform_index ? uniform_index->cbuf_offset : 0,
-        .index_shift = uniform_index ? uniform_index->shift : 0,
     };
 }
 
 TextureInst MakeInst(Environment& env, IR::Block* block, IR::Inst& inst, const HostTranslateInfo& host_info) {
     ConstBufferAddr addr;
     if (IsBindless(inst)) {
-        const std::optional<ConstBufferAddr> track_addr{TrackCached(inst.Arg(0), env, host_info)};
+        const std::optional<ConstBufferAddr> track_addr{TrackCached(inst.Arg(0), env, host_info, block)};
 
         if (!track_addr) {
             throw NotImplementedException("Failed to track bindless texture constant buffer");
@@ -555,13 +551,12 @@ TextureInst MakeInst(Environment& env, IR::Block* block, IR::Inst& inst, const H
 }
 
 u32 GetTextureHandle(Environment& env, const ConstBufferAddr& cbuf) {
-    u32 handle{(env.ReadCbufValue(cbuf.index, cbuf.offset) << cbuf.shift_left) & cbuf.mask};
-    if (cbuf.has_secondary) {
-        handle |= (env.ReadCbufValue(cbuf.secondary_index, cbuf.secondary_offset)
-                   << cbuf.secondary_shift_left) &
-                  cbuf.secondary_mask;
-    }
-    return handle;
+    const u32 secondary_index{cbuf.has_secondary ? cbuf.secondary_index : cbuf.index};
+    const u32 secondary_offset{cbuf.has_secondary ? cbuf.secondary_offset : cbuf.offset};
+    const u32 lhs_raw{env.ReadCbufValue(cbuf.index, cbuf.offset) << cbuf.shift_left};
+    const u32 rhs_raw{env.ReadCbufValue(secondary_index, secondary_offset)
+                      << cbuf.secondary_shift_left};
+    return lhs_raw | rhs_raw;
 }
 
 [[maybe_unused]] TextureType ReadTextureType(Environment& env, const ConstBufferAddr& cbuf) {
@@ -595,8 +590,7 @@ public:
                    desc.secondary_cbuf_offset == existing.secondary_cbuf_offset &&
                    desc.secondary_shift_left == existing.secondary_shift_left &&
                    desc.count == existing.count && desc.size_shift == existing.size_shift &&
-                   desc.has_secondary == existing.has_secondary && desc.mask == existing.mask &&
-                   desc.secondary_mask == existing.secondary_mask;
+                   desc.has_secondary == existing.has_secondary;
         });
     }
 
@@ -622,12 +616,7 @@ public:
                    desc.secondary_cbuf_index == existing.secondary_cbuf_index &&
                    desc.secondary_cbuf_offset == existing.secondary_cbuf_offset &&
                    desc.secondary_shift_left == existing.secondary_shift_left &&
-                   desc.count == existing.count && desc.size_shift == existing.size_shift &&
-                   desc.has_index_cbuf == existing.has_index_cbuf &&
-                   desc.index_cbuf_index == existing.index_cbuf_index &&
-                   desc.index_cbuf_offset == existing.index_cbuf_offset &&
-                   desc.index_shift == existing.index_shift && desc.mask == existing.mask &&
-                   desc.secondary_mask == existing.secondary_mask;
+                   desc.count == existing.count && desc.size_shift == existing.size_shift;
         })};
         // TODO: Read this from TIC
         texture_descriptors[index].is_multisample |= desc.is_multisample;
@@ -736,6 +725,7 @@ void TexturePass(Environment& env, IR::Program& program, const HostTranslateInfo
     env.track_cache.clear();
     TextureInstVector to_replace;
     for (IR::Block* const block : program.post_order_blocks) {
+        env.track_cache.clear();
         for (IR::Inst& inst : block->Instructions()) {
             if (!IsTextureInstruction(inst)) {
                 continue;
@@ -873,12 +863,9 @@ void TexturePass(Environment& env, IR::Program& program, const HostTranslateInfo
                     .secondary_shift_left = cbuf.secondary_shift_left,
                     .count = count,
                     .size_shift = size_shift,
-                    .mask = cbuf.mask,
-                    .secondary_mask = cbuf.secondary_mask,
                 });
             } else {
-                // A uniform index is resolved when drawing, so a single descriptor is enough
-                count = cbuf.has_index_cbuf ? 1 : std::min(count, sampled_dynamic_cap);
+                count = std::min(count, sampled_dynamic_cap);
                 const bool is_integer{is_texture_pixel_format_integer(cbuf)};
                 index = descriptors.Add(TextureDescriptor{
                     .type = flags.type,
@@ -894,12 +881,6 @@ void TexturePass(Environment& env, IR::Program& program, const HostTranslateInfo
                     .secondary_shift_left = cbuf.secondary_shift_left,
                     .count = count,
                     .size_shift = size_shift,
-                    .has_index_cbuf = cbuf.has_index_cbuf,
-                    .index_cbuf_index = cbuf.index_cbuf_index,
-                    .index_cbuf_offset = cbuf.index_cbuf_offset,
-                    .index_shift = cbuf.index_shift,
-                    .mask = cbuf.mask,
-                    .secondary_mask = cbuf.secondary_mask,
                 });
                 flags.is_integer.Assign(is_integer ? 1 : 0);
             }

src/shader_recompiler/shader_info.h

@@ -186,8 +186,6 @@ struct TextureBufferDescriptor {
     u32 secondary_shift_left;
     u32 count;
     u32 size_shift;
-    u32 mask = 0xffffffff;
-    u32 secondary_mask = 0xffffffff;
 
     auto operator<=>(const TextureBufferDescriptor&) const = default;
 };
@@ -221,13 +219,6 @@ struct TextureDescriptor {
     u32 secondary_shift_left;
     u32 count;
     u32 size_shift;
-    /// The handle offset is moved by a constant buffer value read when drawing
-    bool has_index_cbuf;
-    u32 index_cbuf_index;
-    u32 index_cbuf_offset;
-    u32 index_shift;
-    u32 mask = 0xffffffff;
-    u32 secondary_mask = 0xffffffff;
 
     auto operator<=>(const TextureDescriptor&) const = default;
 };

src/video_core/buffer_cache/buffer_cache.h

@@ -948,14 +948,9 @@ void BufferCache<P>::BindHostGraphicsUniformBuffer(size_t stage, u32 index, u32
             return alignment > 1 && (offset % alignment) != 0;
         }
     }();
-    // Uniform data rewritten by the CPU between draws would need a transfer copy into the cached
-    // buffer, which ends the current render pass. Read it from the stream buffer instead.
-    static constexpr u32 MAX_STREAMED_UNIFORM_BYTES = static_cast<u32>(16_KiB);
     const bool use_fast_buffer = needs_alignment_stream
-        || (has_host_buffer && !memory_tracker.IsRegionGpuModified(device_addr, size)
-            && (size <= channel_state->uniform_buffer_skip_cache_size
-                || (size <= MAX_STREAMED_UNIFORM_BYTES
-                    && memory_tracker.IsRegionCpuModified(device_addr, size))));
+        || (has_host_buffer && size <= channel_state->uniform_buffer_skip_cache_size
+            && !memory_tracker.IsRegionGpuModified(device_addr, size));
     if (use_fast_buffer) {
         if constexpr (IS_OPENGL) {
             if (runtime.HasFastBufferSubData()) {
@@ -1743,36 +1738,64 @@ void BufferCache<P>::TouchBuffer(Buffer& buffer, BufferId buffer_id) noexcept {
 
 template <class P>
 bool BufferCache<P>::SynchronizeBuffer(Buffer& buffer, DAddr device_addr, u32 size) {
-    upload_copies.clear();
-    u64 total_size_bytes = 0;
-    u64 largest_copy = 0;
-    const DAddr buffer_start = buffer.cpu_addr_cached;
-    const auto add_upload = [&](DAddr start, DAddr end) {
-        if (start == end) return;
-        const u64 range_size = end - start;
-        upload_copies.push_back(BufferCopy{
-            .src_offset = total_size_bytes,
-            .dst_offset = start - buffer_start,
-            .size = range_size,
+#ifdef __ANDROID__
+    const bool use_optimized_vertex_buffers = Settings::values.use_optimized_vertex_buffers.GetValue();
+#else
+    constexpr bool use_optimized_vertex_buffers = true;
+#endif
+
+    if (use_optimized_vertex_buffers) {
+        upload_copies.clear();
+        u64 total_size_bytes = 0;
+        u64 largest_copy = 0;
+        const DAddr buffer_start = buffer.cpu_addr_cached;
+        const auto add_upload = [&](DAddr start, DAddr end) {
+            if (start == end) return;
+            const u64 range_size = end - start;
+            upload_copies.push_back(BufferCopy{
+                .src_offset = total_size_bytes,
+                .dst_offset = start - buffer_start,
+                .size = range_size,
+            });
+            total_size_bytes += range_size;
+            largest_copy = (std::max)(largest_copy, range_size);
+        };
+        memory_tracker.ForEachUploadRange(device_addr, size, [&](u64 device_addr_out, u64 range_size) {
+            DAddr upload_start = device_addr_out;
+            gpu_modified_ranges.ForEachInRange(device_addr_out, range_size, [&](DAddr gpu_start, DAddr gpu_end) {
+                add_upload(upload_start, gpu_start);
+                upload_start = gpu_end;
+            });
+            add_upload(upload_start, device_addr_out + range_size);
         });
-        total_size_bytes += range_size;
-        largest_copy = (std::max)(largest_copy, range_size);
-    };
-    memory_tracker.ForEachUploadRange(device_addr, size, [&](u64 device_addr_out, u64 range_size) {
-        DAddr upload_start = device_addr_out;
-        gpu_modified_ranges.ForEachInRange(device_addr_out, range_size, [&](DAddr gpu_start, DAddr gpu_end) {
-            add_upload(upload_start, gpu_start);
-            upload_start = gpu_end;
+        if (total_size_bytes == 0) {
+            return true;
+        }
+        const std::span<BufferCopy> copies_span(upload_copies.data(), upload_copies.size());
+        UploadMemory(buffer, total_size_bytes, largest_copy, copies_span);
+        any_buffer_uploaded = true;
+        return false;
+    } else {
+        boost::container::small_vector<BufferCopy, 4> copies;
+        u64 total_size_bytes = 0;
+        u64 largest_copy = 0;
+        DAddr buffer_start = buffer.CpuAddr();
+        memory_tracker.ForEachUploadRange(device_addr, size, [&](u64 device_addr_out, u64 range_size) {
+            copies.push_back(BufferCopy{
+                .src_offset = total_size_bytes,
+                .dst_offset = device_addr_out - buffer_start,
+                .size = range_size,
+            });
+            total_size_bytes += range_size;
+            largest_copy = (std::max)(largest_copy, range_size);
         });
-        add_upload(upload_start, device_addr_out + range_size);
-    });
-    if (total_size_bytes == 0) {
-        return true;
+        if (total_size_bytes == 0) {
+            return true;
+        }
+        const std::span<BufferCopy> copies_span(copies.data(), copies.size());
+        UploadMemory(buffer, total_size_bytes, largest_copy, copies_span);
+        return false;
     }
-    const std::span<BufferCopy> copies_span(upload_copies.data(), upload_copies.size());
-    UploadMemory(buffer, total_size_bytes, largest_copy, copies_span);
-    any_buffer_uploaded = true;
-    return false;
 }
 
 template <class P>

src/video_core/buffer_cache/buffer_cache_base.h

@@ -259,23 +259,6 @@ public:
 
     void BindHostComputeBuffers();
 
-    /// Calls func(buffer, offset, device_addr, size) for each storage buffer the current compute
-    /// dispatch writes through a single host buffer
-    template <typename Func>
-    void ForEachWrittenComputeStorageBuffer(Func&& func) {
-        const u32 written = channel_state->enabled_compute_storage_buffers &
-                            channel_state->written_compute_storage_buffers;
-        ForEachEnabledBit(written, [&](u32 index) {
-            const Binding& binding = channel_state->compute_storage_buffers[index];
-            if (binding.buffer_id == NULL_BUFFER_ID || binding.segment_count != 0 ||
-                binding.size == 0) {
-                return;
-            }
-            Buffer& buffer = slot_buffers[binding.buffer_id];
-            func(buffer, buffer.Offset(binding.device_addr), binding.device_addr, binding.size);
-        });
-    }
-
     void SetUniformBuffersState(const std::array<u32, NUM_STAGES>& mask,
                                 const UniformBufferSizes* sizes);
 

src/video_core/gpu.cpp

@@ -227,17 +227,17 @@ struct GPU::Impl {
 
     /// Notify rasterizer that any caches of the specified region should be invalidated
     void InvalidateRegion(DAddr addr, u64 size, bool preserve_gpu_writes) {
-        if (Settings::values.nce_invalidation_gpu_readback.GetValue()) {
+        if (preserve_gpu_writes) {
             VideoCore::RasterizerInterface* rasterizer = renderer->ReadRasterizer();
-            if (preserve_gpu_writes && rasterizer->MustFlushRegion(addr, size, VideoCommon::CacheType::BufferCache)) {
+            if (!rasterizer->TryCacheInvalidation(addr, size)) {
                 const u64 fence = RequestSyncOperation([rasterizer, addr, size] {
                     rasterizer->FlushRegion(addr, size, VideoCommon::CacheType::BufferCache);
                     rasterizer->OnCacheInvalidation(addr, size);
                 });
                 gpu_thread.TickGPU(is_async);
                 WaitForSyncOperation(fence);
-                return;
             }
+            return;
         }
         gpu_thread.InvalidateRegion(addr, size);
     }

src/video_core/host_shaders/CMakeLists.txt

@@ -83,7 +83,6 @@ set(SHADER_FILES
     ${CMAKE_CURRENT_SOURCE_DIR}/vulkan_present.vert
     ${CMAKE_CURRENT_SOURCE_DIR}/vulkan_present_scaleforce_fp16.frag
     ${CMAKE_CURRENT_SOURCE_DIR}/vulkan_present_scaleforce_fp32.frag
-    ${CMAKE_CURRENT_SOURCE_DIR}/vulkan_block_linear_unswizzle_2d_buffer.comp
     ${CMAKE_CURRENT_SOURCE_DIR}/vulkan_quad_indexed.comp
     ${CMAKE_CURRENT_SOURCE_DIR}/vulkan_turbo_mode.comp
     ${CMAKE_CURRENT_SOURCE_DIR}/vulkan_uint8.comp

src/video_core/host_shaders/vulkan_block_linear_unswizzle_2d_buffer.comp

@@ -1,62 +0,0 @@
-// SPDX-FileCopyrightText: Copyright 2026 Eden Emulator Project
-// SPDX-License-Identifier: GPL-3.0-or-later
-
-#version 460 core
-
-// Copies a block linear 2D image from a buffer into a tightly packed linear buffer, 4 bytes at a
-// time. 16 byte chunks of a row are contiguous in a GOB, so any texel size works as long as the
-// row size is a multiple of 4 bytes.
-
-layout(local_size_x = 64, local_size_y = 1, local_size_z = 1) in;
-
-layout(std430, set = 0, binding = 0) readonly buffer InputBuffer {
-    uint swizzled[];
-};
-
-layout(std430, set = 0, binding = 1) writeonly buffer OutputBuffer {
-    uint linear[];
-};
-
-layout(push_constant) uniform PushConstants {
-    uint src_word_offset;
-    uint row_words;
-    uint height;
-    uint layer_stride;
-    uint block_size;
-    uint x_shift;
-    uint block_height;
-    uint block_height_mask;
-};
-
-const uint GOB_SIZE_X_SHIFT = 6;
-const uint GOB_SIZE_Y_SHIFT = 3;
-const uint GOB_SIZE_SHIFT = GOB_SIZE_X_SHIFT + GOB_SIZE_Y_SHIFT;
-
-uint SwizzleTable(uint pos) {
-    const uint t[8] = uint[](
-        0x12100200, 0x13110301, 0x16140604, 0x17150705,
-        0x1a180a08, 0x1b190b09, 0x1e1c0e0c, 0x1f1d0f0d
-    );
-    const uint i = pos >> 4;
-    const uint h = (t[i / 4] >> ((i % 4) * 8)) & 0xff;
-    return (h << 4) | (pos & 0xf);
-}
-
-void main() {
-    const uint word = gl_GlobalInvocationID.x;
-    const uint y = gl_GlobalInvocationID.y;
-    const uint layer = gl_GlobalInvocationID.z;
-    if (word >= row_words || y >= height) {
-        return;
-    }
-    const uint x = word * 4;
-    const uint block_y = y >> GOB_SIZE_Y_SHIFT;
-
-    uint offset = layer * layer_stride;
-    offset += (block_y >> block_height) * block_size;
-    offset += (block_y & block_height_mask) << GOB_SIZE_SHIFT;
-    offset += (x >> GOB_SIZE_X_SHIFT) << x_shift;
-    offset += SwizzleTable((y & 7) * 64 + (x & 63));
-
-    linear[(layer * height + y) * row_words + word] = swizzled[src_word_offset + offset / 4];
-}

src/video_core/rasterizer_interface.h

@@ -1,3 +1,6 @@
+// SPDX-FileCopyrightText: Copyright 2026 Eden Emulator Project
+// SPDX-License-Identifier: GPL-3.0-or-later
+
 // SPDX-FileCopyrightText: Copyright 2018 yuzu Emulator Project
 // SPDX-License-Identifier: GPL-2.0-or-later
 
@@ -111,6 +114,12 @@ public:
     /// Notify rasterizer that any caches of the specified region are desync with guest
     virtual void OnCacheInvalidation(PAddr addr, u64 size) = 0;
 
+    virtual bool TryCacheInvalidation(DAddr addr, u64 size) {
+        if (MustFlushRegion(addr, size, VideoCommon::CacheType::BufferCache)) return false;
+        OnCacheInvalidation(addr, size);
+        return true;
+    }
+
     virtual bool OnCPUWrite(PAddr addr, u64 size) = 0;
 
     /// Sync memory between guest and host.

src/video_core/renderer_opengl/gl_compute_pipeline.cpp

@@ -107,36 +107,22 @@ void ComputePipeline::Configure() {
     const auto read_handle{[&](const auto& desc, u32 index) {
         ASSERT(((qmd.const_buffer_enable_mask >> desc.cbuf_index) & 1) != 0);
         const u32 index_offset{index << desc.size_shift};
-        u32 offset{desc.cbuf_offset + index_offset};
-        if constexpr (std::is_same_v<decltype(desc), const Shader::TextureDescriptor&>) {
-            if (desc.has_index_cbuf) {
-                ASSERT(((qmd.const_buffer_enable_mask >> desc.index_cbuf_index) & 1) != 0);
-                offset += gpu_memory->Read<u32>(cbufs[desc.index_cbuf_index].Address() +
-                                                desc.index_cbuf_offset)
-                          << desc.index_shift;
-            }
-        }
+        const u32 offset{desc.cbuf_offset + index_offset};
         const GPUVAddr addr{cbufs[desc.cbuf_index].Address() + offset};
         if constexpr (std::is_same_v<decltype(desc), const Shader::TextureDescriptor&> ||
                       std::is_same_v<decltype(desc), const Shader::TextureBufferDescriptor&>) {
             if (desc.has_secondary) {
                 ASSERT(((qmd.const_buffer_enable_mask >> desc.secondary_cbuf_index) & 1) != 0);
-                const u32 secondary_offset{desc.secondary_cbuf_offset};
+                const u32 secondary_offset{desc.secondary_cbuf_offset + index_offset};
                 const GPUVAddr separate_addr{cbufs[desc.secondary_cbuf_index].Address() +
                                              secondary_offset};
-                const u32 lhs_raw{(gpu_memory->Read<u32>(addr) << desc.shift_left) & desc.mask};
-                const u32 rhs_raw{(gpu_memory->Read<u32>(separate_addr)
-                                   << desc.secondary_shift_left) &
-                                  desc.secondary_mask};
+                const u32 lhs_raw{gpu_memory->Read<u32>(addr) << desc.shift_left};
+                const u32 rhs_raw{gpu_memory->Read<u32>(separate_addr)
+                                  << desc.secondary_shift_left};
                 return TexturePair(lhs_raw | rhs_raw, via_header_index);
             }
         }
-        u32 raw{gpu_memory->Read<u32>(addr)};
-        if constexpr (std::is_same_v<decltype(desc), const Shader::TextureDescriptor&> ||
-                      std::is_same_v<decltype(desc), const Shader::TextureBufferDescriptor&>) {
-            raw &= desc.mask;
-        }
-        return TexturePair(raw, via_header_index);
+        return TexturePair(gpu_memory->Read<u32>(addr), via_header_index);
     }};
     const auto add_image{[&](const auto& desc, bool blacklist) {
         for (u32 index = 0; index < desc.count; ++index) {

src/video_core/renderer_opengl/gl_graphics_pipeline.cpp

@@ -308,37 +308,23 @@ bool GraphicsPipeline::ConfigureImpl(bool is_indexed) {
         const auto read_handle{[&](const auto& desc, u32 index) {
             ASSERT(cbufs[desc.cbuf_index].enabled);
             const u32 index_offset{index << desc.size_shift};
-            u32 offset{desc.cbuf_offset + index_offset};
-            if constexpr (std::is_same_v<decltype(desc), const TextureDescriptor&>) {
-                if (desc.has_index_cbuf) {
-                    ASSERT(cbufs[desc.index_cbuf_index].enabled);
-                    offset += gpu_memory->Read<u32>(cbufs[desc.index_cbuf_index].address +
-                                                    desc.index_cbuf_offset)
-                              << desc.index_shift;
-                }
-            }
+            const u32 offset{desc.cbuf_offset + index_offset};
             const GPUVAddr addr{cbufs[desc.cbuf_index].address + offset};
             if constexpr (std::is_same_v<decltype(desc), const TextureDescriptor&> ||
                           std::is_same_v<decltype(desc), const TextureBufferDescriptor&>) {
                 if (desc.has_secondary) {
                     ASSERT(cbufs[desc.secondary_cbuf_index].enabled);
-                    const u32 second_offset{desc.secondary_cbuf_offset};
+                    const u32 second_offset{desc.secondary_cbuf_offset + index_offset};
                     const GPUVAddr separate_addr{cbufs[desc.secondary_cbuf_index].address +
                                                  second_offset};
-                    const u32 lhs_raw{(gpu_memory->Read<u32>(addr) << desc.shift_left) & desc.mask};
-                    const u32 rhs_raw{(gpu_memory->Read<u32>(separate_addr)
-                                       << desc.secondary_shift_left) &
-                                      desc.secondary_mask};
+                    const u32 lhs_raw{gpu_memory->Read<u32>(addr) << desc.shift_left};
+                    const u32 rhs_raw{gpu_memory->Read<u32>(separate_addr)
+                                      << desc.secondary_shift_left};
                     const u32 raw{lhs_raw | rhs_raw};
                     return TexturePair(raw, via_header_index);
                 }
             }
-            u32 raw{gpu_memory->Read<u32>(addr)};
-            if constexpr (std::is_same_v<decltype(desc), const TextureDescriptor&> ||
-                          std::is_same_v<decltype(desc), const TextureBufferDescriptor&>) {
-                raw &= desc.mask;
-            }
-            return TexturePair(raw, via_header_index);
+            return TexturePair(gpu_memory->Read<u32>(addr), via_header_index);
         }};
         const auto add_image{[&](const auto& desc, bool blacklist) LAMBDA_FORCEINLINE {
             for (u32 index = 0; index < desc.count; ++index) {

src/video_core/renderer_opengl/gl_shader_cache.cpp

@@ -414,6 +414,11 @@ GraphicsPipeline* ShaderCache::BuiltPipeline(GraphicsPipeline* pipeline) const n
     if (!use_asynchronous_shaders) {
         return pipeline;
     }
+    // If something is using depth, we can assume that games are not rendering anything which
+    // will be used one time.
+    if (maxwell3d->regs.zeta_enable) {
+        return nullptr;
+    }
     // If games are using a small index count, we can assume these are full screen quads.
     // Usually these shaders are only used once for building textures so we can assume they
     // can't be built async

src/video_core/renderer_vulkan/fixed_pipeline_state.cpp

@@ -129,13 +129,7 @@ void FixedPipelineState::Refresh(Tegra::Engines::Maxwell3D& maxwell3d, DynamicFe
     topology.Assign(can_collapse_topology_class
                         ? TOPOLOGY_CLASS_REPRESENTATIVE_LUT[static_cast<size_t>(topology_)]
                         : topology_);
-    // Target-independent rasterization rasterizes (and depth tests) at the raster sample count
-    if (regs.tir_mode == Maxwell::TIRMode::RasterNTargetM &&
-        regs.anti_alias_raster != Maxwell::AntiAliasRaster::Mode1x1) {
-        msaa_mode.Assign(static_cast<Tegra::Texture::MsaaMode>(regs.anti_alias_raster));
-    } else {
-        msaa_mode.Assign(regs.anti_alias_samples_mode);
-    }
+    msaa_mode.Assign(regs.anti_alias_samples_mode);
     attachment0_dual_source_blend.Assign(ComputeAttachment0DualSourceBlend(regs) ? 1 : 0);
 
     raw2 = 0;

src/video_core/renderer_vulkan/vk_compute_pass.cpp

@@ -20,7 +20,6 @@
 #include "video_core/host_shaders/queries_prefix_scan_sum_comp_spv.h"
 #include "video_core/host_shaders/queries_prefix_scan_sum_nosubgroups_comp_spv.h"
 #include "video_core/host_shaders/resolve_conditional_render_comp_spv.h"
-#include "video_core/host_shaders/vulkan_block_linear_unswizzle_2d_buffer_comp_spv.h"
 #include "video_core/host_shaders/vulkan_quad_indexed_comp_spv.h"
 #include "video_core/host_shaders/vulkan_uint8_comp_spv.h"
 #include "video_core/host_shaders/block_linear_unswizzle_3d_bcn_comp_spv.h"
@@ -32,7 +31,6 @@
 #include "video_core/renderer_vulkan/vk_update_descriptor.h"
 #include "video_core/texture_cache/accelerated_swizzle.h"
 #include "video_core/texture_cache/types.h"
-#include "video_core/texture_cache/util.h"
 #include "video_core/textures/decoders.h"
 #include "video_core/vulkan_common/vulkan_device.h"
 #include "video_core/vulkan_common/vulkan_wrapper.h"
@@ -642,143 +640,6 @@ struct alignas(16) BlockLinearUnswizzle3DPushConstants {
 };
 static_assert(sizeof(BlockLinearUnswizzle3DPushConstants) <= 128);
 
-struct BlockLinearUnswizzle2DBufferPushConstants {
-    u32 src_word_offset;
-    u32 row_words;
-    u32 height;
-    u32 layer_stride;
-    u32 block_size;
-    u32 x_shift;
-    u32 block_height;
-    u32 block_height_mask;
-};
-
-BlockLinearUnswizzle2DBufferPass::BlockLinearUnswizzle2DBufferPass(
-    const Device& device_, Scheduler& scheduler_, DescriptorPool& descriptor_pool_,
-    StagingBufferPool& staging_buffer_pool_,
-    ComputePassDescriptorQueue& compute_pass_descriptor_queue_)
-    : ComputePass(device_, scheduler_, descriptor_pool_, INPUT_OUTPUT_DESCRIPTOR_SET_BINDINGS,
-                  INPUT_OUTPUT_DESCRIPTOR_UPDATE_TEMPLATE, INPUT_OUTPUT_BANK_INFO,
-                  COMPUTE_PUSH_CONSTANT_RANGE<sizeof(BlockLinearUnswizzle2DBufferPushConstants)>,
-                  VULKAN_BLOCK_LINEAR_UNSWIZZLE_2D_BUFFER_COMP_SPV),
-      scheduler{scheduler_}, staging_buffer_pool{staging_buffer_pool_},
-      compute_pass_descriptor_queue{compute_pass_descriptor_queue_} {}
-
-BlockLinearUnswizzle2DBufferPass::~BlockLinearUnswizzle2DBufferPass() = default;
-
-bool BlockLinearUnswizzle2DBufferPass::IsSupported(const Image& image) {
-    const VideoCommon::ImageInfo& info = image.info;
-    if (info.type != VideoCommon::ImageType::e2D || info.num_samples != 1 ||
-        VideoCore::Surface::GetFormatType(info.format) != VideoCore::Surface::SurfaceType::ColorTexture ||
-        VideoCore::Surface::DefaultBlockWidth(info.format) != 1 ||
-        VideoCore::Surface::DefaultBlockHeight(info.format) != 1 ||
-        True(image.flags & VideoCommon::ImageFlagBits::Converted) ||
-        True(image.flags & VideoCommon::ImageFlagBits::Rescaled)) {
-        return false;
-    }
-    const u32 bytes_per_block = VideoCore::Surface::BytesPerBlock(info.format);
-    for (const VideoCommon::SwizzleParameters& swizzle : VideoCommon::FullUploadSwizzles(info)) {
-        if ((swizzle.num_tiles.width * bytes_per_block) % 4 != 0 || swizzle.num_tiles.depth != 1) {
-            return false;
-        }
-    }
-    return true;
-}
-
-std::pair<VkBuffer, boost::container::small_vector<VkBufferImageCopy, 16>>
-BlockLinearUnswizzle2DBufferPass::Unswizzle(Image& image, VkBuffer src_buffer,
-                                            VkDeviceSize src_offset) {
-    const VideoCommon::ImageInfo& info = image.info;
-    const auto swizzles = VideoCommon::FullUploadSwizzles(info);
-    const u32 bytes_per_block = VideoCore::Surface::BytesPerBlock(info.format);
-    const u32 layers = static_cast<u32>(info.resources.layers);
-    const VkDeviceSize alignment = (std::max<VkDeviceSize>)(device.GetStorageBufferAlignment(), 4);
-
-    VkDeviceSize total_size = 0;
-    for (const VideoCommon::SwizzleParameters& swizzle : swizzles) {
-        total_size = Common::AlignUp(total_size, alignment);
-        total_size += VkDeviceSize{swizzle.num_tiles.width} * bytes_per_block *
-                      swizzle.num_tiles.height * layers;
-    }
-    const StagingBufferRef linear = staging_buffer_pool.Request(total_size, MemoryUsage::DeviceLocal);
-
-    scheduler.RequestOutsideRenderPassOperationContext();
-    scheduler.Record([](vk::CommandBuffer cmdbuf) {
-        static constexpr VkMemoryBarrier READ_BARRIER{
-            .sType = VK_STRUCTURE_TYPE_MEMORY_BARRIER,
-            .pNext = nullptr,
-            .srcAccessMask = VK_ACCESS_SHADER_WRITE_BIT | VK_ACCESS_TRANSFER_WRITE_BIT,
-            .dstAccessMask = VK_ACCESS_SHADER_READ_BIT,
-        };
-        cmdbuf.PipelineBarrier(VK_PIPELINE_STAGE_ALL_COMMANDS_BIT,
-                               VK_PIPELINE_STAGE_COMPUTE_SHADER_BIT, 0, READ_BARRIER);
-    });
-
-    boost::container::small_vector<VkBufferImageCopy, 16> copies;
-    VkDeviceSize linear_offset = 0;
-    for (const VideoCommon::SwizzleParameters& swizzle : swizzles) {
-        linear_offset = Common::AlignUp(linear_offset, alignment);
-        const auto params = VideoCommon::Accelerated::MakeBlockLinearSwizzle2DParams(swizzle, info);
-        const u32 row_words = swizzle.num_tiles.width * bytes_per_block / 4;
-        const u32 height = swizzle.num_tiles.height;
-        const VkDeviceSize level_size = VkDeviceSize{row_words} * 4 * height * layers;
-        const VkDeviceSize level_src = src_offset + swizzle.buffer_offset;
-        const VkDeviceSize aligned_src = level_src & ~(alignment - 1);
-        const VkDeviceSize src_size = image.guest_size_bytes - swizzle.buffer_offset +
-                                      (level_src - aligned_src);
-
-        compute_pass_descriptor_queue.Acquire(scheduler, 2);
-        compute_pass_descriptor_queue.AddBuffer(src_buffer, aligned_src, src_size);
-        compute_pass_descriptor_queue.AddBuffer(linear.buffer, linear.offset + linear_offset,
-                                                level_size);
-        const void* const descriptor_data{compute_pass_descriptor_queue.UpdateData()};
-        const BlockLinearUnswizzle2DBufferPushConstants push{
-            .src_word_offset = static_cast<u32>((level_src - aligned_src) / 4),
-            .row_words = row_words,
-            .height = height,
-            .layer_stride = params.layer_stride,
-            .block_size = params.block_size,
-            .x_shift = params.x_shift,
-            .block_height = params.block_height,
-            .block_height_mask = params.block_height_mask,
-        };
-        scheduler.Record([this, descriptor_data, push, layers](vk::CommandBuffer cmdbuf) {
-            static constexpr u32 WORKGROUP_SIZE_X = 64;
-            const VkDescriptorSet set = descriptor_allocator.Commit();
-            device.GetLogical().UpdateDescriptorSet(set, *descriptor_template, descriptor_data);
-            cmdbuf.BindPipeline(VK_PIPELINE_BIND_POINT_COMPUTE, *pipeline);
-            cmdbuf.BindDescriptorSets(VK_PIPELINE_BIND_POINT_COMPUTE, *layout, 0, set, {});
-            cmdbuf.PushConstants(*layout, VK_SHADER_STAGE_COMPUTE_BIT, push);
-            cmdbuf.Dispatch(Common::DivCeil(push.row_words, WORKGROUP_SIZE_X), push.height, layers);
-        });
-        copies.push_back(VkBufferImageCopy{
-            .bufferOffset = linear.offset + linear_offset,
-            .bufferRowLength = 0,
-            .bufferImageHeight = 0,
-            .imageSubresource{
-                .aspectMask = VK_IMAGE_ASPECT_COLOR_BIT,
-                .mipLevel = static_cast<u32>(swizzle.level),
-                .baseArrayLayer = 0,
-                .layerCount = layers,
-            },
-            .imageOffset{0, 0, 0},
-            .imageExtent{swizzle.num_tiles.width, height, 1},
-        });
-        linear_offset += level_size;
-    }
-    scheduler.Record([](vk::CommandBuffer cmdbuf) {
-        static constexpr VkMemoryBarrier WRITE_BARRIER{
-            .sType = VK_STRUCTURE_TYPE_MEMORY_BARRIER,
-            .pNext = nullptr,
-            .srcAccessMask = VK_ACCESS_SHADER_WRITE_BIT,
-            .dstAccessMask = VK_ACCESS_TRANSFER_READ_BIT,
-        };
-        cmdbuf.PipelineBarrier(VK_PIPELINE_STAGE_COMPUTE_SHADER_BIT,
-                               VK_PIPELINE_STAGE_TRANSFER_BIT, 0, WRITE_BARRIER);
-    });
-    return {linear.buffer, std::move(copies)};
-}
-
 BlockLinearUnswizzle3DPass::BlockLinearUnswizzle3DPass(
     const Device& device_, Scheduler& scheduler_,
     DescriptorPool& descriptor_pool_,

src/video_core/renderer_vulkan/vk_compute_pass.h

@@ -10,8 +10,6 @@
 #include <span>
 #include <utility>
 
-#include <boost/container/small_vector.hpp>
-
 #include "common/common_types.h"
 #include "video_core/engines/maxwell_3d.h"
 #include "video_core/renderer_vulkan/vk_descriptor_pool.h"
@@ -139,28 +137,6 @@ private:
     MemoryAllocator& memory_allocator;
 };
 
-class BlockLinearUnswizzle2DBufferPass final : public ComputePass {
-public:
-    explicit BlockLinearUnswizzle2DBufferPass(
-        const Device& device_, Scheduler& scheduler_, DescriptorPool& descriptor_pool_,
-        StagingBufferPool& staging_buffer_pool_,
-        ComputePassDescriptorQueue& compute_pass_descriptor_queue_);
-    ~BlockLinearUnswizzle2DBufferPass();
-
-    /// Returns true when the guest layout of the image can be unswizzled by this pass
-    [[nodiscard]] static bool IsSupported(const Image& image);
-
-    /// Unswizzles the block linear guest data of the image, read from a GPU buffer, into a linear
-    /// buffer. Returns the linear buffer and the copies that upload it into the image.
-    std::pair<VkBuffer, boost::container::small_vector<VkBufferImageCopy, 16>> Unswizzle(
-        Image& image, VkBuffer src_buffer, VkDeviceSize src_offset);
-
-private:
-    Scheduler& scheduler;
-    StagingBufferPool& staging_buffer_pool;
-    ComputePassDescriptorQueue& compute_pass_descriptor_queue;
-};
-
 class BlockLinearUnswizzle3DPass final : public ComputePass {
 public:
     explicit BlockLinearUnswizzle3DPass(const Device& device_, Scheduler& scheduler_,

src/video_core/renderer_vulkan/vk_compute_pipeline.cpp

@@ -174,36 +174,21 @@ bool ComputePipeline::Configure(Tegra::Engines::KeplerCompute& kepler_compute,
     const auto read_handle{[&](const auto& desc, u32 index) {
         ASSERT(((qmd.const_buffer_enable_mask >> desc.cbuf_index) & 1) != 0);
         const u32 index_offset{index << desc.size_shift};
-        u32 offset{desc.cbuf_offset + index_offset};
-        if constexpr (std::is_same_v<decltype(desc), const Shader::TextureDescriptor&>) {
-            if (desc.has_index_cbuf) {
-                ASSERT(((qmd.const_buffer_enable_mask >> desc.index_cbuf_index) & 1) != 0);
-                offset += gpu_memory.Read<u32>(cbufs[desc.index_cbuf_index].Address() +
-                                               desc.index_cbuf_offset)
-                          << desc.index_shift;
-            }
-        }
+        const u32 offset{desc.cbuf_offset + index_offset};
         const GPUVAddr addr{cbufs[desc.cbuf_index].Address() + offset};
         if constexpr (std::is_same_v<decltype(desc), const Shader::TextureDescriptor&> ||
                       std::is_same_v<decltype(desc), const Shader::TextureBufferDescriptor&>) {
             if (desc.has_secondary) {
                 ASSERT(((qmd.const_buffer_enable_mask >> desc.secondary_cbuf_index) & 1) != 0);
-                const u32 secondary_offset{desc.secondary_cbuf_offset};
+                const u32 secondary_offset{desc.secondary_cbuf_offset + index_offset};
                 const GPUVAddr separate_addr{cbufs[desc.secondary_cbuf_index].Address() +
                                              secondary_offset};
-                const u32 lhs_raw{(gpu_memory.Read<u32>(addr) << desc.shift_left) & desc.mask};
-                const u32 rhs_raw{(gpu_memory.Read<u32>(separate_addr)
-                                   << desc.secondary_shift_left) &
-                                  desc.secondary_mask};
+                const u32 lhs_raw{gpu_memory.Read<u32>(addr) << desc.shift_left};
+                const u32 rhs_raw{gpu_memory.Read<u32>(separate_addr) << desc.secondary_shift_left};
                 return TexturePair(lhs_raw | rhs_raw, via_header_index);
             }
         }
-        u32 raw{gpu_memory.Read<u32>(addr)};
-        if constexpr (std::is_same_v<decltype(desc), const Shader::TextureDescriptor&> ||
-                      std::is_same_v<decltype(desc), const Shader::TextureBufferDescriptor&>) {
-            raw &= desc.mask;
-        }
-        return TexturePair(raw, via_header_index);
+        return TexturePair(gpu_memory.Read<u32>(addr), via_header_index);
     }};
     const auto add_image{[&](const auto& desc, bool blacklist) {
         for (u32 index = 0; index < desc.count; ++index) {

src/video_core/renderer_vulkan/vk_graphics_pipeline.cpp

@@ -377,37 +377,23 @@ bool GraphicsPipeline::ConfigureImpl(bool is_indexed) {
         const auto read_handle{[&](const auto& desc, u32 index) {
             ASSERT(cbufs[desc.cbuf_index].enabled);
             const u32 index_offset{index << desc.size_shift};
-            u32 offset{desc.cbuf_offset + index_offset};
-            if constexpr (std::is_same_v<decltype(desc), const Shader::TextureDescriptor&>) {
-                if (desc.has_index_cbuf) {
-                    ASSERT(cbufs[desc.index_cbuf_index].enabled);
-                    offset += gpu_memory->Read<u32>(cbufs[desc.index_cbuf_index].address +
-                                                    desc.index_cbuf_offset)
-                              << desc.index_shift;
-                }
-            }
+            const u32 offset{desc.cbuf_offset + index_offset};
             const GPUVAddr addr{cbufs[desc.cbuf_index].address + offset};
             if constexpr (std::is_same_v<decltype(desc), const Shader::TextureDescriptor&> ||
                           std::is_same_v<decltype(desc), const Shader::TextureBufferDescriptor&>) {
                 if (desc.has_secondary) {
                     ASSERT(cbufs[desc.secondary_cbuf_index].enabled);
-                    const u32 second_offset{desc.secondary_cbuf_offset};
+                    const u32 second_offset{desc.secondary_cbuf_offset + index_offset};
                     const GPUVAddr separate_addr{cbufs[desc.secondary_cbuf_index].address +
                                                  second_offset};
-                    const u32 lhs_raw{(gpu_memory->Read<u32>(addr) << desc.shift_left) & desc.mask};
-                    const u32 rhs_raw{(gpu_memory->Read<u32>(separate_addr)
-                                       << desc.secondary_shift_left) &
-                                      desc.secondary_mask};
+                    const u32 lhs_raw{gpu_memory->Read<u32>(addr) << desc.shift_left};
+                    const u32 rhs_raw{gpu_memory->Read<u32>(separate_addr)
+                                      << desc.secondary_shift_left};
                     const u32 raw{lhs_raw | rhs_raw};
                     return TexturePair(raw, via_header_index);
                 }
             }
-            u32 raw{gpu_memory->Read<u32>(addr)};
-            if constexpr (std::is_same_v<decltype(desc), const Shader::TextureDescriptor&> ||
-                          std::is_same_v<decltype(desc), const Shader::TextureBufferDescriptor&>) {
-                raw &= desc.mask;
-            }
-            return TexturePair(raw, via_header_index);
+            return TexturePair(gpu_memory->Read<u32>(addr), via_header_index);
         }};
         const auto add_image{[&](const auto& desc, bool blacklist) LAMBDA_FORCEINLINE {
             for (u32 index = 0; index < desc.count; ++index) {

src/video_core/renderer_vulkan/vk_pipeline_cache.cpp

@@ -447,7 +447,9 @@ PipelineCache::PipelineCache(Tegra::MaxwellDeviceMemoryManager& device_memory_,
                                        driver_id == VK_DRIVER_ID_AMD_OPEN_SOURCE ||
                                        driver_id == VK_DRIVER_ID_MESA_RADV ||
                                        driver_id == VK_DRIVER_ID_INTEL_PROPRIETARY_WINDOWS ||
-                                       driver_id == VK_DRIVER_ID_INTEL_OPEN_SOURCE_MESA,
+                                       driver_id == VK_DRIVER_ID_INTEL_OPEN_SOURCE_MESA ||
+                                       driver_id == VK_DRIVER_ID_QUALCOMM_PROPRIETARY ||
+                                       driver_id == VK_DRIVER_ID_MESA_TURNIP,
 
         .has_broken_spirv_clamp = driver_id == VK_DRIVER_ID_INTEL_PROPRIETARY_WINDOWS,
         .has_broken_spirv_position_input = driver_id == VK_DRIVER_ID_QUALCOMM_PROPRIETARY,
@@ -787,6 +789,11 @@ GraphicsPipeline* PipelineCache::BuiltPipeline(GraphicsPipeline* pipeline) const
     if (!use_asynchronous_shaders) {
         return pipeline;
     }
+    // If something is using depth, we can assume that games are not rendering anything which
+    // will be used one time.
+    if (maxwell3d->regs.zeta_enable) {
+        return nullptr;
+    }
     // If games are using a small index count, we can assume these are full screen quads.
     // Usually these shaders are only used once for building textures so we can assume they
     // can't be built async

src/video_core/renderer_vulkan/vk_rasterizer.cpp

@@ -596,12 +596,6 @@ void RasterizerVulkan::DispatchCompute() {
         return;
     }
 
-    const auto update_images_written_by_buffers = [this] {
-        buffer_cache.ForEachWrittenComputeStorageBuffer(
-            [this](auto& buffer, u32 offset, DAddr device_addr, u32 size) {
-                texture_cache.UpdateImagesFromBuffer(device_addr, size, buffer.Handle(), offset);
-            });
-    };
     const auto& qmd{kepler_compute->launch_description};
     auto indirect_address = kepler_compute->GetIndirectComputeAddress();
     if (indirect_address) {
@@ -618,7 +612,6 @@ void RasterizerVulkan::DispatchCompute() {
             }
             cmdbuf.DispatchIndirect(indirect_buffer, indirect_offset);
         });
-        update_images_written_by_buffers();
         return;
     }
     const std::array<u32, 3> dim{qmd.grid_dim_x, qmd.grid_dim_y, qmd.grid_dim_z};
@@ -641,7 +634,6 @@ void RasterizerVulkan::DispatchCompute() {
         }
         cmdbuf.Dispatch(dim[0], dim[1], dim[2]);
     });
-    update_images_written_by_buffers();
 
     // Log compute dispatch
     if (GPU::Logging::IsActive() &&
@@ -806,6 +798,18 @@ void RasterizerVulkan::OnCacheInvalidation(DAddr addr, u64 size) {
     pipeline_cache.InvalidateRegion(addr, size);
 }
 
+bool RasterizerVulkan::TryCacheInvalidation(DAddr addr, u64 size) {
+    if (addr == 0 || size == 0) return true;
+    {
+        std::scoped_lock lock{buffer_cache.mutex, texture_cache.mutex};
+        if (buffer_cache.IsRegionGpuModified(addr, size)) return false;
+        texture_cache.WriteMemory(addr, size);
+        buffer_cache.WriteMemory(addr, size);
+    }
+    pipeline_cache.InvalidateRegion(addr, size);
+    return true;
+}
+
 void RasterizerVulkan::InvalidateGPUCache() {
     gpu.InvalidateGPUCache();
 }

src/video_core/renderer_vulkan/vk_rasterizer.h

@@ -106,6 +106,7 @@ public:
                           VideoCommon::CacheType which = VideoCommon::CacheType::All) override;
     void InnerInvalidation(std::span<const std::pair<DAddr, std::size_t>> sequences) override;
     void OnCacheInvalidation(DAddr addr, u64 size) override;
+    bool TryCacheInvalidation(DAddr addr, u64 size) override;
     bool OnCPUWrite(DAddr addr, u64 size) override;
     void InvalidateGPUCache() override;
     void UnmapMemory(DAddr addr, u64 size) override;

src/video_core/renderer_vulkan/vk_texture_cache.cpp

@@ -946,8 +946,6 @@ TextureCacheRuntime::TextureCacheRuntime(const Device& device_, Scheduler& sched
         astc_decoder_pass.emplace(device, scheduler, descriptor_pool, staging_buffer_pool,
                                   compute_pass_descriptor_queue, memory_allocator);
     }
-    bl2d_buffer_unswizzle_pass.emplace(device, scheduler, descriptor_pool, staging_buffer_pool,
-                                       compute_pass_descriptor_queue);
     if (!device.IsKhrImageFormatListSupported()) {
         return;
     }
@@ -1091,58 +1089,6 @@ VkImageView TextureCacheRuntime::GetOrCreateResolveShadow(VkImage msaa_image, Vk
     return *shadow.view;
 }
 
-VkImageView TextureCacheRuntime::GetOrCreateTirColorScratch(VkImage color_image, VkFormat format,
-                                                            VkExtent2D extent, u32 layers,
-                                                            VkSampleCountFlagBits samples) {
-    TirColorScratch& scratch = tir_color_scratch[color_image];
-    if (scratch.image && scratch.format == format && scratch.extent.width == extent.width &&
-        scratch.extent.height == extent.height && scratch.layers == layers &&
-        scratch.samples == samples) {
-        return *scratch.view;
-    }
-    if (scratch.image) {
-        scheduler.Finish();
-    }
-    scratch.image = memory_allocator.CreateImage(VkImageCreateInfo{
-        .sType = VK_STRUCTURE_TYPE_IMAGE_CREATE_INFO,
-        .pNext = nullptr,
-        .flags = 0,
-        .imageType = VK_IMAGE_TYPE_2D,
-        .format = format,
-        .extent = {extent.width, extent.height, 1},
-        .mipLevels = 1,
-        .arrayLayers = layers,
-        .samples = samples,
-        .tiling = VK_IMAGE_TILING_OPTIMAL,
-        .usage = VK_IMAGE_USAGE_COLOR_ATTACHMENT_BIT | VK_IMAGE_USAGE_TRANSFER_SRC_BIT,
-        .sharingMode = VK_SHARING_MODE_EXCLUSIVE,
-        .queueFamilyIndexCount = 0,
-        .pQueueFamilyIndices = nullptr,
-        .initialLayout = VK_IMAGE_LAYOUT_UNDEFINED,
-    });
-    scratch.view = device.GetLogical().CreateImageView(VkImageViewCreateInfo{
-        .sType = VK_STRUCTURE_TYPE_IMAGE_VIEW_CREATE_INFO,
-        .pNext = nullptr,
-        .flags = 0,
-        .image = *scratch.image,
-        .viewType = layers > 1 ? VK_IMAGE_VIEW_TYPE_2D_ARRAY : VK_IMAGE_VIEW_TYPE_2D,
-        .format = format,
-        .components{},
-        .subresourceRange{
-            .aspectMask = VK_IMAGE_ASPECT_COLOR_BIT,
-            .baseMipLevel = 0,
-            .levelCount = 1,
-            .baseArrayLayer = 0,
-            .layerCount = layers,
-        },
-    });
-    scratch.format = format;
-    scratch.extent = extent;
-    scratch.layers = layers;
-    scratch.samples = samples;
-    return *scratch.view;
-}
-
 const TextureCacheRuntime::ResolveShadow* TextureCacheRuntime::GetValidResolveShadow(
     VkImage msaa_image) const {
     const auto it = resolve_shadows.find(msaa_image);
@@ -2122,27 +2068,6 @@ void Image::UploadMemory(VkBuffer buffer, VkDeviceSize offset,
     }
 }
 
-bool Image::UploadSwizzledFromBuffer(VkBuffer buffer, VkDeviceSize offset) {
-    if (runtime == nullptr || !runtime->bl2d_buffer_unswizzle_pass || offset % 4 != 0 ||
-        guest_size_bytes > runtime->device.GetMaxStorageBufferRange() ||
-        !BlockLinearUnswizzle2DBufferPass::IsSupported(*this)) {
-        return false;
-    }
-    if (ENABLE_MSAA_RESOLVE_CONSUME) {
-        runtime->InvalidateResolveShadow(Handle());
-    }
-    auto [linear_buffer, copies] =
-        runtime->bl2d_buffer_unswizzle_pass->Unswizzle(*this, buffer, offset);
-    const VkImage vk_image = *original_image;
-    const VkImageAspectFlags vk_aspect_mask = aspect_mask;
-    const bool was_initialized = std::exchange(initialized, true);
-    scheduler->Record([linear_buffer, vk_image, vk_aspect_mask, was_initialized,
-                       copies = std::move(copies)](vk::CommandBuffer cmdbuf) {
-        CopyBufferToImage(cmdbuf, linear_buffer, vk_image, vk_aspect_mask, was_initialized, copies);
-    });
-    return true;
-}
-
 void Image::UploadMemory(const StagingBufferRef& map, std::span<const BufferImageCopy> copies) {
     UploadMemory(map.buffer, map.offset, copies);
 }
@@ -3045,26 +2970,6 @@ Framebuffer::Framebuffer(TextureCacheRuntime& runtime, ImageView* color_buffer,
 
 Framebuffer::~Framebuffer() = default;
 
-/// Single-sampled color targets with a multisampled depth buffer, as used by target-independent
-/// rasterization. Vulkan requires matching sample counts for these attachments.
-static bool IsMixedSamplesTarget(std::span<ImageView*, NUM_RT> color_buffers,
-                                 const ImageView* depth_buffer) {
-    if (depth_buffer == nullptr || depth_buffer->Samples() == VK_SAMPLE_COUNT_1_BIT) {
-        return false;
-    }
-    bool has_color = false;
-    for (const ImageView* const color_buffer : color_buffers) {
-        if (color_buffer == nullptr) {
-            continue;
-        }
-        if (color_buffer->Samples() != VK_SAMPLE_COUNT_1_BIT) {
-            return false;
-        }
-        has_color = true;
-    }
-    return has_color;
-}
-
 void Framebuffer::CreateFramebuffer(TextureCacheRuntime& runtime,
                                     std::span<ImageView*, NUM_RT> color_buffers,
                                     ImageView* depth_buffer, bool is_rescaled_) {
@@ -3119,13 +3024,6 @@ void Framebuffer::CreateFramebuffer(TextureCacheRuntime& runtime,
     } else {
         renderpass_key.depth_format = PixelFormat::Invalid;
     }
-    if (IsMixedSamplesTarget(color_buffers, depth_buffer)) {
-        const VkExtent2D extent{(std::min)(render_area.width, width),
-                                (std::min)(render_area.height, height)};
-        CreateMixedSamplesFramebuffer(runtime, color_buffers, attachments, renderpass_key,
-                                      static_cast<u32>((std::max)(num_layers, 1)), extent);
-        return;
-    }
     renderpass_key.samples = samples;
     const bool do_resolve_color = ENABLE_MSAA_TILER_RESOLVE &&
         samples != VK_SAMPLE_COUNT_1_BIT && num_colors > 0 && runtime.device.IsTiler();
@@ -3142,7 +3040,10 @@ void Framebuffer::CreateFramebuffer(TextureCacheRuntime& runtime,
     discard_msaa_depth_stencil = ENABLE_MSAA_RESOLVE_CONSUME &&
                                  ENABLE_MSAA_DEPTH_STENCIL_DISCARD && do_resolve_depth_stencil;
 
-    UseRenderPass(runtime, renderpass_key);
+    renderpass = runtime.render_pass_cache.Get(renderpass_key);
+    render_pass_key = renderpass_key;
+    render_pass_cache = &runtime.render_pass_cache;
+    runtime_ptr = &runtime;
     render_area.width = (std::min)(render_area.width, width);
     render_area.height = (std::min)(render_area.height, height);
 
@@ -3174,53 +3075,6 @@ void Framebuffer::CreateFramebuffer(TextureCacheRuntime& runtime,
     }
 
     num_color_buffers = static_cast<u32>(num_colors);
-    CreateVkFramebuffer(runtime, attachments, static_cast<u32>((std::max)(num_layers, 1)));
-}
-
-void Framebuffer::CreateMixedSamplesFramebuffer(TextureCacheRuntime& runtime,
-                                                std::span<ImageView*, NUM_RT> color_buffers,
-                                                std::span<const VkImageView> attachments,
-                                                RenderPassKey renderpass_key, u32 layers,
-                                                VkExtent2D extent) {
-    // attachments holds the real color targets followed by the depth buffer
-    const size_t num_colors = attachments.size() - 1;
-    boost::container::small_vector<VkImageView, NUM_RT * 2 + 2> mixed_attachments;
-    for (const ImageView* const color_buffer : color_buffers) {
-        if (!color_buffer) {
-            continue;
-        }
-        const VkFormat vk_format = MaxwellToVK::SurfaceFormat(runtime.device, FormatType::Optimal,
-                                                              true, color_buffer->format)
-                                       .format;
-        mixed_attachments.push_back(runtime.GetOrCreateTirColorScratch(
-            color_buffer->ImageHandle(), vk_format, extent, layers, samples));
-    }
-    mixed_attachments.push_back(attachments[num_colors]);
-    for (size_t index = 0; index < num_colors; ++index) {
-        mixed_attachments.push_back(attachments[index]);
-    }
-    // Multisampled scratch color + depth, resolving color into the real single-sampled targets
-    renderpass_key.samples = samples;
-    renderpass_key.resolve_color = true;
-    renderpass_key.resolve_depth_stencil = false;
-    discard_msaa_color = false;
-    discard_msaa_depth_stencil = false;
-    UseRenderPass(runtime, renderpass_key);
-    render_area = extent;
-    num_color_buffers = static_cast<u32>(num_colors);
-    CreateVkFramebuffer(runtime, mixed_attachments, layers);
-}
-
-void Framebuffer::UseRenderPass(TextureCacheRuntime& runtime,
-                                const RenderPassKey& renderpass_key) {
-    renderpass = runtime.render_pass_cache.Get(renderpass_key);
-    render_pass_key = renderpass_key;
-    render_pass_cache = &runtime.render_pass_cache;
-    runtime_ptr = &runtime;
-}
-
-void Framebuffer::CreateVkFramebuffer(TextureCacheRuntime& runtime,
-                                      std::span<const VkImageView> attachments, u32 layers) {
     framebuffer = runtime.device.GetLogical().CreateFramebuffer({
         .sType = VK_STRUCTURE_TYPE_FRAMEBUFFER_CREATE_INFO,
         .pNext = nullptr,
@@ -3230,7 +3084,7 @@ void Framebuffer::CreateVkFramebuffer(TextureCacheRuntime& runtime,
         .pAttachments = attachments.data(),
         .width = render_area.width,
         .height = render_area.height,
-        .layers = layers,
+        .layers = static_cast<u32>((std::max)(num_layers, 1)),
     });
 }
 

src/video_core/renderer_vulkan/vk_texture_cache.h

@@ -133,11 +133,6 @@ public:
 
     [[nodiscard]] const ResolveShadow* GetValidResolveShadow(VkImage msaa_image) const;
 
-    /// Multisampled stand-in for a single-sampled color target rendered with a TIR depth buffer
-    [[nodiscard]] VkImageView GetOrCreateTirColorScratch(VkImage color_image, VkFormat format,
-                                                         VkExtent2D extent, u32 layers,
-                                                         VkSampleCountFlagBits samples);
-
     void InvalidateResolveShadow(VkImage msaa_image);
 
     void MarkResolveShadowUpToDate(VkImage msaa_image);
@@ -164,7 +159,6 @@ public:
     std::optional<ASTCDecoderPass> astc_decoder_pass;
 
     std::optional<BlockLinearUnswizzle3DPass> bl3d_unswizzle_pass;
-    std::optional<BlockLinearUnswizzle2DBufferPass> bl2d_buffer_unswizzle_pass;
     const Settings::ResolutionScalingInfo& resolution;
     std::array<std::vector<VkFormat>, VideoCore::Surface::MaxPixelFormat> view_formats;
 
@@ -193,15 +187,6 @@ public:
 
     std::vector<MsaaScratchImage> msaa_scratch_images;
     ::Common::unordered_map<VkImage, ResolveShadow> resolve_shadows;
-    struct TirColorScratch {
-        vk::Image image;
-        vk::ImageView view;
-        VkFormat format = VK_FORMAT_UNDEFINED;
-        VkExtent2D extent{};
-        u32 layers = 0;
-        VkSampleCountFlagBits samples = VK_SAMPLE_COUNT_1_BIT;
-    };
-    ::Common::unordered_map<VkImage, TirColorScratch> tir_color_scratch;
     std::vector<std::pair<u64, ResolveShadow>> pending_resolve_shadows;
 };
 
@@ -294,22 +279,6 @@ public:
     void MarkResolveShadowsUpToDate() const;
 
 private:
-    /// Builds a framebuffer for single-sampled color targets drawn with a multisampled depth
-    /// buffer (target-independent rasterization). Color is rendered into multisampled scratch
-    /// images that resolve into the real targets at the end of the pass.
-    void CreateMixedSamplesFramebuffer(TextureCacheRuntime& runtime,
-                                       std::span<ImageView*, NUM_RT> color_buffers,
-                                       std::span<const VkImageView> attachments,
-                                       RenderPassKey renderpass_key, u32 layers,
-                                       VkExtent2D extent);
-
-    /// Fetches the render pass for the key from the cache and remembers it for variants.
-    void UseRenderPass(TextureCacheRuntime& runtime, const RenderPassKey& renderpass_key);
-
-    /// Creates the Vulkan framebuffer for the current render pass and render area.
-    void CreateVkFramebuffer(TextureCacheRuntime& runtime,
-                             std::span<const VkImageView> attachments, u32 layers);
-
     static constexpr size_t NUM_MEMOIZED_RENDER_PASS_VARIANTS = 8;
 
     vk::Framebuffer framebuffer;
@@ -356,10 +325,6 @@ public:
     void UploadMemory(const StagingBufferRef& map,
                       std::span<const VideoCommon::BufferImageCopy> copies);
 
-    /// Replaces the image contents with its block linear guest data held in a GPU buffer.
-    /// Returns false when the image layout is not supported.
-    bool UploadSwizzledFromBuffer(VkBuffer buffer, VkDeviceSize offset);
-
     void DownloadMemory(VkBuffer buffer, size_t offset,
                         std::span<const VideoCommon::BufferImageCopy> copies);
 

src/video_core/renderer_vulkan/vk_texture_cache_base.cpp

@@ -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
 
@@ -9,6 +6,4 @@
 
 namespace VideoCommon {
 template class VideoCommon::TextureCache<Vulkan::TextureCacheParams>;
-template void VideoCommon::TextureCache<Vulkan::TextureCacheParams>::UpdateImagesFromBuffer<VkBuffer>(
-    DAddr, size_t, VkBuffer, u64);
 }

src/video_core/texture_cache/texture_cache.h

@@ -595,36 +595,6 @@ void TextureCache<P>::WriteMemory(DAddr cpu_addr, size_t size) {
     });
 }
 
-template <class P>
-template <typename BufferHandle>
-void TextureCache<P>::UpdateImagesFromBuffer(DAddr cpu_addr, size_t size, BufferHandle buffer,
-                                             u64 buffer_offset) {
-    if constexpr (requires(Image& image) { image.UploadSwizzledFromBuffer(buffer, buffer_offset); }) {
-        // The guest can write a render target through a storage buffer bound over its memory
-        // (e.g. a compute clear). Only images the binding matches exactly are replaced, a larger
-        // binding may hold unrelated live images in the parts the shader does not write.
-        static constexpr size_t MAX_BINDING_PADDING = 64 * 1024;
-        ForEachImageInRegion(cpu_addr, size, [&](ImageId image_id, Image& image) {
-            if (image.cpu_addr != cpu_addr || size < image.guest_size_bytes ||
-                size - image.guest_size_bytes > MAX_BINDING_PADDING ||
-                True(image.flags & ImageFlagBits::Remapped) ||
-                True(image.flags & ImageFlagBits::Sparse)) {
-                return;
-            }
-            if (!image.UploadSwizzledFromBuffer(buffer, buffer_offset)) {
-                return;
-            }
-            if (True(image.flags & ImageFlagBits::CpuModified)) {
-                image.flags &= ~ImageFlagBits::CpuModified;
-                if (False(image.flags & ImageFlagBits::Tracked)) {
-                    TrackImage(image, image_id);
-                }
-            }
-            MarkModification(image);
-        });
-    }
-}
-
 template <class P>
 bool TextureCache<P>::IsDownloadable(const ImageBase& image) const noexcept {
     if (!image.IsSafeGpuCopy()) {
@@ -1958,26 +1928,7 @@ ImageViewId TextureCache<P>::FindDepthBuffer() {
     if (gpu_addr == 0) {
         return ImageViewId{};
     }
-    // Target-independent rasterization: the depth buffer keeps the raster sample count while
-    // the color targets use anti_alias_samples_mode.
-    Tegra::Texture::MsaaMode zeta_msaa_mode = regs.anti_alias_samples_mode;
-    if (regs.anti_alias_raster != Tegra::Engines::Maxwell3D::Regs::AntiAliasRaster::Mode1x1) {
-        const auto raster_mode = static_cast<Tegra::Texture::MsaaMode>(regs.anti_alias_raster);
-        bool use_raster = regs.tir_mode == Tegra::Engines::Maxwell3D::Regs::TIRMode::RasterNTargetM;
-        if (!use_raster && regs.anti_alias_samples_mode == Tegra::Texture::MsaaMode::Msaa1x1) {
-            // Clears can be issued before tir_mode is enabled. Only a depth buffer sized for the
-            // raster samples relative to the color target is treated as multisampled, otherwise
-            // it ping-pongs between a 1x and a multisampled image every frame.
-            const auto& rt0 = regs.rt[0];
-            use_raster = rt0.Address() != 0 && rt0.format != Tegra::RenderTargetFormat::NONE &&
-                         regs.zeta_size.width == rt0.width * NumSamplesX(raster_mode) &&
-                         regs.zeta_size.height == rt0.height * NumSamplesY(raster_mode);
-        }
-        if (use_raster) {
-            zeta_msaa_mode = raster_mode;
-        }
-    }
-    const ImageInfo info(regs.zeta, regs.zeta_size, zeta_msaa_mode);
+    const ImageInfo info(regs.zeta, regs.zeta_size, regs.anti_alias_samples_mode);
     return FindRenderTargetView(info, gpu_addr);
 }
 

src/video_core/texture_cache/texture_cache_base.h

@@ -203,12 +203,6 @@ public:
     /// Mark images in a range as modified from the CPU
     void WriteMemory(DAddr cpu_addr, size_t size);
 
-    /// Replaces the contents of the images fully inside a range that a GPU buffer wrote, with the
-    /// data of that buffer
-    template <typename BufferHandle>
-    void UpdateImagesFromBuffer(DAddr cpu_addr, size_t size, BufferHandle buffer,
-                                u64 buffer_offset);
-
     /// Download contents of host images to guest memory in a region
     void DownloadMemory(DAddr cpu_addr, size_t size);
 

src/yuzu/main_window.cpp

@@ -2797,9 +2797,9 @@ void MainWindow::IncrementInstallProgress() {
 
 void MainWindow::OnMenuInstallToNAND() {
     const QString file_filter =
-        tr("Installable Switch File (*.nca *.nsp *.xci);;Nintendo Content Archive "
-           "(*.nca);;Nintendo Submission Package (*.nsp);;NX Cartridge "
-           "Image (*.xci)");
+        tr("Installable Switch File (*.nca *.nsp *.nsz *.xci *.xcz);;Nintendo Content Archive "
+           "(*.nca);;Nintendo Submission Package (*.nsp *.nsz);;NX Cartridge "
+           "Image (*.xci *.xcz)");
 
     QStringList filenames = QFileDialog::getOpenFileNames(
         this, tr("Install Files"), QString::fromStdString(UISettings::values.roms_path),
@@ -2856,7 +2856,10 @@ void MainWindow::OnMenuInstallToNAND() {
         QFuture<ContentManager::InstallResult> future;
         ContentManager::InstallResult result;
 
-        if (file.endsWith(QStringLiteral("nsp"), Qt::CaseInsensitive)) {
+        if (file.endsWith(QStringLiteral("nsp"), Qt::CaseInsensitive) ||
+            file.endsWith(QStringLiteral("nsz"), Qt::CaseInsensitive) ||
+            file.endsWith(QStringLiteral("xci"), Qt::CaseInsensitive) ||
+            file.endsWith(QStringLiteral("xcz"), Qt::CaseInsensitive)) {
             const auto progress_callback = [this](size_t size, size_t progress) {
                 emit UpdateInstallProgress();
                 if (install_progress->wasCanceled()) {
@@ -3661,12 +3664,12 @@ void MainWindow::OnToggleGraphicsAPI() {
         api = Settings::RendererBackend::OpenGL_GLSL;
         break;
     case Settings::RendererBackend::OpenGL_GLSL:
-        api = Settings::RendererBackend::OpenGL_GLSL;
-        break;
-    case Settings::RendererBackend::OpenGL_SPIRV:
         api = Settings::RendererBackend::OpenGL_GLASM;
         break;
     case Settings::RendererBackend::OpenGL_GLASM:
+        api = Settings::RendererBackend::OpenGL_SPIRV;
+        break;
+    case Settings::RendererBackend::OpenGL_SPIRV:
         api = Settings::RendererBackend::Null;
         break;
 #else