BCAT: Stub CreateDeliveryCacheStorageService, add some services

gl_shader_cache: Remove special casing for geometry shaders

.ci/templates/build-testing.yml

@@ -10,7 +10,7 @@ jobs:
         BuildSuffix: 'windows-testing'
         ScriptFolder: 'windows'
   steps:
-  - script: sudo apt upgrade python3-pip && pip install requests urllib3
+  - script: sudo apt-get update && sudo apt-get --only-upgrade -y install python3-pip && pip install requests urllib3
     displayName: 'Prepare Environment'
   - task: PythonScript@0
     condition: eq(variables['Build.Reason'], 'PullRequest')

dist/license.md

@@ -0,0 +1,31 @@
+The icons in this folder and its subfolders have the following licenses:
+
+Icon Name | License | Origin/Author
+--- | --- | ---
+qt_themes/default/icons/16x16/checked.png | Free for non-commercial use
+qt_themes/default/icons/16x16/failed.png | Free for non-commercial use
+qt_themes/default/icons/16x16/lock.png | CC BY-ND 3.0 | https://icons8.com
+qt_themes/default/icons/256x256/plus_folder.png | CC BY-ND 3.0 | https://icons8.com
+qt_themes/default/icons/48x48/bad_folder.png | CC BY-ND 3.0 | https://icons8.com
+qt_themes/default/icons/48x48/chip.png | CC BY-ND 3.0 | https://icons8.com
+qt_themes/default/icons/48x48/folder.png | CC BY-ND 3.0 | https://icons8.com
+qt_themes/default/icons/48x48/plus.png | CC0 1.0 | Designed by BreadFish64 from the Citra team
+qt_themes/default/icons/48x48/sd_card.png | CC BY-ND 3.0 | https://icons8.com
+qt_themes/qdarkstyle/icons/16x16/checked.png | Free for non-commercial use
+qt_themes/qdarkstyle/icons/16x16/failed.png | Free for non-commercial use
+qt_themes/qdarkstyle/icons/16x16/lock.png | CC BY-ND 3.0 | https://icons8.com
+qt_themes/qdarkstyle/icons/256x256/plus_folder.png | CC BY-ND 3.0 | https://icons8.com
+qt_themes/qdarkstyle/icons/48x48/bad_folder.png | CC BY-ND 3.0 | https://icons8.com
+qt_themes/qdarkstyle/icons/48x48/chip.png | CC BY-ND 3.0 | https://icons8.com
+qt_themes/qdarkstyle/icons/48x48/folder.png | CC BY-ND 3.0 | https://icons8.com
+qt_themes/qdarkstyle/icons/48x48/plus.png | CC0 1.0 | Designed by BreadFish64 from the Citra team
+qt_themes/qdarkstyle/icons/48x48/sd_card.png | CC BY-ND 3.0 | https://icons8.com
+qt_themes/colorful/icons/16x16/lock.png | CC BY-ND 3.0 | https://icons8.com
+qt_themes/colorful/icons/256x256/plus_folder.png | CC BY-ND 3.0 | https://icons8.com
+qt_themes/colorful/icons/48x48/bad_folder.png | CC BY-ND 3.0 | https://icons8.com
+qt_themes/colorful/icons/48x48/chip.png | CC BY-ND 3.0 | https://icons8.com
+qt_themes/colorful/icons/48x48/folder.png | CC BY-ND 3.0 | https://icons8.com
+qt_themes/colorful/icons/48x48/plus.png | CC BY-ND 3.0 | https://icons8.com
+qt_themes/colorful/icons/48x48/sd_card.png | CC BY-ND 3.0 | https://icons8.com
+
+<!-- TODO: Add the license of the yuzu icon -->

dist/qt_themes/colorful/icons/index.theme

@@ -0,0 +1,14 @@
+[Icon Theme]
+Name=colorful
+Comment=Colorful theme
+Inherits=default
+Directories=16x16,48x48,256x256
+ 
+[16x16]
+Size=16
+
+[48x48]
+Size=48
+
+[256x256]
+Size=256

dist/qt_themes/colorful/style.qrc

@@ -0,0 +1,15 @@
+<RCC>
+    <qresource prefix="icons/colorful">
+        <file alias="index.theme">icons/index.theme</file>
+        <file alias="16x16/lock.png">icons/16x16/lock.png</file>
+        <file alias="48x48/bad_folder.png">icons/48x48/bad_folder.png</file>
+        <file alias="48x48/chip.png">icons/48x48/chip.png</file>
+        <file alias="48x48/folder.png">icons/48x48/folder.png</file>
+        <file alias="48x48/plus.png">icons/48x48/plus.png</file>
+        <file alias="48x48/sd_card.png">icons/48x48/sd_card.png</file>
+        <file alias="256x256/plus_folder.png">icons/256x256/plus_folder.png</file>
+    </qresource>
+    <qresource prefix="colorful">
+        <file>style.qss</file>
+    </qresource>
+</RCC>

dist/qt_themes/colorful/style.qss

@@ -0,0 +1,4 @@
+/*
+    This file is intentionally left blank.
+    We do not want to apply any stylesheet for colorful, only icons.
+*/

dist/qt_themes/colorful_dark/icons/index.theme

@@ -0,0 +1,8 @@
+[Icon Theme]
+Name=colorful_dark
+Comment=Colorful theme (Dark style)
+Inherits=default
+Directories=16x16
+ 
+[16x16]
+Size=16

dist/qt_themes/colorful_dark/style.qrc

@@ -0,0 +1,57 @@
+<RCC>
+    <qresource prefix="icons/colorful_dark">
+        <file alias="index.theme">icons/index.theme</file>
+        <file alias="16x16/lock.png">icons/16x16/lock.png</file>
+        <file alias="48x48/bad_folder.png">../colorful/icons/48x48/bad_folder.png</file>
+        <file alias="48x48/chip.png">../colorful/icons/48x48/chip.png</file>
+        <file alias="48x48/folder.png">../colorful/icons/48x48/folder.png</file>
+        <file alias="48x48/plus.png">../colorful/icons/48x48/plus.png</file>
+        <file alias="48x48/sd_card.png">../colorful/icons/48x48/sd_card.png</file>
+        <file alias="256x256/plus_folder.png">../colorful/icons/256x256/plus_folder.png</file>
+    </qresource>
+
+    <qresource prefix="qss_icons">
+        <file alias="rc/up_arrow_disabled.png">../qdarkstyle/rc/up_arrow_disabled.png</file>
+        <file alias="rc/Hmovetoolbar.png">../qdarkstyle/rc/Hmovetoolbar.png</file>
+        <file alias="rc/stylesheet-branch-end.png">../qdarkstyle/rc/stylesheet-branch-end.png</file>
+        <file alias="rc/branch_closed-on.png">../qdarkstyle/rc/branch_closed-on.png</file>
+        <file alias="rc/stylesheet-vline.png">../qdarkstyle/rc/stylesheet-vline.png</file>
+        <file alias="rc/branch_closed.png">../qdarkstyle/rc/branch_closed.png</file>
+        <file alias="rc/branch_open-on.png">../qdarkstyle/rc/branch_open-on.png</file>
+        <file alias="rc/transparent.png">../qdarkstyle/rc/transparent.png</file>
+        <file alias="rc/right_arrow_disabled.png">../qdarkstyle/rc/right_arrow_disabled.png</file>
+        <file alias="rc/sizegrip.png">../qdarkstyle/rc/sizegrip.png</file>
+        <file alias="rc/close.png">../qdarkstyle/rc/close.png</file>
+        <file alias="rc/close-hover.png">../qdarkstyle/rc/close-hover.png</file>
+        <file alias="rc/close-pressed.png">../qdarkstyle/rc/close-pressed.png</file>
+        <file alias="rc/down_arrow.png">../qdarkstyle/rc/down_arrow.png</file>
+        <file alias="rc/Vmovetoolbar.png">../qdarkstyle/rc/Vmovetoolbar.png</file>
+        <file alias="rc/left_arrow.png">../qdarkstyle/rc/left_arrow.png</file>
+        <file alias="rc/stylesheet-branch-more.png">../qdarkstyle/rc/stylesheet-branch-more.png</file>
+        <file alias="rc/up_arrow.png">../qdarkstyle/rc/up_arrow.png</file>
+        <file alias="rc/right_arrow.png">../qdarkstyle/rc/right_arrow.png</file>
+        <file alias="rc/left_arrow_disabled.png">../qdarkstyle/rc/left_arrow_disabled.png</file>
+        <file alias="rc/Hsepartoolbar.png">../qdarkstyle/rc/Hsepartoolbar.png</file>
+        <file alias="rc/branch_open.png">../qdarkstyle/rc/branch_open.png</file>
+        <file alias="rc/Vsepartoolbar.png">../qdarkstyle/rc/Vsepartoolbar.png</file>
+        <file alias="rc/down_arrow_disabled.png">../qdarkstyle/rc/down_arrow_disabled.png</file>
+        <file alias="rc/undock.png">../qdarkstyle/rc/undock.png</file>
+        <file alias="rc/checkbox_checked_disabled.png">../qdarkstyle/rc/checkbox_checked_disabled.png</file>
+        <file alias="rc/checkbox_checked_focus.png">../qdarkstyle/rc/checkbox_checked_focus.png</file>
+        <file alias="rc/checkbox_checked.png">../qdarkstyle/rc/checkbox_checked.png</file>
+        <file alias="rc/checkbox_indeterminate.png">../qdarkstyle/rc/checkbox_indeterminate.png</file>
+        <file alias="rc/checkbox_indeterminate_focus.png">../qdarkstyle/rc/checkbox_indeterminate_focus.png</file>
+        <file alias="rc/checkbox_unchecked_disabled.png">../qdarkstyle/rc/checkbox_unchecked_disabled.png</file>
+        <file alias="rc/checkbox_unchecked_focus.png">../qdarkstyle/rc/checkbox_unchecked_focus.png</file>
+        <file alias="rc/checkbox_unchecked.png">../qdarkstyle/rc/checkbox_unchecked.png</file>
+        <file alias="rc/radio_checked_disabled.png">../qdarkstyle/rc/radio_checked_disabled.png</file>
+        <file alias="rc/radio_checked_focus.png">../qdarkstyle/rc/radio_checked_focus.png</file>
+        <file alias="rc/radio_checked.png">../qdarkstyle/rc/radio_checked.png</file>
+        <file alias="rc/radio_unchecked_disabled.png">../qdarkstyle/rc/radio_unchecked_disabled.png</file>
+        <file alias="rc/radio_unchecked_focus.png">../qdarkstyle/rc/radio_unchecked_focus.png</file>
+        <file alias="rc/radio_unchecked.png">../qdarkstyle/rc/radio_unchecked.png</file>
+    </qresource>
+    <qresource prefix="colorful_dark">
+        <file alias="style.qss">../qdarkstyle/style.qss</file>
+    </qresource>
+</RCC>

dist/qt_themes/default/default.qrc

@@ -5,7 +5,21 @@
         <file alias="16x16/checked.png">icons/16x16/checked.png</file>
 
         <file alias="16x16/failed.png">icons/16x16/failed.png</file>
+
+        <file alias="16x16/lock.png">icons/16x16/lock.png</file>
+
+        <file alias="48x48/bad_folder.png">icons/48x48/bad_folder.png</file>
+      
+        <file alias="48x48/chip.png">icons/48x48/chip.png</file>
+
+        <file alias="48x48/folder.png">icons/48x48/folder.png</file>
+
+        <file alias="48x48/plus.png">icons/48x48/plus.png</file>
+      
+        <file alias="48x48/sd_card.png">icons/48x48/sd_card.png</file>
       
         <file alias="256x256/yuzu.png">icons/256x256/yuzu.png</file>
+
+        <file alias="256x256/plus_folder.png">icons/256x256/plus_folder.png</file>
     </qresource>
 </RCC>

dist/qt_themes/default/icons/index.theme

@@ -1,10 +1,13 @@
 [Icon Theme]
 Name=default
 Comment=default theme
-Directories=16x16,256x256
+Directories=16x16,48x48,256x256
  
 [16x16]
 Size=16
+
+[48x48]
+Size=48
  
 [256x256]
 Size=256

dist/qt_themes/qdarkstyle/icons/index.theme

@@ -2,10 +2,13 @@
 Name=qdarkstyle
 Comment=dark theme
 Inherits=default
-Directories=16x16,256x256
+Directories=16x16,48x48,256x256
  
 [16x16]
 Size=16
- 
+
+[48x48]
+Size=48
+
 [256x256]
 Size=256

dist/qt_themes/qdarkstyle/style.qrc

@@ -1,6 +1,13 @@
 <RCC>
   <qresource prefix="icons/qdarkstyle">
     <file alias="index.theme">icons/index.theme</file>
+    <file alias="16x16/lock.png">icons/16x16/lock.png</file>
+    <file alias="48x48/bad_folder.png">icons/48x48/bad_folder.png</file>
+    <file alias="48x48/chip.png">icons/48x48/chip.png</file>
+    <file alias="48x48/folder.png">icons/48x48/folder.png</file>
+    <file alias="48x48/plus.png">icons/48x48/plus.png</file>
+    <file alias="48x48/sd_card.png">icons/48x48/sd_card.png</file>
+    <file alias="256x256/plus_folder.png">icons/256x256/plus_folder.png</file>
   </qresource>
   <qresource prefix="qss_icons">
     <file>rc/up_arrow_disabled.png</file>

license.txt

@@ -337,3 +337,19 @@ proprietary programs.  If your program is a subroutine library, you may
 consider it more useful to permit linking proprietary applications with the
 library.  If this is what you want to do, use the GNU Lesser General
 Public License instead of this License.
+
+
+The icons used in this project have the following licenses:
+
+Icon Name                          | License       | Origin/Author
+---                                | ---           | ---
+checked.png                        | Free for non-commercial use
+failed.png                         | Free for non-commercial use
+lock.png                           | CC BY-ND 3.0  | https://icons8.com
+plus_folder.png                    | CC BY-ND 3.0  | https://icons8.com
+bad_folder.png                     | CC BY-ND 3.0  | https://icons8.com
+chip.png                           | CC BY-ND 3.0  | https://icons8.com
+folder.png                         | CC BY-ND 3.0  | https://icons8.com
+plus.png (Default, Dark)           | CC0 1.0       | Designed by BreadFish64 from the Citra team
+plus.png (Colorful, Colorful Dark) | CC BY-ND 3.0  | https://icons8.com
+sd_card.png | CC BY-ND 3.0 | https://icons8.com

src/core/CMakeLists.txt

@@ -70,6 +70,8 @@ add_library(core STATIC
     file_sys/sdmc_factory.h
     file_sys/submission_package.cpp
     file_sys/submission_package.h
+    file_sys/system_archive/mii_model.cpp
+    file_sys/system_archive/mii_model.h
     file_sys/system_archive/ng_word.cpp
     file_sys/system_archive/ng_word.h
     file_sys/system_archive/system_archive.cpp

src/core/crypto/key_manager.cpp

@@ -37,6 +37,7 @@
 namespace Core::Crypto {
 
 constexpr u64 CURRENT_CRYPTO_REVISION = 0x5;
+constexpr u64 FULL_TICKET_SIZE = 0x400;
 
 using namespace Common;
 
@@ -55,6 +56,99 @@ const std::map<std::pair<S128KeyType, u64>, std::string> KEYS_VARIABLE_LENGTH{
     {{S128KeyType::KeyblobMAC, 0}, "keyblob_mac_key_"},
 };
 
+namespace {
+template <std::size_t Size>
+bool IsAllZeroArray(const std::array<u8, Size>& array) {
+    return std::all_of(array.begin(), array.end(), [](const auto& elem) { return elem == 0; });
+}
+} // namespace
+
+u64 GetSignatureTypeDataSize(SignatureType type) {
+    switch (type) {
+    case SignatureType::RSA_4096_SHA1:
+    case SignatureType::RSA_4096_SHA256:
+        return 0x200;
+    case SignatureType::RSA_2048_SHA1:
+    case SignatureType::RSA_2048_SHA256:
+        return 0x100;
+    case SignatureType::ECDSA_SHA1:
+    case SignatureType::ECDSA_SHA256:
+        return 0x3C;
+    }
+    UNREACHABLE();
+}
+
+u64 GetSignatureTypePaddingSize(SignatureType type) {
+    switch (type) {
+    case SignatureType::RSA_4096_SHA1:
+    case SignatureType::RSA_4096_SHA256:
+    case SignatureType::RSA_2048_SHA1:
+    case SignatureType::RSA_2048_SHA256:
+        return 0x3C;
+    case SignatureType::ECDSA_SHA1:
+    case SignatureType::ECDSA_SHA256:
+        return 0x40;
+    }
+    UNREACHABLE();
+}
+
+SignatureType Ticket::GetSignatureType() const {
+    if (auto ticket = std::get_if<RSA4096Ticket>(&data)) {
+        return ticket->sig_type;
+    }
+    if (auto ticket = std::get_if<RSA2048Ticket>(&data)) {
+        return ticket->sig_type;
+    }
+    if (auto ticket = std::get_if<ECDSATicket>(&data)) {
+        return ticket->sig_type;
+    }
+
+    UNREACHABLE();
+}
+
+TicketData& Ticket::GetData() {
+    if (auto ticket = std::get_if<RSA4096Ticket>(&data)) {
+        return ticket->data;
+    }
+    if (auto ticket = std::get_if<RSA2048Ticket>(&data)) {
+        return ticket->data;
+    }
+    if (auto ticket = std::get_if<ECDSATicket>(&data)) {
+        return ticket->data;
+    }
+
+    UNREACHABLE();
+}
+
+const TicketData& Ticket::GetData() const {
+    if (auto ticket = std::get_if<RSA4096Ticket>(&data)) {
+        return ticket->data;
+    }
+    if (auto ticket = std::get_if<RSA2048Ticket>(&data)) {
+        return ticket->data;
+    }
+    if (auto ticket = std::get_if<ECDSATicket>(&data)) {
+        return ticket->data;
+    }
+
+    UNREACHABLE();
+}
+
+u64 Ticket::GetSize() const {
+    const auto sig_type = GetSignatureType();
+
+    return sizeof(SignatureType) + GetSignatureTypeDataSize(sig_type) +
+           GetSignatureTypePaddingSize(sig_type) + sizeof(TicketData);
+}
+
+Ticket Ticket::SynthesizeCommon(Key128 title_key, const std::array<u8, 16>& rights_id) {
+    RSA2048Ticket out{};
+    out.sig_type = SignatureType::RSA_2048_SHA256;
+    out.data.rights_id = rights_id;
+    out.data.title_key_common = title_key;
+    return Ticket{out};
+}
+
 Key128 GenerateKeyEncryptionKey(Key128 source, Key128 master, Key128 kek_seed, Key128 key_seed) {
     Key128 out{};
 
@@ -135,6 +229,27 @@ void KeyManager::DeriveGeneralPurposeKeys(std::size_t crypto_revision) {
     }
 }
 
+RSAKeyPair<2048> KeyManager::GetETicketRSAKey() const {
+    if (IsAllZeroArray(eticket_extended_kek) || !HasKey(S128KeyType::ETicketRSAKek))
+        return {};
+
+    const auto eticket_final = GetKey(S128KeyType::ETicketRSAKek);
+
+    std::vector<u8> extended_iv(eticket_extended_kek.begin(), eticket_extended_kek.begin() + 0x10);
+    std::array<u8, 0x230> extended_dec{};
+    AESCipher<Key128> rsa_1(eticket_final, Mode::CTR);
+    rsa_1.SetIV(extended_iv);
+    rsa_1.Transcode(eticket_extended_kek.data() + 0x10, eticket_extended_kek.size() - 0x10,
+                    extended_dec.data(), Op::Decrypt);
+
+    RSAKeyPair<2048> rsa_key{};
+    std::memcpy(rsa_key.decryption_key.data(), extended_dec.data(), rsa_key.decryption_key.size());
+    std::memcpy(rsa_key.modulus.data(), extended_dec.data() + 0x100, rsa_key.modulus.size());
+    std::memcpy(rsa_key.exponent.data(), extended_dec.data() + 0x200, rsa_key.exponent.size());
+
+    return rsa_key;
+}
+
 Key128 DeriveKeyblobMACKey(const Key128& keyblob_key, const Key128& mac_source) {
     AESCipher<Key128> mac_cipher(keyblob_key, Mode::ECB);
     Key128 mac_key{};
@@ -237,7 +352,7 @@ Loader::ResultStatus DeriveSDKeys(std::array<Key256, 2>& sd_keys, KeyManager& ke
     return Loader::ResultStatus::Success;
 }
 
-std::vector<TicketRaw> GetTicketblob(const FileUtil::IOFile& ticket_save) {
+std::vector<Ticket> GetTicketblob(const FileUtil::IOFile& ticket_save) {
     if (!ticket_save.IsOpen())
         return {};
 
@@ -246,14 +361,14 @@ std::vector<TicketRaw> GetTicketblob(const FileUtil::IOFile& ticket_save) {
         return {};
     }
 
-    std::vector<TicketRaw> out;
+    std::vector<Ticket> out;
     for (std::size_t offset = 0; offset + 0x4 < buffer.size(); ++offset) {
         if (buffer[offset] == 0x4 && buffer[offset + 1] == 0x0 && buffer[offset + 2] == 0x1 &&
             buffer[offset + 3] == 0x0) {
             out.emplace_back();
             auto& next = out.back();
-            std::memcpy(&next, buffer.data() + offset, sizeof(TicketRaw));
-            offset += next.size();
+            std::memcpy(&next, buffer.data() + offset, sizeof(Ticket));
+            offset += FULL_TICKET_SIZE;
         }
     }
 
@@ -305,29 +420,23 @@ static std::optional<u64> FindTicketOffset(const std::array<u8, size>& data) {
     return offset;
 }
 
-std::optional<std::pair<Key128, Key128>> ParseTicket(const TicketRaw& ticket,
+std::optional<std::pair<Key128, Key128>> ParseTicket(const Ticket& ticket,
                                                      const RSAKeyPair<2048>& key) {
-    u32 cert_authority;
-    std::memcpy(&cert_authority, ticket.data() + 0x140, sizeof(cert_authority));
-    if (cert_authority == 0)
+    const auto issuer = ticket.GetData().issuer;
+    if (issuer == std::array<u8, 0x40>{})
         return {};
-    if (cert_authority != Common::MakeMagic('R', 'o', 'o', 't')) {
-        LOG_INFO(Crypto,
-                 "Attempting to parse ticket with non-standard certificate authority {:08X}.",
-                 cert_authority);
+    if (issuer[0] != 'R' || issuer[1] != 'o' || issuer[2] != 'o' || issuer[3] != 't') {
+        LOG_INFO(Crypto, "Attempting to parse ticket with non-standard certificate authority.");
     }
 
-    Key128 rights_id;
-    std::memcpy(rights_id.data(), ticket.data() + 0x2A0, sizeof(Key128));
+    Key128 rights_id = ticket.GetData().rights_id;
 
     if (rights_id == Key128{})
         return {};
 
-    Key128 key_temp{};
-
-    if (!std::any_of(ticket.begin() + 0x190, ticket.begin() + 0x280, [](u8 b) { return b != 0; })) {
-        std::memcpy(key_temp.data(), ticket.data() + 0x180, key_temp.size());
-        return std::make_pair(rights_id, key_temp);
+    if (!std::any_of(ticket.GetData().title_key_common_pad.begin(),
+                     ticket.GetData().title_key_common_pad.end(), [](u8 b) { return b != 0; })) {
+        return std::make_pair(rights_id, ticket.GetData().title_key_common);
     }
 
     mbedtls_mpi D; // RSA Private Exponent
@@ -342,7 +451,7 @@ std::optional<std::pair<Key128, Key128>> ParseTicket(const TicketRaw& ticket,
 
     mbedtls_mpi_read_binary(&D, key.decryption_key.data(), key.decryption_key.size());
     mbedtls_mpi_read_binary(&N, key.modulus.data(), key.modulus.size());
-    mbedtls_mpi_read_binary(&S, ticket.data() + 0x180, 0x100);
+    mbedtls_mpi_read_binary(&S, ticket.GetData().title_key_block.data(), 0x100);
 
     mbedtls_mpi_exp_mod(&M, &S, &D, &N, nullptr);
 
@@ -366,6 +475,7 @@ std::optional<std::pair<Key128, Key128>> ParseTicket(const TicketRaw& ticket,
         return {};
     ASSERT(*offset > 0);
 
+    Key128 key_temp{};
     std::memcpy(key_temp.data(), m_2.data() + *offset, key_temp.size());
 
     return std::make_pair(rights_id, key_temp);
@@ -450,6 +560,8 @@ void KeyManager::LoadFromFile(const std::string& filename, bool is_title_keys) {
 
                 const auto index = std::stoul(out[0].substr(18, 2), nullptr, 16);
                 encrypted_keyblobs[index] = Common::HexStringToArray<0xB0>(out[1]);
+            } else if (out[0].compare(0, 20, "eticket_extended_kek") == 0) {
+                eticket_extended_kek = Common::HexStringToArray<576>(out[1]);
             } else {
                 for (const auto& kv : KEYS_VARIABLE_LENGTH) {
                     if (!ValidCryptoRevisionString(out[0], kv.second.size(), 2))
@@ -862,20 +974,19 @@ void KeyManager::DeriveETicket(PartitionDataManager& data) {
     // Titlekeys
     data.DecryptProdInfo(GetBISKey(0));
 
-    const auto eticket_extended_kek = data.GetETicketExtendedKek();
+    eticket_extended_kek = data.GetETicketExtendedKek();
+    WriteKeyToFile(KeyCategory::Console, "eticket_extended_kek", eticket_extended_kek);
+    PopulateTickets();
+}
 
-    std::vector<u8> extended_iv(0x10);
-    std::memcpy(extended_iv.data(), eticket_extended_kek.data(), extended_iv.size());
-    std::array<u8, 0x230> extended_dec{};
-    AESCipher<Key128> rsa_1(eticket_final, Mode::CTR);
-    rsa_1.SetIV(extended_iv);
-    rsa_1.Transcode(eticket_extended_kek.data() + 0x10, eticket_extended_kek.size() - 0x10,
-                    extended_dec.data(), Op::Decrypt);
+void KeyManager::PopulateTickets() {
+    const auto rsa_key = GetETicketRSAKey();
 
-    RSAKeyPair<2048> rsa_key{};
-    std::memcpy(rsa_key.decryption_key.data(), extended_dec.data(), rsa_key.decryption_key.size());
-    std::memcpy(rsa_key.modulus.data(), extended_dec.data() + 0x100, rsa_key.modulus.size());
-    std::memcpy(rsa_key.exponent.data(), extended_dec.data() + 0x200, rsa_key.exponent.size());
+    if (rsa_key == RSAKeyPair<2048>{})
+        return;
+
+    if (!common_tickets.empty() && !personal_tickets.empty())
+        return;
 
     const FileUtil::IOFile save1(FileUtil::GetUserPath(FileUtil::UserPath::NANDDir) +
                                      "/system/save/80000000000000e1",
@@ -886,19 +997,41 @@ void KeyManager::DeriveETicket(PartitionDataManager& data) {
 
     const auto blob2 = GetTicketblob(save2);
     auto res = GetTicketblob(save1);
+    const auto idx = res.size();
     res.insert(res.end(), blob2.begin(), blob2.end());
 
-    for (const auto& raw : res) {
-        const auto pair = ParseTicket(raw, rsa_key);
+    for (std::size_t i = 0; i < res.size(); ++i) {
+        const auto common = i < idx;
+        const auto pair = ParseTicket(res[i], rsa_key);
         if (!pair)
             continue;
         const auto& [rid, key] = *pair;
         u128 rights_id;
         std::memcpy(rights_id.data(), rid.data(), rid.size());
+
+        if (common) {
+            common_tickets[rights_id] = res[i];
+        } else {
+            personal_tickets[rights_id] = res[i];
+        }
+
         SetKey(S128KeyType::Titlekey, key, rights_id[1], rights_id[0]);
     }
 }
 
+void KeyManager::SynthesizeTickets() {
+    for (const auto& key : s128_keys) {
+        if (key.first.type != S128KeyType::Titlekey) {
+            continue;
+        }
+        u128 rights_id{key.first.field1, key.first.field2};
+        Key128 rights_id_2;
+        std::memcpy(rights_id_2.data(), rights_id.data(), rights_id_2.size());
+        const auto ticket = Ticket::SynthesizeCommon(key.second, rights_id_2);
+        common_tickets.insert_or_assign(rights_id, ticket);
+    }
+}
+
 void KeyManager::SetKeyWrapped(S128KeyType id, Key128 key, u64 field1, u64 field2) {
     if (key == Key128{})
         return;
@@ -997,6 +1130,46 @@ void KeyManager::PopulateFromPartitionData(PartitionDataManager& data) {
     DeriveBase();
 }
 
+const std::map<u128, Ticket>& KeyManager::GetCommonTickets() const {
+    return common_tickets;
+}
+
+const std::map<u128, Ticket>& KeyManager::GetPersonalizedTickets() const {
+    return personal_tickets;
+}
+
+bool KeyManager::AddTicketCommon(Ticket raw) {
+    const auto rsa_key = GetETicketRSAKey();
+    if (rsa_key == RSAKeyPair<2048>{})
+        return false;
+
+    const auto pair = ParseTicket(raw, rsa_key);
+    if (!pair)
+        return false;
+    const auto& [rid, key] = *pair;
+    u128 rights_id;
+    std::memcpy(rights_id.data(), rid.data(), rid.size());
+    common_tickets[rights_id] = raw;
+    SetKey(S128KeyType::Titlekey, key, rights_id[1], rights_id[0]);
+    return true;
+}
+
+bool KeyManager::AddTicketPersonalized(Ticket raw) {
+    const auto rsa_key = GetETicketRSAKey();
+    if (rsa_key == RSAKeyPair<2048>{})
+        return false;
+
+    const auto pair = ParseTicket(raw, rsa_key);
+    if (!pair)
+        return false;
+    const auto& [rid, key] = *pair;
+    u128 rights_id;
+    std::memcpy(rights_id.data(), rid.data(), rid.size());
+    common_tickets[rights_id] = raw;
+    SetKey(S128KeyType::Titlekey, key, rights_id[1], rights_id[0]);
+    return true;
+}
+
 const boost::container::flat_map<std::string, KeyIndex<S128KeyType>> KeyManager::s128_file_id = {
     {"eticket_rsa_kek", {S128KeyType::ETicketRSAKek, 0, 0}},
     {"eticket_rsa_kek_source",

src/core/crypto/key_manager.h

@@ -9,8 +9,10 @@
 #include <optional>
 #include <string>
 
+#include <variant>
 #include <boost/container/flat_map.hpp>
 #include <fmt/format.h>
+#include "common/common_funcs.h"
 #include "common/common_types.h"
 #include "core/crypto/partition_data_manager.h"
 #include "core/file_sys/vfs_types.h"
@@ -30,7 +32,79 @@ constexpr u64 TICKET_FILE_TITLEKEY_OFFSET = 0x180;
 using Key128 = std::array<u8, 0x10>;
 using Key256 = std::array<u8, 0x20>;
 using SHA256Hash = std::array<u8, 0x20>;
-using TicketRaw = std::array<u8, 0x400>;
+
+enum class SignatureType {
+    RSA_4096_SHA1 = 0x10000,
+    RSA_2048_SHA1 = 0x10001,
+    ECDSA_SHA1 = 0x10002,
+    RSA_4096_SHA256 = 0x10003,
+    RSA_2048_SHA256 = 0x10004,
+    ECDSA_SHA256 = 0x10005,
+};
+
+u64 GetSignatureTypeDataSize(SignatureType type);
+u64 GetSignatureTypePaddingSize(SignatureType type);
+
+enum class TitleKeyType : u8 {
+    Common = 0,
+    Personalized = 1,
+};
+
+struct TicketData {
+    std::array<u8, 0x40> issuer;
+    union {
+        std::array<u8, 0x100> title_key_block;
+
+        struct {
+            Key128 title_key_common;
+            std::array<u8, 0xF0> title_key_common_pad;
+        };
+    };
+
+    INSERT_PADDING_BYTES(0x1);
+    TitleKeyType type;
+    INSERT_PADDING_BYTES(0x3);
+    u8 revision;
+    INSERT_PADDING_BYTES(0xA);
+    u64 ticket_id;
+    u64 device_id;
+    std::array<u8, 0x10> rights_id;
+    u32 account_id;
+    INSERT_PADDING_BYTES(0x14C);
+};
+static_assert(sizeof(TicketData) == 0x2C0, "TicketData has incorrect size.");
+
+struct RSA4096Ticket {
+    SignatureType sig_type;
+    std::array<u8, 0x200> sig_data;
+    INSERT_PADDING_BYTES(0x3C);
+    TicketData data;
+};
+
+struct RSA2048Ticket {
+    SignatureType sig_type;
+    std::array<u8, 0x100> sig_data;
+    INSERT_PADDING_BYTES(0x3C);
+    TicketData data;
+};
+
+struct ECDSATicket {
+    SignatureType sig_type;
+    std::array<u8, 0x3C> sig_data;
+    INSERT_PADDING_BYTES(0x40);
+    TicketData data;
+};
+
+struct Ticket {
+    std::variant<RSA4096Ticket, RSA2048Ticket, ECDSATicket> data;
+
+    SignatureType GetSignatureType() const;
+    TicketData& GetData();
+    const TicketData& GetData() const;
+    u64 GetSize() const;
+
+    static Ticket SynthesizeCommon(Key128 title_key, const std::array<u8, 0x10>& rights_id);
+};
 
 static_assert(sizeof(Key128) == 16, "Key128 must be 128 bytes big.");
 static_assert(sizeof(Key256) == 32, "Key256 must be 256 bytes big.");
@@ -43,6 +117,19 @@ struct RSAKeyPair {
     std::array<u8, 4> exponent;
 };
 
+template <size_t bit_size, size_t byte_size>
+bool operator==(const RSAKeyPair<bit_size, byte_size>& lhs,
+                const RSAKeyPair<bit_size, byte_size>& rhs) {
+    return std::tie(lhs.encryption_key, lhs.decryption_key, lhs.modulus, lhs.exponent) ==
+           std::tie(rhs.encryption_key, rhs.decryption_key, rhs.modulus, rhs.exponent);
+}
+
+template <size_t bit_size, size_t byte_size>
+bool operator!=(const RSAKeyPair<bit_size, byte_size>& lhs,
+                const RSAKeyPair<bit_size, byte_size>& rhs) {
+    return !(lhs == rhs);
+}
+
 enum class KeyCategory : u8 {
     Standard,
     Title,
@@ -151,22 +238,35 @@ public:
 
     static bool KeyFileExists(bool title);
 
-    // Call before using the sd seed to attempt to derive it if it dosen't exist. Needs system save
-    // 8*43 and the private file to exist.
+    // Call before using the sd seed to attempt to derive it if it dosen't exist. Needs system
+    // save 8*43 and the private file to exist.
     void DeriveSDSeedLazy();
 
     bool BaseDeriveNecessary() const;
     void DeriveBase();
     void DeriveETicket(PartitionDataManager& data);
+    void PopulateTickets();
+    void SynthesizeTickets();
 
     void PopulateFromPartitionData(PartitionDataManager& data);
 
+    const std::map<u128, Ticket>& GetCommonTickets() const;
+    const std::map<u128, Ticket>& GetPersonalizedTickets() const;
+
+    bool AddTicketCommon(Ticket raw);
+    bool AddTicketPersonalized(Ticket raw);
+
 private:
     std::map<KeyIndex<S128KeyType>, Key128> s128_keys;
     std::map<KeyIndex<S256KeyType>, Key256> s256_keys;
 
+    // Map from rights ID to ticket
+    std::map<u128, Ticket> common_tickets;
+    std::map<u128, Ticket> personal_tickets;
+
     std::array<std::array<u8, 0xB0>, 0x20> encrypted_keyblobs{};
     std::array<std::array<u8, 0x90>, 0x20> keyblobs{};
+    std::array<u8, 576> eticket_extended_kek{};
 
     bool dev_mode;
     void LoadFromFile(const std::string& filename, bool is_title_keys);
@@ -178,6 +278,8 @@ private:
 
     void DeriveGeneralPurposeKeys(std::size_t crypto_revision);
 
+    RSAKeyPair<2048> GetETicketRSAKey() const;
+
     void SetKeyWrapped(S128KeyType id, Key128 key, u64 field1 = 0, u64 field2 = 0);
     void SetKeyWrapped(S256KeyType id, Key256 key, u64 field1 = 0, u64 field2 = 0);
 
@@ -195,11 +297,11 @@ std::array<u8, 0x90> DecryptKeyblob(const std::array<u8, 0xB0>& encrypted_keyblo
 std::optional<Key128> DeriveSDSeed();
 Loader::ResultStatus DeriveSDKeys(std::array<Key256, 2>& sd_keys, KeyManager& keys);
 
-std::vector<TicketRaw> GetTicketblob(const FileUtil::IOFile& ticket_save);
+std::vector<Ticket> GetTicketblob(const FileUtil::IOFile& ticket_save);
 
-// Returns a pair of {rights_id, titlekey}. Fails if the ticket has no certificate authority (offset
-// 0x140-0x144 is zero)
-std::optional<std::pair<Key128, Key128>> ParseTicket(const TicketRaw& ticket,
+// Returns a pair of {rights_id, titlekey}. Fails if the ticket has no certificate authority
+// (offset 0x140-0x144 is zero)
+std::optional<std::pair<Key128, Key128>> ParseTicket(const Ticket& ticket,
                                                      const RSAKeyPair<2048>& eticket_extended_key);
 
 } // namespace Core::Crypto

src/core/file_sys/system_archive/mii_model.cpp

@@ -0,0 +1,46 @@
+// Copyright 2019 yuzu emulator team
+// Licensed under GPLv2 or any later version
+// Refer to the license.txt file included.
+
+#include "core/file_sys/system_archive/mii_model.h"
+#include "core/file_sys/vfs_vector.h"
+
+namespace FileSys::SystemArchive {
+
+namespace MiiModelData {
+
+constexpr std::array<u8, 0x10> NFTR_STANDARD{'N',  'F',  'T',  'R',  0x01, 0x00, 0x00, 0x00,
+                                             0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00};
+constexpr std::array<u8, 0x10> NFSR_STANDARD{'N',  'F',  'S',  'R',  0x01, 0x00, 0x00, 0x00,
+                                             0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00};
+
+constexpr auto TEXTURE_LOW_LINEAR = NFTR_STANDARD;
+constexpr auto TEXTURE_LOW_SRGB = NFTR_STANDARD;
+constexpr auto TEXTURE_MID_LINEAR = NFTR_STANDARD;
+constexpr auto TEXTURE_MID_SRGB = NFTR_STANDARD;
+constexpr auto SHAPE_HIGH = NFSR_STANDARD;
+constexpr auto SHAPE_MID = NFSR_STANDARD;
+
+} // namespace MiiModelData
+
+VirtualDir MiiModel() {
+    auto out = std::make_shared<VectorVfsDirectory>(std::vector<VirtualFile>{},
+                                                    std::vector<VirtualDir>{}, "data");
+
+    out->AddFile(std::make_shared<ArrayVfsFile<MiiModelData::TEXTURE_LOW_LINEAR.size()>>(
+        MiiModelData::TEXTURE_LOW_LINEAR, "NXTextureLowLinear.dat"));
+    out->AddFile(std::make_shared<ArrayVfsFile<MiiModelData::TEXTURE_LOW_SRGB.size()>>(
+        MiiModelData::TEXTURE_LOW_SRGB, "NXTextureLowSRGB.dat"));
+    out->AddFile(std::make_shared<ArrayVfsFile<MiiModelData::TEXTURE_MID_LINEAR.size()>>(
+        MiiModelData::TEXTURE_MID_LINEAR, "NXTextureMidLinear.dat"));
+    out->AddFile(std::make_shared<ArrayVfsFile<MiiModelData::TEXTURE_MID_SRGB.size()>>(
+        MiiModelData::TEXTURE_MID_SRGB, "NXTextureMidSRGB.dat"));
+    out->AddFile(std::make_shared<ArrayVfsFile<MiiModelData::SHAPE_HIGH.size()>>(
+        MiiModelData::SHAPE_HIGH, "ShapeHigh.dat"));
+    out->AddFile(std::make_shared<ArrayVfsFile<MiiModelData::SHAPE_MID.size()>>(
+        MiiModelData::SHAPE_MID, "ShapeMid.dat"));
+
+    return std::move(out);
+}
+
+} // namespace FileSys::SystemArchive

src/core/file_sys/system_archive/mii_model.h

@@ -0,0 +1,13 @@
+// Copyright 2019 yuzu emulator team
+// Licensed under GPLv2 or any later version
+// Refer to the license.txt file included.
+
+#pragma once
+
+#include "core/file_sys/vfs_types.h"
+
+namespace FileSys::SystemArchive {
+
+VirtualDir MiiModel();
+
+} // namespace FileSys::SystemArchive

src/core/file_sys/system_archive/system_archive.cpp

@@ -4,6 +4,7 @@
 
 #include "common/logging/log.h"
 #include "core/file_sys/romfs.h"
+#include "core/file_sys/system_archive/mii_model.h"
 #include "core/file_sys/system_archive/ng_word.h"
 #include "core/file_sys/system_archive/system_archive.h"
 #include "core/file_sys/system_archive/system_version.h"
@@ -24,7 +25,7 @@ struct SystemArchiveDescriptor {
 constexpr std::array<SystemArchiveDescriptor, SYSTEM_ARCHIVE_COUNT> SYSTEM_ARCHIVES{{
     {0x0100000000000800, "CertStore", nullptr},
     {0x0100000000000801, "ErrorMessage", nullptr},
-    {0x0100000000000802, "MiiModel", nullptr},
+    {0x0100000000000802, "MiiModel", &MiiModel},
     {0x0100000000000803, "BrowserDll", nullptr},
     {0x0100000000000804, "Help", nullptr},
     {0x0100000000000805, "SharedFont", nullptr},

src/core/hle/service/am/am.cpp

@@ -1057,6 +1057,7 @@ IApplicationFunctions::IApplicationFunctions() : ServiceFramework("IApplicationF
         {120, nullptr, "ExecuteProgram"},
         {121, nullptr, "ClearUserChannel"},
         {122, nullptr, "UnpopToUserChannel"},
+        {130, &IApplicationFunctions::GetGpuErrorDetectedSystemEvent, "GetGpuErrorDetectedSystemEvent"},
         {500, nullptr, "StartContinuousRecordingFlushForDebug"},
         {1000, nullptr, "CreateMovieMaker"},
         {1001, nullptr, "PrepareForJit"},
@@ -1064,6 +1065,10 @@ IApplicationFunctions::IApplicationFunctions() : ServiceFramework("IApplicationF
     // clang-format on
 
     RegisterHandlers(functions);
+
+    auto& kernel = Core::System::GetInstance().Kernel();
+    gpu_error_detected_event = Kernel::WritableEvent::CreateEventPair(
+        kernel, Kernel::ResetType::Manual, "IApplicationFunctions:GpuErrorDetectedSystemEvent");
 }
 
 IApplicationFunctions::~IApplicationFunctions() = default;
@@ -1285,6 +1290,14 @@ void IApplicationFunctions::GetSaveDataSize(Kernel::HLERequestContext& ctx) {
     rb.Push(size.journal);
 }
 
+void IApplicationFunctions::GetGpuErrorDetectedSystemEvent(Kernel::HLERequestContext& ctx) {
+    LOG_WARNING(Service_AM, "(STUBBED) called");
+
+    IPC::ResponseBuilder rb{ctx, 2, 1};
+    rb.Push(RESULT_SUCCESS);
+    rb.PushCopyObjects(gpu_error_detected_event.readable);
+}
+
 void InstallInterfaces(SM::ServiceManager& service_manager,
                        std::shared_ptr<NVFlinger::NVFlinger> nvflinger, Core::System& system) {
     auto message_queue = std::make_shared<AppletMessageQueue>();

src/core/hle/service/am/am.h

@@ -242,6 +242,9 @@ private:
     void BeginBlockingHomeButton(Kernel::HLERequestContext& ctx);
     void EndBlockingHomeButton(Kernel::HLERequestContext& ctx);
     void EnableApplicationCrashReport(Kernel::HLERequestContext& ctx);
+    void GetGpuErrorDetectedSystemEvent(Kernel::HLERequestContext& ctx);
+
+    Kernel::EventPair gpu_error_detected_event;
 };
 
 class IHomeMenuFunctions final : public ServiceFramework<IHomeMenuFunctions> {

src/core/hle/service/audio/audren_u.cpp

@@ -165,15 +165,15 @@ public:
         static const FunctionInfo functions[] = {
             {0, &IAudioDevice::ListAudioDeviceName, "ListAudioDeviceName"},
             {1, &IAudioDevice::SetAudioDeviceOutputVolume, "SetAudioDeviceOutputVolume"},
-            {2, nullptr, "GetAudioDeviceOutputVolume"},
+            {2, &IAudioDevice::GetAudioDeviceOutputVolume, "GetAudioDeviceOutputVolume"},
             {3, &IAudioDevice::GetActiveAudioDeviceName, "GetActiveAudioDeviceName"},
             {4, &IAudioDevice::QueryAudioDeviceSystemEvent, "QueryAudioDeviceSystemEvent"},
             {5, &IAudioDevice::GetActiveChannelCount, "GetActiveChannelCount"},
             {6, &IAudioDevice::ListAudioDeviceName, "ListAudioDeviceNameAuto"},
             {7, &IAudioDevice::SetAudioDeviceOutputVolume, "SetAudioDeviceOutputVolumeAuto"},
-            {8, nullptr, "GetAudioDeviceOutputVolumeAuto"},
+            {8, &IAudioDevice::GetAudioDeviceOutputVolume, "GetAudioDeviceOutputVolumeAuto"},
             {10, &IAudioDevice::GetActiveAudioDeviceName, "GetActiveAudioDeviceNameAuto"},
-            {11, nullptr, "QueryAudioDeviceInputEvent"},
+            {11, &IAudioDevice::QueryAudioDeviceInputEvent, "QueryAudioDeviceInputEvent"},
             {12, &IAudioDevice::QueryAudioDeviceOutputEvent, "QueryAudioDeviceOutputEvent"},
             {13, nullptr, "GetAudioSystemMasterVolumeSetting"},
         };
@@ -183,6 +183,10 @@ public:
         buffer_event = Kernel::WritableEvent::CreateEventPair(kernel, Kernel::ResetType::Automatic,
                                                               "IAudioOutBufferReleasedEvent");
 
+        // Should be similar to audio_output_device_switch_event
+        audio_input_device_switch_event = Kernel::WritableEvent::CreateEventPair(
+            kernel, Kernel::ResetType::Automatic, "IAudioDevice:AudioInputDeviceSwitchedEvent");
+
         // Should only be signalled when an audio output device has been changed, example: speaker
         // to headset
         audio_output_device_switch_event = Kernel::WritableEvent::CreateEventPair(
@@ -246,6 +250,19 @@ private:
         rb.Push(RESULT_SUCCESS);
     }
 
+    void GetAudioDeviceOutputVolume(Kernel::HLERequestContext& ctx) {
+        IPC::RequestParser rp{ctx};
+
+        const auto device_name_buffer = ctx.ReadBuffer();
+        const std::string name = Common::StringFromBuffer(device_name_buffer);
+
+        LOG_WARNING(Service_Audio, "(STUBBED) called. name={}", name);
+
+        IPC::ResponseBuilder rb{ctx, 3};
+        rb.Push(RESULT_SUCCESS);
+        rb.Push(1.0f);
+    }
+
     void GetActiveAudioDeviceName(Kernel::HLERequestContext& ctx) {
         LOG_WARNING(Service_Audio, "(STUBBED) called");
 
@@ -279,6 +296,15 @@ private:
         rb.Push<u32>(1);
     }
 
+    // Should be similar to QueryAudioDeviceOutputEvent
+    void QueryAudioDeviceInputEvent(Kernel::HLERequestContext& ctx) {
+        LOG_WARNING(Service_Audio, "(STUBBED) called");
+
+        IPC::ResponseBuilder rb{ctx, 2, 1};
+        rb.Push(RESULT_SUCCESS);
+        rb.PushCopyObjects(audio_input_device_switch_event.readable);
+    }
+
     void QueryAudioDeviceOutputEvent(Kernel::HLERequestContext& ctx) {
         LOG_DEBUG(Service_Audio, "called");
 
@@ -289,6 +315,7 @@ private:
 
     u32_le revision = 0;
     Kernel::EventPair buffer_event;
+    Kernel::EventPair audio_input_device_switch_event;
     Kernel::EventPair audio_output_device_switch_event;
 
 }; // namespace Audio

src/core/hle/service/bcat/bcat.cpp

@@ -8,9 +8,16 @@ namespace Service::BCAT {
 
 BCAT::BCAT(std::shared_ptr<Module> module, const char* name)
     : Module::Interface(std::move(module), name) {
+    // clang-format off
     static const FunctionInfo functions[] = {
         {0, &BCAT::CreateBcatService, "CreateBcatService"},
+        {1, &BCAT::CreateDeliveryCacheStorageService, "CreateDeliveryCacheStorageService"},
+        {2, nullptr, "CreateDeliveryCacheStorageServiceWithApplicationId"},
+        {3, nullptr, "CreateDeliveryCacheProgressService"},
+        {4, nullptr, "CreateDeliveryCacheProgressServiceWithApplicationId"},
     };
+    // clang-format on
+
     RegisterHandlers(functions);
 }
 

src/core/hle/service/bcat/module.cpp

@@ -9,11 +9,26 @@
 
 namespace Service::BCAT {
 
+class IDeliveryCacheProgressService final : public ServiceFramework<IDeliveryCacheProgressService> {
+public:
+    IDeliveryCacheProgressService() : ServiceFramework("IDeliveryCacheProgressService") {
+        // clang-format off
+        static const FunctionInfo functions[] = {
+            {0, nullptr, "GetEvent"},
+            {1, nullptr, "GetImpl"},
+        };
+        // clang-format on
+
+        RegisterHandlers(functions);
+    }
+};
+
 class IBcatService final : public ServiceFramework<IBcatService> {
 public:
     IBcatService() : ServiceFramework("IBcatService") {
+        // clang-format off
         static const FunctionInfo functions[] = {
-            {10100, nullptr, "RequestSyncDeliveryCache"},
+            {10100, &IBcatService::RequestSyncDeliveryCache, "RequestSyncDeliveryCache"},
             {10101, nullptr, "RequestSyncDeliveryCacheWithDirectoryName"},
             {10200, nullptr, "CancelSyncDeliveryCacheRequest"},
             {20100, nullptr, "RequestSyncDeliveryCacheWithApplicationId"},
@@ -28,8 +43,91 @@ public:
             {90201, nullptr, "ClearDeliveryCacheStorage"},
             {90300, nullptr, "GetPushNotificationLog"},
         };
+        // clang-format on
+
         RegisterHandlers(functions);
     }
+
+private:
+    void RequestSyncDeliveryCache(Kernel::HLERequestContext& ctx) {
+        LOG_DEBUG(Service_BCAT, "called");
+
+        IPC::ResponseBuilder rb{ctx, 2, 0, 1};
+        rb.Push(RESULT_SUCCESS);
+        rb.PushIpcInterface<IDeliveryCacheProgressService>();
+    }
+};
+
+class IDeliveryCacheFileService final : public ServiceFramework<IDeliveryCacheFileService> {
+public:
+    IDeliveryCacheFileService() : ServiceFramework("IDeliveryCacheFileService") {
+        // clang-format off
+        static const FunctionInfo functions[] = {
+            {0, nullptr, "Open"},
+            {1, nullptr, "Read"},
+            {2, nullptr, "GetSize"},
+            {4, nullptr, "GetDigest"},
+        };
+        // clang-format on
+
+        RegisterHandlers(functions);
+    }
+};
+
+class IDeliveryCacheDirectoryService final
+    : public ServiceFramework<IDeliveryCacheDirectoryService> {
+public:
+    IDeliveryCacheDirectoryService() : ServiceFramework("IDeliveryCacheDirectoryService") {
+        // clang-format off
+        static const FunctionInfo functions[] = {
+            {0, nullptr, "Open"},
+            {1, nullptr, "Read"},
+            {2, nullptr, "GetCount"},
+        };
+        // clang-format on
+
+        RegisterHandlers(functions);
+    }
+};
+
+class IDeliveryCacheStorageService final : public ServiceFramework<IDeliveryCacheStorageService> {
+public:
+    IDeliveryCacheStorageService() : ServiceFramework("IDeliveryCacheStorageService") {
+        // clang-format off
+        static const FunctionInfo functions[] = {
+            {0, &IDeliveryCacheStorageService::CreateFileService, "CreateFileService"},
+            {1, &IDeliveryCacheStorageService::CreateDirectoryService, "CreateDirectoryService"},
+            {10, &IDeliveryCacheStorageService::EnumerateDeliveryCacheDirectory, "EnumerateDeliveryCacheDirectory"},
+        };
+        // clang-format on
+
+        RegisterHandlers(functions);
+    }
+
+private:
+    void CreateFileService(Kernel::HLERequestContext& ctx) {
+        LOG_DEBUG(Service_BCAT, "called");
+
+        IPC::ResponseBuilder rb{ctx, 2, 0, 1};
+        rb.Push(RESULT_SUCCESS);
+        rb.PushIpcInterface<IDeliveryCacheFileService>();
+    }
+
+    void CreateDirectoryService(Kernel::HLERequestContext& ctx) {
+        LOG_DEBUG(Service_BCAT, "called");
+
+        IPC::ResponseBuilder rb{ctx, 2, 0, 1};
+        rb.Push(RESULT_SUCCESS);
+        rb.PushIpcInterface<IDeliveryCacheDirectoryService>();
+    }
+
+    void EnumerateDeliveryCacheDirectory(Kernel::HLERequestContext& ctx) {
+        LOG_WARNING(Service_BCAT, "(STUBBED) called");
+
+        IPC::ResponseBuilder rb{ctx, 3};
+        rb.Push(RESULT_SUCCESS);
+        rb.Push<u32>(0);
+    }
 };
 
 void Module::Interface::CreateBcatService(Kernel::HLERequestContext& ctx) {
@@ -40,6 +138,14 @@ void Module::Interface::CreateBcatService(Kernel::HLERequestContext& ctx) {
     rb.PushIpcInterface<IBcatService>();
 }
 
+void Module::Interface::CreateDeliveryCacheStorageService(Kernel::HLERequestContext& ctx) {
+    LOG_DEBUG(Service_BCAT, "called");
+
+    IPC::ResponseBuilder rb{ctx, 2, 0, 1};
+    rb.Push(RESULT_SUCCESS);
+    rb.PushIpcInterface<IDeliveryCacheStorageService>();
+}
+
 Module::Interface::Interface(std::shared_ptr<Module> module, const char* name)
     : ServiceFramework(name), module(std::move(module)) {}
 

src/core/hle/service/bcat/module.h

@@ -16,6 +16,7 @@ public:
         ~Interface() override;
 
         void CreateBcatService(Kernel::HLERequestContext& ctx);
+        void CreateDeliveryCacheStorageService(Kernel::HLERequestContext& ctx);
 
     protected:
         std::shared_ptr<Module> module;

src/core/hle/service/es/es.cpp

@@ -2,32 +2,37 @@
 // Licensed under GPLv2 or any later version
 // Refer to the license.txt file included.
 
+#include "core/crypto/key_manager.h"
+#include "core/hle/ipc_helpers.h"
 #include "core/hle/service/service.h"
 
 namespace Service::ES {
 
+constexpr ResultCode ERROR_INVALID_ARGUMENT{ErrorModule::ETicket, 2};
+constexpr ResultCode ERROR_INVALID_RIGHTS_ID{ErrorModule::ETicket, 3};
+
 class ETicket final : public ServiceFramework<ETicket> {
 public:
     explicit ETicket() : ServiceFramework{"es"} {
         // clang-format off
         static const FunctionInfo functions[] = {
-            {1, nullptr, "ImportTicket"},
+            {1, &ETicket::ImportTicket, "ImportTicket"},
             {2, nullptr, "ImportTicketCertificateSet"},
             {3, nullptr, "DeleteTicket"},
             {4, nullptr, "DeletePersonalizedTicket"},
             {5, nullptr, "DeleteAllCommonTicket"},
             {6, nullptr, "DeleteAllPersonalizedTicket"},
             {7, nullptr, "DeleteAllPersonalizedTicketEx"},
-            {8, nullptr, "GetTitleKey"},
-            {9, nullptr, "CountCommonTicket"},
-            {10, nullptr, "CountPersonalizedTicket"},
-            {11, nullptr, "ListCommonTicket"},
-            {12, nullptr, "ListPersonalizedTicket"},
+            {8, &ETicket::GetTitleKey, "GetTitleKey"},
+            {9, &ETicket::CountCommonTicket, "CountCommonTicket"},
+            {10, &ETicket::CountPersonalizedTicket, "CountPersonalizedTicket"},
+            {11, &ETicket::ListCommonTicket, "ListCommonTicket"},
+            {12, &ETicket::ListPersonalizedTicket, "ListPersonalizedTicket"},
             {13, nullptr, "ListMissingPersonalizedTicket"},
-            {14, nullptr, "GetCommonTicketSize"},
-            {15, nullptr, "GetPersonalizedTicketSize"},
-            {16, nullptr, "GetCommonTicketData"},
-            {17, nullptr, "GetPersonalizedTicketData"},
+            {14, &ETicket::GetCommonTicketSize, "GetCommonTicketSize"},
+            {15, &ETicket::GetPersonalizedTicketSize, "GetPersonalizedTicketSize"},
+            {16, &ETicket::GetCommonTicketData, "GetCommonTicketData"},
+            {17, &ETicket::GetPersonalizedTicketData, "GetPersonalizedTicketData"},
             {18, nullptr, "OwnTicket"},
             {19, nullptr, "GetTicketInfo"},
             {20, nullptr, "ListLightTicketInfo"},
@@ -51,7 +56,212 @@ public:
         };
         // clang-format on
         RegisterHandlers(functions);
+
+        keys.PopulateTickets();
+        keys.SynthesizeTickets();
     }
+
+private:
+    bool CheckRightsId(Kernel::HLERequestContext& ctx, const u128& rights_id) {
+        if (rights_id == u128{}) {
+            LOG_ERROR(Service_ETicket, "The rights ID was invalid!");
+            IPC::ResponseBuilder rb{ctx, 2};
+            rb.Push(ERROR_INVALID_RIGHTS_ID);
+            return false;
+        }
+
+        return true;
+    }
+
+    void ImportTicket(Kernel::HLERequestContext& ctx) {
+        IPC::RequestParser rp{ctx};
+        const auto ticket = ctx.ReadBuffer();
+        const auto cert = ctx.ReadBuffer(1);
+
+        if (ticket.size() < sizeof(Core::Crypto::Ticket)) {
+            LOG_ERROR(Service_ETicket, "The input buffer is not large enough!");
+            IPC::ResponseBuilder rb{ctx, 2};
+            rb.Push(ERROR_INVALID_ARGUMENT);
+            return;
+        }
+
+        Core::Crypto::Ticket raw{};
+        std::memcpy(&raw, ticket.data(), sizeof(Core::Crypto::Ticket));
+
+        if (!keys.AddTicketPersonalized(raw)) {
+            LOG_ERROR(Service_ETicket, "The ticket could not be imported!");
+            IPC::ResponseBuilder rb{ctx, 2};
+            rb.Push(ERROR_INVALID_ARGUMENT);
+            return;
+        }
+
+        IPC::ResponseBuilder rb{ctx, 2};
+        rb.Push(RESULT_SUCCESS);
+    }
+
+    void GetTitleKey(Kernel::HLERequestContext& ctx) {
+        IPC::RequestParser rp{ctx};
+        const auto rights_id = rp.PopRaw<u128>();
+
+        LOG_DEBUG(Service_ETicket, "called, rights_id={:016X}{:016X}", rights_id[1], rights_id[0]);
+
+        if (!CheckRightsId(ctx, rights_id))
+            return;
+
+        const auto key =
+            keys.GetKey(Core::Crypto::S128KeyType::Titlekey, rights_id[1], rights_id[0]);
+
+        if (key == Core::Crypto::Key128{}) {
+            LOG_ERROR(Service_ETicket,
+                      "The titlekey doesn't exist in the KeyManager or the rights ID was invalid!");
+            IPC::ResponseBuilder rb{ctx, 2};
+            rb.Push(ERROR_INVALID_RIGHTS_ID);
+            return;
+        }
+
+        ctx.WriteBuffer(key.data(), key.size());
+
+        IPC::ResponseBuilder rb{ctx, 2};
+        rb.Push(RESULT_SUCCESS);
+    }
+
+    void CountCommonTicket(Kernel::HLERequestContext& ctx) {
+        LOG_DEBUG(Service_ETicket, "called");
+
+        const auto count = keys.GetCommonTickets().size();
+
+        IPC::ResponseBuilder rb{ctx, 3};
+        rb.Push(RESULT_SUCCESS);
+        rb.Push<u32>(count);
+    }
+
+    void CountPersonalizedTicket(Kernel::HLERequestContext& ctx) {
+        LOG_DEBUG(Service_ETicket, "called");
+
+        const auto count = keys.GetPersonalizedTickets().size();
+
+        IPC::ResponseBuilder rb{ctx, 3};
+        rb.Push(RESULT_SUCCESS);
+        rb.Push<u32>(count);
+    }
+
+    void ListCommonTicket(Kernel::HLERequestContext& ctx) {
+        u32 out_entries;
+        if (keys.GetCommonTickets().empty())
+            out_entries = 0;
+        else
+            out_entries = ctx.GetWriteBufferSize() / sizeof(u128);
+
+        LOG_DEBUG(Service_ETicket, "called, entries={:016X}", out_entries);
+
+        keys.PopulateTickets();
+        const auto tickets = keys.GetCommonTickets();
+        std::vector<u128> ids;
+        std::transform(tickets.begin(), tickets.end(), std::back_inserter(ids),
+                       [](const auto& pair) { return pair.first; });
+
+        out_entries = std::min<u32>(ids.size(), out_entries);
+        ctx.WriteBuffer(ids.data(), out_entries * sizeof(u128));
+
+        IPC::ResponseBuilder rb{ctx, 3};
+        rb.Push(RESULT_SUCCESS);
+        rb.Push<u32>(out_entries);
+    }
+
+    void ListPersonalizedTicket(Kernel::HLERequestContext& ctx) {
+        u32 out_entries;
+        if (keys.GetPersonalizedTickets().empty())
+            out_entries = 0;
+        else
+            out_entries = ctx.GetWriteBufferSize() / sizeof(u128);
+
+        LOG_DEBUG(Service_ETicket, "called, entries={:016X}", out_entries);
+
+        keys.PopulateTickets();
+        const auto tickets = keys.GetPersonalizedTickets();
+        std::vector<u128> ids;
+        std::transform(tickets.begin(), tickets.end(), std::back_inserter(ids),
+                       [](const auto& pair) { return pair.first; });
+
+        out_entries = std::min<u32>(ids.size(), out_entries);
+        ctx.WriteBuffer(ids.data(), out_entries * sizeof(u128));
+
+        IPC::ResponseBuilder rb{ctx, 3};
+        rb.Push(RESULT_SUCCESS);
+        rb.Push<u32>(out_entries);
+    }
+
+    void GetCommonTicketSize(Kernel::HLERequestContext& ctx) {
+        IPC::RequestParser rp{ctx};
+        const auto rights_id = rp.PopRaw<u128>();
+
+        LOG_DEBUG(Service_ETicket, "called, rights_id={:016X}{:016X}", rights_id[1], rights_id[0]);
+
+        if (!CheckRightsId(ctx, rights_id))
+            return;
+
+        const auto ticket = keys.GetCommonTickets().at(rights_id);
+
+        IPC::ResponseBuilder rb{ctx, 4};
+        rb.Push(RESULT_SUCCESS);
+        rb.Push<u64>(ticket.GetSize());
+    }
+
+    void GetPersonalizedTicketSize(Kernel::HLERequestContext& ctx) {
+        IPC::RequestParser rp{ctx};
+        const auto rights_id = rp.PopRaw<u128>();
+
+        LOG_DEBUG(Service_ETicket, "called, rights_id={:016X}{:016X}", rights_id[1], rights_id[0]);
+
+        if (!CheckRightsId(ctx, rights_id))
+            return;
+
+        const auto ticket = keys.GetPersonalizedTickets().at(rights_id);
+
+        IPC::ResponseBuilder rb{ctx, 4};
+        rb.Push(RESULT_SUCCESS);
+        rb.Push<u64>(ticket.GetSize());
+    }
+
+    void GetCommonTicketData(Kernel::HLERequestContext& ctx) {
+        IPC::RequestParser rp{ctx};
+        const auto rights_id = rp.PopRaw<u128>();
+
+        LOG_DEBUG(Service_ETicket, "called, rights_id={:016X}{:016X}", rights_id[1], rights_id[0]);
+
+        if (!CheckRightsId(ctx, rights_id))
+            return;
+
+        const auto ticket = keys.GetCommonTickets().at(rights_id);
+
+        const auto write_size = std::min<u64>(ticket.GetSize(), ctx.GetWriteBufferSize());
+        ctx.WriteBuffer(&ticket, write_size);
+
+        IPC::ResponseBuilder rb{ctx, 4};
+        rb.Push(RESULT_SUCCESS);
+        rb.Push<u64>(write_size);
+    }
+
+    void GetPersonalizedTicketData(Kernel::HLERequestContext& ctx) {
+        IPC::RequestParser rp{ctx};
+        const auto rights_id = rp.PopRaw<u128>();
+
+        LOG_DEBUG(Service_ETicket, "called, rights_id={:016X}{:016X}", rights_id[1], rights_id[0]);
+
+        if (!CheckRightsId(ctx, rights_id))
+            return;
+
+        const auto ticket = keys.GetPersonalizedTickets().at(rights_id);
+
+        const auto write_size = std::min<u64>(ticket.GetSize(), ctx.GetWriteBufferSize());
+        ctx.WriteBuffer(&ticket, write_size);
+
+        IPC::ResponseBuilder rb{ctx, 4};
+        rb.Push(RESULT_SUCCESS);
+        rb.Push<u64>(write_size);
+    }
+
+    Core::Crypto::KeyManager keys;
 };
 
 void InstallInterfaces(SM::ServiceManager& service_manager) {

src/core/hle/service/hid/controllers/npad.cpp

@@ -636,10 +636,15 @@ Controller_NPad::LedPattern Controller_NPad::GetLedPattern(u32 npad_id) {
         return LedPattern{0, 0, 0, 0};
     };
 }
+
 void Controller_NPad::SetVibrationEnabled(bool can_vibrate) {
     can_controllers_vibrate = can_vibrate;
 }
 
+bool Controller_NPad::IsVibrationEnabled() const {
+    return can_controllers_vibrate;
+}
+
 void Controller_NPad::ClearAllConnectedControllers() {
     for (auto& controller : connected_controllers) {
         if (controller.is_connected && controller.type != NPadControllerType::None) {
@@ -648,6 +653,7 @@ void Controller_NPad::ClearAllConnectedControllers() {
         }
     }
 }
+
 void Controller_NPad::DisconnectAllConnectedControllers() {
     std::for_each(connected_controllers.begin(), connected_controllers.end(),
                   [](ControllerHolder& controller) { controller.is_connected = false; });

src/core/hle/service/hid/controllers/npad.h

@@ -119,6 +119,7 @@ public:
     void DisconnectNPad(u32 npad_id);
     LedPattern GetLedPattern(u32 npad_id);
     void SetVibrationEnabled(bool can_vibrate);
+    bool IsVibrationEnabled() const;
     void ClearAllConnectedControllers();
     void DisconnectAllConnectedControllers();
     void ConnectAllDisconnectedControllers();

src/core/hle/service/hid/hid.cpp

@@ -216,8 +216,8 @@ Hid::Hid() : ServiceFramework("hid") {
         {201, &Hid::SendVibrationValue, "SendVibrationValue"},
         {202, &Hid::GetActualVibrationValue, "GetActualVibrationValue"},
         {203, &Hid::CreateActiveVibrationDeviceList, "CreateActiveVibrationDeviceList"},
-        {204, nullptr, "PermitVibration"},
-        {205, nullptr, "IsVibrationPermitted"},
+        {204, &Hid::PermitVibration, "PermitVibration"},
+        {205, &Hid::IsVibrationPermitted, "IsVibrationPermitted"},
         {206, &Hid::SendVibrationValues, "SendVibrationValues"},
         {207, nullptr, "SendVibrationGcErmCommand"},
         {208, nullptr, "GetActualVibrationGcErmCommand"},
@@ -679,6 +679,27 @@ void Hid::CreateActiveVibrationDeviceList(Kernel::HLERequestContext& ctx) {
     rb.PushIpcInterface<IActiveVibrationDeviceList>();
 }
 
+void Hid::PermitVibration(Kernel::HLERequestContext& ctx) {
+    IPC::RequestParser rp{ctx};
+    const auto can_vibrate{rp.Pop<bool>()};
+    applet_resource->GetController<Controller_NPad>(HidController::NPad)
+        .SetVibrationEnabled(can_vibrate);
+
+    LOG_DEBUG(Service_HID, "called, can_vibrate={}", can_vibrate);
+
+    IPC::ResponseBuilder rb{ctx, 2};
+    rb.Push(RESULT_SUCCESS);
+}
+
+void Hid::IsVibrationPermitted(Kernel::HLERequestContext& ctx) {
+    LOG_DEBUG(Service_HID, "called");
+
+    IPC::ResponseBuilder rb{ctx, 3};
+    rb.Push(RESULT_SUCCESS);
+    rb.Push(
+        applet_resource->GetController<Controller_NPad>(HidController::NPad).IsVibrationEnabled());
+}
+
 void Hid::ActivateConsoleSixAxisSensor(Kernel::HLERequestContext& ctx) {
     IPC::RequestParser rp{ctx};
     const auto applet_resource_user_id{rp.Pop<u64>()};

src/core/hle/service/hid/hid.h

@@ -114,6 +114,8 @@ private:
     void SetNpadHandheldActivationMode(Kernel::HLERequestContext& ctx);
     void GetVibrationDeviceInfo(Kernel::HLERequestContext& ctx);
     void CreateActiveVibrationDeviceList(Kernel::HLERequestContext& ctx);
+    void PermitVibration(Kernel::HLERequestContext& ctx);
+    void IsVibrationPermitted(Kernel::HLERequestContext& ctx);
     void ActivateConsoleSixAxisSensor(Kernel::HLERequestContext& ctx);
     void StartConsoleSixAxisSensor(Kernel::HLERequestContext& ctx);
     void StopSixAxisSensor(Kernel::HLERequestContext& ctx);

src/core/hle/service/mii/mii_manager.cpp

@@ -175,6 +175,10 @@ MiiStoreData ConvertInfoToStoreData(const MiiInfo& info) {
 } // namespace
 
 std::ostream& operator<<(std::ostream& os, Source source) {
+    if (static_cast<std::size_t>(source) >= SOURCE_NAMES.size()) {
+        return os << "[UNKNOWN SOURCE]";
+    }
+
     os << SOURCE_NAMES.at(static_cast<std::size_t>(source));
     return os;
 }

src/core/hle/service/nvdrv/devices/nvdisp_disp0.cpp

@@ -37,7 +37,7 @@ void nvdisp_disp0::flip(u32 buffer_handle, u32 offset, u32 format, u32 width, u3
         transform, crop_rect};
 
     system.GetPerfStats().EndGameFrame();
-    system.GPU().SwapBuffers(framebuffer);
+    system.GPU().SwapBuffers(&framebuffer);
 }
 
 } // namespace Service::Nvidia::Devices

src/video_core/buffer_cache/buffer_block.h

@@ -69,7 +69,6 @@ protected:
 private:
     CacheAddr cache_addr{};
     CacheAddr cache_addr_end{};
-    u64 pages{};
     std::size_t size{};
     u64 epoch{};
 };

src/video_core/buffer_cache/buffer_cache.h

@@ -18,10 +18,7 @@
 #include "video_core/buffer_cache/buffer_block.h"
 #include "video_core/buffer_cache/map_interval.h"
 #include "video_core/memory_manager.h"
-
-namespace VideoCore {
-class RasterizerInterface;
-}
+#include "video_core/rasterizer_interface.h"
 
 namespace VideoCommon {
 
@@ -348,7 +345,6 @@ private:
         const CacheAddr cache_addr_end = cache_addr + size - 1;
         u64 page_start = cache_addr >> block_page_bits;
         const u64 page_end = cache_addr_end >> block_page_bits;
-        const u64 num_pages = page_end - page_start + 1;
         while (page_start <= page_end) {
             auto it = blocks.find(page_start);
             if (it == blocks.end()) {
@@ -417,7 +413,10 @@ private:
         return false;
     }
 
+    VideoCore::RasterizerInterface& rasterizer;
+    Core::System& system;
     std::unique_ptr<StreamBuffer> stream_buffer;
+
     TBufferType stream_buffer_handle{};
 
     bool invalidated = false;
@@ -441,8 +440,7 @@ private:
     std::list<TBuffer> pending_destruction{};
     u64 epoch{};
     u64 modified_ticks{};
-    VideoCore::RasterizerInterface& rasterizer;
-    Core::System& system;
+
     std::recursive_mutex mutex;
 };
 

src/video_core/engines/fermi_2d.cpp

@@ -10,8 +10,7 @@
 
 namespace Tegra::Engines {
 
-Fermi2D::Fermi2D(VideoCore::RasterizerInterface& rasterizer, MemoryManager& memory_manager)
-    : rasterizer{rasterizer}, memory_manager{memory_manager} {}
+Fermi2D::Fermi2D(VideoCore::RasterizerInterface& rasterizer) : rasterizer{rasterizer} {}
 
 void Fermi2D::CallMethod(const GPU::MethodCall& method_call) {
     ASSERT_MSG(method_call.method < Regs::NUM_REGS,

src/video_core/engines/fermi_2d.h

@@ -33,7 +33,7 @@ namespace Tegra::Engines {
 
 class Fermi2D final {
 public:
-    explicit Fermi2D(VideoCore::RasterizerInterface& rasterizer, MemoryManager& memory_manager);
+    explicit Fermi2D(VideoCore::RasterizerInterface& rasterizer);
     ~Fermi2D() = default;
 
     /// Write the value to the register identified by method.
@@ -145,7 +145,6 @@ public:
 
 private:
     VideoCore::RasterizerInterface& rasterizer;
-    MemoryManager& memory_manager;
 
     /// Performs the copy from the source surface to the destination surface as configured in the
     /// registers.

src/video_core/engines/kepler_memory.cpp

@@ -15,7 +15,7 @@
 namespace Tegra::Engines {
 
 KeplerMemory::KeplerMemory(Core::System& system, MemoryManager& memory_manager)
-    : system{system}, memory_manager{memory_manager}, upload_state{memory_manager, regs.upload} {}
+    : system{system}, upload_state{memory_manager, regs.upload} {}
 
 KeplerMemory::~KeplerMemory() = default;
 

src/video_core/engines/kepler_memory.h

@@ -65,7 +65,6 @@ public:
 
 private:
     Core::System& system;
-    MemoryManager& memory_manager;
     Upload::State upload_state;
 };
 

src/video_core/engines/maxwell_3d.cpp

@@ -249,16 +249,10 @@ void Maxwell3D::CallMacroMethod(u32 method, std::vector<u32> parameters) {
     executing_macro = 0;
 
     // Lookup the macro offset
-    const u32 entry{(method - MacroRegistersStart) >> 1};
-    const auto& search{macro_offsets.find(entry)};
-    if (search == macro_offsets.end()) {
-        LOG_CRITICAL(HW_GPU, "macro not found for method 0x{:X}!", method);
-        UNREACHABLE();
-        return;
-    }
+    const u32 entry = ((method - MacroRegistersStart) >> 1) % macro_positions.size();
 
     // Execute the current macro.
-    macro_interpreter.Execute(search->second, std::move(parameters));
+    macro_interpreter.Execute(macro_positions[entry], std::move(parameters));
 }
 
 void Maxwell3D::CallMethod(const GPU::MethodCall& method_call) {
@@ -421,7 +415,7 @@ void Maxwell3D::ProcessMacroUpload(u32 data) {
 }
 
 void Maxwell3D::ProcessMacroBind(u32 data) {
-    macro_offsets[regs.macros.entry] = data;
+    macro_positions[regs.macros.entry++] = data;
 }
 
 void Maxwell3D::ProcessQueryGet() {
@@ -524,7 +518,7 @@ void Maxwell3D::ProcessQueryCondition() {
 void Maxwell3D::ProcessSyncPoint() {
     const u32 sync_point = regs.sync_info.sync_point.Value();
     const u32 increment = regs.sync_info.increment.Value();
-    const u32 cache_flush = regs.sync_info.unknown.Value();
+    [[maybe_unused]] const u32 cache_flush = regs.sync_info.unknown.Value();
     if (increment) {
         system.GPU().IncrementSyncPoint(sync_point);
     }
@@ -626,10 +620,10 @@ Texture::TICEntry Maxwell3D::GetTICEntry(u32 tic_index) const {
     Texture::TICEntry tic_entry;
     memory_manager.ReadBlockUnsafe(tic_address_gpu, &tic_entry, sizeof(Texture::TICEntry));
 
-    const auto r_type{tic_entry.r_type.Value()};
-    const auto g_type{tic_entry.g_type.Value()};
-    const auto b_type{tic_entry.b_type.Value()};
-    const auto a_type{tic_entry.a_type.Value()};
+    [[maybe_unused]] const auto r_type{tic_entry.r_type.Value()};
+    [[maybe_unused]] const auto g_type{tic_entry.g_type.Value()};
+    [[maybe_unused]] const auto b_type{tic_entry.b_type.Value()};
+    [[maybe_unused]] const auto a_type{tic_entry.a_type.Value()};
 
     // TODO(Subv): Different data types for separate components are not supported
     DEBUG_ASSERT(r_type == g_type && r_type == b_type && r_type == a_type);

src/video_core/engines/maxwell_3d.h

@@ -1270,7 +1270,7 @@ private:
     MemoryManager& memory_manager;
 
     /// Start offsets of each macro in macro_memory
-    std::unordered_map<u32, u32> macro_offsets;
+    std::array<u32, 0x80> macro_positions = {};
 
     /// Memory for macro code
     MacroMemory macro_memory;

src/video_core/engines/maxwell_dma.cpp

@@ -5,18 +5,17 @@
 #include "common/assert.h"
 #include "common/logging/log.h"
 #include "core/core.h"
+#include "core/settings.h"
 #include "video_core/engines/maxwell_3d.h"
 #include "video_core/engines/maxwell_dma.h"
 #include "video_core/memory_manager.h"
-#include "video_core/rasterizer_interface.h"
 #include "video_core/renderer_base.h"
 #include "video_core/textures/decoders.h"
 
 namespace Tegra::Engines {
 
-MaxwellDMA::MaxwellDMA(Core::System& system, VideoCore::RasterizerInterface& rasterizer,
-                       MemoryManager& memory_manager)
-    : system{system}, rasterizer{rasterizer}, memory_manager{memory_manager} {}
+MaxwellDMA::MaxwellDMA(Core::System& system, MemoryManager& memory_manager)
+    : system{system}, memory_manager{memory_manager} {}
 
 void MaxwellDMA::CallMethod(const GPU::MethodCall& method_call) {
     ASSERT_MSG(method_call.method < Regs::NUM_REGS,
@@ -84,13 +83,17 @@ void MaxwellDMA::HandleCopy() {
     ASSERT(regs.exec.enable_2d == 1);
 
     if (regs.exec.is_dst_linear && !regs.exec.is_src_linear) {
-        ASSERT(regs.src_params.size_z == 1);
+        ASSERT(regs.src_params.BlockDepth() == 0);
         // If the input is tiled and the output is linear, deswizzle the input and copy it over.
-        const u32 src_bytes_per_pixel = regs.src_pitch / regs.src_params.size_x;
+        const u32 bytes_per_pixel = regs.dst_pitch / regs.x_count;
         const std::size_t src_size = Texture::CalculateSize(
-            true, src_bytes_per_pixel, regs.src_params.size_x, regs.src_params.size_y,
+            true, bytes_per_pixel, regs.src_params.size_x, regs.src_params.size_y,
             regs.src_params.size_z, regs.src_params.BlockHeight(), regs.src_params.BlockDepth());
 
+        const std::size_t src_layer_size = Texture::CalculateSize(
+            true, bytes_per_pixel, regs.src_params.size_x, regs.src_params.size_y, 1,
+            regs.src_params.BlockHeight(), regs.src_params.BlockDepth());
+
         const std::size_t dst_size = regs.dst_pitch * regs.y_count;
 
         if (read_buffer.size() < src_size) {
@@ -104,23 +107,23 @@ void MaxwellDMA::HandleCopy() {
         memory_manager.ReadBlock(source, read_buffer.data(), src_size);
         memory_manager.ReadBlock(dest, write_buffer.data(), dst_size);
 
-        Texture::UnswizzleSubrect(regs.x_count, regs.y_count, regs.dst_pitch,
-                                  regs.src_params.size_x, src_bytes_per_pixel, read_buffer.data(),
-                                  write_buffer.data(), regs.src_params.BlockHeight(),
-                                  regs.src_params.pos_x, regs.src_params.pos_y);
+        Texture::UnswizzleSubrect(
+            regs.x_count, regs.y_count, regs.dst_pitch, regs.src_params.size_x, bytes_per_pixel,
+            read_buffer.data() + src_layer_size * regs.src_params.pos_z, write_buffer.data(),
+            regs.src_params.BlockHeight(), regs.src_params.pos_x, regs.src_params.pos_y);
 
         memory_manager.WriteBlock(dest, write_buffer.data(), dst_size);
     } else {
         ASSERT(regs.dst_params.BlockDepth() == 0);
 
-        const u32 src_bytes_per_pixel = regs.src_pitch / regs.x_count;
+        const u32 bytes_per_pixel = regs.src_pitch / regs.x_count;
 
         const std::size_t dst_size = Texture::CalculateSize(
-            true, src_bytes_per_pixel, regs.dst_params.size_x, regs.dst_params.size_y,
+            true, bytes_per_pixel, regs.dst_params.size_x, regs.dst_params.size_y,
             regs.dst_params.size_z, regs.dst_params.BlockHeight(), regs.dst_params.BlockDepth());
 
         const std::size_t dst_layer_size = Texture::CalculateSize(
-            true, src_bytes_per_pixel, regs.dst_params.size_x, regs.dst_params.size_y, 1,
+            true, bytes_per_pixel, regs.dst_params.size_x, regs.dst_params.size_y, 1,
             regs.dst_params.BlockHeight(), regs.dst_params.BlockDepth());
 
         const std::size_t src_size = regs.src_pitch * regs.y_count;
@@ -133,14 +136,19 @@ void MaxwellDMA::HandleCopy() {
             write_buffer.resize(dst_size);
         }
 
-        memory_manager.ReadBlock(source, read_buffer.data(), src_size);
-        memory_manager.ReadBlock(dest, write_buffer.data(), dst_size);
+        if (Settings::values.use_accurate_gpu_emulation) {
+            memory_manager.ReadBlock(source, read_buffer.data(), src_size);
+            memory_manager.ReadBlock(dest, write_buffer.data(), dst_size);
+        } else {
+            memory_manager.ReadBlockUnsafe(source, read_buffer.data(), src_size);
+            memory_manager.ReadBlockUnsafe(dest, write_buffer.data(), dst_size);
+        }
 
         // If the input is linear and the output is tiled, swizzle the input and copy it over.
-        Texture::SwizzleSubrect(regs.x_count, regs.y_count, regs.src_pitch, regs.dst_params.size_x,
-                                src_bytes_per_pixel,
-                                write_buffer.data() + dst_layer_size * regs.dst_params.pos_z,
-                                read_buffer.data(), regs.dst_params.BlockHeight());
+        Texture::SwizzleSubrect(
+            regs.x_count, regs.y_count, regs.src_pitch, regs.dst_params.size_x, bytes_per_pixel,
+            write_buffer.data() + dst_layer_size * regs.dst_params.pos_z, read_buffer.data(),
+            regs.dst_params.BlockHeight(), regs.dst_params.pos_x, regs.dst_params.pos_y);
 
         memory_manager.WriteBlock(dest, write_buffer.data(), dst_size);
     }

src/video_core/engines/maxwell_dma.h

@@ -20,10 +20,6 @@ namespace Tegra {
 class MemoryManager;
 }
 
-namespace VideoCore {
-class RasterizerInterface;
-}
-
 namespace Tegra::Engines {
 
 /**
@@ -33,8 +29,7 @@ namespace Tegra::Engines {
 
 class MaxwellDMA final {
 public:
-    explicit MaxwellDMA(Core::System& system, VideoCore::RasterizerInterface& rasterizer,
-                        MemoryManager& memory_manager);
+    explicit MaxwellDMA(Core::System& system, MemoryManager& memory_manager);
     ~MaxwellDMA() = default;
 
     /// Write the value to the register identified by method.
@@ -180,8 +175,6 @@ public:
 private:
     Core::System& system;
 
-    VideoCore::RasterizerInterface& rasterizer;
-
     MemoryManager& memory_manager;
 
     std::vector<u8> read_buffer;

src/video_core/engines/shader_bytecode.h

@@ -886,6 +886,7 @@ union Instruction {
     union {
         BitField<0, 3, u64> pred0;
         BitField<3, 3, u64> pred3;
+        BitField<6, 1, u64> neg_b;
         BitField<7, 1, u64> abs_a;
         BitField<39, 3, u64> pred39;
         BitField<42, 1, u64> neg_pred;
@@ -1019,7 +1020,6 @@ union Instruction {
     } iset;
 
     union {
-        BitField<41, 2, u64> selector; // i2i and i2f only
         BitField<45, 1, u64> negate_a;
         BitField<49, 1, u64> abs_a;
         BitField<10, 2, Register::Size> src_size;
@@ -1045,6 +1045,13 @@ union Instruction {
             }
         } f2f;
 
+        union {
+            BitField<41, 2, u64> selector;
+        } int_src;
+
+        union {
+            BitField<41, 1, u64> selector;
+        } float_src;
     } conversion;
 
     union {

src/video_core/gpu.cpp

@@ -17,27 +17,15 @@
 
 namespace Tegra {
 
-u32 FramebufferConfig::BytesPerPixel(PixelFormat format) {
-    switch (format) {
-    case PixelFormat::ABGR8:
-    case PixelFormat::BGRA8:
-        return 4;
-    default:
-        return 4;
-    }
-
-    UNREACHABLE();
-}
-
 GPU::GPU(Core::System& system, VideoCore::RendererBase& renderer, bool is_async)
     : system{system}, renderer{renderer}, is_async{is_async} {
     auto& rasterizer{renderer.Rasterizer()};
     memory_manager = std::make_unique<Tegra::MemoryManager>(system, rasterizer);
     dma_pusher = std::make_unique<Tegra::DmaPusher>(*this);
     maxwell_3d = std::make_unique<Engines::Maxwell3D>(system, rasterizer, *memory_manager);
-    fermi_2d = std::make_unique<Engines::Fermi2D>(rasterizer, *memory_manager);
+    fermi_2d = std::make_unique<Engines::Fermi2D>(rasterizer);
     kepler_compute = std::make_unique<Engines::KeplerCompute>(system, rasterizer, *memory_manager);
-    maxwell_dma = std::make_unique<Engines::MaxwellDMA>(system, rasterizer, *memory_manager);
+    maxwell_dma = std::make_unique<Engines::MaxwellDMA>(system, *memory_manager);
     kepler_memory = std::make_unique<Engines::KeplerMemory>(system, *memory_manager);
 }
 

src/video_core/gpu.h

@@ -95,14 +95,10 @@ class DebugContext;
 struct FramebufferConfig {
     enum class PixelFormat : u32 {
         ABGR8 = 1,
+        RGB565 = 4,
         BGRA8 = 5,
     };
 
-    /**
-     * Returns the number of bytes per pixel.
-     */
-    static u32 BytesPerPixel(PixelFormat format);
-
     VAddr address;
     u32 offset;
     u32 width;
@@ -253,8 +249,7 @@ public:
     virtual void PushGPUEntries(Tegra::CommandList&& entries) = 0;
 
     /// Swap buffers (render frame)
-    virtual void SwapBuffers(
-        std::optional<std::reference_wrapper<const Tegra::FramebufferConfig>> framebuffer) = 0;
+    virtual void SwapBuffers(const Tegra::FramebufferConfig* framebuffer) = 0;
 
     /// Notify rasterizer that any caches of the specified region should be flushed to Switch memory
     virtual void FlushRegion(CacheAddr addr, u64 size) = 0;
@@ -285,8 +280,8 @@ private:
 
 protected:
     std::unique_ptr<Tegra::DmaPusher> dma_pusher;
-    VideoCore::RendererBase& renderer;
     Core::System& system;
+    VideoCore::RendererBase& renderer;
 
 private:
     std::unique_ptr<Tegra::MemoryManager> memory_manager;

src/video_core/gpu_asynch.cpp

@@ -23,9 +23,8 @@ void GPUAsynch::PushGPUEntries(Tegra::CommandList&& entries) {
     gpu_thread.SubmitList(std::move(entries));
 }
 
-void GPUAsynch::SwapBuffers(
-    std::optional<std::reference_wrapper<const Tegra::FramebufferConfig>> framebuffer) {
-    gpu_thread.SwapBuffers(std::move(framebuffer));
+void GPUAsynch::SwapBuffers(const Tegra::FramebufferConfig* framebuffer) {
+    gpu_thread.SwapBuffers(framebuffer);
 }
 
 void GPUAsynch::FlushRegion(CacheAddr addr, u64 size) {

src/video_core/gpu_asynch.h

@@ -14,15 +14,14 @@ class RendererBase;
 namespace VideoCommon {
 
 /// Implementation of GPU interface that runs the GPU asynchronously
-class GPUAsynch : public Tegra::GPU {
+class GPUAsynch final : public Tegra::GPU {
 public:
     explicit GPUAsynch(Core::System& system, VideoCore::RendererBase& renderer);
     ~GPUAsynch() override;
 
     void Start() override;
     void PushGPUEntries(Tegra::CommandList&& entries) override;
-    void SwapBuffers(
-        std::optional<std::reference_wrapper<const Tegra::FramebufferConfig>> framebuffer) override;
+    void SwapBuffers(const Tegra::FramebufferConfig* framebuffer) override;
     void FlushRegion(CacheAddr addr, u64 size) override;
     void InvalidateRegion(CacheAddr addr, u64 size) override;
     void FlushAndInvalidateRegion(CacheAddr addr, u64 size) override;

src/video_core/gpu_synch.cpp

@@ -19,9 +19,8 @@ void GPUSynch::PushGPUEntries(Tegra::CommandList&& entries) {
     dma_pusher->DispatchCalls();
 }
 
-void GPUSynch::SwapBuffers(
-    std::optional<std::reference_wrapper<const Tegra::FramebufferConfig>> framebuffer) {
-    renderer.SwapBuffers(std::move(framebuffer));
+void GPUSynch::SwapBuffers(const Tegra::FramebufferConfig* framebuffer) {
+    renderer.SwapBuffers(framebuffer);
 }
 
 void GPUSynch::FlushRegion(CacheAddr addr, u64 size) {

src/video_core/gpu_synch.h

@@ -13,15 +13,14 @@ class RendererBase;
 namespace VideoCommon {
 
 /// Implementation of GPU interface that runs the GPU synchronously
-class GPUSynch : public Tegra::GPU {
+class GPUSynch final : public Tegra::GPU {
 public:
     explicit GPUSynch(Core::System& system, VideoCore::RendererBase& renderer);
     ~GPUSynch() override;
 
     void Start() override;
     void PushGPUEntries(Tegra::CommandList&& entries) override;
-    void SwapBuffers(
-        std::optional<std::reference_wrapper<const Tegra::FramebufferConfig>> framebuffer) override;
+    void SwapBuffers(const Tegra::FramebufferConfig* framebuffer) override;
     void FlushRegion(CacheAddr addr, u64 size) override;
     void InvalidateRegion(CacheAddr addr, u64 size) override;
     void FlushAndInvalidateRegion(CacheAddr addr, u64 size) override;

src/video_core/gpu_thread.cpp

@@ -39,7 +39,7 @@ static void RunThread(VideoCore::RendererBase& renderer, Tegra::DmaPusher& dma_p
                 dma_pusher.Push(std::move(submit_list->entries));
                 dma_pusher.DispatchCalls();
             } else if (const auto data = std::get_if<SwapBuffersCommand>(&next.data)) {
-                renderer.SwapBuffers(std::move(data->framebuffer));
+                renderer.SwapBuffers(data->framebuffer ? &*data->framebuffer : nullptr);
             } else if (const auto data = std::get_if<FlushRegionCommand>(&next.data)) {
                 renderer.Rasterizer().FlushRegion(data->addr, data->size);
             } else if (const auto data = std::get_if<InvalidateRegionCommand>(&next.data)) {
@@ -78,9 +78,9 @@ void ThreadManager::SubmitList(Tegra::CommandList&& entries) {
     system.CoreTiming().ScheduleEvent(synchronization_ticks, synchronization_event, fence);
 }
 
-void ThreadManager::SwapBuffers(
-    std::optional<std::reference_wrapper<const Tegra::FramebufferConfig>> framebuffer) {
-    PushCommand(SwapBuffersCommand(std::move(framebuffer)));
+void ThreadManager::SwapBuffers(const Tegra::FramebufferConfig* framebuffer) {
+    PushCommand(SwapBuffersCommand(framebuffer ? *framebuffer
+                                               : std::optional<const Tegra::FramebufferConfig>{}));
 }
 
 void ThreadManager::FlushRegion(CacheAddr addr, u64 size) {

src/video_core/gpu_thread.h

@@ -110,8 +110,7 @@ public:
     void SubmitList(Tegra::CommandList&& entries);
 
     /// Swap buffers (render frame)
-    void SwapBuffers(
-        std::optional<std::reference_wrapper<const Tegra::FramebufferConfig>> framebuffer);
+    void SwapBuffers(const Tegra::FramebufferConfig* framebuffer);
 
     /// Notify rasterizer that any caches of the specified region should be flushed to Switch memory
     void FlushRegion(CacheAddr addr, u64 size);

src/video_core/morton.cpp

@@ -25,8 +25,8 @@ static void MortonCopy(u32 stride, u32 block_height, u32 height, u32 block_depth
 
     // With the BCn formats (DXT and DXN), each 4x4 tile is swizzled instead of just individual
     // pixel values.
-    const u32 tile_size_x{GetDefaultBlockWidth(format)};
-    const u32 tile_size_y{GetDefaultBlockHeight(format)};
+    constexpr u32 tile_size_x{GetDefaultBlockWidth(format)};
+    constexpr u32 tile_size_y{GetDefaultBlockHeight(format)};
 
     if constexpr (morton_to_linear) {
         Tegra::Texture::UnswizzleTexture(buffer, addr, tile_size_x, tile_size_y, bytes_per_pixel,
@@ -186,99 +186,6 @@ static MortonCopyFn GetSwizzleFunction(MortonSwizzleMode mode, Surface::PixelFor
     return morton_to_linear_fns[static_cast<std::size_t>(format)];
 }
 
-static u32 MortonInterleave128(u32 x, u32 y) {
-    // 128x128 Z-Order coordinate from 2D coordinates
-    static constexpr u32 xlut[] = {
-        0x0000, 0x0001, 0x0002, 0x0003, 0x0008, 0x0009, 0x000a, 0x000b, 0x0040, 0x0041, 0x0042,
-        0x0043, 0x0048, 0x0049, 0x004a, 0x004b, 0x0800, 0x0801, 0x0802, 0x0803, 0x0808, 0x0809,
-        0x080a, 0x080b, 0x0840, 0x0841, 0x0842, 0x0843, 0x0848, 0x0849, 0x084a, 0x084b, 0x1000,
-        0x1001, 0x1002, 0x1003, 0x1008, 0x1009, 0x100a, 0x100b, 0x1040, 0x1041, 0x1042, 0x1043,
-        0x1048, 0x1049, 0x104a, 0x104b, 0x1800, 0x1801, 0x1802, 0x1803, 0x1808, 0x1809, 0x180a,
-        0x180b, 0x1840, 0x1841, 0x1842, 0x1843, 0x1848, 0x1849, 0x184a, 0x184b, 0x2000, 0x2001,
-        0x2002, 0x2003, 0x2008, 0x2009, 0x200a, 0x200b, 0x2040, 0x2041, 0x2042, 0x2043, 0x2048,
-        0x2049, 0x204a, 0x204b, 0x2800, 0x2801, 0x2802, 0x2803, 0x2808, 0x2809, 0x280a, 0x280b,
-        0x2840, 0x2841, 0x2842, 0x2843, 0x2848, 0x2849, 0x284a, 0x284b, 0x3000, 0x3001, 0x3002,
-        0x3003, 0x3008, 0x3009, 0x300a, 0x300b, 0x3040, 0x3041, 0x3042, 0x3043, 0x3048, 0x3049,
-        0x304a, 0x304b, 0x3800, 0x3801, 0x3802, 0x3803, 0x3808, 0x3809, 0x380a, 0x380b, 0x3840,
-        0x3841, 0x3842, 0x3843, 0x3848, 0x3849, 0x384a, 0x384b, 0x0000, 0x0001, 0x0002, 0x0003,
-        0x0008, 0x0009, 0x000a, 0x000b, 0x0040, 0x0041, 0x0042, 0x0043, 0x0048, 0x0049, 0x004a,
-        0x004b, 0x0800, 0x0801, 0x0802, 0x0803, 0x0808, 0x0809, 0x080a, 0x080b, 0x0840, 0x0841,
-        0x0842, 0x0843, 0x0848, 0x0849, 0x084a, 0x084b, 0x1000, 0x1001, 0x1002, 0x1003, 0x1008,
-        0x1009, 0x100a, 0x100b, 0x1040, 0x1041, 0x1042, 0x1043, 0x1048, 0x1049, 0x104a, 0x104b,
-        0x1800, 0x1801, 0x1802, 0x1803, 0x1808, 0x1809, 0x180a, 0x180b, 0x1840, 0x1841, 0x1842,
-        0x1843, 0x1848, 0x1849, 0x184a, 0x184b, 0x2000, 0x2001, 0x2002, 0x2003, 0x2008, 0x2009,
-        0x200a, 0x200b, 0x2040, 0x2041, 0x2042, 0x2043, 0x2048, 0x2049, 0x204a, 0x204b, 0x2800,
-        0x2801, 0x2802, 0x2803, 0x2808, 0x2809, 0x280a, 0x280b, 0x2840, 0x2841, 0x2842, 0x2843,
-        0x2848, 0x2849, 0x284a, 0x284b, 0x3000, 0x3001, 0x3002, 0x3003, 0x3008, 0x3009, 0x300a,
-        0x300b, 0x3040, 0x3041, 0x3042, 0x3043, 0x3048, 0x3049, 0x304a, 0x304b, 0x3800, 0x3801,
-        0x3802, 0x3803, 0x3808, 0x3809, 0x380a, 0x380b, 0x3840, 0x3841, 0x3842, 0x3843, 0x3848,
-        0x3849, 0x384a, 0x384b, 0x0000, 0x0001, 0x0002, 0x0003, 0x0008, 0x0009, 0x000a, 0x000b,
-        0x0040, 0x0041, 0x0042, 0x0043, 0x0048, 0x0049, 0x004a, 0x004b, 0x0800, 0x0801, 0x0802,
-        0x0803, 0x0808, 0x0809, 0x080a, 0x080b, 0x0840, 0x0841, 0x0842, 0x0843, 0x0848, 0x0849,
-        0x084a, 0x084b, 0x1000, 0x1001, 0x1002, 0x1003, 0x1008, 0x1009, 0x100a, 0x100b, 0x1040,
-        0x1041, 0x1042, 0x1043, 0x1048, 0x1049, 0x104a, 0x104b, 0x1800, 0x1801, 0x1802, 0x1803,
-        0x1808, 0x1809, 0x180a, 0x180b, 0x1840, 0x1841, 0x1842, 0x1843, 0x1848, 0x1849, 0x184a,
-        0x184b, 0x2000, 0x2001, 0x2002, 0x2003, 0x2008, 0x2009, 0x200a, 0x200b, 0x2040, 0x2041,
-        0x2042, 0x2043, 0x2048, 0x2049, 0x204a, 0x204b, 0x2800, 0x2801, 0x2802, 0x2803, 0x2808,
-        0x2809, 0x280a, 0x280b, 0x2840, 0x2841, 0x2842, 0x2843, 0x2848, 0x2849, 0x284a, 0x284b,
-        0x3000, 0x3001, 0x3002, 0x3003, 0x3008, 0x3009, 0x300a, 0x300b, 0x3040, 0x3041, 0x3042,
-        0x3043, 0x3048, 0x3049, 0x304a, 0x304b, 0x3800, 0x3801, 0x3802, 0x3803, 0x3808, 0x3809,
-        0x380a, 0x380b, 0x3840, 0x3841, 0x3842, 0x3843, 0x3848, 0x3849, 0x384a, 0x384b,
-    };
-    static constexpr u32 ylut[] = {
-        0x0000, 0x0004, 0x0010, 0x0014, 0x0020, 0x0024, 0x0030, 0x0034, 0x0080, 0x0084, 0x0090,
-        0x0094, 0x00a0, 0x00a4, 0x00b0, 0x00b4, 0x0100, 0x0104, 0x0110, 0x0114, 0x0120, 0x0124,
-        0x0130, 0x0134, 0x0180, 0x0184, 0x0190, 0x0194, 0x01a0, 0x01a4, 0x01b0, 0x01b4, 0x0200,
-        0x0204, 0x0210, 0x0214, 0x0220, 0x0224, 0x0230, 0x0234, 0x0280, 0x0284, 0x0290, 0x0294,
-        0x02a0, 0x02a4, 0x02b0, 0x02b4, 0x0300, 0x0304, 0x0310, 0x0314, 0x0320, 0x0324, 0x0330,
-        0x0334, 0x0380, 0x0384, 0x0390, 0x0394, 0x03a0, 0x03a4, 0x03b0, 0x03b4, 0x0400, 0x0404,
-        0x0410, 0x0414, 0x0420, 0x0424, 0x0430, 0x0434, 0x0480, 0x0484, 0x0490, 0x0494, 0x04a0,
-        0x04a4, 0x04b0, 0x04b4, 0x0500, 0x0504, 0x0510, 0x0514, 0x0520, 0x0524, 0x0530, 0x0534,
-        0x0580, 0x0584, 0x0590, 0x0594, 0x05a0, 0x05a4, 0x05b0, 0x05b4, 0x0600, 0x0604, 0x0610,
-        0x0614, 0x0620, 0x0624, 0x0630, 0x0634, 0x0680, 0x0684, 0x0690, 0x0694, 0x06a0, 0x06a4,
-        0x06b0, 0x06b4, 0x0700, 0x0704, 0x0710, 0x0714, 0x0720, 0x0724, 0x0730, 0x0734, 0x0780,
-        0x0784, 0x0790, 0x0794, 0x07a0, 0x07a4, 0x07b0, 0x07b4, 0x0000, 0x0004, 0x0010, 0x0014,
-        0x0020, 0x0024, 0x0030, 0x0034, 0x0080, 0x0084, 0x0090, 0x0094, 0x00a0, 0x00a4, 0x00b0,
-        0x00b4, 0x0100, 0x0104, 0x0110, 0x0114, 0x0120, 0x0124, 0x0130, 0x0134, 0x0180, 0x0184,
-        0x0190, 0x0194, 0x01a0, 0x01a4, 0x01b0, 0x01b4, 0x0200, 0x0204, 0x0210, 0x0214, 0x0220,
-        0x0224, 0x0230, 0x0234, 0x0280, 0x0284, 0x0290, 0x0294, 0x02a0, 0x02a4, 0x02b0, 0x02b4,
-        0x0300, 0x0304, 0x0310, 0x0314, 0x0320, 0x0324, 0x0330, 0x0334, 0x0380, 0x0384, 0x0390,
-        0x0394, 0x03a0, 0x03a4, 0x03b0, 0x03b4, 0x0400, 0x0404, 0x0410, 0x0414, 0x0420, 0x0424,
-        0x0430, 0x0434, 0x0480, 0x0484, 0x0490, 0x0494, 0x04a0, 0x04a4, 0x04b0, 0x04b4, 0x0500,
-        0x0504, 0x0510, 0x0514, 0x0520, 0x0524, 0x0530, 0x0534, 0x0580, 0x0584, 0x0590, 0x0594,
-        0x05a0, 0x05a4, 0x05b0, 0x05b4, 0x0600, 0x0604, 0x0610, 0x0614, 0x0620, 0x0624, 0x0630,
-        0x0634, 0x0680, 0x0684, 0x0690, 0x0694, 0x06a0, 0x06a4, 0x06b0, 0x06b4, 0x0700, 0x0704,
-        0x0710, 0x0714, 0x0720, 0x0724, 0x0730, 0x0734, 0x0780, 0x0784, 0x0790, 0x0794, 0x07a0,
-        0x07a4, 0x07b0, 0x07b4, 0x0000, 0x0004, 0x0010, 0x0014, 0x0020, 0x0024, 0x0030, 0x0034,
-        0x0080, 0x0084, 0x0090, 0x0094, 0x00a0, 0x00a4, 0x00b0, 0x00b4, 0x0100, 0x0104, 0x0110,
-        0x0114, 0x0120, 0x0124, 0x0130, 0x0134, 0x0180, 0x0184, 0x0190, 0x0194, 0x01a0, 0x01a4,
-        0x01b0, 0x01b4, 0x0200, 0x0204, 0x0210, 0x0214, 0x0220, 0x0224, 0x0230, 0x0234, 0x0280,
-        0x0284, 0x0290, 0x0294, 0x02a0, 0x02a4, 0x02b0, 0x02b4, 0x0300, 0x0304, 0x0310, 0x0314,
-        0x0320, 0x0324, 0x0330, 0x0334, 0x0380, 0x0384, 0x0390, 0x0394, 0x03a0, 0x03a4, 0x03b0,
-        0x03b4, 0x0400, 0x0404, 0x0410, 0x0414, 0x0420, 0x0424, 0x0430, 0x0434, 0x0480, 0x0484,
-        0x0490, 0x0494, 0x04a0, 0x04a4, 0x04b0, 0x04b4, 0x0500, 0x0504, 0x0510, 0x0514, 0x0520,
-        0x0524, 0x0530, 0x0534, 0x0580, 0x0584, 0x0590, 0x0594, 0x05a0, 0x05a4, 0x05b0, 0x05b4,
-        0x0600, 0x0604, 0x0610, 0x0614, 0x0620, 0x0624, 0x0630, 0x0634, 0x0680, 0x0684, 0x0690,
-        0x0694, 0x06a0, 0x06a4, 0x06b0, 0x06b4, 0x0700, 0x0704, 0x0710, 0x0714, 0x0720, 0x0724,
-        0x0730, 0x0734, 0x0780, 0x0784, 0x0790, 0x0794, 0x07a0, 0x07a4, 0x07b0, 0x07b4,
-    };
-    return xlut[x % 128] + ylut[y % 128];
-}
-
-static u32 GetMortonOffset128(u32 x, u32 y, u32 bytes_per_pixel) {
-    // Calculates the offset of the position of the pixel in Morton order
-    // Framebuffer images are split into 128x128 tiles.
-
-    constexpr u32 block_height = 128;
-    const u32 coarse_x = x & ~127;
-
-    const u32 i = MortonInterleave128(x, y);
-
-    const u32 offset = coarse_x * block_height;
-
-    return (i + offset) * bytes_per_pixel;
-}
-
 void MortonSwizzle(MortonSwizzleMode mode, Surface::PixelFormat format, u32 stride,
                    u32 block_height, u32 height, u32 block_depth, u32 depth, u32 tile_width_spacing,
                    u8* buffer, u8* addr) {
@@ -286,23 +193,4 @@ void MortonSwizzle(MortonSwizzleMode mode, Surface::PixelFormat format, u32 stri
                                      tile_width_spacing, buffer, addr);
 }
 
-void MortonCopyPixels128(MortonSwizzleMode mode, u32 width, u32 height, u32 bytes_per_pixel,
-                         u32 linear_bytes_per_pixel, u8* morton_data, u8* linear_data) {
-    const bool morton_to_linear = mode == MortonSwizzleMode::MortonToLinear;
-    u8* data_ptrs[2];
-    for (u32 y = 0; y < height; ++y) {
-        for (u32 x = 0; x < width; ++x) {
-            const u32 coarse_y = y & ~127;
-            const u32 morton_offset =
-                GetMortonOffset128(x, y, bytes_per_pixel) + coarse_y * width * bytes_per_pixel;
-            const u32 linear_pixel_index = (x + y * width) * linear_bytes_per_pixel;
-
-            data_ptrs[morton_to_linear ? 1 : 0] = morton_data + morton_offset;
-            data_ptrs[morton_to_linear ? 0 : 1] = &linear_data[linear_pixel_index];
-
-            std::memcpy(data_ptrs[0], data_ptrs[1], bytes_per_pixel);
-        }
-    }
-}
-
 } // namespace VideoCore

src/video_core/morton.h

@@ -15,7 +15,4 @@ void MortonSwizzle(MortonSwizzleMode mode, VideoCore::Surface::PixelFormat forma
                    u32 block_height, u32 height, u32 block_depth, u32 depth, u32 tile_width_spacing,
                    u8* buffer, u8* addr);
 
-void MortonCopyPixels128(MortonSwizzleMode mode, u32 width, u32 height, u32 bytes_per_pixel,
-                         u32 linear_bytes_per_pixel, u8* morton_data, u8* linear_data);
-
 } // namespace VideoCore

src/video_core/rasterizer_interface.h

@@ -50,7 +50,7 @@ public:
     /// and invalidated
     virtual void FlushAndInvalidateRegion(CacheAddr addr, u64 size) = 0;
 
-    // Notify the rasterizer to send all written commands to the host GPU.
+    /// Notify the rasterizer to send all written commands to the host GPU.
     virtual void FlushCommands() = 0;
 
     /// Notify rasterizer that a frame is about to finish

src/video_core/renderer_base.h

@@ -36,8 +36,7 @@ public:
     virtual ~RendererBase();
 
     /// Swap buffers (render frame)
-    virtual void SwapBuffers(
-        std::optional<std::reference_wrapper<const Tegra::FramebufferConfig>> framebuffer) = 0;
+    virtual void SwapBuffers(const Tegra::FramebufferConfig* framebuffer) = 0;
 
     /// Initialize the renderer
     virtual bool Init() = 0;

src/video_core/renderer_opengl/gl_buffer_cache.cpp

@@ -8,6 +8,7 @@
 
 #include "common/assert.h"
 #include "common/microprofile.h"
+#include "video_core/rasterizer_interface.h"
 #include "video_core/renderer_opengl/gl_buffer_cache.h"
 #include "video_core/renderer_opengl/gl_rasterizer.h"
 #include "video_core/renderer_opengl/gl_resource_manager.h"

src/video_core/renderer_opengl/gl_device.cpp

@@ -14,12 +14,22 @@
 namespace OpenGL {
 
 namespace {
+
 template <typename T>
 T GetInteger(GLenum pname) {
     GLint temporary;
     glGetIntegerv(pname, &temporary);
     return static_cast<T>(temporary);
 }
+
+bool TestProgram(const GLchar* glsl) {
+    const GLuint shader{glCreateShaderProgramv(GL_VERTEX_SHADER, 1, &glsl)};
+    GLint link_status;
+    glGetProgramiv(shader, GL_LINK_STATUS, &link_status);
+    glDeleteProgram(shader);
+    return link_status == GL_TRUE;
+}
+
 } // Anonymous namespace
 
 Device::Device() {
@@ -32,6 +42,11 @@ Device::Device() {
     has_vertex_viewport_layer = GLAD_GL_ARB_shader_viewport_layer_array;
     has_variable_aoffi = TestVariableAoffi();
     has_component_indexing_bug = TestComponentIndexingBug();
+    has_precise_bug = TestPreciseBug();
+
+    LOG_INFO(Render_OpenGL, "Renderer_VariableAOFFI: {}", has_variable_aoffi);
+    LOG_INFO(Render_OpenGL, "Renderer_ComponentIndexingBug: {}", has_component_indexing_bug);
+    LOG_INFO(Render_OpenGL, "Renderer_PreciseBug: {}", has_precise_bug);
 }
 
 Device::Device(std::nullptr_t) {
@@ -42,30 +57,21 @@ Device::Device(std::nullptr_t) {
     has_vertex_viewport_layer = true;
     has_variable_aoffi = true;
     has_component_indexing_bug = false;
+    has_precise_bug = false;
 }
 
 bool Device::TestVariableAoffi() {
-    const GLchar* AOFFI_TEST = R"(#version 430 core
+    return TestProgram(R"(#version 430 core
 // This is a unit test, please ignore me on apitrace bug reports.
 uniform sampler2D tex;
 uniform ivec2 variable_offset;
 out vec4 output_attribute;
 void main() {
     output_attribute = textureOffset(tex, vec2(0), variable_offset);
-}
-)";
-    const GLuint shader{glCreateShaderProgramv(GL_VERTEX_SHADER, 1, &AOFFI_TEST)};
-    GLint link_status{};
-    glGetProgramiv(shader, GL_LINK_STATUS, &link_status);
-    glDeleteProgram(shader);
-
-    const bool supported{link_status == GL_TRUE};
-    LOG_INFO(Render_OpenGL, "Renderer_VariableAOFFI: {}", supported);
-    return supported;
+})");
 }
 
 bool Device::TestComponentIndexingBug() {
-    constexpr char log_message[] = "Renderer_ComponentIndexingBug: {}";
     const GLchar* COMPONENT_TEST = R"(#version 430 core
 layout (std430, binding = 0) buffer OutputBuffer {
     uint output_value;
@@ -105,12 +111,21 @@ void main() {
         GLuint result;
         glGetNamedBufferSubData(ssbo.handle, 0, sizeof(result), &result);
         if (result != values.at(index)) {
-            LOG_INFO(Render_OpenGL, log_message, true);
             return true;
         }
     }
-    LOG_INFO(Render_OpenGL, log_message, false);
     return false;
 }
 
+bool Device::TestPreciseBug() {
+    return !TestProgram(R"(#version 430 core
+in vec3 coords;
+out float out_value;
+uniform sampler2DShadow tex;
+void main() {
+    precise float tmp_value = vec4(texture(tex, coords)).x;
+    out_value = tmp_value;
+})");
+}
+
 } // namespace OpenGL

src/video_core/renderer_opengl/gl_device.h

@@ -46,9 +46,14 @@ public:
         return has_component_indexing_bug;
     }
 
+    bool HasPreciseBug() const {
+        return has_precise_bug;
+    }
+
 private:
     static bool TestVariableAoffi();
     static bool TestComponentIndexingBug();
+    static bool TestPreciseBug();
 
     std::size_t uniform_buffer_alignment{};
     std::size_t shader_storage_alignment{};
@@ -58,6 +63,7 @@ private:
     bool has_vertex_viewport_layer{};
     bool has_variable_aoffi{};
     bool has_component_indexing_bug{};
+    bool has_precise_bug{};
 };
 
 } // namespace OpenGL

src/video_core/renderer_opengl/gl_rasterizer.cpp

@@ -537,8 +537,7 @@ std::pair<bool, bool> RasterizerOpenGL::ConfigureFramebuffers(
         texture_cache.MarkDepthBufferInUse();
 
         fbkey.zeta = depth_surface;
-        fbkey.stencil_enable = regs.stencil_enable &&
-                               depth_surface->GetSurfaceParams().type == SurfaceType::DepthStencil;
+        fbkey.stencil_enable = depth_surface->GetSurfaceParams().type == SurfaceType::DepthStencil;
     }
 
     texture_cache.GuardRenderTargets(false);
@@ -577,16 +576,15 @@ void RasterizerOpenGL::ConfigureClearFramebuffer(OpenGLState& current_state, boo
     if (depth_surface) {
         const auto& params = depth_surface->GetSurfaceParams();
         switch (params.type) {
-        case VideoCore::Surface::SurfaceType::Depth: {
+        case VideoCore::Surface::SurfaceType::Depth:
             depth_surface->Attach(GL_DEPTH_ATTACHMENT, GL_DRAW_FRAMEBUFFER);
             glFramebufferTexture2D(GL_DRAW_FRAMEBUFFER, GL_STENCIL_ATTACHMENT, GL_TEXTURE_2D, 0, 0);
             break;
-        }
-        case VideoCore::Surface::SurfaceType::DepthStencil: {
-            depth_surface->Attach(GL_DEPTH_ATTACHMENT, GL_DRAW_FRAMEBUFFER);
+        case VideoCore::Surface::SurfaceType::DepthStencil:
+            depth_surface->Attach(GL_DEPTH_STENCIL_ATTACHMENT, GL_DRAW_FRAMEBUFFER);
             break;
-        }
-        default: { UNIMPLEMENTED(); }
+        default:
+            UNIMPLEMENTED();
         }
     } else {
         glFramebufferTexture2D(GL_DRAW_FRAMEBUFFER, GL_DEPTH_STENCIL_ATTACHMENT, GL_TEXTURE_2D, 0,
@@ -639,6 +637,7 @@ void RasterizerOpenGL::Clear() {
         ASSERT_MSG(regs.zeta_enable != 0, "Tried to clear stencil but buffer is not enabled!");
         use_stencil = true;
         clear_state.stencil.test_enabled = true;
+
         if (regs.clear_flags.stencil) {
             // Stencil affects the clear so fill it with the used masks
             clear_state.stencil.front.test_func = GL_ALWAYS;
@@ -708,8 +707,6 @@ void RasterizerOpenGL::DrawArrays() {
         return;
     }
 
-    const auto& regs = gpu.regs;
-
     SyncColorMask();
     SyncFragmentColorClampState();
     SyncMultiSampleState();
@@ -1121,9 +1118,12 @@ void RasterizerOpenGL::SyncStencilTestState() {
     if (!maxwell3d.dirty.stencil_test) {
         return;
     }
-    const auto& regs = maxwell3d.regs;
+    maxwell3d.dirty.stencil_test = false;
 
+    const auto& regs = maxwell3d.regs;
     state.stencil.test_enabled = regs.stencil_enable != 0;
+    state.MarkDirtyStencilState();
+
     if (!regs.stencil_enable) {
         return;
     }
@@ -1152,8 +1152,6 @@ void RasterizerOpenGL::SyncStencilTestState() {
         state.stencil.back.action_depth_fail = GL_KEEP;
         state.stencil.back.action_depth_pass = GL_KEEP;
     }
-    state.MarkDirtyStencilState();
-    maxwell3d.dirty.stencil_test = false;
 }
 
 void RasterizerOpenGL::SyncColorMask() {

src/video_core/renderer_opengl/gl_shader_cache.cpp

@@ -249,20 +249,24 @@ CachedProgram SpecializeShader(const std::string& code, const GLShader::ShaderEn
         if (!texture_buffer_usage.test(i)) {
             continue;
         }
-        source += fmt::format("#define SAMPLER_{}_IS_BUFFER", i);
+        source += fmt::format("#define SAMPLER_{}_IS_BUFFER\n", i);
+    }
+    if (texture_buffer_usage.any()) {
+        source += '\n';
     }
 
     if (program_type == ProgramType::Geometry) {
         const auto [glsl_topology, debug_name, max_vertices] =
             GetPrimitiveDescription(primitive_mode);
 
-        source += "layout (" + std::string(glsl_topology) + ") in;\n";
+        source += "layout (" + std::string(glsl_topology) + ") in;\n\n";
         source += "#define MAX_VERTEX_INPUT " + std::to_string(max_vertices) + '\n';
     }
     if (program_type == ProgramType::Compute) {
         source += "layout (local_size_variable) in;\n";
     }
 
+    source += '\n';
     source += code;
 
     OGLShader shader;
@@ -291,7 +295,7 @@ std::set<GLenum> GetSupportedFormats() {
 
 CachedShader::CachedShader(const ShaderParameters& params, ProgramType program_type,
                            GLShader::ProgramResult result)
-    : RasterizerCacheObject{params.host_ptr}, host_ptr{params.host_ptr}, cpu_addr{params.cpu_addr},
+    : RasterizerCacheObject{params.host_ptr}, cpu_addr{params.cpu_addr},
       unique_identifier{params.unique_identifier}, program_type{program_type},
       disk_cache{params.disk_cache}, precompiled_programs{params.precompiled_programs},
       entries{result.second}, code{std::move(result.first)}, shader_length{entries.shader_length} {}
@@ -344,23 +348,16 @@ Shader CachedShader::CreateKernelFromCache(const ShaderParameters& params,
 }
 
 std::tuple<GLuint, BaseBindings> CachedShader::GetProgramHandle(const ProgramVariant& variant) {
-    GLuint handle{};
-    if (program_type == ProgramType::Geometry) {
-        handle = GetGeometryShader(variant);
-    } else {
-        const auto [entry, is_cache_miss] = programs.try_emplace(variant);
-        auto& program = entry->second;
-        if (is_cache_miss) {
-            program = TryLoadProgram(variant);
-            if (!program) {
-                program = SpecializeShader(code, entries, program_type, variant);
-                disk_cache.SaveUsage(GetUsage(variant));
-            }
-
-            LabelGLObject(GL_PROGRAM, program->handle, cpu_addr);
+    const auto [entry, is_cache_miss] = programs.try_emplace(variant);
+    auto& program = entry->second;
+    if (is_cache_miss) {
+        program = TryLoadProgram(variant);
+        if (!program) {
+            program = SpecializeShader(code, entries, program_type, variant);
+            disk_cache.SaveUsage(GetUsage(variant));
         }
 
-        handle = program->handle;
+        LabelGLObject(GL_PROGRAM, program->handle, cpu_addr);
     }
 
     auto base_bindings = variant.base_bindings;
@@ -371,52 +368,9 @@ std::tuple<GLuint, BaseBindings> CachedShader::GetProgramHandle(const ProgramVar
     base_bindings.gmem += static_cast<u32>(entries.global_memory_entries.size());
     base_bindings.sampler += static_cast<u32>(entries.samplers.size());
 
-    return {handle, base_bindings};
+    return {program->handle, base_bindings};
 }
 
-GLuint CachedShader::GetGeometryShader(const ProgramVariant& variant) {
-    const auto [entry, is_cache_miss] = geometry_programs.try_emplace(variant);
-    auto& programs = entry->second;
-
-    switch (variant.primitive_mode) {
-    case GL_POINTS:
-        return LazyGeometryProgram(programs.points, variant);
-    case GL_LINES:
-    case GL_LINE_STRIP:
-        return LazyGeometryProgram(programs.lines, variant);
-    case GL_LINES_ADJACENCY:
-    case GL_LINE_STRIP_ADJACENCY:
-        return LazyGeometryProgram(programs.lines_adjacency, variant);
-    case GL_TRIANGLES:
-    case GL_TRIANGLE_STRIP:
-    case GL_TRIANGLE_FAN:
-        return LazyGeometryProgram(programs.triangles, variant);
-    case GL_TRIANGLES_ADJACENCY:
-    case GL_TRIANGLE_STRIP_ADJACENCY:
-        return LazyGeometryProgram(programs.triangles_adjacency, variant);
-    default:
-        UNREACHABLE_MSG("Unknown primitive mode.");
-        return LazyGeometryProgram(programs.points, variant);
-    }
-}
-
-GLuint CachedShader::LazyGeometryProgram(CachedProgram& target_program,
-                                         const ProgramVariant& variant) {
-    if (target_program) {
-        return target_program->handle;
-    }
-    const auto [glsl_name, debug_name, vertices] = GetPrimitiveDescription(variant.primitive_mode);
-    target_program = TryLoadProgram(variant);
-    if (!target_program) {
-        target_program = SpecializeShader(code, entries, program_type, variant);
-        disk_cache.SaveUsage(GetUsage(variant));
-    }
-
-    LabelGLObject(GL_PROGRAM, target_program->handle, cpu_addr, debug_name);
-
-    return target_program->handle;
-};
-
 CachedProgram CachedShader::TryLoadProgram(const ProgramVariant& variant) const {
     const auto found = precompiled_programs.find(GetUsage(variant));
     if (found == precompiled_programs.end()) {

src/video_core/renderer_opengl/gl_shader_cache.h

@@ -86,27 +86,10 @@ private:
     explicit CachedShader(const ShaderParameters& params, ProgramType program_type,
                           GLShader::ProgramResult result);
 
-    // Geometry programs. These are needed because GLSL needs an input topology but it's not
-    // declared by the hardware. Workaround this issue by generating a different shader per input
-    // topology class.
-    struct GeometryPrograms {
-        CachedProgram points;
-        CachedProgram lines;
-        CachedProgram lines_adjacency;
-        CachedProgram triangles;
-        CachedProgram triangles_adjacency;
-    };
-
-    GLuint GetGeometryShader(const ProgramVariant& variant);
-
-    /// Generates a geometry shader or returns one that already exists.
-    GLuint LazyGeometryProgram(CachedProgram& target_program, const ProgramVariant& variant);
-
     CachedProgram TryLoadProgram(const ProgramVariant& variant) const;
 
     ShaderDiskCacheUsage GetUsage(const ProgramVariant& variant) const;
 
-    u8* host_ptr{};
     VAddr cpu_addr{};
     u64 unique_identifier{};
     ProgramType program_type{};
@@ -118,11 +101,6 @@ private:
     std::size_t shader_length{};
 
     std::unordered_map<ProgramVariant, CachedProgram> programs;
-    std::unordered_map<ProgramVariant, GeometryPrograms> geometry_programs;
-
-    std::unordered_map<u32, GLuint> cbuf_resource_cache;
-    std::unordered_map<u32, GLuint> gmem_resource_cache;
-    std::unordered_map<u32, GLint> uniform_cache;
 };
 
 class ShaderCacheOpenGL final : public RasterizerCache<Shader> {

src/video_core/renderer_opengl/gl_texture_cache.cpp

@@ -184,6 +184,9 @@ GLint GetSwizzleSource(SwizzleSource source) {
 }
 
 void ApplyTextureDefaults(const SurfaceParams& params, GLuint texture) {
+    if (params.IsBuffer()) {
+        return;
+    }
     glTextureParameteri(texture, GL_TEXTURE_MIN_FILTER, GL_LINEAR);
     glTextureParameteri(texture, GL_TEXTURE_MAG_FILTER, GL_LINEAR);
     glTextureParameteri(texture, GL_TEXTURE_WRAP_S, GL_CLAMP_TO_EDGE);
@@ -208,6 +211,7 @@ OGLTexture CreateTexture(const SurfaceParams& params, GLenum target, GLenum inte
         glNamedBufferStorage(texture_buffer.handle, params.width * params.GetBytesPerPixel(),
                              nullptr, GL_DYNAMIC_STORAGE_BIT);
         glTextureBuffer(texture.handle, internal_format, texture_buffer.handle);
+        break;
     case SurfaceTarget::Texture2D:
     case SurfaceTarget::TextureCubemap:
         glTextureStorage2D(texture.handle, params.emulated_levels, internal_format, params.width,

src/video_core/renderer_opengl/gl_texture_cache.h

@@ -51,7 +51,7 @@ public:
     }
 
 protected:
-    void DecorateSurfaceName();
+    void DecorateSurfaceName() override;
 
     View CreateView(const ViewParams& view_key) override;
     View CreateViewInner(const ViewParams& view_key, bool is_proxy);

src/video_core/renderer_opengl/renderer_opengl.cpp

@@ -101,9 +101,7 @@ RendererOpenGL::RendererOpenGL(Core::Frontend::EmuWindow& emu_window, Core::Syst
 
 RendererOpenGL::~RendererOpenGL() = default;
 
-void RendererOpenGL::SwapBuffers(
-    std::optional<std::reference_wrapper<const Tegra::FramebufferConfig>> framebuffer) {
-
+void RendererOpenGL::SwapBuffers(const Tegra::FramebufferConfig* framebuffer) {
     system.GetPerfStats().EndSystemFrame();
 
     // Maintain the rasterizer's state as a priority
@@ -113,9 +111,9 @@ void RendererOpenGL::SwapBuffers(
 
     if (framebuffer) {
         // If framebuffer is provided, reload it from memory to a texture
-        if (screen_info.texture.width != (GLsizei)framebuffer->get().width ||
-            screen_info.texture.height != (GLsizei)framebuffer->get().height ||
-            screen_info.texture.pixel_format != framebuffer->get().pixel_format) {
+        if (screen_info.texture.width != static_cast<GLsizei>(framebuffer->width) ||
+            screen_info.texture.height != static_cast<GLsizei>(framebuffer->height) ||
+            screen_info.texture.pixel_format != framebuffer->pixel_format) {
             // Reallocate texture if the framebuffer size has changed.
             // This is expected to not happen very often and hence should not be a
             // performance problem.
@@ -149,43 +147,43 @@ void RendererOpenGL::SwapBuffers(
  * Loads framebuffer from emulated memory into the active OpenGL texture.
  */
 void RendererOpenGL::LoadFBToScreenInfo(const Tegra::FramebufferConfig& framebuffer) {
-    const u32 bytes_per_pixel{Tegra::FramebufferConfig::BytesPerPixel(framebuffer.pixel_format)};
-    const u64 size_in_bytes{framebuffer.stride * framebuffer.height * bytes_per_pixel};
-    const VAddr framebuffer_addr{framebuffer.address + framebuffer.offset};
-
     // Framebuffer orientation handling
     framebuffer_transform_flags = framebuffer.transform_flags;
     framebuffer_crop_rect = framebuffer.crop_rect;
 
-    // Ensure no bad interactions with GL_UNPACK_ALIGNMENT, which by default
-    // only allows rows to have a memory alignement of 4.
-    ASSERT(framebuffer.stride % 4 == 0);
-
-    if (!rasterizer->AccelerateDisplay(framebuffer, framebuffer_addr, framebuffer.stride)) {
-        // Reset the screen info's display texture to its own permanent texture
-        screen_info.display_texture = screen_info.texture.resource.handle;
-
-        rasterizer->FlushRegion(ToCacheAddr(Memory::GetPointer(framebuffer_addr)), size_in_bytes);
-
-        constexpr u32 linear_bpp = 4;
-        VideoCore::MortonCopyPixels128(VideoCore::MortonSwizzleMode::MortonToLinear,
-                                       framebuffer.width, framebuffer.height, bytes_per_pixel,
-                                       linear_bpp, Memory::GetPointer(framebuffer_addr),
-                                       gl_framebuffer_data.data());
-
-        glPixelStorei(GL_UNPACK_ROW_LENGTH, static_cast<GLint>(framebuffer.stride));
-
-        // Update existing texture
-        // TODO: Test what happens on hardware when you change the framebuffer dimensions so that
-        //       they differ from the LCD resolution.
-        // TODO: Applications could theoretically crash yuzu here by specifying too large
-        //       framebuffer sizes. We should make sure that this cannot happen.
-        glTextureSubImage2D(screen_info.texture.resource.handle, 0, 0, 0, framebuffer.width,
-                            framebuffer.height, screen_info.texture.gl_format,
-                            screen_info.texture.gl_type, gl_framebuffer_data.data());
-
-        glPixelStorei(GL_UNPACK_ROW_LENGTH, 0);
+    const VAddr framebuffer_addr{framebuffer.address + framebuffer.offset};
+    if (rasterizer->AccelerateDisplay(framebuffer, framebuffer_addr, framebuffer.stride)) {
+        return;
     }
+
+    // Reset the screen info's display texture to its own permanent texture
+    screen_info.display_texture = screen_info.texture.resource.handle;
+
+    const auto pixel_format{
+        VideoCore::Surface::PixelFormatFromGPUPixelFormat(framebuffer.pixel_format)};
+    const u32 bytes_per_pixel{VideoCore::Surface::GetBytesPerPixel(pixel_format)};
+    const u64 size_in_bytes{framebuffer.stride * framebuffer.height * bytes_per_pixel};
+    const auto host_ptr{Memory::GetPointer(framebuffer_addr)};
+    rasterizer->FlushRegion(ToCacheAddr(host_ptr), size_in_bytes);
+
+    // TODO(Rodrigo): Read this from HLE
+    constexpr u32 block_height_log2 = 4;
+    VideoCore::MortonSwizzle(VideoCore::MortonSwizzleMode::MortonToLinear, pixel_format,
+                             framebuffer.stride, block_height_log2, framebuffer.height, 0, 1, 1,
+                             gl_framebuffer_data.data(), host_ptr);
+
+    glPixelStorei(GL_UNPACK_ROW_LENGTH, static_cast<GLint>(framebuffer.stride));
+
+    // Update existing texture
+    // TODO: Test what happens on hardware when you change the framebuffer dimensions so that
+    //       they differ from the LCD resolution.
+    // TODO: Applications could theoretically crash yuzu here by specifying too large
+    //       framebuffer sizes. We should make sure that this cannot happen.
+    glTextureSubImage2D(screen_info.texture.resource.handle, 0, 0, 0, framebuffer.width,
+                        framebuffer.height, screen_info.texture.gl_format,
+                        screen_info.texture.gl_type, gl_framebuffer_data.data());
+
+    glPixelStorei(GL_UNPACK_ROW_LENGTH, 0);
 }
 
 /**
@@ -276,22 +274,29 @@ void RendererOpenGL::ConfigureFramebufferTexture(TextureInfo& texture,
     texture.height = framebuffer.height;
     texture.pixel_format = framebuffer.pixel_format;
 
+    const auto pixel_format{
+        VideoCore::Surface::PixelFormatFromGPUPixelFormat(framebuffer.pixel_format)};
+    const u32 bytes_per_pixel{VideoCore::Surface::GetBytesPerPixel(pixel_format)};
+    gl_framebuffer_data.resize(texture.width * texture.height * bytes_per_pixel);
+
     GLint internal_format;
     switch (framebuffer.pixel_format) {
     case Tegra::FramebufferConfig::PixelFormat::ABGR8:
         internal_format = GL_RGBA8;
         texture.gl_format = GL_RGBA;
         texture.gl_type = GL_UNSIGNED_INT_8_8_8_8_REV;
-        gl_framebuffer_data.resize(texture.width * texture.height * 4);
+        break;
+    case Tegra::FramebufferConfig::PixelFormat::RGB565:
+        internal_format = GL_RGB565;
+        texture.gl_format = GL_RGB;
+        texture.gl_type = GL_UNSIGNED_SHORT_5_6_5;
         break;
     default:
         internal_format = GL_RGBA8;
         texture.gl_format = GL_RGBA;
         texture.gl_type = GL_UNSIGNED_INT_8_8_8_8_REV;
-        gl_framebuffer_data.resize(texture.width * texture.height * 4);
-        LOG_CRITICAL(Render_OpenGL, "Unknown framebuffer pixel format: {}",
-                     static_cast<u32>(framebuffer.pixel_format));
-        UNREACHABLE();
+        UNIMPLEMENTED_MSG("Unknown framebuffer pixel format: {}",
+                          static_cast<u32>(framebuffer.pixel_format));
     }
 
     texture.resource.Release();

src/video_core/renderer_opengl/renderer_opengl.h

@@ -43,14 +43,13 @@ struct ScreenInfo {
     TextureInfo texture;
 };
 
-class RendererOpenGL : public VideoCore::RendererBase {
+class RendererOpenGL final : public VideoCore::RendererBase {
 public:
     explicit RendererOpenGL(Core::Frontend::EmuWindow& emu_window, Core::System& system);
     ~RendererOpenGL() override;
 
     /// Swap buffers (render frame)
-    void SwapBuffers(
-        std::optional<std::reference_wrapper<const Tegra::FramebufferConfig>> framebuffer) override;
+    void SwapBuffers(const Tegra::FramebufferConfig* framebuffer) override;
 
     /// Initialize the renderer
     bool Init() override;

src/video_core/shader/decode/conversion.cpp

@@ -14,6 +14,12 @@ using Tegra::Shader::Instruction;
 using Tegra::Shader::OpCode;
 using Tegra::Shader::Register;
 
+namespace {
+constexpr OperationCode GetFloatSelector(u64 selector) {
+    return selector == 0 ? OperationCode::FCastHalf0 : OperationCode::FCastHalf1;
+}
+} // Anonymous namespace
+
 u32 ShaderIR::DecodeConversion(NodeBlock& bb, u32 pc) {
     const Instruction instr = {program_code[pc]};
     const auto opcode = OpCode::Decode(instr);
@@ -22,7 +28,7 @@ u32 ShaderIR::DecodeConversion(NodeBlock& bb, u32 pc) {
     case OpCode::Id::I2I_R:
     case OpCode::Id::I2I_C:
     case OpCode::Id::I2I_IMM: {
-        UNIMPLEMENTED_IF(instr.conversion.selector);
+        UNIMPLEMENTED_IF(instr.conversion.int_src.selector != 0);
         UNIMPLEMENTED_IF(instr.conversion.dst_size != Register::Size::Word);
         UNIMPLEMENTED_IF(instr.alu.saturate_d);
 
@@ -57,8 +63,8 @@ u32 ShaderIR::DecodeConversion(NodeBlock& bb, u32 pc) {
     case OpCode::Id::I2F_R:
     case OpCode::Id::I2F_C:
     case OpCode::Id::I2F_IMM: {
+        UNIMPLEMENTED_IF(instr.conversion.int_src.selector != 0);
         UNIMPLEMENTED_IF(instr.conversion.dst_size == Register::Size::Long);
-        UNIMPLEMENTED_IF(instr.conversion.selector);
         UNIMPLEMENTED_IF_MSG(instr.generates_cc,
                              "Condition codes generation in I2F is not implemented");
 
@@ -113,8 +119,10 @@ u32 ShaderIR::DecodeConversion(NodeBlock& bb, u32 pc) {
         }();
 
         if (instr.conversion.src_size == Register::Size::Short) {
-            // TODO: figure where extract is sey in the encoding
-            value = Operation(OperationCode::FCastHalf0, PRECISE, value);
+            value = Operation(GetFloatSelector(instr.conversion.float_src.selector), NO_PRECISE,
+                              std::move(value));
+        } else {
+            ASSERT(instr.conversion.float_src.selector == 0);
         }
 
         value = GetOperandAbsNegFloat(value, instr.conversion.abs_a, instr.conversion.negate_a);
@@ -169,8 +177,10 @@ u32 ShaderIR::DecodeConversion(NodeBlock& bb, u32 pc) {
         }();
 
         if (instr.conversion.src_size == Register::Size::Short) {
-            // TODO: figure where extract is sey in the encoding
-            value = Operation(OperationCode::FCastHalf0, PRECISE, value);
+            value = Operation(GetFloatSelector(instr.conversion.float_src.selector), NO_PRECISE,
+                              std::move(value));
+        } else {
+            ASSERT(instr.conversion.float_src.selector == 0);
         }
 
         value = GetOperandAbsNegFloat(value, instr.conversion.abs_a, instr.conversion.negate_a);

src/video_core/shader/decode/float_set.cpp

@@ -15,7 +15,6 @@ using Tegra::Shader::OpCode;
 
 u32 ShaderIR::DecodeFloatSet(NodeBlock& bb, u32 pc) {
     const Instruction instr = {program_code[pc]};
-    const auto opcode = OpCode::Decode(instr);
 
     const Node op_a = GetOperandAbsNegFloat(GetRegister(instr.gpr8), instr.fset.abs_a != 0,
                                             instr.fset.neg_a != 0);

src/video_core/shader/decode/float_set_predicate.cpp

@@ -16,10 +16,9 @@ using Tegra::Shader::Pred;
 
 u32 ShaderIR::DecodeFloatSetPredicate(NodeBlock& bb, u32 pc) {
     const Instruction instr = {program_code[pc]};
-    const auto opcode = OpCode::Decode(instr);
 
-    const Node op_a = GetOperandAbsNegFloat(GetRegister(instr.gpr8), instr.fsetp.abs_a != 0,
-                                            instr.fsetp.neg_a != 0);
+    Node op_a = GetOperandAbsNegFloat(GetRegister(instr.gpr8), instr.fsetp.abs_a != 0,
+                                      instr.fsetp.neg_a != 0);
     Node op_b = [&]() {
         if (instr.is_b_imm) {
             return GetImmediate19(instr);
@@ -29,12 +28,13 @@ u32 ShaderIR::DecodeFloatSetPredicate(NodeBlock& bb, u32 pc) {
             return GetConstBuffer(instr.cbuf34.index, instr.cbuf34.GetOffset());
         }
     }();
-    op_b = GetOperandAbsNegFloat(op_b, instr.fsetp.abs_b, false);
+    op_b = GetOperandAbsNegFloat(std::move(op_b), instr.fsetp.abs_b, instr.fsetp.neg_b);
 
     // We can't use the constant predicate as destination.
     ASSERT(instr.fsetp.pred3 != static_cast<u64>(Pred::UnusedIndex));
 
-    const Node predicate = GetPredicateComparisonFloat(instr.fsetp.cond, op_a, op_b);
+    const Node predicate =
+        GetPredicateComparisonFloat(instr.fsetp.cond, std::move(op_a), std::move(op_b));
     const Node second_pred = GetPredicate(instr.fsetp.pred39, instr.fsetp.neg_pred != 0);
 
     const OperationCode combiner = GetPredicateCombiner(instr.fsetp.op);

src/video_core/shader/decode/half_set_predicate.cpp

@@ -42,9 +42,8 @@ u32 ShaderIR::DecodeHalfSetPredicate(NodeBlock& bb, u32 pc) {
         cond = instr.hsetp2.reg.cond;
         h_and = instr.hsetp2.reg.h_and;
         op_b =
-            UnpackHalfFloat(GetOperandAbsNegHalf(GetRegister(instr.gpr20), instr.hsetp2.reg.abs_b,
-                                                 instr.hsetp2.reg.negate_b),
-                            instr.hsetp2.reg.type_b);
+            GetOperandAbsNegHalf(UnpackHalfFloat(GetRegister(instr.gpr20), instr.hsetp2.reg.type_b),
+                                 instr.hsetp2.reg.abs_b, instr.hsetp2.reg.negate_b);
         break;
     default:
         UNREACHABLE();
@@ -52,22 +51,22 @@ u32 ShaderIR::DecodeHalfSetPredicate(NodeBlock& bb, u32 pc) {
     }
 
     const OperationCode combiner = GetPredicateCombiner(instr.hsetp2.op);
-    const Node combined_pred = GetPredicate(instr.hsetp2.pred3, instr.hsetp2.neg_pred);
+    const Node combined_pred = GetPredicate(instr.hsetp2.pred39, instr.hsetp2.neg_pred);
 
     const auto Write = [&](u64 dest, Node src) {
         SetPredicate(bb, dest, Operation(combiner, std::move(src), combined_pred));
     };
 
     const Node comparison = GetPredicateComparisonHalf(cond, op_a, op_b);
-    const u64 first = instr.hsetp2.pred0;
-    const u64 second = instr.hsetp2.pred39;
+    const u64 first = instr.hsetp2.pred3;
+    const u64 second = instr.hsetp2.pred0;
     if (h_and) {
-        const Node joined = Operation(OperationCode::LogicalAnd2, comparison);
+        Node joined = Operation(OperationCode::LogicalAnd2, comparison);
         Write(first, joined);
-        Write(second, Operation(OperationCode::LogicalNegate, joined));
+        Write(second, Operation(OperationCode::LogicalNegate, std::move(joined)));
     } else {
-        Write(first, Operation(OperationCode::LogicalPick2, comparison, Immediate(0u)));
-        Write(second, Operation(OperationCode::LogicalPick2, comparison, Immediate(1u)));
+        Write(first, Operation(OperationCode::LogicalPick2, comparison, Immediate(0U)));
+        Write(second, Operation(OperationCode::LogicalPick2, comparison, Immediate(1U)));
     }
 
     return pc;

src/video_core/shader/decode/integer_set.cpp

@@ -14,7 +14,6 @@ using Tegra::Shader::OpCode;
 
 u32 ShaderIR::DecodeIntegerSet(NodeBlock& bb, u32 pc) {
     const Instruction instr = {program_code[pc]};
-    const auto opcode = OpCode::Decode(instr);
 
     const Node op_a = GetRegister(instr.gpr8);
     const Node op_b = [&]() {

src/video_core/shader/decode/integer_set_predicate.cpp

@@ -16,7 +16,6 @@ using Tegra::Shader::Pred;
 
 u32 ShaderIR::DecodeIntegerSetPredicate(NodeBlock& bb, u32 pc) {
     const Instruction instr = {program_code[pc]};
-    const auto opcode = OpCode::Decode(instr);
 
     const Node op_a = GetRegister(instr.gpr8);
 

src/video_core/shader/decode/predicate_set_register.cpp

@@ -15,7 +15,6 @@ using Tegra::Shader::OpCode;
 
 u32 ShaderIR::DecodePredicateSetRegister(NodeBlock& bb, u32 pc) {
     const Instruction instr = {program_code[pc]};
-    const auto opcode = OpCode::Decode(instr);
 
     UNIMPLEMENTED_IF_MSG(instr.generates_cc,
                          "Condition codes generation in PSET is not implemented");

src/video_core/surface.cpp

@@ -445,11 +445,12 @@ PixelFormat PixelFormatFromGPUPixelFormat(Tegra::FramebufferConfig::PixelFormat
     switch (format) {
     case Tegra::FramebufferConfig::PixelFormat::ABGR8:
         return PixelFormat::ABGR8U;
+    case Tegra::FramebufferConfig::PixelFormat::RGB565:
+        return PixelFormat::B5G6R5U;
     case Tegra::FramebufferConfig::PixelFormat::BGRA8:
         return PixelFormat::BGRA8;
     default:
-        LOG_CRITICAL(HW_GPU, "Unimplemented format={}", static_cast<u32>(format));
-        UNREACHABLE();
+        UNIMPLEMENTED_MSG("Unimplemented format={}", static_cast<u32>(format));
         return PixelFormat::ABGR8U;
     }
 }

src/video_core/texture_cache/surface_params.h

@@ -58,7 +58,6 @@ public:
     std::size_t GetHostSizeInBytes() const {
         std::size_t host_size_in_bytes;
         if (GetCompressionType() == SurfaceCompression::Converted) {
-            constexpr std::size_t rgb8_bpp = 4ULL;
             // ASTC is uncompressed in software, in emulated as RGBA8
             host_size_in_bytes = 0;
             for (u32 level = 0; level < num_levels; ++level) {

src/video_core/texture_cache/texture_cache.h

@@ -308,8 +308,6 @@ protected:
         if (!guard_render_targets && surface->IsRenderTarget()) {
             ManageRenderTargetUnregister(surface);
         }
-        const GPUVAddr gpu_addr = surface->GetGpuAddr();
-        const CacheAddr cache_ptr = surface->GetCacheAddr();
         const std::size_t size = surface->GetSizeInBytes();
         const VAddr cpu_addr = surface->GetCpuAddr();
         rasterizer.UpdatePagesCachedCount(cpu_addr, size, -1);

src/video_core/textures/decoders.cpp

@@ -257,19 +257,21 @@ std::vector<u8> UnswizzleTexture(u8* address, u32 tile_size_x, u32 tile_size_y,
 
 void SwizzleSubrect(u32 subrect_width, u32 subrect_height, u32 source_pitch, u32 swizzled_width,
                     u32 bytes_per_pixel, u8* swizzled_data, u8* unswizzled_data,
-                    u32 block_height_bit) {
+                    u32 block_height_bit, u32 offset_x, u32 offset_y) {
     const u32 block_height = 1U << block_height_bit;
     const u32 image_width_in_gobs{(swizzled_width * bytes_per_pixel + (gob_size_x - 1)) /
                                   gob_size_x};
     for (u32 line = 0; line < subrect_height; ++line) {
+        const u32 dst_y = line + offset_y;
         const u32 gob_address_y =
-            (line / (gob_size_y * block_height)) * gob_size * block_height * image_width_in_gobs +
-            ((line % (gob_size_y * block_height)) / gob_size_y) * gob_size;
-        const auto& table = legacy_swizzle_table[line % gob_size_y];
+            (dst_y / (gob_size_y * block_height)) * gob_size * block_height * image_width_in_gobs +
+            ((dst_y % (gob_size_y * block_height)) / gob_size_y) * gob_size;
+        const auto& table = legacy_swizzle_table[dst_y % gob_size_y];
         for (u32 x = 0; x < subrect_width; ++x) {
+            const u32 dst_x = x + offset_x;
             const u32 gob_address =
-                gob_address_y + (x * bytes_per_pixel / gob_size_x) * gob_size * block_height;
-            const u32 swizzled_offset = gob_address + table[(x * bytes_per_pixel) % gob_size_x];
+                gob_address_y + (dst_x * bytes_per_pixel / gob_size_x) * gob_size * block_height;
+            const u32 swizzled_offset = gob_address + table[(dst_x * bytes_per_pixel) % gob_size_x];
             u8* source_line = unswizzled_data + line * source_pitch + x * bytes_per_pixel;
             u8* dest_addr = swizzled_data + swizzled_offset;
 

src/video_core/textures/decoders.h

@@ -44,7 +44,8 @@ std::size_t CalculateSize(bool tiled, u32 bytes_per_pixel, u32 width, u32 height
 
 /// Copies an untiled subrectangle into a tiled surface.
 void SwizzleSubrect(u32 subrect_width, u32 subrect_height, u32 source_pitch, u32 swizzled_width,
-                    u32 bytes_per_pixel, u8* swizzled_data, u8* unswizzled_data, u32 block_height);
+                    u32 bytes_per_pixel, u8* swizzled_data, u8* unswizzled_data, u32 block_height,
+                    u32 offset_x, u32 offset_y);
 
 /// Copies a tiled subrectangle into a linear surface.
 void UnswizzleSubrect(u32 subrect_width, u32 subrect_height, u32 dest_pitch, u32 swizzled_width,

src/video_core/textures/texture.h

@@ -213,7 +213,7 @@ struct TICEntry {
         if (header_version != TICHeaderVersion::OneDBuffer) {
             return width_minus_1 + 1;
         }
-        return (buffer_high_width_minus_one << 16) | buffer_low_width_minus_one;
+        return ((buffer_high_width_minus_one << 16) | buffer_low_width_minus_one) + 1;
     }
 
     u32 Height() const {