63 files changed, 1235 insertions, 331 deletions

diff --git a/CMakeLists.txt b/CMakeLists.txt
index 1f0af2d41..4668d4bea 100644
--- a/CMakeLists.txt
+++ b/CMakeLists.txt
@@ -79,6 +79,8 @@ else()
     add_definitions(/D_CRT_SECURE_NO_WARNINGS /D_CRT_NONSTDC_NO_DEPRECATE /D_SCL_SECURE_NO_WARNINGS)
     # Avoid windows.h junk
     add_definitions(/DNOMINMAX)
+    # Avoid windows.h from including some usually unused libs like winsocks.h, since this might cause some redefinition errors.
+    add_definitions(/DWIN32_LEAN_AND_MEAN)
 
     # set up output paths for executable binaries (.exe-files, and .dll-files on DLL-capable platforms)
     set(CMAKE_RUNTIME_OUTPUT_DIRECTORY ${CMAKE_BINARY_DIR}/bin)
@@ -92,10 +94,13 @@ else()
     # /W3 - Level 3 warnings
     # /MP - Multi-threaded compilation
     # /Zi - Output debugging information
-    # /Zo - enahnced debug info for optimized builds
-    set(CMAKE_C_FLAGS   "/W3 /MP /Zi /Zo" CACHE STRING "" FORCE)
+    # /Zo - enhanced debug info for optimized builds
+    # /permissive- - enables stricter C++ standards conformance checks
+    set(CMAKE_C_FLAGS   "/W3 /MP /Zi /Zo /permissive-" CACHE STRING "" FORCE)
     # /EHsc - C++-only exception handling semantics
-    set(CMAKE_CXX_FLAGS "${CMAKE_C_FLAGS} /EHsc" CACHE STRING "" FORCE)
+    # /Zc:throwingNew - let codegen assume `operator new` will never return null
+    # /Zc:inline - let codegen omit inline functions in object files
+    set(CMAKE_CXX_FLAGS "${CMAKE_C_FLAGS} /EHsc /Zc:throwingNew,inline" CACHE STRING "" FORCE)
 
     # /MDd - Multi-threaded Debug Runtime DLL
     set(CMAKE_C_FLAGS_DEBUG   "/Od /MDd" CACHE STRING "" FORCE)
diff --git a/README.md b/README.md
index 8de62d0ef..e766918f7 100644
--- a/README.md
+++ b/README.md
@@ -5,7 +5,7 @@ Citra Emulator
 [![Travis CI Build Status](https://travis-ci.org/citra-emu/citra.svg?branch=master)](https://travis-ci.org/citra-emu/citra)
 [![AppVeyor CI Build Status](https://ci.appveyor.com/api/projects/status/sdf1o4kh3g1e68m9?svg=true)](https://ci.appveyor.com/project/bunnei/citra)
 
-Citra is an experimental open-source Nintendo 3DS emulator/debugger written in C++. It is written with portability in mind, with builds actively maintained for Windows, Linux and macOS. Citra only emulates a subset of 3DS hardware, and therefore is generally only useful for running/debugging homebrew applications. At this time, Citra is even able to boot several commercial games! Most of these do not run to a playable state, but we are working every day to advance the project forward.
+Citra is an experimental open-source Nintendo 3DS emulator/debugger written in C++. It is written with portability in mind, with builds actively maintained for Windows, Linux and macOS. Citra only emulates a subset of 3DS hardware and therefore is generally only useful for running/debugging homebrew applications. At this time, Citra is even able to boot several commercial games! Most of these do not run to a playable state, but we are working every day to advance the project forward.
 
 Citra is licensed under the GPLv2 (or any later version). Refer to the license.txt file included. Please read the [FAQ](https://citra-emu.org/wiki/faq/) before getting started with the project.
 
@@ -27,7 +27,7 @@ If you want to contribute please take a look at the [Contributor's Guide](CONTRI
 
 
 ### Support
-We happily accept monetary donations, or donated games and hardware. Please see our [donations page](https://citra-emu.org/donate/) for more information on how you can contribute to Citra. Any donations received will go towards things like:
+We happily accept monetary donations or donated games and hardware. Please see our [donations page](https://citra-emu.org/donate/) for more information on how you can contribute to Citra. Any donations received will go towards things like:
 * 3DS consoles for developers to explore the hardware
 * 3DS games for testing
 * Any equipment required for homebrew
diff --git a/externals/CMakeLists.txt b/externals/CMakeLists.txt
index 1e04931ee..02e02350c 100644
--- a/externals/CMakeLists.txt
+++ b/externals/CMakeLists.txt
@@ -46,7 +46,5 @@ if (ARCHITECTURE_x86_64)
     # Defined before "dynarmic" above
     # add_library(xbyak INTERFACE)
     target_include_directories(xbyak INTERFACE ./xbyak/xbyak)
-    if (NOT MSVC)
-        target_compile_options(xbyak INTERFACE -fno-operator-names)
-    endif()
+    target_compile_definitions(xbyak INTERFACE XBYAK_NO_OP_NAMES)
 endif()
diff --git a/externals/cryptopp/CMakeLists.txt b/externals/cryptopp/CMakeLists.txt
index 864de18bb..8a626e44a 100644
--- a/externals/cryptopp/CMakeLists.txt
+++ b/externals/cryptopp/CMakeLists.txt
@@ -44,6 +44,11 @@ if(CMAKE_CXX_COMPILER_ID MATCHES "Intel")
     add_definitions(-wd68 -wd186 -wd279 -wd327 -wd161 -wd3180)
 endif()
 
+if(MSVC)
+  # Disable C4390: empty controlled statement found: is this the intent?
+  set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} /wd4390")
+endif()
+
 # Endianness
 TEST_BIG_ENDIAN(IS_BIG_ENDIAN)
 if(IS_BIG_ENDIAN)
diff --git a/externals/dynarmic b/externals/dynarmic
-Subproject 7707ff13e981b0aecf87f3156ee0b641469f7bb
+Subproject 8f15e3f70cb96e56705e5de6ba97b5d09423a56
diff --git a/src/citra/citra.cpp b/src/citra/citra.cpp
index dd357ff72..14574e56c 100644
--- a/src/citra/citra.cpp
+++ b/src/citra/citra.cpp
@@ -18,7 +18,10 @@
 #endif
 
 #ifdef _WIN32
+// windows.h needs to be included before shellapi.h
 #include <windows.h>
+
+#include <shellapi.h>
 #endif
 
 #include "citra/config.h"
diff --git a/src/citra/config.cpp b/src/citra/config.cpp
index f08b4069c..957d8dc86 100644
--- a/src/citra/config.cpp
+++ b/src/citra/config.cpp
@@ -88,9 +88,9 @@ void Config::ReadValues() {
     Settings::values.toggle_framelimit =
         sdl2_config->GetBoolean("Renderer", "toggle_framelimit", true);
 
-    Settings::values.bg_red = (float)sdl2_config->GetReal("Renderer", "bg_red", 1.0);
-    Settings::values.bg_green = (float)sdl2_config->GetReal("Renderer", "bg_green", 1.0);
-    Settings::values.bg_blue = (float)sdl2_config->GetReal("Renderer", "bg_blue", 1.0);
+    Settings::values.bg_red = (float)sdl2_config->GetReal("Renderer", "bg_red", 0.0);
+    Settings::values.bg_green = (float)sdl2_config->GetReal("Renderer", "bg_green", 0.0);
+    Settings::values.bg_blue = (float)sdl2_config->GetReal("Renderer", "bg_blue", 0.0);
 
     // Layout
     Settings::values.layout_option =
diff --git a/src/citra_qt/configuration/config.cpp b/src/citra_qt/configuration/config.cpp
index 2b99447ec..64ffc9152 100644
--- a/src/citra_qt/configuration/config.cpp
+++ b/src/citra_qt/configuration/config.cpp
@@ -70,9 +70,9 @@ void Config::ReadValues() {
     Settings::values.use_vsync = qt_config->value("use_vsync", false).toBool();
     Settings::values.toggle_framelimit = qt_config->value("toggle_framelimit", true).toBool();
 
-    Settings::values.bg_red = qt_config->value("bg_red", 1.0).toFloat();
-    Settings::values.bg_green = qt_config->value("bg_green", 1.0).toFloat();
-    Settings::values.bg_blue = qt_config->value("bg_blue", 1.0).toFloat();
+    Settings::values.bg_red = qt_config->value("bg_red", 0.0).toFloat();
+    Settings::values.bg_green = qt_config->value("bg_green", 0.0).toFloat();
+    Settings::values.bg_blue = qt_config->value("bg_blue", 0.0).toFloat();
     qt_config->endGroup();
 
     qt_config->beginGroup("Layout");
diff --git a/src/citra_qt/configuration/configure_debug.ui b/src/citra_qt/configuration/configure_debug.ui
index bbbb0e3f4..96638ebdb 100644
--- a/src/citra_qt/configuration/configure_debug.ui
+++ b/src/citra_qt/configuration/configure_debug.ui
@@ -23,6 +23,13 @@
        </property>
        <layout class="QVBoxLayout" name="verticalLayout_2">
         <item>
+         <widget class="QLabel">
+          <property name="text">
+           <string>The GDB Stub only works correctly when the CPU JIT is off.</string>
+          </property>
+         </widget>
+        </item>
+        <item>
          <layout class="QHBoxLayout" name="horizontalLayout_3">
           <item>
            <widget class="QCheckBox" name="toggle_gdbstub">
diff --git a/src/core/CMakeLists.txt b/src/core/CMakeLists.txt
index b16a89990..ea09819e5 100644
--- a/src/core/CMakeLists.txt
+++ b/src/core/CMakeLists.txt
@@ -144,6 +144,7 @@ set(SRCS
             hle/service/nwm/nwm_tst.cpp
             hle/service/nwm/nwm_uds.cpp
             hle/service/nwm/uds_beacon.cpp
+            hle/service/nwm/uds_data.cpp
             hle/service/pm_app.cpp
             hle/service/ptm/ptm.cpp
             hle/service/ptm/ptm_gets.cpp
@@ -341,6 +342,7 @@ set(HEADERS
             hle/service/nwm/nwm_tst.h
             hle/service/nwm/nwm_uds.h
             hle/service/nwm/uds_beacon.h
+            hle/service/nwm/uds_data.h
             hle/service/pm_app.h
             hle/service/ptm/ptm.h
             hle/service/ptm/ptm_gets.h
diff --git a/src/core/hle/function_wrappers.h b/src/core/hle/function_wrappers.h
index 2abdfe1b3..5e6002f4e 100644
--- a/src/core/hle/function_wrappers.h
+++ b/src/core/hle/function_wrappers.h
@@ -229,9 +229,18 @@ void Wrap() {
     u32 retval = func(&param_1, &param_2,
                       reinterpret_cast<const char*>(Memory::GetPointer(PARAM(2))), PARAM(3))
                      .raw;
-    // The first out parameter is moved into R2 and the second is moved into R1.
-    Core::CPU().SetReg(1, param_2);
-    Core::CPU().SetReg(2, param_1);
+    Core::CPU().SetReg(1, param_1);
+    Core::CPU().SetReg(2, param_2);
+    FuncReturn(retval);
+}
+
+template <ResultCode func(Kernel::Handle*, Kernel::Handle*)>
+void Wrap() {
+    Kernel::Handle param_1 = 0;
+    Kernel::Handle param_2 = 0;
+    u32 retval = func(&param_1, &param_2).raw;
+    Core::CPU().SetReg(1, param_1);
+    Core::CPU().SetReg(2, param_2);
     FuncReturn(retval);
 }
 
diff --git a/src/core/hle/kernel/errors.h b/src/core/hle/kernel/errors.h
index b3b60e7df..64aa61460 100644
--- a/src/core/hle/kernel/errors.h
+++ b/src/core/hle/kernel/errors.h
@@ -13,6 +13,7 @@ enum {
     OutOfHandles = 19,
     SessionClosedByRemote = 26,
     PortNameTooLong = 30,
+    NoPendingSessions = 35,
     WrongPermission = 46,
     InvalidBufferDescriptor = 48,
     MaxConnectionsReached = 52,
@@ -94,5 +95,9 @@ constexpr ResultCode ERR_OUT_OF_RANGE_KERNEL(ErrorDescription::OutOfRange, Error
                                              ErrorLevel::Permanent); // 0xD8E007FD
 constexpr ResultCode RESULT_TIMEOUT(ErrorDescription::Timeout, ErrorModule::OS,
                                     ErrorSummary::StatusChanged, ErrorLevel::Info);
+/// Returned when Accept() is called on a port with no sessions to be accepted.
+constexpr ResultCode ERR_NO_PENDING_SESSIONS(ErrCodes::NoPendingSessions, ErrorModule::OS,
+                                             ErrorSummary::WouldBlock,
+                                             ErrorLevel::Permanent); // 0xD8401823
 
 } // namespace Kernel
diff --git a/src/core/hle/kernel/hle_ipc.cpp b/src/core/hle/kernel/hle_ipc.cpp
index 6cf1886cf..5ebe2eca4 100644
--- a/src/core/hle/kernel/hle_ipc.cpp
+++ b/src/core/hle/kernel/hle_ipc.cpp
@@ -23,6 +23,11 @@ void SessionRequestHandler::ClientDisconnected(SharedPtr<ServerSession> server_s
     boost::range::remove_erase(connected_sessions, server_session);
 }
 
+HLERequestContext::HLERequestContext(SharedPtr<ServerSession> session)
+    : session(std::move(session)) {
+    cmd_buf[0] = 0;
+}
+
 HLERequestContext::~HLERequestContext() = default;
 
 SharedPtr<Object> HLERequestContext::GetIncomingHandle(u32 id_from_cmdbuf) const {
@@ -62,10 +67,13 @@ ResultCode HLERequestContext::PopulateFromIncomingCommandBuffer(const u32_le* sr
             ASSERT(i + num_handles <= command_size); // TODO(yuriks): Return error
             for (u32 j = 0; j < num_handles; ++j) {
                 Handle handle = src_cmdbuf[i];
-                SharedPtr<Object> object = src_table.GetGeneric(handle);
-                ASSERT(object != nullptr); // TODO(yuriks): Return error
-                if (descriptor == IPC::DescriptorType::MoveHandle) {
-                    src_table.Close(handle);
+                SharedPtr<Object> object = nullptr;
+                if (handle != 0) {
+                    object = src_table.GetGeneric(handle);
+                    ASSERT(object != nullptr); // TODO(yuriks): Return error
+                    if (descriptor == IPC::DescriptorType::MoveHandle) {
+                        src_table.Close(handle);
+                    }
                 }
 
                 cmd_buf[i++] = AddOutgoingHandle(std::move(object));
@@ -107,9 +115,11 @@ ResultCode HLERequestContext::WriteToOutgoingCommandBuffer(u32_le* dst_cmdbuf, P
             ASSERT(i + num_handles <= command_size);
             for (u32 j = 0; j < num_handles; ++j) {
                 SharedPtr<Object> object = GetIncomingHandle(cmd_buf[i]);
-
-                // TODO(yuriks): Figure out the proper error handling for if this fails
-                Handle handle = dst_table.Create(object).Unwrap();
+                Handle handle = 0;
+                if (object != nullptr) {
+                    // TODO(yuriks): Figure out the proper error handling for if this fails
+                    handle = dst_table.Create(object).Unwrap();
+                }
                 dst_cmdbuf[i++] = handle;
             }
             break;
diff --git a/src/core/hle/kernel/hle_ipc.h b/src/core/hle/kernel/hle_ipc.h
index cbb109d8f..35795fc1d 100644
--- a/src/core/hle/kernel/hle_ipc.h
+++ b/src/core/hle/kernel/hle_ipc.h
@@ -84,6 +84,7 @@ protected:
  */
 class HLERequestContext {
 public:
+    HLERequestContext(SharedPtr<ServerSession> session);
     ~HLERequestContext();
 
     /// Returns a pointer to the IPC command buffer for this request.
@@ -118,14 +119,14 @@ public:
      */
     void ClearIncomingObjects();
 
-private:
-    friend class Service::ServiceFrameworkBase;
-
+    /// Populates this context with data from the requesting process/thread.
     ResultCode PopulateFromIncomingCommandBuffer(const u32_le* src_cmdbuf, Process& src_process,
                                                  HandleTable& src_table);
+    /// Writes data from this context back to the requesting process/thread.
     ResultCode WriteToOutgoingCommandBuffer(u32_le* dst_cmdbuf, Process& dst_process,
                                             HandleTable& dst_table) const;
 
+private:
     std::array<u32, IPC::COMMAND_BUFFER_LENGTH> cmd_buf;
     SharedPtr<ServerSession> session;
     // TODO(yuriks): Check common usage of this and optimize size accordingly
diff --git a/src/core/hle/kernel/memory.cpp b/src/core/hle/kernel/memory.cpp
index 804f23b1c..496d07cb5 100644
--- a/src/core/hle/kernel/memory.cpp
+++ b/src/core/hle/kernel/memory.cpp
@@ -166,7 +166,7 @@ void HandleSpecialMapping(VMManager& address_space, const AddressMapping& mappin
     auto vma = address_space
                    .MapBackingMemory(mapping.address, target_pointer + offset_into_region,
                                      mapping.size, memory_state)
-                   .MoveFrom();
+                   .Unwrap();
     address_space.Reprotect(vma,
                             mapping.read_only ? VMAPermission::Read : VMAPermission::ReadWrite);
 }
@@ -176,14 +176,14 @@ void MapSharedPages(VMManager& address_space) {
                            .MapBackingMemory(Memory::CONFIG_MEMORY_VADDR,
                                              reinterpret_cast<u8*>(&ConfigMem::config_mem),
                                              Memory::CONFIG_MEMORY_SIZE, MemoryState::Shared)
-                           .MoveFrom();
+                           .Unwrap();
     address_space.Reprotect(cfg_mem_vma, VMAPermission::Read);
 
     auto shared_page_vma = address_space
                                .MapBackingMemory(Memory::SHARED_PAGE_VADDR,
                                                  reinterpret_cast<u8*>(&SharedPage::shared_page),
                                                  Memory::SHARED_PAGE_SIZE, MemoryState::Shared)
-                               .MoveFrom();
+                               .Unwrap();
     address_space.Reprotect(shared_page_vma, VMAPermission::Read);
 }
 
diff --git a/src/core/hle/kernel/process.cpp b/src/core/hle/kernel/process.cpp
index 1c31ec950..522ad2333 100644
--- a/src/core/hle/kernel/process.cpp
+++ b/src/core/hle/kernel/process.cpp
@@ -151,6 +151,8 @@ void Process::Run(s32 main_thread_priority, u32 stack_size) {
 }
 
 VAddr Process::GetLinearHeapAreaAddress() const {
+    // Starting from system version 8.0.0 a new linear heap layout is supported to allow usage of
+    // the extra RAM in the n3DS.
     return kernel_version < 0x22C ? Memory::LINEAR_HEAP_VADDR : Memory::NEW_LINEAR_HEAP_VADDR;
 }
 
diff --git a/src/core/hle/kernel/server_port.cpp b/src/core/hle/kernel/server_port.cpp
index 4d20c39a1..49a9cdfa3 100644
--- a/src/core/hle/kernel/server_port.cpp
+++ b/src/core/hle/kernel/server_port.cpp
@@ -5,8 +5,10 @@
 #include <tuple>
 #include "common/assert.h"
 #include "core/hle/kernel/client_port.h"
+#include "core/hle/kernel/errors.h"
 #include "core/hle/kernel/kernel.h"
 #include "core/hle/kernel/server_port.h"
+#include "core/hle/kernel/server_session.h"
 #include "core/hle/kernel/thread.h"
 
 namespace Kernel {
@@ -14,6 +16,16 @@ namespace Kernel {
 ServerPort::ServerPort() {}
 ServerPort::~ServerPort() {}
 
+ResultVal<SharedPtr<ServerSession>> ServerPort::Accept() {
+    if (pending_sessions.empty()) {
+        return ERR_NO_PENDING_SESSIONS;
+    }
+
+    auto session = std::move(pending_sessions.back());
+    pending_sessions.pop_back();
+    return MakeResult(std::move(session));
+}
+
 bool ServerPort::ShouldWait(Thread* thread) const {
     // If there are no pending sessions, we wait until a new one is added.
     return pending_sessions.size() == 0;
diff --git a/src/core/hle/kernel/server_port.h b/src/core/hle/kernel/server_port.h
index f1419cd46..6fe7c7f2f 100644
--- a/src/core/hle/kernel/server_port.h
+++ b/src/core/hle/kernel/server_port.h
@@ -14,6 +14,7 @@
 namespace Kernel {
 
 class ClientPort;
+class ServerSession;
 class SessionRequestHandler;
 
 class ServerPort final : public WaitObject {
@@ -41,6 +42,12 @@ public:
     }
 
     /**
+     * Accepts a pending incoming connection on this port. If there are no pending sessions, will
+     * return ERR_NO_PENDING_SESSIONS.
+     */
+    ResultVal<SharedPtr<ServerSession>> Accept();
+
+    /**
      * Sets the HLE handler template for the port. ServerSessions crated by connecting to this port
      * will inherit a reference to this handler.
      */
@@ -50,8 +57,8 @@ public:
 
     std::string name; ///< Name of port (optional)
 
-    std::vector<SharedPtr<WaitObject>>
-        pending_sessions; ///< ServerSessions waiting to be accepted by the port
+    /// ServerSessions waiting to be accepted by the port
+    std::vector<SharedPtr<ServerSession>> pending_sessions;
 
     /// This session's HLE request handler template (optional)
     /// ServerSessions created from this port inherit a reference to this handler.
diff --git a/src/core/hle/kernel/server_session.cpp b/src/core/hle/kernel/server_session.cpp
index 970eac5fe..337896abf 100644
--- a/src/core/hle/kernel/server_session.cpp
+++ b/src/core/hle/kernel/server_session.cpp
@@ -81,7 +81,7 @@ ResultCode ServerSession::HandleSyncRequest(SharedPtr<Thread> thread) {
 
 ServerSession::SessionPair ServerSession::CreateSessionPair(const std::string& name,
                                                             SharedPtr<ClientPort> port) {
-    auto server_session = ServerSession::Create(name + "_Server").MoveFrom();
+    auto server_session = ServerSession::Create(name + "_Server").Unwrap();
     SharedPtr<ClientSession> client_session(new ClientSession);
     client_session->name = name + "_Client";
 
diff --git a/src/core/hle/kernel/thread.cpp b/src/core/hle/kernel/thread.cpp
index 75ce626f8..f5f2eb2f7 100644
--- a/src/core/hle/kernel/thread.cpp
+++ b/src/core/hle/kernel/thread.cpp
@@ -389,7 +389,7 @@ ResultVal<SharedPtr<Thread>> Thread::Create(std::string name, VAddr entry_point,
     thread->wait_objects.clear();
     thread->wait_address = 0;
     thread->name = std::move(name);
-    thread->callback_handle = wakeup_callback_handle_table.Create(thread).MoveFrom();
+    thread->callback_handle = wakeup_callback_handle_table.Create(thread).Unwrap();
     thread->owner_process = g_current_process;
 
     // Find the next available TLS index, and mark it as used
@@ -484,7 +484,7 @@ SharedPtr<Thread> SetupMainThread(u32 entry_point, s32 priority) {
     auto thread_res = Thread::Create("main", entry_point, priority, 0, THREADPROCESSORID_0,
                                      Memory::HEAP_VADDR_END);
 
-    SharedPtr<Thread> thread = thread_res.MoveFrom();
+    SharedPtr<Thread> thread = std::move(thread_res).Unwrap();
 
     thread->context.fpscr =
         FPSCR_DEFAULT_NAN | FPSCR_FLUSH_TO_ZERO | FPSCR_ROUND_TOZERO | FPSCR_IXC; // 0x03C00010
diff --git a/src/core/hle/kernel/timer.cpp b/src/core/hle/kernel/timer.cpp
index 6f2cf3b02..d7ec93672 100644
--- a/src/core/hle/kernel/timer.cpp
+++ b/src/core/hle/kernel/timer.cpp
@@ -30,7 +30,7 @@ SharedPtr<Timer> Timer::Create(ResetType reset_type, std::string name) {
     timer->name = std::move(name);
     timer->initial_delay = 0;
     timer->interval_delay = 0;
-    timer->callback_handle = timer_callback_handle_table.Create(timer).MoveFrom();
+    timer->callback_handle = timer_callback_handle_table.Create(timer).Unwrap();
 
     return timer;
 }
diff --git a/src/core/hle/result.h b/src/core/hle/result.h
index 5f2cdbb96..47b6e2b23 100644
--- a/src/core/hle/result.h
+++ b/src/core/hle/result.h
@@ -388,13 +388,14 @@ public:
     }
 
     /// Asserts that the result succeeded and returns a reference to it.
-    T& Unwrap() {
+    T& Unwrap() & {
         ASSERT_MSG(Succeeded(), "Tried to Unwrap empty ResultVal");
         return **this;
     }
 
-    T&& MoveFrom() {
-        return std::move(Unwrap());
+    T&& Unwrap() && {
+        ASSERT_MSG(Succeeded(), "Tried to Unwrap empty ResultVal");
+        return std::move(**this);
     }
 
 private:
diff --git a/src/core/hle/service/apt/apt.cpp b/src/core/hle/service/apt/apt.cpp
index 4c587e3c8..25e7b777d 100644
--- a/src/core/hle/service/apt/apt.cpp
+++ b/src/core/hle/service/apt/apt.cpp
@@ -55,8 +55,8 @@ void Initialize(Service::Interface* self) {
     u32 flags = rp.Pop<u32>();
     IPC::RequestBuilder rb = rp.MakeBuilder(1, 3);
     rb.Push(RESULT_SUCCESS);
-    rb.PushCopyHandles(Kernel::g_handle_table.Create(notification_event).MoveFrom(),
-                       Kernel::g_handle_table.Create(parameter_event).MoveFrom());
+    rb.PushCopyHandles(Kernel::g_handle_table.Create(notification_event).Unwrap(),
+                       Kernel::g_handle_table.Create(parameter_event).Unwrap());
 
     // TODO(bunnei): Check if these events are cleared every time Initialize is called.
     notification_event->Clear();
@@ -93,7 +93,7 @@ void GetSharedFont(Service::Interface* self) {
     // allocated, the real APT service calculates this address by scanning the entire address space
     // (using svcQueryMemory) and searches for an allocation of the same size as the Shared Font.
     rb.Push(target_address);
-    rb.PushCopyHandles(Kernel::g_handle_table.Create(shared_font_mem).MoveFrom());
+    rb.PushCopyHandles(Kernel::g_handle_table.Create(shared_font_mem).Unwrap());
 }
 
 void NotifyToWait(Service::Interface* self) {
@@ -115,7 +115,7 @@ void GetLockHandle(Service::Interface* self) {
     rb.Push(RESULT_SUCCESS);    // No error
     rb.Push(applet_attributes); // Applet Attributes, this value is passed to Enable.
     rb.Push<u32>(0);            // Least significant bit = power button state
-    Kernel::Handle handle_copy = Kernel::g_handle_table.Create(lock).MoveFrom();
+    Kernel::Handle handle_copy = Kernel::g_handle_table.Create(lock).Unwrap();
     rb.PushCopyHandles(handle_copy);
 
     LOG_WARNING(Service_APT, "(STUBBED) called handle=0x%08X applet_attributes=0x%08X", handle_copy,
@@ -231,7 +231,7 @@ void ReceiveParameter(Service::Interface* self) {
     rb.Push(static_cast<u32>(next_parameter.buffer.size())); // Parameter buffer size
 
     rb.PushMoveHandles((next_parameter.object != nullptr)
-                           ? Kernel::g_handle_table.Create(next_parameter.object).MoveFrom()
+                           ? Kernel::g_handle_table.Create(next_parameter.object).Unwrap()
                            : 0);
     rb.PushStaticBuffer(buffer, static_cast<u32>(next_parameter.buffer.size()), 0);
 
@@ -261,7 +261,7 @@ void GlanceParameter(Service::Interface* self) {
     rb.Push(static_cast<u32>(next_parameter.buffer.size())); // Parameter buffer size
 
     rb.PushCopyHandles((next_parameter.object != nullptr)
-                           ? Kernel::g_handle_table.Create(next_parameter.object).MoveFrom()
+                           ? Kernel::g_handle_table.Create(next_parameter.object).Unwrap()
                            : 0);
     rb.PushStaticBuffer(buffer, static_cast<u32>(next_parameter.buffer.size()), 0);
 
diff --git a/src/core/hle/service/cam/cam.cpp b/src/core/hle/service/cam/cam.cpp
index 7394c844f..c9f9e9d95 100644
--- a/src/core/hle/service/cam/cam.cpp
+++ b/src/core/hle/service/cam/cam.cpp
@@ -347,7 +347,7 @@ void GetVsyncInterruptEvent(Service::Interface* self) {
         int port = *port_select.begin();
         rb.Push(RESULT_SUCCESS);
         rb.PushCopyHandles(
-            Kernel::g_handle_table.Create(ports[port].vsync_interrupt_event).MoveFrom());
+            Kernel::g_handle_table.Create(ports[port].vsync_interrupt_event).Unwrap());
     } else {
         LOG_ERROR(Service_CAM, "invalid port_select=%u", port_select.m_val);
         rb.Push(ERROR_INVALID_ENUM_VALUE);
@@ -366,7 +366,7 @@ void GetBufferErrorInterruptEvent(Service::Interface* self) {
         int port = *port_select.begin();
         rb.Push(RESULT_SUCCESS);
         rb.PushCopyHandles(
-            Kernel::g_handle_table.Create(ports[port].buffer_error_interrupt_event).MoveFrom());
+            Kernel::g_handle_table.Create(ports[port].buffer_error_interrupt_event).Unwrap());
     } else {
         LOG_ERROR(Service_CAM, "invalid port_select=%u", port_select.m_val);
         rb.Push(ERROR_INVALID_ENUM_VALUE);
@@ -400,7 +400,7 @@ void SetReceiving(Service::Interface* self) {
         }
 
         rb.Push(RESULT_SUCCESS);
-        rb.PushCopyHandles(Kernel::g_handle_table.Create(port.completion_event).MoveFrom());
+        rb.PushCopyHandles(Kernel::g_handle_table.Create(port.completion_event).Unwrap());
     } else {
         LOG_ERROR(Service_CAM, "invalid port_select=%u", port_select.m_val);
         rb.Push(ERROR_INVALID_ENUM_VALUE);
diff --git a/src/core/hle/service/cecd/cecd.cpp b/src/core/hle/service/cecd/cecd.cpp
index bd9814244..421006a9e 100644
--- a/src/core/hle/service/cecd/cecd.cpp
+++ b/src/core/hle/service/cecd/cecd.cpp
@@ -31,8 +31,8 @@ void GetCecStateAbbreviated(Service::Interface* self) {
 void GetCecInfoEventHandle(Service::Interface* self) {
     u32* cmd_buff = Kernel::GetCommandBuffer();
 
-    cmd_buff[1] = RESULT_SUCCESS.raw;                                      // No error
-    cmd_buff[3] = Kernel::g_handle_table.Create(cecinfo_event).MoveFrom(); // Event handle
+    cmd_buff[1] = RESULT_SUCCESS.raw;                                    // No error
+    cmd_buff[3] = Kernel::g_handle_table.Create(cecinfo_event).Unwrap(); // Event handle
 
     LOG_WARNING(Service_CECD, "(STUBBED) called");
 }
@@ -40,8 +40,8 @@ void GetCecInfoEventHandle(Service::Interface* self) {
 void GetChangeStateEventHandle(Service::Interface* self) {
     u32* cmd_buff = Kernel::GetCommandBuffer();
 
-    cmd_buff[1] = RESULT_SUCCESS.raw;                                           // No error
-    cmd_buff[3] = Kernel::g_handle_table.Create(change_state_event).MoveFrom(); // Event handle
+    cmd_buff[1] = RESULT_SUCCESS.raw;                                         // No error
+    cmd_buff[3] = Kernel::g_handle_table.Create(change_state_event).Unwrap(); // Event handle
 
     LOG_WARNING(Service_CECD, "(STUBBED) called");
 }
diff --git a/src/core/hle/service/cfg/cfg.cpp b/src/core/hle/service/cfg/cfg.cpp
index 5a7878b31..6624f1711 100644
--- a/src/core/hle/service/cfg/cfg.cpp
+++ b/src/core/hle/service/cfg/cfg.cpp
@@ -406,7 +406,7 @@ ResultCode UpdateConfigNANDSavegame() {
     auto config_result = Service::FS::OpenFileFromArchive(cfg_system_save_data_archive, path, mode);
     ASSERT_MSG(config_result.Succeeded(), "could not open file");
 
-    auto config = config_result.MoveFrom();
+    auto config = std::move(config_result).Unwrap();
     config->backend->Write(0, CONFIG_SAVEFILE_SIZE, 1, cfg_config_file_buffer.data());
 
     return RESULT_SUCCESS;
@@ -560,7 +560,7 @@ ResultCode LoadConfigNANDSaveFile() {
 
     // Read the file if it already exists
     if (config_result.Succeeded()) {
-        auto config = config_result.MoveFrom();
+        auto config = std::move(config_result).Unwrap();
         config->backend->Read(0, CONFIG_SAVEFILE_SIZE, cfg_config_file_buffer.data());
         return RESULT_SUCCESS;
     }
diff --git a/src/core/hle/service/csnd_snd.cpp b/src/core/hle/service/csnd_snd.cpp
index 1455f20ca..9471ec1ef 100644
--- a/src/core/hle/service/csnd_snd.cpp
+++ b/src/core/hle/service/csnd_snd.cpp
@@ -51,8 +51,8 @@ static void Initialize(Interface* self) {
 
     cmd_buff[1] = RESULT_SUCCESS.raw;
     cmd_buff[2] = IPC::CopyHandleDesc(2);
-    cmd_buff[3] = Kernel::g_handle_table.Create(mutex).MoveFrom();
-    cmd_buff[4] = Kernel::g_handle_table.Create(shared_memory).MoveFrom();
+    cmd_buff[3] = Kernel::g_handle_table.Create(mutex).Unwrap();
+    cmd_buff[4] = Kernel::g_handle_table.Create(shared_memory).Unwrap();
 
     LOG_WARNING(Service_CSND, "(STUBBED) called");
 }
diff --git a/src/core/hle/service/dsp_dsp.cpp b/src/core/hle/service/dsp_dsp.cpp
index 363066d14..7d746054f 100644
--- a/src/core/hle/service/dsp_dsp.cpp
+++ b/src/core/hle/service/dsp_dsp.cpp
@@ -168,7 +168,7 @@ static void GetSemaphoreEventHandle(Service::Interface* self) {
     cmd_buff[0] = IPC::MakeHeader(0x16, 1, 2);
     cmd_buff[1] = RESULT_SUCCESS.raw; // No error
     // cmd_buff[2] not set
-    cmd_buff[3] = Kernel::g_handle_table.Create(semaphore_event).MoveFrom(); // Event handle
+    cmd_buff[3] = Kernel::g_handle_table.Create(semaphore_event).Unwrap(); // Event handle
 
     LOG_WARNING(Service_DSP, "(STUBBED) called");
 }
diff --git a/src/core/hle/service/fs/archive.cpp b/src/core/hle/service/fs/archive.cpp
index 3605ef175..033fbc9aa 100644
--- a/src/core/hle/service/fs/archive.cpp
+++ b/src/core/hle/service/fs/archive.cpp
@@ -311,7 +311,7 @@ ResultVal<std::shared_ptr<File>> OpenFileFromArchive(ArchiveHandle archive_handl
     if (backend.Failed())
         return backend.Code();
 
-    auto file = std::shared_ptr<File>(new File(backend.MoveFrom(), path));
+    auto file = std::shared_ptr<File>(new File(std::move(backend).Unwrap(), path));
     return MakeResult<std::shared_ptr<File>>(std::move(file));
 }
 
@@ -401,7 +401,7 @@ ResultVal<std::shared_ptr<Directory>> OpenDirectoryFromArchive(ArchiveHandle arc
     if (backend.Failed())
         return backend.Code();
 
-    auto directory = std::shared_ptr<Directory>(new Directory(backend.MoveFrom(), path));
+    auto directory = std::shared_ptr<Directory>(new Directory(std::move(backend).Unwrap(), path));
     return MakeResult<std::shared_ptr<Directory>>(std::move(directory));
 }
 
diff --git a/src/core/hle/service/fs/fs_user.cpp b/src/core/hle/service/fs/fs_user.cpp
index 34e1783ec..b9eab7838 100644
--- a/src/core/hle/service/fs/fs_user.cpp
+++ b/src/core/hle/service/fs/fs_user.cpp
@@ -87,7 +87,7 @@ static void OpenFile(Service::Interface* self) {
         file->ClientConnected(std::get<SharedPtr<ServerSession>>(sessions));
 
         rb.PushMoveHandles(
-            Kernel::g_handle_table.Create(std::get<SharedPtr<ClientSession>>(sessions)).MoveFrom());
+            Kernel::g_handle_table.Create(std::get<SharedPtr<ClientSession>>(sessions)).Unwrap());
     } else {
         rb.PushMoveHandles(0);
         LOG_ERROR(Service_FS, "failed to get a handle for file %s", file_path.DebugStr().c_str());
@@ -153,7 +153,7 @@ static void OpenFileDirectly(Service::Interface* self) {
         file->ClientConnected(std::get<SharedPtr<ServerSession>>(sessions));
 
         cmd_buff[3] =
-            Kernel::g_handle_table.Create(std::get<SharedPtr<ClientSession>>(sessions)).MoveFrom();
+            Kernel::g_handle_table.Create(std::get<SharedPtr<ClientSession>>(sessions)).Unwrap();
     } else {
         cmd_buff[3] = 0;
         LOG_ERROR(Service_FS, "failed to get a handle for file %s mode=%u attributes=%u",
@@ -420,7 +420,7 @@ static void OpenDirectory(Service::Interface* self) {
         directory->ClientConnected(std::get<SharedPtr<ServerSession>>(sessions));
 
         cmd_buff[3] =
-            Kernel::g_handle_table.Create(std::get<SharedPtr<ClientSession>>(sessions)).MoveFrom();
+            Kernel::g_handle_table.Create(std::get<SharedPtr<ClientSession>>(sessions)).Unwrap();
     } else {
         LOG_ERROR(Service_FS, "failed to get a handle for directory type=%d size=%d data=%s",
                   dirname_type, dirname_size, dir_path.DebugStr().c_str());
diff --git a/src/core/hle/service/gsp_gpu.cpp b/src/core/hle/service/gsp_gpu.cpp
index 6ff0f4812..bc964ec60 100644
--- a/src/core/hle/service/gsp_gpu.cpp
+++ b/src/core/hle/service/gsp_gpu.cpp
@@ -389,8 +389,8 @@ static void RegisterInterruptRelayQueue(Interface* self) {
     } else {
         cmd_buff[1] = RESULT_SUCCESS.raw;
     }
-    cmd_buff[2] = g_thread_id++;                                             // Thread ID
-    cmd_buff[4] = Kernel::g_handle_table.Create(g_shared_memory).MoveFrom(); // GSP shared memory
+    cmd_buff[2] = g_thread_id++;                                           // Thread ID
+    cmd_buff[4] = Kernel::g_handle_table.Create(g_shared_memory).Unwrap(); // GSP shared memory
 
     g_interrupt_event->Signal(); // TODO(bunnei): Is this correct?
 
diff --git a/src/core/hle/service/hid/hid.cpp b/src/core/hle/service/hid/hid.cpp
index 5255f6dc8..2014b8461 100644
--- a/src/core/hle/service/hid/hid.cpp
+++ b/src/core/hle/service/hid/hid.cpp
@@ -253,12 +253,12 @@ void GetIPCHandles(Service::Interface* self) {
     cmd_buff[1] = 0;          // No error
     cmd_buff[2] = 0x14000000; // IPC Command Structure translate-header
     // TODO(yuriks): Return error from SendSyncRequest is this fails (part of IPC marshalling)
-    cmd_buff[3] = Kernel::g_handle_table.Create(Service::HID::shared_mem).MoveFrom();
-    cmd_buff[4] = Kernel::g_handle_table.Create(Service::HID::event_pad_or_touch_1).MoveFrom();
-    cmd_buff[5] = Kernel::g_handle_table.Create(Service::HID::event_pad_or_touch_2).MoveFrom();
-    cmd_buff[6] = Kernel::g_handle_table.Create(Service::HID::event_accelerometer).MoveFrom();
-    cmd_buff[7] = Kernel::g_handle_table.Create(Service::HID::event_gyroscope).MoveFrom();
-    cmd_buff[8] = Kernel::g_handle_table.Create(Service::HID::event_debug_pad).MoveFrom();
+    cmd_buff[3] = Kernel::g_handle_table.Create(Service::HID::shared_mem).Unwrap();
+    cmd_buff[4] = Kernel::g_handle_table.Create(Service::HID::event_pad_or_touch_1).Unwrap();
+    cmd_buff[5] = Kernel::g_handle_table.Create(Service::HID::event_pad_or_touch_2).Unwrap();
+    cmd_buff[6] = Kernel::g_handle_table.Create(Service::HID::event_accelerometer).Unwrap();
+    cmd_buff[7] = Kernel::g_handle_table.Create(Service::HID::event_gyroscope).Unwrap();
+    cmd_buff[8] = Kernel::g_handle_table.Create(Service::HID::event_debug_pad).Unwrap();
 }
 
 void EnableAccelerometer(Service::Interface* self) {
diff --git a/src/core/hle/service/ir/ir_rst.cpp b/src/core/hle/service/ir/ir_rst.cpp
index 0de698003..837413f93 100644
--- a/src/core/hle/service/ir/ir_rst.cpp
+++ b/src/core/hle/service/ir/ir_rst.cpp
@@ -145,8 +145,8 @@ static void GetHandles(Interface* self) {
     IPC::RequestParser rp(Kernel::GetCommandBuffer(), 0x01, 0, 0);
     IPC::RequestBuilder rb = rp.MakeBuilder(1, 3);
     rb.Push(RESULT_SUCCESS);
-    rb.PushMoveHandles(Kernel::g_handle_table.Create(Service::IR::shared_memory).MoveFrom(),
-                       Kernel::g_handle_table.Create(Service::IR::update_event).MoveFrom());
+    rb.PushMoveHandles(Kernel::g_handle_table.Create(Service::IR::shared_memory).Unwrap(),
+                       Kernel::g_handle_table.Create(Service::IR::update_event).Unwrap());
 }
 
 /**
diff --git a/src/core/hle/service/ir/ir_user.cpp b/src/core/hle/service/ir/ir_user.cpp
index fdecdce64..fbdf7a465 100644
--- a/src/core/hle/service/ir/ir_user.cpp
+++ b/src/core/hle/service/ir/ir_user.cpp
@@ -337,7 +337,7 @@ void GetReceiveEvent(Interface* self) {
     IPC::RequestBuilder rb(Kernel::GetCommandBuffer(), 0x0A, 1, 2);
 
     rb.Push(RESULT_SUCCESS);
-    rb.PushCopyHandles(Kernel::g_handle_table.Create(Service::IR::receive_event).MoveFrom());
+    rb.PushCopyHandles(Kernel::g_handle_table.Create(Service::IR::receive_event).Unwrap());
 
     LOG_INFO(Service_IR, "called");
 }
@@ -354,7 +354,7 @@ void GetSendEvent(Interface* self) {
     IPC::RequestBuilder rb(Kernel::GetCommandBuffer(), 0x0B, 1, 2);
 
     rb.Push(RESULT_SUCCESS);
-    rb.PushCopyHandles(Kernel::g_handle_table.Create(Service::IR::send_event).MoveFrom());
+    rb.PushCopyHandles(Kernel::g_handle_table.Create(Service::IR::send_event).Unwrap());
 
     LOG_INFO(Service_IR, "called");
 }
@@ -394,7 +394,7 @@ static void GetConnectionStatusEvent(Interface* self) {
     IPC::RequestBuilder rb(Kernel::GetCommandBuffer(), 0x0C, 1, 2);
 
     rb.Push(RESULT_SUCCESS);
-    rb.PushCopyHandles(Kernel::g_handle_table.Create(Service::IR::conn_status_event).MoveFrom());
+    rb.PushCopyHandles(Kernel::g_handle_table.Create(Service::IR::conn_status_event).Unwrap());
 
     LOG_INFO(Service_IR, "called");
 }
diff --git a/src/core/hle/service/mic_u.cpp b/src/core/hle/service/mic_u.cpp
index 35212b59b..23e1ff094 100644
--- a/src/core/hle/service/mic_u.cpp
+++ b/src/core/hle/service/mic_u.cpp
@@ -160,7 +160,7 @@ static void IsSampling(Interface* self) {
 static void GetBufferFullEvent(Interface* self) {
     u32* cmd_buff = Kernel::GetCommandBuffer();
     cmd_buff[1] = RESULT_SUCCESS.raw; // No error
-    cmd_buff[3] = Kernel::g_handle_table.Create(buffer_full_event).MoveFrom();
+    cmd_buff[3] = Kernel::g_handle_table.Create(buffer_full_event).Unwrap();
     LOG_WARNING(Service_MIC, "(STUBBED) called");
 }
 
diff --git a/src/core/hle/service/nfc/nfc.cpp b/src/core/hle/service/nfc/nfc.cpp
index b44a9f668..cb09ed0b7 100644
--- a/src/core/hle/service/nfc/nfc.cpp
+++ b/src/core/hle/service/nfc/nfc.cpp
@@ -95,7 +95,7 @@ void GetTagInRangeEvent(Interface* self) {
     cmd_buff[0] = IPC::MakeHeader(0xB, 1, 2);
     cmd_buff[1] = RESULT_SUCCESS.raw;
     cmd_buff[2] = IPC::CopyHandleDesc();
-    cmd_buff[3] = Kernel::g_handle_table.Create(tag_in_range_event).MoveFrom();
+    cmd_buff[3] = Kernel::g_handle_table.Create(tag_in_range_event).Unwrap();
     LOG_WARNING(Service_NFC, "(STUBBED) called");
 }
 
@@ -105,7 +105,7 @@ void GetTagOutOfRangeEvent(Interface* self) {
     cmd_buff[0] = IPC::MakeHeader(0xC, 1, 2);
     cmd_buff[1] = RESULT_SUCCESS.raw;
     cmd_buff[2] = IPC::CopyHandleDesc();
-    cmd_buff[3] = Kernel::g_handle_table.Create(tag_out_of_range_event).MoveFrom();
+    cmd_buff[3] = Kernel::g_handle_table.Create(tag_out_of_range_event).Unwrap();
     LOG_WARNING(Service_NFC, "(STUBBED) called");
 }
 
diff --git a/src/core/hle/service/nwm/nwm_uds.cpp b/src/core/hle/service/nwm/nwm_uds.cpp
index e92900d48..6dbdff044 100644
--- a/src/core/hle/service/nwm/nwm_uds.cpp
+++ b/src/core/hle/service/nwm/nwm_uds.cpp
@@ -15,6 +15,7 @@
 #include "core/hle/result.h"
 #include "core/hle/service/nwm/nwm_uds.h"
 #include "core/hle/service/nwm/uds_beacon.h"
+#include "core/hle/service/nwm/uds_data.h"
 #include "core/memory.h"
 
 namespace Service {
@@ -190,7 +191,7 @@ static void InitializeWithVersion(Interface* self) {
 
     IPC::RequestBuilder rb = rp.MakeBuilder(1, 2);
     rb.Push(RESULT_SUCCESS);
-    rb.PushCopyHandles(Kernel::g_handle_table.Create(connection_status_event).MoveFrom());
+    rb.PushCopyHandles(Kernel::g_handle_table.Create(connection_status_event).Unwrap());
 
     LOG_DEBUG(Service_NWM, "called sharedmem_size=0x%08X, version=0x%08X, sharedmem_handle=0x%08X",
               sharedmem_size, version, sharedmem_handle);
@@ -265,7 +266,7 @@ static void Bind(Interface* self) {
     IPC::RequestBuilder rb = rp.MakeBuilder(1, 2);
 
     rb.Push(RESULT_SUCCESS);
-    rb.PushCopyHandles(Kernel::g_handle_table.Create(event).MoveFrom());
+    rb.PushCopyHandles(Kernel::g_handle_table.Create(event).Unwrap());
 }
 
 /**
@@ -373,6 +374,80 @@ static void DestroyNetwork(Interface* self) {
 }
 
 /**
+ * NWM_UDS::SendTo service function.
+ * Sends a data frame to the UDS network we're connected to.
+ *  Inputs:
+ *      0 : Command header.
+ *      1 : Unknown.
+ *      2 : u16 Destination network node id.
+ *      3 : u8 Data channel.
+ *      4 : Buffer size >> 2
+ *      5 : Data size
+ *      6 : Flags
+ *      7 : Input buffer descriptor
+ *      8 : Input buffer address
+ *  Outputs:
+ *      0 : Return header
+ *      1 : Result of function, 0 on success, otherwise error code
+ */
+static void SendTo(Interface* self) {
+    IPC::RequestParser rp(Kernel::GetCommandBuffer(), 0x17, 6, 2);
+
+    rp.Skip(1, false);
+    u16 dest_node_id = rp.Pop<u16>();
+    u8 data_channel = rp.Pop<u8>();
+    rp.Skip(1, false);
+    u32 data_size = rp.Pop<u32>();
+    u32 flags = rp.Pop<u32>();
+
+    size_t desc_size;
+    const VAddr input_address = rp.PopStaticBuffer(&desc_size, false);
+    ASSERT(desc_size == data_size);
+
+    IPC::RequestBuilder rb = rp.MakeBuilder(1, 0);
+
+    if (connection_status.status != static_cast<u32>(NetworkStatus::ConnectedAsClient) &&
+        connection_status.status != static_cast<u32>(NetworkStatus::ConnectedAsHost)) {
+        rb.Push(ResultCode(ErrorDescription::NotAuthorized, ErrorModule::UDS,
+                           ErrorSummary::InvalidState, ErrorLevel::Status));
+        return;
+    }
+
+    if (dest_node_id == connection_status.network_node_id) {
+        rb.Push(ResultCode(ErrorDescription::NotFound, ErrorModule::UDS,
+                           ErrorSummary::WrongArgument, ErrorLevel::Status));
+        return;
+    }
+
+    // TODO(Subv): Do something with the flags.
+
+    constexpr size_t MaxSize = 0x5C6;
+    if (data_size > MaxSize) {
+        rb.Push(ResultCode(ErrorDescription::TooLarge, ErrorModule::UDS,
+                           ErrorSummary::WrongArgument, ErrorLevel::Usage));
+        return;
+    }
+
+    std::vector<u8> data(data_size);
+    Memory::ReadBlock(input_address, data.data(), data.size());
+
+    // TODO(Subv): Increment the sequence number after each sent packet.
+    u16 sequence_number = 0;
+    std::vector<u8> data_payload = GenerateDataPayload(
+        data, data_channel, dest_node_id, connection_status.network_node_id, sequence_number);
+
+    // TODO(Subv): Retrieve the MAC address of the dest_node_id and our own to encrypt
+    // and encapsulate the payload.
+
+    // TODO(Subv): Send the frame.
+
+    rb.Push(RESULT_SUCCESS);
+
+    LOG_WARNING(Service_NWM, "(STUB) called dest_node_id=%u size=%u flags=%u channel=%u",
+                static_cast<u32>(dest_node_id), data_size, flags, static_cast<u32>(data_channel));
+}
+
+/**
  * NWM_UDS::GetChannel service function.
  * Returns the WiFi channel in which the network we're connected to is transmitting.
  *  Inputs:
@@ -543,6 +618,42 @@ static void BeaconBroadcastCallback(u64 userdata, int cycles_late) {
                               beacon_broadcast_event, 0);
 }
 
+/*
+ * Returns an available index in the nodes array for the
+ * currently-hosted UDS network.
+ */
+static u32 GetNextAvailableNodeId() {
+    ASSERT_MSG(connection_status.status == static_cast<u32>(NetworkStatus::ConnectedAsHost),
+               "Can not accept clients if we're not hosting a network");
+
+    for (unsigned index = 0; index < connection_status.max_nodes; ++index) {
+        if ((connection_status.node_bitmask & (1 << index)) == 0)
+            return index;
+    }
+
+    // Any connection attempts to an already full network should have been refused.
+    ASSERT_MSG(false, "No available connection slots in the network");
+}
+
+/*
+ * Called when a client connects to an UDS network we're hosting,
+ * updates the connection status and signals the update event.
+ * @param network_node_id Network Node Id of the connecting client.
+ */
+void OnClientConnected(u16 network_node_id) {
+    ASSERT_MSG(connection_status.status == static_cast<u32>(NetworkStatus::ConnectedAsHost),
+               "Can not accept clients if we're not hosting a network");
+    ASSERT_MSG(connection_status.total_nodes < connection_status.max_nodes,
+               "Can not accept connections on a full network");
+
+    u32 node_id = GetNextAvailableNodeId();
+    connection_status.node_bitmask |= 1 << node_id;
+    connection_status.changed_nodes |= 1 << node_id;
+    connection_status.nodes[node_id] = network_node_id;
+    connection_status.total_nodes++;
+    connection_status_event->Signal();
+}
+
 const Interface::FunctionInfo FunctionTable[] = {
     {0x00010442, nullptr, "Initialize (deprecated)"},
     {0x00020000, nullptr, "Scrap"},
@@ -564,7 +675,7 @@ const Interface::FunctionInfo FunctionTable[] = {
     {0x00130040, nullptr, "Unbind"},
     {0x001400C0, nullptr, "PullPacket"},
     {0x00150080, nullptr, "SetMaxSendDelay"},
-    {0x00170182, nullptr, "SendTo"},
+    {0x00170182, SendTo, "SendTo"},
     {0x001A0000, GetChannel, "GetChannel"},
     {0x001B0302, InitializeWithVersion, "InitializeWithVersion"},
     {0x001D0044, BeginHostingNetwork, "BeginHostingNetwork"},
diff --git a/src/core/hle/service/nwm/uds_data.cpp b/src/core/hle/service/nwm/uds_data.cpp
new file mode 100644
index 000000000..8c6742dba
--- /dev/null
+++ b/src/core/hle/service/nwm/uds_data.cpp
@@ -0,0 +1,278 @@
+// Copyright 2017 Citra Emulator Project
+// Licensed under GPLv2 or any later version
+// Refer to the license.txt file included.
+
+#include <cstring>
+#include <cryptopp/aes.h>
+#include <cryptopp/ccm.h>
+#include <cryptopp/filters.h>
+#include <cryptopp/md5.h>
+#include <cryptopp/modes.h>
+#include "core/hle/service/nwm/nwm_uds.h"
+#include "core/hle/service/nwm/uds_data.h"
+#include "core/hw/aes/key.h"
+
+namespace Service {
+namespace NWM {
+
+using MacAddress = std::array<u8, 6>;
+
+/*
+ * Generates a SNAP-enabled 802.2 LLC header for the specified protocol.
+ * @returns a buffer with the bytes of the generated header.
+ */
+static std::vector<u8> GenerateLLCHeader(EtherType protocol) {
+    LLCHeader header{};
+    header.protocol = static_cast<u16>(protocol);
+
+    std::vector<u8> buffer(sizeof(header));
+    memcpy(buffer.data(), &header, sizeof(header));
+
+    return buffer;
+}
+
+/*
+ * Generates a Nintendo UDS SecureData header with the specified parameters.
+ * @returns a buffer with the bytes of the generated header.
+ */
+static std::vector<u8> GenerateSecureDataHeader(u16 data_size, u8 channel, u16 dest_node_id,
+                                                u16 src_node_id, u16 sequence_number) {
+    SecureDataHeader header{};
+    header.protocol_size = data_size + sizeof(SecureDataHeader);
+    // Note: This size includes everything except the first 4 bytes of the structure,
+    // reinforcing the hypotheses that the first 4 bytes are actually the header of
+    // another container protocol.
+    header.securedata_size = data_size + sizeof(SecureDataHeader) - 4;
+    // Frames sent by the emulated application are never UDS management frames
+    header.is_management = 0;
+    header.data_channel = channel;
+    header.sequence_number = sequence_number;
+    header.dest_node_id = dest_node_id;
+    header.src_node_id = src_node_id;
+
+    std::vector<u8> buffer(sizeof(header));
+    memcpy(buffer.data(), &header, sizeof(header));
+
+    return buffer;
+}
+
+/*
+ * Calculates the CTR used for the AES-CTR process that calculates
+ * the CCMP crypto key for data frames.
+ * @returns The CTR used for data frames crypto key generation.
+ */
+static std::array<u8, CryptoPP::MD5::DIGESTSIZE> GetDataCryptoCTR(const NetworkInfo& network_info) {
+    DataFrameCryptoCTR data{};
+
+    data.host_mac = network_info.host_mac_address;
+    data.wlan_comm_id = network_info.wlan_comm_id;
+    data.id = network_info.id;
+    data.network_id = network_info.network_id;
+
+    std::array<u8, CryptoPP::MD5::DIGESTSIZE> hash;
+    CryptoPP::MD5().CalculateDigest(hash.data(), reinterpret_cast<u8*>(&data), sizeof(data));
+
+    return hash;
+}
+
+/*
+ * Generates the key used for encrypting the 802.11 data frames generated by UDS.
+ * @returns The key used for data frames crypto.
+ */
+static std::array<u8, CryptoPP::AES::BLOCKSIZE> GenerateDataCCMPKey(
+    const std::vector<u8>& passphrase, const NetworkInfo& network_info) {
+    // Calculate the MD5 hash of the input passphrase.
+    std::array<u8, CryptoPP::MD5::DIGESTSIZE> passphrase_hash;
+    CryptoPP::MD5().CalculateDigest(passphrase_hash.data(), passphrase.data(), passphrase.size());
+
+    std::array<u8, CryptoPP::AES::BLOCKSIZE> ccmp_key;
+
+    // The CCMP key is the result of encrypting the MD5 hash of the passphrase with AES-CTR using
+    // keyslot 0x2D.
+    using CryptoPP::AES;
+    std::array<u8, CryptoPP::MD5::DIGESTSIZE> counter = GetDataCryptoCTR(network_info);
+    std::array<u8, AES::BLOCKSIZE> key = HW::AES::GetNormalKey(HW::AES::KeySlotID::UDSDataKey);
+    CryptoPP::CTR_Mode<AES>::Encryption aes;
+    aes.SetKeyWithIV(key.data(), AES::BLOCKSIZE, counter.data());
+    aes.ProcessData(ccmp_key.data(), passphrase_hash.data(), passphrase_hash.size());
+
+    return ccmp_key;
+}
+
+/*
+ * Generates the Additional Authenticated Data (AAD) for an UDS 802.11 encrypted data frame.
+ * @returns a buffer with the bytes of the AAD.
+ */
+static std::vector<u8> GenerateCCMPAAD(const MacAddress& sender, const MacAddress& receiver,
+                                       const MacAddress& bssid, u16 frame_control) {
+    // Reference: IEEE 802.11-2007
+
+    // 8.3.3.3.2 Construct AAD (22-30 bytes)
+    // The AAD is constructed from the MPDU header. The AAD does not include the header Duration
+    // field, because the Duration field value can change due to normal IEEE 802.11 operation (e.g.,
+    // a rate change during retransmission). For similar reasons, several subfields in the Frame
+    // Control field are masked to 0.
+    struct {
+        u16_be FC; // MPDU Frame Control field
+        MacAddress A1;
+        MacAddress A2;
+        MacAddress A3;
+        u16_be SC; // MPDU Sequence Control field
+    } aad_struct{};
+
+    constexpr u16 AADFrameControlMask = 0x8FC7;
+    aad_struct.FC = frame_control & AADFrameControlMask;
+    aad_struct.SC = 0;
+
+    bool to_ds = (frame_control & (1 << 0)) != 0;
+    bool from_ds = (frame_control & (1 << 1)) != 0;
+    // In the 802.11 standard, ToDS = 1 and FromDS = 1 is a valid configuration,
+    // however, the 3DS doesn't seem to transmit frames with such combination.
+    ASSERT_MSG(to_ds != from_ds, "Invalid combination");
+
+    // The meaning of the address fields depends on the ToDS and FromDS fields.
+    if (from_ds) {
+        aad_struct.A1 = receiver;
+        aad_struct.A2 = bssid;
+        aad_struct.A3 = sender;
+    }
+
+    if (to_ds) {
+        aad_struct.A1 = bssid;
+        aad_struct.A2 = sender;
+        aad_struct.A3 = receiver;
+    }
+
+    std::vector<u8> aad(sizeof(aad_struct));
+    std::memcpy(aad.data(), &aad_struct, sizeof(aad_struct));
+
+    return aad;
+}
+
+/*
+ * Decrypts the payload of an encrypted 802.11 data frame using the specified key.
+ * @returns The decrypted payload.
+ */
+static std::vector<u8> DecryptDataFrame(const std::vector<u8>& encrypted_payload,
+                                        const std::array<u8, CryptoPP::AES::BLOCKSIZE>& ccmp_key,
+                                        const MacAddress& sender, const MacAddress& receiver,
+                                        const MacAddress& bssid, u16 sequence_number,
+                                        u16 frame_control) {
+
+    // Reference: IEEE 802.11-2007
+
+    std::vector<u8> aad = GenerateCCMPAAD(sender, receiver, bssid, frame_control);
+
+    std::vector<u8> packet_number{0,
+                                  0,
+                                  0,
+                                  0,
+                                  static_cast<u8>((sequence_number >> 8) & 0xFF),
+                                  static_cast<u8>(sequence_number & 0xFF)};
+
+    // 8.3.3.3.3 Construct CCM nonce (13 bytes)
+    std::vector<u8> nonce;
+    nonce.push_back(0);                                                    // priority
+    nonce.insert(nonce.end(), sender.begin(), sender.end());               // Address 2
+    nonce.insert(nonce.end(), packet_number.begin(), packet_number.end()); // PN
+
+    try {
+        CryptoPP::CCM<CryptoPP::AES, 8>::Decryption d;
+        d.SetKeyWithIV(ccmp_key.data(), ccmp_key.size(), nonce.data(), nonce.size());
+        d.SpecifyDataLengths(aad.size(), encrypted_payload.size() - 8, 0);
+
+        CryptoPP::AuthenticatedDecryptionFilter df(
+            d, nullptr, CryptoPP::AuthenticatedDecryptionFilter::MAC_AT_END |
+                            CryptoPP::AuthenticatedDecryptionFilter::THROW_EXCEPTION);
+        // put aad
+        df.ChannelPut(CryptoPP::AAD_CHANNEL, aad.data(), aad.size());
+
+        // put cipher with mac
+        df.ChannelPut(CryptoPP::DEFAULT_CHANNEL, encrypted_payload.data(),
+                      encrypted_payload.size() - 8);
+        df.ChannelPut(CryptoPP::DEFAULT_CHANNEL,
+                      encrypted_payload.data() + encrypted_payload.size() - 8, 8);
+
+        df.ChannelMessageEnd(CryptoPP::AAD_CHANNEL);
+        df.ChannelMessageEnd(CryptoPP::DEFAULT_CHANNEL);
+        df.SetRetrievalChannel(CryptoPP::DEFAULT_CHANNEL);
+
+        int size = df.MaxRetrievable();
+
+        std::vector<u8> pdata(size);
+        df.Get(pdata.data(), size);
+        return pdata;
+    } catch (CryptoPP::Exception&) {
+        LOG_ERROR(Service_NWM, "failed to decrypt");
+    }
+
+    return {};
+}
+
+/*
+ * Encrypts the payload of an 802.11 data frame using the specified key.
+ * @returns The encrypted payload.
+ */
+static std::vector<u8> EncryptDataFrame(const std::vector<u8>& payload,
+                                        const std::array<u8, CryptoPP::AES::BLOCKSIZE>& ccmp_key,
+                                        const MacAddress& sender, const MacAddress& receiver,
+                                        const MacAddress& bssid, u16 sequence_number,
+                                        u16 frame_control) {
+    // Reference: IEEE 802.11-2007
+
+    std::vector<u8> aad = GenerateCCMPAAD(sender, receiver, bssid, frame_control);
+
+    std::vector<u8> packet_number{0,
+                                  0,
+                                  0,
+                                  0,
+                                  static_cast<u8>((sequence_number >> 8) & 0xFF),
+                                  static_cast<u8>(sequence_number & 0xFF)};
+
+    // 8.3.3.3.3 Construct CCM nonce (13 bytes)
+    std::vector<u8> nonce;
+    nonce.push_back(0);                                                    // priority
+    nonce.insert(nonce.end(), sender.begin(), sender.end());               // Address 2
+    nonce.insert(nonce.end(), packet_number.begin(), packet_number.end()); // PN
+
+    try {
+        CryptoPP::CCM<CryptoPP::AES, 8>::Encryption d;
+        d.SetKeyWithIV(ccmp_key.data(), ccmp_key.size(), nonce.data(), nonce.size());
+        d.SpecifyDataLengths(aad.size(), payload.size(), 0);
+
+        CryptoPP::AuthenticatedEncryptionFilter df(d);
+        // put aad
+        df.ChannelPut(CryptoPP::AAD_CHANNEL, aad.data(), aad.size());
+        df.ChannelMessageEnd(CryptoPP::AAD_CHANNEL);
+
+        // put plaintext
+        df.ChannelPut(CryptoPP::DEFAULT_CHANNEL, payload.data(), payload.size());
+        df.ChannelMessageEnd(CryptoPP::DEFAULT_CHANNEL);
+
+        df.SetRetrievalChannel(CryptoPP::DEFAULT_CHANNEL);
+
+        int size = df.MaxRetrievable();
+
+        std::vector<u8> cipher(size);
+        df.Get(cipher.data(), size);
+        return cipher;
+    } catch (CryptoPP::Exception&) {
+        LOG_ERROR(Service_NWM, "failed to encrypt");
+    }
+
+    return {};
+}
+
+std::vector<u8> GenerateDataPayload(const std::vector<u8>& data, u8 channel, u16 dest_node,
+                                    u16 src_node, u16 sequence_number) {
+    std::vector<u8> buffer = GenerateLLCHeader(EtherType::SecureData);
+    std::vector<u8> securedata_header =
+        GenerateSecureDataHeader(data.size(), channel, dest_node, src_node, sequence_number);
+
+    buffer.insert(buffer.end(), securedata_header.begin(), securedata_header.end());
+    buffer.insert(buffer.end(), data.begin(), data.end());
+    return buffer;
+}
+
+} // namespace NWM
+} // namespace Service
diff --git a/src/core/hle/service/nwm/uds_data.h b/src/core/hle/service/nwm/uds_data.h
new file mode 100644
index 000000000..a23520a41
--- /dev/null
+++ b/src/core/hle/service/nwm/uds_data.h
@@ -0,0 +1,78 @@
+// Copyright 2017 Citra Emulator Project
+// Licensed under GPLv2 or any later version
+// Refer to the license.txt file included.
+
+#pragma once
+
+#include <array>
+#include <vector>
+#include "common/common_types.h"
+#include "common/swap.h"
+#include "core/hle/service/service.h"
+
+namespace Service {
+namespace NWM {
+
+enum class SAP : u8 { SNAPExtensionUsed = 0xAA };
+
+enum class PDUControl : u8 { UnnumberedInformation = 3 };
+
+enum class EtherType : u16 { SecureData = 0x876D, EAPoL = 0x888E };
+
+/*
+ * 802.2 header, UDS packets always use SNAP for these headers,
+ * which means the dsap and ssap are always SNAPExtensionUsed (0xAA)
+ * and the OUI is always 0.
+ */
+struct LLCHeader {
+    u8 dsap = static_cast<u8>(SAP::SNAPExtensionUsed);
+    u8 ssap = static_cast<u8>(SAP::SNAPExtensionUsed);
+    u8 control = static_cast<u8>(PDUControl::UnnumberedInformation);
+    std::array<u8, 3> OUI = {};
+    u16_be protocol;
+};
+
+static_assert(sizeof(LLCHeader) == 8, "LLCHeader has the wrong size");
+
+/*
+ * Nintendo SecureData header, every UDS packet contains one,
+ * it is used to store metadata about the transmission such as
+ * the source and destination network node ids.
+ */
+struct SecureDataHeader {
+    // TODO(Subv): It is likely that the first 4 bytes of this header are
+    // actually part of another container protocol.
+    u16_be protocol_size;
+    INSERT_PADDING_BYTES(2);
+    u16_be securedata_size;
+    u8 is_management;
+    u8 data_channel;
+    u16_be sequence_number;
+    u16_be dest_node_id;
+    u16_be src_node_id;
+};
+
+static_assert(sizeof(SecureDataHeader) == 14, "SecureDataHeader has the wrong size");
+
+/*
+ * The raw bytes of this structure are the CTR used in the encryption (AES-CTR)
+ * process used to generate the CCMP key for data frame encryption.
+ */
+struct DataFrameCryptoCTR {
+    u32_le wlan_comm_id;
+    u32_le network_id;
+    std::array<u8, 6> host_mac;
+    u16_le id;
+};
+
+static_assert(sizeof(DataFrameCryptoCTR) == 16, "DataFrameCryptoCTR has the wrong size");
+
+/**
+ * Generates an unencrypted 802.11 data payload.
+ * @returns The generated frame payload.
+ */
+std::vector<u8> GenerateDataPayload(const std::vector<u8>& data, u8 channel, u16 dest_node,
+                                    u16 src_node, u16 sequence_number);
+
+} // namespace NWM
+} // namespace Service
diff --git a/src/core/hle/service/ptm/ptm.cpp b/src/core/hle/service/ptm/ptm.cpp
index 39382ef09..a0b959797 100644
--- a/src/core/hle/service/ptm/ptm.cpp
+++ b/src/core/hle/service/ptm/ptm.cpp
@@ -152,7 +152,7 @@ void Init() {
         auto gamecoin_result =
             Service::FS::OpenFileFromArchive(*archive_result, gamecoin_path, open_mode);
         if (gamecoin_result.Succeeded()) {
-            auto gamecoin = gamecoin_result.MoveFrom();
+            auto gamecoin = std::move(gamecoin_result).Unwrap();
             gamecoin->backend->Write(0, sizeof(GameCoin), true,
                                      reinterpret_cast<const u8*>(&default_game_coin));
             gamecoin->backend->Close();
diff --git a/src/core/hle/service/service.cpp b/src/core/hle/service/service.cpp
index 791a65c19..aad950e50 100644
--- a/src/core/hle/service/service.cpp
+++ b/src/core/hle/service/service.cpp
@@ -173,8 +173,7 @@ void ServiceFrameworkBase::HandleSyncRequest(SharedPtr<ServerSession> server_ses
 
     // TODO(yuriks): The kernel should be the one handling this as part of translation after
     // everything else is migrated
-    Kernel::HLERequestContext context;
-    context.session = std::move(server_session);
+    Kernel::HLERequestContext context(std::move(server_session));
     context.PopulateFromIncomingCommandBuffer(cmd_buf, *Kernel::g_current_process,
                                               Kernel::g_handle_table);
 
@@ -207,7 +206,7 @@ void AddService(Interface* interface_) {
     auto server_port =
         SM::g_service_manager
             ->RegisterService(interface_->GetPortName(), interface_->GetMaxSessions())
-            .MoveFrom();
+            .Unwrap();
     server_port->SetHleHandler(std::shared_ptr<Interface>(interface_));
 }
 
diff --git a/src/core/hle/service/sm/srv.cpp b/src/core/hle/service/sm/srv.cpp
index 74a1256e0..352941e69 100644
--- a/src/core/hle/service/sm/srv.cpp
+++ b/src/core/hle/service/sm/srv.cpp
@@ -113,13 +113,13 @@ void SRV::GetServiceHandle(Kernel::HLERequestContext& ctx) {
                   (*session)->GetObjectId());
         IPC::RequestBuilder rb = rp.MakeBuilder(1, 2);
         rb.Push(session.Code());
-        rb.PushObjects(session.MoveFrom());
+        rb.PushObjects(std::move(session).Unwrap());
     } else if (session.Code() == Kernel::ERR_MAX_CONNECTIONS_REACHED && return_port_on_failure) {
         LOG_WARNING(Service_SRV, "called service=%s -> ERR_MAX_CONNECTIONS_REACHED, *port*=%u",
                     name.c_str(), (*client_port)->GetObjectId());
         IPC::RequestBuilder rb = rp.MakeBuilder(1, 2);
         rb.Push(ERR_MAX_CONNECTIONS_REACHED);
-        rb.PushObjects(client_port.MoveFrom());
+        rb.PushObjects(std::move(client_port).Unwrap());
     } else {
         LOG_ERROR(Service_SRV, "called service=%s -> error 0x%08X", name.c_str(), session.Code());
         IPC::RequestBuilder rb = rp.MakeBuilder(1, 0);
diff --git a/src/core/hle/service/y2r_u.cpp b/src/core/hle/service/y2r_u.cpp
index bb7bf2d67..e73971d5f 100644
--- a/src/core/hle/service/y2r_u.cpp
+++ b/src/core/hle/service/y2r_u.cpp
@@ -275,7 +275,7 @@ static void GetTransferEndEvent(Interface* self) {
 
     cmd_buff[0] = IPC::MakeHeader(0xF, 2, 0);
     cmd_buff[1] = RESULT_SUCCESS.raw;
-    cmd_buff[3] = Kernel::g_handle_table.Create(completion_event).MoveFrom();
+    cmd_buff[3] = Kernel::g_handle_table.Create(completion_event).Unwrap();
 
     LOG_DEBUG(Service_Y2R, "called");
 }
diff --git a/src/core/hle/svc.cpp b/src/core/hle/svc.cpp
index 0a7f011f3..e4b803046 100644
--- a/src/core/hle/svc.cpp
+++ b/src/core/hle/svc.cpp
@@ -37,8 +37,9 @@
 ////////////////////////////////////////////////////////////////////////////////////////////////////
 // Namespace SVC
 
-using Kernel::SharedPtr;
 using Kernel::ERR_INVALID_HANDLE;
+using Kernel::Handle;
+using Kernel::SharedPtr;
 
 namespace SVC {
 
@@ -1040,7 +1041,6 @@ static ResultCode CreatePort(Kernel::Handle* server_port, Kernel::Handle* client
 
     using Kernel::ServerPort;
     using Kernel::ClientPort;
-    using Kernel::SharedPtr;
 
     auto ports = ServerPort::CreatePortPair(max_sessions);
     CASCADE_RESULT(*client_port, Kernel::g_handle_table.Create(
@@ -1054,6 +1054,41 @@ static ResultCode CreatePort(Kernel::Handle* server_port, Kernel::Handle* client
     return RESULT_SUCCESS;
 }
 
+static ResultCode CreateSessionToPort(Handle* out_client_session, Handle client_port_handle) {
+    using Kernel::ClientPort;
+    SharedPtr<ClientPort> client_port = Kernel::g_handle_table.Get<ClientPort>(client_port_handle);
+    if (client_port == nullptr)
+        return ERR_INVALID_HANDLE;
+
+    CASCADE_RESULT(auto session, client_port->Connect());
+    CASCADE_RESULT(*out_client_session, Kernel::g_handle_table.Create(std::move(session)));
+    return RESULT_SUCCESS;
+}
+
+static ResultCode CreateSession(Handle* server_session, Handle* client_session) {
+    auto sessions = Kernel::ServerSession::CreateSessionPair();
+
+    auto& server = std::get<SharedPtr<Kernel::ServerSession>>(sessions);
+    CASCADE_RESULT(*server_session, Kernel::g_handle_table.Create(std::move(server)));
+
+    auto& client = std::get<SharedPtr<Kernel::ClientSession>>(sessions);
+    CASCADE_RESULT(*client_session, Kernel::g_handle_table.Create(std::move(client)));
+
+    LOG_TRACE(Kernel_SVC, "called");
+    return RESULT_SUCCESS;
+}
+
+static ResultCode AcceptSession(Handle* out_server_session, Handle server_port_handle) {
+    using Kernel::ServerPort;
+    SharedPtr<ServerPort> server_port = Kernel::g_handle_table.Get<ServerPort>(server_port_handle);
+    if (server_port == nullptr)
+        return ERR_INVALID_HANDLE;
+
+    CASCADE_RESULT(auto session, server_port->Accept());
+    CASCADE_RESULT(*out_server_session, Kernel::g_handle_table.Create(std::move(session)));
+    return RESULT_SUCCESS;
+}
+
 static ResultCode GetSystemInfo(s64* out, u32 type, s32 param) {
     using Kernel::MemoryRegion;
 
@@ -1228,9 +1263,9 @@ static const FunctionDef SVC_Table[] = {
     {0x45, nullptr, "Unknown"},
     {0x46, nullptr, "Unknown"},
     {0x47, HLE::Wrap<CreatePort>, "CreatePort"},
-    {0x48, nullptr, "CreateSessionToPort"},
-    {0x49, nullptr, "CreateSession"},
-    {0x4A, nullptr, "AcceptSession"},
+    {0x48, HLE::Wrap<CreateSessionToPort>, "CreateSessionToPort"},
+    {0x49, HLE::Wrap<CreateSession>, "CreateSession"},
+    {0x4A, HLE::Wrap<AcceptSession>, "AcceptSession"},
     {0x4B, nullptr, "ReplyAndReceive1"},
     {0x4C, nullptr, "ReplyAndReceive2"},
     {0x4D, nullptr, "ReplyAndReceive3"},
diff --git a/src/core/hw/aes/key.h b/src/core/hw/aes/key.h
index b01d04f13..c9f1342f4 100644
--- a/src/core/hw/aes/key.h
+++ b/src/core/hw/aes/key.h
@@ -12,6 +12,8 @@ namespace HW {
 namespace AES {
 
 enum KeySlotID : size_t {
+    // AES Keyslot used to generate the UDS data frame CCMP key.
+    UDSDataKey = 0x2D,
     APTWrap = 0x31,
 
     MaxKeySlotID = 0x40,
diff --git a/src/core/hw/gpu.cpp b/src/core/hw/gpu.cpp
index 42809c731..6838e449c 100644
--- a/src/core/hw/gpu.cpp
+++ b/src/core/hw/gpu.cpp
@@ -5,6 +5,7 @@
 #include <cstring>
 #include <numeric>
 #include <type_traits>
+#include "common/alignment.h"
 #include "common/color.h"
 #include "common/common_types.h"
 #include "common/logging/log.h"
@@ -313,7 +314,7 @@ static void TextureCopy(const Regs::DisplayTransferConfig& config) {
     const PAddr src_addr = config.GetPhysicalInputAddress();
     const PAddr dst_addr = config.GetPhysicalOutputAddress();
 
-    // TODO: do hwtest with these cases
+    // TODO: do hwtest with invalid addresses
     if (!Memory::IsValidPhysicalAddress(src_addr)) {
         LOG_CRITICAL(HW_GPU, "invalid input address 0x%08X", src_addr);
         return;
@@ -324,31 +325,36 @@ static void TextureCopy(const Regs::DisplayTransferConfig& config) {
         return;
     }
 
-    if (config.texture_copy.input_width == 0) {
-        LOG_CRITICAL(HW_GPU, "zero input width");
+    if (VideoCore::g_renderer->Rasterizer()->AccelerateTextureCopy(config))
         return;
-    }
 
-    if (config.texture_copy.output_width == 0) {
-        LOG_CRITICAL(HW_GPU, "zero output width");
+    u8* src_pointer = Memory::GetPhysicalPointer(src_addr);
+    u8* dst_pointer = Memory::GetPhysicalPointer(dst_addr);
+
+    u32 remaining_size = Common::AlignDown(config.texture_copy.size, 16);
+
+    if (remaining_size == 0) {
+        LOG_CRITICAL(HW_GPU, "zero size. Real hardware freezes on this.");
         return;
     }
 
-    if (config.texture_copy.size == 0) {
-        LOG_CRITICAL(HW_GPU, "zero size");
+    u32 input_gap = config.texture_copy.input_gap * 16;
+    u32 output_gap = config.texture_copy.output_gap * 16;
+
+    // Zero gap means contiguous input/output even if width = 0. To avoid infinite loop below, width
+    // is assigned with the total size if gap = 0.
+    u32 input_width = input_gap == 0 ? remaining_size : config.texture_copy.input_width * 16;
+    u32 output_width = output_gap == 0 ? remaining_size : config.texture_copy.output_width * 16;
+
+    if (input_width == 0) {
+        LOG_CRITICAL(HW_GPU, "zero input width. Real hardware freezes on this.");
         return;
     }
 
-    if (VideoCore::g_renderer->Rasterizer()->AccelerateTextureCopy(config))
+    if (output_width == 0) {
+        LOG_CRITICAL(HW_GPU, "zero output width. Real hardware freezes on this.");
         return;
-
-    u8* src_pointer = Memory::GetPhysicalPointer(src_addr);
-    u8* dst_pointer = Memory::GetPhysicalPointer(dst_addr);
-
-    u32 input_width = config.texture_copy.input_width * 16;
-    u32 input_gap = config.texture_copy.input_gap * 16;
-    u32 output_width = config.texture_copy.output_width * 16;
-    u32 output_gap = config.texture_copy.output_gap * 16;
+    }
 
     size_t contiguous_input_size =
         config.texture_copy.size / input_width * (input_width + input_gap);
@@ -360,7 +366,6 @@ static void TextureCopy(const Regs::DisplayTransferConfig& config) {
     Memory::RasterizerFlushAndInvalidateRegion(config.GetPhysicalOutputAddress(),
                                                static_cast<u32>(contiguous_output_size));
 
-    u32 remaining_size = config.texture_copy.size;
     u32 remaining_input = input_width;
     u32 remaining_output = output_width;
     while (remaining_size > 0) {
diff --git a/src/core/hw/gpu.h b/src/core/hw/gpu.h
index bdd997b2a..21b127fee 100644
--- a/src/core/hw/gpu.h
+++ b/src/core/hw/gpu.h
@@ -225,7 +225,7 @@ struct Regs {
         INSERT_PADDING_WORDS(0x1);
 
         struct {
-            u32 size;
+            u32 size; // The lower 4 bits are ignored
 
             union {
                 u32 input_size;
diff --git a/src/core/memory.h b/src/core/memory.h
index 802aa465e..71fb278ad 100644
--- a/src/core/memory.h
+++ b/src/core/memory.h
@@ -55,8 +55,10 @@ enum : PAddr {
 
     /// Main FCRAM
     FCRAM_PADDR = 0x20000000,
-    FCRAM_SIZE = 0x08000000, ///< FCRAM size (128MB)
+    FCRAM_SIZE = 0x08000000,      ///< FCRAM size on the Old 3DS (128MB)
+    FCRAM_N3DS_SIZE = 0x10000000, ///< FCRAM size on the New 3DS (256MB)
     FCRAM_PADDR_END = FCRAM_PADDR + FCRAM_SIZE,
+    FCRAM_N3DS_PADDR_END = FCRAM_PADDR + FCRAM_N3DS_SIZE,
 };
 
 /// Virtual user-space memory regions
diff --git a/src/tests/CMakeLists.txt b/src/tests/CMakeLists.txt
index 00d7c636a..a14df325a 100644
--- a/src/tests/CMakeLists.txt
+++ b/src/tests/CMakeLists.txt
@@ -1,8 +1,9 @@
 set(SRCS
-            glad.cpp
-            tests.cpp
             common/param_package.cpp
             core/file_sys/path_parser.cpp
+            core/hle/kernel/hle_ipc.cpp
+            glad.cpp
+            tests.cpp
             )
 
 set(HEADERS
diff --git a/src/tests/core/hle/kernel/hle_ipc.cpp b/src/tests/core/hle/kernel/hle_ipc.cpp
new file mode 100644
index 000000000..52336d027
--- /dev/null
+++ b/src/tests/core/hle/kernel/hle_ipc.cpp
@@ -0,0 +1,216 @@
+// Copyright 2017 Citra Emulator Project
+// Licensed under GPLv2 or any later version
+// Refer to the license.txt file included.
+
+#include <catch.hpp>
+#include "core/hle/ipc.h"
+#include "core/hle/kernel/client_port.h"
+#include "core/hle/kernel/client_session.h"
+#include "core/hle/kernel/event.h"
+#include "core/hle/kernel/handle_table.h"
+#include "core/hle/kernel/hle_ipc.h"
+#include "core/hle/kernel/process.h"
+#include "core/hle/kernel/server_session.h"
+
+namespace Kernel {
+
+static SharedPtr<Object> MakeObject() {
+    return Event::Create(ResetType::OneShot);
+}
+
+TEST_CASE("HLERequestContext::PopulateFromIncomingCommandBuffer", "[core][kernel]") {
+    auto session = std::get<SharedPtr<ServerSession>>(ServerSession::CreateSessionPair());
+    HLERequestContext context(std::move(session));
+
+    auto process = Process::Create(CodeSet::Create("", 0));
+    HandleTable handle_table;
+
+    SECTION("works with empty cmdbuf") {
+        const u32_le input[]{
+            IPC::MakeHeader(0x1234, 0, 0),
+        };
+
+        context.PopulateFromIncomingCommandBuffer(input, *process, handle_table);
+
+        REQUIRE(context.CommandBuffer()[0] == 0x12340000);
+    }
+
+    SECTION("translates regular params") {
+        const u32_le input[]{
+            IPC::MakeHeader(0, 3, 0), 0x12345678, 0x21122112, 0xAABBCCDD,
+        };
+
+        context.PopulateFromIncomingCommandBuffer(input, *process, handle_table);
+
+        auto* output = context.CommandBuffer();
+        REQUIRE(output[1] == 0x12345678);
+        REQUIRE(output[2] == 0x21122112);
+        REQUIRE(output[3] == 0xAABBCCDD);
+    }
+
+    SECTION("translates move handles") {
+        auto a = MakeObject();
+        Handle a_handle = handle_table.Create(a).Unwrap();
+        const u32_le input[]{
+            IPC::MakeHeader(0, 0, 2), IPC::MoveHandleDesc(1), a_handle,
+        };
+
+        context.PopulateFromIncomingCommandBuffer(input, *process, handle_table);
+
+        auto* output = context.CommandBuffer();
+        REQUIRE(context.GetIncomingHandle(output[2]) == a);
+        REQUIRE(handle_table.GetGeneric(a_handle) == nullptr);
+    }
+
+    SECTION("translates copy handles") {
+        auto a = MakeObject();
+        Handle a_handle = handle_table.Create(a).Unwrap();
+        const u32_le input[]{
+            IPC::MakeHeader(0, 0, 2), IPC::CopyHandleDesc(1), a_handle,
+        };
+
+        context.PopulateFromIncomingCommandBuffer(input, *process, handle_table);
+
+        auto* output = context.CommandBuffer();
+        REQUIRE(context.GetIncomingHandle(output[2]) == a);
+        REQUIRE(handle_table.GetGeneric(a_handle) == a);
+    }
+
+    SECTION("translates multi-handle descriptors") {
+        auto a = MakeObject();
+        auto b = MakeObject();
+        auto c = MakeObject();
+        const u32_le input[]{
+            IPC::MakeHeader(0, 0, 5),        IPC::MoveHandleDesc(2),
+            handle_table.Create(a).Unwrap(), handle_table.Create(b).Unwrap(),
+            IPC::MoveHandleDesc(1),          handle_table.Create(c).Unwrap(),
+        };
+
+        context.PopulateFromIncomingCommandBuffer(input, *process, handle_table);
+
+        auto* output = context.CommandBuffer();
+        REQUIRE(context.GetIncomingHandle(output[2]) == a);
+        REQUIRE(context.GetIncomingHandle(output[3]) == b);
+        REQUIRE(context.GetIncomingHandle(output[5]) == c);
+    }
+
+    SECTION("translates null handles") {
+        const u32_le input[]{
+            IPC::MakeHeader(0, 0, 2), IPC::MoveHandleDesc(1), 0,
+        };
+
+        auto result = context.PopulateFromIncomingCommandBuffer(input, *process, handle_table);
+
+        REQUIRE(result == RESULT_SUCCESS);
+        auto* output = context.CommandBuffer();
+        REQUIRE(context.GetIncomingHandle(output[2]) == nullptr);
+    }
+
+    SECTION("translates CallingPid descriptors") {
+        const u32_le input[]{
+            IPC::MakeHeader(0, 0, 2), IPC::CallingPidDesc(), 0x98989898,
+        };
+
+        context.PopulateFromIncomingCommandBuffer(input, *process, handle_table);
+
+        REQUIRE(context.CommandBuffer()[2] == process->process_id);
+    }
+
+    SECTION("translates mixed params") {
+        auto a = MakeObject();
+        const u32_le input[]{
+            IPC::MakeHeader(0, 2, 4),
+            0x12345678,
+            0xABCDEF00,
+            IPC::MoveHandleDesc(1),
+            handle_table.Create(a).Unwrap(),
+            IPC::CallingPidDesc(),
+            0,
+        };
+
+        context.PopulateFromIncomingCommandBuffer(input, *process, handle_table);
+
+        auto* output = context.CommandBuffer();
+        REQUIRE(output[1] == 0x12345678);
+        REQUIRE(output[2] == 0xABCDEF00);
+        REQUIRE(context.GetIncomingHandle(output[4]) == a);
+        REQUIRE(output[6] == process->process_id);
+    }
+}
+
+TEST_CASE("HLERequestContext::WriteToOutgoingCommandBuffer", "[core][kernel]") {
+    auto session = std::get<SharedPtr<ServerSession>>(ServerSession::CreateSessionPair());
+    HLERequestContext context(std::move(session));
+
+    auto process = Process::Create(CodeSet::Create("", 0));
+    HandleTable handle_table;
+    auto* input = context.CommandBuffer();
+    u32_le output[IPC::COMMAND_BUFFER_LENGTH];
+
+    SECTION("works with empty cmdbuf") {
+        input[0] = IPC::MakeHeader(0x1234, 0, 0);
+
+        context.WriteToOutgoingCommandBuffer(output, *process, handle_table);
+
+        REQUIRE(output[0] == 0x12340000);
+    }
+
+    SECTION("translates regular params") {
+        input[0] = IPC::MakeHeader(0, 3, 0);
+        input[1] = 0x12345678;
+        input[2] = 0x21122112;
+        input[3] = 0xAABBCCDD;
+
+        context.WriteToOutgoingCommandBuffer(output, *process, handle_table);
+
+        REQUIRE(output[1] == 0x12345678);
+        REQUIRE(output[2] == 0x21122112);
+        REQUIRE(output[3] == 0xAABBCCDD);
+    }
+
+    SECTION("translates move/copy handles") {
+        auto a = MakeObject();
+        auto b = MakeObject();
+        input[0] = IPC::MakeHeader(0, 0, 4);
+        input[1] = IPC::MoveHandleDesc(1);
+        input[2] = context.AddOutgoingHandle(a);
+        input[3] = IPC::CopyHandleDesc(1);
+        input[4] = context.AddOutgoingHandle(b);
+
+        context.WriteToOutgoingCommandBuffer(output, *process, handle_table);
+
+        REQUIRE(handle_table.GetGeneric(output[2]) == a);
+        REQUIRE(handle_table.GetGeneric(output[4]) == b);
+    }
+
+    SECTION("translates null handles") {
+        input[0] = IPC::MakeHeader(0, 0, 2);
+        input[1] = IPC::MoveHandleDesc(1);
+        input[2] = context.AddOutgoingHandle(nullptr);
+
+        auto result = context.WriteToOutgoingCommandBuffer(output, *process, handle_table);
+
+        REQUIRE(result == RESULT_SUCCESS);
+        REQUIRE(output[2] == 0);
+    }
+
+    SECTION("translates multi-handle descriptors") {
+        auto a = MakeObject();
+        auto b = MakeObject();
+        auto c = MakeObject();
+        input[0] = IPC::MakeHeader(0, 0, 5);
+        input[1] = IPC::MoveHandleDesc(2);
+        input[2] = context.AddOutgoingHandle(a);
+        input[3] = context.AddOutgoingHandle(b);
+        input[4] = IPC::CopyHandleDesc(1);
+        input[5] = context.AddOutgoingHandle(c);
+
+        context.WriteToOutgoingCommandBuffer(output, *process, handle_table);
+
+        REQUIRE(handle_table.GetGeneric(output[2]) == a);
+        REQUIRE(handle_table.GetGeneric(output[3]) == b);
+        REQUIRE(handle_table.GetGeneric(output[5]) == c);
+    }
+}
+
+} // namespace Kernel
diff --git a/src/video_core/pica_state.h b/src/video_core/pica_state.h
index f46db09fb..2d23d34e6 100644
--- a/src/video_core/pica_state.h
+++ b/src/video_core/pica_state.h
@@ -87,12 +87,18 @@ struct State {
             // LUT value, encoded as 12-bit fixed point, with 12 fraction bits
             BitField<0, 12, u32> value; // 0.0.12 fixed point
 
-            // Used by HW for efficient interpolation, Citra does not use these
-            BitField<12, 12, s32> difference; // 1.0.11 fixed point
+            // Used for efficient interpolation.
+            BitField<12, 11, u32> difference; // 0.0.11 fixed point
+            BitField<23, 1, u32> neg_difference;
 
-            float ToFloat() {
+            float ToFloat() const {
                 return static_cast<float>(value) / 4095.f;
             }
+
+            float DiffToFloat() const {
+                float diff = static_cast<float>(difference) / 2047.f;
+                return neg_difference ? -diff : diff;
+            }
         };
 
         std::array<std::array<LutEntry, 256>, 24> luts;
@@ -105,6 +111,14 @@ struct State {
 
             BitField<0, 13, s32> difference; // 1.1.11 fixed point
             BitField<13, 11, u32> value;     // 0.0.11 fixed point
+
+            float ToFloat() const {
+                return static_cast<float>(value) / 2047.0f;
+            }
+
+            float DiffToFloat() const {
+                return static_cast<float>(difference) / 2047.0f;
+            }
         };
 
         std::array<LutEntry, 128> lut;
diff --git a/src/video_core/regs_lighting.h b/src/video_core/regs_lighting.h
index fbfebc0a7..b89709cfe 100644
--- a/src/video_core/regs_lighting.h
+++ b/src/video_core/regs_lighting.h
@@ -26,6 +26,8 @@ struct LightingRegs {
         DistanceAttenuation = 16,
     };
 
+    static constexpr unsigned NumLightingSampler = 24;
+
     static LightingSampler SpotlightAttenuationSampler(unsigned index) {
         return static_cast<LightingSampler>(
             static_cast<unsigned>(LightingSampler::SpotlightAttenuation) + index);
@@ -84,7 +86,7 @@ struct LightingRegs {
         NV = 2, // Cosine of the angle between the normal and the view vector
         LN = 3, // Cosine of the angle between the light and the normal vectors
         SP = 4, // Cosine of the angle between the light and the inverse spotlight vectors
-        CP = 5, // TODO: document and implement
+        CP = 5, // Cosine of the angle between the tangent and projection of half-angle vectors
     };
 
     enum class LightingBumpMode : u32 {
@@ -168,6 +170,8 @@ struct LightingRegs {
         union {
             BitField<0, 1, u32> directional;
             BitField<1, 1, u32> two_sided_diffuse; // When disabled, clamp dot-product to 0
+            BitField<2, 1, u32> geometric_factor_0;
+            BitField<3, 1, u32> geometric_factor_1;
         } config;
 
         BitField<0, 20, u32> dist_atten_bias;
diff --git a/src/video_core/regs_texturing.h b/src/video_core/regs_texturing.h
index 3f5355fa9..0b09f2299 100644
--- a/src/video_core/regs_texturing.h
+++ b/src/video_core/regs_texturing.h
@@ -30,10 +30,10 @@ struct TexturingRegs {
             Repeat = 2,
             MirroredRepeat = 3,
             // Mode 4-7 produces some weird result and may be just invalid:
-            // 4: Positive coord: clamp to edge; negative coord: repeat
-            // 5: Positive coord: clamp to border; negative coord: repeat
-            // 6: Repeat
-            // 7: Repeat
+            ClampToEdge2 = 4,   // Positive coord: clamp to edge; negative coord: repeat
+            ClampToBorder2 = 5, // Positive coord: clamp to border; negative coord: repeat
+            Repeat2 = 6,        // Same as Repeat
+            Repeat3 = 7,        // Same as Repeat
         };
 
         enum TextureFilter : u32 {
diff --git a/src/video_core/renderer_opengl/gl_rasterizer.cpp b/src/video_core/renderer_opengl/gl_rasterizer.cpp
index e6cccebf6..ff3f69ba3 100644
--- a/src/video_core/renderer_opengl/gl_rasterizer.cpp
+++ b/src/video_core/renderer_opengl/gl_rasterizer.cpp
@@ -49,9 +49,7 @@ RasterizerOpenGL::RasterizerOpenGL() : shader_dirty(true) {
 
     uniform_block_data.dirty = true;
 
-    for (unsigned index = 0; index < lighting_luts.size(); index++) {
-        uniform_block_data.lut_dirty[index] = true;
-    }
+    uniform_block_data.lut_dirty.fill(true);
 
     uniform_block_data.fog_lut_dirty = true;
 
@@ -96,36 +94,32 @@ RasterizerOpenGL::RasterizerOpenGL() : shader_dirty(true) {
     framebuffer.Create();
 
     // Allocate and bind lighting lut textures
-    for (size_t i = 0; i < lighting_luts.size(); ++i) {
-        lighting_luts[i].Create();
-        state.lighting_luts[i].texture_1d = lighting_luts[i].handle;
-    }
+    lighting_lut.Create();
+    state.lighting_lut.texture_buffer = lighting_lut.handle;
     state.Apply();
-
-    for (size_t i = 0; i < lighting_luts.size(); ++i) {
-        glActiveTexture(static_cast<GLenum>(GL_TEXTURE3 + i));
-        glTexImage1D(GL_TEXTURE_1D, 0, GL_RGBA32F, 256, 0, GL_RGBA, GL_FLOAT, nullptr);
-        glTexParameteri(GL_TEXTURE_1D, GL_TEXTURE_MIN_FILTER, GL_LINEAR);
-        glTexParameteri(GL_TEXTURE_1D, GL_TEXTURE_MAG_FILTER, GL_LINEAR);
-    }
+    lighting_lut_buffer.Create();
+    glBindBuffer(GL_TEXTURE_BUFFER, lighting_lut_buffer.handle);
+    glBufferData(GL_TEXTURE_BUFFER,
+                 sizeof(GLfloat) * 2 * 256 * Pica::LightingRegs::NumLightingSampler, nullptr,
+                 GL_DYNAMIC_DRAW);
+    glActiveTexture(TextureUnits::LightingLUT.Enum());
+    glTexBuffer(GL_TEXTURE_BUFFER, GL_RG32F, lighting_lut_buffer.handle);
 
     // Setup the LUT for the fog
-    {
-        fog_lut.Create();
-        state.fog_lut.texture_1d = fog_lut.handle;
-    }
+    fog_lut.Create();
+    state.fog_lut.texture_buffer = fog_lut.handle;
     state.Apply();
-
-    glActiveTexture(GL_TEXTURE9);
-    glTexImage1D(GL_TEXTURE_1D, 0, GL_R32UI, 128, 0, GL_RED_INTEGER, GL_UNSIGNED_INT, nullptr);
-    glTexParameteri(GL_TEXTURE_1D, GL_TEXTURE_MIN_FILTER, GL_NEAREST);
-    glTexParameteri(GL_TEXTURE_1D, GL_TEXTURE_MAG_FILTER, GL_NEAREST);
+    fog_lut_buffer.Create();
+    glBindBuffer(GL_TEXTURE_BUFFER, fog_lut_buffer.handle);
+    glBufferData(GL_TEXTURE_BUFFER, sizeof(GLfloat) * 2 * 128, nullptr, GL_DYNAMIC_DRAW);
+    glActiveTexture(TextureUnits::FogLUT.Enum());
+    glTexBuffer(GL_TEXTURE_BUFFER, GL_RG32F, fog_lut_buffer.handle);
 
     // Setup the noise LUT for proctex
     proctex_noise_lut.Create();
     state.proctex_noise_lut.texture_1d = proctex_noise_lut.handle;
     state.Apply();
-    glActiveTexture(GL_TEXTURE10);
+    glActiveTexture(TextureUnits::ProcTexNoiseLUT.Enum());
     glTexImage1D(GL_TEXTURE_1D, 0, GL_RG32F, 128, 0, GL_RG, GL_FLOAT, nullptr);
     glTexParameteri(GL_TEXTURE_1D, GL_TEXTURE_MIN_FILTER, GL_NEAREST);
     glTexParameteri(GL_TEXTURE_1D, GL_TEXTURE_MAG_FILTER, GL_NEAREST);
@@ -134,7 +128,7 @@ RasterizerOpenGL::RasterizerOpenGL() : shader_dirty(true) {
     proctex_color_map.Create();
     state.proctex_color_map.texture_1d = proctex_color_map.handle;
     state.Apply();
-    glActiveTexture(GL_TEXTURE11);
+    glActiveTexture(TextureUnits::ProcTexColorMap.Enum());
     glTexImage1D(GL_TEXTURE_1D, 0, GL_RG32F, 128, 0, GL_RG, GL_FLOAT, nullptr);
     glTexParameteri(GL_TEXTURE_1D, GL_TEXTURE_MIN_FILTER, GL_NEAREST);
     glTexParameteri(GL_TEXTURE_1D, GL_TEXTURE_MAG_FILTER, GL_NEAREST);
@@ -143,7 +137,7 @@ RasterizerOpenGL::RasterizerOpenGL() : shader_dirty(true) {
     proctex_alpha_map.Create();
     state.proctex_alpha_map.texture_1d = proctex_alpha_map.handle;
     state.Apply();
-    glActiveTexture(GL_TEXTURE12);
+    glActiveTexture(TextureUnits::ProcTexAlphaMap.Enum());
     glTexImage1D(GL_TEXTURE_1D, 0, GL_RG32F, 128, 0, GL_RG, GL_FLOAT, nullptr);
     glTexParameteri(GL_TEXTURE_1D, GL_TEXTURE_MIN_FILTER, GL_NEAREST);
     glTexParameteri(GL_TEXTURE_1D, GL_TEXTURE_MAG_FILTER, GL_NEAREST);
@@ -152,7 +146,7 @@ RasterizerOpenGL::RasterizerOpenGL() : shader_dirty(true) {
     proctex_lut.Create();
     state.proctex_lut.texture_1d = proctex_lut.handle;
     state.Apply();
-    glActiveTexture(GL_TEXTURE13);
+    glActiveTexture(TextureUnits::ProcTexLUT.Enum());
     glTexImage1D(GL_TEXTURE_1D, 0, GL_RGBA32F, 256, 0, GL_RGBA, GL_FLOAT, nullptr);
     glTexParameteri(GL_TEXTURE_1D, GL_TEXTURE_MIN_FILTER, GL_NEAREST);
     glTexParameteri(GL_TEXTURE_1D, GL_TEXTURE_MAG_FILTER, GL_NEAREST);
@@ -161,7 +155,7 @@ RasterizerOpenGL::RasterizerOpenGL() : shader_dirty(true) {
     proctex_diff_lut.Create();
     state.proctex_diff_lut.texture_1d = proctex_diff_lut.handle;
     state.Apply();
-    glActiveTexture(GL_TEXTURE14);
+    glActiveTexture(TextureUnits::ProcTexDiffLUT.Enum());
     glTexImage1D(GL_TEXTURE_1D, 0, GL_RGBA32F, 256, 0, GL_RGBA, GL_FLOAT, nullptr);
     glTexParameteri(GL_TEXTURE_1D, GL_TEXTURE_MIN_FILTER, GL_NEAREST);
     glTexParameteri(GL_TEXTURE_1D, GL_TEXTURE_MAG_FILTER, GL_NEAREST);
@@ -313,7 +307,7 @@ void RasterizerOpenGL::DrawTriangles() {
     }
 
     // Sync the lighting luts
-    for (unsigned index = 0; index < lighting_luts.size(); index++) {
+    for (unsigned index = 0; index < uniform_block_data.lut_dirty.size(); index++) {
         if (uniform_block_data.lut_dirty[index]) {
             SyncLightingLUT(index);
             uniform_block_data.lut_dirty[index] = false;
@@ -851,7 +845,7 @@ void RasterizerOpenGL::NotifyPicaRegisterChanged(u32 id) {
     case PICA_REG_INDEX_WORKAROUND(lighting.lut_data[6], 0x1ce):
     case PICA_REG_INDEX_WORKAROUND(lighting.lut_data[7], 0x1cf): {
         auto& lut_config = regs.lighting.lut_config;
-        uniform_block_data.lut_dirty[lut_config.type / 4] = true;
+        uniform_block_data.lut_dirty[lut_config.type] = true;
         break;
     }
     }
@@ -1187,77 +1181,57 @@ void RasterizerOpenGL::SetShader() {
         state.Apply();
 
         // Set the texture samplers to correspond to different texture units
-        GLuint uniform_tex = glGetUniformLocation(shader->shader.handle, "tex[0]");
+        GLint uniform_tex = glGetUniformLocation(shader->shader.handle, "tex[0]");
         if (uniform_tex != -1) {
-            glUniform1i(uniform_tex, 0);
+            glUniform1i(uniform_tex, TextureUnits::PicaTexture(0).id);
         }
         uniform_tex = glGetUniformLocation(shader->shader.handle, "tex[1]");
         if (uniform_tex != -1) {
-            glUniform1i(uniform_tex, 1);
+            glUniform1i(uniform_tex, TextureUnits::PicaTexture(1).id);
         }
         uniform_tex = glGetUniformLocation(shader->shader.handle, "tex[2]");
         if (uniform_tex != -1) {
-            glUniform1i(uniform_tex, 2);
+            glUniform1i(uniform_tex, TextureUnits::PicaTexture(2).id);
         }
 
         // Set the texture samplers to correspond to different lookup table texture units
-        GLuint uniform_lut = glGetUniformLocation(shader->shader.handle, "lut[0]");
-        if (uniform_lut != -1) {
-            glUniform1i(uniform_lut, 3);
-        }
-        uniform_lut = glGetUniformLocation(shader->shader.handle, "lut[1]");
-        if (uniform_lut != -1) {
-            glUniform1i(uniform_lut, 4);
-        }
-        uniform_lut = glGetUniformLocation(shader->shader.handle, "lut[2]");
-        if (uniform_lut != -1) {
-            glUniform1i(uniform_lut, 5);
-        }
-        uniform_lut = glGetUniformLocation(shader->shader.handle, "lut[3]");
-        if (uniform_lut != -1) {
-            glUniform1i(uniform_lut, 6);
-        }
-        uniform_lut = glGetUniformLocation(shader->shader.handle, "lut[4]");
+        GLint uniform_lut = glGetUniformLocation(shader->shader.handle, "lighting_lut");
         if (uniform_lut != -1) {
-            glUniform1i(uniform_lut, 7);
-        }
-        uniform_lut = glGetUniformLocation(shader->shader.handle, "lut[5]");
-        if (uniform_lut != -1) {
-            glUniform1i(uniform_lut, 8);
+            glUniform1i(uniform_lut, TextureUnits::LightingLUT.id);
         }
 
-        GLuint uniform_fog_lut = glGetUniformLocation(shader->shader.handle, "fog_lut");
+        GLint uniform_fog_lut = glGetUniformLocation(shader->shader.handle, "fog_lut");
         if (uniform_fog_lut != -1) {
-            glUniform1i(uniform_fog_lut, 9);
+            glUniform1i(uniform_fog_lut, TextureUnits::FogLUT.id);
         }
 
-        GLuint uniform_proctex_noise_lut =
+        GLint uniform_proctex_noise_lut =
             glGetUniformLocation(shader->shader.handle, "proctex_noise_lut");
         if (uniform_proctex_noise_lut != -1) {
-            glUniform1i(uniform_proctex_noise_lut, 10);
+            glUniform1i(uniform_proctex_noise_lut, TextureUnits::ProcTexNoiseLUT.id);
         }
 
-        GLuint uniform_proctex_color_map =
+        GLint uniform_proctex_color_map =
             glGetUniformLocation(shader->shader.handle, "proctex_color_map");
         if (uniform_proctex_color_map != -1) {
-            glUniform1i(uniform_proctex_color_map, 11);
+            glUniform1i(uniform_proctex_color_map, TextureUnits::ProcTexColorMap.id);
         }
 
-        GLuint uniform_proctex_alpha_map =
+        GLint uniform_proctex_alpha_map =
             glGetUniformLocation(shader->shader.handle, "proctex_alpha_map");
         if (uniform_proctex_alpha_map != -1) {
-            glUniform1i(uniform_proctex_alpha_map, 12);
+            glUniform1i(uniform_proctex_alpha_map, TextureUnits::ProcTexAlphaMap.id);
         }
 
-        GLuint uniform_proctex_lut = glGetUniformLocation(shader->shader.handle, "proctex_lut");
+        GLint uniform_proctex_lut = glGetUniformLocation(shader->shader.handle, "proctex_lut");
         if (uniform_proctex_lut != -1) {
-            glUniform1i(uniform_proctex_lut, 13);
+            glUniform1i(uniform_proctex_lut, TextureUnits::ProcTexLUT.id);
         }
 
-        GLuint uniform_proctex_diff_lut =
+        GLint uniform_proctex_diff_lut =
             glGetUniformLocation(shader->shader.handle, "proctex_diff_lut");
         if (uniform_proctex_diff_lut != -1) {
-            glUniform1i(uniform_proctex_diff_lut, 14);
+            glUniform1i(uniform_proctex_diff_lut, TextureUnits::ProcTexDiffLUT.id);
         }
 
         current_shader = shader_cache.emplace(config, std::move(shader)).first->second.get();
@@ -1380,16 +1354,17 @@ void RasterizerOpenGL::SyncFogColor() {
 }
 
 void RasterizerOpenGL::SyncFogLUT() {
-    std::array<GLuint, 128> new_data;
+    std::array<GLvec2, 128> new_data;
 
     std::transform(Pica::g_state.fog.lut.begin(), Pica::g_state.fog.lut.end(), new_data.begin(),
-                   [](const auto& entry) { return entry.raw; });
+                   [](const auto& entry) {
+                       return GLvec2{entry.ToFloat(), entry.DiffToFloat()};
+                   });
 
     if (new_data != fog_lut_data) {
         fog_lut_data = new_data;
-        glActiveTexture(GL_TEXTURE9);
-        glTexSubImage1D(GL_TEXTURE_1D, 0, 0, 128, GL_RED_INTEGER, GL_UNSIGNED_INT,
-                        fog_lut_data.data());
+        glBindBuffer(GL_TEXTURE_BUFFER, fog_lut_buffer.handle);
+        glBufferSubData(GL_TEXTURE_BUFFER, 0, new_data.size() * sizeof(GLvec2), new_data.data());
     }
 }
 
@@ -1426,17 +1401,18 @@ static void SyncProcTexValueLUT(const std::array<Pica::State::ProcTex::ValueEntr
 }
 
 void RasterizerOpenGL::SyncProcTexNoiseLUT() {
-    SyncProcTexValueLUT(Pica::g_state.proctex.noise_table, proctex_noise_lut_data, GL_TEXTURE10);
+    SyncProcTexValueLUT(Pica::g_state.proctex.noise_table, proctex_noise_lut_data,
+                        TextureUnits::ProcTexNoiseLUT.Enum());
 }
 
 void RasterizerOpenGL::SyncProcTexColorMap() {
     SyncProcTexValueLUT(Pica::g_state.proctex.color_map_table, proctex_color_map_data,
-                        GL_TEXTURE11);
+                        TextureUnits::ProcTexColorMap.Enum());
 }
 
 void RasterizerOpenGL::SyncProcTexAlphaMap() {
     SyncProcTexValueLUT(Pica::g_state.proctex.alpha_map_table, proctex_alpha_map_data,
-                        GL_TEXTURE12);
+                        TextureUnits::ProcTexAlphaMap.Enum());
 }
 
 void RasterizerOpenGL::SyncProcTexLUT() {
@@ -1451,7 +1427,7 @@ void RasterizerOpenGL::SyncProcTexLUT() {
 
     if (new_data != proctex_lut_data) {
         proctex_lut_data = new_data;
-        glActiveTexture(GL_TEXTURE13);
+        glActiveTexture(TextureUnits::ProcTexLUT.Enum());
         glTexSubImage1D(GL_TEXTURE_1D, 0, 0, 256, GL_RGBA, GL_FLOAT, proctex_lut_data.data());
     }
 }
@@ -1468,7 +1444,7 @@ void RasterizerOpenGL::SyncProcTexDiffLUT() {
 
     if (new_data != proctex_diff_lut_data) {
         proctex_diff_lut_data = new_data;
-        glActiveTexture(GL_TEXTURE14);
+        glActiveTexture(TextureUnits::ProcTexDiffLUT.Enum());
         glTexSubImage1D(GL_TEXTURE_1D, 0, 0, 256, GL_RGBA, GL_FLOAT, proctex_diff_lut_data.data());
     }
 }
@@ -1571,20 +1547,17 @@ void RasterizerOpenGL::SyncGlobalAmbient() {
 }
 
 void RasterizerOpenGL::SyncLightingLUT(unsigned lut_index) {
-    std::array<GLvec4, 256> new_data;
-
-    for (unsigned offset = 0; offset < new_data.size(); ++offset) {
-        new_data[offset][0] = Pica::g_state.lighting.luts[(lut_index * 4) + 0][offset].ToFloat();
-        new_data[offset][1] = Pica::g_state.lighting.luts[(lut_index * 4) + 1][offset].ToFloat();
-        new_data[offset][2] = Pica::g_state.lighting.luts[(lut_index * 4) + 2][offset].ToFloat();
-        new_data[offset][3] = Pica::g_state.lighting.luts[(lut_index * 4) + 3][offset].ToFloat();
-    }
+    std::array<GLvec2, 256> new_data;
+    const auto& source_lut = Pica::g_state.lighting.luts[lut_index];
+    std::transform(source_lut.begin(), source_lut.end(), new_data.begin(), [](const auto& entry) {
+        return GLvec2{entry.ToFloat(), entry.DiffToFloat()};
+    });
 
     if (new_data != lighting_lut_data[lut_index]) {
         lighting_lut_data[lut_index] = new_data;
-        glActiveTexture(GL_TEXTURE3 + lut_index);
-        glTexSubImage1D(GL_TEXTURE_1D, 0, 0, 256, GL_RGBA, GL_FLOAT,
-                        lighting_lut_data[lut_index].data());
+        glBindBuffer(GL_TEXTURE_BUFFER, lighting_lut_buffer.handle);
+        glBufferSubData(GL_TEXTURE_BUFFER, lut_index * new_data.size() * sizeof(GLvec2),
+                        new_data.size() * sizeof(GLvec2), new_data.data());
     }
 }
 
diff --git a/src/video_core/renderer_opengl/gl_rasterizer.h b/src/video_core/renderer_opengl/gl_rasterizer.h
index d9a3e9d1c..a433c1d4a 100644
--- a/src/video_core/renderer_opengl/gl_rasterizer.h
+++ b/src/video_core/renderer_opengl/gl_rasterizer.h
@@ -263,7 +263,7 @@ private:
 
     struct {
         UniformData data;
-        bool lut_dirty[6];
+        std::array<bool, Pica::LightingRegs::NumLightingSampler> lut_dirty;
         bool fog_lut_dirty;
         bool proctex_noise_lut_dirty;
         bool proctex_color_map_dirty;
@@ -279,11 +279,13 @@ private:
     OGLBuffer uniform_buffer;
     OGLFramebuffer framebuffer;
 
-    std::array<OGLTexture, 6> lighting_luts;
-    std::array<std::array<GLvec4, 256>, 6> lighting_lut_data{};
+    OGLBuffer lighting_lut_buffer;
+    OGLTexture lighting_lut;
+    std::array<std::array<GLvec2, 256>, Pica::LightingRegs::NumLightingSampler> lighting_lut_data{};
 
+    OGLBuffer fog_lut_buffer;
     OGLTexture fog_lut;
-    std::array<GLuint, 128> fog_lut_data{};
+    std::array<GLvec2, 128> fog_lut_data{};
 
     OGLTexture proctex_noise_lut;
     std::array<GLvec2, 128> proctex_noise_lut_data{};
diff --git a/src/video_core/renderer_opengl/gl_shader_gen.cpp b/src/video_core/renderer_opengl/gl_shader_gen.cpp
index db53710aa..c93b108fb 100644
--- a/src/video_core/renderer_opengl/gl_shader_gen.cpp
+++ b/src/video_core/renderer_opengl/gl_shader_gen.cpp
@@ -73,6 +73,8 @@ PicaShaderConfig PicaShaderConfig::BuildFromRegs(const Pica::Regs& regs) {
         state.lighting.light[light_index].num = num;
         state.lighting.light[light_index].directional = light.config.directional != 0;
         state.lighting.light[light_index].two_sided_diffuse = light.config.two_sided_diffuse != 0;
+        state.lighting.light[light_index].geometric_factor_0 = light.config.geometric_factor_0 != 0;
+        state.lighting.light[light_index].geometric_factor_1 = light.config.geometric_factor_1 != 0;
         state.lighting.light[light_index].dist_atten_enable =
             !regs.lighting.IsDistAttenDisabled(num);
         state.lighting.light[light_index].spot_atten_enable =
@@ -518,14 +520,16 @@ static void WriteLighting(std::string& out, const PicaShaderConfig& config) {
            "vec4 specular_sum = vec4(0.0, 0.0, 0.0, 1.0);\n"
            "vec3 light_vector = vec3(0.0);\n"
            "vec3 refl_value = vec3(0.0);\n"
-           "vec3 spot_dir = vec3(0.0);\n;";
+           "vec3 spot_dir = vec3(0.0);\n"
+           "vec3 half_vector = vec3(0.0);\n"
+           "float geo_factor = 1.0;\n";
 
-    // Compute fragment normals
+    // Compute fragment normals and tangents
+    const std::string pertubation =
+        "2.0 * (" + SampleTexture(config, lighting.bump_selector) + ").rgb - 1.0";
     if (lighting.bump_mode == LightingRegs::LightingBumpMode::NormalMap) {
-        // Bump mapping is enabled using a normal map, read perturbation vector from the selected
-        // texture
-        out += "vec3 surface_normal = 2.0 * (" + SampleTexture(config, lighting.bump_selector) +
-               ").rgb - 1.0;\n";
+        // Bump mapping is enabled using a normal map
+        out += "vec3 surface_normal = " + pertubation + ";\n";
 
         // Recompute Z-component of perturbation if 'renorm' is enabled, this provides a higher
         // precision result
@@ -534,31 +538,41 @@ static void WriteLighting(std::string& out, const PicaShaderConfig& config) {
                 "(1.0 - (surface_normal.x*surface_normal.x + surface_normal.y*surface_normal.y))";
             out += "surface_normal.z = sqrt(max(" + val + ", 0.0));\n";
         }
+
+        // The tangent vector is not perturbed by the normal map and is just a unit vector.
+        out += "vec3 surface_tangent = vec3(1.0, 0.0, 0.0);\n";
     } else if (lighting.bump_mode == LightingRegs::LightingBumpMode::TangentMap) {
         // Bump mapping is enabled using a tangent map
-        LOG_CRITICAL(HW_GPU, "unimplemented bump mapping mode (tangent mapping)");
-        UNIMPLEMENTED();
+        out += "vec3 surface_tangent = " + pertubation + ";\n";
+        // Mathematically, recomputing Z-component of the tangent vector won't affect the relevant
+        // computation below, which is also confirmed on 3DS. So we don't bother recomputing here
+        // even if 'renorm' is enabled.
+
+        // The normal vector is not perturbed by the tangent map and is just a unit vector.
+        out += "vec3 surface_normal = vec3(0.0, 0.0, 1.0);\n";
     } else {
-        // No bump mapping - surface local normal is just a unit normal
+        // No bump mapping - surface local normal and tangent are just unit vectors
         out += "vec3 surface_normal = vec3(0.0, 0.0, 1.0);\n";
+        out += "vec3 surface_tangent = vec3(1.0, 0.0, 0.0);\n";
     }
 
     // Rotate the surface-local normal by the interpolated normal quaternion to convert it to
     // eyespace.
-    out += "vec3 normal = quaternion_rotate(normalize(normquat), surface_normal);\n";
+    out += "vec4 normalized_normquat = normalize(normquat);\n";
+    out += "vec3 normal = quaternion_rotate(normalized_normquat, surface_normal);\n";
+    out += "vec3 tangent = quaternion_rotate(normalized_normquat, surface_tangent);\n";
 
-    // Gets the index into the specified lookup table for specular lighting
-    auto GetLutIndex = [&lighting](unsigned light_num, LightingRegs::LightingLutInput input,
-                                   bool abs) {
-        const std::string half_angle = "normalize(normalize(view) + light_vector)";
+    // Samples the specified lookup table for specular lighting
+    auto GetLutValue = [&lighting](LightingRegs::LightingSampler sampler, unsigned light_num,
+                                   LightingRegs::LightingLutInput input, bool abs) {
         std::string index;
         switch (input) {
         case LightingRegs::LightingLutInput::NH:
-            index = "dot(normal, " + half_angle + ")";
+            index = "dot(normal, normalize(half_vector))";
             break;
 
         case LightingRegs::LightingLutInput::VH:
-            index = std::string("dot(normalize(view), " + half_angle + ")");
+            index = std::string("dot(normalize(view), normalize(half_vector))");
             break;
 
         case LightingRegs::LightingLutInput::NV:
@@ -573,6 +587,22 @@ static void WriteLighting(std::string& out, const PicaShaderConfig& config) {
             index = std::string("dot(light_vector, spot_dir)");
             break;
 
+        case LightingRegs::LightingLutInput::CP:
+            // CP input is only available with configuration 7
+            if (lighting.config == LightingRegs::LightingConfig::Config7) {
+                // Note: even if the normal vector is modified by normal map, which is not the
+                // normal of the tangent plane anymore, the half angle vector is still projected
+                // using the modified normal vector.
+                std::string half_angle_proj = "normalize(half_vector) - normal / dot(normal, "
+                                              "normal) * dot(normal, normalize(half_vector))";
+                // Note: the half angle vector projection is confirmed not normalized before the dot
+                // product. The result is in fact not cos(phi) as the name suggested.
+                index = "dot(" + half_angle_proj + ", tangent)";
+            } else {
+                index = "0.0";
+            }
+            break;
+
         default:
             LOG_CRITICAL(HW_GPU, "Unknown lighting LUT input %d\n", (int)input);
             UNIMPLEMENTED();
@@ -580,22 +610,18 @@ static void WriteLighting(std::string& out, const PicaShaderConfig& config) {
             break;
         }
 
+        std::string sampler_string = std::to_string(static_cast<unsigned>(sampler));
+
         if (abs) {
             // LUT index is in the range of (0.0, 1.0)
             index = lighting.light[light_num].two_sided_diffuse ? "abs(" + index + ")"
                                                                 : "max(" + index + ", 0.0)";
+            return "LookupLightingLUTUnsigned(" + sampler_string + ", " + index + ")";
         } else {
             // LUT index is in the range of (-1.0, 1.0)
-            index = "((" + index + " < 0) ? " + index + " + 2.0 : " + index + ") / 2.0";
+            return "LookupLightingLUTSigned(" + sampler_string + ", " + index + ")";
         }
 
-        return "(OFFSET_256 + SCALE_256 * clamp(" + index + ", 0.0, 1.0))";
-    };
-
-    // Gets the lighting lookup table value given the specified sampler and index
-    auto GetLutValue = [](LightingRegs::LightingSampler sampler, std::string lut_index) {
-        return std::string("texture(lut[" + std::to_string((unsigned)sampler / 4) + "], " +
-                           lut_index + ")[" + std::to_string((unsigned)sampler & 3) + "]");
     };
 
     // Write the code to emulate each enabled light
@@ -610,6 +636,7 @@ static void WriteLighting(std::string& out, const PicaShaderConfig& config) {
             out += "light_vector = normalize(" + light_src + ".position + view);\n";
 
         out += "spot_dir = " + light_src + ".spot_direction;\n";
+        out += "half_vector = normalize(view) + light_vector;\n";
 
         // Compute dot product of light_vector and normal, adjust if lighting is one-sided or
         // two-sided
@@ -622,48 +649,57 @@ static void WriteLighting(std::string& out, const PicaShaderConfig& config) {
         if (light_config.spot_atten_enable &&
             LightingRegs::IsLightingSamplerSupported(
                 lighting.config, LightingRegs::LightingSampler::SpotlightAttenuation)) {
-            std::string index =
-                GetLutIndex(light_config.num, lighting.lut_sp.type, lighting.lut_sp.abs_input);
-            auto sampler = LightingRegs::SpotlightAttenuationSampler(light_config.num);
-            spot_atten = "(" + std::to_string(lighting.lut_sp.scale) + " * " +
-                         GetLutValue(sampler, index) + ")";
+            std::string value =
+                GetLutValue(LightingRegs::SpotlightAttenuationSampler(light_config.num),
+                            light_config.num, lighting.lut_sp.type, lighting.lut_sp.abs_input);
+            spot_atten = "(" + std::to_string(lighting.lut_sp.scale) + " * " + value + ")";
         }
 
         // If enabled, compute distance attenuation value
         std::string dist_atten = "1.0";
         if (light_config.dist_atten_enable) {
-            std::string index = "(" + light_src + ".dist_atten_scale * length(-view - " +
-                                light_src + ".position) + " + light_src + ".dist_atten_bias)";
-            index = "(OFFSET_256 + SCALE_256 * clamp(" + index + ", 0.0, 1.0))";
+            std::string index = "clamp(" + light_src + ".dist_atten_scale * length(-view - " +
+                                light_src + ".position) + " + light_src +
+                                ".dist_atten_bias, 0.0, 1.0)";
             auto sampler = LightingRegs::DistanceAttenuationSampler(light_config.num);
-            dist_atten = GetLutValue(sampler, index);
+            dist_atten = "LookupLightingLUTUnsigned(" +
+                         std::to_string(static_cast<unsigned>(sampler)) + "," + index + ")";
         }
 
         // If enabled, clamp specular component if lighting result is negative
         std::string clamp_highlights =
             lighting.clamp_highlights ? "(dot(light_vector, normal) <= 0.0 ? 0.0 : 1.0)" : "1.0";
 
+        if (light_config.geometric_factor_0 || light_config.geometric_factor_1) {
+            out += "geo_factor = dot(half_vector, half_vector);\n"
+                   "geo_factor = geo_factor == 0.0 ? 0.0 : min(" +
+                   dot_product + " / geo_factor, 1.0);\n";
+        }
+
         // Specular 0 component
         std::string d0_lut_value = "1.0";
         if (lighting.lut_d0.enable &&
             LightingRegs::IsLightingSamplerSupported(
                 lighting.config, LightingRegs::LightingSampler::Distribution0)) {
             // Lookup specular "distribution 0" LUT value
-            std::string index =
-                GetLutIndex(light_config.num, lighting.lut_d0.type, lighting.lut_d0.abs_input);
-            d0_lut_value = "(" + std::to_string(lighting.lut_d0.scale) + " * " +
-                           GetLutValue(LightingRegs::LightingSampler::Distribution0, index) + ")";
+            std::string value =
+                GetLutValue(LightingRegs::LightingSampler::Distribution0, light_config.num,
+                            lighting.lut_d0.type, lighting.lut_d0.abs_input);
+            d0_lut_value = "(" + std::to_string(lighting.lut_d0.scale) + " * " + value + ")";
         }
         std::string specular_0 = "(" + d0_lut_value + " * " + light_src + ".specular_0)";
+        if (light_config.geometric_factor_0) {
+            specular_0 = "(" + specular_0 + " * geo_factor)";
+        }
 
         // If enabled, lookup ReflectRed value, otherwise, 1.0 is used
         if (lighting.lut_rr.enable &&
             LightingRegs::IsLightingSamplerSupported(lighting.config,
                                                      LightingRegs::LightingSampler::ReflectRed)) {
-            std::string index =
-                GetLutIndex(light_config.num, lighting.lut_rr.type, lighting.lut_rr.abs_input);
-            std::string value = "(" + std::to_string(lighting.lut_rr.scale) + " * " +
-                                GetLutValue(LightingRegs::LightingSampler::ReflectRed, index) + ")";
+            std::string value =
+                GetLutValue(LightingRegs::LightingSampler::ReflectRed, light_config.num,
+                            lighting.lut_rr.type, lighting.lut_rr.abs_input);
+            value = "(" + std::to_string(lighting.lut_rr.scale) + " * " + value + ")";
             out += "refl_value.r = " + value + ";\n";
         } else {
             out += "refl_value.r = 1.0;\n";
@@ -673,11 +709,10 @@ static void WriteLighting(std::string& out, const PicaShaderConfig& config) {
         if (lighting.lut_rg.enable &&
             LightingRegs::IsLightingSamplerSupported(lighting.config,
                                                      LightingRegs::LightingSampler::ReflectGreen)) {
-            std::string index =
-                GetLutIndex(light_config.num, lighting.lut_rg.type, lighting.lut_rg.abs_input);
-            std::string value = "(" + std::to_string(lighting.lut_rg.scale) + " * " +
-                                GetLutValue(LightingRegs::LightingSampler::ReflectGreen, index) +
-                                ")";
+            std::string value =
+                GetLutValue(LightingRegs::LightingSampler::ReflectGreen, light_config.num,
+                            lighting.lut_rg.type, lighting.lut_rg.abs_input);
+            value = "(" + std::to_string(lighting.lut_rg.scale) + " * " + value + ")";
             out += "refl_value.g = " + value + ";\n";
         } else {
             out += "refl_value.g = refl_value.r;\n";
@@ -687,11 +722,10 @@ static void WriteLighting(std::string& out, const PicaShaderConfig& config) {
         if (lighting.lut_rb.enable &&
             LightingRegs::IsLightingSamplerSupported(lighting.config,
                                                      LightingRegs::LightingSampler::ReflectBlue)) {
-            std::string index =
-                GetLutIndex(light_config.num, lighting.lut_rb.type, lighting.lut_rb.abs_input);
-            std::string value = "(" + std::to_string(lighting.lut_rb.scale) + " * " +
-                                GetLutValue(LightingRegs::LightingSampler::ReflectBlue, index) +
-                                ")";
+            std::string value =
+                GetLutValue(LightingRegs::LightingSampler::ReflectBlue, light_config.num,
+                            lighting.lut_rb.type, lighting.lut_rb.abs_input);
+            value = "(" + std::to_string(lighting.lut_rb.scale) + " * " + value + ")";
             out += "refl_value.b = " + value + ";\n";
         } else {
             out += "refl_value.b = refl_value.r;\n";
@@ -703,23 +737,26 @@ static void WriteLighting(std::string& out, const PicaShaderConfig& config) {
             LightingRegs::IsLightingSamplerSupported(
                 lighting.config, LightingRegs::LightingSampler::Distribution1)) {
             // Lookup specular "distribution 1" LUT value
-            std::string index =
-                GetLutIndex(light_config.num, lighting.lut_d1.type, lighting.lut_d1.abs_input);
-            d1_lut_value = "(" + std::to_string(lighting.lut_d1.scale) + " * " +
-                           GetLutValue(LightingRegs::LightingSampler::Distribution1, index) + ")";
+            std::string value =
+                GetLutValue(LightingRegs::LightingSampler::Distribution1, light_config.num,
+                            lighting.lut_d1.type, lighting.lut_d1.abs_input);
+            d1_lut_value = "(" + std::to_string(lighting.lut_d1.scale) + " * " + value + ")";
         }
         std::string specular_1 =
             "(" + d1_lut_value + " * refl_value * " + light_src + ".specular_1)";
+        if (light_config.geometric_factor_1) {
+            specular_1 = "(" + specular_1 + " * geo_factor)";
+        }
 
         // Fresnel
         if (lighting.lut_fr.enable &&
             LightingRegs::IsLightingSamplerSupported(lighting.config,
                                                      LightingRegs::LightingSampler::Fresnel)) {
             // Lookup fresnel LUT value
-            std::string index =
-                GetLutIndex(light_config.num, lighting.lut_fr.type, lighting.lut_fr.abs_input);
-            std::string value = "(" + std::to_string(lighting.lut_fr.scale) + " * " +
-                                GetLutValue(LightingRegs::LightingSampler::Fresnel, index) + ")";
+            std::string value =
+                GetLutValue(LightingRegs::LightingSampler::Fresnel, light_config.num,
+                            lighting.lut_fr.type, lighting.lut_fr.abs_input);
+            value = "(" + std::to_string(lighting.lut_fr.scale) + " * " + value + ")";
 
             // Enabled for difffuse lighting alpha component
             if (lighting.fresnel_selector == LightingRegs::LightingFresnelSelector::PrimaryAlpha ||
@@ -973,10 +1010,6 @@ std::string GenerateFragmentShader(const PicaShaderConfig& config) {
 #define NUM_TEV_STAGES 6
 #define NUM_LIGHTS 8
 
-// Texture coordinate offsets and scales
-#define OFFSET_256 (0.5 / 256.0)
-#define SCALE_256 (255.0 / 256.0)
-
 in vec4 primary_color;
 in vec2 texcoord[3];
 in float texcoord0_w;
@@ -1018,8 +1051,8 @@ layout (std140) uniform shader_data {
 };
 
 uniform sampler2D tex[3];
-uniform sampler1D lut[6];
-uniform usampler1D fog_lut;
+uniform samplerBuffer lighting_lut;
+uniform samplerBuffer fog_lut;
 uniform sampler1D proctex_noise_lut;
 uniform sampler1D proctex_color_map;
 uniform sampler1D proctex_alpha_map;
@@ -1031,6 +1064,24 @@ vec3 quaternion_rotate(vec4 q, vec3 v) {
     return v + 2.0 * cross(q.xyz, cross(q.xyz, v) + q.w * v);
 }
 
+float LookupLightingLUT(int lut_index, int index, float delta) {
+    vec2 entry = texelFetch(lighting_lut, lut_index * 256 + index).rg;
+    return entry.r + entry.g * delta;
+}
+
+float LookupLightingLUTUnsigned(int lut_index, float pos) {
+    int index = clamp(int(pos * 256.0), 0, 255);
+    float delta = pos * 256.0 - index;
+    return LookupLightingLUT(lut_index, index, delta);
+}
+
+float LookupLightingLUTSigned(int lut_index, float pos) {
+    int index = clamp(int(pos * 128.0), -128, 127);
+    float delta = pos * 128.0 - index;
+    if (index < 0) index += 256;
+    return LookupLightingLUT(lut_index, index, delta);
+}
+
 )";
 
     if (config.state.proctex.enable)
@@ -1094,12 +1145,8 @@ vec4 secondary_fragment_color = vec4(0.0);
         // Generate clamped fog factor from LUT for given fog index
         out += "float fog_i = clamp(floor(fog_index), 0.0, 127.0);\n";
         out += "float fog_f = fog_index - fog_i;\n";
-        out += "uint fog_lut_entry = texelFetch(fog_lut, int(fog_i), 0).r;\n";
-        out += "float fog_lut_entry_difference = float(int((fog_lut_entry & 0x1FFFU) << 19U) >> "
-               "19);\n"; // Extract signed difference
-        out += "float fog_lut_entry_value = float((fog_lut_entry >> 13U) & 0x7FFU);\n";
-        out += "float fog_factor = (fog_lut_entry_value + fog_lut_entry_difference * fog_f) / "
-               "2047.0;\n";
+        out += "vec2 fog_lut_entry = texelFetch(fog_lut, int(fog_i)).rg;\n";
+        out += "float fog_factor = fog_lut_entry.r + fog_lut_entry.g * fog_f;\n";
         out += "fog_factor = clamp(fog_factor, 0.0, 1.0);\n";
 
         // Blend the fog
diff --git a/src/video_core/renderer_opengl/gl_shader_gen.h b/src/video_core/renderer_opengl/gl_shader_gen.h
index 9c90eadf9..2302ae453 100644
--- a/src/video_core/renderer_opengl/gl_shader_gen.h
+++ b/src/video_core/renderer_opengl/gl_shader_gen.h
@@ -94,6 +94,8 @@ union PicaShaderConfig {
                 bool two_sided_diffuse;
                 bool dist_atten_enable;
                 bool spot_atten_enable;
+                bool geometric_factor_0;
+                bool geometric_factor_1;
             } light[8];
 
             bool enable;
diff --git a/src/video_core/renderer_opengl/gl_state.cpp b/src/video_core/renderer_opengl/gl_state.cpp
index bf837a7fb..eface2dea 100644
--- a/src/video_core/renderer_opengl/gl_state.cpp
+++ b/src/video_core/renderer_opengl/gl_state.cpp
@@ -52,11 +52,9 @@ OpenGLState::OpenGLState() {
         texture_unit.sampler = 0;
     }
 
-    for (auto& lut : lighting_luts) {
-        lut.texture_1d = 0;
-    }
+    lighting_lut.texture_buffer = 0;
 
-    fog_lut.texture_1d = 0;
+    fog_lut.texture_buffer = 0;
 
     proctex_lut.texture_1d = 0;
     proctex_diff_lut.texture_1d = 0;
@@ -185,7 +183,7 @@ void OpenGLState::Apply() const {
     // Textures
     for (unsigned i = 0; i < ARRAY_SIZE(texture_units); ++i) {
         if (texture_units[i].texture_2d != cur_state.texture_units[i].texture_2d) {
-            glActiveTexture(GL_TEXTURE0 + i);
+            glActiveTexture(TextureUnits::PicaTexture(i).Enum());
             glBindTexture(GL_TEXTURE_2D, texture_units[i].texture_2d);
         }
         if (texture_units[i].sampler != cur_state.texture_units[i].sampler) {
@@ -194,46 +192,44 @@ void OpenGLState::Apply() const {
     }
 
     // Lighting LUTs
-    for (unsigned i = 0; i < ARRAY_SIZE(lighting_luts); ++i) {
-        if (lighting_luts[i].texture_1d != cur_state.lighting_luts[i].texture_1d) {
-            glActiveTexture(GL_TEXTURE3 + i);
-            glBindTexture(GL_TEXTURE_1D, lighting_luts[i].texture_1d);
-        }
+    if (lighting_lut.texture_buffer != cur_state.lighting_lut.texture_buffer) {
+        glActiveTexture(TextureUnits::LightingLUT.Enum());
+        glBindTexture(GL_TEXTURE_BUFFER, cur_state.lighting_lut.texture_buffer);
     }
 
     // Fog LUT
-    if (fog_lut.texture_1d != cur_state.fog_lut.texture_1d) {
-        glActiveTexture(GL_TEXTURE9);
-        glBindTexture(GL_TEXTURE_1D, fog_lut.texture_1d);
+    if (fog_lut.texture_buffer != cur_state.fog_lut.texture_buffer) {
+        glActiveTexture(TextureUnits::FogLUT.Enum());
+        glBindTexture(GL_TEXTURE_BUFFER, fog_lut.texture_buffer);
     }
 
     // ProcTex Noise LUT
     if (proctex_noise_lut.texture_1d != cur_state.proctex_noise_lut.texture_1d) {
-        glActiveTexture(GL_TEXTURE10);
+        glActiveTexture(TextureUnits::ProcTexNoiseLUT.Enum());
         glBindTexture(GL_TEXTURE_1D, proctex_noise_lut.texture_1d);
     }
 
     // ProcTex Color Map
     if (proctex_color_map.texture_1d != cur_state.proctex_color_map.texture_1d) {
-        glActiveTexture(GL_TEXTURE11);
+        glActiveTexture(TextureUnits::ProcTexColorMap.Enum());
         glBindTexture(GL_TEXTURE_1D, proctex_color_map.texture_1d);
     }
 
     // ProcTex Alpha Map
     if (proctex_alpha_map.texture_1d != cur_state.proctex_alpha_map.texture_1d) {
-        glActiveTexture(GL_TEXTURE12);
+        glActiveTexture(TextureUnits::ProcTexAlphaMap.Enum());
         glBindTexture(GL_TEXTURE_1D, proctex_alpha_map.texture_1d);
     }
 
     // ProcTex LUT
     if (proctex_lut.texture_1d != cur_state.proctex_lut.texture_1d) {
-        glActiveTexture(GL_TEXTURE13);
+        glActiveTexture(TextureUnits::ProcTexLUT.Enum());
         glBindTexture(GL_TEXTURE_1D, proctex_lut.texture_1d);
     }
 
     // ProcTex Diff LUT
     if (proctex_diff_lut.texture_1d != cur_state.proctex_diff_lut.texture_1d) {
-        glActiveTexture(GL_TEXTURE14);
+        glActiveTexture(TextureUnits::ProcTexDiffLUT.Enum());
         glBindTexture(GL_TEXTURE_1D, proctex_diff_lut.texture_1d);
     }
 
@@ -274,6 +270,20 @@ void OpenGLState::ResetTexture(GLuint handle) {
             unit.texture_2d = 0;
         }
     }
+    if (cur_state.lighting_lut.texture_buffer == handle)
+        cur_state.lighting_lut.texture_buffer = 0;
+    if (cur_state.fog_lut.texture_buffer == handle)
+        cur_state.fog_lut.texture_buffer = 0;
+    if (cur_state.proctex_noise_lut.texture_1d == handle)
+        cur_state.proctex_noise_lut.texture_1d = 0;
+    if (cur_state.proctex_color_map.texture_1d == handle)
+        cur_state.proctex_color_map.texture_1d = 0;
+    if (cur_state.proctex_alpha_map.texture_1d == handle)
+        cur_state.proctex_alpha_map.texture_1d = 0;
+    if (cur_state.proctex_lut.texture_1d == handle)
+        cur_state.proctex_lut.texture_1d = 0;
+    if (cur_state.proctex_diff_lut.texture_1d == handle)
+        cur_state.proctex_diff_lut.texture_1d = 0;
 }
 
 void OpenGLState::ResetSampler(GLuint handle) {
diff --git a/src/video_core/renderer_opengl/gl_state.h b/src/video_core/renderer_opengl/gl_state.h
index 7dcc03bd5..1efcf0811 100644
--- a/src/video_core/renderer_opengl/gl_state.h
+++ b/src/video_core/renderer_opengl/gl_state.h
@@ -6,6 +6,29 @@
 
 #include <glad/glad.h>
 
+namespace TextureUnits {
+
+struct TextureUnit {
+    GLint id;
+    constexpr GLenum Enum() const {
+        return static_cast<GLenum>(GL_TEXTURE0 + id);
+    }
+};
+
+constexpr TextureUnit PicaTexture(int unit) {
+    return TextureUnit{unit};
+}
+
+constexpr TextureUnit LightingLUT{3};
+constexpr TextureUnit FogLUT{4};
+constexpr TextureUnit ProcTexNoiseLUT{5};
+constexpr TextureUnit ProcTexColorMap{6};
+constexpr TextureUnit ProcTexAlphaMap{7};
+constexpr TextureUnit ProcTexLUT{8};
+constexpr TextureUnit ProcTexDiffLUT{9};
+
+} // namespace TextureUnits
+
 class OpenGLState {
 public:
     struct {
@@ -64,11 +87,11 @@ public:
     } texture_units[3];
 
     struct {
-        GLuint texture_1d; // GL_TEXTURE_BINDING_1D
-    } lighting_luts[6];
+        GLuint texture_buffer; // GL_TEXTURE_BINDING_BUFFER
+    } lighting_lut;
 
     struct {
-        GLuint texture_1d; // GL_TEXTURE_BINDING_1D
+        GLuint texture_buffer; // GL_TEXTURE_BINDING_BUFFER
     } fog_lut;
 
     struct {
diff --git a/src/video_core/renderer_opengl/pica_to_gl.h b/src/video_core/renderer_opengl/pica_to_gl.h
index 93d7b0b71..70298e211 100644
--- a/src/video_core/renderer_opengl/pica_to_gl.h
+++ b/src/video_core/renderer_opengl/pica_to_gl.h
@@ -55,6 +55,12 @@ inline GLenum WrapMode(Pica::TexturingRegs::TextureConfig::WrapMode mode) {
         GL_CLAMP_TO_BORDER, // WrapMode::ClampToBorder
         GL_REPEAT,          // WrapMode::Repeat
         GL_MIRRORED_REPEAT, // WrapMode::MirroredRepeat
+        // TODO(wwylele): ClampToEdge2 and ClampToBorder2 are not properly implemented here. See the
+        // comments in enum WrapMode.
+        GL_CLAMP_TO_EDGE,   // WrapMode::ClampToEdge2
+        GL_CLAMP_TO_BORDER, // WrapMode::ClampToBorder2
+        GL_REPEAT,          // WrapMode::Repeat2
+        GL_REPEAT,          // WrapMode::Repeat3
     };
 
     // Range check table for input
@@ -65,6 +71,13 @@ inline GLenum WrapMode(Pica::TexturingRegs::TextureConfig::WrapMode mode) {
         return GL_CLAMP_TO_EDGE;
     }
 
+    if (static_cast<u32>(mode) > 3) {
+        // It is still unclear whether mode 4-7 are valid, so log it if a game uses them.
+        // TODO(wwylele): telemetry should be added here so we can collect more info about which
+        // game uses this.
+        LOG_WARNING(Render_OpenGL, "Using texture wrap mode %u", static_cast<u32>(mode));
+    }
+
     GLenum gl_mode = wrap_mode_table[mode];
 
     // Check for dummy values indicating an unknown mode
diff --git a/src/video_core/shader/shader_jit_x64_compiler.cpp b/src/video_core/shader/shader_jit_x64_compiler.cpp
index 5d9b6448c..42a57aab1 100644
--- a/src/video_core/shader/shader_jit_x64_compiler.cpp
+++ b/src/video_core/shader/shader_jit_x64_compiler.cpp
@@ -321,27 +321,27 @@ void JitShader::Compile_EvaluateCondition(Instruction instr) {
     case Instruction::FlowControlType::Or:
         mov(eax, COND0);
         mov(ebx, COND1);
-        xor(eax, (instr.flow_control.refx.Value() ^ 1));
-        xor(ebx, (instr.flow_control.refy.Value() ^ 1));
-        or (eax, ebx);
+        xor_(eax, (instr.flow_control.refx.Value() ^ 1));
+        xor_(ebx, (instr.flow_control.refy.Value() ^ 1));
+        or_(eax, ebx);
         break;
 
     case Instruction::FlowControlType::And:
         mov(eax, COND0);
         mov(ebx, COND1);
-        xor(eax, (instr.flow_control.refx.Value() ^ 1));
-        xor(ebx, (instr.flow_control.refy.Value() ^ 1));
-        and(eax, ebx);
+        xor_(eax, (instr.flow_control.refx.Value() ^ 1));
+        xor_(ebx, (instr.flow_control.refy.Value() ^ 1));
+        and_(eax, ebx);
         break;
 
     case Instruction::FlowControlType::JustX:
         mov(eax, COND0);
-        xor(eax, (instr.flow_control.refx.Value() ^ 1));
+        xor_(eax, (instr.flow_control.refx.Value() ^ 1));
         break;
 
     case Instruction::FlowControlType::JustY:
         mov(eax, COND1);
-        xor(eax, (instr.flow_control.refy.Value() ^ 1));
+        xor_(eax, (instr.flow_control.refy.Value() ^ 1));
         break;
     }
 }
@@ -734,10 +734,10 @@ void JitShader::Compile_LOOP(Instruction instr) {
     mov(LOOPCOUNT, dword[SETUP + offset]);
     mov(LOOPCOUNT_REG, LOOPCOUNT);
     shr(LOOPCOUNT_REG, 4);
-    and(LOOPCOUNT_REG, 0xFF0); // Y-component is the start
+    and_(LOOPCOUNT_REG, 0xFF0); // Y-component is the start
     mov(LOOPINC, LOOPCOUNT);
     shr(LOOPINC, 12);
-    and(LOOPINC, 0xFF0);                // Z-component is the incrementer
+    and_(LOOPINC, 0xFF0);               // Z-component is the incrementer
     movzx(LOOPCOUNT, LOOPCOUNT.cvt8()); // X-component is iteration count
     add(LOOPCOUNT, 1);                  // Iteration count is X-component + 1
 
@@ -858,9 +858,9 @@ void JitShader::Compile(const std::array<u32, MAX_PROGRAM_CODE_LENGTH>* program_
     mov(STATE, ABI_PARAM2);
 
     // Zero address/loop  registers
-    xor(ADDROFFS_REG_0.cvt32(), ADDROFFS_REG_0.cvt32());
-    xor(ADDROFFS_REG_1.cvt32(), ADDROFFS_REG_1.cvt32());
-    xor(LOOPCOUNT_REG, LOOPCOUNT_REG);
+    xor_(ADDROFFS_REG_0.cvt32(), ADDROFFS_REG_0.cvt32());
+    xor_(ADDROFFS_REG_1.cvt32(), ADDROFFS_REG_1.cvt32());
+    xor_(LOOPCOUNT_REG, LOOPCOUNT_REG);
 
     // Used to set a register to one
     static const __m128 one = {1.f, 1.f, 1.f, 1.f};
diff --git a/src/video_core/swrasterizer/rasterizer.cpp b/src/video_core/swrasterizer/rasterizer.cpp
index 8b7b1defb..512e81c08 100644
--- a/src/video_core/swrasterizer/rasterizer.cpp
+++ b/src/video_core/swrasterizer/rasterizer.cpp
@@ -357,10 +357,22 @@ static void ProcessTriangleInternal(const Vertex& v0, const Vertex& v1, const Ve
                 int t = (int)(v * float24::FromFloat32(static_cast<float>(texture.config.height)))
                             .ToFloat32();
 
-                if ((texture.config.wrap_s == TexturingRegs::TextureConfig::ClampToBorder &&
-                     (s < 0 || static_cast<u32>(s) >= texture.config.width)) ||
-                    (texture.config.wrap_t == TexturingRegs::TextureConfig::ClampToBorder &&
-                     (t < 0 || static_cast<u32>(t) >= texture.config.height))) {
+                bool use_border_s = false;
+                bool use_border_t = false;
+
+                if (texture.config.wrap_s == TexturingRegs::TextureConfig::ClampToBorder) {
+                    use_border_s = s < 0 || s >= static_cast<int>(texture.config.width);
+                } else if (texture.config.wrap_s == TexturingRegs::TextureConfig::ClampToBorder2) {
+                    use_border_s = s >= static_cast<int>(texture.config.width);
+                }
+
+                if (texture.config.wrap_t == TexturingRegs::TextureConfig::ClampToBorder) {
+                    use_border_t = t < 0 || t >= static_cast<int>(texture.config.height);
+                } else if (texture.config.wrap_t == TexturingRegs::TextureConfig::ClampToBorder2) {
+                    use_border_t = t >= static_cast<int>(texture.config.height);
+                }
+
+                if (use_border_s || use_border_t) {
                     auto border_color = texture.config.border_color;
                     texture_color[i] = {border_color.r, border_color.g, border_color.b,
                                         border_color.a};
@@ -572,8 +584,7 @@ static void ProcessTriangleInternal(const Vertex& v0, const Vertex& v1, const Ve
                 float fog_i = MathUtil::Clamp(floorf(fog_index), 0.0f, 127.0f);
                 float fog_f = fog_index - fog_i;
                 const auto& fog_lut_entry = g_state.fog.lut[static_cast<unsigned int>(fog_i)];
-                float fog_factor = (fog_lut_entry.value + fog_lut_entry.difference * fog_f) /
-                                   2047.0f; // This is signed fixed point 1.11
+                float fog_factor = fog_lut_entry.ToFloat() + fog_lut_entry.DiffToFloat() * fog_f;
                 fog_factor = MathUtil::Clamp(fog_factor, 0.0f, 1.0f);
 
                 // Blend the fog
diff --git a/src/video_core/swrasterizer/texturing.cpp b/src/video_core/swrasterizer/texturing.cpp
index aeb6aeb8c..4f02b93f2 100644
--- a/src/video_core/swrasterizer/texturing.cpp
+++ b/src/video_core/swrasterizer/texturing.cpp
@@ -18,22 +18,33 @@ using TevStageConfig = TexturingRegs::TevStageConfig;
 
 int GetWrappedTexCoord(TexturingRegs::TextureConfig::WrapMode mode, int val, unsigned size) {
     switch (mode) {
+    case TexturingRegs::TextureConfig::ClampToEdge2:
+        // For negative coordinate, ClampToEdge2 behaves the same as Repeat
+        if (val < 0) {
+            return static_cast<int>(static_cast<unsigned>(val) % size);
+        }
+    // [[fallthrough]]
     case TexturingRegs::TextureConfig::ClampToEdge:
         val = std::max(val, 0);
-        val = std::min(val, (int)size - 1);
+        val = std::min(val, static_cast<int>(size) - 1);
         return val;
 
     case TexturingRegs::TextureConfig::ClampToBorder:
         return val;
 
+    case TexturingRegs::TextureConfig::ClampToBorder2:
+    // For ClampToBorder2, the case of positive coordinate beyond the texture size is already
+    // handled outside. Here we only handle the negative coordinate in the same way as Repeat.
+    case TexturingRegs::TextureConfig::Repeat2:
+    case TexturingRegs::TextureConfig::Repeat3:
     case TexturingRegs::TextureConfig::Repeat:
-        return (int)((unsigned)val % size);
+        return static_cast<int>(static_cast<unsigned>(val) % size);
 
     case TexturingRegs::TextureConfig::MirroredRepeat: {
-        unsigned int coord = ((unsigned)val % (2 * size));
+        unsigned int coord = (static_cast<unsigned>(val) % (2 * size));
         if (coord >= size)
             coord = 2 * size - 1 - coord;
-        return (int)coord;
+        return static_cast<int>(coord);
     }
 
     default: