Merge pull request #3478 from bunnei/a32

Refactoring to boot A32 games
2020-03-04 20:37:51 -05:00 · 2020-03-04 20:37:51 -05:00 · 1e84d22275
parent 67e7186d79 dba112e510
commit 1e84d22275
26 changed files with 1068 additions and 235 deletions
--- a/src/core/CMakeLists.txt
+++ b/src/core/CMakeLists.txt
@ -595,8 +595,12 @@ endif()

 if (ARCHITECTURE_x86_64)
    target_sources(core PRIVATE
-        arm/dynarmic/arm_dynarmic.cpp
-        arm/dynarmic/arm_dynarmic.h
+        arm/dynarmic/arm_dynarmic_32.cpp
+        arm/dynarmic/arm_dynarmic_32.h
+        arm/dynarmic/arm_dynarmic_64.cpp
+        arm/dynarmic/arm_dynarmic_64.h
+        arm/dynarmic/arm_dynarmic_cp15.cpp
+        arm/dynarmic/arm_dynarmic_cp15.h
    )
    target_link_libraries(core PRIVATE dynarmic)
 endif()
--- a/src/core/arm/arm_interface.h
+++ b/src/core/arm/arm_interface.h
@ -25,7 +25,20 @@ public:
    explicit ARM_Interface(System& system_) : system{system_} {}
    virtual ~ARM_Interface() = default;

-    struct ThreadContext {
+    struct ThreadContext32 {
+        std::array<u32, 16> cpu_registers;
+        u32 cpsr;
+        std::array<u8, 4> padding;
+        std::array<u64, 32> fprs;
+        u32 fpscr;
+        u32 fpexc;
+        u32 tpidr;
+    };
+    // Internally within the kernel, it expects the AArch32 version of the
+    // thread context to be 344 bytes in size.
+    static_assert(sizeof(ThreadContext32) == 0x158);
+
+    struct ThreadContext64 {
        std::array<u64, 31> cpu_registers;
        u64 sp;
        u64 pc;
@ -38,7 +51,7 @@ public:
    };
    // Internally within the kernel, it expects the AArch64 version of the
    // thread context to be 800 bytes in size.
-    static_assert(sizeof(ThreadContext) == 0x320);
+    static_assert(sizeof(ThreadContext64) == 0x320);

    /// Runs the CPU until an event happens
    virtual void Run() = 0;
@ -130,17 +143,10 @@ public:
     */
    virtual void SetTPIDR_EL0(u64 value) = 0;

-    /**
-     * Saves the current CPU context
-     * @param ctx Thread context to save
-     */
-    virtual void SaveContext(ThreadContext& ctx) = 0;
-
-    /**
-     * Loads a CPU context
-     * @param ctx Thread context to load
-     */
-    virtual void LoadContext(const ThreadContext& ctx) = 0;
+    virtual void SaveContext(ThreadContext32& ctx) = 0;
+    virtual void SaveContext(ThreadContext64& ctx) = 0;
+    virtual void LoadContext(const ThreadContext32& ctx) = 0;
+    virtual void LoadContext(const ThreadContext64& ctx) = 0;

    /// Clears the exclusive monitor's state.
    virtual void ClearExclusiveState() = 0;
--- a/src/core/arm/dynarmic/arm_dynarmic_32.cpp
+++ b/src/core/arm/dynarmic/arm_dynarmic_32.cpp
@ -0,0 +1,208 @@
+// Copyright 2020 yuzu emulator team
+// Licensed under GPLv2 or any later version
+// Refer to the license.txt file included.
+
+#include <cinttypes>
+#include <memory>
+#include <dynarmic/A32/a32.h>
+#include <dynarmic/A32/config.h>
+#include <dynarmic/A32/context.h>
+#include "common/microprofile.h"
+#include "core/arm/dynarmic/arm_dynarmic_32.h"
+#include "core/arm/dynarmic/arm_dynarmic_64.h"
+#include "core/arm/dynarmic/arm_dynarmic_cp15.h"
+#include "core/core.h"
+#include "core/core_manager.h"
+#include "core/core_timing.h"
+#include "core/hle/kernel/svc.h"
+#include "core/memory.h"
+
+namespace Core {
+
+class DynarmicCallbacks32 : public Dynarmic::A32::UserCallbacks {
+public:
+    explicit DynarmicCallbacks32(ARM_Dynarmic_32& parent) : parent(parent) {}
+
+    u8 MemoryRead8(u32 vaddr) override {
+        return parent.system.Memory().Read8(vaddr);
+    }
+    u16 MemoryRead16(u32 vaddr) override {
+        return parent.system.Memory().Read16(vaddr);
+    }
+    u32 MemoryRead32(u32 vaddr) override {
+        return parent.system.Memory().Read32(vaddr);
+    }
+    u64 MemoryRead64(u32 vaddr) override {
+        return parent.system.Memory().Read64(vaddr);
+    }
+
+    void MemoryWrite8(u32 vaddr, u8 value) override {
+        parent.system.Memory().Write8(vaddr, value);
+    }
+    void MemoryWrite16(u32 vaddr, u16 value) override {
+        parent.system.Memory().Write16(vaddr, value);
+    }
+    void MemoryWrite32(u32 vaddr, u32 value) override {
+        parent.system.Memory().Write32(vaddr, value);
+    }
+    void MemoryWrite64(u32 vaddr, u64 value) override {
+        parent.system.Memory().Write64(vaddr, value);
+    }
+
+    void InterpreterFallback(u32 pc, std::size_t num_instructions) override {
+        UNIMPLEMENTED();
+    }
+
+    void ExceptionRaised(u32 pc, Dynarmic::A32::Exception exception) override {
+        switch (exception) {
+        case Dynarmic::A32::Exception::UndefinedInstruction:
+        case Dynarmic::A32::Exception::UnpredictableInstruction:
+            break;
+        case Dynarmic::A32::Exception::Breakpoint:
+            break;
+        }
+        LOG_CRITICAL(HW_GPU, "ExceptionRaised(exception = {}, pc = {:08X}, code = {:08X})",
+                     static_cast<std::size_t>(exception), pc, MemoryReadCode(pc));
+        UNIMPLEMENTED();
+    }
+
+    void CallSVC(u32 swi) override {
+        Kernel::CallSVC(parent.system, swi);
+    }
+
+    void AddTicks(u64 ticks) override {
+        // Divide the number of ticks by the amount of CPU cores. TODO(Subv): This yields only a
+        // rough approximation of the amount of executed ticks in the system, it may be thrown off
+        // if not all cores are doing a similar amount of work. Instead of doing this, we should
+        // device a way so that timing is consistent across all cores without increasing the ticks 4
+        // times.
+        u64 amortized_ticks = (ticks - num_interpreted_instructions) / Core::NUM_CPU_CORES;
+        // Always execute at least one tick.
+        amortized_ticks = std::max<u64>(amortized_ticks, 1);
+
+        parent.system.CoreTiming().AddTicks(amortized_ticks);
+        num_interpreted_instructions = 0;
+    }
+    u64 GetTicksRemaining() override {
+        return std::max(parent.system.CoreTiming().GetDowncount(), {});
+    }
+
+    ARM_Dynarmic_32& parent;
+    std::size_t num_interpreted_instructions{};
+    u64 tpidrro_el0{};
+    u64 tpidr_el0{};
+};
+
+std::shared_ptr<Dynarmic::A32::Jit> ARM_Dynarmic_32::MakeJit(Common::PageTable& page_table,
+                                                             std::size_t address_space_bits) const {
+    Dynarmic::A32::UserConfig config;
+    config.callbacks = cb.get();
+    // TODO(bunnei): Implement page table for 32-bit
+    // config.page_table = &page_table.pointers;
+    config.coprocessors[15] = std::make_shared<DynarmicCP15>((u32*)&CP15_regs[0]);
+    config.define_unpredictable_behaviour = true;
+    return std::make_unique<Dynarmic::A32::Jit>(config);
+}
+
+MICROPROFILE_DEFINE(ARM_Jit_Dynarmic_32, "ARM JIT", "Dynarmic", MP_RGB(255, 64, 64));
+
+void ARM_Dynarmic_32::Run() {
+    MICROPROFILE_SCOPE(ARM_Jit_Dynarmic_32);
+    jit->Run();
+}
+
+void ARM_Dynarmic_32::Step() {
+    cb->InterpreterFallback(jit->Regs()[15], 1);
+}
+
+ARM_Dynarmic_32::ARM_Dynarmic_32(System& system, ExclusiveMonitor& exclusive_monitor,
+                                 std::size_t core_index)
+    : ARM_Interface{system},
+      cb(std::make_unique<DynarmicCallbacks32>(*this)), core_index{core_index},
+      exclusive_monitor{dynamic_cast<DynarmicExclusiveMonitor&>(exclusive_monitor)} {}
+
+ARM_Dynarmic_32::~ARM_Dynarmic_32() = default;
+
+void ARM_Dynarmic_32::SetPC(u64 pc) {
+    jit->Regs()[15] = static_cast<u32>(pc);
+}
+
+u64 ARM_Dynarmic_32::GetPC() const {
+    return jit->Regs()[15];
+}
+
+u64 ARM_Dynarmic_32::GetReg(int index) const {
+    return jit->Regs()[index];
+}
+
+void ARM_Dynarmic_32::SetReg(int index, u64 value) {
+    jit->Regs()[index] = static_cast<u32>(value);
+}
+
+u128 ARM_Dynarmic_32::GetVectorReg(int index) const {
+    return {};
+}
+
+void ARM_Dynarmic_32::SetVectorReg(int index, u128 value) {}
+
+u32 ARM_Dynarmic_32::GetPSTATE() const {
+    return jit->Cpsr();
+}
+
+void ARM_Dynarmic_32::SetPSTATE(u32 cpsr) {
+    jit->SetCpsr(cpsr);
+}
+
+u64 ARM_Dynarmic_32::GetTlsAddress() const {
+    return CP15_regs[static_cast<std::size_t>(CP15Register::CP15_THREAD_URO)];
+}
+
+void ARM_Dynarmic_32::SetTlsAddress(VAddr address) {
+    CP15_regs[static_cast<std::size_t>(CP15Register::CP15_THREAD_URO)] = static_cast<u32>(address);
+}
+
+u64 ARM_Dynarmic_32::GetTPIDR_EL0() const {
+    return cb->tpidr_el0;
+}
+
+void ARM_Dynarmic_32::SetTPIDR_EL0(u64 value) {
+    cb->tpidr_el0 = value;
+}
+
+void ARM_Dynarmic_32::SaveContext(ThreadContext32& ctx) {
+    Dynarmic::A32::Context context;
+    jit->SaveContext(context);
+    ctx.cpu_registers = context.Regs();
+    ctx.cpsr = context.Cpsr();
+}
+
+void ARM_Dynarmic_32::LoadContext(const ThreadContext32& ctx) {
+    Dynarmic::A32::Context context;
+    context.Regs() = ctx.cpu_registers;
+    context.SetCpsr(ctx.cpsr);
+    jit->LoadContext(context);
+}
+
+void ARM_Dynarmic_32::PrepareReschedule() {
+    jit->HaltExecution();
+}
+
+void ARM_Dynarmic_32::ClearInstructionCache() {
+    jit->ClearCache();
+}
+
+void ARM_Dynarmic_32::ClearExclusiveState() {}
+
+void ARM_Dynarmic_32::PageTableChanged(Common::PageTable& page_table,
+                                       std::size_t new_address_space_size_in_bits) {
+    auto key = std::make_pair(&page_table, new_address_space_size_in_bits);
+    auto iter = jit_cache.find(key);
+    if (iter != jit_cache.end()) {
+        jit = iter->second;
+        return;
+    }
+    jit = MakeJit(page_table, new_address_space_size_in_bits);
+    jit_cache.emplace(key, jit);
+}
+
+} // namespace Core
--- a/src/core/arm/dynarmic/arm_dynarmic_32.h
+++ b/src/core/arm/dynarmic/arm_dynarmic_32.h
@ -0,0 +1,77 @@
+// Copyright 2020 yuzu emulator team
+// Licensed under GPLv2 or any later version
+// Refer to the license.txt file included.
+
+#pragma once
+
+#include <memory>
+#include <unordered_map>
+
+#include <dynarmic/A32/a32.h>
+#include <dynarmic/A64/a64.h>
+#include <dynarmic/A64/exclusive_monitor.h>
+#include "common/common_types.h"
+#include "common/hash.h"
+#include "core/arm/arm_interface.h"
+#include "core/arm/exclusive_monitor.h"
+
+namespace Memory {
+class Memory;
+}
+
+namespace Core {
+
+class DynarmicCallbacks32;
+class DynarmicExclusiveMonitor;
+class System;
+
+class ARM_Dynarmic_32 final : public ARM_Interface {
+public:
+    ARM_Dynarmic_32(System& system, ExclusiveMonitor& exclusive_monitor, std::size_t core_index);
+    ~ARM_Dynarmic_32() override;
+
+    void SetPC(u64 pc) override;
+    u64 GetPC() const override;
+    u64 GetReg(int index) const override;
+    void SetReg(int index, u64 value) override;
+    u128 GetVectorReg(int index) const override;
+    void SetVectorReg(int index, u128 value) override;
+    u32 GetPSTATE() const override;
+    void SetPSTATE(u32 pstate) override;
+    void Run() override;
+    void Step() override;
+    VAddr GetTlsAddress() const override;
+    void SetTlsAddress(VAddr address) override;
+    void SetTPIDR_EL0(u64 value) override;
+    u64 GetTPIDR_EL0() const override;
+
+    void SaveContext(ThreadContext32& ctx) override;
+    void SaveContext(ThreadContext64& ctx) override {}
+    void LoadContext(const ThreadContext32& ctx) override;
+    void LoadContext(const ThreadContext64& ctx) override {}
+
+    void PrepareReschedule() override;
+    void ClearExclusiveState() override;
+
+    void ClearInstructionCache() override;
+    void PageTableChanged(Common::PageTable& new_page_table,
+                          std::size_t new_address_space_size_in_bits) override;
+
+private:
+    std::shared_ptr<Dynarmic::A32::Jit> MakeJit(Common::PageTable& page_table,
+                                                std::size_t address_space_bits) const;
+
+    using JitCacheKey = std::pair<Common::PageTable*, std::size_t>;
+    using JitCacheType =
+        std::unordered_map<JitCacheKey, std::shared_ptr<Dynarmic::A32::Jit>, Common::PairHash>;
+
+    friend class DynarmicCallbacks32;
+    std::unique_ptr<DynarmicCallbacks32> cb;
+    JitCacheType jit_cache;
+    std::shared_ptr<Dynarmic::A32::Jit> jit;
+    std::size_t core_index;
+    DynarmicExclusiveMonitor& exclusive_monitor;
+    std::array<u32, 84> CP15_regs{};
+};
+
+} // namespace Core
--- a/src/core/arm/dynarmic/arm_dynarmic_64.cpp
+++ b/src/core/arm/dynarmic/arm_dynarmic_64.cpp
@ -8,7 +8,7 @@
 #include <dynarmic/A64/config.h>
 #include "common/logging/log.h"
 #include "common/microprofile.h"
-#include "core/arm/dynarmic/arm_dynarmic.h"
+#include "core/arm/dynarmic/arm_dynarmic_64.h"
 #include "core/core.h"
 #include "core/core_manager.h"
 #include "core/core_timing.h"
@ -25,9 +25,9 @@ namespace Core {

 using Vector = Dynarmic::A64::Vector;

-class ARM_Dynarmic_Callbacks : public Dynarmic::A64::UserCallbacks {
+class DynarmicCallbacks64 : public Dynarmic::A64::UserCallbacks {
 public:
-    explicit ARM_Dynarmic_Callbacks(ARM_Dynarmic& parent) : parent(parent) {}
+    explicit DynarmicCallbacks64(ARM_Dynarmic_64& parent) : parent(parent) {}

    u8 MemoryRead8(u64 vaddr) override {
        return parent.system.Memory().Read8(vaddr);
@ -68,7 +68,7 @@ public:
        LOG_INFO(Core_ARM, "Unicorn fallback @ 0x{:X} for {} instructions (instr = {:08X})", pc,
                 num_instructions, MemoryReadCode(pc));

-        ARM_Interface::ThreadContext ctx;
+        ARM_Interface::ThreadContext64 ctx;
        parent.SaveContext(ctx);
        parent.inner_unicorn.LoadContext(ctx);
        parent.inner_unicorn.ExecuteInstructions(num_instructions);
@ -90,7 +90,7 @@ public:
                parent.jit->HaltExecution();
                parent.SetPC(pc);
                Kernel::Thread* const thread = parent.system.CurrentScheduler().GetCurrentThread();
-                parent.SaveContext(thread->GetContext());
+                parent.SaveContext(thread->GetContext64());
                GDBStub::Break();
                GDBStub::SendTrap(thread, 5);
                return;
@ -126,14 +126,14 @@ public:
        return Timing::CpuCyclesToClockCycles(parent.system.CoreTiming().GetTicks());
    }

-    ARM_Dynarmic& parent;
+    ARM_Dynarmic_64& parent;
    std::size_t num_interpreted_instructions = 0;
    u64 tpidrro_el0 = 0;
    u64 tpidr_el0 = 0;
 };

-std::shared_ptr<Dynarmic::A64::Jit> ARM_Dynarmic::MakeJit(Common::PageTable& page_table,
-                                                          std::size_t address_space_bits) const {
+std::shared_ptr<Dynarmic::A64::Jit> ARM_Dynarmic_64::MakeJit(Common::PageTable& page_table,
+                                                             std::size_t address_space_bits) const {
    Dynarmic::A64::UserConfig config;

    // Callbacks
@ -162,76 +162,76 @@ std::shared_ptr<Dynarmic::A64::Jit> ARM_Dynarmic::MakeJit(Common::PageTable& pag
    return std::make_shared<Dynarmic::A64::Jit>(config);
 }

-MICROPROFILE_DEFINE(ARM_Jit_Dynarmic, "ARM JIT", "Dynarmic", MP_RGB(255, 64, 64));
+MICROPROFILE_DEFINE(ARM_Jit_Dynarmic_64, "ARM JIT", "Dynarmic", MP_RGB(255, 64, 64));

-void ARM_Dynarmic::Run() {
-    MICROPROFILE_SCOPE(ARM_Jit_Dynarmic);
+void ARM_Dynarmic_64::Run() {
+    MICROPROFILE_SCOPE(ARM_Jit_Dynarmic_64);

    jit->Run();
 }

-void ARM_Dynarmic::Step() {
+void ARM_Dynarmic_64::Step() {
    cb->InterpreterFallback(jit->GetPC(), 1);
 }

-ARM_Dynarmic::ARM_Dynarmic(System& system, ExclusiveMonitor& exclusive_monitor,
-                           std::size_t core_index)
+ARM_Dynarmic_64::ARM_Dynarmic_64(System& system, ExclusiveMonitor& exclusive_monitor,
+                                 std::size_t core_index)
    : ARM_Interface{system},
-      cb(std::make_unique<ARM_Dynarmic_Callbacks>(*this)), inner_unicorn{system},
+      cb(std::make_unique<DynarmicCallbacks64>(*this)), inner_unicorn{system},
      core_index{core_index}, exclusive_monitor{
                                  dynamic_cast<DynarmicExclusiveMonitor&>(exclusive_monitor)} {}

-ARM_Dynarmic::~ARM_Dynarmic() = default;
+ARM_Dynarmic_64::~ARM_Dynarmic_64() = default;

-void ARM_Dynarmic::SetPC(u64 pc) {
+void ARM_Dynarmic_64::SetPC(u64 pc) {
    jit->SetPC(pc);
 }

-u64 ARM_Dynarmic::GetPC() const {
+u64 ARM_Dynarmic_64::GetPC() const {
    return jit->GetPC();
 }

-u64 ARM_Dynarmic::GetReg(int index) const {
+u64 ARM_Dynarmic_64::GetReg(int index) const {
    return jit->GetRegister(index);
 }

-void ARM_Dynarmic::SetReg(int index, u64 value) {
+void ARM_Dynarmic_64::SetReg(int index, u64 value) {
    jit->SetRegister(index, value);
 }

-u128 ARM_Dynarmic::GetVectorReg(int index) const {
+u128 ARM_Dynarmic_64::GetVectorReg(int index) const {
    return jit->GetVector(index);
 }

-void ARM_Dynarmic::SetVectorReg(int index, u128 value) {
+void ARM_Dynarmic_64::SetVectorReg(int index, u128 value) {
    jit->SetVector(index, value);
 }

-u32 ARM_Dynarmic::GetPSTATE() const {
+u32 ARM_Dynarmic_64::GetPSTATE() const {
    return jit->GetPstate();
 }

-void ARM_Dynarmic::SetPSTATE(u32 pstate) {
+void ARM_Dynarmic_64::SetPSTATE(u32 pstate) {
    jit->SetPstate(pstate);
 }

-u64 ARM_Dynarmic::GetTlsAddress() const {
+u64 ARM_Dynarmic_64::GetTlsAddress() const {
    return cb->tpidrro_el0;
 }

-void ARM_Dynarmic::SetTlsAddress(VAddr address) {
+void ARM_Dynarmic_64::SetTlsAddress(VAddr address) {
    cb->tpidrro_el0 = address;
 }

-u64 ARM_Dynarmic::GetTPIDR_EL0() const {
+u64 ARM_Dynarmic_64::GetTPIDR_EL0() const {
    return cb->tpidr_el0;
 }

-void ARM_Dynarmic::SetTPIDR_EL0(u64 value) {
+void ARM_Dynarmic_64::SetTPIDR_EL0(u64 value) {
    cb->tpidr_el0 = value;
 }

-void ARM_Dynarmic::SaveContext(ThreadContext& ctx) {
+void ARM_Dynarmic_64::SaveContext(ThreadContext64& ctx) {
    ctx.cpu_registers = jit->GetRegisters();
    ctx.sp = jit->GetSP();
    ctx.pc = jit->GetPC();
@ -242,7 +242,7 @@ void ARM_Dynarmic::SaveContext(ThreadContext& ctx) {
    ctx.tpidr = cb->tpidr_el0;
 }

-void ARM_Dynarmic::LoadContext(const ThreadContext& ctx) {
+void ARM_Dynarmic_64::LoadContext(const ThreadContext64& ctx) {
    jit->SetRegisters(ctx.cpu_registers);
    jit->SetSP(ctx.sp);
    jit->SetPC(ctx.pc);
@ -253,20 +253,20 @@ void ARM_Dynarmic::LoadContext(const ThreadContext& ctx) {
    SetTPIDR_EL0(ctx.tpidr);
 }

-void ARM_Dynarmic::PrepareReschedule() {
+void ARM_Dynarmic_64::PrepareReschedule() {
    jit->HaltExecution();
 }

-void ARM_Dynarmic::ClearInstructionCache() {
+void ARM_Dynarmic_64::ClearInstructionCache() {
    jit->ClearCache();
 }

-void ARM_Dynarmic::ClearExclusiveState() {
+void ARM_Dynarmic_64::ClearExclusiveState() {
    jit->ClearExclusiveState();
 }

-void ARM_Dynarmic::PageTableChanged(Common::PageTable& page_table,
-                                    std::size_t new_address_space_size_in_bits) {
+void ARM_Dynarmic_64::PageTableChanged(Common::PageTable& page_table,
+                                       std::size_t new_address_space_size_in_bits) {
    auto key = std::make_pair(&page_table, new_address_space_size_in_bits);
    auto iter = jit_cache.find(key);
    if (iter != jit_cache.end()) {
@ -277,8 +277,8 @@ void ARM_Dynarmic::PageTableChanged(Common::PageTable& page_table,
    jit_cache.emplace(key, jit);
 }

-DynarmicExclusiveMonitor::DynarmicExclusiveMonitor(Memory::Memory& memory_, std::size_t core_count)
-    : monitor(core_count), memory{memory_} {}
+DynarmicExclusiveMonitor::DynarmicExclusiveMonitor(Memory::Memory& memory, std::size_t core_count)
+    : monitor(core_count), memory{memory} {}

 DynarmicExclusiveMonitor::~DynarmicExclusiveMonitor() = default;

--- a/src/core/arm/dynarmic/arm_dynarmic_64.h
+++ b/src/core/arm/dynarmic/arm_dynarmic_64.h
@ -21,18 +21,14 @@ class Memory;

 namespace Core {

-class ARM_Dynarmic_Callbacks;
+class DynarmicCallbacks64;
 class DynarmicExclusiveMonitor;
 class System;

-using JitCacheKey = std::pair<Common::PageTable*, std::size_t>;
-using JitCacheType =
-    std::unordered_map<JitCacheKey, std::shared_ptr<Dynarmic::A64::Jit>, Common::PairHash>;
-
-class ARM_Dynarmic final : public ARM_Interface {
+class ARM_Dynarmic_64 final : public ARM_Interface {
 public:
-    ARM_Dynarmic(System& system, ExclusiveMonitor& exclusive_monitor, std::size_t core_index);
-    ~ARM_Dynarmic() override;
+    ARM_Dynarmic_64(System& system, ExclusiveMonitor& exclusive_monitor, std::size_t core_index);
+    ~ARM_Dynarmic_64() override;

    void SetPC(u64 pc) override;
    u64 GetPC() const override;
@ -49,8 +45,10 @@ public:
    void SetTPIDR_EL0(u64 value) override;
    u64 GetTPIDR_EL0() const override;

-    void SaveContext(ThreadContext& ctx) override;
-    void LoadContext(const ThreadContext& ctx) override;
+    void SaveContext(ThreadContext32& ctx) override {}
+    void SaveContext(ThreadContext64& ctx) override;
+    void LoadContext(const ThreadContext32& ctx) override {}
+    void LoadContext(const ThreadContext64& ctx) override;

    void PrepareReschedule() override;
    void ClearExclusiveState() override;
@ -63,8 +61,12 @@ private:
    std::shared_ptr<Dynarmic::A64::Jit> MakeJit(Common::PageTable& page_table,
                                                std::size_t address_space_bits) const;

-    friend class ARM_Dynarmic_Callbacks;
-    std::unique_ptr<ARM_Dynarmic_Callbacks> cb;
+    using JitCacheKey = std::pair<Common::PageTable*, std::size_t>;
+    using JitCacheType =
+        std::unordered_map<JitCacheKey, std::shared_ptr<Dynarmic::A64::Jit>, Common::PairHash>;
+
+    friend class DynarmicCallbacks64;
+    std::unique_ptr<DynarmicCallbacks64> cb;
    JitCacheType jit_cache;
    std::shared_ptr<Dynarmic::A64::Jit> jit;
    ARM_Unicorn inner_unicorn;
@ -75,7 +77,7 @@ private:

 class DynarmicExclusiveMonitor final : public ExclusiveMonitor {
 public:
-    explicit DynarmicExclusiveMonitor(Memory::Memory& memory_, std::size_t core_count);
+    explicit DynarmicExclusiveMonitor(Memory::Memory& memory, std::size_t core_count);
    ~DynarmicExclusiveMonitor() override;

    void SetExclusive(std::size_t core_index, VAddr addr) override;
@ -88,7 +90,7 @@ public:
    bool ExclusiveWrite128(std::size_t core_index, VAddr vaddr, u128 value) override;

 private:
-    friend class ARM_Dynarmic;
+    friend class ARM_Dynarmic_64;
    Dynarmic::A64::ExclusiveMonitor monitor;
    Memory::Memory& memory;
 };
--- a/src/core/arm/dynarmic/arm_dynarmic_cp15.cpp
+++ b/src/core/arm/dynarmic/arm_dynarmic_cp15.cpp
@ -0,0 +1,80 @@
+// Copyright 2017 Citra Emulator Project
+// Licensed under GPLv2 or any later version
+// Refer to the license.txt file included.
+
+#include "core/arm/dynarmic/arm_dynarmic_cp15.h"
+
+using Callback = Dynarmic::A32::Coprocessor::Callback;
+using CallbackOrAccessOneWord = Dynarmic::A32::Coprocessor::CallbackOrAccessOneWord;
+using CallbackOrAccessTwoWords = Dynarmic::A32::Coprocessor::CallbackOrAccessTwoWords;
+
+std::optional<Callback> DynarmicCP15::CompileInternalOperation(bool two, unsigned opc1,
+                                                               CoprocReg CRd, CoprocReg CRn,
+                                                               CoprocReg CRm, unsigned opc2) {
+    return {};
+}
+
+CallbackOrAccessOneWord DynarmicCP15::CompileSendOneWord(bool two, unsigned opc1, CoprocReg CRn,
+                                                         CoprocReg CRm, unsigned opc2) {
+    // TODO(merry): Privileged CP15 registers
+
+    if (!two && CRn == CoprocReg::C7 && opc1 == 0 && CRm == CoprocReg::C5 && opc2 == 4) {
+        // This is a dummy write, we ignore the value written here.
+        return &CP15[static_cast<std::size_t>(CP15Register::CP15_FLUSH_PREFETCH_BUFFER)];
+    }
+
+    if (!two && CRn == CoprocReg::C7 && opc1 == 0 && CRm == CoprocReg::C10) {
+        switch (opc2) {
+        case 4:
+            // This is a dummy write, we ignore the value written here.
+            return &CP15[static_cast<std::size_t>(CP15Register::CP15_DATA_SYNC_BARRIER)];
+        case 5:
+            // This is a dummy write, we ignore the value written here.
+            return &CP15[static_cast<std::size_t>(CP15Register::CP15_DATA_MEMORY_BARRIER)];
+        default:
+            return {};
+        }
+    }
+
+    if (!two && CRn == CoprocReg::C13 && opc1 == 0 && CRm == CoprocReg::C0 && opc2 == 2) {
+        return &CP15[static_cast<std::size_t>(CP15Register::CP15_THREAD_UPRW)];
+    }
+
+    return {};
+}
+
+CallbackOrAccessTwoWords DynarmicCP15::CompileSendTwoWords(bool two, unsigned opc, CoprocReg CRm) {
+    return {};
+}
+
+CallbackOrAccessOneWord DynarmicCP15::CompileGetOneWord(bool two, unsigned opc1, CoprocReg CRn,
+                                                        CoprocReg CRm, unsigned opc2) {
+    // TODO(merry): Privileged CP15 registers
+
+    if (!two && CRn == CoprocReg::C13 && opc1 == 0 && CRm == CoprocReg::C0) {
+        switch (opc2) {
+        case 2:
+            return &CP15[static_cast<std::size_t>(CP15Register::CP15_THREAD_UPRW)];
+        case 3:
+            return &CP15[static_cast<std::size_t>(CP15Register::CP15_THREAD_URO)];
+        default:
+            return {};
+        }
+    }
+
+    return {};
+}
+
+CallbackOrAccessTwoWords DynarmicCP15::CompileGetTwoWords(bool two, unsigned opc, CoprocReg CRm) {
+    return {};
+}
+
+std::optional<Callback> DynarmicCP15::CompileLoadWords(bool two, bool long_transfer, CoprocReg CRd,
+                                                       std::optional<u8> option) {
+    return {};
+}
+
+std::optional<Callback> DynarmicCP15::CompileStoreWords(bool two, bool long_transfer, CoprocReg CRd,
+                                                        std::optional<u8> option) {
+    return {};
+}
--- a/src/core/arm/dynarmic/arm_dynarmic_cp15.h
+++ b/src/core/arm/dynarmic/arm_dynarmic_cp15.h
@ -0,0 +1,152 @@
+// Copyright 2017 Citra Emulator Project
+// Licensed under GPLv2 or any later version
+// Refer to the license.txt file included.
+
+#pragma once
+
+#include <memory>
+#include <optional>
+
+#include <dynarmic/A32/coprocessor.h>
+#include "common/common_types.h"
+
+enum class CP15Register {
+    // c0 - Information registers
+    CP15_MAIN_ID,
+    CP15_CACHE_TYPE,
+    CP15_TCM_STATUS,
+    CP15_TLB_TYPE,
+    CP15_CPU_ID,
+    CP15_PROCESSOR_FEATURE_0,
+    CP15_PROCESSOR_FEATURE_1,
+    CP15_DEBUG_FEATURE_0,
+    CP15_AUXILIARY_FEATURE_0,
+    CP15_MEMORY_MODEL_FEATURE_0,
+    CP15_MEMORY_MODEL_FEATURE_1,
+    CP15_MEMORY_MODEL_FEATURE_2,
+    CP15_MEMORY_MODEL_FEATURE_3,
+    CP15_ISA_FEATURE_0,
+    CP15_ISA_FEATURE_1,
+    CP15_ISA_FEATURE_2,
+    CP15_ISA_FEATURE_3,
+    CP15_ISA_FEATURE_4,
+
+    // c1 - Control registers
+    CP15_CONTROL,
+    CP15_AUXILIARY_CONTROL,
+    CP15_COPROCESSOR_ACCESS_CONTROL,
+
+    // c2 - Translation table registers
+    CP15_TRANSLATION_BASE_TABLE_0,
+    CP15_TRANSLATION_BASE_TABLE_1,
+    CP15_TRANSLATION_BASE_CONTROL,
+    CP15_DOMAIN_ACCESS_CONTROL,
+    CP15_RESERVED,
+
+    // c5 - Fault status registers
+    CP15_FAULT_STATUS,
+    CP15_INSTR_FAULT_STATUS,
+    CP15_COMBINED_DATA_FSR = CP15_FAULT_STATUS,
+    CP15_INST_FSR,
+
+    // c6 - Fault Address registers
+    CP15_FAULT_ADDRESS,
+    CP15_COMBINED_DATA_FAR = CP15_FAULT_ADDRESS,
+    CP15_WFAR,
+    CP15_IFAR,
+
+    // c7 - Cache operation registers
+    CP15_WAIT_FOR_INTERRUPT,
+    CP15_PHYS_ADDRESS,
+    CP15_INVALIDATE_INSTR_CACHE,
+    CP15_INVALIDATE_INSTR_CACHE_USING_MVA,
+    CP15_INVALIDATE_INSTR_CACHE_USING_INDEX,
+    CP15_FLUSH_PREFETCH_BUFFER,
+    CP15_FLUSH_BRANCH_TARGET_CACHE,
+    CP15_FLUSH_BRANCH_TARGET_CACHE_ENTRY,
+    CP15_INVALIDATE_DATA_CACHE,
+    CP15_INVALIDATE_DATA_CACHE_LINE_USING_MVA,
+    CP15_INVALIDATE_DATA_CACHE_LINE_USING_INDEX,
+    CP15_INVALIDATE_DATA_AND_INSTR_CACHE,
+    CP15_CLEAN_DATA_CACHE,
+    CP15_CLEAN_DATA_CACHE_LINE_USING_MVA,
+    CP15_CLEAN_DATA_CACHE_LINE_USING_INDEX,
+    CP15_DATA_SYNC_BARRIER,
+    CP15_DATA_MEMORY_BARRIER,
+    CP15_CLEAN_AND_INVALIDATE_DATA_CACHE,
+    CP15_CLEAN_AND_INVALIDATE_DATA_CACHE_LINE_USING_MVA,
+    CP15_CLEAN_AND_INVALIDATE_DATA_CACHE_LINE_USING_INDEX,
+
+    // c8 - TLB operations
+    CP15_INVALIDATE_ITLB,
+    CP15_INVALIDATE_ITLB_SINGLE_ENTRY,
+    CP15_INVALIDATE_ITLB_ENTRY_ON_ASID_MATCH,
+    CP15_INVALIDATE_ITLB_ENTRY_ON_MVA,
+    CP15_INVALIDATE_DTLB,
+    CP15_INVALIDATE_DTLB_SINGLE_ENTRY,
+    CP15_INVALIDATE_DTLB_ENTRY_ON_ASID_MATCH,
+    CP15_INVALIDATE_DTLB_ENTRY_ON_MVA,
+    CP15_INVALIDATE_UTLB,
+    CP15_INVALIDATE_UTLB_SINGLE_ENTRY,
+    CP15_INVALIDATE_UTLB_ENTRY_ON_ASID_MATCH,
+    CP15_INVALIDATE_UTLB_ENTRY_ON_MVA,
+
+    // c9 - Data cache lockdown register
+    CP15_DATA_CACHE_LOCKDOWN,
+
+    // c10 - TLB/Memory map registers
+    CP15_TLB_LOCKDOWN,
+    CP15_PRIMARY_REGION_REMAP,
+    CP15_NORMAL_REGION_REMAP,
+
+    // c13 - Thread related registers
+    CP15_PID,
+    CP15_CONTEXT_ID,
+    CP15_THREAD_UPRW, // Thread ID register - User/Privileged Read/Write
+    CP15_THREAD_URO,  // Thread ID register - User Read Only (Privileged R/W)
+    CP15_THREAD_PRW,  // Thread ID register - Privileged R/W only.
+
+    // c15 - Performance and TLB lockdown registers
+    CP15_PERFORMANCE_MONITOR_CONTROL,
+    CP15_CYCLE_COUNTER,
+    CP15_COUNT_0,
+    CP15_COUNT_1,
+    CP15_READ_MAIN_TLB_LOCKDOWN_ENTRY,
+    CP15_WRITE_MAIN_TLB_LOCKDOWN_ENTRY,
+    CP15_MAIN_TLB_LOCKDOWN_VIRT_ADDRESS,
+    CP15_MAIN_TLB_LOCKDOWN_PHYS_ADDRESS,
+    CP15_MAIN_TLB_LOCKDOWN_ATTRIBUTE,
+    CP15_TLB_DEBUG_CONTROL,
+
+    // Skyeye defined
+    CP15_TLB_FAULT_ADDR,
+    CP15_TLB_FAULT_STATUS,
+
+    // Not an actual register.
+    // All registers should be defined above this.
+    CP15_REGISTER_COUNT,
+};
+
+class DynarmicCP15 final : public Dynarmic::A32::Coprocessor {
+public:
+    using CoprocReg = Dynarmic::A32::CoprocReg;
+
+    explicit DynarmicCP15(u32* cp15) : CP15(cp15){};
+
+    std::optional<Callback> CompileInternalOperation(bool two, unsigned opc1, CoprocReg CRd,
+                                                     CoprocReg CRn, CoprocReg CRm,
+                                                     unsigned opc2) override;
+    CallbackOrAccessOneWord CompileSendOneWord(bool two, unsigned opc1, CoprocReg CRn,
+                                               CoprocReg CRm, unsigned opc2) override;
+    CallbackOrAccessTwoWords CompileSendTwoWords(bool two, unsigned opc, CoprocReg CRm) override;
+    CallbackOrAccessOneWord CompileGetOneWord(bool two, unsigned opc1, CoprocReg CRn, CoprocReg CRm,
+                                              unsigned opc2) override;
+    CallbackOrAccessTwoWords CompileGetTwoWords(bool two, unsigned opc, CoprocReg CRm) override;
+    std::optional<Callback> CompileLoadWords(bool two, bool long_transfer, CoprocReg CRd,
+                                             std::optional<u8> option) override;
+    std::optional<Callback> CompileStoreWords(bool two, bool long_transfer, CoprocReg CRd,
+                                              std::optional<u8> option) override;
+
+private:
+    u32* CP15{};
+};
--- a/src/core/arm/exclusive_monitor.cpp
+++ b/src/core/arm/exclusive_monitor.cpp
@ -3,7 +3,7 @@
 // Refer to the license.txt file included.

 #ifdef ARCHITECTURE_x86_64
-#include "core/arm/dynarmic/arm_dynarmic.h"
+#include "core/arm/dynarmic/arm_dynarmic_64.h"
 #endif
 #include "core/arm/exclusive_monitor.h"
 #include "core/memory.h"
--- a/src/core/arm/unicorn/arm_unicorn.cpp
+++ b/src/core/arm/unicorn/arm_unicorn.cpp
@ -53,7 +53,7 @@ static bool UnmappedMemoryHook(uc_engine* uc, uc_mem_type type, u64 addr, int si
                               void* user_data) {
    auto* const system = static_cast<System*>(user_data);

-    ARM_Interface::ThreadContext ctx{};
+    ARM_Interface::ThreadContext64 ctx{};
    system->CurrentArmInterface().SaveContext(ctx);
    ASSERT_MSG(false, "Attempted to read from unmapped memory: 0x{:X}, pc=0x{:X}, lr=0x{:X}", addr,
               ctx.pc, ctx.cpu_registers[30]);
@ -179,7 +179,7 @@ void ARM_Unicorn::ExecuteInstructions(std::size_t num_instructions) {
        }

        Kernel::Thread* const thread = system.CurrentScheduler().GetCurrentThread();
-        SaveContext(thread->GetContext());
+        SaveContext(thread->GetContext64());
        if (last_bkpt_hit || GDBStub::IsMemoryBreak() || GDBStub::GetCpuStepFlag()) {
            last_bkpt_hit = false;
            GDBStub::Break();
@ -188,7 +188,7 @@ void ARM_Unicorn::ExecuteInstructions(std::size_t num_instructions) {
    }
 }

-void ARM_Unicorn::SaveContext(ThreadContext& ctx) {
+void ARM_Unicorn::SaveContext(ThreadContext64& ctx) {
    int uregs[32];
    void* tregs[32];

@ -215,7 +215,7 @@ void ARM_Unicorn::SaveContext(ThreadContext& ctx) {
    CHECKED(uc_reg_read_batch(uc, uregs, tregs, 32));
 }

-void ARM_Unicorn::LoadContext(const ThreadContext& ctx) {
+void ARM_Unicorn::LoadContext(const ThreadContext64& ctx) {
    int uregs[32];
    void* tregs[32];

--- a/src/core/arm/unicorn/arm_unicorn.h
+++ b/src/core/arm/unicorn/arm_unicorn.h
@ -30,8 +30,6 @@ public:
    void SetTlsAddress(VAddr address) override;
    void SetTPIDR_EL0(u64 value) override;
    u64 GetTPIDR_EL0() const override;
-    void SaveContext(ThreadContext& ctx) override;
-    void LoadContext(const ThreadContext& ctx) override;
    void PrepareReschedule() override;
    void ClearExclusiveState() override;
    void ExecuteInstructions(std::size_t num_instructions);
@ -41,6 +39,11 @@ public:
    void PageTableChanged(Common::PageTable&, std::size_t) override {}
    void RecordBreak(GDBStub::BreakpointAddress bkpt);

+    void SaveContext(ThreadContext32& ctx) override {}
+    void SaveContext(ThreadContext64& ctx) override;
+    void LoadContext(const ThreadContext32& ctx) override {}
+    void LoadContext(const ThreadContext64& ctx) override;
+
 private:
    static void InterruptHook(uc_engine* uc, u32 int_no, void* user_data);

--- a/src/core/core_manager.cpp
+++ b/src/core/core_manager.cpp
@ -6,9 +6,6 @@
 #include <mutex>

 #include "common/logging/log.h"
-#ifdef ARCHITECTURE_x86_64
-#include "core/arm/dynarmic/arm_dynarmic.h"
-#endif
 #include "core/arm/exclusive_monitor.h"
 #include "core/arm/unicorn/arm_unicorn.h"
 #include "core/core.h"
--- a/src/core/gdbstub/gdbstub.cpp
+++ b/src/core/gdbstub/gdbstub.cpp
@ -217,7 +217,7 @@ static u64 RegRead(std::size_t id, Kernel::Thread* thread = nullptr) {
        return 0;
    }

-    const auto& thread_context = thread->GetContext();
+    const auto& thread_context = thread->GetContext64();

    if (id < SP_REGISTER) {
        return thread_context.cpu_registers[id];
@ -239,7 +239,7 @@ static void RegWrite(std::size_t id, u64 val, Kernel::Thread* thread = nullptr)
        return;
    }

-    auto& thread_context = thread->GetContext();
+    auto& thread_context = thread->GetContext64();

    if (id < SP_REGISTER) {
        thread_context.cpu_registers[id] = val;
@ -259,7 +259,7 @@ static u128 FpuRead(std::size_t id, Kernel::Thread* thread = nullptr) {
        return u128{0};
    }

-    auto& thread_context = thread->GetContext();
+    auto& thread_context = thread->GetContext64();

    if (id >= UC_ARM64_REG_Q0 && id < FPCR_REGISTER) {
        return thread_context.vector_registers[id - UC_ARM64_REG_Q0];
@ -275,7 +275,7 @@ static void FpuWrite(std::size_t id, u128 val, Kernel::Thread* thread = nullptr)
        return;
    }

-    auto& thread_context = thread->GetContext();
+    auto& thread_context = thread->GetContext64();

    if (id >= UC_ARM64_REG_Q0 && id < FPCR_REGISTER) {
        thread_context.vector_registers[id - UC_ARM64_REG_Q0] = val;
@ -916,7 +916,7 @@ static void WriteRegister() {
    // Update ARM context, skipping scheduler - no running threads at this point
    Core::System::GetInstance()
        .ArmInterface(current_core)
-        .LoadContext(current_thread->GetContext());
+        .LoadContext(current_thread->GetContext64());

    SendReply("OK");
 }
@ -947,7 +947,7 @@ static void WriteRegisters() {
    // Update ARM context, skipping scheduler - no running threads at this point
    Core::System::GetInstance()
        .ArmInterface(current_core)
-        .LoadContext(current_thread->GetContext());
+        .LoadContext(current_thread->GetContext64());

    SendReply("OK");
 }
@ -1019,7 +1019,7 @@ static void Step() {
        // Update ARM context, skipping scheduler - no running threads at this point
        Core::System::GetInstance()
            .ArmInterface(current_core)
-            .LoadContext(current_thread->GetContext());
+            .LoadContext(current_thread->GetContext64());
    }
    step_loop = true;
    halt_loop = true;
--- a/src/core/hle/kernel/kernel.cpp
+++ b/src/core/hle/kernel/kernel.cpp
@ -186,6 +186,10 @@ struct KernelCore::Impl {
            return;
        }

+        for (auto& core : cores) {
+            core.SetIs64Bit(process->Is64BitProcess());
+        }
+
        system.Memory().SetCurrentPageTable(*process);
    }

--- a/src/core/hle/kernel/physical_core.cpp
+++ b/src/core/hle/kernel/physical_core.cpp
@ -5,7 +5,8 @@
 #include "common/logging/log.h"
 #include "core/arm/arm_interface.h"
 #ifdef ARCHITECTURE_x86_64
-#include "core/arm/dynarmic/arm_dynarmic.h"
+#include "core/arm/dynarmic/arm_dynarmic_32.h"
+#include "core/arm/dynarmic/arm_dynarmic_64.h"
 #endif
 #include "core/arm/exclusive_monitor.h"
 #include "core/arm/unicorn/arm_unicorn.h"
@ -20,13 +21,17 @@ PhysicalCore::PhysicalCore(Core::System& system, std::size_t id,
                           Core::ExclusiveMonitor& exclusive_monitor)
    : core_index{id} {
 #ifdef ARCHITECTURE_x86_64
-    arm_interface = std::make_unique<Core::ARM_Dynarmic>(system, exclusive_monitor, core_index);
+    arm_interface_32 =
+        std::make_unique<Core::ARM_Dynarmic_32>(system, exclusive_monitor, core_index);
+    arm_interface_64 =
+        std::make_unique<Core::ARM_Dynarmic_64>(system, exclusive_monitor, core_index);
+
 #else
    arm_interface = std::make_shared<Core::ARM_Unicorn>(system);
    LOG_WARNING(Core, "CPU JIT requested, but Dynarmic not available");
 #endif

-    scheduler = std::make_unique<Kernel::Scheduler>(system, *arm_interface, core_index);
+    scheduler = std::make_unique<Kernel::Scheduler>(system, core_index);
 }

 PhysicalCore::~PhysicalCore() = default;
@ -48,4 +53,12 @@ void PhysicalCore::Shutdown() {
    scheduler->Shutdown();
 }

+void PhysicalCore::SetIs64Bit(bool is_64_bit) {
+    if (is_64_bit) {
+        arm_interface = arm_interface_64.get();
+    } else {
+        arm_interface = arm_interface_32.get();
+    }
+}
+
 } // namespace Kernel
--- a/src/core/hle/kernel/physical_core.h
+++ b/src/core/hle/kernel/physical_core.h
@ -68,10 +68,14 @@ public:
        return *scheduler;
    }

+    void SetIs64Bit(bool is_64_bit);
+
 private:
    std::size_t core_index;
-    std::unique_ptr<Core::ARM_Interface> arm_interface;
+    std::unique_ptr<Core::ARM_Interface> arm_interface_32;
+    std::unique_ptr<Core::ARM_Interface> arm_interface_64;
    std::unique_ptr<Kernel::Scheduler> scheduler;
+    Core::ARM_Interface* arm_interface{};
 };

 } // namespace Kernel
--- a/src/core/hle/kernel/process.cpp
+++ b/src/core/hle/kernel/process.cpp
@ -42,7 +42,8 @@ void SetupMainThread(Process& owner_process, KernelCore& kernel, u32 priority) {

    // Register 1 must be a handle to the main thread
    const Handle thread_handle = owner_process.GetHandleTable().Create(thread).Unwrap();
-    thread->GetContext().cpu_registers[1] = thread_handle;
+    thread->GetContext32().cpu_registers[1] = thread_handle;
+    thread->GetContext64().cpu_registers[1] = thread_handle;

    // Threads by default are dormant, wake up the main thread so it runs when the scheduler fires
    thread->ResumeFromWait();
--- a/src/core/hle/kernel/scheduler.cpp
+++ b/src/core/hle/kernel/scheduler.cpp
@ -383,8 +383,8 @@ void GlobalScheduler::Unlock() {
    // TODO(Blinkhawk): Setup the interrupts and change context on current core.
 }

-Scheduler::Scheduler(Core::System& system, Core::ARM_Interface& cpu_core, std::size_t core_id)
-    : system(system), cpu_core(cpu_core), core_id(core_id) {}
+Scheduler::Scheduler(Core::System& system, std::size_t core_id)
+    : system{system}, core_id{core_id} {}

 Scheduler::~Scheduler() = default;

@ -422,9 +422,10 @@ void Scheduler::UnloadThread() {

    // Save context for previous thread
    if (previous_thread) {
-        cpu_core.SaveContext(previous_thread->GetContext());
+        system.ArmInterface(core_id).SaveContext(previous_thread->GetContext32());
+        system.ArmInterface(core_id).SaveContext(previous_thread->GetContext64());
        // Save the TPIDR_EL0 system register in case it was modified.
-        previous_thread->SetTPIDR_EL0(cpu_core.GetTPIDR_EL0());
+        previous_thread->SetTPIDR_EL0(system.ArmInterface(core_id).GetTPIDR_EL0());

        if (previous_thread->GetStatus() == ThreadStatus::Running) {
            // This is only the case when a reschedule is triggered without the current thread
@ -451,9 +452,10 @@ void Scheduler::SwitchContext() {

    // Save context for previous thread
    if (previous_thread) {
-        cpu_core.SaveContext(previous_thread->GetContext());
+        system.ArmInterface(core_id).SaveContext(previous_thread->GetContext32());
+        system.ArmInterface(core_id).SaveContext(previous_thread->GetContext64());
        // Save the TPIDR_EL0 system register in case it was modified.
-        previous_thread->SetTPIDR_EL0(cpu_core.GetTPIDR_EL0());
+        previous_thread->SetTPIDR_EL0(system.ArmInterface(core_id).GetTPIDR_EL0());

        if (previous_thread->GetStatus() == ThreadStatus::Running) {
            // This is only the case when a reschedule is triggered without the current thread
@ -481,9 +483,10 @@ void Scheduler::SwitchContext() {
            system.Kernel().MakeCurrentProcess(thread_owner_process);
        }

-        cpu_core.LoadContext(new_thread->GetContext());
-        cpu_core.SetTlsAddress(new_thread->GetTLSAddress());
-        cpu_core.SetTPIDR_EL0(new_thread->GetTPIDR_EL0());
+        system.ArmInterface(core_id).LoadContext(new_thread->GetContext32());
+        system.ArmInterface(core_id).LoadContext(new_thread->GetContext64());
+        system.ArmInterface(core_id).SetTlsAddress(new_thread->GetTLSAddress());
+        system.ArmInterface(core_id).SetTPIDR_EL0(new_thread->GetTPIDR_EL0());
    } else {
        current_thread = nullptr;
        // Note: We do not reset the current process and current page table when idling because
--- a/src/core/hle/kernel/scheduler.h
+++ b/src/core/hle/kernel/scheduler.h
@ -181,7 +181,7 @@ private:

 class Scheduler final {
 public:
-    explicit Scheduler(Core::System& system, Core::ARM_Interface& cpu_core, std::size_t core_id);
+    explicit Scheduler(Core::System& system, std::size_t core_id);
    ~Scheduler();

    /// Returns whether there are any threads that are ready to run.
@ -235,7 +235,6 @@ private:
    std::shared_ptr<Thread> selected_thread = nullptr;

    Core::System& system;
-    Core::ARM_Interface& cpu_core;
    u64 last_context_switch_time = 0;
    u64 idle_selection_count = 0;
    const std::size_t core_id;
--- a/src/core/hle/kernel/svc.cpp
+++ b/src/core/hle/kernel/svc.cpp
@ -187,6 +187,13 @@ static ResultCode SetHeapSize(Core::System& system, VAddr* heap_addr, u64 heap_s
    return RESULT_SUCCESS;
 }

+static ResultCode SetHeapSize32(Core::System& system, u32* heap_addr, u32 heap_size) {
+    VAddr temp_heap_addr{};
+    const ResultCode result{SetHeapSize(system, &temp_heap_addr, heap_size)};
+    *heap_addr = static_cast<u32>(temp_heap_addr);
+    return result;
+}
+
 static ResultCode SetMemoryPermission(Core::System& system, VAddr addr, u64 size, u32 prot) {
    LOG_TRACE(Kernel_SVC, "called, addr=0x{:X}, size=0x{:X}, prot=0x{:X}", addr, size, prot);

@ -371,6 +378,12 @@ static ResultCode ConnectToNamedPort(Core::System& system, Handle* out_handle,
    return RESULT_SUCCESS;
 }

+static ResultCode ConnectToNamedPort32(Core::System& system, Handle* out_handle,
+                                       u32 port_name_address) {
+
+    return ConnectToNamedPort(system, out_handle, port_name_address);
+}
+
 /// Makes a blocking IPC call to an OS service.
 static ResultCode SendSyncRequest(Core::System& system, Handle handle) {
    const auto& handle_table = system.Kernel().CurrentProcess()->GetHandleTable();
@ -390,6 +403,10 @@ static ResultCode SendSyncRequest(Core::System& system, Handle handle) {
    return session->SendSyncRequest(SharedFrom(thread), system.Memory());
 }

+static ResultCode SendSyncRequest32(Core::System& system, Handle handle) {
+    return SendSyncRequest(system, handle);
+}
+
 /// Get the ID for the specified thread.
 static ResultCode GetThreadId(Core::System& system, u64* thread_id, Handle thread_handle) {
    LOG_TRACE(Kernel_SVC, "called thread=0x{:08X}", thread_handle);
@ -405,6 +422,17 @@ static ResultCode GetThreadId(Core::System& system, u64* thread_id, Handle threa
    return RESULT_SUCCESS;
 }

+static ResultCode GetThreadId32(Core::System& system, u32* thread_id_low, u32* thread_id_high,
+                                Handle thread_handle) {
+    u64 thread_id{};
+    const ResultCode result{GetThreadId(system, &thread_id, thread_handle)};
+
+    *thread_id_low = static_cast<u32>(thread_id >> 32);
+    *thread_id_high = static_cast<u32>(thread_id & std::numeric_limits<u32>::max());
+
+    return result;
+}
+
 /// Gets the ID of the specified process or a specified thread's owning process.
 static ResultCode GetProcessId(Core::System& system, u64* process_id, Handle handle) {
    LOG_DEBUG(Kernel_SVC, "called handle=0x{:08X}", handle);
@ -479,6 +507,12 @@ static ResultCode WaitSynchronization(Core::System& system, Handle* index, VAddr
    return result;
 }

+static ResultCode WaitSynchronization32(Core::System& system, u32 timeout_low, u32 handles_address,
+                                        s32 handle_count, u32 timeout_high, Handle* index) {
+    const s64 nano_seconds{(static_cast<s64>(timeout_high) << 32) | static_cast<s64>(timeout_low)};
+    return WaitSynchronization(system, index, handles_address, handle_count, nano_seconds);
+}
+
 /// Resumes a thread waiting on WaitSynchronization
 static ResultCode CancelSynchronization(Core::System& system, Handle thread_handle) {
    LOG_TRACE(Kernel_SVC, "called thread=0x{:X}", thread_handle);
@ -917,6 +951,18 @@ static ResultCode GetInfo(Core::System& system, u64* result, u64 info_id, u64 ha
    }
 }

+static ResultCode GetInfo32(Core::System& system, u32* result_low, u32* result_high, u32 sub_id_low,
+                            u32 info_id, u32 handle, u32 sub_id_high) {
+    const u64 sub_id{static_cast<u64>(sub_id_low | (static_cast<u64>(sub_id_high) << 32))};
+    u64 res_value{};
+
+    const ResultCode result{GetInfo(system, &res_value, info_id, handle, sub_id)};
+    *result_high = static_cast<u32>(res_value >> 32);
+    *result_low = static_cast<u32>(res_value & std::numeric_limits<u32>::max());
+
+    return result;
+}
+
 /// Maps memory at a desired address
 static ResultCode MapPhysicalMemory(Core::System& system, VAddr addr, u64 size) {
    LOG_DEBUG(Kernel_SVC, "called, addr=0x{:016X}, size=0x{:X}", addr, size);
@ -1058,7 +1104,7 @@ static ResultCode GetThreadContext(Core::System& system, VAddr thread_context, H
        return ERR_BUSY;
    }

-    Core::ARM_Interface::ThreadContext ctx = thread->GetContext();
+    Core::ARM_Interface::ThreadContext64 ctx = thread->GetContext64();
    // Mask away mode bits, interrupt bits, IL bit, and other reserved bits.
    ctx.pstate &= 0xFF0FFE20;

@ -1088,6 +1134,10 @@ static ResultCode GetThreadPriority(Core::System& system, u32* priority, Handle
    return RESULT_SUCCESS;
 }

+static ResultCode GetThreadPriority32(Core::System& system, u32* priority, Handle handle) {
+    return GetThreadPriority(system, priority, handle);
+}
+
 /// Sets the priority for the specified thread
 static ResultCode SetThreadPriority(Core::System& system, Handle handle, u32 priority) {
    LOG_TRACE(Kernel_SVC, "called");
@ -1259,6 +1309,11 @@ static ResultCode QueryMemory(Core::System& system, VAddr memory_info_address,
                              query_address);
 }

+static ResultCode QueryMemory32(Core::System& system, u32 memory_info_address,
+                                u32 page_info_address, u32 query_address) {
+    return QueryMemory(system, memory_info_address, page_info_address, query_address);
+}
+
 static ResultCode MapProcessCodeMemory(Core::System& system, Handle process_handle, u64 dst_address,
                                       u64 src_address, u64 size) {
    LOG_DEBUG(Kernel_SVC,
@ -1675,6 +1730,10 @@ static void SignalProcessWideKey(Core::System& system, VAddr condition_variable_
    }
 }

+static void SignalProcessWideKey32(Core::System& system, u32 condition_variable_addr, s32 target) {
+    SignalProcessWideKey(system, condition_variable_addr, target);
+}
+
 // Wait for an address (via Address Arbiter)
 static ResultCode WaitForAddress(Core::System& system, VAddr address, u32 type, s32 value,
                                 s64 timeout) {
@ -1760,6 +1819,10 @@ static ResultCode CloseHandle(Core::System& system, Handle handle) {
    return handle_table.Close(handle);
 }

+static ResultCode CloseHandle32(Core::System& system, Handle handle) {
+    return CloseHandle(system, handle);
+}
+
 /// Clears the signaled state of an event or process.
 static ResultCode ResetSignal(Core::System& system, Handle handle) {
    LOG_DEBUG(Kernel_SVC, "called handle 0x{:08X}", handle);
@ -2317,69 +2380,196 @@ struct FunctionDef {
 };
 } // namespace

-static const FunctionDef SVC_Table[] = {
+static const FunctionDef SVC_Table_32[] = {
    {0x00, nullptr, "Unknown"},
-    {0x01, SvcWrap<SetHeapSize>, "SetHeapSize"},
-    {0x02, SvcWrap<SetMemoryPermission>, "SetMemoryPermission"},
-    {0x03, SvcWrap<SetMemoryAttribute>, "SetMemoryAttribute"},
-    {0x04, SvcWrap<MapMemory>, "MapMemory"},
-    {0x05, SvcWrap<UnmapMemory>, "UnmapMemory"},
-    {0x06, SvcWrap<QueryMemory>, "QueryMemory"},
-    {0x07, SvcWrap<ExitProcess>, "ExitProcess"},
-    {0x08, SvcWrap<CreateThread>, "CreateThread"},
-    {0x09, SvcWrap<StartThread>, "StartThread"},
-    {0x0A, SvcWrap<ExitThread>, "ExitThread"},
-    {0x0B, SvcWrap<SleepThread>, "SleepThread"},
-    {0x0C, SvcWrap<GetThreadPriority>, "GetThreadPriority"},
-    {0x0D, SvcWrap<SetThreadPriority>, "SetThreadPriority"},
-    {0x0E, SvcWrap<GetThreadCoreMask>, "GetThreadCoreMask"},
-    {0x0F, SvcWrap<SetThreadCoreMask>, "SetThreadCoreMask"},
-    {0x10, SvcWrap<GetCurrentProcessorNumber>, "GetCurrentProcessorNumber"},
-    {0x11, SvcWrap<SignalEvent>, "SignalEvent"},
-    {0x12, SvcWrap<ClearEvent>, "ClearEvent"},
-    {0x13, SvcWrap<MapSharedMemory>, "MapSharedMemory"},
-    {0x14, SvcWrap<UnmapSharedMemory>, "UnmapSharedMemory"},
-    {0x15, SvcWrap<CreateTransferMemory>, "CreateTransferMemory"},
-    {0x16, SvcWrap<CloseHandle>, "CloseHandle"},
-    {0x17, SvcWrap<ResetSignal>, "ResetSignal"},
-    {0x18, SvcWrap<WaitSynchronization>, "WaitSynchronization"},
-    {0x19, SvcWrap<CancelSynchronization>, "CancelSynchronization"},
-    {0x1A, SvcWrap<ArbitrateLock>, "ArbitrateLock"},
-    {0x1B, SvcWrap<ArbitrateUnlock>, "ArbitrateUnlock"},
-    {0x1C, SvcWrap<WaitProcessWideKeyAtomic>, "WaitProcessWideKeyAtomic"},
-    {0x1D, SvcWrap<SignalProcessWideKey>, "SignalProcessWideKey"},
-    {0x1E, SvcWrap<GetSystemTick>, "GetSystemTick"},
-    {0x1F, SvcWrap<ConnectToNamedPort>, "ConnectToNamedPort"},
+    {0x01, SvcWrap32<SetHeapSize32>, "SetHeapSize32"},
+    {0x02, nullptr, "Unknown"},
+    {0x03, nullptr, "SetMemoryAttribute32"},
+    {0x04, nullptr, "MapMemory32"},
+    {0x05, nullptr, "UnmapMemory32"},
+    {0x06, SvcWrap32<QueryMemory32>, "QueryMemory32"},
+    {0x07, nullptr, "ExitProcess32"},
+    {0x08, nullptr, "CreateThread32"},
+    {0x09, nullptr, "StartThread32"},
+    {0x0a, nullptr, "ExitThread32"},
+    {0x0b, nullptr, "SleepThread32"},
+    {0x0c, SvcWrap32<GetThreadPriority32>, "GetThreadPriority32"},
+    {0x0d, nullptr, "SetThreadPriority32"},
+    {0x0e, nullptr, "GetThreadCoreMask32"},
+    {0x0f, nullptr, "SetThreadCoreMask32"},
+    {0x10, nullptr, "GetCurrentProcessorNumber32"},
+    {0x11, nullptr, "SignalEvent32"},
+    {0x12, nullptr, "ClearEvent32"},
+    {0x13, nullptr, "MapSharedMemory32"},
+    {0x14, nullptr, "UnmapSharedMemory32"},
+    {0x15, nullptr, "CreateTransferMemory32"},
+    {0x16, SvcWrap32<CloseHandle32>, "CloseHandle32"},
+    {0x17, nullptr, "ResetSignal32"},
+    {0x18, SvcWrap32<WaitSynchronization32>, "WaitSynchronization32"},
+    {0x19, nullptr, "CancelSynchronization32"},
+    {0x1a, nullptr, "ArbitrateLock32"},
+    {0x1b, nullptr, "ArbitrateUnlock32"},
+    {0x1c, nullptr, "WaitProcessWideKeyAtomic32"},
+    {0x1d, SvcWrap32<SignalProcessWideKey32>, "SignalProcessWideKey32"},
+    {0x1e, nullptr, "GetSystemTick32"},
+    {0x1f, SvcWrap32<ConnectToNamedPort32>, "ConnectToNamedPort32"},
+    {0x20, nullptr, "Unknown"},
+    {0x21, SvcWrap32<SendSyncRequest32>, "SendSyncRequest32"},
+    {0x22, nullptr, "SendSyncRequestWithUserBuffer32"},
+    {0x23, nullptr, "Unknown"},
+    {0x24, nullptr, "GetProcessId32"},
+    {0x25, SvcWrap32<GetThreadId32>, "GetThreadId32"},
+    {0x26, nullptr, "Break32"},
+    {0x27, nullptr, "OutputDebugString32"},
+    {0x28, nullptr, "Unknown"},
+    {0x29, SvcWrap32<GetInfo32>, "GetInfo32"},
+    {0x2a, nullptr, "Unknown"},
+    {0x2b, nullptr, "Unknown"},
+    {0x2c, nullptr, "MapPhysicalMemory32"},
+    {0x2d, nullptr, "UnmapPhysicalMemory32"},
+    {0x2e, nullptr, "Unknown"},
+    {0x2f, nullptr, "Unknown"},
+    {0x30, nullptr, "Unknown"},
+    {0x31, nullptr, "Unknown"},
+    {0x32, nullptr, "SetThreadActivity32"},
+    {0x33, nullptr, "GetThreadContext32"},
+    {0x34, nullptr, "WaitForAddress32"},
+    {0x35, nullptr, "SignalToAddress32"},
+    {0x36, nullptr, "Unknown"},
+    {0x37, nullptr, "Unknown"},
+    {0x38, nullptr, "Unknown"},
+    {0x39, nullptr, "Unknown"},
+    {0x3a, nullptr, "Unknown"},
+    {0x3b, nullptr, "Unknown"},
+    {0x3c, nullptr, "Unknown"},
+    {0x3d, nullptr, "Unknown"},
+    {0x3e, nullptr, "Unknown"},
+    {0x3f, nullptr, "Unknown"},
+    {0x40, nullptr, "CreateSession32"},
+    {0x41, nullptr, "AcceptSession32"},
+    {0x42, nullptr, "Unknown"},
+    {0x43, nullptr, "ReplyAndReceive32"},
+    {0x44, nullptr, "Unknown"},
+    {0x45, nullptr, "CreateEvent32"},
+    {0x46, nullptr, "Unknown"},
+    {0x47, nullptr, "Unknown"},
+    {0x48, nullptr, "Unknown"},
+    {0x49, nullptr, "Unknown"},
+    {0x4a, nullptr, "Unknown"},
+    {0x4b, nullptr, "Unknown"},
+    {0x4c, nullptr, "Unknown"},
+    {0x4d, nullptr, "Unknown"},
+    {0x4e, nullptr, "Unknown"},
+    {0x4f, nullptr, "Unknown"},
+    {0x50, nullptr, "Unknown"},
+    {0x51, nullptr, "Unknown"},
+    {0x52, nullptr, "Unknown"},
+    {0x53, nullptr, "Unknown"},
+    {0x54, nullptr, "Unknown"},
+    {0x55, nullptr, "Unknown"},
+    {0x56, nullptr, "Unknown"},
+    {0x57, nullptr, "Unknown"},
+    {0x58, nullptr, "Unknown"},
+    {0x59, nullptr, "Unknown"},
+    {0x5a, nullptr, "Unknown"},
+    {0x5b, nullptr, "Unknown"},
+    {0x5c, nullptr, "Unknown"},
+    {0x5d, nullptr, "Unknown"},
+    {0x5e, nullptr, "Unknown"},
+    {0x5F, nullptr, "FlushProcessDataCache32"},
+    {0x60, nullptr, "Unknown"},
+    {0x61, nullptr, "Unknown"},
+    {0x62, nullptr, "Unknown"},
+    {0x63, nullptr, "Unknown"},
+    {0x64, nullptr, "Unknown"},
+    {0x65, nullptr, "GetProcessList32"},
+    {0x66, nullptr, "Unknown"},
+    {0x67, nullptr, "Unknown"},
+    {0x68, nullptr, "Unknown"},
+    {0x69, nullptr, "Unknown"},
+    {0x6A, nullptr, "Unknown"},
+    {0x6B, nullptr, "Unknown"},
+    {0x6C, nullptr, "Unknown"},
+    {0x6D, nullptr, "Unknown"},
+    {0x6E, nullptr, "Unknown"},
+    {0x6f, nullptr, "GetSystemInfo32"},
+    {0x70, nullptr, "CreatePort32"},
+    {0x71, nullptr, "ManageNamedPort32"},
+    {0x72, nullptr, "ConnectToPort32"},
+    {0x73, nullptr, "SetProcessMemoryPermission32"},
+    {0x74, nullptr, "Unknown"},
+    {0x75, nullptr, "Unknown"},
+    {0x76, nullptr, "Unknown"},
+    {0x77, nullptr, "MapProcessCodeMemory32"},
+    {0x78, nullptr, "UnmapProcessCodeMemory32"},
+    {0x79, nullptr, "Unknown"},
+    {0x7A, nullptr, "Unknown"},
+    {0x7B, nullptr, "TerminateProcess32"},
+};
+
+static const FunctionDef SVC_Table_64[] = {
+    {0x00, nullptr, "Unknown"},
+    {0x01, SvcWrap64<SetHeapSize>, "SetHeapSize"},
+    {0x02, SvcWrap64<SetMemoryPermission>, "SetMemoryPermission"},
+    {0x03, SvcWrap64<SetMemoryAttribute>, "SetMemoryAttribute"},
+    {0x04, SvcWrap64<MapMemory>, "MapMemory"},
+    {0x05, SvcWrap64<UnmapMemory>, "UnmapMemory"},
+    {0x06, SvcWrap64<QueryMemory>, "QueryMemory"},
+    {0x07, SvcWrap64<ExitProcess>, "ExitProcess"},
+    {0x08, SvcWrap64<CreateThread>, "CreateThread"},
+    {0x09, SvcWrap64<StartThread>, "StartThread"},
+    {0x0A, SvcWrap64<ExitThread>, "ExitThread"},
+    {0x0B, SvcWrap64<SleepThread>, "SleepThread"},
+    {0x0C, SvcWrap64<GetThreadPriority>, "GetThreadPriority"},
+    {0x0D, SvcWrap64<SetThreadPriority>, "SetThreadPriority"},
+    {0x0E, SvcWrap64<GetThreadCoreMask>, "GetThreadCoreMask"},
+    {0x0F, SvcWrap64<SetThreadCoreMask>, "SetThreadCoreMask"},
+    {0x10, SvcWrap64<GetCurrentProcessorNumber>, "GetCurrentProcessorNumber"},
+    {0x11, SvcWrap64<SignalEvent>, "SignalEvent"},
+    {0x12, SvcWrap64<ClearEvent>, "ClearEvent"},
+    {0x13, SvcWrap64<MapSharedMemory>, "MapSharedMemory"},
+    {0x14, SvcWrap64<UnmapSharedMemory>, "UnmapSharedMemory"},
+    {0x15, SvcWrap64<CreateTransferMemory>, "CreateTransferMemory"},
+    {0x16, SvcWrap64<CloseHandle>, "CloseHandle"},
+    {0x17, SvcWrap64<ResetSignal>, "ResetSignal"},
+    {0x18, SvcWrap64<WaitSynchronization>, "WaitSynchronization"},
+    {0x19, SvcWrap64<CancelSynchronization>, "CancelSynchronization"},
+    {0x1A, SvcWrap64<ArbitrateLock>, "ArbitrateLock"},
+    {0x1B, SvcWrap64<ArbitrateUnlock>, "ArbitrateUnlock"},
+    {0x1C, SvcWrap64<WaitProcessWideKeyAtomic>, "WaitProcessWideKeyAtomic"},
+    {0x1D, SvcWrap64<SignalProcessWideKey>, "SignalProcessWideKey"},
+    {0x1E, SvcWrap64<GetSystemTick>, "GetSystemTick"},
+    {0x1F, SvcWrap64<ConnectToNamedPort>, "ConnectToNamedPort"},
    {0x20, nullptr, "SendSyncRequestLight"},
-    {0x21, SvcWrap<SendSyncRequest>, "SendSyncRequest"},
+    {0x21, SvcWrap64<SendSyncRequest>, "SendSyncRequest"},
    {0x22, nullptr, "SendSyncRequestWithUserBuffer"},
    {0x23, nullptr, "SendAsyncRequestWithUserBuffer"},
-    {0x24, SvcWrap<GetProcessId>, "GetProcessId"},
-    {0x25, SvcWrap<GetThreadId>, "GetThreadId"},
-    {0x26, SvcWrap<Break>, "Break"},
-    {0x27, SvcWrap<OutputDebugString>, "OutputDebugString"},
+    {0x24, SvcWrap64<GetProcessId>, "GetProcessId"},
+    {0x25, SvcWrap64<GetThreadId>, "GetThreadId"},
+    {0x26, SvcWrap64<Break>, "Break"},
+    {0x27, SvcWrap64<OutputDebugString>, "OutputDebugString"},
    {0x28, nullptr, "ReturnFromException"},
-    {0x29, SvcWrap<GetInfo>, "GetInfo"},
+    {0x29, SvcWrap64<GetInfo>, "GetInfo"},
    {0x2A, nullptr, "FlushEntireDataCache"},
    {0x2B, nullptr, "FlushDataCache"},
-    {0x2C, SvcWrap<MapPhysicalMemory>, "MapPhysicalMemory"},
-    {0x2D, SvcWrap<UnmapPhysicalMemory>, "UnmapPhysicalMemory"},
+    {0x2C, SvcWrap64<MapPhysicalMemory>, "MapPhysicalMemory"},
+    {0x2D, SvcWrap64<UnmapPhysicalMemory>, "UnmapPhysicalMemory"},
    {0x2E, nullptr, "GetFutureThreadInfo"},
    {0x2F, nullptr, "GetLastThreadInfo"},
-    {0x30, SvcWrap<GetResourceLimitLimitValue>, "GetResourceLimitLimitValue"},
-    {0x31, SvcWrap<GetResourceLimitCurrentValue>, "GetResourceLimitCurrentValue"},
-    {0x32, SvcWrap<SetThreadActivity>, "SetThreadActivity"},
-    {0x33, SvcWrap<GetThreadContext>, "GetThreadContext"},
-    {0x34, SvcWrap<WaitForAddress>, "WaitForAddress"},
-    {0x35, SvcWrap<SignalToAddress>, "SignalToAddress"},
+    {0x30, SvcWrap64<GetResourceLimitLimitValue>, "GetResourceLimitLimitValue"},
+    {0x31, SvcWrap64<GetResourceLimitCurrentValue>, "GetResourceLimitCurrentValue"},
+    {0x32, SvcWrap64<SetThreadActivity>, "SetThreadActivity"},
+    {0x33, SvcWrap64<GetThreadContext>, "GetThreadContext"},
+    {0x34, SvcWrap64<WaitForAddress>, "WaitForAddress"},
+    {0x35, SvcWrap64<SignalToAddress>, "SignalToAddress"},
    {0x36, nullptr, "SynchronizePreemptionState"},
    {0x37, nullptr, "Unknown"},
    {0x38, nullptr, "Unknown"},
    {0x39, nullptr, "Unknown"},
    {0x3A, nullptr, "Unknown"},
    {0x3B, nullptr, "Unknown"},
-    {0x3C, SvcWrap<KernelDebug>, "KernelDebug"},
-    {0x3D, SvcWrap<ChangeKernelTraceState>, "ChangeKernelTraceState"},
+    {0x3C, SvcWrap64<KernelDebug>, "KernelDebug"},
+    {0x3D, SvcWrap64<ChangeKernelTraceState>, "ChangeKernelTraceState"},
    {0x3E, nullptr, "Unknown"},
    {0x3F, nullptr, "Unknown"},
    {0x40, nullptr, "CreateSession"},
@ -2387,7 +2577,7 @@ static const FunctionDef SVC_Table[] = {
    {0x42, nullptr, "ReplyAndReceiveLight"},
    {0x43, nullptr, "ReplyAndReceive"},
    {0x44, nullptr, "ReplyAndReceiveWithUserBuffer"},
-    {0x45, SvcWrap<CreateEvent>, "CreateEvent"},
+    {0x45, SvcWrap64<CreateEvent>, "CreateEvent"},
    {0x46, nullptr, "Unknown"},
    {0x47, nullptr, "Unknown"},
    {0x48, nullptr, "MapPhysicalMemoryUnsafe"},
@ -2398,9 +2588,9 @@ static const FunctionDef SVC_Table[] = {
    {0x4D, nullptr, "SleepSystem"},
    {0x4E, nullptr, "ReadWriteRegister"},
    {0x4F, nullptr, "SetProcessActivity"},
-    {0x50, SvcWrap<CreateSharedMemory>, "CreateSharedMemory"},
-    {0x51, SvcWrap<MapTransferMemory>, "MapTransferMemory"},
-    {0x52, SvcWrap<UnmapTransferMemory>, "UnmapTransferMemory"},
+    {0x50, SvcWrap64<CreateSharedMemory>, "CreateSharedMemory"},
+    {0x51, SvcWrap64<MapTransferMemory>, "MapTransferMemory"},
+    {0x52, SvcWrap64<UnmapTransferMemory>, "UnmapTransferMemory"},
    {0x53, nullptr, "CreateInterruptEvent"},
    {0x54, nullptr, "QueryPhysicalAddress"},
    {0x55, nullptr, "QueryIoMapping"},
@ -2419,8 +2609,8 @@ static const FunctionDef SVC_Table[] = {
    {0x62, nullptr, "TerminateDebugProcess"},
    {0x63, nullptr, "GetDebugEvent"},
    {0x64, nullptr, "ContinueDebugEvent"},
-    {0x65, SvcWrap<GetProcessList>, "GetProcessList"},
-    {0x66, SvcWrap<GetThreadList>, "GetThreadList"},
+    {0x65, SvcWrap64<GetProcessList>, "GetProcessList"},
+    {0x66, SvcWrap64<GetThreadList>, "GetThreadList"},
    {0x67, nullptr, "GetDebugThreadContext"},
    {0x68, nullptr, "SetDebugThreadContext"},
    {0x69, nullptr, "QueryDebugProcessMemory"},
@ -2436,24 +2626,32 @@ static const FunctionDef SVC_Table[] = {
    {0x73, nullptr, "SetProcessMemoryPermission"},
    {0x74, nullptr, "MapProcessMemory"},
    {0x75, nullptr, "UnmapProcessMemory"},
-    {0x76, SvcWrap<QueryProcessMemory>, "QueryProcessMemory"},
-    {0x77, SvcWrap<MapProcessCodeMemory>, "MapProcessCodeMemory"},
-    {0x78, SvcWrap<UnmapProcessCodeMemory>, "UnmapProcessCodeMemory"},
+    {0x76, SvcWrap64<QueryProcessMemory>, "QueryProcessMemory"},
+    {0x77, SvcWrap64<MapProcessCodeMemory>, "MapProcessCodeMemory"},
+    {0x78, SvcWrap64<UnmapProcessCodeMemory>, "UnmapProcessCodeMemory"},
    {0x79, nullptr, "CreateProcess"},
    {0x7A, nullptr, "StartProcess"},
    {0x7B, nullptr, "TerminateProcess"},
-    {0x7C, SvcWrap<GetProcessInfo>, "GetProcessInfo"},
-    {0x7D, SvcWrap<CreateResourceLimit>, "CreateResourceLimit"},
-    {0x7E, SvcWrap<SetResourceLimitLimitValue>, "SetResourceLimitLimitValue"},
+    {0x7C, SvcWrap64<GetProcessInfo>, "GetProcessInfo"},
+    {0x7D, SvcWrap64<CreateResourceLimit>, "CreateResourceLimit"},
+    {0x7E, SvcWrap64<SetResourceLimitLimitValue>, "SetResourceLimitLimitValue"},
    {0x7F, nullptr, "CallSecureMonitor"},
 };

-static const FunctionDef* GetSVCInfo(u32 func_num) {
-    if (func_num >= std::size(SVC_Table)) {
+static const FunctionDef* GetSVCInfo32(u32 func_num) {
+    if (func_num >= std::size(SVC_Table_32)) {
        LOG_ERROR(Kernel_SVC, "Unknown svc=0x{:02X}", func_num);
        return nullptr;
    }
-    return &SVC_Table[func_num];
+    return &SVC_Table_32[func_num];
+}
+
+static const FunctionDef* GetSVCInfo64(u32 func_num) {
+    if (func_num >= std::size(SVC_Table_64)) {
+        LOG_ERROR(Kernel_SVC, "Unknown svc=0x{:02X}", func_num);
+        return nullptr;
+    }
+    return &SVC_Table_64[func_num];
 }

 MICROPROFILE_DEFINE(Kernel_SVC, "Kernel", "SVC", MP_RGB(70, 200, 70));
@ -2464,7 +2662,8 @@ void CallSVC(Core::System& system, u32 immediate) {
    // Lock the global kernel mutex when we enter the kernel HLE.
    std::lock_guard lock{HLE::g_hle_lock};

-    const FunctionDef* info = GetSVCInfo(immediate);
+    const FunctionDef* info = system.CurrentProcess()->Is64BitProcess() ? GetSVCInfo64(immediate)
+                                                                        : GetSVCInfo32(immediate);
    if (info) {
        if (info->func) {
            info->func(system);
--- a/src/core/hle/kernel/svc_wrap.h
+++ b/src/core/hle/kernel/svc_wrap.h
@ -15,6 +15,10 @@ static inline u64 Param(const Core::System& system, int n) {
    return system.CurrentArmInterface().GetReg(n);
 }

+static inline u32 Param32(const Core::System& system, int n) {
+    return static_cast<u32>(system.CurrentArmInterface().GetReg(n));
+}
+
 /**
 * HLE a function return from the current ARM userland process
 * @param system System context
@ -24,40 +28,44 @@ static inline void FuncReturn(Core::System& system, u64 result) {
    system.CurrentArmInterface().SetReg(0, result);
 }

+static inline void FuncReturn32(Core::System& system, u32 result) {
+    system.CurrentArmInterface().SetReg(0, (u64)result);
+}
+
 ////////////////////////////////////////////////////////////////////////////////////////////////////
 // Function wrappers that return type ResultCode

 template <ResultCode func(Core::System&, u64)>
-void SvcWrap(Core::System& system) {
+void SvcWrap64(Core::System& system) {
    FuncReturn(system, func(system, Param(system, 0)).raw);
 }

 template <ResultCode func(Core::System&, u64, u64)>
-void SvcWrap(Core::System& system) {
+void SvcWrap64(Core::System& system) {
    FuncReturn(system, func(system, Param(system, 0), Param(system, 1)).raw);
 }

 template <ResultCode func(Core::System&, u32)>
-void SvcWrap(Core::System& system) {
+void SvcWrap64(Core::System& system) {
    FuncReturn(system, func(system, static_cast<u32>(Param(system, 0))).raw);
 }

 template <ResultCode func(Core::System&, u32, u32)>
-void SvcWrap(Core::System& system) {
+void SvcWrap64(Core::System& system) {
    FuncReturn(
        system,
        func(system, static_cast<u32>(Param(system, 0)), static_cast<u32>(Param(system, 1))).raw);
 }

 template <ResultCode func(Core::System&, u32, u64, u64, u64)>
-void SvcWrap(Core::System& system) {
+void SvcWrap64(Core::System& system) {
    FuncReturn(system, func(system, static_cast<u32>(Param(system, 0)), Param(system, 1),
                            Param(system, 2), Param(system, 3))
                           .raw);
 }

 template <ResultCode func(Core::System&, u32*)>
-void SvcWrap(Core::System& system) {
+void SvcWrap64(Core::System& system) {
    u32 param = 0;
    const u32 retval = func(system, &param).raw;
    system.CurrentArmInterface().SetReg(1, param);
@ -65,7 +73,7 @@ void SvcWrap(Core::System& system) {
 }

 template <ResultCode func(Core::System&, u32*, u32)>
-void SvcWrap(Core::System& system) {
+void SvcWrap64(Core::System& system) {
    u32 param_1 = 0;
    const u32 retval = func(system, &param_1, static_cast<u32>(Param(system, 1))).raw;
    system.CurrentArmInterface().SetReg(1, param_1);
@ -73,7 +81,7 @@ void SvcWrap(Core::System& system) {
 }

 template <ResultCode func(Core::System&, u32*, u32*)>
-void SvcWrap(Core::System& system) {
+void SvcWrap64(Core::System& system) {
    u32 param_1 = 0;
    u32 param_2 = 0;
    const u32 retval = func(system, &param_1, &param_2).raw;
@ -86,7 +94,7 @@ void SvcWrap(Core::System& system) {
 }

 template <ResultCode func(Core::System&, u32*, u64)>
-void SvcWrap(Core::System& system) {
+void SvcWrap64(Core::System& system) {
    u32 param_1 = 0;
    const u32 retval = func(system, &param_1, Param(system, 1)).raw;
    system.CurrentArmInterface().SetReg(1, param_1);
@ -94,7 +102,7 @@ void SvcWrap(Core::System& system) {
 }

 template <ResultCode func(Core::System&, u32*, u64, u32)>
-void SvcWrap(Core::System& system) {
+void SvcWrap64(Core::System& system) {
    u32 param_1 = 0;
    const u32 retval =
        func(system, &param_1, Param(system, 1), static_cast<u32>(Param(system, 2))).raw;
@ -104,7 +112,7 @@ void SvcWrap(Core::System& system) {
 }

 template <ResultCode func(Core::System&, u64*, u32)>
-void SvcWrap(Core::System& system) {
+void SvcWrap64(Core::System& system) {
    u64 param_1 = 0;
    const u32 retval = func(system, &param_1, static_cast<u32>(Param(system, 1))).raw;

@ -113,12 +121,12 @@ void SvcWrap(Core::System& system) {
 }

 template <ResultCode func(Core::System&, u64, u32)>
-void SvcWrap(Core::System& system) {
+void SvcWrap64(Core::System& system) {
    FuncReturn(system, func(system, Param(system, 0), static_cast<u32>(Param(system, 1))).raw);
 }

 template <ResultCode func(Core::System&, u64*, u64)>
-void SvcWrap(Core::System& system) {
+void SvcWrap64(Core::System& system) {
    u64 param_1 = 0;
    const u32 retval = func(system, &param_1, Param(system, 1)).raw;

@ -127,7 +135,7 @@ void SvcWrap(Core::System& system) {
 }

 template <ResultCode func(Core::System&, u64*, u32, u32)>
-void SvcWrap(Core::System& system) {
+void SvcWrap64(Core::System& system) {
    u64 param_1 = 0;
    const u32 retval = func(system, &param_1, static_cast<u32>(Param(system, 1)),
                            static_cast<u32>(Param(system, 2)))
@ -138,19 +146,19 @@ void SvcWrap(Core::System& system) {
 }

 template <ResultCode func(Core::System&, u32, u64)>
-void SvcWrap(Core::System& system) {
+void SvcWrap64(Core::System& system) {
    FuncReturn(system, func(system, static_cast<u32>(Param(system, 0)), Param(system, 1)).raw);
 }

 template <ResultCode func(Core::System&, u32, u32, u64)>
-void SvcWrap(Core::System& system) {
+void SvcWrap64(Core::System& system) {
    FuncReturn(system, func(system, static_cast<u32>(Param(system, 0)),
                            static_cast<u32>(Param(system, 1)), Param(system, 2))
                           .raw);
 }

 template <ResultCode func(Core::System&, u32, u32*, u64*)>
-void SvcWrap(Core::System& system) {
+void SvcWrap64(Core::System& system) {
    u32 param_1 = 0;
    u64 param_2 = 0;
    const ResultCode retval = func(system, static_cast<u32>(Param(system, 2)), &param_1, &param_2);
@ -161,54 +169,54 @@ void SvcWrap(Core::System& system) {
 }

 template <ResultCode func(Core::System&, u64, u64, u32, u32)>
-void SvcWrap(Core::System& system) {
+void SvcWrap64(Core::System& system) {
    FuncReturn(system, func(system, Param(system, 0), Param(system, 1),
                            static_cast<u32>(Param(system, 2)), static_cast<u32>(Param(system, 3)))
                           .raw);
 }

 template <ResultCode func(Core::System&, u64, u64, u32, u64)>
-void SvcWrap(Core::System& system) {
+void SvcWrap64(Core::System& system) {
    FuncReturn(system, func(system, Param(system, 0), Param(system, 1),
                            static_cast<u32>(Param(system, 2)), Param(system, 3))
                           .raw);
 }

 template <ResultCode func(Core::System&, u32, u64, u32)>
-void SvcWrap(Core::System& system) {
+void SvcWrap64(Core::System& system) {
    FuncReturn(system, func(system, static_cast<u32>(Param(system, 0)), Param(system, 1),
                            static_cast<u32>(Param(system, 2)))
                           .raw);
 }

 template <ResultCode func(Core::System&, u64, u64, u64)>
-void SvcWrap(Core::System& system) {
+void SvcWrap64(Core::System& system) {
    FuncReturn(system, func(system, Param(system, 0), Param(system, 1), Param(system, 2)).raw);
 }

 template <ResultCode func(Core::System&, u64, u64, u32)>
-void SvcWrap(Core::System& system) {
+void SvcWrap64(Core::System& system) {
    FuncReturn(
        system,
        func(system, Param(system, 0), Param(system, 1), static_cast<u32>(Param(system, 2))).raw);
 }

 template <ResultCode func(Core::System&, u32, u64, u64, u32)>
-void SvcWrap(Core::System& system) {
+void SvcWrap64(Core::System& system) {
    FuncReturn(system, func(system, static_cast<u32>(Param(system, 0)), Param(system, 1),
                            Param(system, 2), static_cast<u32>(Param(system, 3)))
                           .raw);
 }

 template <ResultCode func(Core::System&, u32, u64, u64)>
-void SvcWrap(Core::System& system) {
+void SvcWrap64(Core::System& system) {
    FuncReturn(
        system,
        func(system, static_cast<u32>(Param(system, 0)), Param(system, 1), Param(system, 2)).raw);
 }

 template <ResultCode func(Core::System&, u32*, u64, u64, s64)>
-void SvcWrap(Core::System& system) {
+void SvcWrap64(Core::System& system) {
    u32 param_1 = 0;
    const u32 retval = func(system, &param_1, Param(system, 1), static_cast<u32>(Param(system, 2)),
                            static_cast<s64>(Param(system, 3)))
@ -219,14 +227,14 @@ void SvcWrap(Core::System& system) {
 }

 template <ResultCode func(Core::System&, u64, u64, u32, s64)>
-void SvcWrap(Core::System& system) {
+void SvcWrap64(Core::System& system) {
    FuncReturn(system, func(system, Param(system, 0), Param(system, 1),
                            static_cast<u32>(Param(system, 2)), static_cast<s64>(Param(system, 3)))
                           .raw);
 }

 template <ResultCode func(Core::System&, u64*, u64, u64, u64)>
-void SvcWrap(Core::System& system) {
+void SvcWrap64(Core::System& system) {
    u64 param_1 = 0;
    const u32 retval =
        func(system, &param_1, Param(system, 1), Param(system, 2), Param(system, 3)).raw;
@ -236,7 +244,7 @@ void SvcWrap(Core::System& system) {
 }

 template <ResultCode func(Core::System&, u32*, u64, u64, u64, u32, s32)>
-void SvcWrap(Core::System& system) {
+void SvcWrap64(Core::System& system) {
    u32 param_1 = 0;
    const u32 retval = func(system, &param_1, Param(system, 1), Param(system, 2), Param(system, 3),
                            static_cast<u32>(Param(system, 4)), static_cast<s32>(Param(system, 5)))
@ -247,7 +255,7 @@ void SvcWrap(Core::System& system) {
 }

 template <ResultCode func(Core::System&, u32*, u64, u64, u32)>
-void SvcWrap(Core::System& system) {
+void SvcWrap64(Core::System& system) {
    u32 param_1 = 0;
    const u32 retval = func(system, &param_1, Param(system, 1), Param(system, 2),
                            static_cast<u32>(Param(system, 3)))
@ -258,7 +266,7 @@ void SvcWrap(Core::System& system) {
 }

 template <ResultCode func(Core::System&, Handle*, u64, u32, u32)>
-void SvcWrap(Core::System& system) {
+void SvcWrap64(Core::System& system) {
    u32 param_1 = 0;
    const u32 retval = func(system, &param_1, Param(system, 1), static_cast<u32>(Param(system, 2)),
                            static_cast<u32>(Param(system, 3)))
@ -269,14 +277,14 @@ void SvcWrap(Core::System& system) {
 }

 template <ResultCode func(Core::System&, u64, u32, s32, s64)>
-void SvcWrap(Core::System& system) {
+void SvcWrap64(Core::System& system) {
    FuncReturn(system, func(system, Param(system, 0), static_cast<u32>(Param(system, 1)),
                            static_cast<s32>(Param(system, 2)), static_cast<s64>(Param(system, 3)))
                           .raw);
 }

 template <ResultCode func(Core::System&, u64, u32, s32, s32)>
-void SvcWrap(Core::System& system) {
+void SvcWrap64(Core::System& system) {
    FuncReturn(system, func(system, Param(system, 0), static_cast<u32>(Param(system, 1)),
                            static_cast<s32>(Param(system, 2)), static_cast<s32>(Param(system, 3)))
                           .raw);
@ -286,7 +294,7 @@ void SvcWrap(Core::System& system) {
 // Function wrappers that return type u32

 template <u32 func(Core::System&)>
-void SvcWrap(Core::System& system) {
+void SvcWrap64(Core::System& system) {
    FuncReturn(system, func(system));
 }

@ -294,7 +302,7 @@ void SvcWrap(Core::System& system) {
 // Function wrappers that return type u64

 template <u64 func(Core::System&)>
-void SvcWrap(Core::System& system) {
+void SvcWrap64(Core::System& system) {
    FuncReturn(system, func(system));
 }

@ -302,44 +310,110 @@ void SvcWrap(Core::System& system) {
 /// Function wrappers that return type void

 template <void func(Core::System&)>
-void SvcWrap(Core::System& system) {
+void SvcWrap64(Core::System& system) {
    func(system);
 }

 template <void func(Core::System&, u32)>
-void SvcWrap(Core::System& system) {
+void SvcWrap64(Core::System& system) {
    func(system, static_cast<u32>(Param(system, 0)));
 }

 template <void func(Core::System&, u32, u64, u64, u64)>
-void SvcWrap(Core::System& system) {
+void SvcWrap64(Core::System& system) {
    func(system, static_cast<u32>(Param(system, 0)), Param(system, 1), Param(system, 2),
         Param(system, 3));
 }

 template <void func(Core::System&, s64)>
-void SvcWrap(Core::System& system) {
+void SvcWrap64(Core::System& system) {
    func(system, static_cast<s64>(Param(system, 0)));
 }

 template <void func(Core::System&, u64, s32)>
-void SvcWrap(Core::System& system) {
+void SvcWrap64(Core::System& system) {
    func(system, Param(system, 0), static_cast<s32>(Param(system, 1)));
 }

 template <void func(Core::System&, u64, u64)>
-void SvcWrap(Core::System& system) {
+void SvcWrap64(Core::System& system) {
    func(system, Param(system, 0), Param(system, 1));
 }

 template <void func(Core::System&, u64, u64, u64)>
-void SvcWrap(Core::System& system) {
+void SvcWrap64(Core::System& system) {
    func(system, Param(system, 0), Param(system, 1), Param(system, 2));
 }

 template <void func(Core::System&, u32, u64, u64)>
-void SvcWrap(Core::System& system) {
+void SvcWrap64(Core::System& system) {
    func(system, static_cast<u32>(Param(system, 0)), Param(system, 1), Param(system, 2));
 }

+// Used by QueryMemory32
+template <ResultCode func(Core::System&, u32, u32, u32)>
+void SvcWrap32(Core::System& system) {
+    FuncReturn32(system,
+                 func(system, Param32(system, 0), Param32(system, 1), Param32(system, 2)).raw);
+}
+
+// Used by GetInfo32
+template <ResultCode func(Core::System&, u32*, u32*, u32, u32, u32, u32)>
+void SvcWrap32(Core::System& system) {
+    u32 param_1 = 0;
+    u32 param_2 = 0;
+
+    const u32 retval = func(system, &param_1, &param_2, Param32(system, 0), Param32(system, 1),
+                            Param32(system, 2), Param32(system, 3))
+                           .raw;
+
+    system.CurrentArmInterface().SetReg(1, param_1);
+    system.CurrentArmInterface().SetReg(2, param_2);
+    FuncReturn(system, retval);
+}
+
+// Used by GetThreadPriority32, ConnectToNamedPort32
+template <ResultCode func(Core::System&, u32*, u32)>
+void SvcWrap32(Core::System& system) {
+    u32 param_1 = 0;
+    const u32 retval = func(system, &param_1, Param32(system, 1)).raw;
+    system.CurrentArmInterface().SetReg(1, param_1);
+    FuncReturn(system, retval);
+}
+
+// Used by GetThreadId32
+template <ResultCode func(Core::System&, u32*, u32*, u32)>
+void SvcWrap32(Core::System& system) {
+    u32 param_1 = 0;
+    u32 param_2 = 0;
+
+    const u32 retval = func(system, &param_1, &param_2, Param32(system, 1)).raw;
+    system.CurrentArmInterface().SetReg(1, param_1);
+    system.CurrentArmInterface().SetReg(2, param_2);
+    FuncReturn(system, retval);
+}
+
+// Used by SignalProcessWideKey32
+template <void func(Core::System&, u32, s32)>
+void SvcWrap32(Core::System& system) {
+    func(system, static_cast<u32>(Param(system, 0)), static_cast<s32>(Param(system, 1)));
+}
+
+// Used by SendSyncRequest32
+template <ResultCode func(Core::System&, u32)>
+void SvcWrap32(Core::System& system) {
+    FuncReturn(system, func(system, static_cast<u32>(Param(system, 0))).raw);
+}
+
+// Used by WaitSynchronization32
+template <ResultCode func(Core::System&, u32, u32, s32, u32, Handle*)>
+void SvcWrap32(Core::System& system) {
+    u32 param_1 = 0;
+    const u32 retval = func(system, Param32(system, 0), Param32(system, 1), Param32(system, 2),
+                            Param32(system, 3), &param_1)
+                           .raw;
+    system.CurrentArmInterface().SetReg(1, param_1);
+    FuncReturn(system, retval);
+}
+
 } // namespace Kernel
--- a/src/core/hle/kernel/thread.cpp
+++ b/src/core/hle/kernel/thread.cpp
@ -133,15 +133,16 @@ void Thread::CancelWait() {
    ResumeFromWait();
 }

-/**
- * Resets a thread context, making it ready to be scheduled and run by the CPU
- * @param context Thread context to reset
- * @param stack_top Address of the top of the stack
- * @param entry_point Address of entry point for execution
- * @param arg User argument for thread
- */
-static void ResetThreadContext(Core::ARM_Interface::ThreadContext& context, VAddr stack_top,
-                               VAddr entry_point, u64 arg) {
+static void ResetThreadContext32(Core::ARM_Interface::ThreadContext32& context, u32 stack_top,
+                                 u32 entry_point, u32 arg) {
+    context = {};
+    context.cpu_registers[0] = arg;
+    context.cpu_registers[15] = entry_point;
+    context.cpu_registers[13] = stack_top;
+}
+
+static void ResetThreadContext64(Core::ARM_Interface::ThreadContext64& context, VAddr stack_top,
+                                 VAddr entry_point, u64 arg) {
    context = {};
    context.cpu_registers[0] = arg;
    context.pc = entry_point;
@ -198,9 +199,9 @@ ResultVal<std::shared_ptr<Thread>> Thread::Create(KernelCore& kernel, std::strin

    thread->owner_process->RegisterThread(thread.get());

-    // TODO(peachum): move to ScheduleThread() when scheduler is added so selected core is used
-    // to initialize the context
-    ResetThreadContext(thread->context, stack_top, entry_point, arg);
+    ResetThreadContext32(thread->context_32, static_cast<u32>(stack_top),
+                         static_cast<u32>(entry_point), static_cast<u32>(arg));
+    ResetThreadContext64(thread->context_64, stack_top, entry_point, arg);

    return MakeResult<std::shared_ptr<Thread>>(std::move(thread));
 }
@ -213,11 +214,13 @@ void Thread::SetPriority(u32 priority) {
 }

 void Thread::SetWaitSynchronizationResult(ResultCode result) {
-    context.cpu_registers[0] = result.raw;
+    context_32.cpu_registers[0] = result.raw;
+    context_64.cpu_registers[0] = result.raw;
 }

 void Thread::SetWaitSynchronizationOutput(s32 output) {
-    context.cpu_registers[1] = output;
+    context_32.cpu_registers[1] = output;
+    context_64.cpu_registers[1] = output;
 }

 s32 Thread::GetSynchronizationObjectIndex(std::shared_ptr<SynchronizationObject> object) const {
--- a/src/core/hle/kernel/thread.h
+++ b/src/core/hle/kernel/thread.h
@ -102,7 +102,8 @@ public:

    using MutexWaitingThreads = std::vector<std::shared_ptr<Thread>>;

-    using ThreadContext = Core::ARM_Interface::ThreadContext;
+    using ThreadContext32 = Core::ARM_Interface::ThreadContext32;
+    using ThreadContext64 = Core::ARM_Interface::ThreadContext64;

    using ThreadSynchronizationObjects = std::vector<std::shared_ptr<SynchronizationObject>>;

@ -273,12 +274,20 @@ public:
        return status == ThreadStatus::WaitSynch;
    }

-    ThreadContext& GetContext() {
-        return context;
+    ThreadContext32& GetContext32() {
+        return context_32;
    }

-    const ThreadContext& GetContext() const {
-        return context;
+    const ThreadContext32& GetContext32() const {
+        return context_32;
+    }
+
+    ThreadContext64& GetContext64() {
+        return context_64;
+    }
+
+    const ThreadContext64& GetContext64() const {
+        return context_64;
    }

    ThreadStatus GetStatus() const {
@ -466,7 +475,8 @@ private:
    void AdjustSchedulingOnPriority(u32 old_priority);
    void AdjustSchedulingOnAffinity(u64 old_affinity_mask, s32 old_core);

-    Core::ARM_Interface::ThreadContext context{};
+    ThreadContext32 context_32{};
+    ThreadContext64 context_64{};

    u64 thread_id = 0;

--- a/src/core/loader/deconstructed_rom_directory.cpp
+++ b/src/core/loader/deconstructed_rom_directory.cpp
@ -129,12 +129,6 @@ AppLoader_DeconstructedRomDirectory::LoadResult AppLoader_DeconstructedRomDirect
    }
    metadata.Print();

-    const FileSys::ProgramAddressSpaceType arch_bits{metadata.GetAddressSpaceType()};
-    if (arch_bits == FileSys::ProgramAddressSpaceType::Is32Bit ||
-        arch_bits == FileSys::ProgramAddressSpaceType::Is32BitNoMap) {
-        return {ResultStatus::Error32BitISA, {}};
-    }
-
    if (process.LoadFromMetadata(metadata).IsError()) {
        return {ResultStatus::ErrorUnableToParseKernelMetadata, {}};
    }
--- a/src/core/reporter.cpp
+++ b/src/core/reporter.cpp
@ -111,7 +111,7 @@ json GetProcessorStateDataAuto(Core::System& system) {
    const auto& vm_manager{process->VMManager()};
    auto& arm{system.CurrentArmInterface()};

-    Core::ARM_Interface::ThreadContext context{};
+    Core::ARM_Interface::ThreadContext64 context{};
    arm.SaveContext(context);

    return GetProcessorStateData(process->Is64BitProcess() ? "AArch64" : "AArch32",
--- a/src/yuzu/debugger/wait_tree.cpp
+++ b/src/yuzu/debugger/wait_tree.cpp
@ -116,7 +116,7 @@ std::vector<std::unique_ptr<WaitTreeItem>> WaitTreeCallstack::GetChildren() cons

    constexpr std::size_t BaseRegister = 29;
    auto& memory = Core::System::GetInstance().Memory();
-    u64 base_pointer = thread.GetContext().cpu_registers[BaseRegister];
+    u64 base_pointer = thread.GetContext64().cpu_registers[BaseRegister];

    while (base_pointer != 0) {
        const u64 lr = memory.Read64(base_pointer + sizeof(u64));
@ -240,7 +240,7 @@ QString WaitTreeThread::GetText() const {
        break;
    }

-    const auto& context = thread.GetContext();
+    const auto& context = thread.GetContext64();
    const QString pc_info = tr(" PC = 0x%1 LR = 0x%2")
                                .arg(context.pc, 8, 16, QLatin1Char{'0'})
                                .arg(context.cpu_registers[30], 8, 16, QLatin1Char{'0'});