diff --git a/Source/Core/Core/HLE/HLE_OS.cpp b/Source/Core/Core/HLE/HLE_OS.cpp
@@ -3,9 +3,15 @@
 
 #include "Core/HLE/HLE_OS.h"
 
+#include <cinttypes>
 #include <memory>
 #include <string>
+#include <string_view>
 
+#include <fmt/format.h>
+#include <fmt/printf.h>
+
+#include "Common/BitUtils.h"
 #include "Common/CommonTypes.h"
 #include "Common/Logging/Log.h"
 #include "Common/MsgHandler.h"
@@ -204,82 +210,365 @@ void HLE_LogVFPrint(const Core::CPUThreadGuard& guard)
   HLE_LogFPrint(guard, ParameterType::VariableArgumentList);
 }
 
+namespace
+{
+class HLEPrintArgsVAList final : public HLEPrintArgs
+{
+public:
+  HLEPrintArgsVAList(const Core::CPUThreadGuard& guard, HLE::SystemVABI::VAList* va_list)
+      : m_guard(guard), m_va_list(va_list)
+  {
+  }
+
+  u32 GetU32() override { return m_va_list->GetArgT<u32>(); }
+  u64 GetU64() override { return m_va_list->GetArgT<u64>(); }
+  double GetF64() override { return m_va_list->GetArgT<double>(); }
+  std::string GetString(std::optional<u32> max_length) override
+  {
+    return max_length == 0u ? std::string() :
+                              PowerPC::MMU::HostGetString(m_guard, m_va_list->GetArgT<u32>(),
+                                                          max_length.value_or(0u));
+  }
+  std::u16string GetU16String(std::optional<u32> max_length) override
+  {
+    return max_length == 0u ? std::u16string() :
+                              PowerPC::MMU::HostGetU16String(m_guard, m_va_list->GetArgT<u32>(),
+                                                             max_length.value_or(0u));
+  }
+
+private:
+  const Core::CPUThreadGuard& m_guard;
+  HLE::SystemVABI::VAList* m_va_list;
+};
+}  // namespace
+
 std::string GetStringVA(Core::System& system, const Core::CPUThreadGuard& guard, u32 str_reg,
                         ParameterType parameter_type)
 {
   auto& ppc_state = system.GetPPCState();
 
-  std::string ArgumentBuffer;
-  std::string result;
   std::string string = PowerPC::MMU::HostGetString(guard, ppc_state.gpr[str_reg]);
-  auto ap =
+  std::unique_ptr<HLE::SystemVABI::VAList> ap =
       parameter_type == ParameterType::VariableArgumentList ?
-          std::make_unique<HLE::SystemVABI::VAListStruct>(system, guard,
-                                                          ppc_state.gpr[str_reg + 1]) :
-          std::make_unique<HLE::SystemVABI::VAList>(system, ppc_state.gpr[1] + 0x8, str_reg + 1);
+          std::make_unique<HLE::SystemVABI::VAListStruct>(guard, ppc_state.gpr[str_reg + 1]) :
+          std::make_unique<HLE::SystemVABI::VAList>(guard, ppc_state.gpr[1] + 0x8, str_reg + 1);
+
+  HLEPrintArgsVAList args(guard, ap.get());
+  return GetStringVA(&args, string);
+}
 
-  for (size_t i = 0; i < string.size(); i++)
+std::string GetStringVA(HLEPrintArgs* args, std::string_view string)
+{
+  std::string result;
+  for (size_t i = 0; i < string.size(); ++i)
   {
-    if (string[i] == '%')
+    if (string[i] != '%')
+    {
+      result += string[i];
+      continue;
+    }
+
+    const size_t formatting_start_position = i;
+    ++i;
+    if (i < string.size() && string[i] == '%')
+    {
+      result += '%';
+      continue;
+    }
+
+    bool left_justified_flag = false;
+    bool sign_prepended_flag = false;
+    bool space_prepended_flag = false;
+    bool alternative_form_flag = false;
+    bool padding_zeroes_flag = false;
+    while (i < string.size())
     {
-      ArgumentBuffer = '%';
-      i++;
-      if (string[i] == '%')
+      if (string[i] == '-')
+        left_justified_flag = true;
+      else if (string[i] == '+')
+        sign_prepended_flag = true;
+      else if (string[i] == ' ')
+        space_prepended_flag = true;
+      else if (string[i] == '#')
+        alternative_form_flag = true;
+      else if (string[i] == '0')
+        padding_zeroes_flag = true;
+      else
+        break;
+      ++i;
+    }
+
+    const auto take_field_or_precision = [&](bool* left_justified_flag_ptr) -> std::optional<u32> {
+      if (i >= string.size())
+        return std::nullopt;
+
+      if (string[i] == '*')
       {
-        result += '%';
-        continue;
+        ++i;
+        const s32 result = Common::BitCast<s32>(args->GetU32());
+        if (result >= 0)
+          return static_cast<u32>(result);
+
+        if (left_justified_flag_ptr)
+        {
+          // field width; this results in positive field width and left_justified flag set
+          *left_justified_flag_ptr = true;
+          return static_cast<u32>(-static_cast<s64>(result));
+        }
+
+        // precision; this is ignored
+        return std::nullopt;
       }
 
-      while (i < string.size() &&
-             (string[i] < 'A' || string[i] > 'z' || string[i] == 'l' || string[i] == '-'))
+      size_t start = i;
+      while (i < string.size() && string[i] >= '0' && string[i] <= '9')
+        ++i;
+      if (start != i)
       {
-        ArgumentBuffer += string[i++];
+        while (start < i && string[start] == '0')
+          ++start;
+        if (start == i)
+          return 0;
+        u32 result = 0;
+        const std::string tmp(string, start, i - start);
+        if (TryParse(tmp, &result, 10))
+          return result;
       }
-      if (i >= string.size())
-        break;
-
-      ArgumentBuffer += string[i];
 
-      switch (string[i])
-      {
-      case 's':
-        result +=
-            StringFromFormat(ArgumentBuffer.c_str(),
-                             PowerPC::MMU::HostGetString(guard, ap->GetArgT<u32>(guard)).c_str());
-        break;
+      return std::nullopt;
+    };
 
-      case 'a':
-      case 'A':
-      case 'e':
-      case 'E':
-      case 'f':
-      case 'F':
-      case 'g':
-      case 'G':
-        result += StringFromFormat(ArgumentBuffer.c_str(), ap->GetArgT<double>(guard));
-        break;
+    const std::optional<u32> field_width = take_field_or_precision(&left_justified_flag);
+    std::optional<u32> precision = std::nullopt;
+    if (i < string.size() && string[i] == '.')
+    {
+      ++i;
+      precision = take_field_or_precision(nullptr).value_or(0);
+    }
 
-      case 'p':
-        // Override, so 64bit Dolphin prints 32bit pointers, since the ppc is 32bit :)
-        result += StringFromFormat("%x", ap->GetArgT<u32>(guard));
-        break;
+    enum class LengthModifier
+    {
+      None,
+      hh,
+      h,
+      l,
+      ll,
+      L,
+    };
+    auto length_modifier = LengthModifier::None;
+
+    if (i < string.size() && (string[i] == 'h' || string[i] == 'l' || string[i] == 'L'))
+    {
+      ++i;
+      if (i < string.size() && string[i - 1] == 'h' && string[i] == 'h')
+      {
+        ++i;
+        length_modifier = LengthModifier::hh;
+      }
+      else if (i < string.size() && string[i - 1] == 'l' && string[i] == 'l')
+      {
+        ++i;
+        length_modifier = LengthModifier::ll;
+      }
+      else if (string[i - 1] == 'h')
+      {
+        length_modifier = LengthModifier::h;
+      }
+      else if (string[i - 1] == 'l')
+      {
+        length_modifier = LengthModifier::l;
+      }
+      else if (string[i - 1] == 'L')
+      {
+        length_modifier = LengthModifier::L;
+      }
+    }
 
-      case 'n':
-        // %n doesn't output anything, so the result variable is untouched
-        // the actual PPC function will take care of the memory write
-        break;
+    if (i >= string.size())
+    {
+      // not a valid formatting string, print the formatting string as-is
+      result += string.substr(formatting_start_position);
+      break;
+    }
 
-      default:
-        if (string[i - 1] == 'l' && string[i - 2] == 'l')
-          result += StringFromFormat(ArgumentBuffer.c_str(), ap->GetArgT<u64>(guard));
-        else
-          result += StringFromFormat(ArgumentBuffer.c_str(), ap->GetArgT<u32>(guard));
-        break;
+    const char format_specifier = string[i];
+    switch (format_specifier)
+    {
+    case 's':
+    {
+      if (length_modifier == LengthModifier::l)
+      {
+        // This is a bit of a mess... wchar_t could be 16 bits or 32 bits per character depending
+        // on the software. Retail software seems usually to use 16 bits and homebrew 32 bits, but
+        // that's really just a guess. Ideally we can figure out a way to autodetect this, but if
+        // not we should probably just expose a setting for it in the debugger somewhere. For now
+        // we just assume 16 bits.
+        fmt::format_to(
+            std::back_inserter(result), fmt::runtime(left_justified_flag ? "{0:<{1}}" : "{0:>{1}}"),
+            UTF8ToSHIFTJIS(UTF16ToUTF8(args->GetU16String(precision))), field_width.value_or(0));
+      }
+      else
+      {
+        fmt::format_to(std::back_inserter(result),
+                       fmt::runtime(left_justified_flag ? "{0:<{1}}" : "{0:>{1}}"),
+                       args->GetString(precision), field_width.value_or(0));
       }
+      break;
     }
-    else
+    case 'c':
     {
-      result += string[i];
+      const s32 value = Common::BitCast<s32>(args->GetU32());
+      if (length_modifier == LengthModifier::l)
+      {
+        // Same problem as with wide strings here.
+        const char16_t wide_char = static_cast<char16_t>(value);
+        fmt::format_to(std::back_inserter(result),
+                       fmt::runtime(left_justified_flag ? "{0:<{1}}" : "{0:>{1}}"),
+                       UTF8ToSHIFTJIS(UTF16ToUTF8(std::u16string_view(&wide_char, 1))),
+                       field_width.value_or(0));
+      }
+      else
+      {
+        fmt::format_to(std::back_inserter(result),
+                       fmt::runtime(left_justified_flag ? "{0:<{1}}" : "{0:>{1}}"),
+                       static_cast<char>(value), field_width.value_or(0));
+      }
+      break;
+    }
+    case 'd':
+    case 'i':
+    {
+      const auto options = fmt::format(
+          "{}{}{}{}{}{}", left_justified_flag ? "-" : "", sign_prepended_flag ? "+" : "",
+          space_prepended_flag ? " " : "", padding_zeroes_flag ? "0" : "",
+          field_width ? fmt::format("{}", *field_width) : "",
+          precision ? fmt::format(".{}", *precision) : "");
+      if (length_modifier == LengthModifier::ll)
+      {
+        const s64 value = Common::BitCast<s64>(args->GetU64());
+        result += fmt::sprintf(fmt::format("%{}" PRId64, options).c_str(), value);
+      }
+      else
+      {
+        s32 value = Common::BitCast<s32>(args->GetU32());
+        if (length_modifier == LengthModifier::h)
+          value = static_cast<s16>(value);
+        else if (length_modifier == LengthModifier::hh)
+          value = static_cast<s8>(value);
+        result += fmt::sprintf(fmt::format("%{}" PRId32, options).c_str(), value);
+      }
+      break;
+    }
+    case 'o':
+    {
+      const auto options = fmt::format(
+          "{}{}{}{}{}{}{}", left_justified_flag ? "-" : "", sign_prepended_flag ? "+" : "",
+          space_prepended_flag ? " " : "", alternative_form_flag ? "#" : "",
+          padding_zeroes_flag ? "0" : "", field_width ? fmt::format("{}", *field_width) : "",
+          precision ? fmt::format(".{}", *precision) : "");
+      if (length_modifier == LengthModifier::ll)
+      {
+        const u64 value = args->GetU64();
+        result += fmt::sprintf(fmt::format("%{}" PRIo64, options).c_str(), value);
+      }
+      else
+      {
+        u32 value = args->GetU32();
+        if (length_modifier == LengthModifier::h)
+          value = static_cast<u16>(value);
+        else if (length_modifier == LengthModifier::hh)
+          value = static_cast<u8>(value);
+        result += fmt::sprintf(fmt::format("%{}" PRIo32, options).c_str(), value);
+      }
+      break;
+    }
+    case 'x':
+    case 'X':
+    {
+      const auto options = fmt::format(
+          "{}{}{}{}{}{}{}", left_justified_flag ? "-" : "", sign_prepended_flag ? "+" : "",
+          space_prepended_flag ? " " : "", alternative_form_flag ? "#" : "",
+          padding_zeroes_flag ? "0" : "", field_width ? fmt::format("{}", *field_width) : "",
+          precision ? fmt::format(".{}", *precision) : "");
+      if (length_modifier == LengthModifier::ll)
+      {
+        const u64 value = args->GetU64();
+        result += fmt::sprintf(
+            fmt::format("%{}{}", options, format_specifier == 'x' ? PRIx64 : PRIX64).c_str(),
+            value);
+      }
+      else
+      {
+        u32 value = args->GetU32();
+        if (length_modifier == LengthModifier::h)
+          value = static_cast<u16>(value);
+        else if (length_modifier == LengthModifier::hh)
+          value = static_cast<u8>(value);
+        result += fmt::sprintf(
+            fmt::format("%{}{}", options, format_specifier == 'x' ? PRIx32 : PRIX32).c_str(),
+            value);
+      }
+      break;
+    }
+    case 'u':
+    {
+      const auto options = fmt::format(
+          "{}{}{}{}{}{}", left_justified_flag ? "-" : "", sign_prepended_flag ? "+" : "",
+          space_prepended_flag ? " " : "", padding_zeroes_flag ? "0" : "",
+          field_width ? fmt::format("{}", *field_width) : "",
+          precision ? fmt::format(".{}", *precision) : "");
+      if (length_modifier == LengthModifier::ll)
+      {
+        const u64 value = args->GetU64();
+        result += fmt::sprintf(fmt::format("%{}" PRIu64, options).c_str(), value);
+      }
+      else
+      {
+        u32 value = args->GetU32();
+        if (length_modifier == LengthModifier::h)
+          value = static_cast<u16>(value);
+        else if (length_modifier == LengthModifier::hh)
+          value = static_cast<u8>(value);
+        result += fmt::sprintf(fmt::format("%{}" PRIu32, options).c_str(), value);
+      }
+      break;
+    }
+    case 'f':
+    case 'F':
+    case 'e':
+    case 'E':
+    case 'a':
+    case 'A':
+    case 'g':
+    case 'G':
+    {
+      const auto options = fmt::format(
+          "{}{}{}{}{}{}{}", left_justified_flag ? "-" : "", sign_prepended_flag ? "+" : "",
+          space_prepended_flag ? " " : "", alternative_form_flag ? "#" : "",
+          padding_zeroes_flag ? "0" : "", field_width ? fmt::format("{}", *field_width) : "",
+          precision ? fmt::format(".{}", *precision) : "");
+      double value = args->GetF64();
+      result += fmt::sprintf(fmt::format("%{}{}", options, format_specifier).c_str(), value);
+      break;
+    }
+    case 'n':
+      // %n doesn't output anything, so the result variable is untouched
+      // the actual PPC function will take care of the memory write
+      break;
+    case 'p':
+    {
+      const auto options =
+          fmt::format("{}{}{}{}{}", left_justified_flag ? "-" : "", sign_prepended_flag ? "+" : "",
+                      space_prepended_flag ? " " : "", padding_zeroes_flag ? "0" : "",
+                      field_width ? fmt::format("{}", *field_width) : "");
+      const u32 value = args->GetU32();
+      result += fmt::sprintf(fmt::format("%{}" PRIx32, options).c_str(), value);
+      break;
+    }
+    default:
+      // invalid conversion specifier, print the formatting string as-is
+      result += string.substr(formatting_start_position, formatting_start_position - i + 1);
+      break;
     }
   }
 

diff --git a/Source/Core/Core/HLE/HLE_OS.h b/Source/Core/Core/HLE/HLE_OS.h
@@ -3,13 +3,30 @@
 
 #pragma once
 
+#include <optional>
+#include <string>
+
+#include "Common/CommonTypes.h"
+
 namespace Core
 {
 class CPUThreadGuard;
 };
 
 namespace HLE_OS
 {
+class HLEPrintArgs
+{
+public:
+  virtual u32 GetU32() = 0;
+  virtual u64 GetU64() = 0;
+  virtual double GetF64() = 0;
+  virtual std::string GetString(std::optional<u32> max_length) = 0;
+  virtual std::u16string GetU16String(std::optional<u32> max_length) = 0;
+};
+
+std::string GetStringVA(HLEPrintArgs* args, std::string_view string);
+
 void HLE_GeneralDebugPrint(const Core::CPUThreadGuard& guard);
 void HLE_GeneralDebugVPrint(const Core::CPUThreadGuard& guard);
 void HLE_write_console(const Core::CPUThreadGuard& guard);

diff --git a/Source/Core/Core/HLE/HLE_VarArgs.cpp b/Source/Core/Core/HLE/HLE_VarArgs.cpp
@@ -2,29 +2,29 @@
 // SPDX-License-Identifier: GPL-2.0-or-later
 
 #include "Core/HLE/HLE_VarArgs.h"
+#include "Core/Core.h"
 #include "Core/System.h"
 
 #include "Common/Logging/Log.h"
 
 HLE::SystemVABI::VAList::~VAList() = default;
 
-u32 HLE::SystemVABI::VAList::GetGPR(const Core::CPUThreadGuard&, u32 gpr) const
+u32 HLE::SystemVABI::VAList::GetGPR(u32 gpr) const
 {
-  return m_system.GetPPCState().gpr[gpr];
+  return m_guard.GetSystem().GetPPCState().gpr[gpr];
 }
 
-double HLE::SystemVABI::VAList::GetFPR(const Core::CPUThreadGuard&, u32 fpr) const
+double HLE::SystemVABI::VAList::GetFPR(u32 fpr) const
 {
-  return m_system.GetPPCState().ps[fpr].PS0AsDouble();
+  return m_guard.GetSystem().GetPPCState().ps[fpr].PS0AsDouble();
 }
 
-HLE::SystemVABI::VAListStruct::VAListStruct(Core::System& system, const Core::CPUThreadGuard& guard,
-                                            u32 address)
-    : VAList(system, 0), m_va_list{PowerPC::MMU::HostRead_U8(guard, address),
-                                   PowerPC::MMU::HostRead_U8(guard, address + 1),
-                                   PowerPC::MMU::HostRead_U32(guard, address + 4),
-                                   PowerPC::MMU::HostRead_U32(guard, address + 8)},
-      m_address(address), m_has_fpr_area(system.GetPPCState().cr.GetBit(6) == 1)
+HLE::SystemVABI::VAListStruct::VAListStruct(const Core::CPUThreadGuard& guard, u32 address)
+    : VAList(guard, 0), m_va_list{PowerPC::MMU::HostRead_U8(guard, address),
+                                  PowerPC::MMU::HostRead_U8(guard, address + 1),
+                                  PowerPC::MMU::HostRead_U32(guard, address + 4),
+                                  PowerPC::MMU::HostRead_U32(guard, address + 8)},
+      m_address(address), m_has_fpr_area(guard.GetSystem().GetPPCState().cr.GetBit(6) == 1)
 {
   m_stack = m_va_list.overflow_arg_area;
   m_gpr += m_va_list.gpr;
@@ -41,24 +41,24 @@ u32 HLE::SystemVABI::VAListStruct::GetFPRArea() const
   return GetGPRArea() + 4 * 8;
 }
 
-u32 HLE::SystemVABI::VAListStruct::GetGPR(const Core::CPUThreadGuard& guard, u32 gpr) const
+u32 HLE::SystemVABI::VAListStruct::GetGPR(u32 gpr) const
 {
   if (gpr < 3 || gpr > 10)
   {
     ERROR_LOG_FMT(OSHLE, "VAListStruct at {:08x} doesn't have GPR{}!", m_address, gpr);
     return 0;
   }
   const u32 gpr_address = Common::AlignUp(GetGPRArea() + 4 * (gpr - 3), 4);
-  return PowerPC::MMU::HostRead_U32(guard, gpr_address);
+  return PowerPC::MMU::HostRead_U32(m_guard, gpr_address);
 }
 
-double HLE::SystemVABI::VAListStruct::GetFPR(const Core::CPUThreadGuard& guard, u32 fpr) const
+double HLE::SystemVABI::VAListStruct::GetFPR(u32 fpr) const
 {
   if (!m_has_fpr_area || fpr < 1 || fpr > 8)
   {
     ERROR_LOG_FMT(OSHLE, "VAListStruct at {:08x} doesn't have FPR{}!", m_address, fpr);
     return 0.0;
   }
   const u32 fpr_address = Common::AlignUp(GetFPRArea() + 8 * (fpr - 1), 8);
-  return PowerPC::MMU::HostRead_F64(guard, fpr_address);
+  return PowerPC::MMU::HostRead_F64(m_guard, fpr_address);
 }
diff --git a/Source/Core/Core/HLE/HLE_VarArgs.h b/Source/Core/Core/HLE/HLE_VarArgs.h
@@ -3,14 +3,14 @@
 
 #pragma once
 
+#include <type_traits>
+
 #include "Common/Align.h"
 #include "Common/CommonTypes.h"
 
 #include "Core/PowerPC/MMU.h"
 #include "Core/PowerPC/PowerPC.h"
 
-#include <type_traits>
-
 namespace Core
 {
 class CPUThreadGuard;
@@ -38,45 +38,45 @@ constexpr bool IS_ARG_REAL = std::is_floating_point<T>();
 class VAList
 {
 public:
-  explicit VAList(Core::System& system, u32 stack, u32 gpr = 3, u32 fpr = 1, u32 gpr_max = 10,
-                  u32 fpr_max = 8)
-      : m_system(system), m_gpr(gpr), m_fpr(fpr), m_gpr_max(gpr_max), m_fpr_max(fpr_max),
+  explicit VAList(const Core::CPUThreadGuard& guard, u32 stack, u32 gpr = 3, u32 fpr = 1,
+                  u32 gpr_max = 10, u32 fpr_max = 8)
+      : m_guard(guard), m_gpr(gpr), m_fpr(fpr), m_gpr_max(gpr_max), m_fpr_max(fpr_max),
         m_stack(stack)
   {
   }
   virtual ~VAList();
 
   // 0 - arg_ARGPOINTER
   template <typename T, typename std::enable_if_t<IS_ARG_POINTER<T>>* = nullptr>
-  T GetArg(const Core::CPUThreadGuard& guard)
+  T GetArg()
   {
     T obj;
-    u32 addr = GetArg<u32>(guard);
+    u32 addr = GetArg<u32>();
 
     for (size_t i = 0; i < sizeof(T); i += 1, addr += 1)
     {
-      reinterpret_cast<u8*>(&obj)[i] = PowerPC::MMU::HostRead_U8(guard, addr);
+      reinterpret_cast<u8*>(&obj)[i] = PowerPC::MMU::HostRead_U8(m_guard, addr);
     }
 
     return obj;
   }
 
   // 1 - arg_WORD
   template <typename T, typename std::enable_if_t<IS_WORD<T>>* = nullptr>
-  T GetArg(const Core::CPUThreadGuard& guard)
+  T GetArg()
   {
     static_assert(!std::is_pointer<T>(), "VAList doesn't support pointers");
     u64 value;
 
     if (m_gpr <= m_gpr_max)
     {
-      value = GetGPR(guard, m_gpr);
+      value = GetGPR(m_gpr);
       m_gpr += 1;
     }
     else
     {
       m_stack = Common::AlignUp(m_stack, 4);
-      value = PowerPC::MMU::HostRead_U32(guard, m_stack);
+      value = PowerPC::MMU::HostRead_U32(m_guard, m_stack);
       m_stack += 4;
     }
 
@@ -85,21 +85,21 @@ class VAList
 
   // 2 - arg_DOUBLEWORD
   template <typename T, typename std::enable_if_t<IS_DOUBLE_WORD<T>>* = nullptr>
-  T GetArg(const Core::CPUThreadGuard& guard)
+  T GetArg()
   {
     u64 value;
 
     if (m_gpr % 2 == 0)
       m_gpr += 1;
     if (m_gpr < m_gpr_max)
     {
-      value = static_cast<u64>(GetGPR(guard, m_gpr)) << 32 | GetGPR(guard, m_gpr + 1);
+      value = static_cast<u64>(GetGPR(m_gpr)) << 32 | GetGPR(m_gpr + 1);
       m_gpr += 2;
     }
     else
     {
       m_stack = Common::AlignUp(m_stack, 8);
-      value = PowerPC::MMU::HostRead_U64(guard, m_stack);
+      value = PowerPC::MMU::HostRead_U64(m_guard, m_stack);
       m_stack += 8;
     }
 
@@ -108,19 +108,19 @@ class VAList
 
   // 3 - arg_ARGREAL
   template <typename T, typename std::enable_if_t<IS_ARG_REAL<T>>* = nullptr>
-  T GetArg(const Core::CPUThreadGuard& guard)
+  T GetArg()
   {
     double value;
 
     if (m_fpr <= m_fpr_max)
     {
-      value = GetFPR(guard, m_fpr);
+      value = GetFPR(m_fpr);
       m_fpr += 1;
     }
     else
     {
       m_stack = Common::AlignUp(m_stack, 8);
-      value = PowerPC::MMU::HostRead_F64(guard, m_stack);
+      value = PowerPC::MMU::HostRead_F64(m_guard, m_stack);
       m_stack += 8;
     }
 
@@ -129,22 +129,22 @@ class VAList
 
   // Helper
   template <typename T>
-  T GetArgT(const Core::CPUThreadGuard& guard)
+  T GetArgT()
   {
-    return static_cast<T>(GetArg<T>(guard));
+    return static_cast<T>(GetArg<T>());
   }
 
 protected:
-  Core::System& m_system;
+  const Core::CPUThreadGuard& m_guard;
   u32 m_gpr = 3;
   u32 m_fpr = 1;
   const u32 m_gpr_max = 10;
   const u32 m_fpr_max = 8;
   u32 m_stack;
 
 private:
-  virtual u32 GetGPR(const Core::CPUThreadGuard& guard, u32 gpr) const;
-  virtual double GetFPR(const Core::CPUThreadGuard& guard, u32 fpr) const;
+  virtual u32 GetGPR(u32 gpr) const;
+  virtual double GetFPR(u32 fpr) const;
 };
 
 // See System V ABI (SVR4) for more details
@@ -156,7 +156,7 @@ class VAList
 class VAListStruct : public VAList
 {
 public:
-  explicit VAListStruct(Core::System& system, const Core::CPUThreadGuard& guard, u32 address);
+  explicit VAListStruct(const Core::CPUThreadGuard& guard, u32 address);
   ~VAListStruct() = default;
 
 private:
@@ -174,8 +174,8 @@ class VAListStruct : public VAList
   u32 GetGPRArea() const;
   u32 GetFPRArea() const;
 
-  u32 GetGPR(const Core::CPUThreadGuard& guard, u32 gpr) const override;
-  double GetFPR(const Core::CPUThreadGuard& guard, u32 fpr) const override;
+  u32 GetGPR(u32 gpr) const override;
+  double GetFPR(u32 fpr) const override;
 };
 
 }  // namespace HLE::SystemVABI
diff --git a/Source/Core/Core/PowerPC/MMU.cpp b/Source/Core/Core/PowerPC/MMU.cpp
@@ -867,6 +867,22 @@ std::string MMU::HostGetString(const Core::CPUThreadGuard& guard, u32 address, s
   return s;
 }
 
+std::u16string MMU::HostGetU16String(const Core::CPUThreadGuard& guard, u32 address, size_t size)
+{
+  std::u16string s;
+  do
+  {
+    if (!HostIsRAMAddress(guard, address) || !HostIsRAMAddress(guard, address + 1))
+      break;
+    const u16 res = HostRead_U16(guard, address);
+    if (!res)
+      break;
+    s += static_cast<char16_t>(res);
+    address += 2;
+  } while (size == 0 || s.length() < size);
+  return s;
+}
+
 std::optional<ReadResult<std::string>> MMU::HostTryReadString(const Core::CPUThreadGuard& guard,
                                                               u32 address, size_t size,
                                                               RequestedAddressSpace space)

diff --git a/Source/Core/Core/PowerPC/MMU.h b/Source/Core/Core/PowerPC/MMU.h
@@ -132,6 +132,8 @@ class MMU
   static double HostRead_F64(const Core::CPUThreadGuard& guard, u32 address);
   static u32 HostRead_Instruction(const Core::CPUThreadGuard& guard, u32 address);
   static std::string HostGetString(const Core::CPUThreadGuard& guard, u32 address, size_t size = 0);
+  static std::u16string HostGetU16String(const Core::CPUThreadGuard& guard, u32 address,
+                                         size_t size = 0);
 
   // Try to read a value from emulated memory at the given address in the given memory space.
   // If the read succeeds, the returned value will be present and the ReadResult contains the read