Use ScratchBuffer to reduce memory allocations in audio_core

Remove fake vertex bindings when dynamic state is enabled

.ci/scripts/format/script.sh

@@ -10,7 +10,7 @@ if grep -nrI '\s$' src *.yml *.txt *.md Doxyfile .gitignore .gitmodules .ci* dis
 fi
 
 # Default clang-format points to default 3.5 version one
-CLANG_FORMAT=clang-format-12
+CLANG_FORMAT=${CLANG_FORMAT:-clang-format-15}
 $CLANG_FORMAT --version
 
 if [ "$TRAVIS_EVENT_TYPE" = "pull_request" ]; then

.ci/scripts/merge/apply-patches-by-label.py

@@ -2,15 +2,12 @@
 # SPDX-License-Identifier: GPL-2.0-or-later
 
 # Download all pull requests as patches that match a specific label
-# Usage: python download-patches-by-label.py <Label to Match> <Root Path Folder to DL to>
+# Usage: python apply-patches-by-label.py <Label to Match>
 
-import requests, sys, json, urllib3.request, shutil, subprocess, os, traceback
+import json, requests, subprocess, sys, traceback
 
 tagline = sys.argv[2]
 
-http = urllib3.PoolManager()
-dl_list = {}
-
 def check_individual(labels):
     for label in labels:
         if (label["name"] == sys.argv[1]):
@@ -18,8 +15,9 @@ def check_individual(labels):
     return False
 
 def do_page(page):
-    url = 'https://api.github.com/repos/yuzu-emu/yuzu/pulls?page=%s' % page
+    url = f"https://api.github.com/repos/yuzu-emu/yuzu/pulls?page={page}"
     response = requests.get(url)
+    response.raise_for_status()
     if (response.ok):
         j = json.loads(response.content)
         if j == []:
@@ -27,13 +25,13 @@ def do_page(page):
         for pr in j:
             if (check_individual(pr["labels"])):
                 pn = pr["number"]
-                print("Matched PR# %s" % pn)
-                print(subprocess.check_output(["git", "fetch", "https://github.com/yuzu-emu/yuzu.git", "pull/%s/head:pr-%s" % (pn, pn), "-f", "--no-recurse-submodules"]))
-                print(subprocess.check_output(["git", "merge", "--squash", "pr-%s" % pn]))
-                print(subprocess.check_output(["git", "commit", "-m\"Merge %s PR %s\"" % (tagline, pn)]))
+                print(f"Matched PR# {pn}")
+                print(subprocess.check_output(["git", "fetch", "https://github.com/yuzu-emu/yuzu.git", f"pull/{pn}/head:pr-{pn}", "-f", "--no-recurse-submodules"]))
+                print(subprocess.check_output(["git", "merge", "--squash", f"pr-{pn}"]))
+                print(subprocess.check_output(["git", "commit", f"-m\"Merge {tagline} PR {pn}\""]))
 
 try:
-    for i in range(1,30):
+    for i in range(1,10):
         do_page(i)
 except:
     traceback.print_exc(file=sys.stdout)

CMakeLists.txt

@@ -3,17 +3,14 @@
 
 cmake_minimum_required(VERSION 3.22)
 
-# Dynarmic has cmake_minimum_required(3.12) and we may want to override
-# some of its variables, which is only possible in 3.13+
-set(CMAKE_POLICY_DEFAULT_CMP0077 NEW)
+project(yuzu)
 
 list(APPEND CMAKE_MODULE_PATH "${CMAKE_CURRENT_SOURCE_DIR}/CMakeModules")
 list(APPEND CMAKE_MODULE_PATH "${CMAKE_CURRENT_SOURCE_DIR}/externals/cmake-modules")
-list(APPEND CMAKE_MODULE_PATH "${CMAKE_CURRENT_SOURCE_DIR}/externals/find-modules")
+
 include(DownloadExternals)
 include(CMakeDependentOption)
-
-project(yuzu)
+include(CTest)
 
 # Set bundled sdl2/qt as dependent options.
 # OFF by default, but if ENABLE_SDL2 and MSVC are true then ON
@@ -47,7 +44,7 @@ option(ENABLE_CUBEB "Enables the cubeb audio backend" ON)
 
 option(USE_DISCORD_PRESENCE "Enables Discord Rich Presence" OFF)
 
-option(YUZU_TESTS "Compile tests" ON)
+option(YUZU_TESTS "Compile tests" "${BUILD_TESTING}")
 
 option(YUZU_USE_PRECOMPILED_HEADERS "Use precompiled headers" ON)
 
@@ -207,46 +204,48 @@ set(CMAKE_RUNTIME_OUTPUT_DIRECTORY ${PROJECT_BINARY_DIR}/bin)
 # System imported libraries
 # =======================================================================
 
-find_package(enet 1.3)
+# Enforce the search mode of non-required packages for better and shorter failure messages
+find_package(enet 1.3 MODULE)
 find_package(fmt 9 REQUIRED)
-find_package(inih)
+find_package(inih MODULE)
+find_package(LLVM MODULE)
 find_package(lz4 REQUIRED)
 find_package(nlohmann_json 3.8 REQUIRED)
-find_package(Opus 1.3)
+find_package(Opus 1.3 MODULE)
 find_package(ZLIB 1.2 REQUIRED)
 find_package(zstd 1.5 REQUIRED)
 
 if (NOT YUZU_USE_EXTERNAL_VULKAN_HEADERS)
-    find_package(Vulkan 1.3.238)
+    find_package(Vulkan 1.3.238 REQUIRED)
 endif()
 
 if (ENABLE_LIBUSB)
-    find_package(libusb 1.0.24)
+    find_package(libusb 1.0.24 MODULE)
 endif()
 
 if (ARCHITECTURE_x86 OR ARCHITECTURE_x86_64)
-    find_package(xbyak 6 QUIET)
+    find_package(xbyak 6 CONFIG)
 endif()
 
 if (ARCHITECTURE_x86_64 OR ARCHITECTURE_arm64)
-    find_package(dynarmic 6.4.0 QUIET)
+    find_package(dynarmic 6.4.0 CONFIG)
 endif()
 
 if (ENABLE_CUBEB)
-    find_package(cubeb QUIET)
+    find_package(cubeb CONFIG)
 endif()
 
 if (USE_DISCORD_PRESENCE)
-    find_package(DiscordRPC QUIET)
+    find_package(DiscordRPC MODULE)
 endif()
 
 if (ENABLE_WEB_SERVICE)
-    find_package(cpp-jwt 1.4 QUIET)
-    find_package(httplib 0.11 QUIET)
+    find_package(cpp-jwt 1.4 CONFIG)
+    find_package(httplib 0.11 MODULE)
 endif()
 
 if (YUZU_TESTS)
-    find_package(Catch2 2.13.7 REQUIRED)
+    find_package(Catch2 3.0.1 REQUIRED)
 endif()
 
 find_package(Boost 1.73.0 COMPONENTS context)
@@ -515,7 +514,7 @@ endif()
 # against all the src files. This should be used before making a pull request.
 # =======================================================================
 
-set(CLANG_FORMAT_POSTFIX "-12")
+set(CLANG_FORMAT_POSTFIX "-15")
 find_program(CLANG_FORMAT
     NAMES clang-format${CLANG_FORMAT_POSTFIX}
           clang-format
@@ -610,7 +609,6 @@ if (YUZU_USE_FASTER_LD AND CMAKE_CXX_COMPILER_ID STREQUAL "GNU")
     endif()
 endif()
 
-enable_testing()
 add_subdirectory(externals)
 add_subdirectory(src)
 

CMakeModules/FindLLVM.cmake

@@ -0,0 +1,16 @@
+# SPDX-FileCopyrightText: 2023 Alexandre Bouvier <contact@amb.tf>
+#
+# SPDX-License-Identifier: GPL-3.0-or-later
+
+find_package(LLVM QUIET CONFIG)
+
+include(FindPackageHandleStandardArgs)
+find_package_handle_standard_args(LLVM CONFIG_MODE)
+
+if (LLVM_FOUND AND NOT TARGET LLVM::Demangle)
+    add_library(LLVM::Demangle INTERFACE IMPORTED)
+    llvm_map_components_to_libnames(LLVM_LIBRARIES demangle)
+    target_compile_definitions(LLVM::Demangle INTERFACE ${LLVM_DEFINITIONS})
+    target_include_directories(LLVM::Demangle INTERFACE ${LLVM_INCLUDE_DIRS})
+    target_link_libraries(LLVM::Demangle INTERFACE ${LLVM_LIBRARIES})
+endif()

CMakeModules/Findhttplib.cmake

@@ -5,7 +5,7 @@
 include(FindPackageHandleStandardArgs)
 
 find_package(httplib QUIET CONFIG)
-if (httplib_FOUND)
+if (httplib_CONSIDERED_CONFIGS)
     find_package_handle_standard_args(httplib CONFIG_MODE)
 else()
     find_package(PkgConfig QUIET)

CMakeModules/Findlz4.cmake

@@ -4,7 +4,7 @@
 include(FindPackageHandleStandardArgs)
 
 find_package(lz4 QUIET CONFIG)
-if (lz4_FOUND)
+if (lz4_CONSIDERED_CONFIGS)
     find_package_handle_standard_args(lz4 CONFIG_MODE)
 else()
     find_package(PkgConfig QUIET)

CMakeModules/Findzstd.cmake

@@ -4,7 +4,7 @@
 include(FindPackageHandleStandardArgs)
 
 find_package(zstd QUIET CONFIG)
-if (zstd_FOUND)
+if (zstd_CONSIDERED_CONFIGS)
     find_package_handle_standard_args(zstd CONFIG_MODE)
 else()
     find_package(PkgConfig QUIET)

LICENSES/LLVM-exception.txt

@@ -0,0 +1,15 @@
+---- LLVM Exceptions to the Apache 2.0 License ----
+
+   As an exception, if, as a result of your compiling your source code, portions
+   of this Software are embedded into an Object form of such source code, you
+   may redistribute such embedded portions in such Object form without complying
+   with the conditions of Sections 4(a), 4(b) and 4(d) of the License.
+
+   In addition, if you combine or link compiled forms of this Software with
+   software that is licensed under the GPLv2 ("Combined Software") and if a
+   court of competent jurisdiction determines that the patent provision (Section
+   3), the indemnity provision (Section 9) or other Section of the License
+   conflicts with the conditions of the GPLv2, you may retroactively and
+   prospectively choose to deem waived or otherwise exclude such Section(s) of
+   the License, but only in their entirety and only with respect to the Combined
+   Software.

dist/org.yuzu_emu.yuzu.metainfo.xml

@@ -53,10 +53,10 @@ SPDX-License-Identifier: CC0-1.0
   <developer_name>yuzu Emulator Team</developer_name>
   <content_rating type="oars-1.0"/>
   <screenshots>
-    <screenshot type="default">https://raw.githubusercontent.com/yuzu-emu/yuzu-emu.github.io/master/images/screenshots/001-Super%20Mario%20Odyssey%20.png</screenshot>
-    <screenshot>https://raw.githubusercontent.com/yuzu-emu/yuzu-emu.github.io/master/images/screenshots/004-The%20Legend%20of%20Zelda%20Skyward%20Sword%20HD.png</screenshot>
-    <screenshot>https://raw.githubusercontent.com/yuzu-emu/yuzu-emu.github.io/master/images/screenshots/007-Pokemon%20Sword.png</screenshot>
-    <screenshot>https://raw.githubusercontent.com/yuzu-emu/yuzu-emu.github.io/master/images/screenshots/010-Hyrule%20Warriors%20Age%20of%20Calamity.png</screenshot>
-    <screenshot>https://raw.githubusercontent.com/yuzu-emu/yuzu-emu.github.io/master/images/screenshots/039-Pok%C3%A9mon%20Mystery%20Dungeon%20Rescue%20Team%20DX.png.png.png</screenshot>
+    <screenshot type="default"><image>https://raw.githubusercontent.com/yuzu-emu/yuzu-emu.github.io/master/images/screenshots/001-Super%20Mario%20Odyssey%20.png</image></screenshot>
+    <screenshot><image>https://raw.githubusercontent.com/yuzu-emu/yuzu-emu.github.io/master/images/screenshots/004-The%20Legend%20of%20Zelda%20Skyward%20Sword%20HD.png</image></screenshot>
+    <screenshot><image>https://raw.githubusercontent.com/yuzu-emu/yuzu-emu.github.io/master/images/screenshots/007-Pokemon%20Sword.png</image></screenshot>
+    <screenshot><image>https://raw.githubusercontent.com/yuzu-emu/yuzu-emu.github.io/master/images/screenshots/010-Hyrule%20Warriors%20Age%20of%20Calamity.png</image></screenshot>
+    <screenshot><image>https://raw.githubusercontent.com/yuzu-emu/yuzu-emu.github.io/master/images/screenshots/039-Pok%C3%A9mon%20Mystery%20Dungeon%20Rescue%20Team%20DX.png.png.png</image></screenshot>
   </screenshots>
 </component>

externals/CMakeLists.txt

@@ -1,9 +1,12 @@
 # SPDX-FileCopyrightText: 2016 Citra Emulator Project
 # SPDX-License-Identifier: GPL-2.0-or-later
 
-list(APPEND CMAKE_MODULE_PATH "${PROJECT_SOURCE_DIR}/CMakeModules")
-list(APPEND CMAKE_MODULE_PATH "${PROJECT_SOURCE_DIR}/externals/find-modules")
-include(DownloadExternals)
+# Dynarmic has cmake_minimum_required(3.12) and we may want to override
+# some of its variables, which is only possible in 3.13+
+set(CMAKE_POLICY_DEFAULT_CMP0077 NEW)
+
+# Disable tests in all externals supporting the standard option name
+set(BUILD_TESTING OFF)
 
 # xbyak
 if ((ARCHITECTURE_x86 OR ARCHITECTURE_x86_64) AND NOT TARGET xbyak::xbyak)
@@ -12,8 +15,7 @@ endif()
 
 # Dynarmic
 if ((ARCHITECTURE_x86_64 OR ARCHITECTURE_arm64) AND NOT TARGET dynarmic::dynarmic)
-    set(DYNARMIC_NO_BUNDLED_FMT ON)
-    set(DYNARMIC_IGNORE_ASSERTS ON CACHE BOOL "" FORCE)
+    set(DYNARMIC_IGNORE_ASSERTS ON)
     add_subdirectory(dynarmic EXCLUDE_FROM_ALL)
     add_library(dynarmic::dynarmic ALIAS dynarmic)
 endif()
@@ -60,10 +62,10 @@ if (YUZU_USE_EXTERNAL_SDL2)
             Locale Power Render)
         foreach(_SUB ${SDL_UNUSED_SUBSYSTEMS})
           string(TOUPPER ${_SUB} _OPT)
-          option(SDL_${_OPT} "" OFF)
+          set(SDL_${_OPT} OFF)
         endforeach()
 
-        option(HIDAPI "" ON)
+        set(HIDAPI ON)
     endif()
     set(SDL_STATIC ON)
     set(SDL_SHARED OFF)
@@ -83,7 +85,7 @@ endif()
 
 # Cubeb
 if (ENABLE_CUBEB AND NOT TARGET cubeb::cubeb)
-    set(BUILD_TESTS OFF CACHE BOOL "")
+    set(BUILD_TESTS OFF)
     add_subdirectory(cubeb EXCLUDE_FROM_ALL)
     add_library(cubeb::cubeb ALIAS cubeb)
 endif()
@@ -98,6 +100,7 @@ endif()
 # Sirit
 add_subdirectory(sirit EXCLUDE_FROM_ALL)
 
+# httplib
 if (ENABLE_WEB_SERVICE AND NOT TARGET httplib::httplib)
     if (NOT WIN32)
         find_package(OpenSSL 1.1)
@@ -108,7 +111,7 @@ if (ENABLE_WEB_SERVICE AND NOT TARGET httplib::httplib)
 
     if (WIN32 OR NOT OPENSSL_FOUND)
         # LibreSSL
-        set(LIBRESSL_SKIP_INSTALL ON CACHE BOOL "")
+        set(LIBRESSL_SKIP_INSTALL ON)
         set(OPENSSLDIR "/etc/ssl/")
         add_subdirectory(libressl EXCLUDE_FROM_ALL)
         target_include_directories(ssl INTERFACE ./libressl/include)
@@ -118,7 +121,6 @@ if (ENABLE_WEB_SERVICE AND NOT TARGET httplib::httplib)
             DEFINITION OPENSSL_LIBS)
     endif()
 
-    # httplib
     add_library(httplib INTERFACE)
     target_include_directories(httplib INTERFACE ./cpp-httplib)
     target_compile_definitions(httplib INTERFACE -DCPPHTTPLIB_OPENSSL_SUPPORT)
@@ -155,3 +157,10 @@ endif()
 if (YUZU_USE_EXTERNAL_VULKAN_HEADERS)
     add_subdirectory(Vulkan-Headers EXCLUDE_FROM_ALL)
 endif()
+
+if (NOT TARGET LLVM::Demangle)
+    add_library(demangle STATIC)
+    target_include_directories(demangle PUBLIC ./demangle)
+    target_sources(demangle PRIVATE demangle/ItaniumDemangle.cpp)
+    add_library(LLVM::Demangle ALIAS demangle)
+endif()

externals/demangle/ItaniumDemangle.cpp

@@ -0,0 +1,588 @@
+//===------------------------- ItaniumDemangle.cpp ------------------------===//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-FileCopyrightText: Part of the LLVM Project
+// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+//
+//===----------------------------------------------------------------------===//
+
+// FIXME: (possibly) incomplete list of features that clang mangles that this
+// file does not yet support:
+//   - C++ modules TS
+
+#include "llvm/Demangle/Demangle.h"
+#include "llvm/Demangle/ItaniumDemangle.h"
+
+#include <cassert>
+#include <cctype>
+#include <cstdio>
+#include <cstdlib>
+#include <cstring>
+#include <functional>
+#include <numeric>
+#include <utility>
+#include <vector>
+
+using namespace llvm;
+using namespace llvm::itanium_demangle;
+
+constexpr const char *itanium_demangle::FloatData<float>::spec;
+constexpr const char *itanium_demangle::FloatData<double>::spec;
+constexpr const char *itanium_demangle::FloatData<long double>::spec;
+
+// <discriminator> := _ <non-negative number>      # when number < 10
+//                 := __ <non-negative number> _   # when number >= 10
+//  extension      := decimal-digit+               # at the end of string
+const char *itanium_demangle::parse_discriminator(const char *first,
+                                                  const char *last) {
+  // parse but ignore discriminator
+  if (first != last) {
+    if (*first == '_') {
+      const char *t1 = first + 1;
+      if (t1 != last) {
+        if (std::isdigit(*t1))
+          first = t1 + 1;
+        else if (*t1 == '_') {
+          for (++t1; t1 != last && std::isdigit(*t1); ++t1)
+            ;
+          if (t1 != last && *t1 == '_')
+            first = t1 + 1;
+        }
+      }
+    } else if (std::isdigit(*first)) {
+      const char *t1 = first + 1;
+      for (; t1 != last && std::isdigit(*t1); ++t1)
+        ;
+      if (t1 == last)
+        first = last;
+    }
+  }
+  return first;
+}
+
+#ifndef NDEBUG
+namespace {
+struct DumpVisitor {
+  unsigned Depth = 0;
+  bool PendingNewline = false;
+
+  template<typename NodeT> static constexpr bool wantsNewline(const NodeT *) {
+    return true;
+  }
+  static bool wantsNewline(NodeArray A) { return !A.empty(); }
+  static constexpr bool wantsNewline(...) { return false; }
+
+  template<typename ...Ts> static bool anyWantNewline(Ts ...Vs) {
+    for (bool B : {wantsNewline(Vs)...})
+      if (B)
+        return true;
+    return false;
+  }
+
+  void printStr(const char *S) { fprintf(stderr, "%s", S); }
+  void print(StringView SV) {
+    fprintf(stderr, "\"%.*s\"", (int)SV.size(), SV.begin());
+  }
+  void print(const Node *N) {
+    if (N)
+      N->visit(std::ref(*this));
+    else
+      printStr("<null>");
+  }
+  void print(NodeOrString NS) {
+    if (NS.isNode())
+      print(NS.asNode());
+    else if (NS.isString())
+      print(NS.asString());
+    else
+      printStr("NodeOrString()");
+  }
+  void print(NodeArray A) {
+    ++Depth;
+    printStr("{");
+    bool First = true;
+    for (const Node *N : A) {
+      if (First)
+        print(N);
+      else
+        printWithComma(N);
+      First = false;
+    }
+    printStr("}");
+    --Depth;
+  }
+
+  // Overload used when T is exactly 'bool', not merely convertible to 'bool'.
+  void print(bool B) { printStr(B ? "true" : "false"); }
+
+  template <class T>
+  typename std::enable_if<std::is_unsigned<T>::value>::type print(T N) {
+    fprintf(stderr, "%llu", (unsigned long long)N);
+  }
+
+  template <class T>
+  typename std::enable_if<std::is_signed<T>::value>::type print(T N) {
+    fprintf(stderr, "%lld", (long long)N);
+  }
+
+  void print(ReferenceKind RK) {
+    switch (RK) {
+    case ReferenceKind::LValue:
+      return printStr("ReferenceKind::LValue");
+    case ReferenceKind::RValue:
+      return printStr("ReferenceKind::RValue");
+    }
+  }
+  void print(FunctionRefQual RQ) {
+    switch (RQ) {
+    case FunctionRefQual::FrefQualNone:
+      return printStr("FunctionRefQual::FrefQualNone");
+    case FunctionRefQual::FrefQualLValue:
+      return printStr("FunctionRefQual::FrefQualLValue");
+    case FunctionRefQual::FrefQualRValue:
+      return printStr("FunctionRefQual::FrefQualRValue");
+    }
+  }
+  void print(Qualifiers Qs) {
+    if (!Qs) return printStr("QualNone");
+    struct QualName { Qualifiers Q; const char *Name; } Names[] = {
+      {QualConst, "QualConst"},
+      {QualVolatile, "QualVolatile"},
+      {QualRestrict, "QualRestrict"},
+    };
+    for (QualName Name : Names) {
+      if (Qs & Name.Q) {
+        printStr(Name.Name);
+        Qs = Qualifiers(Qs & ~Name.Q);
+        if (Qs) printStr(" | ");
+      }
+    }
+  }
+  void print(SpecialSubKind SSK) {
+    switch (SSK) {
+    case SpecialSubKind::allocator:
+      return printStr("SpecialSubKind::allocator");
+    case SpecialSubKind::basic_string:
+      return printStr("SpecialSubKind::basic_string");
+    case SpecialSubKind::string:
+      return printStr("SpecialSubKind::string");
+    case SpecialSubKind::istream:
+      return printStr("SpecialSubKind::istream");
+    case SpecialSubKind::ostream:
+      return printStr("SpecialSubKind::ostream");
+    case SpecialSubKind::iostream:
+      return printStr("SpecialSubKind::iostream");
+    }
+  }
+  void print(TemplateParamKind TPK) {
+    switch (TPK) {
+    case TemplateParamKind::Type:
+      return printStr("TemplateParamKind::Type");
+    case TemplateParamKind::NonType:
+      return printStr("TemplateParamKind::NonType");
+    case TemplateParamKind::Template:
+      return printStr("TemplateParamKind::Template");
+    }
+  }
+
+  void newLine() {
+    printStr("\n");
+    for (unsigned I = 0; I != Depth; ++I)
+      printStr(" ");
+    PendingNewline = false;
+  }
+
+  template<typename T> void printWithPendingNewline(T V) {
+    print(V);
+    if (wantsNewline(V))
+      PendingNewline = true;
+  }
+
+  template<typename T> void printWithComma(T V) {
+    if (PendingNewline || wantsNewline(V)) {
+      printStr(",");
+      newLine();
+    } else {
+      printStr(", ");
+    }
+
+    printWithPendingNewline(V);
+  }
+
+  struct CtorArgPrinter {
+    DumpVisitor &Visitor;
+
+    template<typename T, typename ...Rest> void operator()(T V, Rest ...Vs) {
+      if (Visitor.anyWantNewline(V, Vs...))
+        Visitor.newLine();
+      Visitor.printWithPendingNewline(V);
+      int PrintInOrder[] = { (Visitor.printWithComma(Vs), 0)..., 0 };
+      (void)PrintInOrder;
+    }
+  };
+
+  template<typename NodeT> void operator()(const NodeT *Node) {
+    Depth += 2;
+    fprintf(stderr, "%s(", itanium_demangle::NodeKind<NodeT>::name());
+    Node->match(CtorArgPrinter{*this});
+    fprintf(stderr, ")");
+    Depth -= 2;
+  }
+
+  void operator()(const ForwardTemplateReference *Node) {
+    Depth += 2;
+    fprintf(stderr, "ForwardTemplateReference(");
+    if (Node->Ref && !Node->Printing) {
+      Node->Printing = true;
+      CtorArgPrinter{*this}(Node->Ref);
+      Node->Printing = false;
+    } else {
+      CtorArgPrinter{*this}(Node->Index);
+    }
+    fprintf(stderr, ")");
+    Depth -= 2;
+  }
+};
+}
+
+void itanium_demangle::Node::dump() const {
+  DumpVisitor V;
+  visit(std::ref(V));
+  V.newLine();
+}
+#endif
+
+namespace {
+class BumpPointerAllocator {
+  struct BlockMeta {
+    BlockMeta* Next;
+    size_t Current;
+  };
+
+  static constexpr size_t AllocSize = 4096;
+  static constexpr size_t UsableAllocSize = AllocSize - sizeof(BlockMeta);
+
+  alignas(long double) char InitialBuffer[AllocSize];
+  BlockMeta* BlockList = nullptr;
+
+  void grow() {
+    char* NewMeta = static_cast<char *>(std::malloc(AllocSize));
+    if (NewMeta == nullptr)
+      std::terminate();
+    BlockList = new (NewMeta) BlockMeta{BlockList, 0};
+  }
+
+  void* allocateMassive(size_t NBytes) {
+    NBytes += sizeof(BlockMeta);
+    BlockMeta* NewMeta = reinterpret_cast<BlockMeta*>(std::malloc(NBytes));
+    if (NewMeta == nullptr)
+      std::terminate();
+    BlockList->Next = new (NewMeta) BlockMeta{BlockList->Next, 0};
+    return static_cast<void*>(NewMeta + 1);
+  }
+
+public:
+  BumpPointerAllocator()
+      : BlockList(new (InitialBuffer) BlockMeta{nullptr, 0}) {}
+
+  void* allocate(size_t N) {
+    N = (N + 15u) & ~15u;
+    if (N + BlockList->Current >= UsableAllocSize) {
+      if (N > UsableAllocSize)
+        return allocateMassive(N);
+      grow();
+    }
+    BlockList->Current += N;
+    return static_cast<void*>(reinterpret_cast<char*>(BlockList + 1) +
+                              BlockList->Current - N);
+  }
+
+  void reset() {
+    while (BlockList) {
+      BlockMeta* Tmp = BlockList;
+      BlockList = BlockList->Next;
+      if (reinterpret_cast<char*>(Tmp) != InitialBuffer)
+        std::free(Tmp);
+    }
+    BlockList = new (InitialBuffer) BlockMeta{nullptr, 0};
+  }
+
+  ~BumpPointerAllocator() { reset(); }
+};
+
+class DefaultAllocator {
+  BumpPointerAllocator Alloc;
+
+public:
+  void reset() { Alloc.reset(); }
+
+  template<typename T, typename ...Args> T *makeNode(Args &&...args) {
+    return new (Alloc.allocate(sizeof(T)))
+        T(std::forward<Args>(args)...);
+  }
+
+  void *allocateNodeArray(size_t sz) {
+    return Alloc.allocate(sizeof(Node *) * sz);
+  }
+};
+}  // unnamed namespace
+
+//===----------------------------------------------------------------------===//
+// Code beyond this point should not be synchronized with libc++abi.
+//===----------------------------------------------------------------------===//
+
+using Demangler = itanium_demangle::ManglingParser<DefaultAllocator>;
+
+char *llvm::itaniumDemangle(const char *MangledName, char *Buf,
+                            size_t *N, int *Status) {
+  if (MangledName == nullptr || (Buf != nullptr && N == nullptr)) {
+    if (Status)
+      *Status = demangle_invalid_args;
+    return nullptr;
+  }
+
+  int InternalStatus = demangle_success;
+  Demangler Parser(MangledName, MangledName + std::strlen(MangledName));
+  OutputStream S;
+
+  Node *AST = Parser.parse();
+
+  if (AST == nullptr)
+    InternalStatus = demangle_invalid_mangled_name;
+  else if (!initializeOutputStream(Buf, N, S, 1024))
+    InternalStatus = demangle_memory_alloc_failure;
+  else {
+    assert(Parser.ForwardTemplateRefs.empty());
+    AST->print(S);
+    S += '\0';
+    if (N != nullptr)
+      *N = S.getCurrentPosition();
+    Buf = S.getBuffer();
+  }
+
+  if (Status)
+    *Status = InternalStatus;
+  return InternalStatus == demangle_success ? Buf : nullptr;
+}
+
+ItaniumPartialDemangler::ItaniumPartialDemangler()
+    : RootNode(nullptr), Context(new Demangler{nullptr, nullptr}) {}
+
+ItaniumPartialDemangler::~ItaniumPartialDemangler() {
+  delete static_cast<Demangler *>(Context);
+}
+
+ItaniumPartialDemangler::ItaniumPartialDemangler(
+    ItaniumPartialDemangler &&Other)
+    : RootNode(Other.RootNode), Context(Other.Context) {
+  Other.Context = Other.RootNode = nullptr;
+}
+
+ItaniumPartialDemangler &ItaniumPartialDemangler::
+operator=(ItaniumPartialDemangler &&Other) {
+  std::swap(RootNode, Other.RootNode);
+  std::swap(Context, Other.Context);
+  return *this;
+}
+
+// Demangle MangledName into an AST, storing it into this->RootNode.
+bool ItaniumPartialDemangler::partialDemangle(const char *MangledName) {
+  Demangler *Parser = static_cast<Demangler *>(Context);
+  size_t Len = std::strlen(MangledName);
+  Parser->reset(MangledName, MangledName + Len);
+  RootNode = Parser->parse();
+  return RootNode == nullptr;
+}
+
+static char *printNode(const Node *RootNode, char *Buf, size_t *N) {
+  OutputStream S;
+  if (!initializeOutputStream(Buf, N, S, 128))
+    return nullptr;
+  RootNode->print(S);
+  S += '\0';
+  if (N != nullptr)
+    *N = S.getCurrentPosition();
+  return S.getBuffer();
+}
+
+char *ItaniumPartialDemangler::getFunctionBaseName(char *Buf, size_t *N) const {
+  if (!isFunction())
+    return nullptr;
+
+  const Node *Name = static_cast<const FunctionEncoding *>(RootNode)->getName();
+
+  while (true) {
+    switch (Name->getKind()) {
+    case Node::KAbiTagAttr:
+      Name = static_cast<const AbiTagAttr *>(Name)->Base;
+      continue;
+    case Node::KStdQualifiedName:
+      Name = static_cast<const StdQualifiedName *>(Name)->Child;
+      continue;
+    case Node::KNestedName:
+      Name = static_cast<const NestedName *>(Name)->Name;
+      continue;
+    case Node::KLocalName:
+      Name = static_cast<const LocalName *>(Name)->Entity;
+      continue;
+    case Node::KNameWithTemplateArgs:
+      Name = static_cast<const NameWithTemplateArgs *>(Name)->Name;
+      continue;
+    default:
+      return printNode(Name, Buf, N);
+    }
+  }
+}
+
+char *ItaniumPartialDemangler::getFunctionDeclContextName(char *Buf,
+                                                          size_t *N) const {
+  if (!isFunction())
+    return nullptr;
+  const Node *Name = static_cast<const FunctionEncoding *>(RootNode)->getName();
+
+  OutputStream S;
+  if (!initializeOutputStream(Buf, N, S, 128))
+    return nullptr;
+
+ KeepGoingLocalFunction:
+  while (true) {
+    if (Name->getKind() == Node::KAbiTagAttr) {
+      Name = static_cast<const AbiTagAttr *>(Name)->Base;
+      continue;
+    }
+    if (Name->getKind() == Node::KNameWithTemplateArgs) {
+      Name = static_cast<const NameWithTemplateArgs *>(Name)->Name;
+      continue;
+    }
+    break;
+  }
+
+  switch (Name->getKind()) {
+  case Node::KStdQualifiedName:
+    S += "std";
+    break;
+  case Node::KNestedName:
+    static_cast<const NestedName *>(Name)->Qual->print(S);
+    break;
+  case Node::KLocalName: {
+    auto *LN = static_cast<const LocalName *>(Name);
+    LN->Encoding->print(S);
+    S += "::";
+    Name = LN->Entity;
+    goto KeepGoingLocalFunction;
+  }
+  default:
+    break;
+  }
+  S += '\0';
+  if (N != nullptr)
+    *N = S.getCurrentPosition();
+  return S.getBuffer();
+}
+
+char *ItaniumPartialDemangler::getFunctionName(char *Buf, size_t *N) const {
+  if (!isFunction())
+    return nullptr;
+  auto *Name = static_cast<FunctionEncoding *>(RootNode)->getName();
+  return printNode(Name, Buf, N);
+}
+
+char *ItaniumPartialDemangler::getFunctionParameters(char *Buf,
+                                                     size_t *N) const {
+  if (!isFunction())
+    return nullptr;
+  NodeArray Params = static_cast<FunctionEncoding *>(RootNode)->getParams();
+
+  OutputStream S;
+  if (!initializeOutputStream(Buf, N, S, 128))
+    return nullptr;
+
+  S += '(';
+  Params.printWithComma(S);
+  S += ')';
+  S += '\0';
+  if (N != nullptr)
+    *N = S.getCurrentPosition();
+  return S.getBuffer();
+}
+
+char *ItaniumPartialDemangler::getFunctionReturnType(
+    char *Buf, size_t *N) const {
+  if (!isFunction())
+    return nullptr;
+
+  OutputStream S;
+  if (!initializeOutputStream(Buf, N, S, 128))
+    return nullptr;
+
+  if (const Node *Ret =
+          static_cast<const FunctionEncoding *>(RootNode)->getReturnType())
+    Ret->print(S);
+
+  S += '\0';
+  if (N != nullptr)
+    *N = S.getCurrentPosition();
+  return S.getBuffer();
+}
+
+char *ItaniumPartialDemangler::finishDemangle(char *Buf, size_t *N) const {
+  assert(RootNode != nullptr && "must call partialDemangle()");
+  return printNode(static_cast<Node *>(RootNode), Buf, N);
+}
+
+bool ItaniumPartialDemangler::hasFunctionQualifiers() const {
+  assert(RootNode != nullptr && "must call partialDemangle()");
+  if (!isFunction())
+    return false;
+  auto *E = static_cast<const FunctionEncoding *>(RootNode);
+  return E->getCVQuals() != QualNone || E->getRefQual() != FrefQualNone;
+}
+
+bool ItaniumPartialDemangler::isCtorOrDtor() const {
+  const Node *N = static_cast<const Node *>(RootNode);
+  while (N) {
+    switch (N->getKind()) {
+    default:
+      return false;
+    case Node::KCtorDtorName:
+      return true;
+
+    case Node::KAbiTagAttr:
+      N = static_cast<const AbiTagAttr *>(N)->Base;
+      break;
+    case Node::KFunctionEncoding:
+      N = static_cast<const FunctionEncoding *>(N)->getName();
+      break;
+    case Node::KLocalName:
+      N = static_cast<const LocalName *>(N)->Entity;
+      break;
+    case Node::KNameWithTemplateArgs:
+      N = static_cast<const NameWithTemplateArgs *>(N)->Name;
+      break;
+    case Node::KNestedName:
+      N = static_cast<const NestedName *>(N)->Name;
+      break;
+    case Node::KStdQualifiedName:
+      N = static_cast<const StdQualifiedName *>(N)->Child;
+      break;
+    }
+  }
+  return false;
+}
+
+bool ItaniumPartialDemangler::isFunction() const {
+  assert(RootNode != nullptr && "must call partialDemangle()");
+  return static_cast<const Node *>(RootNode)->getKind() ==
+         Node::KFunctionEncoding;
+}
+
+bool ItaniumPartialDemangler::isSpecialName() const {
+  assert(RootNode != nullptr && "must call partialDemangle()");
+  auto K = static_cast<const Node *>(RootNode)->getKind();
+  return K == Node::KSpecialName || K == Node::KCtorVtableSpecialName;
+}
+
+bool ItaniumPartialDemangler::isData() const {
+  return !isFunction() && !isSpecialName();
+}

externals/demangle/llvm/Demangle/Demangle.h

@@ -0,0 +1,104 @@
+//===--- Demangle.h ---------------------------------------------*- C++ -*-===//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-FileCopyrightText: Part of the LLVM Project
+// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+//
+//===----------------------------------------------------------------------===//
+
+#ifndef LLVM_DEMANGLE_DEMANGLE_H
+#define LLVM_DEMANGLE_DEMANGLE_H
+
+#include <cstddef>
+#include <string>
+
+namespace llvm {
+/// This is a llvm local version of __cxa_demangle. Other than the name and
+/// being in the llvm namespace it is identical.
+///
+/// The mangled_name is demangled into buf and returned. If the buffer is not
+/// large enough, realloc is used to expand it.
+///
+/// The *status will be set to a value from the following enumeration
+enum : int {
+  demangle_unknown_error = -4,
+  demangle_invalid_args = -3,
+  demangle_invalid_mangled_name = -2,
+  demangle_memory_alloc_failure = -1,
+  demangle_success = 0,
+};
+
+char *itaniumDemangle(const char *mangled_name, char *buf, size_t *n,
+                      int *status);
+
+
+enum MSDemangleFlags {
+  MSDF_None = 0,
+  MSDF_DumpBackrefs = 1 << 0,
+  MSDF_NoAccessSpecifier = 1 << 1,
+  MSDF_NoCallingConvention = 1 << 2,
+  MSDF_NoReturnType = 1 << 3,
+  MSDF_NoMemberType = 1 << 4,
+};
+char *microsoftDemangle(const char *mangled_name, char *buf, size_t *n,
+                        int *status, MSDemangleFlags Flags = MSDF_None);
+
+/// "Partial" demangler. This supports demangling a string into an AST
+/// (typically an intermediate stage in itaniumDemangle) and querying certain
+/// properties or partially printing the demangled name.
+struct ItaniumPartialDemangler {
+  ItaniumPartialDemangler();
+
+  ItaniumPartialDemangler(ItaniumPartialDemangler &&Other);
+  ItaniumPartialDemangler &operator=(ItaniumPartialDemangler &&Other);
+
+  /// Demangle into an AST. Subsequent calls to the rest of the member functions
+  /// implicitly operate on the AST this produces.
+  /// \return true on error, false otherwise
+  bool partialDemangle(const char *MangledName);
+
+  /// Just print the entire mangled name into Buf. Buf and N behave like the
+  /// second and third parameters to itaniumDemangle.
+  char *finishDemangle(char *Buf, size_t *N) const;
+
+  /// Get the base name of a function. This doesn't include trailing template
+  /// arguments, ie for "a::b<int>" this function returns "b".
+  char *getFunctionBaseName(char *Buf, size_t *N) const;
+
+  /// Get the context name for a function. For "a::b::c", this function returns
+  /// "a::b".
+  char *getFunctionDeclContextName(char *Buf, size_t *N) const;
+
+  /// Get the entire name of this function.
+  char *getFunctionName(char *Buf, size_t *N) const;
+
+  /// Get the parameters for this function.
+  char *getFunctionParameters(char *Buf, size_t *N) const;
+  char *getFunctionReturnType(char *Buf, size_t *N) const;
+
+  /// If this function has any any cv or reference qualifiers. These imply that
+  /// the function is a non-static member function.
+  bool hasFunctionQualifiers() const;
+
+  /// If this symbol describes a constructor or destructor.
+  bool isCtorOrDtor() const;
+
+  /// If this symbol describes a function.
+  bool isFunction() const;
+
+  /// If this symbol describes a variable.
+  bool isData() const;
+
+  /// If this symbol is a <special-name>. These are generally implicitly
+  /// generated by the implementation, such as vtables and typeinfo names.
+  bool isSpecialName() const;
+
+  ~ItaniumPartialDemangler();
+private:
+  void *RootNode;
+  void *Context;
+};
+} // namespace llvm
+
+#endif

externals/demangle/llvm/Demangle/DemangleConfig.h

@@ -0,0 +1,93 @@
+//===--- DemangleConfig.h ---------------------------------------*- C++ -*-===//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-FileCopyrightText: Part of the LLVM Project
+// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+//
+//===----------------------------------------------------------------------===//
+//
+// This file contains a variety of feature test macros copied from
+// include/llvm/Support/Compiler.h so that LLVMDemangle does not need to take
+// a dependency on LLVMSupport.
+//
+//===----------------------------------------------------------------------===//
+
+#ifndef LLVM_DEMANGLE_COMPILER_H
+#define LLVM_DEMANGLE_COMPILER_H
+
+#ifndef __has_feature
+#define __has_feature(x) 0
+#endif
+
+#ifndef __has_cpp_attribute
+#define __has_cpp_attribute(x) 0
+#endif
+
+#ifndef __has_attribute
+#define __has_attribute(x) 0
+#endif
+
+#ifndef __has_builtin
+#define __has_builtin(x) 0
+#endif
+
+#ifndef DEMANGLE_GNUC_PREREQ
+#if defined(__GNUC__) && defined(__GNUC_MINOR__) && defined(__GNUC_PATCHLEVEL__)
+#define DEMANGLE_GNUC_PREREQ(maj, min, patch)                           \
+  ((__GNUC__ << 20) + (__GNUC_MINOR__ << 10) + __GNUC_PATCHLEVEL__ >=          \
+   ((maj) << 20) + ((min) << 10) + (patch))
+#elif defined(__GNUC__) && defined(__GNUC_MINOR__)
+#define DEMANGLE_GNUC_PREREQ(maj, min, patch)                           \
+  ((__GNUC__ << 20) + (__GNUC_MINOR__ << 10) >= ((maj) << 20) + ((min) << 10))
+#else
+#define DEMANGLE_GNUC_PREREQ(maj, min, patch) 0
+#endif
+#endif
+
+#if __has_attribute(used) || DEMANGLE_GNUC_PREREQ(3, 1, 0)
+#define DEMANGLE_ATTRIBUTE_USED __attribute__((__used__))
+#else
+#define DEMANGLE_ATTRIBUTE_USED
+#endif
+
+#if __has_builtin(__builtin_unreachable) || DEMANGLE_GNUC_PREREQ(4, 5, 0)
+#define DEMANGLE_UNREACHABLE __builtin_unreachable()
+#elif defined(_MSC_VER)
+#define DEMANGLE_UNREACHABLE __assume(false)
+#else
+#define DEMANGLE_UNREACHABLE
+#endif
+
+#if __has_attribute(noinline) || DEMANGLE_GNUC_PREREQ(3, 4, 0)
+#define DEMANGLE_ATTRIBUTE_NOINLINE __attribute__((noinline))
+#elif defined(_MSC_VER)
+#define DEMANGLE_ATTRIBUTE_NOINLINE __declspec(noinline)
+#else
+#define DEMANGLE_ATTRIBUTE_NOINLINE
+#endif
+
+#if !defined(NDEBUG)
+#define DEMANGLE_DUMP_METHOD DEMANGLE_ATTRIBUTE_NOINLINE DEMANGLE_ATTRIBUTE_USED
+#else
+#define DEMANGLE_DUMP_METHOD DEMANGLE_ATTRIBUTE_NOINLINE
+#endif
+
+#if __cplusplus > 201402L && __has_cpp_attribute(fallthrough)
+#define DEMANGLE_FALLTHROUGH [[fallthrough]]
+#elif __has_cpp_attribute(gnu::fallthrough)
+#define DEMANGLE_FALLTHROUGH [[gnu::fallthrough]]
+#elif !__cplusplus
+// Workaround for llvm.org/PR23435, since clang 3.6 and below emit a spurious
+// error when __has_cpp_attribute is given a scoped attribute in C mode.
+#define DEMANGLE_FALLTHROUGH
+#elif __has_cpp_attribute(clang::fallthrough)
+#define DEMANGLE_FALLTHROUGH [[clang::fallthrough]]
+#else
+#define DEMANGLE_FALLTHROUGH
+#endif
+
+#define DEMANGLE_NAMESPACE_BEGIN namespace llvm { namespace itanium_demangle {
+#define DEMANGLE_NAMESPACE_END } }
+
+#endif

externals/demangle/llvm/Demangle/StringView.h

@@ -0,0 +1,127 @@
+//===--- StringView.h -------------------------------------------*- C++ -*-===//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-FileCopyrightText: Part of the LLVM Project
+// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+//
+//===----------------------------------------------------------------------===//
+//
+// FIXME: Use std::string_view instead when we support C++17.
+//
+//===----------------------------------------------------------------------===//
+
+#ifndef DEMANGLE_STRINGVIEW_H
+#define DEMANGLE_STRINGVIEW_H
+
+#include "DemangleConfig.h"
+#include <algorithm>
+#include <cassert>
+#include <cstring>
+
+DEMANGLE_NAMESPACE_BEGIN
+
+class StringView {
+  const char *First;
+  const char *Last;
+
+public:
+  static const size_t npos = ~size_t(0);
+
+  template <size_t N>
+  StringView(const char (&Str)[N]) : First(Str), Last(Str + N - 1) {}
+  StringView(const char *First_, const char *Last_)
+      : First(First_), Last(Last_) {}
+  StringView(const char *First_, size_t Len)
+      : First(First_), Last(First_ + Len) {}
+  StringView(const char *Str) : First(Str), Last(Str + std::strlen(Str)) {}
+  StringView() : First(nullptr), Last(nullptr) {}
+
+  StringView substr(size_t From) const {
+    return StringView(begin() + From, size() - From);
+  }
+
+  size_t find(char C, size_t From = 0) const {
+    size_t FindBegin = std::min(From, size());
+    // Avoid calling memchr with nullptr.
+    if (FindBegin < size()) {
+      // Just forward to memchr, which is faster than a hand-rolled loop.
+      if (const void *P = ::memchr(First + FindBegin, C, size() - FindBegin))
+        return size_t(static_cast<const char *>(P) - First);
+    }
+    return npos;
+  }
+
+  StringView substr(size_t From, size_t To) const {
+    if (To >= size())
+      To = size() - 1;
+    if (From >= size())
+      From = size() - 1;
+    return StringView(First + From, First + To);
+  }
+
+  StringView dropFront(size_t N = 1) const {
+    if (N >= size())
+      N = size();
+    return StringView(First + N, Last);
+  }
+
+  StringView dropBack(size_t N = 1) const {
+    if (N >= size())
+      N = size();
+    return StringView(First, Last - N);
+  }
+
+  char front() const {
+    assert(!empty());
+    return *begin();
+  }
+
+  char back() const {
+    assert(!empty());
+    return *(end() - 1);
+  }
+
+  char popFront() {
+    assert(!empty());
+    return *First++;
+  }
+
+  bool consumeFront(char C) {
+    if (!startsWith(C))
+      return false;
+    *this = dropFront(1);
+    return true;
+  }
+
+  bool consumeFront(StringView S) {
+    if (!startsWith(S))
+      return false;
+    *this = dropFront(S.size());
+    return true;
+  }
+
+  bool startsWith(char C) const { return !empty() && *begin() == C; }
+
+  bool startsWith(StringView Str) const {
+    if (Str.size() > size())
+      return false;
+    return std::equal(Str.begin(), Str.end(), begin());
+  }
+
+  const char &operator[](size_t Idx) const { return *(begin() + Idx); }
+
+  const char *begin() const { return First; }
+  const char *end() const { return Last; }
+  size_t size() const { return static_cast<size_t>(Last - First); }
+  bool empty() const { return First == Last; }
+};
+
+inline bool operator==(const StringView &LHS, const StringView &RHS) {
+  return LHS.size() == RHS.size() &&
+         std::equal(LHS.begin(), LHS.end(), RHS.begin());
+}
+
+DEMANGLE_NAMESPACE_END
+
+#endif

externals/demangle/llvm/Demangle/Utility.h

@@ -0,0 +1,192 @@
+//===--- Utility.h ----------------------------------------------*- C++ -*-===//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-FileCopyrightText: Part of the LLVM Project
+// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+//
+//===----------------------------------------------------------------------===//
+//
+// Provide some utility classes for use in the demangler(s).
+//
+//===----------------------------------------------------------------------===//
+
+#ifndef DEMANGLE_UTILITY_H
+#define DEMANGLE_UTILITY_H
+
+#include "StringView.h"
+#include <cstdint>
+#include <cstdlib>
+#include <cstring>
+#include <iterator>
+#include <limits>
+
+DEMANGLE_NAMESPACE_BEGIN
+
+// Stream that AST nodes write their string representation into after the AST
+// has been parsed.
+class OutputStream {
+  char *Buffer;
+  size_t CurrentPosition;
+  size_t BufferCapacity;
+
+  // Ensure there is at least n more positions in buffer.
+  void grow(size_t N) {
+    if (N + CurrentPosition >= BufferCapacity) {
+      BufferCapacity *= 2;
+      if (BufferCapacity < N + CurrentPosition)
+        BufferCapacity = N + CurrentPosition;
+      Buffer = static_cast<char *>(std::realloc(Buffer, BufferCapacity));
+      if (Buffer == nullptr)
+        std::terminate();
+    }
+  }
+
+  void writeUnsigned(uint64_t N, bool isNeg = false) {
+    // Handle special case...
+    if (N == 0) {
+      *this << '0';
+      return;
+    }
+
+    char Temp[21];
+    char *TempPtr = std::end(Temp);
+
+    while (N) {
+      *--TempPtr = '0' + char(N % 10);
+      N /= 10;
+    }
+
+    // Add negative sign...
+    if (isNeg)
+      *--TempPtr = '-';
+    this->operator<<(StringView(TempPtr, std::end(Temp)));
+  }
+
+public:
+  OutputStream(char *StartBuf, size_t Size)
+      : Buffer(StartBuf), CurrentPosition(0), BufferCapacity(Size) {}
+  OutputStream() = default;
+  void reset(char *Buffer_, size_t BufferCapacity_) {
+    CurrentPosition = 0;
+    Buffer = Buffer_;
+    BufferCapacity = BufferCapacity_;
+  }
+
+  /// If a ParameterPackExpansion (or similar type) is encountered, the offset
+  /// into the pack that we're currently printing.
+  unsigned CurrentPackIndex = std::numeric_limits<unsigned>::max();
+  unsigned CurrentPackMax = std::numeric_limits<unsigned>::max();
+
+  OutputStream &operator+=(StringView R) {
+    size_t Size = R.size();
+    if (Size == 0)
+      return *this;
+    grow(Size);
+    std::memmove(Buffer + CurrentPosition, R.begin(), Size);
+    CurrentPosition += Size;
+    return *this;
+  }
+
+  OutputStream &operator+=(char C) {
+    grow(1);
+    Buffer[CurrentPosition++] = C;
+    return *this;
+  }
+
+  OutputStream &operator<<(StringView R) { return (*this += R); }
+
+  OutputStream &operator<<(char C) { return (*this += C); }
+
+  OutputStream &operator<<(long long N) {
+    if (N < 0)
+      writeUnsigned(static_cast<unsigned long long>(-N), true);
+    else
+      writeUnsigned(static_cast<unsigned long long>(N));
+    return *this;
+  }
+
+  OutputStream &operator<<(unsigned long long N) {
+    writeUnsigned(N, false);
+    return *this;
+  }
+
+  OutputStream &operator<<(long N) {
+    return this->operator<<(static_cast<long long>(N));
+  }
+
+  OutputStream &operator<<(unsigned long N) {
+    return this->operator<<(static_cast<unsigned long long>(N));
+  }
+
+  OutputStream &operator<<(int N) {
+    return this->operator<<(static_cast<long long>(N));
+  }
+
+  OutputStream &operator<<(unsigned int N) {
+    return this->operator<<(static_cast<unsigned long long>(N));
+  }
+
+  size_t getCurrentPosition() const { return CurrentPosition; }
+  void setCurrentPosition(size_t NewPos) { CurrentPosition = NewPos; }
+
+  char back() const {
+    return CurrentPosition ? Buffer[CurrentPosition - 1] : '\0';
+  }
+
+  bool empty() const { return CurrentPosition == 0; }
+
+  char *getBuffer() { return Buffer; }
+  char *getBufferEnd() { return Buffer + CurrentPosition - 1; }
+  size_t getBufferCapacity() { return BufferCapacity; }
+};
+
+template <class T> class SwapAndRestore {
+  T &Restore;
+  T OriginalValue;
+  bool ShouldRestore = true;
+
+public:
+  SwapAndRestore(T &Restore_) : SwapAndRestore(Restore_, Restore_) {}
+
+  SwapAndRestore(T &Restore_, T NewVal)
+      : Restore(Restore_), OriginalValue(Restore) {
+    Restore = std::move(NewVal);
+  }
+  ~SwapAndRestore() {
+    if (ShouldRestore)
+      Restore = std::move(OriginalValue);
+  }
+
+  void shouldRestore(bool ShouldRestore_) { ShouldRestore = ShouldRestore_; }
+
+  void restoreNow(bool Force) {
+    if (!Force && !ShouldRestore)
+      return;
+
+    Restore = std::move(OriginalValue);
+    ShouldRestore = false;
+  }
+
+  SwapAndRestore(const SwapAndRestore &) = delete;
+  SwapAndRestore &operator=(const SwapAndRestore &) = delete;
+};
+
+inline bool initializeOutputStream(char *Buf, size_t *N, OutputStream &S,
+                                   size_t InitSize) {
+  size_t BufferSize;
+  if (Buf == nullptr) {
+    Buf = static_cast<char *>(std::malloc(InitSize));
+    if (Buf == nullptr)
+      return false;
+    BufferSize = InitSize;
+  } else
+    BufferSize = *N;
+
+  S.reset(Buf, BufferSize);
+  return true;
+}
+
+DEMANGLE_NAMESPACE_END
+
+#endif

src/audio_core/device/audio_buffers.h

@@ -48,7 +48,7 @@ public:
      *
      * @param out_buffers - The buffers which were registered.
      */
-    void RegisterBuffers(std::vector<AudioBuffer>& out_buffers) {
+    void RegisterBuffers(std::span<AudioBuffer> out_buffers, u32& out_size) {
         std::scoped_lock l{lock};
         const s32 to_register{std::min(std::min(appended_count, BufferAppendLimit),
                                        BufferAppendLimit - registered_count)};
@@ -59,7 +59,7 @@ public:
                 index += N;
             }
 
-            out_buffers.push_back(buffers[index]);
+            out_buffers[out_size++] = buffers[index];
             registered_count++;
             registered_index = (registered_index + 1) % append_limit;
 
@@ -162,7 +162,7 @@ public:
      * @param max_buffers     - Maximum number of buffers to released.
      * @return The number of buffers released.
      */
-    u32 GetRegisteredAppendedBuffers(std::vector<AudioBuffer>& buffers_flushed, u32 max_buffers) {
+    u32 GetRegisteredAppendedBuffers(std::span<AudioBuffer> out_buffers, u32 max_buffers) {
         std::scoped_lock l{lock};
         if (registered_count + appended_count == 0) {
             return 0;
@@ -174,19 +174,20 @@ public:
             return 0;
         }
 
+        u32 buffers_flushed{0};
         while (registered_count > 0) {
             auto index{registered_index - registered_count};
             if (index < 0) {
                 index += N;
             }
 
-            buffers_flushed.push_back(buffers[index]);
+            out_buffers[buffers_flushed++] = buffers[index];
 
             registered_count--;
             released_count++;
             released_index = (released_index + 1) % append_limit;
 
-            if (buffers_flushed.size() >= buffers_to_flush) {
+            if (buffers_flushed >= buffers_to_flush) {
                 break;
             }
         }
@@ -197,18 +198,18 @@ public:
                 index += N;
             }
 
-            buffers_flushed.push_back(buffers[index]);
+            out_buffers[buffers_flushed++] = buffers[index];
 
             appended_count--;
             released_count++;
             released_index = (released_index + 1) % append_limit;
 
-            if (buffers_flushed.size() >= buffers_to_flush) {
+            if (buffers_flushed >= buffers_to_flush) {
                 break;
             }
         }
 
-        return static_cast<u32>(buffers_flushed.size());
+        return buffers_flushed;
     }
 
     /**
@@ -270,8 +271,9 @@ public:
      */
     bool FlushBuffers(u32& buffers_released) {
         std::scoped_lock l{lock};
-        std::vector<AudioBuffer> buffers_flushed{};
 
+        static Common::ScratchBuffer<AudioBuffer> buffers_flushed{};
+        buffers_flushed.resize_destructive(append_limit);
         buffers_released = GetRegisteredAppendedBuffers(buffers_flushed, append_limit);
 
         if (registered_count > 0) {

src/audio_core/device/device_session.cpp

@@ -6,6 +6,7 @@
 #include "audio_core/device/audio_buffer.h"
 #include "audio_core/device/device_session.h"
 #include "audio_core/sink/sink_stream.h"
+#include "common/scratch_buffer.h"
 #include "core/core.h"
 #include "core/core_timing.h"
 #include "core/memory.h"
@@ -79,21 +80,22 @@ void DeviceSession::ClearBuffers() {
     }
 }
 
-void DeviceSession::AppendBuffers(std::span<const AudioBuffer> buffers) const {
-    for (const auto& buffer : buffers) {
+void DeviceSession::AppendBuffers(std::span<const AudioBuffer> buffers, u32 out_size) const {
+    static Common::ScratchBuffer<s16> samples{};
+
+    for (u32 i = 0; i < out_size; i++) {
         Sink::SinkBuffer new_buffer{
-            .frames = buffer.size / (channel_count * sizeof(s16)),
+            .frames = buffers[i].size / (channel_count * sizeof(s16)),
             .frames_played = 0,
-            .tag = buffer.tag,
+            .tag = buffers[i].tag,
             .consumed = false,
         };
 
         if (type == Sink::StreamType::In) {
-            std::vector<s16> samples{};
             stream->AppendBuffer(new_buffer, samples);
         } else {
-            std::vector<s16> samples(buffer.size / sizeof(s16));
-            system.Memory().ReadBlockUnsafe(buffer.samples, samples.data(), buffer.size);
+            samples.resize_destructive(buffers[i].size / sizeof(s16));
+            system.Memory().ReadBlockUnsafe(buffers[i].samples, samples.data(), buffers[i].size);
             stream->AppendBuffer(new_buffer, samples);
         }
     }

src/audio_core/device/device_session.h

@@ -62,7 +62,7 @@ public:
      *
      * @param buffers - The buffers to play.
      */
-    void AppendBuffers(std::span<const AudioBuffer> buffers) const;
+    void AppendBuffers(std::span<const AudioBuffer> buffers, u32 out_size) const;
 
     /**
      * (Audio In only) Pop samples from the backend, and write them back to this buffer's address.

src/audio_core/in/audio_in_system.cpp

@@ -89,9 +89,10 @@ Result System::Start() {
     session->Start();
     state = State::Started;
 
-    std::vector<AudioBuffer> buffers_to_flush{};
-    buffers.RegisterBuffers(buffers_to_flush);
-    session->AppendBuffers(buffers_to_flush);
+    std::array<AudioBuffer, BufferAppendLimit> buffers_to_flush{};
+    u32 out_size{0};
+    buffers.RegisterBuffers(buffers_to_flush, out_size);
+    session->AppendBuffers(buffers_to_flush, out_size);
     session->SetRingSize(static_cast<u32>(buffers_to_flush.size()));
 
     return ResultSuccess;
@@ -134,9 +135,10 @@ bool System::AppendBuffer(const AudioInBuffer& buffer, const u64 tag) {
 
 void System::RegisterBuffers() {
     if (state == State::Started) {
-        std::vector<AudioBuffer> registered_buffers{};
-        buffers.RegisterBuffers(registered_buffers);
-        session->AppendBuffers(registered_buffers);
+        std::array<AudioBuffer, BufferAppendLimit> registered_buffers{};
+        u32 out_size{0};
+        buffers.RegisterBuffers(registered_buffers, out_size);
+        session->AppendBuffers(registered_buffers, out_size);
     }
 }
 

src/audio_core/out/audio_out_system.cpp

@@ -89,9 +89,10 @@ Result System::Start() {
     session->Start();
     state = State::Started;
 
-    std::vector<AudioBuffer> buffers_to_flush{};
-    buffers.RegisterBuffers(buffers_to_flush);
-    session->AppendBuffers(buffers_to_flush);
+    std::array<AudioBuffer, BufferAppendLimit> buffers_to_flush{};
+    u32 out_size{0};
+    buffers.RegisterBuffers(buffers_to_flush, out_size);
+    session->AppendBuffers(buffers_to_flush, out_size);
     session->SetRingSize(static_cast<u32>(buffers_to_flush.size()));
 
     return ResultSuccess;
@@ -134,9 +135,10 @@ bool System::AppendBuffer(const AudioOutBuffer& buffer, u64 tag) {
 
 void System::RegisterBuffers() {
     if (state == State::Started) {
-        std::vector<AudioBuffer> registered_buffers{};
-        buffers.RegisterBuffers(registered_buffers);
-        session->AppendBuffers(registered_buffers);
+        std::array<AudioBuffer, BufferAppendLimit> registered_buffers{};
+        u32 out_size{0};
+        buffers.RegisterBuffers(registered_buffers, out_size);
+        session->AppendBuffers(registered_buffers, out_size);
     }
 }
 

src/audio_core/renderer/command/data_source/decode.cpp

@@ -8,6 +8,7 @@
 #include "audio_core/renderer/command/resample/resample.h"
 #include "common/fixed_point.h"
 #include "common/logging/log.h"
+#include "common/scratch_buffer.h"
 #include "core/memory.h"
 
 namespace AudioCore::AudioRenderer {
@@ -29,6 +30,7 @@ static u32 DecodePcm(Core::Memory::Memory& memory, std::span<s16> out_buffer,
                      const DecodeArg& req) {
     constexpr s32 min{std::numeric_limits<s16>::min()};
     constexpr s32 max{std::numeric_limits<s16>::max()};
+    static Common::ScratchBuffer<T> samples;
 
     if (req.buffer == 0 || req.buffer_size == 0) {
         return 0;
@@ -49,7 +51,7 @@ static u32 DecodePcm(Core::Memory::Memory& memory, std::span<s16> out_buffer,
         const u64 size{channel_count * samples_to_decode};
         const u64 size_bytes{size * sizeof(T)};
 
-        std::vector<T> samples(size);
+        samples.resize_destructive(size);
         memory.ReadBlockUnsafe(source, samples.data(), size_bytes);
 
         if constexpr (std::is_floating_point_v<T>) {
@@ -73,7 +75,7 @@ static u32 DecodePcm(Core::Memory::Memory& memory, std::span<s16> out_buffer,
         }
 
         const VAddr source{req.buffer + ((req.start_offset + req.offset) * sizeof(T))};
-        std::vector<T> samples(samples_to_decode);
+        samples.resize_destructive(samples_to_decode);
         memory.ReadBlockUnsafe(source, samples.data(), samples_to_decode * sizeof(T));
 
         if constexpr (std::is_floating_point_v<T>) {
@@ -103,6 +105,7 @@ static u32 DecodeAdpcm(Core::Memory::Memory& memory, std::span<s16> out_buffer,
                        const DecodeArg& req) {
     constexpr u32 SamplesPerFrame{14};
     constexpr u32 NibblesPerFrame{16};
+    static Common::ScratchBuffer<u8> wavebuffer;
 
     if (req.buffer == 0 || req.buffer_size == 0) {
         return 0;
@@ -138,7 +141,7 @@ static u32 DecodeAdpcm(Core::Memory::Memory& memory, std::span<s16> out_buffer,
     }
 
     const auto size{std::max((samples_to_process / 8U) * SamplesPerFrame, 8U)};
-    std::vector<u8> wavebuffer(size);
+    wavebuffer.resize_destructive(size);
     memory.ReadBlockUnsafe(req.buffer + position_in_frame / 2, wavebuffer.data(),
                            wavebuffer.size());
 
@@ -227,6 +230,8 @@ static u32 DecodeAdpcm(Core::Memory::Memory& memory, std::span<s16> out_buffer,
  * @param args   - The wavebuffer data, and information for how to decode it.
  */
 void DecodeFromWaveBuffers(Core::Memory::Memory& memory, const DecodeFromWaveBuffersArgs& args) {
+    Common::ScratchBuffer<s16> temp_buffer(TempBufferSize);
+
     auto& voice_state{*args.voice_state};
     auto remaining_sample_count{args.sample_count};
     auto fraction{voice_state.fraction};
@@ -256,9 +261,8 @@ void DecodeFromWaveBuffers(Core::Memory::Memory& memory, const DecodeFromWaveBuf
 
     bool is_buffer_starved{false};
     u32 offset{voice_state.offset};
-
+    std::memset(temp_buffer.data(), 0, temp_buffer.size() * sizeof(s16));
     auto output_buffer{args.output};
-    std::vector<s16> temp_buffer(TempBufferSize, 0);
 
     while (remaining_sample_count > 0) {
         const auto samples_to_write{std::min(remaining_sample_count, max_remaining_sample_count)};

src/audio_core/renderer/command/effect/compressor.cpp

@@ -8,6 +8,7 @@
 #include "audio_core/renderer/adsp/command_list_processor.h"
 #include "audio_core/renderer/command/effect/compressor.h"
 #include "audio_core/renderer/effect/compressor.h"
+#include "common/scratch_buffer.h"
 
 namespace AudioCore::AudioRenderer {
 
@@ -44,8 +45,8 @@ static void InitializeCompressorEffect(const CompressorInfo::ParameterVersion2&
 
 static void ApplyCompressorEffect(const CompressorInfo::ParameterVersion2& params,
                                   CompressorInfo::State& state, bool enabled,
-                                  std::vector<std::span<const s32>> input_buffers,
-                                  std::vector<std::span<s32>> output_buffers, u32 sample_count) {
+                                  std::span<std::span<const s32>> input_buffers,
+                                  std::span<std::span<s32>> output_buffers, u32 sample_count) {
     if (enabled) {
         auto state_00{state.unk_00};
         auto state_04{state.unk_04};
@@ -124,8 +125,10 @@ void CompressorCommand::Dump([[maybe_unused]] const ADSP::CommandListProcessor&
 }
 
 void CompressorCommand::Process(const ADSP::CommandListProcessor& processor) {
-    std::vector<std::span<const s32>> input_buffers(parameter.channel_count);
-    std::vector<std::span<s32>> output_buffers(parameter.channel_count);
+    static Common::ScratchBuffer<std::span<const s32>> input_buffers{};
+    static Common::ScratchBuffer<std::span<s32>> output_buffers{};
+    input_buffers.resize_destructive(parameter.channel_count);
+    output_buffers.resize_destructive(parameter.channel_count);
 
     for (s16 i = 0; i < parameter.channel_count; i++) {
         input_buffers[i] = processor.mix_buffers.subspan(inputs[i] * processor.sample_count,

src/audio_core/renderer/command/effect/delay.cpp

@@ -3,6 +3,7 @@
 
 #include "audio_core/renderer/adsp/command_list_processor.h"
 #include "audio_core/renderer/command/effect/delay.h"
+#include "common/scratch_buffer.h"
 
 namespace AudioCore::AudioRenderer {
 /**
@@ -74,8 +75,8 @@ static void InitializeDelayEffect(const DelayInfo::ParameterVersion1& params,
  */
 template <size_t NumChannels>
 static void ApplyDelay(const DelayInfo::ParameterVersion1& params, DelayInfo::State& state,
-                       std::vector<std::span<const s32>>& inputs,
-                       std::vector<std::span<s32>>& outputs, const u32 sample_count) {
+                       std::span<std::span<const s32>> inputs, std::span<std::span<s32>> outputs,
+                       const u32 sample_count) {
     for (u32 sample_index = 0; sample_index < sample_count; sample_index++) {
         std::array<Common::FixedPoint<50, 14>, NumChannels> input_samples{};
         for (u32 channel = 0; channel < NumChannels; channel++) {
@@ -153,8 +154,8 @@ static void ApplyDelay(const DelayInfo::ParameterVersion1& params, DelayInfo::St
  * @param sample_count - Number of samples to process.
  */
 static void ApplyDelayEffect(const DelayInfo::ParameterVersion1& params, DelayInfo::State& state,
-                             const bool enabled, std::vector<std::span<const s32>>& inputs,
-                             std::vector<std::span<s32>>& outputs, const u32 sample_count) {
+                             const bool enabled, std::span<std::span<const s32>> inputs,
+                             std::span<std::span<s32>> outputs, const u32 sample_count) {
 
     if (!IsChannelCountValid(params.channel_count)) {
         LOG_ERROR(Service_Audio, "Invalid delay channels {}", params.channel_count);
@@ -208,8 +209,10 @@ void DelayCommand::Dump([[maybe_unused]] const ADSP::CommandListProcessor& proce
 }
 
 void DelayCommand::Process(const ADSP::CommandListProcessor& processor) {
-    std::vector<std::span<const s32>> input_buffers(parameter.channel_count);
-    std::vector<std::span<s32>> output_buffers(parameter.channel_count);
+    static Common::ScratchBuffer<std::span<const s32>> input_buffers{};
+    static Common::ScratchBuffer<std::span<s32>> output_buffers{};
+    input_buffers.resize_destructive(parameter.channel_count);
+    output_buffers.resize_destructive(parameter.channel_count);
 
     for (s16 i = 0; i < parameter.channel_count; i++) {
         input_buffers[i] = processor.mix_buffers.subspan(inputs[i] * processor.sample_count,

src/audio_core/renderer/command/effect/i3dl2_reverb.cpp

@@ -6,6 +6,7 @@
 #include "audio_core/renderer/adsp/command_list_processor.h"
 #include "audio_core/renderer/command/effect/i3dl2_reverb.h"
 #include "common/polyfill_ranges.h"
+#include "common/scratch_buffer.h"
 
 namespace AudioCore::AudioRenderer {
 
@@ -408,8 +409,10 @@ void I3dl2ReverbCommand::Dump([[maybe_unused]] const ADSP::CommandListProcessor&
 }
 
 void I3dl2ReverbCommand::Process(const ADSP::CommandListProcessor& processor) {
-    std::vector<std::span<const s32>> input_buffers(parameter.channel_count);
-    std::vector<std::span<s32>> output_buffers(parameter.channel_count);
+    static Common::ScratchBuffer<std::span<const s32>> input_buffers{};
+    static Common::ScratchBuffer<std::span<s32>> output_buffers{};
+    input_buffers.resize_destructive(parameter.channel_count);
+    output_buffers.resize_destructive(parameter.channel_count);
 
     for (u32 i = 0; i < parameter.channel_count; i++) {
         input_buffers[i] = processor.mix_buffers.subspan(inputs[i] * processor.sample_count,

src/audio_core/renderer/command/effect/light_limiter.cpp

@@ -3,6 +3,7 @@
 
 #include "audio_core/renderer/adsp/command_list_processor.h"
 #include "audio_core/renderer/command/effect/light_limiter.h"
+#include "common/scratch_buffer.h"
 
 namespace AudioCore::AudioRenderer {
 /**
@@ -47,8 +48,8 @@ static void InitializeLightLimiterEffect(const LightLimiterInfo::ParameterVersio
  */
 static void ApplyLightLimiterEffect(const LightLimiterInfo::ParameterVersion2& params,
                                     LightLimiterInfo::State& state, const bool enabled,
-                                    std::vector<std::span<const s32>>& inputs,
-                                    std::vector<std::span<s32>>& outputs, const u32 sample_count,
+                                    std::span<std::span<const s32>> inputs,
+                                    std::span<std::span<s32>> outputs, const u32 sample_count,
                                     LightLimiterInfo::StatisticsInternal* statistics) {
     constexpr s64 min{std::numeric_limits<s32>::min()};
     constexpr s64 max{std::numeric_limits<s32>::max()};
@@ -147,8 +148,10 @@ void LightLimiterVersion1Command::Dump([[maybe_unused]] const ADSP::CommandListP
 }
 
 void LightLimiterVersion1Command::Process(const ADSP::CommandListProcessor& processor) {
-    std::vector<std::span<const s32>> input_buffers(parameter.channel_count);
-    std::vector<std::span<s32>> output_buffers(parameter.channel_count);
+    static Common::ScratchBuffer<std::span<const s32>> input_buffers{};
+    static Common::ScratchBuffer<std::span<s32>> output_buffers{};
+    input_buffers.resize_destructive(parameter.channel_count);
+    output_buffers.resize_destructive(parameter.channel_count);
 
     for (u32 i = 0; i < parameter.channel_count; i++) {
         input_buffers[i] = processor.mix_buffers.subspan(inputs[i] * processor.sample_count,
@@ -190,8 +193,10 @@ void LightLimiterVersion2Command::Dump([[maybe_unused]] const ADSP::CommandListP
 }
 
 void LightLimiterVersion2Command::Process(const ADSP::CommandListProcessor& processor) {
-    std::vector<std::span<const s32>> input_buffers(parameter.channel_count);
-    std::vector<std::span<s32>> output_buffers(parameter.channel_count);
+    static Common::ScratchBuffer<std::span<const s32>> input_buffers{};
+    static Common::ScratchBuffer<std::span<s32>> output_buffers{};
+    input_buffers.resize_destructive(parameter.channel_count);
+    output_buffers.resize_destructive(parameter.channel_count);
 
     for (u32 i = 0; i < parameter.channel_count; i++) {
         input_buffers[i] = processor.mix_buffers.subspan(inputs[i] * processor.sample_count,

src/audio_core/renderer/command/effect/reverb.cpp

@@ -7,6 +7,7 @@
 #include "audio_core/renderer/adsp/command_list_processor.h"
 #include "audio_core/renderer/command/effect/reverb.h"
 #include "common/polyfill_ranges.h"
+#include "common/scratch_buffer.h"
 
 namespace AudioCore::AudioRenderer {
 
@@ -250,8 +251,8 @@ static Common::FixedPoint<50, 14> Axfx2AllPassTick(ReverbInfo::ReverbDelayLine&
  */
 template <size_t NumChannels>
 static void ApplyReverbEffect(const ReverbInfo::ParameterVersion2& params, ReverbInfo::State& state,
-                              std::vector<std::span<const s32>>& inputs,
-                              std::vector<std::span<s32>>& outputs, const u32 sample_count) {
+                              std::span<std::span<const s32>> inputs,
+                              std::span<std::span<s32>> outputs, const u32 sample_count) {
     constexpr std::array<u8, ReverbInfo::MaxDelayTaps> OutTapIndexes1Ch{
         0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
     };
@@ -368,8 +369,8 @@ static void ApplyReverbEffect(const ReverbInfo::ParameterVersion2& params, Rever
  * @param sample_count - Number of samples to process.
  */
 static void ApplyReverbEffect(const ReverbInfo::ParameterVersion2& params, ReverbInfo::State& state,
-                              const bool enabled, std::vector<std::span<const s32>>& inputs,
-                              std::vector<std::span<s32>>& outputs, const u32 sample_count) {
+                              const bool enabled, std::span<std::span<const s32>> inputs,
+                              std::span<std::span<s32>> outputs, const u32 sample_count) {
     if (enabled) {
         switch (params.channel_count) {
         case 0:
@@ -411,8 +412,10 @@ void ReverbCommand::Dump([[maybe_unused]] const ADSP::CommandListProcessor& proc
 }
 
 void ReverbCommand::Process(const ADSP::CommandListProcessor& processor) {
-    std::vector<std::span<const s32>> input_buffers(parameter.channel_count);
-    std::vector<std::span<s32>> output_buffers(parameter.channel_count);
+    static Common::ScratchBuffer<std::span<const s32>> input_buffers{};
+    static Common::ScratchBuffer<std::span<s32>> output_buffers{};
+    input_buffers.resize_destructive(parameter.channel_count);
+    output_buffers.resize_destructive(parameter.channel_count);
 
     for (u32 i = 0; i < parameter.channel_count; i++) {
         input_buffers[i] = processor.mix_buffers.subspan(inputs[i] * processor.sample_count,

src/audio_core/renderer/command/resample/upsample.cpp

@@ -20,25 +20,25 @@ static void SrcProcessFrame(std::span<s32> output, std::span<const s32> input,
                             const u32 target_sample_count, const u32 source_sample_count,
                             UpsamplerState* state) {
     constexpr u32 WindowSize = 10;
-    constexpr std::array<Common::FixedPoint<24, 8>, WindowSize> SincWindow1{
-        51.93359375f, -18.80078125f, 9.73046875f, -5.33203125f, 2.84375f,
-        -1.41015625f, 0.62109375f,   -0.2265625f, 0.0625f,      -0.00390625f,
+    constexpr std::array<Common::FixedPoint<17, 15>, WindowSize> WindowedSinc1{
+        0.95376587f,   -0.12872314f, 0.060028076f,  -0.032470703f, 0.017669678f,
+        -0.009124756f, 0.004272461f, -0.001739502f, 0.000579834f,  -0.000091552734f,
     };
-    constexpr std::array<Common::FixedPoint<24, 8>, WindowSize> SincWindow2{
-        105.35546875f, -24.52734375f, 11.9609375f,  -6.515625f, 3.52734375f,
-        -1.796875f,    0.828125f,     -0.32421875f, 0.1015625f, -0.015625f,
+    constexpr std::array<Common::FixedPoint<17, 15>, WindowSize> WindowedSinc2{
+        0.8230896f,    -0.19161987f,  0.093444824f,  -0.05090332f,   0.027557373f,
+        -0.014038086f, 0.0064697266f, -0.002532959f, 0.00079345703f, -0.00012207031f,
     };
-    constexpr std::array<Common::FixedPoint<24, 8>, WindowSize> SincWindow3{
-        122.08203125f, -16.47656250f, 7.68359375f,  -4.15625000f, 2.26171875f,
-        -1.16796875f,  0.54687500f,   -0.22265625f, 0.07421875f,  -0.01171875f,
+    constexpr std::array<Common::FixedPoint<17, 15>, WindowSize> WindowedSinc3{
+        0.6298828f,    -0.19274902f, 0.09725952f,    -0.05319214f,  0.028625488f,
+        -0.014373779f, 0.006500244f, -0.0024719238f, 0.0007324219f, -0.000091552734f,
     };
-    constexpr std::array<Common::FixedPoint<24, 8>, WindowSize> SincWindow4{
-        23.73437500f, -9.62109375f, 5.07812500f,  -2.78125000f, 1.46875000f,
-        -0.71484375f, 0.30859375f,  -0.10546875f, 0.02734375f,  0.00000000f,
+    constexpr std::array<Common::FixedPoint<17, 15>, WindowSize> WindowedSinc4{
+        0.4057312f,    -0.1468811f,  0.07601929f,    -0.041656494f,  0.022216797f,
+        -0.011016846f, 0.004852295f, -0.0017700195f, 0.00048828125f, -0.000030517578f,
     };
-    constexpr std::array<Common::FixedPoint<24, 8>, WindowSize> SincWindow5{
-        80.62500000f, -24.67187500f, 12.44921875f, -6.80859375f, 3.66406250f,
-        -1.83984375f, 0.83203125f,   -0.31640625f, 0.09375000f,  -0.01171875f,
+    constexpr std::array<Common::FixedPoint<17, 15>, WindowSize> WindowedSinc5{
+        0.1854248f,    -0.075164795f, 0.03967285f,    -0.021728516f,  0.011474609f,
+        -0.005584717f, 0.0024108887f, -0.0008239746f, 0.00021362305f, 0.0f,
     };
 
     if (!state->initialized) {
@@ -91,52 +91,31 @@ static void SrcProcessFrame(std::span<s32> output, std::span<const s32> input,
             static_cast<u16>((state->history_output_index + 1) % UpsamplerState::HistorySize);
     };
 
-    auto calculate_sample = [&state](std::span<const Common::FixedPoint<24, 8>> coeffs1,
-                                     std::span<const Common::FixedPoint<24, 8>> coeffs2) -> s32 {
+    auto calculate_sample = [&state](std::span<const Common::FixedPoint<17, 15>> coeffs1,
+                                     std::span<const Common::FixedPoint<17, 15>> coeffs2) -> s32 {
         auto output_index{state->history_output_index};
-        auto start_pos{output_index - state->history_start_index + 1U};
-        auto end_pos{10U};
+        u64 result{0};
 
-        if (start_pos < 10) {
-            end_pos = start_pos;
-        }
+        for (u32 coeff_index = 0; coeff_index < 10; coeff_index++) {
+            result += static_cast<u64>(state->history[output_index].to_raw()) *
+                      coeffs1[coeff_index].to_raw();
 
-        u64 prev_contrib{0};
-        u32 coeff_index{0};
-        for (; coeff_index < end_pos; coeff_index++, output_index--) {
-            prev_contrib += static_cast<u64>(state->history[output_index].to_raw()) *
-                            coeffs1[coeff_index].to_raw();
-        }
-
-        auto end_index{state->history_end_index};
-        for (; start_pos < 9; start_pos++, coeff_index++, end_index--) {
-            prev_contrib += static_cast<u64>(state->history[end_index].to_raw()) *
-                            coeffs1[coeff_index].to_raw();
+            output_index = output_index == state->history_start_index ? state->history_end_index
+                                                                      : output_index - 1;
         }
 
         output_index =
             static_cast<u16>((state->history_output_index + 1) % UpsamplerState::HistorySize);
-        start_pos = state->history_end_index - output_index + 1U;
-        end_pos = 10U;
 
-        if (start_pos < 10) {
-            end_pos = start_pos;
+        for (u32 coeff_index = 0; coeff_index < 10; coeff_index++) {
+            result += static_cast<u64>(state->history[output_index].to_raw()) *
+                      coeffs2[coeff_index].to_raw();
+
+            output_index = output_index == state->history_end_index ? state->history_start_index
+                                                                    : output_index + 1;
         }
 
-        u64 next_contrib{0};
-        coeff_index = 0;
-        for (; coeff_index < end_pos; coeff_index++, output_index++) {
-            next_contrib += static_cast<u64>(state->history[output_index].to_raw()) *
-                            coeffs2[coeff_index].to_raw();
-        }
-
-        auto start_index{state->history_start_index};
-        for (; start_pos < 9; start_pos++, start_index++, coeff_index++) {
-            next_contrib += static_cast<u64>(state->history[start_index].to_raw()) *
-                            coeffs2[coeff_index].to_raw();
-        }
-
-        return static_cast<s32>(((prev_contrib >> 15) + (next_contrib >> 15)) >> 8);
+        return static_cast<s32>(result >> (8 + 15));
     };
 
     switch (state->ratio.to_int_floor()) {
@@ -150,23 +129,23 @@ static void SrcProcessFrame(std::span<s32> output, std::span<const s32> input,
                 break;
 
             case 1:
-                output[write_index] = calculate_sample(SincWindow3, SincWindow4);
+                output[write_index] = calculate_sample(WindowedSinc1, WindowedSinc5);
                 break;
 
             case 2:
-                output[write_index] = calculate_sample(SincWindow2, SincWindow1);
+                output[write_index] = calculate_sample(WindowedSinc2, WindowedSinc4);
                 break;
 
             case 3:
-                output[write_index] = calculate_sample(SincWindow5, SincWindow5);
+                output[write_index] = calculate_sample(WindowedSinc3, WindowedSinc3);
                 break;
 
             case 4:
-                output[write_index] = calculate_sample(SincWindow1, SincWindow2);
+                output[write_index] = calculate_sample(WindowedSinc4, WindowedSinc2);
                 break;
 
             case 5:
-                output[write_index] = calculate_sample(SincWindow4, SincWindow3);
+                output[write_index] = calculate_sample(WindowedSinc5, WindowedSinc1);
                 break;
             }
             state->sample_index = static_cast<u8>((state->sample_index + 1) % 6);
@@ -183,11 +162,11 @@ static void SrcProcessFrame(std::span<s32> output, std::span<const s32> input,
                 break;
 
             case 1:
-                output[write_index] = calculate_sample(SincWindow2, SincWindow1);
+                output[write_index] = calculate_sample(WindowedSinc2, WindowedSinc4);
                 break;
 
             case 2:
-                output[write_index] = calculate_sample(SincWindow1, SincWindow2);
+                output[write_index] = calculate_sample(WindowedSinc4, WindowedSinc2);
                 break;
             }
             state->sample_index = static_cast<u8>((state->sample_index + 1) % 3);
@@ -204,12 +183,12 @@ static void SrcProcessFrame(std::span<s32> output, std::span<const s32> input,
                 break;
 
             case 1:
-                output[write_index] = calculate_sample(SincWindow1, SincWindow2);
+                output[write_index] = calculate_sample(WindowedSinc4, WindowedSinc2);
                 break;
 
             case 2:
                 increment();
-                output[write_index] = calculate_sample(SincWindow2, SincWindow1);
+                output[write_index] = calculate_sample(WindowedSinc2, WindowedSinc4);
                 break;
             }
             state->sample_index = static_cast<u8>((state->sample_index + 1) % 3);

src/audio_core/renderer/command/sink/circular_buffer.cpp

@@ -5,6 +5,7 @@
 
 #include "audio_core/renderer/adsp/command_list_processor.h"
 #include "audio_core/renderer/command/sink/circular_buffer.h"
+#include "common/scratch_buffer.h"
 #include "core/memory.h"
 
 namespace AudioCore::AudioRenderer {
@@ -24,7 +25,9 @@ void CircularBufferSinkCommand::Process(const ADSP::CommandListProcessor& proces
     constexpr s32 min{std::numeric_limits<s16>::min()};
     constexpr s32 max{std::numeric_limits<s16>::max()};
 
-    std::vector<s16> output(processor.sample_count);
+    static Common::ScratchBuffer<s16> output{};
+    output.resize_destructive(processor.sample_count);
+
     for (u32 channel = 0; channel < input_count; channel++) {
         auto input{processor.mix_buffers.subspan(inputs[channel] * processor.sample_count,
                                                  processor.sample_count)};

src/audio_core/renderer/command/sink/device.cpp

@@ -33,7 +33,8 @@ void DeviceSinkCommand::Process(const ADSP::CommandListProcessor& processor) {
         .consumed{false},
     };
 
-    std::vector<s16> samples(out_buffer.frames * input_count);
+    static Common::ScratchBuffer<s16> samples{};
+    samples.resize_destructive(out_buffer.frames * input_count);
 
     for (u32 channel = 0; channel < input_count; channel++) {
         const auto offset{inputs[channel] * out_buffer.frames};

src/audio_core/sink/cubeb_sink.cpp

@@ -302,11 +302,21 @@ std::vector<std::string> ListCubebSinkDevices(bool capture) {
     std::vector<std::string> device_list;
     cubeb* ctx;
 
+#ifdef _WIN32
+    auto com_init_result = CoInitializeEx(nullptr, COINIT_MULTITHREADED);
+#endif
+
     if (cubeb_init(&ctx, "yuzu Device Enumerator", nullptr) != CUBEB_OK) {
         LOG_CRITICAL(Audio_Sink, "cubeb_init failed");
         return {};
     }
 
+#ifdef _WIN32
+    if (SUCCEEDED(com_init_result)) {
+        CoUninitialize();
+    }
+#endif
+
     auto type{capture ? CUBEB_DEVICE_TYPE_INPUT : CUBEB_DEVICE_TYPE_OUTPUT};
     cubeb_device_collection collection;
     if (cubeb_enumerate_devices(ctx, type, &collection) != CUBEB_OK) {
@@ -329,12 +339,22 @@ std::vector<std::string> ListCubebSinkDevices(bool capture) {
 u32 GetCubebLatency() {
     cubeb* ctx;
 
+#ifdef _WIN32
+    auto com_init_result = CoInitializeEx(nullptr, COINIT_MULTITHREADED);
+#endif
+
     if (cubeb_init(&ctx, "yuzu Latency Getter", nullptr) != CUBEB_OK) {
         LOG_CRITICAL(Audio_Sink, "cubeb_init failed");
         // Return a large latency so we choose SDL instead.
         return 10000u;
     }
 
+#ifdef _WIN32
+    if (SUCCEEDED(com_init_result)) {
+        CoUninitialize();
+    }
+#endif
+
     cubeb_stream_params params{};
     params.rate = TargetSampleRate;
     params.channels = 2;

src/audio_core/sink/null_sink.h

@@ -9,6 +9,7 @@
 
 #include "audio_core/sink/sink.h"
 #include "audio_core/sink/sink_stream.h"
+#include "common/scratch_buffer.h"
 
 namespace Core {
 class System;
@@ -20,7 +21,7 @@ public:
     explicit NullSinkStreamImpl(Core::System& system_, StreamType type_)
         : SinkStream{system_, type_} {}
     ~NullSinkStreamImpl() override {}
-    void AppendBuffer(SinkBuffer&, std::vector<s16>&) override {}
+    void AppendBuffer(SinkBuffer&, Common::ScratchBuffer<s16>&) override {}
     std::vector<s16> ReleaseBuffer(u64) override {
         return {};
     }

src/audio_core/sink/sink_stream.cpp

@@ -17,7 +17,7 @@
 
 namespace AudioCore::Sink {
 
-void SinkStream::AppendBuffer(SinkBuffer& buffer, std::vector<s16>& samples) {
+void SinkStream::AppendBuffer(SinkBuffer& buffer, Common::ScratchBuffer<s16>& samples) {
     if (type == StreamType::In) {
         queue.enqueue(buffer);
         queued_buffers++;
@@ -71,7 +71,9 @@ void SinkStream::AppendBuffer(SinkBuffer& buffer, std::vector<s16>& samples) {
         // We need moar samples! Not all games will provide 6 channel audio.
         // TODO: Implement some upmixing here. Currently just passthrough, with other
         // channels left as silence.
-        std::vector<s16> new_samples(samples.size() / system_channels * device_channels, 0);
+        static Common::ScratchBuffer<s16> new_samples{};
+        new_samples.resize_destructive(samples.size() / system_channels * device_channels);
+        std::memset(new_samples.data(), 0, new_samples.size() * sizeof(s16));
 
         for (u32 read_index = 0, write_index = 0; read_index < samples.size();
              read_index += system_channels, write_index += device_channels) {
@@ -100,7 +102,7 @@ void SinkStream::AppendBuffer(SinkBuffer& buffer, std::vector<s16>& samples) {
         }
     }
 
-    samples_buffer.Push(samples);
+    samples_buffer.Push(samples.data(), samples.size());
     queue.enqueue(buffer);
     queued_buffers++;
 }

src/audio_core/sink/sink_stream.h

@@ -14,6 +14,7 @@
 #include "common/common_types.h"
 #include "common/reader_writer_queue.h"
 #include "common/ring_buffer.h"
+#include "common/scratch_buffer.h"
 
 namespace Core {
 class System;
@@ -169,7 +170,7 @@ public:
      * @param buffer  - Audio buffer information to be queued.
      * @param samples - The s16 samples to be queue for playback.
      */
-    virtual void AppendBuffer(SinkBuffer& buffer, std::vector<s16>& samples);
+    virtual void AppendBuffer(SinkBuffer& buffer, Common::ScratchBuffer<s16>& samples);
 
     /**
      * Release a buffer. Audio In only, will fill a buffer with recorded samples.

src/common/CMakeLists.txt

@@ -38,6 +38,8 @@ add_library(common STATIC
     common_precompiled_headers.h
     common_types.h
     concepts.h
+    demangle.cpp
+    demangle.h
     div_ceil.h
     dynamic_library.cpp
     dynamic_library.h
@@ -97,6 +99,7 @@ add_library(common STATIC
     point.h
     precompiled_headers.h
     quaternion.h
+    range_map.h
     reader_writer_queue.h
     ring_buffer.h
     ${CMAKE_CURRENT_BINARY_DIR}/scm_rev.cpp
@@ -174,7 +177,7 @@ endif()
 create_target_directory_groups(common)
 
 target_link_libraries(common PUBLIC ${Boost_LIBRARIES} fmt::fmt microprofile Threads::Threads)
-target_link_libraries(common PRIVATE lz4::lz4 zstd::zstd)
+target_link_libraries(common PRIVATE lz4::lz4 zstd::zstd LLVM::Demangle)
 
 if (YUZU_USE_PRECOMPILED_HEADERS)
     target_precompile_headers(common PRIVATE precompiled_headers.h)

src/common/address_space.h

@@ -12,7 +12,8 @@
 
 namespace Common {
 template <typename VaType, size_t AddressSpaceBits>
-concept AddressSpaceValid = std::is_unsigned_v<VaType> && sizeof(VaType) * 8 >= AddressSpaceBits;
+concept AddressSpaceValid = std::is_unsigned_v<VaType> && sizeof(VaType) * 8 >=
+AddressSpaceBits;
 
 struct EmptyStruct {};
 
@@ -21,7 +22,7 @@ struct EmptyStruct {};
  */
 template <typename VaType, VaType UnmappedVa, typename PaType, PaType UnmappedPa,
           bool PaContigSplit, size_t AddressSpaceBits, typename ExtraBlockInfo = EmptyStruct>
-requires AddressSpaceValid<VaType, AddressSpaceBits>
+    requires AddressSpaceValid<VaType, AddressSpaceBits>
 class FlatAddressSpaceMap {
 public:
     /// The maximum VA that this AS can technically reach
@@ -109,7 +110,7 @@ private:
  * initial, fast linear pass and a subsequent slower pass that iterates until it finds a free block
  */
 template <typename VaType, VaType UnmappedVa, size_t AddressSpaceBits>
-requires AddressSpaceValid<VaType, AddressSpaceBits>
+    requires AddressSpaceValid<VaType, AddressSpaceBits>
 class FlatAllocator
     : public FlatAddressSpaceMap<VaType, UnmappedVa, bool, false, false, AddressSpaceBits> {
 private:

src/common/alignment.h

@@ -10,7 +10,7 @@
 namespace Common {
 
 template <typename T>
-requires std::is_unsigned_v<T>
+    requires std::is_unsigned_v<T>
 [[nodiscard]] constexpr T AlignUp(T value, size_t size) {
     auto mod{static_cast<T>(value % size)};
     value -= mod;
@@ -18,31 +18,31 @@ requires std::is_unsigned_v<T>
 }
 
 template <typename T>
-requires std::is_unsigned_v<T>
+    requires std::is_unsigned_v<T>
 [[nodiscard]] constexpr T AlignUpLog2(T value, size_t align_log2) {
     return static_cast<T>((value + ((1ULL << align_log2) - 1)) >> align_log2 << align_log2);
 }
 
 template <typename T>
-requires std::is_unsigned_v<T>
+    requires std::is_unsigned_v<T>
 [[nodiscard]] constexpr T AlignDown(T value, size_t size) {
     return static_cast<T>(value - value % size);
 }
 
 template <typename T>
-requires std::is_unsigned_v<T>
+    requires std::is_unsigned_v<T>
 [[nodiscard]] constexpr bool Is4KBAligned(T value) {
     return (value & 0xFFF) == 0;
 }
 
 template <typename T>
-requires std::is_unsigned_v<T>
+    requires std::is_unsigned_v<T>
 [[nodiscard]] constexpr bool IsWordAligned(T value) {
     return (value & 0b11) == 0;
 }
 
 template <typename T>
-requires std::is_integral_v<T>
+    requires std::is_integral_v<T>
 [[nodiscard]] constexpr bool IsAligned(T value, size_t alignment) {
     using U = typename std::make_unsigned_t<T>;
     const U mask = static_cast<U>(alignment - 1);
@@ -50,7 +50,7 @@ requires std::is_integral_v<T>
 }
 
 template <typename T, typename U>
-requires std::is_integral_v<T>
+    requires std::is_integral_v<T>
 [[nodiscard]] constexpr T DivideUp(T x, U y) {
     return (x + (y - 1)) / y;
 }
@@ -73,11 +73,11 @@ public:
     constexpr AlignmentAllocator(const AlignmentAllocator<T2, Align>&) noexcept {}
 
     [[nodiscard]] T* allocate(size_type n) {
-        return static_cast<T*>(::operator new (n * sizeof(T), std::align_val_t{Align}));
+        return static_cast<T*>(::operator new(n * sizeof(T), std::align_val_t{Align}));
     }
 
     void deallocate(T* p, size_type n) {
-        ::operator delete (p, n * sizeof(T), std::align_val_t{Align});
+        ::operator delete(p, n * sizeof(T), std::align_val_t{Align});
     }
 
     template <typename T2>

src/common/atomic_helpers.h

@@ -75,7 +75,7 @@ extern "C" void AnnotateHappensAfter(const char*, int, void*);
 #if defined(AE_VCPP) || defined(AE_ICC)
 #define AE_FORCEINLINE __forceinline
 #elif defined(AE_GCC)
-//#define AE_FORCEINLINE __attribute__((always_inline))
+// #define AE_FORCEINLINE __attribute__((always_inline))
 #define AE_FORCEINLINE inline
 #else
 #define AE_FORCEINLINE inline

src/common/bit_util.h

@@ -45,19 +45,19 @@ template <typename T>
 }
 
 template <typename T>
-requires std::is_unsigned_v<T>
+    requires std::is_unsigned_v<T>
 [[nodiscard]] constexpr bool IsPow2(T value) {
     return std::has_single_bit(value);
 }
 
 template <typename T>
-requires std::is_integral_v<T>
+    requires std::is_integral_v<T>
 [[nodiscard]] T NextPow2(T value) {
     return static_cast<T>(1ULL << ((8U * sizeof(T)) - std::countl_zero(value - 1U)));
 }
 
 template <size_t bit_index, typename T>
-requires std::is_integral_v<T>
+    requires std::is_integral_v<T>
 [[nodiscard]] constexpr bool Bit(const T value) {
     static_assert(bit_index < BitSize<T>(), "bit_index must be smaller than size of T");
     return ((value >> bit_index) & T(1)) == T(1);

src/common/concepts.h

@@ -16,9 +16,9 @@ concept IsContiguousContainer = std::contiguous_iterator<typename T::iterator>;
 //       is available on all supported platforms.
 template <typename Derived, typename Base>
 concept DerivedFrom = requires {
-    std::is_base_of_v<Base, Derived>;
-    std::is_convertible_v<const volatile Derived*, const volatile Base*>;
-};
+                          std::is_base_of_v<Base, Derived>;
+                          std::is_convertible_v<const volatile Derived*, const volatile Base*>;
+                      };
 
 // TODO: Replace with std::convertible_to when libc++ implements it.
 template <typename From, typename To>

src/common/demangle.cpp

@@ -0,0 +1,35 @@
+// SPDX-FileCopyrightText: Copyright 2020 yuzu Emulator Project
+// SPDX-License-Identifier: GPL-2.0-or-later
+
+#include <llvm/Demangle/Demangle.h>
+
+#include "common/demangle.h"
+#include "common/scope_exit.h"
+
+namespace Common {
+
+std::string DemangleSymbol(const std::string& mangled) {
+    auto is_itanium = [](const std::string& name) -> bool {
+        // A valid Itanium encoding requires 1-4 leading underscores, followed by 'Z'.
+        auto pos = name.find_first_not_of('_');
+        return pos > 0 && pos <= 4 && pos < name.size() && name[pos] == 'Z';
+    };
+
+    if (mangled.empty()) {
+        return mangled;
+    }
+
+    char* demangled = nullptr;
+    SCOPE_EXIT({ std::free(demangled); });
+
+    if (is_itanium(mangled)) {
+        demangled = llvm::itaniumDemangle(mangled.c_str(), nullptr, nullptr, nullptr);
+    }
+
+    if (!demangled) {
+        return mangled;
+    }
+    return demangled;
+}
+
+} // namespace Common

src/common/demangle.h

@@ -0,0 +1,12 @@
+// SPDX-FileCopyrightText: Copyright 2020 yuzu Emulator Project
+// SPDX-License-Identifier: GPL-2.0-or-later
+
+#pragma once
+
+#include <string>
+
+namespace Common {
+
+std::string DemangleSymbol(const std::string& mangled);
+
+} // namespace Common

src/common/div_ceil.h

@@ -10,14 +10,14 @@ namespace Common {
 
 /// Ceiled integer division.
 template <typename N, typename D>
-requires std::is_integral_v<N> && std::is_unsigned_v<D>
+    requires std::is_integral_v<N> && std::is_unsigned_v<D>
 [[nodiscard]] constexpr N DivCeil(N number, D divisor) {
     return static_cast<N>((static_cast<D>(number) + divisor - 1) / divisor);
 }
 
 /// Ceiled integer division with logarithmic divisor in base 2
 template <typename N, typename D>
-requires std::is_integral_v<N> && std::is_unsigned_v<D>
+    requires std::is_integral_v<N> && std::is_unsigned_v<D>
 [[nodiscard]] constexpr N DivCeilLog2(N value, D alignment_log2) {
     return static_cast<N>((static_cast<D>(value) + (D(1) << alignment_log2) - 1) >> alignment_log2);
 }

src/common/expected.h

@@ -64,7 +64,7 @@ struct no_init_t {
  * Additionally, this requires E to be trivially destructible
  */
 template <typename T, typename E, bool = std::is_trivially_destructible_v<T>>
-requires std::is_trivially_destructible_v<E>
+    requires std::is_trivially_destructible_v<E>
 struct expected_storage_base {
     constexpr expected_storage_base() : m_val{T{}}, m_has_val{true} {}
 
@@ -111,7 +111,7 @@ struct expected_storage_base {
  * Additionally, this requires E to be trivially destructible
  */
 template <typename T, typename E>
-requires std::is_trivially_destructible_v<E>
+    requires std::is_trivially_destructible_v<E>
 struct expected_storage_base<T, E, true> {
     constexpr expected_storage_base() : m_val{T{}}, m_has_val{true} {}
 
@@ -251,7 +251,7 @@ struct expected_operations_base : expected_storage_base<T, E> {
  * Additionally, this requires E to be trivially copy constructible
  */
 template <typename T, typename E, bool = std::is_trivially_copy_constructible_v<T>>
-requires std::is_trivially_copy_constructible_v<E>
+    requires std::is_trivially_copy_constructible_v<E>
 struct expected_copy_base : expected_operations_base<T, E> {
     using expected_operations_base<T, E>::expected_operations_base;
 };
@@ -261,7 +261,7 @@ struct expected_copy_base : expected_operations_base<T, E> {
  * Additionally, this requires E to be trivially copy constructible
  */
 template <typename T, typename E>
-requires std::is_trivially_copy_constructible_v<E>
+    requires std::is_trivially_copy_constructible_v<E>
 struct expected_copy_base<T, E, false> : expected_operations_base<T, E> {
     using expected_operations_base<T, E>::expected_operations_base;
 
@@ -289,7 +289,7 @@ struct expected_copy_base<T, E, false> : expected_operations_base<T, E> {
  * Additionally, this requires E to be trivially move constructible
  */
 template <typename T, typename E, bool = std::is_trivially_move_constructible_v<T>>
-requires std::is_trivially_move_constructible_v<E>
+    requires std::is_trivially_move_constructible_v<E>
 struct expected_move_base : expected_copy_base<T, E> {
     using expected_copy_base<T, E>::expected_copy_base;
 };
@@ -299,7 +299,7 @@ struct expected_move_base : expected_copy_base<T, E> {
  * Additionally, this requires E to be trivially move constructible
  */
 template <typename T, typename E>
-requires std::is_trivially_move_constructible_v<E>
+    requires std::is_trivially_move_constructible_v<E>
 struct expected_move_base<T, E, false> : expected_copy_base<T, E> {
     using expected_copy_base<T, E>::expected_copy_base;
 
@@ -330,9 +330,9 @@ template <typename T, typename E,
           bool = std::conjunction_v<std::is_trivially_copy_assignable<T>,
                                     std::is_trivially_copy_constructible<T>,
                                     std::is_trivially_destructible<T>>>
-requires std::conjunction_v<std::is_trivially_copy_assignable<E>,
-                            std::is_trivially_copy_constructible<E>,
-                            std::is_trivially_destructible<E>>
+    requires std::conjunction_v<std::is_trivially_copy_assignable<E>,
+                                std::is_trivially_copy_constructible<E>,
+                                std::is_trivially_destructible<E>>
 struct expected_copy_assign_base : expected_move_base<T, E> {
     using expected_move_base<T, E>::expected_move_base;
 };
@@ -342,9 +342,9 @@ struct expected_copy_assign_base : expected_move_base<T, E> {
  * Additionally, this requires E to be trivially copy assignable
  */
 template <typename T, typename E>
-requires std::conjunction_v<std::is_trivially_copy_assignable<E>,
-                            std::is_trivially_copy_constructible<E>,
-                            std::is_trivially_destructible<E>>
+    requires std::conjunction_v<std::is_trivially_copy_assignable<E>,
+                                std::is_trivially_copy_constructible<E>,
+                                std::is_trivially_destructible<E>>
 struct expected_copy_assign_base<T, E, false> : expected_move_base<T, E> {
     using expected_move_base<T, E>::expected_move_base;
 
@@ -371,9 +371,9 @@ template <typename T, typename E,
           bool = std::conjunction_v<std::is_trivially_move_assignable<T>,
                                     std::is_trivially_move_constructible<T>,
                                     std::is_trivially_destructible<T>>>
-requires std::conjunction_v<std::is_trivially_move_assignable<E>,
-                            std::is_trivially_move_constructible<E>,
-                            std::is_trivially_destructible<E>>
+    requires std::conjunction_v<std::is_trivially_move_assignable<E>,
+                                std::is_trivially_move_constructible<E>,
+                                std::is_trivially_destructible<E>>
 struct expected_move_assign_base : expected_copy_assign_base<T, E> {
     using expected_copy_assign_base<T, E>::expected_copy_assign_base;
 };
@@ -383,9 +383,9 @@ struct expected_move_assign_base : expected_copy_assign_base<T, E> {
  * Additionally, this requires E to be trivially move assignable
  */
 template <typename T, typename E>
-requires std::conjunction_v<std::is_trivially_move_assignable<E>,
-                            std::is_trivially_move_constructible<E>,
-                            std::is_trivially_destructible<E>>
+    requires std::conjunction_v<std::is_trivially_move_assignable<E>,
+                                std::is_trivially_move_constructible<E>,
+                                std::is_trivially_destructible<E>>
 struct expected_move_assign_base<T, E, false> : expected_copy_assign_base<T, E> {
     using expected_copy_assign_base<T, E>::expected_copy_assign_base;
 
@@ -412,7 +412,7 @@ struct expected_move_assign_base<T, E, false> : expected_copy_assign_base<T, E>
  */
 template <typename T, typename E, bool EnableCopy = std::is_copy_constructible_v<T>,
           bool EnableMove = std::is_move_constructible_v<T>>
-requires std::conjunction_v<std::is_copy_constructible<E>, std::is_move_constructible<E>>
+    requires std::conjunction_v<std::is_copy_constructible<E>, std::is_move_constructible<E>>
 struct expected_delete_ctor_base {
     expected_delete_ctor_base() = default;
     expected_delete_ctor_base(const expected_delete_ctor_base&) = default;
@@ -422,7 +422,7 @@ struct expected_delete_ctor_base {
 };
 
 template <typename T, typename E>
-requires std::conjunction_v<std::is_copy_constructible<E>, std::is_move_constructible<E>>
+    requires std::conjunction_v<std::is_copy_constructible<E>, std::is_move_constructible<E>>
 struct expected_delete_ctor_base<T, E, true, false> {
     expected_delete_ctor_base() = default;
     expected_delete_ctor_base(const expected_delete_ctor_base&) = default;
@@ -432,7 +432,7 @@ struct expected_delete_ctor_base<T, E, true, false> {
 };
 
 template <typename T, typename E>
-requires std::conjunction_v<std::is_copy_constructible<E>, std::is_move_constructible<E>>
+    requires std::conjunction_v<std::is_copy_constructible<E>, std::is_move_constructible<E>>
 struct expected_delete_ctor_base<T, E, false, true> {
     expected_delete_ctor_base() = default;
     expected_delete_ctor_base(const expected_delete_ctor_base&) = delete;
@@ -442,7 +442,7 @@ struct expected_delete_ctor_base<T, E, false, true> {
 };
 
 template <typename T, typename E>
-requires std::conjunction_v<std::is_copy_constructible<E>, std::is_move_constructible<E>>
+    requires std::conjunction_v<std::is_copy_constructible<E>, std::is_move_constructible<E>>
 struct expected_delete_ctor_base<T, E, false, false> {
     expected_delete_ctor_base() = default;
     expected_delete_ctor_base(const expected_delete_ctor_base&) = delete;
@@ -460,8 +460,8 @@ template <
     typename T, typename E,
     bool EnableCopy = std::conjunction_v<std::is_copy_constructible<T>, std::is_copy_assignable<T>>,
     bool EnableMove = std::conjunction_v<std::is_move_constructible<T>, std::is_move_assignable<T>>>
-requires std::conjunction_v<std::is_copy_constructible<E>, std::is_move_constructible<E>,
-                            std::is_copy_assignable<E>, std::is_move_assignable<E>>
+    requires std::conjunction_v<std::is_copy_constructible<E>, std::is_move_constructible<E>,
+                                std::is_copy_assignable<E>, std::is_move_assignable<E>>
 struct expected_delete_assign_base {
     expected_delete_assign_base() = default;
     expected_delete_assign_base(const expected_delete_assign_base&) = default;
@@ -471,8 +471,8 @@ struct expected_delete_assign_base {
 };
 
 template <typename T, typename E>
-requires std::conjunction_v<std::is_copy_constructible<E>, std::is_move_constructible<E>,
-                            std::is_copy_assignable<E>, std::is_move_assignable<E>>
+    requires std::conjunction_v<std::is_copy_constructible<E>, std::is_move_constructible<E>,
+                                std::is_copy_assignable<E>, std::is_move_assignable<E>>
 struct expected_delete_assign_base<T, E, true, false> {
     expected_delete_assign_base() = default;
     expected_delete_assign_base(const expected_delete_assign_base&) = default;
@@ -482,8 +482,8 @@ struct expected_delete_assign_base<T, E, true, false> {
 };
 
 template <typename T, typename E>
-requires std::conjunction_v<std::is_copy_constructible<E>, std::is_move_constructible<E>,
-                            std::is_copy_assignable<E>, std::is_move_assignable<E>>
+    requires std::conjunction_v<std::is_copy_constructible<E>, std::is_move_constructible<E>,
+                                std::is_copy_assignable<E>, std::is_move_assignable<E>>
 struct expected_delete_assign_base<T, E, false, true> {
     expected_delete_assign_base() = default;
     expected_delete_assign_base(const expected_delete_assign_base&) = default;
@@ -493,8 +493,8 @@ struct expected_delete_assign_base<T, E, false, true> {
 };
 
 template <typename T, typename E>
-requires std::conjunction_v<std::is_copy_constructible<E>, std::is_move_constructible<E>,
-                            std::is_copy_assignable<E>, std::is_move_assignable<E>>
+    requires std::conjunction_v<std::is_copy_constructible<E>, std::is_move_constructible<E>,
+                                std::is_copy_assignable<E>, std::is_move_assignable<E>>
 struct expected_delete_assign_base<T, E, false, false> {
     expected_delete_assign_base() = default;
     expected_delete_assign_base(const expected_delete_assign_base&) = default;

src/common/input.h

@@ -51,6 +51,8 @@ enum class PollingMode {
     NFC,
     // Enable infrared camera polling
     IR,
+    // Enable ring controller polling
+    Ring,
 };
 
 enum class CameraFormat {
@@ -62,21 +64,22 @@ enum class CameraFormat {
     None,
 };
 
-// Vibration reply from the controller
-enum class VibrationError {
-    None,
+// Different results that can happen from a device request
+enum class DriverResult {
+    Success,
+    WrongReply,
+    Timeout,
+    UnsupportedControllerType,
+    HandleInUse,
+    ErrorReadingData,
+    ErrorWritingData,
+    NoDeviceDetected,
+    InvalidHandle,
     NotSupported,
     Disabled,
     Unknown,
 };
 
-// Polling mode reply from the controller
-enum class PollingError {
-    None,
-    NotSupported,
-    Unknown,
-};
-
 // Nfc reply from the controller
 enum class NfcState {
     Success,
@@ -90,13 +93,6 @@ enum class NfcState {
     Unknown,
 };
 
-// Ir camera reply from the controller
-enum class CameraError {
-    None,
-    NotSupported,
-    Unknown,
-};
-
 // Hint for amplification curve to be used
 enum class VibrationAmplificationType {
     Linear,
@@ -134,6 +130,8 @@ struct ButtonStatus {
     bool inverted{};
     // Press once to activate, press again to release
     bool toggle{};
+    // Spams the button when active
+    bool turbo{};
     // Internal lock for the toggle status
     bool locked{};
 };
@@ -190,6 +188,8 @@ struct TouchStatus {
 struct BodyColorStatus {
     u32 body{};
     u32 buttons{};
+    u32 left_grip{};
+    u32 right_grip{};
 };
 
 // HD rumble data
@@ -228,17 +228,31 @@ enum class ButtonNames {
     Engine,
     // This will display the button by value instead of the button name
     Value,
+
+    // Joycon button names
     ButtonLeft,
     ButtonRight,
     ButtonDown,
     ButtonUp,
-    TriggerZ,
-    TriggerR,
-    TriggerL,
     ButtonA,
     ButtonB,
     ButtonX,
     ButtonY,
+    ButtonPlus,
+    ButtonMinus,
+    ButtonHome,
+    ButtonCapture,
+    ButtonStickL,
+    ButtonStickR,
+    TriggerL,
+    TriggerZL,
+    TriggerSL,
+    TriggerR,
+    TriggerZR,
+    TriggerSR,
+
+    // GC button names
+    TriggerZ,
     ButtonStart,
 
     // DS4 button names
@@ -292,9 +306,6 @@ class InputDevice {
 public:
     virtual ~InputDevice() = default;
 
-    // Request input device to update if necessary
-    virtual void SoftUpdate() {}
-
     // Force input device to update data regardless of the current state
     virtual void ForceUpdate() {}
 
@@ -319,22 +330,24 @@ class OutputDevice {
 public:
     virtual ~OutputDevice() = default;
 
-    virtual void SetLED([[maybe_unused]] const LedStatus& led_status) {}
+    virtual DriverResult SetLED([[maybe_unused]] const LedStatus& led_status) {
+        return DriverResult::NotSupported;
+    }
 
-    virtual VibrationError SetVibration([[maybe_unused]] const VibrationStatus& vibration_status) {
-        return VibrationError::NotSupported;
+    virtual DriverResult SetVibration([[maybe_unused]] const VibrationStatus& vibration_status) {
+        return DriverResult::NotSupported;
     }
 
     virtual bool IsVibrationEnabled() {
         return false;
     }
 
-    virtual PollingError SetPollingMode([[maybe_unused]] PollingMode polling_mode) {
-        return PollingError::NotSupported;
+    virtual DriverResult SetPollingMode([[maybe_unused]] PollingMode polling_mode) {
+        return DriverResult::NotSupported;
     }
 
-    virtual CameraError SetCameraFormat([[maybe_unused]] CameraFormat camera_format) {
-        return CameraError::NotSupported;
+    virtual DriverResult SetCameraFormat([[maybe_unused]] CameraFormat camera_format) {
+        return DriverResult::NotSupported;
     }
 
     virtual NfcState SupportsNfc() const {

src/common/intrusive_red_black_tree.h

@@ -242,19 +242,21 @@ public:
 
 template <typename T>
 concept HasRedBlackKeyType = requires {
-    { std::is_same<typename T::RedBlackKeyType, void>::value } -> std::convertible_to<bool>;
-};
+                                 {
+                                     std::is_same<typename T::RedBlackKeyType, void>::value
+                                     } -> std::convertible_to<bool>;
+                             };
 
 namespace impl {
 
-    template <typename T, typename Default>
-    consteval auto* GetRedBlackKeyType() {
-        if constexpr (HasRedBlackKeyType<T>) {
-            return static_cast<typename T::RedBlackKeyType*>(nullptr);
-        } else {
-            return static_cast<Default*>(nullptr);
-        }
+template <typename T, typename Default>
+consteval auto* GetRedBlackKeyType() {
+    if constexpr (HasRedBlackKeyType<T>) {
+        return static_cast<typename T::RedBlackKeyType*>(nullptr);
+    } else {
+        return static_cast<Default*>(nullptr);
     }
+}
 
 } // namespace impl
 

src/common/make_unique_for_overwrite.h

@@ -9,17 +9,19 @@
 namespace Common {
 
 template <class T>
-requires(!std::is_array_v<T>) std::unique_ptr<T> make_unique_for_overwrite() {
+    requires(!std::is_array_v<T>)
+std::unique_ptr<T> make_unique_for_overwrite() {
     return std::unique_ptr<T>(new T);
 }
 
 template <class T>
-requires std::is_unbounded_array_v<T> std::unique_ptr<T> make_unique_for_overwrite(std::size_t n) {
+    requires std::is_unbounded_array_v<T>
+std::unique_ptr<T> make_unique_for_overwrite(std::size_t n) {
     return std::unique_ptr<T>(new std::remove_extent_t<T>[n]);
 }
 
 template <class T, class... Args>
-requires std::is_bounded_array_v<T>
+    requires std::is_bounded_array_v<T>
 void make_unique_for_overwrite(Args&&...) = delete;
 
 } // namespace Common

src/common/polyfill_ranges.h

@@ -18,9 +18,9 @@ namespace ranges {
 
 template <typename T>
 concept range = requires(T& t) {
-    begin(t);
-    end(t);
-};
+                    begin(t);
+                    end(t);
+                };
 
 template <typename T>
 concept input_range = range<T>;
@@ -421,7 +421,7 @@ struct generate_fn {
     }
 
     template <typename R, std::copy_constructible F>
-    requires std::invocable<F&> && ranges::output_range<R>
+        requires std::invocable<F&> && ranges::output_range<R>
     constexpr ranges::iterator_t<R> operator()(R&& r, F gen) const {
         return operator()(ranges::begin(r), ranges::end(r), std::move(gen));
     }

src/common/polyfill_thread.h

@@ -11,6 +11,8 @@
 
 #ifdef __cpp_lib_jthread
 
+#include <chrono>
+#include <condition_variable>
 #include <stop_token>
 #include <thread>
 
@@ -21,23 +23,36 @@ void CondvarWait(Condvar& cv, Lock& lock, std::stop_token token, Pred&& pred) {
     cv.wait(lock, token, std::move(pred));
 }
 
+template <typename Rep, typename Period>
+bool StoppableTimedWait(std::stop_token token, const std::chrono::duration<Rep, Period>& rel_time) {
+    std::condition_variable_any cv;
+    std::mutex m;
+
+    // Perform the timed wait.
+    std::unique_lock lk{m};
+    return !cv.wait_for(lk, token, rel_time, [&] { return token.stop_requested(); });
+}
+
 } // namespace Common
 
 #else
 
 #include <atomic>
+#include <chrono>
+#include <condition_variable>
 #include <functional>
-#include <list>
+#include <map>
 #include <memory>
 #include <mutex>
 #include <optional>
 #include <thread>
 #include <type_traits>
+#include <utility>
 
 namespace std {
 namespace polyfill {
 
-using stop_state_callbacks = list<function<void()>>;
+using stop_state_callback = size_t;
 
 class stop_state {
 public:
@@ -45,61 +60,69 @@ public:
     ~stop_state() = default;
 
     bool request_stop() {
-        stop_state_callbacks callbacks;
+        unique_lock lk{m_lock};
 
-        {
-            scoped_lock lk{m_lock};
-
-            if (m_stop_requested.load()) {
-                // Already set, nothing to do
-                return false;
-            }
-
-            // Set as requested
-            m_stop_requested = true;
-
-            // Copy callback list
-            callbacks = m_callbacks;
+        if (m_stop_requested) {
+            // Already set, nothing to do.
+            return false;
         }
 
-        for (auto callback : callbacks) {
-            callback();
+        // Mark stop requested.
+        m_stop_requested = true;
+
+        while (!m_callbacks.empty()) {
+            // Get an iterator to the first element.
+            const auto it = m_callbacks.begin();
+
+            // Move the callback function out of the map.
+            function<void()> f;
+            swap(it->second, f);
+
+            // Erase the now-empty map element.
+            m_callbacks.erase(it);
+
+            // Run the callback.
+            if (f) {
+                f();
+            }
         }
 
         return true;
     }
 
     bool stop_requested() const {
-        return m_stop_requested.load();
+        unique_lock lk{m_lock};
+        return m_stop_requested;
     }
 
-    stop_state_callbacks::const_iterator insert_callback(function<void()> f) {
-        stop_state_callbacks::const_iterator ret{};
-        bool should_run{};
+    stop_state_callback insert_callback(function<void()> f) {
+        unique_lock lk{m_lock};
 
-        {
-            scoped_lock lk{m_lock};
-            should_run = m_stop_requested.load();
-            m_callbacks.push_front(f);
-            ret = m_callbacks.begin();
-        }
-
-        if (should_run) {
-            f();
+        if (m_stop_requested) {
+            // Stop already requested. Don't insert anything,
+            // just run the callback synchronously.
+            if (f) {
+                f();
+            }
+            return 0;
         }
 
+        // Insert the callback.
+        stop_state_callback ret = ++m_next_callback;
+        m_callbacks.emplace(ret, move(f));
         return ret;
     }
 
-    void remove_callback(stop_state_callbacks::const_iterator it) {
-        scoped_lock lk{m_lock};
-        m_callbacks.erase(it);
+    void remove_callback(stop_state_callback cb) {
+        unique_lock lk{m_lock};
+        m_callbacks.erase(cb);
     }
 
 private:
-    mutex m_lock;
-    atomic<bool> m_stop_requested;
-    stop_state_callbacks m_callbacks;
+    mutable recursive_mutex m_lock;
+    map<stop_state_callback, function<void()>> m_callbacks;
+    stop_state_callback m_next_callback{0};
+    bool m_stop_requested{false};
 };
 
 } // namespace polyfill
@@ -190,7 +213,7 @@ public:
     using callback_type = Callback;
 
     template <typename C>
-    requires constructible_from<Callback, C>
+        requires constructible_from<Callback, C>
     explicit stop_callback(const stop_token& st,
                            C&& cb) noexcept(is_nothrow_constructible_v<Callback, C>)
         : m_stop_state(st.m_stop_state) {
@@ -199,7 +222,7 @@ public:
         }
     }
     template <typename C>
-    requires constructible_from<Callback, C>
+        requires constructible_from<Callback, C>
     explicit stop_callback(stop_token&& st,
                            C&& cb) noexcept(is_nothrow_constructible_v<Callback, C>)
         : m_stop_state(move(st.m_stop_state)) {
@@ -209,7 +232,7 @@ public:
     }
     ~stop_callback() {
         if (m_stop_state && m_callback) {
-            m_stop_state->remove_callback(*m_callback);
+            m_stop_state->remove_callback(m_callback);
         }
     }
 
@@ -220,7 +243,7 @@ public:
 
 private:
     shared_ptr<polyfill::stop_state> m_stop_state;
-    optional<polyfill::stop_state_callbacks::const_iterator> m_callback;
+    polyfill::stop_state_callback m_callback;
 };
 
 template <typename Callback>
@@ -318,6 +341,28 @@ void CondvarWait(Condvar& cv, Lock& lock, std::stop_token token, Pred pred) {
     cv.wait(lock, [&] { return pred() || token.stop_requested(); });
 }
 
+template <typename Rep, typename Period>
+bool StoppableTimedWait(std::stop_token token, const std::chrono::duration<Rep, Period>& rel_time) {
+    if (token.stop_requested()) {
+        return false;
+    }
+
+    bool stop_requested = false;
+    std::condition_variable cv;
+    std::mutex m;
+
+    std::stop_callback cb(token, [&] {
+        // Wake up the waiting thread.
+        std::unique_lock lk{m};
+        stop_requested = true;
+        cv.notify_one();
+    });
+
+    // Perform the timed wait.
+    std::unique_lock lk{m};
+    return !cv.wait_for(lk, rel_time, [&] { return stop_requested; });
+}
+
 } // namespace Common
 
 #endif

src/common/range_map.h

@@ -0,0 +1,139 @@
+// SPDX-FileCopyrightText: Copyright 2022 yuzu Emulator Project
+// SPDX-License-Identifier: GPL-3.0-or-later
+
+#pragma once
+
+#include <map>
+#include <type_traits>
+
+#include "common/common_types.h"
+
+namespace Common {
+
+template <typename KeyTBase, typename ValueT>
+class RangeMap {
+private:
+    using KeyT =
+        std::conditional_t<std::is_signed_v<KeyTBase>, KeyTBase, std::make_signed_t<KeyTBase>>;
+
+public:
+    explicit RangeMap(ValueT null_value_) : null_value{null_value_} {
+        container.emplace(std::numeric_limits<KeyT>::min(), null_value);
+    };
+    ~RangeMap() = default;
+
+    void Map(KeyTBase address, KeyTBase address_end, ValueT value) {
+        KeyT new_address = static_cast<KeyT>(address);
+        KeyT new_address_end = static_cast<KeyT>(address_end);
+        if (new_address < 0) {
+            new_address = 0;
+        }
+        if (new_address_end < 0) {
+            new_address_end = 0;
+        }
+        InternalMap(new_address, new_address_end, value);
+    }
+
+    void Unmap(KeyTBase address, KeyTBase address_end) {
+        Map(address, address_end, null_value);
+    }
+
+    [[nodiscard]] size_t GetContinousSizeFrom(KeyTBase address) const {
+        const KeyT new_address = static_cast<KeyT>(address);
+        if (new_address < 0) {
+            return 0;
+        }
+        return ContinousSizeInternal(new_address);
+    }
+
+    [[nodiscard]] ValueT GetValueAt(KeyT address) const {
+        const KeyT new_address = static_cast<KeyT>(address);
+        if (new_address < 0) {
+            return null_value;
+        }
+        return GetValueInternal(new_address);
+    }
+
+private:
+    using MapType = std::map<KeyT, ValueT>;
+    using IteratorType = typename MapType::iterator;
+    using ConstIteratorType = typename MapType::const_iterator;
+
+    size_t ContinousSizeInternal(KeyT address) const {
+        const auto it = GetFirstElementBeforeOrOn(address);
+        if (it == container.end() || it->second == null_value) {
+            return 0;
+        }
+        const auto it_end = std::next(it);
+        if (it_end == container.end()) {
+            return std::numeric_limits<KeyT>::max() - address;
+        }
+        return it_end->first - address;
+    }
+
+    ValueT GetValueInternal(KeyT address) const {
+        const auto it = GetFirstElementBeforeOrOn(address);
+        if (it == container.end()) {
+            return null_value;
+        }
+        return it->second;
+    }
+
+    ConstIteratorType GetFirstElementBeforeOrOn(KeyT address) const {
+        auto it = container.lower_bound(address);
+        if (it == container.begin()) {
+            return it;
+        }
+        if (it != container.end() && (it->first == address)) {
+            return it;
+        }
+        --it;
+        return it;
+    }
+
+    ValueT GetFirstValueWithin(KeyT address) {
+        auto it = container.lower_bound(address);
+        if (it == container.begin()) {
+            return it->second;
+        }
+        if (it == container.end()) [[unlikely]] { // this would be a bug
+            return null_value;
+        }
+        --it;
+        return it->second;
+    }
+
+    ValueT GetLastValueWithin(KeyT address) {
+        auto it = container.upper_bound(address);
+        if (it == container.end()) {
+            return null_value;
+        }
+        if (it == container.begin()) [[unlikely]] { // this would be a bug
+            return it->second;
+        }
+        --it;
+        return it->second;
+    }
+
+    void InternalMap(KeyT address, KeyT address_end, ValueT value) {
+        const bool must_add_start = GetFirstValueWithin(address) != value;
+        const ValueT last_value = GetLastValueWithin(address_end);
+        const bool must_add_end = last_value != value;
+        auto it = container.lower_bound(address);
+        const auto it_end = container.upper_bound(address_end);
+        while (it != it_end) {
+            it = container.erase(it);
+        }
+        if (must_add_start) {
+            container.emplace(address, value);
+        }
+        if (must_add_end) {
+            container.emplace(address_end, last_value);
+        }
+    }
+
+    ValueT null_value;
+    MapType container;
+};
+
+} // namespace Common

src/common/scratch_buffer.h

@@ -24,6 +24,24 @@ public:
 
     ~ScratchBuffer() = default;
 
+    ScratchBuffer(ScratchBuffer&& rhs) {
+        last_requested_size = rhs.last_requested_size;
+        buffer_capacity = rhs.buffer_capacity;
+        buffer = std::move(rhs.buffer);
+
+        rhs.last_requested_size = 0;
+        rhs.buffer_capacity = 0;
+    }
+
+    void operator=(ScratchBuffer&& rhs) {
+        last_requested_size = rhs.last_requested_size;
+        buffer_capacity = rhs.buffer_capacity;
+        buffer = std::move(rhs.buffer);
+
+        rhs.last_requested_size = 0;
+        rhs.buffer_capacity = 0;
+    }
+
     /// This will only grow the buffer's capacity if size is greater than the current capacity.
     /// The previously held data will remain intact.
     void resize(size_t size) {

src/common/settings.cpp

@@ -129,6 +129,10 @@ void UpdateRescalingInfo() {
         info.up_scale = 1;
         info.down_shift = 0;
         break;
+    case ResolutionSetup::Res3_2X:
+        info.up_scale = 3;
+        info.down_shift = 1;
+        break;
     case ResolutionSetup::Res2X:
         info.up_scale = 2;
         info.down_shift = 0;
@@ -149,6 +153,14 @@ void UpdateRescalingInfo() {
         info.up_scale = 6;
         info.down_shift = 0;
         break;
+    case ResolutionSetup::Res7X:
+        info.up_scale = 7;
+        info.down_shift = 0;
+        break;
+    case ResolutionSetup::Res8X:
+        info.up_scale = 8;
+        info.down_shift = 0;
+        break;
     default:
         ASSERT(false);
         info.up_scale = 1;
@@ -185,6 +197,7 @@ void RestoreGlobalState(bool is_powered_on) {
     // Renderer
     values.fsr_sharpening_slider.SetGlobal(true);
     values.renderer_backend.SetGlobal(true);
+    values.renderer_force_max_clock.SetGlobal(true);
     values.vulkan_device.SetGlobal(true);
     values.aspect_ratio.SetGlobal(true);
     values.max_anisotropy.SetGlobal(true);
@@ -200,6 +213,7 @@ void RestoreGlobalState(bool is_powered_on) {
     values.use_asynchronous_shaders.SetGlobal(true);
     values.use_fast_gpu_time.SetGlobal(true);
     values.use_pessimistic_flushes.SetGlobal(true);
+    values.use_vulkan_driver_pipeline_cache.SetGlobal(true);
     values.bg_red.SetGlobal(true);
     values.bg_green.SetGlobal(true);
     values.bg_blue.SetGlobal(true);

src/common/settings.h

@@ -56,11 +56,14 @@ enum class ResolutionSetup : u32 {
     Res1_2X = 0,
     Res3_4X = 1,
     Res1X = 2,
-    Res2X = 3,
-    Res3X = 4,
-    Res4X = 5,
-    Res5X = 6,
-    Res6X = 7,
+    Res3_2X = 3,
+    Res2X = 4,
+    Res3X = 5,
+    Res4X = 6,
+    Res5X = 7,
+    Res6X = 8,
+    Res7X = 9,
+    Res8X = 10,
 };
 
 enum class ScalingFilter : u32 {
@@ -128,7 +131,8 @@ public:
      * @param default_val Intial value of the setting, and default value of the setting
      * @param name Label for the setting
      */
-    explicit Setting(const Type& default_val, const std::string& name) requires(!ranged)
+    explicit Setting(const Type& default_val, const std::string& name)
+        requires(!ranged)
         : value{default_val}, default_value{default_val}, label{name} {}
     virtual ~Setting() = default;
 
@@ -141,7 +145,8 @@ public:
      * @param name Label for the setting
      */
     explicit Setting(const Type& default_val, const Type& min_val, const Type& max_val,
-                     const std::string& name) requires(ranged)
+                     const std::string& name)
+        requires(ranged)
         : value{default_val},
           default_value{default_val}, maximum{max_val}, minimum{min_val}, label{name} {}
 
@@ -229,7 +234,8 @@ public:
      * @param default_val Intial value of the setting, and default value of the setting
      * @param name Label for the setting
      */
-    explicit SwitchableSetting(const Type& default_val, const std::string& name) requires(!ranged)
+    explicit SwitchableSetting(const Type& default_val, const std::string& name)
+        requires(!ranged)
         : Setting<Type>{default_val, name} {}
     virtual ~SwitchableSetting() = default;
 
@@ -242,7 +248,8 @@ public:
      * @param name Label for the setting
      */
     explicit SwitchableSetting(const Type& default_val, const Type& min_val, const Type& max_val,
-                               const std::string& name) requires(ranged)
+                               const std::string& name)
+        requires(ranged)
         : Setting<Type, true>{default_val, min_val, max_val, name} {}
 
     /**
@@ -415,6 +422,7 @@ struct Values {
     // Renderer
     SwitchableSetting<RendererBackend, true> renderer_backend{
         RendererBackend::Vulkan, RendererBackend::OpenGL, RendererBackend::Null, "backend"};
+    SwitchableSetting<bool> renderer_force_max_clock{false, "force_max_clock"};
     Setting<bool> renderer_debug{false, "debug"};
     Setting<bool> renderer_shader_feedback{false, "shader_feedback"};
     Setting<bool> enable_nsight_aftermath{false, "nsight_aftermath"};
@@ -451,6 +459,8 @@ struct Values {
     SwitchableSetting<bool> use_asynchronous_shaders{false, "use_asynchronous_shaders"};
     SwitchableSetting<bool> use_fast_gpu_time{true, "use_fast_gpu_time"};
     SwitchableSetting<bool> use_pessimistic_flushes{false, "use_pessimistic_flushes"};
+    SwitchableSetting<bool> use_vulkan_driver_pipeline_cache{true,
+                                                             "use_vulkan_driver_pipeline_cache"};
 
     SwitchableSetting<u8> bg_red{0, "bg_red"};
     SwitchableSetting<u8> bg_green{0, "bg_green"};
@@ -477,6 +487,7 @@ struct Values {
 
     Setting<bool> enable_raw_input{false, "enable_raw_input"};
     Setting<bool> controller_navigation{true, "controller_navigation"};
+    Setting<bool> enable_joycon_driver{true, "enable_joycon_driver"};
 
     SwitchableSetting<bool> vibration_enabled{true, "vibration_enabled"};
     SwitchableSetting<bool> enable_accurate_vibrations{false, "enable_accurate_vibrations"};
@@ -531,6 +542,7 @@ struct Values {
     Setting<bool> reporting_services{false, "reporting_services"};
     Setting<bool> quest_flag{false, "quest_flag"};
     Setting<bool> disable_macro_jit{false, "disable_macro_jit"};
+    Setting<bool> disable_macro_hle{false, "disable_macro_hle"};
     Setting<bool> extended_logging{false, "extended_logging"};
     Setting<bool> use_debug_asserts{false, "use_debug_asserts"};
     Setting<bool> use_auto_stub{false, "use_auto_stub"};

src/common/string_util.cpp

@@ -30,7 +30,7 @@ std::string ToUpper(std::string str) {
     return str;
 }
 
-std::string StringFromBuffer(const std::vector<u8>& data) {
+std::string StringFromBuffer(std::span<const u8> data) {
     return std::string(data.begin(), std::find(data.begin(), data.end(), '\0'));
 }
 

src/common/string_util.h

@@ -5,6 +5,7 @@
 #pragma once
 
 #include <cstddef>
+#include <span>
 #include <string>
 #include <vector>
 #include "common/common_types.h"
@@ -17,7 +18,7 @@ namespace Common {
 /// Make a string uppercase
 [[nodiscard]] std::string ToUpper(std::string str);
 
-[[nodiscard]] std::string StringFromBuffer(const std::vector<u8>& data);
+[[nodiscard]] std::string StringFromBuffer(std::span<const u8> data);
 
 [[nodiscard]] std::string StripSpaces(const std::string& s);
 [[nodiscard]] std::string StripQuotes(const std::string& s);

src/common/tree.h

@@ -103,12 +103,12 @@ concept IsRBEntry = CheckRBEntry<T>::value;
 
 template <typename T>
 concept HasRBEntry = requires(T& t, const T& ct) {
-    { t.GetRBEntry() } -> std::same_as<RBEntry<T>&>;
-    { ct.GetRBEntry() } -> std::same_as<const RBEntry<T>&>;
-};
+                         { t.GetRBEntry() } -> std::same_as<RBEntry<T>&>;
+                         { ct.GetRBEntry() } -> std::same_as<const RBEntry<T>&>;
+                     };
 
 template <typename T>
-requires HasRBEntry<T>
+    requires HasRBEntry<T>
 class RBHead {
 private:
     T* m_rbh_root = nullptr;
@@ -130,90 +130,90 @@ public:
 };
 
 template <typename T>
-requires HasRBEntry<T>
+    requires HasRBEntry<T>
 [[nodiscard]] constexpr RBEntry<T>& RB_ENTRY(T* t) {
     return t->GetRBEntry();
 }
 template <typename T>
-requires HasRBEntry<T>
+    requires HasRBEntry<T>
 [[nodiscard]] constexpr const RBEntry<T>& RB_ENTRY(const T* t) {
     return t->GetRBEntry();
 }
 
 template <typename T>
-requires HasRBEntry<T>
+    requires HasRBEntry<T>
 [[nodiscard]] constexpr T* RB_LEFT(T* t) {
     return RB_ENTRY(t).Left();
 }
 template <typename T>
-requires HasRBEntry<T>
+    requires HasRBEntry<T>
 [[nodiscard]] constexpr const T* RB_LEFT(const T* t) {
     return RB_ENTRY(t).Left();
 }
 
 template <typename T>
-requires HasRBEntry<T>
+    requires HasRBEntry<T>
 [[nodiscard]] constexpr T* RB_RIGHT(T* t) {
     return RB_ENTRY(t).Right();
 }
 template <typename T>
-requires HasRBEntry<T>
+    requires HasRBEntry<T>
 [[nodiscard]] constexpr const T* RB_RIGHT(const T* t) {
     return RB_ENTRY(t).Right();
 }
 
 template <typename T>
-requires HasRBEntry<T>
+    requires HasRBEntry<T>
 [[nodiscard]] constexpr T* RB_PARENT(T* t) {
     return RB_ENTRY(t).Parent();
 }
 template <typename T>
-requires HasRBEntry<T>
+    requires HasRBEntry<T>
 [[nodiscard]] constexpr const T* RB_PARENT(const T* t) {
     return RB_ENTRY(t).Parent();
 }
 
 template <typename T>
-requires HasRBEntry<T>
+    requires HasRBEntry<T>
 constexpr void RB_SET_LEFT(T* t, T* e) {
     RB_ENTRY(t).SetLeft(e);
 }
 template <typename T>
-requires HasRBEntry<T>
+    requires HasRBEntry<T>
 constexpr void RB_SET_RIGHT(T* t, T* e) {
     RB_ENTRY(t).SetRight(e);
 }
 template <typename T>
-requires HasRBEntry<T>
+    requires HasRBEntry<T>
 constexpr void RB_SET_PARENT(T* t, T* e) {
     RB_ENTRY(t).SetParent(e);
 }
 
 template <typename T>
-requires HasRBEntry<T>
+    requires HasRBEntry<T>
 [[nodiscard]] constexpr bool RB_IS_BLACK(const T* t) {
     return RB_ENTRY(t).IsBlack();
 }
 template <typename T>
-requires HasRBEntry<T>
+    requires HasRBEntry<T>
 [[nodiscard]] constexpr bool RB_IS_RED(const T* t) {
     return RB_ENTRY(t).IsRed();
 }
 
 template <typename T>
-requires HasRBEntry<T>
+    requires HasRBEntry<T>
 [[nodiscard]] constexpr RBColor RB_COLOR(const T* t) {
     return RB_ENTRY(t).Color();
 }
 
 template <typename T>
-requires HasRBEntry<T>
+    requires HasRBEntry<T>
 constexpr void RB_SET_COLOR(T* t, RBColor c) {
     RB_ENTRY(t).SetColor(c);
 }
 
 template <typename T>
-requires HasRBEntry<T>
+    requires HasRBEntry<T>
 constexpr void RB_SET(T* elm, T* parent) {
     auto& rb_entry = RB_ENTRY(elm);
     rb_entry.SetParent(parent);
@@ -223,14 +223,14 @@ constexpr void RB_SET(T* elm, T* parent) {
 }
 
 template <typename T>
-requires HasRBEntry<T>
+    requires HasRBEntry<T>
 constexpr void RB_SET_BLACKRED(T* black, T* red) {
     RB_SET_COLOR(black, RBColor::RB_BLACK);
     RB_SET_COLOR(red, RBColor::RB_RED);
 }
 
 template <typename T>
-requires HasRBEntry<T>
+    requires HasRBEntry<T>
 constexpr void RB_ROTATE_LEFT(RBHead<T>& head, T* elm, T*& tmp) {
     tmp = RB_RIGHT(elm);
     if (RB_SET_RIGHT(elm, RB_LEFT(tmp)); RB_RIGHT(elm) != nullptr) {
@@ -252,7 +252,7 @@ constexpr void RB_ROTATE_LEFT(RBHead<T>& head, T* elm, T*& tmp) {
 }
 
 template <typename T>
-requires HasRBEntry<T>
+    requires HasRBEntry<T>
 constexpr void RB_ROTATE_RIGHT(RBHead<T>& head, T* elm, T*& tmp) {
     tmp = RB_LEFT(elm);
     if (RB_SET_LEFT(elm, RB_RIGHT(tmp)); RB_LEFT(elm) != nullptr) {
@@ -274,7 +274,7 @@ constexpr void RB_ROTATE_RIGHT(RBHead<T>& head, T* elm, T*& tmp) {
 }
 
 template <typename T>
-requires HasRBEntry<T>
+    requires HasRBEntry<T>
 constexpr void RB_REMOVE_COLOR(RBHead<T>& head, T* parent, T* elm) {
     T* tmp;
     while ((elm == nullptr || RB_IS_BLACK(elm)) && elm != head.Root()) {
@@ -358,7 +358,7 @@ constexpr void RB_REMOVE_COLOR(RBHead<T>& head, T* parent, T* elm) {
 }
 
 template <typename T>
-requires HasRBEntry<T>
+    requires HasRBEntry<T>
 constexpr T* RB_REMOVE(RBHead<T>& head, T* elm) {
     T* child = nullptr;
     T* parent = nullptr;
@@ -451,7 +451,7 @@ constexpr T* RB_REMOVE(RBHead<T>& head, T* elm) {
 }
 
 template <typename T>
-requires HasRBEntry<T>
+    requires HasRBEntry<T>
 constexpr void RB_INSERT_COLOR(RBHead<T>& head, T* elm) {
     T *parent = nullptr, *tmp = nullptr;
     while ((parent = RB_PARENT(elm)) != nullptr && RB_IS_RED(parent)) {
@@ -499,7 +499,7 @@ constexpr void RB_INSERT_COLOR(RBHead<T>& head, T* elm) {
 }
 
 template <typename T, typename Compare>
-requires HasRBEntry<T>
+    requires HasRBEntry<T>
 constexpr T* RB_INSERT(RBHead<T>& head, T* elm, Compare cmp) {
     T* parent = nullptr;
     T* tmp = head.Root();
@@ -534,7 +534,7 @@ constexpr T* RB_INSERT(RBHead<T>& head, T* elm, Compare cmp) {
 }
 
 template <typename T, typename Compare>
-requires HasRBEntry<T>
+    requires HasRBEntry<T>
 constexpr T* RB_FIND(RBHead<T>& head, T* elm, Compare cmp) {
     T* tmp = head.Root();
 
@@ -553,7 +553,7 @@ constexpr T* RB_FIND(RBHead<T>& head, T* elm, Compare cmp) {
 }
 
 template <typename T, typename Compare>
-requires HasRBEntry<T>
+    requires HasRBEntry<T>
 constexpr T* RB_NFIND(RBHead<T>& head, T* elm, Compare cmp) {
     T* tmp = head.Root();
     T* res = nullptr;
@@ -574,7 +574,7 @@ constexpr T* RB_NFIND(RBHead<T>& head, T* elm, Compare cmp) {
 }
 
 template <typename T, typename U, typename Compare>
-requires HasRBEntry<T>
+    requires HasRBEntry<T>
 constexpr T* RB_FIND_KEY(RBHead<T>& head, const U& key, Compare cmp) {
     T* tmp = head.Root();
 
@@ -593,7 +593,7 @@ constexpr T* RB_FIND_KEY(RBHead<T>& head, const U& key, Compare cmp) {
 }
 
 template <typename T, typename U, typename Compare>
-requires HasRBEntry<T>
+    requires HasRBEntry<T>
 constexpr T* RB_NFIND_KEY(RBHead<T>& head, const U& key, Compare cmp) {
     T* tmp = head.Root();
     T* res = nullptr;
@@ -614,7 +614,7 @@ constexpr T* RB_NFIND_KEY(RBHead<T>& head, const U& key, Compare cmp) {
 }
 
 template <typename T, typename Compare>
-requires HasRBEntry<T>
+    requires HasRBEntry<T>
 constexpr T* RB_FIND_EXISTING(RBHead<T>& head, T* elm, Compare cmp) {
     T* tmp = head.Root();
 
@@ -631,7 +631,7 @@ constexpr T* RB_FIND_EXISTING(RBHead<T>& head, T* elm, Compare cmp) {
 }
 
 template <typename T, typename U, typename Compare>
-requires HasRBEntry<T>
+    requires HasRBEntry<T>
 constexpr T* RB_FIND_EXISTING_KEY(RBHead<T>& head, const U& key, Compare cmp) {
     T* tmp = head.Root();
 
@@ -648,7 +648,7 @@ constexpr T* RB_FIND_EXISTING_KEY(RBHead<T>& head, const U& key, Compare cmp) {
 }
 
 template <typename T>
-requires HasRBEntry<T>
+    requires HasRBEntry<T>
 constexpr T* RB_NEXT(T* elm) {
     if (RB_RIGHT(elm)) {
         elm = RB_RIGHT(elm);
@@ -669,7 +669,7 @@ constexpr T* RB_NEXT(T* elm) {
 }
 
 template <typename T>
-requires HasRBEntry<T>
+    requires HasRBEntry<T>
 constexpr T* RB_PREV(T* elm) {
     if (RB_LEFT(elm)) {
         elm = RB_LEFT(elm);
@@ -690,7 +690,7 @@ constexpr T* RB_PREV(T* elm) {
 }
 
 template <typename T>
-requires HasRBEntry<T>
+    requires HasRBEntry<T>
 constexpr T* RB_MIN(RBHead<T>& head) {
     T* tmp = head.Root();
     T* parent = nullptr;
@@ -704,7 +704,7 @@ constexpr T* RB_MIN(RBHead<T>& head) {
 }
 
 template <typename T>
-requires HasRBEntry<T>
+    requires HasRBEntry<T>
 constexpr T* RB_MAX(RBHead<T>& head) {
     T* tmp = head.Root();
     T* parent = nullptr;

src/common/vector_math.h

@@ -348,9 +348,7 @@ public:
 // _DEFINE_SWIZZLER2 defines a single such function, DEFINE_SWIZZLER2 defines all of them for all
 // component names (x<->r) and permutations (xy<->yx)
 #define _DEFINE_SWIZZLER2(a, b, name)                                                              \
-    [[nodiscard]] constexpr Vec2<T> name() const {                                                 \
-        return Vec2<T>(a, b);                                                                      \
-    }
+    [[nodiscard]] constexpr Vec2<T> name() const { return Vec2<T>(a, b); }
 #define DEFINE_SWIZZLER2(a, b, a2, b2, a3, b3, a4, b4)                                             \
     _DEFINE_SWIZZLER2(a, b, a##b);                                                                 \
     _DEFINE_SWIZZLER2(a, b, a2##b2);                                                               \
@@ -543,9 +541,7 @@ public:
 // DEFINE_SWIZZLER2_COMP2 defines two component functions for all component names (x<->r) and
 // permutations (xy<->yx)
 #define _DEFINE_SWIZZLER2(a, b, name)                                                              \
-    [[nodiscard]] constexpr Vec2<T> name() const {                                                 \
-        return Vec2<T>(a, b);                                                                      \
-    }
+    [[nodiscard]] constexpr Vec2<T> name() const { return Vec2<T>(a, b); }
 #define DEFINE_SWIZZLER2_COMP1(a, a2)                                                              \
     _DEFINE_SWIZZLER2(a, a, a##a);                                                                 \
     _DEFINE_SWIZZLER2(a, a, a2##a2)
@@ -570,9 +566,7 @@ public:
 #undef _DEFINE_SWIZZLER2
 
 #define _DEFINE_SWIZZLER3(a, b, c, name)                                                           \
-    [[nodiscard]] constexpr Vec3<T> name() const {                                                 \
-        return Vec3<T>(a, b, c);                                                                   \
-    }
+    [[nodiscard]] constexpr Vec3<T> name() const { return Vec3<T>(a, b, c); }
 #define DEFINE_SWIZZLER3_COMP1(a, a2)                                                              \
     _DEFINE_SWIZZLER3(a, a, a, a##a##a);                                                           \
     _DEFINE_SWIZZLER3(a, a, a, a2##a2##a2)
@@ -641,8 +635,8 @@ template <typename T>
 
 // linear interpolation via float: 0.0=begin, 1.0=end
 template <typename X>
-[[nodiscard]] constexpr decltype(X{} * float{} + X{} * float{})
-    Lerp(const X& begin, const X& end, const float t) {
+[[nodiscard]] constexpr decltype(X{} * float{} + X{} * float{}) Lerp(const X& begin, const X& end,
+                                                                     const float t) {
     return begin * (1.f - t) + end * t;
 }
 

src/core/CMakeLists.txt

@@ -182,6 +182,8 @@ add_library(core STATIC
     hle/kernel/k_auto_object_container.cpp
     hle/kernel/k_auto_object_container.h
     hle/kernel/k_affinity_mask.h
+    hle/kernel/k_capabilities.cpp
+    hle/kernel/k_capabilities.h
     hle/kernel/k_class_token.cpp
     hle/kernel/k_class_token.h
     hle/kernel/k_client_port.cpp
@@ -193,6 +195,8 @@ add_library(core STATIC
     hle/kernel/k_condition_variable.cpp
     hle/kernel/k_condition_variable.h
     hle/kernel/k_debug.h
+    hle/kernel/k_device_address_space.cpp
+    hle/kernel/k_device_address_space.h
     hle/kernel/k_dynamic_page_manager.h
     hle/kernel/k_dynamic_resource_manager.h
     hle/kernel/k_dynamic_slab_heap.h
@@ -294,7 +298,42 @@ add_library(core STATIC
     hle/kernel/svc.h
     hle/kernel/svc_common.h
     hle/kernel/svc_types.h
-    hle/kernel/svc_wrap.h
+    hle/kernel/svc/svc_activity.cpp
+    hle/kernel/svc/svc_address_arbiter.cpp
+    hle/kernel/svc/svc_address_translation.cpp
+    hle/kernel/svc/svc_cache.cpp
+    hle/kernel/svc/svc_code_memory.cpp
+    hle/kernel/svc/svc_condition_variable.cpp
+    hle/kernel/svc/svc_debug.cpp
+    hle/kernel/svc/svc_debug_string.cpp
+    hle/kernel/svc/svc_device_address_space.cpp
+    hle/kernel/svc/svc_event.cpp
+    hle/kernel/svc/svc_exception.cpp
+    hle/kernel/svc/svc_info.cpp
+    hle/kernel/svc/svc_interrupt_event.cpp
+    hle/kernel/svc/svc_io_pool.cpp
+    hle/kernel/svc/svc_ipc.cpp
+    hle/kernel/svc/svc_kernel_debug.cpp
+    hle/kernel/svc/svc_light_ipc.cpp
+    hle/kernel/svc/svc_lock.cpp
+    hle/kernel/svc/svc_memory.cpp
+    hle/kernel/svc/svc_physical_memory.cpp
+    hle/kernel/svc/svc_port.cpp
+    hle/kernel/svc/svc_power_management.cpp
+    hle/kernel/svc/svc_process.cpp
+    hle/kernel/svc/svc_process_memory.cpp
+    hle/kernel/svc/svc_processor.cpp
+    hle/kernel/svc/svc_query_memory.cpp
+    hle/kernel/svc/svc_register.cpp
+    hle/kernel/svc/svc_resource_limit.cpp
+    hle/kernel/svc/svc_secure_monitor_call.cpp
+    hle/kernel/svc/svc_session.cpp
+    hle/kernel/svc/svc_shared_memory.cpp
+    hle/kernel/svc/svc_synchronization.cpp
+    hle/kernel/svc/svc_thread.cpp
+    hle/kernel/svc/svc_thread_profiler.cpp
+    hle/kernel/svc/svc_tick.cpp
+    hle/kernel/svc/svc_transfer_memory.cpp
     hle/result.h
     hle/service/acc/acc.cpp
     hle/service/acc/acc.h

src/core/arm/arm_interface.cpp

@@ -1,14 +1,12 @@
 // SPDX-FileCopyrightText: Copyright 2018 yuzu Emulator Project
 // SPDX-License-Identifier: GPL-2.0-or-later
 
-#ifndef _MSC_VER
-#include <cxxabi.h>
-#endif
-
 #include <map>
 #include <optional>
+
 #include "common/bit_field.h"
 #include "common/common_types.h"
+#include "common/demangle.h"
 #include "common/logging/log.h"
 #include "core/arm/arm_interface.h"
 #include "core/arm/symbols.h"
@@ -71,20 +69,8 @@ void ARM_Interface::SymbolicateBacktrace(Core::System& system, std::vector<Backt
         const auto symbol_set = symbols.find(entry.module);
         if (symbol_set != symbols.end()) {
             const auto symbol = Symbols::GetSymbolName(symbol_set->second, entry.offset);
-            if (symbol.has_value()) {
-#ifdef _MSC_VER
-                // TODO(DarkLordZach): Add demangling of symbol names.
-                entry.name = *symbol;
-#else
-                int status{-1};
-                char* demangled{abi::__cxa_demangle(symbol->c_str(), nullptr, nullptr, &status)};
-                if (status == 0 && demangled != nullptr) {
-                    entry.name = demangled;
-                    std::free(demangled);
-                } else {
-                    entry.name = *symbol;
-                }
-#endif
+            if (symbol) {
+                entry.name = Common::DemangleSymbol(*symbol);
             }
         }
     }

src/core/arm/dynarmic/arm_dynarmic_32.cpp

@@ -229,7 +229,11 @@ std::shared_ptr<Dynarmic::A32::Jit> ARM_Dynarmic_32::MakeJit(Common::PageTable*
     config.enable_cycle_counting = true;
 
     // Code cache size
+#ifdef ARCHITECTURE_arm64
+    config.code_cache_size = 128_MiB;
+#else
     config.code_cache_size = 512_MiB;
+#endif
 
     // Allow memory fault handling to work
     if (system.DebuggerEnabled()) {

src/core/arm/dynarmic/arm_dynarmic_64.cpp

@@ -288,7 +288,11 @@ std::shared_ptr<Dynarmic::A64::Jit> ARM_Dynarmic_64::MakeJit(Common::PageTable*
     config.enable_cycle_counting = true;
 
     // Code cache size
+#ifdef ARCHITECTURE_arm64
+    config.code_cache_size = 128_MiB;
+#else
     config.code_cache_size = 512_MiB;
+#endif
 
     // Allow memory fault handling to work
     if (system.DebuggerEnabled()) {

src/core/core_timing.cpp

@@ -142,16 +142,24 @@ void CoreTiming::ScheduleLoopingEvent(std::chrono::nanoseconds start_time,
 }
 
 void CoreTiming::UnscheduleEvent(const std::shared_ptr<EventType>& event_type,
-                                 std::uintptr_t user_data) {
-    std::scoped_lock scope{basic_lock};
-    const auto itr = std::remove_if(event_queue.begin(), event_queue.end(), [&](const Event& e) {
-        return e.type.lock().get() == event_type.get() && e.user_data == user_data;
-    });
+                                 std::uintptr_t user_data, bool wait) {
+    {
+        std::scoped_lock lk{basic_lock};
+        const auto itr =
+            std::remove_if(event_queue.begin(), event_queue.end(), [&](const Event& e) {
+                return e.type.lock().get() == event_type.get() && e.user_data == user_data;
+            });
 
-    // Removing random items breaks the invariant so we have to re-establish it.
-    if (itr != event_queue.end()) {
-        event_queue.erase(itr, event_queue.end());
-        std::make_heap(event_queue.begin(), event_queue.end(), std::greater<>());
+        // Removing random items breaks the invariant so we have to re-establish it.
+        if (itr != event_queue.end()) {
+            event_queue.erase(itr, event_queue.end());
+            std::make_heap(event_queue.begin(), event_queue.end(), std::greater<>());
+        }
+    }
+
+    // Force any in-progress events to finish
+    if (wait) {
+        std::scoped_lock lk{advance_lock};
     }
 }
 
@@ -190,20 +198,6 @@ u64 CoreTiming::GetClockTicks() const {
     return CpuCyclesToClockCycles(ticks);
 }
 
-void CoreTiming::RemoveEvent(const std::shared_ptr<EventType>& event_type) {
-    std::scoped_lock lock{basic_lock};
-
-    const auto itr = std::remove_if(event_queue.begin(), event_queue.end(), [&](const Event& e) {
-        return e.type.lock().get() == event_type.get();
-    });
-
-    // Removing random items breaks the invariant so we have to re-establish it.
-    if (itr != event_queue.end()) {
-        event_queue.erase(itr, event_queue.end());
-        std::make_heap(event_queue.begin(), event_queue.end(), std::greater<>());
-    }
-}
-
 std::optional<s64> CoreTiming::Advance() {
     std::scoped_lock lock{advance_lock, basic_lock};
     global_timer = GetGlobalTimeNs().count();

src/core/core_timing.h

@@ -98,10 +98,13 @@ public:
                               const std::shared_ptr<EventType>& event_type,
                               std::uintptr_t user_data = 0, bool absolute_time = false);
 
-    void UnscheduleEvent(const std::shared_ptr<EventType>& event_type, std::uintptr_t user_data);
+    void UnscheduleEvent(const std::shared_ptr<EventType>& event_type, std::uintptr_t user_data,
+                         bool wait = true);
 
-    /// We only permit one event of each type in the queue at a time.
-    void RemoveEvent(const std::shared_ptr<EventType>& event_type);
+    void UnscheduleEventWithoutWait(const std::shared_ptr<EventType>& event_type,
+                                    std::uintptr_t user_data) {
+        UnscheduleEvent(event_type, user_data, false);
+    }
 
     void AddTicks(u64 ticks_to_add);
 

src/core/debugger/gdbstub.cpp

@@ -11,6 +11,7 @@
 #include "common/hex_util.h"
 #include "common/logging/log.h"
 #include "common/scope_exit.h"
+#include "common/settings.h"
 #include "core/arm/arm_interface.h"
 #include "core/core.h"
 #include "core/debugger/gdbstub.h"
@@ -731,7 +732,25 @@ void GDBStub::HandleRcmd(const std::vector<u8>& command) {
     auto* process = system.CurrentProcess();
     auto& page_table = process->PageTable();
 
-    if (command_str == "get info") {
+    const char* commands = "Commands:\n"
+                           "  get fastmem\n"
+                           "  get info\n"
+                           "  get mappings\n";
+
+    if (command_str == "get fastmem") {
+        if (Settings::IsFastmemEnabled()) {
+            const auto& impl = page_table.PageTableImpl();
+            const auto region = reinterpret_cast<uintptr_t>(impl.fastmem_arena);
+            const auto region_bits = impl.current_address_space_width_in_bits;
+            const auto region_size = 1ULL << region_bits;
+
+            reply = fmt::format("Region bits:  {}\n"
+                                "Host address: {:#x} - {:#x}\n",
+                                region_bits, region, region + region_size - 1);
+        } else {
+            reply = "Fastmem is not enabled.\n";
+        }
+    } else if (command_str == "get info") {
         Loader::AppLoader::Modules modules;
         system.GetAppLoader().ReadNSOModules(modules);
 
@@ -787,9 +806,10 @@ void GDBStub::HandleRcmd(const std::vector<u8>& command) {
             cur_addr = next_address;
         }
     } else if (command_str == "help") {
-        reply = "Commands:\n  get info\n  get mappings\n";
+        reply = commands;
     } else {
-        reply = "Unknown command.\nCommands:\n  get info\n  get mappings\n";
+        reply = "Unknown command.\n";
+        reply += commands;
     }
 
     std::span<const u8> reply_span{reinterpret_cast<u8*>(&reply.front()), reply.size()};

src/core/file_sys/vfs.cpp

@@ -194,9 +194,9 @@ std::size_t VfsFile::WriteBytes(const std::vector<u8>& data, std::size_t offset)
 
 std::string VfsFile::GetFullPath() const {
     if (GetContainingDirectory() == nullptr)
-        return "/" + GetName();
+        return '/' + GetName();
 
-    return GetContainingDirectory()->GetFullPath() + "/" + GetName();
+    return GetContainingDirectory()->GetFullPath() + '/' + GetName();
 }
 
 VirtualFile VfsDirectory::GetFileRelative(std::string_view path) const {
@@ -435,7 +435,7 @@ std::string VfsDirectory::GetFullPath() const {
     if (IsRoot())
         return GetName();
 
-    return GetParentDirectory()->GetFullPath() + "/" + GetName();
+    return GetParentDirectory()->GetFullPath() + '/' + GetName();
 }
 
 bool ReadOnlyVfsDirectory::IsWritable() const {

src/core/hardware_properties.h

@@ -25,6 +25,26 @@ constexpr std::array<s32, Common::BitSize<u64>()> VirtualToPhysicalCoreMap{
     0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 3,
 };
 
+static constexpr inline size_t NumVirtualCores = Common::BitSize<u64>();
+
+static constexpr inline u64 VirtualCoreMask = [] {
+    u64 mask = 0;
+    for (size_t i = 0; i < NumVirtualCores; ++i) {
+        mask |= (UINT64_C(1) << i);
+    }
+    return mask;
+}();
+
+static constexpr inline u64 ConvertVirtualCoreMaskToPhysical(u64 v_core_mask) {
+    u64 p_core_mask = 0;
+    while (v_core_mask != 0) {
+        const u64 next = std::countr_zero(v_core_mask);
+        v_core_mask &= ~(static_cast<u64>(1) << next);
+        p_core_mask |= (static_cast<u64>(1) << VirtualToPhysicalCoreMap[next]);
+    }
+    return p_core_mask;
+}
+
 // Cortex-A57 supports 4 memory watchpoints
 constexpr u64 NUM_WATCHPOINTS = 4;