Compare commits

..

1 Commits

Author SHA1 Message Date
Bartosz Taudul
9618131ce9 Histogram zooming prototype. 2018-09-03 23:18:26 +02:00
561 changed files with 73389 additions and 466274 deletions

19
.appveyor.yml Normal file
View File

@@ -0,0 +1,19 @@
version: '{build}'
platform:
- x64
image:
- Visual Studio 2017
- Ubuntu
before_build:
- cmd: cd profiler\build\win32
- cmd: nuget restore
- cmd: cd ..\..\..
build_script:
- cmd: msbuild .\update\build\win32\update.vcxproj
- cmd: msbuild .\profiler\build\win32\Tracy.vcxproj
- cmd: msbuild .\capture\build\win32\capture.vcxproj
- sh: sudo apt-get update && sudo apt-get -y install libglfw3-dev libgtk2.0-dev
- sh: make -C update/build/unix debug release
- sh: make -C profiler/build/unix debug release
- sh: make -C capture/build/unix debug release
test: off

View File

@@ -1,62 +0,0 @@
---
Checks:
'
clang-diagnostic-*,
clang-analyzer-*,
bugprone-*,
google-*,
misc-*,
modernize-*,
performance-*,
readability-*,
-bugprone-easily-swappable-parameters,
-bugprone-implicit-widening-of-multiplication-result,
-bugprone-narrowing-conversions,
-bugprone-reserved-identifier,
-google-readability-braces-around-statements,
-google-readability-casting,
-google-readability-function-size,
-google-readability-todo,
-google-readability-namespace-comments,
-misc-confusable-identifiers,
-misc-no-recursion,
-modernize-avoid-c-arrays,
-modernize-deprecated-headers,
-modernize-use-default-member-init,
-modernize-use-trailing-return-type,
-performance-no-int-to-ptr,
-readability-braces-around-statements,
-readability-else-after-return,
-readability-function-cognitive-complexity,
-readability-identifier-length,
-readability-implicit-bool-conversion,
-readability-isolate-declaration,
-readability-magic-numbers,
-readability-qualified-auto,
-readability-uppercase-literal-suffix
'
WarningsAsErrors: ''
HeaderFilterRegex: ''
AnalyzeTemporaryDtors: false
FormatStyle: none
CheckOptions:
llvm-else-after-return.WarnOnConditionVariables: 'false'
modernize-loop-convert.MinConfidence: reasonable
modernize-replace-auto-ptr.IncludeStyle: llvm
modernize-pass-by-value.IncludeStyle: llvm
google-readability-namespace-comments.ShortNamespaceLines: '10'
google-readability-namespace-comments.SpacesBeforeComments: '2'
cppcoreguidelines-non-private-member-variables-in-classes.IgnoreClassesWithAllMemberVariablesBeingPublic: 'true'
google-readability-braces-around-statements.ShortStatementLines: '1'
cert-err33-c.CheckedFunctions: '::aligned_alloc;::asctime_s;::at_quick_exit;::atexit;::bsearch;::bsearch_s;::btowc;::c16rtomb;::c32rtomb;::calloc;::clock;::cnd_broadcast;::cnd_init;::cnd_signal;::cnd_timedwait;::cnd_wait;::ctime_s;::fclose;::fflush;::fgetc;::fgetpos;::fgets;::fgetwc;::fopen;::fopen_s;::fprintf;::fprintf_s;::fputc;::fputs;::fputwc;::fputws;::fread;::freopen;::freopen_s;::fscanf;::fscanf_s;::fseek;::fsetpos;::ftell;::fwprintf;::fwprintf_s;::fwrite;::fwscanf;::fwscanf_s;::getc;::getchar;::getenv;::getenv_s;::gets_s;::getwc;::getwchar;::gmtime;::gmtime_s;::localtime;::localtime_s;::malloc;::mbrtoc16;::mbrtoc32;::mbsrtowcs;::mbsrtowcs_s;::mbstowcs;::mbstowcs_s;::memchr;::mktime;::mtx_init;::mtx_lock;::mtx_timedlock;::mtx_trylock;::mtx_unlock;::printf_s;::putc;::putwc;::raise;::realloc;::remove;::rename;::scanf;::scanf_s;::setlocale;::setvbuf;::signal;::snprintf;::snprintf_s;::sprintf;::sprintf_s;::sscanf;::sscanf_s;::strchr;::strerror_s;::strftime;::strpbrk;::strrchr;::strstr;::strtod;::strtof;::strtoimax;::strtok;::strtok_s;::strtol;::strtold;::strtoll;::strtoul;::strtoull;::strtoumax;::strxfrm;::swprintf;::swprintf_s;::swscanf;::swscanf_s;::thrd_create;::thrd_detach;::thrd_join;::thrd_sleep;::time;::timespec_get;::tmpfile;::tmpfile_s;::tmpnam;::tmpnam_s;::tss_create;::tss_get;::tss_set;::ungetc;::ungetwc;::vfprintf;::vfprintf_s;::vfscanf;::vfscanf_s;::vfwprintf;::vfwprintf_s;::vfwscanf;::vfwscanf_s;::vprintf_s;::vscanf;::vscanf_s;::vsnprintf;::vsnprintf_s;::vsprintf;::vsprintf_s;::vsscanf;::vsscanf_s;::vswprintf;::vswprintf_s;::vswscanf;::vswscanf_s;::vwprintf_s;::vwscanf;::vwscanf_s;::wcrtomb;::wcschr;::wcsftime;::wcspbrk;::wcsrchr;::wcsrtombs;::wcsrtombs_s;::wcsstr;::wcstod;::wcstof;::wcstoimax;::wcstok;::wcstok_s;::wcstol;::wcstold;::wcstoll;::wcstombs;::wcstombs_s;::wcstoul;::wcstoull;::wcstoumax;::wcsxfrm;::wctob;::wctrans;::wctype;::wmemchr;::wprintf_s;::wscanf;::wscanf_s;'
modernize-loop-convert.MaxCopySize: '16'
cert-dcl16-c.NewSuffixes: 'L;LL;LU;LLU'
cert-oop54-cpp.WarnOnlyIfThisHasSuspiciousField: 'false'
cert-str34-c.DiagnoseSignedUnsignedCharComparisons: 'false'
modernize-use-nullptr.NullMacros: 'NULL'
llvm-qualified-auto.AddConstToQualified: 'false'
modernize-loop-convert.NamingStyle: CamelCase
llvm-else-after-return.WarnOnUnfixable: 'false'
google-readability-function-size.StatementThreshold: '800'
...

1
.github/FUNDING.yml vendored
View File

@@ -1 +0,0 @@
github: wolfpld

BIN
.github/sponsor.png vendored

Binary file not shown.

Before

Width:  |  Height:  |  Size: 1.0 KiB

View File

@@ -1,28 +0,0 @@
name: Manual
on:
push:
branches: [ master ]
pull_request:
branches: [ master ]
jobs:
build:
runs-on: ubuntu-latest
steps:
- uses: actions/checkout@v2
- name: Fix stupidity
run: |
cp AUTHORS AUTHORS.
cp LICENSE LICENSE.
- name: Compile LaTeX
uses: xu-cheng/latex-action@v2
with:
working_directory: manual
root_file: tracy.tex
- uses: actions/upload-artifact@v2
with:
name: manual
path: manual/tracy.pdf

View File

@@ -1,41 +0,0 @@
name: linux
on:
push:
branches: [ master ]
pull_request:
branches: [ master ]
jobs:
build:
runs-on: ubuntu-22.04
container: archlinux:base-devel
steps:
- uses: actions/checkout@v2
- name: Install linux libraries
run: pacman -Syu --noconfirm && pacman -S --noconfirm --needed freetype2 capstone tbb debuginfod wayland dbus libxkbcommon libglvnd meson
- name: Profiler GUI
run: make -j`nproc` -C profiler/build/unix debug release
- name: Update utility
run: make -j`nproc` -C update/build/unix debug release
- name: Capture utility
run: make -j`nproc` -C capture/build/unix debug release
- name: Csvexport utility
run: make -j`nproc` -C csvexport/build/unix debug release
- name: Import-chrome utility
run: make -j`nproc` -C import-chrome/build/unix debug release
- name: Import-fuchsia utility
run: make -j`nproc` -C import-fuchsia/build/unix debug release
- name: Library
run: make -j`nproc` -C library/unix debug release
- name: Library (meson)
run: meson setup -Dprefix=$PWD/install build && meson compile -C build && meson install -C build
- name: Test application
run: |
make -j`nproc` -C test
make -j`nproc` -C test clean
make -j`nproc` -C test TRACYFLAGS=-DTRACY_ON_DEMAND
make -j`nproc` -C test clean
make -j`nproc` -C test TRACYFLAGS="-DTRACY_DELAYED_INIT -DTRACY_MANUAL_LIFETIME"
make -C test -B ../public/TracyClient.o DEFINES='-DTRACY_DEMANGLE'

View File

@@ -1,40 +0,0 @@
name: macos
on:
push:
branches: [ master ]
pull_request:
branches: [ master ]
jobs:
build:
runs-on: macOS-latest
steps:
- uses: actions/checkout@v2
- name: Install macos libraries
run: brew install capstone tbb pkg-config glfw meson
- name: Profiler GUI
run: make -j`nproc` -C profiler/build/unix debug release
- name: Update utility
run: make -j`nproc` -C update/build/unix debug release
- name: Capture utility
run: make -j`nproc` -C capture/build/unix debug release
- name: Csvexport utility
run: make -j`nproc` -C csvexport/build/unix debug release
- name: Import-chrome utility
run: make -j`nproc` -C import-chrome/build/unix debug release
- name: Import-fuchsia utility
run: make -j`nproc` -C import-fuchsia/build/unix debug release
- name: Library
run: make -j`nproc` -C library/unix debug release
- name: Library (meson)
run: meson setup -Dprefix=$PWD/install build && meson compile -C build && meson install -C build
- name: Test application
run: |
make -j`nproc` -C test
make -j`nproc` -C test clean
make -j`nproc` -C test TRACYFLAGS=-DTRACY_ON_DEMAND
make -j`nproc` -C test clean
make -j`nproc` -C test TRACYFLAGS="-DTRACY_DELAYED_INIT -DTRACY_MANUAL_LIFETIME"
make -C test -B ../public/TracyClient.o DEFINES='-DTRACY_DEMANGLE'

View File

@@ -1,76 +0,0 @@
name: MSVC
on:
push:
branches: [ master ]
pull_request:
branches: [ master ]
jobs:
build-with-vcpkg-integration:
runs-on: windows-2022
steps:
- uses: actions/checkout@v2
- uses: microsoft/setup-msbuild@v1.0.2
- name: Integrate vcpkg
run: vcpkg integrate install
- name: Profiler GUI Debug
run: msbuild .\profiler\build\win32\Tracy.vcxproj /property:Configuration=Debug /property:Platform=x64
- name: Profiler GUI Release
run: msbuild .\profiler\build\win32\Tracy.vcxproj /property:Configuration=Release /property:Platform=x64
- name: Update utility Debug
run: msbuild .\update\build\win32\update.vcxproj /property:Configuration=Debug /property:Platform=x64
- name: Update utility Release
run: msbuild .\update\build\win32\update.vcxproj /property:Configuration=Release /property:Platform=x64
- name: Capture utility Debug
run: msbuild .\capture\build\win32\capture.vcxproj /property:Configuration=Debug /property:Platform=x64
- name: Capture utility Release
run: msbuild .\capture\build\win32\capture.vcxproj /property:Configuration=Release /property:Platform=x64
- name: Csvexport utility Debug
run: msbuild .\csvexport\build\win32\csvexport.vcxproj /property:Configuration=Debug /property:Platform=x64
- name: Csvexport utility Release
run: msbuild .\csvexport\build\win32\csvexport.vcxproj /property:Configuration=Release /property:Platform=x64
- name: Import-chrome utility Debug
run: msbuild .\import-chrome\build\win32\import-chrome.vcxproj /property:Configuration=Debug /property:Platform=x64
- name: Import-chrome utility Release
run: msbuild .\import-chrome\build\win32\import-chrome.vcxproj /property:Configuration=Release /property:Platform=x64
- name: Import-fuchsia utility Debug
run: msbuild .\import-fuchsia\build\win32\import-fuchsia.vcxproj /property:Configuration=Debug /property:Platform=x64
- name: Import-fuchsia utility Release
run: msbuild .\import-fuchsia\build\win32\import-fuchsia.vcxproj /property:Configuration=Release /property:Platform=x64
- name: Library
run: msbuild .\library\win32\TracyProfiler.vcxproj /property:Configuration=Release /property:Platform=x64
- name: Package binaries
run: |
mkdir bin
mkdir bin\dev
copy profiler\build\win32\x64\Release\Tracy.exe bin
copy update\build\win32\x64\Release\update.exe bin
copy capture\build\win32\x64\Release\capture.exe bin
copy import-chrome\build\win32\x64\Release\import-chrome.exe bin
copy csvexport\build\win32\x64\Release\csvexport.exe bin
copy library\win32\x64\Release\TracyProfiler.dll bin\dev
copy library\win32\x64\Release\TracyProfiler.lib bin\dev
7z a Tracy.7z bin
- uses: actions/upload-artifact@v2
with:
path: Tracy.7z
build-without-vcpkg-integration:
runs-on: windows-2022
env:
VCPKG_ROOT: ''
steps:
- uses: actions/checkout@v2
- uses: microsoft/setup-msbuild@v1.0.2
- name: Install vcpkg dependencies
run: vcpkg/install_vcpkg_dependencies.bat
- name: Profiler GUI Release
run: msbuild .\profiler\build\win32\Tracy.vcxproj /property:Configuration=Release /property:Platform=x64
- name: Capture utility Release
run: msbuild .\capture\build\win32\capture.vcxproj /property:Configuration=Release /property:Platform=x64

37
.gitignore vendored
View File

@@ -5,42 +5,17 @@
x64
Release
Debug
_build
_compiler
tools/*
*.d
*.o
*.so
*.swp
*.obj
imgui.ini
test/tracy_test
test/tracy_test.exe
*/build/unix/*-*
manual/t*.aux
manual/t*.log
manual/t*.out
manual/t*.pdf
manual/t*.synctex.gz
manual/t*.toc
manual/t*.bbl
manual/t*.blg
manual/tracy.aux
manual/tracy.log
manual/tracy.out
manual/tracy.pdf
manual/tracy.synctex.gz
manual/tracy.toc
profiler/build/win32/packages
profiler/build/win32/Tracy.aps
# include the vcpkg install script but not the files it produces
vcpkg/*
!vcpkg/install_vcpkg_dependencies.bat
/vcpkg_installed
.deps/
.dirstamp
.vscode/
/_*/**
/**/__pycache__/**
extra/vswhere.exe
extra/tracy-build
/.cache
compile_commands.json
profiler/build/wasm/Tracy-release.*
profiler/build/wasm/Tracy-debug.*
profiler/build/wasm/embed.tracy

View File

@@ -1,17 +0,0 @@
{
"configurations": [
{
"name": "Linux",
"includePath": [
"${workspaceFolder}/**",
"/usr/include/freetype2"
],
"defines": [],
"compilerPath": "/usr/bin/clang++",
"cStandard": "c11",
"cppStandard": "c++17",
"intelliSenseMode": "clang-x64"
}
],
"version": 4
}

58
.vscode/launch.json vendored
View File

@@ -1,58 +0,0 @@
{
// Use IntelliSense to learn about possible attributes.
// Hover to view descriptions of existing attributes.
// For more information, visit: https://go.microsoft.com/fwlink/?linkid=830387
"version": "0.2.0",
"configurations": [
{
"name": "Launch Profiler GUI",
"type": "lldb",
"request": "launch",
"program": "${workspaceFolder}/profiler/build/unix/Tracy-debug",
"args": [],
"stopAtEntry": false,
"cwd": "${workspaceFolder}",
"environment": [],
"externalConsole": false,
"preLaunchTask": "Build Profiler GUI",
"console": "internalConsole",
"internalConsoleOptions": "neverOpen"
},
{
"name": "Launch capture tool",
"type": "lldb",
"request": "launch",
"program": "${workspaceFolder}/capture/build/unix/capture-debug",
"args": [],
"stopAtEntry": false,
"cwd": "${workspaceFolder}",
"environment": [],
"externalConsole": false,
"preLaunchTask": "Build capture tool"
},
{
"name": "Launch update tool",
"type": "lldb",
"request": "launch",
"program": "${workspaceFolder}/update/build/unix/update-debug",
"args": [],
"stopAtEntry": false,
"cwd": "${workspaceFolder}",
"environment": [],
"externalConsole": false,
"preLaunchTask": "Build update tool"
},
{
"name": "Launch test application",
"type": "lldb",
"request": "launch",
"program": "${workspaceFolder}/test/tracy_test",
"args": [],
"stopAtEntry": false,
"cwd": "${workspaceFolder}/test",
"environment": [],
"externalConsole": false,
"preLaunchTask": "Build test application"
},
]
}

88
.vscode/tasks.json vendored
View File

@@ -1,88 +0,0 @@
{
// See https://go.microsoft.com/fwlink/?LinkId=733558
// for the documentation about the tasks.json format
"version": "2.0.0",
"tasks": [
{
"label": "Build Profiler GUI",
"type": "shell",
"command": "CC=clang CXX=clang++ make debug -C profiler/build/unix -j `nproc`",
"group": {
"kind": "build",
"isDefault": true
},
"presentation": {
"close": true
}
},
{
"label": "Build capture tool",
"type": "shell",
"command": "CC=clang CXX=clang++ make debug -C capture/build/unix -j `nproc`",
"group": {
"kind": "build",
"isDefault": true
},
"presentation": {
"close": true
}
},
{
"label": "Build csvexport tool",
"type": "shell",
"command": "CC=clang CXX=clang++ make debug -C csvexport/build/unix -j `nproc`",
"group": {
"kind": "build",
"isDefault": true
},
"presentation": {
"close": true
}
},
{
"label": "Build import-chrome tool",
"type": "shell",
"command": "CC=clang CXX=clang++ make debug -C import-chrome/build/unix -j `nproc`",
"group": {
"kind": "build",
"isDefault": true
}
},
{
"label": "Build update tool",
"type": "shell",
"command": "CC=clang CXX=clang++ make debug -C update/build/unix -j `nproc`",
"group": {
"kind": "build",
"isDefault": true
},
"presentation": {
"close": true
}
},
{
"label": "Build client library",
"type": "shell",
"command": "CC=clang CXX=clang++ make debug -C library/unix -j `nproc`",
"group": {
"kind": "build",
"isDefault": true
},
"presentation": {
"close": true
}
},
{
"label": "Build test application",
"type": "shell",
"command": "CC=clang CXX=clang++ make OPTFLAGS=-DTRACY_VERBOSE\\ -fno-omit-frame-pointer\\ -march=native\\ -g -C test -j `nproc`",
"group": {
"kind": "build",
"isDefault": true
},
"presentation": {
"close": true
}
}
]
}

26
AUTHORS
View File

@@ -1,20 +1,6 @@
Bartosz Taudul <wolf@nereid.pl>
Kamil Klimek <kamil.klimek@sharkbits.com> (initial find zone implementation)
Bartosz Szreder <zgredder@gmail.com> (view/worker split)
Arvid Gerstmann <dev@arvid-g.de> (compatibility fixes)
Rokas Kupstys <rokups@zoho.com> (compatibility fixes, initial CI work, MingW support)
Till Rathmann <till.rathmann@gmx.de> (DLL support)
Sherief Farouk <sherief.personal@gmail.com> (compatibility fixes)
Dedmen Miller <dedmen@dedmen.de> (find zone bug fixes, improvements)
Michał Cichoń <michcic@gmail.com> (OSX call stack decoding backport)
Thales Sabino <thales@codeplay.com> (OpenCL support)
Andrew Depke <andrewdepke@gmail.com> (Direct3D 12 support)
Simonas Kazlauskas <git@kazlauskas.me> (OSX CI, external bindings)
Jakub Žádník <kubouch@gmail.com> (csvexport utility)
Andrey Voroshilov <andrew.voroshilov@gmail.com> (multi-DLL fixes)
Benoit Jacob <benoitjacob@google.com> (Android improvements)
David Farrel <dafarrel@adobe.com> (Direct3D 11 support)
Terence Rokop <rokopt@sharpears.net> (Non-reentrant zones)
Lukas Berbuer <lukas.berbuer@gmail.com> (CMake integration)
Xavier Bouchoux <xavierb@gmail.com> (sample data in find zone)
Balazs Kovacsics <kovab93@gmail.com> (Universal Windows Platform)
Bartosz Taudul <wolf.pld@gmail.com>
Kamil Klimek <kamil.klimek@sharkbits.com>
Bartosz Szreder <zgredder@gmail.com>
Arvid Gerstmann <dev@arvid-g.de>
Rokas Kupstys <rokups@zoho.com>
Till Rathmann <till.rathmann@gmx.de>

View File

@@ -1,169 +0,0 @@
cmake_minimum_required(VERSION 3.10)
# Run version helper script
include(version.cmake)
set(TRACY_VERSION_MAJOR ${major})
set(TRACY_VERSION_MINOR ${minor})
set(TRACY_VERSION_PATCH ${patch})
set(TRACY_VERSION_STRING "${TRACY_VERSION_MAJOR}.${TRACY_VERSION_MINOR}.${TRACY_VERSION_PATCH}")
project(Tracy LANGUAGES CXX VERSION ${TRACY_VERSION_STRING})
if(${BUILD_SHARED_LIBS})
set(DEFAULT_STATIC OFF)
else()
set(DEFAULT_STATIC ON)
endif()
option(TRACY_STATIC "Whether to build Tracy as a static library" ${DEFAULT_STATIC})
find_package(Threads REQUIRED)
set(TRACY_PUBLIC_DIR ${CMAKE_CURRENT_SOURCE_DIR}/public)
if(TRACY_STATIC)
set(TRACY_VISIBILITY "STATIC")
else()
set(TRACY_VISIBILITY "SHARED")
endif()
add_library(TracyClient ${TRACY_VISIBILITY} "${TRACY_PUBLIC_DIR}/TracyClient.cpp")
target_compile_features(TracyClient PUBLIC cxx_std_11)
target_include_directories(TracyClient SYSTEM PUBLIC
$<BUILD_INTERFACE:${TRACY_PUBLIC_DIR}>
$<INSTALL_INTERFACE:include>)
target_link_libraries(
TracyClient
PUBLIC
Threads::Threads
${CMAKE_DL_LIBS}
)
# Public dependency on some libraries required when using Mingw
if(WIN32 AND ${CMAKE_CXX_COMPILER_ID} MATCHES "GNU")
target_link_libraries(TracyClient PUBLIC ws2_32 dbghelp)
endif()
if(CMAKE_SYSTEM_NAME MATCHES "FreeBSD")
find_library(EXECINFO_LIBRARY NAMES execinfo REQUIRED)
target_link_libraries(TracyClient PUBLIC ${EXECINFO_LIBRARY})
endif()
add_library(Tracy::TracyClient ALIAS TracyClient)
macro(set_option option help value)
option(${option} ${help} ${value})
if(${option})
message(STATUS "${option}: ON")
target_compile_definitions(TracyClient PUBLIC ${option})
else()
message(STATUS "${option}: OFF")
endif()
endmacro()
set_option(TRACY_ENABLE "Enable profiling" ON)
set_option(TRACY_ON_DEMAND "On-demand profiling" OFF)
set_option(TRACY_CALLSTACK "Enforce callstack collection for tracy regions" OFF)
set_option(TRACY_NO_CALLSTACK "Disable all callstack related functionality" OFF)
set_option(TRACY_NO_CALLSTACK_INLINES "Disables the inline functions in callstacks" OFF)
set_option(TRACY_ONLY_LOCALHOST "Only listen on the localhost interface" OFF)
set_option(TRACY_NO_BROADCAST "Disable client discovery by broadcast to local network" OFF)
set_option(TRACY_ONLY_IPV4 "Tracy will only accept connections on IPv4 addresses (disable IPv6)" OFF)
set_option(TRACY_NO_CODE_TRANSFER "Disable collection of source code" OFF)
set_option(TRACY_NO_CONTEXT_SWITCH "Disable capture of context switches" OFF)
set_option(TRACY_NO_EXIT "Client executable does not exit until all profile data is sent to server" OFF)
set_option(TRACY_NO_SAMPLING "Disable call stack sampling" OFF)
set_option(TRACY_NO_VERIFY "Disable zone validation for C API" OFF)
set_option(TRACY_NO_VSYNC_CAPTURE "Disable capture of hardware Vsync events" OFF)
set_option(TRACY_NO_FRAME_IMAGE "Disable the frame image support and its thread" OFF)
set_option(TRACY_NO_SYSTEM_TRACING "Disable systrace sampling" OFF)
set_option(TRACY_PATCHABLE_NOPSLEDS "Enable nopsleds for efficient patching by system-level tools (e.g. rr)" OFF)
set_option(TRACY_DELAYED_INIT "Enable delayed initialization of the library (init on first call)" OFF)
set_option(TRACY_MANUAL_LIFETIME "Enable the manual lifetime management of the profile" OFF)
set_option(TRACY_FIBERS "Enable fibers support" OFF)
set_option(TRACY_NO_CRASH_HANDLER "Disable crash handling" OFF)
set_option(TRACY_TIMER_FALLBACK "Use lower resolution timers" OFF)
set_option(TRACY_LIBUNWIND_BACKTRACE "Use libunwind backtracing where supported" OFF)
set_option(TRACY_SYMBOL_OFFLINE_RESOLVE "Instead of full runtime symbol resolution, only resolve the image path and offset to enable offline symbol resolution" OFF)
set_option(TRACY_LIBBACKTRACE_ELF_DYNLOAD_SUPPORT "Enable libbacktrace to support dynamically loaded elfs in symbol resolution resolution after the first symbol resolve operation" OFF)
if(NOT TRACY_STATIC)
target_compile_definitions(TracyClient PRIVATE TRACY_EXPORTS)
target_compile_definitions(TracyClient PUBLIC TRACY_IMPORTS)
endif()
include(CMakePackageConfigHelpers)
include(GNUInstallDirs)
set_target_properties(TracyClient PROPERTIES VERSION ${PROJECT_VERSION})
set(tracy_includes
${TRACY_PUBLIC_DIR}/tracy/TracyC.h
${TRACY_PUBLIC_DIR}/tracy/Tracy.hpp
${TRACY_PUBLIC_DIR}/tracy/TracyD3D11.hpp
${TRACY_PUBLIC_DIR}/tracy/TracyD3D12.hpp
${TRACY_PUBLIC_DIR}/tracy/TracyLua.hpp
${TRACY_PUBLIC_DIR}/tracy/TracyOpenCL.hpp
${TRACY_PUBLIC_DIR}/tracy/TracyOpenGL.hpp
${TRACY_PUBLIC_DIR}/tracy/TracyVulkan.hpp)
set(client_includes
${TRACY_PUBLIC_DIR}/client/tracy_concurrentqueue.h
${TRACY_PUBLIC_DIR}/client/tracy_rpmalloc.hpp
${TRACY_PUBLIC_DIR}/client/tracy_SPSCQueue.h
${TRACY_PUBLIC_DIR}/client/TracyArmCpuTable.hpp
${TRACY_PUBLIC_DIR}/client/TracyCallstack.h
${TRACY_PUBLIC_DIR}/client/TracyCallstack.hpp
${TRACY_PUBLIC_DIR}/client/TracyCpuid.hpp
${TRACY_PUBLIC_DIR}/client/TracyDebug.hpp
${TRACY_PUBLIC_DIR}/client/TracyDxt1.hpp
${TRACY_PUBLIC_DIR}/client/TracyFastVector.hpp
${TRACY_PUBLIC_DIR}/client/TracyLock.hpp
${TRACY_PUBLIC_DIR}/client/TracyProfiler.hpp
${TRACY_PUBLIC_DIR}/client/TracyRingBuffer.hpp
${TRACY_PUBLIC_DIR}/client/TracyScoped.hpp
${TRACY_PUBLIC_DIR}/client/TracyStringHelpers.hpp
${TRACY_PUBLIC_DIR}/client/TracySysPower.hpp
${TRACY_PUBLIC_DIR}/client/TracySysTime.hpp
${TRACY_PUBLIC_DIR}/client/TracySysTrace.hpp
${TRACY_PUBLIC_DIR}/client/TracyThread.hpp)
set(common_includes
${TRACY_PUBLIC_DIR}/common/tracy_lz4.hpp
${TRACY_PUBLIC_DIR}/common/tracy_lz4hc.hpp
${TRACY_PUBLIC_DIR}/common/TracyAlign.hpp
${TRACY_PUBLIC_DIR}/common/TracyAlloc.hpp
${TRACY_PUBLIC_DIR}/common/TracyApi.h
${TRACY_PUBLIC_DIR}/common/TracyColor.hpp
${TRACY_PUBLIC_DIR}/common/TracyForceInline.hpp
${TRACY_PUBLIC_DIR}/common/TracyMutex.hpp
${TRACY_PUBLIC_DIR}/common/TracyProtocol.hpp
${TRACY_PUBLIC_DIR}/common/TracyQueue.hpp
${TRACY_PUBLIC_DIR}/common/TracySocket.hpp
${TRACY_PUBLIC_DIR}/common/TracyStackFrames.hpp
${TRACY_PUBLIC_DIR}/common/TracySystem.hpp
${TRACY_PUBLIC_DIR}/common/TracyUwp.hpp
${TRACY_PUBLIC_DIR}/common/TracyYield.hpp)
install(TARGETS TracyClient
EXPORT TracyConfig
RUNTIME DESTINATION ${CMAKE_INSTALL_BINDIR}
LIBRARY DESTINATION ${CMAKE_INSTALL_LIBDIR}
ARCHIVE DESTINATION ${CMAKE_INSTALL_LIBDIR})
install(FILES ${tracy_includes}
DESTINATION ${CMAKE_INSTALL_INCLUDEDIR}/tracy)
install(FILES ${client_includes}
DESTINATION ${CMAKE_INSTALL_INCLUDEDIR}/client)
install(FILES ${common_includes}
DESTINATION ${CMAKE_INSTALL_INCLUDEDIR}/common)
install(EXPORT TracyConfig
NAMESPACE Tracy::
FILE TracyTargets.cmake
DESTINATION ${CMAKE_INSTALL_DATAROOTDIR}/Tracy)
include(CMakePackageConfigHelpers)
configure_package_config_file(${CMAKE_CURRENT_SOURCE_DIR}/Config.cmake.in
"${CMAKE_CURRENT_BINARY_DIR}/TracyConfig.cmake"
INSTALL_DESTINATION ${CMAKE_INSTALL_DATAROOTDIR}/Tracy)
install(FILES ${CMAKE_CURRENT_BINARY_DIR}/TracyConfig.cmake
DESTINATION ${CMAKE_INSTALL_DATAROOTDIR}/Tracy)

View File

@@ -1,6 +0,0 @@
@PACKAGE_INIT@
include(CMakeFindDependencyMacro)
find_dependency(Threads REQUIRED)
include("${CMAKE_CURRENT_LIST_DIR}/TracyTargets.cmake")

77
FAQ.md Normal file
View File

@@ -0,0 +1,77 @@
# A quick tracy FAQ
### I already use VTune/perf/Very Sleepy/callgrind/MSVC profiler.
These are statistical profilers, which can be used to find hot spots in the code. This is very useful, but it won't show you the underlying reason for semi-random frame stutter that may occur every couple of seconds.
### You can use Telemetry for that.
Telemetry license costs about 8000 $ per year. Tracy is open source software. Telemetry doesn't have Lua bindings.
### You can use the free Brofiler. Crytek does use it, so it has to be good.
After a cursory look at the Brofiler code I can tell that the timer resolution there is at 300 ns. Tracy can achieve 5 ns timer resolution. Brofiler event logging infrastructure seems to be over-engineered. Brofiler can't track lock contention, nor does it have Lua bindings.
### So tracy is supposedly faster?
My measurements show that logging a single zone with tracy takes only 15 ns. In theory, if the program was doing nothing else, tracy should be able to log 66 million zones per second.
### Bullshit, RAD is advertising that they are able only to log about a million zones, over the network nevertheless: "Capture over a million timing zones per second in real-time!"
Tracy can perform network transfer of 15 million zones per second. Should the client and server be on separate machines, this number will be even higher, but you will need more than a gigabit link to achieve the maximum throughput. [Click here for a video of a max-throughput capture.](https://www.youtube.com/watch?v=DSMIHShKGAc)
### Can I connect to my application at any time and start profiling at this moment?
By default no, all events are registered from the beginning of program execution and are waiting in a queue. There's a separate on-demand mode, enabled by using a `TRACY_ON_DEMAND` macro.
### Am I seeing correctly that the profiler allocates one gigabyte of memory per second?
Only in extreme cases. Normal usage has much lower memory pressure.
### Why do you do magic with the static initialization order? Everyone says that's a bad practice.
It allows tracking construction of static objects and memory allocations performed before main() is entered.
### There's no support for consoles.
Welp. But there's mobile support.
### I do need console support.
The code is open. Write your own, then send a patch.
### I don't believe you can capture a zone in 15 ns. Show me the code!
Following is the annotated assembly code (generated from C++ sources) that's responsible for logging start of the zone:
```
call qword ptr [__imp_GetCurrentThreadId]
mov r14d,eax
mov qword ptr [rsp+0F0h],r14 // save thread id for later use
mov r12d,10h
mov rax,qword ptr gs:[58h] // TLS
mov r15,qword ptr [rax] // queue address
mov rdi,qword ptr [r12+r15] // data address
mov rbp,qword ptr [rdi+20h] // buffer counter
mov rbx,rbp
and ebx,7Fh // 128 item buffer
jne Application::InnerLoop+66h --+
mov rdx,rbp |
mov rcx,rdi |
call enqueue_begin_alloc | // reclaim/alloc next buffer
shl rbx,5 <---------------------+ // buffer items are 32 bytes
add rbx,qword ptr [rdi+40h]
mov byte ptr [rbx],4 // queue item type
rdtscp
mov dword ptr [rbx+19h],ecx // cpu id
shl rdx,20h
or rax,rdx // 64 bit timestamp
mov qword ptr [rbx+1],rax
mov qword ptr [rbx+9],r14 // thread id
lea rax,[__tracy_source_location] // static struct address
mov qword ptr [rbx+11h],rax
lea rax,[rbp+1] // increment buffer counter
mov qword ptr [rdi+20h],rax
```
There's also a second code block, for the end of the zone. It's similar, but a bit smaller, as it can use some of the variables that were retrieved above.

View File

@@ -1,7 +1,7 @@
Tracy Profiler (https://github.com/wolfpld/tracy) is licensed under the
Tracy Profiler (https://bitbucket.org/wolfpld/tracy) is licensed under the
3-clause BSD license.
Copyright (c) 2017-2024, Bartosz Taudul <wolf@nereid.pl>
Copyright (c) 2017-2018, Bartosz Taudul <wolf.pld@gmail.com>
All rights reserved.
Redistribution and use in source and binary forms, with or without

932
NEWS
View File

@@ -2,902 +2,11 @@ Note: There is no guarantee that version mismatched client and server will
be able to talk with each other. Network protocol breakages won't be listed
here.
vx.xx.x (xxxx-xx-xx)
--------------------
Note: Release numbers are nothing more than numbers. There are some
"missing" versions due to trace file changes during development. This is not
a mistake.
- Support for pre-0.9 traces has been dropped.
- Various Meson fixes.
- Proper way of loading Vulkan calibrated timestamps extension.
- Fixed C API support for GPU tracing when on demand mode is enabled.
- Added a way to resynchronize CPU and GPU timestamps.
- Using calibrated contexts should always be preferred.
- Each synchronization event requires a sync of CPU and GPU, which is
something you always want to avoid.
- This is not exposed as an easy-to-use API available through the GPU
wrappers.
- Added TracyIsStarted macro to check if the profiler has been started.
Using this functionality only makes sense in the manual lifetime mode,
and will always return true in any other mode of operation.
- Added basic QNX support.
- Zmmword is now recognized as an assemble size directive.
- Libunwind can be used for call stack capture on Linux if you build with
the TRACY_LIBUNWIND_BACKTRACE define.
- Preloading symbols for all modules on Windows, which is always performed
on program init, and which can be quite slow, may now be omitted through
the TRACY_NO_DBGHELP_INIT_LOAD define. In this mode, symbols will be
loaded as needed.
- Validation of discontinuous frames has been disabled in on-demand mode.
It's quite likely to connect in the middle of a discontinuous frame,
which resulted in frame end event for a frame that hasn't been started.
- Symbols can be now resolved offline on Windows and Linux.
- Enabled with the TRACY_SYMBOL_OFFLINE_RESOLVE define or env variable.
- The update utility has two additional options:
- -r, which enables resolving symbol and patching stack frames in the
trace.
- -p, which you can use to modify the paths used for symbol resolution.
- Some functionality will be missing if this mode is used. For example,
symbol statistics are unavailable.
- Resolving symbol names on Linux will now use image cache to reduce the
number of dladdr() calls.
- Compiling with the TRACY_LIBBACKTRACE_ELF_DYNLOAD_SUPPORT define will
enable support for run-time updating of known elf ranges in libbacktrace
on Linux. Previously, shared objects dlopened() after libbacktrace init
would not be visible during symbol resolution.
- Zone group count in the Find zone window is now explicitly displayed.
- Instrumentation statistics now display in how many threads each source
location has appeared in.
- Added import tool for fuchsia traces.
- https://fuchsia.dev/fuchsia-src/reference/tracing/trace-format
- Added checks for overflow of source locations.
- As a reminder, Tracy only allows to have 64K unique source locations,
split in half between static and dynamic locations.
- Runtime checks are active during capture and will stop a trace that
goes beyond the limit.
- Load-time checks will stop any broken trace file from loading.
- Opening the source code view that has no associated address in code
(i.e., from the list of instrumented zones, or from the find zone
window) will now search the list of symbols for a function name match.
- In many cases this will result in displaying the full disassembly view
where previously you would only see the source code.
- Matching is performed by string comparisons, which in rare cases may
result in showing false data.
- Press ctrl key while opening source view to keep the old behavior.
- If more than one matching symbol is found (e.g., if two classes have
methods with the same name, or if a template is instantiated in multiple
places in code), it is not possible to tell which of the code locations
the source location corresponds to and only the source code will be
displayed.
v0.10.0 (2023-10-16)
--------------------
- Missed frames region of on-demand captures will be now ignored when
calculating trace time span, zone percentages, etc.
- Due to technicalities information about locks, frame statistics in trace
information window and csvexport utility still include the missed frames
time.
- When source location dynamic zone coloring mode is enabled, collapsed
zones will be now gray-colored. Previously such regions falled back to
showing thread colors, which may have been confusing to users.
- Vulkan contexts can now use VK_EXT_host_query_reset extension.
- System power usage is now reported on x86 Linux.
- Program name displayed in broadcast messages can be now changed with the
TracySetProgramName() macro.
- Zone error markers (red regions and error bars) have been removed for
consistency with how all other profiling events are displayed.
- It is now possible to export messages in the csvexport utility.
- Major overhaul of how timeline items are processed in GUI.
- The process of figuring out what needs to be drawn on the timeline has
been heavily parallelized.
- The impact is especially visible with traces containing large amounts
of data. The framerate improvement in such cases can be ~30x.
- Consequently, the profiler GUI will now produce multi-core spikes when
rendering frames. This may have impact on the profiled application's
performance, if both the application and the profiler GUI are running
on the same machine. If this is a problem, you may consider the capture
utility instead, which is not affected by these changes. Alternatively,
you may disable parallelization in the options menu.
- Most of the timeline item logic has been written from scratch, which
may have taken care of some elusive bugs.
- Added global configuration settings dialog. You can find it in the
profiler's about menu (the wrench icon in the welcome dialog).
- List of found zones in the Find zone menu can be filtered by user text.
- Fixed div-by-zero in cvsexport utility when there was only one zone of
a kind.
- Fixed compatibility problems with FreeBSD.
- Added support for dynamically loaded Vulkan symbols.
- Trace description or filename is now displayed on the window title bar.
- The csvexport utility will now export thread id data.
- Improved compatibility with MSVC projects not defining NOMINMAX.
- Improved compatibility with Linux setups targeting musl as libc.
- Thread safety of Vulkan instrumentation has been reviewed.
- D3D11 and D3D12 instrumentation was rewritten.
- Added support for efficient profiling when running under rr, the record-
replaying debugger. This is enabled with TRACY_PATCHABLE_NOPSLEDS define.
- History of viewed symbols is now preserved and you can go back to
previously displayed entries.
v0.9.1 (2023-02-26)
-------------------
- Support for pre-0.8 traces has been dropped.
- Profiled programs will ignore dlclose() calls.
- Added warning when the profiler interface is run with privilege elevation.
Advice is given to instead run the client with admin rights.
- Switched to official ZEN4 uarch data.
- Handle cases when thread name is set, but not through Tracy facilities.
- Allow customization of source location data through the following macros:
- TracyFunction - defaults to __FUNCTION__
- TracyFile - defaults to __FILE__
- TracyLine - defaults to __LINE__
- Tracy on Linux now targets and requires Wayland by default.
- Please don't ask about window decorations on Gnome. Current behavior is
the intended behavior. Gnome does not want windows to have decorations,
and Tracy respects this choice. If you find this problematic, use a
desktop environment that actually listens to its users.
- Pass LEGACY=1 parameter to make, if you want to instead rely on the GLFW
library, like before.
- Other platforms still use GLFW.
- Compare traces menu can now display source code differences between two
traces.
- Assembly listings saved to files have been improved.
- Listings are now annotated with source line information.
- To improve compatibility with external tools comments are now prefixed
with '#' instead of ';'.
- Histogram tooltip will now also show left/right counts.
- Tracy now actively manages timeline vertical scroll offset in order to keep
the thread under the mouse cursor in the same place on screen.
- Removed support for AT&T assembly syntax.
- Tracy will not display notification if the file selector can't be used.
Possible reasons for failure include lack of xdg-desktop-portal.
- Using the TRACY_NO_CRASH_HANDLER define will disable handling of
application crashes by the profiler.
- Tracy will now query jump and call target addresses. This enables discovery
of target function names, even if such function has no samples and is not
present in any call stack.
v0.9.0 (2022-10-26)
-------------------
- Attention! All the header and source files used for integrating Tracy with
applications were moved to the public/ directory. This will break your
integration!
- To fix this, update the source and include directories lists to point to
the new location.
- Tracy include files directly referenced by the client were moved to
tracy/ subdirectory, to facilitate setups which previously had Tracy
checkout parent directory in the include paths list (i.e. when you
included "tracy/Tracy.hpp").
- Previously, if you have included the Tracy checkout directory in your
project include directories list (i.e. you could include "Tracy.hpp"),
this could result in third-party library conflicts, e.g. with ImGui.
Such scenarios are no longer the case.
- Tracy macros now require to be terminated with a semicolon.
- The undocumented ___tracy_demangle() function API has been changed. Please
refer to the source code for further instructions.
- The parameter callback and its registration macro have been extended to
include user data pointer. You will need to update your code accordingly.
- Plots visualization has been improved.
- Each plot now has its own color, which can also be defined by the user.
- The area below the plot is now optionally filled with a color.
- Plots can now also be configured to be staircase instead of smooth. This
new setting is appropriate for many inputs where only discrete values
make sense, e.g. the memory allocation plot.
- The API for TracyPlotConfig() macro has been changed. Please refer to
the manual to see how you can fix this.
- Some text labels in the user interface are now more easy to read.
- The profiler will now instruct the user in the UI on what can be done, if
the send queue is slow to process (typically due to symbol resolution).
- If a client with an incompatible protocol is discovered, Tracy will now
try to show which versions can be used to handle the connection.
- Messages list in zone info window can now show messages exclusive to the
zone, filtering out the messages emitted from child zones.
- Added capture of vertical synchronization timings on Linux.
- The range of frame bar colors in the frames overview on top of the screen
can be now controlled with the "Target FPS" entry box in the options menu.
- The "Draw frame targets" option does not need to be selected.
- Previously the hardcoded FPS target thresholds were: 30, 60, 144 FPS.
- Currently the FPS target threshold is: half of target, target, twice the
target.
- Reworked the way zone names are shortened.
- Previously shortening supported only namespace removal, in a way that
didn't consider function parameters or template arguments.
- Shortening to one-letter namespace chains is no longer available.
- The new shortening rules first perform normalization of the function name.
- The function const qualifier is removed.
- Common return types are removed.
- All function parameters and all template arguments are removed.
- The next steps consist of repeated removal of namespaces, starting with
the most outermost one.
- While the old process was all or nothing, the new implementation by
default will dynamically adjust to the space available, trying to show
the most context possible.
- It is also possible to completely disable shortening, or require that it
is always performed in full.
- Function name normalization is enabled by default, even if there is space
to show full function name. This can be changed in options.
- Previously shortening was only applied to the zone names displayed on the
timeline. Currently this process will also apply to all other places in
the UI where function names are displayed. However, in these cases the
function names will only be normalized.
- Full function names are still available as tooltips, or in fine print if
the normalized name is already displayed in a tooltip.
- This functionality is disabled if zone name shortening is disabled.
- Added context menu for timeline labels. Currently the only option is to hide
the selected thread, plot, etc.
- You can now provide custom source file contents through a profiler callback.
- Exposed Tracy version to client applications (available through the
common/TracyVersion.hpp header file).
- D3D12 instrumentation is now thread-safe.
- Timeline can be now navigated with WASD keys.
- Symbol file paths are now normalized on libbacktrace systems. For example,
instead of "/usr/bin/../lib64/gcc/x86_64-pc-linux-gnu/12.2.0/../../../../
include/c++/12.2.0/bits/std_mutex.h" Tracy will now report such file as
"/usr/include/c++/12.2.0/bits/std_mutex.h".
- The import-chrome utility interprets Instant (`i`/`I`) events where the
`name` field contains the word `frame` as a frame event. The `name` is the
frame set name.
- Frame data won't be displayed if there was no frame instrumentation in the
profiling session.
- Note that some automated functionality (e.g. vertical synchronization
capture) may automatically generate frame data, which will force frames to
be displayed.
- Tracy threads will now be collapsed by default on the timeline.
- Clicking on a local thread in the CPU data view will make the thread visible
and uncollapsed on the timeline.
- Assembly view is now in color.
- The profiler UI will no longer unnecessarily redraw the screen if nothing
was changed. This should have a profound impact on power usage.
- Added microarchitecture data for Zen 4.
- Implemented optional propagation of inline cost down the local call stack.
- This feature may be useful when trying to get a general outlook of the
cost at the top-level function in the symbol.
- It is possible to get nonsense data when this is enabled, for example
total cost exceeding 100%. This is by design.
- Assembly line costs are not affected.
- Available clients now also broadcast their PID.
- Reversed mouse button assignments for jumping to source / assembly line in
symbol view. The left mouse button will now focus the target line.
- Assembly lines tooltip will now display local call stack of inline functions
(within the symbol).
- Right-clicking the source location entry in assembly line will show the
local call stack, along with source code preview of each entry and ability
to navigate to any selected inline function.
- The profiler UI will now indicate that it needs attention if the window is
not focused and something interesting happens. For example when a connection
is established, or when a saved trace finishes loading, etc. How the
attention request is indicated depends on the operating system.
- Clicking on the red microarchitecture icon in the symbol view assembly pane
will switch the selected microarchitecture to one the profiled application
was running on.
- Removed option to display instruction latencies in a graphical form. Latency
data is still available in instruction tooltip.
v0.8.2 (2022-06-28)
-------------------
- Added support for debuginfod debug information services. Note that
since this depends on proper system configuration, vendors providing
the debug information, and network retrieval, it is disabled by
default. To enable, compile the profiled application with the
TRACY_DEBUGINFOD define and link with libdebuginfod.
- When Tracy server-side utilities are build with MSVC, the required
libraries will be now automatically retrieved and built with vcpkg.
- Added microarchitecture data for: Bonnell, Airmont, Goldmont, Goldmont
Plus, Tremont.
- Recognize additional CPUIDs of Zen 3, Alder Lake, Ice Lake
microarchitectures.
- Assembly line width will be now extended, if needed. Previously the line
width was calculated for the initial layout and changing amount of
displayed data (especially listing the read/written registers) didn't
affect this, which may have made some lines partially unreadable.
- Added ability to filter call stacks in memory tab by inactive allocations.
Filtering by inactive allocations helps to pinpoint wasteful allocations
in the program.
- Plot graph will no longer display min/max values interpolated for
animation, but rather true values.
- The CPU topology tree structure was replaced by a CPU schematic showing
the same thing in a more concise way.
v0.8.1 (2022-04-21)
-------------------
- Support for pre-0.7 traces has been dropped.
- Update utility can now scan for source files missing in the trace cache,
if the '-c' parameter is given. Found files will be added to the cache.
- Added high-priority queue for fast queries to bypass slow symbol queries.
- Fixed Android documentation to show how to enable context switch tracing.
- Workaround MSVC 2015 stupidity which prevented compilation as C++11.
- Added support for showing branch cost data for CPUs that don't report
branch retirement events (but do report branch misses).
- The right-click context menu available for jump arrows in the symbol view
window will now additionally display jump context, i.e. jump sources and
jump target source code fragments.
- Added freedesktop.org compliant desktop entry and MIME type definition.
- The call stack column in list of messages will now be only displayed when
at least one message on the list has call stack data.
- File dialogs on Unix will be now native to the desktop environment you are
using. Note that this relies on xdg-desktop-portal and dbus.
v0.8.0 (2022-03-28)
-------------------
- Support for Cygwin has been dropped. It was not working for a very long
time and nobody had complained about it.
- Mingw is deprecated due to lack of interest.
- Added TRACY_NO_CALLSTACK_INLINES macro to disable inline functions
resolution in call stacks on Windows.
- Improved function matching algorithm in compare traces view.
- Added CMake integration.
- Reworked rpmalloc initialization.
- Fixed display of messages with newlines on messages list.
- Excluded some uninteresting wrapper functions from call stacks (for
example SIMD pass-through intrinsics to the compiler built-ins).
- Adjusted coloring of instruction hotness in symbol view.
- Properly handle rare cases when sampling on Linux is momentary not able to
resolve time stamps.
- Added Rocket Lake microarchitectural data.
- Updated CPU identifier lists.
- Implemented GPU timer overflow handling heuristics.
- Assembly instructions are now assigned to inline symbols.
- You can not only see the assembly source file and line, but also the
originating function.
- If symbol view is restricted to a single inline function, all assembly
instructions not in this context will be dimmed out.
- Likewise, the navigation in assembly code will be limited just to the
inline context, if a single function is selected.
- Kernel call stacks will be now properly captured and displayed in the
profiler. Kernel functions are marked with the red color.
- The CPU hardware performance counters can be now sampled on Linux.
- Three inferred statistics are displayed for lines in both source and
assembly code in the symbol view window:
- Instructions executed per cycle.
- Branch miss rate.
- Cache miss rate.
- Instruction cost estimation method is no longer tied to software call
stack sampling.
- The image name filter entry field is now providing a list of available
images.
- Reentrant function calls may be now excluded from calculations in the
statistics view.
- Crash handler is now properly removed during profiler destruction.
- Repeatedly right-clicking on the same source line in the symbol view
window will now cycle through assembly blocks associated with this source
line.
- Vulkan headers must be now explicitly included before including
TracyVulkan.hpp.
- The capture utility may now limit capture time to a specified number of
seconds.
- Fixed message thread assignment in the import-chrome utility.
- Sampling data can be now also found in the find zone menu.
- Instrumentation failures may now display their context, e.g. the zone text
that was to be set.
- A warning is now displayed when sampling data is out-of-order.
- Average value for plots can be now viewed.
- Moved symbol resolution to a separate thread. Profiling will no longer be
stuck when there is a large number of symbols to resolve. This not only
improves user experience, but also prevents buildup of data (and memory
consumption) on the client side.
- Android device name will be now reported.
- Added support for capturing fibers.
- Fibers require additional processing, which has to be enabled by adding
the TRACY_FIBERS define on the client side.
- Client code requires additional instrumentation using the new macros
TracyFiberEnter and TracyFiberLeave (or the corresponding C API
variants).
- Fibers are represented in traces as separate threads, and are
distinguished by green color. Faux context switch regions are used to
indicate when a fiber is being run by the worker thread.
- Continuous frame marks no longer need to be issued from a single thread.
- Context switch call stacks are now captured on Windows and Linux.
- Hovering the context switch wait region will now display wait stack,
which may provide additional insight into why the switch happened.
- Wait stacks inspection can be performed in a new view.
- Stacks can be limited to certain threads and to a selected time range.
- Stacks are presented either as a sorted list, or as a bottom-up and
top-down trees.
- Entry call stacks can be now also viewed as a bottom-up and top-down
trees.
- Updated project build files to MSVC 2022.
- Call stack tooltips now also show the executable image name.
- Playback frames can be now changed by interacting with the frame image
slider using the mouse wheel.
- Signal used to handle crashes on Linux can be now redefined.
- Various DPI scaling improvements.
- User interface can be now scaled in run time.
- Symbol code retrieval now also supports kernel on Windows.
- Added low-level C API interface for GPU zones.
- Symbol child calls can be now listed.
- Replaced "restrict time" in memory window with a proper time range limit.
- Added Alder Lake microarchitectural data.
- Added GPU zone statistics.
- Universal Windows Platform support.
- All call stack related functionality can be now disabled with the
TRACY_NO_CALLSTACK macro.
- Added ability to add full-view annotations from the annotations list
window.
v0.7.8 (2021-05-19)
-------------------
- Updated Zen 3 and added Tiger Lake microarchitectural data.
- Manually disconnecting from the server will no longer display erroneous
warning message.
- Added ability to display sample time spent in child function calls.
- Fixed issue which may have prevented sampling on ARM64.
- Added TRACY_NO_FRAME_IMAGE macro to disable frame image compression
thread.
- Ctrl and shift keys will now modify mouse wheel zoom speed.
- Improved user experience in the symbol view window.
- Added support for Direct3D 11 instrumentation.
- Vulkan contexts can be now calibrated on Linux.
- Support loading zstd-compressed chrome traces.
- Chrome traces with multiple PID entries (and possibly conflicting TIDs)
can be now imported.
- Added support for custom source location tag ("loc") in chrome traces.
- Sampling frequency can be now controlled using TRACY_SAMPLING_HZ macro.
- Trace compression can be now selected when saving a trace.
- If a trace cannot be saved, a failure dialog will be displayed.
- Run-time memory usage of frame images can be reduced by calculating
a compression dictionary. This can be only performed when a trace is saved
or through the update utility.
v0.7.7 (2021-04-01)
-------------------
- Linux crash handler will now also catch SIGABRT.
- Fixed invalid name assignment to source files discovered client-side.
- Added ability to check if a zone is active (which may be used to avoid
preparing zone text, etc., as it wouldn't be used anyway).
- Improved sorting behavior of internal vectors.
- Some data will now be always properly displayed during live capture.
This was not particularly visible before, as it mainly concerns edge
cases.
- Sorting is performed only as needed.
- In case of plots the performance during live capture may be decreased,
as these were sorted with at least 0.25 second intervals before. Now
the sorting is performed every frame.
- Some other data, which previously was not sorted, is sorted now.
- In headless capture mode sorting will be only performed when the trace
is saved to disk.
- Fixed some typos in macros.
- Fixed handling of non-ANSI file names on Windows. You can now name your
traces 'ęśąćż.tracy' and it should work as intended. This is supported on
Windows 10 release 1903 and newer.
- Fixed sending GPU context name in on-demand mode.
- Fixed color channel order in ZoneColor() macro.
- Handle failure state when a memory pointer allocation is reported twice,
without an intermediate free.
- Renamed "call stack parents" to "entry call stacks".
- Display number of entry call stacks in assembly line sample count tooltip.
- Added tooltips with preview of source code in various places in the UI.
v0.7.6 (2021-02-06)
-------------------
- Various fixes in build scripts.
- Fixed a faulty rpmalloc initialization path when the first thing the
thread did was sending a message with call stack.
- Added fallback timer define for various virtualized environments, which
may not be able to access the hardware timer registers. This will result
in usage of timer provided by the standard library, with reduced
resolution.
- Further OpenCL improvements.
- Updated libbacktrace.
- Adds Mach-O 64-bit FAT support.
- Fixes memory corruption when processing Mach-O data.
- Fixes missing matching entries during binary search.
- Adds support for MiniDebugInfo.
- Adds fallback to ELF symbol table if no debug info is available.
- Various other fixes.
- Store build time of profiled program in captures.
- GPU contexts can be now named.
- Implemented client -> server source code transfer.
v0.7.5 (2021-01-23)
-------------------
- More robust handling of system tracing on Android.
- Added warning dialog when the connection is lost before all needed data
can be retrieved.
- Fixed handling of NaN plot entries (by skipping them).
- Dynamic zone colors are now supported through the ZoneColor() macro.
- Fixed Arm machine code printout to match the one printed by objdump.
- Fixed client memory corruption when using colored messages.
- Switched to the next-gen ImGui table UI.
- Table columns can have their order rearranged, can be hidden, can be
sorted both in ascending and descending order (where appropriate).
- Table columns state is now preserved between runs.
- Various fixes related to restricting listening to localhost.
- Improved compatibility of ETW tracing with non-MSVC compilers.
- Fixed Vulkan call stack transfer.
- Added support for transient GPU zones (OpenGL, Vulkan, Direct3D 12).
- OpenCL fixes for assert-less builds and non-active zones.
- Added support for thread names and title bar description in traces
imported from chrome tracing format.
v0.7.4 (2020-11-15)
-------------------
- Added support for user-provided locks to keep dbghelp calls thread-safe.
- Call stacks can be now copied to clipboard.
- Allow more control over which automated captures are performed.
- Added textual descriptions for some assembly instructions.
- Profiler memory usage is now also displayed as a percentage of available
physical memory.
- Microarchitecture mismatch is now clearly displayed in the source view
window.
- Added Zen 3 and Cascade Lake microarchitectural data.
- Ghost zones are now supporting all zone coloring modes and namespace
shortening.
- Extend C API to support memory pools.
- Frame rate targets can be now visually represented on the timeline view.
v0.7.3 (2020-10-06)
-------------------
- Properly support DPI scaling on Linux (requires GLFW 3.3).
- Added early checks for output file validity in the capture utility.
- Improvements to presence broadcast handling.
- Custom zone colors can be optionally ignored.
- Added support for tracking multiple memory pools.
- Memory free failure dialog can now show call stack pointing to the failure
location.
- Added support for Wayland on Linux.
- If during the first 5 seconds of the trace there are no frames being
reported, the profiler will switch to following last 5 seconds of the
trace, instead of displaying three last frames.
v0.7.2 (2020-09-14)
-------------------
- Note: the bitbucket repository is obsolete and will soon stop receiving
updates. Migrate to https://github.com/wolfpld/tracy, if you haven't
already.
- The "waiting for connection" dialog no longer has "cancel" button. To
abort connection attempt just use the "close window" button.
- Added update notification.
- The most recent traced events can be now viewed regardless of timeline
zoom level.
- Fixed going-to-line in source view (again).
- Crash handling on client is now not performed, if there is no active
connection.
- Added ability to listen only on IPv4 interfaces.
v0.7.1 (2020-08-24)
-------------------
- Dropped support for pre-v0.6 traces.
- Fixed regression on non-AVX2 CPUs.
- Fixed incorrect calculation of some ghost zones.
- Added list of cached source files.
- Added import of plot data.
- Secure versions of alloc/free macros.
- Automated tracing of vertical synchronization on Windows.
- Fixed attachment of postponed frame images.
- Source location data can be now copied to clipboard from zone info window.
- Zones in find zones menu can be now grouped by zone name.
- Vulkan and D3D12 GPU contexts can be now calibrated.
- Added CSV export utility.
- "Go to frame" popup no longer has a dedicated button. To show it, click on
the frame counter.
- Added macro for checking if profiler is connected.
- Implemented optional data removal from traces in the update utility.
- Allow manual management of profiler lifetime.
- Adjusted priority of ETW threads to time critical.
- Annotations can be now freely adjusted on the timeline.
- Limiting time range for find zone functionality has been significantly
improved.
- Added time range limits for statistics and symbol view.
- Implemented call stack sampling on Linux (including Android).
- Exact time from start of profiling session can be now viewed by hovering
the mouse over the time scale.
- Code transfer can be now compiled-out.
- Added support for zone markup in unloadable modules.
- Added image name filter to sampling statistics results window.
v0.7 (2020-06-11)
-----------------
This is the last release which will be able to load pre-v0.6 traces. Use the
update utility to convert your old traces now!
- chrome:tracing importer now imports zone metadata from "args" key.
- Added display of statistical mode to find zone menu.
- Automatic stack sampling is now available on windows.
- Properly handle tracing on long-running systems.
- Message list entries can now show associated frame image.
- Call stack window will now display module names.
- Symbol location in call stack window may now also display symbol address.
- Statistics menu can now be used to display call stack sampling data or
list available symbols.
- All call paths leading to the sampled instruction in a call stack can be
now displayed.
- Frame image compression ratio (lossless in-memory compression, not taking
into account DXT compression) is displayed in playback window.
- Allow reconnection straight from the discard data dialog.
- Added ability to set custom names for locks.
- Improved handling of network ports.
- Added time percentage display to instrumentation statistics.
- Display of ghost zones (generated from automated call stack sampling).
- Notify when empty labels display is enabled.
- Small fragments of executable code will be now sent from client to server.
- Added notification about query backlog.
- Fixed performance problem with query backlog.
- Display number of in-flight queries, in addition to query backlog.
- Improved failure reports.
- The capture utility will connect to localhost by default.
- Added optional support for QPC timer on windows.
- Complete rewrite of source file viewer. It is now 100% reliable when going
to a source location.
- Symbol source view was added.
- Extension of source file viewer.
- Can display source file, assembly view, or both at the same time.
- May include display of statistical profiling data.
- Ability to switch between source files which were used to build the
symbol.
- Ability to switch between inlined functions which are incorporated into
the symbol.
- Graphical representation of control flow in program.
- Display of micro-architectural data for each assembly instruction.
- Tracking register dependencies between assembly instructions.
- Disassembly may be saved to a file, in order to be processed by external
tools.
- If the default listening port is occupied, profiler will now try listening
on other ports.
- Added possibility to perform source file names substitution.
- Profiler windows can be now docked.
- CPU usage tooltip now displays a list of running threads.
- Added possibility to filter discovered clients list.
- Source files are now cached during capture.
- Profiler will now display a popup when application crashes.
- Added ability to send simple integral values as extra payload for zones.
- Per-frame zone times on the frames plot can now display self time.
- Ability to bind only on localhost interface.
- OpenCL profiling.
- Direct3D 12 profiling.
v0.6.3 (2020-02-13)
-------------------
- Fixed performance issues with loading saved traces on Ryzen CPUs.
- Profiler window contents are now properly updated during window resize.
- Improved tid to pid mapping on windows.
- Zero length and unfinished zones are no longer taken into account for
statistics.
- Build files for shared library are now available (experimental).
- GPU zones now also have "active" parameter.
- Further reduction of memory usage and on-disk trace size.
- Replaced ska::flat_hash_map with robin-hood-hashing.
- Speed-up rendering of long lists of items.
- Exact event time is displayed in some places in the UI.
- Memory allocation lists can now be sorted.
- Added display of trace file compression ratio.
- Optional Zstd compression of trace files.
- Frame images are now internally compressed using Zstd (instead of LZ4).
- Fix display of continuous frame set tooltips.
v0.6.2 (2019-12-30)
-------------------
- Improved call stack decoding on OSX.
- Collection of CPU topology data.
- C API now supports allocated source locations.
- Added chrome:tracing importer.
- Allow merging of ZoneText() strings.
- Time distribution can now show both exclusive and inclusive times.
- Display proper value of selection time in find zone menu.
- Implemented limiting find zone search to a specified time range.
- Highlight hovered zone from find zone menu zone list on the histogram.
- Allow copying user data directory location to the clipboard.
v0.6.1 (2019-11-28)
-------------------
- Dropped support for pre-v0.5 traces.
- Improve BSD support.
- GPU zone CPU thread highlight will now highlight whole thread, not only
the thread name.
- Added CPU thread highlight for CPU data items.
- Client parameters may be now set from the server.
- Minor UI fixes.
v0.6 (2019-11-17)
-----------------
This is the last release which will be able to load pre-v0.5 traces. Use the
update utility to convert your old traces now!
- Dropped support for pre-v0.4 traces.
- Major memory usage decrease.
- Significant network bandwidth decrease.
- Implemented context switch capture on selected platforms.
- Zone timings in various UI places can now take into account only the
time when the thread was executing.
- Zone information window can now display regions in which thread was
suspended by the operating system.
- CPUs on which the zone was running are enumerated.
- Thread activity regions can be graphed on the timeline.
- API breakage: SetThreadName() now only works on current thread.
- Fixed thread name retrieval after thread is destroyed.
- Added number of CPU cores to host info.
- Limited number of possible source locations to 64K.
- Limited supported capture length to 1.6 days.
- CPU cores are now displayed on the timeline.
- Thread execution workload is displayed, including threads from external
programs.
- Thread migrations across CPU cores can be graphed.
- System-wide workload distribution is now plotted on the timeline.
- Added "CPU data" window showing programs competing for CPU during the
capture.
- Switched to using native thread identifiers (relatively small numbers), as
opposed to pthreads identifiers, which in reality were pointers.
- Improved thread name discovery if context switch capture is enabled.
- Per-trace state is now preserved between profiling sessions:
- Timeline view position.
- Item categories draw/hide settings.
- Timeline zones will be highlighted using a different color, when a
matching time range is selected on histogram.
- Per-frame zone times are now displayed on the frames plot when a zone is
selected in the find zone menu.
- Zone color is now displayed in zone information window.
- Zone colors can now be determined basing on depth and thread or source
location.
- Thread colors are displayed across the profiler application.
- Frame times can be now compared.
- Expose more lock handling functionality.
- Network port can be now specified by the user.
- Proper handling of multithreaded Vulkan code.
- Added extreme compression level in update utility.
- Added time distribution data in the zone information window.
- Trace file name is now displayed in trace information window.
- Annotations can be now added to the timeline.
- Server now performs network data retrieval and decompression on a dedicated
thread.
- Added examples of Tracy integration.
- Allow grouping of zones in the find zone menu by zone parent or with no
grouping.
- Zone list in the statistics window can be now filtered.
- Implemented configuration of plots.
- Messages can now collect call stacks.
v0.5 (2019-08-10)
-----------------
This is the last release which will be able to load pre-v0.4 traces. Use the
update utility to convert your old traces now!
- Major decrease of trace dump file size.
- Major optimizations across the board.
- Vcpkg is now used for library management on Windows.
- Display dump file size change in the update utility.
- Added notification area.
- Display trace loading time.
- Display background processing tasks after trace is loaded.
- Display trace save notification.
- Show crash icon, if there was a crash.
- Added C API.
- Profiling session may now gracefully terminate, due to incorrect
instrumentation. A popup with termination reason will be displayed.
- Call stack improvements.
- Call stack frames now have a proper source file and file line
information on Linux.
- Single call stack frame may now have multiple entries, representing
inlined function calls.
- Call stack grouping in the find zone menu now has a special display
mode.
- Call stack memory allocations tree improvements:
- Add top-down variant to complement the previously available bottom-up
one.
- Add ability to group tree nodes by function name.
- Allow restricting tree to display only active allocations.
- Added support for Lua call stack capture.
- Self time of zones may be now displayed in the find zone menu.
- Added ability to disconnect from a client.
- Find zone groups can now be sorted by mean time per call.
- Zones displayed in the find zone menu can be now grouped by order of
appearance, execution time or name.
- Time is now displayed without trailing fractional zeros (e.g. "2.5 ms"
instead of "2.50 ms").
- Child zones displayed in zone info window can be now grouped by source
location.
- Selected or hovered lock is now highlighted on the timeline.
- Locks are now grouped into single and multithreaded (contended and
uncontended) in the options menu locks list.
- On broken platforms the profiler can now be initialized as needed (and
possible), taking a performance and functionality hit.
- User experience improvements in the graphical profiler.
- Thread position and height is now animated, to eliminate flickering that
was happening when depth of displayed zones was changing.
- Zooming in/out using the mouse wheel is now animated.
- Plot range adjustment is now animated.
- Various other UI improvements.
- System CPU usage is now being monitored.
- Threads that have nothing to display in the current view are now hidden by
default.
- Dimmed-out the timeline outside the profiling area.
- Source file view can now be opened also from statistics menu.
- Display standard deviation in find zone and compare traces menus.
- Display zone messages in zone information window.
- Display order of threads can be changed in the options menu.
- Prevent deadlocks by querying socket send buffer size.
- Frame set statistics can be now limited to frames visible on the screen.
- Messages can be now colored.
- Zone selection in compare traces menu can be now linked to the other
trace.
- Added support for frame image (screen shot) storage.
- Implemented ability to cut off outliers on histograms.
- Zone or frame that is currently hovered by the mouse cursor will be
highlighted on the histogram.
- Server now displays available clients in the local network.
- Source code whitespace visibility can now be enabled or disabled.
- Profiler will now check if proper timer readings can be performed on
x86/x64.
- Application can now log app-specific information, similarly to how the
host info reports system information.
- Message list will automatically scroll down to the most recent message.
- Feature will disable when the list is scrolled by user.
- To re-enable, scroll to the bottom of the list.
- Message list can be now filtered.
- A notification popup will be displayed during trace cleanup.
- Source file view won't be available if a source file is newer than the
capture.
- Added ability to set custom trace descriptions.
- Added frame time target lines.
- FPS counts are now displayed next to frame times.
- GPU drift value can be now automatically measured.
- Connection window is now a popup hidden under a dedicated button.
v0.4.1 (2018-12-30)
-------------------
- Active frame set can be now switched by clicking on a frame set on the
timeline.
- Add ability to go to a specified frame.
- Most commonly used addresses can be now selected from the drop-down menu.
- Fixed corner case problem with profiler initialization on Windows.
- Added third state (stopped) to the pause/resume button. It will be used
after the connection to the client is terminated.
- Active trace can be discarded.
- Call stack capture may be forced through TRACY_CALLSTACK define.
- Lock info window has been added.
- Time of lock creation and termination is now being tracked.
- Menu bar buttons are now toggles that can also close their corresponding
windows.
- Find zone and compare menu improvements.
- Ability to ignore case during search.
- Pressing enter key will now start search, just like pressing the "find"
button.
- Using the ^F keyboard shortcut will open the find zone menu and focus
the input box.
- Added ability to automatically connect to an IP address in the graphical
profiler application (use "-a address" argument to enable).
- Pressing enter key after entering client address in the welcome dialog
will now automatically begin connection process.
v0.4 (2018-10-09)
v0.4 (xxxx-xx-xx)
-----------------
- Renamed "standalone" utility to "profiler".
@@ -922,19 +31,14 @@ v0.4 (2018-10-09)
environmental variable to 1 to get the same effect.
- Added ability to show/hide all threads and plots.
- Performance improvements.
- Improvements to memory data presentation.
- Added memory allocation info window.
- Selecting memory allocation on a plot will draw time range of the
allocation.
- Middle clicking on an memory allocation address (or on a button in
memory allocation info window) will zoom the view to the allocation
range.
- Added memory allocation info window.
- Selecting memory allocation on a plot will draw time range of the
allocation.
- Find zone menu improvements:
- Zones can be now also grouped by call stacks.
- Zone groups can be now also sorted by time spend in each zone.
- Zone groups list now displays group times.
- Average and median zone times are now displayed on the histogram.
- Selected zones will be highlighted on the timeline view.
- Added named versions of tracing macros that allow specifying scoped
variable name.
- The main profiler window is now kept at the bottom of windows stack.
@@ -953,15 +57,12 @@ v0.4 (2018-10-09)
- Added support for cross-DLL profiling.
- Items in options menu (locks, threads, etc.) are now described with number
of events.
- Source location of lock declaration is also provided.
- Created an extensive user manual for the profiler.
- Added ability to capture multiple frame sets.
- Viewer will display multiple frame ranges at once.
- Only one frame set can be active at once. The selected one is used for
the frame navigation graph, frame navigation buttons and drawing frame
separators.
- The active frame set will be highlighted, and the rest will be dimmed
out.
- Frames can now also be discontinuous.
- Frames and zones too small to be displayed will be marked with a zig-zag
pattern.
@@ -975,7 +76,6 @@ v0.4 (2018-10-09)
- This includes frame time statistics and histogram.
- Displayed memory sizes are now properly formatted.
- Added call stack tree for memory allocations.
- You can display allocations list for each call stack tree entry.
- The source code of the profiled application may now be viewed in the
profiler.
- BIG FAT WARNING: The actual profiled program source code is not known to
@@ -994,11 +94,8 @@ v0.4 (2018-10-09)
- Fonts should now be less blurry.
- "Go to parent" button in zone info window won't be displayed if there is
no parent to go to.
- Improvements to the compare traces menu.
- There are now colored markers to make it easier to distinguish "this" and
"external" traces.
- The amount of saved time is now displayed (a difference between total
run times of both traces).
- Compare menu now has colored markers to make it easier to distinguish
"this" and "external" traces.
- Tracy will now collect host information, like CPU name, amount of system
memory, etc.
- Windows builds of the "profiler" utility will perform a check of supported
@@ -1019,17 +116,10 @@ v0.4 (2018-10-09)
and restored.
- The size and position of internal windows now doesn't depend on the
runtime directory of the profiler executable.
- Added connection handshake.
- Server won't be able to connect to client if there's a protocol version
mismatch.
- Client not in on-demand mode will refuse connections after the first
connection was made and the initial event buffers were cleared.
- A single server will no longer try to connect to multiple clients.
- The capture utility will now display time span of the ongoing capture.
v0.3 (2018-07-03)
-----------------
v0.3.3 (2018-07-03)
-------------------
- Breaking change: the format of trace files has changed.
- Previous tracy version will crash when trying to open new traces.

View File

@@ -1,26 +1,68 @@
# Tracy Profiler
[![Sponsor](.github/sponsor.png)](https://github.com/sponsors/wolfpld/)
[![Build status](https://ci.appveyor.com/api/projects/status/968a88arq06gm3el/branch/master?svg=true)](https://ci.appveyor.com/project/wolfpld/tracy/branch/master)
### A real time, nanosecond resolution, remote telemetry, hybrid frame and sampling profiler for games and other applications.
Tracy supports profiling CPU (Direct support is provided for C, C++, and Lua integration. At the same time, third-party bindings to many other languages exist on the internet, such as [Rust](https://github.com/nagisa/rust_tracy_client), [Zig](https://github.com/nektro/zig-tracy), [C#](https://github.com/clibequilibrium/Tracy-CSharp), [OCaml](https://github.com/imandra-ai/ocaml-tracy), [Odin](https://github.com/oskarnp/odin-tracy), etc.), GPU (All major graphic APIs: OpenGL, Vulkan, Direct3D 11/12, OpenCL.), memory allocations, locks, context switches, automatically attribute screenshots to captured frames, and much more.
- [Documentation](https://github.com/wolfpld/tracy/releases/latest/download/tracy.pdf) for usage and build process instructions
- [Releases](https://github.com/wolfpld/tracy/releases) containing the documentation (`tracy.pdf`) and compiled Windows x64 binaries (`Tracy-<version>.7z`) as assets
- [Changelog](NEWS)
- [Interactive demo](https://tracy.nereid.pl/)
Tracy is a real time, nanosecond resolution frame profiler that can be used for remote or embedded telemetry of your application. It can profile CPU (C++, Lua), GPU (OpenGL, Vulkan) and memory. It also can display locks held by threads and their interactions with each other.
![](doc/profiler.png)
![](doc/profiler2.png)
Tracy requires compiler support for C++11, Thread Local Storage and a way to workaround static initialization order fiasco. There are no other requirements. The following platforms are confirmed to be working (this is not a complete list):
![](doc/profiler3.png)
- Windows (x86, x64)
- Linux (x86, x64, ARM, ARM64)
- Android (ARM, x86)
- FreeBSD (x64)
- Cygwin (x64)
- WSL (x64)
- OSX (x64)
The following compilers are supported:
- MSVC
- gcc
- clang
[Introduction to Tracy Profiler v0.2](https://www.youtube.com/watch?v=fB5B46lbapc)
[New features in Tracy Profiler v0.3](https://www.youtube.com/watch?v=3SXpDpDh2Uo)
[New features in Tracy Profiler v0.4](https://www.youtube.com/watch?v=eAkgkaO8B9o)
[New features in Tracy Profiler v0.5](https://www.youtube.com/watch?v=P6E7qLMmzTQ)
[New features in Tracy Profiler v0.6](https://www.youtube.com/watch?v=uJkrFgriuOo)
[New features in Tracy Profiler v0.7](https://www.youtube.com/watch?v=_hU7vw00MZ4)
[New features in Tracy Profiler v0.8](https://www.youtube.com/watch?v=30wpRpHTTag)
[New features in Tracy Profiler v0.3](https://www.youtube.com/watch?v=3SXpDpDh2Uo)
[A quick FAQ.](FAQ.md)
[List of changes.](NEWS)
### High-level overview
![](doc/design.svg)
Tracy is split into client and server side. The client side collects events using a high-efficiency queue and awaits for an incoming connection. The server part connects to client and receives collected data from the client, which is then reconstructed into a viewable timeline. The transfer is performed using a TCP connection.
### Performance impact
To check how much slowdown is introduced by using Tracy, I have profiled [etcpak](https://bitbucket.org/wolfpld/etcpak), which is the fastest ETC texture compression utility there is. I used an 8192×8192 test image as input data and instrumented everything down to the 4×4 pixel block compression function (that's 4 million blocks to compress). It should be noted that Tracy needs to calibrate its internal timers at each run. This introduces a delay of 115 ms (on my machine), which is negligible when doing lengthy profiling runs, but it skews the results of etcpak timing. The following times have this delay subtracted, to give focus on zone collection impact, which is the thing that really matters here.
| Scenario | Zones | Clean run | Profiling run | Difference |
|-------------------------------------------------------|---------|-----------|---------------|------------|
| Compression of an image to ETC1 format | 4194568 | 0.94 s | 1.003 s | +0.063 s |
| Compression of an image to ETC2 format, with mip-maps | 5592822 | 1.034 s | 1.119 s | +0.085 s |
In both scenarios the per-zone time cost is at ~15 ns. This is in line with the measured 8 ns single event collection time (each zone has to report start and end event).
## Usage instructions
The user manual for Tracy is available [at the following address](https://bitbucket.org/wolfpld/tracy/downloads/tracy.pdf). It provides information about the integration process, required code markup and so on.
## Features
#### Marking locks
![](doc/locks.png)
#### Plotting data
![](doc/plot.png)
#### Message log
![](doc/messages.png)
#### Approximation of capture cost
![](doc/cost.png)

126
Tracy.hpp Normal file
View File

@@ -0,0 +1,126 @@
#ifndef __TRACY_HPP__
#define __TRACY_HPP__
#include "common/TracyColor.hpp"
#include "common/TracySystem.hpp"
#ifndef TRACY_ENABLE
#define ZoneNamed(x,y)
#define ZoneNamedN(x,y,z)
#define ZoneNamedC(x,y,z)
#define ZoneNamedNC(x,y,z,w)
#define ZoneScoped
#define ZoneScopedN(x)
#define ZoneScopedC(x)
#define ZoneScopedNC(x,y)
#define ZoneText(x,y)
#define ZoneName(x,y)
#define FrameMark
#define FrameMarkNamed(x)
#define FrameMarkStart(x)
#define FrameMarkEnd(x)
#define TracyLockable( type, varname ) type varname;
#define TracyLockableN( type, varname, desc ) type varname;
#define TracySharedLockable( type, varname ) type varname;
#define TracySharedLockableN( type, varname, desc ) type varname;
#define LockableBase( type ) type
#define SharedLockableBase( type ) type
#define LockMark(x) (void)x;
#define TracyPlot(x,y)
#define TracyMessage(x,y)
#define TracyMessageL(x)
#define TracyAlloc(x,y)
#define TracyFree(x)
#define ZoneNamedS(x,y,z)
#define ZoneNamedNS(x,y,z,w)
#define ZoneNamedCS(x,y,z,w)
#define ZoneNamedNCS(x,y,z,w,a)
#define ZoneScopedS(x)
#define ZoneScopedNS(x,y)
#define ZoneScopedCS(x,y)
#define ZoneScopedNCS(x,y,z)
#define TracyAllocS(x,y,z)
#define TracyFreeS(x,y)
#else
#include "client/TracyLock.hpp"
#include "client/TracyProfiler.hpp"
#include "client/TracyScoped.hpp"
#define ZoneNamed( varname, active ) static const tracy::SourceLocationData TracyConcat(__tracy_source_location,__LINE__) { nullptr, __FUNCTION__, __FILE__, (uint32_t)__LINE__, 0 }; tracy::ScopedZone varname( &TracyConcat(__tracy_source_location,__LINE__), active );
#define ZoneNamedN( varname, name, active ) static const tracy::SourceLocationData TracyConcat(__tracy_source_location,__LINE__) { name, __FUNCTION__, __FILE__, (uint32_t)__LINE__, 0 }; tracy::ScopedZone varname( &TracyConcat(__tracy_source_location,__LINE__), active );
#define ZoneNamedC( varname, color, active ) static const tracy::SourceLocationData TracyConcat(__tracy_source_location,__LINE__) { nullptr, __FUNCTION__, __FILE__, (uint32_t)__LINE__, color }; tracy::ScopedZone varname( &TracyConcat(__tracy_source_location,__LINE__), active );
#define ZoneNamedNC( varname, name, color, active ) static const tracy::SourceLocationData TracyConcat(__tracy_source_location,__LINE__) { name, __FUNCTION__, __FILE__, (uint32_t)__LINE__, color }; tracy::ScopedZone varname( &TracyConcat(__tracy_source_location,__LINE__), active );
#define ZoneScoped ZoneNamed( ___tracy_scoped_zone, true )
#define ZoneScopedN( name ) ZoneNamedN( ___tracy_scoped_zone, name, true )
#define ZoneScopedC( color ) ZoneNamedC( ___tracy_scoped_zone, color, true )
#define ZoneScopedNC( name, color ) ZoneNamedNC( ___tracy_scoped_zone, name, color, true )
#define ZoneText( txt, size ) ___tracy_scoped_zone.Text( txt, size );
#define ZoneName( txt, size ) ___tracy_scoped_zone.Name( txt, size );
#define FrameMark tracy::Profiler::SendFrameMark();
#define FrameMarkNamed( name ) tracy::Profiler::SendFrameMark( name, tracy::QueueType::FrameMarkMsg );
#define FrameMarkStart( name ) tracy::Profiler::SendFrameMark( name, tracy::QueueType::FrameMarkMsgStart );
#define FrameMarkEnd( name ) tracy::Profiler::SendFrameMark( name, tracy::QueueType::FrameMarkMsgEnd );
#define TracyLockable( type, varname ) tracy::Lockable<type> varname { [] () -> const tracy::SourceLocationData* { static const tracy::SourceLocationData srcloc { nullptr, #type " " #varname, __FILE__, __LINE__, 0 }; return &srcloc; }() };
#define TracyLockableN( type, varname, desc ) tracy::Lockable<type> varname { [] () -> const tracy::SourceLocationData* { static const tracy::SourceLocationData srcloc { nullptr, desc, __FILE__, __LINE__, 0 }; return &srcloc; }() };
#define TracySharedLockable( type, varname ) tracy::SharedLockable<type> varname { [] () -> const tracy::SourceLocationData* { static const tracy::SourceLocationData srcloc { nullptr, #type " " #varname, __FILE__, __LINE__, 0 }; return &srcloc; }() };
#define TracySharedLockableN( type, varname, desc ) tracy::SharedLockable<type> varname { [] () -> const tracy::SourceLocationData* { static const tracy::SourceLocationData srcloc { nullptr, desc, __FILE__, __LINE__, 0 }; return &srcloc; }() };
#define LockableBase( type ) tracy::Lockable<type>
#define SharedLockableBase( type ) tracy::SharedLockable<type>
#define LockMark( varname ) static const tracy::SourceLocationData __tracy_lock_location_##varname { nullptr, __FUNCTION__, __FILE__, (uint32_t)__LINE__, 0 }; varname.Mark( &__tracy_lock_location_##varname );
#define TracyPlot( name, val ) tracy::Profiler::PlotData( name, val );
#define TracyMessage( txt, size ) tracy::Profiler::Message( txt, size );
#define TracyMessageL( txt ) tracy::Profiler::Message( txt );
#define TracyAlloc( ptr, size ) tracy::Profiler::MemAlloc( ptr, size );
#define TracyFree( ptr ) tracy::Profiler::MemFree( ptr );
#ifdef TRACY_HAS_CALLSTACK
# define ZoneNamedS( varname, depth, active ) static const tracy::SourceLocationData TracyConcat(__tracy_source_location,__LINE__) { nullptr, __FUNCTION__, __FILE__, (uint32_t)__LINE__, 0 }; tracy::ScopedZone varname( &TracyConcat(__tracy_source_location,__LINE__), depth, active );
# define ZoneNamedNS( varname, name, depth, active ) static const tracy::SourceLocationData TracyConcat(__tracy_source_location,__LINE__) { name, __FUNCTION__, __FILE__, (uint32_t)__LINE__, 0 }; tracy::ScopedZone varname( &TracyConcat(__tracy_source_location,__LINE__), depth, active );
# define ZoneNamedCS( varname, color, depth, active ) static const tracy::SourceLocationData TracyConcat(__tracy_source_location,__LINE__) { nullptr, __FUNCTION__, __FILE__, (uint32_t)__LINE__, color }; tracy::ScopedZone varname( &TracyConcat(__tracy_source_location,__LINE__), depth, active );
# define ZoneNamedNCS( varname, name, color, depth, active ) static const tracy::SourceLocationData TracyConcat(__tracy_source_location,__LINE__) { name, __FUNCTION__, __FILE__, (uint32_t)__LINE__, color }; tracy::ScopedZone varname( &TracyConcat(__tracy_source_location,__LINE__), depth, active );
# define ZoneScopedS( depth ) ZoneNamedS( ___tracy_scoped_zone, depth, true )
# define ZoneScopedNS( name, depth ) ZoneNamedNS( ___tracy_scoped_zone, name, depth, true )
# define ZoneScopedCS( color, depth ) ZoneNamedCS( ___tracy_scoped_zone, color, depth, true )
# define ZoneScopedNCS( name, color, depth ) ZoneNamedNCS( ___tracy_scoped_zone, name, color depth, true )
# define TracyAllocS( ptr, size, depth ) tracy::Profiler::MemAllocCallstack( ptr, size, depth );
# define TracyFreeS( ptr, depth ) tracy::Profiler::MemFreeCallstack( ptr, depth );
#else
# define ZoneNamedS( varname, depth, active ) ZoneNamed( varname, active )
# define ZoneNamedNS( varname, name, depth, active ) ZoneNamedN( varname, name, active )
# define ZoneNamedCS( varname, color, depth, active ) ZoneNamedC( varname, color, active )
# define ZoneNamedNCS( varname, name, color, depth, active ) ZoneNamedNC( varname, name, color, active )
# define ZoneScopedS( depth ) ZoneScoped
# define ZoneScopedNS( name, depth ) ZoneScopedN( name )
# define ZoneScopedCS( color, depth ) ZoneScopedC( color )
# define ZoneScopedNCS( name, color, depth ) ZoneScopedNC( name, color )
# define TracyAllocS( ptr, size, depth ) TracyAlloc( ptr, size )
# define TracyFreeS( ptr, depth ) TracyFree( ptr )
#endif
#endif
#endif

29
TracyClient.cpp Normal file
View File

@@ -0,0 +1,29 @@
//
// Tracy profiler
// ----------------
//
// For fast integration, compile and
// link with this source file (and none
// other) in your executable (or in the
// main DLL / shared object on multi-DLL
// projects).
//
// Define TRACY_ENABLE to enable profiler.
#include "common/TracySystem.cpp"
#ifdef TRACY_ENABLE
#include "client/TracyProfiler.cpp"
#include "client/TracyCallstack.cpp"
#include "common/tracy_lz4.cpp"
#include "common/TracySocket.cpp"
#include "client/tracy_rpmalloc.cpp"
#ifdef _MSC_VER
# pragma comment(lib, "ws2_32.lib")
# pragma comment(lib, "dbghelp.lib")
#endif
#endif

73
TracyClientDLL.cpp Normal file
View File

@@ -0,0 +1,73 @@
//
// Tracy profiler
// ----------------
//
// On multi-DLL projects compile and
// link with this source file (and none
// other) in the executable and in
// DLLs / shared objects that link to
// the main DLL.
//
// Define TRACY_ENABLE to enable profiler.
#include "common/TracySystem.cpp"
#ifdef TRACY_ENABLE
#include "client/TracyProfiler.hpp"
#include "client/concurrentqueue.h"
#include "common/TracyQueue.hpp"
namespace tracy
{
#ifdef _MSC_VER
# define DLL_IMPORT __declspec(dllimport)
#else
# define DLL_IMPORT
#endif
DLL_IMPORT moodycamel::ConcurrentQueue<QueueItem>::ExplicitProducer* get_token();
DLL_IMPORT void*(*get_rpmalloc())(size_t size);
DLL_IMPORT void(*get_rpfree())(void* ptr);
DLL_IMPORT Profiler& get_profiler();
#if defined TRACY_HW_TIMER && __ARM_ARCH >= 6
DLL_IMPORT int64_t(*get_GetTimeImpl())();
int64_t(*GetTimeImpl)() = get_GetTimeImpl();
#endif
#ifdef TRACY_COLLECT_THREAD_NAMES
DLL_IMPORT std::atomic<ThreadNameData*>& get_threadNameData();
DLL_IMPORT void(*get_rpmalloc_thread_initialize())();
std::atomic<ThreadNameData*>& s_threadNameData = get_threadNameData();
void(*rpmalloc_thread_initialize_fpt)() = get_rpmalloc_thread_initialize();
void rpmalloc_thread_initialize(void)
{
rpmalloc_thread_initialize_fpt();
}
#endif
static void*(*rpmalloc_fpt)(size_t size) = get_rpmalloc();
static void(*rpfree_fpt)(void* ptr) = get_rpfree();
RPMALLOC_RESTRICT void* rpmalloc(size_t size)
{
return rpmalloc_fpt(size);
}
void rpfree(void* ptr)
{
rpfree_fpt(ptr);
}
Profiler& s_profiler = get_profiler();
thread_local ProducerWrapper s_token { get_token() };
}
#endif

376
TracyLua.hpp Normal file
View File

@@ -0,0 +1,376 @@
#ifndef __TRACYLUA_HPP__
#define __TRACYLUA_HPP__
// Include this file after you include lua headers.
#ifndef TRACY_ENABLE
#include <string.h>
namespace tracy
{
namespace detail
{
static inline int noop( lua_State* L ) { return 0; }
}
static inline void LuaRegister( lua_State* L )
{
lua_newtable( L );
lua_pushcfunction( L, detail::noop );
lua_setfield( L, -2, "ZoneBegin" );
lua_pushcfunction( L, detail::noop );
lua_setfield( L, -2, "ZoneBeginN" );
lua_pushcfunction( L, detail::noop );
lua_setfield( L, -2, "ZoneEnd" );
lua_pushcfunction( L, detail::noop );
lua_setfield( L, -2, "ZoneText" );
lua_pushcfunction( L, detail::noop );
lua_setfield( L, -2, "ZoneName" );
lua_pushcfunction( L, detail::noop );
lua_setfield( L, -2, "Message" );
lua_setglobal( L, "tracy" );
}
static inline char* FindEnd( char* ptr )
{
unsigned int cnt = 1;
while( cnt != 0 )
{
if( *ptr == '(' ) cnt++;
else if( *ptr == ')' ) cnt--;
ptr++;
}
return ptr;
}
static inline void LuaRemove( char* script )
{
while( *script )
{
if( strncmp( script, "tracy.", 6 ) == 0 )
{
if( strncmp( script + 6, "Zone", 4 ) == 0 )
{
if( strncmp( script + 10, "End()", 5 ) == 0 )
{
memset( script, ' ', 15 );
script += 15;
}
else if( strncmp( script + 10, "Begin()", 7 ) == 0 )
{
memset( script, ' ', 17 );
script += 17;
}
else if( strncmp( script + 10, "Text(", 5 ) == 0 )
{
auto end = FindEnd( script + 15 );
memset( script, ' ', end - script );
script = end;
}
else if( strncmp( script + 10, "Name(", 5 ) == 0 )
{
auto end = FindEnd( script + 15 );
memset( script, ' ', end - script );
script = end;
}
else if( strncmp( script + 10, "BeginN(", 7 ) == 0 )
{
auto end = FindEnd( script + 17 );
memset( script, ' ', end - script );
script = end;
}
else
{
script += 10;
}
}
else if( strncmp( script + 6, "Message(", 8 ) == 0 )
{
auto end = FindEnd( script + 14 );
memset( script, ' ', end - script );
script = end;
}
else
{
script += 6;
}
}
else
{
script++;
}
}
}
}
#else
#include <assert.h>
#include "common/TracyColor.hpp"
#include "common/TracyAlign.hpp"
#include "common/TracySystem.hpp"
#include "client/TracyProfiler.hpp"
namespace tracy
{
#ifdef TRACY_ON_DEMAND
extern thread_local LuaZoneState s_luaZoneState;
#endif
namespace detail
{
static inline int LuaZoneBegin( lua_State* L )
{
#ifdef TRACY_ON_DEMAND
const auto zoneCnt = s_luaZoneState.counter++;
if( zoneCnt != 0 && !s_luaZoneState.active ) return 0;
s_luaZoneState.active = s_profiler.IsConnected();
if( !s_luaZoneState.active ) return 0;
#endif
const uint32_t color = Color::DeepSkyBlue3;
lua_Debug dbg;
lua_getstack( L, 1, &dbg );
lua_getinfo( L, "Snl", &dbg );
const uint32_t line = dbg.currentline;
const auto func = dbg.name ? dbg.name : dbg.short_src;
const auto fsz = strlen( func );
const auto ssz = strlen( dbg.source );
// Data layout:
// 4b payload size
// 4b color
// 4b source line
// fsz function name
// 1b null terminator
// ssz source file name
// 1b null terminator
const uint32_t sz = uint32_t( 4 + 4 + 4 + fsz + 1 + ssz + 1 );
auto ptr = (char*)tracy_malloc( sz );
memcpy( ptr, &sz, 4 );
memcpy( ptr + 4, &color, 4 );
memcpy( ptr + 8, &line, 4 );
memcpy( ptr + 12, func, fsz+1 );
memcpy( ptr + 12 + fsz + 1, dbg.source, ssz + 1 );
Magic magic;
auto& token = s_token.ptr;
auto& tail = token->get_tail_index();
auto item = token->enqueue_begin<tracy::moodycamel::CanAlloc>( magic );
MemWrite( &item->hdr.type, QueueType::ZoneBeginAllocSrcLoc );
#ifdef TRACY_RDTSCP_OPT
MemWrite( &item->zoneBegin.time, Profiler::GetTime( item->zoneBegin.cpu ) );
#else
uint32_t cpu;
MemWrite( &item->zoneBegin.time, Profiler::GetTime( cpu ) );
MemWrite( &item->zoneBegin.cpu, cpu );
#endif
MemWrite( &item->zoneBegin.thread, GetThreadHandle() );
MemWrite( &item->zoneBegin.srcloc, (uint64_t)ptr );
tail.store( magic + 1, std::memory_order_release );
return 0;
}
static inline int LuaZoneBeginN( lua_State* L )
{
#ifdef TRACY_ON_DEMAND
const auto zoneCnt = s_luaZoneState.counter++;
if( zoneCnt != 0 && !s_luaZoneState.active ) return 0;
s_luaZoneState.active = s_profiler.IsConnected();
if( !s_luaZoneState.active ) return 0;
#endif
const uint32_t color = Color::DeepSkyBlue3;
lua_Debug dbg;
lua_getstack( L, 1, &dbg );
lua_getinfo( L, "Snl", &dbg );
const uint32_t line = dbg.currentline;
const auto func = dbg.name ? dbg.name : dbg.short_src;
size_t nsz;
const auto name = lua_tolstring( L, 1, &nsz );
const auto fsz = strlen( func );
const auto ssz = strlen( dbg.source );
// Data layout:
// 4b payload size
// 4b color
// 4b source line
// fsz function name
// 1b null terminator
// ssz source file name
// 1b null terminator
// nsz zone name
const uint32_t sz = uint32_t( 4 + 4 + 4 + fsz + 1 + ssz + 1 + nsz );
auto ptr = (char*)tracy_malloc( sz );
memcpy( ptr, &sz, 4 );
memcpy( ptr + 4, &color, 4 );
memcpy( ptr + 8, &line, 4 );
memcpy( ptr + 12, func, fsz+1 );
memcpy( ptr + 12 + fsz + 1, dbg.source, ssz + 1 );
memcpy( ptr + 12 + fsz + 1 + ssz + 1, name, nsz );
Magic magic;
auto& token = s_token.ptr;
auto& tail = token->get_tail_index();
auto item = token->enqueue_begin<tracy::moodycamel::CanAlloc>( magic );
MemWrite( &item->hdr.type, QueueType::ZoneBeginAllocSrcLoc );
#ifdef TRACY_RDTSCP_OPT
MemWrite( &item->zoneBegin.time, Profiler::GetTime( item->zoneBegin.cpu ) );
#else
uint32_t cpu;
MemWrite( &item->zoneBegin.time, Profiler::GetTime( cpu ) );
MemWrite( &item->zoneBegin.cpu, cpu );
#endif
MemWrite( &item->zoneBegin.thread, GetThreadHandle() );
MemWrite( &item->zoneBegin.srcloc, (uint64_t)ptr );
tail.store( magic + 1, std::memory_order_release );
return 0;
}
static inline int LuaZoneEnd( lua_State* L )
{
#ifdef TRACY_ON_DEMAND
assert( s_luaZoneState.counter != 0 );
s_luaZoneState.counter--;
if( !s_luaZoneState.active ) return 0;
if( !s_profiler.IsConnected() )
{
s_luaZoneState.active = false;
return 0;
}
#endif
Magic magic;
auto& token = s_token.ptr;
auto& tail = token->get_tail_index();
auto item = token->enqueue_begin<tracy::moodycamel::CanAlloc>( magic );
MemWrite( &item->hdr.type, QueueType::ZoneEnd );
#ifdef TRACY_RDTSCP_OPT
MemWrite( &item->zoneEnd.time, Profiler::GetTime( item->zoneEnd.cpu ) );
#else
uint32_t cpu;
MemWrite( &item->zoneEnd.time, Profiler::GetTime( cpu ) );
MemWrite( &item->zoneEnd.cpu, cpu );
#endif
MemWrite( &item->zoneEnd.thread, GetThreadHandle() );
tail.store( magic + 1, std::memory_order_release );
return 0;
}
static inline int LuaZoneText( lua_State* L )
{
#ifdef TRACY_ON_DEMAND
if( !s_luaZoneState.active ) return 0;
if( !s_profiler.IsConnected() )
{
s_luaZoneState.active = false;
return 0;
}
#endif
auto txt = lua_tostring( L, 1 );
const auto size = strlen( txt );
Magic magic;
auto& token = s_token.ptr;
auto ptr = (char*)tracy_malloc( size+1 );
memcpy( ptr, txt, size );
ptr[size] = '\0';
auto& tail = token->get_tail_index();
auto item = token->enqueue_begin<tracy::moodycamel::CanAlloc>( magic );
MemWrite( &item->hdr.type, QueueType::ZoneText );
MemWrite( &item->zoneText.thread, GetThreadHandle() );
MemWrite( &item->zoneText.text, (uint64_t)ptr );
tail.store( magic + 1, std::memory_order_release );
return 0;
}
static inline int LuaZoneName( lua_State* L )
{
#ifdef TRACY_ON_DEMAND
if( !s_luaZoneState.active ) return 0;
if( !s_profiler.IsConnected() )
{
s_luaZoneState.active = false;
return 0;
}
#endif
auto txt = lua_tostring( L, 1 );
const auto size = strlen( txt );
Magic magic;
auto& token = s_token.ptr;
auto ptr = (char*)tracy_malloc( size+1 );
memcpy( ptr, txt, size );
ptr[size] = '\0';
auto& tail = token->get_tail_index();
auto item = token->enqueue_begin<tracy::moodycamel::CanAlloc>( magic );
MemWrite( &item->hdr.type, QueueType::ZoneName );
MemWrite( &item->zoneText.thread, GetThreadHandle() );
MemWrite( &item->zoneText.text, (uint64_t)ptr );
tail.store( magic + 1, std::memory_order_release );
return 0;
}
static inline int LuaMessage( lua_State* L )
{
#ifdef TRACY_ON_DEMAND
if( !s_profiler.IsConnected() ) return 0;
#endif
auto txt = lua_tostring( L, 1 );
const auto size = strlen( txt );
Magic magic;
auto& token = s_token.ptr;
auto ptr = (char*)tracy_malloc( size+1 );
memcpy( ptr, txt, size );
ptr[size] = '\0';
auto& tail = token->get_tail_index();
auto item = token->enqueue_begin<tracy::moodycamel::CanAlloc>( magic );
MemWrite( &item->hdr.type, QueueType::Message );
MemWrite( &item->message.time, Profiler::GetTime() );
MemWrite( &item->message.thread, GetThreadHandle() );
MemWrite( &item->message.text, (uint64_t)ptr );
tail.store( magic + 1, std::memory_order_release );
return 0;
}
}
static inline void LuaRegister( lua_State* L )
{
lua_newtable( L );
lua_pushcfunction( L, detail::LuaZoneBegin );
lua_setfield( L, -2, "ZoneBegin" );
lua_pushcfunction( L, detail::LuaZoneBeginN );
lua_setfield( L, -2, "ZoneBeginN" );
lua_pushcfunction( L, detail::LuaZoneEnd );
lua_setfield( L, -2, "ZoneEnd" );
lua_pushcfunction( L, detail::LuaZoneText );
lua_setfield( L, -2, "ZoneText" );
lua_pushcfunction( L, detail::LuaZoneName );
lua_setfield( L, -2, "ZoneName" );
lua_pushcfunction( L, detail::LuaMessage );
lua_setfield( L, -2, "Message" );
lua_setglobal( L, "tracy" );
}
static inline void LuaRemove( char* script ) {}
}
#endif
#endif

265
TracyOpenGL.hpp Normal file
View File

@@ -0,0 +1,265 @@
#ifndef __TRACYOPENGL_HPP__
#define __TRACYOPENGL_HPP__
// Include this file after you include OpenGL 3.2 headers.
#if !defined TRACY_ENABLE || defined __APPLE__
#define TracyGpuContext
#define TracyGpuNamedZone(x,y)
#define TracyGpuNamedZoneC(x,y,z)
#define TracyGpuZone(x)
#define TracyGpuZoneC(x,y)
#define TracyGpuCollect
#define TracyGpuNamedZoneS(x,y,z)
#define TracyGpuNamedZoneCS(x,y,z,w)
#define TracyGpuZoneS(x,y)
#define TracyGpuZoneCS(x,y,z)
#else
#include <atomic>
#include <assert.h>
#include <stdlib.h>
#include "Tracy.hpp"
#include "client/TracyProfiler.hpp"
#include "client/TracyCallstack.hpp"
#include "common/TracyAlign.hpp"
#include "common/TracyAlloc.hpp"
#define TracyGpuContext tracy::s_gpuCtx.ptr = (tracy::GpuCtx*)tracy::tracy_malloc( sizeof( tracy::GpuCtx ) ); new(tracy::s_gpuCtx.ptr) tracy::GpuCtx;
#define TracyGpuNamedZone( varname, name ) static const tracy::SourceLocationData TracyConcat(__tracy_gpu_source_location,__LINE__) { name, __FUNCTION__, __FILE__, (uint32_t)__LINE__, 0 }; tracy::GpuCtxScope varname( &TracyConcat(__tracy_gpu_source_location,__LINE__) );
#define TracyGpuNamedZoneC( varname, name, color ) static const tracy::SourceLocationData TracyConcat(__tracy_gpu_source_location,__LINE__) { name, __FUNCTION__, __FILE__, (uint32_t)__LINE__, color }; tracy::GpuCtxScope varname( &TracyConcat(__tracy_gpu_source_location,__LINE__) );
#define TracyGpuZone( name ) TracyGpuNamedZone( ___tracy_gpu_zone, name )
#define TracyGpuZoneC( name, color ) TracyGpuNamedZoneC( ___tracy_gpu_zone, name, color )
#define TracyGpuCollect tracy::s_gpuCtx.ptr->Collect();
#ifdef TRACY_HAS_CALLSTACK
# define TracyGpuNamedZoneS( varname, name, depth ) static const tracy::SourceLocationData TracyConcat(__tracy_gpu_source_location,__LINE__) { name, __FUNCTION__, __FILE__, (uint32_t)__LINE__, 0 }; tracy::GpuCtxScope varname( &TracyConcat(__tracy_gpu_source_location,__LINE__), depth );
# define TracyGpuNamedZoneCS( varname, name, color, depth ) static const tracy::SourceLocationData TracyConcat(__tracy_gpu_source_location,__LINE__) { name, __FUNCTION__, __FILE__, (uint32_t)__LINE__, color }; tracy::GpuCtxScope varname( &TracyConcat(__tracy_gpu_source_location,__LINE__), depth );
# define TracyGpuZoneS( name, depth ) TracyGpuNamedZoneS( ___tracy_gpu_zone, name, depth )
# define TracyGpuZoneCS( name, color, depth ) TracyGpuNamedZoneCS( ___tracy_gpu_zone, name, color, depth )
#else
# define TracyGpuNamedZoneS( varname, name, depth ) TracyGpuNamedZone( varname, name )
# define TracyGpuNamedZoneCS( varname, name, color, depth ) TracyGpuNamedZoneC( varname, name, color )
# define TracyGpuZoneS( name, depth ) TracyGpuZone( name )
# define TracyGpuZoneCS( name, color, depth ) TracyGpuZoneC( name, color )
#endif
namespace tracy
{
extern std::atomic<uint8_t> s_gpuCtxCounter;
class GpuCtx
{
friend class GpuCtxScope;
enum { QueryCount = 64 * 1024 };
public:
GpuCtx()
: m_context( s_gpuCtxCounter.fetch_add( 1, std::memory_order_relaxed ) )
, m_head( 0 )
, m_tail( 0 )
{
assert( m_context != 255 );
glGenQueries( QueryCount, m_query );
int64_t tgpu;
glGetInteger64v( GL_TIMESTAMP, &tgpu );
int64_t tcpu = Profiler::GetTime();
GLint bits;
glGetQueryiv( GL_TIMESTAMP, GL_QUERY_COUNTER_BITS, &bits );
const float period = 1.f;
Magic magic;
auto& token = s_token.ptr;
auto& tail = token->get_tail_index();
auto item = token->enqueue_begin<tracy::moodycamel::CanAlloc>( magic );
MemWrite( &item->hdr.type, QueueType::GpuNewContext );
MemWrite( &item->gpuNewContext.cpuTime, tcpu );
MemWrite( &item->gpuNewContext.gpuTime, tgpu );
MemWrite( &item->gpuNewContext.thread, GetThreadHandle() );
MemWrite( &item->gpuNewContext.period, period );
MemWrite( &item->gpuNewContext.context, m_context );
MemWrite( &item->gpuNewContext.accuracyBits, (uint8_t)bits );
#ifdef TRACY_ON_DEMAND
s_profiler.DeferItem( *item );
#endif
tail.store( magic + 1, std::memory_order_release );
}
void Collect()
{
ZoneScopedC( Color::Red4 );
if( m_tail == m_head ) return;
#ifdef TRACY_ON_DEMAND
if( !s_profiler.IsConnected() )
{
m_head = m_tail = 0;
return;
}
#endif
auto start = m_tail;
auto end = m_head + QueryCount;
auto cnt = ( end - start ) % QueryCount;
while( cnt > 1 )
{
auto mid = start + cnt / 2;
GLint available;
glGetQueryObjectiv( m_query[mid % QueryCount], GL_QUERY_RESULT_AVAILABLE, &available );
if( available )
{
start = mid;
}
else
{
end = mid;
}
cnt = ( end - start ) % QueryCount;
}
start %= QueryCount;
Magic magic;
auto& token = s_token.ptr;
auto& tail = token->get_tail_index();
while( m_tail != start )
{
uint64_t time;
glGetQueryObjectui64v( m_query[m_tail], GL_QUERY_RESULT, &time );
auto item = token->enqueue_begin<tracy::moodycamel::CanAlloc>( magic );
MemWrite( &item->hdr.type, QueueType::GpuTime );
MemWrite( &item->gpuTime.gpuTime, (int64_t)time );
MemWrite( &item->gpuTime.queryId, (uint16_t)m_tail );
MemWrite( &item->gpuTime.context, m_context );
tail.store( magic + 1, std::memory_order_release );
m_tail = ( m_tail + 1 ) % QueryCount;
}
}
private:
tracy_force_inline unsigned int NextQueryId()
{
const auto id = m_head;
m_head = ( m_head + 1 ) % QueryCount;
assert( m_head != m_tail );
return id;
}
tracy_force_inline unsigned int TranslateOpenGlQueryId( unsigned int id )
{
return m_query[id];
}
tracy_force_inline uint8_t GetId() const
{
return m_context;
}
unsigned int m_query[QueryCount];
uint8_t m_context;
unsigned int m_head;
unsigned int m_tail;
};
extern thread_local GpuCtxWrapper s_gpuCtx;
class GpuCtxScope
{
public:
tracy_force_inline GpuCtxScope( const SourceLocationData* srcloc )
#ifdef TRACY_ON_DEMAND
: m_active( s_profiler.IsConnected() )
#endif
{
#ifdef TRACY_ON_DEMAND
if( !m_active ) return;
#endif
const auto queryId = s_gpuCtx.ptr->NextQueryId();
glQueryCounter( s_gpuCtx.ptr->TranslateOpenGlQueryId( queryId ), GL_TIMESTAMP );
Magic magic;
auto& token = s_token.ptr;
auto& tail = token->get_tail_index();
auto item = token->enqueue_begin<tracy::moodycamel::CanAlloc>( magic );
MemWrite( &item->hdr.type, QueueType::GpuZoneBegin );
MemWrite( &item->gpuZoneBegin.cpuTime, Profiler::GetTime() );
MemWrite( &item->gpuZoneBegin.srcloc, (uint64_t)srcloc );
memset( &item->gpuZoneBegin.thread, 0, sizeof( item->gpuZoneBegin.thread ) );
MemWrite( &item->gpuZoneBegin.queryId, uint16_t( queryId ) );
MemWrite( &item->gpuZoneBegin.context, s_gpuCtx.ptr->GetId() );
tail.store( magic + 1, std::memory_order_release );
}
tracy_force_inline GpuCtxScope( const SourceLocationData* srcloc, int depth )
#ifdef TRACY_ON_DEMAND
: m_active( s_profiler.IsConnected() )
#endif
{
#ifdef TRACY_ON_DEMAND
if( !m_active ) return;
#endif
const auto queryId = s_gpuCtx.ptr->NextQueryId();
glQueryCounter( s_gpuCtx.ptr->TranslateOpenGlQueryId( queryId ), GL_TIMESTAMP );
const auto thread = GetThreadHandle();
Magic magic;
auto& token = s_token.ptr;
auto& tail = token->get_tail_index();
auto item = token->enqueue_begin<tracy::moodycamel::CanAlloc>( magic );
MemWrite( &item->hdr.type, QueueType::GpuZoneBeginCallstack );
MemWrite( &item->gpuZoneBegin.cpuTime, Profiler::GetTime() );
MemWrite( &item->gpuZoneBegin.srcloc, (uint64_t)srcloc );
MemWrite( &item->gpuZoneBegin.thread, thread );
MemWrite( &item->gpuZoneBegin.queryId, uint16_t( queryId ) );
MemWrite( &item->gpuZoneBegin.context, s_gpuCtx.ptr->GetId() );
tail.store( magic + 1, std::memory_order_release );
s_profiler.SendCallstack( depth, thread );
}
tracy_force_inline ~GpuCtxScope()
{
#ifdef TRACY_ON_DEMAND
if( !m_active ) return;
#endif
const auto queryId = s_gpuCtx.ptr->NextQueryId();
glQueryCounter( s_gpuCtx.ptr->TranslateOpenGlQueryId( queryId ), GL_TIMESTAMP );
Magic magic;
auto& token = s_token.ptr;
auto& tail = token->get_tail_index();
auto item = token->enqueue_begin<tracy::moodycamel::CanAlloc>( magic );
MemWrite( &item->hdr.type, QueueType::GpuZoneEnd );
MemWrite( &item->gpuZoneEnd.cpuTime, Profiler::GetTime() );
MemWrite( &item->gpuZoneEnd.queryId, uint16_t( queryId ) );
MemWrite( &item->gpuZoneEnd.context, s_gpuCtx.ptr->GetId() );
tail.store( magic + 1, std::memory_order_release );
}
private:
#ifdef TRACY_ON_DEMAND
const bool m_active;
#endif
};
}
#endif
#endif

305
TracyVulkan.hpp Normal file
View File

@@ -0,0 +1,305 @@
#ifndef __TRACYVULKAN_HPP__
#define __TRACYVULKAN_HPP__
#if !defined TRACY_ENABLE
#define TracyVkContext(x,y,z,w)
#define TracyVkDestroy
#define TracyVkNamedZone(x,y,z)
#define TracyVkNamedZoneC(x,y,z,w)
#define TracyVkZone(x,y)
#define TracyVkZoneC(x,y,z)
#define TracyVkCollect(x)
#define TracyVkNamedZoneS(x,y,z,w)
#define TracyVkNamedZoneCS(x,y,z,w,v)
#define TracyVkZoneS(x,y,z)
#define TracyVkZoneCS(x,y,z,w)
#else
#include <assert.h>
#include <stdlib.h>
#include <vulkan/vulkan.h>
#include "Tracy.hpp"
#include "client/TracyProfiler.hpp"
#include "client/TracyCallstack.hpp"
#define TracyVkContext( physdev, device, queue, cmdbuf ) tracy::s_vkCtx.ptr = (tracy::VkCtx*)tracy::tracy_malloc( sizeof( tracy::VkCtx ) ); new(tracy::s_vkCtx.ptr) tracy::VkCtx( physdev, device, queue, cmdbuf );
#define TracyVkDestroy tracy::s_vkCtx.ptr->~VkCtx(); tracy::tracy_free( tracy::s_vkCtx.ptr ); tracy::s_vkCtx.ptr = nullptr;
#define TracyVkNamedZone( varname, cmdbuf, name ) static const tracy::SourceLocationData TracyConcat(__tracy_gpu_source_location,__LINE__) { name, __FUNCTION__, __FILE__, (uint32_t)__LINE__, 0 }; tracy::VkCtxScope varname( &TracyConcat(__tracy_gpu_source_location,__LINE__), cmdbuf );
#define TracyVkNamedZoneC( varname, cmdbuf, name, color ) static const tracy::SourceLocationData TracyConcat(__tracy_gpu_source_location,__LINE__) { name, __FUNCTION__, __FILE__, (uint32_t)__LINE__, color }; tracy::VkCtxScope varname( &TracyConcat(__tracy_gpu_source_location,__LINE__), cmdbuf );
#define TracyVkZone( cmdbuf, name ) TracyVkNamedZone( ___tracy_gpu_zone, cmdbuf, name )
#define TracyVkZoneC( cmdbuf, name, color ) TracyVkNamedZoneC( ___tracy_gpu_zone, cmdbuf, name, color )
#define TracyVkCollect( cmdbuf ) tracy::s_vkCtx.ptr->Collect( cmdbuf );
#ifdef TRACY_HAS_CALLSTACK
# define TracyVkNamedZoneS( varname, cmdbuf, name, depth ) static const tracy::SourceLocationData TracyConcat(__tracy_gpu_source_location,__LINE__) { name, __FUNCTION__, __FILE__, (uint32_t)__LINE__, 0 }; tracy::VkCtxScope varname( &TracyConcat(__tracy_gpu_source_location,__LINE__), cmdbuf, depth );
# define TracyVkNamedZoneCS( varname, cmdbuf, name, color, depth ) static const tracy::SourceLocationData TracyConcat(__tracy_gpu_source_location,__LINE__) { name, __FUNCTION__, __FILE__, (uint32_t)__LINE__, color }; tracy::VkCtxScope varname( &TracyConcat(__tracy_gpu_source_location,__LINE__), cmdbuf, depth );
# define TracyVkZoneS( cmdbuf, name, depth ) TracyVkNamedZoneS( ___tracy_gpu_zone, cmdbuf, name, depth )
# define TracyVkZoneCS( cmdbuf, name, color, depth ) TracyVkNamedZoneCS( ___tracy_gpu_zone, cmdbuf, name, color, depth )
#else
# define TracyVkNamedZoneS( varname, cmdbuf, name, depth ) TracyVkNamedZone( varname, cmdbuf, name )
# define TracyVkNamedZoneCS( varname, cmdbuf, name, color, depth ) TracyVkNamedZoneC( varname, cmdbuf, name, color )
# define TracyVkZoneS( cmdbuf, name, depth ) TracyVkZone( cmdbuf, name )
# define TracyVkZoneCS( cmdbuf, name, color, depth ) TracyVkZoneC( cmdbuf, name, color )
#endif
namespace tracy
{
extern std::atomic<uint8_t> s_gpuCtxCounter;
class VkCtx
{
friend class VkCtxScope;
enum { QueryCount = 64 * 1024 };
public:
VkCtx( VkPhysicalDevice physdev, VkDevice device, VkQueue queue, VkCommandBuffer cmdbuf )
: m_device( device )
, m_queue( queue )
, m_context( s_gpuCtxCounter.fetch_add( 1, std::memory_order_relaxed ) )
, m_head( 0 )
, m_tail( 0 )
, m_oldCnt( 0 )
{
assert( m_context != 255 );
VkPhysicalDeviceProperties prop;
vkGetPhysicalDeviceProperties( physdev, &prop );
const float period = prop.limits.timestampPeriod;
VkQueryPoolCreateInfo poolInfo = {};
poolInfo.sType = VK_STRUCTURE_TYPE_QUERY_POOL_CREATE_INFO;
poolInfo.queryCount = QueryCount;
poolInfo.queryType = VK_QUERY_TYPE_TIMESTAMP;
vkCreateQueryPool( device, &poolInfo, nullptr, &m_query );
VkCommandBufferBeginInfo beginInfo = {};
beginInfo.sType = VK_STRUCTURE_TYPE_COMMAND_BUFFER_BEGIN_INFO;
beginInfo.flags = VK_COMMAND_BUFFER_USAGE_ONE_TIME_SUBMIT_BIT;
VkSubmitInfo submitInfo = {};
submitInfo.sType = VK_STRUCTURE_TYPE_SUBMIT_INFO;
submitInfo.commandBufferCount = 1;
submitInfo.pCommandBuffers = &cmdbuf;
vkBeginCommandBuffer( cmdbuf, &beginInfo );
vkCmdResetQueryPool( cmdbuf, m_query, 0, QueryCount );
vkEndCommandBuffer( cmdbuf );
vkQueueSubmit( queue, 1, &submitInfo, VK_NULL_HANDLE );
vkQueueWaitIdle( queue );
vkBeginCommandBuffer( cmdbuf, &beginInfo );
vkCmdWriteTimestamp( cmdbuf, VK_PIPELINE_STAGE_TOP_OF_PIPE_BIT, m_query, 0 );
vkEndCommandBuffer( cmdbuf );
vkQueueSubmit( queue, 1, &submitInfo, VK_NULL_HANDLE );
vkQueueWaitIdle( queue );
int64_t tcpu = Profiler::GetTime();
int64_t tgpu;
vkGetQueryPoolResults( device, m_query, 0, 1, sizeof( tgpu ), &tgpu, sizeof( tgpu ), VK_QUERY_RESULT_64_BIT | VK_QUERY_RESULT_WAIT_BIT );
vkBeginCommandBuffer( cmdbuf, &beginInfo );
vkCmdResetQueryPool( cmdbuf, m_query, 0, 1 );
vkEndCommandBuffer( cmdbuf );
vkQueueSubmit( queue, 1, &submitInfo, VK_NULL_HANDLE );
vkQueueWaitIdle( queue );
Magic magic;
auto& token = s_token.ptr;
auto& tail = token->get_tail_index();
auto item = token->enqueue_begin<tracy::moodycamel::CanAlloc>( magic );
MemWrite( &item->hdr.type, QueueType::GpuNewContext );
MemWrite( &item->gpuNewContext.cpuTime, tcpu );
MemWrite( &item->gpuNewContext.gpuTime, tgpu );
memset( &item->gpuNewContext.thread, 0, sizeof( item->gpuNewContext.thread ) );
MemWrite( &item->gpuNewContext.period, period );
MemWrite( &item->gpuNewContext.context, m_context );
MemWrite( &item->gpuNewContext.accuracyBits, uint8_t( 0 ) );
#ifdef TRACY_ON_DEMAND
s_profiler.DeferItem( *item );
#endif
tail.store( magic + 1, std::memory_order_release );
}
~VkCtx()
{
vkDestroyQueryPool( m_device, m_query, nullptr );
}
void Collect( VkCommandBuffer cmdbuf )
{
ZoneScopedC( Color::Red4 );
if( m_tail == m_head ) return;
#ifdef TRACY_ON_DEMAND
if( !s_profiler.IsConnected() )
{
vkCmdResetQueryPool( cmdbuf, m_query, 0, QueryCount );
m_head = m_tail = 0;
return;
}
#endif
unsigned int cnt;
if( m_oldCnt != 0 )
{
cnt = m_oldCnt;
m_oldCnt = 0;
}
else
{
cnt = m_head < m_tail ? QueryCount - m_tail : m_head - m_tail;
}
int64_t res[QueryCount];
if( vkGetQueryPoolResults( m_device, m_query, m_tail, cnt, sizeof( res ), res, sizeof( *res ), VK_QUERY_RESULT_64_BIT ) == VK_NOT_READY )
{
m_oldCnt = cnt;
return;
}
Magic magic;
auto& token = s_token.ptr;
auto& tail = token->get_tail_index();
for( unsigned int idx=0; idx<cnt; idx++ )
{
auto item = token->enqueue_begin<tracy::moodycamel::CanAlloc>( magic );
MemWrite( &item->hdr.type, QueueType::GpuTime );
MemWrite( &item->gpuTime.gpuTime, res[idx] );
MemWrite( &item->gpuTime.queryId, uint16_t( m_tail + idx ) );
MemWrite( &item->gpuTime.context, m_context );
tail.store( magic + 1, std::memory_order_release );
}
vkCmdResetQueryPool( cmdbuf, m_query, m_tail, cnt );
m_tail += cnt;
if( m_tail == QueryCount ) m_tail = 0;
}
private:
tracy_force_inline unsigned int NextQueryId()
{
const auto id = m_head;
m_head = ( m_head + 1 ) % QueryCount;
assert( m_head != m_tail );
return id;
}
tracy_force_inline uint8_t GetId() const
{
return m_context;
}
VkDevice m_device;
VkQueue m_queue;
VkQueryPool m_query;
uint8_t m_context;
unsigned int m_head;
unsigned int m_tail;
unsigned int m_oldCnt;
};
extern VkCtxWrapper s_vkCtx;
class VkCtxScope
{
public:
tracy_force_inline VkCtxScope( const SourceLocationData* srcloc, VkCommandBuffer cmdbuf )
: m_cmdbuf( cmdbuf )
#ifdef TRACY_ON_DEMAND
, m_active( s_profiler.IsConnected() )
#endif
{
#ifdef TRACY_ON_DEMAND
if( !m_active ) return;
#endif
auto ctx = s_vkCtx.ptr;
const auto queryId = ctx->NextQueryId();
vkCmdWriteTimestamp( cmdbuf, VK_PIPELINE_STAGE_TOP_OF_PIPE_BIT, ctx->m_query, queryId );
Magic magic;
auto& token = s_token.ptr;
auto& tail = token->get_tail_index();
auto item = token->enqueue_begin<tracy::moodycamel::CanAlloc>( magic );
MemWrite( &item->hdr.type, QueueType::GpuZoneBegin );
MemWrite( &item->gpuZoneBegin.cpuTime, Profiler::GetTime() );
MemWrite( &item->gpuZoneBegin.srcloc, (uint64_t)srcloc );
MemWrite( &item->gpuZoneBegin.thread, GetThreadHandle() );
MemWrite( &item->gpuZoneBegin.queryId, uint16_t( queryId ) );
MemWrite( &item->gpuZoneBegin.context, ctx->GetId() );
tail.store( magic + 1, std::memory_order_release );
}
tracy_force_inline VkCtxScope( const SourceLocationData* srcloc, VkCommandBuffer cmdbuf, int depth )
: m_cmdbuf( cmdbuf )
#ifdef TRACY_ON_DEMAND
, m_active( s_profiler.IsConnected() )
#endif
{
#ifdef TRACY_ON_DEMAND
if( !m_active ) return;
#endif
const auto thread = GetThreadHandle();
auto ctx = s_vkCtx.ptr;
const auto queryId = ctx->NextQueryId();
vkCmdWriteTimestamp( cmdbuf, VK_PIPELINE_STAGE_TOP_OF_PIPE_BIT, ctx->m_query, queryId );
Magic magic;
auto& token = s_token.ptr;
auto& tail = token->get_tail_index();
auto item = token->enqueue_begin<tracy::moodycamel::CanAlloc>( magic );
MemWrite( &item->hdr.type, QueueType::GpuZoneBeginCallstack );
MemWrite( &item->gpuZoneBegin.cpuTime, Profiler::GetTime() );
MemWrite( &item->gpuZoneBegin.srcloc, (uint64_t)srcloc );
MemWrite( &item->gpuZoneBegin.thread, thread );
MemWrite( &item->gpuZoneBegin.queryId, uint16_t( queryId ) );
MemWrite( &item->gpuZoneBegin.context, ctx->GetId() );
tail.store( magic + 1, std::memory_order_release );
s_profiler.SendCallstack( depth, thread );
}
tracy_force_inline ~VkCtxScope()
{
#ifdef TRACY_ON_DEMAND
if( !m_active ) return;
#endif
auto ctx = s_vkCtx.ptr;
const auto queryId = ctx->NextQueryId();
vkCmdWriteTimestamp( m_cmdbuf, VK_PIPELINE_STAGE_TOP_OF_PIPE_BIT, ctx->m_query, queryId );
Magic magic;
auto& token = s_token.ptr;
auto& tail = token->get_tail_index();
auto item = token->enqueue_begin<tracy::moodycamel::CanAlloc>( magic );
MemWrite( &item->hdr.type, QueueType::GpuZoneEnd );
MemWrite( &item->gpuZoneEnd.cpuTime, Profiler::GetTime() );
MemWrite( &item->gpuZoneEnd.queryId, uint16_t( queryId ) );
MemWrite( &item->gpuZoneEnd.context, ctx->GetId() );
tail.store( magic + 1, std::memory_order_release );
}
private:
VkCommandBuffer m_cmdbuf;
#ifdef TRACY_ON_DEMAND
const bool m_active;
#endif
};
}
#endif
#endif

View File

@@ -1,16 +1,12 @@
all: release
all: debug
debug:
@$(MAKE) -f debug.mk all
@+make -f debug.mk all
release:
@$(MAKE) -f release.mk all
@+make -f release.mk all
clean:
@$(MAKE) -f build.mk clean
@+make -f build.mk clean
db: clean
@bear -- $(MAKE) -f debug.mk all
@mv -f compile_commands.json ../../../
.PHONY: all clean debug release db
.PHONY: all clean debug release

View File

@@ -1,12 +1,55 @@
CFLAGS +=
CXXFLAGS := $(CFLAGS) -std=gnu++17
DEFINES += -DTRACY_NO_STATISTICS
INCLUDES := $(shell pkg-config --cflags capstone)
LIBS += $(shell pkg-config --libs capstone) -lpthread
INCLUDES :=
LIBS := -lpthread
PROJECT := capture
IMAGE := $(PROJECT)-$(BUILD)
FILTER := ../../../getopt/getopt.c
include ../../../common/src-from-vcxproj.mk
FILTER :=
include ../../../common/unix.mk
BASE := $(shell egrep 'ClCompile.*cpp"' ../win32/$(PROJECT).vcxproj | sed -e 's/.*\"\(.*\)\".*/\1/' | sed -e 's@\\@/@g')
BASE2 := $(shell egrep 'ClCompile.*c"' ../win32/$(PROJECT).vcxproj | sed -e 's/.*\"\(.*\)\".*/\1/' | sed -e 's@\\@/@g')
SRC := $(filter-out $(FILTER),$(BASE))
SRC2 := $(filter-out $(FILTER),$(BASE2))
OBJDIRBASE := obj/$(BUILD)
OBJDIR := $(OBJDIRBASE)/o/o/o
OBJ := $(addprefix $(OBJDIR)/,$(SRC:%.cpp=%.o))
OBJ2 := $(addprefix $(OBJDIR)/,$(SRC2:%.c=%.o))
all: $(IMAGE)
$(OBJDIR)/%.o: %.cpp
$(CXX) -c $(INCLUDES) $(CXXFLAGS) $(DEFINES) $< -o $@
$(OBJDIR)/%.d : %.cpp
@echo Resolving dependencies of $<
@mkdir -p $(@D)
@$(CXX) -MM $(INCLUDES) $(CXXFLAGS) $(DEFINES) $< > $@.$$$$; \
sed 's,.*\.o[ :]*,$(OBJDIR)/$(<:.cpp=.o) $@ : ,g' < $@.$$$$ > $@; \
rm -f $@.$$$$
$(OBJDIR)/%.o: %.c
$(CC) -c $(INCLUDES) $(CFLAGS) $(DEFINES) $< -o $@
$(OBJDIR)/%.d : %.c
@echo Resolving dependencies of $<
@mkdir -p $(@D)
@$(CC) -MM $(INCLUDES) $(CFLAGS) $(DEFINES) $< > $@.$$$$; \
sed 's,.*\.o[ :]*,$(OBJDIR)/$(<:.c=.o) $@ : ,g' < $@.$$$$ > $@; \
rm -f $@.$$$$
$(IMAGE): $(OBJ) $(OBJ2)
$(CXX) $(CXXFLAGS) $(DEFINES) $(OBJ) $(OBJ2) $(LIBS) -o $@
ifneq "$(MAKECMDGOALS)" "clean"
-include $(addprefix $(OBJDIR)/,$(SRC:.cpp=.d)) %(addprefix $(OBJDIR)/,$(SRC2:.c=.d))
endif
clean:
rm -rf $(OBJDIRBASE) $(IMAGE)*
.PHONY: clean all

View File

@@ -1,6 +1,11 @@
ARCH := $(shell uname -m)
CFLAGS := -g3 -Wall
DEFINES := -DDEBUG
BUILD := debug
include ../../../common/unix-debug.mk
ifeq ($(ARCH),x86_64)
CFLAGS += -msse4.1
endif
include build.mk

View File

@@ -1,9 +1,11 @@
CFLAGS := -O3
ifndef TRACY_NO_LTO
CFLAGS += -flto
endif
ARCH := $(shell uname -m)
CFLAGS := -O3 -s -fomit-frame-pointer
DEFINES := -DNDEBUG
BUILD := release
include ../../../common/unix-release.mk
ifeq ($(ARCH),x86_64)
CFLAGS += -msse4.1
endif
include build.mk

View File

@@ -1,7 +1,7 @@
Microsoft Visual Studio Solution File, Format Version 12.00
# Visual Studio Version 16
VisualStudioVersion = 16.0.30907.101
# Visual Studio 15
VisualStudioVersion = 15.0.27428.2002
MinimumVisualStudioVersion = 10.0.40219.1
Project("{8BC9CEB8-8B4A-11D0-8D11-00A0C91BC942}") = "capture", "capture.vcxproj", "{447D58BF-94CD-4469-BB90-549C05D03E00}"
EndProject

View File

@@ -1,6 +1,14 @@
<?xml version="1.0" encoding="utf-8"?>
<Project DefaultTargets="Build" ToolsVersion="15.0" xmlns="http://schemas.microsoft.com/developer/msbuild/2003">
<ItemGroup Label="ProjectConfigurations">
<ProjectConfiguration Include="Debug|Win32">
<Configuration>Debug</Configuration>
<Platform>Win32</Platform>
</ProjectConfiguration>
<ProjectConfiguration Include="Release|Win32">
<Configuration>Release</Configuration>
<Platform>Win32</Platform>
</ProjectConfiguration>
<ProjectConfiguration Include="Debug|x64">
<Configuration>Debug</Configuration>
<Platform>x64</Platform>
@@ -14,20 +22,32 @@
<VCProjectVersion>15.0</VCProjectVersion>
<ProjectGuid>{447D58BF-94CD-4469-BB90-549C05D03E00}</ProjectGuid>
<RootNamespace>capture</RootNamespace>
<WindowsTargetPlatformVersion>10.0</WindowsTargetPlatformVersion>
<VcpkgTriplet>x64-windows-static</VcpkgTriplet>
<WindowsTargetPlatformVersion>10.0.16299.0</WindowsTargetPlatformVersion>
</PropertyGroup>
<Import Project="$(VCTargetsPath)\Microsoft.Cpp.Default.props" />
<PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Debug|Win32'" Label="Configuration">
<ConfigurationType>Application</ConfigurationType>
<UseDebugLibraries>true</UseDebugLibraries>
<PlatformToolset>v141</PlatformToolset>
<CharacterSet>MultiByte</CharacterSet>
</PropertyGroup>
<PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Release|Win32'" Label="Configuration">
<ConfigurationType>Application</ConfigurationType>
<UseDebugLibraries>false</UseDebugLibraries>
<PlatformToolset>v141</PlatformToolset>
<WholeProgramOptimization>true</WholeProgramOptimization>
<CharacterSet>MultiByte</CharacterSet>
</PropertyGroup>
<PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Debug|x64'" Label="Configuration">
<ConfigurationType>Application</ConfigurationType>
<UseDebugLibraries>true</UseDebugLibraries>
<PlatformToolset>v143</PlatformToolset>
<PlatformToolset>v141</PlatformToolset>
<CharacterSet>MultiByte</CharacterSet>
</PropertyGroup>
<PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Release|x64'" Label="Configuration">
<ConfigurationType>Application</ConfigurationType>
<UseDebugLibraries>false</UseDebugLibraries>
<PlatformToolset>v143</PlatformToolset>
<PlatformToolset>v141</PlatformToolset>
<WholeProgramOptimization>true</WholeProgramOptimization>
<CharacterSet>MultiByte</CharacterSet>
</PropertyGroup>
@@ -36,6 +56,12 @@
</ImportGroup>
<ImportGroup Label="Shared">
</ImportGroup>
<ImportGroup Label="PropertySheets" Condition="'$(Configuration)|$(Platform)'=='Debug|Win32'">
<Import Project="$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props" Condition="exists('$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props')" Label="LocalAppDataPlatform" />
</ImportGroup>
<ImportGroup Label="PropertySheets" Condition="'$(Configuration)|$(Platform)'=='Release|Win32'">
<Import Project="$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props" Condition="exists('$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props')" Label="LocalAppDataPlatform" />
</ImportGroup>
<ImportGroup Label="PropertySheets" Condition="'$(Configuration)|$(Platform)'=='Debug|x64'">
<Import Project="$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props" Condition="exists('$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props')" Label="LocalAppDataPlatform" />
</ImportGroup>
@@ -44,9 +70,14 @@
</ImportGroup>
<PropertyGroup Label="UserMacros" />
<PropertyGroup />
<PropertyGroup Label="Vcpkg">
<VcpkgEnableManifest>true</VcpkgEnableManifest>
</PropertyGroup>
<ItemDefinitionGroup Condition="'$(Configuration)|$(Platform)'=='Debug|Win32'">
<ClCompile>
<WarningLevel>Level3</WarningLevel>
<Optimization>Disabled</Optimization>
<SDLCheck>true</SDLCheck>
<ConformanceMode>true</ConformanceMode>
</ClCompile>
</ItemDefinitionGroup>
<ItemDefinitionGroup Condition="'$(Configuration)|$(Platform)'=='Debug|x64'">
<ClCompile>
<WarningLevel>Level3</WarningLevel>
@@ -57,12 +88,24 @@
<PreprocessorDefinitions>TRACY_NO_STATISTICS;_CRT_SECURE_NO_DEPRECATE;_CRT_NONSTDC_NO_DEPRECATE;WIN32_LEAN_AND_MEAN;NOMINMAX;_USE_MATH_DEFINES;%(PreprocessorDefinitions)</PreprocessorDefinitions>
<EnableEnhancedInstructionSet>AdvancedVectorExtensions2</EnableEnhancedInstructionSet>
<LanguageStandard>stdcpplatest</LanguageStandard>
<AdditionalIncludeDirectories>$(ProjectDir)..\..\..\vcpkg_installed\$(VcpkgTriplet)\include;$(ProjectDir)..\..\..\vcpkg_installed\$(VcpkgTriplet)\include\capstone;$(VcpkgManifestRoot)\vcpkg_installed\$(VcpkgTriplet)\$(VcpkgTriplet)\include\capstone;$(VcpkgRoot)\installed\$(VcpkgTriplet)\include\capstone</AdditionalIncludeDirectories>
</ClCompile>
<Link>
<AdditionalDependencies>ws2_32.lib;capstone.lib;%(AdditionalDependencies)</AdditionalDependencies>
<AdditionalDependencies>ws2_32.lib;%(AdditionalDependencies)</AdditionalDependencies>
<SubSystem>Console</SubSystem>
<AdditionalLibraryDirectories>$(ProjectDir)..\..\..\vcpkg_installed\$(VcpkgTriplet)\debug\lib</AdditionalLibraryDirectories>
</Link>
</ItemDefinitionGroup>
<ItemDefinitionGroup Condition="'$(Configuration)|$(Platform)'=='Release|Win32'">
<ClCompile>
<WarningLevel>Level3</WarningLevel>
<Optimization>MaxSpeed</Optimization>
<FunctionLevelLinking>true</FunctionLevelLinking>
<IntrinsicFunctions>true</IntrinsicFunctions>
<SDLCheck>true</SDLCheck>
<ConformanceMode>true</ConformanceMode>
</ClCompile>
<Link>
<EnableCOMDATFolding>true</EnableCOMDATFolding>
<OptimizeReferences>true</OptimizeReferences>
</Link>
</ItemDefinitionGroup>
<ItemDefinitionGroup Condition="'$(Configuration)|$(Platform)'=='Release|x64'">
@@ -77,132 +120,47 @@
<PreprocessorDefinitions>TRACY_NO_STATISTICS;NDEBUG;_CRT_SECURE_NO_DEPRECATE;_CRT_NONSTDC_NO_DEPRECATE;WIN32_LEAN_AND_MEAN;NOMINMAX;_USE_MATH_DEFINES;%(PreprocessorDefinitions)</PreprocessorDefinitions>
<EnableEnhancedInstructionSet>AdvancedVectorExtensions2</EnableEnhancedInstructionSet>
<LanguageStandard>stdcpplatest</LanguageStandard>
<AdditionalIncludeDirectories>$(ProjectDir)..\..\..\vcpkg_installed\$(VcpkgTriplet)\include;$(ProjectDir)..\..\..\vcpkg_installed\$(VcpkgTriplet)\include\capstone;$(VcpkgManifestRoot)\vcpkg_installed\$(VcpkgTriplet)\$(VcpkgTriplet)\include\capstone;$(VcpkgRoot)\installed\$(VcpkgTriplet)\include\capstone</AdditionalIncludeDirectories>
</ClCompile>
<Link>
<EnableCOMDATFolding>true</EnableCOMDATFolding>
<OptimizeReferences>true</OptimizeReferences>
<AdditionalDependencies>ws2_32.lib;capstone.lib;%(AdditionalDependencies)</AdditionalDependencies>
<AdditionalDependencies>ws2_32.lib;%(AdditionalDependencies)</AdditionalDependencies>
<SubSystem>Console</SubSystem>
<AdditionalLibraryDirectories>$(ProjectDir)..\..\..\vcpkg_installed\$(VcpkgTriplet)\lib</AdditionalLibraryDirectories>
</Link>
</ItemDefinitionGroup>
<ItemGroup>
<ClCompile Include="..\..\..\getopt\getopt.c" />
<ClCompile Include="..\..\..\public\common\TracySocket.cpp" />
<ClCompile Include="..\..\..\public\common\TracyStackFrames.cpp" />
<ClCompile Include="..\..\..\public\common\TracySystem.cpp" />
<ClCompile Include="..\..\..\public\common\tracy_lz4.cpp" />
<ClCompile Include="..\..\..\public\common\tracy_lz4hc.cpp" />
<ClCompile Include="..\..\..\common\TracySocket.cpp" />
<ClCompile Include="..\..\..\common\TracySystem.cpp" />
<ClCompile Include="..\..\..\common\tracy_lz4.cpp" />
<ClCompile Include="..\..\..\common\tracy_lz4hc.cpp" />
<ClCompile Include="..\..\..\server\TracyMemory.cpp" />
<ClCompile Include="..\..\..\server\TracyMmap.cpp" />
<ClCompile Include="..\..\..\server\TracyPrint.cpp" />
<ClCompile Include="..\..\..\server\TracyTaskDispatch.cpp" />
<ClCompile Include="..\..\..\server\TracyTextureCompression.cpp" />
<ClCompile Include="..\..\..\server\TracyThreadCompress.cpp" />
<ClCompile Include="..\..\..\server\TracyWorker.cpp" />
<ClCompile Include="..\..\..\zstd\common\debug.c" />
<ClCompile Include="..\..\..\zstd\common\entropy_common.c" />
<ClCompile Include="..\..\..\zstd\common\error_private.c" />
<ClCompile Include="..\..\..\zstd\common\fse_decompress.c" />
<ClCompile Include="..\..\..\zstd\common\pool.c" />
<ClCompile Include="..\..\..\zstd\common\threading.c" />
<ClCompile Include="..\..\..\zstd\common\xxhash.c" />
<ClCompile Include="..\..\..\zstd\common\zstd_common.c" />
<ClCompile Include="..\..\..\zstd\compress\fse_compress.c" />
<ClCompile Include="..\..\..\zstd\compress\hist.c" />
<ClCompile Include="..\..\..\zstd\compress\huf_compress.c" />
<ClCompile Include="..\..\..\zstd\compress\zstdmt_compress.c" />
<ClCompile Include="..\..\..\zstd\compress\zstd_compress.c" />
<ClCompile Include="..\..\..\zstd\compress\zstd_compress_literals.c" />
<ClCompile Include="..\..\..\zstd\compress\zstd_compress_sequences.c" />
<ClCompile Include="..\..\..\zstd\compress\zstd_compress_superblock.c" />
<ClCompile Include="..\..\..\zstd\compress\zstd_double_fast.c" />
<ClCompile Include="..\..\..\zstd\compress\zstd_fast.c" />
<ClCompile Include="..\..\..\zstd\compress\zstd_lazy.c" />
<ClCompile Include="..\..\..\zstd\compress\zstd_ldm.c" />
<ClCompile Include="..\..\..\zstd\compress\zstd_opt.c" />
<ClCompile Include="..\..\..\zstd\decompress\huf_decompress.c" />
<ClCompile Include="..\..\..\zstd\decompress\zstd_ddict.c" />
<ClCompile Include="..\..\..\zstd\decompress\zstd_decompress.c" />
<ClCompile Include="..\..\..\zstd\decompress\zstd_decompress_block.c" />
<ClCompile Include="..\..\..\zstd\dictBuilder\cover.c" />
<ClCompile Include="..\..\..\zstd\dictBuilder\divsufsort.c" />
<ClCompile Include="..\..\..\zstd\dictBuilder\fastcover.c" />
<ClCompile Include="..\..\..\zstd\dictBuilder\zdict.c" />
<ClCompile Include="..\..\src\capture.cpp" />
<ClCompile Include="..\..\src\getopt.c" />
</ItemGroup>
<ItemGroup>
<ClInclude Include="..\..\..\getopt\getopt.h" />
<ClInclude Include="..\..\..\public\common\TracyAlign.hpp" />
<ClInclude Include="..\..\..\public\common\TracyAlloc.hpp" />
<ClInclude Include="..\..\..\public\common\TracyApi.h" />
<ClInclude Include="..\..\..\public\common\TracyColor.hpp" />
<ClInclude Include="..\..\..\public\common\TracyForceInline.hpp" />
<ClInclude Include="..\..\..\public\common\TracyMutex.hpp" />
<ClInclude Include="..\..\..\public\common\TracyProtocol.hpp" />
<ClInclude Include="..\..\..\public\common\TracyQueue.hpp" />
<ClInclude Include="..\..\..\public\common\TracySocket.hpp" />
<ClInclude Include="..\..\..\public\common\TracyStackFrames.hpp" />
<ClInclude Include="..\..\..\public\common\TracySystem.hpp" />
<ClInclude Include="..\..\..\public\common\TracyUwp.hpp" />
<ClInclude Include="..\..\..\public\common\TracyYield.hpp" />
<ClInclude Include="..\..\..\public\common\tracy_lz4.hpp" />
<ClInclude Include="..\..\..\public\common\tracy_lz4hc.hpp" />
<ClInclude Include="..\..\..\common\TracyAlign.hpp" />
<ClInclude Include="..\..\..\common\TracyAlloc.hpp" />
<ClInclude Include="..\..\..\common\TracyColor.hpp" />
<ClInclude Include="..\..\..\common\TracyForceInline.hpp" />
<ClInclude Include="..\..\..\common\TracyProtocol.hpp" />
<ClInclude Include="..\..\..\common\TracyQueue.hpp" />
<ClInclude Include="..\..\..\common\TracySocket.hpp" />
<ClInclude Include="..\..\..\common\TracySystem.hpp" />
<ClInclude Include="..\..\..\common\tracy_benaphore.h" />
<ClInclude Include="..\..\..\common\tracy_lz4.hpp" />
<ClInclude Include="..\..\..\common\tracy_lz4hc.hpp" />
<ClInclude Include="..\..\..\common\tracy_sema.h" />
<ClInclude Include="..\..\..\server\TracyCharUtil.hpp" />
<ClInclude Include="..\..\..\server\TracyEvent.hpp" />
<ClInclude Include="..\..\..\server\TracyFileRead.hpp" />
<ClInclude Include="..\..\..\server\TracyFileWrite.hpp" />
<ClInclude Include="..\..\..\server\TracyMemory.hpp" />
<ClInclude Include="..\..\..\server\TracyMmap.hpp" />
<ClInclude Include="..\..\..\server\TracyPopcnt.hpp" />
<ClInclude Include="..\..\..\server\TracyPrint.hpp" />
<ClInclude Include="..\..\..\server\TracySlab.hpp" />
<ClInclude Include="..\..\..\server\TracyTaskDispatch.hpp" />
<ClInclude Include="..\..\..\server\TracyTextureCompression.hpp" />
<ClInclude Include="..\..\..\server\TracyThreadCompress.hpp" />
<ClInclude Include="..\..\..\server\TracyVector.hpp" />
<ClInclude Include="..\..\..\server\TracyWorker.hpp" />
<ClInclude Include="..\..\..\zstd\common\bitstream.h" />
<ClInclude Include="..\..\..\zstd\common\compiler.h" />
<ClInclude Include="..\..\..\zstd\common\cpu.h" />
<ClInclude Include="..\..\..\zstd\common\debug.h" />
<ClInclude Include="..\..\..\zstd\common\error_private.h" />
<ClInclude Include="..\..\..\zstd\common\fse.h" />
<ClInclude Include="..\..\..\zstd\common\huf.h" />
<ClInclude Include="..\..\..\zstd\common\mem.h" />
<ClInclude Include="..\..\..\zstd\common\pool.h" />
<ClInclude Include="..\..\..\zstd\common\portability_macros.h" />
<ClInclude Include="..\..\..\zstd\common\threading.h" />
<ClInclude Include="..\..\..\zstd\common\xxhash.h" />
<ClInclude Include="..\..\..\zstd\common\zstd_deps.h" />
<ClInclude Include="..\..\..\zstd\common\zstd_internal.h" />
<ClInclude Include="..\..\..\zstd\common\zstd_trace.h" />
<ClInclude Include="..\..\..\zstd\compress\clevels.h" />
<ClInclude Include="..\..\..\zstd\compress\hist.h" />
<ClInclude Include="..\..\..\zstd\compress\zstdmt_compress.h" />
<ClInclude Include="..\..\..\zstd\compress\zstd_compress_internal.h" />
<ClInclude Include="..\..\..\zstd\compress\zstd_compress_literals.h" />
<ClInclude Include="..\..\..\zstd\compress\zstd_compress_sequences.h" />
<ClInclude Include="..\..\..\zstd\compress\zstd_compress_superblock.h" />
<ClInclude Include="..\..\..\zstd\compress\zstd_cwksp.h" />
<ClInclude Include="..\..\..\zstd\compress\zstd_double_fast.h" />
<ClInclude Include="..\..\..\zstd\compress\zstd_fast.h" />
<ClInclude Include="..\..\..\zstd\compress\zstd_lazy.h" />
<ClInclude Include="..\..\..\zstd\compress\zstd_ldm.h" />
<ClInclude Include="..\..\..\zstd\compress\zstd_ldm_geartab.h" />
<ClInclude Include="..\..\..\zstd\compress\zstd_opt.h" />
<ClInclude Include="..\..\..\zstd\decompress\zstd_ddict.h" />
<ClInclude Include="..\..\..\zstd\decompress\zstd_decompress_block.h" />
<ClInclude Include="..\..\..\zstd\decompress\zstd_decompress_internal.h" />
<ClInclude Include="..\..\..\zstd\dictBuilder\cover.h" />
<ClInclude Include="..\..\..\zstd\dictBuilder\divsufsort.h" />
<ClInclude Include="..\..\..\zstd\zdict.h" />
<ClInclude Include="..\..\..\zstd\zstd.h" />
<ClInclude Include="..\..\..\zstd\zstd_errors.h" />
</ItemGroup>
<ItemGroup>
<None Include="..\..\..\zstd\decompress\huf_decompress_amd64.S" />
<ClInclude Include="..\..\..\server\tracy_flat_hash_map.hpp" />
<ClInclude Include="..\..\src\getopt.h" />
</ItemGroup>
<Import Project="$(VCTargetsPath)\Microsoft.Cpp.targets" />
<ImportGroup Label="ExtensionTargets">

View File

@@ -10,26 +10,17 @@
<Filter Include="common">
<UniqueIdentifier>{e39d3623-47cd-4752-8da9-3ea324f964c1}</UniqueIdentifier>
</Filter>
<Filter Include="getopt">
<UniqueIdentifier>{ee9737d2-69c7-44da-b9c7-539d18f9d4b4}</UniqueIdentifier>
</Filter>
<Filter Include="zstd">
<UniqueIdentifier>{f201463b-5e69-46fe-bdfc-1b5eed86c7f7}</UniqueIdentifier>
</Filter>
<Filter Include="zstd\common">
<UniqueIdentifier>{7e93ae33-6543-4bca-b05b-50818dbf24cc}</UniqueIdentifier>
</Filter>
<Filter Include="zstd\compress">
<UniqueIdentifier>{3b0c32f5-9efb-4503-9394-5ab95909fb1c}</UniqueIdentifier>
</Filter>
<Filter Include="zstd\decompress">
<UniqueIdentifier>{c1f99170-d904-4af1-8010-0a3ded5736c8}</UniqueIdentifier>
</Filter>
<Filter Include="zstd\dictBuilder">
<UniqueIdentifier>{456e6786-ea57-42b8-ae38-829cd2d918bd}</UniqueIdentifier>
</Filter>
</ItemGroup>
<ItemGroup>
<ClCompile Include="..\..\..\common\tracy_lz4.cpp">
<Filter>common</Filter>
</ClCompile>
<ClCompile Include="..\..\..\common\TracySocket.cpp">
<Filter>common</Filter>
</ClCompile>
<ClCompile Include="..\..\..\common\TracySystem.cpp">
<Filter>common</Filter>
</ClCompile>
<ClCompile Include="..\..\..\server\TracyMemory.cpp">
<Filter>server</Filter>
</ClCompile>
@@ -39,128 +30,41 @@
<ClCompile Include="..\..\src\capture.cpp">
<Filter>src</Filter>
</ClCompile>
<ClCompile Include="..\..\..\server\TracyPrint.cpp">
<Filter>server</Filter>
<ClCompile Include="..\..\src\getopt.c">
<Filter>src</Filter>
</ClCompile>
<ClCompile Include="..\..\..\server\TracyThreadCompress.cpp">
<Filter>server</Filter>
</ClCompile>
<ClCompile Include="..\..\..\server\TracyTaskDispatch.cpp">
<Filter>server</Filter>
</ClCompile>
<ClCompile Include="..\..\..\server\TracyMmap.cpp">
<Filter>server</Filter>
</ClCompile>
<ClCompile Include="..\..\..\server\TracyTextureCompression.cpp">
<Filter>server</Filter>
</ClCompile>
<ClCompile Include="..\..\..\getopt\getopt.c">
<Filter>getopt</Filter>
</ClCompile>
<ClCompile Include="..\..\..\zstd\common\debug.c">
<Filter>zstd\common</Filter>
</ClCompile>
<ClCompile Include="..\..\..\zstd\common\entropy_common.c">
<Filter>zstd\common</Filter>
</ClCompile>
<ClCompile Include="..\..\..\zstd\common\error_private.c">
<Filter>zstd\common</Filter>
</ClCompile>
<ClCompile Include="..\..\..\zstd\common\fse_decompress.c">
<Filter>zstd\common</Filter>
</ClCompile>
<ClCompile Include="..\..\..\zstd\common\pool.c">
<Filter>zstd\common</Filter>
</ClCompile>
<ClCompile Include="..\..\..\zstd\common\threading.c">
<Filter>zstd\common</Filter>
</ClCompile>
<ClCompile Include="..\..\..\zstd\common\xxhash.c">
<Filter>zstd\common</Filter>
</ClCompile>
<ClCompile Include="..\..\..\zstd\common\zstd_common.c">
<Filter>zstd\common</Filter>
</ClCompile>
<ClCompile Include="..\..\..\zstd\compress\fse_compress.c">
<Filter>zstd\compress</Filter>
</ClCompile>
<ClCompile Include="..\..\..\zstd\compress\hist.c">
<Filter>zstd\compress</Filter>
</ClCompile>
<ClCompile Include="..\..\..\zstd\compress\huf_compress.c">
<Filter>zstd\compress</Filter>
</ClCompile>
<ClCompile Include="..\..\..\zstd\compress\zstd_compress.c">
<Filter>zstd\compress</Filter>
</ClCompile>
<ClCompile Include="..\..\..\zstd\compress\zstd_compress_literals.c">
<Filter>zstd\compress</Filter>
</ClCompile>
<ClCompile Include="..\..\..\zstd\compress\zstd_compress_sequences.c">
<Filter>zstd\compress</Filter>
</ClCompile>
<ClCompile Include="..\..\..\zstd\compress\zstd_compress_superblock.c">
<Filter>zstd\compress</Filter>
</ClCompile>
<ClCompile Include="..\..\..\zstd\compress\zstd_double_fast.c">
<Filter>zstd\compress</Filter>
</ClCompile>
<ClCompile Include="..\..\..\zstd\compress\zstd_fast.c">
<Filter>zstd\compress</Filter>
</ClCompile>
<ClCompile Include="..\..\..\zstd\compress\zstd_lazy.c">
<Filter>zstd\compress</Filter>
</ClCompile>
<ClCompile Include="..\..\..\zstd\compress\zstd_ldm.c">
<Filter>zstd\compress</Filter>
</ClCompile>
<ClCompile Include="..\..\..\zstd\compress\zstd_opt.c">
<Filter>zstd\compress</Filter>
</ClCompile>
<ClCompile Include="..\..\..\zstd\compress\zstdmt_compress.c">
<Filter>zstd\compress</Filter>
</ClCompile>
<ClCompile Include="..\..\..\zstd\decompress\huf_decompress.c">
<Filter>zstd\decompress</Filter>
</ClCompile>
<ClCompile Include="..\..\..\zstd\decompress\zstd_ddict.c">
<Filter>zstd\decompress</Filter>
</ClCompile>
<ClCompile Include="..\..\..\zstd\decompress\zstd_decompress.c">
<Filter>zstd\decompress</Filter>
</ClCompile>
<ClCompile Include="..\..\..\zstd\decompress\zstd_decompress_block.c">
<Filter>zstd\decompress</Filter>
</ClCompile>
<ClCompile Include="..\..\..\zstd\dictBuilder\cover.c">
<Filter>zstd\dictBuilder</Filter>
</ClCompile>
<ClCompile Include="..\..\..\zstd\dictBuilder\divsufsort.c">
<Filter>zstd\dictBuilder</Filter>
</ClCompile>
<ClCompile Include="..\..\..\zstd\dictBuilder\fastcover.c">
<Filter>zstd\dictBuilder</Filter>
</ClCompile>
<ClCompile Include="..\..\..\zstd\dictBuilder\zdict.c">
<Filter>zstd\dictBuilder</Filter>
</ClCompile>
<ClCompile Include="..\..\..\public\common\tracy_lz4.cpp">
<Filter>common</Filter>
</ClCompile>
<ClCompile Include="..\..\..\public\common\tracy_lz4hc.cpp">
<Filter>common</Filter>
</ClCompile>
<ClCompile Include="..\..\..\public\common\TracySocket.cpp">
<Filter>common</Filter>
</ClCompile>
<ClCompile Include="..\..\..\public\common\TracyStackFrames.cpp">
<Filter>common</Filter>
</ClCompile>
<ClCompile Include="..\..\..\public\common\TracySystem.cpp">
<ClCompile Include="..\..\..\common\tracy_lz4hc.cpp">
<Filter>common</Filter>
</ClCompile>
</ItemGroup>
<ItemGroup>
<ClInclude Include="..\..\..\common\tracy_lz4.hpp">
<Filter>common</Filter>
</ClInclude>
<ClInclude Include="..\..\..\common\TracyAlloc.hpp">
<Filter>common</Filter>
</ClInclude>
<ClInclude Include="..\..\..\common\TracyColor.hpp">
<Filter>common</Filter>
</ClInclude>
<ClInclude Include="..\..\..\common\TracyForceInline.hpp">
<Filter>common</Filter>
</ClInclude>
<ClInclude Include="..\..\..\common\TracyProtocol.hpp">
<Filter>common</Filter>
</ClInclude>
<ClInclude Include="..\..\..\common\TracyQueue.hpp">
<Filter>common</Filter>
</ClInclude>
<ClInclude Include="..\..\..\common\TracySocket.hpp">
<Filter>common</Filter>
</ClInclude>
<ClInclude Include="..\..\..\common\TracySystem.hpp">
<Filter>common</Filter>
</ClInclude>
<ClInclude Include="..\..\..\server\tracy_flat_hash_map.hpp">
<Filter>server</Filter>
</ClInclude>
<ClInclude Include="..\..\..\server\TracyCharUtil.hpp">
<Filter>server</Filter>
</ClInclude>
@@ -185,187 +89,20 @@
<ClInclude Include="..\..\..\server\TracyWorker.hpp">
<Filter>server</Filter>
</ClInclude>
<ClInclude Include="..\..\..\server\TracyPrint.hpp">
<Filter>server</Filter>
<ClInclude Include="..\..\src\getopt.h">
<Filter>src</Filter>
</ClInclude>
<ClInclude Include="..\..\..\server\TracyThreadCompress.hpp">
<Filter>server</Filter>
</ClInclude>
<ClInclude Include="..\..\..\server\TracyTaskDispatch.hpp">
<Filter>server</Filter>
</ClInclude>
<ClInclude Include="..\..\..\server\TracyFileRead.hpp">
<Filter>server</Filter>
</ClInclude>
<ClInclude Include="..\..\..\server\TracyMmap.hpp">
<Filter>server</Filter>
</ClInclude>
<ClInclude Include="..\..\..\server\TracyTextureCompression.hpp">
<Filter>server</Filter>
</ClInclude>
<ClInclude Include="..\..\..\getopt\getopt.h">
<Filter>getopt</Filter>
</ClInclude>
<ClInclude Include="..\..\..\zstd\zstd.h">
<Filter>zstd</Filter>
</ClInclude>
<ClInclude Include="..\..\..\zstd\zstd_errors.h">
<Filter>zstd</Filter>
</ClInclude>
<ClInclude Include="..\..\..\zstd\common\bitstream.h">
<Filter>zstd\common</Filter>
</ClInclude>
<ClInclude Include="..\..\..\zstd\common\compiler.h">
<Filter>zstd\common</Filter>
</ClInclude>
<ClInclude Include="..\..\..\zstd\common\cpu.h">
<Filter>zstd\common</Filter>
</ClInclude>
<ClInclude Include="..\..\..\zstd\common\debug.h">
<Filter>zstd\common</Filter>
</ClInclude>
<ClInclude Include="..\..\..\zstd\common\error_private.h">
<Filter>zstd\common</Filter>
</ClInclude>
<ClInclude Include="..\..\..\zstd\common\fse.h">
<Filter>zstd\common</Filter>
</ClInclude>
<ClInclude Include="..\..\..\zstd\common\huf.h">
<Filter>zstd\common</Filter>
</ClInclude>
<ClInclude Include="..\..\..\zstd\common\mem.h">
<Filter>zstd\common</Filter>
</ClInclude>
<ClInclude Include="..\..\..\zstd\common\pool.h">
<Filter>zstd\common</Filter>
</ClInclude>
<ClInclude Include="..\..\..\zstd\common\threading.h">
<Filter>zstd\common</Filter>
</ClInclude>
<ClInclude Include="..\..\..\zstd\common\xxhash.h">
<Filter>zstd\common</Filter>
</ClInclude>
<ClInclude Include="..\..\..\zstd\common\zstd_deps.h">
<Filter>zstd\common</Filter>
</ClInclude>
<ClInclude Include="..\..\..\zstd\common\zstd_internal.h">
<Filter>zstd\common</Filter>
</ClInclude>
<ClInclude Include="..\..\..\zstd\common\zstd_trace.h">
<Filter>zstd\common</Filter>
</ClInclude>
<ClInclude Include="..\..\..\zstd\compress\hist.h">
<Filter>zstd\compress</Filter>
</ClInclude>
<ClInclude Include="..\..\..\zstd\compress\zstd_compress_internal.h">
<Filter>zstd\compress</Filter>
</ClInclude>
<ClInclude Include="..\..\..\zstd\compress\zstd_compress_literals.h">
<Filter>zstd\compress</Filter>
</ClInclude>
<ClInclude Include="..\..\..\zstd\compress\zstd_compress_sequences.h">
<Filter>zstd\compress</Filter>
</ClInclude>
<ClInclude Include="..\..\..\zstd\compress\zstd_compress_superblock.h">
<Filter>zstd\compress</Filter>
</ClInclude>
<ClInclude Include="..\..\..\zstd\compress\zstd_cwksp.h">
<Filter>zstd\compress</Filter>
</ClInclude>
<ClInclude Include="..\..\..\zstd\compress\zstd_double_fast.h">
<Filter>zstd\compress</Filter>
</ClInclude>
<ClInclude Include="..\..\..\zstd\compress\zstd_fast.h">
<Filter>zstd\compress</Filter>
</ClInclude>
<ClInclude Include="..\..\..\zstd\compress\zstd_lazy.h">
<Filter>zstd\compress</Filter>
</ClInclude>
<ClInclude Include="..\..\..\zstd\compress\zstd_ldm.h">
<Filter>zstd\compress</Filter>
</ClInclude>
<ClInclude Include="..\..\..\zstd\compress\zstd_ldm_geartab.h">
<Filter>zstd\compress</Filter>
</ClInclude>
<ClInclude Include="..\..\..\zstd\compress\zstd_opt.h">
<Filter>zstd\compress</Filter>
</ClInclude>
<ClInclude Include="..\..\..\zstd\compress\zstdmt_compress.h">
<Filter>zstd\compress</Filter>
</ClInclude>
<ClInclude Include="..\..\..\zstd\decompress\zstd_ddict.h">
<Filter>zstd\decompress</Filter>
</ClInclude>
<ClInclude Include="..\..\..\zstd\decompress\zstd_decompress_block.h">
<Filter>zstd\decompress</Filter>
</ClInclude>
<ClInclude Include="..\..\..\zstd\decompress\zstd_decompress_internal.h">
<Filter>zstd\decompress</Filter>
</ClInclude>
<ClInclude Include="..\..\..\zstd\dictBuilder\cover.h">
<Filter>zstd\dictBuilder</Filter>
</ClInclude>
<ClInclude Include="..\..\..\zstd\dictBuilder\divsufsort.h">
<Filter>zstd\dictBuilder</Filter>
</ClInclude>
<ClInclude Include="..\..\..\zstd\zdict.h">
<Filter>zstd</Filter>
</ClInclude>
<ClInclude Include="..\..\..\zstd\common\portability_macros.h">
<Filter>zstd\common</Filter>
</ClInclude>
<ClInclude Include="..\..\..\zstd\compress\clevels.h">
<Filter>zstd\compress</Filter>
</ClInclude>
<ClInclude Include="..\..\..\public\common\tracy_lz4.hpp">
<ClInclude Include="..\..\..\common\TracyAlign.hpp">
<Filter>common</Filter>
</ClInclude>
<ClInclude Include="..\..\..\public\common\tracy_lz4hc.hpp">
<ClInclude Include="..\..\..\common\tracy_benaphore.h">
<Filter>common</Filter>
</ClInclude>
<ClInclude Include="..\..\..\public\common\TracyAlign.hpp">
<ClInclude Include="..\..\..\common\tracy_sema.h">
<Filter>common</Filter>
</ClInclude>
<ClInclude Include="..\..\..\public\common\TracyAlloc.hpp">
<ClInclude Include="..\..\..\common\tracy_lz4hc.hpp">
<Filter>common</Filter>
</ClInclude>
<ClInclude Include="..\..\..\public\common\TracyApi.h">
<Filter>common</Filter>
</ClInclude>
<ClInclude Include="..\..\..\public\common\TracyColor.hpp">
<Filter>common</Filter>
</ClInclude>
<ClInclude Include="..\..\..\public\common\TracyForceInline.hpp">
<Filter>common</Filter>
</ClInclude>
<ClInclude Include="..\..\..\public\common\TracyMutex.hpp">
<Filter>common</Filter>
</ClInclude>
<ClInclude Include="..\..\..\public\common\TracyProtocol.hpp">
<Filter>common</Filter>
</ClInclude>
<ClInclude Include="..\..\..\public\common\TracyQueue.hpp">
<Filter>common</Filter>
</ClInclude>
<ClInclude Include="..\..\..\public\common\TracySocket.hpp">
<Filter>common</Filter>
</ClInclude>
<ClInclude Include="..\..\..\public\common\TracyStackFrames.hpp">
<Filter>common</Filter>
</ClInclude>
<ClInclude Include="..\..\..\public\common\TracySystem.hpp">
<Filter>common</Filter>
</ClInclude>
<ClInclude Include="..\..\..\public\common\TracyUwp.hpp">
<Filter>common</Filter>
</ClInclude>
<ClInclude Include="..\..\..\public\common\TracyYield.hpp">
<Filter>common</Filter>
</ClInclude>
</ItemGroup>
<ItemGroup>
<None Include="..\..\..\zstd\decompress\huf_decompress_amd64.S">
<Filter>zstd\decompress</Filter>
</None>
</ItemGroup>
</Project>

View File

@@ -1,96 +1,104 @@
#ifdef _WIN32
# include <windows.h>
# include <io.h>
#else
# include <unistd.h>
#endif
#include <atomic>
#include <chrono>
#include <inttypes.h>
#include <mutex>
#include <signal.h>
#include <stdarg.h>
#include <stdint.h>
#include <stdio.h>
#include <stdlib.h>
#include <sys/stat.h>
#include "../../public/common/TracyProtocol.hpp"
#include "../../public/common/TracyStackFrames.hpp"
#include "../../server/TracyFileWrite.hpp"
#include "../../server/TracyMemory.hpp"
#include "../../server/TracyPrint.hpp"
#include "../../server/TracyWorker.hpp"
#include "getopt.h"
#ifdef _WIN32
# include "../../getopt/getopt.h"
#endif
// This atomic is written by a signal handler (SigInt). Traditionally that would
// have had to be `volatile sig_atomic_t`, and annoyingly, `bool` was
// technically not allowed there, even though in practice it would work.
// The good thing with C++11 atomics is that we can use atomic<bool> instead
// here and be on the actually supported path.
static std::atomic<bool> s_disconnect { false };
void SigInt( int )
static const char* TimeToString( int64_t ns )
{
// Relaxed order is closest to a traditional `volatile` write.
// We don't need stronger ordering since this signal handler doesn't do
// anything else that would need to be ordered relatively to this.
s_disconnect.store(true, std::memory_order_relaxed);
}
enum { Pool = 8 };
static char bufpool[Pool][64];
static int bufsel = 0;
char* buf = bufpool[bufsel];
bufsel = ( bufsel + 1 ) % Pool;
static bool s_isStdoutATerminal = false;
void InitIsStdoutATerminal() {
#ifdef _WIN32
s_isStdoutATerminal = _isatty( fileno( stdout ) );
#else
s_isStdoutATerminal = isatty( fileno( stdout ) );
#endif
}
bool IsStdoutATerminal() { return s_isStdoutATerminal; }
#define ANSI_RESET "\033[0m"
#define ANSI_BOLD "\033[1m"
#define ANSI_BLACK "\033[30m"
#define ANSI_RED "\033[31m"
#define ANSI_GREEN "\033[32m"
#define ANSI_YELLOW "\033[33m"
#define ANSI_MAGENTA "\033[35m"
#define ANSI_CYAN "\033[36m"
#define ANSI_ERASE_LINE "\033[2K"
// Like printf, but if stdout is a terminal, prepends the output with
// the given `ansiEscape` and appends ANSI_RESET.
void AnsiPrintf( const char* ansiEscape, const char* format, ... ) {
if( IsStdoutATerminal() )
const char* sign = "";
if( ns < 0 )
{
// Prepend ansiEscape and append ANSI_RESET.
char buf[256];
va_list args;
va_start( args, format );
vsnprintf( buf, sizeof buf, format, args );
va_end( args );
printf( "%s%s" ANSI_RESET, ansiEscape, buf );
sign = "-";
ns = -ns;
}
if( ns < 1000 )
{
sprintf( buf, "%s%" PRIi64 " ns", sign, ns );
}
else if( ns < 1000ll * 1000 )
{
sprintf( buf, "%s%.2f us", sign, ns / 1000. );
}
else if( ns < 1000ll * 1000 * 1000 )
{
sprintf( buf, "%s%.2f ms", sign, ns / ( 1000. * 1000. ) );
}
else if( ns < 1000ll * 1000 * 1000 * 60 )
{
sprintf( buf, "%s%.2f s", sign, ns / ( 1000. * 1000. * 1000. ) );
}
else
{
// Just a normal printf.
va_list args;
va_start( args, format );
vfprintf( stdout, format, args );
va_end( args );
const auto m = int64_t( ns / ( 1000ll * 1000 * 1000 * 60 ) );
const auto s = int64_t( ns - m * ( 1000ll * 1000 * 1000 * 60 ) );
sprintf( buf, "%s%" PRIi64 ":%04.1f", sign, m, s / ( 1000. * 1000. * 1000. ) );
}
return buf;
}
[[noreturn]] void Usage()
static const char* RealToString( double val, bool separator )
{
printf( "Usage: capture -o output.tracy [-a address] [-p port] [-f] [-s seconds]\n" );
enum { Pool = 8 };
static char bufpool[Pool][64];
static int bufsel = 0;
char* buf = bufpool[bufsel];
bufsel = ( bufsel + 1 ) % Pool;
sprintf( buf, "%f", val );
auto ptr = buf;
if( *ptr == '-' ) ptr++;
const auto vbegin = ptr;
if( separator )
{
while( *ptr != '\0' && *ptr != ',' && *ptr != '.' ) ptr++;
auto end = ptr;
while( *end != '\0' ) end++;
auto sz = end - ptr;
while( ptr - vbegin > 3 )
{
ptr -= 3;
memmove( ptr+1, ptr, sz );
*ptr = ',';
sz += 4;
}
}
while( *ptr != '\0' && *ptr != ',' && *ptr != '.' ) ptr++;
if( *ptr == '\0' ) return buf;
while( *ptr != '\0' ) ptr++;
ptr--;
while( *ptr == '0' && *ptr != ',' && *ptr != '.' ) ptr--;
if( *ptr != '.' && *ptr != ',' ) ptr++;
*ptr = '\0';
return buf;
}
void Usage()
{
printf( "Usage: capture -a address -o output.tracy\n" );
exit( 1 );
}
@@ -104,16 +112,11 @@ int main( int argc, char** argv )
}
#endif
InitIsStdoutATerminal();
bool overwrite = false;
const char* address = "127.0.0.1";
const char* address = nullptr;
const char* output = nullptr;
int port = 8086;
int seconds = -1;
int c;
while( ( c = getopt( argc, argv, "a:o:p:fs:" ) ) != -1 )
while( ( c = getopt( argc, argv, "a:o:" ) ) != -1 )
{
switch( c )
{
@@ -123,15 +126,6 @@ int main( int argc, char** argv )
case 'o':
output = optarg;
break;
case 'p':
port = atoi( optarg );
break;
case 'f':
overwrite = true;
break;
case 's':
seconds = atoi (optarg);
break;
default:
Usage();
break;
@@ -140,213 +134,46 @@ int main( int argc, char** argv )
if( !address || !output ) Usage();
struct stat st;
if( stat( output, &st ) == 0 && !overwrite )
{
printf( "Output file %s already exists! Use -f to force overwrite.\n", output );
return 4;
}
FILE* test = fopen( output, "wb" );
if( !test )
{
printf( "Cannot open output file %s for writing!\n", output );
return 5;
}
fclose( test );
unlink( output );
printf( "Connecting to %s:%i...", address, port );
printf( "Connecting to %s...", address );
fflush( stdout );
tracy::Worker worker( address, port );
while( !worker.HasData() )
{
const auto handshake = worker.GetHandshakeStatus();
if( handshake == tracy::HandshakeProtocolMismatch )
{
printf( "\nThe client you are trying to connect to uses incompatible protocol version.\nMake sure you are using the same Tracy version on both client and server.\n" );
return 1;
}
if( handshake == tracy::HandshakeNotAvailable )
{
printf( "\nThe client you are trying to connect to is no longer able to sent profiling data,\nbecause another server was already connected to it.\nYou can do the following:\n\n 1. Restart the client application.\n 2. Rebuild the client application with on-demand mode enabled.\n" );
return 2;
}
if( handshake == tracy::HandshakeDropped )
{
printf( "\nThe client you are trying to connect to has disconnected during the initial\nconnection handshake. Please check your network configuration.\n" );
return 3;
}
std::this_thread::sleep_for( std::chrono::milliseconds( 100 ) );
}
printf( "\nQueue delay: %s\nTimer resolution: %s\n", tracy::TimeToString( worker.GetDelay() ), tracy::TimeToString( worker.GetResolution() ) );
tracy::Worker worker( address );
while( !worker.HasData() ) std::this_thread::sleep_for( std::chrono::milliseconds( 100 ) );
printf( "\nQueue delay: %s\nTimer resolution: %s\n", TimeToString( worker.GetDelay() ), TimeToString( worker.GetResolution() ) );
#ifdef _WIN32
signal( SIGINT, SigInt );
#else
struct sigaction sigint, oldsigint;
memset( &sigint, 0, sizeof( sigint ) );
sigint.sa_handler = SigInt;
sigaction( SIGINT, &sigint, &oldsigint );
#endif
const auto firstTime = worker.GetFirstTime();
auto& lock = worker.GetMbpsDataLock();
const auto t0 = std::chrono::high_resolution_clock::now();
while( worker.IsConnected() )
{
// Relaxed order is sufficient here because `s_disconnect` is only ever
// set by this thread or by the SigInt handler, and that handler does
// nothing else than storing `s_disconnect`.
if( s_disconnect.load( std::memory_order_relaxed ) )
{
worker.Disconnect();
// Relaxed order is sufficient because only this thread ever reads
// this value.
s_disconnect.store(false, std::memory_order_relaxed );
break;
}
lock.lock();
const auto mbps = worker.GetMbpsData().back();
const auto compRatio = worker.GetCompRatio();
const auto netTotal = worker.GetDataTransferred();
lock.unlock();
// Output progress info only if destination is a TTY to avoid bloating
// log files (so this is not just about usage of ANSI color codes).
if( IsStdoutATerminal() )
if( mbps < 0.1f )
{
const char* unit = "Mbps";
float unitsPerMbps = 1.f;
if( mbps < 0.1f )
{
unit = "Kbps";
unitsPerMbps = 1000.f;
}
AnsiPrintf( ANSI_ERASE_LINE ANSI_CYAN ANSI_BOLD, "\r%7.2f %s", mbps * unitsPerMbps, unit );
printf( " /");
AnsiPrintf( ANSI_CYAN ANSI_BOLD, "%5.1f%%", compRatio * 100.f );
printf( " =");
AnsiPrintf( ANSI_YELLOW ANSI_BOLD, "%7.2f Mbps", mbps / compRatio );
printf( " | ");
AnsiPrintf( ANSI_YELLOW, "Tx: ");
AnsiPrintf( ANSI_GREEN, "%s", tracy::MemSizeToString( netTotal ) );
printf( " | ");
AnsiPrintf( ANSI_RED ANSI_BOLD, "%s", tracy::MemSizeToString( tracy::memUsage ) );
printf( " | ");
AnsiPrintf( ANSI_RED, "%s", tracy::TimeToString( worker.GetLastTime() - firstTime ) );
fflush( stdout );
printf( "\33[2K\r\033[36;1m%7.2f Kbps", mbps * 1000.f );
}
else
{
printf( "\33[2K\r\033[36;1m%7.2f Mbps", mbps );
}
printf( " \033[0m| Ratio: \033[36;1m%5.1f%% \033[0m| Real: \033[33;1m%7.2f Mbps \033[0m| Mem: \033[31;1m%.2f MB\033[0m", compRatio * 100.f, mbps / compRatio, tracy::memUsage.load( std::memory_order_relaxed ) / ( 1024.f * 1024.f ) );
fflush( stdout );
std::this_thread::sleep_for( std::chrono::milliseconds( 100 ) );
if( seconds != -1 )
{
const auto dur = std::chrono::high_resolution_clock::now() - t0;
if( std::chrono::duration_cast<std::chrono::seconds>(dur).count() >= seconds )
{
// Relaxed order is sufficient because only this thread ever reads
// this value.
s_disconnect.store(true, std::memory_order_relaxed );
}
}
}
const auto t1 = std::chrono::high_resolution_clock::now();
const auto& failure = worker.GetFailureType();
if( failure != tracy::Worker::Failure::None )
{
AnsiPrintf( ANSI_RED ANSI_BOLD, "\nInstrumentation failure: %s", tracy::Worker::GetFailureString( failure ) );
auto& fd = worker.GetFailureData();
if( !fd.message.empty() )
{
printf( "\nContext: %s", fd.message.c_str() );
}
if( fd.callstack != 0 )
{
AnsiPrintf( ANSI_BOLD, "\n%sFailure callstack:%s\n" );
auto& cs = worker.GetCallstack( fd.callstack );
int fidx = 0;
for( auto& entry : cs )
{
auto frameData = worker.GetCallstackFrame( entry );
if( !frameData )
{
printf( "%3i. %p\n", fidx++, (void*)worker.GetCanonicalPointer( entry ) );
}
else
{
const auto fsz = frameData->size;
for( uint8_t f=0; f<fsz; f++ )
{
const auto& frame = frameData->data[f];
auto txt = worker.GetString( frame.name );
if( fidx == 0 && f != fsz-1 )
{
auto test = tracy::s_tracyStackFrames;
bool match = false;
do
{
if( strcmp( txt, *test ) == 0 )
{
match = true;
break;
}
}
while( *++test );
if( match ) continue;
}
if( f == fsz-1 )
{
printf( "%3i. ", fidx++ );
}
else
{
AnsiPrintf( ANSI_BLACK ANSI_BOLD, "inl. " );
}
AnsiPrintf( ANSI_CYAN, "%s ", txt );
txt = worker.GetString( frame.file );
if( frame.line == 0 )
{
AnsiPrintf( ANSI_YELLOW, "(%s)", txt );
}
else
{
AnsiPrintf( ANSI_YELLOW, "(%s:%" PRIu32 ")", txt, frame.line );
}
if( frameData->imageName.Active() )
{
AnsiPrintf( ANSI_MAGENTA, " %s\n", worker.GetString( frameData->imageName ) );
}
else
{
printf( "\n" );
}
}
}
}
}
}
printf( "\nFrames: %" PRIu64 "\nTime span: %s\nZones: %s\nElapsed time: %s\nSaving trace...",
worker.GetFrameCount( *worker.GetFramesBase() ), tracy::TimeToString( worker.GetLastTime() - firstTime ), tracy::RealToString( worker.GetZoneCount() ),
tracy::TimeToString( std::chrono::duration_cast<std::chrono::nanoseconds>( t1 - t0 ).count() ) );
printf( "\nFrames: %" PRIu64 "\nTime span: %s\nZones: %s\nSaving trace...", worker.GetFrameCount( *worker.GetFramesBase() ), TimeToString( worker.GetLastTime() - worker.GetTimeBegin() ), RealToString( worker.GetZoneCount(), true ) );
fflush( stdout );
auto f = std::unique_ptr<tracy::FileWrite>( tracy::FileWrite::Open( output ) );
if( f )
{
worker.Write( *f, false );
AnsiPrintf( ANSI_GREEN ANSI_BOLD, " done!\n" );
f->Finish();
const auto stats = f->GetCompressionStatistics();
printf( "Trace size %s (%.2f%% ratio)\n", tracy::MemSizeToString( stats.second ), 100.f * stats.second / stats.first );
worker.Write( *f );
printf( " \033[32;1mdone!\033[0m\n" );
}
else
{
AnsiPrintf( ANSI_RED ANSI_BOLD, " failed!\n");
printf( " \033[31;1failed!\033[0m\n" );
}
return 0;

View File

@@ -1,5 +1,5 @@
/*******************************************************************************
* Copyright (c) 2012-2023, Kim Grasman <kim.grasman@gmail.com>
* Copyright (c) 2012-2017, Kim Grasman <kim.grasman@gmail.com>
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
@@ -159,7 +159,6 @@ int getopt_long(int argc, char* const argv[], const char* optstring,
const struct option* match = NULL;
int num_matches = 0;
size_t argument_name_length = 0;
size_t option_length = 0;
const char* current_argument = NULL;
int retval = -1;
@@ -176,16 +175,6 @@ int getopt_long(int argc, char* const argv[], const char* optstring,
current_argument = argv[optind] + 2;
argument_name_length = strcspn(current_argument, "=");
for (; o->name; ++o) {
/* Check for exact match first. */
option_length = strlen(o->name);
if (option_length == argument_name_length &&
strncmp(o->name, current_argument, option_length) == 0) {
match = o;
num_matches = 1;
break;
}
/* If not exact, count the number of abbreviated matches. */
if (strncmp(o->name, current_argument, argument_name_length) == 0) {
match = o;
++num_matches;

View File

@@ -1,5 +1,5 @@
/*******************************************************************************
* Copyright (c) 2012-2023, Kim Grasman <kim.grasman@gmail.com>
* Copyright (c) 2012-2017, Kim Grasman <kim.grasman@gmail.com>
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without

253
client/TracyCallstack.cpp Normal file
View File

@@ -0,0 +1,253 @@
#include <stdio.h>
#include "TracyCallstack.hpp"
#ifdef TRACY_HAS_CALLSTACK
#if TRACY_HAS_CALLSTACK == 1
# include <windows.h>
# pragma warning( push )
# pragma warning( disable : 4091 )
# include <dbghelp.h>
# pragma warning( pop )
#elif TRACY_HAS_CALLSTACK >= 2
# include <dlfcn.h>
# include <cxxabi.h>
#endif
namespace tracy
{
#if TRACY_HAS_CALLSTACK == 1
extern "C" t_RtlWalkFrameChain RtlWalkFrameChain = 0;
void InitCallstack()
{
#ifdef UNICODE
RtlWalkFrameChain = (t_RtlWalkFrameChain)GetProcAddress( GetModuleHandle( L"ntdll.dll" ), "RtlWalkFrameChain" );
#else
RtlWalkFrameChain = (t_RtlWalkFrameChain)GetProcAddress( GetModuleHandle( "ntdll.dll" ), "RtlWalkFrameChain" );
#endif
SymInitialize( GetCurrentProcess(), nullptr, true );
SymSetOptions( SYMOPT_LOAD_LINES );
}
CallstackEntry DecodeCallstackPtr( uint64_t ptr )
{
CallstackEntry ret;
const auto proc = GetCurrentProcess();
char buf[sizeof( SYMBOL_INFO ) + 1024];
auto si = (SYMBOL_INFO*)buf;
si->SizeOfStruct = sizeof( SYMBOL_INFO );
si->MaxNameLen = 1024;
if( SymFromAddr( proc, ptr, nullptr, si ) == 0 )
{
memcpy( si->Name, "[unknown]", 10 );
si->NameLen = 9;
}
auto name = (char*)tracy_malloc( si->NameLen + 1 );
memcpy( name, si->Name, si->NameLen );
name[si->NameLen] = '\0';
ret.name = name;
const char* filename;
IMAGEHLP_LINE64 line;
DWORD displacement = 0;
line.SizeOfStruct = sizeof( IMAGEHLP_LINE64 );
if( SymGetLineFromAddr64( proc, ptr, &displacement, &line ) == 0 )
{
filename = "[unknown]";
ret.line = 0;
}
else
{
filename = line.FileName;
ret.line = line.LineNumber;
}
const auto fsz = strlen( filename );
auto file = (char*)tracy_malloc( fsz + 1 );
memcpy( file, filename, fsz );
file[fsz] = '\0';
ret.file = file;
return ret;
}
#elif TRACY_HAS_CALLSTACK == 2
CallstackEntry DecodeCallstackPtr( uint64_t ptr )
{
CallstackEntry ret;
ret.line = 0;
char* demangled = nullptr;
const char* symname = nullptr;
const char* symloc = nullptr;
auto vptr = (void*)ptr;
char** sym = nullptr;
ptrdiff_t symoff = 0;
Dl_info dlinfo;
if( dladdr( vptr, &dlinfo ) )
{
symloc = dlinfo.dli_fname;
symname = dlinfo.dli_sname;
symoff = (char*)ptr - (char*)dlinfo.dli_saddr;
if( symname && symname[0] == '_' )
{
size_t len = 0;
int status;
demangled = abi::__cxa_demangle( symname, nullptr, &len, &status );
if( status == 0 )
{
symname = demangled;
}
}
}
if( !symname )
{
symname = "[unknown]";
}
if( !symloc )
{
symloc = "[unknown]";
}
if( symoff == 0 )
{
const auto namelen = strlen( symname );
auto name = (char*)tracy_malloc( namelen + 1 );
memcpy( name, symname, namelen );
name[namelen] = '\0';
ret.name = name;
}
else
{
char buf[32];
sprintf( buf, " + %td", symoff );
const auto offlen = strlen( buf );
const auto namelen = strlen( symname );
auto name = (char*)tracy_malloc( namelen + offlen + 1 );
memcpy( name, symname, namelen );
memcpy( name + namelen, buf, offlen );
name[namelen + offlen] = '\0';
ret.name = name;
}
char buf[32];
sprintf( buf, " [%p]", (void*)ptr );
const auto addrlen = strlen( buf );
const auto loclen = strlen( symloc );
auto loc = (char*)tracy_malloc( loclen + addrlen + 1 );
memcpy( loc, symloc, loclen );
memcpy( loc + loclen, buf, addrlen );
loc[loclen + addrlen] = '\0';
ret.file = loc;
if( sym ) free( sym );
if( demangled ) free( demangled );
return ret;
}
#elif TRACY_HAS_CALLSTACK == 3
CallstackEntry DecodeCallstackPtr( uint64_t ptr )
{
CallstackEntry ret;
ret.line = 0;
char* demangled = nullptr;
const char* symname = nullptr;
const char* symloc = nullptr;
auto vptr = (void*)ptr;
char** sym = nullptr;
ptrdiff_t symoff = 0;
Dl_info dlinfo;
if( dladdr( vptr, &dlinfo ) )
{
symloc = dlinfo.dli_fname;
symname = dlinfo.dli_sname;
symoff = (char*)ptr - (char*)dlinfo.dli_saddr;
if( symname && symname[0] == '_' )
{
size_t len = 0;
int status;
demangled = abi::__cxa_demangle( symname, nullptr, &len, &status );
if( status == 0 )
{
symname = demangled;
}
}
}
if( !symname )
{
sym = backtrace_symbols( &vptr, 1 );
if( !sym )
{
symname = "[unknown]";
}
else
{
symname = *sym;
}
}
if( !symloc )
{
symloc = "[unknown]";
}
if( symoff == 0 )
{
const auto namelen = strlen( symname );
auto name = (char*)tracy_malloc( namelen + 1 );
memcpy( name, symname, namelen );
name[namelen] = '\0';
ret.name = name;
}
else
{
char buf[32];
sprintf( buf, " + %td", symoff );
const auto offlen = strlen( buf );
const auto namelen = strlen( symname );
auto name = (char*)tracy_malloc( namelen + offlen + 1 );
memcpy( name, symname, namelen );
memcpy( name + namelen, buf, offlen );
name[namelen + offlen] = '\0';
ret.name = name;
}
char buf[32];
sprintf( buf, " [%p]", (void*)ptr );
const auto addrlen = strlen( buf );
const auto loclen = strlen( symloc );
auto loc = (char*)tracy_malloc( loclen + addrlen + 1 );
memcpy( loc, symloc, loclen );
memcpy( loc + loclen, buf, addrlen );
loc[loclen + addrlen] = '\0';
ret.file = loc;
if( sym ) free( sym );
if( demangled ) free( demangled );
return ret;
}
#endif
}
#endif

117
client/TracyCallstack.hpp Normal file
View File

@@ -0,0 +1,117 @@
#ifndef __TRACYCALLSTACK_HPP__
#define __TRACYCALLSTACK_HPP__
#if defined _WIN32 || defined __CYGWIN__
# define TRACY_HAS_CALLSTACK 1
extern "C"
{
typedef unsigned long (__stdcall *t_RtlWalkFrameChain)( void**, unsigned long, unsigned long );
extern t_RtlWalkFrameChain RtlWalkFrameChain;
}
#elif defined __ANDROID__
# define TRACY_HAS_CALLSTACK 2
# include <unwind.h>
#elif defined __linux
# if defined _GNU_SOURCE && defined __GLIBC__
# define TRACY_HAS_CALLSTACK 3
# include <execinfo.h>
# else
# define TRACY_HAS_CALLSTACK 2
# include <unwind.h>
# endif
#endif
#ifdef TRACY_HAS_CALLSTACK
#include <assert.h>
#include <stdint.h>
#include <string.h>
#include "../common/TracyAlloc.hpp"
#include "../common/TracyForceInline.hpp"
namespace tracy
{
struct CallstackEntry
{
const char* name;
const char* file;
uint32_t line;
};
CallstackEntry DecodeCallstackPtr( uint64_t ptr );
#if TRACY_HAS_CALLSTACK == 1
void InitCallstack();
static tracy_force_inline void* Callstack( int depth )
{
assert( depth >= 1 && depth < 63 );
auto trace = (uintptr_t*)tracy_malloc( ( 1 + depth ) * sizeof( uintptr_t ) );
const auto num = RtlWalkFrameChain( (void**)( trace + 1 ), depth, 0 );
*trace = num;
return trace;
}
#elif TRACY_HAS_CALLSTACK == 2
static tracy_force_inline void InitCallstack() {}
struct BacktraceState
{
void** current;
void** end;
};
static _Unwind_Reason_Code tracy_unwind_callback( struct _Unwind_Context* ctx, void* arg )
{
auto state = (BacktraceState*)arg;
uintptr_t pc = _Unwind_GetIP( ctx );
if( pc )
{
if( state->current == state->end ) return _URC_END_OF_STACK;
*state->current++ = (void*)pc;
}
return _URC_NO_REASON;
}
static tracy_force_inline void* Callstack( int depth )
{
assert( depth >= 1 && depth < 63 );
auto trace = (uintptr_t*)tracy_malloc( ( 1 + depth ) * sizeof( uintptr_t ) );
BacktraceState state = { (void**)(trace+1), (void**)(trace+1+depth) };
_Unwind_Backtrace( tracy_unwind_callback, &state );
*trace = (uintptr_t*)state.current - trace + 1;
return trace;
}
#elif TRACY_HAS_CALLSTACK == 3
static tracy_force_inline void InitCallstack() {}
static tracy_force_inline void* Callstack( int depth )
{
assert( depth >= 1 );
auto trace = (uintptr_t*)tracy_malloc( ( 1 + depth ) * sizeof( uintptr_t ) );
const auto num = backtrace( (void**)(trace+1), depth );
*trace = num;
return trace;
}
#endif
}
#endif
#endif

View File

@@ -1,7 +1,6 @@
#ifndef __TRACYFASTVECTOR_HPP__
#define __TRACYFASTVECTOR_HPP__
#include <assert.h>
#include <stddef.h>
#include "../common/TracyAlloc.hpp"
@@ -22,7 +21,6 @@ public:
, m_write( m_ptr )
, m_end( m_ptr + capacity )
{
assert( capacity != 0 );
}
FastVector( const FastVector& ) = delete;
@@ -98,11 +96,11 @@ public:
private:
tracy_no_inline void AllocMore()
{
const auto cap = size_t( m_end - m_ptr ) * 2;
const auto size = size_t( m_write - m_ptr );
const auto cap = ( m_end - m_ptr ) * 2;
const auto size = m_write - m_ptr;
T* ptr = (T*)tracy_malloc( sizeof( T ) * cap );
memcpy( ptr, m_ptr, size * sizeof( T ) );
tracy_free_fast( m_ptr );
tracy_free( m_ptr );
m_ptr = ptr;
m_write = m_ptr + size;
m_end = m_ptr + cap;

473
client/TracyLock.hpp Normal file
View File

@@ -0,0 +1,473 @@
#ifndef __TRACYLOCK_HPP__
#define __TRACYLOCK_HPP__
#include <atomic>
#include <limits>
#include "../common/TracySystem.hpp"
#include "../common/TracyAlign.hpp"
#include "TracyProfiler.hpp"
namespace tracy
{
extern std::atomic<uint32_t> s_lockCounter;
template<class T>
class Lockable
{
public:
tracy_force_inline Lockable( const SourceLocationData* srcloc )
: m_id( s_lockCounter.fetch_add( 1, std::memory_order_relaxed ) )
#ifdef TRACY_ON_DEMAND
, m_lockCount( 0 )
, m_active( false )
#endif
{
assert( m_id != std::numeric_limits<uint32_t>::max() );
Magic magic;
auto& token = s_token.ptr;
auto& tail = token->get_tail_index();
auto item = token->enqueue_begin<tracy::moodycamel::CanAlloc>( magic );
MemWrite( &item->hdr.type, QueueType::LockAnnounce );
MemWrite( &item->lockAnnounce.id, m_id );
MemWrite( &item->lockAnnounce.lckloc, (uint64_t)srcloc );
MemWrite( &item->lockAnnounce.type, LockType::Lockable );
#ifdef TRACY_ON_DEMAND
s_profiler.DeferItem( *item );
#endif
tail.store( magic + 1, std::memory_order_release );
}
Lockable( const Lockable& ) = delete;
Lockable& operator=( const Lockable& ) = delete;
tracy_force_inline void lock()
{
#ifdef TRACY_ON_DEMAND
bool queue = false;
const auto locks = m_lockCount.fetch_add( 1, std::memory_order_relaxed );
const auto active = m_active.load( std::memory_order_relaxed );
if( locks == 0 || active )
{
const bool connected = s_profiler.IsConnected();
if( active != connected ) m_active.store( connected, std::memory_order_relaxed );
if( connected ) queue = true;
}
if( !queue )
{
m_lockable.lock();
return;
}
#endif
const auto thread = GetThreadHandle();
{
Magic magic;
auto& token = s_token.ptr;
auto& tail = token->get_tail_index();
auto item = token->enqueue_begin<tracy::moodycamel::CanAlloc>( magic );
MemWrite( &item->hdr.type, QueueType::LockWait );
MemWrite( &item->lockWait.id, m_id );
MemWrite( &item->lockWait.thread, thread );
MemWrite( &item->lockWait.time, Profiler::GetTime() );
MemWrite( &item->lockWait.type, LockType::Lockable );
tail.store( magic + 1, std::memory_order_release );
}
m_lockable.lock();
{
Magic magic;
auto& token = s_token.ptr;
auto& tail = token->get_tail_index();
auto item = token->enqueue_begin<tracy::moodycamel::CanAlloc>( magic );
MemWrite( &item->hdr.type, QueueType::LockObtain );
MemWrite( &item->lockObtain.id, m_id );
MemWrite( &item->lockObtain.thread, thread );
MemWrite( &item->lockObtain.time, Profiler::GetTime() );
tail.store( magic + 1, std::memory_order_release );
}
}
tracy_force_inline void unlock()
{
m_lockable.unlock();
#ifdef TRACY_ON_DEMAND
m_lockCount.fetch_sub( 1, std::memory_order_relaxed );
if( !m_active.load( std::memory_order_relaxed ) ) return;
if( !s_profiler.IsConnected() )
{
m_active.store( false, std::memory_order_relaxed );
return;
}
#endif
Magic magic;
auto& token = s_token.ptr;
auto& tail = token->get_tail_index();
auto item = token->enqueue_begin<tracy::moodycamel::CanAlloc>( magic );
MemWrite( &item->hdr.type, QueueType::LockRelease );
MemWrite( &item->lockRelease.id, m_id );
MemWrite( &item->lockRelease.thread, GetThreadHandle() );
MemWrite( &item->lockRelease.time, Profiler::GetTime() );
tail.store( magic + 1, std::memory_order_release );
}
tracy_force_inline bool try_lock()
{
const auto ret = m_lockable.try_lock();
#ifdef TRACY_ON_DEMAND
if( !ret ) return ret;
bool queue = false;
const auto locks = m_lockCount.fetch_add( 1, std::memory_order_relaxed );
const auto active = m_active.load( std::memory_order_relaxed );
if( locks == 0 || active )
{
const bool connected = s_profiler.IsConnected();
if( active != connected ) m_active.store( connected, std::memory_order_relaxed );
if( connected ) queue = true;
}
if( !queue ) return ret;
#endif
if( ret )
{
Magic magic;
auto& token = s_token.ptr;
auto& tail = token->get_tail_index();
auto item = token->enqueue_begin<tracy::moodycamel::CanAlloc>( magic );
MemWrite( &item->hdr.type, QueueType::LockObtain );
MemWrite( &item->lockObtain.id, m_id );
MemWrite( &item->lockObtain.thread, GetThreadHandle() );
MemWrite( &item->lockObtain.time, Profiler::GetTime() );
tail.store( magic + 1, std::memory_order_release );
}
return ret;
}
tracy_force_inline void Mark( const SourceLocationData* srcloc )
{
#ifdef TRACY_ON_DEMAND
const auto active = m_active.load( std::memory_order_relaxed );
if( !active ) return;
const auto connected = s_profiler.IsConnected();
if( !connected )
{
if( active ) m_active.store( false, std::memory_order_relaxed );
return;
}
#endif
Magic magic;
auto& token = s_token.ptr;
auto& tail = token->get_tail_index();
auto item = token->enqueue_begin<tracy::moodycamel::CanAlloc>( magic );
MemWrite( &item->hdr.type, QueueType::LockMark );
MemWrite( &item->lockMark.id, m_id );
MemWrite( &item->lockMark.thread, GetThreadHandle() );
MemWrite( &item->lockMark.srcloc, (uint64_t)srcloc );
tail.store( magic + 1, std::memory_order_release );
}
private:
T m_lockable;
uint32_t m_id;
#ifdef TRACY_ON_DEMAND
std::atomic<uint32_t> m_lockCount;
std::atomic<bool> m_active;
#endif
};
template<class T>
class SharedLockable
{
public:
tracy_force_inline SharedLockable( const SourceLocationData* srcloc )
: m_id( s_lockCounter.fetch_add( 1, std::memory_order_relaxed ) )
#ifdef TRACY_ON_DEMAND
, m_lockCount( 0 )
, m_active( false )
#endif
{
assert( m_id != std::numeric_limits<uint32_t>::max() );
Magic magic;
auto& token = s_token.ptr;
auto& tail = token->get_tail_index();
auto item = token->enqueue_begin<tracy::moodycamel::CanAlloc>( magic );
MemWrite( &item->hdr.type, QueueType::LockAnnounce );
MemWrite( &item->lockAnnounce.id, m_id );
MemWrite( &item->lockAnnounce.lckloc, (uint64_t)srcloc );
MemWrite( &item->lockAnnounce.type, LockType::SharedLockable );
#ifdef TRACY_ON_DEMAND
s_profiler.DeferItem( *item );
#endif
tail.store( magic + 1, std::memory_order_release );
}
SharedLockable( const SharedLockable& ) = delete;
SharedLockable& operator=( const SharedLockable& ) = delete;
tracy_force_inline void lock()
{
#ifdef TRACY_ON_DEMAND
bool queue = false;
const auto locks = m_lockCount.fetch_add( 1, std::memory_order_relaxed );
const auto active = m_active.load( std::memory_order_relaxed );
if( locks == 0 || active )
{
const bool connected = s_profiler.IsConnected();
if( active != connected ) m_active.store( connected, std::memory_order_relaxed );
if( connected ) queue = true;
}
if( !queue )
{
m_lockable.lock();
return;
}
#endif
const auto thread = GetThreadHandle();
{
Magic magic;
auto& token = s_token.ptr;
auto& tail = token->get_tail_index();
auto item = token->enqueue_begin<tracy::moodycamel::CanAlloc>( magic );
MemWrite( &item->hdr.type, QueueType::LockWait );
MemWrite( &item->lockWait.id, m_id );
MemWrite( &item->lockWait.thread, thread );
MemWrite( &item->lockWait.time, Profiler::GetTime() );
MemWrite( &item->lockWait.type, LockType::SharedLockable );
tail.store( magic + 1, std::memory_order_release );
}
m_lockable.lock();
{
Magic magic;
auto& token = s_token.ptr;
auto& tail = token->get_tail_index();
auto item = token->enqueue_begin<tracy::moodycamel::CanAlloc>( magic );
MemWrite( &item->hdr.type, QueueType::LockObtain );
MemWrite( &item->lockObtain.id, m_id );
MemWrite( &item->lockObtain.thread, thread );
MemWrite( &item->lockObtain.time, Profiler::GetTime() );
tail.store( magic + 1, std::memory_order_release );
}
}
tracy_force_inline void unlock()
{
m_lockable.unlock();
#ifdef TRACY_ON_DEMAND
m_lockCount.fetch_sub( 1, std::memory_order_relaxed );
if( !m_active.load( std::memory_order_relaxed ) ) return;
if( !s_profiler.IsConnected() )
{
m_active.store( false, std::memory_order_relaxed );
return;
}
#endif
Magic magic;
auto& token = s_token.ptr;
auto& tail = token->get_tail_index();
auto item = token->enqueue_begin<tracy::moodycamel::CanAlloc>( magic );
MemWrite( &item->hdr.type, QueueType::LockRelease );
MemWrite( &item->lockRelease.id, m_id );
MemWrite( &item->lockRelease.thread, GetThreadHandle() );
MemWrite( &item->lockRelease.time, Profiler::GetTime() );
tail.store( magic + 1, std::memory_order_release );
}
tracy_force_inline bool try_lock()
{
const auto ret = m_lockable.try_lock();
#ifdef TRACY_ON_DEMAND
if( !ret ) return ret;
bool queue = false;
const auto locks = m_lockCount.fetch_add( 1, std::memory_order_relaxed );
const auto active = m_active.load( std::memory_order_relaxed );
if( locks == 0 || active )
{
const bool connected = s_profiler.IsConnected();
if( active != connected ) m_active.store( connected, std::memory_order_relaxed );
if( connected ) queue = true;
}
if( !queue ) return ret;
#endif
if( ret )
{
Magic magic;
auto& token = s_token.ptr;
auto& tail = token->get_tail_index();
auto item = token->enqueue_begin<tracy::moodycamel::CanAlloc>( magic );
MemWrite( &item->hdr.type, QueueType::LockObtain );
MemWrite( &item->lockObtain.id, m_id );
MemWrite( &item->lockObtain.thread, GetThreadHandle() );
MemWrite( &item->lockObtain.time, Profiler::GetTime() );
tail.store( magic + 1, std::memory_order_release );
}
return ret;
}
tracy_force_inline void lock_shared()
{
#ifdef TRACY_ON_DEMAND
bool queue = false;
const auto locks = m_lockCount.fetch_add( 1, std::memory_order_relaxed );
const auto active = m_active.load( std::memory_order_relaxed );
if( locks == 0 || active )
{
const bool connected = s_profiler.IsConnected();
if( active != connected ) m_active.store( connected, std::memory_order_relaxed );
if( connected ) queue = true;
}
if( !queue )
{
m_lockable.lock_shared();
return;
}
#endif
const auto thread = GetThreadHandle();
{
Magic magic;
auto& token = s_token.ptr;
auto& tail = token->get_tail_index();
auto item = token->enqueue_begin<tracy::moodycamel::CanAlloc>( magic );
MemWrite( &item->hdr.type, QueueType::LockSharedWait );
MemWrite( &item->lockWait.id, m_id );
MemWrite( &item->lockWait.thread, thread );
MemWrite( &item->lockWait.time, Profiler::GetTime() );
MemWrite( &item->lockWait.type, LockType::SharedLockable );
tail.store( magic + 1, std::memory_order_release );
}
m_lockable.lock_shared();
{
Magic magic;
auto& token = s_token.ptr;
auto& tail = token->get_tail_index();
auto item = token->enqueue_begin<tracy::moodycamel::CanAlloc>( magic );
MemWrite( &item->hdr.type, QueueType::LockSharedObtain );
MemWrite( &item->lockObtain.id, m_id );
MemWrite( &item->lockObtain.thread, thread );
MemWrite( &item->lockObtain.time, Profiler::GetTime() );
tail.store( magic + 1, std::memory_order_release );
}
}
tracy_force_inline void unlock_shared()
{
m_lockable.unlock_shared();
#ifdef TRACY_ON_DEMAND
m_lockCount.fetch_sub( 1, std::memory_order_relaxed );
if( !m_active.load( std::memory_order_relaxed ) ) return;
if( !s_profiler.IsConnected() )
{
m_active.store( false, std::memory_order_relaxed );
return;
}
#endif
Magic magic;
auto& token = s_token.ptr;
auto& tail = token->get_tail_index();
auto item = token->enqueue_begin<tracy::moodycamel::CanAlloc>( magic );
MemWrite( &item->hdr.type, QueueType::LockSharedRelease );
MemWrite( &item->lockRelease.id, m_id );
MemWrite( &item->lockRelease.thread, GetThreadHandle() );
MemWrite( &item->lockRelease.time, Profiler::GetTime() );
tail.store( magic + 1, std::memory_order_release );
}
tracy_force_inline bool try_lock_shared()
{
const auto ret = m_lockable.try_lock_shared();
#ifdef TRACY_ON_DEMAND
if( !ret ) return ret;
bool queue = false;
const auto locks = m_lockCount.fetch_add( 1, std::memory_order_relaxed );
const auto active = m_active.load( std::memory_order_relaxed );
if( locks == 0 || active )
{
const bool connected = s_profiler.IsConnected();
if( active != connected ) m_active.store( connected, std::memory_order_relaxed );
if( connected ) queue = true;
}
if( !queue ) return ret;
#endif
if( ret )
{
Magic magic;
auto& token = s_token.ptr;
auto& tail = token->get_tail_index();
auto item = token->enqueue_begin<tracy::moodycamel::CanAlloc>( magic );
MemWrite( &item->hdr.type, QueueType::LockSharedObtain );
MemWrite( &item->lockObtain.id, m_id );
MemWrite( &item->lockObtain.thread, GetThreadHandle() );
MemWrite( &item->lockObtain.time, Profiler::GetTime() );
tail.store( magic + 1, std::memory_order_release );
}
return ret;
}
tracy_force_inline void Mark( const SourceLocationData* srcloc )
{
#ifdef TRACY_ON_DEMAND
const auto active = m_active.load( std::memory_order_relaxed );
if( !active ) return;
const auto connected = s_profiler.IsConnected();
if( !connected )
{
if( active ) m_active.store( false, std::memory_order_relaxed );
return;
}
#endif
Magic magic;
auto& token = s_token.ptr;
auto& tail = token->get_tail_index();
auto item = token->enqueue_begin<tracy::moodycamel::CanAlloc>( magic );
MemWrite( &item->hdr.type, QueueType::LockMark );
MemWrite( &item->lockMark.id, m_id );
MemWrite( &item->lockMark.thread, GetThreadHandle() );
MemWrite( &item->lockMark.srcloc, (uint64_t)srcloc );
tail.store( magic + 1, std::memory_order_release );
}
private:
T m_lockable;
uint32_t m_id;
#ifdef TRACY_ON_DEMAND
std::atomic<uint32_t> m_lockCount;
std::atomic<bool> m_active;
#endif
};
};
#endif

1588
client/TracyProfiler.cpp Normal file

File diff suppressed because it is too large Load Diff

459
client/TracyProfiler.hpp Normal file
View File

@@ -0,0 +1,459 @@
#ifndef __TRACYPROFILER_HPP__
#define __TRACYPROFILER_HPP__
#include <assert.h>
#include <atomic>
#include <chrono>
#include <stdint.h>
#include <string.h>
#include "concurrentqueue.h"
#include "TracyCallstack.hpp"
#include "TracyFastVector.hpp"
#include "../common/tracy_lz4.hpp"
#include "../common/TracyQueue.hpp"
#include "../common/TracyAlign.hpp"
#include "../common/TracyAlloc.hpp"
#include "../common/TracyMutex.hpp"
#include "../common/TracySystem.hpp"
#if defined _MSC_VER || defined __CYGWIN__
# include <intrin.h>
#endif
#if defined _MSC_VER || defined __CYGWIN__ || ( ( defined __i386 || defined _M_IX86 || defined __x86_64__ || defined _M_X64 ) && !defined __ANDROID__ ) || __ARM_ARCH >= 6
# define TRACY_HW_TIMER
# if defined _MSC_VER || defined __CYGWIN__
// Enable optimization for MSVC __rdtscp() intrin, saving one LHS of a cpu value on the stack.
// This comes at the cost of an unaligned memory write.
# define TRACY_RDTSCP_OPT
# endif
#endif
#define TracyConcat(x,y) TracyConcatIndirect(x,y)
#define TracyConcatIndirect(x,y) x##y
namespace tracy
{
class Socket;
struct SourceLocationData
{
const char* name;
const char* function;
const char* file;
uint32_t line;
uint32_t color;
};
struct ProducerWrapper
{
tracy::moodycamel::ConcurrentQueue<QueueItem>::ExplicitProducer* ptr;
};
extern thread_local ProducerWrapper s_token;
class GpuCtx;
struct GpuCtxWrapper
{
GpuCtx* ptr;
};
class VkCtx;
struct VkCtxWrapper
{
VkCtx* ptr;
};
#ifdef TRACY_ON_DEMAND
struct LuaZoneState
{
uint32_t counter;
bool active;
};
#endif
using Magic = tracy::moodycamel::ConcurrentQueueDefaultTraits::index_t;
#if __ARM_ARCH >= 6
extern int64_t (*GetTimeImpl)();
#endif
class Profiler;
extern Profiler& s_profiler;
class Profiler
{
public:
Profiler();
~Profiler();
static tracy_force_inline int64_t GetTime( uint32_t& cpu )
{
#ifdef TRACY_HW_TIMER
# if __ARM_ARCH >= 6
cpu = 0xFFFFFFFF;
return GetTimeImpl();
# elif defined _MSC_VER || defined __CYGWIN__
const auto t = int64_t( __rdtscp( &cpu ) );
return t;
# elif defined __i386 || defined _M_IX86 || defined __x86_64__ || defined _M_X64
uint32_t eax, edx;
asm volatile ( "rdtscp" : "=a" (eax), "=d" (edx), "=c" (cpu) :: );
return ( uint64_t( edx ) << 32 ) + uint64_t( eax );
# endif
#else
cpu = 0xFFFFFFFF;
return std::chrono::duration_cast<std::chrono::nanoseconds>( std::chrono::high_resolution_clock::now().time_since_epoch() ).count();
#endif
}
static tracy_force_inline int64_t GetTime()
{
#ifdef TRACY_HW_TIMER
# if __ARM_ARCH >= 6
return GetTimeImpl();
# elif defined _MSC_VER || defined __CYGWIN__
unsigned int dontcare;
const auto t = int64_t( __rdtscp( &dontcare ) );
return t;
# elif defined __i386 || defined _M_IX86 || defined __x86_64__ || defined _M_X64
uint32_t eax, edx;
asm volatile ( "rdtscp" : "=a" (eax), "=d" (edx) :: "%ecx" );
return ( uint64_t( edx ) << 32 ) + uint64_t( eax );
# endif
#else
return std::chrono::duration_cast<std::chrono::nanoseconds>( std::chrono::high_resolution_clock::now().time_since_epoch() ).count();
#endif
}
static tracy_force_inline void SendFrameMark()
{
#ifdef TRACY_ON_DEMAND
s_profiler.m_frameCount.fetch_add( 1, std::memory_order_relaxed );
if( !s_profiler.IsConnected() ) return;
#endif
Magic magic;
auto& token = s_token.ptr;
auto& tail = token->get_tail_index();
auto item = token->enqueue_begin<tracy::moodycamel::CanAlloc>( magic );
MemWrite( &item->hdr.type, QueueType::FrameMarkMsg );
MemWrite( &item->frameMark.time, GetTime() );
MemWrite( &item->frameMark.name, uint64_t( 0 ) );
tail.store( magic + 1, std::memory_order_release );
}
static tracy_force_inline void SendFrameMark( const char* name, QueueType type )
{
assert( type == QueueType::FrameMarkMsg || type == QueueType::FrameMarkMsgStart || type == QueueType::FrameMarkMsgEnd );
#ifdef TRACY_ON_DEMAND
if( !s_profiler.IsConnected() ) return;
#endif
Magic magic;
auto& token = s_token.ptr;
auto& tail = token->get_tail_index();
auto item = token->enqueue_begin<tracy::moodycamel::CanAlloc>( magic );
MemWrite( &item->hdr.type, type );
MemWrite( &item->frameMark.time, GetTime() );
MemWrite( &item->frameMark.name, uint64_t( name ) );
tail.store( magic + 1, std::memory_order_release );
}
static tracy_force_inline void PlotData( const char* name, int64_t val )
{
#ifdef TRACY_ON_DEMAND
if( !s_profiler.IsConnected() ) return;
#endif
Magic magic;
auto& token = s_token.ptr;
auto& tail = token->get_tail_index();
auto item = token->enqueue_begin<tracy::moodycamel::CanAlloc>( magic );
MemWrite( &item->hdr.type, QueueType::PlotData );
MemWrite( &item->plotData.name, (uint64_t)name );
MemWrite( &item->plotData.time, GetTime() );
MemWrite( &item->plotData.type, PlotDataType::Int );
MemWrite( &item->plotData.data.i, val );
tail.store( magic + 1, std::memory_order_release );
}
static tracy_force_inline void PlotData( const char* name, float val )
{
#ifdef TRACY_ON_DEMAND
if( !s_profiler.IsConnected() ) return;
#endif
Magic magic;
auto& token = s_token.ptr;
auto& tail = token->get_tail_index();
auto item = token->enqueue_begin<tracy::moodycamel::CanAlloc>( magic );
MemWrite( &item->hdr.type, QueueType::PlotData );
MemWrite( &item->plotData.name, (uint64_t)name );
MemWrite( &item->plotData.time, GetTime() );
MemWrite( &item->plotData.type, PlotDataType::Float );
MemWrite( &item->plotData.data.f, val );
tail.store( magic + 1, std::memory_order_release );
}
static tracy_force_inline void PlotData( const char* name, double val )
{
#ifdef TRACY_ON_DEMAND
if( !s_profiler.IsConnected() ) return;
#endif
Magic magic;
auto& token = s_token.ptr;
auto& tail = token->get_tail_index();
auto item = token->enqueue_begin<tracy::moodycamel::CanAlloc>( magic );
MemWrite( &item->hdr.type, QueueType::PlotData );
MemWrite( &item->plotData.name, (uint64_t)name );
MemWrite( &item->plotData.time, GetTime() );
MemWrite( &item->plotData.type, PlotDataType::Double );
MemWrite( &item->plotData.data.d, val );
tail.store( magic + 1, std::memory_order_release );
}
static tracy_force_inline void Message( const char* txt, size_t size )
{
#ifdef TRACY_ON_DEMAND
if( !s_profiler.IsConnected() ) return;
#endif
Magic magic;
auto& token = s_token.ptr;
auto ptr = (char*)tracy_malloc( size+1 );
memcpy( ptr, txt, size );
ptr[size] = '\0';
auto& tail = token->get_tail_index();
auto item = token->enqueue_begin<tracy::moodycamel::CanAlloc>( magic );
MemWrite( &item->hdr.type, QueueType::Message );
MemWrite( &item->message.time, GetTime() );
MemWrite( &item->message.thread, GetThreadHandle() );
MemWrite( &item->message.text, (uint64_t)ptr );
tail.store( magic + 1, std::memory_order_release );
}
static tracy_force_inline void Message( const char* txt )
{
#ifdef TRACY_ON_DEMAND
if( !s_profiler.IsConnected() ) return;
#endif
Magic magic;
auto& token = s_token.ptr;
auto& tail = token->get_tail_index();
auto item = token->enqueue_begin<tracy::moodycamel::CanAlloc>( magic );
MemWrite( &item->hdr.type, QueueType::MessageLiteral );
MemWrite( &item->message.time, GetTime() );
MemWrite( &item->message.thread, GetThreadHandle() );
MemWrite( &item->message.text, (uint64_t)txt );
tail.store( magic + 1, std::memory_order_release );
}
static tracy_force_inline void MemAlloc( const void* ptr, size_t size )
{
#ifdef TRACY_ON_DEMAND
if( !s_profiler.IsConnected() ) return;
#endif
const auto thread = GetThreadHandle();
s_profiler.m_serialLock.lock();
SendMemAlloc( QueueType::MemAlloc, thread, ptr, size );
s_profiler.m_serialLock.unlock();
}
static tracy_force_inline void MemFree( const void* ptr )
{
#ifdef TRACY_ON_DEMAND
if( !s_profiler.IsConnected() ) return;
#endif
const auto thread = GetThreadHandle();
s_profiler.m_serialLock.lock();
SendMemFree( QueueType::MemFree, thread, ptr );
s_profiler.m_serialLock.unlock();
}
static tracy_force_inline void MemAllocCallstack( const void* ptr, size_t size, int depth )
{
#ifdef TRACY_HAS_CALLSTACK
# ifdef TRACY_ON_DEMAND
if( !s_profiler.IsConnected() ) return;
# endif
const auto thread = GetThreadHandle();
rpmalloc_thread_initialize();
auto callstack = Callstack( depth );
s_profiler.m_serialLock.lock();
SendMemAlloc( QueueType::MemAllocCallstack, thread, ptr, size );
SendCallstackMemory( callstack );
s_profiler.m_serialLock.unlock();
#else
MemAlloc( ptr, size );
#endif
}
static tracy_force_inline void MemFreeCallstack( const void* ptr, int depth )
{
#ifdef TRACY_HAS_CALLSTACK
# ifdef TRACY_ON_DEMAND
if( !s_profiler.IsConnected() ) return;
# endif
const auto thread = GetThreadHandle();
rpmalloc_thread_initialize();
auto callstack = Callstack( depth );
s_profiler.m_serialLock.lock();
SendMemFree( QueueType::MemFreeCallstack, thread, ptr );
SendCallstackMemory( callstack );
s_profiler.m_serialLock.unlock();
#else
MemFree( ptr );
#endif
}
static tracy_force_inline void SendCallstack( int depth, uint64_t thread )
{
#ifdef TRACY_HAS_CALLSTACK
auto ptr = Callstack( depth );
Magic magic;
auto& token = s_token.ptr;
auto& tail = token->get_tail_index();
auto item = token->enqueue_begin<tracy::moodycamel::CanAlloc>( magic );
MemWrite( &item->hdr.type, QueueType::Callstack );
MemWrite( &item->callstack.ptr, ptr );
MemWrite( &item->callstack.thread, thread );
tail.store( magic + 1, std::memory_order_release );
#endif
}
void SendCallstack( int depth, uint64_t thread, const char* skipBefore );
static bool ShouldExit();
#ifdef TRACY_ON_DEMAND
tracy_force_inline bool IsConnected()
{
return m_isConnected.load( std::memory_order_relaxed );
}
tracy_force_inline void DeferItem( const QueueItem& item )
{
m_deferredLock.lock();
auto dst = m_deferredQueue.push_next();
memcpy( dst, &item, sizeof( item ) );
m_deferredLock.unlock();
}
#endif
void RequestShutdown() { m_shutdown.store( true, std::memory_order_relaxed ); m_shutdownManual.store( true, std::memory_order_relaxed ); }
bool HasShutdownFinished() const { return m_shutdownFinished.load( std::memory_order_relaxed ); }
private:
enum DequeueStatus { Success, ConnectionLost, QueueEmpty };
static void LaunchWorker( void* ptr ) { ((Profiler*)ptr)->Worker(); }
void Worker();
void ClearQueues( tracy::moodycamel::ConsumerToken& token );
DequeueStatus Dequeue( tracy::moodycamel::ConsumerToken& token );
DequeueStatus DequeueSerial();
bool AppendData( const void* data, size_t len );
bool CommitData();
bool NeedDataSize( size_t len );
tracy_force_inline void AppendDataUnsafe( const void* data, size_t len )
{
memcpy( m_buffer + m_bufferOffset, data, len );
m_bufferOffset += int( len );
}
bool SendData( const char* data, size_t len );
void SendString( uint64_t ptr, const char* str, QueueType type );
void SendSourceLocation( uint64_t ptr );
void SendSourceLocationPayload( uint64_t ptr );
void SendCallstackPayload( uint64_t ptr );
void SendCallstackFrame( uint64_t ptr );
bool HandleServerQuery();
void CalibrateTimer();
void CalibrateDelay();
static tracy_force_inline void SendCallstackMemory( void* ptr )
{
#ifdef TRACY_HAS_CALLSTACK
auto item = s_profiler.m_serialQueue.prepare_next();
MemWrite( &item->hdr.type, QueueType::CallstackMemory );
MemWrite( &item->callstackMemory.ptr, (uint64_t)ptr );
s_profiler.m_serialQueue.commit_next();
#endif
}
static tracy_force_inline void SendMemAlloc( QueueType type, const uint64_t thread, const void* ptr, size_t size )
{
assert( type == QueueType::MemAlloc || type == QueueType::MemAllocCallstack );
auto item = s_profiler.m_serialQueue.prepare_next();
MemWrite( &item->hdr.type, type );
MemWrite( &item->memAlloc.time, GetTime() );
MemWrite( &item->memAlloc.thread, thread );
MemWrite( &item->memAlloc.ptr, (uint64_t)ptr );
if( compile_time_condition<sizeof( size ) == 4>::value )
{
memcpy( &item->memAlloc.size, &size, 4 );
memset( &item->memAlloc.size + 4, 0, 2 );
}
else
{
assert( sizeof( size ) == 8 );
memcpy( &item->memAlloc.size, &size, 6 );
}
s_profiler.m_serialQueue.commit_next();
}
static tracy_force_inline void SendMemFree( QueueType type, const uint64_t thread, const void* ptr )
{
assert( type == QueueType::MemFree || type == QueueType::MemFreeCallstack );
auto item = s_profiler.m_serialQueue.prepare_next();
MemWrite( &item->hdr.type, type );
MemWrite( &item->memFree.time, GetTime() );
MemWrite( &item->memFree.thread, thread );
MemWrite( &item->memFree.ptr, (uint64_t)ptr );
s_profiler.m_serialQueue.commit_next();
}
double m_timerMul;
uint64_t m_resolution;
uint64_t m_delay;
std::atomic<int64_t> m_timeBegin;
uint64_t m_mainThread;
uint64_t m_epoch;
std::atomic<bool> m_shutdown;
std::atomic<bool> m_shutdownManual;
std::atomic<bool> m_shutdownFinished;
Socket* m_sock;
bool m_noExit;
LZ4_stream_t* m_stream;
char* m_buffer;
int m_bufferOffset;
int m_bufferStart;
QueueItem* m_itemBuf;
char* m_lz4Buf;
FastVector<QueueItem> m_serialQueue, m_serialDequeue;
TracyMutex m_serialLock;
#ifdef TRACY_ON_DEMAND
std::atomic<bool> m_isConnected;
std::atomic<uint64_t> m_frameCount;
TracyMutex m_deferredLock;
FastVector<QueueItem> m_deferredQueue;
#endif
};
};
#endif

132
client/TracyScoped.hpp Normal file
View File

@@ -0,0 +1,132 @@
#ifndef __TRACYSCOPED_HPP__
#define __TRACYSCOPED_HPP__
#include <stdint.h>
#include <string.h>
#include "../common/TracySystem.hpp"
#include "../common/TracyAlign.hpp"
#include "../common/TracyAlloc.hpp"
#include "TracyProfiler.hpp"
namespace tracy
{
class ScopedZone
{
public:
tracy_force_inline ScopedZone( const SourceLocationData* srcloc, bool is_active = true )
#ifdef TRACY_ON_DEMAND
: m_active( s_profiler.IsConnected() )
#else
: m_active( is_active )
#endif
{
if( !m_active ) return;
const auto thread = GetThreadHandle();
m_thread = thread;
Magic magic;
auto& token = s_token.ptr;
auto& tail = token->get_tail_index();
auto item = token->enqueue_begin<tracy::moodycamel::CanAlloc>( magic );
MemWrite( &item->hdr.type, QueueType::ZoneBegin );
#ifdef TRACY_RDTSCP_OPT
MemWrite( &item->zoneBegin.time, Profiler::GetTime( item->zoneBegin.cpu ) );
#else
uint32_t cpu;
MemWrite( &item->zoneBegin.time, Profiler::GetTime( cpu ) );
MemWrite( &item->zoneBegin.cpu, cpu );
#endif
MemWrite( &item->zoneBegin.thread, thread );
MemWrite( &item->zoneBegin.srcloc, (uint64_t)srcloc );
tail.store( magic + 1, std::memory_order_release );
}
tracy_force_inline ScopedZone( const SourceLocationData* srcloc, int depth, bool is_active = true )
#ifdef TRACY_ON_DEMAND
: m_active( s_profiler.IsConnected() )
#else
: m_active( is_active )
#endif
{
if( !m_active ) return;
const auto thread = GetThreadHandle();
m_thread = thread;
Magic magic;
auto& token = s_token.ptr;
auto& tail = token->get_tail_index();
auto item = token->enqueue_begin<tracy::moodycamel::CanAlloc>( magic );
MemWrite( &item->hdr.type, QueueType::ZoneBeginCallstack );
#ifdef TRACY_RDTSCP_OPT
MemWrite( &item->zoneBegin.time, Profiler::GetTime( item->zoneBegin.cpu ) );
#else
uint32_t cpu;
MemWrite( &item->zoneBegin.time, Profiler::GetTime( cpu ) );
MemWrite( &item->zoneBegin.cpu, cpu );
#endif
MemWrite( &item->zoneBegin.thread, thread );
MemWrite( &item->zoneBegin.srcloc, (uint64_t)srcloc );
tail.store( magic + 1, std::memory_order_release );
s_profiler.SendCallstack( depth, thread );
}
tracy_force_inline ~ScopedZone()
{
if( !m_active ) return;
Magic magic;
auto& token = s_token.ptr;
auto& tail = token->get_tail_index();
auto item = token->enqueue_begin<tracy::moodycamel::CanAlloc>( magic );
MemWrite( &item->hdr.type, QueueType::ZoneEnd );
#ifdef TRACY_RDTSCP_OPT
MemWrite( &item->zoneEnd.time, Profiler::GetTime( item->zoneEnd.cpu ) );
#else
uint32_t cpu;
MemWrite( &item->zoneEnd.time, Profiler::GetTime( cpu ) );
MemWrite( &item->zoneEnd.cpu, cpu );
#endif
MemWrite( &item->zoneEnd.thread, m_thread );
tail.store( magic + 1, std::memory_order_release );
}
tracy_force_inline void Text( const char* txt, size_t size )
{
if( !m_active ) return;
Magic magic;
auto& token = s_token.ptr;
auto ptr = (char*)tracy_malloc( size+1 );
memcpy( ptr, txt, size );
ptr[size] = '\0';
auto& tail = token->get_tail_index();
auto item = token->enqueue_begin<tracy::moodycamel::CanAlloc>( magic );
MemWrite( &item->hdr.type, QueueType::ZoneText );
MemWrite( &item->zoneText.thread, m_thread );
MemWrite( &item->zoneText.text, (uint64_t)ptr );
tail.store( magic + 1, std::memory_order_release );
}
tracy_force_inline void Name( const char* txt, size_t size )
{
if( !m_active ) return;
Magic magic;
auto& token = s_token.ptr;
auto ptr = (char*)tracy_malloc( size+1 );
memcpy( ptr, txt, size );
ptr[size] = '\0';
auto& tail = token->get_tail_index();
auto item = token->enqueue_begin<tracy::moodycamel::CanAlloc>( magic );
MemWrite( &item->hdr.type, QueueType::ZoneName );
MemWrite( &item->zoneText.thread, m_thread );
MemWrite( &item->zoneText.text, (uint64_t)ptr );
tail.store( magic + 1, std::memory_order_release );
}
private:
uint64_t m_thread;
const bool m_active;
};
}
#endif

View File

@@ -1,36 +1,16 @@
#ifndef __TRACYTHREAD_HPP__
#define __TRACYTHREAD_HPP__
#if defined _WIN32
#ifdef _MSC_VER
# include <windows.h>
#else
# include <pthread.h>
#endif
#ifdef TRACY_MANUAL_LIFETIME
# include "tracy_rpmalloc.hpp"
#endif
namespace tracy
{
#ifdef TRACY_MANUAL_LIFETIME
extern thread_local bool RpThreadInitDone;
#endif
class ThreadExitHandler
{
public:
~ThreadExitHandler()
{
#ifdef TRACY_MANUAL_LIFETIME
rpmalloc_thread_finalize( 1 );
RpThreadInitDone = false;
#endif
}
};
#if defined _WIN32
#ifdef _MSC_VER
class Thread
{

3697
client/concurrentqueue.h Normal file

File diff suppressed because it is too large Load Diff

2096
client/tracy_rpmalloc.cpp Normal file

File diff suppressed because it is too large Load Diff

151
client/tracy_rpmalloc.hpp Normal file
View File

@@ -0,0 +1,151 @@
/* rpmalloc.h - Memory allocator - Public Domain - 2016 Mattias Jansson / Rampant Pixels
*
* This library provides a cross-platform lock free thread caching malloc implementation in C11.
* The latest source code is always available at
*
* https://github.com/rampantpixels/rpmalloc
*
* This library is put in the public domain; you can redistribute it and/or modify it without any restrictions.
*
*/
#pragma once
#include <stddef.h>
namespace tracy
{
#if defined(__clang__) || defined(__GNUC__)
# define RPMALLOC_ATTRIBUTE __attribute__((__malloc__))
# define RPMALLOC_RESTRICT
# define RPMALLOC_CDECL
#elif defined(_MSC_VER)
# define RPMALLOC_ATTRIBUTE
# define RPMALLOC_RESTRICT __declspec(restrict)
# define RPMALLOC_CDECL __cdecl
#else
# define RPMALLOC_ATTRIBUTE
# define RPMALLOC_RESTRICT
# define RPMALLOC_CDECL
#endif
//! Flag to rpaligned_realloc to not preserve content in reallocation
#define RPMALLOC_NO_PRESERVE 1
typedef struct rpmalloc_global_statistics_t {
//! Current amount of virtual memory mapped (only if ENABLE_STATISTICS=1)
size_t mapped;
//! Current amount of memory in global caches for small and medium sizes (<64KiB)
size_t cached;
//! Total amount of memory mapped (only if ENABLE_STATISTICS=1)
size_t mapped_total;
//! Total amount of memory unmapped (only if ENABLE_STATISTICS=1)
size_t unmapped_total;
} rpmalloc_global_statistics_t;
typedef struct rpmalloc_thread_statistics_t {
//! Current number of bytes available for allocation from active spans
size_t active;
//! Current number of bytes available in thread size class caches
size_t sizecache;
//! Current number of bytes available in thread span caches
size_t spancache;
//! Current number of bytes in pending deferred deallocations
size_t deferred;
//! Total number of bytes transitioned from thread cache to global cache
size_t thread_to_global;
//! Total number of bytes transitioned from global cache to thread cache
size_t global_to_thread;
} rpmalloc_thread_statistics_t;
typedef struct rpmalloc_config_t {
//! Map memory pages for the given number of bytes. The returned address MUST be
// aligned to the rpmalloc span size, which will always be a power of two.
// Optionally the function can store an alignment offset in the offset variable
// in case it performs alignment and the returned pointer is offset from the
// actual start of the memory region due to this alignment. The alignment offset
// will be passed to the memory unmap function. The alignment offset MUST NOT be
// larger than 65535 (storable in an uint16_t), if it is you must use natural
// alignment to shift it into 16 bits.
void* (*memory_map)(size_t size, size_t* offset);
//! Unmap the memory pages starting at address and spanning the given number of bytes.
// If release is set to 1, the unmap is for an entire span range as returned by
// a previous call to memory_map and that the entire range should be released.
// If release is set to 0, the unmap is a partial decommit of a subset of the mapped
// memory range.
void (*memory_unmap)(void* address, size_t size, size_t offset, int release);
//! Size of memory pages. The page size MUST be a power of two in [512,16384] range
// (2^9 to 2^14) unless 0 - set to 0 to use system page size. All memory mapping
// requests to memory_map will be made with size set to a multiple of the page size.
size_t page_size;
//! Size of a span of memory pages. MUST be a multiple of page size, and in [4096,262144]
// range (unless 0 - set to 0 to use the default span size).
size_t span_size;
//! Number of spans to map at each request to map new virtual memory blocks. This can
// be used to minimize the system call overhead at the cost of virtual memory address
// space. The extra mapped pages will not be written until actually used, so physical
// committed memory should not be affected in the default implementation.
size_t span_map_count;
//! Debug callback if memory guards are enabled. Called if a memory overwrite is detected
void (*memory_overwrite)(void* address);
} rpmalloc_config_t;
extern int
rpmalloc_initialize(void);
extern int
rpmalloc_initialize_config(const rpmalloc_config_t* config);
extern const rpmalloc_config_t*
rpmalloc_config(void);
extern void
rpmalloc_finalize(void);
extern void
rpmalloc_thread_initialize(void);
extern void
rpmalloc_thread_finalize(void);
extern void
rpmalloc_thread_collect(void);
extern int
rpmalloc_is_thread_initialized(void);
extern void
rpmalloc_thread_statistics(rpmalloc_thread_statistics_t* stats);
extern void
rpmalloc_global_statistics(rpmalloc_global_statistics_t* stats);
extern RPMALLOC_RESTRICT void*
rpmalloc(size_t size) RPMALLOC_ATTRIBUTE;
extern void
rpfree(void* ptr);
extern RPMALLOC_RESTRICT void*
rpcalloc(size_t num, size_t size) RPMALLOC_ATTRIBUTE;
extern void*
rprealloc(void* ptr, size_t size);
extern void*
rpaligned_realloc(void* ptr, size_t alignment, size_t size, size_t oldsize, unsigned int flags);
extern RPMALLOC_RESTRICT void*
rpaligned_alloc(size_t alignment, size_t size) RPMALLOC_ATTRIBUTE;
extern RPMALLOC_RESTRICT void*
rpmemalign(size_t alignment, size_t size) RPMALLOC_ATTRIBUTE;
extern int
rpposix_memalign(void **memptr, size_t alignment, size_t size);
extern size_t
rpmalloc_usable_size(void* ptr);
}

33
common/TracyAlloc.hpp Normal file
View File

@@ -0,0 +1,33 @@
#ifndef __TRACYALLOC_HPP__
#define __TRACYALLOC_HPP__
#include <cstdlib>
#ifdef TRACY_ENABLE
# include "../client/tracy_rpmalloc.hpp"
#endif
namespace tracy
{
static inline void* tracy_malloc( size_t size )
{
#ifdef TRACY_ENABLE
return rpmalloc( size );
#else
return malloc( size );
#endif
}
static inline void tracy_free( void* ptr )
{
#ifdef TRACY_ENABLE
rpfree( ptr );
#else
free( ptr );
#endif
}
}
#endif

View File

@@ -10,6 +10,15 @@ namespace tracy
using TracyMutex = std::shared_mutex;
}
#elif defined __CYGWIN__
#include "tracy_benaphore.h"
namespace tracy
{
using TracyMutex = NonRecursiveBenaphore;
}
#else
#include <mutex>

62
common/TracyProtocol.hpp Normal file
View File

@@ -0,0 +1,62 @@
#ifndef __TRACYPROTOCOL_HPP__
#define __TRACYPROTOCOL_HPP__
#include <limits>
#include <stdint.h>
#include "../common/tracy_lz4.hpp"
namespace tracy
{
using lz4sz_t = uint32_t;
enum { TargetFrameSize = 256 * 1024 };
enum { LZ4Size = LZ4_COMPRESSBOUND( TargetFrameSize ) };
static_assert( LZ4Size <= std::numeric_limits<lz4sz_t>::max(), "LZ4Size greater than lz4sz_t" );
static_assert( TargetFrameSize * 2 >= 64 * 1024, "Not enough space for LZ4 stream buffer" );
enum ServerQuery : uint8_t
{
ServerQueryTerminate,
ServerQueryString,
ServerQueryThreadString,
ServerQuerySourceLocation,
ServerQueryPlotName,
ServerQueryCallstackFrame,
ServerQueryFrameName,
};
enum { WelcomeMessageProgramNameSize = 64 };
enum { WelcomeMessageHostInfoSize = 1024 };
#pragma pack( 1 )
struct WelcomeMessage
{
double timerMul;
int64_t initBegin;
int64_t initEnd;
uint64_t delay;
uint64_t resolution;
uint64_t epoch;
uint8_t onDemand;
char programName[WelcomeMessageProgramNameSize];
char hostInfo[WelcomeMessageHostInfoSize];
};
enum { WelcomeMessageSize = sizeof( WelcomeMessage ) };
struct OnDemandPayloadMessage
{
uint64_t frames;
};
enum { OnDemandPayloadMessageSize = sizeof( OnDemandPayloadMessage ) };
#pragma pack()
}
#endif

346
common/TracyQueue.hpp Normal file
View File

@@ -0,0 +1,346 @@
#ifndef __TRACYQUEUE_HPP__
#define __TRACYQUEUE_HPP__
#include <stdint.h>
namespace tracy
{
enum class QueueType : uint8_t
{
ZoneText,
ZoneName,
Message,
ZoneBeginAllocSrcLoc,
CallstackMemory,
Callstack,
Terminate,
KeepAlive,
Crash,
CrashReport,
ZoneBegin,
ZoneBeginCallstack,
ZoneEnd,
FrameMarkMsg,
FrameMarkMsgStart,
FrameMarkMsgEnd,
SourceLocation,
LockAnnounce,
LockWait,
LockObtain,
LockRelease,
LockSharedWait,
LockSharedObtain,
LockSharedRelease,
LockMark,
PlotData,
MessageLiteral,
GpuNewContext,
GpuZoneBegin,
GpuZoneBeginCallstack,
GpuZoneEnd,
GpuTime,
MemAlloc,
MemFree,
MemAllocCallstack,
MemFreeCallstack,
CallstackFrame,
StringData,
ThreadName,
CustomStringData,
PlotName,
SourceLocationPayload,
CallstackPayload,
FrameName,
NUM_TYPES
};
#pragma pack( 1 )
struct QueueZoneBegin
{
int64_t time;
uint64_t thread;
uint64_t srcloc; // ptr
uint32_t cpu;
};
struct QueueZoneEnd
{
int64_t time;
uint64_t thread;
uint32_t cpu;
};
struct QueueStringTransfer
{
uint64_t ptr;
};
struct QueueFrameMark
{
int64_t time;
uint64_t name; // ptr
};
struct QueueSourceLocation
{
uint64_t name;
uint64_t function; // ptr
uint64_t file; // ptr
uint32_t line;
uint8_t r;
uint8_t g;
uint8_t b;
};
struct QueueZoneText
{
uint64_t thread;
uint64_t text; // ptr
};
enum class LockType : uint8_t
{
Lockable,
SharedLockable
};
struct QueueLockAnnounce
{
uint32_t id;
uint64_t lckloc; // ptr
LockType type;
};
struct QueueLockWait
{
uint32_t id;
int64_t time;
uint64_t thread;
LockType type;
};
struct QueueLockObtain
{
uint32_t id;
int64_t time;
uint64_t thread;
};
struct QueueLockRelease
{
uint32_t id;
int64_t time;
uint64_t thread;
};
struct QueueLockMark
{
uint32_t id;
uint64_t thread;
uint64_t srcloc; // ptr
};
enum class PlotDataType : uint8_t
{
Float,
Double,
Int
};
struct QueuePlotData
{
uint64_t name; // ptr
int64_t time;
PlotDataType type;
union
{
double d;
float f;
int64_t i;
} data;
};
struct QueueMessage
{
int64_t time;
uint64_t thread;
uint64_t text; // ptr
};
struct QueueGpuNewContext
{
int64_t cpuTime;
int64_t gpuTime;
uint64_t thread;
float period;
uint8_t context;
uint8_t accuracyBits;
};
struct QueueGpuZoneBegin
{
int64_t cpuTime;
uint64_t srcloc;
uint64_t thread;
uint16_t queryId;
uint8_t context;
};
struct QueueGpuZoneEnd
{
int64_t cpuTime;
uint16_t queryId;
uint8_t context;
};
struct QueueGpuTime
{
int64_t gpuTime;
uint16_t queryId;
uint8_t context;
};
struct QueueMemAlloc
{
int64_t time;
uint64_t thread;
uint64_t ptr;
char size[6];
};
struct QueueMemFree
{
int64_t time;
uint64_t thread;
uint64_t ptr;
};
struct QueueCallstackMemory
{
uint64_t ptr;
};
struct QueueCallstack
{
uint64_t ptr;
uint64_t thread;
};
struct QueueCallstackFrame
{
uint64_t ptr;
uint64_t name;
uint64_t file;
uint32_t line;
};
struct QueueCrashReport
{
int64_t time;
uint64_t thread;
uint64_t text; // ptr
};
struct QueueHeader
{
union
{
QueueType type;
uint8_t idx;
};
};
struct QueueItem
{
QueueHeader hdr;
union
{
QueueZoneBegin zoneBegin;
QueueZoneEnd zoneEnd;
QueueStringTransfer stringTransfer;
QueueFrameMark frameMark;
QueueSourceLocation srcloc;
QueueZoneText zoneText;
QueueLockAnnounce lockAnnounce;
QueueLockWait lockWait;
QueueLockObtain lockObtain;
QueueLockRelease lockRelease;
QueueLockMark lockMark;
QueuePlotData plotData;
QueueMessage message;
QueueGpuNewContext gpuNewContext;
QueueGpuZoneBegin gpuZoneBegin;
QueueGpuZoneEnd gpuZoneEnd;
QueueGpuTime gpuTime;
QueueMemAlloc memAlloc;
QueueMemFree memFree;
QueueCallstackMemory callstackMemory;
QueueCallstack callstack;
QueueCallstackFrame callstackFrame;
QueueCrashReport crashReport;
};
};
#pragma pack()
enum { QueueItemSize = sizeof( QueueItem ) };
static const size_t QueueDataSize[] = {
sizeof( QueueHeader ) + sizeof( QueueZoneText ),
sizeof( QueueHeader ) + sizeof( QueueZoneText ), // zone name
sizeof( QueueHeader ) + sizeof( QueueMessage ),
sizeof( QueueHeader ) + sizeof( QueueZoneBegin ), // allocated source location
sizeof( QueueHeader ) + sizeof( QueueCallstackMemory ),
sizeof( QueueHeader ) + sizeof( QueueCallstack ),
// above items must be first
sizeof( QueueHeader ), // terminate
sizeof( QueueHeader ), // keep alive
sizeof( QueueHeader ), // crash
sizeof( QueueHeader ) + sizeof( QueueCrashReport ),
sizeof( QueueHeader ) + sizeof( QueueZoneBegin ),
sizeof( QueueHeader ) + sizeof( QueueZoneBegin ), // callstack
sizeof( QueueHeader ) + sizeof( QueueZoneEnd ),
sizeof( QueueHeader ) + sizeof( QueueFrameMark ), // continuous frames
sizeof( QueueHeader ) + sizeof( QueueFrameMark ), // start
sizeof( QueueHeader ) + sizeof( QueueFrameMark ), // end
sizeof( QueueHeader ) + sizeof( QueueSourceLocation ),
sizeof( QueueHeader ) + sizeof( QueueLockAnnounce ),
sizeof( QueueHeader ) + sizeof( QueueLockWait ),
sizeof( QueueHeader ) + sizeof( QueueLockObtain ),
sizeof( QueueHeader ) + sizeof( QueueLockRelease ),
sizeof( QueueHeader ) + sizeof( QueueLockWait ),
sizeof( QueueHeader ) + sizeof( QueueLockObtain ),
sizeof( QueueHeader ) + sizeof( QueueLockRelease ),
sizeof( QueueHeader ) + sizeof( QueueLockMark ),
sizeof( QueueHeader ) + sizeof( QueuePlotData ),
sizeof( QueueHeader ) + sizeof( QueueMessage ), // literal
sizeof( QueueHeader ) + sizeof( QueueGpuNewContext ),
sizeof( QueueHeader ) + sizeof( QueueGpuZoneBegin ),
sizeof( QueueHeader ) + sizeof( QueueGpuZoneBegin ), // callstack
sizeof( QueueHeader ) + sizeof( QueueGpuZoneEnd ),
sizeof( QueueHeader ) + sizeof( QueueGpuTime ),
sizeof( QueueHeader ) + sizeof( QueueMemAlloc ),
sizeof( QueueHeader ) + sizeof( QueueMemFree ),
sizeof( QueueHeader ) + sizeof( QueueMemAlloc ), // callstack
sizeof( QueueHeader ) + sizeof( QueueMemFree ), // callstack
sizeof( QueueHeader ) + sizeof( QueueCallstackFrame ),
// keep all QueueStringTransfer below
sizeof( QueueHeader ) + sizeof( QueueStringTransfer ), // string data
sizeof( QueueHeader ) + sizeof( QueueStringTransfer ), // thread name
sizeof( QueueHeader ) + sizeof( QueueStringTransfer ), // custom string data
sizeof( QueueHeader ) + sizeof( QueueStringTransfer ), // plot name
sizeof( QueueHeader ) + sizeof( QueueStringTransfer ), // allocated source location payload
sizeof( QueueHeader ) + sizeof( QueueStringTransfer ), // callstack payload
sizeof( QueueHeader ) + sizeof( QueueStringTransfer ), // frame name
};
static_assert( QueueItemSize == 32, "Queue item size not 32 bytes" );
static_assert( sizeof( QueueDataSize ) / sizeof( size_t ) == (uint8_t)QueueType::NUM_TYPES, "QueueDataSize mismatch" );
static_assert( sizeof( void* ) <= sizeof( uint64_t ), "Pointer size > 8 bytes" );
static_assert( sizeof( void* ) == sizeof( uintptr_t ), "Pointer size != uintptr_t" );
};
#endif

335
common/TracySocket.cpp Normal file
View File

@@ -0,0 +1,335 @@
#include <algorithm>
#include <assert.h>
#include <new>
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <sys/types.h>
#include "TracyAlloc.hpp"
#include "TracySocket.hpp"
#ifdef _MSC_VER
# include <winsock2.h>
# include <ws2tcpip.h>
# pragma warning(disable:4244)
# pragma warning(disable:4267)
#else
# include <sys/socket.h>
# include <netdb.h>
# include <unistd.h>
#endif
#ifndef MSG_NOSIGNAL
# define MSG_NOSIGNAL 0
#endif
namespace tracy
{
#ifdef _MSC_VER
typedef SOCKET socket_t;
#else
typedef int socket_t;
#endif
#ifdef _MSC_VER
struct __wsinit
{
__wsinit()
{
WSADATA wsaData;
if( WSAStartup( MAKEWORD( 2, 2 ), &wsaData ) != 0 )
{
fprintf( stderr, "Cannot init winsock.\n" );
exit( 1 );
}
}
};
void InitWinSock()
{
static __wsinit init;
}
#endif
Socket::Socket()
: m_sock( -1 )
, m_buf( (char*)tracy_malloc( BufSize ) )
, m_bufPtr( nullptr )
, m_bufLeft( 0 )
{
#ifdef _MSC_VER
InitWinSock();
#endif
}
Socket::Socket( int sock )
: m_sock( sock )
, m_buf( (char*)tracy_malloc( BufSize ) )
, m_bufPtr( nullptr )
, m_bufLeft( 0 )
{
}
Socket::~Socket()
{
tracy_free( m_buf );
if( m_sock != -1 )
{
Close();
}
}
bool Socket::Connect( const char* addr, const char* port )
{
assert( m_sock == -1 );
struct addrinfo hints;
struct addrinfo *res, *ptr;
memset( &hints, 0, sizeof( hints ) );
hints.ai_family = AF_UNSPEC;
hints.ai_socktype = SOCK_STREAM;
if( getaddrinfo( addr, port, &hints, &res ) != 0 ) return false;
int sock = 0;
for( ptr = res; ptr; ptr = ptr->ai_next )
{
if( ( sock = socket( ptr->ai_family, ptr->ai_socktype, ptr->ai_protocol ) ) == -1 ) continue;
#if defined __APPLE__
int val = 1;
setsockopt( m_sock, SOL_SOCKET, SO_NOSIGPIPE, &val, sizeof( val ) );
#endif
if( connect( sock, ptr->ai_addr, ptr->ai_addrlen ) == -1 )
{
#ifdef _MSC_VER
closesocket( sock );
#else
close( sock );
#endif
continue;
}
break;
}
freeaddrinfo( res );
if( !ptr ) return false;
m_sock = sock;
return true;
}
void Socket::Close()
{
assert( m_sock != -1 );
#ifdef _MSC_VER
closesocket( m_sock );
#else
close( m_sock );
#endif
m_sock = -1;
}
int Socket::Send( const void* _buf, int len )
{
auto buf = (const char*)_buf;
assert( m_sock != -1 );
auto start = buf;
while( len > 0 )
{
auto ret = send( m_sock, buf, len, MSG_NOSIGNAL );
if( ret == -1 ) return -1;
len -= ret;
buf += ret;
}
return int( buf - start );
}
int Socket::RecvBuffered( void* buf, int len, const timeval* tv )
{
if( len <= m_bufLeft )
{
memcpy( buf, m_bufPtr, len );
m_bufPtr += len;
m_bufLeft -= len;
return len;
}
if( m_bufLeft > 0 )
{
memcpy( buf, m_bufPtr, m_bufLeft );
const auto ret = m_bufLeft;
m_bufLeft = 0;
return ret;
}
if( len >= BufSize ) return Recv( buf, len, tv );
m_bufLeft = Recv( m_buf, BufSize, tv );
if( m_bufLeft <= 0 ) return m_bufLeft;
const auto sz = std::min( len, m_bufLeft );
memcpy( buf, m_buf, sz );
m_bufPtr = m_buf + sz;
m_bufLeft -= sz;
return sz;
}
int Socket::Recv( void* _buf, int len, const timeval* tv )
{
auto buf = (char*)_buf;
fd_set fds;
FD_ZERO( &fds );
FD_SET( static_cast<socket_t>(m_sock), &fds );
#ifndef _WIN32
timeval _tv = *tv;
select( m_sock+1, &fds, nullptr, nullptr, &_tv );
#else
select( m_sock+1, &fds, nullptr, nullptr, tv );
#endif
if( FD_ISSET( m_sock, &fds ) )
{
return recv( m_sock, buf, len, 0 );
}
else
{
return -1;
}
}
bool Socket::Read( void* _buf, int len, const timeval* tv, std::function<bool()> exitCb )
{
auto buf = (char*)_buf;
while( len > 0 )
{
if( exitCb() ) return false;
const auto sz = RecvBuffered( buf, len, tv );
switch( sz )
{
case 0:
return false;
case -1:
#ifdef _WIN32
{
auto err = WSAGetLastError();
if( err == WSAECONNABORTED || err == WSAECONNRESET ) return false;
}
#endif
break;
default:
len -= sz;
buf += sz;
break;
}
}
return true;
}
bool Socket::HasData()
{
if( m_bufLeft > 0 ) return true;
struct timeval tv;
memset( &tv, 0, sizeof( tv ) );
fd_set fds;
FD_ZERO( &fds );
FD_SET( static_cast<socket_t>(m_sock), &fds );
return select( m_sock+1, &fds, nullptr, nullptr, &tv ) > 0;
}
ListenSocket::ListenSocket()
: m_sock( -1 )
{
#ifdef _MSC_VER
InitWinSock();
#endif
}
ListenSocket::~ListenSocket()
{
}
bool ListenSocket::Listen( const char* port, int backlog )
{
assert( m_sock == -1 );
struct addrinfo* res;
struct addrinfo hints;
memset( &hints, 0, sizeof( hints ) );
hints.ai_family = AF_INET6;
hints.ai_socktype = SOCK_STREAM;
hints.ai_flags = AI_PASSIVE;
if( getaddrinfo( nullptr, port, &hints, &res ) != 0 ) return false;
m_sock = socket( res->ai_family, res->ai_socktype, res->ai_protocol );
#if defined _MSC_VER || defined __CYGWIN__
unsigned long val = 0;
setsockopt( m_sock, IPPROTO_IPV6, IPV6_V6ONLY, (const char*)&val, sizeof( val ) );
#else
int val = 1;
setsockopt( m_sock, SOL_SOCKET, SO_REUSEADDR, &val, sizeof( val ) );
#endif
if( bind( m_sock, res->ai_addr, res->ai_addrlen ) == -1 ) return false;
if( listen( m_sock, backlog ) == -1 ) return false;
return true;
}
Socket* ListenSocket::Accept()
{
struct sockaddr_storage remote;
socklen_t sz = sizeof( remote );
struct timeval tv;
tv.tv_sec = 0;
tv.tv_usec = 10000;
fd_set fds;
FD_ZERO( &fds );
FD_SET( static_cast<socket_t>(m_sock), &fds );
select( m_sock+1, &fds, nullptr, nullptr, &tv );
if( FD_ISSET( m_sock, &fds ) )
{
int sock = accept( m_sock, (sockaddr*)&remote, &sz);
#if defined __APPLE__
int val = 1;
setsockopt( sock, SOL_SOCKET, SO_NOSIGPIPE, &val, sizeof( val ) );
#endif
if( sock == -1 )
{
return nullptr;
}
else
{
auto ptr = (Socket*)tracy_malloc( sizeof( Socket ) );
new(ptr) Socket( sock );
return ptr;
}
}
else
{
return nullptr;
}
}
void ListenSocket::Close()
{
assert( m_sock != -1 );
#ifdef _MSC_VER
closesocket( m_sock );
#else
close( m_sock );
#endif
m_sock = -1;
}
}

69
common/TracySocket.hpp Normal file
View File

@@ -0,0 +1,69 @@
#ifndef __TRACYSOCKET_HPP__
#define __TRACYSOCKET_HPP__
#include <functional>
struct timeval;
namespace tracy
{
#ifdef _MSC_VER
void InitWinSock();
#endif
class Socket
{
enum { BufSize = 128 * 1024 };
public:
Socket();
Socket( int sock );
~Socket();
bool Connect( const char* addr, const char* port );
void Close();
int Send( const void* buf, int len );
bool Read( void* buf, int len, const timeval* tv, std::function<bool()> exitCb );
bool HasData();
Socket( const Socket& ) = delete;
Socket( Socket&& ) = delete;
Socket& operator=( const Socket& ) = delete;
Socket& operator=( Socket&& ) = delete;
private:
int RecvBuffered( void* buf, int len, const timeval* tv );
int Recv( void* buf, int len, const timeval* tv );
int m_sock;
char* m_buf;
char* m_bufPtr;
int m_bufLeft;
};
class ListenSocket
{
public:
ListenSocket();
~ListenSocket();
bool Listen( const char* port, int backlog );
Socket* Accept();
void Close();
ListenSocket( const ListenSocket& ) = delete;
ListenSocket( ListenSocket&& ) = delete;
ListenSocket& operator=( const ListenSocket& ) = delete;
ListenSocket& operator=( ListenSocket&& ) = delete;
private:
int m_sock;
};
}
#endif

183
common/TracySystem.cpp Normal file
View File

@@ -0,0 +1,183 @@
#if defined _MSC_VER || defined __CYGWIN__ || defined _WIN32
# ifndef WIN32_LEAN_AND_MEAN
# define WIN32_LEAN_AND_MEAN
# endif
# ifndef NOMINMAX
# define NOMINMAX
# endif
#endif
#ifdef _WIN32
# include <windows.h>
#else
# include <pthread.h>
# include <string.h>
# include <unistd.h>
#endif
#ifdef __linux__
# ifndef __ANDROID__
# include <syscall.h>
# endif
# include <fcntl.h>
#endif
#include <inttypes.h>
#include <stdio.h>
#include "TracySystem.hpp"
#ifdef TRACY_COLLECT_THREAD_NAMES
# include <atomic>
# include "TracyAlloc.hpp"
#endif
namespace tracy
{
#ifdef TRACY_COLLECT_THREAD_NAMES
struct ThreadNameData
{
uint64_t id;
const char* name;
ThreadNameData* next;
};
extern std::atomic<ThreadNameData*>& s_threadNameData;
#endif
void SetThreadName( std::thread& thread, const char* name )
{
SetThreadName( thread.native_handle(), name );
}
void SetThreadName( std::thread::native_handle_type handle, const char* name )
{
#ifdef _WIN32
# if defined NTDDI_WIN10_RS2 && NTDDI_VERSION >= NTDDI_WIN10_RS2
wchar_t buf[256];
mbstowcs( buf, name, 256 );
SetThreadDescription( static_cast<HANDLE>( handle ), buf );
# else
const DWORD MS_VC_EXCEPTION=0x406D1388;
# pragma pack( push, 8 )
struct THREADNAME_INFO
{
DWORD dwType;
LPCSTR szName;
DWORD dwThreadID;
DWORD dwFlags;
};
# pragma pack(pop)
DWORD ThreadId = GetThreadId( static_cast<HANDLE>( handle ) );
THREADNAME_INFO info;
info.dwType = 0x1000;
info.szName = name;
info.dwThreadID = ThreadId;
info.dwFlags = 0;
__try
{
RaiseException( MS_VC_EXCEPTION, 0, sizeof(info)/sizeof(ULONG_PTR), (ULONG_PTR*)&info );
}
__except(EXCEPTION_EXECUTE_HANDLER)
{
}
# endif
#elif defined _GNU_SOURCE && !defined __EMSCRIPTEN__
{
const auto sz = strlen( name );
if( sz <= 15 )
{
pthread_setname_np( handle, name );
}
else
{
char buf[16];
memcpy( buf, name, 15 );
buf[15] = '\0';
pthread_setname_np( handle, buf );
}
}
#endif
#ifdef TRACY_COLLECT_THREAD_NAMES
{
rpmalloc_thread_initialize();
const auto sz = strlen( name );
char* buf = (char*)tracy_malloc( sz+1 );
memcpy( buf, name, sz );
buf[sz+1] = '\0';
auto data = (ThreadNameData*)tracy_malloc( sizeof( ThreadNameData ) );
# ifdef _WIN32
data->id = GetThreadId( static_cast<HANDLE>( handle ) );
# elif defined __APPLE__
pthread_threadid_np( handle, &data->id );
# else
data->id = (uint64_t)handle;
# endif
data->name = buf;
data->next = s_threadNameData.load( std::memory_order_relaxed );
while( !s_threadNameData.compare_exchange_weak( data->next, data, std::memory_order_release, std::memory_order_relaxed ) ) {}
}
#endif
}
const char* GetThreadName( uint64_t id )
{
static char buf[256];
#ifdef TRACY_COLLECT_THREAD_NAMES
auto ptr = s_threadNameData.load( std::memory_order_relaxed );
while( ptr )
{
if( ptr->id == id )
{
return ptr->name;
}
ptr = ptr->next;
}
#else
# ifdef _WIN32
# if defined NTDDI_WIN10_RS2 && NTDDI_VERSION >= NTDDI_WIN10_RS2
auto hnd = OpenThread( THREAD_QUERY_LIMITED_INFORMATION, FALSE, (DWORD)id );
if( hnd != 0 )
{
PWSTR tmp;
GetThreadDescription( hnd, &tmp );
auto ret = wcstombs( buf, tmp, 256 );
CloseHandle( hnd );
if( ret != 0 )
{
return buf;
}
}
# endif
# elif defined __GLIBC__ && !defined __ANDROID__ && !defined __EMSCRIPTEN__
if( pthread_getname_np( (pthread_t)id, buf, 256 ) == 0 )
{
return buf;
}
# elif defined __linux__
int cs, fd;
char path[32];
# ifdef __ANDROID__
int tid = gettid();
# else
int tid = (int) syscall( SYS_gettid );
# endif
snprintf( path, sizeof( path ), "/proc/self/task/%d/comm", tid );
sprintf( buf, "%" PRIu64, id );
pthread_setcancelstate( PTHREAD_CANCEL_DISABLE, &cs );
if ( ( fd = open( path, O_RDONLY ) ) > 0) {
int len = read( fd, buf, 255 );
if ( len > 0 )
buf[len] = 0;
close( fd );
}
pthread_setcancelstate( cs, 0 );
return buf;
# endif
#endif
sprintf( buf, "%" PRIu64, id );
return buf;
}
}

45
common/TracySystem.hpp Normal file
View File

@@ -0,0 +1,45 @@
#ifndef __TRACYSYSTEM_HPP__
#define __TRACYSYSTEM_HPP__
#ifdef TRACY_ENABLE
# if defined __ANDROID__ || defined __CYGWIN__ || defined __APPLE__ || defined _GNU_SOURCE || ( defined _WIN32 && ( !defined NTDDI_WIN10_RS2 || NTDDI_VERSION < NTDDI_WIN10_RS2 ) )
# define TRACY_COLLECT_THREAD_NAMES
# endif
#endif
#ifdef _WIN32
#ifndef _WINDOWS_
extern "C" __declspec(dllimport) unsigned long __stdcall GetCurrentThreadId(void);
#endif
#else
# include <pthread.h>
#endif
#include <stdint.h>
#include <thread>
namespace tracy
{
static inline uint64_t GetThreadHandle()
{
#ifdef _WIN32
static_assert( sizeof( decltype( GetCurrentThreadId() ) ) <= sizeof( uint64_t ), "Thread handle too big to fit in protocol" );
return uint64_t( GetCurrentThreadId() );
#elif defined __APPLE__
uint64_t id;
pthread_threadid_np( pthread_self(), &id );
return id;
#else
static_assert( sizeof( decltype( pthread_self() ) ) <= sizeof( uint64_t ), "Thread handle too big to fit in protocol" );
return uint64_t( pthread_self() );
#endif
}
void SetThreadName( std::thread& thread, const char* name );
void SetThreadName( std::thread::native_handle_type handle, const char* name );
const char* GetThreadName( uint64_t id );
}
#endif

View File

@@ -1,21 +0,0 @@
# Extract the actual list of source files from a sibling Visual Studio project.
# Ensure these are simply-substituted variables, without changing their values.
SRC := $(SRC)
SRC2 := $(SRC2)
SRC3 := $(SRC3)
SRC4 := $(SRC4)
# Paths here are relative to the directory in which make was invoked, not to
# this file, so ../win32/$(PROJECT).vcxproj refers to the Visual Studio project
# of whichever tool is including this makefile fragment.
BASE := $(shell grep -E 'ClCompile.*cpp"' ../win32/$(PROJECT).vcxproj | sed -e 's/.*\"\(.*\)\".*/\1/' | sed -e 's@\\@/@g')
BASE2 := $(shell grep -E 'ClCompile.*c"' ../win32/$(PROJECT).vcxproj | sed -e 's/.*\"\(.*\)\".*/\1/' | sed -e 's@\\@/@g')
BASE4 := $(shell grep -E 'None.*S"' ../win32/$(PROJECT).vcxproj | sed -e 's/.*\"\(.*\)\".*/\1/' | sed -e 's@\\@/@g')
# The tool-specific makefile may request that certain files be omitted.
SRC += $(filter-out $(FILTER),$(BASE))
SRC2 += $(filter-out $(FILTER),$(BASE2))
SRC3 += $(filter-out $(FILTER),$(BASE3))
SRC4 += $(filter-out $(FILTER),$(BASE4))

68
common/tracy_benaphore.h Normal file
View File

@@ -0,0 +1,68 @@
// Copyright (c) 2015 Jeff Preshing
//
// This software is provided 'as-is', without any express or implied
// warranty. In no event will the authors be held liable for any damages
// arising from the use of this software.
//
// Permission is granted to anyone to use this software for any purpose,
// including commercial applications, and to alter it and redistribute it
// freely, subject to the following restrictions:
//
// 1. The origin of this software must not be misrepresented; you must not
// claim that you wrote the original software. If you use this software
// in a product, an acknowledgement in the product documentation would be
// appreciated but is not required.
// 2. Altered source versions must be plainly marked as such, and must not be
// misrepresented as being the original software.
// 3. This notice may not be removed or altered from any source distribution.
#ifndef __TRACY_CPP11OM_BENAPHORE_H__
#define __TRACY_CPP11OM_BENAPHORE_H__
#include <cassert>
#include <thread>
#include <atomic>
#include "tracy_sema.h"
namespace tracy
{
class NonRecursiveBenaphore
{
private:
std::atomic<int> m_contentionCount;
DefaultSemaphoreType m_sema;
public:
NonRecursiveBenaphore() : m_contentionCount(0) {}
void lock()
{
if (m_contentionCount.fetch_add(1, std::memory_order_acquire) > 0)
{
m_sema.wait();
}
}
bool try_lock()
{
if (m_contentionCount.load(std::memory_order_relaxed) != 0)
return false;
int expected = 0;
return m_contentionCount.compare_exchange_strong(expected, 1, std::memory_order_acquire);
}
void unlock()
{
int oldCount = m_contentionCount.fetch_sub(1, std::memory_order_release);
assert(oldCount > 0);
if (oldCount > 1)
{
m_sema.signal();
}
}
};
}
#endif // __CPP11OM_BENAPHORE_H__

1918
common/tracy_lz4.cpp Normal file

File diff suppressed because it is too large Load Diff

604
common/tracy_lz4.hpp Normal file
View File

@@ -0,0 +1,604 @@
/*
* LZ4 - Fast LZ compression algorithm
* Header File
* Copyright (C) 2011-2017, Yann Collet.
BSD 2-Clause License (http://www.opensource.org/licenses/bsd-license.php)
Redistribution and use in source and binary forms, with or without
modification, are permitted provided that the following conditions are
met:
* Redistributions of source code must retain the above copyright
notice, this list of conditions and the following disclaimer.
* Redistributions in binary form must reproduce the above
copyright notice, this list of conditions and the following disclaimer
in the documentation and/or other materials provided with the
distribution.
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
You can contact the author at :
- LZ4 homepage : http://www.lz4.org
- LZ4 source repository : https://github.com/lz4/lz4
*/
#ifndef TRACY_LZ4_H_2983827168210
#define TRACY_LZ4_H_2983827168210
/* --- Dependency --- */
#include <stddef.h> /* size_t */
#include <stdint.h>
namespace tracy
{
/**
Introduction
LZ4 is lossless compression algorithm, providing compression speed at 400 MB/s per core,
scalable with multi-cores CPU. It features an extremely fast decoder, with speed in
multiple GB/s per core, typically reaching RAM speed limits on multi-core systems.
The LZ4 compression library provides in-memory compression and decompression functions.
Compression can be done in:
- a single step (described as Simple Functions)
- a single step, reusing a context (described in Advanced Functions)
- unbounded multiple steps (described as Streaming compression)
lz4.h provides block compression functions. It gives full buffer control to user.
Decompressing an lz4-compressed block also requires metadata (such as compressed size).
Each application is free to encode such metadata in whichever way it wants.
An additional format, called LZ4 frame specification (doc/lz4_Frame_format.md),
take care of encoding standard metadata alongside LZ4-compressed blocks.
If your application requires interoperability, it's recommended to use it.
A library is provided to take care of it, see lz4frame.h.
*/
/*^***************************************************************
* Export parameters
*****************************************************************/
/*
* LZ4_DLL_EXPORT :
* Enable exporting of functions when building a Windows DLL
* LZ4LIB_VISIBILITY :
* Control library symbols visibility.
*/
#ifndef LZ4LIB_VISIBILITY
# if defined(__GNUC__) && (__GNUC__ >= 4)
# define LZ4LIB_VISIBILITY __attribute__ ((visibility ("default")))
# else
# define LZ4LIB_VISIBILITY
# endif
#endif
#if defined(LZ4_DLL_EXPORT) && (LZ4_DLL_EXPORT==1)
# define LZ4LIB_API __declspec(dllexport) LZ4LIB_VISIBILITY
#elif defined(LZ4_DLL_IMPORT) && (LZ4_DLL_IMPORT==1)
# define LZ4LIB_API __declspec(dllimport) LZ4LIB_VISIBILITY /* It isn't required but allows to generate better code, saving a function pointer load from the IAT and an indirect jump.*/
#else
# define LZ4LIB_API LZ4LIB_VISIBILITY
#endif
/*------ Version ------*/
#define LZ4_VERSION_MAJOR 1 /* for breaking interface changes */
#define LZ4_VERSION_MINOR 8 /* for new (non-breaking) interface capabilities */
#define LZ4_VERSION_RELEASE 2 /* for tweaks, bug-fixes, or development */
#define LZ4_VERSION_NUMBER (LZ4_VERSION_MAJOR *100*100 + LZ4_VERSION_MINOR *100 + LZ4_VERSION_RELEASE)
#define LZ4_LIB_VERSION LZ4_VERSION_MAJOR.LZ4_VERSION_MINOR.LZ4_VERSION_RELEASE
#define LZ4_QUOTE(str) #str
#define LZ4_EXPAND_AND_QUOTE(str) LZ4_QUOTE(str)
#define LZ4_VERSION_STRING LZ4_EXPAND_AND_QUOTE(LZ4_LIB_VERSION)
LZ4LIB_API int LZ4_versionNumber (void); /**< library version number; useful to check dll version */
LZ4LIB_API const char* LZ4_versionString (void); /**< library version string; unseful to check dll version */
/*-************************************
* Tuning parameter
**************************************/
/*!
* LZ4_MEMORY_USAGE :
* Memory usage formula : N->2^N Bytes (examples : 10 -> 1KB; 12 -> 4KB ; 16 -> 64KB; 20 -> 1MB; etc.)
* Increasing memory usage improves compression ratio
* Reduced memory usage may improve speed, thanks to cache effect
* Default value is 14, for 16KB, which nicely fits into Intel x86 L1 cache
*/
#ifndef LZ4_MEMORY_USAGE
# define LZ4_MEMORY_USAGE 12
#endif
/*-************************************
* Simple Functions
**************************************/
/*! LZ4_compress_default() :
Compresses 'srcSize' bytes from buffer 'src'
into already allocated 'dst' buffer of size 'dstCapacity'.
Compression is guaranteed to succeed if 'dstCapacity' >= LZ4_compressBound(srcSize).
It also runs faster, so it's a recommended setting.
If the function cannot compress 'src' into a more limited 'dst' budget,
compression stops *immediately*, and the function result is zero.
Note : as a consequence, 'dst' content is not valid.
Note 2 : This function is protected against buffer overflow scenarios (never writes outside 'dst' buffer, nor read outside 'source' buffer).
srcSize : max supported value is LZ4_MAX_INPUT_SIZE.
dstCapacity : size of buffer 'dst' (which must be already allocated)
return : the number of bytes written into buffer 'dst' (necessarily <= dstCapacity)
or 0 if compression fails */
LZ4LIB_API int LZ4_compress_default(const char* src, char* dst, int srcSize, int dstCapacity);
/*! LZ4_decompress_safe() :
compressedSize : is the exact complete size of the compressed block.
dstCapacity : is the size of destination buffer, which must be already allocated.
return : the number of bytes decompressed into destination buffer (necessarily <= dstCapacity)
If destination buffer is not large enough, decoding will stop and output an error code (negative value).
If the source stream is detected malformed, the function will stop decoding and return a negative result.
This function is protected against malicious data packets.
*/
LZ4LIB_API int LZ4_decompress_safe (const char* src, char* dst, int compressedSize, int dstCapacity);
/*-************************************
* Advanced Functions
**************************************/
#define LZ4_MAX_INPUT_SIZE 0x7E000000 /* 2 113 929 216 bytes */
#define LZ4_COMPRESSBOUND(isize) ((unsigned)(isize) > (unsigned)LZ4_MAX_INPUT_SIZE ? 0 : (isize) + ((isize)/255) + 16)
/*!
LZ4_compressBound() :
Provides the maximum size that LZ4 compression may output in a "worst case" scenario (input data not compressible)
This function is primarily useful for memory allocation purposes (destination buffer size).
Macro LZ4_COMPRESSBOUND() is also provided for compilation-time evaluation (stack memory allocation for example).
Note that LZ4_compress_default() compresses faster when dstCapacity is >= LZ4_compressBound(srcSize)
inputSize : max supported value is LZ4_MAX_INPUT_SIZE
return : maximum output size in a "worst case" scenario
or 0, if input size is incorrect (too large or negative)
*/
LZ4LIB_API int LZ4_compressBound(int inputSize);
/*!
LZ4_compress_fast() :
Same as LZ4_compress_default(), but allows selection of "acceleration" factor.
The larger the acceleration value, the faster the algorithm, but also the lesser the compression.
It's a trade-off. It can be fine tuned, with each successive value providing roughly +~3% to speed.
An acceleration value of "1" is the same as regular LZ4_compress_default()
Values <= 0 will be replaced by ACCELERATION_DEFAULT (currently == 1, see lz4.c).
*/
LZ4LIB_API int LZ4_compress_fast (const char* src, char* dst, int srcSize, int dstCapacity, int acceleration);
/*!
LZ4_compress_fast_extState() :
Same compression function, just using an externally allocated memory space to store compression state.
Use LZ4_sizeofState() to know how much memory must be allocated,
and allocate it on 8-bytes boundaries (using malloc() typically).
Then, provide it as 'void* state' to compression function.
*/
LZ4LIB_API int LZ4_sizeofState(void);
LZ4LIB_API int LZ4_compress_fast_extState (void* state, const char* src, char* dst, int srcSize, int dstCapacity, int acceleration);
/*!
LZ4_compress_destSize() :
Reverse the logic : compresses as much data as possible from 'src' buffer
into already allocated buffer 'dst' of size 'targetDestSize'.
This function either compresses the entire 'src' content into 'dst' if it's large enough,
or fill 'dst' buffer completely with as much data as possible from 'src'.
*srcSizePtr : will be modified to indicate how many bytes where read from 'src' to fill 'dst'.
New value is necessarily <= old value.
return : Nb bytes written into 'dst' (necessarily <= targetDestSize)
or 0 if compression fails
*/
LZ4LIB_API int LZ4_compress_destSize (const char* src, char* dst, int* srcSizePtr, int targetDstSize);
/*!
LZ4_decompress_fast() : **unsafe!**
This function is a bit faster than LZ4_decompress_safe(),
but it may misbehave on malformed input because it doesn't perform full validation of compressed data.
originalSize : is the uncompressed size to regenerate
Destination buffer must be already allocated, and its size must be >= 'originalSize' bytes.
return : number of bytes read from source buffer (== compressed size).
If the source stream is detected malformed, the function stops decoding and return a negative result.
note : This function is only usable if the originalSize of uncompressed data is known in advance.
The caller should also check that all the compressed input has been consumed properly,
i.e. that the return value matches the size of the buffer with compressed input.
The function never writes past the output buffer. However, since it doesn't know its 'src' size,
it may read past the intended input. Also, because match offsets are not validated during decoding,
reads from 'src' may underflow. Use this function in trusted environment **only**.
*/
LZ4LIB_API int LZ4_decompress_fast (const char* src, char* dst, int originalSize);
/*!
LZ4_decompress_safe_partial() :
This function decompress a compressed block of size 'srcSize' at position 'src'
into destination buffer 'dst' of size 'dstCapacity'.
The function will decompress a minimum of 'targetOutputSize' bytes, and stop after that.
However, it's not accurate, and may write more than 'targetOutputSize' (but always <= dstCapacity).
@return : the number of bytes decoded in the destination buffer (necessarily <= dstCapacity)
Note : this number can also be < targetOutputSize, if compressed block contains less data.
Therefore, always control how many bytes were decoded.
If source stream is detected malformed, function returns a negative result.
This function is protected against malicious data packets.
*/
LZ4LIB_API int LZ4_decompress_safe_partial (const char* src, char* dst, int srcSize, int targetOutputSize, int dstCapacity);
/*-*********************************************
* Streaming Compression Functions
***********************************************/
typedef union LZ4_stream_u LZ4_stream_t; /* incomplete type (defined later) */
/*! LZ4_createStream() and LZ4_freeStream() :
* LZ4_createStream() will allocate and initialize an `LZ4_stream_t` structure.
* LZ4_freeStream() releases its memory.
*/
LZ4LIB_API LZ4_stream_t* LZ4_createStream(void);
LZ4LIB_API int LZ4_freeStream (LZ4_stream_t* streamPtr);
/*! LZ4_resetStream() :
* An LZ4_stream_t structure can be allocated once and re-used multiple times.
* Use this function to start compressing a new stream.
*/
LZ4LIB_API void LZ4_resetStream (LZ4_stream_t* streamPtr);
/*! LZ4_loadDict() :
* Use this function to load a static dictionary into LZ4_stream_t.
* Any previous data will be forgotten, only 'dictionary' will remain in memory.
* Loading a size of 0 is allowed, and is the same as reset.
* @return : dictionary size, in bytes (necessarily <= 64 KB)
*/
LZ4LIB_API int LZ4_loadDict (LZ4_stream_t* streamPtr, const char* dictionary, int dictSize);
/*! LZ4_compress_fast_continue() :
* Compress 'src' content using data from previously compressed blocks, for better compression ratio.
* 'dst' buffer must be already allocated.
* If dstCapacity >= LZ4_compressBound(srcSize), compression is guaranteed to succeed, and runs faster.
*
* Important : The previous 64KB of compressed data is assumed to remain present and unmodified in memory!
*
* Special 1 : When input is a double-buffer, they can have any size, including < 64 KB.
* Make sure that buffers are separated by at least one byte.
* This way, each block only depends on previous block.
* Special 2 : If input buffer is a ring-buffer, it can have any size, including < 64 KB.
*
* @return : size of compressed block
* or 0 if there is an error (typically, cannot fit into 'dst').
* After an error, the stream status is invalid, it can only be reset or freed.
*/
LZ4LIB_API int LZ4_compress_fast_continue (LZ4_stream_t* streamPtr, const char* src, char* dst, int srcSize, int dstCapacity, int acceleration);
/*! LZ4_saveDict() :
* If last 64KB data cannot be guaranteed to remain available at its current memory location,
* save it into a safer place (char* safeBuffer).
* This is schematically equivalent to a memcpy() followed by LZ4_loadDict(),
* but is much faster, because LZ4_saveDict() doesn't need to rebuild tables.
* @return : saved dictionary size in bytes (necessarily <= maxDictSize), or 0 if error.
*/
LZ4LIB_API int LZ4_saveDict (LZ4_stream_t* streamPtr, char* safeBuffer, int maxDictSize);
/*-**********************************************
* Streaming Decompression Functions
* Bufferless synchronous API
************************************************/
typedef union LZ4_streamDecode_u LZ4_streamDecode_t; /* tracking context */
/*! LZ4_createStreamDecode() and LZ4_freeStreamDecode() :
* creation / destruction of streaming decompression tracking context.
* A tracking context can be re-used multiple times.
*/
LZ4LIB_API LZ4_streamDecode_t* LZ4_createStreamDecode(void);
LZ4LIB_API int LZ4_freeStreamDecode (LZ4_streamDecode_t* LZ4_stream);
/*! LZ4_setStreamDecode() :
* An LZ4_streamDecode_t context can be allocated once and re-used multiple times.
* Use this function to start decompression of a new stream of blocks.
* A dictionary can optionnally be set. Use NULL or size 0 for a reset order.
* Dictionary is presumed stable : it must remain accessible and unmodified during next decompression.
* @return : 1 if OK, 0 if error
*/
LZ4LIB_API int LZ4_setStreamDecode (LZ4_streamDecode_t* LZ4_streamDecode, const char* dictionary, int dictSize);
/*! LZ4_decoderRingBufferSize() : v1.8.2
* Note : in a ring buffer scenario (optional),
* blocks are presumed decompressed next to each other
* up to the moment there is not enough remaining space for next block (remainingSize < maxBlockSize),
* at which stage it resumes from beginning of ring buffer.
* When setting such a ring buffer for streaming decompression,
* provides the minimum size of this ring buffer
* to be compatible with any source respecting maxBlockSize condition.
* @return : minimum ring buffer size,
* or 0 if there is an error (invalid maxBlockSize).
*/
LZ4LIB_API int LZ4_decoderRingBufferSize(int maxBlockSize);
#define LZ4_DECODER_RING_BUFFER_SIZE(mbs) (65536 + 14 + (mbs)) /* for static allocation; mbs presumed valid */
/*! LZ4_decompress_*_continue() :
* These decoding functions allow decompression of consecutive blocks in "streaming" mode.
* A block is an unsplittable entity, it must be presented entirely to a decompression function.
* Decompression functions only accepts one block at a time.
* The last 64KB of previously decoded data *must* remain available and unmodified at the memory position where they were decoded.
* If less than 64KB of data has been decoded, all the data must be present.
*
* Special : if decompression side sets a ring buffer, it must respect one of the following conditions :
* - Decompression buffer size is _at least_ LZ4_decoderRingBufferSize(maxBlockSize).
* maxBlockSize is the maximum size of any single block. It can have any value > 16 bytes.
* In which case, encoding and decoding buffers do not need to be synchronized.
* Actually, data can be produced by any source compliant with LZ4 format specification, and respecting maxBlockSize.
* - Synchronized mode :
* Decompression buffer size is _exactly_ the same as compression buffer size,
* and follows exactly same update rule (block boundaries at same positions),
* and decoding function is provided with exact decompressed size of each block (exception for last block of the stream),
* _then_ decoding & encoding ring buffer can have any size, including small ones ( < 64 KB).
* - Decompression buffer is larger than encoding buffer, by a minimum of maxBlockSize more bytes.
* In which case, encoding and decoding buffers do not need to be synchronized,
* and encoding ring buffer can have any size, including small ones ( < 64 KB).
*
* Whenever these conditions are not possible,
* save the last 64KB of decoded data into a safe buffer where it can't be modified during decompression,
* then indicate where this data is saved using LZ4_setStreamDecode(), before decompressing next block.
*/
LZ4LIB_API int LZ4_decompress_safe_continue (LZ4_streamDecode_t* LZ4_streamDecode, const char* src, char* dst, int srcSize, int dstCapacity);
LZ4LIB_API int LZ4_decompress_fast_continue (LZ4_streamDecode_t* LZ4_streamDecode, const char* src, char* dst, int originalSize);
/*! LZ4_decompress_*_usingDict() :
* These decoding functions work the same as
* a combination of LZ4_setStreamDecode() followed by LZ4_decompress_*_continue()
* They are stand-alone, and don't need an LZ4_streamDecode_t structure.
* Dictionary is presumed stable : it must remain accessible and unmodified during next decompression.
*/
LZ4LIB_API int LZ4_decompress_safe_usingDict (const char* src, char* dst, int srcSize, int dstCapcity, const char* dictStart, int dictSize);
LZ4LIB_API int LZ4_decompress_fast_usingDict (const char* src, char* dst, int originalSize, const char* dictStart, int dictSize);
/*^**********************************************
* !!!!!! STATIC LINKING ONLY !!!!!!
***********************************************/
/*-************************************
* Unstable declarations
**************************************
* Declarations in this section should be considered unstable.
* Use at your own peril, etc., etc.
* They may be removed in the future.
* Their signatures may change.
**************************************/
#ifdef LZ4_STATIC_LINKING_ONLY
/*! LZ4_resetStream_fast() :
* Use this, like LZ4_resetStream(), to prepare a context for a new chain of
* calls to a streaming API (e.g., LZ4_compress_fast_continue()).
*
* Note:
* Using this in advance of a non- streaming-compression function is redundant,
* and potentially bad for performance, since they all perform their own custom
* reset internally.
*
* Differences from LZ4_resetStream():
* When an LZ4_stream_t is known to be in a internally coherent state,
* it can often be prepared for a new compression with almost no work, only
* sometimes falling back to the full, expensive reset that is always required
* when the stream is in an indeterminate state (i.e., the reset performed by
* LZ4_resetStream()).
*
* LZ4_streams are guaranteed to be in a valid state when:
* - returned from LZ4_createStream()
* - reset by LZ4_resetStream()
* - memset(stream, 0, sizeof(LZ4_stream_t)), though this is discouraged
* - the stream was in a valid state and was reset by LZ4_resetStream_fast()
* - the stream was in a valid state and was then used in any compression call
* that returned success
* - the stream was in an indeterminate state and was used in a compression
* call that fully reset the state (e.g., LZ4_compress_fast_extState()) and
* that returned success
*
* When a stream isn't known to be in a valid state, it is not safe to pass to
* any fastReset or streaming function. It must first be cleansed by the full
* LZ4_resetStream().
*/
LZ4LIB_API void LZ4_resetStream_fast (LZ4_stream_t* streamPtr);
/*! LZ4_compress_fast_extState_fastReset() :
* A variant of LZ4_compress_fast_extState().
*
* Using this variant avoids an expensive initialization step. It is only safe
* to call if the state buffer is known to be correctly initialized already
* (see above comment on LZ4_resetStream_fast() for a definition of "correctly
* initialized"). From a high level, the difference is that this function
* initializes the provided state with a call to something like
* LZ4_resetStream_fast() while LZ4_compress_fast_extState() starts with a
* call to LZ4_resetStream().
*/
LZ4LIB_API int LZ4_compress_fast_extState_fastReset (void* state, const char* src, char* dst, int srcSize, int dstCapacity, int acceleration);
/*! LZ4_attach_dictionary() :
* This is an experimental API that allows for the efficient use of a
* static dictionary many times.
*
* Rather than re-loading the dictionary buffer into a working context before
* each compression, or copying a pre-loaded dictionary's LZ4_stream_t into a
* working LZ4_stream_t, this function introduces a no-copy setup mechanism,
* in which the working stream references the dictionary stream in-place.
*
* Several assumptions are made about the state of the dictionary stream.
* Currently, only streams which have been prepared by LZ4_loadDict() should
* be expected to work.
*
* Alternatively, the provided dictionary stream pointer may be NULL, in which
* case any existing dictionary stream is unset.
*
* If a dictionary is provided, it replaces any pre-existing stream history.
* The dictionary contents are the only history that can be referenced and
* logically immediately precede the data compressed in the first subsequent
* compression call.
*
* The dictionary will only remain attached to the working stream through the
* first compression call, at the end of which it is cleared. The dictionary
* stream (and source buffer) must remain in-place / accessible / unchanged
* through the completion of the first compression call on the stream.
*/
LZ4LIB_API void LZ4_attach_dictionary(LZ4_stream_t *working_stream, const LZ4_stream_t *dictionary_stream);
#endif
/*-************************************
* Private definitions
**************************************
* Do not use these definitions.
* They are exposed to allow static allocation of `LZ4_stream_t` and `LZ4_streamDecode_t`.
* Using these definitions will expose code to API and/or ABI break in future versions of the library.
**************************************/
#define LZ4_HASHLOG (LZ4_MEMORY_USAGE-2)
#define LZ4_HASHTABLESIZE (1 << LZ4_MEMORY_USAGE)
#define LZ4_HASH_SIZE_U32 (1 << LZ4_HASHLOG) /* required as macro for static allocation */
#if defined(__cplusplus) || (defined (__STDC_VERSION__) && (__STDC_VERSION__ >= 199901L) /* C99 */)
typedef struct LZ4_stream_t_internal LZ4_stream_t_internal;
struct LZ4_stream_t_internal {
uint32_t hashTable[LZ4_HASH_SIZE_U32];
uint32_t currentOffset;
uint16_t initCheck;
uint16_t tableType;
const uint8_t* dictionary;
const LZ4_stream_t_internal* dictCtx;
uint32_t dictSize;
};
typedef struct {
const uint8_t* externalDict;
size_t extDictSize;
const uint8_t* prefixEnd;
size_t prefixSize;
} LZ4_streamDecode_t_internal;
#else
typedef struct LZ4_stream_t_internal LZ4_stream_t_internal;
struct LZ4_stream_t_internal {
unsigned int hashTable[LZ4_HASH_SIZE_U32];
unsigned int currentOffset;
unsigned short initCheck;
unsigned short tableType;
const unsigned char* dictionary;
const LZ4_stream_t_internal* dictCtx;
unsigned int dictSize;
};
typedef struct {
const unsigned char* externalDict;
size_t extDictSize;
const unsigned char* prefixEnd;
size_t prefixSize;
} LZ4_streamDecode_t_internal;
#endif
/*!
* LZ4_stream_t :
* information structure to track an LZ4 stream.
* init this structure before first use.
* note : only use in association with static linking !
* this definition is not API/ABI safe,
* it may change in a future version !
*/
#define LZ4_STREAMSIZE_U64 ((1 << (LZ4_MEMORY_USAGE-3)) + 4)
#define LZ4_STREAMSIZE (LZ4_STREAMSIZE_U64 * sizeof(unsigned long long))
union LZ4_stream_u {
unsigned long long table[LZ4_STREAMSIZE_U64];
LZ4_stream_t_internal internal_donotuse;
} ; /* previously typedef'd to LZ4_stream_t */
/*!
* LZ4_streamDecode_t :
* information structure to track an LZ4 stream during decompression.
* init this structure using LZ4_setStreamDecode (or memset()) before first use
* note : only use in association with static linking !
* this definition is not API/ABI safe,
* and may change in a future version !
*/
#define LZ4_STREAMDECODESIZE_U64 4
#define LZ4_STREAMDECODESIZE (LZ4_STREAMDECODESIZE_U64 * sizeof(unsigned long long))
union LZ4_streamDecode_u {
unsigned long long table[LZ4_STREAMDECODESIZE_U64];
LZ4_streamDecode_t_internal internal_donotuse;
} ; /* previously typedef'd to LZ4_streamDecode_t */
/*-************************************
* Obsolete Functions
**************************************/
/*! Deprecation warnings
Should deprecation warnings be a problem,
it is generally possible to disable them,
typically with -Wno-deprecated-declarations for gcc
or _CRT_SECURE_NO_WARNINGS in Visual.
Otherwise, it's also possible to define LZ4_DISABLE_DEPRECATE_WARNINGS */
#ifdef LZ4_DISABLE_DEPRECATE_WARNINGS
# define LZ4_DEPRECATED(message) /* disable deprecation warnings */
#else
# define LZ4_GCC_VERSION (__GNUC__ * 100 + __GNUC_MINOR__)
# if defined (__cplusplus) && (__cplusplus >= 201402) /* C++14 or greater */
# define LZ4_DEPRECATED(message) [[deprecated(message)]]
# elif (LZ4_GCC_VERSION >= 405) || defined(__clang__)
# define LZ4_DEPRECATED(message) __attribute__((deprecated(message)))
# elif (LZ4_GCC_VERSION >= 301)
# define LZ4_DEPRECATED(message) __attribute__((deprecated))
# elif defined(_MSC_VER)
# define LZ4_DEPRECATED(message) __declspec(deprecated(message))
# else
# pragma message("WARNING: You need to implement LZ4_DEPRECATED for this compiler")
# define LZ4_DEPRECATED(message)
# endif
#endif /* LZ4_DISABLE_DEPRECATE_WARNINGS */
/* Obsolete compression functions */
LZ4_DEPRECATED("use LZ4_compress_default() instead") LZ4LIB_API int LZ4_compress (const char* source, char* dest, int sourceSize);
LZ4_DEPRECATED("use LZ4_compress_default() instead") LZ4LIB_API int LZ4_compress_limitedOutput (const char* source, char* dest, int sourceSize, int maxOutputSize);
LZ4_DEPRECATED("use LZ4_compress_fast_extState() instead") LZ4LIB_API int LZ4_compress_withState (void* state, const char* source, char* dest, int inputSize);
LZ4_DEPRECATED("use LZ4_compress_fast_extState() instead") LZ4LIB_API int LZ4_compress_limitedOutput_withState (void* state, const char* source, char* dest, int inputSize, int maxOutputSize);
LZ4_DEPRECATED("use LZ4_compress_fast_continue() instead") LZ4LIB_API int LZ4_compress_continue (LZ4_stream_t* LZ4_streamPtr, const char* source, char* dest, int inputSize);
LZ4_DEPRECATED("use LZ4_compress_fast_continue() instead") LZ4LIB_API int LZ4_compress_limitedOutput_continue (LZ4_stream_t* LZ4_streamPtr, const char* source, char* dest, int inputSize, int maxOutputSize);
/* Obsolete decompression functions */
LZ4_DEPRECATED("use LZ4_decompress_fast() instead") LZ4LIB_API int LZ4_uncompress (const char* source, char* dest, int outputSize);
LZ4_DEPRECATED("use LZ4_decompress_safe() instead") LZ4LIB_API int LZ4_uncompress_unknownOutputSize (const char* source, char* dest, int isize, int maxOutputSize);
/* Obsolete streaming functions; degraded functionality; do not use!
*
* In order to perform streaming compression, these functions depended on data
* that is no longer tracked in the state. They have been preserved as well as
* possible: using them will still produce a correct output. However, they don't
* actually retain any history between compression calls. The compression ratio
* achieved will therefore be no better than compressing each chunk
* independently.
*/
LZ4_DEPRECATED("Use LZ4_createStream() instead") LZ4LIB_API void* LZ4_create (char* inputBuffer);
LZ4_DEPRECATED("Use LZ4_createStream() instead") LZ4LIB_API int LZ4_sizeofStreamState(void);
LZ4_DEPRECATED("Use LZ4_resetStream() instead") LZ4LIB_API int LZ4_resetStreamState(void* state, char* inputBuffer);
LZ4_DEPRECATED("Use LZ4_saveDict() instead") LZ4LIB_API char* LZ4_slideInputBuffer (void* state);
/* Obsolete streaming decoding functions */
/*LZ4_DEPRECATED("use LZ4_decompress_safe_usingDict() instead")*/ LZ4LIB_API int LZ4_decompress_safe_withPrefix64k (const char* src, char* dst, int compressedSize, int maxDstSize);
LZ4_DEPRECATED("use LZ4_decompress_fast_usingDict() instead") LZ4LIB_API int LZ4_decompress_fast_withPrefix64k (const char* src, char* dst, int originalSize);
}
#endif /* LZ4_H_2983827168210 */

File diff suppressed because it is too large Load Diff

View File

@@ -1,7 +1,7 @@
/*
LZ4 HC - High Compression Mode of LZ4
Header File
Copyright (C) 2011-2020, Yann Collet.
Copyright (C) 2011-2017, Yann Collet.
BSD 2-Clause License (http://www.opensource.org/licenses/bsd-license.php)
Redistribution and use in source and binary forms, with or without
@@ -31,13 +31,15 @@
- LZ4 source repository : https://github.com/lz4/lz4
- LZ4 public forum : https://groups.google.com/forum/#!forum/lz4c
*/
#ifndef TRACY_LZ4_HC_H_19834876238432
#define TRACY_LZ4_HC_H_19834876238432
#ifndef LZ4_HC_H_19834876238432
#define LZ4_HC_H_19834876238432
/* --- Dependency --- */
/* note : lz4hc requires lz4.h/lz4.c for compilation */
#include "tracy_lz4.hpp" /* stddef, LZ4LIB_API, LZ4_DEPRECATED */
namespace tracy
{
/* --- Useful constants --- */
#define LZ4HC_CLEVEL_MIN 3
@@ -45,14 +47,12 @@
#define LZ4HC_CLEVEL_OPT_MIN 10
#define LZ4HC_CLEVEL_MAX 12
namespace tracy
{
/*-************************************
* Block Compression
**************************************/
/*! LZ4_compress_HC() :
* Compress data from `src` into `dst`, using the powerful but slower "HC" algorithm.
* Compress data from `src` into `dst`, using the more powerful but slower "HC" algorithm.
* `dst` must be already allocated.
* Compression is guaranteed to succeed if `dstCapacity >= LZ4_compressBound(srcSize)` (see "lz4.h")
* Max supported `srcSize` value is LZ4_MAX_INPUT_SIZE (see "lz4.h")
@@ -75,21 +75,7 @@ LZ4LIB_API int LZ4_compress_HC (const char* src, char* dst, int srcSize, int dst
* Memory segment must be aligned on 8-bytes boundaries (which a normal malloc() should do properly).
*/
LZ4LIB_API int LZ4_sizeofStateHC(void);
LZ4LIB_API int LZ4_compress_HC_extStateHC(void* stateHC, const char* src, char* dst, int srcSize, int maxDstSize, int compressionLevel);
/*! LZ4_compress_HC_destSize() : v1.9.0+
* Will compress as much data as possible from `src`
* to fit into `targetDstSize` budget.
* Result is provided in 2 parts :
* @return : the number of bytes written into 'dst' (necessarily <= targetDstSize)
* or 0 if compression fails.
* `srcSizePtr` : on success, *srcSizePtr is updated to indicate how much bytes were read from `src`
*/
LZ4LIB_API int LZ4_compress_HC_destSize(void* stateHC,
const char* src, char* dst,
int* srcSizePtr, int targetDstSize,
int compressionLevel);
LZ4LIB_API int LZ4_compress_HC_extStateHC(void* state, const char* src, char* dst, int srcSize, int maxDstSize, int compressionLevel);
/*-************************************
@@ -101,92 +87,46 @@ LZ4LIB_API int LZ4_compress_HC_destSize(void* stateHC,
/*! LZ4_createStreamHC() and LZ4_freeStreamHC() :
* These functions create and release memory for LZ4 HC streaming state.
* Newly created states are automatically initialized.
* A same state can be used multiple times consecutively,
* starting with LZ4_resetStreamHC_fast() to start a new stream of blocks.
* Existing states can be re-used several times, using LZ4_resetStreamHC().
* These methods are API and ABI stable, they can be used in combination with a DLL.
*/
LZ4LIB_API LZ4_streamHC_t* LZ4_createStreamHC(void);
LZ4LIB_API int LZ4_freeStreamHC (LZ4_streamHC_t* streamHCPtr);
/*
These functions compress data in successive blocks of any size,
using previous blocks as dictionary, to improve compression ratio.
One key assumption is that previous blocks (up to 64 KB) remain read-accessible while compressing next blocks.
There is an exception for ring buffers, which can be smaller than 64 KB.
Ring-buffer scenario is automatically detected and handled within LZ4_compress_HC_continue().
Before starting compression, state must be allocated and properly initialized.
LZ4_createStreamHC() does both, though compression level is set to LZ4HC_CLEVEL_DEFAULT.
Selecting the compression level can be done with LZ4_resetStreamHC_fast() (starts a new stream)
or LZ4_setCompressionLevel() (anytime, between blocks in the same stream) (experimental).
LZ4_resetStreamHC_fast() only works on states which have been properly initialized at least once,
which is automatically the case when state is created using LZ4_createStreamHC().
After reset, a first "fictional block" can be designated as initial dictionary,
using LZ4_loadDictHC() (Optional).
Invoke LZ4_compress_HC_continue() to compress each successive block.
The number of blocks is unlimited.
Previous input blocks, including initial dictionary when present,
must remain accessible and unmodified during compression.
It's allowed to update compression level anytime between blocks,
using LZ4_setCompressionLevel() (experimental).
'dst' buffer should be sized to handle worst case scenarios
(see LZ4_compressBound(), it ensures compression success).
In case of failure, the API does not guarantee recovery,
so the state _must_ be reset.
To ensure compression success
whenever `dst` buffer size cannot be made >= LZ4_compressBound(),
consider using LZ4_compress_HC_continue_destSize().
Whenever previous input blocks can't be preserved unmodified in-place during compression of next blocks,
it's possible to copy the last blocks into a more stable memory space, using LZ4_saveDictHC().
Return value of LZ4_saveDictHC() is the size of dictionary effectively saved into 'safeBuffer' (<= 64 KB)
After completing a streaming compression,
it's possible to start a new stream of blocks, using the same LZ4_streamHC_t state,
just by resetting it, using LZ4_resetStreamHC_fast().
*/
LZ4LIB_API void LZ4_resetStreamHC_fast(LZ4_streamHC_t* streamHCPtr, int compressionLevel); /* v1.9.0+ */
LZ4LIB_API void LZ4_resetStreamHC (LZ4_streamHC_t* streamHCPtr, int compressionLevel);
LZ4LIB_API int LZ4_loadDictHC (LZ4_streamHC_t* streamHCPtr, const char* dictionary, int dictSize);
LZ4LIB_API int LZ4_compress_HC_continue (LZ4_streamHC_t* streamHCPtr,
const char* src, char* dst,
int srcSize, int maxDstSize);
/*! LZ4_compress_HC_continue_destSize() : v1.9.0+
* Similar to LZ4_compress_HC_continue(),
* but will read as much data as possible from `src`
* to fit into `targetDstSize` budget.
* Result is provided into 2 parts :
* @return : the number of bytes written into 'dst' (necessarily <= targetDstSize)
* or 0 if compression fails.
* `srcSizePtr` : on success, *srcSizePtr will be updated to indicate how much bytes were read from `src`.
* Note that this function may not consume the entire input.
*/
LZ4LIB_API int LZ4_compress_HC_continue_destSize(LZ4_streamHC_t* LZ4_streamHCPtr,
const char* src, char* dst,
int* srcSizePtr, int targetDstSize);
LZ4LIB_API int LZ4_compress_HC_continue (LZ4_streamHC_t* streamHCPtr, const char* src, char* dst, int srcSize, int maxDstSize);
LZ4LIB_API int LZ4_saveDictHC (LZ4_streamHC_t* streamHCPtr, char* safeBuffer, int maxDictSize);
/*
These functions compress data in successive blocks of any size, using previous blocks as dictionary.
One key assumption is that previous blocks (up to 64 KB) remain read-accessible while compressing next blocks.
There is an exception for ring buffers, which can be smaller than 64 KB.
Ring buffers scenario is automatically detected and handled by LZ4_compress_HC_continue().
Before starting compression, state must be properly initialized, using LZ4_resetStreamHC().
A first "fictional block" can then be designated as initial dictionary, using LZ4_loadDictHC() (Optional).
Then, use LZ4_compress_HC_continue() to compress each successive block.
Previous memory blocks (including initial dictionary when present) must remain accessible and unmodified during compression.
'dst' buffer should be sized to handle worst case scenarios (see LZ4_compressBound()), to ensure operation success.
Because in case of failure, the API does not guarantee context recovery, and context will have to be reset.
If `dst` buffer budget cannot be >= LZ4_compressBound(), consider using LZ4_compress_HC_continue_destSize() instead.
If, for any reason, previous data block can't be preserved unmodified in memory for next compression block,
you can save it to a more stable memory space, using LZ4_saveDictHC().
Return value of LZ4_saveDictHC() is the size of dictionary effectively saved into 'safeBuffer'.
*/
/*^**********************************************
* !!!!!! STATIC LINKING ONLY !!!!!!
***********************************************/
/*-******************************************************************
/*-**************************************************************
* PRIVATE DEFINITIONS :
* Do not use these definitions directly.
* They are merely exposed to allow static allocation of `LZ4_streamHC_t`.
* Declare an `LZ4_streamHC_t` directly, rather than any type below.
* Even then, only do so in the context of static linking, as definitions may change between versions.
********************************************************************/
* Do not use these definitions.
* They are exposed to allow static allocation of `LZ4_streamHC_t`.
* Using these definitions makes the code vulnerable to potential API break when upgrading LZ4
****************************************************************/
#define LZ4HC_DICTIONARY_LOGSIZE 16
#define LZ4HC_MAXD (1<<LZ4HC_DICTIONARY_LOGSIZE)
#define LZ4HC_MAXD_MASK (LZ4HC_MAXD - 1)
@@ -196,51 +136,59 @@ LZ4LIB_API int LZ4_saveDictHC (LZ4_streamHC_t* streamHCPtr, char* safeBuffer, in
#define LZ4HC_HASH_MASK (LZ4HC_HASHTABLESIZE - 1)
/* Never ever use these definitions directly !
* Declare or allocate an LZ4_streamHC_t instead.
**/
#if defined(__cplusplus) || (defined (__STDC_VERSION__) && (__STDC_VERSION__ >= 199901L) /* C99 */)
#include <stdint.h>
typedef struct LZ4HC_CCtx_internal LZ4HC_CCtx_internal;
struct LZ4HC_CCtx_internal
{
LZ4_u32 hashTable[LZ4HC_HASHTABLESIZE];
LZ4_u16 chainTable[LZ4HC_MAXD];
const LZ4_byte* end; /* next block here to continue on current prefix */
const LZ4_byte* prefixStart; /* Indexes relative to this position */
const LZ4_byte* dictStart; /* alternate reference for extDict */
LZ4_u32 dictLimit; /* below that point, need extDict */
LZ4_u32 lowLimit; /* below that point, no more dict */
LZ4_u32 nextToUpdate; /* index from which to continue dictionary update */
short compressionLevel;
LZ4_i8 favorDecSpeed; /* favor decompression speed if this flag set,
otherwise, favor compression ratio */
LZ4_i8 dirty; /* stream has to be fully reset if this flag is set */
uint32_t hashTable[LZ4HC_HASHTABLESIZE];
uint16_t chainTable[LZ4HC_MAXD];
const uint8_t* end; /* next block here to continue on current prefix */
const uint8_t* base; /* All index relative to this position */
const uint8_t* dictBase; /* alternate base for extDict */
uint32_t dictLimit; /* below that point, need extDict */
uint32_t lowLimit; /* below that point, no more dict */
uint32_t nextToUpdate; /* index from which to continue dictionary update */
short compressionLevel;
short favorDecSpeed;
const LZ4HC_CCtx_internal* dictCtx;
};
#define LZ4_STREAMHC_MINSIZE 262200 /* static size, for inter-version compatibility */
#else
typedef struct LZ4HC_CCtx_internal LZ4HC_CCtx_internal;
struct LZ4HC_CCtx_internal
{
unsigned int hashTable[LZ4HC_HASHTABLESIZE];
unsigned short chainTable[LZ4HC_MAXD];
const unsigned char* end; /* next block here to continue on current prefix */
const unsigned char* base; /* All index relative to this position */
const unsigned char* dictBase; /* alternate base for extDict */
unsigned int dictLimit; /* below that point, need extDict */
unsigned int lowLimit; /* below that point, no more dict */
unsigned int nextToUpdate; /* index from which to continue dictionary update */
short compressionLevel;
short favorDecSpeed;
const LZ4HC_CCtx_internal* dictCtx;
};
#endif
#define LZ4_STREAMHCSIZE (4*LZ4HC_HASHTABLESIZE + 2*LZ4HC_MAXD + 56) /* 262200 */
#define LZ4_STREAMHCSIZE_SIZET (LZ4_STREAMHCSIZE / sizeof(size_t))
union LZ4_streamHC_u {
char minStateSize[LZ4_STREAMHC_MINSIZE];
size_t table[LZ4_STREAMHCSIZE_SIZET];
LZ4HC_CCtx_internal internal_donotuse;
}; /* previously typedef'd to LZ4_streamHC_t */
}; /* previously typedef'd to LZ4_streamHC_t */
/*
LZ4_streamHC_t :
This structure allows static allocation of LZ4 HC streaming state.
State must be initialized using LZ4_resetStreamHC() before first use.
/* LZ4_streamHC_t :
* This structure allows static allocation of LZ4 HC streaming state.
* This can be used to allocate statically on stack, or as part of a larger structure.
*
* Such state **must** be initialized using LZ4_initStreamHC() before first use.
*
* Note that invoking LZ4_initStreamHC() is not required when
* the state was created using LZ4_createStreamHC() (which is recommended).
* Using the normal builder, a newly created state is automatically initialized.
*
* Static allocation shall only be used in combination with static linking.
*/
/* LZ4_initStreamHC() : v1.9.0+
* Required before first use of a statically allocated LZ4_streamHC_t.
* Before v1.9.0 : use LZ4_resetStreamHC() instead
*/
LZ4LIB_API LZ4_streamHC_t* LZ4_initStreamHC(void* buffer, size_t size);
Static allocation shall only be used in combination with static linking.
When invoking LZ4 from a DLL, use create/free functions instead, which are API and ABI stable.
*/
/*-************************************
@@ -251,11 +199,11 @@ LZ4LIB_API LZ4_streamHC_t* LZ4_initStreamHC(void* buffer, size_t size);
/* deprecated compression functions */
LZ4_DEPRECATED("use LZ4_compress_HC() instead") LZ4LIB_API int LZ4_compressHC (const char* source, char* dest, int inputSize);
LZ4_DEPRECATED("use LZ4_compress_HC() instead") LZ4LIB_API int LZ4_compressHC_limitedOutput (const char* source, char* dest, int inputSize, int maxOutputSize);
LZ4_DEPRECATED("use LZ4_compress_HC() instead") LZ4LIB_API int LZ4_compressHC2 (const char* source, char* dest, int inputSize, int compressionLevel);
LZ4_DEPRECATED("use LZ4_compress_HC() instead") LZ4LIB_API int LZ4_compressHC2_limitedOutput(const char* source, char* dest, int inputSize, int maxOutputSize, int compressionLevel);
LZ4_DEPRECATED("use LZ4_compress_HC() instead") LZ4LIB_API int LZ4_compressHC2 (const char* source, char* dest, int inputSize, int compressionLevel);
LZ4_DEPRECATED("use LZ4_compress_HC() instead") LZ4LIB_API int LZ4_compressHC2_limitedOutput (const char* source, char* dest, int inputSize, int maxOutputSize, int compressionLevel);
LZ4_DEPRECATED("use LZ4_compress_HC_extStateHC() instead") LZ4LIB_API int LZ4_compressHC_withStateHC (void* state, const char* source, char* dest, int inputSize);
LZ4_DEPRECATED("use LZ4_compress_HC_extStateHC() instead") LZ4LIB_API int LZ4_compressHC_limitedOutput_withStateHC (void* state, const char* source, char* dest, int inputSize, int maxOutputSize);
LZ4_DEPRECATED("use LZ4_compress_HC_extStateHC() instead") LZ4LIB_API int LZ4_compressHC2_withStateHC (void* state, const char* source, char* dest, int inputSize, int compressionLevel);
LZ4_DEPRECATED("use LZ4_compress_HC_extStateHC() instead") LZ4LIB_API int LZ4_compressHC2_withStateHC (void* state, const char* source, char* dest, int inputSize, int compressionLevel);
LZ4_DEPRECATED("use LZ4_compress_HC_extStateHC() instead") LZ4LIB_API int LZ4_compressHC2_limitedOutput_withStateHC(void* state, const char* source, char* dest, int inputSize, int maxOutputSize, int compressionLevel);
LZ4_DEPRECATED("use LZ4_compress_HC_continue() instead") LZ4LIB_API int LZ4_compressHC_continue (LZ4_streamHC_t* LZ4_streamHCPtr, const char* source, char* dest, int inputSize);
LZ4_DEPRECATED("use LZ4_compress_HC_continue() instead") LZ4LIB_API int LZ4_compressHC_limitedOutput_continue (LZ4_streamHC_t* LZ4_streamHCPtr, const char* source, char* dest, int inputSize, int maxOutputSize);
@@ -268,26 +216,13 @@ LZ4_DEPRECATED("use LZ4_compress_HC_continue() instead") LZ4LIB_API int LZ4_comp
* LZ4_slideInputBufferHC() will truncate the history of the stream, rather
* than preserve a window-sized chunk of history.
*/
#if !defined(LZ4_STATIC_LINKING_ONLY_DISABLE_MEMORY_ALLOCATION)
LZ4_DEPRECATED("use LZ4_createStreamHC() instead") LZ4LIB_API void* LZ4_createHC (const char* inputBuffer);
LZ4_DEPRECATED("use LZ4_freeStreamHC() instead") LZ4LIB_API int LZ4_freeHC (void* LZ4HC_Data);
#endif
LZ4_DEPRECATED("use LZ4_saveDictHC() instead") LZ4LIB_API char* LZ4_slideInputBufferHC (void* LZ4HC_Data);
LZ4_DEPRECATED("use LZ4_compress_HC_continue() instead") LZ4LIB_API int LZ4_compressHC2_continue (void* LZ4HC_Data, const char* source, char* dest, int inputSize, int compressionLevel);
LZ4_DEPRECATED("use LZ4_freeStreamHC() instead") LZ4LIB_API int LZ4_freeHC (void* LZ4HC_Data);
LZ4_DEPRECATED("use LZ4_compress_HC_continue() instead") LZ4LIB_API int LZ4_compressHC2_continue (void* LZ4HC_Data, const char* source, char* dest, int inputSize, int compressionLevel);
LZ4_DEPRECATED("use LZ4_compress_HC_continue() instead") LZ4LIB_API int LZ4_compressHC2_limitedOutput_continue (void* LZ4HC_Data, const char* source, char* dest, int inputSize, int maxOutputSize, int compressionLevel);
LZ4_DEPRECATED("use LZ4_createStreamHC() instead") LZ4LIB_API int LZ4_sizeofStreamStateHC(void);
LZ4_DEPRECATED("use LZ4_initStreamHC() instead") LZ4LIB_API int LZ4_resetStreamStateHC(void* state, char* inputBuffer);
/* LZ4_resetStreamHC() is now replaced by LZ4_initStreamHC().
* The intention is to emphasize the difference with LZ4_resetStreamHC_fast(),
* which is now the recommended function to start a new stream of blocks,
* but cannot be used to initialize a memory segment containing arbitrary garbage data.
*
* It is recommended to switch to LZ4_initStreamHC().
* LZ4_resetStreamHC() will generate deprecation warnings in a future version.
*/
LZ4LIB_API void LZ4_resetStreamHC (LZ4_streamHC_t* streamHCPtr, int compressionLevel);
LZ4_DEPRECATED("use LZ4_resetStreamHC() instead") LZ4LIB_API int LZ4_resetStreamStateHC(void* state, char* inputBuffer);
}
@@ -300,34 +235,53 @@ LZ4LIB_API void LZ4_resetStreamHC (LZ4_streamHC_t* streamHCPtr, int compressionL
* They should not be linked from DLL,
* as there is no guarantee of API stability yet.
* Prototypes will be promoted to "stable" status
* after successful usage in real-life scenarios.
* after successfull usage in real-life scenarios.
***************************************************/
#ifdef LZ4_HC_STATIC_LINKING_ONLY /* protection macro */
#ifndef TRACY_LZ4_HC_SLO_098092834
#define TRACY_LZ4_HC_SLO_098092834
#define LZ4_STATIC_LINKING_ONLY /* LZ4LIB_STATIC_API */
#include "tracy_lz4.hpp"
#ifndef LZ4_HC_SLO_098092834
#define LZ4_HC_SLO_098092834
namespace tracy
{
/*! LZ4_setCompressionLevel() : v1.8.0+ (experimental)
* It's possible to change compression level
* between successive invocations of LZ4_compress_HC_continue*()
* for dynamic adaptation.
/*! LZ4_compress_HC_destSize() : v1.8.0 (experimental)
* Will try to compress as much data from `src` as possible
* that can fit into `targetDstSize` budget.
* Result is provided in 2 parts :
* @return : the number of bytes written into 'dst'
* or 0 if compression fails.
* `srcSizePtr` : value will be updated to indicate how much bytes were read from `src`
*/
LZ4LIB_STATIC_API void LZ4_setCompressionLevel(
LZ4_streamHC_t* LZ4_streamHCPtr, int compressionLevel);
int LZ4_compress_HC_destSize(void* LZ4HC_Data,
const char* src, char* dst,
int* srcSizePtr, int targetDstSize,
int compressionLevel);
/*! LZ4_favorDecompressionSpeed() : v1.8.2+ (experimental)
* Opt. Parser will favor decompression speed over compression ratio.
* Only applicable to levels >= LZ4HC_CLEVEL_OPT_MIN.
/*! LZ4_compress_HC_continue_destSize() : v1.8.0 (experimental)
* Similar as LZ4_compress_HC_continue(),
* but will read a variable nb of bytes from `src`
* to fit into `targetDstSize` budget.
* Result is provided in 2 parts :
* @return : the number of bytes written into 'dst'
* or 0 if compression fails.
* `srcSizePtr` : value will be updated to indicate how much bytes were read from `src`.
*/
LZ4LIB_STATIC_API void LZ4_favorDecompressionSpeed(
LZ4_streamHC_t* LZ4_streamHCPtr, int favor);
int LZ4_compress_HC_continue_destSize(LZ4_streamHC_t* LZ4_streamHCPtr,
const char* src, char* dst,
int* srcSizePtr, int targetDstSize);
/*! LZ4_resetStreamHC_fast() : v1.9.0+
/*! LZ4_setCompressionLevel() : v1.8.0 (experimental)
* It's possible to change compression level between 2 invocations of LZ4_compress_HC_continue*()
*/
void LZ4_setCompressionLevel(LZ4_streamHC_t* LZ4_streamHCPtr, int compressionLevel);
/*! LZ4_favorDecompressionSpeed() : v1.8.2 (experimental)
* Parser will select decisions favoring decompression over compression ratio.
* Only work at highest compression settings (level >= LZ4HC_CLEVEL_OPT_MIN)
*/
void LZ4_favorDecompressionSpeed(LZ4_streamHC_t* LZ4_streamHCPtr, int favor);
/*! LZ4_resetStreamHC_fast() :
* When an LZ4_streamHC_t is known to be in a internally coherent state,
* it can often be prepared for a new compression with almost no work, only
* sometimes falling back to the full, expensive reset that is always required
@@ -344,14 +298,8 @@ LZ4LIB_STATIC_API void LZ4_favorDecompressionSpeed(
* - the stream was in an indeterminate state and was used in a compression
* call that fully reset the state (LZ4_compress_HC_extStateHC()) and that
* returned success
*
* Note:
* A stream that was last used in a compression call that returned an error
* may be passed to this function. However, it will be fully reset, which will
* clear any existing history and settings from the context.
*/
LZ4LIB_STATIC_API void LZ4_resetStreamHC_fast(
LZ4_streamHC_t* LZ4_streamHCPtr, int compressionLevel);
void LZ4_resetStreamHC_fast(LZ4_streamHC_t* LZ4_streamHCPtr, int compressionLevel);
/*! LZ4_compress_HC_extStateHC_fastReset() :
* A variant of LZ4_compress_HC_extStateHC().
@@ -364,11 +312,7 @@ LZ4LIB_STATIC_API void LZ4_resetStreamHC_fast(
* LZ4_resetStreamHC_fast() while LZ4_compress_HC_extStateHC() starts with a
* call to LZ4_resetStreamHC().
*/
LZ4LIB_STATIC_API int LZ4_compress_HC_extStateHC_fastReset (
void* state,
const char* src, char* dst,
int srcSize, int dstCapacity,
int compressionLevel);
int LZ4_compress_HC_extStateHC_fastReset (void* state, const char* src, char* dst, int srcSize, int dstCapacity, int compressionLevel);
/*! LZ4_attach_HC_dictionary() :
* This is an experimental API that allows for the efficient use of a
@@ -395,9 +339,7 @@ LZ4LIB_STATIC_API int LZ4_compress_HC_extStateHC_fastReset (
* stream (and source buffer) must remain in-place / accessible / unchanged
* through the lifetime of the stream session.
*/
LZ4LIB_STATIC_API void LZ4_attach_HC_dictionary(
LZ4_streamHC_t *working_stream,
const LZ4_streamHC_t *dictionary_stream);
LZ4LIB_API void LZ4_attach_HC_dictionary(LZ4_streamHC_t *working_stream, const LZ4_streamHC_t *dictionary_stream);
}

255
common/tracy_sema.h Normal file
View File

@@ -0,0 +1,255 @@
// Copyright (c) 2015 Jeff Preshing
//
// This software is provided 'as-is', without any express or implied
// warranty. In no event will the authors be held liable for any damages
// arising from the use of this software.
//
// Permission is granted to anyone to use this software for any purpose,
// including commercial applications, and to alter it and redistribute it
// freely, subject to the following restrictions:
//
// 1. The origin of this software must not be misrepresented; you must not
// claim that you wrote the original software. If you use this software
// in a product, an acknowledgement in the product documentation would be
// appreciated but is not required.
// 2. Altered source versions must be plainly marked as such, and must not be
// misrepresented as being the original software.
// 3. This notice may not be removed or altered from any source distribution.
#ifndef __TRACY_CPP11OM_SEMAPHORE_H__
#define __TRACY_CPP11OM_SEMAPHORE_H__
#include <atomic>
#include <cassert>
#if defined(__MACH__)
#include <mach/mach.h>
#elif defined(__unix__)
#include <semaphore.h>
#endif
namespace tracy
{
#if defined(_WIN32)
//---------------------------------------------------------
// Semaphore (Windows)
//---------------------------------------------------------
#ifndef MAXLONG
enum { MAXLONG = 0x7fffffff };
#endif
#ifndef INFINITE
enum { INFINITE = 0xFFFFFFFF };
#endif
#ifndef _WINDOWS_
typedef void* HANDLE;
extern "C" __declspec(dllimport) HANDLE __stdcall CreateSemaphoreA( void*, long, long, const char* );
extern "C" __declspec(dllimport) int __stdcall CloseHandle( HANDLE );
extern "C" __declspec(dllimport) unsigned long __stdcall WaitForSingleObject( HANDLE, unsigned long );
extern "C" __declspec(dllimport) int __stdcall ReleaseSemaphore( HANDLE, long, long* );
#endif
class Semaphore
{
private:
HANDLE m_hSema;
Semaphore(const Semaphore& other) = delete;
Semaphore& operator=(const Semaphore& other) = delete;
public:
Semaphore(int initialCount = 0)
{
assert(initialCount >= 0);
m_hSema = CreateSemaphoreA(NULL, initialCount, MAXLONG, NULL);
}
~Semaphore()
{
CloseHandle(m_hSema);
}
void wait()
{
WaitForSingleObject(m_hSema, INFINITE);
}
void signal(int count = 1)
{
ReleaseSemaphore(m_hSema, count, NULL);
}
};
#elif defined(__MACH__)
//---------------------------------------------------------
// Semaphore (Apple iOS and OSX)
// Can't use POSIX semaphores due to http://lists.apple.com/archives/darwin-kernel/2009/Apr/msg00010.html
//---------------------------------------------------------
class Semaphore
{
private:
semaphore_t m_sema;
Semaphore(const Semaphore& other) = delete;
Semaphore& operator=(const Semaphore& other) = delete;
public:
Semaphore(int initialCount = 0)
{
assert(initialCount >= 0);
semaphore_create(mach_task_self(), &m_sema, SYNC_POLICY_FIFO, initialCount);
}
~Semaphore()
{
semaphore_destroy(mach_task_self(), m_sema);
}
void wait()
{
semaphore_wait(m_sema);
}
void signal()
{
semaphore_signal(m_sema);
}
void signal(int count)
{
while (count-- > 0)
{
semaphore_signal(m_sema);
}
}
};
#elif defined(__unix__)
//---------------------------------------------------------
// Semaphore (POSIX, Linux)
//---------------------------------------------------------
class Semaphore
{
private:
sem_t m_sema;
Semaphore(const Semaphore& other) = delete;
Semaphore& operator=(const Semaphore& other) = delete;
public:
Semaphore(int initialCount = 0)
{
assert(initialCount >= 0);
sem_init(&m_sema, 0, initialCount);
}
~Semaphore()
{
sem_destroy(&m_sema);
}
void wait()
{
// http://stackoverflow.com/questions/2013181/gdb-causes-sem-wait-to-fail-with-eintr-error
int rc;
do
{
rc = sem_wait(&m_sema);
}
while (rc == -1 && errno == EINTR);
}
void signal()
{
sem_post(&m_sema);
}
void signal(int count)
{
while (count-- > 0)
{
sem_post(&m_sema);
}
}
};
#else
#error Unsupported platform!
#endif
//---------------------------------------------------------
// LightweightSemaphore
//---------------------------------------------------------
class LightweightSemaphore
{
private:
std::atomic<int> m_count;
Semaphore m_sema;
void waitWithPartialSpinning()
{
int oldCount;
// Is there a better way to set the initial spin count?
// If we lower it to 1000, testBenaphore becomes 15x slower on my Core i7-5930K Windows PC,
// as threads start hitting the kernel semaphore.
int spin = 10000;
while (spin--)
{
oldCount = m_count.load(std::memory_order_relaxed);
if ((oldCount > 0) && m_count.compare_exchange_strong(oldCount, oldCount - 1, std::memory_order_acquire))
return;
std::atomic_signal_fence(std::memory_order_acquire); // Prevent the compiler from collapsing the loop.
}
oldCount = m_count.fetch_sub(1, std::memory_order_acquire);
if (oldCount <= 0)
{
m_sema.wait();
}
}
public:
LightweightSemaphore(int initialCount = 0) : m_count(initialCount)
{
assert(initialCount >= 0);
}
bool tryWait()
{
int oldCount = m_count.load(std::memory_order_relaxed);
return (oldCount > 0 && m_count.compare_exchange_strong(oldCount, oldCount - 1, std::memory_order_acquire));
}
void wait()
{
if (!tryWait())
waitWithPartialSpinning();
}
void signal(int count = 1)
{
int oldCount = m_count.fetch_add(count, std::memory_order_release);
int toRelease = -oldCount < count ? -oldCount : count;
if (toRelease > 0)
{
m_sema.signal(toRelease);
}
}
};
typedef LightweightSemaphore DefaultSemaphoreType;
}
#endif // __CPP11OM_SEMAPHORE_H__

View File

@@ -1,15 +0,0 @@
ARCH := $(shell uname -m)
ifeq (1,$(shell $(CC) --version | grep clang > /dev/null && echo 1 || echo 0))
ifeq (1,$(shell ld.mold --version > /dev/null 2> /dev/null && echo 1 || echo 0))
LDFLAGS := -fuse-ld=mold
endif
endif
ifndef TRACY_NO_ISA_EXTENSIONS
ifneq (,$(filter $(ARCH),aarch64 arm64))
CFLAGS += -mcpu=native
else
CFLAGS += -march=native
endif
endif

View File

@@ -1,19 +0,0 @@
ARCH := $(shell uname -m)
ifeq (0,$(shell $(CC) --version | grep clang > /dev/null && echo 1 || echo 0))
CFLAGS += -s
else
ifeq (1,$(shell ld.mold --version > /dev/null 2> /dev/null && echo 1 || echo 0))
LDFLAGS := -s -fuse-ld=mold
else
LDFLAGS := -s
endif
endif
ifndef TRACY_NO_ISA_EXTENSIONS
ifneq (,$(filter $(ARCH),aarch64 arm64))
CFLAGS += -mcpu=native
else
CFLAGS += -march=native
endif
endif

View File

@@ -1,84 +0,0 @@
# Common code needed by most Tracy Unix Makefiles.
# Ensure these are simply-substituted variables, without changing their values.
LIBS := $(LIBS)
ifndef NO_TBB
# Tracy does not use TBB directly, but the implementation of parallel algorithms
# in some versions of libstdc++ depends on TBB. When it does, you must
# explicitly link against -ltbb.
#
# Some distributions have pgk-config files for TBB, others don't.
ifeq (0,$(shell pkg-config --libs tbb >/dev/null 2>&1; echo $$?))
LIBS += $(shell pkg-config --libs tbb)
else ifeq (0,$(shell ld -ltbb -o /dev/null 2>/dev/null; echo $$?))
LIBS += -ltbb
endif
endif
OBJDIRBASE := obj/$(BUILD)
OBJDIR := $(OBJDIRBASE)/o/o/o
OBJ := $(addprefix $(OBJDIR)/,$(SRC:%.cpp=%.o))
OBJ2 := $(addprefix $(OBJDIR)/,$(SRC2:%.c=%.o))
OBJ3 := $(addprefix $(OBJDIR)/,$(SRC3:%.m=%.o))
OBJ4 := $(addprefix $(OBJDIR)/,$(SRC4:%.S=%.o))
all: $(IMAGE)
$(OBJDIR)/%.o: %.cpp
$(CXX) -c $(INCLUDES) $(CXXFLAGS) $(DEFINES) $< -o $@
$(OBJDIR)/%.d : %.cpp
@echo Resolving dependencies of $<
@mkdir -p $(@D)
@$(CXX) -MM $(INCLUDES) $(CXXFLAGS) $(DEFINES) $< > $@.$$$$; \
sed 's,.*\.o[ :]*,$(OBJDIR)/$(<:.cpp=.o) $@ : ,g' < $@.$$$$ > $@; \
rm -f $@.$$$$
$(OBJDIR)/%.o: %.c
$(CC) -c $(INCLUDES) $(CFLAGS) $(DEFINES) $< -o $@
$(OBJDIR)/%.d : %.c
@echo Resolving dependencies of $<
@mkdir -p $(@D)
@$(CC) -MM $(INCLUDES) $(CFLAGS) $(DEFINES) $< > $@.$$$$; \
sed 's,.*\.o[ :]*,$(OBJDIR)/$(<:.c=.o) $@ : ,g' < $@.$$$$ > $@; \
rm -f $@.$$$$
$(OBJDIR)/%.o: %.m
$(CC) -c $(INCLUDES) $(CFLAGS) $(DEFINES) $< -o $@
$(OBJDIR)/%.d : %.m
@echo Resolving dependencies of $<
@mkdir -p $(@D)
@$(CC) -MM $(INCLUDES) $(CFLAGS) $(DEFINES) $< > $@.$$$$; \
sed 's,.*\.o[ :]*,$(OBJDIR)/$(<:.m=.o) $@ : ,g' < $@.$$$$ > $@; \
rm -f $@.$$$$
$(OBJDIR)/%.o: %.S
$(CC) -c $(INCLUDES) $(CFLAGS) $(DEFINES) $< -o $@
$(OBJDIR)/%.d : %.S
@echo Resolving dependencies of $<
@mkdir -p $(@D)
@$(CC) -MM $(INCLUDES) $(CFLAGS) $(DEFINES) $< > $@.$$$$; \
sed 's,.*\.o[ :]*,$(OBJDIR)/$(<:.m=.o) $@ : ,g' < $@.$$$$ > $@; \
rm -f $@.$$$$
ifeq (yes,$(SHARED_LIBRARY))
$(IMAGE): $(OBJ) $(OBJ2) $(OBJ4)
$(CXX) $(CXXFLAGS) $(LDFLAGS) $(DEFINES) $(OBJ) $(OBJ2) $(OBJ4) $(LIBS) -shared -o $@
else
$(IMAGE): $(OBJ) $(OBJ2) $(OBJ3) $(OBJ4)
$(CXX) $(CXXFLAGS) $(LDFLAGS) $(DEFINES) $(OBJ) $(OBJ2) $(OBJ3) $(OBJ4) $(LIBS) -o $@
endif
ifneq "$(MAKECMDGOALS)" "clean"
-include $(addprefix $(OBJDIR)/,$(SRC:.cpp=.d)) $(addprefix $(OBJDIR)/,$(SRC2:.c=.d)) $(addprefix $(OBJDIR)/,$(SRC3:.m=.d)) $(addprefix $(OBJDIR)/,$(SRC4:.S=.d))
endif
clean:
rm -rf $(OBJDIRBASE) $(IMAGE)*
.PHONY: clean all

View File

@@ -1,16 +0,0 @@
all: release
debug:
@$(MAKE) -f debug.mk all
release:
@$(MAKE) -f release.mk all
clean:
@$(MAKE) -f build.mk clean
db: clean
@bear -- $(MAKE) -f debug.mk all
@mv -f compile_commands.json ../../../
.PHONY: all clean debug release db

View File

@@ -1,12 +0,0 @@
CFLAGS +=
CXXFLAGS := $(CFLAGS) -std=gnu++17
# DEFINES += -DTRACY_NO_STATISTICS
INCLUDES := $(shell pkg-config --cflags capstone)
LIBS := $(shell pkg-config --libs capstone) -lpthread
PROJECT := csvexport
IMAGE := $(PROJECT)-$(BUILD)
FILTER :=
include ../../../common/src-from-vcxproj.mk
include ../../../common/unix.mk

View File

@@ -1,6 +0,0 @@
CFLAGS := -g3 -Wall
DEFINES := -DDEBUG
BUILD := debug
include ../../../common/unix-debug.mk
include build.mk

View File

@@ -1,9 +0,0 @@
CFLAGS := -O3
ifndef TRACY_NO_LTO
CFLAGS += -flto
endif
DEFINES := -DNDEBUG
BUILD := release
include ../../../common/unix-release.mk
include build.mk

View File

@@ -1,25 +0,0 @@
Microsoft Visual Studio Solution File, Format Version 12.00
# Visual Studio Version 16
VisualStudioVersion = 16.0.30907.101
MinimumVisualStudioVersion = 10.0.40219.1
Project("{8BC9CEB8-8B4A-11D0-8D11-00A0C91BC942}") = "csvexport", "csvexport.vcxproj", "{447D58BF-94CD-4469-BB90-549C05D03E00}"
EndProject
Global
GlobalSection(SolutionConfigurationPlatforms) = preSolution
Debug|x64 = Debug|x64
Release|x64 = Release|x64
EndGlobalSection
GlobalSection(ProjectConfigurationPlatforms) = postSolution
{447D58BF-94CD-4469-BB90-549C05D03E00}.Debug|x64.ActiveCfg = Debug|x64
{447D58BF-94CD-4469-BB90-549C05D03E00}.Debug|x64.Build.0 = Debug|x64
{447D58BF-94CD-4469-BB90-549C05D03E00}.Release|x64.ActiveCfg = Release|x64
{447D58BF-94CD-4469-BB90-549C05D03E00}.Release|x64.Build.0 = Release|x64
EndGlobalSection
GlobalSection(SolutionProperties) = preSolution
HideSolutionNode = FALSE
EndGlobalSection
GlobalSection(ExtensibilityGlobals) = postSolution
SolutionGuid = {3E51386C-43EA-44AC-9F24-AFAFE4D63ADE}
EndGlobalSection
EndGlobal

View File

@@ -1,210 +0,0 @@
<?xml version="1.0" encoding="utf-8"?>
<Project DefaultTargets="Build" ToolsVersion="15.0" xmlns="http://schemas.microsoft.com/developer/msbuild/2003">
<ItemGroup Label="ProjectConfigurations">
<ProjectConfiguration Include="Debug|x64">
<Configuration>Debug</Configuration>
<Platform>x64</Platform>
</ProjectConfiguration>
<ProjectConfiguration Include="Release|x64">
<Configuration>Release</Configuration>
<Platform>x64</Platform>
</ProjectConfiguration>
</ItemGroup>
<PropertyGroup Label="Globals">
<VCProjectVersion>15.0</VCProjectVersion>
<ProjectGuid>{447D58BF-94CD-4469-BB90-549C05D03E00}</ProjectGuid>
<RootNamespace>capture</RootNamespace>
<WindowsTargetPlatformVersion>10.0</WindowsTargetPlatformVersion>
<VcpkgTriplet>x64-windows-static</VcpkgTriplet>
</PropertyGroup>
<Import Project="$(VCTargetsPath)\Microsoft.Cpp.Default.props" />
<PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Debug|x64'" Label="Configuration">
<ConfigurationType>Application</ConfigurationType>
<UseDebugLibraries>true</UseDebugLibraries>
<PlatformToolset>v143</PlatformToolset>
<CharacterSet>MultiByte</CharacterSet>
</PropertyGroup>
<PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Release|x64'" Label="Configuration">
<ConfigurationType>Application</ConfigurationType>
<UseDebugLibraries>false</UseDebugLibraries>
<PlatformToolset>v143</PlatformToolset>
<WholeProgramOptimization>true</WholeProgramOptimization>
<CharacterSet>MultiByte</CharacterSet>
</PropertyGroup>
<Import Project="$(VCTargetsPath)\Microsoft.Cpp.props" />
<ImportGroup Label="ExtensionSettings">
</ImportGroup>
<ImportGroup Label="Shared">
</ImportGroup>
<ImportGroup Label="PropertySheets" Condition="'$(Configuration)|$(Platform)'=='Debug|x64'">
<Import Project="$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props" Condition="exists('$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props')" Label="LocalAppDataPlatform" />
</ImportGroup>
<ImportGroup Label="PropertySheets" Condition="'$(Configuration)|$(Platform)'=='Release|x64'">
<Import Project="$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props" Condition="exists('$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props')" Label="LocalAppDataPlatform" />
</ImportGroup>
<PropertyGroup Label="UserMacros" />
<PropertyGroup />
<PropertyGroup Label="Vcpkg">
<VcpkgEnableManifest>true</VcpkgEnableManifest>
</PropertyGroup>
<ItemDefinitionGroup Condition="'$(Configuration)|$(Platform)'=='Debug|x64'">
<ClCompile>
<WarningLevel>Level3</WarningLevel>
<Optimization>Disabled</Optimization>
<SDLCheck>true</SDLCheck>
<ConformanceMode>true</ConformanceMode>
<MultiProcessorCompilation>true</MultiProcessorCompilation>
<PreprocessorDefinitions>_CRT_SECURE_NO_DEPRECATE;_CRT_NONSTDC_NO_DEPRECATE;WIN32_LEAN_AND_MEAN;NOMINMAX;_USE_MATH_DEFINES;%(PreprocessorDefinitions)</PreprocessorDefinitions>
<EnableEnhancedInstructionSet>AdvancedVectorExtensions2</EnableEnhancedInstructionSet>
<LanguageStandard>stdcpplatest</LanguageStandard>
<AdditionalIncludeDirectories>$(ProjectDir)..\..\..\vcpkg_installed\$(VcpkgTriplet)\include;$(ProjectDir)..\..\..\vcpkg_installed\$(VcpkgTriplet)\include\capstone;$(VcpkgManifestRoot)\vcpkg_installed\$(VcpkgTriplet)\$(VcpkgTriplet)\include\capstone;$(VcpkgRoot)\installed\$(VcpkgTriplet)\include\capstone</AdditionalIncludeDirectories>
</ClCompile>
<Link>
<AdditionalDependencies>ws2_32.lib;capstone.lib;%(AdditionalDependencies)</AdditionalDependencies>
<SubSystem>Console</SubSystem>
<AdditionalLibraryDirectories>$(ProjectDir)..\..\..\vcpkg_installed\$(VcpkgTriplet)\debug\lib</AdditionalLibraryDirectories>
</Link>
</ItemDefinitionGroup>
<ItemDefinitionGroup Condition="'$(Configuration)|$(Platform)'=='Release|x64'">
<ClCompile>
<WarningLevel>Level3</WarningLevel>
<Optimization>MaxSpeed</Optimization>
<FunctionLevelLinking>true</FunctionLevelLinking>
<IntrinsicFunctions>true</IntrinsicFunctions>
<SDLCheck>true</SDLCheck>
<ConformanceMode>true</ConformanceMode>
<MultiProcessorCompilation>true</MultiProcessorCompilation>
<PreprocessorDefinitions>NDEBUG;_CRT_SECURE_NO_DEPRECATE;_CRT_NONSTDC_NO_DEPRECATE;WIN32_LEAN_AND_MEAN;NOMINMAX;_USE_MATH_DEFINES;%(PreprocessorDefinitions)</PreprocessorDefinitions>
<EnableEnhancedInstructionSet>AdvancedVectorExtensions2</EnableEnhancedInstructionSet>
<LanguageStandard>stdcpplatest</LanguageStandard>
<AdditionalIncludeDirectories>$(ProjectDir)..\..\..\vcpkg_installed\$(VcpkgTriplet)\include;$(ProjectDir)..\..\..\vcpkg_installed\$(VcpkgTriplet)\include\capstone;$(VcpkgManifestRoot)\vcpkg_installed\$(VcpkgTriplet)\$(VcpkgTriplet)\include\capstone;$(VcpkgRoot)\installed\$(VcpkgTriplet)\include\capstone</AdditionalIncludeDirectories>
</ClCompile>
<Link>
<EnableCOMDATFolding>true</EnableCOMDATFolding>
<OptimizeReferences>true</OptimizeReferences>
<AdditionalDependencies>ws2_32.lib;capstone.lib;%(AdditionalDependencies)</AdditionalDependencies>
<SubSystem>Console</SubSystem>
<AdditionalLibraryDirectories>$(ProjectDir)..\..\..\vcpkg_installed\$(VcpkgTriplet)\lib</AdditionalLibraryDirectories>
</Link>
</ItemDefinitionGroup>
<ItemGroup>
<ClCompile Include="..\..\..\getopt\getopt.c" />
<ClCompile Include="..\..\..\public\common\TracySocket.cpp" />
<ClCompile Include="..\..\..\public\common\TracyStackFrames.cpp" />
<ClCompile Include="..\..\..\public\common\TracySystem.cpp" />
<ClCompile Include="..\..\..\public\common\tracy_lz4.cpp" />
<ClCompile Include="..\..\..\public\common\tracy_lz4hc.cpp" />
<ClCompile Include="..\..\..\server\TracyMemory.cpp" />
<ClCompile Include="..\..\..\server\TracyMmap.cpp" />
<ClCompile Include="..\..\..\server\TracyPrint.cpp" />
<ClCompile Include="..\..\..\server\TracyTaskDispatch.cpp" />
<ClCompile Include="..\..\..\server\TracyTextureCompression.cpp" />
<ClCompile Include="..\..\..\server\TracyThreadCompress.cpp" />
<ClCompile Include="..\..\..\server\TracyWorker.cpp" />
<ClCompile Include="..\..\..\zstd\common\debug.c" />
<ClCompile Include="..\..\..\zstd\common\entropy_common.c" />
<ClCompile Include="..\..\..\zstd\common\error_private.c" />
<ClCompile Include="..\..\..\zstd\common\fse_decompress.c" />
<ClCompile Include="..\..\..\zstd\common\pool.c" />
<ClCompile Include="..\..\..\zstd\common\threading.c" />
<ClCompile Include="..\..\..\zstd\common\xxhash.c" />
<ClCompile Include="..\..\..\zstd\common\zstd_common.c" />
<ClCompile Include="..\..\..\zstd\compress\fse_compress.c" />
<ClCompile Include="..\..\..\zstd\compress\hist.c" />
<ClCompile Include="..\..\..\zstd\compress\huf_compress.c" />
<ClCompile Include="..\..\..\zstd\compress\zstdmt_compress.c" />
<ClCompile Include="..\..\..\zstd\compress\zstd_compress.c" />
<ClCompile Include="..\..\..\zstd\compress\zstd_compress_literals.c" />
<ClCompile Include="..\..\..\zstd\compress\zstd_compress_sequences.c" />
<ClCompile Include="..\..\..\zstd\compress\zstd_compress_superblock.c" />
<ClCompile Include="..\..\..\zstd\compress\zstd_double_fast.c" />
<ClCompile Include="..\..\..\zstd\compress\zstd_fast.c" />
<ClCompile Include="..\..\..\zstd\compress\zstd_lazy.c" />
<ClCompile Include="..\..\..\zstd\compress\zstd_ldm.c" />
<ClCompile Include="..\..\..\zstd\compress\zstd_opt.c" />
<ClCompile Include="..\..\..\zstd\decompress\huf_decompress.c" />
<ClCompile Include="..\..\..\zstd\decompress\zstd_ddict.c" />
<ClCompile Include="..\..\..\zstd\decompress\zstd_decompress.c" />
<ClCompile Include="..\..\..\zstd\decompress\zstd_decompress_block.c" />
<ClCompile Include="..\..\..\zstd\dictBuilder\cover.c" />
<ClCompile Include="..\..\..\zstd\dictBuilder\divsufsort.c" />
<ClCompile Include="..\..\..\zstd\dictBuilder\fastcover.c" />
<ClCompile Include="..\..\..\zstd\dictBuilder\zdict.c" />
<ClCompile Include="..\..\src\csvexport.cpp" />
</ItemGroup>
<ItemGroup>
<ClInclude Include="..\..\..\getopt\getopt.h" />
<ClInclude Include="..\..\..\public\common\TracyAlign.hpp" />
<ClInclude Include="..\..\..\public\common\TracyAlloc.hpp" />
<ClInclude Include="..\..\..\public\common\TracyApi.h" />
<ClInclude Include="..\..\..\public\common\TracyColor.hpp" />
<ClInclude Include="..\..\..\public\common\TracyForceInline.hpp" />
<ClInclude Include="..\..\..\public\common\TracyMutex.hpp" />
<ClInclude Include="..\..\..\public\common\TracyProtocol.hpp" />
<ClInclude Include="..\..\..\public\common\TracyQueue.hpp" />
<ClInclude Include="..\..\..\public\common\TracySocket.hpp" />
<ClInclude Include="..\..\..\public\common\TracyStackFrames.hpp" />
<ClInclude Include="..\..\..\public\common\TracySystem.hpp" />
<ClInclude Include="..\..\..\public\common\TracyUwp.hpp" />
<ClInclude Include="..\..\..\public\common\TracyYield.hpp" />
<ClInclude Include="..\..\..\public\common\tracy_lz4.hpp" />
<ClInclude Include="..\..\..\public\common\tracy_lz4hc.hpp" />
<ClInclude Include="..\..\..\server\TracyCharUtil.hpp" />
<ClInclude Include="..\..\..\server\TracyEvent.hpp" />
<ClInclude Include="..\..\..\server\TracyFileRead.hpp" />
<ClInclude Include="..\..\..\server\TracyFileWrite.hpp" />
<ClInclude Include="..\..\..\server\TracyMemory.hpp" />
<ClInclude Include="..\..\..\server\TracyMmap.hpp" />
<ClInclude Include="..\..\..\server\TracyPopcnt.hpp" />
<ClInclude Include="..\..\..\server\TracyPrint.hpp" />
<ClInclude Include="..\..\..\server\TracySlab.hpp" />
<ClInclude Include="..\..\..\server\TracyTaskDispatch.hpp" />
<ClInclude Include="..\..\..\server\TracyTextureCompression.hpp" />
<ClInclude Include="..\..\..\server\TracyThreadCompress.hpp" />
<ClInclude Include="..\..\..\server\TracyVector.hpp" />
<ClInclude Include="..\..\..\server\TracyWorker.hpp" />
<ClInclude Include="..\..\..\zstd\common\bitstream.h" />
<ClInclude Include="..\..\..\zstd\common\compiler.h" />
<ClInclude Include="..\..\..\zstd\common\cpu.h" />
<ClInclude Include="..\..\..\zstd\common\debug.h" />
<ClInclude Include="..\..\..\zstd\common\error_private.h" />
<ClInclude Include="..\..\..\zstd\common\fse.h" />
<ClInclude Include="..\..\..\zstd\common\huf.h" />
<ClInclude Include="..\..\..\zstd\common\mem.h" />
<ClInclude Include="..\..\..\zstd\common\pool.h" />
<ClInclude Include="..\..\..\zstd\common\portability_macros.h" />
<ClInclude Include="..\..\..\zstd\common\threading.h" />
<ClInclude Include="..\..\..\zstd\common\xxhash.h" />
<ClInclude Include="..\..\..\zstd\common\zstd_deps.h" />
<ClInclude Include="..\..\..\zstd\common\zstd_internal.h" />
<ClInclude Include="..\..\..\zstd\common\zstd_trace.h" />
<ClInclude Include="..\..\..\zstd\compress\clevels.h" />
<ClInclude Include="..\..\..\zstd\compress\hist.h" />
<ClInclude Include="..\..\..\zstd\compress\zstdmt_compress.h" />
<ClInclude Include="..\..\..\zstd\compress\zstd_compress_internal.h" />
<ClInclude Include="..\..\..\zstd\compress\zstd_compress_literals.h" />
<ClInclude Include="..\..\..\zstd\compress\zstd_compress_sequences.h" />
<ClInclude Include="..\..\..\zstd\compress\zstd_compress_superblock.h" />
<ClInclude Include="..\..\..\zstd\compress\zstd_cwksp.h" />
<ClInclude Include="..\..\..\zstd\compress\zstd_double_fast.h" />
<ClInclude Include="..\..\..\zstd\compress\zstd_fast.h" />
<ClInclude Include="..\..\..\zstd\compress\zstd_lazy.h" />
<ClInclude Include="..\..\..\zstd\compress\zstd_ldm.h" />
<ClInclude Include="..\..\..\zstd\compress\zstd_ldm_geartab.h" />
<ClInclude Include="..\..\..\zstd\compress\zstd_opt.h" />
<ClInclude Include="..\..\..\zstd\decompress\zstd_ddict.h" />
<ClInclude Include="..\..\..\zstd\decompress\zstd_decompress_block.h" />
<ClInclude Include="..\..\..\zstd\decompress\zstd_decompress_internal.h" />
<ClInclude Include="..\..\..\zstd\dictBuilder\cover.h" />
<ClInclude Include="..\..\..\zstd\dictBuilder\divsufsort.h" />
<ClInclude Include="..\..\..\zstd\zdict.h" />
<ClInclude Include="..\..\..\zstd\zstd.h" />
<ClInclude Include="..\..\..\zstd\zstd_errors.h" />
</ItemGroup>
<ItemGroup>
<None Include="..\..\..\zstd\decompress\huf_decompress_amd64.S" />
</ItemGroup>
<Import Project="$(VCTargetsPath)\Microsoft.Cpp.targets" />
<ImportGroup Label="ExtensionTargets">
</ImportGroup>
</Project>

View File

@@ -1,371 +0,0 @@
<?xml version="1.0" encoding="utf-8"?>
<Project ToolsVersion="4.0" xmlns="http://schemas.microsoft.com/developer/msbuild/2003">
<ItemGroup>
<Filter Include="src">
<UniqueIdentifier>{729c80ee-4d26-4a5e-8f1f-6c075783eb56}</UniqueIdentifier>
</Filter>
<Filter Include="server">
<UniqueIdentifier>{cf23ef7b-7694-4154-830b-00cf053350ea}</UniqueIdentifier>
</Filter>
<Filter Include="common">
<UniqueIdentifier>{e39d3623-47cd-4752-8da9-3ea324f964c1}</UniqueIdentifier>
</Filter>
<Filter Include="getopt">
<UniqueIdentifier>{ee9737d2-69c7-44da-b9c7-539d18f9d4b4}</UniqueIdentifier>
</Filter>
<Filter Include="zstd">
<UniqueIdentifier>{44ea1742-0fd9-40ba-879c-031868509600}</UniqueIdentifier>
</Filter>
<Filter Include="zstd\common">
<UniqueIdentifier>{2d065bba-d78e-4e44-87f7-b44cf3248260}</UniqueIdentifier>
</Filter>
<Filter Include="zstd\compress">
<UniqueIdentifier>{a19ca3bc-2a17-49c5-a0d9-5916213de774}</UniqueIdentifier>
</Filter>
<Filter Include="zstd\decompress">
<UniqueIdentifier>{d4181058-2198-4931-ae31-b7eda0312458}</UniqueIdentifier>
</Filter>
<Filter Include="zstd\dictBuilder">
<UniqueIdentifier>{873c22fe-b4d7-480d-ad67-48271296f4c1}</UniqueIdentifier>
</Filter>
</ItemGroup>
<ItemGroup>
<ClCompile Include="..\..\..\server\TracyMemory.cpp">
<Filter>server</Filter>
</ClCompile>
<ClCompile Include="..\..\..\server\TracyWorker.cpp">
<Filter>server</Filter>
</ClCompile>
<ClCompile Include="..\..\..\server\TracyPrint.cpp">
<Filter>server</Filter>
</ClCompile>
<ClCompile Include="..\..\..\server\TracyThreadCompress.cpp">
<Filter>server</Filter>
</ClCompile>
<ClCompile Include="..\..\..\server\TracyTaskDispatch.cpp">
<Filter>server</Filter>
</ClCompile>
<ClCompile Include="..\..\..\server\TracyMmap.cpp">
<Filter>server</Filter>
</ClCompile>
<ClCompile Include="..\..\..\server\TracyTextureCompression.cpp">
<Filter>server</Filter>
</ClCompile>
<ClCompile Include="..\..\..\getopt\getopt.c">
<Filter>getopt</Filter>
</ClCompile>
<ClCompile Include="..\..\src\csvexport.cpp">
<Filter>src</Filter>
</ClCompile>
<ClCompile Include="..\..\..\zstd\common\debug.c">
<Filter>zstd\common</Filter>
</ClCompile>
<ClCompile Include="..\..\..\zstd\common\entropy_common.c">
<Filter>zstd\common</Filter>
</ClCompile>
<ClCompile Include="..\..\..\zstd\common\error_private.c">
<Filter>zstd\common</Filter>
</ClCompile>
<ClCompile Include="..\..\..\zstd\common\fse_decompress.c">
<Filter>zstd\common</Filter>
</ClCompile>
<ClCompile Include="..\..\..\zstd\common\pool.c">
<Filter>zstd\common</Filter>
</ClCompile>
<ClCompile Include="..\..\..\zstd\common\threading.c">
<Filter>zstd\common</Filter>
</ClCompile>
<ClCompile Include="..\..\..\zstd\common\xxhash.c">
<Filter>zstd\common</Filter>
</ClCompile>
<ClCompile Include="..\..\..\zstd\common\zstd_common.c">
<Filter>zstd\common</Filter>
</ClCompile>
<ClCompile Include="..\..\..\zstd\decompress\huf_decompress.c">
<Filter>zstd\decompress</Filter>
</ClCompile>
<ClCompile Include="..\..\..\zstd\decompress\zstd_ddict.c">
<Filter>zstd\decompress</Filter>
</ClCompile>
<ClCompile Include="..\..\..\zstd\decompress\zstd_decompress.c">
<Filter>zstd\decompress</Filter>
</ClCompile>
<ClCompile Include="..\..\..\zstd\decompress\zstd_decompress_block.c">
<Filter>zstd\decompress</Filter>
</ClCompile>
<ClCompile Include="..\..\..\zstd\compress\fse_compress.c">
<Filter>zstd\compress</Filter>
</ClCompile>
<ClCompile Include="..\..\..\zstd\compress\hist.c">
<Filter>zstd\compress</Filter>
</ClCompile>
<ClCompile Include="..\..\..\zstd\compress\huf_compress.c">
<Filter>zstd\compress</Filter>
</ClCompile>
<ClCompile Include="..\..\..\zstd\compress\zstd_compress.c">
<Filter>zstd\compress</Filter>
</ClCompile>
<ClCompile Include="..\..\..\zstd\compress\zstd_compress_literals.c">
<Filter>zstd\compress</Filter>
</ClCompile>
<ClCompile Include="..\..\..\zstd\compress\zstd_compress_sequences.c">
<Filter>zstd\compress</Filter>
</ClCompile>
<ClCompile Include="..\..\..\zstd\compress\zstd_compress_superblock.c">
<Filter>zstd\compress</Filter>
</ClCompile>
<ClCompile Include="..\..\..\zstd\compress\zstd_double_fast.c">
<Filter>zstd\compress</Filter>
</ClCompile>
<ClCompile Include="..\..\..\zstd\compress\zstd_fast.c">
<Filter>zstd\compress</Filter>
</ClCompile>
<ClCompile Include="..\..\..\zstd\compress\zstd_lazy.c">
<Filter>zstd\compress</Filter>
</ClCompile>
<ClCompile Include="..\..\..\zstd\compress\zstd_ldm.c">
<Filter>zstd\compress</Filter>
</ClCompile>
<ClCompile Include="..\..\..\zstd\compress\zstd_opt.c">
<Filter>zstd\compress</Filter>
</ClCompile>
<ClCompile Include="..\..\..\zstd\compress\zstdmt_compress.c">
<Filter>zstd\compress</Filter>
</ClCompile>
<ClCompile Include="..\..\..\zstd\dictBuilder\cover.c">
<Filter>zstd\dictBuilder</Filter>
</ClCompile>
<ClCompile Include="..\..\..\zstd\dictBuilder\divsufsort.c">
<Filter>zstd\dictBuilder</Filter>
</ClCompile>
<ClCompile Include="..\..\..\zstd\dictBuilder\fastcover.c">
<Filter>zstd\dictBuilder</Filter>
</ClCompile>
<ClCompile Include="..\..\..\zstd\dictBuilder\zdict.c">
<Filter>zstd\dictBuilder</Filter>
</ClCompile>
<ClCompile Include="..\..\..\public\common\tracy_lz4.cpp">
<Filter>common</Filter>
</ClCompile>
<ClCompile Include="..\..\..\public\common\tracy_lz4hc.cpp">
<Filter>common</Filter>
</ClCompile>
<ClCompile Include="..\..\..\public\common\TracySocket.cpp">
<Filter>common</Filter>
</ClCompile>
<ClCompile Include="..\..\..\public\common\TracyStackFrames.cpp">
<Filter>common</Filter>
</ClCompile>
<ClCompile Include="..\..\..\public\common\TracySystem.cpp">
<Filter>common</Filter>
</ClCompile>
</ItemGroup>
<ItemGroup>
<ClInclude Include="..\..\..\server\TracyCharUtil.hpp">
<Filter>server</Filter>
</ClInclude>
<ClInclude Include="..\..\..\server\TracyEvent.hpp">
<Filter>server</Filter>
</ClInclude>
<ClInclude Include="..\..\..\server\TracyFileWrite.hpp">
<Filter>server</Filter>
</ClInclude>
<ClInclude Include="..\..\..\server\TracyMemory.hpp">
<Filter>server</Filter>
</ClInclude>
<ClInclude Include="..\..\..\server\TracyPopcnt.hpp">
<Filter>server</Filter>
</ClInclude>
<ClInclude Include="..\..\..\server\TracySlab.hpp">
<Filter>server</Filter>
</ClInclude>
<ClInclude Include="..\..\..\server\TracyVector.hpp">
<Filter>server</Filter>
</ClInclude>
<ClInclude Include="..\..\..\server\TracyWorker.hpp">
<Filter>server</Filter>
</ClInclude>
<ClInclude Include="..\..\..\server\TracyPrint.hpp">
<Filter>server</Filter>
</ClInclude>
<ClInclude Include="..\..\..\server\TracyThreadCompress.hpp">
<Filter>server</Filter>
</ClInclude>
<ClInclude Include="..\..\..\server\TracyTaskDispatch.hpp">
<Filter>server</Filter>
</ClInclude>
<ClInclude Include="..\..\..\server\TracyFileRead.hpp">
<Filter>server</Filter>
</ClInclude>
<ClInclude Include="..\..\..\server\TracyMmap.hpp">
<Filter>server</Filter>
</ClInclude>
<ClInclude Include="..\..\..\server\TracyTextureCompression.hpp">
<Filter>server</Filter>
</ClInclude>
<ClInclude Include="..\..\..\getopt\getopt.h">
<Filter>getopt</Filter>
</ClInclude>
<ClInclude Include="..\..\..\zstd\zstd.h">
<Filter>zstd</Filter>
</ClInclude>
<ClInclude Include="..\..\..\zstd\zstd_errors.h">
<Filter>zstd</Filter>
</ClInclude>
<ClInclude Include="..\..\..\zstd\common\bitstream.h">
<Filter>zstd\common</Filter>
</ClInclude>
<ClInclude Include="..\..\..\zstd\common\compiler.h">
<Filter>zstd\common</Filter>
</ClInclude>
<ClInclude Include="..\..\..\zstd\common\cpu.h">
<Filter>zstd\common</Filter>
</ClInclude>
<ClInclude Include="..\..\..\zstd\common\debug.h">
<Filter>zstd\common</Filter>
</ClInclude>
<ClInclude Include="..\..\..\zstd\common\error_private.h">
<Filter>zstd\common</Filter>
</ClInclude>
<ClInclude Include="..\..\..\zstd\common\fse.h">
<Filter>zstd\common</Filter>
</ClInclude>
<ClInclude Include="..\..\..\zstd\common\huf.h">
<Filter>zstd\common</Filter>
</ClInclude>
<ClInclude Include="..\..\..\zstd\common\mem.h">
<Filter>zstd\common</Filter>
</ClInclude>
<ClInclude Include="..\..\..\zstd\common\pool.h">
<Filter>zstd\common</Filter>
</ClInclude>
<ClInclude Include="..\..\..\zstd\common\threading.h">
<Filter>zstd\common</Filter>
</ClInclude>
<ClInclude Include="..\..\..\zstd\common\xxhash.h">
<Filter>zstd\common</Filter>
</ClInclude>
<ClInclude Include="..\..\..\zstd\common\zstd_deps.h">
<Filter>zstd\common</Filter>
</ClInclude>
<ClInclude Include="..\..\..\zstd\common\zstd_internal.h">
<Filter>zstd\common</Filter>
</ClInclude>
<ClInclude Include="..\..\..\zstd\common\zstd_trace.h">
<Filter>zstd\common</Filter>
</ClInclude>
<ClInclude Include="..\..\..\zstd\decompress\zstd_ddict.h">
<Filter>zstd\decompress</Filter>
</ClInclude>
<ClInclude Include="..\..\..\zstd\decompress\zstd_decompress_block.h">
<Filter>zstd\decompress</Filter>
</ClInclude>
<ClInclude Include="..\..\..\zstd\decompress\zstd_decompress_internal.h">
<Filter>zstd\decompress</Filter>
</ClInclude>
<ClInclude Include="..\..\..\zstd\compress\hist.h">
<Filter>zstd\compress</Filter>
</ClInclude>
<ClInclude Include="..\..\..\zstd\compress\zstd_compress_internal.h">
<Filter>zstd\compress</Filter>
</ClInclude>
<ClInclude Include="..\..\..\zstd\compress\zstd_compress_literals.h">
<Filter>zstd\compress</Filter>
</ClInclude>
<ClInclude Include="..\..\..\zstd\compress\zstd_compress_sequences.h">
<Filter>zstd\compress</Filter>
</ClInclude>
<ClInclude Include="..\..\..\zstd\compress\zstd_compress_superblock.h">
<Filter>zstd\compress</Filter>
</ClInclude>
<ClInclude Include="..\..\..\zstd\compress\zstd_cwksp.h">
<Filter>zstd\compress</Filter>
</ClInclude>
<ClInclude Include="..\..\..\zstd\compress\zstd_double_fast.h">
<Filter>zstd\compress</Filter>
</ClInclude>
<ClInclude Include="..\..\..\zstd\compress\zstd_fast.h">
<Filter>zstd\compress</Filter>
</ClInclude>
<ClInclude Include="..\..\..\zstd\compress\zstd_lazy.h">
<Filter>zstd\compress</Filter>
</ClInclude>
<ClInclude Include="..\..\..\zstd\compress\zstd_ldm.h">
<Filter>zstd\compress</Filter>
</ClInclude>
<ClInclude Include="..\..\..\zstd\compress\zstd_ldm_geartab.h">
<Filter>zstd\compress</Filter>
</ClInclude>
<ClInclude Include="..\..\..\zstd\compress\zstd_opt.h">
<Filter>zstd\compress</Filter>
</ClInclude>
<ClInclude Include="..\..\..\zstd\compress\zstdmt_compress.h">
<Filter>zstd\compress</Filter>
</ClInclude>
<ClInclude Include="..\..\..\zstd\zdict.h">
<Filter>zstd</Filter>
</ClInclude>
<ClInclude Include="..\..\..\zstd\dictBuilder\cover.h">
<Filter>zstd\dictBuilder</Filter>
</ClInclude>
<ClInclude Include="..\..\..\zstd\dictBuilder\divsufsort.h">
<Filter>zstd\dictBuilder</Filter>
</ClInclude>
<ClInclude Include="..\..\..\zstd\common\portability_macros.h">
<Filter>zstd\common</Filter>
</ClInclude>
<ClInclude Include="..\..\..\zstd\compress\clevels.h">
<Filter>zstd\compress</Filter>
</ClInclude>
<ClInclude Include="..\..\..\public\common\tracy_lz4.hpp">
<Filter>common</Filter>
</ClInclude>
<ClInclude Include="..\..\..\public\common\tracy_lz4hc.hpp">
<Filter>common</Filter>
</ClInclude>
<ClInclude Include="..\..\..\public\common\TracyAlign.hpp">
<Filter>common</Filter>
</ClInclude>
<ClInclude Include="..\..\..\public\common\TracyAlloc.hpp">
<Filter>common</Filter>
</ClInclude>
<ClInclude Include="..\..\..\public\common\TracyApi.h">
<Filter>common</Filter>
</ClInclude>
<ClInclude Include="..\..\..\public\common\TracyColor.hpp">
<Filter>common</Filter>
</ClInclude>
<ClInclude Include="..\..\..\public\common\TracyForceInline.hpp">
<Filter>common</Filter>
</ClInclude>
<ClInclude Include="..\..\..\public\common\TracyMutex.hpp">
<Filter>common</Filter>
</ClInclude>
<ClInclude Include="..\..\..\public\common\TracyProtocol.hpp">
<Filter>common</Filter>
</ClInclude>
<ClInclude Include="..\..\..\public\common\TracyQueue.hpp">
<Filter>common</Filter>
</ClInclude>
<ClInclude Include="..\..\..\public\common\TracySocket.hpp">
<Filter>common</Filter>
</ClInclude>
<ClInclude Include="..\..\..\public\common\TracyStackFrames.hpp">
<Filter>common</Filter>
</ClInclude>
<ClInclude Include="..\..\..\public\common\TracySystem.hpp">
<Filter>common</Filter>
</ClInclude>
<ClInclude Include="..\..\..\public\common\TracyUwp.hpp">
<Filter>common</Filter>
</ClInclude>
<ClInclude Include="..\..\..\public\common\TracyYield.hpp">
<Filter>common</Filter>
</ClInclude>
</ItemGroup>
<ItemGroup>
<None Include="..\..\..\zstd\decompress\huf_decompress_amd64.S">
<Filter>zstd\decompress</Filter>
</None>
</ItemGroup>
</Project>

View File

@@ -1,353 +0,0 @@
#ifdef _WIN32
# include <windows.h>
#endif
#include <algorithm>
#include <cctype>
#include <fstream>
#include <iostream>
#include <sstream>
#include <string>
#include <math.h>
#include <stdio.h>
#include <stdint.h>
#include "../../server/TracyFileRead.hpp"
#include "../../server/TracyWorker.hpp"
#include "../../getopt/getopt.h"
void print_usage_exit(int e)
{
fprintf(stderr, "Extract statistics from a trace to a CSV format\n");
fprintf(stderr, "Usage:\n");
fprintf(stderr, " extract [OPTION...] <trace file>\n");
fprintf(stderr, "\n");
fprintf(stderr, " -h, --help Print usage\n");
fprintf(stderr, " -f, --filter arg Filter zone names (default: "")\n");
fprintf(stderr, " -s, --sep arg CSV separator (default: ,)\n");
fprintf(stderr, " -c, --case Case sensitive filtering\n");
fprintf(stderr, " -e, --self Get self times\n");
fprintf(stderr, " -u, --unwrap Report each zone event\n");
fprintf(stderr, " -m, --messages Report only messages\n");
exit(e);
}
struct Args {
const char* filter;
const char* separator;
const char* trace_file;
bool case_sensitive;
bool self_time;
bool unwrap;
bool unwrapMessages;
};
Args parse_args(int argc, char** argv)
{
if (argc == 1)
{
print_usage_exit(1);
}
Args args = { "", ",", "", false, false, false, false };
struct option long_opts[] = {
{ "help", no_argument, NULL, 'h' },
{ "filter", optional_argument, NULL, 'f' },
{ "sep", optional_argument, NULL, 's' },
{ "case", no_argument, NULL, 'c' },
{ "self", no_argument, NULL, 'e' },
{ "unwrap", no_argument, NULL, 'u' },
{ "messages", no_argument, NULL, 'm' },
{ NULL, 0, NULL, 0 }
};
int c;
while ((c = getopt_long(argc, argv, "hf:s:ceum", long_opts, NULL)) != -1)
{
switch (c)
{
case 'h':
print_usage_exit(0);
break;
case 'f':
args.filter = optarg;
break;
case 's':
args.separator = optarg;
break;
case 'c':
args.case_sensitive = true;
break;
case 'e':
args.self_time = true;
break;
case 'u':
args.unwrap = true;
break;
case 'm':
args.unwrapMessages = true;
break;
default:
print_usage_exit(1);
break;
}
}
if (argc != optind + 1)
{
print_usage_exit(1);
}
args.trace_file = argv[optind];
return args;
}
bool is_substring(
const char* term,
const char* s,
bool case_sensitive = false
){
auto new_term = std::string(term);
auto new_s = std::string(s);
if (!case_sensitive) {
std::transform(
new_term.begin(),
new_term.end(),
new_term.begin(),
[](unsigned char c){ return std::tolower(c); }
);
std::transform(
new_s.begin(),
new_s.end(),
new_s.begin(),
[](unsigned char c){ return std::tolower(c); }
);
}
return new_s.find(new_term) != std::string::npos;
}
const char* get_name(int32_t id, const tracy::Worker& worker)
{
auto& srcloc = worker.GetSourceLocation(id);
return worker.GetString(srcloc.name.active ? srcloc.name : srcloc.function);
}
template <typename T>
std::string join(const T& v, const char* sep) {
std::ostringstream s;
for (const auto& i : v) {
if (&i != &v[0]) {
s << sep;
}
s << i;
}
return s.str();
}
// From TracyView.cpp
int64_t GetZoneChildTimeFast(
const tracy::Worker& worker,
const tracy::ZoneEvent& zone
){
int64_t time = 0;
if( zone.HasChildren() )
{
auto& children = worker.GetZoneChildren( zone.Child() );
if( children.is_magic() )
{
auto& vec = *(tracy::Vector<tracy::ZoneEvent>*)&children;
for( auto& v : vec )
{
assert( v.IsEndValid() );
time += v.End() - v.Start();
}
}
else
{
for( auto& v : children )
{
assert( v->IsEndValid() );
time += v->End() - v->Start();
}
}
}
return time;
}
int main(int argc, char** argv)
{
#ifdef _WIN32
if (!AttachConsole(ATTACH_PARENT_PROCESS))
{
AllocConsole();
SetConsoleMode(GetStdHandle(STD_OUTPUT_HANDLE), 0x07);
}
#endif
Args args = parse_args(argc, argv);
auto f = std::unique_ptr<tracy::FileRead>(
tracy::FileRead::Open(args.trace_file)
);
if (!f)
{
fprintf(stderr, "Could not open file %s\n", args.trace_file);
return 1;
}
auto worker = tracy::Worker(*f);
if (args.unwrapMessages)
{
const auto& msgs = worker.GetMessages();
if (msgs.size() > 0)
{
std::vector<const char*> columnsForMessages;
columnsForMessages = {
"MessageName", "total_ns"
};
std::string headerForMessages = join(columnsForMessages, args.separator);
printf("%s\n", headerForMessages.data());
for(auto& it : msgs)
{
std::vector<std::string> values(columnsForMessages.size());
values[0] = worker.GetString(it->ref);
values[1] = std::to_string(it->time);
std::string row = join(values, args.separator);
printf("%s\n", row.data());
}
}
else
{
printf("There are currently no messages!\n");
}
return 0;
}
while (!worker.AreSourceLocationZonesReady())
{
std::this_thread::sleep_for(std::chrono::milliseconds(10));
}
auto& slz = worker.GetSourceLocationZones();
tracy::Vector<decltype(slz.begin())> slz_selected;
slz_selected.reserve(slz.size());
uint32_t total_cnt = 0;
for(auto it = slz.begin(); it != slz.end(); ++it)
{
if(it->second.total != 0)
{
++total_cnt;
if(args.filter[0] == '\0')
{
slz_selected.push_back_no_space_check(it);
}
else
{
auto name = get_name(it->first, worker);
if(is_substring(args.filter, name, args.case_sensitive))
{
slz_selected.push_back_no_space_check(it);
}
}
}
}
std::vector<const char*> columns;
if (args.unwrap)
{
columns = {
"name", "src_file", "src_line", "ns_since_start", "exec_time_ns", "thread"
};
}
else
{
columns = {
"name", "src_file", "src_line", "total_ns", "total_perc",
"counts", "mean_ns", "min_ns", "max_ns", "std_ns"
};
}
std::string header = join(columns, args.separator);
printf("%s\n", header.data());
const auto last_time = worker.GetLastTime();
for(auto& it : slz_selected)
{
std::vector<std::string> values(columns.size());
values[0] = get_name(it->first, worker);
const auto& srcloc = worker.GetSourceLocation(it->first);
values[1] = worker.GetString(srcloc.file);
values[2] = std::to_string(srcloc.line);
const auto& zone_data = it->second;
if (args.unwrap)
{
int i = 0;
for (const auto& zone_thread_data : zone_data.zones) {
const auto zone_event = zone_thread_data.Zone();
const auto tId = zone_thread_data.Thread();
const auto start = zone_event->Start();
const auto end = zone_event->End();
values[3] = std::to_string(start);
auto timespan = end - start;
if (args.self_time) {
timespan -= GetZoneChildTimeFast(worker, *zone_event);
}
values[4] = std::to_string(timespan);
values[5] = std::to_string(tId);
std::string row = join(values, args.separator);
printf("%s\n", row.data());
}
}
else
{
const auto time = args.self_time ? zone_data.selfTotal : zone_data.total;
values[3] = std::to_string(time);
values[4] = std::to_string(100. * time / last_time);
values[5] = std::to_string(zone_data.zones.size());
const auto avg = (args.self_time ? zone_data.selfTotal : zone_data.total)
/ zone_data.zones.size();
values[6] = std::to_string(avg);
const auto tmin = args.self_time ? zone_data.selfMin : zone_data.min;
const auto tmax = args.self_time ? zone_data.selfMax : zone_data.max;
values[7] = std::to_string(tmin);
values[8] = std::to_string(tmax);
const auto sz = zone_data.zones.size();
const auto ss = zone_data.sumSq
- 2. * zone_data.total * avg
+ avg * avg * sz;
double std = 0;
if( sz > 1 )
std = sqrt(ss / (sz - 1));
values[9] = std::to_string(std);
std::string row = join(values, args.separator);
printf("%s\n", row.data());
}
}
return 0;
}

BIN
doc/cost.png Normal file

Binary file not shown.

After

Width:  |  Height:  |  Size: 2.4 KiB

Binary file not shown.

Before

Width:  |  Height:  |  Size: 17 KiB

BIN
doc/locks.png Normal file

Binary file not shown.

After

Width:  |  Height:  |  Size: 8.3 KiB

BIN
doc/messages.png Normal file

Binary file not shown.

After

Width:  |  Height:  |  Size: 4.6 KiB

BIN
doc/plot.png Normal file

Binary file not shown.

After

Width:  |  Height:  |  Size: 9.3 KiB

Binary file not shown.

Before

Width:  |  Height:  |  Size: 213 KiB

After

Width:  |  Height:  |  Size: 28 KiB

Binary file not shown.

Before

Width:  |  Height:  |  Size: 250 KiB

Binary file not shown.

Before

Width:  |  Height:  |  Size: 234 KiB

View File

@@ -1,706 +0,0 @@
/**
dtl -- Diff Template Library
In short, Diff Template Library is distributed under so called "BSD license",
Copyright (c) 2015 Tatsuhiko Kubo <cubicdaiya@gmail.com>
All rights reserved.
Redistribution and use in source and binary forms, with or without modification,
are permitted provided that the following conditions are met:
* Redistributions of source code must retain the above copyright notice,
this list of conditions and the following disclaimer.
* Redistributions in binary form must reproduce the above copyright notice,
this list of conditions and the following disclaimer in the documentation
and/or other materials provided with the distribution.
* Neither the name of the authors nor the names of its contributors
may be used to endorse or promote products derived from this software
without specific prior written permission.
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED
TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF
LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING
NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
*/
/* If you use this library, you must include dtl.hpp only. */
#ifndef DTL_DIFF_H
#define DTL_DIFF_H
namespace dtl {
/**
* diff class template
* sequence must support random_access_iterator.
*/
template <typename elem, typename sequence = vector< elem >, typename comparator = Compare< elem > >
class Diff
{
private :
dtl_typedefs(elem, sequence)
sequence A;
sequence B;
size_t M;
size_t N;
size_t delta;
size_t offset;
long long *fp;
long long editDistance;
Lcs< elem > lcs;
Ses< elem > ses;
editPath path;
editPathCordinates pathCordinates;
bool swapped;
bool huge;
bool trivial;
bool editDistanceOnly;
uniHunkVec uniHunks;
comparator cmp;
long long ox;
long long oy;
public :
Diff () {}
Diff (const sequence& a,
const sequence& b) : A(a), B(b), ses(false) {
init();
}
Diff (const sequence& a,
const sequence& b,
bool deletesFirst) : A(a), B(b), ses(deletesFirst) {
init();
}
Diff (const sequence& a,
const sequence& b,
const comparator& comp) : A(a), B(b), ses(false), cmp(comp) {
init();
}
Diff (const sequence& a,
const sequence& b,
bool deleteFirst,
const comparator& comp) : A(a), B(b), ses(deleteFirst), cmp(comp) {
init();
}
~Diff() {}
long long getEditDistance () const {
return editDistance;
}
Lcs< elem > getLcs () const {
return lcs;
}
elemVec getLcsVec () const {
return lcs.getSequence();
}
Ses< elem > getSes () const {
return ses;
}
uniHunkVec getUniHunks () const {
return uniHunks;
}
/* These should be deprecated */
bool isHuge () const {
return huge;
}
void onHuge () {
this->huge = true;
}
void offHuge () {
this->huge = false;
}
bool isUnserious () const {
return trivial;
}
void onUnserious () {
this->trivial = true;
}
void offUnserious () {
this->trivial = false;
}
void onOnlyEditDistance () {
this->editDistanceOnly = true;
}
/* These are the replacements for the above */
bool hugeEnabled () const {
return huge;
}
void enableHuge () {
this->huge = true;
}
void disableHuge () {
this->huge = false;
}
bool trivialEnabled () const {
return trivial;
}
void enableTrivial () const {
this->trivial = true;
}
void disableTrivial () {
this->trivial = false;
}
void editDistanceOnlyEnabled () {
this->editDistanceOnly = true;
}
/**
* patching with Unified Format Hunks
*/
sequence uniPatch (const sequence& seq) {
elemList seqLst(seq.begin(), seq.end());
sesElemVec shunk;
sesElemVec_iter vsesIt;
elemList_iter lstIt = seqLst.begin();
long long inc_dec_total = 0;
long long gap = 1;
for (uniHunkVec_iter it=uniHunks.begin();it!=uniHunks.end();++it) {
joinSesVec(shunk, it->common[0]);
joinSesVec(shunk, it->change);
joinSesVec(shunk, it->common[1]);
it->a += inc_dec_total;
inc_dec_total += it->inc_dec_count;
for (long long i=0;i<it->a - gap;++i) {
++lstIt;
}
gap = it->a + it->b + it->inc_dec_count;
vsesIt = shunk.begin();
while (vsesIt!=shunk.end()) {
switch (vsesIt->second.type) {
case SES_ADD :
seqLst.insert(lstIt, vsesIt->first);
break;
case SES_DELETE :
if (lstIt != seqLst.end()) {
lstIt = seqLst.erase(lstIt);
}
break;
case SES_COMMON :
if (lstIt != seqLst.end()) {
++lstIt;
}
break;
default :
// no fall-through
break;
}
++vsesIt;
}
shunk.clear();
}
sequence patchedSeq(seqLst.begin(), seqLst.end());
return patchedSeq;
}
/**
* patching with Shortest Edit Script (SES)
*/
sequence patch (const sequence& seq) const {
sesElemVec sesSeq = ses.getSequence();
elemList seqLst(seq.begin(), seq.end());
elemList_iter lstIt = seqLst.begin();
for (sesElemVec_iter sesIt=sesSeq.begin();sesIt!=sesSeq.end();++sesIt) {
switch (sesIt->second.type) {
case SES_ADD :
seqLst.insert(lstIt, sesIt->first);
break;
case SES_DELETE :
lstIt = seqLst.erase(lstIt);
break;
case SES_COMMON :
++lstIt;
break;
default :
// no through
break;
}
}
sequence patchedSeq(seqLst.begin(), seqLst.end());
return patchedSeq;
}
/**
* compose Longest Common Subsequence and Shortest Edit Script.
* The algorithm implemented here is based on "An O(NP) Sequence Comparison Algorithm"
* described by Sun Wu, Udi Manber and Gene Myers
*/
void compose() {
if (isHuge()) {
pathCordinates.reserve(MAX_CORDINATES_SIZE);
}
ox = 0;
oy = 0;
long long p = -1;
fp = new long long[M + N + 3];
fill(&fp[0], &fp[M + N + 3], -1);
path = editPath(M + N + 3);
fill(path.begin(), path.end(), -1);
ONP:
do {
++p;
for (long long k=-p;k<=static_cast<long long>(delta)-1;++k) {
fp[k+offset] = snake(k, fp[k-1+offset]+1, fp[k+1+offset]);
}
for (long long k=static_cast<long long>(delta)+p;k>=static_cast<long long>(delta)+1;--k) {
fp[k+offset] = snake(k, fp[k-1+offset]+1, fp[k+1+offset]);
}
fp[delta+offset] = snake(static_cast<long long>(delta), fp[delta-1+offset]+1, fp[delta+1+offset]);
} while (fp[delta+offset] != static_cast<long long>(N) && pathCordinates.size() < MAX_CORDINATES_SIZE);
editDistance += static_cast<long long>(delta) + 2 * p;
long long r = path[delta+offset];
P cordinate;
editPathCordinates epc(0);
// recording edit distance only
if (editDistanceOnly) {
delete[] this->fp;
return;
}
while(r != -1) {
cordinate.x = pathCordinates[(size_t)r].x;
cordinate.y = pathCordinates[(size_t)r].y;
epc.push_back(cordinate);
r = pathCordinates[(size_t)r].k;
}
// record Longest Common Subsequence & Shortest Edit Script
if (!recordSequence(epc)) {
pathCordinates.resize(0);
epc.resize(0);
p = -1;
goto ONP;
}
delete[] this->fp;
}
/**
* print difference between A and B as an SES
*/
template < typename stream >
void printSES (stream& out) const {
sesElemVec ses_v = ses.getSequence();
for_each(ses_v.begin(), ses_v.end(), ChangePrinter< sesElem, stream >(out));
}
void printSES (ostream& out = cout) const {
printSES< ostream >(out);
}
/**
* print differences given an SES
*/
template < typename stream >
static void printSES (const Ses< elem >& s, stream& out) {
sesElemVec ses_v = s.getSequence();
for_each(ses_v.begin(), ses_v.end(), ChangePrinter< sesElem, stream >(out));
}
static void printSES (const Ses< elem >& s, ostream& out = cout) {
printSES< ostream >(s, out);
}
/**
* print difference between A and B as an SES with custom printer
*/
template < typename stream, template < typename SEET, typename STRT > class PT >
void printSES (stream& out) const {
sesElemVec ses_v = ses.getSequence ();
for_each (ses_v.begin (), ses_v.end(), PT < sesElem, stream > (out));
}
/**
* store difference between A and B as an SES with custom storage
*/
template < typename storedData, template < typename SEET, typename STRT > class ST >
void storeSES(storedData& sd) const {
sesElemVec ses_v = ses.getSequence();
for_each(ses_v.begin(), ses_v.end(), ST < sesElem, storedData >(sd));
}
/**
* print difference between A and B in the Unified Format
*/
template < typename stream >
void printUnifiedFormat (stream& out) const {
for_each(uniHunks.begin(), uniHunks.end(), UniHunkPrinter< sesElem, stream >(out));
}
void printUnifiedFormat (ostream& out = cout) const {
printUnifiedFormat< ostream >(out);
}
/**
* print unified format difference with given unified format hunks
*/
template < typename stream >
static void printUnifiedFormat (const uniHunkVec& hunks, stream& out) {
for_each(hunks.begin(), hunks.end(), UniHunkPrinter< sesElem >(out));
}
static void printUnifiedFormat (const uniHunkVec& hunks, ostream& out = cout) {
printUnifiedFormat< ostream >(hunks, out);
}
/**
* compose Unified Format Hunks from Shortest Edit Script
*/
void composeUnifiedHunks () {
sesElemVec common[2];
sesElemVec change;
sesElemVec ses_v = ses.getSequence();
long long l_cnt = 1;
long long length = distance(ses_v.begin(), ses_v.end());
long long middle = 0;
bool isMiddle, isAfter;
elemInfo einfo;
long long a, b, c, d; // @@ -a,b +c,d @@
long long inc_dec_count = 0;
uniHunk< sesElem > hunk;
sesElemVec adds;
sesElemVec deletes;
isMiddle = isAfter = false;
a = b = c = d = 0;
for (sesElemVec_iter it=ses_v.begin();it!=ses_v.end();++it, ++l_cnt) {
einfo = it->second;
switch (einfo.type) {
case SES_ADD :
middle = 0;
++inc_dec_count;
adds.push_back(*it);
if (!isMiddle) isMiddle = true;
if (isMiddle) ++d;
if (l_cnt >= length) {
joinSesVec(change, deletes);
joinSesVec(change, adds);
isAfter = true;
}
break;
case SES_DELETE :
middle = 0;
--inc_dec_count;
deletes.push_back(*it);
if (!isMiddle) isMiddle = true;
if (isMiddle) ++b;
if (l_cnt >= length) {
joinSesVec(change, deletes);
joinSesVec(change, adds);
isAfter = true;
}
break;
case SES_COMMON :
++b;++d;
if (common[1].empty() && adds.empty() && deletes.empty() && change.empty()) {
if (static_cast<long long>(common[0].size()) < DTL_CONTEXT_SIZE) {
if (a == 0 && c == 0) {
if (!wasSwapped()) {
a = einfo.beforeIdx;
c = einfo.afterIdx;
} else {
a = einfo.afterIdx;
c = einfo.beforeIdx;
}
}
common[0].push_back(*it);
} else {
rotate(common[0].begin(), common[0].begin() + 1, common[0].end());
common[0].pop_back();
common[0].push_back(*it);
++a;++c;
--b;--d;
}
}
if (isMiddle && !isAfter) {
++middle;
joinSesVec(change, deletes);
joinSesVec(change, adds);
change.push_back(*it);
if (middle >= DTL_SEPARATE_SIZE || l_cnt >= length) {
isAfter = true;
}
adds.clear();
deletes.clear();
}
break;
default :
// no through
break;
}
// compose unified format hunk
if (isAfter && !change.empty()) {
sesElemVec_iter cit = it;
long long cnt = 0;
for (long long i=0;i<DTL_SEPARATE_SIZE && (cit != ses_v.end());++i, ++cit) {
if (cit->second.type == SES_COMMON) {
++cnt;
}
}
if (cnt < DTL_SEPARATE_SIZE && l_cnt < length) {
middle = 0;
isAfter = false;
continue;
}
if (static_cast<long long>(common[0].size()) >= DTL_SEPARATE_SIZE) {
long long c0size = static_cast<long long>(common[0].size());
rotate(common[0].begin(),
common[0].begin() + (size_t)c0size - DTL_SEPARATE_SIZE,
common[0].end());
for (long long i=0;i<c0size - DTL_SEPARATE_SIZE;++i) {
common[0].pop_back();
}
a += c0size - DTL_SEPARATE_SIZE;
c += c0size - DTL_SEPARATE_SIZE;
}
if (a == 0) ++a;
if (c == 0) ++c;
if (wasSwapped()) swap(a, c);
hunk.a = a;
hunk.b = b;
hunk.c = c;
hunk.d = d;
hunk.common[0] = common[0];
hunk.change = change;
hunk.common[1] = common[1];
hunk.inc_dec_count = inc_dec_count;
uniHunks.push_back(hunk);
isMiddle = false;
isAfter = false;
common[0].clear();
common[1].clear();
adds.clear();
deletes.clear();
change.clear();
a = b = c = d = middle = inc_dec_count = 0;
}
}
}
/**
* compose ses from stream
*/
template <typename stream>
static Ses< elem > composeSesFromStream (stream& st)
{
elem line;
Ses< elem > ret;
long long x_idx, y_idx;
x_idx = y_idx = 1;
while (getline(st, line)) {
elem mark(line.begin(), line.begin() + 1);
elem e(line.begin() + 1, line.end());
if (mark == SES_MARK_DELETE) {
ret.addSequence(e, x_idx, 0, SES_DELETE);
++x_idx;
} else if (mark == SES_MARK_ADD) {
ret.addSequence(e, y_idx, 0, SES_ADD);
++y_idx;
} else if (mark == SES_MARK_COMMON) {
ret.addSequence(e, x_idx, y_idx, SES_COMMON);
++x_idx;
++y_idx;
}
}
return ret;
}
private :
/**
* initialize
*/
void init () {
M = distance(A.begin(), A.end());
N = distance(B.begin(), B.end());
if (M < N) {
swapped = false;
} else {
swap(A, B);
swap(M, N);
swapped = true;
}
editDistance = 0;
delta = N - M;
offset = M + 1;
huge = false;
trivial = false;
editDistanceOnly = false;
fp = NULL;
}
/**
* search shortest path and record the path
*/
long long snake(const long long& k, const long long& above, const long long& below) {
long long r = above > below ? path[(size_t)k-1+offset] : path[(size_t)k+1+offset];
long long y = max(above, below);
long long x = y - k;
while ((size_t)x < M && (size_t)y < N && (swapped ? cmp.impl(B[(size_t)y], A[(size_t)x]) : cmp.impl(A[(size_t)x], B[(size_t)y]))) {
++x;++y;
}
path[(size_t)k+offset] = static_cast<long long>(pathCordinates.size());
if (!editDistanceOnly) {
P p;
p.x = x;p.y = y;p.k = r;
pathCordinates.push_back(p);
}
return y;
}
/**
* record SES and LCS
*/
bool recordSequence (const editPathCordinates& v) {
sequence_const_iter x(A.begin());
sequence_const_iter y(B.begin());
long long x_idx, y_idx; // line number for Unified Format
long long px_idx, py_idx; // cordinates
bool complete = false;
x_idx = y_idx = 1;
px_idx = py_idx = 0;
for (size_t i=v.size()-1;!complete;--i) {
while(px_idx < v[i].x || py_idx < v[i].y) {
if (v[i].y - v[i].x > py_idx - px_idx) {
if (!wasSwapped()) {
ses.addSequence(*y, 0, y_idx + oy, SES_ADD);
} else {
ses.addSequence(*y, y_idx + oy, 0, SES_DELETE);
}
++y;
++y_idx;
++py_idx;
} else if (v[i].y - v[i].x < py_idx - px_idx) {
if (!wasSwapped()) {
ses.addSequence(*x, x_idx + ox, 0, SES_DELETE);
} else {
ses.addSequence(*x, 0, x_idx + ox, SES_ADD);
}
++x;
++x_idx;
++px_idx;
} else {
if (!wasSwapped()) {
lcs.addSequence(*x);
ses.addSequence(*x, x_idx + ox, y_idx + oy, SES_COMMON);
} else {
lcs.addSequence(*y);
ses.addSequence(*y, y_idx + oy, x_idx + ox, SES_COMMON);
}
++x;
++y;
++x_idx;
++y_idx;
++px_idx;
++py_idx;
}
}
if (i == 0) complete = true;
}
if (x_idx > static_cast<long long>(M) && y_idx > static_cast<long long>(N)) {
// all recording succeeded
} else {
// trivial difference
if (trivialEnabled()) {
if (!wasSwapped()) {
recordOddSequence(x_idx, M, x, SES_DELETE);
recordOddSequence(y_idx, N, y, SES_ADD);
} else {
recordOddSequence(x_idx, M, x, SES_ADD);
recordOddSequence(y_idx, N, y, SES_DELETE);
}
return true;
}
// nontrivial difference
sequence A_(A.begin() + (size_t)x_idx - 1, A.end());
sequence B_(B.begin() + (size_t)y_idx - 1, B.end());
A = A_;
B = B_;
M = distance(A.begin(), A.end());
N = distance(B.begin(), B.end());
delta = N - M;
offset = M + 1;
delete[] fp;
fp = new long long[M + N + 3];
fill(&fp[0], &fp[M + N + 3], -1);
fill(path.begin(), path.end(), -1);
ox = x_idx - 1;
oy = y_idx - 1;
return false;
}
return true;
}
/**
* record odd sequence in SES
*/
void inline recordOddSequence (long long idx, long long length, sequence_const_iter it, const edit_t et) {
while(idx < length){
ses.addSequence(*it, idx, 0, et);
++it;
++idx;
++editDistance;
}
ses.addSequence(*it, idx, 0, et);
++editDistance;
}
/**
* join SES vectors
*/
void inline joinSesVec (sesElemVec& s1, sesElemVec& s2) const {
if (!s2.empty()) {
for (sesElemVec_iter vit=s2.begin();vit!=s2.end();++vit) {
s1.push_back(*vit);
}
}
}
/**
* check if the sequences have been swapped
*/
bool inline wasSwapped () const {
return swapped;
}
};
}
#endif // DTL_DIFF_H

View File

@@ -1,245 +0,0 @@
/**
dtl -- Diff Template Library
In short, Diff Template Library is distributed under so called "BSD license",
Copyright (c) 2015 Tatsuhiko Kubo <cubicdaiya@gmail.com>
All rights reserved.
Redistribution and use in source and binary forms, with or without modification,
are permitted provided that the following conditions are met:
* Redistributions of source code must retain the above copyright notice,
this list of conditions and the following disclaimer.
* Redistributions in binary form must reproduce the above copyright notice,
this list of conditions and the following disclaimer in the documentation
and/or other materials provided with the distribution.
* Neither the name of the authors nor the names of its contributors
may be used to endorse or promote products derived from this software
without specific prior written permission.
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED
TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF
LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING
NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
*/
/* If you use this library, you must include dtl.hpp only. */
#ifndef DTL_DIFF3_H
#define DTL_DIFF3_H
namespace dtl {
/**
* diff3 class template
* sequence must support random_access_iterator.
*/
template <typename elem, typename sequence = vector< elem >, typename comparator = Compare< elem > >
class Diff3
{
private:
dtl_typedefs(elem, sequence)
sequence A;
sequence B;
sequence C;
sequence S;
Diff< elem, sequence, comparator > diff_ba;
Diff< elem, sequence, comparator > diff_bc;
bool conflict;
elem csepabegin;
elem csepa;
elem csepaend;
public :
Diff3 () {}
Diff3 (const sequence& a,
const sequence& b,
const sequence& c) : A(a), B(b), C(c),
diff_ba(b, a), diff_bc(b, c),
conflict(false) {}
~Diff3 () {}
bool isConflict () const {
return conflict;
}
sequence getMergedSequence () const {
return S;
}
/**
* merge changes B and C into A
*/
bool merge () {
if (diff_ba.getEditDistance() == 0) { // A == B
if (diff_bc.getEditDistance() == 0) { // A == B == C
S = B;
return true;
}
S = C;
return true;
} else { // A != B
if (diff_bc.getEditDistance() == 0) { // A != B == C
S = A;
return true;
} else { // A != B != C
S = merge_();
if (isConflict()) { // conflict occured
return false;
}
}
}
return true;
}
/**
* compose differences
*/
void compose () {
diff_ba.compose();
diff_bc.compose();
}
private :
/**
* merge implementation
*/
sequence merge_ () {
elemVec seq;
Ses< elem > ses_ba = diff_ba.getSes();
Ses< elem > ses_bc = diff_bc.getSes();
sesElemVec ses_ba_v = ses_ba.getSequence();
sesElemVec ses_bc_v = ses_bc.getSequence();
sesElemVec_iter ba_it = ses_ba_v.begin();
sesElemVec_iter bc_it = ses_bc_v.begin();
sesElemVec_iter ba_end = ses_ba_v.end();
sesElemVec_iter bc_end = ses_bc_v.end();
while (!isEnd(ba_end, ba_it) || !isEnd(bc_end, bc_it)) {
while (true) {
if (!isEnd(ba_end, ba_it) &&
!isEnd(bc_end, bc_it) &&
ba_it->first == bc_it->first &&
ba_it->second.type == SES_COMMON &&
bc_it->second.type == SES_COMMON) {
// do nothing
} else {
break;
}
if (!isEnd(ba_end, ba_it)) seq.push_back(ba_it->first);
else if (!isEnd(bc_end, bc_it)) seq.push_back(bc_it->first);
forwardUntilEnd(ba_end, ba_it);
forwardUntilEnd(bc_end, bc_it);
}
if (isEnd(ba_end, ba_it) || isEnd(bc_end, bc_it)) break;
if ( ba_it->second.type == SES_COMMON
&& bc_it->second.type == SES_DELETE) {
forwardUntilEnd(ba_end, ba_it);
forwardUntilEnd(bc_end, bc_it);
} else if (ba_it->second.type == SES_COMMON &&
bc_it->second.type == SES_ADD) {
seq.push_back(bc_it->first);
forwardUntilEnd(bc_end, bc_it);
} else if (ba_it->second.type == SES_DELETE &&
bc_it->second.type == SES_COMMON) {
forwardUntilEnd(ba_end, ba_it);
forwardUntilEnd(bc_end, bc_it);
} else if (ba_it->second.type == SES_DELETE &&
bc_it->second.type == SES_DELETE) {
if (ba_it->first == bc_it->first) {
forwardUntilEnd(ba_end, ba_it);
forwardUntilEnd(bc_end, bc_it);
} else {
// conflict
conflict = true;
return B;
}
} else if (ba_it->second.type == SES_DELETE &&
bc_it->second.type == SES_ADD) {
// conflict
conflict = true;
return B;
} else if (ba_it->second.type == SES_ADD &&
bc_it->second.type == SES_COMMON) {
seq.push_back(ba_it->first);
forwardUntilEnd(ba_end, ba_it);
} else if (ba_it->second.type == SES_ADD &&
bc_it->second.type == SES_DELETE) {
// conflict
conflict = true;
return B;
} else if (ba_it->second.type == SES_ADD &&
bc_it->second.type == SES_ADD) {
if (ba_it->first == bc_it->first) {
seq.push_back(ba_it->first);
forwardUntilEnd(ba_end, ba_it);
forwardUntilEnd(bc_end, bc_it);
} else {
// conflict
conflict = true;
return B;
}
}
}
if (isEnd(ba_end, ba_it)) {
addDecentSequence(bc_end, bc_it, seq);
} else if (isEnd(bc_end, bc_it)) {
addDecentSequence(ba_end, ba_it, seq);
}
sequence mergedSeq(seq.begin(), seq.end());
return mergedSeq;
}
/**
* join elem vectors
*/
void inline joinElemVec (elemVec& s1, elemVec& s2) const {
if (!s2.empty()) {
for (elemVec_iter vit=s2.begin();vit!=s2.end();++vit) {
s1.push_back(*vit);
}
}
}
/**
* check if sequence is at end
*/
template <typename T_iter>
bool inline isEnd (const T_iter& end, const T_iter& it) const {
return it == end ? true : false;
}
/**
* increment iterator until iterator is at end
*/
template <typename T_iter>
void inline forwardUntilEnd (const T_iter& end, T_iter& it) const {
if (!isEnd(end, it)) ++it;
}
/**
* add elements whose SES's type is ADD
*/
void inline addDecentSequence (const sesElemVec_iter& end, sesElemVec_iter& it, elemVec& seq) const {
while (!isEnd(end, it)) {
if (it->second.type == SES_ADD) seq.push_back(it->first);
++it;
}
}
};
}
#endif // DTL_DIFF3_H

View File

@@ -1,55 +0,0 @@
/**
dtl -- Diff Template Library
In short, Diff Template Library is distributed under so called "BSD license",
Copyright (c) 2015 Tatsuhiko Kubo <cubicdaiya@gmail.com>
All rights reserved.
Redistribution and use in source and binary forms, with or without modification,
are permitted provided that the following conditions are met:
* Redistributions of source code must retain the above copyright notice,
this list of conditions and the following disclaimer.
* Redistributions in binary form must reproduce the above copyright notice,
this list of conditions and the following disclaimer in the documentation
and/or other materials provided with the distribution.
* Neither the name of the authors nor the names of its contributors
may be used to endorse or promote products derived from this software
without specific prior written permission.
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED
TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF
LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING
NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
*/
/* If you use this library, you must include dtl.hpp only. */
#ifndef DTL_LCS_H
#define DTL_LCS_H
namespace dtl {
/**
* Longest Common Subsequence template class
*/
template <typename elem>
class Lcs : public Sequence< elem >
{
public :
Lcs () {}
~Lcs () {}
};
}
#endif // DTL_LCS_H

View File

@@ -1,65 +0,0 @@
/**
dtl -- Diff Template Library
In short, Diff Template Library is distributed under so called "BSD license",
Copyright (c) 2015 Tatsuhiko Kubo <cubicdaiya@gmail.com>
All rights reserved.
Redistribution and use in source and binary forms, with or without modification,
are permitted provided that the following conditions are met:
* Redistributions of source code must retain the above copyright notice,
this list of conditions and the following disclaimer.
* Redistributions in binary form must reproduce the above copyright notice,
this list of conditions and the following disclaimer in the documentation
and/or other materials provided with the distribution.
* Neither the name of the authors nor the names of its contributors
may be used to endorse or promote products derived from this software
without specific prior written permission.
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED
TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF
LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING
NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
*/
/* If you use this library, you must include dtl.hpp only. */
#ifndef DTL_SEQUENCE_H
#define DTL_SEQUENCE_H
namespace dtl {
/**
* sequence class template
*/
template <typename elem>
class Sequence
{
public :
typedef vector< elem > elemVec;
Sequence () {}
virtual ~Sequence () {}
elemVec getSequence () const {
return sequence;
}
void addSequence (elem e) {
sequence.push_back(e);
}
protected :
elemVec sequence;
};
}
#endif // DTL_SEQUENCE_H

View File

@@ -1,132 +0,0 @@
/**
dtl -- Diff Template Library
In short, Diff Template Library is distributed under so called "BSD license",
Copyright (c) 2015 Tatsuhiko Kubo <cubicdaiya@gmail.com>
All rights reserved.
Redistribution and use in source and binary forms, with or without modification,
are permitted provided that the following conditions are met:
* Redistributions of source code must retain the above copyright notice,
this list of conditions and the following disclaimer.
* Redistributions in binary form must reproduce the above copyright notice,
this list of conditions and the following disclaimer in the documentation
and/or other materials provided with the distribution.
* Neither the name of the authors nor the names of its contributors
may be used to endorse or promote products derived from this software
without specific prior written permission.
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED
TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF
LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING
NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
*/
/* If you use this library, you must include dtl.hpp only. */
#ifndef DTL_SES_H
#define DTL_SES_H
namespace dtl {
/**
* Shortest Edit Script template class
*/
template <typename elem>
class Ses : public Sequence< elem >
{
private :
typedef pair< elem, elemInfo > sesElem;
typedef vector< sesElem > sesElemVec;
public :
Ses () : onlyAdd(true), onlyDelete(true), onlyCopy(true), deletesFirst(false) {
nextDeleteIdx = 0;
}
Ses (bool moveDel) : onlyAdd(true), onlyDelete(true), onlyCopy(true), deletesFirst(moveDel) {
nextDeleteIdx = 0;
}
~Ses () {}
bool isOnlyAdd () const {
return onlyAdd;
}
bool isOnlyDelete () const {
return onlyDelete;
}
bool isOnlyCopy () const {
return onlyCopy;
}
bool isOnlyOneOperation () const {
return isOnlyAdd() || isOnlyDelete() || isOnlyCopy();
}
bool isChange () const {
return !onlyCopy;
}
using Sequence< elem >::addSequence;
void addSequence (elem e, long long beforeIdx, long long afterIdx, const edit_t type) {
elemInfo info;
info.beforeIdx = beforeIdx;
info.afterIdx = afterIdx;
info.type = type;
sesElem pe(e, info);
if (!deletesFirst) {
sequence.push_back(pe);
}
switch (type) {
case SES_DELETE:
onlyCopy = false;
onlyAdd = false;
if (deletesFirst) {
sequence.insert(sequence.begin() + nextDeleteIdx, pe);
nextDeleteIdx++;
}
break;
case SES_COMMON:
onlyAdd = false;
onlyDelete = false;
if (deletesFirst) {
sequence.push_back(pe);
nextDeleteIdx = sequence.size();
}
break;
case SES_ADD:
onlyDelete = false;
onlyCopy = false;
if (deletesFirst) {
sequence.push_back(pe);
}
break;
}
}
sesElemVec getSequence () const {
return sequence;
}
private :
sesElemVec sequence;
bool onlyAdd;
bool onlyDelete;
bool onlyCopy;
bool deletesFirst;
size_t nextDeleteIdx;
};
}
#endif // DTL_SES_H

View File

@@ -1,47 +0,0 @@
/**
dtl -- Diff Template Library
In short, Diff Template Library is distributed under so called "BSD license",
Copyright (c) 2015 Tatsuhiko Kubo <cubicdaiya@gmail.com>
All rights reserved.
Redistribution and use in source and binary forms, with or without modification,
are permitted provided that the following conditions are met:
* Redistributions of source code must retain the above copyright notice,
this list of conditions and the following disclaimer.
* Redistributions in binary form must reproduce the above copyright notice,
this list of conditions and the following disclaimer in the documentation
and/or other materials provided with the distribution.
* Neither the name of the authors nor the names of its contributors
may be used to endorse or promote products derived from this software
without specific prior written permission.
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED
TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF
LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING
NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
*/
#ifndef DTL_H
#define DTL_H
#include "variables.hpp"
#include "functors.hpp"
#include "Sequence.hpp"
#include "Lcs.hpp"
#include "Ses.hpp"
#include "Diff.hpp"
#include "Diff3.hpp"
#endif // DTL_H

View File

@@ -1,151 +0,0 @@
/**
dtl -- Diff Template Library
In short, Diff Template Library is distributed under so called "BSD license",
Copyright (c) 2015 Tatsuhiko Kubo <cubicdaiya@gmail.com>
All rights reserved.
Redistribution and use in source and binary forms, with or without modification,
are permitted provided that the following conditions are met:
* Redistributions of source code must retain the above copyright notice,
this list of conditions and the following disclaimer.
* Redistributions in binary form must reproduce the above copyright notice,
this list of conditions and the following disclaimer in the documentation
and/or other materials provided with the distribution.
* Neither the name of the authors nor the names of its contributors
may be used to endorse or promote products derived from this software
without specific prior written permission.
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED
TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF
LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING
NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
*/
/* If you use this library, you must include dtl.hpp only. */
#ifndef DTL_FUNCTORS_H
#define DTL_FUNCTORS_H
namespace dtl {
/**
* printer class template
*/
template <typename sesElem, typename stream = ostream >
class Printer
{
public :
Printer () : out_(cout) {}
Printer (stream& out) : out_(out) {}
virtual ~Printer () {}
virtual void operator() (const sesElem& se) const = 0;
protected :
stream& out_;
};
/**
* common element printer class template
*/
template <typename sesElem, typename stream = ostream >
class CommonPrinter : public Printer < sesElem, stream >
{
public :
CommonPrinter () : Printer < sesElem, stream > () {}
CommonPrinter (stream& out) : Printer < sesElem, stream > (out) {}
~CommonPrinter () {}
void operator() (const sesElem& se) const {
this->out_ << SES_MARK_COMMON << se.first << endl;
}
};
/**
* ses element printer class template
*/
template <typename sesElem, typename stream = ostream >
class ChangePrinter : public Printer < sesElem, stream >
{
public :
ChangePrinter () : Printer < sesElem, stream > () {}
ChangePrinter (stream& out) : Printer < sesElem, stream > (out) {}
~ChangePrinter () {}
void operator() (const sesElem& se) const {
switch (se.second.type) {
case SES_ADD:
this->out_ << SES_MARK_ADD << se.first << endl;
break;
case SES_DELETE:
this->out_ << SES_MARK_DELETE << se.first << endl;
break;
case SES_COMMON:
this->out_ << SES_MARK_COMMON << se.first << endl;
break;
}
}
};
/**
* unified format element printer class template
*/
template <typename sesElem, typename stream = ostream >
class UniHunkPrinter
{
public :
UniHunkPrinter () : out_(cout) {}
UniHunkPrinter (stream& out) : out_(out) {}
~UniHunkPrinter () {}
void operator() (const uniHunk< sesElem >& hunk) const {
out_ << "@@"
<< " -" << hunk.a << "," << hunk.b
<< " +" << hunk.c << "," << hunk.d
<< " @@" << endl;
for_each(hunk.common[0].begin(), hunk.common[0].end(), CommonPrinter< sesElem, stream >(out_));
for_each(hunk.change.begin(), hunk.change.end(), ChangePrinter< sesElem, stream >(out_));
for_each(hunk.common[1].begin(), hunk.common[1].end(), CommonPrinter< sesElem, stream >(out_));
}
private :
stream& out_;
};
/**
* storage class template
*/
template <typename sesElem, typename storedData >
class Storage
{
public:
Storage(storedData& sd) : storedData_(sd) {}
virtual ~Storage() {}
virtual void operator() (const sesElem& se) const = 0;
protected:
storedData& storedData_;
};
/**
* compare class template
*/
template <typename elem>
class Compare
{
public :
Compare () {}
virtual ~Compare () {}
virtual inline bool impl (const elem& e1, const elem& e2) const {
return e1 == e2;
}
};
}
#endif // DTL_FUNCTORS_H

View File

@@ -1,142 +0,0 @@
/**
dtl -- Diff Template Library
In short, Diff Template Library is distributed under so called "BSD license",
Copyright (c) 2015 Tatsuhiko Kubo <cubicdaiya@gmail.com>
All rights reserved.
Redistribution and use in source and binary forms, with or without modification,
are permitted provided that the following conditions are met:
* Redistributions of source code must retain the above copyright notice,
this list of conditions and the following disclaimer.
* Redistributions in binary form must reproduce the above copyright notice,
this list of conditions and the following disclaimer in the documentation
and/or other materials provided with the distribution.
* Neither the name of the authors nor the names of its contributors
may be used to endorse or promote products derived from this software
without specific prior written permission.
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED
TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF
LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING
NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
*/
/* If you use this library, you must include dtl.hpp only. */
#ifndef DTL_VARIABLES_H
#define DTL_VARIABLES_H
#include <vector>
#include <list>
#include <string>
#include <algorithm>
#include <iostream>
namespace dtl {
using std::vector;
using std::string;
using std::pair;
using std::ostream;
using std::list;
using std::for_each;
using std::distance;
using std::fill;
using std::cout;
using std::endl;
using std::rotate;
using std::swap;
using std::max;
/**
* version string
*/
const string version = "1.20";
/**
* type of edit for SES
*/
typedef int edit_t;
const edit_t SES_DELETE = -1;
const edit_t SES_COMMON = 0;
const edit_t SES_ADD = 1;
/**
* mark of SES
*/
#define SES_MARK_DELETE "-"
#define SES_MARK_COMMON " "
#define SES_MARK_ADD "+"
/**
* info for Unified Format
*/
typedef struct eleminfo {
long long beforeIdx; // index of prev sequence
long long afterIdx; // index of after sequence
edit_t type; // type of edit(Add, Delete, Common)
bool operator==(const eleminfo& other) const{
return (this->beforeIdx == other.beforeIdx && this->afterIdx == other.afterIdx && this->type == other.type);
}
} elemInfo;
const long long DTL_SEPARATE_SIZE = 3;
const long long DTL_CONTEXT_SIZE = 3;
/**
* cordinate for registering route
*/
typedef struct Point {
long long x; // x cordinate
long long y; // y cordinate
long long k; // vertex
} P;
/**
* limit of cordinate size
*/
const unsigned long long MAX_CORDINATES_SIZE = 2000000;
typedef vector< long long > editPath;
typedef vector< P > editPathCordinates;
/**
* Structure of Unified Format Hunk
*/
template <typename sesElem>
struct uniHunk {
long long a, b, c, d; // @@ -a,b +c,d @@
vector< sesElem > common[2]; // anteroposterior commons on changes
vector< sesElem > change; // changes
long long inc_dec_count; // count of increace and decrease
};
#define dtl_typedefs(elem, sequence) \
typedef pair< elem, elemInfo > sesElem; \
typedef vector< sesElem > sesElemVec; \
typedef vector< uniHunk< sesElem > > uniHunkVec; \
typedef list< elem > elemList; \
typedef vector< elem > elemVec; \
typedef typename uniHunkVec::iterator uniHunkVec_iter; \
typedef typename sesElemVec::iterator sesElemVec_iter; \
typedef typename elemList::iterator elemList_iter; \
typedef typename sequence::iterator sequence_iter; \
typedef typename sequence::const_iterator sequence_const_iter; \
typedef typename elemVec::iterator elemVec_iter;
}
#endif // DTL_VARIABLES_H

View File

@@ -1,17 +0,0 @@
cmake_minimum_required(VERSION 3.10)
project(OpenCLVectorAdd)
set(THREADS_PREFER_PTHREAD_FLAG ON)
find_package(Threads REQUIRED)
find_package(OpenCL REQUIRED)
add_executable(OpenCLVectorAdd OpenCLVectorAdd.cpp)
add_library(TracyClient STATIC ../../public/TracyClient.cpp
../../public/tracy/TracyOpenCL.hpp)
target_include_directories(TracyClient PUBLIC ../../public/tracy)
target_compile_definitions(TracyClient PUBLIC TRACY_ENABLE=1)
target_link_libraries(OpenCLVectorAdd PUBLIC OpenCL::OpenCL TracyClient ${CMAKE_DL_LIBS} Threads::Threads)

View File

@@ -1,220 +0,0 @@
#include <algorithm>
#include <iostream>
#include <cassert>
#include <string>
#include <vector>
#include <numeric>
#include <CL/cl.h>
#include <Tracy.hpp>
#include <TracyOpenCL.hpp>
#define CL_ASSERT(err) \
if((err) != CL_SUCCESS) \
{ \
std::cerr << "OpenCL Call Returned " << err << std::endl; \
assert(false); \
}
const char kernelSource[] =
" void __kernel vectorAdd(global float* C, global float* A, global float* B, int N) "
" { "
" int i = get_global_id(0); "
" if (i < N) { "
" C[i] = A[i] + B[i]; "
" } "
" } ";
int main()
{
cl_platform_id platform;
cl_device_id device;
cl_context context;
cl_command_queue commandQueue;
cl_kernel vectorAddKernel;
cl_program program;
cl_int err;
cl_mem bufferA, bufferB, bufferC;
TracyCLCtx tracyCLCtx;
{
ZoneScopedN("OpenCL Init");
cl_uint numPlatforms = 0;
CL_ASSERT(clGetPlatformIDs(0, nullptr, &numPlatforms));
if (numPlatforms == 0)
{
std::cerr << "Cannot find OpenCL platform to run this application" << std::endl;
return 1;
}
CL_ASSERT(clGetPlatformIDs(1, &platform, nullptr));
size_t platformNameBufferSize = 0;
CL_ASSERT(clGetPlatformInfo(platform, CL_PLATFORM_NAME, 0, nullptr, &platformNameBufferSize));
std::string platformName(platformNameBufferSize, '\0');
CL_ASSERT(clGetPlatformInfo(platform, CL_PLATFORM_NAME, platformNameBufferSize, &platformName[0], nullptr));
std::cout << "OpenCL Platform: " << platformName << std::endl;
CL_ASSERT(clGetDeviceIDs(platform, CL_DEVICE_TYPE_ALL, 1, &device, nullptr));
size_t deviceNameBufferSize = 0;
CL_ASSERT(clGetDeviceInfo(device, CL_DEVICE_NAME, 0, nullptr, &deviceNameBufferSize));
std::string deviceName(deviceNameBufferSize, '\0');
CL_ASSERT(clGetDeviceInfo(device, CL_DEVICE_NAME, deviceNameBufferSize, &deviceName[0], nullptr));
std::cout << "OpenCL Device: " << deviceName << std::endl;
err = CL_SUCCESS;
context = clCreateContext(nullptr, 1, &device, nullptr, nullptr, &err);
CL_ASSERT(err);
size_t kernelSourceLength = sizeof(kernelSource);
const char* kernelSourceArray = { kernelSource };
program = clCreateProgramWithSource(context, 1, &kernelSourceArray, &kernelSourceLength, &err);
CL_ASSERT(err);
if (clBuildProgram(program, 1, &device, nullptr, nullptr, nullptr) != CL_SUCCESS)
{
size_t programBuildLogBufferSize = 0;
CL_ASSERT(clGetProgramBuildInfo(program, device, CL_PROGRAM_BUILD_LOG, 0, nullptr, &programBuildLogBufferSize));
std::string programBuildLog(programBuildLogBufferSize, '\0');
CL_ASSERT(clGetProgramBuildInfo(program, device, CL_PROGRAM_BUILD_LOG, programBuildLogBufferSize, &programBuildLog[0], nullptr));
std::clog << programBuildLog << std::endl;
return 1;
}
vectorAddKernel = clCreateKernel(program, "vectorAdd", &err);
CL_ASSERT(err);
commandQueue = clCreateCommandQueue(context, device, CL_QUEUE_PROFILING_ENABLE, &err);
CL_ASSERT(err);
}
tracyCLCtx = TracyCLContext(context, device);
size_t N = 10 * 1024 * 1024 / sizeof(float); // 10MB of floats
std::vector<float> hostA, hostB, hostC;
{
ZoneScopedN("Host Data Init");
hostA.resize(N);
hostB.resize(N);
hostC.resize(N);
std::iota(std::begin(hostA), std::end(hostA), 0.0f);
std::iota(std::begin(hostB), std::end(hostB), 0.0f);
}
{
ZoneScopedN("Host to Device Memory Copy");
bufferA = clCreateBuffer(context, CL_MEM_READ_WRITE, N * sizeof(float), nullptr, &err);
CL_ASSERT(err);
bufferB = clCreateBuffer(context, CL_MEM_READ_WRITE, N * sizeof(float), nullptr, &err);
CL_ASSERT(err);
bufferC = clCreateBuffer(context, CL_MEM_READ_WRITE, N * sizeof(float), nullptr, &err);
CL_ASSERT(err);
cl_event writeBufferAEvent, writeBufferBEvent;
{
ZoneScopedN("Write Buffer A");
TracyCLZoneS(tracyCLCtx, "Write BufferA", 5);
CL_ASSERT(clEnqueueWriteBuffer(commandQueue, bufferA, CL_FALSE, 0, N * sizeof(float), hostA.data(), 0, nullptr, &writeBufferAEvent));
TracyCLZoneSetEvent(writeBufferAEvent);
}
{
ZoneScopedN("Write Buffer B");
TracyCLZone(tracyCLCtx, "Write BufferB");
CL_ASSERT(clEnqueueWriteBuffer(commandQueue, bufferB, CL_FALSE, 0, N * sizeof(float), hostB.data(), 0, nullptr, &writeBufferBEvent));
TracyCLZoneSetEvent(writeBufferBEvent);
}
}
cl_int clN = static_cast<cl_int>(N);
const int numFrames = 10;
const int launchsPerFrame = 10;
constexpr int numLaunchs = numFrames * launchsPerFrame;
std::vector<cl_event> kernelLaunchEvts;
kernelLaunchEvts.reserve(numLaunchs);
for (int i = 0; i < numFrames; ++i)
{
FrameMark;
for (int j = 0; j < launchsPerFrame; ++j) {
ZoneScopedN("VectorAdd Kernel Launch");
TracyCLZoneC(tracyCLCtx, "VectorAdd Kernel", tracy::Color::Blue4);
CL_ASSERT(clSetKernelArg(vectorAddKernel, 0, sizeof(cl_mem), &bufferC));
CL_ASSERT(clSetKernelArg(vectorAddKernel, 1, sizeof(cl_mem), &bufferA));
CL_ASSERT(clSetKernelArg(vectorAddKernel, 2, sizeof(cl_mem), &bufferB));
CL_ASSERT(clSetKernelArg(vectorAddKernel, 3, sizeof(cl_int), &clN));
cl_event vectorAddKernelEvent;
CL_ASSERT(clEnqueueNDRangeKernel(commandQueue, vectorAddKernel, 1, nullptr, &N, nullptr, 0, nullptr, &vectorAddKernelEvent));
TracyCLZoneSetEvent(vectorAddKernelEvent);
CL_ASSERT(clRetainEvent(vectorAddKernelEvent));
kernelLaunchEvts.push_back(vectorAddKernelEvent);
std::cout << "VectorAdd Kernel Enqueued" << std::endl;
}
{
// Wait frame events to be finished
ZoneScopedN("clFinish");
CL_ASSERT(clFinish(commandQueue));
}
// You should collect on each 'frame' ends, so that streaming can be achieved.
TracyCLCollect(tracyCLCtx);
}
{
ZoneScopedN("Device to Host Memory Copy");
TracyCLZone(tracyCLCtx, "Read Buffer C");
cl_event readbufferCEvent;
CL_ASSERT(clEnqueueReadBuffer(commandQueue, bufferC, CL_TRUE, 0, N * sizeof(float), hostC.data(), 0, nullptr, &readbufferCEvent));
TracyCLZoneSetEvent(readbufferCEvent);
}
CL_ASSERT(clFinish(commandQueue));
std::vector<float> durations(kernelLaunchEvts.size());
for (int i=0; i<kernelLaunchEvts.size(); i++) {
cl_event evt = kernelLaunchEvts[i];
cl_ulong start;
cl_ulong end;
CL_ASSERT(clGetEventProfilingInfo(evt, CL_PROFILING_COMMAND_START, sizeof(cl_ulong), &start, nullptr));
CL_ASSERT(clGetEventProfilingInfo(evt, CL_PROFILING_COMMAND_END, sizeof(cl_ulong), &end, nullptr));
CL_ASSERT(clReleaseEvent(evt));
durations[i] = (end - start) * 0.001f;
std::cout << "VectorAdd Kernel " << i << " tooks " << static_cast<int>(durations[i]) << "us" << std::endl;
};
float avg = std::accumulate(durations.cbegin(), durations.cend(), 0.0f) / durations.size();
float stddev2 = std::accumulate(durations.cbegin(), durations.cend(), 0.0f, [avg](const float& acc, const float& v) {
auto d = v - avg;
return acc + d*d;
}) / (durations.size() - 1.0f);
std::cout << "VectorAdd runtime avg: " << avg << "us, std: " << sqrt(stddev2) << "us over " << numLaunchs << " runs." << std::endl;
// User should ensure all events are finished, in this case, collect after the clFinish will do the trick.
TracyCLCollect(tracyCLCtx);
{
ZoneScopedN("Checking results");
for (int i = 0; i < N; ++i)
{
assert(hostC[i] == hostA[i] + hostB[i]);
}
}
std::cout << "Results are correct!" << std::endl;
TracyCLDestroy(tracyCLCtx);
return 0;
}

View File

@@ -1 +0,0 @@
Windows/Compiled*Shader.h

View File

@@ -1,4 +0,0 @@
https://github.com/aras-p/ToyPathTracer
Modified to render only 10 frames. Client part requires 12 GB, server part
requires 6.4 GB.

View File

@@ -1,33 +0,0 @@
#if defined(__APPLE__) && !defined(__METAL_VERSION__)
#include <TargetConditionals.h>
#endif
#define kBackbufferWidth 1280
#define kBackbufferHeight 720
#if defined(__EMSCRIPTEN__)
#define CPU_CAN_DO_SIMD 0
#define CPU_CAN_DO_THREADS 0
#else
#define CPU_CAN_DO_SIMD 1
#define CPU_CAN_DO_THREADS 1
#endif
#define DO_SAMPLES_PER_PIXEL 4
#define DO_ANIMATE_SMOOTHING 0.9f
#define DO_LIGHT_SAMPLING 1
#define DO_MITSUBA_COMPARE 0
// Should path tracing be done on the GPU with a compute shader?
#define DO_COMPUTE_GPU 0
#define kCSGroupSizeX 8
#define kCSGroupSizeY 8
#define kCSMaxObjects 64
// Should float3 struct use SSE/NEON?
#define DO_FLOAT3_WITH_SIMD (!(DO_COMPUTE_GPU) && CPU_CAN_DO_SIMD && 1)
// Should HitSpheres function use SSE/NEON?
#define DO_HIT_SPHERES_SIMD (CPU_CAN_DO_SIMD && 1)

View File

@@ -1,192 +0,0 @@
#pragma once
#if defined(_MSC_VER)
#define VM_INLINE __forceinline
#else
#define VM_INLINE __attribute__((unused, always_inline, nodebug)) inline
#endif
#define kSimdWidth 4
#if !defined(__arm__) && !defined(__arm64__) && !defined(__EMSCRIPTEN__)
// ---- SSE implementation
#include <xmmintrin.h>
#include <emmintrin.h>
#include <smmintrin.h>
#define SHUFFLE4(V, X,Y,Z,W) float4(_mm_shuffle_ps((V).m, (V).m, _MM_SHUFFLE(W,Z,Y,X)))
struct float4
{
VM_INLINE float4() {}
VM_INLINE explicit float4(const float *p) { m = _mm_loadu_ps(p); }
VM_INLINE explicit float4(float x, float y, float z, float w) { m = _mm_set_ps(w, z, y, x); }
VM_INLINE explicit float4(float v) { m = _mm_set_ps1(v); }
VM_INLINE explicit float4(__m128 v) { m = v; }
VM_INLINE float getX() const { return _mm_cvtss_f32(m); }
VM_INLINE float getY() const { return _mm_cvtss_f32(_mm_shuffle_ps(m, m, _MM_SHUFFLE(1, 1, 1, 1))); }
VM_INLINE float getZ() const { return _mm_cvtss_f32(_mm_shuffle_ps(m, m, _MM_SHUFFLE(2, 2, 2, 2))); }
VM_INLINE float getW() const { return _mm_cvtss_f32(_mm_shuffle_ps(m, m, _MM_SHUFFLE(3, 3, 3, 3))); }
__m128 m;
};
typedef float4 bool4;
VM_INLINE float4 operator+ (float4 a, float4 b) { a.m = _mm_add_ps(a.m, b.m); return a; }
VM_INLINE float4 operator- (float4 a, float4 b) { a.m = _mm_sub_ps(a.m, b.m); return a; }
VM_INLINE float4 operator* (float4 a, float4 b) { a.m = _mm_mul_ps(a.m, b.m); return a; }
VM_INLINE bool4 operator==(float4 a, float4 b) { a.m = _mm_cmpeq_ps(a.m, b.m); return a; }
VM_INLINE bool4 operator!=(float4 a, float4 b) { a.m = _mm_cmpneq_ps(a.m, b.m); return a; }
VM_INLINE bool4 operator< (float4 a, float4 b) { a.m = _mm_cmplt_ps(a.m, b.m); return a; }
VM_INLINE bool4 operator> (float4 a, float4 b) { a.m = _mm_cmpgt_ps(a.m, b.m); return a; }
VM_INLINE bool4 operator<=(float4 a, float4 b) { a.m = _mm_cmple_ps(a.m, b.m); return a; }
VM_INLINE bool4 operator>=(float4 a, float4 b) { a.m = _mm_cmpge_ps(a.m, b.m); return a; }
VM_INLINE bool4 operator&(bool4 a, bool4 b) { a.m = _mm_and_ps(a.m, b.m); return a; }
VM_INLINE bool4 operator|(bool4 a, bool4 b) { a.m = _mm_or_ps(a.m, b.m); return a; }
VM_INLINE float4 operator- (float4 a) { a.m = _mm_xor_ps(a.m, _mm_set1_ps(-0.0f)); return a; }
VM_INLINE float4 min(float4 a, float4 b) { a.m = _mm_min_ps(a.m, b.m); return a; }
VM_INLINE float4 max(float4 a, float4 b) { a.m = _mm_max_ps(a.m, b.m); return a; }
VM_INLINE float hmin(float4 v)
{
v = min(v, SHUFFLE4(v, 2, 3, 0, 0));
v = min(v, SHUFFLE4(v, 1, 0, 0, 0));
return v.getX();
}
// Returns a 4-bit code where bit0..bit3 is X..W
VM_INLINE unsigned mask(float4 v) { return _mm_movemask_ps(v.m); }
// Once we have a comparison, we can branch based on its results:
VM_INLINE bool any(bool4 v) { return mask(v) != 0; }
VM_INLINE bool all(bool4 v) { return mask(v) == 15; }
// "select", i.e. hibit(cond) ? b : a
// on SSE4.1 and up this can be done easily via "blend" instruction;
// on older SSEs has to do a bunch of hoops, see
// https://fgiesen.wordpress.com/2016/04/03/sse-mind-the-gap/
VM_INLINE float4 select(float4 a, float4 b, bool4 cond)
{
#if defined(__SSE4_1__) || defined(_MSC_VER) // on windows assume we always have SSE4.1
a.m = _mm_blendv_ps(a.m, b.m, cond.m);
#else
__m128 d = _mm_castsi128_ps(_mm_srai_epi32(_mm_castps_si128(cond.m), 31));
a.m = _mm_or_ps(_mm_and_ps(d, b.m), _mm_andnot_ps(d, a.m));
#endif
return a;
}
VM_INLINE __m128i select(__m128i a, __m128i b, bool4 cond)
{
#if defined(__SSE4_1__) || defined(_MSC_VER) // on windows assume we always have SSE4.1
return _mm_blendv_epi8(a, b, _mm_castps_si128(cond.m));
#else
__m128i d = _mm_srai_epi32(_mm_castps_si128(cond.m), 31);
return _mm_or_si128(_mm_and_si128(d, b), _mm_andnot_si128(d, a));
#endif
}
VM_INLINE float4 sqrtf(float4 v) { return float4(_mm_sqrt_ps(v.m)); }
#elif !defined(__EMSCRIPTEN__)
// ---- NEON implementation
#define USE_NEON 1
#include <arm_neon.h>
struct float4
{
VM_INLINE float4() {}
VM_INLINE explicit float4(const float *p) { m = vld1q_f32(p); }
VM_INLINE explicit float4(float x, float y, float z, float w) { float v[4] = {x, y, z, w}; m = vld1q_f32(v); }
VM_INLINE explicit float4(float v) { m = vdupq_n_f32(v); }
VM_INLINE explicit float4(float32x4_t v) { m = v; }
VM_INLINE float getX() const { return vgetq_lane_f32(m, 0); }
VM_INLINE float getY() const { return vgetq_lane_f32(m, 1); }
VM_INLINE float getZ() const { return vgetq_lane_f32(m, 2); }
VM_INLINE float getW() const { return vgetq_lane_f32(m, 3); }
float32x4_t m;
};
typedef float4 bool4;
VM_INLINE float4 operator+ (float4 a, float4 b) { a.m = vaddq_f32(a.m, b.m); return a; }
VM_INLINE float4 operator- (float4 a, float4 b) { a.m = vsubq_f32(a.m, b.m); return a; }
VM_INLINE float4 operator* (float4 a, float4 b) { a.m = vmulq_f32(a.m, b.m); return a; }
VM_INLINE bool4 operator==(float4 a, float4 b) { a.m = vceqq_f32(a.m, b.m); return a; }
VM_INLINE bool4 operator!=(float4 a, float4 b) { a.m = a.m = vmvnq_u32(vceqq_f32(a.m, b.m)); return a; }
VM_INLINE bool4 operator< (float4 a, float4 b) { a.m = vcltq_f32(a.m, b.m); return a; }
VM_INLINE bool4 operator> (float4 a, float4 b) { a.m = vcgtq_f32(a.m, b.m); return a; }
VM_INLINE bool4 operator<=(float4 a, float4 b) { a.m = vcleq_f32(a.m, b.m); return a; }
VM_INLINE bool4 operator>=(float4 a, float4 b) { a.m = vcgeq_f32(a.m, b.m); return a; }
VM_INLINE bool4 operator&(bool4 a, bool4 b) { a.m = vandq_u32(a.m, b.m); return a; }
VM_INLINE bool4 operator|(bool4 a, bool4 b) { a.m = vorrq_u32(a.m, b.m); return a; }
VM_INLINE float4 operator- (float4 a) { a.m = vnegq_f32(a.m); return a; }
VM_INLINE float4 min(float4 a, float4 b) { a.m = vminq_f32(a.m, b.m); return a; }
VM_INLINE float4 max(float4 a, float4 b) { a.m = vmaxq_f32(a.m, b.m); return a; }
VM_INLINE float hmin(float4 v)
{
float32x2_t minOfHalfs = vpmin_f32(vget_low_f32(v.m), vget_high_f32(v.m));
float32x2_t minOfMinOfHalfs = vpmin_f32(minOfHalfs, minOfHalfs);
return vget_lane_f32(minOfMinOfHalfs, 0);
}
// Returns a 4-bit code where bit0..bit3 is X..W
VM_INLINE unsigned mask(float4 v)
{
static const uint32x4_t movemask = { 1, 2, 4, 8 };
static const uint32x4_t highbit = { 0x80000000, 0x80000000, 0x80000000, 0x80000000 };
uint32x4_t t0 = vreinterpretq_u32_f32(v.m);
uint32x4_t t1 = vtstq_u32(t0, highbit);
uint32x4_t t2 = vandq_u32(t1, movemask);
uint32x2_t t3 = vorr_u32(vget_low_u32(t2), vget_high_u32(t2));
return vget_lane_u32(t3, 0) | vget_lane_u32(t3, 1);
}
// Once we have a comparison, we can branch based on its results:
VM_INLINE bool any(bool4 v) { return mask(v) != 0; }
VM_INLINE bool all(bool4 v) { return mask(v) == 15; }
// "select", i.e. hibit(cond) ? b : a
// on SSE4.1 and up this can be done easily via "blend" instruction;
// on older SSEs has to do a bunch of hoops, see
// https://fgiesen.wordpress.com/2016/04/03/sse-mind-the-gap/
VM_INLINE float4 select(float4 a, float4 b, bool4 cond)
{
a.m = vbslq_f32(cond.m, b.m, a.m);
return a;
}
VM_INLINE int32x4_t select(int32x4_t a, int32x4_t b, bool4 cond)
{
return vbslq_f32(cond.m, b, a);
}
VM_INLINE float4 sqrtf(float4 v)
{
float32x4_t V = v.m;
float32x4_t S0 = vrsqrteq_f32(V);
float32x4_t P0 = vmulq_f32( V, S0 );
float32x4_t R0 = vrsqrtsq_f32( P0, S0 );
float32x4_t S1 = vmulq_f32( S0, R0 );
float32x4_t P1 = vmulq_f32( V, S1 );
float32x4_t R1 = vrsqrtsq_f32( P1, S1 );
float32x4_t S2 = vmulq_f32( S1, R1 );
float32x4_t P2 = vmulq_f32( V, S2 );
float32x4_t R2 = vrsqrtsq_f32( P2, S2 );
float32x4_t S3 = vmulq_f32( S2, R2 );
return float4(vmulq_f32(V, S3));
}
VM_INLINE float4 splatX(float32x4_t v) { return float4(vdupq_lane_f32(vget_low_f32(v), 0)); }
VM_INLINE float4 splatY(float32x4_t v) { return float4(vdupq_lane_f32(vget_low_f32(v), 1)); }
VM_INLINE float4 splatZ(float32x4_t v) { return float4(vdupq_lane_f32(vget_high_f32(v), 0)); }
VM_INLINE float4 splatW(float32x4_t v) { return float4(vdupq_lane_f32(vget_high_f32(v), 1)); }
#endif

View File

@@ -1,203 +0,0 @@
#include "Maths.h"
#include <stdlib.h>
#include <stdint.h>
static uint32_t XorShift32(uint32_t& state)
{
uint32_t x = state;
x ^= x << 13;
x ^= x >> 17;
x ^= x << 15;
state = x;
return x;
}
float RandomFloat01(uint32_t& state)
{
return (XorShift32(state) & 0xFFFFFF) / 16777216.0f;
}
float3 RandomInUnitDisk(uint32_t& state)
{
float3 p;
do
{
p = 2.0 * float3(RandomFloat01(state),RandomFloat01(state),0) - float3(1,1,0);
} while (dot(p,p) >= 1.0);
return p;
}
float3 RandomInUnitSphere(uint32_t& state)
{
float3 p;
do {
p = 2.0*float3(RandomFloat01(state),RandomFloat01(state),RandomFloat01(state)) - float3(1,1,1);
} while (sqLength(p) >= 1.0);
return p;
}
float3 RandomUnitVector(uint32_t& state)
{
float z = RandomFloat01(state) * 2.0f - 1.0f;
float a = RandomFloat01(state) * 2.0f * kPI;
float r = sqrtf(1.0f - z * z);
float x = r * cosf(a);
float y = r * sinf(a);
return float3(x, y, z);
}
int HitSpheres(const Ray& r, const SpheresSoA& spheres, float tMin, float tMax, Hit& outHit)
{
#if DO_HIT_SPHERES_SIMD
float4 hitT = float4(tMax);
#if USE_NEON
int32x4_t id = vdupq_n_s32(-1);
#else
__m128i id = _mm_set1_epi32(-1);
#endif
#if DO_FLOAT3_WITH_SIMD && !USE_NEON
float4 rOrigX = SHUFFLE4(r.orig, 0, 0, 0, 0);
float4 rOrigY = SHUFFLE4(r.orig, 1, 1, 1, 1);
float4 rOrigZ = SHUFFLE4(r.orig, 2, 2, 2, 2);
float4 rDirX = SHUFFLE4(r.dir, 0, 0, 0, 0);
float4 rDirY = SHUFFLE4(r.dir, 1, 1, 1, 1);
float4 rDirZ = SHUFFLE4(r.dir, 2, 2, 2, 2);
#elif DO_FLOAT3_WITH_SIMD
float4 rOrigX = splatX(r.orig.m);
float4 rOrigY = splatY(r.orig.m);
float4 rOrigZ = splatZ(r.orig.m);
float4 rDirX = splatX(r.dir.m);
float4 rDirY = splatY(r.dir.m);
float4 rDirZ = splatZ(r.dir.m);
#else
float4 rOrigX = float4(r.orig.x);
float4 rOrigY = float4(r.orig.y);
float4 rOrigZ = float4(r.orig.z);
float4 rDirX = float4(r.dir.x);
float4 rDirY = float4(r.dir.y);
float4 rDirZ = float4(r.dir.z);
#endif
float4 tMin4 = float4(tMin);
#if USE_NEON
int32x4_t curId = vcombine_u32(vcreate_u32(0ULL | (1ULL<<32)), vcreate_u32(2ULL | (3ULL<<32)));
#else
__m128i curId = _mm_set_epi32(3, 2, 1, 0);
#endif
// process 4 spheres at once
for (int i = 0; i < spheres.simdCount; i += kSimdWidth)
{
// load data for 4 spheres
float4 sCenterX = float4(spheres.centerX + i);
float4 sCenterY = float4(spheres.centerY + i);
float4 sCenterZ = float4(spheres.centerZ + i);
float4 sSqRadius = float4(spheres.sqRadius + i);
// note: we flip this vector and calculate -b (nb) since that happens to be slightly preferable computationally
float4 coX = sCenterX - rOrigX;
float4 coY = sCenterY - rOrigY;
float4 coZ = sCenterZ - rOrigZ;
float4 nb = coX * rDirX + coY * rDirY + coZ * rDirZ;
float4 c = coX * coX + coY * coY + coZ * coZ - sSqRadius;
float4 discr = nb * nb - c;
bool4 discrPos = discr > float4(0.0f);
// if ray hits any of the 4 spheres
if (any(discrPos))
{
float4 discrSq = sqrtf(discr);
// ray could hit spheres at t0 & t1
float4 t0 = nb - discrSq;
float4 t1 = nb + discrSq;
float4 t = select(t1, t0, t0 > tMin4); // if t0 is above min, take it (since it's the earlier hit); else try t1.
bool4 msk = discrPos & (t > tMin4) & (t < hitT);
// if hit, take it
id = select(id, curId, msk);
hitT = select(hitT, t, msk);
}
#if USE_NEON
curId = vaddq_s32(curId, vdupq_n_s32(kSimdWidth));
#else
curId = _mm_add_epi32(curId, _mm_set1_epi32(kSimdWidth));
#endif
}
// now we have up to 4 hits, find and return closest one
float minT = hmin(hitT);
if (minT < tMax) // any actual hits?
{
int minMask = mask(hitT == float4(minT));
if (minMask != 0)
{
int id_scalar[4];
float hitT_scalar[4];
#if USE_NEON
vst1q_s32(id_scalar, id);
vst1q_f32(hitT_scalar, hitT.m);
#else
_mm_storeu_si128((__m128i *)id_scalar, id);
_mm_storeu_ps(hitT_scalar, hitT.m);
#endif
// In general, you would do this with a bit scan (first set/trailing zero count).
// But who cares, it's only 16 options.
static const int laneId[16] =
{
0, 0, 1, 0, // 00xx
2, 0, 1, 0, // 01xx
3, 0, 1, 0, // 10xx
2, 0, 1, 0, // 11xx
};
int lane = laneId[minMask];
int hitId = id_scalar[lane];
float finalHitT = hitT_scalar[lane];
outHit.pos = r.pointAt(finalHitT);
outHit.normal = (outHit.pos - float3(spheres.centerX[hitId], spheres.centerY[hitId], spheres.centerZ[hitId])) * spheres.invRadius[hitId];
outHit.t = finalHitT;
return hitId;
}
}
return -1;
#else // #if DO_HIT_SPHERES_SIMD
float hitT = tMax;
int id = -1;
for (int i = 0; i < spheres.count; ++i)
{
float coX = spheres.centerX[i] - r.orig.getX();
float coY = spheres.centerY[i] - r.orig.getY();
float coZ = spheres.centerZ[i] - r.orig.getZ();
float nb = coX * r.dir.getX() + coY * r.dir.getY() + coZ * r.dir.getZ();
float c = coX * coX + coY * coY + coZ * coZ - spheres.sqRadius[i];
float discr = nb * nb - c;
if (discr > 0)
{
float discrSq = sqrtf(discr);
// Try earlier t
float t = nb - discrSq;
if (t <= tMin) // before min, try later t!
t = nb + discrSq;
if (t > tMin && t < hitT)
{
id = i;
hitT = t;
}
}
}
if (id != -1)
{
outHit.pos = r.pointAt(hitT);
outHit.normal = (outHit.pos - float3(spheres.centerX[id], spheres.centerY[id], spheres.centerZ[id])) * spheres.invRadius[id];
outHit.t = hitT;
return id;
}
else
return -1;
#endif // #else of #if DO_HIT_SPHERES_SIMD
}

View File

@@ -1,436 +0,0 @@
#pragma once
#include <math.h>
#include <assert.h>
#include <stdint.h>
#include "Config.h"
#include "MathSimd.h"
#define kPI 3.1415926f
// SSE/SIMD vector largely based on http://www.codersnotes.com/notes/maths-lib-2016/
#if DO_FLOAT3_WITH_SIMD
#if !defined(__arm__) && !defined(__arm64__)
// ---- SSE implementation
// SHUFFLE3(v, 0,1,2) leaves the vector unchanged (v.xyz).
// SHUFFLE3(v, 0,0,0) splats the X (v.xxx).
#define SHUFFLE3(V, X,Y,Z) float3(_mm_shuffle_ps((V).m, (V).m, _MM_SHUFFLE(Z,Z,Y,X)))
struct float3
{
VM_INLINE float3() {}
VM_INLINE explicit float3(const float *p) { m = _mm_set_ps(p[2], p[2], p[1], p[0]); }
VM_INLINE explicit float3(float x, float y, float z) { m = _mm_set_ps(z, z, y, x); }
VM_INLINE explicit float3(float v) { m = _mm_set1_ps(v); }
VM_INLINE explicit float3(__m128 v) { m = v; }
VM_INLINE float getX() const { return _mm_cvtss_f32(m); }
VM_INLINE float getY() const { return _mm_cvtss_f32(_mm_shuffle_ps(m, m, _MM_SHUFFLE(1, 1, 1, 1))); }
VM_INLINE float getZ() const { return _mm_cvtss_f32(_mm_shuffle_ps(m, m, _MM_SHUFFLE(2, 2, 2, 2))); }
VM_INLINE float3 yzx() const { return SHUFFLE3(*this, 1, 2, 0); }
VM_INLINE float3 zxy() const { return SHUFFLE3(*this, 2, 0, 1); }
VM_INLINE void store(float *p) const { p[0] = getX(); p[1] = getY(); p[2] = getZ(); }
void setX(float x)
{
m = _mm_move_ss(m, _mm_set_ss(x));
}
void setY(float y)
{
__m128 t = _mm_move_ss(m, _mm_set_ss(y));
t = _mm_shuffle_ps(t, t, _MM_SHUFFLE(3, 2, 0, 0));
m = _mm_move_ss(t, m);
}
void setZ(float z)
{
__m128 t = _mm_move_ss(m, _mm_set_ss(z));
t = _mm_shuffle_ps(t, t, _MM_SHUFFLE(3, 0, 1, 0));
m = _mm_move_ss(t, m);
}
__m128 m;
};
typedef float3 bool3;
VM_INLINE float3 operator+ (float3 a, float3 b) { a.m = _mm_add_ps(a.m, b.m); return a; }
VM_INLINE float3 operator- (float3 a, float3 b) { a.m = _mm_sub_ps(a.m, b.m); return a; }
VM_INLINE float3 operator* (float3 a, float3 b) { a.m = _mm_mul_ps(a.m, b.m); return a; }
VM_INLINE float3 operator/ (float3 a, float3 b) { a.m = _mm_div_ps(a.m, b.m); return a; }
VM_INLINE float3 operator* (float3 a, float b) { a.m = _mm_mul_ps(a.m, _mm_set1_ps(b)); return a; }
VM_INLINE float3 operator/ (float3 a, float b) { a.m = _mm_div_ps(a.m, _mm_set1_ps(b)); return a; }
VM_INLINE float3 operator* (float a, float3 b) { b.m = _mm_mul_ps(_mm_set1_ps(a), b.m); return b; }
VM_INLINE float3 operator/ (float a, float3 b) { b.m = _mm_div_ps(_mm_set1_ps(a), b.m); return b; }
VM_INLINE float3& operator+= (float3 &a, float3 b) { a = a + b; return a; }
VM_INLINE float3& operator-= (float3 &a, float3 b) { a = a - b; return a; }
VM_INLINE float3& operator*= (float3 &a, float3 b) { a = a * b; return a; }
VM_INLINE float3& operator/= (float3 &a, float3 b) { a = a / b; return a; }
VM_INLINE float3& operator*= (float3 &a, float b) { a = a * b; return a; }
VM_INLINE float3& operator/= (float3 &a, float b) { a = a / b; return a; }
VM_INLINE bool3 operator==(float3 a, float3 b) { a.m = _mm_cmpeq_ps(a.m, b.m); return a; }
VM_INLINE bool3 operator!=(float3 a, float3 b) { a.m = _mm_cmpneq_ps(a.m, b.m); return a; }
VM_INLINE bool3 operator< (float3 a, float3 b) { a.m = _mm_cmplt_ps(a.m, b.m); return a; }
VM_INLINE bool3 operator> (float3 a, float3 b) { a.m = _mm_cmpgt_ps(a.m, b.m); return a; }
VM_INLINE bool3 operator<=(float3 a, float3 b) { a.m = _mm_cmple_ps(a.m, b.m); return a; }
VM_INLINE bool3 operator>=(float3 a, float3 b) { a.m = _mm_cmpge_ps(a.m, b.m); return a; }
VM_INLINE float3 min(float3 a, float3 b) { a.m = _mm_min_ps(a.m, b.m); return a; }
VM_INLINE float3 max(float3 a, float3 b) { a.m = _mm_max_ps(a.m, b.m); return a; }
VM_INLINE float3 operator- (float3 a) { return float3(_mm_setzero_ps()) - a; }
VM_INLINE float hmin(float3 v)
{
v = min(v, SHUFFLE3(v, 1, 0, 2));
return min(v, SHUFFLE3(v, 2, 0, 1)).getX();
}
VM_INLINE float hmax(float3 v)
{
v = max(v, SHUFFLE3(v, 1, 0, 2));
return max(v, SHUFFLE3(v, 2, 0, 1)).getX();
}
VM_INLINE float3 cross(float3 a, float3 b)
{
// x <- a.y*b.z - a.z*b.y
// y <- a.z*b.x - a.x*b.z
// z <- a.x*b.y - a.y*b.x
// We can save a shuffle by grouping it in this wacky order:
return (a.zxy()*b - a*b.zxy()).zxy();
}
// Returns a 3-bit code where bit0..bit2 is X..Z
VM_INLINE unsigned mask(float3 v) { return _mm_movemask_ps(v.m) & 7; }
// Once we have a comparison, we can branch based on its results:
VM_INLINE bool any(bool3 v) { return mask(v) != 0; }
VM_INLINE bool all(bool3 v) { return mask(v) == 7; }
VM_INLINE float3 clamp(float3 t, float3 a, float3 b) { return min(max(t, a), b); }
VM_INLINE float sum(float3 v) { return v.getX() + v.getY() + v.getZ(); }
VM_INLINE float dot(float3 a, float3 b) { return sum(a*b); }
#else // #if !defined(__arm__) && !defined(__arm64__)
// ---- NEON implementation
#include <arm_neon.h>
struct float3
{
VM_INLINE float3() {}
VM_INLINE explicit float3(const float *p) { float v[4] = {p[0], p[1], p[2], 0}; m = vld1q_f32(v); }
VM_INLINE explicit float3(float x, float y, float z) { float v[4] = {x, y, z, 0}; m = vld1q_f32(v); }
VM_INLINE explicit float3(float v) { m = vdupq_n_f32(v); }
VM_INLINE explicit float3(float32x4_t v) { m = v; }
VM_INLINE float getX() const { return vgetq_lane_f32(m, 0); }
VM_INLINE float getY() const { return vgetq_lane_f32(m, 1); }
VM_INLINE float getZ() const { return vgetq_lane_f32(m, 2); }
VM_INLINE float3 yzx() const
{
float32x2_t low = vget_low_f32(m);
float32x4_t yzx = vcombine_f32(vext_f32(low, vget_high_f32(m), 1), low);
return float3(yzx);
}
VM_INLINE float3 zxy() const
{
float32x4_t p = m;
p = vuzpq_f32(vreinterpretq_f32_s32(vextq_s32(vreinterpretq_s32_f32(p), vreinterpretq_s32_f32(p), 1)), p).val[1];
return float3(p);
}
VM_INLINE void store(float *p) const { p[0] = getX(); p[1] = getY(); p[2] = getZ(); }
void setX(float x)
{
m = vsetq_lane_f32(x, m, 0);
}
void setY(float y)
{
m = vsetq_lane_f32(y, m, 1);
}
void setZ(float z)
{
m = vsetq_lane_f32(z, m, 2);
}
float32x4_t m;
};
typedef float3 bool3;
VM_INLINE float32x4_t rcp_2(float32x4_t v)
{
float32x4_t e = vrecpeq_f32(v);
e = vmulq_f32(vrecpsq_f32(e, v), e);
e = vmulq_f32(vrecpsq_f32(e, v), e);
return e;
}
VM_INLINE float3 operator+ (float3 a, float3 b) { a.m = vaddq_f32(a.m, b.m); return a; }
VM_INLINE float3 operator- (float3 a, float3 b) { a.m = vsubq_f32(a.m, b.m); return a; }
VM_INLINE float3 operator* (float3 a, float3 b) { a.m = vmulq_f32(a.m, b.m); return a; }
VM_INLINE float3 operator/ (float3 a, float3 b) { float32x4_t recip = rcp_2(b.m); a.m = vmulq_f32(a.m, recip); return a; }
VM_INLINE float3 operator* (float3 a, float b) { a.m = vmulq_f32(a.m, vdupq_n_f32(b)); return a; }
VM_INLINE float3 operator/ (float3 a, float b) { float32x4_t recip = rcp_2(vdupq_n_f32(b)); a.m = vmulq_f32(a.m, recip); return a; }
VM_INLINE float3 operator* (float a, float3 b) { b.m = vmulq_f32(vdupq_n_f32(a), b.m); return b; }
VM_INLINE float3 operator/ (float a, float3 b) { float32x4_t recip = rcp_2(b.m); b.m = vmulq_f32(vdupq_n_f32(a), recip); return b; }
VM_INLINE float3& operator+= (float3 &a, float3 b) { a = a + b; return a; }
VM_INLINE float3& operator-= (float3 &a, float3 b) { a = a - b; return a; }
VM_INLINE float3& operator*= (float3 &a, float3 b) { a = a * b; return a; }
VM_INLINE float3& operator/= (float3 &a, float3 b) { a = a / b; return a; }
VM_INLINE float3& operator*= (float3 &a, float b) { a = a * b; return a; }
VM_INLINE float3& operator/= (float3 &a, float b) { a = a / b; return a; }
VM_INLINE bool3 operator==(float3 a, float3 b) { a.m = vceqq_f32(a.m, b.m); return a; }
VM_INLINE bool3 operator!=(float3 a, float3 b) { a.m = vmvnq_u32(vceqq_f32(a.m, b.m)); return a; }
VM_INLINE bool3 operator< (float3 a, float3 b) { a.m = vcltq_f32(a.m, b.m); return a; }
VM_INLINE bool3 operator> (float3 a, float3 b) { a.m = vcgtq_f32(a.m, b.m); return a; }
VM_INLINE bool3 operator<=(float3 a, float3 b) { a.m = vcleq_f32(a.m, b.m); return a; }
VM_INLINE bool3 operator>=(float3 a, float3 b) { a.m = vcgeq_f32(a.m, b.m); return a; }
VM_INLINE float3 min(float3 a, float3 b) { a.m = vminq_f32(a.m, b.m); return a; }
VM_INLINE float3 max(float3 a, float3 b) { a.m = vmaxq_f32(a.m, b.m); return a; }
VM_INLINE float3 operator- (float3 a) { a.m = vnegq_f32(a.m); return a; }
VM_INLINE float hmin(float3 v)
{
float32x2_t minOfHalfs = vpmin_f32(vget_low_f32(v.m), vget_high_f32(v.m));
float32x2_t minOfMinOfHalfs = vpmin_f32(minOfHalfs, minOfHalfs);
return vget_lane_f32(minOfMinOfHalfs, 0);
}
VM_INLINE float hmax(float3 v)
{
float32x2_t maxOfHalfs = vpmax_f32(vget_low_f32(v.m), vget_high_f32(v.m));
float32x2_t maxOfMaxOfHalfs = vpmax_f32(maxOfHalfs, maxOfHalfs);
return vget_lane_f32(maxOfMaxOfHalfs, 0);
}
VM_INLINE float3 cross(float3 a, float3 b)
{
// x <- a.y*b.z - a.z*b.y
// y <- a.z*b.x - a.x*b.z
// z <- a.x*b.y - a.y*b.x
// We can save a shuffle by grouping it in this wacky order:
return (a.zxy()*b - a*b.zxy()).zxy();
}
// Returns a 3-bit code where bit0..bit2 is X..Z
VM_INLINE unsigned mask(float3 v)
{
static const uint32x4_t movemask = { 1, 2, 4, 8 };
static const uint32x4_t highbit = { 0x80000000, 0x80000000, 0x80000000, 0x80000000 };
uint32x4_t t0 = vreinterpretq_u32_f32(v.m);
uint32x4_t t1 = vtstq_u32(t0, highbit);
uint32x4_t t2 = vandq_u32(t1, movemask);
uint32x2_t t3 = vorr_u32(vget_low_u32(t2), vget_high_u32(t2));
return vget_lane_u32(t3, 0) | vget_lane_u32(t3, 1);
}
// Once we have a comparison, we can branch based on its results:
VM_INLINE bool any(bool3 v) { return mask(v) != 0; }
VM_INLINE bool all(bool3 v) { return mask(v) == 7; }
VM_INLINE float3 clamp(float3 t, float3 a, float3 b) { return min(max(t, a), b); }
VM_INLINE float sum(float3 v) { return v.getX() + v.getY() + v.getZ(); }
VM_INLINE float dot(float3 a, float3 b) { return sum(a*b); }
#endif // #else of #if !defined(__arm__) && !defined(__arm64__)
#else // #if DO_FLOAT3_WITH_SIMD
// ---- Simple scalar C implementation
struct float3
{
float3() : x(0), y(0), z(0) {}
float3(float x_, float y_, float z_) : x(x_), y(y_), z(z_) {}
float3 operator-() const { return float3(-x, -y, -z); }
float3& operator+=(const float3& o) { x+=o.x; y+=o.y; z+=o.z; return *this; }
float3& operator-=(const float3& o) { x-=o.x; y-=o.y; z-=o.z; return *this; }
float3& operator*=(const float3& o) { x*=o.x; y*=o.y; z*=o.z; return *this; }
float3& operator*=(float o) { x*=o; y*=o; z*=o; return *this; }
VM_INLINE float getX() const { return x; }
VM_INLINE float getY() const { return y; }
VM_INLINE float getZ() const { return z; }
VM_INLINE void setX(float x_) { x = x_; }
VM_INLINE void setY(float y_) { y = y_; }
VM_INLINE void setZ(float z_) { z = z_; }
VM_INLINE void store(float *p) const { p[0] = getX(); p[1] = getY(); p[2] = getZ(); }
float x, y, z;
};
VM_INLINE float3 operator+(const float3& a, const float3& b) { return float3(a.x+b.x,a.y+b.y,a.z+b.z); }
VM_INLINE float3 operator-(const float3& a, const float3& b) { return float3(a.x-b.x,a.y-b.y,a.z-b.z); }
VM_INLINE float3 operator*(const float3& a, const float3& b) { return float3(a.x*b.x,a.y*b.y,a.z*b.z); }
VM_INLINE float3 operator*(const float3& a, float b) { return float3(a.x*b,a.y*b,a.z*b); }
VM_INLINE float3 operator*(float a, const float3& b) { return float3(a*b.x,a*b.y,a*b.z); }
VM_INLINE float dot(const float3& a, const float3& b) { return a.x*b.x+a.y*b.y+a.z*b.z; }
VM_INLINE float3 cross(const float3& a, const float3& b)
{
return float3(
a.y*b.z - a.z*b.y,
-(a.x*b.z - a.z*b.x),
a.x*b.y - a.y*b.x
);
}
#endif // #else of #if DO_FLOAT3_WITH_SIMD
VM_INLINE float length(float3 v) { return sqrtf(dot(v, v)); }
VM_INLINE float sqLength(float3 v) { return dot(v, v); }
VM_INLINE float3 normalize(float3 v) { return v * (1.0f / length(v)); }
VM_INLINE float3 lerp(float3 a, float3 b, float t) { return a + (b-a)*t; }
inline void AssertUnit(float3 v)
{
assert(fabsf(sqLength(v) - 1.0f) < 0.01f);
}
inline float3 reflect(float3 v, float3 n)
{
return v - 2*dot(v,n)*n;
}
inline bool refract(float3 v, float3 n, float nint, float3& outRefracted)
{
AssertUnit(v);
float dt = dot(v, n);
float discr = 1.0f - nint*nint*(1-dt*dt);
if (discr > 0)
{
outRefracted = nint * (v - n*dt) - n*sqrtf(discr);
return true;
}
return false;
}
inline float schlick(float cosine, float ri)
{
float r0 = (1-ri) / (1+ri);
r0 = r0*r0;
return r0 + (1-r0)*powf(1-cosine, 5);
}
struct Ray
{
Ray() {}
Ray(float3 orig_, float3 dir_) : orig(orig_), dir(dir_) { AssertUnit(dir); }
float3 pointAt(float t) const { return orig + dir * t; }
float3 orig;
float3 dir;
};
struct Hit
{
float3 pos;
float3 normal;
float t;
};
struct Sphere
{
Sphere() : radius(1.0f), invRadius(0.0f) {}
Sphere(float3 center_, float radius_) : center(center_), radius(radius_), invRadius(0.0f) {}
void UpdateDerivedData() { invRadius = 1.0f/radius; }
float3 center;
float radius;
float invRadius;
};
// data for all spheres in a "structure of arrays" layout
struct SpheresSoA
{
SpheresSoA(int c)
{
count = c;
// we'll be processing spheres in kSimdWidth chunks, so make sure to allocate
// enough space
simdCount = (c + (kSimdWidth - 1)) / kSimdWidth * kSimdWidth;
centerX = new float[simdCount];
centerY = new float[simdCount];
centerZ = new float[simdCount];
sqRadius = new float[simdCount];
invRadius = new float[simdCount];
// set all data to "impossible sphere" state
for (int i = count; i < simdCount; ++i)
{
centerX[i] = centerY[i] = centerZ[i] = 10000.0f;
sqRadius[i] = 0.0f;
invRadius[i] = 0.0f;
}
}
~SpheresSoA()
{
delete[] centerX;
delete[] centerY;
delete[] centerZ;
delete[] sqRadius;
delete[] invRadius;
}
float* centerX;
float* centerY;
float* centerZ;
float* sqRadius;
float* invRadius;
int simdCount;
int count;
};
int HitSpheres(const Ray& r, const SpheresSoA& spheres, float tMin, float tMax, Hit& outHit);
float RandomFloat01(uint32_t& state);
float3 RandomInUnitDisk(uint32_t& state);
float3 RandomInUnitSphere(uint32_t& state);
float3 RandomUnitVector(uint32_t& state);
struct Camera
{
Camera() {}
// vfov is top to bottom in degrees
Camera(const float3& lookFrom, const float3& lookAt, const float3& vup, float vfov, float aspect, float aperture, float focusDist)
{
lensRadius = aperture / 2;
float theta = vfov*kPI/180;
float halfHeight = tanf(theta/2);
float halfWidth = aspect * halfHeight;
origin = lookFrom;
w = normalize(lookFrom - lookAt);
u = normalize(cross(vup, w));
v = cross(w, u);
lowerLeftCorner = origin - halfWidth*focusDist*u - halfHeight*focusDist*v - focusDist*w;
horizontal = 2*halfWidth*focusDist*u;
vertical = 2*halfHeight*focusDist*v;
}
Ray GetRay(float s, float t, uint32_t& state) const
{
float3 rd = lensRadius * RandomInUnitDisk(state);
float3 offset = u * rd.getX() + v * rd.getY();
return Ray(origin + offset, normalize(lowerLeftCorner + s*horizontal + t*vertical - origin - offset));
}
float3 origin;
float3 lowerLeftCorner;
float3 horizontal;
float3 vertical;
float3 u, v, w;
float lensRadius;
};

View File

@@ -1,392 +0,0 @@
#include "Config.h"
#include "Test.h"
#include "Maths.h"
#include <algorithm>
#if CPU_CAN_DO_THREADS
#include "enkiTS/TaskScheduler_c.h"
#include <thread>
#endif
#include <atomic>
#include "../../../public/tracy/Tracy.hpp"
// 46 spheres (2 emissive) when enabled; 9 spheres (1 emissive) when disabled
#define DO_BIG_SCENE 1
static Sphere s_Spheres[] =
{
{float3(0,-100.5,-1), 100},
{float3(2,0,-1), 0.5f},
{float3(0,0,-1), 0.5f},
{float3(-2,0,-1), 0.5f},
{float3(2,0,1), 0.5f},
{float3(0,0,1), 0.5f},
{float3(-2,0,1), 0.5f},
{float3(0.5f,1,0.5f), 0.5f},
{float3(-1.5f,1.5f,0.f), 0.3f},
#if DO_BIG_SCENE
{float3(4,0,-3), 0.5f}, {float3(3,0,-3), 0.5f}, {float3(2,0,-3), 0.5f}, {float3(1,0,-3), 0.5f}, {float3(0,0,-3), 0.5f}, {float3(-1,0,-3), 0.5f}, {float3(-2,0,-3), 0.5f}, {float3(-3,0,-3), 0.5f}, {float3(-4,0,-3), 0.5f},
{float3(4,0,-4), 0.5f}, {float3(3,0,-4), 0.5f}, {float3(2,0,-4), 0.5f}, {float3(1,0,-4), 0.5f}, {float3(0,0,-4), 0.5f}, {float3(-1,0,-4), 0.5f}, {float3(-2,0,-4), 0.5f}, {float3(-3,0,-4), 0.5f}, {float3(-4,0,-4), 0.5f},
{float3(4,0,-5), 0.5f}, {float3(3,0,-5), 0.5f}, {float3(2,0,-5), 0.5f}, {float3(1,0,-5), 0.5f}, {float3(0,0,-5), 0.5f}, {float3(-1,0,-5), 0.5f}, {float3(-2,0,-5), 0.5f}, {float3(-3,0,-5), 0.5f}, {float3(-4,0,-5), 0.5f},
{float3(4,0,-6), 0.5f}, {float3(3,0,-6), 0.5f}, {float3(2,0,-6), 0.5f}, {float3(1,0,-6), 0.5f}, {float3(0,0,-6), 0.5f}, {float3(-1,0,-6), 0.5f}, {float3(-2,0,-6), 0.5f}, {float3(-3,0,-6), 0.5f}, {float3(-4,0,-6), 0.5f},
{float3(1.5f,1.5f,-2), 0.3f},
#endif // #if DO_BIG_SCENE
};
const int kSphereCount = sizeof(s_Spheres) / sizeof(s_Spheres[0]);
static SpheresSoA s_SpheresSoA(kSphereCount);
struct Material
{
enum Type { Lambert, Metal, Dielectric };
Type type;
float3 albedo;
float3 emissive;
float roughness;
float ri;
};
static Material s_SphereMats[kSphereCount] =
{
{ Material::Lambert, float3(0.8f, 0.8f, 0.8f), float3(0,0,0), 0, 0, },
{ Material::Lambert, float3(0.8f, 0.4f, 0.4f), float3(0,0,0), 0, 0, },
{ Material::Lambert, float3(0.4f, 0.8f, 0.4f), float3(0,0,0), 0, 0, },
{ Material::Metal, float3(0.4f, 0.4f, 0.8f), float3(0,0,0), 0, 0 },
{ Material::Metal, float3(0.4f, 0.8f, 0.4f), float3(0,0,0), 0, 0 },
{ Material::Metal, float3(0.4f, 0.8f, 0.4f), float3(0,0,0), 0.2f, 0 },
{ Material::Metal, float3(0.4f, 0.8f, 0.4f), float3(0,0,0), 0.6f, 0 },
{ Material::Dielectric, float3(0.4f, 0.4f, 0.4f), float3(0,0,0), 0, 1.5f },
{ Material::Lambert, float3(0.8f, 0.6f, 0.2f), float3(30,25,15), 0, 0 },
#if DO_BIG_SCENE
{ Material::Lambert, float3(0.1f, 0.1f, 0.1f), float3(0,0,0), 0, 0, }, { Material::Lambert, float3(0.2f, 0.2f, 0.2f), float3(0,0,0), 0, 0, }, { Material::Lambert, float3(0.3f, 0.3f, 0.3f), float3(0,0,0), 0, 0, }, { Material::Lambert, float3(0.4f, 0.4f, 0.4f), float3(0,0,0), 0, 0, }, { Material::Lambert, float3(0.5f, 0.5f, 0.5f), float3(0,0,0), 0, 0, }, { Material::Lambert, float3(0.6f, 0.6f, 0.6f), float3(0,0,0), 0, 0, }, { Material::Lambert, float3(0.7f, 0.7f, 0.7f), float3(0,0,0), 0, 0, }, { Material::Lambert, float3(0.8f, 0.8f, 0.8f), float3(0,0,0), 0, 0, }, { Material::Lambert, float3(0.9f, 0.9f, 0.9f), float3(0,0,0), 0, 0, },
{ Material::Metal, float3(0.1f, 0.1f, 0.1f), float3(0,0,0), 0, 0, }, { Material::Metal, float3(0.2f, 0.2f, 0.2f), float3(0,0,0), 0, 0, }, { Material::Metal, float3(0.3f, 0.3f, 0.3f), float3(0,0,0), 0, 0, }, { Material::Metal, float3(0.4f, 0.4f, 0.4f), float3(0,0,0), 0, 0, }, { Material::Metal, float3(0.5f, 0.5f, 0.5f), float3(0,0,0), 0, 0, }, { Material::Metal, float3(0.6f, 0.6f, 0.6f), float3(0,0,0), 0, 0, }, { Material::Metal, float3(0.7f, 0.7f, 0.7f), float3(0,0,0), 0, 0, }, { Material::Metal, float3(0.8f, 0.8f, 0.8f), float3(0,0,0), 0, 0, }, { Material::Metal, float3(0.9f, 0.9f, 0.9f), float3(0,0,0), 0, 0, },
{ Material::Metal, float3(0.8f, 0.1f, 0.1f), float3(0,0,0), 0, 0, }, { Material::Metal, float3(0.8f, 0.5f, 0.1f), float3(0,0,0), 0, 0, }, { Material::Metal, float3(0.8f, 0.8f, 0.1f), float3(0,0,0), 0, 0, }, { Material::Metal, float3(0.4f, 0.8f, 0.1f), float3(0,0,0), 0, 0, }, { Material::Metal, float3(0.1f, 0.8f, 0.1f), float3(0,0,0), 0, 0, }, { Material::Metal, float3(0.1f, 0.8f, 0.5f), float3(0,0,0), 0, 0, }, { Material::Metal, float3(0.1f, 0.8f, 0.8f), float3(0,0,0), 0, 0, }, { Material::Metal, float3(0.1f, 0.1f, 0.8f), float3(0,0,0), 0, 0, }, { Material::Metal, float3(0.5f, 0.1f, 0.8f), float3(0,0,0), 0, 0, },
{ Material::Lambert, float3(0.8f, 0.1f, 0.1f), float3(0,0,0), 0, 0, }, { Material::Lambert, float3(0.8f, 0.5f, 0.1f), float3(0,0,0), 0, 0, }, { Material::Lambert, float3(0.8f, 0.8f, 0.1f), float3(0,0,0), 0, 0, }, { Material::Lambert, float3(0.4f, 0.8f, 0.1f), float3(0,0,0), 0, 0, }, { Material::Lambert, float3(0.1f, 0.8f, 0.1f), float3(0,0,0), 0, 0, }, { Material::Lambert, float3(0.1f, 0.8f, 0.5f), float3(0,0,0), 0, 0, }, { Material::Lambert, float3(0.1f, 0.8f, 0.8f), float3(0,0,0), 0, 0, }, { Material::Lambert, float3(0.1f, 0.1f, 0.8f), float3(0,0,0), 0, 0, }, { Material::Metal, float3(0.5f, 0.1f, 0.8f), float3(0,0,0), 0, 0, },
{ Material::Lambert, float3(0.1f, 0.2f, 0.5f), float3(3,10,20), 0, 0 },
#endif
};
static int s_EmissiveSpheres[kSphereCount];
static int s_EmissiveSphereCount;
static Camera s_Cam;
const float kMinT = 0.001f;
const float kMaxT = 1.0e7f;
const int kMaxDepth = 10;
bool HitWorld(const Ray& r, float tMin, float tMax, Hit& outHit, int& outID)
{
outID = HitSpheres(r, s_SpheresSoA, tMin, tMax, outHit);
return outID != -1;
}
static bool Scatter(const Material& mat, const Ray& r_in, const Hit& rec, float3& attenuation, Ray& scattered, float3& outLightE, int& inoutRayCount, uint32_t& state)
{
ZoneScoped;
outLightE = float3(0,0,0);
if (mat.type == Material::Lambert)
{
// random point on unit sphere that is tangent to the hit point
float3 target = rec.pos + rec.normal + RandomUnitVector(state);
scattered = Ray(rec.pos, normalize(target - rec.pos));
attenuation = mat.albedo;
// sample lights
#if DO_LIGHT_SAMPLING
for (int j = 0; j < s_EmissiveSphereCount; ++j)
{
int i = s_EmissiveSpheres[j];
const Material& smat = s_SphereMats[i];
if (&mat == &smat)
continue; // skip self
const Sphere& s = s_Spheres[i];
// create a random direction towards sphere
// coord system for sampling: sw, su, sv
float3 sw = normalize(s.center - rec.pos);
float3 su = normalize(cross(fabs(sw.getX())>0.01f ? float3(0,1,0):float3(1,0,0), sw));
float3 sv = cross(sw, su);
// sample sphere by solid angle
float cosAMax = sqrtf(1.0f - s.radius*s.radius / sqLength(rec.pos-s.center));
float eps1 = RandomFloat01(state), eps2 = RandomFloat01(state);
float cosA = 1.0f - eps1 + eps1 * cosAMax;
float sinA = sqrtf(1.0f - cosA*cosA);
float phi = 2 * kPI * eps2;
float3 l = su * (cosf(phi) * sinA) + sv * (sinf(phi) * sinA) + sw * cosA;
//l = normalize(l); // NOTE(fg): This is already normalized, by construction.
// shoot shadow ray
Hit lightHit;
int hitID;
++inoutRayCount;
if (HitWorld(Ray(rec.pos, l), kMinT, kMaxT, lightHit, hitID) && hitID == i)
{
float omega = 2 * kPI * (1-cosAMax);
float3 rdir = r_in.dir;
AssertUnit(rdir);
float3 nl = dot(rec.normal, rdir) < 0 ? rec.normal : -rec.normal;
outLightE += (mat.albedo * smat.emissive) * (std::max(0.0f, dot(l, nl)) * omega / kPI);
}
}
#endif
return true;
}
else if (mat.type == Material::Metal)
{
AssertUnit(r_in.dir); AssertUnit(rec.normal);
float3 refl = reflect(r_in.dir, rec.normal);
// reflected ray, and random inside of sphere based on roughness
float roughness = mat.roughness;
#if DO_MITSUBA_COMPARE
roughness = 0; // until we get better BRDF for metals
#endif
scattered = Ray(rec.pos, normalize(refl + roughness*RandomInUnitSphere(state)));
attenuation = mat.albedo;
return dot(scattered.dir, rec.normal) > 0;
}
else if (mat.type == Material::Dielectric)
{
AssertUnit(r_in.dir); AssertUnit(rec.normal);
float3 outwardN;
float3 rdir = r_in.dir;
float3 refl = reflect(rdir, rec.normal);
float nint;
attenuation = float3(1,1,1);
float3 refr;
float reflProb;
float cosine;
if (dot(rdir, rec.normal) > 0)
{
outwardN = -rec.normal;
nint = mat.ri;
cosine = mat.ri * dot(rdir, rec.normal);
}
else
{
outwardN = rec.normal;
nint = 1.0f / mat.ri;
cosine = -dot(rdir, rec.normal);
}
if (refract(rdir, outwardN, nint, refr))
{
reflProb = schlick(cosine, mat.ri);
}
else
{
reflProb = 1;
}
if (RandomFloat01(state) < reflProb)
scattered = Ray(rec.pos, normalize(refl));
else
scattered = Ray(rec.pos, normalize(refr));
}
else
{
attenuation = float3(1,0,1);
return false;
}
return true;
}
static float3 Trace(const Ray& r, int depth, int& inoutRayCount, uint32_t& state, bool doMaterialE = true)
{
ZoneScoped;
Hit rec;
int id = 0;
++inoutRayCount;
if (HitWorld(r, kMinT, kMaxT, rec, id))
{
Ray scattered;
float3 attenuation;
float3 lightE;
const Material& mat = s_SphereMats[id];
float3 matE = mat.emissive;
if (depth < kMaxDepth && Scatter(mat, r, rec, attenuation, scattered, lightE, inoutRayCount, state))
{
#if DO_LIGHT_SAMPLING
if (!doMaterialE) matE = float3(0,0,0); // don't add material emission if told so
// dor Lambert materials, we just did explicit light (emissive) sampling and already
// for their contribution, so if next ray bounce hits the light again, don't add
// emission
doMaterialE = (mat.type != Material::Lambert);
#endif
return matE + lightE + attenuation * Trace(scattered, depth+1, inoutRayCount, state, doMaterialE);
}
else
{
return matE;
}
}
else
{
// sky
#if DO_MITSUBA_COMPARE
return float3(0.15f,0.21f,0.3f); // easier compare with Mitsuba's constant environment light
#else
float3 unitDir = r.dir;
float t = 0.5f*(unitDir.getY() + 1.0f);
return ((1.0f-t)*float3(1.0f, 1.0f, 1.0f) + t*float3(0.5f, 0.7f, 1.0f)) * 0.3f;
#endif
}
}
#if CPU_CAN_DO_THREADS
static enkiTaskScheduler* g_TS;
#endif
void InitializeTest()
{
ZoneScoped;
#if CPU_CAN_DO_THREADS
g_TS = enkiNewTaskScheduler();
enkiInitTaskSchedulerNumThreads(g_TS, std::max<int>( 2, std::thread::hardware_concurrency() - 2));
#endif
}
void ShutdownTest()
{
ZoneScoped;
#if CPU_CAN_DO_THREADS
enkiDeleteTaskScheduler(g_TS);
#endif
}
struct JobData
{
float time;
int frameCount;
int screenWidth, screenHeight;
float* backbuffer;
Camera* cam;
std::atomic<int> rayCount;
unsigned testFlags;
};
static void TraceRowJob(uint32_t start, uint32_t end, uint32_t threadnum, void* data_)
{
ZoneScoped;
JobData& data = *(JobData*)data_;
float* backbuffer = data.backbuffer + start * data.screenWidth * 4;
float invWidth = 1.0f / data.screenWidth;
float invHeight = 1.0f / data.screenHeight;
float lerpFac = float(data.frameCount) / float(data.frameCount+1);
if (data.testFlags & kFlagAnimate)
lerpFac *= DO_ANIMATE_SMOOTHING;
if (!(data.testFlags & kFlagProgressive))
lerpFac = 0;
int rayCount = 0;
for (uint32_t y = start; y < end; ++y)
{
uint32_t state = (y * 9781 + data.frameCount * 6271) | 1;
for (int x = 0; x < data.screenWidth; ++x)
{
float3 col(0, 0, 0);
for (int s = 0; s < DO_SAMPLES_PER_PIXEL; s++)
{
float u = float(x + RandomFloat01(state)) * invWidth;
float v = float(y + RandomFloat01(state)) * invHeight;
Ray r = data.cam->GetRay(u, v, state);
col += Trace(r, 0, rayCount, state);
}
col *= 1.0f / float(DO_SAMPLES_PER_PIXEL);
float3 prev(backbuffer[0], backbuffer[1], backbuffer[2]);
col = prev * lerpFac + col * (1-lerpFac);
col.store(backbuffer);
backbuffer += 4;
}
}
data.rayCount += rayCount;
}
void UpdateTest(float time, int frameCount, int screenWidth, int screenHeight, unsigned testFlags)
{
ZoneScoped;
if (testFlags & kFlagAnimate)
{
s_Spheres[1].center.setY(cosf(time) + 1.0f);
s_Spheres[8].center.setZ(sinf(time)*0.3f);
}
float3 lookfrom(0, 2, 3);
float3 lookat(0, 0, 0);
float distToFocus = 3;
#if DO_MITSUBA_COMPARE
float aperture = 0.0f;
#else
float aperture = 0.1f;
#endif
#if DO_BIG_SCENE
aperture *= 0.2f;
#endif
s_EmissiveSphereCount = 0;
for (int i = 0; i < kSphereCount; ++i)
{
Sphere& s = s_Spheres[i];
s.UpdateDerivedData();
s_SpheresSoA.centerX[i] = s.center.getX();
s_SpheresSoA.centerY[i] = s.center.getY();
s_SpheresSoA.centerZ[i] = s.center.getZ();
s_SpheresSoA.sqRadius[i] = s.radius * s.radius;
s_SpheresSoA.invRadius[i] = s.invRadius;
// Remember IDs of emissive spheres (light sources)
const Material& smat = s_SphereMats[i];
if (smat.emissive.getX() > 0 || smat.emissive.getY() > 0 || smat.emissive.getZ() > 0)
{
s_EmissiveSpheres[s_EmissiveSphereCount] = i;
s_EmissiveSphereCount++;
}
}
s_Cam = Camera(lookfrom, lookat, float3(0, 1, 0), 60, float(screenWidth) / float(screenHeight), aperture, distToFocus);
}
void DrawTest(float time, int frameCount, int screenWidth, int screenHeight, float* backbuffer, int& outRayCount, unsigned testFlags)
{
ZoneScoped;
JobData args;
args.time = time;
args.frameCount = frameCount;
args.screenWidth = screenWidth;
args.screenHeight = screenHeight;
args.backbuffer = backbuffer;
args.cam = &s_Cam;
args.testFlags = testFlags;
args.rayCount = 0;
#if CPU_CAN_DO_THREADS
enkiTaskSet* task = enkiCreateTaskSet(g_TS, TraceRowJob);
bool threaded = true;
enkiAddTaskSetToPipeMinRange(g_TS, task, &args, screenHeight, threaded ? 4 : screenHeight);
enkiWaitForTaskSet(g_TS, task);
enkiDeleteTaskSet(task);
#else
TraceRowJob(0, screenHeight, 0, &args);
#endif
outRayCount = args.rayCount;
}
void GetObjectCount(int& outCount, int& outObjectSize, int& outMaterialSize, int& outCamSize)
{
ZoneScoped;
outCount = kSphereCount;
outObjectSize = sizeof(Sphere);
outMaterialSize = sizeof(Material);
outCamSize = sizeof(Camera);
}
void GetSceneDesc(void* outObjects, void* outMaterials, void* outCam, void* outEmissives, int* outEmissiveCount)
{
ZoneScoped;
memcpy(outObjects, s_Spheres, kSphereCount * sizeof(s_Spheres[0]));
memcpy(outMaterials, s_SphereMats, kSphereCount * sizeof(s_SphereMats[0]));
memcpy(outCam, &s_Cam, sizeof(s_Cam));
memcpy(outEmissives, s_EmissiveSpheres, s_EmissiveSphereCount * sizeof(s_EmissiveSpheres[0]));
*outEmissiveCount = s_EmissiveSphereCount;
}

Some files were not shown because too many files have changed in this diff Show More