settings.yml
The input settings
for packages in Conan are predefined in ~/.conan/settings.yml
file, so only a few like os
or compiler
are possible. These are the default values, but it is possible to customize them, see Customizing settings.
# Only for cross building, 'os_build/arch_build' is the system that runs Conan
os_build: [Windows, WindowsStore, Linux, Macos, FreeBSD, SunOS, AIX]
arch_build: [x86, x86_64, ppc32be, ppc32, ppc64le, ppc64, armv5el, armv5hf, armv6, armv7, armv7hf, armv7s, armv7k, armv8, armv8_32, armv8.3, sparc, sparcv9, mips, mips64, avr, s390, s390x, sh4le, e2k-v2, e2k-v3, e2k-v4, e2k-v5, e2k-v6, e2k-v7]
# Only for building cross compilation tools, 'os_target/arch_target' is the system for
# which the tools generate code
os_target: [Windows, Linux, Macos, Android, iOS, watchOS, tvOS, FreeBSD, SunOS, AIX, Arduino, Neutrino]
arch_target: [x86, x86_64, ppc32be, ppc32, ppc64le, ppc64, armv5el, armv5hf, armv6, armv7, armv7hf, armv7s, armv7k, armv8, armv8_32, armv8.3, sparc, sparcv9, mips, mips64, avr, s390, s390x, asm.js, wasm, sh4le, e2k-v2, e2k-v3, e2k-v4, e2k-v5, e2k-v6, e2k-v7]
# Rest of the settings are "host" settings:
# - For native building/cross building: Where the library/program will run.
# - For building cross compilation tools: Where the cross compiler will run.
os:
Windows:
subsystem: [None, cygwin, msys, msys2, wsl]
WindowsStore:
version: ["8.1", "10.0"]
WindowsCE:
platform: ANY
version: ["5.0", "6.0", "7.0", "8.0"]
Linux:
Macos:
version: [None, "10.6", "10.7", "10.8", "10.9", "10.10", "10.11", "10.12", "10.13", "10.14", "10.15", "11.0"]
sdk: [None, "macosx"]
subsystem: [None, catalyst]
Android:
api_level: ANY
iOS:
version: ["7.0", "7.1", "8.0", "8.1", "8.2", "8.3", "9.0", "9.1", "9.2", "9.3", "10.0", "10.1", "10.2", "10.3", "11.0", "11.1", "11.2", "11.3", "11.4", "12.0", "12.1", "12.2", "12.3", "12.4", "13.0", "13.1", "13.2", "13.3", "13.4", "13.5", "13.6"]
sdk: [None, "iphoneos", "iphonesimulator"]
watchOS:
version: ["4.0", "4.1", "4.2", "4.3", "5.0", "5.1", "5.2", "5.3", "6.0", "6.1"]
sdk: [None, "watchos", "watchsimulator"]
tvOS:
version: ["11.0", "11.1", "11.2", "11.3", "11.4", "12.0", "12.1", "12.2", "12.3", "12.4", "13.0"]
sdk: [None, "appletvos", "appletvsimulator"]
FreeBSD:
SunOS:
AIX:
Arduino:
board: ANY
Emscripten:
Neutrino:
version: ["6.4", "6.5", "6.6", "7.0"]
arch: [x86, x86_64, ppc32be, ppc32, ppc64le, ppc64, armv4, armv4i, armv5el, armv5hf, armv6, armv7, armv7hf, armv7s, armv7k, armv8, armv8_32, armv8.3, sparc, sparcv9, mips, mips64, avr, s390, s390x, asm.js, wasm, sh4le, e2k-v2, e2k-v3, e2k-v4, e2k-v5, e2k-v6, e2k-v7]
compiler:
sun-cc:
version: ["5.10", "5.11", "5.12", "5.13", "5.14"]
threads: [None, posix]
libcxx: [libCstd, libstdcxx, libstlport, libstdc++]
gcc: &gcc
version: ["4.1", "4.4", "4.5", "4.6", "4.7", "4.8", "4.9",
"5", "5.1", "5.2", "5.3", "5.4", "5.5",
"6", "6.1", "6.2", "6.3", "6.4", "6.5",
"7", "7.1", "7.2", "7.3", "7.4", "7.5",
"8", "8.1", "8.2", "8.3", "8.4",
"9", "9.1", "9.2", "9.3",
"10", "10.1"]
libcxx: [libstdc++, libstdc++11]
threads: [None, posix, win32] # Windows MinGW
exception: [None, dwarf2, sjlj, seh] # Windows MinGW
cppstd: [None, 98, gnu98, 11, gnu11, 14, gnu14, 17, gnu17, 20, gnu20]
Visual Studio: &visual_studio
runtime: [MD, MT, MTd, MDd]
version: ["8", "9", "10", "11", "12", "14", "15", "16"]
toolset: [None, v90, v100, v110, v110_xp, v120, v120_xp,
v140, v140_xp, v140_clang_c2, LLVM-vs2012, LLVM-vs2012_xp,
LLVM-vs2013, LLVM-vs2013_xp, LLVM-vs2014, LLVM-vs2014_xp,
LLVM-vs2017, LLVM-vs2017_xp, v141, v141_xp, v141_clang_c2, v142,
llvm, ClangCL]
cppstd: [None, 14, 17, 20]
msvc:
version: ["19.0",
"19.1", "19.10", "19.11", "19.12", "19.13", "19.14", "19.15", "19.16",
"19.2", "19.20", "19.21", "19.22", "19.23", "19.24", "19.25", "19.26", "19.27", "19.28"]
runtime: [static, dynamic]
runtime_type: [Debug, Release]
cppstd: [14, 17, 20]
clang:
version: ["3.3", "3.4", "3.5", "3.6", "3.7", "3.8", "3.9", "4.0",
"5.0", "6.0", "7.0", "7.1",
"8", "9", "10", "11"]
libcxx: [None, libstdc++, libstdc++11, libc++, c++_shared, c++_static]
cppstd: [None, 98, gnu98, 11, gnu11, 14, gnu14, 17, gnu17, 20, gnu20]
runtime: [None, MD, MT, MTd, MDd]
apple-clang: &apple_clang
version: ["5.0", "5.1", "6.0", "6.1", "7.0", "7.3", "8.0", "8.1", "9.0", "9.1", "10.0", "11.0", "12.0"]
libcxx: [libstdc++, libc++]
cppstd: [None, 98, gnu98, 11, gnu11, 14, gnu14, 17, gnu17, 20, gnu20]
intel:
version: ["11", "12", "13", "14", "15", "16", "17", "18", "19"]
base:
gcc:
<<: *gcc
threads: [None]
exception: [None]
Visual Studio:
<<: *visual_studio
apple-clang:
<<: *apple_clang
qcc:
version: ["4.4", "5.4"]
libcxx: [cxx, gpp, cpp, cpp-ne, accp, acpp-ne, ecpp, ecpp-ne]
mcst-lcc:
version: ["1.19", "1.20", "1.21", "1.22", "1.23", "1.24", "1.25"]
base:
gcc:
<<: *gcc
threads: [None]
exceptions: [None]
build_type: [None, Debug, Release, RelWithDebInfo, MinSizeRel]
cppstd: [None, 98, gnu98, 11, gnu11, 14, gnu14, 17, gnu17, 20, gnu20] # Deprecated, use compiler.cppstd
As you can see, the possible values settings
can take are restricted in the same file. This is done to ensure matching naming and
spelling as well as defining a common settings model among users and the OSS community.
If a setting is allowed to be set to any value, you can use ANY
.
If a setting is allowed to be set to any value or it can also be unset, you can use [None, ANY]
.
However, this configuration file can be modified to any needs, including new settings or subsettings and their values. If you want to distribute a unified settings.yml file you can use the conan config install command.
Note
The settings.yml file is not perfect nor definitive and surely incomplete. Please share any suggestion in the Conan issue tracker with any missing settings and values that could make sense for other users.
To force the creation of the settings.yml the command conan config init
is available.
Compilers
Some notes about different compilers:
msvc
The new msvc
compiler is a new, experimental one, that is intended to deprecate the Visual Studio
one in Conan 2.0:
It uses the compiler version, that is 19.0, 19.1, etc, instead of the Visual Studio IDE (15, 16, etc).
It is only used by the new build integrations in conan.tools.cmake and conan.tools.microsoft, but not the previous ones.
At the moment it implements a
compatible_packages
fallback to Visual Studio compiled packages, that is, previous existing binaries compiled withsettings.compiler="Visual Studio"
can be used for themsvc
compiler if no binaries exist for it yet. This behavior can be opted-out withcore.package_id:msvc_visual_incompatible
global.conf configuration.It is not detected by the profile auto-detect, it needs to explicitly be defined in profiles.
When using the msvc
compiler, the Visual Studio toolset version (the actual vcvars
activation and MSBuild
location) will be
defined by the default provide of that compiler version:
msvc
compiler version ‘19.0’: Visual Studio 14 2015msvc
compiler version ‘19.1’: Visual Studio 15 2017msvc
compiler version ‘19.2’: Visual Studio 16 2019
This can be configured in your profiles with the tools.microsoft.msbuild:vs_version
configuration:
[settings]
compiler=msvc
compiler.version=19.0
[conf]
tools.microsoft.msbuild:vs_version = 16
In this case, the vcvars
will activate the Visual Studio 16 installation, but the 19.0
compiler version will still be used
because the necessary toolset=v140
will be set.
Architectures
Here you can find a brief explanation of each of the architectures defined as arch
, arch_build
and arch_target
settings.
x86: The popular 32 bit x86 architecture.
x86_64: The popular 64 bit x64 architecture.
ppc64le: The PowerPC 64 bit Big Endian architecture.
ppc32: The PowerPC 32 bit architecture.
ppc64le: The PowerPC 64 bit Little Endian architecture.
ppc64: The PowerPC 64 bit Big Endian architecture.
armv5el: The ARM 32 bit version 5 architecture, soft-float.
armv5hf: The ARM 32 bit version 5 architecture, hard-float.
armv6: The ARM 32 bit version 6 architecture.
armv7: The ARM 32 bit version 7 architecture.
armv7hf: The ARM 32 bit version 7 hard-float architecture.
armv7s: The ARM 32 bit version 7 swift architecture mostly used in Apple’s A6 and A6X chips on iPhone 5, iPhone 5C and iPad 4.
armv7k: The ARM 32 bit version 7 k architecture mostly used in Apple’s WatchOS.
armv8: The ARM 64 bit and 32 bit compatible version 8 architecture. It covers only the
aarch64
instruction set.armv8_32: The ARM 32 bit version 8 architecture. It covers only the
aarch32
instruction set (a.k.a.ILP32
).armv8.3: The ARM 64 bit and 32 bit compatible version 8.3 architecture. Also known as
arm64e
, it is used on the A12 chipset added in the latest iPhone models (XS/XS Max/XR).sparc: The SPARC (Scalable Processor Architecture) originally developed by Sun Microsystems.
sparcv9: The SPARC version 9 architecture.
mips: The 32 bit MIPS (Microprocessor without Interlocked Pipelined Stages) developed by MIPS Technologies (formerly MIPS Computer Systems).
mips64: The 64 bit MIPS (Microprocessor without Interlocked Pipelined Stages) developed by MIPS Technologies (formerly MIPS Computer Systems).
avr: The 8 bit AVR microcontroller architecture developed by Atmel (Microchip Technology).
s390: The 32 bit address Enterprise Systems Architecture 390 from IBM.
s390x: The 64 bit address Enterprise Systems Architecture 390 from IBM.
asm.js: The subset of JavaScript that can be used as low-level target for compilers, not really a processor architecture, it’s produced by Emscripten. Conan treats it as an architecture to align with build systems design (e.g. GNU auto tools and CMake).
wasm: The Web Assembly, not really a processor architecture, but byte-code format for Web, it’s produced by Emscripten. Conan treats it as an architecture to align with build systems design (e.g. GNU auto tools and CMake).
sh4le: The Hitachi SH-4 SuperH architecture.
e2k-v2: The Elbrus 2000 v2 512 bit VLIW (Very Long Instruction Word) architecture (Elbrus 2CM, Elbrus 2C+ CPUs) originally developed by MCST (Moscow Center of SPARC Technologies).
e2k-v3: The Elbrus 2000 v3 512 bit VLIW (Very Long Instruction Word) architecture (Elbrus 2S, aka Elbrus 4C, CPU) originally developed by MCST (Moscow Center of SPARC Technologies).
e2k-v4: The Elbrus 2000 v4 512 bit VLIW (Very Long Instruction Word) architecture (Elbrus 8C, Elbrus 8C1, Elbrus 1C+ and Elbrus 1CK CPUs) originally developed by MCST (Moscow Center of SPARC Technologies).
e2k-v5: The Elbrus 2000 v5 512 bit VLIW (Very Long Instruction Word) architecture (Elbrus 8C2 ,aka Elbrus 8CB, CPU) originally developed by MCST (Moscow Center of SPARC Technologies).
e2k-v6: The Elbrus 2000 v6 512 bit VLIW (Very Long Instruction Word) architecture (Elbrus 2C3, Elbrus 12C and Elbrus 16C CPUs) originally developed by MCST (Moscow Center of SPARC Technologies).
e2k-v7: The Elbrus 2000 v7 512 bit VLIW (Very Long Instruction Word) architecture (Elbrus 32C CPU) originally developed by MCST (Moscow Center of SPARC Technologies).
C++ standard libraries (aka compiler.libcxx)
compiler.libcxx
sub-setting defines C++ standard libraries implementation to be used. The sub-setting applies only to certain compilers,
e.g. it applies to clang, apple-clang and gcc, but doesn’t apply to Visual Studio.
libstdc++ (gcc, clang, apple-clang, sun-cc): The GNU C++ Library. NOTE that this implicitly defines _GLIBCXX_USE_CXX11_ABI=0 to use old ABI. See How to manage the GCC >= 5 ABI for the additional details. Might be a wise choice for old systems, such as CentOS 6. On Linux systems, you may need to install libstdc++-dev (package name could be different in various distros) in order to use the standard library. NOTE that on Apple systems usage of libstdc++ has been deprecated.
libstdc++11 (gcc, clang, apple-clang): The GNU C++ Library. NOTE that this implicitly defines _GLIBCXX_USE_CXX11_ABI=1 to use new ABI. See How to manage the GCC >= 5 ABI for the additional details. Might be a wise choice for newer systems, such as Ubuntu 20. On Linux systems, you may need to install libstdc++-dev (package name could be different in various distros) in order to use the standard library. NOTE that on Apple systems usage of libstdc++ has been deprecated.
libc++ (clang, apple-clang): LLVM libc++. On Linux systems, you may need to install libc++-dev (package name could be different in various distros) in order to use the standard library.
c++_shared (clang, Android only): use LLVM libc++ as a shared library. Refer to the C++ Library Support for the additiona details.
c++_static (clang, Android only): use LLVM libc++ as a static library. Refer to the C++ Library Support for the additiona details.
libCstd (sun-cc): Rogue Wave’s stdlib. See Comparing C++ Standard Libraries libCstd, libstlport, and libstdcxx.
libstlport (sun-cc): STLport. See Comparing C++ Standard Libraries libCstd, libstlport, and libstdcxx.
libstdcxx (sun-cc): Apache C++ Standard Library. See Comparing C++ Standard Libraries libCstd, libstlport, and libstdcxx.
gpp (qcc): GNU C++ lib. See QCC documentation.
cpp (qcc): Dinkum C++ lib. See QCC documentation.
cpp-ne (qcc): Dinkum C++ lib (no exceptions). See QCC documentation.
acpp (qcc): Dinkum Abridged C++ lib. See QCC documentation.
acpp-ne (qcc): Dinkum Abridged C++ lib (no exceptions). See QCC documentation.
ecpp (qcc): Embedded Dinkum C++ lib. See QCC documentation.
ecpp-ne (qcc): Embedded Dinkum C++ lib (no exceptions). See QCC documentation.
cxx (qcc): LLVM C++. See QCC documentation.