Working with Intel compilers
intel
Note
This compiler is aimed to manage legacy Intel Parallel Studio XE compiler versions. For new Intel oneAPI, check the
information about the intel-cc
compiler below.
The Intel
compiler is a particular case, as it uses Visual Studio
compiler in Windows environments
and gcc
in Linux environments. If you are wondering how to manage the compatibility between the packages generated
with intel
and the generated with the pure base compiler (gcc
or Visual Studio
) check the
Compatible Packages and Compatible Compilers sections.
intel-cc
Warning
The support for this compiler is experimental and subject to breaking changes.
Available since: 1.41.0
This new compiler is defined to manage the different Intel oneAPI DPC++/C++ and Classic ones.
Warning
macOS is not supported for the Intel oneAPI DPC++/C++ (icx/icpx or dpcpp) compilers. For macOS or Xcode support, you’ll have to use the Intel C++ Classic Compiler.
It can be declared into your local profile like any other compiler as follows:
[settings]
...
compiler=intel-cc
compiler.mode=dpcpp
compiler.version=2021.3
compiler.libcxx=libstdc++
build_type=Release
[options]
[build_requires]
[env]
CC=dpcpp
CXX=dpcpp
[conf]
tools.intel:installation_path=/opt/intel/oneapi
Important
Remember to put this [conf]
entry to find out the root path of your Intel oneAPI folder. Normally, it’ll be installed by default:
* Linux, Darwin: /opt/intel/oneapi
* Windows: C:\Program Files (x86)\Intel\oneAPI
We’re specifying the CC
and CXX
compilers and the compiler.mode
subsetting. The possible values for compiler.mode
are:
icx
for Intel oneAPI C++ (icx/icpx compilers).dpcpp
for Intel oneAPI DPC++ (dpcpp compiler and dpcpp-cl for Windows only).classic
for Intel C++ Classic (icc for Linux and icl for Windows).
To set up the compiler without Conan you need to run an Intel official script to set all the proper variables to use those compilers called setvars.sh|bat
script.
If you are using either the CMakeToolChain
or the MSBuildToolchain
, when using the intel-cc
compiler, Conan automatically calls the setvars
script.
Otherwise, you can use the IntelCC generator.
This is an example of a Conan package called hello/1.0
using the CMakeToolchain
. Remember you can use the command conan new hello/1.0 -m cmake_lib
to create a simple project like this one:
from conans import ConanFile
from conan.tools.cmake import CMakeToolchain
class HelloConan(ConanFile):
name = "hello"
version = "1.0"
# more code here...
def generate(self):
tc = CMakeToolchain(self)
tc.generate()
Running conan create . -pr intelprofile -pr:b intelprofile, you’ll see something like this output:
......
hello/1.0: Generating the package
hello/1.0: Package folder /home/franchuti/.conan/data/hello/1.0/_/_/package/7d9c7d5fa3c48c9705c2cb864656c00fa8672524
hello/1.0: Calling package()
hello/1.0: CMake command: cmake --build '/home/franchuti/.conan/data/hello/1.0/_/_/build/7d9c7d5fa3c48c9705c2cb864656c00fa8672524/cmake-build-release' '--target' 'install'
:: initializing oneAPI environment ...
dash: SH_VERSION = unknown
:: advisor -- latest
:: ccl -- latest
:: clck -- latest
:: compiler -- latest
:: dal -- latest
:: debugger -- latest
:: dev-utilities -- latest
:: dnnl -- latest
:: dpcpp-ct -- latest
:: dpl -- latest
:: inspector -- latest
:: intelpython -- latest
:: ipp -- latest
:: ippcp -- latest
:: ipp -- latest
:: itac -- latest
:: mkl -- latest
:: mpi -- latest
:: tbb -- latest
:: vpl -- latest
:: vtune -- latest
:: oneAPI environment initialized ::
Using Conan toolchain through /home/franchuti/.conan/data/hello/1.0/_/_/build/7d9c7d5fa3c48c9705c2cb864656c00fa8672524/cmake-build-release/conan/conan_toolchain.cmake.
-- Conan toolchain: Setting CMAKE_POSITION_INDEPENDENT_CODE=ON (options.fPIC)
-- Conan toolchain: Setting BUILD_SHARED_LIBS= OFF
-- The CXX compiler identification is Clang 13.0.0
-- Check for working CXX compiler: /opt/intel/oneapi/compiler/2021.3.0/linux/bin/dpcpp
Using Conan toolchain through .
-- Check for working CXX compiler: /opt/intel/oneapi/compiler/2021.3.0/linux/bin/dpcpp -- works
-- Detecting CXX compiler ABI info
Using Conan toolchain through .
-- Detecting CXX compiler ABI info - done
-- Detecting CXX compile features
-- Detecting CXX compile features - done
-- Configuring done
-- Generating done
.......
As you can observe, you have used one of these Intel compilers, the DPC++ one and successfully generated the libhello.a
file.
intel-cc and Microsoft Visual Studio
Note
Ensure you have installed the Intel plugins for Microsoft Visual Studio before reading this section.
If you’re working on a Microsoft Visual Studio project, you can add the Intel Toolset as a new .props file. Let’s suppose you have defined these files into your current project folder:
[settings]
os=Windows
os_build=Windows
arch=x86_64
arch_build=x86_64
compiler=intel-cc
compiler.mode=classic
compiler.version=2021.3
compiler.runtime=dynamic
build_type=Release
[options]
[build_requires]
[env]
[conf]
tools.intel:installation_path="C:\Program Files (x86)\Intel\oneAPI"
from conans import ConanFile
from conan.tools.microsoft import MSBuildToolchain
class App(ConanFile):
settings = "os", "arch", "compiler", "build_type"
def generate(self):
tc = MSBuildToolchain(self)
tc.generate()
Running a conan install . -pr intelprofile, a file conantoolchain_release_x64.props is generated in your current folder:
<?xml version="1.0" encoding="utf-8"?>
<Project xmlns="http://schemas.microsoft.com/developer/msbuild/2003">
<ItemDefinitionGroup>
<ClCompile>
<PreprocessorDefinitions>
;%(PreprocessorDefinitions)
</PreprocessorDefinitions>
<RuntimeLibrary>MultiThreadedDLL</RuntimeLibrary>
<LanguageStandard></LanguageStandard>
</ClCompile>
</ItemDefinitionGroup>
<PropertyGroup Label="Configuration">
<PlatformToolset>Intel C++ Compiler 19.2</PlatformToolset>
</PropertyGroup>
</Project>
Note that a PlatformToolset
is set to Intel C++ Compiler 19.2
. You can import that file to your project or solution of Visual Studio.
Read more about the MSBuildToolchain here.
Note
See the complete IntelCC reference for more information about that tool.