How does one check the version of NVCC, with CMake 3.15 (not 3.17 or later)?
I suppose I could write my own module to run it with --version, but is there an easier way to do it?
In CMake 3.15 with CUDA language support enabled, the CMAKE_CUDA_COMPILER_VERSION variable is defined and contains the correct version. MRE:
cmake_minimum_required(VERSION 3.15)
project(test LANGUAGES CUDA)
message(STATUS "CMAKE_CUDA_COMPILER = ${CMAKE_CUDA_COMPILER}")
message(STATUS "CMAKE_CUDA_COMPILER_ID = ${CMAKE_CUDA_COMPILER_ID}")
message(STATUS "CMAKE_CUDA_COMPILER_VERSION = ${CMAKE_CUDA_COMPILER_VERSION}")
Output:
$ cmake-3.15 -S . -B build
-- The CUDA compiler identification is NVIDIA 10.1.243
-- Check for working CUDA compiler: /usr/bin/nvcc
-- Check for working CUDA compiler: /usr/bin/nvcc -- works
-- Detecting CUDA compiler ABI info
-- Detecting CUDA compiler ABI info - done
-- CMAKE_CUDA_COMPILER = /usr/bin/nvcc
-- CMAKE_CUDA_COMPILER_ID = NVIDIA
-- CMAKE_CUDA_COMPILER_VERSION = 10.1.243
-- Configuring done
-- Generating done
-- Build files have been written to: /home/alex/test/build
I have various versions of CMake installed as cmake-3.XX, I assume you would simply run cmake.
Related
We are using flatbuffer v1.3.0 and seeing error building flatc compiler.
Same error for flatbuffer version v1.3.0 to v1.12.1
Build is successful for version 2.0.0
Should we use a particular version of make to build the flatc compiler? If so how to do it?
Steps
git clone https://github.com/google/flatbuffers/tree/v1.3.0
cd flatbuffers
cmake -G "Unix Makefiles" -DCMAKE_BUILD_TYPE=Release -Success
make --Error
Versions installed on MacOS BigSur Version 11.6
$ cmake --version
cmake version 3.19.4
CMake suite maintained and supported by Kitware (kitware.com/cmake).
$ make --version
GNU Make 3.81
Copyright (C) 2006 Free Software Foundation, Inc.
This is free software; see the source for copying conditions.
There is NO warranty; not even for MERCHANTABILITY or FITNESS FOR A
PARTICULAR PURPOSE.
This program built for i386-apple-darwin11.3.0
$ gcc --version
Configured with: --prefix=/Library/Developer/CommandLineTools/usr --with-gxx-include-dir=/Library/Developer/CommandLineTools/SDKs/MacOSX.sdk/usr/include/c++/4.2.1
Apple clang version 13.0.0 (clang-1300.0.29.3)
Target: x86_64-apple-darwin20.6.0
Thread model: posix
InstalledDir: /Library/Developer/CommandLineTools/usr/bin
$ cmake -G "Unix Makefiles" -DCMAKE_BUILD_TYPE=Release
CMake Warning:
No source or binary directory provided. Both will be assumed to be the
same as the current working directory, but note that this warning will
become a fatal error in future CMake releases.
CMake Deprecation Warning at CMakeLists.txt:1 (cmake_minimum_required):
Compatibility with CMake < 2.8.12 will be removed from a future version of
CMake.
Update the VERSION argument <min> value or use a ...<max> suffix to tell
CMake that the project does not need compatibility with older versions.
-- The C compiler identification is AppleClang 13.0.0.13000029
-- The CXX compiler identification is AppleClang 13.0.0.13000029
-- Detecting C compiler ABI info
-- Detecting C compiler ABI info - done
-- Check for working C compiler: /Library/Developer/CommandLineTools/usr/bin/cc - skipped
-- Detecting C compile features
-- Detecting C compile features - done
-- Detecting CXX compiler ABI info
-- Detecting CXX compiler ABI info - done
-- Check for working CXX compiler: /Library/Developer/CommandLineTools/usr/bin/c++ - skipped
-- Detecting CXX compile features
-- Detecting CXX compile features - done
-- Configuring done
-- Generating done
-- Build files have been written to: /Users/abcd/Documents/FlatBuffer/flatbuffers
$ make
Scanning dependencies of target flatc
[ 2%] Building CXX object CMakeFiles/flatc.dir/src/idl_parser.cpp.o
In file included from /Users/abcd/Documents/FlatBuffer/flatbuffers/src/idl_parser.cpp:28:
In file included from /Users/abcd/Documents/FlatBuffer/flatbuffers/include/flatbuffers/idl.h:25:
/Users/abcd/Documents/FlatBuffer/flatbuffers/include/flatbuffers/flatbuffers.h:1086:25: error: definition of implicit copy constructor for 'TableKeyComparator<reflection::Object>' is deprecated because it has a user-declared copy assignment operator [-Werror,-Wdeprecated-copy]
TableKeyComparator& operator= (const TableKeyComparator&);
^
/Users/abcd/Documents/FlatBuffer/flatbuffers/include/flatbuffers/flatbuffers.h:1100:27: note: in implicit copy constructor for 'flatbuffers::FlatBufferBuilder::TableKeyComparator<reflection::Object>' first required here
std::sort(v, v + len, TableKeyComparator<T>(buf_));
^
/Users/abcd/Documents/FlatBuffer/flatbuffers/include/flatbuffers/flatbuffers.h:1113:12: note: in instantiation of function template specialization 'flatbuffers::FlatBufferBuilder::CreateVectorOfSortedTables<reflection::Object>' requested here
return CreateVectorOfSortedTables(v->data(), v->size());
^
/Users/abcd/Documents/FlatBuffer/flatbuffers/src/idl_parser.cpp:2008:35: note: in instantiation of function template specialization 'flatbuffers::FlatBufferBuilder::CreateVectorOfSortedTables<reflection::Object>' requested here```
Flatbuffer requires a specific version of clang to compile the 1.12.1.
Installing the required version.
sudo apt-get install clang-6.0 lld-6.0 clang-tidy-6.0
use update-alternatives to link the clang.
sudo update-alternatives --install /usr/bin/clang clang /usr/bin/clang-6.0 10
sudo update-alternatives --install /usr/bin/clang++ clang++
/usr/bin/clang++-6.0 10
I am trying to use CMake to compile a library using cuda.
I want cuda to be optional, i.e. I want my library to compile if nvcc is not available on the system.
Here is how I am checking if cuda is available in my CMake:
include(CheckLanguage)
check_language(CUDA)
if (CMAKE_CUDA_COMPILER)
enable_language(CUDA)
else()
message(WARNING "CUDA not found: GPU features won't be available.")
endif ()
Here are the environment variables I am setting:
$> env | grep CXX
CUDAHOSTCXX=/usr/bin/g++-6
CUDACXX=/usr/bin/nvcc
CXX=/usr/bin/g++
Here is the output of the cmake command:
$> cmake ..
-- The CXX compiler identification is GNU 7.4.0
-- Check for working CXX compiler: /usr/bin/g++
-- Check for working CXX compiler: /usr/bin/g++ -- works
-- Detecting CXX compiler ABI info
-- Detecting CXX compiler ABI info - done
-- Detecting CXX compile features
-- Detecting CXX compile features - done
-- Looking for a CUDA compiler
-- Looking for a CUDA compiler - /usr/bin/nvcc
-- The CUDA compiler identification is NVIDIA 9.1.85
-- Check for working CUDA compiler: /usr/bin/nvcc
-- Check for working CUDA compiler: /usr/bin/nvcc -- works
-- Detecting CUDA compiler ABI info
-- Detecting CUDA compiler ABI info - done
-- Configuring done
CMake Error: Error required internal CMake variable not set, cmake may not be built correctly.
Missing variable is:
CMAKE_CUDA_COMPILE_WHOLE_COMPILATION
-- Generating done
CMake Generate step failed. Build files cannot be regenerated correctly.
I am not finding anything usefull conserning CMAKE_CUDA_COMPILE_WHOLE_COMPILATION.
How can I make this cmake works? What did I missed?
I just fingure out what the problem was...
I was calling a function
function(enable_cuda_if_available)
# Here CUDA is properly found and variable are correctly set
include(CheckLanguage)
check_language(CUDA)
if (CMAKE_CUDA_COMPILER)
enable_language(CUDA)
else ()
message(WARNING "CUDA not found: GPU features won't be available.")
endif ()
endfunction()
enable_cuda_if_available()
# Here CUDA's variables have not been forwarded to the parent scope and leading to an error
I still don't know how to automatically forward all variables set by enable_cuda_if_available to the parent scope
I am trying to compile a library with both, C++ and CUDA source files. I am using GNU make with CMake. My compiler of choice is clang, since CUDA only supports gcc up to version 5 and Debian 9 has gcc 6 as its oldest version and I have to use software that is provided by the Debian 9 or 10 repositories.
CMake version is 3.9.0
clang version is 3.8.1
cc and c++ in /usr/bin correctly link to clang and clang++ ẃhich also link to the correct files.
Unfortunately the initial checks of CMake for CUDA fail although everything, as far as I can see, seems to be set up correctly. It looks like the arguments aren't passed correctly to the CUDA compiler.
This is a part of my project's main CMake file:
cmake_minimum_required (VERSION 3.9.0 FATAL_ERROR)
project (dev)
find_package(CUDA REQUIRED)
set(CUDA_HOST_COMPILATION_CPP ON)
set(CUDA_NVCC_FLAGS "${CUDA_NVCC_FLAGS} -Wno-deprecated-gpu-targets -ccbin clang-3.8")
The library's CMake file looks like this:
cmake_minimum_required(VERSION 3.9.0 FATAL_ERROR)
project (libname LANGUAGES CXX CUDA)
file(GLOB SOURCES "*.cu" "*.cpp")
add_library(libname ${SOURCES})
set_target_properties(libname PROPERTIES CUDA_SEPARABLE_COMPILATION ON)
set_target_properties(libname PROPERTIES POSITION_INDEPENDENT_CODE ON)
This is the output of CMake:
-- The C compiler identification is Clang 3.8.1
-- The CXX compiler identification is Clang 3.8.1
-- Check for working C compiler: /usr/bin/cc
-- Check for working C compiler: /usr/bin/cc -- works
-- Detecting C compiler ABI info
-- Detecting C compiler ABI info - done
-- Detecting C compile features
-- Detecting C compile features - done
-- Check for working CXX compiler: /usr/bin/c++
-- Check for working CXX compiler: /usr/bin/c++ -- works
-- Detecting CXX compiler ABI info
-- Detecting CXX compiler ABI info - done
-- Detecting CXX compile features
-- Detecting CXX compile features - done
-- Looking for pthread.h
-- Looking for pthread.h - found
-- Looking for pthread_create
-- Looking for pthread_create - not found
-- Looking for pthread_create in pthreads
-- Looking for pthread_create in pthreads - not found
-- Looking for pthread_create in pthread
-- Looking for pthread_create in pthread - found
-- Found Threads: TRUE
-- Found CUDA: /usr (found version "8.0")
-- The CUDA compiler identification is unknown
-- Check for working CUDA compiler: /usr/bin/nvcc
-- Check for working CUDA compiler: /usr/bin/nvcc -- broken
The CUDA/nvcc tests of CMake fails with the following errors:
Change Dir: /home/user/projects/hamonIC-linux-experimental/current_state/working_copy/code/build/CMakeFiles/CMakeTmp
Run Build Command:"/usr/bin/make" "cmTC_6d1a9/fast"
/usr/bin/make -f CMakeFiles/cmTC_6d1a9.dir/build.make
CMakeFiles/cmTC_6d1a9.dir/build
make[1]: Entering directory
'/home/user/projects/hamonIC-linux-experimental/current_state/working_copy/code/build/CMakeFiles/CMakeTmp'
Building CUDA object CMakeFiles/cmTC_6d1a9.dir/main.cu.o
/usr/bin/nvcc -x cu -c
/home/user/projects/hamonIC-linux-experimental/current_state/working_copy/code/build/CMakeFiles/CMakeTmp/main.cu
-o CMakeFiles/cmTC_6d1a9.dir/main.cu.o
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are
deprecated, and may be removed in a future release (Use
-Wno-deprecated-gpu-targets to suppress warning).
ERROR: No supported gcc/g++ host compiler found, but clang-3.8 is
available.
Use 'nvcc -ccbin clang-3.8' to use that instead.
CMakeFiles/cmTC_6d1a9.dir/build.make:65: recipe for target
'CMakeFiles/cmTC_6d1a9.dir/main.cu.o' failed
make[1]: *** [CMakeFiles/cmTC_6d1a9.dir/main.cu.o] Error 1
make[1]: Leaving directory
'/home/user/projects/hamonIC-linux-experimental/current_state/working_copy/code/build/CMakeFiles/CMakeTmp'
Makefile:126: recipe for target 'cmTC_6d1a9/fast' failed
make: *** [cmTC_6d1a9/fast] Error 2
CMake will not be able to correctly generate this project.
Call Stack (most recent call first):
hic_iplibrary/source/ciccommon/CMakeLists.txt:7 (project)
The CUDA NVCC flags in CMake have to be semicolon delimited and not whitespace delimited.
Change the flags in your CMakeLists.txt to use:
set(CUDA_NVCC_FLAGS "${CUDA_NVCC_FLAGS};-Wno-deprecated-gpu-targets;-ccbin=clang-3.8")
and that should allow CMake to pass your flags to NVCC.
I am attempting to build the ev3dev C++ bindings (see https://github.com/ddemidov/ev3dev-lang-cpp) for my EV3. But CMake isn't using my cross-compiler toolchain. I've successfully installed the code sourcery toolchain, built a simple hello world app, and run it on my brick. That works just fine. The trouble is that CMake uses my host's g++ compiler instead of the toolchain. I'm familiar with linux and cross-compilers, but I'm a total CMake noob. So I suspect CMake is where I'm going wrong.
I've added the following lines into CMakeLists.txt:
set(CMAKE_C_COMPILER "/cygdrive/c/Users/me/MentorGraphics/Sourcery_CodeBench_Lite_for_ARM_GNU_Linux/bin/arm-none-linux-gnueabi-gcc.exe")
set(CMAKE_CXX_COMPILER "/cygdrive/c/Users/me/MentorGraphics/Sourcery_CodeBench_Lite_for_ARM_GNU_Linux/bin/arm-none-linux-gnueabi-g++.exe")
Here's the host gcc compiler:
$ gcc --version
gcc (GCC) 5.4.0
Copyright (C) 2015 Free Software Foundation, Inc.
This is free software; see the source for copying conditions. There is NO
warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
And the target toolchain gcc compiler:
$ /cygdrive/c/Users/me/MentorGraphics/Sourcery_CodeBench_Lite_for_ARM_GNU_Linux/bin/arm-none-linux-gnueabi-gcc.exe --version
arm-none-linux-gnueabi-gcc.exe (Sourcery CodeBench Lite 2014.05-29) 4.8.3 20140320 (prerelease)
Copyright (C) 2013 Free Software Foundation, Inc.
This is free software; see the source for copying conditions. There is NO
warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
The compilers are different versions, so they are really easy to distinguish.
Now here's the terminal session for the cmake build:
$ mkdir build
$ cd build
$ cmake ../ -DEV3DEV_PLATFORM=EV3
-- No build type selected, default to RelWithDebInfo
-- The C compiler identification is GNU 5.4.0
-- The CXX compiler identification is GNU 5.4.0
CMake Warning at /usr/share/cmake-3.6.2/Modules/Platform/CYGWIN.cmake:15 (message):
CMake no longer defines WIN32 on Cygwin!
(1) If you are just trying to build this project, ignore this warning or
quiet it by setting CMAKE_LEGACY_CYGWIN_WIN32=0 in your environment or in
the CMake cache. If later configuration or build errors occur then this
project may have been written under the assumption that Cygwin is WIN32.
In that case, set CMAKE_LEGACY_CYGWIN_WIN32=1 instead.
(2) If you are developing this project, add the line
set(CMAKE_LEGACY_CYGWIN_WIN32 0) # Remove when CMake >= 2.8.4 is required
at the top of your top-level CMakeLists.txt file or set the minimum
required version of CMake to 2.8.4 or higher. Then teach your project to
build on Cygwin without WIN32.
Call Stack (most recent call first):
/usr/share/cmake-3.6.2/Modules/CMakeSystemSpecificInformation.cmake:36 (include)
CMakeLists.txt:7 (project)
-- Check for working C compiler: /usr/bin/cc
-- Check for working C compiler: /usr/bin/cc -- works
-- Detecting C compiler ABI info
-- Detecting C compiler ABI info - done
-- Detecting C compile features
-- Detecting C compile features - done
-- Check for working CXX compiler: /usr/bin/c++.exe
-- Check for working CXX compiler: /usr/bin/c++.exe -- works
-- Detecting CXX compiler ABI info
-- Detecting CXX compiler ABI info - done
-- Detecting CXX compile features
-- Detecting CXX compile features - done
-- Configuring done
-- Generating done
-- Build files have been written to: /cygdrive/c/Users/me/Documents/AppProjects/ev3dev-lang-cpp/build
$
CMake is clearly using the cygwin host gcc/g++ compiler even though I've attempted to configure it to use the toolchain. Where am I going wrong?
EDIT: This is not a duplicate of cmake: problems specifying the compiler (2). I did actually use google, and the cross-compiling instructions on the CMake wiki (https://cmake.org/Wiki/CMake_Cross_Compiling) explicitly dictate using SET(). The problem turned out to be that CMake requires the compiler to be set before PROJECT(). I just moved the SET() lines up above PROJECT() and then CMake found the compiler.
For demonstration, I use the 3-line CMakeLists.txt:
cmake_minimum_required(VERSION 2.8)
find_package( BLAS REQUIRED )
message( STATUS BLAS found: ${BLAS_LIBRARIES} )
I have cblas, ATLAS and OpenBLAS including developer packages installed on a Debian Linux system, plus CMake 2.8.9. When I call
cmake . -DBLA_VENDOR=ATLAS -DCMAKE_PREFIX_PATH=/usr/lib/atlas-base
the ATLAS library nicely appears found:
-- The C compiler identification is GNU 4.7.2
-- The CXX compiler identification is GNU 4.7.2
-- Check for working C compiler: /usr/bin/gcc
-- Check for working C compiler: /usr/bin/gcc -- works
-- Detecting C compiler ABI info
-- Detecting C compiler ABI info - done
-- Check for working CXX compiler: /usr/bin/c++
-- Check for working CXX compiler: /usr/bin/c++ -- works
-- Detecting CXX compiler ABI info
-- Detecting CXX compiler ABI info - done
-- Looking for dgemm_
-- Looking for dgemm_ - found
-- A library with BLAS API found.
-- BLASfound:/usr/lib/atlas-base/libf77blas.so/usr/lib/atlas-base/libatlas.so
-- Configuring done
-- Generating done
-- Build files have been written to: /tmp
Similarly, just
cmake .
will find /usr/lib/libblas.so for me. (I do not forget to remove the cache files before the second call.)
When I look into /usr/share/cmake-2.8/Modules/FindBLAS.cmake, I read as the permitted values of BLA_VENDOR:
## Goto,ATLAS PhiPACK,CXML,DXML,SunPerf,SCSL,SGIMATH,IBMESSL,Intel10_32 (intel mkl v10 32 bit),Intel10_64lp (intel mkl v10 64 bit,lp thread model, lp64 model),
## Intel10_64lp_seq (intel mkl v10 64 bit,sequential code, lp64 model),
## Intel( older versions of mkl 32 and 64 bit), ACML,ACML_MP,ACML_GPU,Apple, NAS, Generic
That is, OpenBLAS is not listed. And a few random trials like
cmake . -DBLA_VENDOR=open -DCMAKE_PREFIX_PATH=/usr/lib/openblas-base
do not work either. Do I have to write my own FindBLAS in order to link to OpenBLAS with CMake?
CMake supports finding OpenBLAS using FindBLAS since CMake 3.6.0, as this commit finally made it into the release.
NB: OpenBLAS can also be used to substitute LAPACK, for which you should use the FindLAPACK command, that is also available since 3.6.0.
The "FindBLAS" and "FindLAPACK" modules learned to support OpenBLAS.
(See: https://blog.kitware.com/cmake-3-6-0-rc3-is-now-ready/)
I think you have to write your own CMake module for that.
You can check there and there maybe for inspirations
According to caffe, you can use it this way,below is my project structure:
FindOpenBLAS.cmake contents as below:
SET(Open_BLAS_INCLUDE_SEARCH_PATHS
/usr/include
/usr/include/openblas
/usr/include/openblas-base
/usr/local/include
/usr/local/include/openblas
/usr/local/include/openblas-base
/opt/OpenBLAS/include
$ENV{OpenBLAS_HOME}
$ENV{OpenBLAS_HOME}/include
)
SET(Open_BLAS_LIB_SEARCH_PATHS
/lib/
/lib/openblas-base
/lib64/
/usr/lib
/usr/lib/openblas-base
/usr/lib64
/usr/local/lib
/usr/local/lib64
/opt/OpenBLAS/lib
$ENV{OpenBLAS}cd
$ENV{OpenBLAS}/lib
$ENV{OpenBLAS_HOME}
$ENV{OpenBLAS_HOME}/lib
)
FIND_PATH(OpenBLAS_INCLUDE_DIR NAMES
cblas.h PATHS ${Open_BLAS_INCLUDE_SEARCH_PATHS}
)
FIND_LIBRARY(OpenBLAS_LIB NAMES
openblas PATHS ${Open_BLAS_LIB_SEARCH_PATHS}
)
message("OpenBLAS_INCLUDE_DIR: ${OpenBLAS_INCLUDE_DIR}, OpenBLAS_LIB: ${OpenBLAS_LIB}")
when use cmake it outputs :
The main CMakeLists.txt use the above FindOpenBLAS.cmake to find blas:
cmake_minimum_required(VERSION 3.5)
project(my_caffe)
set(CMAKE_CXX_STANDARD 11)
#dependence lib
set(CMAKE_MODULE_PATH ${PROJECT_SOURCE_DIR}/cmake)
find_package(OpenBLAS REQUIRED)
set(SOURCE_FILES main.cpp)
add_executable(my_caffe_test ${SOURCE_FILES})
# link blas
target_link_libraries(my_caffe_test
${OpenBLAS_LIB}
)