I've recently downloaded CUDA 7 and set it up to work with my project. On Mac, CUDA 7 requires clang to be the host compiler.
Now, I'm using a number of C++11 features. I've enabled these with -std=c++11 passed to nvcc this works. However, if I pass -Xcompiler -std=c++11 to nvcc, I get the following error regardless of if I also passed -std=c++11 by itself. The error is:
"invalid argument '-std=c++11' not allowed with 'C/ObjC'"
It seems like this should work, it certainly does with GCC. Anyone have a workaround. Otherwise, I'll file a bug report with Nvidia.
nvcc -dryrun ... will show what commands nvcc will execute. I don't currently have access to CUDA 7, only 6.5, but mine issues, among other things, two commands that compiles generated C source. This code generated by cudafe and have to be compiled by C compiler, but -Xcompiler adds options for both C and C++ modes.
I guess difference with my gcc situation is that I'm getting a warning while you have an error (this is exactly how gcc and clang differs in that case). For both compilers I don't see any way to suppress it, so I guess you have to fix your .cmake files to omit -Xcompiler options. This options shouldn't be used for language standard, just some very compiler-specific things.
Of course it doesn't work. You are specifying C++ options while using a C or Objective-C compiler. The source files must be C++ or Objective-C++.
Related
I am trying to build cmake source, which requires C++11.
The build halts and apparently the complaint is that C++11 is not detected. The g++ mode is actually set to -std=gnu++17
This is part of the console log
---------------------------------------------
CMake 3.18.20200919, Copyright 2000-2020 Kitware, Inc. and Contributors
Found GNU toolchain
C compiler on this system is: gcc
C++ compiler on this system is: g++ -std=gnu++17
Makefile processor on this system is: make
g++ has setenv
g++ has unsetenv
g++ does not have environ in stdlib.h
g++ has stl wstring
g++ has <ext/stdio_filebuf.h>
---------------------------------------------
g++ -std=gnu++17 -DCMAKE_BOOTSTRAP -DCMake_HAVE_CXX_MAKE_UNIQUE=1 -c $HOME/Apps/CMake-master/Source/cmAddCustomCommandCommand.cxx -o cmAddCustomCommandCommand.o
This is part of the error in the log file...
In file included from /usr/include/c++/5/unordered_map:35:0,
from cmake_bootstrap_11920_test.cxx:4:
/usr/include/c++/5/bits/c++0x_warning.h:32:2: error: #error This file requires compiler and library support for the ISO C++ 2011 standard. This support must be enabled with the -std=c++11 or -std=gnu++11 compiler options.
#error This file requires compiler and library support \
^
cmake_bootstrap_11920_test.cxx:7:2: error: #error "Compiler is not in a mode aware of C++11."
#error "Compiler is not in a mode aware of C++11."
^
cmake_bootstrap_11920_test.cxx:70:16: warning: non-static data member initializers only available with -std=c++11 or -std=gnu++11
int Member = 1;
Looking around on the web, I noticed that C++11 is only available after gcc version 4.6.
I checked my version, and it seems to be above.
g++ --version
g++ (Ubuntu 5.4.0-6ubuntu1~16.04.12) 5.4.0 20160609
I understand the -std=c++11 flag is used to enable the C++11 features in g++, but I don't seem to know what I am doing in this case.
I tried editing the CompileFlags.cmake file, but no change occurs.
I came upon this page which points to the cmake source I am using.
It says...
bootstrap: Require compiler mode aware of C++11
Some compilers have enough features enabled in their default modes to
pass our simple C++11 unique_ptr check but do not enable enough to build
CMake. Poison this case so that we choose one of the explicit `-std=`
options for such compilers.
Not sure what that means exactly.
How exactly do I change the g++ mode, to C++11, so that on running the bootstrap command, C++11 is used?
Or, in other words, how do I change std to point to C++11 (-std=c++11)?
First of all, you have g++ version 5.4.0 in your host PC installed, which is good, cause it means this is also supports the C++11, which you want to use.
To set it up, you could define it in your CMakeList.txt file:
set (CMAKE_CXX_STANDARD 11)
that should do the trick.
Please also check the documentation:
https://cmake.org/cmake/help/v3.1/variable/CMAKE_CXX_STANDARD.html
Usually, I would suggest to use the latest standard that you compiler is supporting (https://gcc.gnu.org/projects/cxx-status.html), cause you'll get also the latest features introduced in that standard. Exception for this rather in case you are working with legacy codes.
I'm using the Windows version of Clang (LLVM) 8 under Windows.
I'm compiling a code which uses OpenMP.
Under the lib folder of Clang there are 2 files which are OpenMP related:
libomp.lib.
libiomp5md.dll.
My questions are:
When I compile the code I use the flags -Xclang -fopenmp for the compiler. In in GCC and ICC using the flags tell the compiler to link the OpenMP library automatically. What about Clang? Does it do it automatically or must I link with libomp.lib manually? Is there a way to trigger automatic linking to the OpenMP library?
Answer: This was answered in Michael Klemm's answer below - Use the clang driver both for compiling and linking and then the -fopenmp will work as in GCC.
When I link with libomp.lib manually (Defining as a library for the linker) the output exe requires libomp.dll while the supplied OpenMP Dynamic Library is libiomp5md.dll. Is that a bug or is it because I link manually?
Answer: The libomp.dll is supplied in the bin folder and not the lib folder.
What's the proper way to utilize OpenMP in Clang under Windows? The clang-cl driver doesn't work with /openmp or -openmp as the MSVC's cl compiler.
Answer: Currently it can be done either with clang -fopenmp ..., clang-cl -Xclang -fopenmp ... or clang-cl /clang:-fopenmp ... (Which is equivalent of -Xclang -fopenmp).
Remark
On Windows I use Windows Driver of Clang using clang-cl.
Adding clarity to what the OpenMP libraries actually are, and how to use them on Windows with clang-cl
libomp.dll and libiomp5md.dll ARE THE SAME FILES!
When compiling for Windows, you link against libomp.lib OR libiomp5md.lib which will link to the same-named DLL at runtime, i.e. libomp.dll OR libiomp5md.dll respectively.
If you load 2 files that use the "different-name DLL," the interpreter will crash and give you a nasty error like: OMP: Error #15: Initializing libiomp5md.dll, but found libomp.dll already initialized.
Why? Because the program has no idea they are the same DLL, they have different names, so it assumes they are different. And it crashes. For this reason only, you can choose to swap which OpenMP DLL you link to in your program.
If your program doesn't crash and give you an error, you can keep using the same link to OpenMP. Otherwise, to silence the error, link to the one that is loaded by another program already.
If using clang-cl.exe which is the "drop-in" Clang replacement for MSVC cl.exe you should pass a compiler argument such as -Xclang -fopenmp which will convert the argument over to "Clang language." Don't forget to still pass to the linker the OpenMP LIB you chose, because on Windows, it won't be automatic.
That's all I've learned as brief as possible about OpenMP linking on Windows.
To compile and link OpenMP code with clang on Windows, you will have to pass -fopenmp to both the compiler and the linker:
clang -fopenmp -o bla.obj -c bla.c
clang -fopenmp -o bla.exe bla.obj
I am trying to configure the last version of clang (6.0) to use the arm-none-eabi linker instead of the ld.lld but clang is always ignoring everything and keep asking for the ld.lld one. I am trying to build for cortex-m3 (lpx1769 board). How do I force clang to use the linker I want.
-fuse-ld=ld is also not working, so does clang no longer allow the use of any other linker?
So the answer was to use the flag:
-fuse-ld=path/to/linker-to-be-used
Remember that if you passing this flag to clang it will cause a warning that clang will not use this flag (only the linker stage will do). Thus if you compiling with -Werror, the warning will be turned into an error.
Moreover, because you are cross-compiling probably you will need to let the linker know where to find the target-specific libraries needed using the -L option. See this for more info:
https://clang.llvm.org/docs/CrossCompilation.html
I followed instructions at http://clang.llvm.org/get_started.html
I compiled latest trunk of llvm and clang with MSVC 2010. Now I can compile simple programs with Clang but when I tried to compile this program I got a lot of errors.
Here is program:
#include <algorithm>
int main(){ return 0; }
And here are some of errors:
In file included from hello.cpp:1:
In file included from C:\Program Files\Microsoft Visual Studio 10.0\VC\include\algorithm:6:
In file included from C:\Program Files\Microsoft Visual Studio 10.0\VC\include\memory:987:
In file included from C:\Program Files\Microsoft Visual Studio 10.0\VC\include\intrin.h:24:
In file included from H:/LLVM/build/bin/Debug/../lib/clang/3.3/include\immintrin.h:32:
In file included from H:/LLVM/build/bin/Debug/../lib/clang/3.3/include\xmmintrin.h:988:
H:/LLVM/build/bin/Debug/../lib/clang/3.3/include\emmintrin.h:1384:22: error: expected expression
return (__m128)__in;
^
H:/LLVM/build/bin/Debug/../lib/clang/3.3/include\emmintrin.h:1390:23: error: expected expression
return (__m128i)__in;
^
H:/LLVM/build/bin/Debug/../lib/clang/3.3/include\emmintrin.h:1396:23: error: expected expression
return (__m128d)__in;
^
Complete output from Clang: http://pastebin.com/qi87K8qr
Clang tries to use MSVC headers but it doesn't work. Maybe I should use libc++ or libstdc++ instead, but how to do that?
Note I'm not interested in precompiled clang binaries
Yes, clang simply does not support all of Microsoft's extended C++ syntax, and therefore cannot parse Microsoft's C++ headers which use that syntax. Not only that but Clang also doesn't have complete support for Microsoft's C++ ABI, name mangling, etc. I believe Clang on Windows works alright with C, however.
To use a different C++ standard library instead you can make clang ignore the normal header and library directories with, IIRC, -nostdinc++ and -nostdlib++. Then you can tell clang the include and library directories you want to use (using -isystem or -I or whatever). However I'm not sure whether libc++ or libstdc++ work under those circumstances, since they probably depend on things that the Windows C runtime library does not have.
Chandler Carruth mentioned at Going Native 2013 that there are now alpha builds of clang for Windows with Visual Studio integration. Lots of stuff is broken, for example, streams (so good old hello world won't work). However, there is a lot of effort being put into making clang work on Windows, so expect it to get pretty good pretty fast.
Errors were in the header supplied with clang itself. Looks like it can't handle MMX/SSE types properly. Try to add -msse -msse2 switches to the command line.
I'm using libstdc++ and built clang using VS2012Express for desktop. The cmake string was "Visual Studio 11 Win64" and the essential dirs. are specified using -I argument.
My guess you program could work if I used mingw headers for Windows.
When running nvcc, it always uses the Visual C++ compiler (cl.exe). How can I make it use the GCC compiler?
Setting the CC environment-variable to gcc didn't fix it. I also couldn't find any option for this in the executeables help-output.
On Windows, NVCC only supports the Visual C++ compiler (cl.exe) for host compilation.
You can of course compile .cpp (non-CUDA) code using GCC and link the objects with objects generated by nvcc.