[cfe-dev] Clang and CUDA with C++11 features

[Peter Colberg]

On Wed, Jun 13, 2012 at 07:30:25PM -0700, Richard Smith wrote:
> On Wed, Jun 13, 2012 at 6:57 PM, Peter Colberg <peter at colberg.org> wrote:
> > Hi,
> >
> > I am experimenting with CUDA language support in Clang, and so far
> > the kernel-call test in Clang trunk (r158426) compiles and runs, on
> > a Tesla C2050.
> >
> > Now I would like to enable C++11 support in Clang to use compile-time
> > C++11 features in GPU code, primarily variadic templates and lambda
> > functions.
> > …
> > Is there a way to tell Clang to enable C++11 extensions?
> 
> Sadly no, it appears we enable CUDA features based on whether we're
> using -std=cuda, not whether we're using -x cuda, so even though we
> accept -x cuda -std=c++11, that diasbles CUDA support! I expect we'd
> accept a patch to fix that :-)
> 
> For your own experimentation, try modifying
> include/clang/Frontend/LangStandards.def as follows:
> 
> --- include/clang/Frontend/LangStandards.def    (revision 158416)
> +++ include/clang/Frontend/LangStandards.def    (working copy)
> @@ -115,6 +115,6 @@
>  // CUDA
>  LANGSTANDARD(cuda, "cuda",
>               "NVIDIA CUDA(tm)",
> -             BCPLComment | CPlusPlus | Digraphs)
> +             BCPLComment | CPlusPlus | CPlusPlus0x | Digraphs)
> 
>  #undef LANGSTANDARD

Thanks, Richard and Sean, indeed that works! I could compile a
__attribute__((global)) function calling a local lambda function,
and a variadic function with __attribute__((device)).

There seems to be an issue with C++11 headers, e.g. <iostream>

    clang++ -I/usr/local/cuda-4.2/cuda/include -L/usr/local/cuda-4.2/cuda/lib64 -lcudart -o kernel-call kernel-call.cu 
    In file included from kernel-call.cu:7:
    In file included from /home/peter/usr/rhel6-x86_64/gcc-4.7.0/lib/gcc/x86_64-unknown-linux-gnu/4.7.0/../../../../include/c++/4.7.0/iostream:39:
    In file included from /home/peter/usr/rhel6-x86_64/gcc-4.7.0/lib/gcc/x86_64-unknown-linux-gnu/4.7.0/../../../../include/c++/4.7.0/ostream:39:
    In file included from /home/peter/usr/rhel6-x86_64/gcc-4.7.0/lib/gcc/x86_64-unknown-linux-gnu/4.7.0/../../../../include/c++/4.7.0/ios:40:
    In file included from /home/peter/usr/rhel6-x86_64/gcc-4.7.0/lib/gcc/x86_64-unknown-linux-gnu/4.7.0/../../../../include/c++/4.7.0/bits/char_traits.h:40:
    In file included from /home/peter/usr/rhel6-x86_64/gcc-4.7.0/lib/gcc/x86_64-unknown-linux-gnu/4.7.0/../../../../include/c++/4.7.0/bits/stl_algobase.h:65:
    In file included from /home/peter/usr/rhel6-x86_64/gcc-4.7.0/lib/gcc/x86_64-unknown-linux-gnu/4.7.0/../../../../include/c++/4.7.0/bits/stl_pair.h:61:
    In file included from /home/peter/usr/rhel6-x86_64/gcc-4.7.0/lib/gcc/x86_64-unknown-linux-gnu/4.7.0/../../../../include/c++/4.7.0/bits/move.h:57:
    /home/peter/usr/rhel6-x86_64/gcc-4.7.0/lib/gcc/x86_64-unknown-linux-gnu/4.7.0/../../../../include/c++/4.7.0/type_traits:441:38: error: expected '>'
                              is_void<_Tp>>>::type
                                         ^
    /home/peter/usr/rhel6-x86_64/gcc-4.7.0/lib/gcc/x86_64-unknown-linux-gnu/4.7.0/../../../../include/c++/4.7.0/type_traits:441:38: error: expected a type
    /home/peter/usr/rhel6-x86_64/gcc-4.7.0/lib/gcc/x86_64-unknown-linux-gnu/4.7.0/../../../../include/c++/4.7.0/type_traits:442:8: error: expected a type
        { };
           ^
    /home/peter/usr/rhel6-x86_64/gcc-4.7.0/lib/gcc/x86_64-unknown-linux-gnu/4.7.0/../../../../include/c++/4.7.0/type_traits:442:8: error: expected class
          name
    /home/peter/usr/rhel6-x86_64/gcc-4.7.0/lib/gcc/x86_64-unknown-linux-gnu/4.7.0/../../../../include/c++/4.7.0/type_traits:442:8: error: expected
          '{' after base class list
    /home/peter/usr/rhel6-x86_64/gcc-4.7.0/lib/gcc/x86_64-unknown-linux-gnu/4.7.0/../../../../include/c++/4.7.0/type_traits:563:61: error: expected '>'
        : public __and_<is_arithmetic<_Tp>, __not_<is_signed<_Tp>>>::type
                                                                ^
    /home/peter/usr/rhel6-x86_64/gcc-4.7.0/lib/gcc/x86_64-unknown-linux-gnu/4.7.0/../../../../include/c++/4.7.0/type_traits:563:61: error: expected a type
    /home/peter/usr/rhel6-x86_64/gcc-4.7.0/lib/gcc/x86_64-unknown-linux-gnu/4.7.0/../../../../include/c++/4.7.0/type_traits:564:8: error: expected a type
        { };
    …


The parser interprets the compressed C++11 template parameter syntax
as a call to a CUDA kernel function. Is there a way to disable parsing
of the CUDA call syntax <<< >>>? I would be using a C++ wrapper around
cudaConfigureCall, cudaSetupArgument and cudaLaunch anyway.


Besides, this is very impressive. C++11 on GPUs appears to be near :-).

Peter