Skip to content

CUDA nvcc compilation error #10

Open

Description

@Char-Aznable

Hi, I got this very simple Taocpp code:

#include <iostream>
#include <tao/tuple/tuple.hpp>

int main(int argc, char* argv[]) {
   
  auto t2 = tao::tuple< int, double, int >( 1, 2, 3 );

  
  tao::get< 0 >( t2 ) = 0;
  tao::get< 1 >( t2 ) = 2;
  tao::get< 2 >( t2 ) = 3;

}

with the dependency: https://github.com/taocpp/tuple.git cloned to ./_deps/taotuple-src. Then I compile it with nvcc 11.4.48 and g++ 9.3.0 by

nvcc -ccbin g++-9 -I_deps/taotuple-src/include -x cu testTuple.cpp -o testTuple

and get this compilation error:

testTuple.cpp(19): error: no instance of overloaded function "tao::get" matches the argument list
            argument types are: (tao::tuple<int, double, int>)

testTuple.cpp(20): error: no instance of overloaded function "tao::get" matches the argument list
            argument types are: (tao::tuple<int, double, int>)

testTuple.cpp(21): error: no instance of overloaded function "tao::get" matches the argument list
            argument types are: (tao::tuple<int, double, int>)

I am using a5f3b87 on the main branch.

Activity

  1. d-frey commented on Sep 4, 2021

    @d-frey
    Member

    Since I'm not using CUDA myself, there isn't much that I can do. You could try to update to the latest commit, see if that helps. Or maybe @allisonvacanti might help? She developed the original PR for CUDA support. Allison: Let me know if it is OK for you if I tag you when CUDA-related issues with taocpp/tuple come up. If not, I will stop doing so.

  2. Char-Aznable commented on Sep 4, 2021

    @Char-Aznable
    Author

    The problem persists in the latest commit a5f3b87

  3. DarthRubik commented on Sep 6, 2021

    @DarthRubik
    Contributor

    @d-frey It appears maybe somebody fixed it in this fork?:

    https://github.com/maudoin/tuple

  4. d-frey commented on Sep 6, 2021

    @d-frey
    Member

    @Char-Aznable Could you send the complete output from the compiler, please?

  5. d-frey commented on Sep 6, 2021

    @d-frey
    Member

    @DarthRubik That is #6, where I hoped that my patch also fixed the issue. Apparently that is not the case, so I'll see what I can do. Can you also reproduce the error? If so, I'd like to know which version of GCC, nvcc, etc. you are using and also the complete error message from the compiler with all the details (candidates, why rejected, etc.)

  6. DarthRubik commented on Sep 6, 2021

    @DarthRubik
    Contributor

    @d-frey I do not have a CUDA environment set up and I cannot reproduce the error

  7. Char-Aznable commented on Sep 9, 2021

    @Char-Aznable
    Author

    Could you send the complete output from the compiler, please?

    Those I posted were all I got from it

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions