ambiguous call to ref in vector_of_vector.hpp when using bind.hpp

[Will Moss <wmoss@…>]

When bind.hpp and numeric/ublas/vector_of_vector.hpp are both included, instantiating and using a generalized_vector_of_vector causes an ambiguous call to the overloaded function 'boost::numeric::ublas::ref'. See attached code for a simple example and attached patch for the simple proposed fix.

[Gunter]

Sorry, I cannot confirm this using g++ 4.1.2. Which compiler did you use?

Actually the correct ref() function should be found by Koenig-lookup.

[Gunter]

I close this issue because I did not get any comment on this and I cannot recreate it.

[sspa9999@…]

compiler error log

[anonymous]

Hi there,

I can reproduce the error with Visual C++ 2012. Attached gvov.log is the VC++ error log for compiling the previously attached ublas_bug.cpp. The issue exists even without bind.hpp in ublas_bug.cpp. Note the same code actually compiles successfully with Visual C++ 2008, or recent g++.

Thanks, Steven

[anonymous]

I have this problem both on Windows with MSVC 2013 and on Mac with the default clang. Since the fix is trivial, I don't see why it shouldn't be implemented.

[harrysummerxr@…]

I'm using Boost 1.56 + MSVC2013. I also this problem, even without including bind.hpp. The patch vector_of_vector.patch works like a charm. Here is my test code:

#include "boost/numeric/ublas/matrix.hpp"
#include "boost/numeric/ublas/matrix_sparse.hpp"
#include "boost/numeric/ublas/vector_of_vector.hpp"

int main(int argc, char** argv)
{
    boost::numeric::ublas::generalized_vector_of_vector<double,
                                                        boost::numeric::ublas::row_major,
                                                        boost::numeric::ublas::vector<boost::numeric::ublas::compressed_vector<double> > > gvov(10, 10);
    gvov.clear();
    return 0;
}

My first compile error:

C:\Boost\include\boost-1_56\boost/numeric/ublas/vector_of_vector.hpp(326) : error C2668: 'boost::numeric::ublas::ref' : ambiguous call to overloaded function
        C:\Boost\include\boost-1_56\boost/numeric/ublas/vector_sparse.hpp(178): could be 'boost::numeric::ublas::compressed_vector<double,0x00,boost::numeric::ublas::unbounded_array<size_t,std::allocator<T>>,boost::numeric::ublas::unbounded_array<double,std::allocator<double>>> &boost::numeric::ublas::ref<boost::numeric::ublas::compressed_vector<double,0x00,boost::numeric::ublas::unbounded_array<T,std::allocator<T>>,boost::numeric::ublas::unbounded_array<double,std::allocator<double>>>>(const VER &)'
        with
        [
            T=size_t
,            VER=boost::numeric::ublas::compressed_vector<double,0x00,boost::numeric::ublas::unbounded_array<size_t,std::allocator<size_t>>,boost::numeric::ublas::unbounded_array<double,std::allocator<double>>>
        ]
        C:\Boost\include\boost-1_56\boost/numeric/ublas/vector_sparse.hpp(174): or       'boost::numeric::ublas::compressed_vector<double,0x00,boost::numeric::ublas::unbounded_array<size_t,std::allocator<T>>,boost::numeric::ublas::unbounded_array<double,std::allocator<double>>> &boost::numeric::ublas::ref<boost::numeric::ublas::compressed_vector<double,0x00,boost::numeric::ublas::unbounded_array<T,std::allocator<T>>,boost::numeric::ublas::unbounded_array<double,std::allocator<double>>>>(VER &)'
        with
        [
            T=size_t
,            VER=boost::numeric::ublas::compressed_vector<double,0x00,boost::numeric::ublas::unbounded_array<size_t,std::allocator<size_t>>,boost::numeric::ublas::unbounded_array<double,std::allocator<double>>>
        ]
        C:\Program Files (x86)\Microsoft Visual Studio 12.0\VC\INCLUDE\xrefwrap(697): or       'void std::ref<T>(const _Ty &&)' [found using argument-dependent lookup]
        with
        [
            T=boost::numeric::ublas::compressed_vector<double,0x00,boost::numeric::ublas::unbounded_array<size_t,std::allocator<size_t>>,boost::numeric::ublas::unbounded_array<double,std::allocator<double>>>
,            _Ty=boost::numeric::ublas::compressed_vector<double,0x00,boost::numeric::ublas::unbounded_array<size_t,std::allocator<size_t>>,boost::numeric::ublas::unbounded_array<double,std::allocator<double>>>
        ]
        C:\Program Files (x86)\Microsoft Visual Studio 12.0\VC\INCLUDE\xrefwrap(691): or       'std::reference_wrapper<T> std::ref<T>(_Ty &) throw()' [found using argument-dependent lookup]
        with
        [
            T=boost::numeric::ublas::compressed_vector<double,0x00,boost::numeric::ublas::unbounded_array<size_t,std::allocator<size_t>>,boost::numeric::ublas::unbounded_array<double,std::allocator<double>>>
,            _Ty=boost::numeric::ublas::compressed_vector<double,0x00,boost::numeric::ublas::unbounded_array<size_t,std::allocator<size_t>>,boost::numeric::ublas::unbounded_array<double,std::allocator<double>>>
        ]
        while trying to match the argument list '(boost::numeric::ublas::compressed_vector<double,0x00,boost::numeric::ublas::unbounded_array<size_t,std::allocator<T>>,boost::numeric::ublas::unbounded_array<double,std::allocator<double>>>)'
        with
        [
            T=size_t
        ]
        C:\Boost\include\boost-1_56\boost/numeric/ublas/vector_of_vector.hpp(322) : while compiling class template member function 'void boost::numeric::ublas::generalized_vector_of_vector<double,boost::numeric::ublas::row_major,boost::numeric::ublas::vector<boost::numeric::ublas::compressed_vector<double,0x00,boost::numeric::ublas::unbounded_array<size_t,std::allocator<T>>,boost::numeric::ublas::unbounded_array<double,std::allocator<double>>>,boost::numeric::ublas::unbounded_array<boost::numeric::ublas::compressed_vector<double,0x00,boost::numeric::ublas::unbounded_array<T,std::allocator<T>>,boost::numeric::ublas::unbounded_array<double,std::allocator<double>>>,std::allocator<boost::numeric::ublas::compressed_vector<double,0x00,boost::numeric::ublas::unbounded_array<T,std::allocator<T>>,boost::numeric::ublas::unbounded_array<double,std::allocator<double>>>>>>>::clear(void)'
        with
        [
            T=size_t
        ]
        D:\test\main.cpp(10) : see reference to function template instantiation 'void boost::numeric::ublas::generalized_vector_of_vector<double,boost::numeric::ublas::row_major,boost::numeric::ublas::vector<boost::numeric::ublas::compressed_vector<double,0x00,boost::numeric::ublas::unbounded_array<size_t,std::allocator<T>>,boost::numeric::ublas::unbounded_array<double,std::allocator<double>>>,boost::numeric::ublas::unbounded_array<boost::numeric::ublas::compressed_vector<double,0x00,boost::numeric::ublas::unbounded_array<T,std::allocator<T>>,boost::numeric::ublas::unbounded_array<double,std::allocator<double>>>,std::allocator<boost::numeric::ublas::compressed_vector<double,0x00,boost::numeric::ublas::unbounded_array<T,std::allocator<T>>,boost::numeric::ublas::unbounded_array<double,std::allocator<double>>>>>>>::clear(void)' being compiled
        with
        [
            T=size_t
        ]
        D:\test\main.cpp(9) : see reference to class template instantiation 'boost::numeric::ublas::generalized_vector_of_vector<double,boost::numeric::ublas::row_major,boost::numeric::ublas::vector<boost::numeric::ublas::compressed_vector<double,0x00,boost::numeric::ublas::unbounded_array<size_t,std::allocator<T>>,boost::numeric::ublas::unbounded_array<double,std::allocator<double>>>,boost::numeric::ublas::unbounded_array<boost::numeric::ublas::compressed_vector<double,0x00,boost::numeric::ublas::unbounded_array<T,std::allocator<T>>,boost::numeric::ublas::unbounded_array<double,std::allocator<double>>>,std::allocator<boost::numeric::ublas::compressed_vector<double,0x00,boost::numeric::ublas::unbounded_array<T,std::allocator<T>>,boost::numeric::ublas::unbounded_array<double,std::allocator<double>>>>>>>' being compiled
        with
        [
            T=size_t
        ]

[pa@…]

Just a heads up: The bug is also present in 1.59 compiled with

$ gcc -v
Configured with: --prefix=/Library/Developer/CommandLineTools/usr --with-gxx-include-dir=/usr/include/c++/4.2.1
Apple LLVM version 7.0.2 (clang-700.1.81)
Target: x86_64-apple-darwin15.2.0
Thread model: posix

(the default "gcc" in MacOS X El Capitain).

The conflict here is again with a system library

/Library/Developer/CommandLineTools/usr/bin/../include/c++/v1/__functional_base_03:1115:1: note: 
      candidate function [with _Tp =
      boost::numeric::ublas::mapped_vector<double,
      boost::numeric::ublas::map_std<unsigned long, double,
      std::__1::allocator<std::__1::pair<const unsigned long, double> > > >]
ref(_Tp& __t)

The simple 8 year old patch applies flawlessly, and solves the problem.

[anonymous]

Please implement this simple fix for this annoying problem. The same thing is flagged by g++ versions 4.8.5 and 5.3.0 on gnu/linux systems.

[anonymous]

I can confirm that this bug exists on Windows using the latest ms compiler. I can also confirm that the provided patch DOES fix the issue. Any info on when this will be fixed?