C++ Standard library stack usage. Trouble pushing float array - c++

Question

I am having trouble pushing an array to the stack. I thought this was pretty straightforward, but I've already spent too much time trying to figure this one out.
I expected to be able to push arrays just as I push ints or floats, but that is not happening.
The push command is giving me the issue. Here is my code:
#include <iostream>
#include <stack>
struct Matrix4x4
{
float data[16];
};
int main(int argc, char **argv)
{
// My original code
typedef std::stack<float[16]> myStack;
myStack modelViewStack;
myStack projectionStack;
float testMat[16] = {0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15};
modelViewStack.push(testMat); // THIS LINE GIVES ME ERRORS
////Stack initialization - This is thokra's solution
////typedef std::stack<std::vector<float[16]>> myStack;
//typedef std::stack<Matrix4x4> myStack;
//myStack modelViewStack;
//myStack projectionStack;
//
//Matrix4x4 m = {0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15};
////std::vector<float> testMat2(testMat, testMat + sizeof(testMat) / sizeof(float));
//modelViewStack.push(m);
//for(int i = 0; i<16 ; i++)
//{
// std::cout << "m data: " << m.data[i] << std::endl;
//}
//system("pause");
return 0;
}
Thanks for your help!
Here are the errors. I can't decipher them. Maybe an explanation of how to read these would be helpful too.
1>------ Build started: Project: opengl4_4, Configuration: Release Win32 ------
1> main.cpp
1>C:\Program Files (x86)\Microsoft Visual Studio 11.0\VC\include\xmemory0(606): error C2075: 'Target of operator new()' : array initialization needs curly braces
1> C:\Program Files (x86)\Microsoft Visual Studio 11.0\VC\include\xmemory0(605) : while compiling class template member function 'void std::allocator<_Ty>::construct(_Ty (*),const _Ty (&))'
1> with
1> [
1> _Ty=float [16]
1> ]
1> C:\Program Files (x86)\Microsoft Visual Studio 11.0\VC\include\xmemory0(751) : see reference to function template instantiation 'void std::allocator<_Ty>::construct(_Ty (*),const _Ty (&))' being compiled
1> with
1> [
1> _Ty=float [16]
1> ]
1> C:\Program Files (x86)\Microsoft Visual Studio 11.0\VC\include\type_traits(743) : see reference to class template instantiation 'std::allocator<_Ty>' being compiled
1> with
1> [
1> _Ty=float [16]
1> ]
1> C:\Program Files (x86)\Microsoft Visual Studio 11.0\VC\include\deque(925) : see reference to class template instantiation 'std::is_empty<_Ty>' being compiled
1> with
1> [
1> _Ty=std::allocator<float [16]>
1> ]
1> C:\Program Files (x86)\Microsoft Visual Studio 11.0\VC\include\stack(21) : see reference to class template instantiation 'std::deque<_Ty>' being compiled
1> with
1> [
1> _Ty=float [16]
1> ]
1> main.cpp(14) : see reference to class template instantiation 'std::stack<_Ty>' being compiled
1> with
1> [
1> _Ty=float [16]
1> ]
1>C:\Program Files (x86)\Microsoft Visual Studio 11.0\VC\include\xmemory0(606): fatal error C1903: unable to recover from previous error(s); stopping compilation
========== Build: 0 succeeded, 1 failed, 0 up-to-date, 0 skipped ==========

Answer 1

Just encapsulate the data store holding the elements of the matrix in a suitable type:
#include <stack>
struct Matrix4x4
{
float data[16];
};
int main()
{
typedef std::stack<Matrix4x4> myStack;
myStack modelViewStack;
Matrix4x4 m = {0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15};
modelViewStack.push(m);
return 0;
}
More to the point: std::stack::push will internally call push_back on the std::deque that's used as the container when you don't change the default. Essentially, when trying to construct a new element at the end of the deque the container tries to place the new element at the address which currently marks the end of the container with a placement-new. For instance, g++ implements it as follows:
template<typename _Up, typename... _Args>
void
construct(_Up* __p, _Args&&... __args)
{ ::new((void *)__p) _Up(std::forward<_Args>(__args)...); }
This effectively boils down to:
::new((void *)__p) float[16](_args); // where __p is a pointer to float[16] and _args is testMat
Trying to copy- or move-initialize a C-array is simply not legal. Even if construction succeeded for some reason, the container would try to call the destructor on an element of type float[16] when going out of scope. One can easily see, a destructor ~T[n] does not exist.
In C++11, you can push a std::array<float,16> instead of defining an additional type.