#pragma clang diagnostic ignored "-Wmissing-prototypes"

#include <metal_stdlib>
#include <simd/simd.h>

using namespace metal;

template<typename T, uint A>
inline void spvArrayCopyFromConstantToStack1(thread T (&dst)[A], constant T (&src)[A])
{
    for (uint i = 0; i < A; i++)
    {
        dst[i] = src[i];
    }
}

template<typename T, uint A>
inline void spvArrayCopyFromConstantToThreadGroup1(threadgroup T (&dst)[A], constant T (&src)[A])
{
    for (uint i = 0; i < A; i++)
    {
        dst[i] = src[i];
    }
}

template<typename T, uint A>
inline void spvArrayCopyFromStackToStack1(thread T (&dst)[A], thread const T (&src)[A])
{
    for (uint i = 0; i < A; i++)
    {
        dst[i] = src[i];
    }
}

template<typename T, uint A>
inline void spvArrayCopyFromStackToThreadGroup1(threadgroup T (&dst)[A], thread const T (&src)[A])
{
    for (uint i = 0; i < A; i++)
    {
        dst[i] = src[i];
    }
}

template<typename T, uint A>
inline void spvArrayCopyFromThreadGroupToStack1(thread T (&dst)[A], threadgroup const T (&src)[A])
{
    for (uint i = 0; i < A; i++)
    {
        dst[i] = src[i];
    }
}

template<typename T, uint A>
inline void spvArrayCopyFromThreadGroupToThreadGroup1(threadgroup T (&dst)[A], threadgroup const T (&src)[A])
{
    for (uint i = 0; i < A; i++)
    {
        dst[i] = src[i];
    }
}

template<typename T, uint A>
inline void spvArrayCopyFromDeviceToDevice1(device T (&dst)[A], device const T (&src)[A])
{
    for (uint i = 0; i < A; i++)
    {
        dst[i] = src[i];
    }
}

template<typename T, uint A>
inline void spvArrayCopyFromConstantToDevice1(device T (&dst)[A], constant T (&src)[A])
{
    for (uint i = 0; i < A; i++)
    {
        dst[i] = src[i];
    }
}

template<typename T, uint A>
inline void spvArrayCopyFromStackToDevice1(device T (&dst)[A], thread const T (&src)[A])
{
    for (uint i = 0; i < A; i++)
    {
        dst[i] = src[i];
    }
}

template<typename T, uint A>
inline void spvArrayCopyFromThreadGroupToDevice1(device T (&dst)[A], threadgroup const T (&src)[A])
{
    for (uint i = 0; i < A; i++)
    {
        dst[i] = src[i];
    }
}

template<typename T, uint A>
inline void spvArrayCopyFromDeviceToStack1(thread T (&dst)[A], device const T (&src)[A])
{
    for (uint i = 0; i < A; i++)
    {
        dst[i] = src[i];
    }
}

template<typename T, uint A>
inline void spvArrayCopyFromDeviceToThreadGroup1(threadgroup T (&dst)[A], device const T (&src)[A])
{
    for (uint i = 0; i < A; i++)
    {
        dst[i] = src[i];
    }
}

struct Block
{
    uint2 _m0[2];
    uint2 _m1[2];
};

struct SSBO
{
    Block _m0[3];
};

kernel void main0(device SSBO& ssbo [[buffer(0)]], constant SSBO& ubo [[buffer(1)]])
{
    threadgroup uint2 _18[2];
    spvArrayCopyFromDeviceToDevice1(ssbo._m0[0u]._m0, ssbo._m0[0u]._m1);
    spvArrayCopyFromConstantToDevice1(ssbo._m0[0u]._m0, ubo._m0[0u]._m1);
    uint2 _23[2];
    spvArrayCopyFromStackToDevice1(ssbo._m0[0u]._m0, _23);
    spvArrayCopyFromThreadGroupToDevice1(ssbo._m0[0u]._m0, _18);
    spvArrayCopyFromDeviceToThreadGroup1(_18, ssbo._m0[0u]._m1);
    spvArrayCopyFromDeviceToStack1(_23, ssbo._m0[0u]._m1);
    spvArrayCopyFromConstantToThreadGroup1(_18, ubo._m0[0u]._m1);
    spvArrayCopyFromConstantToStack1(_23, ubo._m0[0u]._m1);
}