#version 450 core
#extension GL_KHR_memory_scope_semantics : enable
#extension GL_EXT_long_vector : enable
#extension GL_EXT_shader_explicit_arithmetic_types : enable
#extension GL_EXT_buffer_reference : enable
#extension GL_EXT_expect_assume : enable
#extension GL_EXT_bfloat16 : enable
#extension GL_KHR_shader_subgroup_arithmetic : enable
#extension GL_KHR_shader_subgroup_clustered : enable
#extension GL_NV_shader_subgroup_partitioned : enable
#extension GL_KHR_shader_subgroup_quad : enable
#extension GL_KHR_shader_subgroup_shuffle : enable
#extension GL_KHR_shader_subgroup_shuffle_relative : enable
#extension GL_KHR_shader_subgroup_rotate : enable
#extension GL_KHR_shader_subgroup_vote : enable
#extension GL_KHR_shader_subgroup_ballot : enable
#extension GL_EXT_shader_subgroup_extended_types_int64 : enable




layout (local_size_x = 64, local_size_y = 1, local_size_z = 1) in;

void main()
{
    vector<float16_t, 5> vf16;
    vector<float, 5> vf32;
    vector<float64_t, 5> vf64;
    vector<uint16_t, 5> vu16;
    vector<int16_t, 5> vi16;
    vector<uint32_t, 5> vu32;
    vector<int32_t, 5> vi32;
    vector<uint64_t, 5> vu64;
    vector<int64_t, 5> vi64;
    vector<bool, 5> vb;

    bool b;
    uint u;
    uvec4 vu4;

    b = subgroupAllEqual(vu32);
    vu32 = subgroupBroadcast(vu32, 1u);
    vu32 = subgroupBroadcastFirst(vu32);
    vu32 = subgroupShuffle(vu32, u);
    vu32 = subgroupShuffleXor(vu32, u);
    vu32 = subgroupShuffleUp(vu32, u);
    vu32 = subgroupShuffleDown(vu32, u);
    vu32 = subgroupRotate(vu32, u);
    vu32 = subgroupClusteredRotate(vu32, u, 4);
    vu32 = subgroupAdd(vu32);
    vu32 = subgroupMul(vu32);
    vu32 = subgroupMin(vu32);
    vu32 = subgroupMax(vu32);
    vu32 = subgroupAnd(vu32);
    vu32 = subgroupOr(vu32);
    vu32 = subgroupXor(vu32);
    vu32 = subgroupInclusiveAdd(vu32);
    vu32 = subgroupInclusiveMul(vu32);
    vu32 = subgroupInclusiveMin(vu32);
    vu32 = subgroupInclusiveMax(vu32);
    vu32 = subgroupInclusiveAnd(vu32);
    vu32 = subgroupInclusiveOr(vu32);
    vu32 = subgroupInclusiveXor(vu32);
    vu32 = subgroupExclusiveAdd(vu32);
    vu32 = subgroupExclusiveMul(vu32);
    vu32 = subgroupExclusiveMin(vu32);
    vu32 = subgroupExclusiveMax(vu32);
    vu32 = subgroupExclusiveAnd(vu32);
    vu32 = subgroupExclusiveOr(vu32);
    vu32 = subgroupExclusiveXor(vu32);
    vu32 = subgroupClusteredAdd(vu32, 4);
    vu32 = subgroupClusteredMul(vu32, 4);
    vu32 = subgroupClusteredMin(vu32, 4);
    vu32 = subgroupClusteredMax(vu32, 4);
    vu32 = subgroupClusteredAnd(vu32, 4);
    vu32 = subgroupClusteredOr(vu32, 4);
    vu32 = subgroupClusteredXor(vu32, 4);
    vu32 = subgroupQuadBroadcast(vu32, 1);
    vu32 = subgroupQuadSwapHorizontal(vu32);
    vu32 = subgroupQuadSwapVertical(vu32);
    vu32 = subgroupQuadSwapDiagonal(vu32);
    vu4 = subgroupPartitionNV(vu32);
    vu32 = subgroupPartitionedAddNV(vu32, vu4);
    vu32 = subgroupPartitionedMulNV(vu32, vu4);
    vu32 = subgroupPartitionedMinNV(vu32, vu4);
    vu32 = subgroupPartitionedMaxNV(vu32, vu4);
    vu32 = subgroupPartitionedAndNV(vu32, vu4);
    vu32 = subgroupPartitionedOrNV(vu32, vu4);
    vu32 = subgroupPartitionedXorNV(vu32, vu4);
    vu32 = subgroupPartitionedInclusiveAddNV(vu32, vu4);
    vu32 = subgroupPartitionedInclusiveMulNV(vu32, vu4);
    vu32 = subgroupPartitionedInclusiveMinNV(vu32, vu4);
    vu32 = subgroupPartitionedInclusiveMaxNV(vu32, vu4);
    vu32 = subgroupPartitionedInclusiveAndNV(vu32, vu4);
    vu32 = subgroupPartitionedInclusiveOrNV(vu32, vu4);
    vu32 = subgroupPartitionedInclusiveXorNV(vu32, vu4);
    vu32 = subgroupPartitionedExclusiveAddNV(vu32, vu4);
    vu32 = subgroupPartitionedExclusiveMulNV(vu32, vu4);
    vu32 = subgroupPartitionedExclusiveMinNV(vu32, vu4);
    vu32 = subgroupPartitionedExclusiveMaxNV(vu32, vu4);
    vu32 = subgroupPartitionedExclusiveAndNV(vu32, vu4);
    vu32 = subgroupPartitionedExclusiveOrNV(vu32, vu4);
    vu32 = subgroupPartitionedExclusiveXorNV(vu32, vu4);

    b = subgroupAllEqual(vf32);
    vf32 = subgroupBroadcast(vf32, 1u);
    vf32 = subgroupBroadcastFirst(vf32);
    vf32 = subgroupShuffle(vf32, u);
    vf32 = subgroupShuffleXor(vf32, u);
    vf32 = subgroupShuffleUp(vf32, u);
    vf32 = subgroupShuffleDown(vf32, u);
    vf32 = subgroupRotate(vf32, u);
    vf32 = subgroupClusteredRotate(vf32, u, 4);
    vf32 = subgroupAdd(vf32);
    vf32 = subgroupMul(vf32);
    vf32 = subgroupMin(vf32);
    vf32 = subgroupMax(vf32);
    //vf32 = subgroupAnd(vf32);
    //vf32 = subgroupOr(vf32);
    //vf32 = subgroupXor(vf32);
    vf32 = subgroupInclusiveAdd(vf32);
    vf32 = subgroupInclusiveMul(vf32);
    vf32 = subgroupInclusiveMin(vf32);
    vf32 = subgroupInclusiveMax(vf32);
    //vf32 = subgroupInclusiveAnd(vf32);
    //vf32 = subgroupInclusiveOr(vf32);
    //vf32 = subgroupInclusiveXor(vf32);
    vf32 = subgroupExclusiveAdd(vf32);
    vf32 = subgroupExclusiveMul(vf32);
    vf32 = subgroupExclusiveMin(vf32);
    vf32 = subgroupExclusiveMax(vf32);
    //vf32 = subgroupExclusiveAnd(vf32);
    //vf32 = subgroupExclusiveOr(vf32);
    //vf32 = subgroupExclusiveXor(vf32);
    vf32 = subgroupClusteredAdd(vf32, 4);
    vf32 = subgroupClusteredMul(vf32, 4);
    vf32 = subgroupClusteredMin(vf32, 4);
    vf32 = subgroupClusteredMax(vf32, 4);
    //vf32 = subgroupClusteredAnd(vf32, 4);
    //vf32 = subgroupClusteredOr(vf32, 4);
    //vf32 = subgroupClusteredXor(vf32, 4);
    vf32 = subgroupQuadBroadcast(vf32, 1);
    vf32 = subgroupQuadSwapHorizontal(vf32);
    vf32 = subgroupQuadSwapVertical(vf32);
    vf32 = subgroupQuadSwapDiagonal(vf32);
    vu4 = subgroupPartitionNV(vf32);
    vf32 = subgroupPartitionedAddNV(vf32, vu4);
    vf32 = subgroupPartitionedMulNV(vf32, vu4);
    vf32 = subgroupPartitionedMinNV(vf32, vu4);
    vf32 = subgroupPartitionedMaxNV(vf32, vu4);
    //vf32 = subgroupPartitionedAndNV(vf32, vu4);
    //vf32 = subgroupPartitionedOrNV(vf32, vu4);
    //vf32 = subgroupPartitionedXorNV(vf32, vu4);
    vf32 = subgroupPartitionedInclusiveAddNV(vf32, vu4);
    vf32 = subgroupPartitionedInclusiveMulNV(vf32, vu4);
    vf32 = subgroupPartitionedInclusiveMinNV(vf32, vu4);
    vf32 = subgroupPartitionedInclusiveMaxNV(vf32, vu4);
    //vf32 = subgroupPartitionedInclusiveAndNV(vf32, vu4);
    //vf32 = subgroupPartitionedInclusiveOrNV(vf32, vu4);
    //vf32 = subgroupPartitionedInclusiveXorNV(vf32, vu4);
    vf32 = subgroupPartitionedExclusiveAddNV(vf32, vu4);
    vf32 = subgroupPartitionedExclusiveMulNV(vf32, vu4);
    vf32 = subgroupPartitionedExclusiveMinNV(vf32, vu4);
    vf32 = subgroupPartitionedExclusiveMaxNV(vf32, vu4);
    //vf32 = subgroupPartitionedExclusiveAndNV(vf32, vu4);
    //vf32 = subgroupPartitionedExclusiveOrNV(vf32, vu4);
    //vf32 = subgroupPartitionedExclusiveXorNV(vf32, vu4);

    b = subgroupAllEqual(vf64);
    vf64 = subgroupBroadcast(vf64, 1u);
    vf64 = subgroupBroadcastFirst(vf64);
    vf64 = subgroupShuffle(vf64, u);
    vf64 = subgroupShuffleXor(vf64, u);
    vf64 = subgroupShuffleUp(vf64, u);
    vf64 = subgroupShuffleDown(vf64, u);
    vf64 = subgroupRotate(vf64, u);
    vf64 = subgroupClusteredRotate(vf64, u, 4);
    vf64 = subgroupAdd(vf64);
    vf64 = subgroupMul(vf64);
    vf64 = subgroupMin(vf64);
    vf64 = subgroupMax(vf64);
    //vf64 = subgroupAnd(vf64);
    //vf64 = subgroupOr(vf64);
    //vf64 = subgroupXor(vf64);
    vf64 = subgroupInclusiveAdd(vf64);
    vf64 = subgroupInclusiveMul(vf64);
    vf64 = subgroupInclusiveMin(vf64);
    vf64 = subgroupInclusiveMax(vf64);
    //vf64 = subgroupInclusiveAnd(vf64);
    //vf64 = subgroupInclusiveOr(vf64);
    //vf64 = subgroupInclusiveXor(vf64);
    vf64 = subgroupExclusiveAdd(vf64);
    vf64 = subgroupExclusiveMul(vf64);
    vf64 = subgroupExclusiveMin(vf64);
    vf64 = subgroupExclusiveMax(vf64);
    //vf64 = subgroupExclusiveAnd(vf64);
    //vf64 = subgroupExclusiveOr(vf64);
    //vf64 = subgroupExclusiveXor(vf64);
    vf64 = subgroupClusteredAdd(vf64, 4);
    vf64 = subgroupClusteredMul(vf64, 4);
    vf64 = subgroupClusteredMin(vf64, 4);
    vf64 = subgroupClusteredMax(vf64, 4);
    //vf64 = subgroupClusteredAnd(vf64, 4);
    //vf64 = subgroupClusteredOr(vf64, 4);
    //vf64 = subgroupClusteredXor(vf64, 4);
    vf64 = subgroupQuadBroadcast(vf64, 1);
    vf64 = subgroupQuadSwapHorizontal(vf64);
    vf64 = subgroupQuadSwapVertical(vf64);
    vf64 = subgroupQuadSwapDiagonal(vf64);
    vu4 = subgroupPartitionNV(vf64);
    vf64 = subgroupPartitionedAddNV(vf64, vu4);
    vf64 = subgroupPartitionedMulNV(vf64, vu4);
    vf64 = subgroupPartitionedMinNV(vf64, vu4);
    vf64 = subgroupPartitionedMaxNV(vf64, vu4);
    //vf64 = subgroupPartitionedAndNV(vf64, vu4);
    //vf64 = subgroupPartitionedOrNV(vf64, vu4);
    //vf64 = subgroupPartitionedXorNV(vf64, vu4);
    vf64 = subgroupPartitionedInclusiveAddNV(vf64, vu4);
    vf64 = subgroupPartitionedInclusiveMulNV(vf64, vu4);
    vf64 = subgroupPartitionedInclusiveMinNV(vf64, vu4);
    vf64 = subgroupPartitionedInclusiveMaxNV(vf64, vu4);
    //vf64 = subgroupPartitionedInclusiveAndNV(vf64, vu4);
    //vf64 = subgroupPartitionedInclusiveOrNV(vf64, vu4);
    //vf64 = subgroupPartitionedInclusiveXorNV(vf64, vu4);
    vf64 = subgroupPartitionedExclusiveAddNV(vf64, vu4);
    vf64 = subgroupPartitionedExclusiveMulNV(vf64, vu4);
    vf64 = subgroupPartitionedExclusiveMinNV(vf64, vu4);
    vf64 = subgroupPartitionedExclusiveMaxNV(vf64, vu4);
    //vf64 = subgroupPartitionedExclusiveAndNV(vf64, vu4);
    //vf64 = subgroupPartitionedExclusiveOrNV(vf64, vu4);
    //vf64 = subgroupPartitionedExclusiveXorNV(vf64, vu4);

    b = subgroupAllEqual(vb);
    vb = subgroupBroadcast(vb, 1u);
    vb = subgroupBroadcastFirst(vb);
    vb = subgroupShuffle(vb, u);
    vb = subgroupShuffleXor(vb, u);
    vb = subgroupShuffleUp(vb, u);
    vb = subgroupShuffleDown(vb, u);
    vb = subgroupRotate(vb, u);
    vb = subgroupClusteredRotate(vb, u, 4);
    //vb = subgroupAdd(vb);
    //vb = subgroupMul(vb);
    //vb = subgroupMin(vb);
    //vb = subgroupMax(vb);
    vb = subgroupAnd(vb);
    vb = subgroupOr(vb);
    vb = subgroupXor(vb);
    //vb = subgroupInclusiveAdd(vb);
    //vb = subgroupInclusiveMul(vb);
    //vb = subgroupInclusiveMin(vb);
    //vb = subgroupInclusiveMax(vb);
    vb = subgroupInclusiveAnd(vb);
    vb = subgroupInclusiveOr(vb);
    vb = subgroupInclusiveXor(vb);
    //vb = subgroupExclusiveAdd(vb);
    //vb = subgroupExclusiveMul(vb);
    //vb = subgroupExclusiveMin(vb);
    //vb = subgroupExclusiveMax(vb);
    vb = subgroupExclusiveAnd(vb);
    vb = subgroupExclusiveOr(vb);
    vb = subgroupExclusiveXor(vb);
    //vb = subgroupClusteredAdd(vb, 4);
    //vb = subgroupClusteredMul(vb, 4);
    //vb = subgroupClusteredMin(vb, 4);
    //vb = subgroupClusteredMax(vb, 4);
    vb = subgroupClusteredAnd(vb, 4);
    vb = subgroupClusteredOr(vb, 4);
    vb = subgroupClusteredXor(vb, 4);
    vb = subgroupQuadBroadcast(vb, 1);
    vb = subgroupQuadSwapHorizontal(vb);
    vb = subgroupQuadSwapVertical(vb);
    vb = subgroupQuadSwapDiagonal(vb);
    vu4 = subgroupPartitionNV(vb);
    //vb = subgroupPartitionedAddNV(vb, vu4);
    //vb = subgroupPartitionedMulNV(vb, vu4);
    //vb = subgroupPartitionedMinNV(vb, vu4);
    //vb = subgroupPartitionedMaxNV(vb, vu4);
    vb = subgroupPartitionedAndNV(vb, vu4);
    vb = subgroupPartitionedOrNV(vb, vu4);
    vb = subgroupPartitionedXorNV(vb, vu4);
    //vb = subgroupPartitionedInclusiveAddNV(vb, vu4);
    //vb = subgroupPartitionedInclusiveMulNV(vb, vu4);
    //vb = subgroupPartitionedInclusiveMinNV(vb, vu4);
    //vb = subgroupPartitionedInclusiveMaxNV(vb, vu4);
    vb = subgroupPartitionedInclusiveAndNV(vb, vu4);
    vb = subgroupPartitionedInclusiveOrNV(vb, vu4);
    vb = subgroupPartitionedInclusiveXorNV(vb, vu4);
    //vb = subgroupPartitionedExclusiveAddNV(vb, vu4);
    //vb = subgroupPartitionedExclusiveMulNV(vb, vu4);
    //vb = subgroupPartitionedExclusiveMinNV(vb, vu4);
    //vb = subgroupPartitionedExclusiveMaxNV(vb, vu4);
    vb = subgroupPartitionedExclusiveAndNV(vb, vu4);
    vb = subgroupPartitionedExclusiveOrNV(vb, vu4);
    vb = subgroupPartitionedExclusiveXorNV(vb, vu4);

    b = subgroupAllEqual(vu64);
    vu64 = subgroupBroadcast(vu64, 1u);
    vu64 = subgroupBroadcastFirst(vu64);
    vu64 = subgroupShuffle(vu64, u);
    vu64 = subgroupShuffleXor(vu64, u);
    vu64 = subgroupShuffleUp(vu64, u);
    vu64 = subgroupShuffleDown(vu64, u);
    vu64 = subgroupRotate(vu64, u);
    vu64 = subgroupClusteredRotate(vu64, u, 4);
    vu64 = subgroupAdd(vu64);
    vu64 = subgroupMul(vu64);
    vu64 = subgroupMin(vu64);
    vu64 = subgroupMax(vu64);
    vu64 = subgroupAnd(vu64);
    vu64 = subgroupOr(vu64);
    vu64 = subgroupXor(vu64);
    vu64 = subgroupInclusiveAdd(vu64);
    vu64 = subgroupInclusiveMul(vu64);
    vu64 = subgroupInclusiveMin(vu64);
    vu64 = subgroupInclusiveMax(vu64);
    vu64 = subgroupInclusiveAnd(vu64);
    vu64 = subgroupInclusiveOr(vu64);
    vu64 = subgroupInclusiveXor(vu64);
    vu64 = subgroupExclusiveAdd(vu64);
    vu64 = subgroupExclusiveMul(vu64);
    vu64 = subgroupExclusiveMin(vu64);
    vu64 = subgroupExclusiveMax(vu64);
    vu64 = subgroupExclusiveAnd(vu64);
    vu64 = subgroupExclusiveOr(vu64);
    vu64 = subgroupExclusiveXor(vu64);
    vu64 = subgroupClusteredAdd(vu64, 4);
    vu64 = subgroupClusteredMul(vu64, 4);
    vu64 = subgroupClusteredMin(vu64, 4);
    vu64 = subgroupClusteredMax(vu64, 4);
    vu64 = subgroupClusteredAnd(vu64, 4);
    vu64 = subgroupClusteredOr(vu64, 4);
    vu64 = subgroupClusteredXor(vu64, 4);
    vu64 = subgroupQuadBroadcast(vu64, 1);
    vu64 = subgroupQuadSwapHorizontal(vu64);
    vu64 = subgroupQuadSwapVertical(vu64);
    vu64 = subgroupQuadSwapDiagonal(vu64);
    vu4 = subgroupPartitionNV(vu64);
    vu64 = subgroupPartitionedAddNV(vu64, vu4);
    vu64 = subgroupPartitionedMulNV(vu64, vu4);
    vu64 = subgroupPartitionedMinNV(vu64, vu4);
    vu64 = subgroupPartitionedMaxNV(vu64, vu4);
    vu64 = subgroupPartitionedAndNV(vu64, vu4);
    vu64 = subgroupPartitionedOrNV(vu64, vu4);
    vu64 = subgroupPartitionedXorNV(vu64, vu4);
    vu64 = subgroupPartitionedInclusiveAddNV(vu64, vu4);
    vu64 = subgroupPartitionedInclusiveMulNV(vu64, vu4);
    vu64 = subgroupPartitionedInclusiveMinNV(vu64, vu4);
    vu64 = subgroupPartitionedInclusiveMaxNV(vu64, vu4);
    vu64 = subgroupPartitionedInclusiveAndNV(vu64, vu4);
    vu64 = subgroupPartitionedInclusiveOrNV(vu64, vu4);
    vu64 = subgroupPartitionedInclusiveXorNV(vu64, vu4);
    vu64 = subgroupPartitionedExclusiveAddNV(vu64, vu4);
    vu64 = subgroupPartitionedExclusiveMulNV(vu64, vu4);
    vu64 = subgroupPartitionedExclusiveMinNV(vu64, vu4);
    vu64 = subgroupPartitionedExclusiveMaxNV(vu64, vu4);
    vu64 = subgroupPartitionedExclusiveAndNV(vu64, vu4);
    vu64 = subgroupPartitionedExclusiveOrNV(vu64, vu4);
    vu64 = subgroupPartitionedExclusiveXorNV(vu64, vu4);

}