[require]
shader model >= 6.0
wave-ops
format r32-uint uav-load

[uav 0]
format r32-uint
size (buffer, 4)

8 15 8 10

[uav 1]
format r32-uint
size (buffer, 8)

0 0 0 0 0 0 0 0

[compute shader]
RWBuffer<uint> u1 : register(u1);

    [numthreads(4, 1, 1)]
void main(uint id : SV_GroupIndex)
{
    u1[id] = WaveIsFirstLane();
}

[test]
dispatch 4 1 1
probe uav 1 (0) rui (1)
probe uav 1 (1) rui (0)
probe uav 1 (2) rui (0)
probe uav 1 (3) rui (0)


[compute shader]
RWBuffer<uint> u1 : register(u1);

    [numthreads(4, 1, 1)]
void main(uint id : SV_GroupIndex)
{
    if (!id)
    {
        u1[0] = 0;
        return;
    }
    u1[id] = WaveIsFirstLane();
}

[test]
dispatch 4 1 1
probe uav 1 (0) rui (0)
probe uav 1 (1) rui (1)
probe uav 1 (2) rui (0)
probe uav 1 (3) rui (0)


[compute shader]
RWBuffer<uint> u1 : register(u1);

    [numthreads(4, 1, 1)]
void main(uint id : SV_GroupIndex)
{
    u1[id] = WaveGetLaneIndex();
}

[test]
dispatch 4 1 1
probe uav 1 (0) rui (0)
probe uav 1 (1) rui (1)
probe uav 1 (2) rui (2)
probe uav 1 (3) rui (3)


[compute shader]
RWBuffer<uint> u1 : register(u1);

    [numthreads(4, 1, 1)]
void main(uint id : SV_GroupIndex)
{
    // Between 4 and 128, and includes inactive and/or helper lanes.
    uint count = WaveGetLaneCount();
    u1[id] = count >= 4 && count <= 128;
}

[test]
dispatch 4 1 1
probe uav 1 (0) rui (1)
probe uav 1 (1) rui (1)
probe uav 1 (2) rui (1)
probe uav 1 (3) rui (1)


[compute shader]
RWBuffer<uint> u0;
RWBuffer<uint> u1;

    [numthreads(4, 1, 1)]
void main(uint id : SV_GroupIndex)
{
    uint i = (id & 1) ? -u0[id] : u0[id];
    u1[id] = WaveActiveMin(i);
}

[test]
dispatch 4 1 1
probe uav 1 (0) rui (8)
probe uav 1 (1) rui (8)
probe uav 1 (2) rui (8)
probe uav 1 (3) rui (8)


[compute shader]
RWBuffer<uint> u0;
RWBuffer<uint> u1;

    [numthreads(4, 1, 1)]
void main(uint id : SV_GroupIndex)
{
    bool b = u0[id] == 8;
    u1[id] = WaveActiveBallot(b).x;
}

[test]
dispatch 4 1 1
probe uav 1 (0) rui (5)
probe uav 1 (1) rui (5)
probe uav 1 (2) rui (5)
probe uav 1 (3) rui (5)


[compute shader]
RWBuffer<uint> u0;
RWBuffer<uint> u1;

    [numthreads(4, 1, 1)]
void main(uint id : SV_GroupIndex)
{
    uint i = u0[id] + u0[id ^ 1];
    u1[id] = WaveReadLaneFirst(i);
}

[test]
dispatch 4 1 1
probe uav 1 (0) rui (23)
probe uav 1 (1) rui (23)
probe uav 1 (2) rui (23)
probe uav 1 (3) rui (23)


[compute shader]
RWBuffer<uint> u0;
RWBuffer<uint> u1;

    [numthreads(4, 1, 1)]
void main(uint id : SV_GroupIndex)
{
    uint i = u0[id] + u0[id ^ 1];
    u1[id] = WaveReadLaneAt(i, 3);
}

[test]
dispatch 4 1 1
probe uav 1 (0) rui (18)
probe uav 1 (1) rui (18)
probe uav 1 (2) rui (18)
probe uav 1 (3) rui (18)


[compute shader]
RWBuffer<uint> u0;
RWBuffer<uint> u1;

    [numthreads(4, 1, 1)]
void main(uint id : SV_GroupIndex)
{
    uint i = u0[id] + u0[id ^ 1];
    // Non-uniform lane index
    u1[id] = WaveReadLaneAt(i, (id + 2) & 3);
}

[test]
dispatch 4 1 1
probe uav 1 (0) rui (18)
probe uav 1 (1) rui (18)
probe uav 1 (2) rui (23)
probe uav 1 (3) rui (23)


[compute shader]
RWBuffer<uint> u0;
RWBuffer<uint> u1;

    [numthreads(4, 1, 1)]
void main(uint id : SV_GroupIndex)
{
    bool b = u0[id] == 8;
    u1[id] = WaveActiveCountBits(b);
}

[test]
dispatch 4 1 1
probe uav 1 (0) rui (2)
probe uav 1 (1) rui (2)
probe uav 1 (2) rui (2)
probe uav 1 (3) rui (2)


[compute shader]
RWBuffer<uint> u0;
RWBuffer<uint> u1;

    [numthreads(4, 1, 1)]
void main(uint id : SV_GroupIndex)
{
    bool b = u0[id] == 8;
    u1[id] = WavePrefixCountBits(b);
}

[test]
dispatch 4 1 1
probe uav 1 (0) rui (0)
probe uav 1 (1) rui (1)
probe uav 1 (2) rui (1)
probe uav 1 (3) rui (2)


[compute shader]
RWBuffer<uint> u0;
RWBuffer<uint> u1;

    [numthreads(4, 1, 1)]
void main(uint id : SV_GroupIndex)
{
    u1[id] = WaveActiveSum(u0[id]);
}

[test]
dispatch 4 1 1
probe uav 1 (0) rui (41)
probe uav 1 (1) rui (41)
probe uav 1 (2) rui (41)
probe uav 1 (3) rui (41)


[compute shader]
RWBuffer<uint> u0;
RWBuffer<uint> u1;

    [numthreads(4, 1, 1)]
void main(uint id : SV_GroupIndex)
{
    u1[id] = WaveActiveBitAnd(u0[id]);
}

[test]
dispatch 4 1 1
probe uav 1 (0) rui (8)
probe uav 1 (1) rui (8)
probe uav 1 (2) rui (8)
probe uav 1 (3) rui (8)


[compute shader]
RWBuffer<uint> u0;
RWBuffer<uint> u1;

    [numthreads(4, 1, 1)]
void main(uint id : SV_GroupIndex)
{
    u1[id] = WaveActiveBitOr(u0[id]);
}

[test]
dispatch 4 1 1
probe uav 1 (0) rui (15)
probe uav 1 (1) rui (15)
probe uav 1 (2) rui (15)
probe uav 1 (3) rui (15)


[compute shader]
RWBuffer<uint> u0;
RWBuffer<uint> u1;

    [numthreads(4, 1, 1)]
void main(uint id : SV_GroupIndex)
{
    u1[id] = WaveActiveBitXor(u0[id]);
}

[test]
dispatch 4 1 1
probe uav 1 (0) rui (5)
probe uav 1 (1) rui (5)
probe uav 1 (2) rui (5)
probe uav 1 (3) rui (5)


[compute shader]
uniform uint2 u;

RWBuffer<uint> u0;
RWBuffer<uint> u1;

    [numthreads(4, 1, 1)]
void main(uint id : SV_GroupIndex)
{
    uint table[] = {u.x, u.y};
    uint i;

    // Value depends on control flow.
    if (id & 1)
        i = table[id / 2u];
    else
        i = u0[id / 2u];

    u1[id] = WaveActiveSum(i);
}

[test]
uniform 0 uint4 1 5 0 0
dispatch 4 1 1
probe uav 1 (0) rui (29)
probe uav 1 (1) rui (29)
probe uav 1 (2) rui (29)
probe uav 1 (3) rui (29)