// Copyright 2024 Tencent
// SPDX-License-Identifier: BSD-3-Clause

#include "testutil.h"

#if NCNN_INT8
static void RandomizeA(ncnn::Mat& m, int transA, float absmax)
{
    if (transA == 0)
    {
        const int h = m.dims == 3 ? m.c : m.h;
        for (int i = 0; i < h; i++)
        {
            float* p = m.dims == 3 ? m.channel(i) : m.row(i);
            float randabsmax = RandomFloat(absmax * 0.5f, absmax);
            randabsmax = ncnn::float16_to_float32(ncnn::float32_to_float16(randabsmax));
            randabsmax = ncnn::bfloat16_to_float32(ncnn::float32_to_bfloat16(randabsmax));

            for (int j = 0; j < m.w; j++)
            {
                p[j] = RandomFloat(-randabsmax, randabsmax);
            }

            // set random a and b
            p[RandomInt(0, m.w - 1)] = -randabsmax;
            p[RandomInt(0, m.w - 1)] = randabsmax;

            // drop 0.45 ~ 0.55
            for (int j = 0; j < m.w; j++)
            {
                float v = p[j] * (127.f / randabsmax);
                float vv = fabs(v - (int)v);

                float hp = ncnn::float16_to_float32(ncnn::float32_to_float16(p[j]));
                float hv = hp * (127.f / randabsmax);
                float hvv = fabs(hv - (int)hv);

                float bp = ncnn::bfloat16_to_float32(ncnn::float32_to_bfloat16(p[j]));
                float bv = bp * (127.f / randabsmax);
                float bvv = fabs(bv - (int)bv);

                while ((vv > 0.45f && vv < 0.55f) || (hvv > 0.45f && hvv < 0.55f) || (bvv > 0.45f && bvv < 0.55f))
                {
                    p[j] = RandomFloat(-randabsmax, randabsmax);
                    v = p[j] * (127.f / randabsmax);
                    vv = fabs(v - (int)v);

                    hp = ncnn::float16_to_float32(ncnn::float32_to_float16(p[j]));
                    hv = hp * (127.f / randabsmax);
                    hvv = fabs(hv - (int)hv);

                    bp = ncnn::bfloat16_to_float32(ncnn::float32_to_bfloat16(p[j]));
                    bv = bp * (127.f / randabsmax);
                    bvv = fabs(bv - (int)bv);
                }
            }
        }
    }
    else // if (transA == 1)
    {
        std::vector<float> randabsmaxes(m.w);
        for (int j = 0; j < m.w; j++)
        {
            float randabsmax = RandomFloat(absmax * 0.5f, absmax);
            randabsmax = ncnn::float16_to_float32(ncnn::float32_to_float16(randabsmax));
            randabsmax = ncnn::bfloat16_to_float32(ncnn::float32_to_bfloat16(randabsmax));
            randabsmaxes[j] = randabsmax;
        }

        const int h = m.dims == 3 ? m.c : m.h;
        for (int i = 0; i < h; i++)
        {
            float* p = m.dims == 3 ? m.channel(i) : m.row(i);
            for (int j = 0; j < m.w; j++)
            {
                const float randabsmax = randabsmaxes[j];
                p[j] = RandomFloat(-randabsmax, randabsmax);
            }

            // drop 0.45 ~ 0.55
            for (int j = 0; j < m.w; j++)
            {
                const float randabsmax = randabsmaxes[j];
                float v = p[j] * (127.f / randabsmax);
                float vv = fabs(v - (int)v);

                float hp = ncnn::float16_to_float32(ncnn::float32_to_float16(p[j]));
                float hv = hp * (127.f / randabsmax);
                float hvv = fabs(hv - (int)hv);

                float bp = ncnn::bfloat16_to_float32(ncnn::float32_to_bfloat16(p[j]));
                float bv = bp * (127.f / randabsmax);
                float bvv = fabs(bv - (int)bv);

                while ((vv > 0.45f && vv < 0.55f) || (hvv > 0.45f && hvv < 0.55f) || (bvv > 0.45f && bvv < 0.55f))
                {
                    p[j] = RandomFloat(-randabsmax, randabsmax);
                    v = p[j] * (127.f / randabsmax);
                    vv = fabs(v - (int)v);

                    hp = ncnn::float16_to_float32(ncnn::float32_to_float16(p[j]));
                    hv = hp * (127.f / randabsmax);
                    hvv = fabs(hv - (int)hv);

                    bp = ncnn::bfloat16_to_float32(ncnn::float32_to_bfloat16(p[j]));
                    bv = bp * (127.f / randabsmax);
                    bvv = fabs(bv - (int)bv);
                }
            }
        }

        for (int j = 0; j < m.w; j++)
        {
            const int randi0 = RandomInt(0, h - 1);
            const int randi1 = RandomInt(0, h - 1);
            float* p0 = m.dims == 3 ? m.channel(randi0) : m.row(randi0);
            float* p1 = m.dims == 3 ? m.channel(randi1) : m.row(randi1);

            const float randabsmax = randabsmaxes[j];

            // set random a and b
            p0[j] = -randabsmax;
            p1[j] = randabsmax;
        }
    }
}

static void RandomizeB(ncnn::Mat& m, float absmax)
{
    absmax = ncnn::float16_to_float32(ncnn::float32_to_float16(absmax));
    absmax = ncnn::bfloat16_to_float32(ncnn::float32_to_bfloat16(absmax));

    const int h = m.dims == 3 ? m.c : m.h;
    float* p = m;
    for (int i = 0; i < h; i++)
    {
        float* p = m.dims == 3 ? m.channel(i) : m.row(i);
        for (int j = 0; j < m.w; j++)
        {
            p[j] = RandomFloat(-absmax, absmax);

            // drop 0.45 ~ 0.55
            float v = p[j] * (127.f / absmax);
            float vv = fabs(v - (int)v);

            float hp = ncnn::float16_to_float32(ncnn::float32_to_float16(p[j]));
            float hv = hp * (127.f / absmax);
            float hvv = fabs(hv - (int)hv);

            float bp = ncnn::bfloat16_to_float32(ncnn::float32_to_bfloat16(p[j]));
            float bv = bp * (127.f / absmax);
            float bvv = fabs(bv - (int)bv);

            while ((vv > 0.45f && vv < 0.55f) || (hvv > 0.45f && hvv < 0.55f) || (bvv > 0.45f && bvv < 0.55f))
            {
                p[j] = RandomFloat(-absmax, absmax);
                v = p[j] * (127.f / absmax);
                vv = fabs(v - (int)v);

                hp = ncnn::float16_to_float32(ncnn::float32_to_float16(p[j]));
                hv = hp * (127.f / absmax);
                hvv = fabs(hv - (int)hv);

                bp = ncnn::bfloat16_to_float32(ncnn::float32_to_bfloat16(p[j]));
                bv = bp * (127.f / absmax);
                bvv = fabs(bv - (int)bv);
            }
        }
    }

    // set random a and b
    if (m.dims == 3)
    {
        m.channel(RandomInt(0, h - 1))[RandomInt(0, m.w - 1)] = -absmax;
        m.channel(RandomInt(0, h - 1))[RandomInt(0, m.w - 1)] = absmax;
    }
    else
    {
        m.row(RandomInt(0, h - 1))[RandomInt(0, m.w - 1)] = -absmax;
        m.row(RandomInt(0, h - 1))[RandomInt(0, m.w - 1)] = absmax;
    }
}

static int test_gemm_int8(int M, int N, int K, float alpha, int transA, int transB, int output_elemtype, int output_transpose, int constantA, int constantB, int output_N1M, int flag = 0)
{
    ncnn::ParamDict pd;
    pd.set(0, alpha);
    pd.set(1, 1.f); // beta
    pd.set(2, transA);
    pd.set(3, transB);
    pd.set(4, constantA);
    pd.set(5, constantB);
    pd.set(6, 1);
    pd.set(7, M);
    pd.set(8, N);
    pd.set(9, K);
    pd.set(10, -1);
    pd.set(11, output_N1M);
    pd.set(13, output_elemtype);
    pd.set(14, output_transpose);
    pd.set(18, 2); // int8_scale_term

    std::vector<ncnn::Mat> weights;
    if (constantA) weights.push_back(transA ? RandomS8Mat(M, K) : RandomS8Mat(K, M));
    if (constantB) weights.push_back(transB ? RandomS8Mat(K, N) : RandomS8Mat(N, K));
    if (constantA) weights.push_back(RandomMat(M, 10.f, 20.f));
    if (constantB) weights.push_back(RandomMat(1, 10.f, 20.f));

    std::vector<ncnn::Mat> a;
    if (!constantA)
    {
        a.push_back(transA ? (output_N1M ? ncnn::Mat(M, 1, K) : ncnn::Mat(M, K)) : (output_N1M ? ncnn::Mat(K, 1, M) : ncnn::Mat(K, M)));
        RandomizeA(a[a.size() - 1], transA, 10.f);
    }
    if (!constantB)
    {
        a.push_back(transB ? (output_N1M ? ncnn::Mat(K, 1, N) : ncnn::Mat(K, N)) : (output_N1M ? ncnn::Mat(N, 1, K) : ncnn::Mat(N, K)));
        RandomizeB(a[a.size() - 1], 10.f);
    }

    ncnn::Option opt;
    opt.num_threads = 1;
    opt.use_packing_layout = true;
    opt.use_fp16_packed = false;
    opt.use_fp16_storage = false;
    opt.use_fp16_arithmetic = false;
    opt.use_bf16_packed = false;
    opt.use_bf16_storage = false;

    int ret = test_layer_opt("Gemm", pd, weights, opt, a, 1, 0.001, flag);
    if (ret != 0)
    {
        fprintf(stderr, "test_gemm_int8 failed M=%d N=%d K=%d alpha=%f transA=%d transB=%d output_elemtype=%d output_transpose=%d constantA=%d constantB=%d output_N1M=%d\n", M, N, K, alpha, transA, transB, output_elemtype, output_transpose, constantA, constantB, output_N1M);
    }

    return ret;
}

static int test_gemm_int8_bias(int M, int N, int K, const ncnn::Mat& C, float alpha, float beta, int transA, int transB, int output_elemtype, int output_transpose, int constantA, int constantB, int constantC, bool use_bf16 = false, int output_N1M = 0, int flag = 0)
{
    int broadcast_type_C = 0;
    if (C.dims == 1 && C.w == 1)
    {
        // scalar
        broadcast_type_C = 0;
    }
    if (C.dims == 1 && C.w == M)
    {
        // M
        // auto broadcast from h to w is the ncnn-style convention
        broadcast_type_C = 1;
    }
    if (C.dims == 1 && C.w == N)
    {
        // N
        broadcast_type_C = 4;
    }
    if (C.dims == 2 && C.w == 1 && C.h == M)
    {
        // Mx1
        broadcast_type_C = 2;
    }
    if (C.dims == 2 && C.w == N && C.h == M)
    {
        // MxN
        broadcast_type_C = 3;
    }
    if (C.dims == 2 && C.w == N && C.h == 1)
    {
        // 1xN
        broadcast_type_C = 4;
    }

    ncnn::ParamDict pd;
    pd.set(0, alpha);
    pd.set(1, beta);
    pd.set(2, transA);
    pd.set(3, transB);
    pd.set(4, constantA);
    pd.set(5, constantB);
    pd.set(6, constantC);
    pd.set(7, M);
    pd.set(8, N);
    pd.set(9, K);
    pd.set(10, broadcast_type_C);
    pd.set(11, output_N1M);
    // pd.set(12, 1);                  // output_elempack
    pd.set(13, output_elemtype);
    pd.set(14, output_transpose);
    pd.set(18, 2); // int8_scale_term

    std::vector<ncnn::Mat> weights;
    if (constantA) weights.push_back(transA ? RandomS8Mat(M, K) : RandomS8Mat(K, M));
    if (constantB) weights.push_back(transB ? RandomS8Mat(K, N) : RandomS8Mat(N, K));
    if (constantC) weights.push_back(C);
    if (constantA) weights.push_back(RandomMat(M, 10.f, 20.f));
    if (constantB) weights.push_back(RandomMat(1, 10.f, 20.f));

    std::vector<ncnn::Mat> a;
    if (!constantA)
    {
        a.push_back(transA ? (output_N1M ? ncnn::Mat(M, 1, K) : ncnn::Mat(M, K)) : (output_N1M ? ncnn::Mat(K, 1, M) : ncnn::Mat(K, M)));
        RandomizeA(a[a.size() - 1], transA, 10.f);
    }
    if (!constantB)
    {
        a.push_back(transB ? (output_N1M ? ncnn::Mat(K, 1, N) : ncnn::Mat(K, N)) : (output_N1M ? ncnn::Mat(N, 1, K) : ncnn::Mat(N, K)));
        RandomizeB(a[a.size() - 1], 10.f);
    }
    if (!constantC) a.push_back(C);

    ncnn::Option opt;
    opt.num_threads = 1;
    opt.use_packing_layout = true;
    opt.use_fp16_packed = false;
    opt.use_fp16_storage = false;
    opt.use_fp16_arithmetic = false;
    opt.use_bf16_packed = use_bf16;
    opt.use_bf16_storage = use_bf16;

    int ret = test_layer_opt("Gemm", pd, weights, opt, a, 1, 0.001, flag);
    if (ret != 0)
    {
        fprintf(stderr, "test_gemm_int8_bias failed M=%d N=%d K=%d C.dims=%d C=(%d %d %d) alpha=%f beta=%f transA=%d transB=%d output_elemtype=%d output_transpose=%d constantA=%d constantB=%d constantC=%d use_bf16=%d output_N1M=%d\n", M, N, K, C.dims, C.w, C.h, C.c, alpha, beta, transA, transB, output_elemtype, output_transpose, constantA, constantB, constantC, use_bf16, output_N1M);
    }

    return ret;
}

static int test_gemm_int8_fp16s(int M, int N, int K, float alpha, int transA, int transB, int output_elemtype, int output_transpose, int constantA, int constantB, int output_N1M)
{
    ncnn::ParamDict pd;
    pd.set(0, alpha);
    pd.set(1, 1.f); // beta
    pd.set(2, transA);
    pd.set(3, transB);
    pd.set(4, constantA);
    pd.set(5, constantB);
    pd.set(6, 1);
    pd.set(7, M);
    pd.set(8, N);
    pd.set(9, K);
    pd.set(10, -1);
    pd.set(11, output_N1M);
    pd.set(13, output_elemtype);
    pd.set(14, output_transpose);
    pd.set(18, 2); // int8_scale_term

    std::vector<ncnn::Mat> weights;
    if (constantA) weights.push_back(transA ? RandomS8Mat(M, K) : RandomS8Mat(K, M));
    if (constantB) weights.push_back(transB ? RandomS8Mat(K, N) : RandomS8Mat(N, K));
    if (constantA) weights.push_back(RandomMat(M, 10.f, 20.f));
    if (constantB) weights.push_back(RandomMat(1, 10.f, 20.f));

    std::vector<ncnn::Mat> a;
    if (!constantA)
    {
        a.push_back(transA ? (output_N1M ? ncnn::Mat(M, 1, K) : ncnn::Mat(M, K)) : (output_N1M ? ncnn::Mat(K, 1, M) : ncnn::Mat(K, M)));
        RandomizeA(a[a.size() - 1], transA, 10.f);
    }
    if (!constantB)
    {
        a.push_back(transB ? (output_N1M ? ncnn::Mat(K, 1, N) : ncnn::Mat(K, N)) : (output_N1M ? ncnn::Mat(N, 1, K) : ncnn::Mat(N, K)));
        RandomizeB(a[a.size() - 1], 10.f);
    }

    ncnn::Option opt;
    opt.num_threads = 1;
    opt.use_packing_layout = true;
    opt.use_fp16_packed = false;
    opt.use_fp16_storage = true;
    opt.use_fp16_arithmetic = false;
    opt.use_bf16_packed = false;
    opt.use_bf16_storage = false;

    float epsilon = 0.001;

    int ret = test_layer_opt("Gemm", pd, weights, opt, a, 1, epsilon);
    if (ret != 0)
    {
        fprintf(stderr, "test_gemm_int8_fp16s failed M=%d N=%d K=%d alpha=%f transA=%d transB=%d output_elemtype=%d output_transpose=%d constantA=%d constantB=%d output_N1M=%d\n", M, N, K, alpha, transA, transB, output_elemtype, output_transpose, constantA, constantB, output_N1M);
        return ret;
    }

    return 0;
}

static int test_gemm_int8_bf16s(int M, int N, int K, float alpha, int transA, int transB, int output_elemtype, int output_transpose, int constantA, int constantB, int output_N1M)
{
    ncnn::ParamDict pd;
    pd.set(0, alpha);
    pd.set(1, 1.f); // beta
    pd.set(2, transA);
    pd.set(3, transB);
    pd.set(4, constantA);
    pd.set(5, constantB);
    pd.set(6, 1);
    pd.set(7, M);
    pd.set(8, N);
    pd.set(9, K);
    pd.set(10, -1);
    pd.set(11, output_N1M);
    pd.set(13, output_elemtype);
    pd.set(14, output_transpose);
    pd.set(18, 2); // int8_scale_term

    std::vector<ncnn::Mat> weights;
    if (constantA) weights.push_back(transA ? RandomS8Mat(M, K) : RandomS8Mat(K, M));
    if (constantB) weights.push_back(transB ? RandomS8Mat(K, N) : RandomS8Mat(N, K));
    if (constantA) weights.push_back(RandomMat(M, 10.f, 20.f));
    if (constantB) weights.push_back(RandomMat(1, 10.f, 20.f));

    std::vector<ncnn::Mat> a;
    if (!constantA)
    {
        a.push_back(transA ? (output_N1M ? ncnn::Mat(M, 1, K) : ncnn::Mat(M, K)) : (output_N1M ? ncnn::Mat(K, 1, M) : ncnn::Mat(K, M)));
        RandomizeA(a[a.size() - 1], transA, 10.f);
    }
    if (!constantB)
    {
        a.push_back(transB ? (output_N1M ? ncnn::Mat(K, 1, N) : ncnn::Mat(K, N)) : (output_N1M ? ncnn::Mat(N, 1, K) : ncnn::Mat(N, K)));
        RandomizeB(a[a.size() - 1], 10.f);
    }

    ncnn::Option opt;
    opt.num_threads = 1;
    opt.use_packing_layout = true;
    opt.use_fp16_packed = false;
    opt.use_fp16_storage = false;
    opt.use_fp16_arithmetic = false;
    opt.use_bf16_packed = true;
    opt.use_bf16_storage = true;

    float epsilon = 0.001;

    int ret = test_layer_opt("Gemm", pd, weights, opt, a, 1, epsilon);
    if (ret != 0)
    {
        fprintf(stderr, "test_gemm_int8_bf16s failed M=%d N=%d K=%d alpha=%f transA=%d transB=%d output_elemtype=%d output_transpose=%d constantA=%d constantB=%d output_N1M=%d\n", M, N, K, alpha, transA, transB, output_elemtype, output_transpose, constantA, constantB, output_N1M);
        return ret;
    }

    return 0;
}

static int test_gemm_0(int M, int N, int K, int constant_flag = 0)
{
    return 0
           || test_gemm_int8(M, N, K, 2.1f, 0, 1, 0, 0, 0, 0, 0)
           || test_gemm_int8(M, N, K, 3.1f, 1, 1, 0, 0, 0, 0, 0)
           || test_gemm_int8(M, N, K, 4.1f, 0, 0, 0, 0, 0, 0, 1)
           || test_gemm_int8(M, N, K, 5.1f, 1, 0, 0, 0, 0, 0, 1)

           || test_gemm_int8(M, N, K, 0.2f, 0, 1, 0, 0, 1, 0, 1, constant_flag)
           || test_gemm_int8(M, N, K, 0.3f, 1, 1, 0, 0, 1, 0, 1, constant_flag)
           || test_gemm_int8(M, N, K, 0.4f, 0, 0, 0, 0, 0, 1, 0, constant_flag)
           || test_gemm_int8(M, N, K, 0.5f, 0, 1, 0, 0, 0, 1, 0, constant_flag)

           || test_gemm_int8(M, N, K, 1.2f, 0, 1, 0, 0, 1, 1, 0, constant_flag)
           || test_gemm_int8(M, N, K, 1.3f, 1, 1, 0, 0, 1, 1, 1, constant_flag)
           || test_gemm_int8(M, N, K, 1.4f, 0, 0, 0, 0, 1, 0, 0, constant_flag)
           || test_gemm_int8(M, N, K, 1.5f, 1, 0, 0, 0, 1, 0, 1, constant_flag)

           || test_gemm_int8(M, N, K, -1.2f, 0, 1, 0, 1, 1, 1, 0, constant_flag)
           || test_gemm_int8(M, N, K, -1.3f, 1, 1, 0, 1, 1, 1, 0, constant_flag)
           || test_gemm_int8(M, N, K, -1.4f, 0, 0, 0, 1, 1, 1, 1, constant_flag)
           || test_gemm_int8(M, N, K, -1.5f, 1, 0, 0, 1, 1, 1, 1, constant_flag)

           || test_gemm_int8(M, N, K, -2.0f, 0, 1, 0, 1, 0, 1, 1, constant_flag)
           || test_gemm_int8(M, N, K, -3.0f, 1, 1, 0, 1, 0, 1, 1, constant_flag)
           || test_gemm_int8(M, N, K, -4.0f, 0, 0, 0, 1, 1, 0, 0, constant_flag)
           || test_gemm_int8(M, N, K, -5.0f, 0, 1, 0, 1, 1, 0, 0, constant_flag)

           || test_gemm_int8(M, N, K, -2.1f, 0, 1, 0, 1, 0, 0, 0)
           || test_gemm_int8(M, N, K, -3.1f, 1, 1, 0, 1, 0, 0, 1)
           || test_gemm_int8(M, N, K, -4.1f, 0, 0, 0, 1, 0, 1, 0, constant_flag)
           || test_gemm_int8(M, N, K, -5.1f, 1, 0, 0, 1, 0, 1, 1, constant_flag)

           || test_gemm_int8_fp16s(M, N, K, 1.f, 0, 1, 0, 0, 0, 0, 0)
           || test_gemm_int8_fp16s(M, N, K, 1.f, 1, 0, 0, 1, 0, 0, 0);
}

// dynamic matrix geometries cover transposes, rank-three output and fp16 storage
static int test_gemm_dynamic_geometry(int M, int N, int K)
{
    return 0
           || test_gemm_int8(M, N, K, 2.1f, 0, 1, 0, 0, 0, 0, 0)
           || test_gemm_int8(M, N, K, 3.1f, 1, 1, 0, 0, 0, 0, 0)
           || test_gemm_int8(M, N, K, 4.1f, 0, 0, 0, 0, 0, 0, 1)
           || test_gemm_int8(M, N, K, 5.1f, 1, 0, 0, 0, 0, 0, 1)
           || test_gemm_int8(M, N, K, -2.1f, 0, 1, 0, 1, 0, 0, 0)
           || test_gemm_int8(M, N, K, -3.1f, 1, 1, 0, 1, 0, 0, 1)
           || test_gemm_int8_fp16s(M, N, K, 1.f, 0, 1, 0, 0, 0, 0, 0)
           || test_gemm_int8_fp16s(M, N, K, 1.f, 1, 0, 0, 1, 0, 0, 0);
}

static int test_gemm_1(int M, int N, int K, int constant_flag = 0)
{
    return 0
           || test_gemm_int8_bias(M, N, K, RandomMat(1), 2.1f, 0.5f, 0, 0, 0, 0, 0, 0, 0)
           || test_gemm_int8_bias(M, N, K, RandomMat(1), 2.1f, 0.5f, 0, 0, 1, 1, 0, 0, 0)
           || test_gemm_int8_bias(M, N, K, RandomMat(M), 3.1f, 0.6f, 0, 1, 2, 0, 0, 0, 0)
           || test_gemm_int8_bias(M, N, K, RandomMat(M), 3.1f, 0.6f, 0, 1, 3, 1, 0, 0, 0)
           || test_gemm_int8_bias(M, N, K, RandomMat(1, M), 4.1f, 0.7f, 1, 0, 0, 0, 0, 0, 0)
           || test_gemm_int8_bias(M, N, K, RandomMat(1, M), 4.1f, 0.7f, 1, 0, 1, 1, 0, 0, 0)
           || test_gemm_int8_bias(M, N, K, RandomMat(N, M), 5.1f, -0.8f, 1, 1, 2, 0, 0, 0, 0)
           || test_gemm_int8_bias(M, N, K, RandomMat(N, M), 5.1f, -0.8f, 1, 1, 3, 1, 0, 0, 0)
           || test_gemm_int8_bias(M, N, K, RandomMat(N, M), 1.f, 1.f, 1, 1, 0, 0, 0, 0, 0)
           || test_gemm_int8_bias(M, N, K, RandomMat(N, M), 1.f, 1.f, 1, 1, 1, 1, 0, 0, 0)
           || test_gemm_int8_bias(M, N, K, RandomMat(N, 1), 2.1f, -0.5f, 0, 0, 2, 0, 0, 0, 0)
           || test_gemm_int8_bias(M, N, K, RandomMat(N, 1), 2.1f, -0.5f, 0, 0, 3, 1, 0, 0, 0)
           || test_gemm_int8_bias(M, N, K, RandomMat(N, 1), 0.8f, 1.f, 0, 0, 0, 0, 0, 0, 0)
           || test_gemm_int8_bias(M, N, K, RandomMat(N), 0.8f, 1.f, 0, 0, 1, 1, 0, 0, 0)
           || test_gemm_int8_bias(M, N, K, RandomMat(N), 3.1f, -0.6f, 0, 1, 2, 0, 0, 0, 0)
           || test_gemm_int8_bias(M, N, K, RandomMat(N), 3.1f, -0.6f, 0, 1, 3, 1, 0, 0, 0)
           || test_gemm_int8_bias(M, N, K, RandomMat(1), 1.7f, -0.4f, 0, 0, 0, 0, 0, 0, 0, false, 1)
           || test_gemm_int8_bias(M, N, K, RandomMat(M), -1.3f, 0.6f, 1, 0, 0, 1, 0, 0, 0, false, 1)
           || test_gemm_int8_bias(M, N, K, RandomMat(N, M), 0.8f, 0.5f, 0, 1, 0, 0, 0, 0, 0, false, 1)

           || test_gemm_int8_bias(M, N, K, RandomMat(1), -2.1f, 0.5f, 0, 0, 0, 0, 1, 1, 1, false, 0, constant_flag)
           || test_gemm_int8_bias(M, N, K, RandomMat(1), -2.1f, 0.5f, 0, 0, 1, 1, 1, 1, 1, false, 0, constant_flag)
           || test_gemm_int8_bias(M, N, K, RandomMat(M), -3.1f, 0.6f, 0, 1, 2, 0, 1, 1, 1, false, 0, constant_flag)
           || test_gemm_int8_bias(M, N, K, RandomMat(M), -3.1f, 0.6f, 0, 1, 3, 1, 1, 1, 1, false, 0, constant_flag)
           || test_gemm_int8_bias(M, N, K, RandomMat(1, M), -4.1f, 0.7f, 1, 0, 0, 0, 1, 1, 1, false, 0, constant_flag)
           || test_gemm_int8_bias(M, N, K, RandomMat(1, M), -4.1f, 0.7f, 1, 0, 1, 1, 1, 1, 1, false, 0, constant_flag)
           || test_gemm_int8_bias(M, N, K, RandomMat(N, M), -5.1f, -0.8f, 1, 1, 2, 0, 1, 1, 1, false, 0, constant_flag)
           || test_gemm_int8_bias(M, N, K, RandomMat(N, M), -5.1f, -0.8f, 1, 1, 3, 1, 1, 1, 1, false, 0, constant_flag)
           || test_gemm_int8_bias(M, N, K, RandomMat(N, M), 1.f, 1.f, 1, 1, 0, 0, 1, 1, 1, false, 0, constant_flag)
           || test_gemm_int8_bias(M, N, K, RandomMat(N, M), 1.f, 1.f, 1, 1, 1, 1, 1, 1, 1, false, 0, constant_flag)
           || test_gemm_int8_bias(M, N, K, RandomMat(N, 1), -2.1f, -0.5f, 0, 0, 2, 0, 1, 1, 1, false, 0, constant_flag)
           || test_gemm_int8_bias(M, N, K, RandomMat(N, 1), -2.1f, -0.5f, 0, 0, 3, 1, 1, 1, 1, false, 0, constant_flag)
           || test_gemm_int8_bias(M, N, K, RandomMat(N, 1), 0.8f, 1.f, 0, 0, 0, 0, 1, 1, 1, false, 0, constant_flag)
           || test_gemm_int8_bias(M, N, K, RandomMat(N), 0.8f, 1.f, 0, 0, 1, 1, 1, 1, 1, false, 0, constant_flag)
           || test_gemm_int8_bias(M, N, K, RandomMat(N), -3.1f, -0.6f, 0, 1, 2, 0, 1, 1, 1, false, 0, constant_flag)
           || test_gemm_int8_bias(M, N, K, RandomMat(N), -3.1f, -0.6f, 0, 1, 3, 1, 1, 1, 1, false, 0, constant_flag);
}
// bias geometries cover output types, row and column layouts, and rank-three inputs
static int test_gemm_bias_geometry(int M, int N, int K)
{
    return 0
           || test_gemm_int8_bias(M, N, K, RandomMat(1), 2.1f, 0.5f, 0, 0, 0, 0, 0, 0, 0)
           || test_gemm_int8_bias(M, N, K, RandomMat(1), 2.1f, 0.5f, 0, 0, 1, 1, 0, 0, 0)
           || test_gemm_int8_bias(M, N, K, RandomMat(M), 3.1f, 0.6f, 0, 1, 2, 0, 0, 0, 0)
           || test_gemm_int8_bias(M, N, K, RandomMat(M), 3.1f, 0.6f, 0, 1, 3, 1, 0, 0, 0)
           || test_gemm_int8_bias(M, N, K, RandomMat(1, M), 4.1f, 0.7f, 1, 0, 0, 0, 0, 0, 0)
           || test_gemm_int8_bias(M, N, K, RandomMat(1, M), 4.1f, 0.7f, 1, 0, 1, 1, 0, 0, 0)
           || test_gemm_int8_bias(M, N, K, RandomMat(N, M), 5.1f, -0.8f, 1, 1, 2, 0, 0, 0, 0)
           || test_gemm_int8_bias(M, N, K, RandomMat(N, M), 5.1f, -0.8f, 1, 1, 3, 1, 0, 0, 0)
           || test_gemm_int8_bias(M, N, K, RandomMat(N, 1), 2.1f, -0.5f, 0, 0, 2, 0, 0, 0, 0)
           || test_gemm_int8_bias(M, N, K, RandomMat(N, 1), 2.1f, -0.5f, 0, 0, 3, 1, 0, 0, 0)
           || test_gemm_int8_bias(M, N, K, RandomMat(1), 1.7f, -0.4f, 0, 0, 0, 0, 0, 0, 0, false, 1)
           || test_gemm_int8_bias(M, N, K, RandomMat(M), -1.3f, 0.6f, 1, 0, 0, 1, 0, 0, 0, false, 1)
           || test_gemm_int8_bias(M, N, K, RandomMat(N, M), 0.8f, 0.5f, 0, 1, 0, 0, 0, 0, 0, false, 1);
}
// dynamic input geometries cover four transpose directions and fp16 storage
static int test_gemm_dynamic_pack_geometry(int M, int N, int K)
{
    return 0
           || test_gemm_int8(M, N, K, 2.1f, 0, 1, 0, 0, 0, 0, 0)
           || test_gemm_int8(M, N, K, 3.1f, 1, 1, 0, 0, 0, 0, 0)
           || test_gemm_int8(M, N, K, 4.1f, 0, 0, 0, 0, 0, 0, 1)
           || test_gemm_int8(M, N, K, 5.1f, 1, 0, 0, 0, 0, 0, 1)
           || test_gemm_int8_fp16s(M, N, K, 1.f, 0, 1, 0, 0, 0, 0, 0)
           || test_gemm_int8_fp16s(M, N, K, 1.f, 1, 0, 0, 1, 0, 0, 0);
}

// bias geometries cover scalar, row, column, matrix and rank-three layouts
static int test_gemm_bias_pack_geometry(int M, int N, int K)
{
    return 0
           || test_gemm_int8_bias(M, N, K, RandomMat(1), 2.1f, 0.5f, 0, 0, 0, 0, 0, 0, 0)
           || test_gemm_int8_bias(M, N, K, RandomMat(M), 3.1f, 0.6f, 0, 1, 2, 0, 0, 0, 0)
           || test_gemm_int8_bias(M, N, K, RandomMat(1, M), 4.1f, 0.7f, 1, 0, 0, 0, 0, 0, 0)
           || test_gemm_int8_bias(M, N, K, RandomMat(N, M), 5.1f, -0.8f, 1, 1, 3, 1, 0, 0, 0)
           || test_gemm_int8_bias(M, N, K, RandomMat(N, 1), 2.1f, -0.5f, 0, 0, 3, 1, 0, 0, 0)
           || test_gemm_int8_bias(M, N, K, RandomMat(N, M), 0.8f, 0.5f, 0, 1, 0, 0, 0, 0, 0, false, 1);
}

// original cpu-only constant operands retain every packed geometry
static int test_gemm_constant_operands(int M, int N, int K, int constant_flag)
{
    return 0
           || test_gemm_int8(M, N, K, 0.2f, 0, 1, 0, 0, 1, 0, 1, constant_flag)
           || test_gemm_int8(M, N, K, 0.3f, 1, 1, 0, 0, 1, 0, 1, constant_flag)
           || test_gemm_int8(M, N, K, 0.4f, 0, 0, 0, 0, 0, 1, 0, constant_flag)
           || test_gemm_int8(M, N, K, 0.5f, 0, 1, 0, 0, 0, 1, 0, constant_flag)
           || test_gemm_int8(M, N, K, 1.2f, 0, 1, 0, 0, 1, 1, 0, constant_flag)
           || test_gemm_int8(M, N, K, 1.3f, 1, 1, 0, 0, 1, 1, 1, constant_flag)
           || test_gemm_int8(M, N, K, 1.4f, 0, 0, 0, 0, 1, 0, 0, constant_flag)
           || test_gemm_int8(M, N, K, 1.5f, 1, 0, 0, 0, 1, 0, 1, constant_flag)
           || test_gemm_int8(M, N, K, -1.2f, 0, 1, 0, 1, 1, 1, 0, constant_flag)
           || test_gemm_int8(M, N, K, -1.3f, 1, 1, 0, 1, 1, 1, 0, constant_flag)
           || test_gemm_int8(M, N, K, -1.4f, 0, 0, 0, 1, 1, 1, 1, constant_flag)
           || test_gemm_int8(M, N, K, -1.5f, 1, 0, 0, 1, 1, 1, 1, constant_flag)
           || test_gemm_int8(M, N, K, -2.0f, 0, 1, 0, 1, 0, 1, 1, constant_flag)
           || test_gemm_int8(M, N, K, -3.0f, 1, 1, 0, 1, 0, 1, 1, constant_flag)
           || test_gemm_int8(M, N, K, -4.0f, 0, 0, 0, 1, 1, 0, 0, constant_flag)
           || test_gemm_int8(M, N, K, -5.0f, 0, 1, 0, 1, 1, 0, 0, constant_flag)
           || test_gemm_int8(M, N, K, -4.1f, 0, 0, 0, 1, 0, 1, 0, constant_flag)
           || test_gemm_int8(M, N, K, -5.1f, 1, 0, 0, 1, 0, 1, 1, constant_flag)
           || test_gemm_int8_bias(M, N, K, RandomMat(1), -2.1f, 0.5f, 0, 0, 0, 0, 1, 1, 1, false, 0, constant_flag)
           || test_gemm_int8_bias(M, N, K, RandomMat(1), -2.1f, 0.5f, 0, 0, 1, 1, 1, 1, 1, false, 0, constant_flag)
           || test_gemm_int8_bias(M, N, K, RandomMat(M), -3.1f, 0.6f, 0, 1, 2, 0, 1, 1, 1, false, 0, constant_flag)
           || test_gemm_int8_bias(M, N, K, RandomMat(M), -3.1f, 0.6f, 0, 1, 3, 1, 1, 1, 1, false, 0, constant_flag)
           || test_gemm_int8_bias(M, N, K, RandomMat(1, M), -4.1f, 0.7f, 1, 0, 0, 0, 1, 1, 1, false, 0, constant_flag)
           || test_gemm_int8_bias(M, N, K, RandomMat(1, M), -4.1f, 0.7f, 1, 0, 1, 1, 1, 1, 1, false, 0, constant_flag)
           || test_gemm_int8_bias(M, N, K, RandomMat(N, M), -5.1f, -0.8f, 1, 1, 2, 0, 1, 1, 1, false, 0, constant_flag)
           || test_gemm_int8_bias(M, N, K, RandomMat(N, M), -5.1f, -0.8f, 1, 1, 3, 1, 1, 1, 1, false, 0, constant_flag)
           || test_gemm_int8_bias(M, N, K, RandomMat(N, M), 1.f, 1.f, 1, 1, 0, 0, 1, 1, 1, false, 0, constant_flag)
           || test_gemm_int8_bias(M, N, K, RandomMat(N, M), 1.f, 1.f, 1, 1, 1, 1, 1, 1, 1, false, 0, constant_flag)
           || test_gemm_int8_bias(M, N, K, RandomMat(N, 1), -2.1f, -0.5f, 0, 0, 2, 0, 1, 1, 1, false, 0, constant_flag)
           || test_gemm_int8_bias(M, N, K, RandomMat(N, 1), -2.1f, -0.5f, 0, 0, 3, 1, 1, 1, 1, false, 0, constant_flag)
           || test_gemm_int8_bias(M, N, K, RandomMat(N, 1), 0.8f, 1.f, 0, 0, 0, 0, 1, 1, 1, false, 0, constant_flag)
           || test_gemm_int8_bias(M, N, K, RandomMat(N), 0.8f, 1.f, 0, 0, 1, 1, 1, 1, 1, false, 0, constant_flag)
           || test_gemm_int8_bias(M, N, K, RandomMat(N), -3.1f, -0.6f, 0, 1, 2, 0, 1, 1, 1, false, 0, constant_flag)
           || test_gemm_int8_bias(M, N, K, RandomMat(N), -3.1f, -0.6f, 0, 1, 3, 1, 1, 1, 1, false, 0, constant_flag);
}

// packed matrix bias with unit beta covers vector blocks and scalar column tails
static int test_gemm_matrix_bias_unity_beta()
{
    return 0
           || test_gemm_int8_bias(24, 47, 24, RandomMat(47, 24), 1.f, 1.f, 1, 1, 0, 0, 0, 0, 0)
           || test_gemm_int8_bias(24, 47, 24, RandomMat(47, 24), 1.f, 1.f, 1, 1, 1, 1, 0, 0, 0);
}

#endif // NCNN_INT8

int main()
{
    SRAND(7767517);

#if NCNN_INT8
    // the flag applies to constant operands; dynamic operands retain every matrix size
    int ret = 0
              || test_gemm_0(1, 1, 1)
              || test_gemm_1(1, 1, 1)
              || test_gemm_dynamic_pack_geometry(1, 1, 23)
              || test_gemm_bias_pack_geometry(1, 1, 23)
              || test_gemm_constant_operands(1, 1, 47, TEST_LAYER_DISABLE_GPU_TESTING)
              || test_gemm_dynamic_pack_geometry(1, 1, 47)
              || test_gemm_bias_pack_geometry(1, 1, 47)
              || test_gemm_dynamic_pack_geometry(1, 23, 1)
              || test_gemm_bias_pack_geometry(1, 23, 1)
              || test_gemm_dynamic_pack_geometry(23, 1, 1)
              || test_gemm_bias_pack_geometry(23, 1, 1)
              || test_gemm_dynamic_pack_geometry(1, 23, 23)
              || test_gemm_bias_pack_geometry(1, 23, 23)
              || test_gemm_dynamic_pack_geometry(23, 1, 23)
              || test_gemm_bias_pack_geometry(23, 1, 23)
              || test_gemm_dynamic_pack_geometry(1, 31, 1)
              || test_gemm_bias_pack_geometry(1, 31, 1)
              || test_gemm_dynamic_pack_geometry(31, 1, 1)
              || test_gemm_bias_pack_geometry(31, 1, 1)
              || test_gemm_constant_operands(1, 35, 1, TEST_LAYER_DISABLE_GPU_TESTING)
              || test_gemm_dynamic_pack_geometry(1, 35, 1)
              || test_gemm_bias_pack_geometry(1, 35, 1)
              || test_gemm_constant_operands(35, 1, 1, TEST_LAYER_DISABLE_GPU_TESTING)
              || test_gemm_dynamic_pack_geometry(35, 1, 1)
              || test_gemm_bias_pack_geometry(35, 1, 1)
              || test_gemm_dynamic_pack_geometry(1, 35, 47)
              || test_gemm_bias_pack_geometry(1, 35, 47)
              || test_gemm_dynamic_pack_geometry(35, 1, 47)
              || test_gemm_bias_pack_geometry(35, 1, 47)
              || test_gemm_constant_operands(1, 47, 1, TEST_LAYER_DISABLE_GPU_TESTING)
              || test_gemm_dynamic_pack_geometry(1, 47, 1)
              || test_gemm_bias_pack_geometry(1, 47, 1)
              || test_gemm_constant_operands(47, 1, 1, TEST_LAYER_DISABLE_GPU_TESTING)
              || test_gemm_dynamic_pack_geometry(47, 1, 1)
              || test_gemm_bias_pack_geometry(47, 1, 1)
              || test_gemm_constant_operands(2, 2, 2, TEST_LAYER_DISABLE_GPU_TESTING)
              || test_gemm_dynamic_pack_geometry(2, 2, 2)
              || test_gemm_bias_pack_geometry(2, 2, 2)
              || test_gemm_constant_operands(3, 3, 3, TEST_LAYER_DISABLE_GPU_TESTING)
              || test_gemm_dynamic_pack_geometry(3, 3, 3)
              || test_gemm_bias_pack_geometry(3, 3, 3)
              || test_gemm_dynamic_geometry(4, 4, 4)
              || test_gemm_bias_geometry(4, 4, 4)
              || test_gemm_constant_operands(5, 5, 5, TEST_LAYER_DISABLE_GPU_TESTING)
              || test_gemm_dynamic_pack_geometry(5, 5, 5)
              || test_gemm_bias_pack_geometry(5, 5, 5)
              || test_gemm_constant_operands(6, 6, 6, TEST_LAYER_DISABLE_GPU_TESTING)
              || test_gemm_dynamic_pack_geometry(6, 6, 6)
              || test_gemm_bias_pack_geometry(6, 6, 6)
              || test_gemm_constant_operands(7, 7, 7, TEST_LAYER_DISABLE_GPU_TESTING)
              || test_gemm_dynamic_pack_geometry(7, 7, 7)
              || test_gemm_bias_pack_geometry(7, 7, 7)
              || test_gemm_constant_operands(7, 31, 3, TEST_LAYER_DISABLE_GPU_TESTING)
              || test_gemm_dynamic_pack_geometry(7, 31, 3)
              || test_gemm_bias_pack_geometry(7, 31, 3)
              || test_gemm_constant_operands(31, 7, 3, TEST_LAYER_DISABLE_GPU_TESTING)
              || test_gemm_dynamic_pack_geometry(31, 7, 3)
              || test_gemm_bias_pack_geometry(31, 7, 3)
              || test_gemm_constant_operands(8, 8, 8, TEST_LAYER_DISABLE_GPU_TESTING)
              || test_gemm_dynamic_pack_geometry(8, 8, 8)
              || test_gemm_bias_pack_geometry(8, 8, 8)
              || test_gemm_dynamic_pack_geometry(12, 12, 23)
              || test_gemm_bias_pack_geometry(12, 12, 23)
              || test_gemm_dynamic_pack_geometry(12, 23, 12)
              || test_gemm_bias_pack_geometry(12, 23, 12)
              || test_gemm_dynamic_pack_geometry(23, 12, 12)
              || test_gemm_bias_pack_geometry(23, 12, 12)
              || test_gemm_dynamic_pack_geometry(12, 31, 12)
              || test_gemm_bias_pack_geometry(12, 31, 12)
              || test_gemm_dynamic_pack_geometry(31, 12, 12)
              || test_gemm_bias_pack_geometry(31, 12, 12)
              || test_gemm_dynamic_pack_geometry(15, 15, 15)
              || test_gemm_bias_pack_geometry(15, 15, 15)
              || test_gemm_dynamic_geometry(16, 16, 16)
              || test_gemm_bias_geometry(16, 16, 16)
              || test_gemm_constant_operands(19, 44, 7, TEST_LAYER_DISABLE_GPU_TESTING)
              || test_gemm_dynamic_pack_geometry(19, 44, 7)
              || test_gemm_bias_pack_geometry(19, 44, 7)
              || test_gemm_constant_operands(44, 19, 7, TEST_LAYER_DISABLE_GPU_TESTING)
              || test_gemm_dynamic_pack_geometry(44, 19, 7)
              || test_gemm_bias_pack_geometry(44, 19, 7)
              || test_gemm_dynamic_pack_geometry(20, 28, 7)
              || test_gemm_bias_pack_geometry(20, 28, 7)
              || test_gemm_dynamic_pack_geometry(28, 20, 7)
              || test_gemm_bias_pack_geometry(28, 20, 7)
              || test_gemm_dynamic_pack_geometry(23, 31, 1)
              || test_gemm_bias_pack_geometry(23, 31, 1)
              || test_gemm_dynamic_pack_geometry(31, 23, 1)
              || test_gemm_bias_pack_geometry(31, 23, 1)
              || test_gemm_0(23, 31, 23)
              || test_gemm_1(23, 31, 23)
              || test_gemm_dynamic_pack_geometry(31, 23, 23)
              || test_gemm_bias_pack_geometry(31, 23, 23)
              || test_gemm_constant_operands(24, 24, 47, TEST_LAYER_DISABLE_GPU_TESTING)
              || test_gemm_dynamic_pack_geometry(24, 24, 47)
              || test_gemm_bias_pack_geometry(24, 24, 47)
              || test_gemm_constant_operands(24, 35, 24, TEST_LAYER_DISABLE_GPU_TESTING)
              || test_gemm_dynamic_pack_geometry(24, 35, 24)
              || test_gemm_bias_pack_geometry(24, 35, 24)
              || test_gemm_constant_operands(35, 24, 24, TEST_LAYER_DISABLE_GPU_TESTING)
              || test_gemm_dynamic_pack_geometry(35, 24, 24)
              || test_gemm_bias_pack_geometry(35, 24, 24)
              || test_gemm_constant_operands(24, 47, 24, TEST_LAYER_DISABLE_GPU_TESTING)
              || test_gemm_dynamic_pack_geometry(24, 47, 24)
              || test_gemm_bias_pack_geometry(24, 47, 24)
              || test_gemm_constant_operands(47, 24, 24, TEST_LAYER_DISABLE_GPU_TESTING)
              || test_gemm_dynamic_pack_geometry(47, 24, 24)
              || test_gemm_bias_pack_geometry(47, 24, 24)
              || test_gemm_constant_operands(31, 31, 31, TEST_LAYER_DISABLE_GPU_TESTING)
              || test_gemm_dynamic_pack_geometry(31, 31, 31)
              || test_gemm_bias_pack_geometry(31, 31, 31)
              || test_gemm_constant_operands(32, 32, 9, TEST_LAYER_DISABLE_GPU_TESTING)
              || test_gemm_dynamic_pack_geometry(32, 32, 9)
              || test_gemm_bias_pack_geometry(32, 32, 9)
              || test_gemm_dynamic_pack_geometry(35, 47, 48)
              || test_gemm_bias_pack_geometry(35, 47, 48)
              || test_gemm_0(47, 35, 48)
              || test_gemm_1(47, 35, 48)
              || test_gemm_0(35, 48, 47)
              || test_gemm_1(35, 48, 47)
              || test_gemm_0(48, 35, 47)
              || test_gemm_bias_geometry(48, 35, 47)
              || test_gemm_constant_operands(40, 40, 40, TEST_LAYER_DISABLE_GPU_TESTING)
              || test_gemm_dynamic_pack_geometry(40, 40, 40)
              || test_gemm_bias_pack_geometry(40, 40, 40)
              || test_gemm_dynamic_geometry(47, 48, 47)
              || test_gemm_bias_geometry(47, 48, 47)
              || test_gemm_dynamic_geometry(48, 47, 47)
              || test_gemm_bias_pack_geometry(48, 47, 47)
              || test_gemm_matrix_bias_unity_beta();
    if (ret != 0)
        return ret;

    if (test_gemm_int8_bf16s(12, 23, 12, 1.f, 0, 1, 0, 0, 0, 0, 0)
            || test_gemm_int8_bf16s(12, 23, 12, 1.f, 1, 0, 0, 1, 0, 0, 0)
            || test_gemm_int8_bias(12, 23, 12, RandomMat(23), 1.7f, -0.4f, 0, 1, 0, 1, 0, 0, 0, true))
    {
        return -1;
    }
#else
    // test nothing for non-int8 build
#endif

    return 0;
}