Your task is to convert a given C++ ARM NEON SIMD to WASM SIMD. Here is an example of another function:

```cpp
void ggml_vec_dot_q5_0_q8_0(int n, float * restrict s, size_t bs, const void * restrict vx, size_t bx, const void * restrict vy, size_t by, int nrc) {
    const int qk = QK8_0;
    const int nb = n / qk;

    int ib = 0;
    float sumf = 0;

    assert(n % qk == 0);
    assert(qk == QK5_0);
    assert(nrc == 1);
    UNUSED(nrc);
    UNUSED(bx);
    UNUSED(by);
    UNUSED(bs);

    const block_q5_0 * restrict x = vx;
    const block_q8_0 * restrict y = vy;

#if defined(__ARM_NEON)
    float32x4_t sumv0 = vdupq_n_f32(0.0f);
    float32x4_t sumv1 = vdupq_n_f32(0.0f);

    uint32_t qh0;
    uint32_t qh1;

    uint64_t tmp0[4];
    uint64_t tmp1[4];

    for (; ib + 1 < nb; ib += 2) {
        const block_q5_0 * restrict x0 = &x[ib];
        const block_q5_0 * restrict x1 = &x[ib + 1];
        const block_q8_0 * restrict y0 = &y[ib];
        const block_q8_0 * restrict y1 = &y[ib + 1];

        const uint8x16_t m4b = vdupq_n_u8(0x0F);

        // extract the 5th bit via lookup table ((!b) << 4)
        memcpy(&qh0, x0->qh, sizeof(qh0));
        memcpy(&qh1, x1->qh, sizeof(qh1));

        tmp0[0] = table_b2b_1[(qh0 >>  0) & 0xFF];
        tmp0[1] = table_b2b_1[(qh0 >>  8) & 0xFF];
        tmp0[2] = table_b2b_1[(qh0 >> 16) & 0xFF];
        tmp0[3] = table_b2b_1[(qh0 >> 24)       ];

        tmp1[0] = table_b2b_1[(qh1 >>  0) & 0xFF];
        tmp1[1] = table_b2b_1[(qh1 >>  8) & 0xFF];
        tmp1[2] = table_b2b_1[(qh1 >> 16) & 0xFF];
        tmp1[3] = table_b2b_1[(qh1 >> 24)       ];

        const int8x16_t qhl0 = vld1q_s8((const int8_t *)(tmp0 + 0));
        const int8x16_t qhh0 = vld1q_s8((const int8_t *)(tmp0 + 2));
        const int8x16_t qhl1 = vld1q_s8((const int8_t *)(tmp1 + 0));
        const int8x16_t qhh1 = vld1q_s8((const int8_t *)(tmp1 + 2));

        const uint8x16_t v0_0 = vld1q_u8(x0->qs);
        const uint8x16_t v0_1 = vld1q_u8(x1->qs);

        // 4-bit -> 8-bit
        int8x16_t v0_0l = vreinterpretq_s8_u8(vandq_u8  (v0_0, m4b));
        int8x16_t v0_0h = vreinterpretq_s8_u8(vshrq_n_u8(v0_0, 4));
        int8x16_t v0_1l = vreinterpretq_s8_u8(vandq_u8  (v0_1, m4b));
        int8x16_t v0_1h = vreinterpretq_s8_u8(vshrq_n_u8(v0_1, 4));

        // add high bit and sub 16 (equivalent to sub 0x10 when bit is zero)
        const int8x16_t v0_0lf = vsubq_s8(v0_0l, qhl0);
        const int8x16_t v0_0hf = vsubq_s8(v0_0h, qhh0);
        const int8x16_t v0_1lf = vsubq_s8(v0_1l, qhl1);
        const int8x16_t v0_1hf = vsubq_s8(v0_1h, qhh1);

        // load y
        const int8x16_t v1_0l = vld1q_s8(y0->qs);
        const int8x16_t v1_0h = vld1q_s8(y0->qs + 16);
        const int8x16_t v1_1l = vld1q_s8(y1->qs);
        const int8x16_t v1_1h = vld1q_s8(y1->qs + 16);

        sumv0 = vmlaq_n_f32(sumv0, vcvtq_f32_s32(vaddq_s32(
                        ggml_vdotq_s32(vdupq_n_s32(0), v0_0lf, v1_0l),
                        ggml_vdotq_s32(vdupq_n_s32(0), v0_0hf, v1_0h))), GGML_FP16_TO_FP32(x0->d)*GGML_FP16_TO_FP32(y0->d));
        sumv1 = vmlaq_n_f32(sumv1, vcvtq_f32_s32(vaddq_s32(
                        ggml_vdotq_s32(vdupq_n_s32(0), v0_1lf, v1_1l),
                        ggml_vdotq_s32(vdupq_n_s32(0), v0_1hf, v1_1h))), GGML_FP16_TO_FP32(x1->d)*GGML_FP16_TO_FP32(y1->d));
    }

    sumf = vaddvq_f32(sumv0) + vaddvq_f32(sumv1);
#elif defined(__wasm_simd128__)
    v128_t sumv = wasm_f32x4_splat(0.0f);

    uint32_t qh;
    uint64_t tmp[4];

    // TODO: check if unrolling this is better
    for (; ib < nb; ++ib) {
        const block_q5_0 * restrict x0 = &x[ib];
        const block_q8_0 * restrict y0 = &y[ib];

        const v128_t m4b  = wasm_i8x16_splat(0x0F);

        // extract the 5th bit
        memcpy(&qh, x0->qh, sizeof(qh));

        tmp[0] = table_b2b_1[(qh >>  0) & 0xFF];
        tmp[1] = table_b2b_1[(qh >>  8) & 0xFF];
        tmp[2] = table_b2b_1[(qh >> 16) & 0xFF];
        tmp[3] = table_b2b_1[(qh >> 24)       ];

        const v128_t qhl = wasm_v128_load(tmp + 0);
        const v128_t qhh = wasm_v128_load(tmp + 2);

        const v128_t v0 = wasm_v128_load(x0->qs);

        // 4-bit -> 8-bit
        const v128_t v0l = wasm_v128_and (v0, m4b);
        const v128_t v0h = wasm_u8x16_shr(v0, 4);

        // add high bit and sub 16 (equivalent to sub 0x10 when bit is zero)
        const v128_t v0lf = wasm_i8x16_sub(v0l, qhl);
        const v128_t v0hf = wasm_i8x16_sub(v0h, qhh);

        // load y
        const v128_t v1l = wasm_v128_load(y0->qs);
        const v128_t v1h = wasm_v128_load(y0->qs + 16);

        // int8x16 -> int16x8
        const v128_t v0lfl = wasm_i16x8_extend_low_i8x16 (v0lf);
        const v128_t v0lfh = wasm_i16x8_extend_high_i8x16(v0lf);
        const v128_t v0hfl = wasm_i16x8_extend_low_i8x16 (v0hf);
        const v128_t v0hfh = wasm_i16x8_extend_high_i8x16(v0hf);

        const v128_t v1ll = wasm_i16x8_extend_low_i8x16 (v1l);
        const v128_t v1lh = wasm_i16x8_extend_high_i8x16(v1l);
        const v128_t v1hl = wasm_i16x8_extend_low_i8x16 (v1h);
        const v128_t v1hh = wasm_i16x8_extend_high_i8x16(v1h);

        // dot product
        sumv = wasm_f32x4_add(sumv, wasm_f32x4_mul(wasm_f32x4_convert_i32x4(
                        wasm_i32x4_add(
                            wasm_i32x4_add(wasm_i32x4_dot_i16x8(v0lfl, v1ll),
                                           wasm_i32x4_dot_i16x8(v0lfh, v1lh)),
                            wasm_i32x4_add(wasm_i32x4_dot_i16x8(v0hfl, v1hl),
                                           wasm_i32x4_dot_i16x8(v0hfh, v1hh)))),
                    wasm_f32x4_splat(GGML_FP16_TO_FP32(x0->d) * GGML_FP16_TO_FP32(y0->d))));
    }

    sumf = wasm_f32x4_extract_lane(sumv, 0) + wasm_f32x4_extract_lane(sumv, 1) +
           wasm_f32x4_extract_lane(sumv, 2) + wasm_f32x4_extract_lane(sumv, 3);
#endif

    for (; ib < nb; ++ib) {
        uint32_t qh;
        memcpy(&qh, x[ib].qh, sizeof(qh));

        int sumi0 = 0;
        int sumi1 = 0;

        for (int j = 0; j < qk/2; ++j) {
            const uint8_t xh_0 = ((qh & (1u << (j + 0 ))) >> (j + 0 )) << 4;
            const uint8_t xh_1 = ((qh & (1u << (j + 16))) >> (j + 12));

            const int32_t x0 = (int8_t)(((x[ib].qs[j] & 0x0F) | xh_0) - 16);
            const int32_t x1 = (int8_t)(((x[ib].qs[j] >>   4) | xh_1) - 16);

            sumi0 += (x0 * y[ib].qs[j]);
            sumi1 += (x1 * y[ib].qs[j + qk/2]);
        }

        int sumi = sumi0 + sumi1;
        sumf += (GGML_FP16_TO_FP32(x[ib].d)*GGML_FP16_TO_FP32(y[ib].d)) * sumi;
    }

    *s = sumf;
}
````

Here is the function that you need to convert:


```cpp
void ggml_vec_dot_q8_0_q8_0(int n, float * restrict s, size_t bs, const void * restrict vx, size_t bx, const void * restrict vy, size_t by, int nrc) {
    const int qk = QK8_0;
    const int nb = n / qk;

    assert(n % qk == 0);
    assert(nrc == 1);
    UNUSED(nrc);
    UNUSED(bx);
    UNUSED(by);
    UNUSED(bs);

    const block_q8_0 * restrict x = vx;
    const block_q8_0 * restrict y = vy;

    int ib = 0;
    float sumf = 0;

#if defined(__ARM_NEON)
    float32x4_t sumv0 = vdupq_n_f32(0.0f);
    float32x4_t sumv1 = vdupq_n_f32(0.0f);

    for (; ib + 1 < nb; ib += 2) {
        const block_q8_0 * restrict x0 = &x[ib + 0];
        const block_q8_0 * restrict x1 = &x[ib + 1];
        const block_q8_0 * restrict y0 = &y[ib + 0];
        const block_q8_0 * restrict y1 = &y[ib + 1];

        const int8x16_t x0_0 = vld1q_s8(x0->qs);
        const int8x16_t x0_1 = vld1q_s8(x0->qs + 16);
        const int8x16_t x1_0 = vld1q_s8(x1->qs);
        const int8x16_t x1_1 = vld1q_s8(x1->qs + 16);

        // load y
        const int8x16_t y0_0 = vld1q_s8(y0->qs);
        const int8x16_t y0_1 = vld1q_s8(y0->qs + 16);
        const int8x16_t y1_0 = vld1q_s8(y1->qs);
        const int8x16_t y1_1 = vld1q_s8(y1->qs + 16);

        sumv0 = vmlaq_n_f32(sumv0, vcvtq_f32_s32(vaddq_s32(
                        ggml_vdotq_s32(vdupq_n_s32(0), x0_0, y0_0),
                        ggml_vdotq_s32(vdupq_n_s32(0), x0_1, y0_1))), GGML_FP16_TO_FP32(x0->d)*GGML_FP16_TO_FP32(y0->d));

        sumv1 = vmlaq_n_f32(sumv1, vcvtq_f32_s32(vaddq_s32(
                        ggml_vdotq_s32(vdupq_n_s32(0), x1_0, y1_0),
                        ggml_vdotq_s32(vdupq_n_s32(0), x1_1, y1_1))), GGML_FP16_TO_FP32(x1->d)*GGML_FP16_TO_FP32(y1->d));
    }

    sumf = vaddvq_f32(sumv0) + vaddvq_f32(sumv1);
#endif
    for (; ib < nb; ++ib) {
        int sumi = 0;

        for (int j = 0; j < qk; j++) {
            sumi += x[ib].qs[j]*y[ib].qs[j];
        }

        sumf += sumi*(GGML_FP16_TO_FP32(x[ib].d)*GGML_FP16_TO_FP32(y[ib].d));
    }

    *s = sumf;
}
```

You must start your code with `#elif defined(__wasm_simd128__)`