// 
// sub_100928 from 0x100928 to 0x100aa8 (384 bytes)
// 
// 1 xrefs:
// > <???> @ 0x118b08
// 
// This function is hookable on this platform!
// 
// Using BromaIDA 8.0.0 @ https://github.com/Stazzical/BromaIDA
// Using bindings at commit ba9f177b at Thu Jul 16 22:52:06 2026 from https://github.com/geode-sdk/bindings
// 

float32x4_t *__fastcall sub_100928(
        float32x4_t *result,
        __int64 a2,
        unsigned int a3,
        __int64 a4,
        __int64 a5,
        __int64 a6,
        int a7)
{
  float32x4_t v7; // q0
  float v8; // s3
  float v9; // s5
  float v10; // s4
  float v11; // s1
  float v12; // s2
  float32x4_t v13; // q6
  float32x4_t v14; // q7
  float32x2_t v15; // d17
  unsigned int v16; // w8
  float32x4_t v17; // t1
  __int64 v18; // x8
  float v19; // s6
  float32x4_t v20; // t1
  unsigned __int64 v21; // d20
  unsigned __int64 v22; // d21
  float32x4_t v23; // q22
  float32x4_t v24; // q23
  unsigned __int64 v25; // d24
  unsigned __int64 v26; // d25
  float32x4_t v27; // q26
  __int64 v28; // x8
  float v29; // s3
  float32x4_t v30; // q4
  float v31; // s5
  float v32; // s3

  v7.n128_u32[0] = *(_DWORD *)a4;
  v8 = *(float *)(a4 + 256);
  v9 = *(float *)(a4 + 128);
  v10 = *(float *)(a4 + 384);
  v11 = *(float *)(a4 + 512);
  v12 = *(float *)(a4 + 640);
  v13.n128_u32[0] = *(_DWORD *)a4;
  v14.n128_u64[0] = __PAIR64__(LODWORD(v10), LODWORD(v8));
  v13.n128_f32[1] = v9;
  v15.n64_u32[0] = *(_DWORD *)a4;
  v13.n128_u64[1] = __PAIR64__(LODWORD(v10), LODWORD(v8));
  v14.n128_u64[1] = __PAIR64__(LODWORD(v12), LODWORD(v11));
  v15.n64_f32[1] = v9;
  v16 = a3 >> 2;
  if ( a7 )
  {
    for ( ; v16; a2 += 96 )
    {
      v17 = *result++;
      --v16;
      *(float32x4_t *)a2 = vmulq_n_f32(v13, v17.n128_f32[0]);
      *(float32x2_t *)(a2 + 16) = vmul_n_f32((float32x2_t)__PAIR64__(LODWORD(v12), LODWORD(v11)), v17.n128_f32[0]);
      *(float32x2_t *)(a2 + 24) = vmul_lane_f32(v15, (float32x2_t)v17.n128_u64[0], 1);
      *(float32x4_t *)(a2 + 32) = vmulq_lane_f32(v14, (float32x2_t)v17.n128_u64[0], 1);
      *(float32x4_t *)(a2 + 48) = vmulq_laneq_f32(v13, v17, 2);
      *(float32x2_t *)(a2 + 64) = vmul_laneq_f32((float32x2_t)__PAIR64__(LODWORD(v12), LODWORD(v11)), v17, 2);
      *(float32x2_t *)(a2 + 72) = vmul_laneq_f32(v15, v17, 3);
      *(float32x4_t *)(a2 + 80) = vmulq_laneq_f32(v14, v17, 3);
    }
    v18 = 0;
    while ( (a3 & 3) != (_DWORD)v18 )
    {
      v19 = result->n128_f32[v18++];
      *(float *)a2 = v19 * v7.n128_f32[0];
      *(float *)(a2 + 4) = v19 * v9;
      *(float *)(a2 + 8) = v19 * v8;
      *(float *)(a2 + 12) = v19 * v10;
      *(float *)(a2 + 16) = v19 * v11;
      *(float *)(a2 + 20) = v19 * v12;
      a2 += 24;
    }
  }
  else
  {
    while ( v16 )
    {
      v20 = *result++;
      v21 = vmla_n_f32(*(float32x2_t *)(a2 + 16), (float32x2_t)__PAIR64__(LODWORD(v12), LODWORD(v11)), v20.n128_f32[0]).n64_u64[0];
      v22 = vmla_lane_f32(*(float32x2_t *)(a2 + 24), v15, (float32x2_t)v20.n128_u64[0], 1).n64_u64[0];
      v23 = vmlaq_lane_f32(*(float32x4_t *)(a2 + 32), v14, (float32x2_t)v20.n128_u64[0], 1);
      v24 = vmlaq_laneq_f32(*(float32x4_t *)(a2 + 48), v13, v20, 2);
      v25 = vmla_laneq_f32(*(float32x2_t *)(a2 + 64), (float32x2_t)__PAIR64__(LODWORD(v12), LODWORD(v11)), v20, 2).n64_u64[0];
      v26 = vmla_laneq_f32(*(float32x2_t *)(a2 + 72), v15, v20, 3).n64_u64[0];
      v27 = vmlaq_laneq_f32(*(float32x4_t *)(a2 + 80), v14, v20, 3);
      *(float32x4_t *)a2 = vmlaq_n_f32(*(float32x4_t *)a2, v13, v20.n128_f32[0]);
      *(_QWORD *)(a2 + 16) = v21;
      *(_QWORD *)(a2 + 24) = v22;
      *(float32x4_t *)(a2 + 32) = v23;
      *(float32x4_t *)(a2 + 48) = v24;
      *(_QWORD *)(a2 + 64) = v25;
      *(_QWORD *)(a2 + 72) = v26;
      *(float32x4_t *)(a2 + 80) = v27;
      a2 += 96;
      --v16;
    }
    v7.n128_f32[1] = v9;
    v7.n128_u64[1] = __PAIR64__(LODWORD(v10), LODWORD(v8));
    v28 = 0;
    while ( (a3 & 3) != (_DWORD)v28 )
    {
      v29 = result->n128_f32[v28++];
      v30 = vmlaq_n_f32(*(float32x4_t *)a2, v7, v29);
      v31 = *(float *)(a2 + 16) + (float)(v29 * v11);
      v32 = *(float *)(a2 + 20) + (float)(v29 * v12);
      *(float32x4_t *)a2 = v30;
      *(float *)(a2 + 16) = v31;
      *(float *)(a2 + 20) = v32;
      a2 += 24;
    }
  }
  return result;
}