// 
// sub_100DEC from 0x100dec to 0x101080 (660 bytes)
// 
// 1 xrefs:
// > <???> @ 0x118b18
// 
// 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
// 

float __fastcall sub_100DEC(float *a1, float *s, unsigned int a3, float32x4_t *a4, __int64 a5, __int64 a6, int a7)
{
  float32x4_t v7; // q24
  float32x4_t v8; // q25
  float32x4_t v9; // q26
  float32x4_t v10; // q27
  float32x4_t v11; // q28
  float32x4_t v12; // q29
  float *v15; // x20
  float32x4_t v17; // q0
  unsigned int v18; // w8
  float32x4_t *v19; // x9
  float32x4_t *i; // x10
  float32x4_t v21; // q3
  float32x4_t v22; // q4
  int8x16_t v23; // q5
  int8x16_t v24; // q6
  int8x16_t v25; // q7
  int8x16_t v26; // q16
  int32x4_t v27; // q23
  int8x16_t v28; // q17
  int8x16_t v29; // q4
  int32x4_t v30; // q6
  int8x16_t v31; // q18
  int8x16_t v32; // q19
  float32x4_t v33; // q16
  float32x4_t v34; // q6
  int32x4_t v35; // q23
  int8x16_t v36; // q20
  int8x16_t v37; // q21
  int32x4_t v38; // q4
  int8x16_t v39; // q22
  int8x16_t v40; // q3
  float32x4_t v41; // q19
  float32x4_t v42; // q4
  int32x4_t v43; // q23
  int32x4_t v44; // q21
  float32x4_t v45; // q1
  float32x4_t v46; // q2
  int v47; // w8
  float v48; // s1
  float v49; // s4
  float v50; // s5
  float v51; // s3
  float v52; // s6
  float v53; // s2
  float v54; // s0
  float v55; // s7
  float v56; // s1
  float v57; // s3
  float v58; // s4
  float32x4_t v60; // [xsp+0h] [xbp-80h]
  float32x4_t v61; // [xsp+10h] [xbp-70h]
  float32x4_t v62; // [xsp+20h] [xbp-60h]
  float32x4_t v63; // [xsp+30h] [xbp-50h]
  float32x4_t v64; // [xsp+40h] [xbp-40h]
  float32x4_t v65; // [xsp+50h] [xbp-30h]

  v7 = *a4;
  v8 = a4[8];
  v9 = a4[16];
  v10 = a4[24];
  v11 = a4[32];
  v12 = a4[40];
  v15 = s;
  if ( a7 )
  {
    v65 = *a4;
    v63 = a4[16];
    v64 = a4[8];
    v61 = a4[32];
    v62 = a4[24];
    v60 = a4[40];
    memset(s, 0, 24LL * a3);
    v12 = v60;
    v11 = v61;
    v10 = v62;
    v9 = v63;
    v8 = v64;
    v7 = v65;
  }
  v18 = a3 >> 1;
  v19 = (float32x4_t *)v15;
  for ( i = (float32x4_t *)a1; v18; v19 += 3 )
  {
    v22 = *i;
    v21 = i[1];
    i += 2;
    --v18;
    v23 = vmulq_f32(v22, v7);
    v24 = vmulq_f32(v22, v8);
    v25 = vmulq_f32(v22, v9);
    v26 = vmulq_f32(v22, v10);
    v27 = vextq_s8(v23, vextq_s8(v24, v23, 8u), 8u);
    v23.n128_u64[1] = v24.n128_u64[0];
    v28 = vmulq_f32(v22, v11);
    v29 = vmulq_f32(v22, v12);
    v30 = vextq_s8(v25, vextq_s8(v26, v25, 8u), 8u);
    v31 = vmulq_f32(v21, v7);
    v32 = vmulq_f32(v21, v8);
    v25.n128_u64[1] = v26.n128_u64[0];
    v33 = vuzp1q_s32(v27, v30);
    v34 = vuzp2q_s32(v27, v30);
    v35 = vextq_s8(v28, vextq_s8(v29, v28, 8u), 8u);
    v28.n128_u64[1] = v29.n128_u64[0];
    v36 = vmulq_f32(v21, v9);
    v37 = vmulq_f32(v21, v10);
    v38 = vextq_s8(v31, vextq_s8(v32, v31, 8u), 8u);
    v39 = vmulq_f32(v21, v11);
    v40 = vmulq_f32(v21, v12);
    v31.n128_u64[1] = v32.n128_u64[0];
    v41 = vuzp1q_s32(v35, v38);
    v42 = vuzp2q_s32(v35, v38);
    v43 = vextq_s8(v36, vextq_s8(v37, v36, 8u), 8u);
    v36.n128_u64[1] = v37.n128_u64[0];
    v44 = vextq_s8(v39, vextq_s8(v40, v39, 8u), 8u);
    v39.n128_u64[1] = v40.n128_u64[0];
    v17 = vaddq_f32(vuzp1q_s32(v23, v25), vaddq_f32(vaddq_f32(vaddq_f32(*v19, vuzp2q_s32(v23, v25)), v33), v34));
    v45 = vaddq_f32(vuzp1q_s32(v28, v31), vaddq_f32(vaddq_f32(vaddq_f32(v19[1], vuzp2q_s32(v28, v31)), v41), v42));
    v46 = vaddq_f32(
            vuzp1q_s32(v36, v39),
            vaddq_f32(vaddq_f32(vaddq_f32(v19[2], vuzp2q_s32(v36, v39)), vuzp1q_s32(v43, v44)), vuzp2q_s32(v43, v44)));
    *v19 = v17;
    v19[1] = v45;
    v19[2] = v46;
  }
  v47 = a3 & 1;
  if ( (a3 & 1) != 0 )
  {
    do
    {
      v48 = a1[1];
      v49 = a1[2];
      v50 = a1[3];
      v51 = (float)((float)(a4[24].n128_f32[0] * *a1) + (float)(a4[24].n128_f32[1] * v48))
          + (float)(a4[24].n128_f32[2] * v49);
      v52 = (float)((float)(a4[32].n128_f32[0] * *a1) + (float)(a4[32].n128_f32[1] * v48))
          + (float)(a4[32].n128_f32[2] * v49);
      v53 = (float)((float)((float)(a4->n128_f32[0] * *a1) + (float)(a4->n128_f32[1] * v48))
                  + (float)(a4->n128_f32[2] * v49))
          + (float)(a4->n128_f32[3] * v50);
      v54 = (float)((float)((float)(a4[40].n128_f32[0] * *a1) + (float)(a4[40].n128_f32[1] * v48))
                  + (float)(a4[40].n128_f32[2] * v49))
          + (float)(a4[40].n128_f32[3] * v50);
      v55 = (float)((float)((float)((float)(a4[8].n128_f32[0] * *a1) + (float)(a4[8].n128_f32[1] * v48))
                          + (float)(a4[8].n128_f32[2] * v49))
                  + (float)(a4[8].n128_f32[3] * v50))
          + v15[1];
      v56 = (float)((float)((float)((float)(a4[16].n128_f32[0] * *a1) + (float)(a4[16].n128_f32[1] * v48))
                          + (float)(a4[16].n128_f32[2] * v49))
                  + (float)(a4[16].n128_f32[3] * v50))
          + v15[2];
      v57 = (float)(v51 + (float)(a4[24].n128_f32[3] * v50)) + v15[3];
      --v47;
      v58 = (float)(v52 + (float)(a4[32].n128_f32[3] * v50)) + v15[4];
      v17.n128_f32[0] = v54 + v15[5];
      a1 += 4;
      *v15 = v53 + *v15;
      v15[1] = v55;
      v15[2] = v56;
      v15[3] = v57;
      v15[4] = v58;
      v15[5] = v17.n128_f32[0];
      v15 += 6;
    }
    while ( v47 );
  }
  return v17.n128_f32[0];
}