// 
// sub_180004FC0 from 0x180004fc0 to 0x18000521b (603 bytes)
// 
// 1 xrefs:
// > sub_180004520 @ 0x1800045a2
// 
// 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
// 

__m128 *__fastcall sub_180004FC0(
        __m128 *a1,
        __m128 *a2,
        __m128 *a3,
        __m128 *a4,
        __m128 *a5,
        int a6,
        int a7,
        int a8,
        int a9)
{
  int i; // eax
  __m128 v10; // xmm1
  __m128 v11; // xmm0
  int v12; // r10d
  __int32 v13; // eax
  __m128 *result; // rax
  int j; // r11d
  __m128 v16; // xmm1
  __m128 v17; // xmm0
  __m128 v18; // xmm4
  __m128 v19; // xmm3
  __m128 v20; // xmm1
  int v21; // r10d
  float v22; // xmm0_4
  int k; // r11d
  __m128 v24; // xmm4
  __m128 v25; // xmm0
  __m128 v26; // xmm3
  __m128 v27; // xmm4
  __m128 v28; // xmm0
  __m128 v29; // xmm2
  int v30; // r10d
  signed __int64 v31; // r9
  signed __int64 v32; // rdx
  signed __int64 v33; // r8
  int m; // r8d
  __m128i si128; // xmm0
  __m128i v36; // xmm2
  int v37; // edx
  float v38; // xmm0_4

  for ( i = a6 >> 3; i; --i )
  {
    v10 = a4[-2];
    a4 -= 2;
    v11 = a4[1];
    a1[1] = _mm_shuffle_ps(v10, v10, 27);
    *a1 = _mm_shuffle_ps(v11, v11, 27);
    a1 += 2;
  }
  v12 = a6 & 7;
  if ( (a6 & 7) != 0 )
  {
    do
    {
      v13 = a4[-1].m128_i32[3];
      a4 = (__m128 *)((char *)a4 - 4);
      a1->m128_i32[0] = v13;
      a1 = (__m128 *)((char *)a1 + 4);
      --v12;
    }
    while ( v12 );
  }
  result = a5;
  for ( j = a7 >> 3; j; --j )
  {
    a2 -= 2;
    a4 -= 2;
    v16 = _mm_mul_ps(a3[1], result[1]);
    v17 = *result;
    result += 2;
    v18 = _mm_add_ps(_mm_mul_ps(_mm_shuffle_ps(*a2, *a2, 27), _mm_shuffle_ps(*a4, *a4, 27)), v16);
    v19 = _mm_mul_ps(_mm_shuffle_ps(a2[1], a2[1], 27), _mm_shuffle_ps(a4[1], a4[1], 27));
    v20 = *a3;
    a3 += 2;
    a1[1] = v18;
    *a1 = _mm_add_ps(v19, _mm_mul_ps(v20, v17));
    a1 += 2;
  }
  v21 = a7 & 7;
  if ( (a7 & 7) != 0 )
  {
    do
    {
      a4 = (__m128 *)((char *)a4 - 4);
      v22 = a3->m128_f32[0] * result->m128_f32[0];
      a2 = (__m128 *)((char *)a2 - 4);
      a3 = (__m128 *)((char *)a3 + 4);
      result = (__m128 *)((char *)result + 4);
      a1->m128_f32[0] = (float)(a4->m128_f32[0] * a2->m128_f32[0]) + v22;
      a1 = (__m128 *)((char *)a1 + 4);
      --v21;
    }
    while ( v21 );
  }
  for ( k = a8 >> 3; k; --k )
  {
    result -= 2;
    a2 -= 2;
    v24 = _mm_mul_ps(_mm_shuffle_ps(*a2, *a2, 27), a4[1]);
    v25 = *a4;
    a4 += 2;
    v26 = _mm_mul_ps(_mm_shuffle_ps(a2[1], a2[1], 27), v25);
    v27 = _mm_sub_ps(v24, _mm_mul_ps(_mm_shuffle_ps(*result, *result, 27), a3[1]));
    v28 = *a3;
    a3 += 2;
    v29 = _mm_mul_ps(_mm_shuffle_ps(result[1], result[1], 27), v28);
    a1[1] = v27;
    *a1 = _mm_sub_ps(v26, v29);
    a1 += 2;
  }
  v30 = a8 & 7;
  if ( (a8 & 7) != 0 )
  {
    v31 = (char *)a4 - (char *)a1;
    v32 = (char *)a2 - (char *)result;
    v33 = (char *)a3 - (char *)a1;
    do
    {
      result = (__m128 *)((char *)result - 4);
      a1->m128_f32[0] = (float)(*(float *)((char *)result->m128_f32 + v32) * *(float *)((char *)a1->m128_f32 + v31))
                      - (float)(*(float *)((char *)a1->m128_f32 + v33) * result->m128_f32[0]);
      a1 = (__m128 *)((char *)a1 + 4);
      --v30;
    }
    while ( v30 );
  }
  for ( m = a9 >> 3; m; --m )
  {
    result -= 2;
    si128 = _mm_load_si128(xmmword_180144630);
    v36 = _mm_xor_si128(si128, (__m128i)_mm_shuffle_ps(*result, *result, 27));
    *(__m128i *)a1 = _mm_xor_si128(si128, (__m128i)_mm_shuffle_ps(result[1], result[1], 27));
    a1[1] = (__m128)v36;
    a1 += 2;
  }
  v37 = a9 & 7;
  if ( (a9 & 7) != 0 )
  {
    do
    {
      v38 = result[-1].m128_f32[3];
      result = (__m128 *)((char *)result - 4);
      a1 = (__m128 *)((char *)a1 + 4);
      a1[-1].m128_f32[3] = -v38;
      --v37;
    }
    while ( v37 );
  }
  return result;
}