#include "CFTTBatchModelManager.h"

#include <arm_neon.h>

#include "CFTTBatchModelBin.h"
#include "CFTTDevice.h"
#include "CFTTGraphicsDevice.h"
#include "CFTTJobQueue.h"
#include "CFTTMaterialManager.h"
#include "CFTTMatrix32.h"
#include "CFTTShaderBuiltInConstants.h"
#include "CFTTSkinningTask.h"
#include "TFTTCacheLookup.h"
#include "global_func.h"
#include "memctrl.h"
CFTTBatchModelManager *GFXPLAYER_pBatch = nullptr;

struct UseBy_dword_5C9840_Sub {
  CFTTJobQueue *p_cfttjobqueue_0;
  CFTTBatchModelManager *field_4;
  CFTTSkinningTask *field_8;
  int field_c;
  int field_10;
};

struct UseBy_dword_5C9840 {
  int field_0;
  int field_4;
  int field_8;
  int field_c;
  int field_10;
  UseBy_dword_5C9840_Sub useby_dword_5c9840_sub_14[4];
};

MYSTATIC UseBy_dword_5C9840 dword_5C9840;
MYSTATIC CFTTSkinningTask cfttskinningtask_5AA834[1024];

// 0020FEFC
//乱写的
MYSTATIC void sub_20FEFC(CFTTJob *p_cfttjob_r0_arg, void *p_v_r1_arg) {}

//乱写的
struct TFTTBatchVertexShadow {
  int un_0;
};

//乱写的
// 0x20 00211460 080   ADDS            R2, #0x20
struct TFTTBatchVertexBump {
  float f_0;
       // 002111DC
      float f_4;
       // 00211224
      float f_8;
       // 0021126C
      short us_C;
     // 002112BC
      short us_E;
     // 00211302
      short us_10;
     // 00211348
      short us_12;
     // 没有看到给这儿赋值的地方
      short us_14;
     // 00211444
      short us_16;
     // 0021145E
      short us_18;
     // 0021138E
      short us_1A;
     // 002113D4
      short us_1C;
     // 0021141A
      short us_1E;
     // 没有看到给这儿赋值的地方，可能仅仅是为了内存对齐
};

// 0x14 00211E24 由此处确定
struct VertBone {
  short us_0;
  short us_2;
  short us_4;
  short us_6;
  short us_8;
  short us_A;
  short us_C;
  short us_E;
  short us_10;
  short us_12;
};
struct VertNorTex {
  short us_0;
  short us_2;
  short us_4;
  short us_6;
  short us_8;
  short us_A;
  short us_C;
  short us_E;
  short us_10;
  short us_12;
};
struct VertNorTexTan : public VertNorTex {
  short us_14;
  short us_16;
  short us_18;
  ushort us_1A;
};
// 0x20
struct VertNorTexBone : public VertBone {
  // VertBone tVerBone_0;
  short us_14;
  short us_16;
  short us_18;
  ushort us_1A;
  uchar uc_1C;
  // 00210FB2   LDRB.W          R4, [R9,#0x1C]
  uchar uc_1D;
  // 0021108A   LDRB.W          R3, [R9,#0x1D]
  uchar uc_1E;
  // 00211086   LDRB.W          R5, [R9,#0x1E]
  uchar uc_1F;
  // 00210FD6   LDRB.W          R1, [R9,#0x1F]
};

// 0x28 00210FA0 由此处确定
struct VertNorTexTanBone : public VertNorTexBone {
  // VertNorTexBone tVerNorTexBone_0;
  ushort us_20;
     //此处一定是ushort
      ushort us_22;
  ushort us_24;
  ushort us_26;
};
// VertNorTexTanBone
//  template <typename T>
//  void ProcessRigidVertsBumpCount(CFTTMatrix32 const *pMatrix320,
//                                  CFTTMatrix32 const *pMatrix321,
//                                  TFTTBatchVertexBump *pVertexbump2, T const
//                                  *pT3, uint ui4, TUV const *pTUV5) {
//    if (g_bNeonAvailable) {
//      // loc_2137D0
//      while (ui4) {
//        // loc_213686
//      }
//      return;
//    } else {
//      // loc_2137DA
//      // loc_213AE4

//     for(int r9_i=0;r9_i!=ui4;r9_i++) {
//       // loc_2137F8
//       float s2_f = pT3[r9_i].us_2;
//       float s6_f = pT3[r9_i].us_4;
//       float s12_f = pT3[r9_i].us_0;
//       s2_f = pMatrix320->d[1][0] * s2_f;
//       float s4_f = pMatrix320->d[0][0];
//       float s8_f = pMatrix320->d[2][0];
//       float s10_f = pMatrix320->d[3][0];
//       s2_f += s4_f * s12_f;
//       s2_f += s8_f * s6_f;
//       s2_f += s10_f;
//       pVertexbump2[r9_i].f_0 = s2_f;

//       s2_f = pT3[r9_i].us_2;
//       s6_f = pT3[r9_i].us_4;
//       s4_f = pMatrix320->d[0][1];
//       s8_f = pMatrix320->d[2][1];
//       s10_f = pMatrix320->d[3][1];
//       s2_f = pMatrix320->[1][1] * s2_f;
//       s2_f += s4_f * s12_f;
//       s2_f += s8_f * s6_f;
//       s2_f += s10_f;
//       pVertexbump2[r9_i].f_4 = s2_f;
//       // 00213896
//       s2_f = pT3[r9_i].us_2;
//       s6_f = pT3[r9_i].us_4;

//       s4_f = pMatrix320->d[0][2];
//       s8_f = pMatrix320->d[2][2];
//       s10_f = pMatrix320->d[3][2];
//       s2_f = pMatrix320->[1][2] * s2_f;
//       s2_f += s4_f * s12_f;
//       s2_f += s8_f * s6_f;
//       s2_f += s10_f;
//       pVertexbump2[r9_i].f_8 = s2_f;
//       // 002138DE
//       s2_f = pT3[r9_i].us_A;
//       s10_f = pT3[r9_i].us_8;
//       s6_f = pT3[r9_i].us_C;
//       s4_f = pMatrix321->d[0][0];
//       s8_f = pMatrix321->d[2][0];
//       s2_f = pMatrix321->d[1][0] * s2_f;
//       s2_f += s4_f * s10_f;
//       s2_f += s8_f * s6_f;
//       pVertexbump2[r9_i].us_C = s2_f;
//       // 00213926
//       s2_f = pT3[r9_i].us_A;
//       s10_f = pT3[r9_i].us_8;
//       s6_f = pT3[r9_i].us_C;
//       s4_f = pMatrix321->d[0][1];
//       s8_f = pMatrix321->d[2][1];
//       s2_f = pMatrix321->d[1][1] * s2_f;
//       s2_f += s4_f * s10_f;
//       s2_f += s8_f * s6_f;
//       pVertexbump2[r9_i].us_E = s2_f;
//       //0021396E
//       s2_f = pT3[r9_i].us_A;
//       s10_f = pT3[r9_i].us_8;
//       s6_f = pT3[r9_i].us_C;
//       s4_f = pMatrix321->d[0][2];
//       s8_f = pMatrix321->d[2][2];
//       s2_f = pMatrix321->d[1][2] * s2_f;
//       s2_f += s4_f * s10_f;
//       s2_f += s8_f * s6_f;
//       pVertexbump2[r9_i].us_10 = s2_f;
//       //002139B6
//       s2_f = pT3[r9_i].us_16;
//       s10_f = pT3[r9_i].us_14;
//       s6_f = pT3[r9_i].us_18;
//       s4_f = pMatrix321->d[0][0];
//       s8_f = pMatrix321->d[2][0];
//       s2_f = pMatrix321->d[1][0] * s2_f;
//       s2_f += s4_f * s10_f;
//       s2_f += s8_f * s6_f;
//       pVertexbump2[r9_i].us_18 = s2_f;
//       //002139FE
//       s2_f = pT3[r9_i].us_16;
//       s10_f = pT3[r9_i].us_14;
//       s6_f = pT3[r9_i].us_18;
//       s4_f = pMatrix321->d[0][1];
//       s8_f = pMatrix321->d[2][1];
//       s2_f = pMatrix321->d[1][1] * s2_f;
//       s2_f += s4_f * s10_f;
//       s2_f += s8_f * s6_f;
//       pVertexbump2[r9_i].us_1A = s2_f;
//       //00213A46
//       s2_f = pT3[r9_i].us_16;
//       s10_f = pT3[r9_i].us_14;
//       s6_f = pT3[r9_i].us_18;
//       s4_f = pMatrix321->d[0][2];
//       s8_f = pMatrix321->d[2][2];
//       s2_f = pMatrix321->d[1][2] * s2_f;
//       s2_f += s4_f * s10_f;
//       s2_f += s8_f * s6_f;
//       pVertexbump2[r9_i].us_1C = s2_f;
//       if(pTUV5){
//         //00213A90
//        pVertexbump2[r9_i].us_14= pTUV5[r9_i].x*32767.0f;
//        pVertexbump2[r9_i].us_16=  pTUV5[r9_i].y*32767.0f;
//       }else{
//         //loc_213AB4
//         pVertexbump2[r9_i].us_14= pT3[r9_i].us_10;
//        pVertexbump2[r9_i].us_16=  pT3[r9_i].us_12;;
//       }
//     }
//   }
// }

// VertBone
template <typename T>
void ProcessRigidVertsBumpCount(CFTTMatrix32 const *pMatrix320,
                                CFTTMatrix32 const *pMatrix321,
                                TFTTBatchVertexBump *pVertexbump2, T const *pT3,
                                uint ui4, TUV const *pTUV5) {
  for (int r9_i = 0; r9_i != ui4; r9_i++) {
    float s2_f = pT3[r9_i].us_2;
    float s6_f = pT3[r9_i].us_4;
    float s12_f = pT3[r9_i].us_0;
    s2_f = pMatrix320->d[1][0] * s2_f;
    float s4_f = pMatrix320->d[0][0];
    float s8_f = pMatrix320->d[2][0];
    float s10_f = pMatrix320->d[3][0];
    s2_f += s4_f * s12_f;
    s2_f += s8_f * s6_f;
    s2_f += s10_f;
    pVertexbump2[r9_i].f_0 = s2_f;
    s2_f = pT3[r9_i].us_2;
    s6_f = pT3[r9_i].us_4;
    s4_f = pMatrix320->d[0][1];
    s8_f = pMatrix320->d[2][1];
    s10_f = pMatrix320->d[3][1];
    s2_f = pMatrix320->d[1][1] * s2_f;
    s2_f += s4_f * s12_f;
    s2_f += s8_f * s6_f;
    s2_f += s10_f;
    pVertexbump2[r9_i].f_4 = s2_f;

    s2_f = pT3[r9_i].us_2;
    s6_f = pT3[r9_i].us_4;

    s4_f = pMatrix320->d[0][2];
    s8_f = pMatrix320->d[2][2];
    s10_f = pMatrix320->d[3][2];
    s2_f = pMatrix320->d[1][2] * s2_f;
    s2_f += s4_f * s12_f;
    s2_f += s8_f * s6_f;
    s2_f += s10_f;
    pVertexbump2[r9_i].f_8 = s2_f;

    if (sizeof(*pT3) == sizeof(VertBone)) {
      // 00214B00
      s2_f = pT3[r9_i + 1].us_2;
      s10_f = pT3[r9_i + 1].us_0;
      s6_f = pT3[r9_i + 1].us_4;
      s4_f = pMatrix321->d[0][0];
      s8_f = pMatrix321->d[2][0];
      s2_f = pMatrix321->d[1][0] * s2_f;
      s2_f += s4_f * s10_f;
      s2_f += s8_f * s6_f;
      pVertexbump2[r9_i].us_C = s2_f;

      s2_f = pT3[r9_i + 1].us_2;
      s10_f = pT3[r9_i + 1].us_0;
      s6_f = pT3[r9_i + 1].us_4;
      s4_f = pMatrix321->d[0][1];
      s8_f = pMatrix321->d[2][1];
      s2_f = pMatrix321->d[1][1] * s2_f;
      s2_f += s4_f * s10_f;
      s2_f += s8_f * s6_f;
      pVertexbump2[r9_i].us_E = s2_f;

      s2_f = pT3[r9_i + 1].us_2;
      s10_f = pT3[r9_i + 1].us_0;
      s6_f = pT3[r9_i + 1].us_4;
      s4_f = pMatrix321->d[0][2];
      s8_f = pMatrix321->d[2][2];
      s2_f = pMatrix321->d[1][2] * s2_f;
      s2_f += s4_f * s10_f;
      s2_f += s8_f * s6_f;
      pVertexbump2[r9_i].us_10 = s2_f;
      s2_f = pT3[r9_i + 1].us_A;
      s10_f = pT3[r9_i + 1].us_8;
      s6_f = pT3[r9_i + 1].us_C;
      s4_f = pMatrix321->d[0][0];
      s8_f = pMatrix321->d[2][0];
      s2_f = pMatrix321->d[1][0] * s2_f;
      s2_f += s4_f * s10_f;
      s2_f += s8_f * s6_f;
      pVertexbump2[r9_i].us_18 = s2_f;
      s2_f = pT3[r9_i + 1].us_A;
      s10_f = pT3[r9_i + 1].us_8;
      s6_f = pT3[r9_i + 1].us_C;
      s4_f = pMatrix321->d[0][1];
      s8_f = pMatrix321->d[2][1];
      s2_f = pMatrix321->d[1][1] * s2_f;
      s2_f += s4_f * s10_f;
      s2_f += s8_f * s6_f;
      pVertexbump2[r9_i].us_1A = s2_f;

      s2_f = pT3[r9_i + 1].us_A;
      s10_f = pT3[r9_i + 1].us_8;
      s6_f = pT3[r9_i + 1].us_C;
      s4_f = pMatrix321->d[0][2];
      s8_f = pMatrix321->d[2][2];
      s2_f = pMatrix321->d[1][2] * s2_f;
      s2_f += s4_f * s10_f;
      s2_f += s8_f * s6_f;
      pVertexbump2[r9_i].us_1C = s2_f;
    } else if (sizeof(*pT3) == sizeof(VertNorTexTanBone)) {
      VertNorTexTanBone *pVertNorTexTanBone3 = (VertNorTexTanBone *)pT3;
      s2_f = pVertNorTexTanBone3[r9_i].us_A;
      s10_f = pVertNorTexTanBone3[r9_i].us_8;
      s6_f = pVertNorTexTanBone3[r9_i].us_C;
      s4_f = pMatrix321->d[0][0];
      s8_f = pMatrix321->d[2][0];
      s2_f = pMatrix321->d[1][0] * s2_f;
      s2_f += s4_f * s10_f;
      s2_f += s8_f * s6_f;
      pVertexbump2[r9_i].us_C = s2_f;
      // 00213926
      s2_f = pVertNorTexTanBone3[r9_i].us_A;
      s10_f = pVertNorTexTanBone3[r9_i].us_8;
      s6_f = pVertNorTexTanBone3[r9_i].us_C;
      s4_f = pMatrix321->d[0][1];
      s8_f = pMatrix321->d[2][1];
      s2_f = pMatrix321->d[1][1] * s2_f;
      s2_f += s4_f * s10_f;
      s2_f += s8_f * s6_f;
      pVertexbump2[r9_i].us_E = s2_f;
      // 0021396E
      s2_f = pVertNorTexTanBone3[r9_i].us_A;
      s10_f = pVertNorTexTanBone3[r9_i].us_8;
      s6_f = pVertNorTexTanBone3[r9_i].us_C;
      s4_f = pMatrix321->d[0][2];
      s8_f = pMatrix321->d[2][2];
      s2_f = pMatrix321->d[1][2] * s2_f;
      s2_f += s4_f * s10_f;
      s2_f += s8_f * s6_f;
      pVertexbump2[r9_i].us_10 = s2_f;
      // 002139B6
      s2_f = pVertNorTexTanBone3[r9_i].us_16;
      s10_f = pVertNorTexTanBone3[r9_i].us_14;
      s6_f = pVertNorTexTanBone3[r9_i].us_18;
      s4_f = pMatrix321->d[0][0];
      s8_f = pMatrix321->d[2][0];
      s2_f = pMatrix321->d[1][0] * s2_f;
      s2_f += s4_f * s10_f;
      s2_f += s8_f * s6_f;
      pVertexbump2[r9_i].us_18 = s2_f;
      // 002139FE
      s2_f = pVertNorTexTanBone3[r9_i].us_16;
      s10_f = pVertNorTexTanBone3[r9_i].us_14;
      s6_f = pVertNorTexTanBone3[r9_i].us_18;
      s4_f = pMatrix321->d[0][1];
      s8_f = pMatrix321->d[2][1];
      s2_f = pMatrix321->d[1][1] * s2_f;
      s2_f += s4_f * s10_f;
      s2_f += s8_f * s6_f;
      pVertexbump2[r9_i].us_1A = s2_f;
      // 00213A46
      s2_f = pVertNorTexTanBone3[r9_i].us_16;
      s10_f = pVertNorTexTanBone3[r9_i].us_14;
      s6_f = pVertNorTexTanBone3[r9_i].us_18;
      s4_f = pMatrix321->d[0][2];
      s8_f = pMatrix321->d[2][2];
      s2_f = pMatrix321->d[1][2] * s2_f;
      s2_f += s4_f * s10_f;
      s2_f += s8_f * s6_f;
      pVertexbump2[r9_i].us_1C = s2_f;
    }

    if (pTUV5) {
      // 00213A90
      pVertexbump2[r9_i].us_14 = pTUV5[r9_i].x * 32767.0f;
      pVertexbump2[r9_i].us_16 = pTUV5[r9_i].y * 32767.0f;
    } else {
      if (sizeof(*pT3) == sizeof(VertBone)) {
        pVertexbump2[r9_i].us_14 = pT3[r9_i + 1].us_10;
        pVertexbump2[r9_i].us_16 = pT3[r9_i + 1].us_12;
      } else if (sizeof(*pT3) == sizeof(VertNorTexTanBone)) {
        // loc_213AB4
        pVertexbump2[r9_i].us_14 = pT3[r9_i].us_10;
        pVertexbump2[r9_i].us_16 = pT3[r9_i].us_12;
      }
    }
  }
}

struct TFTTBatchVertex {
  float f_0;
  float f_4;
  float f_8;
  short us_C;
  short us_E;
  short us_10;
  short us_12;
  short us_14;
  short us_16;
};

template <typename T>
void ProcessRigidVertsCount(CFTTMatrix32 const *pMatrix320,
                            CFTTMatrix32 const *pMatrix321,
                            TFTTBatchVertex *pVertex2, T const *pT3, uint u4,
                            TUV const *pTUV5) {
  //LOGE("g_bNeonAvailable%x", g_bNeonAvailable);
  if (g_bNeonAvailable) {
    float32x4_t Q8 = vld1q_f32((float *)pMatrix321);
    float32x4_t Q12 = vld1q_f32(((float *)pMatrix321) + 4);
    float32x4_t Q10 = vld1q_f32(((float *)pMatrix321) + 8);

    float32x4_t Q9 = vld1q_f32((float *)pMatrix320);
    float32x4_t Q14 = vld1q_f32(((float *)pMatrix320) + 4);
    float32x4_t Q13 = vld1q_f32(((float *)pMatrix320) + 8);
    float32x4_t Q11 = vld1q_f32(((float *)pMatrix320) + 12);
    for (int i = 0; i != u4; i++) {
      //loc_215B78
      float32x4_t Q2;
      memcpy(&Q2, &Q11, sizeof(Q2));
      int16x4_t d30 = vld1_dup_s16((short *)&pT3[i]);
      float32x4_t Q15 = vmovl_s16(d30);
      Q15 = vcvtq_f32_s32(Q15);
      Q2 = vmlaq_f32(Q2, Q9, Q15);
      int16x4_t D1 = vld1_dup_s16(((short *)&pT3[i]) + 1);
      float32x4_t Q1 = vmovl_s16(D1);
      Q15 = vcvtq_f32_s32(Q1);
      Q2 = vmlaq_f32(Q2, Q14, Q15);
      D1 = vld1_dup_s16(((short *)&pT3[i]) + 2);
      Q1 = vmovl_s16(D1);
      Q15 = vcvtq_f32_s32(Q1);
      Q2 = vmlaq_f32(Q2, Q13, Q15);
      memcpy(&pVertex2[i], &Q2, sizeof(Q2));
      if (typeid(T) == typeid(VertBone)) {
        d30 = vld1_dup_s16(((short *)&pT3[i]) + 0xA);
        Q15 = vmovl_s16(d30);
        Q15 = vcvtq_f32_s32(Q15);
        Q15 = vmulq_f32(Q15, Q8);
        D1 = vld1_dup_s16(((short *)&pT3[i]) + 0xB);
        Q1 = vmovl_s16(D1);
        Q1 = vcvtq_f32_s32(Q1);
        Q15 = vmlaq_f32(Q15, Q12, Q1);
        D1 = vld1_dup_s16(((short *)&pT3[i]) + 0xC);
        Q2 = vmovl_s16(D1);
        Q1 = vcvtq_f32_s32(Q2);
        Q15 = vmlaq_f32(Q15, Q10, Q1);
        Q15 = vcvtq_s32_f32(Q15);
        d30 = vmovn_s32(Q15);
        memcpy(&pVertex2[i].us_C, &d30, sizeof(d30));
      }else {
        d30 = vld1_dup_s16(((short *)&pT3[i]) + 0x4);
        Q15 = vmovl_s16(d30);
        Q15 = vcvtq_f32_s32(Q15);
        Q15 = vmulq_f32(Q15, Q8);

        D1 = vld1_dup_s16(((short *)&pT3[i]) + 0x5);
        Q1 = vmovl_s16(D1);
        Q1 = vcvtq_f32_s32(Q1);
        Q15 = vmlaq_f32(Q15, Q12, Q1);
        D1 = vld1_dup_s16(((short *)&pT3[i]) + 0x6);
        Q2 = vmovl_s16(D1);
        Q1 = vcvtq_f32_s32(Q2);
        Q15 = vmlaq_f32(Q15, Q10, Q1);
        Q15 = vcvtq_s32_f32(Q15);
        d30 = vmovn_s32(Q15);
        memcpy(&pVertex2[i].us_C, &d30, sizeof(d30));
      }
      if (pTUV5) {
        pVertex2[i].us_14 = pTUV5[i].x * 32767.0f;
        pVertex2[i].us_16 = pTUV5[i].x * 32767.0f;
      } else {
        if (typeid(T) == typeid(VertBone)) {
          pVertex2[i].us_14 = pT3[i + 1].us_10;
          pVertex2[i].us_16 = pT3[i + 1].us_12;
        } else {
          pVertex2[i].us_14 = pT3[i].us_10;
          pVertex2[i].us_16 = pT3[i].us_12;
        }
      }
    }
  } else {
    for (int r9_i = 0; r9_i != u4; r9_i++) {
      float s2_f = pT3[r9_i].us_2;
      float s6_f = pT3[r9_i].us_4;
      float s12_f = pT3[r9_i].us_0;
      s2_f = pMatrix320->d[1][0] * s2_f;
      float s4_f = pMatrix320->d[0][0];
      float s8_f = pMatrix320->d[2][0];
      float s10_f = pMatrix320->d[3][0];
      s2_f += s4_f * s12_f;
      s2_f += s8_f * s6_f;
      s2_f += s10_f;
      pVertex2[r9_i].f_0 = s2_f;
      s2_f = pT3[r9_i].us_2;
      s6_f = pT3[r9_i].us_4;
      s4_f = pMatrix320->d[0][1];
      s8_f = pMatrix320->d[2][1];
      s10_f = pMatrix320->d[3][1];
      s2_f = pMatrix320->d[1][1] * s2_f;
      s2_f += s4_f * s12_f;
      s2_f += s8_f * s6_f;
      s2_f += s10_f;
      pVertex2[r9_i].f_4 = s2_f;
      s2_f = pT3[r9_i].us_2;
      s6_f = pT3[r9_i].us_4;
      s4_f = pMatrix320->d[0][2];
      s8_f = pMatrix320->d[2][2];
      s10_f = pMatrix320->d[3][2];
      s2_f = pMatrix320->d[1][2] * s2_f;
      s2_f += s4_f * s12_f;
      s2_f += s8_f * s6_f;
      s2_f += s10_f;
      pVertex2[r9_i].f_8 = s2_f;
      if (typeid(T) == typeid(VertBone)) {
        s2_f = pT3[r9_i + 1].us_2;
        s10_f = pT3[r9_i + 1].us_0;
        s6_f = pT3[r9_i + 1].us_4;
        s4_f = pMatrix321->d[0][0];
        s8_f = pMatrix321->d[2][0];
        s2_f = pMatrix321->d[1][0] * s2_f;
        s2_f += s4_f * s10_f;
        s2_f += s8_f * s6_f;
        pVertex2[r9_i].us_C = s2_f;
        s2_f = pT3[r9_i + 1].us_2;
        s10_f = pT3[r9_i + 1].us_0;
        s6_f = pT3[r9_i + 1].us_4;
        s4_f = pMatrix321->d[0][1];
        s8_f = pMatrix321->d[2][1];
        s2_f = pMatrix321->d[1][1] * s2_f;
        s2_f += s4_f * s10_f;
        s2_f += s8_f * s6_f;
        pVertex2[r9_i].us_E = s2_f;

        s2_f = pT3[r9_i + 1].us_2;
        s10_f = pT3[r9_i + 1].us_0;
        s6_f = pT3[r9_i + 1].us_4;
        s4_f = pMatrix321->d[0][2];
        s8_f = pMatrix321->d[2][2];
        s2_f = pMatrix321->d[1][2] * s2_f;
        s2_f += s4_f * s10_f;
        s2_f += s8_f * s6_f;
        pVertex2[r9_i].us_10 = s2_f;
        if (typeid(T) == typeid(VertBone)) {
          pVertex2[r9_i].us_14 = pT3[r9_i + 1].us_10;
          pVertex2[r9_i].us_16 = pT3[r9_i + 1].us_12;
        } else {
          pVertex2[r9_i].us_14 = pT3[r9_i].us_10;
          pVertex2[r9_i].us_16 = pT3[r9_i].us_12;
        }

      } else {
        T *pVertNorTexTanBone3 = (T *)pT3;
        s2_f = pVertNorTexTanBone3[r9_i].us_A;
        s10_f = pVertNorTexTanBone3[r9_i].us_8;
        s6_f = pVertNorTexTanBone3[r9_i].us_C;
        s4_f = pMatrix321->d[0][0];
        s8_f = pMatrix321->d[2][0];
        s2_f = pMatrix321->d[1][0] * s2_f;
        s2_f += s4_f * s10_f;
        s2_f += s8_f * s6_f;
        pVertex2[r9_i].us_C = s2_f;

        s2_f = pVertNorTexTanBone3[r9_i].us_A;
        s10_f = pVertNorTexTanBone3[r9_i].us_8;
        s6_f = pVertNorTexTanBone3[r9_i].us_C;
        s4_f = pMatrix321->d[0][1];
        s8_f = pMatrix321->d[2][1];
        s2_f = pMatrix321->d[1][1] * s2_f;
        s2_f += s4_f * s10_f;
        s2_f += s8_f * s6_f;
        pVertex2[r9_i].us_E = s2_f;

        s2_f = pVertNorTexTanBone3[r9_i].us_A;
        s10_f = pVertNorTexTanBone3[r9_i].us_8;
        s6_f = pVertNorTexTanBone3[r9_i].us_C;
        s4_f = pMatrix321->d[0][2];
        s8_f = pMatrix321->d[2][2];
        s2_f = pMatrix321->d[1][2] * s2_f;
        s2_f += s4_f * s10_f;
        s2_f += s8_f * s6_f;
        pVertex2[r9_i].us_10 = s2_f;
      }
      // else if(typeid(T) == typeid(VertNorTex)){
      //   //002159FA
      //   VertNorTex *pVertNorTex3 = (VertNorTex *)pT3;
      //   s2_f = pVertNorTex3[r9_i].us_A;
      //   s10_f = pVertNorTex3[r9_i].us_8;
      //   s6_f = pVertNorTex3[r9_i].us_C;
      //   s4_f = pMatrix321->d[0][0];
      //   s8_f = pMatrix321->d[2][0];
      //   s2_f = pMatrix321->d[1][0] * s2_f;
      //   s2_f += s4_f * s10_f;
      //   s2_f += s8_f * s6_f;
      //   pVertex2[r9_i].us_C = s2_f;

      //   s2_f = pVertNorTex3[r9_i].us_A;
      //   s10_f = pVertNorTex3[r9_i].us_8;
      //   s6_f = pVertNorTex3[r9_i].us_C;
      //   s4_f = pMatrix321->d[0][1];
      //   s8_f = pMatrix321->d[2][1];
      //   s2_f = pMatrix321->d[1][1] * s2_f;
      //   s2_f += s4_f * s10_f;
      //   s2_f += s8_f * s6_f;
      //   pVertex2[r9_i].us_E = s2_f;

      //   s2_f = pVertNorTex3[r9_i].us_A;
      //   s10_f = pVertNorTex3[r9_i].us_8;
      //   s6_f = pVertNorTex3[r9_i].us_C;
      //   s4_f = pMatrix321->d[0][2];
      //   s8_f = pMatrix321->d[2][2];
      //   s2_f = pMatrix321->d[1][2] * s2_f;
      //   s2_f += s4_f * s10_f;
      //   s2_f += s8_f * s6_f;
      //   pVertex2[r9_i].us_10 = s2_f;
      // }
      if (pTUV5) {
        // 00213A90
        pVertex2[r9_i].us_14 = pTUV5[r9_i].x * 32767.0f;
        pVertex2[r9_i].us_16 = pTUV5[r9_i].y * 32767.0f;
      } else {
        if (typeid(T) == typeid(VertBone)) {
          pVertex2[r9_i].us_14 = pT3[r9_i + 1].us_10;
          pVertex2[r9_i].us_16 = pT3[r9_i + 1].us_12;
        } else {
          pVertex2[r9_i].us_14 = pT3[r9_i].us_10;
          pVertex2[r9_i].us_16 = pT3[r9_i].us_12;
        }
        // }else if(typeid(T) == typeid(VertNorTex)){
        //   //loc_215AF0
        //   pVertex2[r9_i].us_14 = pT3[r9_i].us_10;
        //   pVertex2[r9_i].us_16 = pT3[r9_i].us_12;
        // }
      }
    }
  }
}
void testProcessRigidVertsBumpCount() {
  CFTTMatrix32 a1;
  CFTTMatrix32 a2;
  TFTTBatchVertexBump a3;
  VertBone a4;
  TUV a6;
  VertNorTexTanBone a4_1;
  VertNorTex a4_2;
  TFTTBatchVertex a3_1;
  VertNorTexTan a4_3;
  ProcessRigidVertsCount(&a1, &a2, &a3_1, &a4_3, 5, &a6);
  ProcessRigidVertsCount(&a1, &a2, &a3_1, &a4_2, 5, &a6);
  ProcessRigidVertsBumpCount(&a1, &a2, &a3, &a4, 5, &a6);
  ProcessRigidVertsBumpCount(&a1, &a2, &a3, &a4_1, 5, &a6);
}

// // 00213594
// //乱写的
// void ProcessRigidVertsBumpCount(
//     uint ui_r0_arg, CFTTMatrix32 const *p_cfttmatrix32_r1_arg,
//     CFTTMatrix32 const *p_cfttmatrix32_r2_arg,
//     TFTTBatchVertexBump *p_tfttbatchvertexbump_r3_arg, void const *p_v_arg_0,
//     uint ui_arg_4, TUV const *p_tuv_arg_8) {
//   float tmp = ui_r0_arg + p_cfttmatrix32_r1_arg->d[0][0] +
//               p_cfttmatrix32_r2_arg->d[0][0] +
//               p_tfttbatchvertexbump_r3_arg->un_0 + ((int)p_v_arg_0) +
//               ui_arg_4 + p_tuv_arg_8->x;
//   LOGI("ProcessRigidVertsBumpCount %f", tmp);
// }

//-------------------------------------------------------------------------------------------------
// 0020F240 //^_- 经过单元测试，功能正常
// 0020F2B6 指令顺序不同
// 0020F2DE 指令顺序不同
// CFTTBatchModelManager::CFTTBatchModelManager(int d_r1_arg, bool b_r2_arg,
//                                              int d_r3_arg,
//                                              unsigned int ud_arg_0,
//                                              bool b_arg_4, bool b_arg_8)
//     : cfttbmmmatrixcache_0(b_arg_8, 0x200, 0) {}
//-------------------------------------------------------------------------------------------------
// 0020F378 //^_^
CFTTBatchModelManager::~CFTTBatchModelManager() {}
//-------------------------------------------------------------------------------------------------
// 0020F414 //^_^
bool CFTTBatchModelManager::PerformFlushOnFull() { return false; }
//-------------------------------------------------------------------------------------------------
// 0020F450 //^_^
void CFTTBatchModelManager::End() { LOGE("End%x", this); }
//-------------------------------------------------------------------------------------------------
// 0020F480 //^_^
void CFTTBatchModelManager::Begin() { LOGE("Begin%x", this); }
//-------------------------------------------------------------------------------------------------
// 0020F4EC //^_-
//单元测试等级: 目测游戏表现
//单元测试内容: 启动游戏到第2关（包括第2关）
//单元测试结果: 正常
//已测试分支:
//  0020F524
//  0020F558
void CFTTBatchModelManager::FlushThreaded() {}
//-------------------------------------------------------------------------------------------------
// 0020F5B4 //^_^
void CFTTBatchModelManager::Flush(int i_r1_arg, bool b_r2_arg, bool b_r3_arg) {}
//-------------------------------------------------------------------------------------------------
// 0020F670 //^_^
void CFTTBatchModelManager::ResetCache(bool b_r1_arg) {
  LOGE("%x", this);
  LOGE("%x", b_r1_arg);
}
//-------------------------------------------------------------------------------------------------
// 0020F68C //^_^
void CFTTBatchModelManager::SetMatrix(
    CFTTMatrix32 const &p_cfttmatrix32_r1_arg) {}
//-------------------------------------------------------------------------------------------------
// 0020F694 //^_-
//单元测试等级: 目测游戏表现
//单元测试内容: 启动游戏到第2关（包括第2关）
//单元测试结果: 正常
//已测试分支:
//  0020F6D0
//  0020F6FA
//  0020F726
//  0020F72E
//  0020F79E
//  0020F6BC
//  0020F6C6
//  0020F7E4
int CFTTBatchModelManager::AddModelVerts(
    int i_r1_arg, CFTTModel *p_cfttmodel_r2_arg, int i_r3_arg,
    CFTTMatrix32 const *p_cfttmatrix32_arg_0, TUV *p_tuv_arg_4,
    TFTTCacheLookup tfttcachelookup_arg_8, callback_1 callback_arg_10,
    CFTTMatrix32 const *p_cfttmatrix32_arg_14) {
  return 0;
}
//-------------------------------------------------------------------------------------------------
// 0020F8E4 //^_-
//单元测试等级: 目测游戏表现
//单元测试内容: 启动游戏到第2关（包括第2关）
//单元测试结果: 正常
//已测试分支:
//  0020F944
//  0020F97C
//  0020F994
//  0020F9B6
//  0020F91E
//  0020FA14
//  0020FA24
//  0020FA00
//  0020FA04
void CFTTBatchModelManager::SkinVertsBoned(
    CFTTModel *p_cfttmodel_r1_arg, int i_r2_arg, float *p_f_r3_arg,
    void *p_v_arg_0, CFTTMatrix32 const *p_cfttmatrix32_arg_4,
    CFTTBMMMatrixCache::StartCount cfttbmmmatrixcache_startcount_arg_8,
    CFTTBMMMatrixCache::StartCount cfttbmmmatrixcache_startcount_arg_10,
    TUV *p_tuv_arg_18, TFTTCacheLookup tfttcachelookup_arg_1c,
    callback_1 callback_1_arg_24) {}
//-------------------------------------------------------------------------------------------------
// 0020FA32 //^_^
void CFTTBatchModelManager::CopyVerts(
    CFTTModel *p_cfttmodel_r1_arg, int i_r2_arg, float *p_f_r3_arg,
    void *p_v_arg_0, CFTTMatrix32 const *p_cfttmatrix32_arg_4,
    CFTTBMMMatrixCache::StartCount cfttbmmmatrixcache_startcount_arg_8,
    CFTTBMMMatrixCache::StartCount cfttbmmmatrixcache_startcount_arg_10,
    TUV *p_tuv_arg_18, TFTTCacheLookup tfttcachelookup_arg_1c,
    callback_1 callback_1_arg_24) {}
//-------------------------------------------------------------------------------------------------
// 0020FAD8 //^_-
//单元测试等级: 目测游戏表现
//单元测试内容: 启动游戏到第2关（包括第2关）
//单元测试结果: 正常
//已测试分支:
//  0020FB00
//  0020FB0A
//  0020FB3C
//  0020FB4E
//  0020FB60
//  0020FCB6
//  0020FCC0
//  0020FCCE
//  0020FCD4
//  0020FCD8
//  0020FB76
//  0020FB80
//  0020FBCC
//  0020FBA2
//  0020FBD8
//  0020FB20
void CFTTBatchModelManager::AddModel(
    CFTTModel *p_cfttmodel_r1_arg, CFTTMatrix32 const *p_cfttmatrix32_r2_arg,
    TUV *p_tuv_r3_arg, TFTTCacheInstance tfttcacheinstance_arg_0, int i_arg_4,
    int i_arg_8, callback_1 callback_arg_c,
    CFTTMatrix32 const *p_cfttmatrix32_arg_10) {

  LOGE(" CFTTBatchModelManager::AddModel%p%p%p%p%x%x%p%p",p_cfttmodel_r1_arg,p_cfttmatrix32_r2_arg,p_tuv_r3_arg,
  tfttcacheinstance_arg_0,i_arg_4,i_arg_8,callback_arg_c,p_cfttmatrix32_arg_10);
  // __android_log_print(ANDROID_LOG_INFO, "sh_log",
  // "CFTTBatchModelManager::AddModel end");
}
//-------------------------------------------------------------------------------------------------
// 0020FD4A //^_-
//单元测试等级: 目测游戏表现
//单元测试内容: 启动游戏到第2关（包括第2关）
//单元测试结果: 正常
//已测试分支:
//  0020FD6A
//  0020FD76
//  0020FD8A
//  0020FD90
//  0020FD9C
//  0020FDA4
//  0020FDD6
void CFTTBatchModelManager::AddModelX(
    CFTTModel *p_cfttmodel_r1_arg, int i_r2_arg,
    CFTTMatrix32 const *p_cfttmatrix32_r3_arg, TUV *p_tuv_arg_0,
    callback_1 callback_1_arg_4, CFTTMatrix32 const *p_cfttmatrix32_arg_8) {}
//-------------------------------------------------------------------------------------------------
// 0020FE74
void CFTTBatchModelManager::SkinVertsBlend(
    CFTTModel *, int, float *, void *, CFTTMatrix32 const *,
    CFTTBMMMatrixCache::StartCount, CFTTBMMMatrixCache::StartCount, TUV *,
    TFTTCacheLookup, void (*)(void *, CFTTModel *, int, CFTTMatrix32 const *)) {
}
//-------------------------------------------------------------------------------------------------
// 0020FE76
void CFTTBatchModelManager::SkinVerts(
    CFTTModel *, int, float *, void *, CFTTMatrix32 const *,
    CFTTBMMMatrixCache::StartCount, CFTTBMMMatrixCache::StartCount, TUV *,
    TFTTCacheLookup, void (*)(void *, CFTTModel *, int, CFTTMatrix32 const *)) {
}
//-------------------------------------------------------------------------------------------------
// 0020FE6E //^_^
void CFTTBatchModelManager::SetFlushOnFullCB(
    pFlushOnFullCBCallBack pFlushOnFullCBCallBack_r1, void *p_v_r2) {
  this->field_D4 = pFlushOnFullCBCallBack_r1;
  this->field_D8 = p_v_r2;
}
//-------------------------------------------------------------------------------------------------
// 0020F494 经过单元测试，功能正常
//^_^-
// 0020F4D8处我的代码只取了一次this->field_70，而原始代码取了两次
CFTTBatchModelBin *CFTTBatchModelManager::AddBin(
    CFTTBatchModelBin *p_cfttbatchmodelbin_r1_arg, int d_r2_arg, uint ud_r3_arg,
    int d_arg_0,  // shader mat id
    int d_arg_4, bool b_arg_8, bool b_arg_c) {
  return 0;
}
//-------------------------------------------------------------------------------------------------
//目前还没有实现给dword_5C9840.field_4、dword_5C9840.field_8等字段赋值的函数
//接上，编译器优化会导致这些字段没了，所以自己写一个函数防止编译器优化
void TestUseBy_dword_5C9840() {
  printf("%p", &dword_5C9840.field_0);
  printf("%p", &dword_5C9840.field_4);
  printf("%p", &dword_5C9840.field_8);
  printf("%p", &dword_5C9840.field_c);
  // printf("%p", &dword_5C9840.useby_dword_5c9840_sub_14[0].p_cfttjobqueue_0);
}
//-------------------------------------------------------------------------------------------------
// 0020F1EC //^_^
void CFTTBatchModelManager::InitialiseCodePath() {}
//-------------------------------------------------------------------------------------------------