C代码及其C ++翻译的初始化方式不同

我有一个用C ++ 11重写的旧C代码。在较早的C代码（这是一个控制台应用程序）struct中，TC.C文件中的全局变量为。有问题的结构是

typedef struct 
{
   char* m_szTabName;         
   char* m_szCWName;          
   FILE* m_fpCWFile;          
   double* m_dNatCost;        
   double* m_dStateCost;      
   double* m_dLocalCost;      
   short m_nDateRanges;       
   short m_sNumCodes;         
   GTIME* m_gStart;           
   GTIME* m_gEnd;             
   short* m_sIcdVsn;          
   short* m_sIcdEdn;          
   MapRecord** m_tIcdMap;     
   FileHeader m_tHdr;         
   CodeHeader m_tDxTab;       
   CodeHeader m_tPrTab;       
   CCMatrixHeader m_tCCMatTab;
   NeoHeader m_tNeoTab;       
   DRGHeader m_tDRGTab;       
   CCHeader m_tCCTab;         
   CCExclHeader m_tCCExclTab; 
   LogicHeader m_tLogic;        
   CodeRecord* m_tCodeRec;    
   char* m_szLabStore;        
} Grouper;

在C代码这struct是不初始化并用于如下

#include "tabcomp.h"
#include "OtherHeaderFiles"
...
Grouper tgLoc;
...
void main(int argc, char *argv[])
{
    FILE *fpOut, *fpIn;
    ...
    CreateCodeTable(fpIn, &tgLoc.m_tDxTab, (int)DX_LAB);   
    ...
}

因此，Grouper tgLoc不会初始化，并且在调用CreateCodeTableintgLoc中的值时

tgLoc  Grouper:
+       m_szTabName     0x00000000 <NULL> char *
+       m_szCWName      0x00000000 <NULL> char *
+       m_fpCWFile      0x00000000 <NULL> _iobuf *
+       m_dNatCost      0x00000000 {???} double *
+       m_dStateCost    0x00000000 {???} double *
+       m_dLocalCost    0x00000000 {???} double *
+       m_nDateRanges   0 short
+       m_sNumCodes     0 short
+       m_gStart        0x00000000 {???} long *
+       m_gEnd          0x00000000 {???} long *
+       m_sIcdVsn       0x00000000 {???} short *
+       m_sIcdEdn       0x00000000 {???} short *
+       m_tIcdMap       0x00000000 {???} TMapRec * *
+       m_tHdr          {m_sCreateDate=0x00e80150 "" m_usRefID=0 m_cHiByte=0 '\0' ...} FileHeader
+       m_tDxTab        {m_ucCodeLen=0 '\0' m_usNumHdr=0 m_usNumCodes=0 ...} CodeHeader
+       m_tPrTab        {m_ucCodeLen=0 '\0' m_usNumHdr=0 m_usNumCodes=0 ...} CodeHeader
+       m_tCCMatTab     {m_ucNRow=0 '\0' m_ucNCol=0 '\0' m_cpCCMatrix=0x00000000 <NULL> } CCMatrixHeader
+       m_tNeoTab       {m_usNumCodes=0 m_ucNeoVal=0x00000000 <NULL> } TNeoHdr
+       m_tDRGTab       {m_usNumDRG=0 m_usNumBits=0 m_bBitVals=0x00000000 {???} ...} DRGHeader
+       m_tCCTab        {m_usNumCodes=0 m_ucSep=0x00000000 <NULL> m_useCCRow=0x00000000 {???} } CCHeader
+       m_tCCExclTab    {m_usNumCCHdr=0 m_usNumCCTail=0 m_usLowPDX=0x00000000 {???} ...} CCExclHeader
+       m_tLogic        {wFldErr=0 wDefLen=0 wFldLen=0 ...} LogicHeader
+       m_tCodeRec      0x00000000 <NULL> CodeRecord *
+       m_szLabStore    0x00000000 <NULL> char *

现在，我将代码从C应用程序移到了C ++ 11应用程序。代码可以正常运行，但是tgLoc调用的值CreateCodeTable现在是

tgLoc Grouper:
+       m_szTabName     0xcdcdcdcd <Error reading characters of string.> char *
+       m_szCWName      0xcdcdcdcd <Error reading characters of string.> char *
+       m_fpCWFile      0xcdcdcdcd {_ptr=??? _cnt=??? _base=??? ...} _iobuf *
+       m_dNatCost      0xcdcdcdcd {???} double *
+       m_dStateCost    0xcdcdcdcd {???} double *
+       m_dLocalCost    0xcdcdcdcd {???} double *
        m_nDateRanges   -12851 short
        m_sNumCodes     -12851 short
+       m_gStart        0xcdcdcdcd {???} long *
+       m_gEnd          0xcdcdcdcd {???} long *
+       m_sIcdVsn       0xcdcdcdcd {???} short *
+       m_sIcdEdn       0xcdcdcdcd {???} short *
+       m_tIcdMap       0xcdcdcdcd {???} MapRecord * *
+       m_tHdr          {m_sCreateDate=0x27570cc0 "ÍÍÍÍÍÍ... m_usRefID=52685 m_cHiByte=-51 'Í' ...} FileHeader
+       m_tDxTab        {m_ucCodeLen=205 'Í' m_usNumHdr=52685 m_usNumCodes=52685 ...} CodeHeader
+       m_tPrTab        {m_ucCodeLen=205 'Í' m_usNumHdr=52685 m_usNumCodes=52685 ...} CodeHeader
+       m_tCCMatTab     {m_ucNRow=205 'Í' m_ucNCol=205 'Í' m_cpCCMatrix=0xcdcdcdcd <Error reading characters of string.> } CCMatrixHeader
+       m_tNeoTab       {m_usNumCodes=52685 m_ucNeoVal=0xcdcdcdcd <Error reading characters of string.> } NeoHeader
+       m_tDRGTab       {m_usNumDRG=52685 m_usNumBits=52685 m_bBitVals=0xcdcdcdcd {???} ...} DRGHeader
+       m_tCCTab        {m_usNumCodes=52685 m_ucSep=0xcdcdcdcd <Error reading characters of string.> m_useCCRow=0xcdcdcdcd {...} } CCHeader
+       m_tCCExclTab    {m_usNumCCHdr=52685 m_usNumCCTail=52685 m_usLowPDX=0xcdcdcdcd {???} ...} CCExclHeader
+       m_tLogic        {wFldErr=52685 wDefLen=52685 wFldLen=52685 ...} LogicHeader
+       m_tCodeRec      0xcdcdcdcd {szCode=0xcdcdcdcd <Error reading characters of string.> ucFldNum=??? ucFldType=??? ...} CodeRecord *
+       m_szLabStore    0xcdcdcdcd <Error reading characters of string.> char *

调用之前只有一小段代码，CreateCodeTable并且不会初始化tgLoc，所以我的问题是，为什么在C ++ 11下编译的新代码与旧版C代码的初始化方式不同，如何初始化新代码，所以这tgLoc包含的旧代码，它首先使用相同的价值观？

我知道这0xCDCDCDCD是来自C运行时库的调试值。在调试版本中分配内存块时，会将其初始化为该虚假值，以期捕获错误。0xCDCDCDCD为非NULL，并且永远不是有效的内存指针。但这似乎正在影响该CreateCodeTable方法的输出。我还不明白为什么将其编译为C ++ 11时会像C一样初始化此值？ 参考

谢谢你的时间。