Примеры ошибок, обнаруженных с помощью диагностики V528

V528. Pointer is compared with 'zero' value. Probably meant: *ptr != zero.

Ultimate TCP/IP

V528 It is odd that pointer to 'char' type is compared with the '\0' value. Probably meant: *m_szPassword != '\0'. UTMail ut_crammd5.cpp 333

char *CUT_CramMd5::GetClientResponse(LPCSTR ServerChallenge)
{
  ....
  if (m_szPassword != NULL)
  {
    ....
    if (m_szPassword != '\0')
    {
  ....
}

Most likely this is what should be written here: (*m_szPassword != '\0').

Similar errors can be found in some other places:

• V528 It is odd that pointer to 'char' type is compared with the '\0' value. Probably meant: *m_szPassword != '\0'. UTMail ut_crammd5.cpp 331
• V528 It is odd that pointer to 'char' type is compared with the '\0' value. Probably meant: *m_szUserName != '\0'. UTMail ut_crammd5.cpp 340
• V528 It is odd that pointer to 'char' type is compared with the '\0' value. Probably meant: *m_szUserName != '\0'. UTMail ut_crammd5.cpp 342

Fennec Media

V528 It is odd that pointer to 'char' type is compared with the '\0' value. Probably meant: *pSlash != '\0'. rtphint.cpp 346

void MP4RtpHintTrack::GetPayload(....)
{
  ....
  if (pSlash != NULL) {
    pSlash++;
    if (pSlash != '\0') {
      length = strlen(pRtpMap) - (pSlash - pRtpMap);
      *ppEncodingParams = (char *)MP4Calloc(length + 1);
      strncpy(*ppEncodingParams, pSlash, length);
    }
}

This is what should have been written here: (*pSlash != '\0').

CxImage

V528 It is odd that pointer to 'char' type is compared with the '\0' value. Probably meant: *cp != '\0'. jasper jpc_util.c 154

int jpc_atoaf(char *s, int *numvalues, double **values)
{
  ....
  while ((cp = strtok(0, delim))) {
    if (cp != '\0') {
      ++n;
    }
  }
  ....
  while ((cp = strtok(0, delim))) {
    if (cp != '\0') {
      vs[n] = atof(cp);
      ++n;
    }
  }
  ....
}

This is an example of a Copy-Paste code with an error. This is an example of potential vulnerability. Pointer dereferencing is missing. This is what should have been written here: (*cp != '\0').

Similar errors can be found in some other places:

• V528 It is odd that pointer to 'char' type is compared with the '\0' value. Probably meant: *cp != '\0'. jasper jpc_util.c 172

Notepad++

V528 It is odd that pointer to 'char' type is compared with the '\0' value. Probably meant: *headerM != '\0'. notepadPlus printer.cpp 378

TCHAR headerM[headerSize] = TEXT("");
....
size_t Printer::doPrint(bool justDoIt)
{
  ....
  if (headerM != '\0')
  ....
}

Miranda IM

V528 It is odd that pointer to 'char' type is compared with the '\0' value. Probably meant: *str != '\0'. clist_modern modern_skinbutton.cpp 282

static char *_skipblank(char * str)
{
  char * endstr=str+strlen(str);
  while ((*str==' ' || *str=='\t') && str!='\0') str++;
  while ((*endstr==' ' || *endstr=='\t') &&
         endstr!='\0' && endstr<str)
    endstr--;
  ....
}

This code is a candidate for access violation: the '*' pointer dereferencing is missing twice.

Similar errors can be found in some other places:

• V528 It is odd that pointer to 'char' type is compared with the '\0' value. Probably meant: *endstr != '\0'. clist_modern modern_skinbutton.cpp 283

Doom 3

V528 It is odd that pointer to 'char' type is compared with the '\0' value. Probably meant: *classname != '\0'. Game game_local.cpp 1250

bool idGameLocal::InitFromSaveGame(....)
{
  ....
  const char *classname =
    mapEnt->epairs.GetString( "classname" );
  if ( classname != '\0' ) {
    FindEntityDef( classname, false );
  }
  ....
}

Most likely this is what should be written here: ( *classname != '\0' )

Similar errors can be found in some other places:

• V528 It is odd that pointer to 'char' type is compared with the '\0' value. Probably meant: *soundShaderName != '\0'. Game game_local.cpp 1619

Mozilla Firefox

V528 It is odd that pointer to 'char' type is compared with the '\0' value. Probably meant: *token == '\0'. svgnumberlist.cpp 96

nsresult
SVGNumberList::SetValueFromString(const nsAString& aValue)
{
  ....
  const char *token = str.get();
  if (token == '\0') {
    return NS_ERROR_DOM_SYNTAX_ERR; // nothing between commas
  }
  ....
}

Most likely this is what should be written here: (*token == '\0')

Trans-Proteomic Pipeline

V528 It is odd that pointer to 'char' type is compared with the '\0' value. Probably meant: *pValue == '\0'. xtandem saxhandler.cpp 323

void SAXSpectraHandler::pushPeaks(....)
{
  ....
  while(*pValue != '\0' && a < m_peaksCount) {
    while(*pValue != '\0' && isspace(*pValue))
      pValue++;
    if(pValue == '\0')
      break;
    m_vfM.push_back((float)atof(pValue));
    ....
}

This is what should have been written here: if(*pValue == '\0')

Similar errors can be found in some other places:

• V528 It is odd that pointer to 'char' type is compared with the '\0' value. Probably meant: *pValue == '\0'. xtandem saxhandler.cpp 335
• V528 It is odd that pointer to 'char' type is compared with the '\0' value. Probably meant: *pValue != '\0'. xtandem loadmspectrum.cpp 727
• V528 It is odd that pointer to 'char' type is compared with the '\0' value. Probably meant: *pValue != '\0'. xtandem loadmspectrum.cpp 918

Visualization Toolkit (VTK)

V528 It is odd that pointer to 'char' type is compared with the '\0' value. Probably meant: *this->GeometryFileName == '\0'. vtkIO vtkbyureader.cxx 109

int vtkBYUReader::RequestData(....)
{
  ....
  if (this->GeometryFileName == NULL ||
      this->GeometryFileName == '\0')
  ....
}

This is what should have been written here: this->GeometryFileName[0] == '\0'

OpenCV

V528 It is odd that pointer to 'char' type is compared with the '\0' value. Probably meant: *cp != '\0'. jpc_util.c 105

int jpc_atoaf(char *s, int *numvalues, double **values)
{
  char *cp;
  ....
  while ((cp = strtok(0, delim))) {
    if (cp != '\0') {
      ++n;
    }
  }
  ....
}

Similar errors can be found in some other places:

• V528 It is odd that pointer to 'char' type is compared with the '\0' value. Probably meant: *cp != '\0'. jpc_util.c 123

Snes9x

V528 It is odd that pointer to 'wchar_t' type is compared with the L'\0' value. Probably meant: *ext == L'\0'. wsnes9x.cpp 6952

bool RegisterExt(TCHAR *ext) {
  LONG  regResult;
  TCHAR  szRegKey[PATH_MAX];
  TCHAR  szExePath[PATH_MAX];
  TCHAR   *szExeName;
  HKEY  hKey;

  if (!ext || ext==TEXT('\0'))
    return false;
  ....
}

Apache HTTP Server

V528 It is odd that pointer to 'char' type is compared with the '\0' value. Probably meant: ** ctx->re_source == '\0'. libhttpd util_expr_eval.c 167

typedef struct {
  ....
  const char **re_source;
  ....
} ap_expr_eval_ctx_t;

static const char *ap_expr_eval_re_backref(
  ap_expr_eval_ctx_t *ctx, unsigned int n)
{
  int len;
  if (!ctx->re_pmatch || !ctx->re_source ||
      *ctx->re_source == '\0' ||
      ctx->re_nmatch < n + 1)
    return "";
  ....
}

Scilab

V528 It is odd that pointer to 'char' type is compared with the '\0' value. Probably meant: ** category == '\0'. sci_xcospalload.cpp 57

int sci_xcosPalLoad(char *fname, unsigned long fname_len)
{
  ....
  char **category = NULL;
  ....
  if (category == NULL ||
      (lenCategory == 1 && *category == '\0'))
  ....
}

Scilab

V528 It is odd that pointer to 'char' type is compared with the '\0' value. Probably meant: *st->start != '\0'. pldstr.c 303

struct PLD_strtok
{
  char *start;
  char delimeter;
};

char *PLD_strtok(....)
{
  ....
  if ((st->start)&&(st->start != '\0'))
  ....
}

Similar errors can be found in some other places:

• V528 It is odd that pointer to 'char' type is compared with the '\0' value. Probably meant: ** category == '\0'. sci_xcospalload.cpp 57

TortoiseGit

V528 It is odd that pointer to 'char' type is compared with the '\0' value. Probably meant: *last_dot + 1 != '\0'. path.c 1258

void
svn_path_splitext(const char **path_root,
                  const char **path_ext,
                  const char *path,
                  apr_pool_t *pool)
{
  const char *last_dot;
  ....
  last_dot = strrchr(path, '.');
  if (last_dot && (last_dot + 1 != '\0'))
  ....
}

Most likely this is what should be written here: "if (last_dot && (*(last_dot + 1) != '\0'))" or "if (last_dot && last_dot[1] != '\0')".

TortoiseGit

V528 It is odd that pointer to 'char' type is compared with the '\0' value. Probably meant: *src_orig == '\0'. utf.c 501

static const char *
fuzzy_escape(const char *src, apr_size_t len, apr_pool_t *pool)
{
  const char *src_orig = src;
  ....
  while (src_orig < src_end)
  {
    if (! svn_ctype_isascii(*src_orig) || src_orig == '\0')
  ....
}

Most likely this is what should be written here: if (! svn_ctype_isascii(*src_orig) || *src_orig == '\0').

Miranda NG

V528 It is odd that pointer to 'wchar_t' type is compared with the L'\0' value. Probably meant: *dbv.ptszVal != L'\0'. SimpleStatusMsg msgbox.cpp 247

HWND WINAPI CreateRecentComboBoxEx(....)
{
  ....
  if (dbv.ptszVal != NULL && dbv.ptszVal != '\0') {
  ....
}

Similar errors can be found in some other places:

• V528 It is odd that pointer to 'char' type is compared with the '\0' value. Probably meant: *fp->trust != '\0'. MirOTR options.cpp 759
• V528 It is odd that pointer to 'char' type is compared with the '\0' value. Probably meant: *(end + 1) != '\0'. DbEditorPP exportimport.cpp 425
• V528 It is odd that pointer to 'char' type is compared with the '\0' value. Probably meant: *(end + 1) != '\0'. DbEditorPP exportimport.cpp 433
• And 1 additional diagnostic messages.

The GTK+ Project

V528 It is odd that pointer to 'char' type is compared with the '\0' value. Probably meant: *data->groups[0] != '\0'. gtkrecentmanager.c 979

struct _GtkRecentData
{
  ....
  gchar **groups;
  ....
};

gboolean
gtk_recent_manager_add_full (GtkRecentManager    *manager,
                             const gchar         *uri,
                             const GtkRecentData *data)
{
  ....
  if (data->groups && data->groups[0] != '\0')
      ....
  ....
}

The GTK+ Project

V528 It is odd that pointer to 'char' type is compared with the '\0' value. Probably meant: *priv->icon_list[0] == '\0'. gtkscalebutton.c 987

struct _GtkScaleButtonPrivate
{
  ....
  gchar **icon_list;
  ....
};

struct _GtkScaleButton
{
  ....
  GtkScaleButtonPrivate *priv;
};

static void
gtk_scale_button_update_icon (GtkScaleButton *button)
{
  GtkScaleButtonPrivate *priv = button->priv;
  ....
  if (!priv->icon_list || priv->icon_list[0] == '\0')
  ....
}

GCC

V528 It is odd that pointer to 'char' type is compared with the '\0' value. Probably meant: *xloc.file == '\0'. ubsan.c 1472

static bool
ubsan_use_new_style_p (location_t loc)
{
  if (loc == UNKNOWN_LOCATION)
    return false;

  expanded_location xloc = expand_location (loc);
  if (xloc.file == NULL || strncmp (xloc.file, "\1", 2) == 0
      || xloc.file == '\0' || xloc.file[0] == '\xff'
      || xloc.file[1] == '\xff')
    return false;

  return true;
}

ReOpenLDAP

V528 It is odd that pointer to 'char' type is compared with the '\0' value. Probably meant: *ludp->lud_filter != '\0'. backend.c 1525

int
fe_acl_group(....)
{
  ....
  if ( ludp->lud_filter != NULL &&
       ludp->lud_filter != '\0') // <=
  {
    ....
  }
}

Similar errors can be found in some other places:

• V528 It is odd that pointer to 'char' type is compared with the '\0' value. Probably meant: *(* lsei)->lsei_values[0] == '\0'. syntax.c 240
• V528 It is odd that pointer to 'char' type is compared with the '\0' value. Probably meant: *(* lsei)->lsei_values[1] != '\0'. syntax.c 241

GDB

V528 It is odd that pointer to 'char' type is compared with the '\0' value. Probably meant: ** argp == '\0'. location.c 527

struct event_location *
string_to_explicit_location (const char **argp, ....)
{
  ....
  if (argp == NULL
      || *argp == '\0'
      || *argp[0] != '-'
      || !isalpha ((*argp)[1])
      || ((*argp)[0] == '-' && (*argp)[1] == 'p'))
    return NULL;
  ....
}

Partio

V528 It is odd that pointer to 'char' type is compared with the '\0' value. Probably meant: *charArray[i] != '\0'. MC.cpp 109

int CharArrayLen(char** charArray)
{
  int i = 0;
  if(charArray != false)
  {
    while(charArray[i] != '\0')   // <=
    {
      i++;
    }
  }
  return i;
}

Aspell

V528 It is odd that pointer to 'char' type is compared with the '\0' value. Probably meant: *word == '\0'. check_funs.cpp 650

static void print_truncate(
  FILE * out, const char * word, int width)
{
  ....
  if (i == width-1) {
    if (word == '\0')
      put(out,' ');
    else if (word[len] == '\0')
      put(out, word, len);
    else
      put(out,'$');
    ++i;
  }
  ....
}