Problem report #109: base/src/emacs/src/coding.c (decode_coding_utf_16); UNINIT

Problem report #109: base/src/emacs/src/coding.c (decode_coding_utf_16); UNINIT

[Dan Nicolaescu]

CID: 109
Checker: UNINIT (help)
File: base/src/emacs/src/coding.c
Function: decode_coding_utf_16
Description: Using uninitialized value "consumed_chars_base"


Event var_decl: Declared variable "consumed_chars_base" without initializer
Also see events: [uninit_use]

1640 	  int consumed_chars = 0, consumed_chars_base;
1641 	  int multibytep = coding->src_multibyte;
1642 	  enum utf_bom_type bom = CODING_UTF_16_BOM (coding);
1643 	  enum utf_16_endian_type endian = CODING_UTF_16_ENDIAN (coding);
1644 	  int surrogate = CODING_UTF_16_SURROGATE (coding);
1645 	  Lisp_Object attr, charset_list;

At conditional (1): "((0), (((0), ((Vcoding_system_hash_table & -8)->key_and_value & -8))->contents[((2 * (coding)->id) + 1)] & -8))->contents[2] == Qdos" taking true path

1646 	  int eol_crlf = EQ (CODING_ID_EOL_TYPE (coding->id), Qdos);
1647 	  int byte_after_cr1 = -1, byte_after_cr2 = -1;
1648 	

At conditional (2): "0" taking false path

1649 	  CODING_GET_INFO (coding, attr, charset_list);
1650 	

At conditional (3): "bom == 2" taking true path

1651 	  if (bom == utf_with_bom)
1652 	    {
1653 	      int c, c1, c2;
1654 	
1655 	      src_base = src;

At conditional (4): "src == src_end" taking true path
At conditional (5): "src_base < src" taking true path

1656 	      ONE_MORE_BYTE (c1);
1657 	      ONE_MORE_BYTE (c2);
1658 	      c = (c1 << 8) | c2;
1659 	
1660 	      if (endian == utf_16_big_endian
1661 		  ? c != 0xFEFF : c != 0xFFFE)
1662 		{
1663 		  /* The first two bytes are not BOM.  Treat them as bytes
1664 		     for a normal character.  */
1665 		  src = src_base;
1666 		  coding->errors++;
1667 		}
1668 	      CODING_UTF_16_BOM (coding) = utf_without_bom;
1669 	    }
1670 	  else if (bom == utf_detect_bom)
1671 	    {
1672 	      /* We have already tried to detect BOM and failed in
1673 		 detect_coding.  */
1674 	      CODING_UTF_16_BOM (coding) = utf_without_bom;
1675 	    }
1676 	
1677 	  while (1)
1678 	    {
1679 	      int c, c1, c2;
1680 	
1681 	      src_base = src;
1682 	      consumed_chars_base = consumed_chars;
1683 	
1684 	      if (charbuf + 2 >= charbuf_end)
1685 		break;
1686 	
1687 	      if (byte_after_cr1 >= 0)
1688 		c1 = byte_after_cr1, byte_after_cr1 = -1;
1689 	      else
1690 		ONE_MORE_BYTE (c1);
1691 	      if (c1 < 0)
1692 		{
1693 		  *charbuf++ = -c1;
1694 		  continue;
1695 		}
1696 	      if (byte_after_cr2 >= 0)
1697 		c2 = byte_after_cr2, byte_after_cr2 = -1;
1698 	      else
1699 		ONE_MORE_BYTE (c2);
1700 	      if (c2 < 0)
1701 		{
1702 		  *charbuf++ = ASCII_BYTE_P (c1) ? c1 : BYTE8_TO_CHAR (c1);
1703 		  *charbuf++ = -c2;
1704 		  continue;
1705 		}
1706 	      c = (endian == utf_16_big_endian
1707 		   ? ((c1 << 8) | c2) : ((c2 << 8) | c1));
1708 	
1709 	      if (surrogate)
1710 		{
1711 		  if (! UTF_16_LOW_SURROGATE_P (c))
1712 		    {
1713 		      if (endian == utf_16_big_endian)
1714 			c1 = surrogate >> 8, c2 = surrogate & 0xFF;
1715 		      else
1716 			c1 = surrogate & 0xFF, c2 = surrogate >> 8;
1717 		      *charbuf++ = c1;
1718 		      *charbuf++ = c2;
1719 		      coding->errors++;
1720 		      if (UTF_16_HIGH_SURROGATE_P (c))
1721 			CODING_UTF_16_SURROGATE (coding) = surrogate = c;
1722 		      else
1723 			*charbuf++ = c;
1724 		    }
1725 		  else
1726 		    {
1727 		      c = ((surrogate - 0xD800) << 10) | (c - 0xDC00);
1728 		      CODING_UTF_16_SURROGATE (coding) = surrogate = 0;
1729 		      *charbuf++ = 0x10000 + c;
1730 		    }
1731 		}
1732 	      else
1733 		{
1734 		  if (UTF_16_HIGH_SURROGATE_P (c))
1735 		    CODING_UTF_16_SURROGATE (coding) = surrogate = c;
1736 		  else
1737 		    {
1738 		      if (eol_crlf && c == '\r')
1739 			{
1740 			  ONE_MORE_BYTE (byte_after_cr1);
1741 			  ONE_MORE_BYTE (byte_after_cr2);
1742 			}
1743 		      *charbuf++ = c;
1744 		    }
1745 		}
1746 	    }
1747 	
1748 	 no_more_source:

Event uninit_use: Using uninitialized value "consumed_chars_base"
Also see events: [var_decl]

1749 	  coding->consumed_char += consumed_chars_base;
1750 	  coding->consumed = src_base - coding->source;

Re: Problem report #109: base/src/emacs/src/coding.c (decode_coding_utf_16); UNINIT

[Kenichi Handa]

In article <200812022217.mB2MH9h6021745@mothra.ics.uci.edu>, Dan Nicolaescu <dann@ics.uci.edu> writes:

> CID: 109
> Checker: UNINIT (help)
> File: base/src/emacs/src/coding.c
> Function: decode_coding_utf_16
> Description: Using uninitialized value "consumed_chars_base"


> Event var_decl: Declared variable "consumed_chars_base" without initializer
> Also see events: [uninit_use]

[...]
> Event uninit_use: Using uninitialized value "consumed_chars_base"
> Also see events: [var_decl]

> 1749 	  coding->consumed_char += consumed_chars_base;
> 1750 	  coding->consumed = src_base - coding->source;

I've just installed a fix.

---
Kenichi Handa
handa@ni.aist.go.jp