Re: struct which is exactly 4 bytes but on 64-bit platform

From:

Carl Barron <cbarron413@adelphia.net>

Newsgroups:

comp.lang.c++.moderated

Date:

Thu, 25 Jan 2007 03:34:19 CST

Message-ID:

<240120072147159812%cbarron413@adelphia.net>

In article <1169681895.722697.253500@v45g2000cwv.googlegroups.com>,
<"andrew_nuss@yahoo.com"> wrote:

Hi,

I have a struct for heap management which I have defined to need
exactly 4 bytes:

struct Header {
    bool free : 1;
    unsigned int chksum : 3;
    unsigned int log : 5;
    unsigned int history : 23;
};

    Safer and guaranteed not to fault. is to create some simple
classes that hold an unsigned char * and have an operator unsigned int()
to convert the field to an unsigned int an an operator = (unsigned int )
to set the appropriate field.
   example
namespace Header
{
    class chksum
    {
       unsigned char *p;
    public:
       chksum(unsigned char *a):p(a){}
       operator unsigned int() { return (*p >>4) & 7;}
       chksum & operator = (unsigned int x)
       {
          unsigned char u = x & 7;
          *p & = '\x8f';
          *p |= u;
          return *this;
       }
    };

// etc.
}

then the bit twiddling is 'hidden' and x = Header::chksum(header)
retrieves the 'chksum field' and Header::chksum(header) = x , stores
it.

almost as concise as the bitfield notation and guaranteed not to depend
on anything but possibly the endianess of an unsigned int, but the
order of bits in the header and alignment problems go away.

--
      [ See http://www.gotw.ca/resources/clcm.htm for info about ]
      [ comp.lang.c++.moderated. First time posters: Do this! ]