Re: [parchive-devel] CRC32

[Michael N. <mi...@na...>]

Peter,
         It looks like you are using a different polynomial.   Does your 
code work with the CCITT polynomial, 104C11DB7?

                 Mike


At 12:32 PM 11/14/2002 +0000, Peter B Clements wrote:
>Hello Michael,
>
>On Thursday, November 14, 2002, 3:33:25 AM, you wrote:
>
>MN> At 11:44 PM 11/13/2002 +0100, michael Minault wrote:
> >>Some minor remark about the spec. :
> >>
> >>The spec. doesn't say anything about the CRC32 parameters chosen.
> >>All implementations of CRC32 i know uses the same parameters as the
> >>Ethernet CRC32 implementation, so I guess it's the same in PAR2, but it
> >>could be a good thing to state it explicitly in the spec.
> >>There is a very good article about CRC code at
> >><http://www.ross.net/crc/crcpaper.html>http://www.ross.net/crc/crcpaper. 
> html
> >>and a list of well known CRC parameters at
> >><http://www.repairfaq.org/filipg/LINK/F_crc_v35.html#CRCV_001>http://www 
> .repairfaq.org/filipg/LINK/F_crc_v35.html#CRCV_001
> >>(including Ethernet CRC32)
> >>
> >>michael
>
>Those links make interesting reading.
>
>MN> Thanks Michael.  I'll make sure that gets into the final version of the
>MN> spec.
>
>MN> I knew there were many 32-bit CRCs and I thought that CRC32 was a "name
>MN> brand" one.  I should know better by now to check the facts.  I'll make
>MN> sure I specify it is the 32-bit CRC specified by CCITT.  (This is how 
>it is
>MN> identified in the 802.11 spec.)   I pasted the polynomial below.
>
>MN> Thanks again,
>MN>          Mike
>
>MN> G(x) = x32 + x26 + x23 + x22 + x16 + x12 + x11 + x10 + x8 + x7 + x5 + 
>x4 +
>MN> x2 + x + 1
>
>The code I use for CRC32 is derived from the following:
>
>#define CRC32_POLYNOMIAL     0xEDB88320L
>
>unsigned long ccitt_32[256];
>
>void build_table_32(void)
>{
>   unsigned int i;
>   unsigned int j;
>   unsigned long crc;
>
>   for ( i = 0; i <= 255 ; i++ )
>   {
>     crc = i;
>     for ( j = 0 ; j < 8; j++ )
>     {
>       crc = ( crc >> 1 ) ^ (( crc & 1 ) ? CRC32_POLYNOMIAL : 0 );
>     }
>     ccitt_32[ i ] = crc;
>   }
>}
>
>unsigned long UpdateCRC32(unsigned long crc,
>                           unsigned int count,
>                           const unsigned char *buffer)
>{
>   const unsigned char *p = buffer;
>   unsigned long temp1;
>   unsigned long temp2;
>
>   while ( count-- != 0 )
>   {
>     temp1 = ( crc >> 8 ) & 0x00FFFFFFL);
>     temp2 = ccitt_32[ ( (int) crc ^ *p++ ) & 0xff ];
>     crc = temp1 ^ temp2;
>   }
>   return( crc );
>}
>
>main()
>{
>   // Calculate the CRC of two blocks of data
>   // joined end to end.
>   unsigned char text1[] = "whatever";
>   unsigned char text2[] = "next";
>
>   // Do the first block of data and start with a CRC of 0xFFFFFFFF.
>   unsigned long crc = UpdateCRC32(~0, strlen(text1), text1);
>
>   // Do the second block and apply post conditioning to get
>   // the final CRC value.
>   crc = ~0 ^ UpdateCRC32(crc, strlen(text2), text2));
>}
>
>I assume this algorithm is what everyone else is using in their PAR2 code.
>
>--
>Best regards,
>  Peter                            mailto:pe...@sh...
>
>
>
>-------------------------------------------------------
>This sf.net email is sponsored by: To learn the basics of securing
>your web site with SSL, click here to get a FREE TRIAL of a Thawte
>Server Certificate: http://www.gothawte.com/rd524.html
>_______________________________________________
>Parchive-devel mailing list
>Par...@li...
>https://lists.sourceforge.net/lists/listinfo/parchive-devel