One infers from the sample below that you did not include the "UDP Pseudo-header" (IP src/dst address etc.) in the calculation. See Comer/Stevens, RFC 768, or a host of other references for the UDP checksum computation. I hope that this helps. From: Qin An [mailto:anqinat_private] Sent: Tuesday, July 29, 2003 11:41 PM To: secprogat_private Subject: [Q] cksum of UDP packet Hi, Glad to meet you all! I got a problem in cksum computation for UDP packet. At the end of this mail, there is a function copied from a textbook. I believe it is a universal cksum calculation function. I captured a UDP packet on the net: 0x07, 0x2c, 0x00, 0x35, 0x00, 0x0c, 0x98, 0x8e, 0x64, 0x69, 0x72, 0x0a The 6th and 7th element is the cksum computed by the sender's kernel. My question is that when I use in_cksum to compute the cksum, it returns 0x1f22, different from 0x98, 0x8e. What's wrong here? Is there any other algorithm to compute cksum? Thank you in advance! Qin An ========================================== unsigned short in_cksum (unsigned short *ptr, int nbytes) { unsigned long sum; /* assumes long == 32 bits */ unsigned short oddbyte, answer; /* assumes short == 16 bits */ sum = 0; while (nbytes > 1){ sum += *ptr++; nbytes -= 2; } /* mop up an odd byte, if necessary */ if (nbytes == 1){ /* make sure top half is zero */ oddbyte = 0; /* one byte only */ * ((unsigned char *) &oddbyte) = *(unsigned char *) ptr; sum += oddbyte; } /* add back carry outs from top 16 bits to low 16 bits */ /* add high-16 to low-16 */ sum = (sum >> 16) + (sum & 0xffff); /* add carry */ sum += (sum >> 16); /* ones-compliment, then truncate to 16 bits */ answer = (unsigned short)(~sum); return answer; } unsigned char s[] = { 0x07, 0x2c, 0x00, 0x35, 0x00, 0x0c, 0x98, 0x8e, 0x64, 0x69, 0x72, 0x0a}; int main (int argc, char *argv[]) { int i; s[6] = s[7] = 0; i = in_cksum ((unsigned short *) &s[0], 12); } _________________________________________________________________ Add photos to your messages with MSN 8. Get 2 months FREE*. http://join.msn.com/?page=features/featuredemail
This archive was generated by hypermail 2b30 : Wed Jul 30 2003 - 13:34:49 PDT