ip_packet_match

[Americo Melara]

I'm trying to understand the beginning execution of 'ipt_do_table()', please let me know if I'm wrong.  When the code enters the 'ip_packet_match()' function inside the 'do loop', the first thing that the firewall does is to check on the source and destination address of the packet against one specific rule, am I right?   For example, if I only have one rule with INPUT ACCEPT: iptables -A INPUT -s 192.168.24.2 -j DROP   if the destination address matches the address it will return 0, then exit 'ip_packet_match()' and jump to 'no_match:' inside 'ipt_do_table()' right?  What if it does not match?   ---- On the other hand...  if I have no rules to filter ip addresses, only one rule to filter one TCP port, then the code for 'ip_packet_match()' will not match anything, return a 1, and enter the "if statement" of 'ipt_do_table' ... 'ipt_do_table'  will execute the IPT_MATCH_ITERATE macro, wich jumps to the function 'do_match'  .  How does IPT_MATCH_ITERATE know that you want to filter a TCP port and not a MAC addresses?  who carries this info (the 'e' structure)?   Thank you! Americo