wireshark/epan/conversation.c

1448 lines
42 KiB
C
Raw Normal View History

Generalize the "ip_src" and "ip_dst" members of the "packet_info" structure to "dl_src"/"dl_dst", "net_src"/"net_dst", and "src"/"dst" addresses, where an address is an address type, an address length in bytes, and a pointer to that many bytes. "dl_{src,dst}" are the link-layer source/destination; "net_{src,dst}" are the network-layer source/destination; "{src,dst}" are the source/destination from the highest of those two layers that we have in the packet. Add a port type to "packet_info" as well, specifying whether it's a TCP or UDP port. Don't set the address and port columns in the dissector functions; just set the address and port members of the "packet_info" structure. Set the columns in "fill_in_columns()"; this means that if we're showing COL_{DEF,RES,UNRES}_SRC" or "COL_{DEF,RES,UNRES}_DST", we only generate the string from "src" or "dst", we don't generate a string for the link-layer address and then overwrite it with a string for the network-layer address (generating those strings costs CPU). Add support for "conversations", where a "conversation" is (at present) a source and destination address and a source and destination port. (In the future, we may support "conversations" above the transport layer, e.g. a TFTP conversation, where the first packet goes from the client to the TFTP server port, but the reply comes back from a different port, and all subsequent packets go between the client address/port and the server address/new port, or an NFS conversation, which might include lock manager, status monitor, and mount packets, as well as NFS packets.) Currently, all we support is a call that takes the source and destination address/port pairs, looks them up in a hash table, and: if nothing is found, creates a new entry in the hash table, and assigns it a unique 32-bit conversation ID, and returns that conversation ID; if an entry is found, returns its conversation ID. Use that in the SMB and AFS code to keep track of individual SMB or AFS conversations. We need to match up requests and replies, as, for certain replies, the operation code for the request to which it's a reply doesn't show up in the reply - you have to find the request with a matching transaction ID. Transaction IDs are per-conversation, so the hash table for requests should include a conversation ID and transaction ID as the key. This allows SMB and AFS decoders to handle IPv4 or IPv6 addresses transparently (and should allow the SMB decoder to handle NetBIOS atop other protocols as well, if the source and destination address and port values in the "packet_info" structure are set appropriately). In the "Follow TCP Connection" code, check to make sure that the addresses are IPv4 addressses; ultimately, that code should be changed to use the conversation code instead, which will let it handle IPv6 transparently. svn path=/trunk/; revision=909
1999-10-22 07:18:23 +00:00
/* conversation.c
* Routines for building lists of packets that are part of a "conversation"
*
* Wireshark - Network traffic analyzer
* By Gerald Combs <gerald@wireshark.org>
Generalize the "ip_src" and "ip_dst" members of the "packet_info" structure to "dl_src"/"dl_dst", "net_src"/"net_dst", and "src"/"dst" addresses, where an address is an address type, an address length in bytes, and a pointer to that many bytes. "dl_{src,dst}" are the link-layer source/destination; "net_{src,dst}" are the network-layer source/destination; "{src,dst}" are the source/destination from the highest of those two layers that we have in the packet. Add a port type to "packet_info" as well, specifying whether it's a TCP or UDP port. Don't set the address and port columns in the dissector functions; just set the address and port members of the "packet_info" structure. Set the columns in "fill_in_columns()"; this means that if we're showing COL_{DEF,RES,UNRES}_SRC" or "COL_{DEF,RES,UNRES}_DST", we only generate the string from "src" or "dst", we don't generate a string for the link-layer address and then overwrite it with a string for the network-layer address (generating those strings costs CPU). Add support for "conversations", where a "conversation" is (at present) a source and destination address and a source and destination port. (In the future, we may support "conversations" above the transport layer, e.g. a TFTP conversation, where the first packet goes from the client to the TFTP server port, but the reply comes back from a different port, and all subsequent packets go between the client address/port and the server address/new port, or an NFS conversation, which might include lock manager, status monitor, and mount packets, as well as NFS packets.) Currently, all we support is a call that takes the source and destination address/port pairs, looks them up in a hash table, and: if nothing is found, creates a new entry in the hash table, and assigns it a unique 32-bit conversation ID, and returns that conversation ID; if an entry is found, returns its conversation ID. Use that in the SMB and AFS code to keep track of individual SMB or AFS conversations. We need to match up requests and replies, as, for certain replies, the operation code for the request to which it's a reply doesn't show up in the reply - you have to find the request with a matching transaction ID. Transaction IDs are per-conversation, so the hash table for requests should include a conversation ID and transaction ID as the key. This allows SMB and AFS decoders to handle IPv4 or IPv6 addresses transparently (and should allow the SMB decoder to handle NetBIOS atop other protocols as well, if the source and destination address and port values in the "packet_info" structure are set appropriately). In the "Follow TCP Connection" code, check to make sure that the addresses are IPv4 addressses; ultimately, that code should be changed to use the conversation code instead, which will let it handle IPv6 transparently. svn path=/trunk/; revision=909
1999-10-22 07:18:23 +00:00
* Copyright 1998 Gerald Combs
*
Generalize the "ip_src" and "ip_dst" members of the "packet_info" structure to "dl_src"/"dl_dst", "net_src"/"net_dst", and "src"/"dst" addresses, where an address is an address type, an address length in bytes, and a pointer to that many bytes. "dl_{src,dst}" are the link-layer source/destination; "net_{src,dst}" are the network-layer source/destination; "{src,dst}" are the source/destination from the highest of those two layers that we have in the packet. Add a port type to "packet_info" as well, specifying whether it's a TCP or UDP port. Don't set the address and port columns in the dissector functions; just set the address and port members of the "packet_info" structure. Set the columns in "fill_in_columns()"; this means that if we're showing COL_{DEF,RES,UNRES}_SRC" or "COL_{DEF,RES,UNRES}_DST", we only generate the string from "src" or "dst", we don't generate a string for the link-layer address and then overwrite it with a string for the network-layer address (generating those strings costs CPU). Add support for "conversations", where a "conversation" is (at present) a source and destination address and a source and destination port. (In the future, we may support "conversations" above the transport layer, e.g. a TFTP conversation, where the first packet goes from the client to the TFTP server port, but the reply comes back from a different port, and all subsequent packets go between the client address/port and the server address/new port, or an NFS conversation, which might include lock manager, status monitor, and mount packets, as well as NFS packets.) Currently, all we support is a call that takes the source and destination address/port pairs, looks them up in a hash table, and: if nothing is found, creates a new entry in the hash table, and assigns it a unique 32-bit conversation ID, and returns that conversation ID; if an entry is found, returns its conversation ID. Use that in the SMB and AFS code to keep track of individual SMB or AFS conversations. We need to match up requests and replies, as, for certain replies, the operation code for the request to which it's a reply doesn't show up in the reply - you have to find the request with a matching transaction ID. Transaction IDs are per-conversation, so the hash table for requests should include a conversation ID and transaction ID as the key. This allows SMB and AFS decoders to handle IPv4 or IPv6 addresses transparently (and should allow the SMB decoder to handle NetBIOS atop other protocols as well, if the source and destination address and port values in the "packet_info" structure are set appropriately). In the "Follow TCP Connection" code, check to make sure that the addresses are IPv4 addressses; ultimately, that code should be changed to use the conversation code instead, which will let it handle IPv6 transparently. svn path=/trunk/; revision=909
1999-10-22 07:18:23 +00:00
* This program is free software; you can redistribute it and/or
* modify it under the terms of the GNU General Public License
* as published by the Free Software Foundation; either version 2
* of the License, or (at your option) any later version.
*
Generalize the "ip_src" and "ip_dst" members of the "packet_info" structure to "dl_src"/"dl_dst", "net_src"/"net_dst", and "src"/"dst" addresses, where an address is an address type, an address length in bytes, and a pointer to that many bytes. "dl_{src,dst}" are the link-layer source/destination; "net_{src,dst}" are the network-layer source/destination; "{src,dst}" are the source/destination from the highest of those two layers that we have in the packet. Add a port type to "packet_info" as well, specifying whether it's a TCP or UDP port. Don't set the address and port columns in the dissector functions; just set the address and port members of the "packet_info" structure. Set the columns in "fill_in_columns()"; this means that if we're showing COL_{DEF,RES,UNRES}_SRC" or "COL_{DEF,RES,UNRES}_DST", we only generate the string from "src" or "dst", we don't generate a string for the link-layer address and then overwrite it with a string for the network-layer address (generating those strings costs CPU). Add support for "conversations", where a "conversation" is (at present) a source and destination address and a source and destination port. (In the future, we may support "conversations" above the transport layer, e.g. a TFTP conversation, where the first packet goes from the client to the TFTP server port, but the reply comes back from a different port, and all subsequent packets go between the client address/port and the server address/new port, or an NFS conversation, which might include lock manager, status monitor, and mount packets, as well as NFS packets.) Currently, all we support is a call that takes the source and destination address/port pairs, looks them up in a hash table, and: if nothing is found, creates a new entry in the hash table, and assigns it a unique 32-bit conversation ID, and returns that conversation ID; if an entry is found, returns its conversation ID. Use that in the SMB and AFS code to keep track of individual SMB or AFS conversations. We need to match up requests and replies, as, for certain replies, the operation code for the request to which it's a reply doesn't show up in the reply - you have to find the request with a matching transaction ID. Transaction IDs are per-conversation, so the hash table for requests should include a conversation ID and transaction ID as the key. This allows SMB and AFS decoders to handle IPv4 or IPv6 addresses transparently (and should allow the SMB decoder to handle NetBIOS atop other protocols as well, if the source and destination address and port values in the "packet_info" structure are set appropriately). In the "Follow TCP Connection" code, check to make sure that the addresses are IPv4 addressses; ultimately, that code should be changed to use the conversation code instead, which will let it handle IPv6 transparently. svn path=/trunk/; revision=909
1999-10-22 07:18:23 +00:00
* This program is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU General Public License for more details.
*
Generalize the "ip_src" and "ip_dst" members of the "packet_info" structure to "dl_src"/"dl_dst", "net_src"/"net_dst", and "src"/"dst" addresses, where an address is an address type, an address length in bytes, and a pointer to that many bytes. "dl_{src,dst}" are the link-layer source/destination; "net_{src,dst}" are the network-layer source/destination; "{src,dst}" are the source/destination from the highest of those two layers that we have in the packet. Add a port type to "packet_info" as well, specifying whether it's a TCP or UDP port. Don't set the address and port columns in the dissector functions; just set the address and port members of the "packet_info" structure. Set the columns in "fill_in_columns()"; this means that if we're showing COL_{DEF,RES,UNRES}_SRC" or "COL_{DEF,RES,UNRES}_DST", we only generate the string from "src" or "dst", we don't generate a string for the link-layer address and then overwrite it with a string for the network-layer address (generating those strings costs CPU). Add support for "conversations", where a "conversation" is (at present) a source and destination address and a source and destination port. (In the future, we may support "conversations" above the transport layer, e.g. a TFTP conversation, where the first packet goes from the client to the TFTP server port, but the reply comes back from a different port, and all subsequent packets go between the client address/port and the server address/new port, or an NFS conversation, which might include lock manager, status monitor, and mount packets, as well as NFS packets.) Currently, all we support is a call that takes the source and destination address/port pairs, looks them up in a hash table, and: if nothing is found, creates a new entry in the hash table, and assigns it a unique 32-bit conversation ID, and returns that conversation ID; if an entry is found, returns its conversation ID. Use that in the SMB and AFS code to keep track of individual SMB or AFS conversations. We need to match up requests and replies, as, for certain replies, the operation code for the request to which it's a reply doesn't show up in the reply - you have to find the request with a matching transaction ID. Transaction IDs are per-conversation, so the hash table for requests should include a conversation ID and transaction ID as the key. This allows SMB and AFS decoders to handle IPv4 or IPv6 addresses transparently (and should allow the SMB decoder to handle NetBIOS atop other protocols as well, if the source and destination address and port values in the "packet_info" structure are set appropriately). In the "Follow TCP Connection" code, check to make sure that the addresses are IPv4 addressses; ultimately, that code should be changed to use the conversation code instead, which will let it handle IPv6 transparently. svn path=/trunk/; revision=909
1999-10-22 07:18:23 +00:00
* You should have received a copy of the GNU General Public License
* along with this program; if not, write to the Free Software
* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
Generalize the "ip_src" and "ip_dst" members of the "packet_info" structure to "dl_src"/"dl_dst", "net_src"/"net_dst", and "src"/"dst" addresses, where an address is an address type, an address length in bytes, and a pointer to that many bytes. "dl_{src,dst}" are the link-layer source/destination; "net_{src,dst}" are the network-layer source/destination; "{src,dst}" are the source/destination from the highest of those two layers that we have in the packet. Add a port type to "packet_info" as well, specifying whether it's a TCP or UDP port. Don't set the address and port columns in the dissector functions; just set the address and port members of the "packet_info" structure. Set the columns in "fill_in_columns()"; this means that if we're showing COL_{DEF,RES,UNRES}_SRC" or "COL_{DEF,RES,UNRES}_DST", we only generate the string from "src" or "dst", we don't generate a string for the link-layer address and then overwrite it with a string for the network-layer address (generating those strings costs CPU). Add support for "conversations", where a "conversation" is (at present) a source and destination address and a source and destination port. (In the future, we may support "conversations" above the transport layer, e.g. a TFTP conversation, where the first packet goes from the client to the TFTP server port, but the reply comes back from a different port, and all subsequent packets go between the client address/port and the server address/new port, or an NFS conversation, which might include lock manager, status monitor, and mount packets, as well as NFS packets.) Currently, all we support is a call that takes the source and destination address/port pairs, looks them up in a hash table, and: if nothing is found, creates a new entry in the hash table, and assigns it a unique 32-bit conversation ID, and returns that conversation ID; if an entry is found, returns its conversation ID. Use that in the SMB and AFS code to keep track of individual SMB or AFS conversations. We need to match up requests and replies, as, for certain replies, the operation code for the request to which it's a reply doesn't show up in the reply - you have to find the request with a matching transaction ID. Transaction IDs are per-conversation, so the hash table for requests should include a conversation ID and transaction ID as the key. This allows SMB and AFS decoders to handle IPv4 or IPv6 addresses transparently (and should allow the SMB decoder to handle NetBIOS atop other protocols as well, if the source and destination address and port values in the "packet_info" structure are set appropriately). In the "Follow TCP Connection" code, check to make sure that the addresses are IPv4 addressses; ultimately, that code should be changed to use the conversation code instead, which will let it handle IPv6 transparently. svn path=/trunk/; revision=909
1999-10-22 07:18:23 +00:00
*/
#include "config.h"
Generalize the "ip_src" and "ip_dst" members of the "packet_info" structure to "dl_src"/"dl_dst", "net_src"/"net_dst", and "src"/"dst" addresses, where an address is an address type, an address length in bytes, and a pointer to that many bytes. "dl_{src,dst}" are the link-layer source/destination; "net_{src,dst}" are the network-layer source/destination; "{src,dst}" are the source/destination from the highest of those two layers that we have in the packet. Add a port type to "packet_info" as well, specifying whether it's a TCP or UDP port. Don't set the address and port columns in the dissector functions; just set the address and port members of the "packet_info" structure. Set the columns in "fill_in_columns()"; this means that if we're showing COL_{DEF,RES,UNRES}_SRC" or "COL_{DEF,RES,UNRES}_DST", we only generate the string from "src" or "dst", we don't generate a string for the link-layer address and then overwrite it with a string for the network-layer address (generating those strings costs CPU). Add support for "conversations", where a "conversation" is (at present) a source and destination address and a source and destination port. (In the future, we may support "conversations" above the transport layer, e.g. a TFTP conversation, where the first packet goes from the client to the TFTP server port, but the reply comes back from a different port, and all subsequent packets go between the client address/port and the server address/new port, or an NFS conversation, which might include lock manager, status monitor, and mount packets, as well as NFS packets.) Currently, all we support is a call that takes the source and destination address/port pairs, looks them up in a hash table, and: if nothing is found, creates a new entry in the hash table, and assigns it a unique 32-bit conversation ID, and returns that conversation ID; if an entry is found, returns its conversation ID. Use that in the SMB and AFS code to keep track of individual SMB or AFS conversations. We need to match up requests and replies, as, for certain replies, the operation code for the request to which it's a reply doesn't show up in the reply - you have to find the request with a matching transaction ID. Transaction IDs are per-conversation, so the hash table for requests should include a conversation ID and transaction ID as the key. This allows SMB and AFS decoders to handle IPv4 or IPv6 addresses transparently (and should allow the SMB decoder to handle NetBIOS atop other protocols as well, if the source and destination address and port values in the "packet_info" structure are set appropriately). In the "Follow TCP Connection" code, check to make sure that the addresses are IPv4 addressses; ultimately, that code should be changed to use the conversation code instead, which will let it handle IPv6 transparently. svn path=/trunk/; revision=909
1999-10-22 07:18:23 +00:00
#include <string.h>
#include <glib.h>
#include "packet.h"
#include "to_str.h"
#include "conversation.h"
Generalize the "ip_src" and "ip_dst" members of the "packet_info" structure to "dl_src"/"dl_dst", "net_src"/"net_dst", and "src"/"dst" addresses, where an address is an address type, an address length in bytes, and a pointer to that many bytes. "dl_{src,dst}" are the link-layer source/destination; "net_{src,dst}" are the network-layer source/destination; "{src,dst}" are the source/destination from the highest of those two layers that we have in the packet. Add a port type to "packet_info" as well, specifying whether it's a TCP or UDP port. Don't set the address and port columns in the dissector functions; just set the address and port members of the "packet_info" structure. Set the columns in "fill_in_columns()"; this means that if we're showing COL_{DEF,RES,UNRES}_SRC" or "COL_{DEF,RES,UNRES}_DST", we only generate the string from "src" or "dst", we don't generate a string for the link-layer address and then overwrite it with a string for the network-layer address (generating those strings costs CPU). Add support for "conversations", where a "conversation" is (at present) a source and destination address and a source and destination port. (In the future, we may support "conversations" above the transport layer, e.g. a TFTP conversation, where the first packet goes from the client to the TFTP server port, but the reply comes back from a different port, and all subsequent packets go between the client address/port and the server address/new port, or an NFS conversation, which might include lock manager, status monitor, and mount packets, as well as NFS packets.) Currently, all we support is a call that takes the source and destination address/port pairs, looks them up in a hash table, and: if nothing is found, creates a new entry in the hash table, and assigns it a unique 32-bit conversation ID, and returns that conversation ID; if an entry is found, returns its conversation ID. Use that in the SMB and AFS code to keep track of individual SMB or AFS conversations. We need to match up requests and replies, as, for certain replies, the operation code for the request to which it's a reply doesn't show up in the reply - you have to find the request with a matching transaction ID. Transaction IDs are per-conversation, so the hash table for requests should include a conversation ID and transaction ID as the key. This allows SMB and AFS decoders to handle IPv4 or IPv6 addresses transparently (and should allow the SMB decoder to handle NetBIOS atop other protocols as well, if the source and destination address and port values in the "packet_info" structure are set appropriately). In the "Follow TCP Connection" code, check to make sure that the addresses are IPv4 addressses; ultimately, that code should be changed to use the conversation code instead, which will let it handle IPv6 transparently. svn path=/trunk/; revision=909
1999-10-22 07:18:23 +00:00
/* define DEBUG_CONVERSATION for pretty debug printing */
/* #define DEBUG_CONVERSATION */
#include "conversation_debug.h"
#ifdef DEBUG_CONVERSATION
int _debug_conversation_indent = 0;
#endif
struct endpoint {
address addr1;
address addr2;
endpoint_type etype;
guint32 port1;
guint32 port2;
guint options;
};
struct conversation_key {
struct conversation_key *next;
address addr1;
address addr2;
endpoint_type etype;
guint32 port1;
guint32 port2;
};
/*
* Hash table for conversations with no wildcards.
*/
static wmem_map_t *conversation_hashtable_exact = NULL;
/*
* Hash table for conversations with one wildcard address.
*/
static wmem_map_t *conversation_hashtable_no_addr2 = NULL;
/*
* Hash table for conversations with one wildcard port.
*/
static wmem_map_t *conversation_hashtable_no_port2 = NULL;
/*
* Hash table for conversations with one wildcard address and port.
*/
static wmem_map_t *conversation_hashtable_no_addr2_or_port2 = NULL;
Generalize the "ip_src" and "ip_dst" members of the "packet_info" structure to "dl_src"/"dl_dst", "net_src"/"net_dst", and "src"/"dst" addresses, where an address is an address type, an address length in bytes, and a pointer to that many bytes. "dl_{src,dst}" are the link-layer source/destination; "net_{src,dst}" are the network-layer source/destination; "{src,dst}" are the source/destination from the highest of those two layers that we have in the packet. Add a port type to "packet_info" as well, specifying whether it's a TCP or UDP port. Don't set the address and port columns in the dissector functions; just set the address and port members of the "packet_info" structure. Set the columns in "fill_in_columns()"; this means that if we're showing COL_{DEF,RES,UNRES}_SRC" or "COL_{DEF,RES,UNRES}_DST", we only generate the string from "src" or "dst", we don't generate a string for the link-layer address and then overwrite it with a string for the network-layer address (generating those strings costs CPU). Add support for "conversations", where a "conversation" is (at present) a source and destination address and a source and destination port. (In the future, we may support "conversations" above the transport layer, e.g. a TFTP conversation, where the first packet goes from the client to the TFTP server port, but the reply comes back from a different port, and all subsequent packets go between the client address/port and the server address/new port, or an NFS conversation, which might include lock manager, status monitor, and mount packets, as well as NFS packets.) Currently, all we support is a call that takes the source and destination address/port pairs, looks them up in a hash table, and: if nothing is found, creates a new entry in the hash table, and assigns it a unique 32-bit conversation ID, and returns that conversation ID; if an entry is found, returns its conversation ID. Use that in the SMB and AFS code to keep track of individual SMB or AFS conversations. We need to match up requests and replies, as, for certain replies, the operation code for the request to which it's a reply doesn't show up in the reply - you have to find the request with a matching transaction ID. Transaction IDs are per-conversation, so the hash table for requests should include a conversation ID and transaction ID as the key. This allows SMB and AFS decoders to handle IPv4 or IPv6 addresses transparently (and should allow the SMB decoder to handle NetBIOS atop other protocols as well, if the source and destination address and port values in the "packet_info" structure are set appropriately). In the "Follow TCP Connection" code, check to make sure that the addresses are IPv4 addressses; ultimately, that code should be changed to use the conversation code instead, which will let it handle IPv6 transparently. svn path=/trunk/; revision=909
1999-10-22 07:18:23 +00:00
static guint32 new_index;
/*
* Creates a new conversation with known endpoints based on a conversation
* created with the CONVERSATION_TEMPLATE option while keeping the
* conversation created with the CONVERSATION_TEMPLATE option so it can still
* match future connections.
*
* Passing a pointer to a conversation whose options mask does not include
* CONVERSATION_TEMPLATE or where the conversation's protocol type (ptype)
* indicates a non-connnection oriented protocol will return the conversation
* without changes.
*
* addr2 and port2 are used in the function if their respective conversation
* options bits are set (NO_ADDR2 and NO_PORT2).
*/
static conversation_t *
conversation_create_from_template(conversation_t *conversation, const address *addr2, const guint32 port2)
{
/*
* Add a new conversation and keep the conversation template only if the
* CONVERSATION_TEMPLATE bit is set for a connection oriented protocol.
*/
if(conversation->options & CONVERSATION_TEMPLATE &&
conversation->key_ptr->etype != ENDPOINT_UDP)
{
/*
* Set up a new options mask where the conversation template bit and the
* bits for absence of a second address and port pair have been removed.
*/
conversation_t *new_conversation_from_template;
guint options = conversation->options & ~(CONVERSATION_TEMPLATE | NO_ADDR2 | NO_PORT2);
/*
* Are both the NO_ADDR2 and NO_PORT2 wildcards set in the options mask?
*/
if(conversation->options & NO_ADDR2 &&
conversation->options & NO_PORT2)
{
/*
* The conversation template was created without knowledge of both
* the second address as well as the second port. Create a new
* conversation with new 2nd address and 2nd port.
*/
new_conversation_from_template =
conversation_new(conversation->setup_frame,
&conversation->key_ptr->addr1, addr2,
conversation->key_ptr->etype, conversation->key_ptr->port1,
port2, options);
}
else if(conversation->options & NO_PORT2)
{
/*
* The conversation template was created without knowledge of port 2
* only. Create a new conversation with new 2nd port.
*/
new_conversation_from_template =
conversation_new(conversation->setup_frame,
&conversation->key_ptr->addr1, &conversation->key_ptr->addr2,
conversation->key_ptr->etype, conversation->key_ptr->port1,
port2, options);
}
else if(conversation->options & NO_ADDR2)
{
/*
* The conversation template was created without knowledge of address
* 2. Create a new conversation with new 2nd address.
*/
new_conversation_from_template =
conversation_new(conversation->setup_frame,
&conversation->key_ptr->addr1, addr2,
conversation->key_ptr->etype, conversation->key_ptr->port1,
conversation->key_ptr->port2, options);
}
else
{
/*
* The CONVERSATION_TEMPLATE bit was set, but no other bit that the
* CONVERSATION_TEMPLATE bit controls is active. Just return the old
* conversation.
*/
return conversation;
}
/*
* Set the protocol dissector used for the template conversation as
* the handler of the new conversation as well.
*/
new_conversation_from_template->dissector_tree = conversation->dissector_tree;
return new_conversation_from_template;
}
else
{
return conversation;
}
}
Generalize the "ip_src" and "ip_dst" members of the "packet_info" structure to "dl_src"/"dl_dst", "net_src"/"net_dst", and "src"/"dst" addresses, where an address is an address type, an address length in bytes, and a pointer to that many bytes. "dl_{src,dst}" are the link-layer source/destination; "net_{src,dst}" are the network-layer source/destination; "{src,dst}" are the source/destination from the highest of those two layers that we have in the packet. Add a port type to "packet_info" as well, specifying whether it's a TCP or UDP port. Don't set the address and port columns in the dissector functions; just set the address and port members of the "packet_info" structure. Set the columns in "fill_in_columns()"; this means that if we're showing COL_{DEF,RES,UNRES}_SRC" or "COL_{DEF,RES,UNRES}_DST", we only generate the string from "src" or "dst", we don't generate a string for the link-layer address and then overwrite it with a string for the network-layer address (generating those strings costs CPU). Add support for "conversations", where a "conversation" is (at present) a source and destination address and a source and destination port. (In the future, we may support "conversations" above the transport layer, e.g. a TFTP conversation, where the first packet goes from the client to the TFTP server port, but the reply comes back from a different port, and all subsequent packets go between the client address/port and the server address/new port, or an NFS conversation, which might include lock manager, status monitor, and mount packets, as well as NFS packets.) Currently, all we support is a call that takes the source and destination address/port pairs, looks them up in a hash table, and: if nothing is found, creates a new entry in the hash table, and assigns it a unique 32-bit conversation ID, and returns that conversation ID; if an entry is found, returns its conversation ID. Use that in the SMB and AFS code to keep track of individual SMB or AFS conversations. We need to match up requests and replies, as, for certain replies, the operation code for the request to which it's a reply doesn't show up in the reply - you have to find the request with a matching transaction ID. Transaction IDs are per-conversation, so the hash table for requests should include a conversation ID and transaction ID as the key. This allows SMB and AFS decoders to handle IPv4 or IPv6 addresses transparently (and should allow the SMB decoder to handle NetBIOS atop other protocols as well, if the source and destination address and port values in the "packet_info" structure are set appropriately). In the "Follow TCP Connection" code, check to make sure that the addresses are IPv4 addressses; ultimately, that code should be changed to use the conversation code instead, which will let it handle IPv6 transparently. svn path=/trunk/; revision=909
1999-10-22 07:18:23 +00:00
/*
* Compute the hash value for two given address/port pairs if the match
* is to be exact.
*/
/* http://eternallyconfuzzled.com/tuts/algorithms/jsw_tut_hashing.aspx#existing
* One-at-a-Time hash
*/
guint
conversation_hash_exact(gconstpointer v)
{
const conversation_key_t key = (const conversation_key_t)v;
guint hash_val;
address tmp_addr;
hash_val = 0;
tmp_addr.len = 4;
hash_val = add_address_to_hash(hash_val, &key->addr1);
tmp_addr.data = &key->port1;
hash_val = add_address_to_hash(hash_val, &tmp_addr);
hash_val = add_address_to_hash(hash_val, &key->addr2);
tmp_addr.data = &key->port2;
hash_val = add_address_to_hash(hash_val, &tmp_addr);
hash_val += ( hash_val << 3 );
hash_val ^= ( hash_val >> 11 );
hash_val += ( hash_val << 15 );
return hash_val;
}
/*
* Compare two conversation keys for an exact match.
Generalize the "ip_src" and "ip_dst" members of the "packet_info" structure to "dl_src"/"dl_dst", "net_src"/"net_dst", and "src"/"dst" addresses, where an address is an address type, an address length in bytes, and a pointer to that many bytes. "dl_{src,dst}" are the link-layer source/destination; "net_{src,dst}" are the network-layer source/destination; "{src,dst}" are the source/destination from the highest of those two layers that we have in the packet. Add a port type to "packet_info" as well, specifying whether it's a TCP or UDP port. Don't set the address and port columns in the dissector functions; just set the address and port members of the "packet_info" structure. Set the columns in "fill_in_columns()"; this means that if we're showing COL_{DEF,RES,UNRES}_SRC" or "COL_{DEF,RES,UNRES}_DST", we only generate the string from "src" or "dst", we don't generate a string for the link-layer address and then overwrite it with a string for the network-layer address (generating those strings costs CPU). Add support for "conversations", where a "conversation" is (at present) a source and destination address and a source and destination port. (In the future, we may support "conversations" above the transport layer, e.g. a TFTP conversation, where the first packet goes from the client to the TFTP server port, but the reply comes back from a different port, and all subsequent packets go between the client address/port and the server address/new port, or an NFS conversation, which might include lock manager, status monitor, and mount packets, as well as NFS packets.) Currently, all we support is a call that takes the source and destination address/port pairs, looks them up in a hash table, and: if nothing is found, creates a new entry in the hash table, and assigns it a unique 32-bit conversation ID, and returns that conversation ID; if an entry is found, returns its conversation ID. Use that in the SMB and AFS code to keep track of individual SMB or AFS conversations. We need to match up requests and replies, as, for certain replies, the operation code for the request to which it's a reply doesn't show up in the reply - you have to find the request with a matching transaction ID. Transaction IDs are per-conversation, so the hash table for requests should include a conversation ID and transaction ID as the key. This allows SMB and AFS decoders to handle IPv4 or IPv6 addresses transparently (and should allow the SMB decoder to handle NetBIOS atop other protocols as well, if the source and destination address and port values in the "packet_info" structure are set appropriately). In the "Follow TCP Connection" code, check to make sure that the addresses are IPv4 addressses; ultimately, that code should be changed to use the conversation code instead, which will let it handle IPv6 transparently. svn path=/trunk/; revision=909
1999-10-22 07:18:23 +00:00
*/
static gint
conversation_match_exact(gconstpointer v, gconstpointer w)
Generalize the "ip_src" and "ip_dst" members of the "packet_info" structure to "dl_src"/"dl_dst", "net_src"/"net_dst", and "src"/"dst" addresses, where an address is an address type, an address length in bytes, and a pointer to that many bytes. "dl_{src,dst}" are the link-layer source/destination; "net_{src,dst}" are the network-layer source/destination; "{src,dst}" are the source/destination from the highest of those two layers that we have in the packet. Add a port type to "packet_info" as well, specifying whether it's a TCP or UDP port. Don't set the address and port columns in the dissector functions; just set the address and port members of the "packet_info" structure. Set the columns in "fill_in_columns()"; this means that if we're showing COL_{DEF,RES,UNRES}_SRC" or "COL_{DEF,RES,UNRES}_DST", we only generate the string from "src" or "dst", we don't generate a string for the link-layer address and then overwrite it with a string for the network-layer address (generating those strings costs CPU). Add support for "conversations", where a "conversation" is (at present) a source and destination address and a source and destination port. (In the future, we may support "conversations" above the transport layer, e.g. a TFTP conversation, where the first packet goes from the client to the TFTP server port, but the reply comes back from a different port, and all subsequent packets go between the client address/port and the server address/new port, or an NFS conversation, which might include lock manager, status monitor, and mount packets, as well as NFS packets.) Currently, all we support is a call that takes the source and destination address/port pairs, looks them up in a hash table, and: if nothing is found, creates a new entry in the hash table, and assigns it a unique 32-bit conversation ID, and returns that conversation ID; if an entry is found, returns its conversation ID. Use that in the SMB and AFS code to keep track of individual SMB or AFS conversations. We need to match up requests and replies, as, for certain replies, the operation code for the request to which it's a reply doesn't show up in the reply - you have to find the request with a matching transaction ID. Transaction IDs are per-conversation, so the hash table for requests should include a conversation ID and transaction ID as the key. This allows SMB and AFS decoders to handle IPv4 or IPv6 addresses transparently (and should allow the SMB decoder to handle NetBIOS atop other protocols as well, if the source and destination address and port values in the "packet_info" structure are set appropriately). In the "Follow TCP Connection" code, check to make sure that the addresses are IPv4 addressses; ultimately, that code should be changed to use the conversation code instead, which will let it handle IPv6 transparently. svn path=/trunk/; revision=909
1999-10-22 07:18:23 +00:00
{
const conversation_key_t v1 = (const conversation_key_t)v;
const conversation_key_t v2 = (const conversation_key_t)w;
Generalize the "ip_src" and "ip_dst" members of the "packet_info" structure to "dl_src"/"dl_dst", "net_src"/"net_dst", and "src"/"dst" addresses, where an address is an address type, an address length in bytes, and a pointer to that many bytes. "dl_{src,dst}" are the link-layer source/destination; "net_{src,dst}" are the network-layer source/destination; "{src,dst}" are the source/destination from the highest of those two layers that we have in the packet. Add a port type to "packet_info" as well, specifying whether it's a TCP or UDP port. Don't set the address and port columns in the dissector functions; just set the address and port members of the "packet_info" structure. Set the columns in "fill_in_columns()"; this means that if we're showing COL_{DEF,RES,UNRES}_SRC" or "COL_{DEF,RES,UNRES}_DST", we only generate the string from "src" or "dst", we don't generate a string for the link-layer address and then overwrite it with a string for the network-layer address (generating those strings costs CPU). Add support for "conversations", where a "conversation" is (at present) a source and destination address and a source and destination port. (In the future, we may support "conversations" above the transport layer, e.g. a TFTP conversation, where the first packet goes from the client to the TFTP server port, but the reply comes back from a different port, and all subsequent packets go between the client address/port and the server address/new port, or an NFS conversation, which might include lock manager, status monitor, and mount packets, as well as NFS packets.) Currently, all we support is a call that takes the source and destination address/port pairs, looks them up in a hash table, and: if nothing is found, creates a new entry in the hash table, and assigns it a unique 32-bit conversation ID, and returns that conversation ID; if an entry is found, returns its conversation ID. Use that in the SMB and AFS code to keep track of individual SMB or AFS conversations. We need to match up requests and replies, as, for certain replies, the operation code for the request to which it's a reply doesn't show up in the reply - you have to find the request with a matching transaction ID. Transaction IDs are per-conversation, so the hash table for requests should include a conversation ID and transaction ID as the key. This allows SMB and AFS decoders to handle IPv4 or IPv6 addresses transparently (and should allow the SMB decoder to handle NetBIOS atop other protocols as well, if the source and destination address and port values in the "packet_info" structure are set appropriately). In the "Follow TCP Connection" code, check to make sure that the addresses are IPv4 addressses; ultimately, that code should be changed to use the conversation code instead, which will let it handle IPv6 transparently. svn path=/trunk/; revision=909
1999-10-22 07:18:23 +00:00
if (v1->etype != v2->etype)
Generalize the "ip_src" and "ip_dst" members of the "packet_info" structure to "dl_src"/"dl_dst", "net_src"/"net_dst", and "src"/"dst" addresses, where an address is an address type, an address length in bytes, and a pointer to that many bytes. "dl_{src,dst}" are the link-layer source/destination; "net_{src,dst}" are the network-layer source/destination; "{src,dst}" are the source/destination from the highest of those two layers that we have in the packet. Add a port type to "packet_info" as well, specifying whether it's a TCP or UDP port. Don't set the address and port columns in the dissector functions; just set the address and port members of the "packet_info" structure. Set the columns in "fill_in_columns()"; this means that if we're showing COL_{DEF,RES,UNRES}_SRC" or "COL_{DEF,RES,UNRES}_DST", we only generate the string from "src" or "dst", we don't generate a string for the link-layer address and then overwrite it with a string for the network-layer address (generating those strings costs CPU). Add support for "conversations", where a "conversation" is (at present) a source and destination address and a source and destination port. (In the future, we may support "conversations" above the transport layer, e.g. a TFTP conversation, where the first packet goes from the client to the TFTP server port, but the reply comes back from a different port, and all subsequent packets go between the client address/port and the server address/new port, or an NFS conversation, which might include lock manager, status monitor, and mount packets, as well as NFS packets.) Currently, all we support is a call that takes the source and destination address/port pairs, looks them up in a hash table, and: if nothing is found, creates a new entry in the hash table, and assigns it a unique 32-bit conversation ID, and returns that conversation ID; if an entry is found, returns its conversation ID. Use that in the SMB and AFS code to keep track of individual SMB or AFS conversations. We need to match up requests and replies, as, for certain replies, the operation code for the request to which it's a reply doesn't show up in the reply - you have to find the request with a matching transaction ID. Transaction IDs are per-conversation, so the hash table for requests should include a conversation ID and transaction ID as the key. This allows SMB and AFS decoders to handle IPv4 or IPv6 addresses transparently (and should allow the SMB decoder to handle NetBIOS atop other protocols as well, if the source and destination address and port values in the "packet_info" structure are set appropriately). In the "Follow TCP Connection" code, check to make sure that the addresses are IPv4 addressses; ultimately, that code should be changed to use the conversation code instead, which will let it handle IPv6 transparently. svn path=/trunk/; revision=909
1999-10-22 07:18:23 +00:00
return 0; /* different types of port */
/*
* Are the first and second port 1 values the same, the first and
* second port 2 values the same, the first and second address
* 1 values the same, and the first and second address 2 values
* the same?
*/
if (v1->port1 == v2->port1 &&
v1->port2 == v2->port2 &&
addresses_equal(&v1->addr1, &v2->addr1) &&
addresses_equal(&v1->addr2, &v2->addr2)) {
Generalize the "ip_src" and "ip_dst" members of the "packet_info" structure to "dl_src"/"dl_dst", "net_src"/"net_dst", and "src"/"dst" addresses, where an address is an address type, an address length in bytes, and a pointer to that many bytes. "dl_{src,dst}" are the link-layer source/destination; "net_{src,dst}" are the network-layer source/destination; "{src,dst}" are the source/destination from the highest of those two layers that we have in the packet. Add a port type to "packet_info" as well, specifying whether it's a TCP or UDP port. Don't set the address and port columns in the dissector functions; just set the address and port members of the "packet_info" structure. Set the columns in "fill_in_columns()"; this means that if we're showing COL_{DEF,RES,UNRES}_SRC" or "COL_{DEF,RES,UNRES}_DST", we only generate the string from "src" or "dst", we don't generate a string for the link-layer address and then overwrite it with a string for the network-layer address (generating those strings costs CPU). Add support for "conversations", where a "conversation" is (at present) a source and destination address and a source and destination port. (In the future, we may support "conversations" above the transport layer, e.g. a TFTP conversation, where the first packet goes from the client to the TFTP server port, but the reply comes back from a different port, and all subsequent packets go between the client address/port and the server address/new port, or an NFS conversation, which might include lock manager, status monitor, and mount packets, as well as NFS packets.) Currently, all we support is a call that takes the source and destination address/port pairs, looks them up in a hash table, and: if nothing is found, creates a new entry in the hash table, and assigns it a unique 32-bit conversation ID, and returns that conversation ID; if an entry is found, returns its conversation ID. Use that in the SMB and AFS code to keep track of individual SMB or AFS conversations. We need to match up requests and replies, as, for certain replies, the operation code for the request to which it's a reply doesn't show up in the reply - you have to find the request with a matching transaction ID. Transaction IDs are per-conversation, so the hash table for requests should include a conversation ID and transaction ID as the key. This allows SMB and AFS decoders to handle IPv4 or IPv6 addresses transparently (and should allow the SMB decoder to handle NetBIOS atop other protocols as well, if the source and destination address and port values in the "packet_info" structure are set appropriately). In the "Follow TCP Connection" code, check to make sure that the addresses are IPv4 addressses; ultimately, that code should be changed to use the conversation code instead, which will let it handle IPv6 transparently. svn path=/trunk/; revision=909
1999-10-22 07:18:23 +00:00
/*
* Yes. It's the same conversation, and the two
* address/port pairs are going in the same direction.
Generalize the "ip_src" and "ip_dst" members of the "packet_info" structure to "dl_src"/"dl_dst", "net_src"/"net_dst", and "src"/"dst" addresses, where an address is an address type, an address length in bytes, and a pointer to that many bytes. "dl_{src,dst}" are the link-layer source/destination; "net_{src,dst}" are the network-layer source/destination; "{src,dst}" are the source/destination from the highest of those two layers that we have in the packet. Add a port type to "packet_info" as well, specifying whether it's a TCP or UDP port. Don't set the address and port columns in the dissector functions; just set the address and port members of the "packet_info" structure. Set the columns in "fill_in_columns()"; this means that if we're showing COL_{DEF,RES,UNRES}_SRC" or "COL_{DEF,RES,UNRES}_DST", we only generate the string from "src" or "dst", we don't generate a string for the link-layer address and then overwrite it with a string for the network-layer address (generating those strings costs CPU). Add support for "conversations", where a "conversation" is (at present) a source and destination address and a source and destination port. (In the future, we may support "conversations" above the transport layer, e.g. a TFTP conversation, where the first packet goes from the client to the TFTP server port, but the reply comes back from a different port, and all subsequent packets go between the client address/port and the server address/new port, or an NFS conversation, which might include lock manager, status monitor, and mount packets, as well as NFS packets.) Currently, all we support is a call that takes the source and destination address/port pairs, looks them up in a hash table, and: if nothing is found, creates a new entry in the hash table, and assigns it a unique 32-bit conversation ID, and returns that conversation ID; if an entry is found, returns its conversation ID. Use that in the SMB and AFS code to keep track of individual SMB or AFS conversations. We need to match up requests and replies, as, for certain replies, the operation code for the request to which it's a reply doesn't show up in the reply - you have to find the request with a matching transaction ID. Transaction IDs are per-conversation, so the hash table for requests should include a conversation ID and transaction ID as the key. This allows SMB and AFS decoders to handle IPv4 or IPv6 addresses transparently (and should allow the SMB decoder to handle NetBIOS atop other protocols as well, if the source and destination address and port values in the "packet_info" structure are set appropriately). In the "Follow TCP Connection" code, check to make sure that the addresses are IPv4 addressses; ultimately, that code should be changed to use the conversation code instead, which will let it handle IPv6 transparently. svn path=/trunk/; revision=909
1999-10-22 07:18:23 +00:00
*/
return 1;
}
/*
* Is the first port 2 the same as the second port 1, the first
* port 1 the same as the second port 2, the first address 2
* the same as the second address 1, and the first address 1
* the same as the second address 2?
*/
if (v1->port2 == v2->port1 &&
v1->port1 == v2->port2 &&
addresses_equal(&v1->addr2, &v2->addr1) &&
addresses_equal(&v1->addr1, &v2->addr2)) {
Generalize the "ip_src" and "ip_dst" members of the "packet_info" structure to "dl_src"/"dl_dst", "net_src"/"net_dst", and "src"/"dst" addresses, where an address is an address type, an address length in bytes, and a pointer to that many bytes. "dl_{src,dst}" are the link-layer source/destination; "net_{src,dst}" are the network-layer source/destination; "{src,dst}" are the source/destination from the highest of those two layers that we have in the packet. Add a port type to "packet_info" as well, specifying whether it's a TCP or UDP port. Don't set the address and port columns in the dissector functions; just set the address and port members of the "packet_info" structure. Set the columns in "fill_in_columns()"; this means that if we're showing COL_{DEF,RES,UNRES}_SRC" or "COL_{DEF,RES,UNRES}_DST", we only generate the string from "src" or "dst", we don't generate a string for the link-layer address and then overwrite it with a string for the network-layer address (generating those strings costs CPU). Add support for "conversations", where a "conversation" is (at present) a source and destination address and a source and destination port. (In the future, we may support "conversations" above the transport layer, e.g. a TFTP conversation, where the first packet goes from the client to the TFTP server port, but the reply comes back from a different port, and all subsequent packets go between the client address/port and the server address/new port, or an NFS conversation, which might include lock manager, status monitor, and mount packets, as well as NFS packets.) Currently, all we support is a call that takes the source and destination address/port pairs, looks them up in a hash table, and: if nothing is found, creates a new entry in the hash table, and assigns it a unique 32-bit conversation ID, and returns that conversation ID; if an entry is found, returns its conversation ID. Use that in the SMB and AFS code to keep track of individual SMB or AFS conversations. We need to match up requests and replies, as, for certain replies, the operation code for the request to which it's a reply doesn't show up in the reply - you have to find the request with a matching transaction ID. Transaction IDs are per-conversation, so the hash table for requests should include a conversation ID and transaction ID as the key. This allows SMB and AFS decoders to handle IPv4 or IPv6 addresses transparently (and should allow the SMB decoder to handle NetBIOS atop other protocols as well, if the source and destination address and port values in the "packet_info" structure are set appropriately). In the "Follow TCP Connection" code, check to make sure that the addresses are IPv4 addressses; ultimately, that code should be changed to use the conversation code instead, which will let it handle IPv6 transparently. svn path=/trunk/; revision=909
1999-10-22 07:18:23 +00:00
/*
* Yes. It's the same conversation, and the two
* address/port pairs are going in opposite directions.
Generalize the "ip_src" and "ip_dst" members of the "packet_info" structure to "dl_src"/"dl_dst", "net_src"/"net_dst", and "src"/"dst" addresses, where an address is an address type, an address length in bytes, and a pointer to that many bytes. "dl_{src,dst}" are the link-layer source/destination; "net_{src,dst}" are the network-layer source/destination; "{src,dst}" are the source/destination from the highest of those two layers that we have in the packet. Add a port type to "packet_info" as well, specifying whether it's a TCP or UDP port. Don't set the address and port columns in the dissector functions; just set the address and port members of the "packet_info" structure. Set the columns in "fill_in_columns()"; this means that if we're showing COL_{DEF,RES,UNRES}_SRC" or "COL_{DEF,RES,UNRES}_DST", we only generate the string from "src" or "dst", we don't generate a string for the link-layer address and then overwrite it with a string for the network-layer address (generating those strings costs CPU). Add support for "conversations", where a "conversation" is (at present) a source and destination address and a source and destination port. (In the future, we may support "conversations" above the transport layer, e.g. a TFTP conversation, where the first packet goes from the client to the TFTP server port, but the reply comes back from a different port, and all subsequent packets go between the client address/port and the server address/new port, or an NFS conversation, which might include lock manager, status monitor, and mount packets, as well as NFS packets.) Currently, all we support is a call that takes the source and destination address/port pairs, looks them up in a hash table, and: if nothing is found, creates a new entry in the hash table, and assigns it a unique 32-bit conversation ID, and returns that conversation ID; if an entry is found, returns its conversation ID. Use that in the SMB and AFS code to keep track of individual SMB or AFS conversations. We need to match up requests and replies, as, for certain replies, the operation code for the request to which it's a reply doesn't show up in the reply - you have to find the request with a matching transaction ID. Transaction IDs are per-conversation, so the hash table for requests should include a conversation ID and transaction ID as the key. This allows SMB and AFS decoders to handle IPv4 or IPv6 addresses transparently (and should allow the SMB decoder to handle NetBIOS atop other protocols as well, if the source and destination address and port values in the "packet_info" structure are set appropriately). In the "Follow TCP Connection" code, check to make sure that the addresses are IPv4 addressses; ultimately, that code should be changed to use the conversation code instead, which will let it handle IPv6 transparently. svn path=/trunk/; revision=909
1999-10-22 07:18:23 +00:00
*/
return 1;
Generalize the "ip_src" and "ip_dst" members of the "packet_info" structure to "dl_src"/"dl_dst", "net_src"/"net_dst", and "src"/"dst" addresses, where an address is an address type, an address length in bytes, and a pointer to that many bytes. "dl_{src,dst}" are the link-layer source/destination; "net_{src,dst}" are the network-layer source/destination; "{src,dst}" are the source/destination from the highest of those two layers that we have in the packet. Add a port type to "packet_info" as well, specifying whether it's a TCP or UDP port. Don't set the address and port columns in the dissector functions; just set the address and port members of the "packet_info" structure. Set the columns in "fill_in_columns()"; this means that if we're showing COL_{DEF,RES,UNRES}_SRC" or "COL_{DEF,RES,UNRES}_DST", we only generate the string from "src" or "dst", we don't generate a string for the link-layer address and then overwrite it with a string for the network-layer address (generating those strings costs CPU). Add support for "conversations", where a "conversation" is (at present) a source and destination address and a source and destination port. (In the future, we may support "conversations" above the transport layer, e.g. a TFTP conversation, where the first packet goes from the client to the TFTP server port, but the reply comes back from a different port, and all subsequent packets go between the client address/port and the server address/new port, or an NFS conversation, which might include lock manager, status monitor, and mount packets, as well as NFS packets.) Currently, all we support is a call that takes the source and destination address/port pairs, looks them up in a hash table, and: if nothing is found, creates a new entry in the hash table, and assigns it a unique 32-bit conversation ID, and returns that conversation ID; if an entry is found, returns its conversation ID. Use that in the SMB and AFS code to keep track of individual SMB or AFS conversations. We need to match up requests and replies, as, for certain replies, the operation code for the request to which it's a reply doesn't show up in the reply - you have to find the request with a matching transaction ID. Transaction IDs are per-conversation, so the hash table for requests should include a conversation ID and transaction ID as the key. This allows SMB and AFS decoders to handle IPv4 or IPv6 addresses transparently (and should allow the SMB decoder to handle NetBIOS atop other protocols as well, if the source and destination address and port values in the "packet_info" structure are set appropriately). In the "Follow TCP Connection" code, check to make sure that the addresses are IPv4 addressses; ultimately, that code should be changed to use the conversation code instead, which will let it handle IPv6 transparently. svn path=/trunk/; revision=909
1999-10-22 07:18:23 +00:00
}
/*
* The addresses or the ports don't match.
*/
Generalize the "ip_src" and "ip_dst" members of the "packet_info" structure to "dl_src"/"dl_dst", "net_src"/"net_dst", and "src"/"dst" addresses, where an address is an address type, an address length in bytes, and a pointer to that many bytes. "dl_{src,dst}" are the link-layer source/destination; "net_{src,dst}" are the network-layer source/destination; "{src,dst}" are the source/destination from the highest of those two layers that we have in the packet. Add a port type to "packet_info" as well, specifying whether it's a TCP or UDP port. Don't set the address and port columns in the dissector functions; just set the address and port members of the "packet_info" structure. Set the columns in "fill_in_columns()"; this means that if we're showing COL_{DEF,RES,UNRES}_SRC" or "COL_{DEF,RES,UNRES}_DST", we only generate the string from "src" or "dst", we don't generate a string for the link-layer address and then overwrite it with a string for the network-layer address (generating those strings costs CPU). Add support for "conversations", where a "conversation" is (at present) a source and destination address and a source and destination port. (In the future, we may support "conversations" above the transport layer, e.g. a TFTP conversation, where the first packet goes from the client to the TFTP server port, but the reply comes back from a different port, and all subsequent packets go between the client address/port and the server address/new port, or an NFS conversation, which might include lock manager, status monitor, and mount packets, as well as NFS packets.) Currently, all we support is a call that takes the source and destination address/port pairs, looks them up in a hash table, and: if nothing is found, creates a new entry in the hash table, and assigns it a unique 32-bit conversation ID, and returns that conversation ID; if an entry is found, returns its conversation ID. Use that in the SMB and AFS code to keep track of individual SMB or AFS conversations. We need to match up requests and replies, as, for certain replies, the operation code for the request to which it's a reply doesn't show up in the reply - you have to find the request with a matching transaction ID. Transaction IDs are per-conversation, so the hash table for requests should include a conversation ID and transaction ID as the key. This allows SMB and AFS decoders to handle IPv4 or IPv6 addresses transparently (and should allow the SMB decoder to handle NetBIOS atop other protocols as well, if the source and destination address and port values in the "packet_info" structure are set appropriately). In the "Follow TCP Connection" code, check to make sure that the addresses are IPv4 addressses; ultimately, that code should be changed to use the conversation code instead, which will let it handle IPv6 transparently. svn path=/trunk/; revision=909
1999-10-22 07:18:23 +00:00
return 0;
}
/*
* Compute the hash value for two given address/port pairs if the match
* has a wildcard address 2.
Generalize the "ip_src" and "ip_dst" members of the "packet_info" structure to "dl_src"/"dl_dst", "net_src"/"net_dst", and "src"/"dst" addresses, where an address is an address type, an address length in bytes, and a pointer to that many bytes. "dl_{src,dst}" are the link-layer source/destination; "net_{src,dst}" are the network-layer source/destination; "{src,dst}" are the source/destination from the highest of those two layers that we have in the packet. Add a port type to "packet_info" as well, specifying whether it's a TCP or UDP port. Don't set the address and port columns in the dissector functions; just set the address and port members of the "packet_info" structure. Set the columns in "fill_in_columns()"; this means that if we're showing COL_{DEF,RES,UNRES}_SRC" or "COL_{DEF,RES,UNRES}_DST", we only generate the string from "src" or "dst", we don't generate a string for the link-layer address and then overwrite it with a string for the network-layer address (generating those strings costs CPU). Add support for "conversations", where a "conversation" is (at present) a source and destination address and a source and destination port. (In the future, we may support "conversations" above the transport layer, e.g. a TFTP conversation, where the first packet goes from the client to the TFTP server port, but the reply comes back from a different port, and all subsequent packets go between the client address/port and the server address/new port, or an NFS conversation, which might include lock manager, status monitor, and mount packets, as well as NFS packets.) Currently, all we support is a call that takes the source and destination address/port pairs, looks them up in a hash table, and: if nothing is found, creates a new entry in the hash table, and assigns it a unique 32-bit conversation ID, and returns that conversation ID; if an entry is found, returns its conversation ID. Use that in the SMB and AFS code to keep track of individual SMB or AFS conversations. We need to match up requests and replies, as, for certain replies, the operation code for the request to which it's a reply doesn't show up in the reply - you have to find the request with a matching transaction ID. Transaction IDs are per-conversation, so the hash table for requests should include a conversation ID and transaction ID as the key. This allows SMB and AFS decoders to handle IPv4 or IPv6 addresses transparently (and should allow the SMB decoder to handle NetBIOS atop other protocols as well, if the source and destination address and port values in the "packet_info" structure are set appropriately). In the "Follow TCP Connection" code, check to make sure that the addresses are IPv4 addressses; ultimately, that code should be changed to use the conversation code instead, which will let it handle IPv6 transparently. svn path=/trunk/; revision=909
1999-10-22 07:18:23 +00:00
*/
static guint
conversation_hash_no_addr2(gconstpointer v)
{
const conversation_key_t key = (const conversation_key_t)v;
guint hash_val;
address tmp_addr;
hash_val = 0;
tmp_addr.len = 4;
hash_val = add_address_to_hash(hash_val, &key->addr1);
tmp_addr.data = &key->port1;
hash_val = add_address_to_hash(hash_val, &tmp_addr);
tmp_addr.data = &key->port2;
hash_val = add_address_to_hash(hash_val, &tmp_addr);
hash_val += ( hash_val << 3 );
hash_val ^= ( hash_val >> 11 );
hash_val += ( hash_val << 15 );
return hash_val;
}
/*
* Compare two conversation keys, except for the address 2 value.
* We don't check both directions of the conversation - the routine
* doing the hash lookup has to do two searches, as the hash key
* will be different for the two directions.
*/
static gint
conversation_match_no_addr2(gconstpointer v, gconstpointer w)
Generalize the "ip_src" and "ip_dst" members of the "packet_info" structure to "dl_src"/"dl_dst", "net_src"/"net_dst", and "src"/"dst" addresses, where an address is an address type, an address length in bytes, and a pointer to that many bytes. "dl_{src,dst}" are the link-layer source/destination; "net_{src,dst}" are the network-layer source/destination; "{src,dst}" are the source/destination from the highest of those two layers that we have in the packet. Add a port type to "packet_info" as well, specifying whether it's a TCP or UDP port. Don't set the address and port columns in the dissector functions; just set the address and port members of the "packet_info" structure. Set the columns in "fill_in_columns()"; this means that if we're showing COL_{DEF,RES,UNRES}_SRC" or "COL_{DEF,RES,UNRES}_DST", we only generate the string from "src" or "dst", we don't generate a string for the link-layer address and then overwrite it with a string for the network-layer address (generating those strings costs CPU). Add support for "conversations", where a "conversation" is (at present) a source and destination address and a source and destination port. (In the future, we may support "conversations" above the transport layer, e.g. a TFTP conversation, where the first packet goes from the client to the TFTP server port, but the reply comes back from a different port, and all subsequent packets go between the client address/port and the server address/new port, or an NFS conversation, which might include lock manager, status monitor, and mount packets, as well as NFS packets.) Currently, all we support is a call that takes the source and destination address/port pairs, looks them up in a hash table, and: if nothing is found, creates a new entry in the hash table, and assigns it a unique 32-bit conversation ID, and returns that conversation ID; if an entry is found, returns its conversation ID. Use that in the SMB and AFS code to keep track of individual SMB or AFS conversations. We need to match up requests and replies, as, for certain replies, the operation code for the request to which it's a reply doesn't show up in the reply - you have to find the request with a matching transaction ID. Transaction IDs are per-conversation, so the hash table for requests should include a conversation ID and transaction ID as the key. This allows SMB and AFS decoders to handle IPv4 or IPv6 addresses transparently (and should allow the SMB decoder to handle NetBIOS atop other protocols as well, if the source and destination address and port values in the "packet_info" structure are set appropriately). In the "Follow TCP Connection" code, check to make sure that the addresses are IPv4 addressses; ultimately, that code should be changed to use the conversation code instead, which will let it handle IPv6 transparently. svn path=/trunk/; revision=909
1999-10-22 07:18:23 +00:00
{
const conversation_key_t v1 = (const conversation_key_t)v;
const conversation_key_t v2 = (const conversation_key_t)w;
if (v1->etype != v2->etype)
return 0; /* different types of port */
/*
* Are the first and second port 1 values the same, the first and
* second port 2 valuess the same, and the first and second
* address 1 values the same?
*/
if (v1->port1 == v2->port1 &&
v1->port2 == v2->port2 &&
addresses_equal(&v1->addr1, &v2->addr1)) {
/*
* Yes. It's the same conversation, and the two
* address/port pairs are going in the same direction.
*/
return 1;
}
/*
* The addresses or the ports don't match.
*/
return 0;
}
/*
* Compute the hash value for two given address/port pairs if the match
* has a wildcard port 2.
*/
static guint
conversation_hash_no_port2(gconstpointer v)
{
const conversation_key_t key = (const conversation_key_t)v;
Generalize the "ip_src" and "ip_dst" members of the "packet_info" structure to "dl_src"/"dl_dst", "net_src"/"net_dst", and "src"/"dst" addresses, where an address is an address type, an address length in bytes, and a pointer to that many bytes. "dl_{src,dst}" are the link-layer source/destination; "net_{src,dst}" are the network-layer source/destination; "{src,dst}" are the source/destination from the highest of those two layers that we have in the packet. Add a port type to "packet_info" as well, specifying whether it's a TCP or UDP port. Don't set the address and port columns in the dissector functions; just set the address and port members of the "packet_info" structure. Set the columns in "fill_in_columns()"; this means that if we're showing COL_{DEF,RES,UNRES}_SRC" or "COL_{DEF,RES,UNRES}_DST", we only generate the string from "src" or "dst", we don't generate a string for the link-layer address and then overwrite it with a string for the network-layer address (generating those strings costs CPU). Add support for "conversations", where a "conversation" is (at present) a source and destination address and a source and destination port. (In the future, we may support "conversations" above the transport layer, e.g. a TFTP conversation, where the first packet goes from the client to the TFTP server port, but the reply comes back from a different port, and all subsequent packets go between the client address/port and the server address/new port, or an NFS conversation, which might include lock manager, status monitor, and mount packets, as well as NFS packets.) Currently, all we support is a call that takes the source and destination address/port pairs, looks them up in a hash table, and: if nothing is found, creates a new entry in the hash table, and assigns it a unique 32-bit conversation ID, and returns that conversation ID; if an entry is found, returns its conversation ID. Use that in the SMB and AFS code to keep track of individual SMB or AFS conversations. We need to match up requests and replies, as, for certain replies, the operation code for the request to which it's a reply doesn't show up in the reply - you have to find the request with a matching transaction ID. Transaction IDs are per-conversation, so the hash table for requests should include a conversation ID and transaction ID as the key. This allows SMB and AFS decoders to handle IPv4 or IPv6 addresses transparently (and should allow the SMB decoder to handle NetBIOS atop other protocols as well, if the source and destination address and port values in the "packet_info" structure are set appropriately). In the "Follow TCP Connection" code, check to make sure that the addresses are IPv4 addressses; ultimately, that code should be changed to use the conversation code instead, which will let it handle IPv6 transparently. svn path=/trunk/; revision=909
1999-10-22 07:18:23 +00:00
guint hash_val;
address tmp_addr;
Generalize the "ip_src" and "ip_dst" members of the "packet_info" structure to "dl_src"/"dl_dst", "net_src"/"net_dst", and "src"/"dst" addresses, where an address is an address type, an address length in bytes, and a pointer to that many bytes. "dl_{src,dst}" are the link-layer source/destination; "net_{src,dst}" are the network-layer source/destination; "{src,dst}" are the source/destination from the highest of those two layers that we have in the packet. Add a port type to "packet_info" as well, specifying whether it's a TCP or UDP port. Don't set the address and port columns in the dissector functions; just set the address and port members of the "packet_info" structure. Set the columns in "fill_in_columns()"; this means that if we're showing COL_{DEF,RES,UNRES}_SRC" or "COL_{DEF,RES,UNRES}_DST", we only generate the string from "src" or "dst", we don't generate a string for the link-layer address and then overwrite it with a string for the network-layer address (generating those strings costs CPU). Add support for "conversations", where a "conversation" is (at present) a source and destination address and a source and destination port. (In the future, we may support "conversations" above the transport layer, e.g. a TFTP conversation, where the first packet goes from the client to the TFTP server port, but the reply comes back from a different port, and all subsequent packets go between the client address/port and the server address/new port, or an NFS conversation, which might include lock manager, status monitor, and mount packets, as well as NFS packets.) Currently, all we support is a call that takes the source and destination address/port pairs, looks them up in a hash table, and: if nothing is found, creates a new entry in the hash table, and assigns it a unique 32-bit conversation ID, and returns that conversation ID; if an entry is found, returns its conversation ID. Use that in the SMB and AFS code to keep track of individual SMB or AFS conversations. We need to match up requests and replies, as, for certain replies, the operation code for the request to which it's a reply doesn't show up in the reply - you have to find the request with a matching transaction ID. Transaction IDs are per-conversation, so the hash table for requests should include a conversation ID and transaction ID as the key. This allows SMB and AFS decoders to handle IPv4 or IPv6 addresses transparently (and should allow the SMB decoder to handle NetBIOS atop other protocols as well, if the source and destination address and port values in the "packet_info" structure are set appropriately). In the "Follow TCP Connection" code, check to make sure that the addresses are IPv4 addressses; ultimately, that code should be changed to use the conversation code instead, which will let it handle IPv6 transparently. svn path=/trunk/; revision=909
1999-10-22 07:18:23 +00:00
hash_val = 0;
tmp_addr.len = 4;
hash_val = add_address_to_hash(hash_val, &key->addr1);
tmp_addr.data = &key->port1;
hash_val = add_address_to_hash(hash_val, &tmp_addr);
hash_val = add_address_to_hash(hash_val, &key->addr2);
hash_val += ( hash_val << 3 );
hash_val ^= ( hash_val >> 11 );
hash_val += ( hash_val << 15 );
return hash_val;
}
/*
* Compare two conversation keys, except for the port 2 value.
* We don't check both directions of the conversation - the routine
* doing the hash lookup has to do two searches, as the hash key
* will be different for the two directions.
*/
static gint
conversation_match_no_port2(gconstpointer v, gconstpointer w)
{
const conversation_key_t v1 = (const conversation_key_t)v;
const conversation_key_t v2 = (const conversation_key_t)w;
if (v1->etype != v2->etype)
return 0; /* different types of port */
/*
* Are the first and second port 1 values the same, the first and
* second address 1 values the same, and the first and second
* address 2 values the same?
*/
if (v1->port1 == v2->port1 &&
addresses_equal(&v1->addr1, &v2->addr1) &&
addresses_equal(&v1->addr2, &v2->addr2)) {
/*
* Yes. It's the same conversation, and the two
* address/port pairs are going in the same direction.
*/
return 1;
}
/*
* The addresses or the ports don't match.
*/
return 0;
}
/*
* Compute the hash value for two given address/port pairs if the match
* has a wildcard address 2 and port 2.
*/
static guint
conversation_hash_no_addr2_or_port2(gconstpointer v)
{
const conversation_key_t key = (const conversation_key_t)v;
guint hash_val;
address tmp_addr;
hash_val = 0;
tmp_addr.len = 4;
hash_val = add_address_to_hash(hash_val, &key->addr1);
tmp_addr.data = &key->port1;
hash_val = add_address_to_hash(hash_val, &tmp_addr);
hash_val += ( hash_val << 3 );
hash_val ^= ( hash_val >> 11 );
hash_val += ( hash_val << 15 );
Generalize the "ip_src" and "ip_dst" members of the "packet_info" structure to "dl_src"/"dl_dst", "net_src"/"net_dst", and "src"/"dst" addresses, where an address is an address type, an address length in bytes, and a pointer to that many bytes. "dl_{src,dst}" are the link-layer source/destination; "net_{src,dst}" are the network-layer source/destination; "{src,dst}" are the source/destination from the highest of those two layers that we have in the packet. Add a port type to "packet_info" as well, specifying whether it's a TCP or UDP port. Don't set the address and port columns in the dissector functions; just set the address and port members of the "packet_info" structure. Set the columns in "fill_in_columns()"; this means that if we're showing COL_{DEF,RES,UNRES}_SRC" or "COL_{DEF,RES,UNRES}_DST", we only generate the string from "src" or "dst", we don't generate a string for the link-layer address and then overwrite it with a string for the network-layer address (generating those strings costs CPU). Add support for "conversations", where a "conversation" is (at present) a source and destination address and a source and destination port. (In the future, we may support "conversations" above the transport layer, e.g. a TFTP conversation, where the first packet goes from the client to the TFTP server port, but the reply comes back from a different port, and all subsequent packets go between the client address/port and the server address/new port, or an NFS conversation, which might include lock manager, status monitor, and mount packets, as well as NFS packets.) Currently, all we support is a call that takes the source and destination address/port pairs, looks them up in a hash table, and: if nothing is found, creates a new entry in the hash table, and assigns it a unique 32-bit conversation ID, and returns that conversation ID; if an entry is found, returns its conversation ID. Use that in the SMB and AFS code to keep track of individual SMB or AFS conversations. We need to match up requests and replies, as, for certain replies, the operation code for the request to which it's a reply doesn't show up in the reply - you have to find the request with a matching transaction ID. Transaction IDs are per-conversation, so the hash table for requests should include a conversation ID and transaction ID as the key. This allows SMB and AFS decoders to handle IPv4 or IPv6 addresses transparently (and should allow the SMB decoder to handle NetBIOS atop other protocols as well, if the source and destination address and port values in the "packet_info" structure are set appropriately). In the "Follow TCP Connection" code, check to make sure that the addresses are IPv4 addressses; ultimately, that code should be changed to use the conversation code instead, which will let it handle IPv6 transparently. svn path=/trunk/; revision=909
1999-10-22 07:18:23 +00:00
return hash_val;
}
/*
* Compare the address 1 and port 1 in the two conversation keys.
* We don't check both directions of the conversation - the routine
* doing the hash lookup has to do two searches, as the hash key
* will be different for the two directions.
*/
static gint
conversation_match_no_addr2_or_port2(gconstpointer v, gconstpointer w)
{
const conversation_key_t v1 = (const conversation_key_t)v;
const conversation_key_t v2 = (const conversation_key_t)w;
if (v1->etype != v2->etype)
return 0; /* different types of port */
/*
* Are the first and second port 1 values the same and the first
* and second address 1 values the same?
*/
if (v1->port1 == v2->port1 &&
addresses_equal(&v1->addr1, &v2->addr1)) {
/*
* Yes. It's the same conversation, and the two
* address/port pairs are going in the same direction.
*/
return 1;
}
/*
* The addresses or the ports don't match.
*/
return 0;
}
/**
* Create a new hash tables for conversations.
*/
void
conversation_init(void)
{
/*
* Free up any space allocated for conversation protocol data
* areas.
*
* We can free the space, as the structures it contains are
* pointed to by conversation data structures that were freed
* above.
*/
conversation_hashtable_exact =
wmem_map_new_autoreset(wmem_epan_scope(), wmem_file_scope(), conversation_hash_exact,
conversation_match_exact);
conversation_hashtable_no_addr2 =
wmem_map_new_autoreset(wmem_epan_scope(), wmem_file_scope(), conversation_hash_no_addr2,
conversation_match_no_addr2);
conversation_hashtable_no_port2 =
wmem_map_new_autoreset(wmem_epan_scope(), wmem_file_scope(), conversation_hash_no_port2,
conversation_match_no_port2);
conversation_hashtable_no_addr2_or_port2 =
wmem_map_new_autoreset(wmem_epan_scope(), wmem_file_scope(), conversation_hash_no_addr2_or_port2,
conversation_match_no_addr2_or_port2);
}
/**
* Initialize some variables every time a file is loaded or re-loaded.
*/
void conversation_epan_reset(void)
{
Generalize the "ip_src" and "ip_dst" members of the "packet_info" structure to "dl_src"/"dl_dst", "net_src"/"net_dst", and "src"/"dst" addresses, where an address is an address type, an address length in bytes, and a pointer to that many bytes. "dl_{src,dst}" are the link-layer source/destination; "net_{src,dst}" are the network-layer source/destination; "{src,dst}" are the source/destination from the highest of those two layers that we have in the packet. Add a port type to "packet_info" as well, specifying whether it's a TCP or UDP port. Don't set the address and port columns in the dissector functions; just set the address and port members of the "packet_info" structure. Set the columns in "fill_in_columns()"; this means that if we're showing COL_{DEF,RES,UNRES}_SRC" or "COL_{DEF,RES,UNRES}_DST", we only generate the string from "src" or "dst", we don't generate a string for the link-layer address and then overwrite it with a string for the network-layer address (generating those strings costs CPU). Add support for "conversations", where a "conversation" is (at present) a source and destination address and a source and destination port. (In the future, we may support "conversations" above the transport layer, e.g. a TFTP conversation, where the first packet goes from the client to the TFTP server port, but the reply comes back from a different port, and all subsequent packets go between the client address/port and the server address/new port, or an NFS conversation, which might include lock manager, status monitor, and mount packets, as well as NFS packets.) Currently, all we support is a call that takes the source and destination address/port pairs, looks them up in a hash table, and: if nothing is found, creates a new entry in the hash table, and assigns it a unique 32-bit conversation ID, and returns that conversation ID; if an entry is found, returns its conversation ID. Use that in the SMB and AFS code to keep track of individual SMB or AFS conversations. We need to match up requests and replies, as, for certain replies, the operation code for the request to which it's a reply doesn't show up in the reply - you have to find the request with a matching transaction ID. Transaction IDs are per-conversation, so the hash table for requests should include a conversation ID and transaction ID as the key. This allows SMB and AFS decoders to handle IPv4 or IPv6 addresses transparently (and should allow the SMB decoder to handle NetBIOS atop other protocols as well, if the source and destination address and port values in the "packet_info" structure are set appropriately). In the "Follow TCP Connection" code, check to make sure that the addresses are IPv4 addressses; ultimately, that code should be changed to use the conversation code instead, which will let it handle IPv6 transparently. svn path=/trunk/; revision=909
1999-10-22 07:18:23 +00:00
/*
* Start the conversation indices over at 0.
*/
new_index = 0;
}
/*
* Does the right thing when inserting into one of the conversation hash tables,
* taking into account ordering and hash chains and all that good stuff.
*
* Mostly adapted from the old conversation_new().
*/
static void
conversation_insert_into_hashtable(wmem_map_t *hashtable, conversation_t *conv)
{
conversation_t *chain_head, *chain_tail, *cur, *prev;
chain_head = (conversation_t *)wmem_map_lookup(hashtable, conv->key_ptr);
if (NULL==chain_head) {
/* New entry */
conv->next = NULL;
conv->last = conv;
wmem_map_insert(hashtable, conv->key_ptr, conv);
DPRINT(("created a new conversation chain"));
}
else {
/* There's an existing chain for this key */
DPRINT(("there's an existing conversation chain"));
chain_tail = chain_head->last;
if(conv->setup_frame >= chain_tail->setup_frame) {
/* This convo belongs at the end of the chain */
conv->next = NULL;
conv->last = NULL;
chain_tail->next = conv;
chain_head->last = conv;
}
else {
/* Loop through the chain to find the right spot */
cur = chain_head;
prev = NULL;
for (; (conv->setup_frame > cur->setup_frame) && cur->next; prev=cur, cur=cur->next)
;
if (NULL==prev) {
/* Changing the head of the chain */
conv->next = chain_head;
conv->last = chain_tail;
chain_head->last = NULL;
wmem_map_insert(hashtable, conv->key_ptr, conv);
}
else {
/* Inserting into the middle of the chain */
conv->next = cur;
conv->last = NULL;
prev->next = conv;
}
}
}
}
/*
* Does the right thing when removing from one of the conversation hash tables,
* taking into account ordering and hash chains and all that good stuff.
*/
static void
conversation_remove_from_hashtable(wmem_map_t *hashtable, conversation_t *conv)
{
conversation_t *chain_head, *cur, *prev;
chain_head = (conversation_t *)wmem_map_lookup(hashtable, conv->key_ptr);
if (conv == chain_head) {
/* We are currently the front of the chain */
if (NULL == conv->next) {
/* We are the only conversation in the chain, no need to
* update next pointer, but do not call
* wmem_map_remove() either because the conv data
* will be re-inserted. */
wmem_map_steal(hashtable, conv->key_ptr);
}
else {
/* Update the head of the chain */
chain_head = conv->next;
chain_head->last = conv->last;
if (conv->latest_found == conv)
chain_head->latest_found = NULL;
else
chain_head->latest_found = conv->latest_found;
wmem_map_insert(hashtable, chain_head->key_ptr, chain_head);
}
}
else {
/* We are not the front of the chain. Loop through to find us.
* Start loop at chain_head->next rather than chain_head because
* we already know we're not at the head. */
cur = chain_head->next;
prev = chain_head;
for (; (cur != conv) && cur->next; prev=cur, cur=cur->next)
;
if (cur != conv) {
/* XXX: Conversation not found. Wrong hashtable? */
return;
}
prev->next = conv->next;
if (NULL == conv->next) {
/* We're at the very end of the list. */
chain_head->last = prev;
}
if (chain_head->latest_found == conv)
chain_head->latest_found = prev;
}
}
Generalize the "ip_src" and "ip_dst" members of the "packet_info" structure to "dl_src"/"dl_dst", "net_src"/"net_dst", and "src"/"dst" addresses, where an address is an address type, an address length in bytes, and a pointer to that many bytes. "dl_{src,dst}" are the link-layer source/destination; "net_{src,dst}" are the network-layer source/destination; "{src,dst}" are the source/destination from the highest of those two layers that we have in the packet. Add a port type to "packet_info" as well, specifying whether it's a TCP or UDP port. Don't set the address and port columns in the dissector functions; just set the address and port members of the "packet_info" structure. Set the columns in "fill_in_columns()"; this means that if we're showing COL_{DEF,RES,UNRES}_SRC" or "COL_{DEF,RES,UNRES}_DST", we only generate the string from "src" or "dst", we don't generate a string for the link-layer address and then overwrite it with a string for the network-layer address (generating those strings costs CPU). Add support for "conversations", where a "conversation" is (at present) a source and destination address and a source and destination port. (In the future, we may support "conversations" above the transport layer, e.g. a TFTP conversation, where the first packet goes from the client to the TFTP server port, but the reply comes back from a different port, and all subsequent packets go between the client address/port and the server address/new port, or an NFS conversation, which might include lock manager, status monitor, and mount packets, as well as NFS packets.) Currently, all we support is a call that takes the source and destination address/port pairs, looks them up in a hash table, and: if nothing is found, creates a new entry in the hash table, and assigns it a unique 32-bit conversation ID, and returns that conversation ID; if an entry is found, returns its conversation ID. Use that in the SMB and AFS code to keep track of individual SMB or AFS conversations. We need to match up requests and replies, as, for certain replies, the operation code for the request to which it's a reply doesn't show up in the reply - you have to find the request with a matching transaction ID. Transaction IDs are per-conversation, so the hash table for requests should include a conversation ID and transaction ID as the key. This allows SMB and AFS decoders to handle IPv4 or IPv6 addresses transparently (and should allow the SMB decoder to handle NetBIOS atop other protocols as well, if the source and destination address and port values in the "packet_info" structure are set appropriately). In the "Follow TCP Connection" code, check to make sure that the addresses are IPv4 addressses; ultimately, that code should be changed to use the conversation code instead, which will let it handle IPv6 transparently. svn path=/trunk/; revision=909
1999-10-22 07:18:23 +00:00
/*
* Given two address/port pairs for a packet, create a new conversation
* to contain packets between those address/port pairs.
*
* The options field is used to specify whether the address 2 value
* and/or port 2 value are not given and any value is acceptable
* when searching for this conversation.
Generalize the "ip_src" and "ip_dst" members of the "packet_info" structure to "dl_src"/"dl_dst", "net_src"/"net_dst", and "src"/"dst" addresses, where an address is an address type, an address length in bytes, and a pointer to that many bytes. "dl_{src,dst}" are the link-layer source/destination; "net_{src,dst}" are the network-layer source/destination; "{src,dst}" are the source/destination from the highest of those two layers that we have in the packet. Add a port type to "packet_info" as well, specifying whether it's a TCP or UDP port. Don't set the address and port columns in the dissector functions; just set the address and port members of the "packet_info" structure. Set the columns in "fill_in_columns()"; this means that if we're showing COL_{DEF,RES,UNRES}_SRC" or "COL_{DEF,RES,UNRES}_DST", we only generate the string from "src" or "dst", we don't generate a string for the link-layer address and then overwrite it with a string for the network-layer address (generating those strings costs CPU). Add support for "conversations", where a "conversation" is (at present) a source and destination address and a source and destination port. (In the future, we may support "conversations" above the transport layer, e.g. a TFTP conversation, where the first packet goes from the client to the TFTP server port, but the reply comes back from a different port, and all subsequent packets go between the client address/port and the server address/new port, or an NFS conversation, which might include lock manager, status monitor, and mount packets, as well as NFS packets.) Currently, all we support is a call that takes the source and destination address/port pairs, looks them up in a hash table, and: if nothing is found, creates a new entry in the hash table, and assigns it a unique 32-bit conversation ID, and returns that conversation ID; if an entry is found, returns its conversation ID. Use that in the SMB and AFS code to keep track of individual SMB or AFS conversations. We need to match up requests and replies, as, for certain replies, the operation code for the request to which it's a reply doesn't show up in the reply - you have to find the request with a matching transaction ID. Transaction IDs are per-conversation, so the hash table for requests should include a conversation ID and transaction ID as the key. This allows SMB and AFS decoders to handle IPv4 or IPv6 addresses transparently (and should allow the SMB decoder to handle NetBIOS atop other protocols as well, if the source and destination address and port values in the "packet_info" structure are set appropriately). In the "Follow TCP Connection" code, check to make sure that the addresses are IPv4 addressses; ultimately, that code should be changed to use the conversation code instead, which will let it handle IPv6 transparently. svn path=/trunk/; revision=909
1999-10-22 07:18:23 +00:00
*/
conversation_t *
conversation_new(const guint32 setup_frame, const address *addr1, const address *addr2,
const endpoint_type etype, const guint32 port1, const guint32 port2, const guint options)
{
/*
DISSECTOR_ASSERT(!(options | CONVERSATION_TEMPLATE) || ((options | (NO_ADDR2 | NO_PORT2 | NO_PORT2_FORCE))) &&
"A conversation template may not be constructed without wildcard options");
*/
wmem_map_t* hashtable;
conversation_t *conversation=NULL;
conversation_key_t new_key;
DPRINT(("creating conversation for frame #%d: %s:%d -> %s:%d (ptype=%d)",
setup_frame, address_to_str(wmem_packet_scope(), addr1), port1,
address_to_str(wmem_packet_scope(), addr2), port2, ptype));
if (options & NO_ADDR2) {
if (options & (NO_PORT2|NO_PORT2_FORCE)) {
hashtable = conversation_hashtable_no_addr2_or_port2;
} else {
hashtable = conversation_hashtable_no_addr2;
}
} else {
if (options & (NO_PORT2|NO_PORT2_FORCE)) {
hashtable = conversation_hashtable_no_port2;
} else {
hashtable = conversation_hashtable_exact;
}
}
new_key = wmem_new(wmem_file_scope(), struct conversation_key);
copy_address_wmem(wmem_file_scope(), &new_key->addr1, addr1);
copy_address_wmem(wmem_file_scope(), &new_key->addr2, addr2);
new_key->etype = etype;
new_key->port1 = port1;
new_key->port2 = port2;
conversation = wmem_new(wmem_file_scope(), conversation_t);
memset(conversation, 0, sizeof(conversation_t));
conversation->conv_index = new_index;
conversation->setup_frame = conversation->last_frame = setup_frame;
conversation->data_list = NULL;
conversation->dissector_tree = wmem_tree_new(wmem_file_scope());
/* set the options and key pointer */
conversation->options = options;
conversation->key_ptr = new_key;
new_index++;
DINDENT();
conversation_insert_into_hashtable(hashtable, conversation);
DENDENT();
return conversation;
}
/*
* Set the port 2 value in a key. Remove the original from table,
* update the options and port values, insert the updated key.
*/
void
conversation_set_port2(conversation_t *conv, const guint32 port)
{
DISSECTOR_ASSERT_HINT(!(conv->options & CONVERSATION_TEMPLATE),
"Use the conversation_create_from_template function when the CONVERSATION_TEMPLATE bit is set in the options mask");
DPRINT(("called for port=%d", port));
/*
* If the port 2 value is not wildcarded, don't set it.
*/
if ((!(conv->options & NO_PORT2)) || (conv->options & NO_PORT2_FORCE))
return;
DINDENT();
if (conv->options & NO_ADDR2) {
conversation_remove_from_hashtable(conversation_hashtable_no_addr2_or_port2, conv);
} else {
conversation_remove_from_hashtable(conversation_hashtable_no_port2, conv);
}
conv->options &= ~NO_PORT2;
conv->key_ptr->port2 = port;
if (conv->options & NO_ADDR2) {
conversation_insert_into_hashtable(conversation_hashtable_no_addr2, conv);
} else {
conversation_insert_into_hashtable(conversation_hashtable_exact, conv);
}
DENDENT();
}
/*
* Set the address 2 value in a key. Remove the original from
* table, update the options and port values, insert the updated key.
*/
void
conversation_set_addr2(conversation_t *conv, const address *addr)
{
char* addr_str;
DISSECTOR_ASSERT_HINT(!(conv->options & CONVERSATION_TEMPLATE),
"Use the conversation_create_from_template function when the CONVERSATION_TEMPLATE bit is set in the options mask");
addr_str = address_to_str(NULL, addr);
DPRINT(("called for addr=%s", addr_str));
wmem_free(NULL, addr_str);
/*
* If the address 2 value is not wildcarded, don't set it.
*/
if (!(conv->options & NO_ADDR2))
return;
DINDENT();
if (conv->options & NO_PORT2) {
conversation_remove_from_hashtable(conversation_hashtable_no_addr2_or_port2, conv);
} else {
conversation_remove_from_hashtable(conversation_hashtable_no_port2, conv);
}
conv->options &= ~NO_ADDR2;
copy_address_wmem(wmem_file_scope(), &conv->key_ptr->addr2, addr);
if (conv->options & NO_PORT2) {
conversation_insert_into_hashtable(conversation_hashtable_no_port2, conv);
} else {
conversation_insert_into_hashtable(conversation_hashtable_exact, conv);
}
DENDENT();
}
/*
* Search a particular hash table for a conversation with the specified
* {addr1, port1, addr2, port2} and set up before frame_num.
*/
static conversation_t *
conversation_lookup_hashtable(wmem_map_t *hashtable, const guint32 frame_num, const address *addr1, const address *addr2,
const endpoint_type etype, const guint32 port1, const guint32 port2)
Generalize the "ip_src" and "ip_dst" members of the "packet_info" structure to "dl_src"/"dl_dst", "net_src"/"net_dst", and "src"/"dst" addresses, where an address is an address type, an address length in bytes, and a pointer to that many bytes. "dl_{src,dst}" are the link-layer source/destination; "net_{src,dst}" are the network-layer source/destination; "{src,dst}" are the source/destination from the highest of those two layers that we have in the packet. Add a port type to "packet_info" as well, specifying whether it's a TCP or UDP port. Don't set the address and port columns in the dissector functions; just set the address and port members of the "packet_info" structure. Set the columns in "fill_in_columns()"; this means that if we're showing COL_{DEF,RES,UNRES}_SRC" or "COL_{DEF,RES,UNRES}_DST", we only generate the string from "src" or "dst", we don't generate a string for the link-layer address and then overwrite it with a string for the network-layer address (generating those strings costs CPU). Add support for "conversations", where a "conversation" is (at present) a source and destination address and a source and destination port. (In the future, we may support "conversations" above the transport layer, e.g. a TFTP conversation, where the first packet goes from the client to the TFTP server port, but the reply comes back from a different port, and all subsequent packets go between the client address/port and the server address/new port, or an NFS conversation, which might include lock manager, status monitor, and mount packets, as well as NFS packets.) Currently, all we support is a call that takes the source and destination address/port pairs, looks them up in a hash table, and: if nothing is found, creates a new entry in the hash table, and assigns it a unique 32-bit conversation ID, and returns that conversation ID; if an entry is found, returns its conversation ID. Use that in the SMB and AFS code to keep track of individual SMB or AFS conversations. We need to match up requests and replies, as, for certain replies, the operation code for the request to which it's a reply doesn't show up in the reply - you have to find the request with a matching transaction ID. Transaction IDs are per-conversation, so the hash table for requests should include a conversation ID and transaction ID as the key. This allows SMB and AFS decoders to handle IPv4 or IPv6 addresses transparently (and should allow the SMB decoder to handle NetBIOS atop other protocols as well, if the source and destination address and port values in the "packet_info" structure are set appropriately). In the "Follow TCP Connection" code, check to make sure that the addresses are IPv4 addressses; ultimately, that code should be changed to use the conversation code instead, which will let it handle IPv6 transparently. svn path=/trunk/; revision=909
1999-10-22 07:18:23 +00:00
{
conversation_t* convo=NULL;
conversation_t* match=NULL;
conversation_t* chain_head=NULL;
struct conversation_key key;
Generalize the "ip_src" and "ip_dst" members of the "packet_info" structure to "dl_src"/"dl_dst", "net_src"/"net_dst", and "src"/"dst" addresses, where an address is an address type, an address length in bytes, and a pointer to that many bytes. "dl_{src,dst}" are the link-layer source/destination; "net_{src,dst}" are the network-layer source/destination; "{src,dst}" are the source/destination from the highest of those two layers that we have in the packet. Add a port type to "packet_info" as well, specifying whether it's a TCP or UDP port. Don't set the address and port columns in the dissector functions; just set the address and port members of the "packet_info" structure. Set the columns in "fill_in_columns()"; this means that if we're showing COL_{DEF,RES,UNRES}_SRC" or "COL_{DEF,RES,UNRES}_DST", we only generate the string from "src" or "dst", we don't generate a string for the link-layer address and then overwrite it with a string for the network-layer address (generating those strings costs CPU). Add support for "conversations", where a "conversation" is (at present) a source and destination address and a source and destination port. (In the future, we may support "conversations" above the transport layer, e.g. a TFTP conversation, where the first packet goes from the client to the TFTP server port, but the reply comes back from a different port, and all subsequent packets go between the client address/port and the server address/new port, or an NFS conversation, which might include lock manager, status monitor, and mount packets, as well as NFS packets.) Currently, all we support is a call that takes the source and destination address/port pairs, looks them up in a hash table, and: if nothing is found, creates a new entry in the hash table, and assigns it a unique 32-bit conversation ID, and returns that conversation ID; if an entry is found, returns its conversation ID. Use that in the SMB and AFS code to keep track of individual SMB or AFS conversations. We need to match up requests and replies, as, for certain replies, the operation code for the request to which it's a reply doesn't show up in the reply - you have to find the request with a matching transaction ID. Transaction IDs are per-conversation, so the hash table for requests should include a conversation ID and transaction ID as the key. This allows SMB and AFS decoders to handle IPv4 or IPv6 addresses transparently (and should allow the SMB decoder to handle NetBIOS atop other protocols as well, if the source and destination address and port values in the "packet_info" structure are set appropriately). In the "Follow TCP Connection" code, check to make sure that the addresses are IPv4 addressses; ultimately, that code should be changed to use the conversation code instead, which will let it handle IPv6 transparently. svn path=/trunk/; revision=909
1999-10-22 07:18:23 +00:00
/*
* We don't make a copy of the address data, we just copy the
* pointer to it, as "key" disappears when we return.
*/
key.addr1 = *addr1;
key.addr2 = *addr2;
key.etype = etype;
key.port1 = port1;
key.port2 = port2;
chain_head = (conversation_t *)wmem_map_lookup(hashtable, &key);
if (chain_head && (chain_head->setup_frame <= frame_num)) {
match = chain_head;
if((chain_head->last)&&(chain_head->last->setup_frame<=frame_num))
return chain_head->last;
if((chain_head->latest_found)&&(chain_head->latest_found->setup_frame<=frame_num))
match = chain_head->latest_found;
for (convo = match; convo && convo->setup_frame <= frame_num; convo = convo->next) {
if (convo->setup_frame > match->setup_frame) {
match = convo;
}
}
}
if (match)
chain_head->latest_found = match;
return match;
}
/*
* Given two address/port pairs for a packet, search for a conversation
* containing packets between those address/port pairs. Returns NULL if
* not found.
*
* We try to find the most exact match that we can, and then proceed to
* try wildcard matches on the "addr_b" and/or "port_b" argument if a more
* exact match failed.
*
* Either or both of the "addr_b" and "port_b" arguments may be specified as
* a wildcard by setting the NO_ADDR_B or NO_PORT_B flags in the "options"
* argument. We do only wildcard matches on addresses and ports specified
* as wildcards.
*
* I.e.:
*
* if neither "addr_b" nor "port_b" were specified as wildcards, we
* do an exact match (addr_a/port_a and addr_b/port_b) and, if that
* succeeds, we return a pointer to the matched conversation;
*
* otherwise, if "port_b" wasn't specified as a wildcard, we try to
* match any address 2 with the specified port 2 (addr_a/port_a and
* {any}/port_b) and, if that succeeds, we return a pointer to the
* matched conversation;
*
* otherwise, if "addr_b" wasn't specified as a wildcard, we try to
* match any port 2 with the specified address 2 (addr_a/port_a and
* addr_b/{any}) and, if that succeeds, we return a pointer to the
* matched conversation;
*
* otherwise, we try to match any address 2 and any port 2
* (addr_a/port_a and {any}/{any}) and, if that succeeds, we return
* a pointer to the matched conversation;
*
* otherwise, we found no matching conversation, and return NULL.
*/
conversation_t *
find_conversation(const guint32 frame_num, const address *addr_a, const address *addr_b, const endpoint_type etype,
const guint32 port_a, const guint32 port_b, const guint options)
{
conversation_t *conversation;
/*
* First try an exact match, if we have two addresses and ports.
*/
if (!(options & (NO_ADDR_B|NO_PORT_B))) {
/*
* Neither search address B nor search port B are wildcarded,
* start out with an exact match.
*/
DPRINT(("trying exact match"));
conversation =
conversation_lookup_hashtable(conversation_hashtable_exact,
frame_num, addr_a, addr_b, etype,
port_a, port_b);
/* Didn't work, try the other direction */
if (conversation == NULL) {
DPRINT(("trying opposite direction"));
conversation =
conversation_lookup_hashtable(conversation_hashtable_exact,
frame_num, addr_b, addr_a, etype,
port_b, port_a);
}
if ((conversation == NULL) && (addr_a->type == AT_FC)) {
/* In Fibre channel, OXID & RXID are never swapped as
* TCP/UDP ports are in TCP/IP.
*/
conversation =
conversation_lookup_hashtable(conversation_hashtable_exact,
frame_num, addr_b, addr_a, etype,
port_a, port_b);
}
DPRINT(("exact match %sfound",conversation?"":"not "));
if (conversation != NULL)
return conversation;
}
/*
* Well, that didn't find anything. Try matches that wildcard
* one of the addresses, if we have two ports.
*/
if (!(options & NO_PORT_B)) {
/*
* Search port B isn't wildcarded.
*
* First try looking for a conversation with the specified
* address A and port A as the first address and port, and
* with any address and the specified port B as the second
* address and port.
* ("addr_b" doesn't take part in this lookup.)
*/
DPRINT(("trying wildcarded dest address"));
conversation =
conversation_lookup_hashtable(conversation_hashtable_no_addr2,
frame_num, addr_a, addr_b, etype, port_a, port_b);
if ((conversation == NULL) && (addr_a->type == AT_FC)) {
/* In Fibre channel, OXID & RXID are never swapped as
* TCP/UDP ports are in TCP/IP.
*/
conversation =
conversation_lookup_hashtable(conversation_hashtable_no_addr2,
frame_num, addr_b, addr_a, etype,
port_a, port_b);
}
if (conversation != NULL) {
/*
* If search address B isn't wildcarded, and this is for a
* connection-oriented protocol, set the second address for this
* conversation to address B, as that's the address that matched the
* wildcarded second address for this conversation.
*
* (This assumes that, for all connection oriented protocols, the
* endpoints of a connection have only one address each, i.e. you
* don't get packets in a given direction coming from more than one
* address, unless the CONVERSATION_TEMPLATE option is set.)
*/
DPRINT(("wildcarded dest address match found"));
if (!(conversation->options & NO_ADDR_B) && etype != ENDPOINT_UDP)
{
if(!(conversation->options & CONVERSATION_TEMPLATE))
{
conversation_set_addr2(conversation, addr_b);
}
else
{
conversation =
conversation_create_from_template(conversation, addr_b, 0);
}
}
return conversation;
}
/*
* Well, that didn't find anything.
* If search address B was specified, try looking for a
* conversation with the specified address B and port B as
* the first address and port, and with any address and the
* specified port A as the second address and port (this
* packet may be going in the opposite direction from the
* first packet in the conversation).
* ("addr_a" doesn't take part in this lookup.)
*/
if (!(options & NO_ADDR_B)) {
DPRINT(("trying dest addr:port as source addr:port with wildcarded dest addr"));
conversation =
conversation_lookup_hashtable(conversation_hashtable_no_addr2,
frame_num, addr_b, addr_a, etype, port_b, port_a);
if (conversation != NULL) {
/*
* If this is for a connection-oriented
* protocol, set the second address for
* this conversation to address A, as
* that's the address that matched the
* wildcarded second address for this
* conversation.
*/
DPRINT(("match found"));
if (etype != ENDPOINT_UDP) {
if(!(conversation->options & CONVERSATION_TEMPLATE))
{
conversation_set_addr2(conversation, addr_a);
}
else
{
conversation =
conversation_create_from_template(conversation, addr_a, 0);
}
}
return conversation;
}
}
}
/*
* Well, that didn't find anything. Try matches that wildcard
* one of the ports, if we have two addresses.
*/
if (!(options & NO_ADDR_B)) {
/*
* Search address B isn't wildcarded.
*
* First try looking for a conversation with the specified
* address A and port A as the first address and port, and
* with the specified address B and any port as the second
* address and port.
* ("port_b" doesn't take part in this lookup.)
*/
DPRINT(("trying wildcarded dest port"));
conversation =
conversation_lookup_hashtable(conversation_hashtable_no_port2,
frame_num, addr_a, addr_b, etype, port_a, port_b);
if ((conversation == NULL) && (addr_a->type == AT_FC)) {
/* In Fibre channel, OXID & RXID are never swapped as
* TCP/UDP ports are in TCP/IP
*/
conversation =
conversation_lookup_hashtable(conversation_hashtable_no_port2,
frame_num, addr_b, addr_a, etype, port_a, port_b);
}
if (conversation != NULL) {
/*
* If search port B isn't wildcarded, and this is for a connection-
* oriented protocol, set the second port for this conversation to
* port B, as that's the port that matched the wildcarded second port
* for this conversation.
*
* (This assumes that, for all connection oriented protocols, the
* endpoints of a connection have only one port each, i.e. you don't
* get packets in a given direction coming from more than one port,
* unless the CONVERSATION_TEMPLATE option is set.)
*/
DPRINT(("match found"));
if (!(conversation->options & NO_PORT_B) && etype != ENDPOINT_UDP)
{
if(!(conversation->options & CONVERSATION_TEMPLATE))
{
conversation_set_port2(conversation, port_b);
}
else
{
conversation =
conversation_create_from_template(conversation, 0, port_b);
}
}
return conversation;
}
/*
* Well, that didn't find anything.
* If search port B was specified, try looking for a
* conversation with the specified address B and port B
* as the first address and port, and with the specified
* address A and any port as the second address and port
* (this packet may be going in the opposite direction
* from the first packet in the conversation).
* ("port_a" doesn't take part in this lookup.)
*/
if (!(options & NO_PORT_B)) {
DPRINT(("trying dest addr:port as source addr:port and wildcarded dest port"));
conversation =
conversation_lookup_hashtable(conversation_hashtable_no_port2,
frame_num, addr_b, addr_a, etype, port_b, port_a);
if (conversation != NULL) {
/*
* If this is for a connection-oriented
* protocol, set the second port for
* this conversation to port A, as
* that's the address that matched the
* wildcarded second address for this
* conversation.
*/
DPRINT(("match found"));
if (etype != ENDPOINT_UDP)
{
if(!(conversation->options & CONVERSATION_TEMPLATE))
{
conversation_set_port2(conversation, port_a);
}
else
{
conversation =
conversation_create_from_template(conversation, 0, port_a);
}
}
return conversation;
}
}
}
/*
* Well, that didn't find anything. Try matches that wildcard
* one address/port pair.
*
* First try looking for a conversation with the specified address A
* and port A as the first address and port.
* (Neither "addr_b" nor "port_b" take part in this lookup.)
*/
DPRINT(("trying wildcarding dest addr:port"));
conversation =
conversation_lookup_hashtable(conversation_hashtable_no_addr2_or_port2,
frame_num, addr_a, addr_b, etype, port_a, port_b);
if (conversation != NULL) {
/*
* If this is for a connection-oriented protocol:
*
* if search address B isn't wildcarded, set the
* second address for this conversation to address
* B, as that's the address that matched the
* wildcarded second address for this conversation;
*
* if search port B isn't wildcarded, set the
* second port for this conversation to port B,
* as that's the port that matched the wildcarded
* second port for this conversation.
*/
DPRINT(("match found"));
if (etype != ENDPOINT_UDP)
{
if(!(conversation->options & CONVERSATION_TEMPLATE))
{
if (!(conversation->options & NO_ADDR_B))
conversation_set_addr2(conversation, addr_b);
if (!(conversation->options & NO_PORT_B))
conversation_set_port2(conversation, port_b);
}
else
{
conversation =
conversation_create_from_template(conversation, addr_b, port_b);
}
}
return conversation;
}
2016-11-29 03:40:55 +00:00
/* for Infiniband, don't try to look in addresses of reverse
* direction, because it could be another different
* valid conversation than what is being searched using
* addr_a, port_a.
*/
if (etype != ENDPOINT_IBQP)
2016-11-29 03:40:55 +00:00
{
/*
2016-11-29 03:40:55 +00:00
* Well, that didn't find anything.
* If search address and port B were specified, try looking for a
* conversation with the specified address B and port B as the
* first address and port, and with any second address and port
* (this packet may be going in the opposite direction from the
* first packet in the conversation).
* (Neither "addr_a" nor "port_a" take part in this lookup.)
*/
2016-11-29 03:40:55 +00:00
DPRINT(("trying dest addr:port as source addr:port and wildcarding dest addr:port"));
if (addr_a->type == AT_FC)
conversation =
conversation_lookup_hashtable(conversation_hashtable_no_addr2_or_port2,
frame_num, addr_b, addr_a, etype, port_a, port_b);
2016-11-29 03:40:55 +00:00
else
conversation =
conversation_lookup_hashtable(conversation_hashtable_no_addr2_or_port2,
frame_num, addr_b, addr_a, etype, port_b, port_a);
2016-11-29 03:40:55 +00:00
if (conversation != NULL) {
/*
* If this is for a connection-oriented protocol, set the
* second address for this conversation to address A, as
* that's the address that matched the wildcarded second
* address for this conversation, and set the second port
* for this conversation to port A, as that's the port
* that matched the wildcarded second port for this
* conversation.
*/
DPRINT(("match found"));
if (etype != ENDPOINT_UDP)
{
2016-11-29 03:40:55 +00:00
if(!(conversation->options & CONVERSATION_TEMPLATE))
{
conversation_set_addr2(conversation, addr_a);
conversation_set_port2(conversation, port_a);
}
else
{
conversation = conversation_create_from_template(conversation, addr_a, port_a);
}
}
2016-11-29 03:40:55 +00:00
return conversation;
}
}
DPRINT(("no matches found"));
/*
* We found no conversation.
*/
return NULL;
Generalize the "ip_src" and "ip_dst" members of the "packet_info" structure to "dl_src"/"dl_dst", "net_src"/"net_dst", and "src"/"dst" addresses, where an address is an address type, an address length in bytes, and a pointer to that many bytes. "dl_{src,dst}" are the link-layer source/destination; "net_{src,dst}" are the network-layer source/destination; "{src,dst}" are the source/destination from the highest of those two layers that we have in the packet. Add a port type to "packet_info" as well, specifying whether it's a TCP or UDP port. Don't set the address and port columns in the dissector functions; just set the address and port members of the "packet_info" structure. Set the columns in "fill_in_columns()"; this means that if we're showing COL_{DEF,RES,UNRES}_SRC" or "COL_{DEF,RES,UNRES}_DST", we only generate the string from "src" or "dst", we don't generate a string for the link-layer address and then overwrite it with a string for the network-layer address (generating those strings costs CPU). Add support for "conversations", where a "conversation" is (at present) a source and destination address and a source and destination port. (In the future, we may support "conversations" above the transport layer, e.g. a TFTP conversation, where the first packet goes from the client to the TFTP server port, but the reply comes back from a different port, and all subsequent packets go between the client address/port and the server address/new port, or an NFS conversation, which might include lock manager, status monitor, and mount packets, as well as NFS packets.) Currently, all we support is a call that takes the source and destination address/port pairs, looks them up in a hash table, and: if nothing is found, creates a new entry in the hash table, and assigns it a unique 32-bit conversation ID, and returns that conversation ID; if an entry is found, returns its conversation ID. Use that in the SMB and AFS code to keep track of individual SMB or AFS conversations. We need to match up requests and replies, as, for certain replies, the operation code for the request to which it's a reply doesn't show up in the reply - you have to find the request with a matching transaction ID. Transaction IDs are per-conversation, so the hash table for requests should include a conversation ID and transaction ID as the key. This allows SMB and AFS decoders to handle IPv4 or IPv6 addresses transparently (and should allow the SMB decoder to handle NetBIOS atop other protocols as well, if the source and destination address and port values in the "packet_info" structure are set appropriately). In the "Follow TCP Connection" code, check to make sure that the addresses are IPv4 addressses; ultimately, that code should be changed to use the conversation code instead, which will let it handle IPv6 transparently. svn path=/trunk/; revision=909
1999-10-22 07:18:23 +00:00
}
void
conversation_add_proto_data(conversation_t *conv, const int proto, void *proto_data)
{
/* Add it to the list of items for this conversation. */
if (conv->data_list == NULL)
conv->data_list = wmem_tree_new(wmem_file_scope());
wmem_tree_insert32(conv->data_list, proto, proto_data);
}
void *
conversation_get_proto_data(const conversation_t *conv, const int proto)
{
/* No tree created yet */
if (conv->data_list == NULL)
return NULL;
return wmem_tree_lookup32(conv->data_list, proto);
}
void
conversation_delete_proto_data(conversation_t *conv, const int proto)
{
if (conv->data_list != NULL)
wmem_tree_remove32(conv->data_list, proto);
}
void
conversation_set_dissector_from_frame_number(conversation_t *conversation,
const guint32 starting_frame_num, const dissector_handle_t handle)
{
wmem_tree_insert32(conversation->dissector_tree, starting_frame_num, (void *)handle);
}
void
conversation_set_dissector(conversation_t *conversation, const dissector_handle_t handle)
{
conversation_set_dissector_from_frame_number(conversation, 0, handle);
}
dissector_handle_t
conversation_get_dissector(conversation_t *conversation, const guint32 frame_num)
{
return (dissector_handle_t)wmem_tree_lookup32_le(conversation->dissector_tree, frame_num);
}
/*
* Given two address/port pairs for a packet, search for a matching
* conversation and, if found and it has a conversation dissector,
* call that dissector and return TRUE, otherwise return FALSE.
*
* This helper uses call_dissector_only which will NOT call the default
* "data" dissector if the packet was rejected.
* Our caller is responsible to call the data dissector explicitly in case
* this function returns FALSE.
*/
Allow either old-style (pre-tvbuff) or new-style (tvbuffified) dissectors to be registered as dissectors for particular ports, registered as heuristic dissectors, and registered as dissectors for conversations, and have routines to be used both by old-style and new-style dissectors to call registered dissectors. Have the code that calls those dissectors translate the arguments as necessary. (For conversation dissectors, replace "find_conversation_dissector()", which just returns a pointer to the dissector, with "old_try_conversation_dissector()" and "try_conversation_dissector()", which actually call the dissector, so that there's a single place at which we can do that translation. Also make "dissector_lookup()" static and, instead of calling it and, if it returns a non-null pointer, calling that dissector, just use "old_dissector_try_port()" or "dissector_try_port()", for the same reason.) This allows some dissectors that took old-style arguments and immediately translated them to new-style arguments to just take new-style arguments; make them do so. It also allows some new-style dissectors not to have to translate arguments before calling routines to look up and call dissectors; make them not do so. Get rid of checks for too-short frames in new-style dissectors - the tvbuff code does those checks for you. Give the routines to register old-style dissectors, and to call dissectors from old-style dissectors, names beginning with "old_", with the routines for new-style dissectors not having the "old_". Update the dissectors that use those routines appropriately. Rename "dissect_data()" to "old_dissect_data()", and "dissect_data_tvb()" to "dissect_data()". svn path=/trunk/; revision=2218
2000-08-07 03:21:25 +00:00
gboolean
try_conversation_dissector(const address *addr_a, const address *addr_b, const endpoint_type etype,
const guint32 port_a, const guint32 port_b, tvbuff_t *tvb, packet_info *pinfo,
proto_tree *tree, void* data)
Allow either old-style (pre-tvbuff) or new-style (tvbuffified) dissectors to be registered as dissectors for particular ports, registered as heuristic dissectors, and registered as dissectors for conversations, and have routines to be used both by old-style and new-style dissectors to call registered dissectors. Have the code that calls those dissectors translate the arguments as necessary. (For conversation dissectors, replace "find_conversation_dissector()", which just returns a pointer to the dissector, with "old_try_conversation_dissector()" and "try_conversation_dissector()", which actually call the dissector, so that there's a single place at which we can do that translation. Also make "dissector_lookup()" static and, instead of calling it and, if it returns a non-null pointer, calling that dissector, just use "old_dissector_try_port()" or "dissector_try_port()", for the same reason.) This allows some dissectors that took old-style arguments and immediately translated them to new-style arguments to just take new-style arguments; make them do so. It also allows some new-style dissectors not to have to translate arguments before calling routines to look up and call dissectors; make them not do so. Get rid of checks for too-short frames in new-style dissectors - the tvbuff code does those checks for you. Give the routines to register old-style dissectors, and to call dissectors from old-style dissectors, names beginning with "old_", with the routines for new-style dissectors not having the "old_". Update the dissectors that use those routines appropriately. Rename "dissect_data()" to "old_dissect_data()", and "dissect_data_tvb()" to "dissect_data()". svn path=/trunk/; revision=2218
2000-08-07 03:21:25 +00:00
{
conversation_t *conversation;
conversation = find_conversation(pinfo->num, addr_a, addr_b, etype, port_a,
port_b, 0);
Allow either old-style (pre-tvbuff) or new-style (tvbuffified) dissectors to be registered as dissectors for particular ports, registered as heuristic dissectors, and registered as dissectors for conversations, and have routines to be used both by old-style and new-style dissectors to call registered dissectors. Have the code that calls those dissectors translate the arguments as necessary. (For conversation dissectors, replace "find_conversation_dissector()", which just returns a pointer to the dissector, with "old_try_conversation_dissector()" and "try_conversation_dissector()", which actually call the dissector, so that there's a single place at which we can do that translation. Also make "dissector_lookup()" static and, instead of calling it and, if it returns a non-null pointer, calling that dissector, just use "old_dissector_try_port()" or "dissector_try_port()", for the same reason.) This allows some dissectors that took old-style arguments and immediately translated them to new-style arguments to just take new-style arguments; make them do so. It also allows some new-style dissectors not to have to translate arguments before calling routines to look up and call dissectors; make them not do so. Get rid of checks for too-short frames in new-style dissectors - the tvbuff code does those checks for you. Give the routines to register old-style dissectors, and to call dissectors from old-style dissectors, names beginning with "old_", with the routines for new-style dissectors not having the "old_". Update the dissectors that use those routines appropriately. Rename "dissect_data()" to "old_dissect_data()", and "dissect_data_tvb()" to "dissect_data()". svn path=/trunk/; revision=2218
2000-08-07 03:21:25 +00:00
if (conversation != NULL) {
int ret;
dissector_handle_t handle = (dissector_handle_t)wmem_tree_lookup32_le(conversation->dissector_tree, pinfo->num);
if (handle == NULL)
return FALSE;
ret=call_dissector_only(handle, tvb, pinfo, tree, data);
if(!ret) {
/* this packet was rejected by the dissector
* so return FALSE in case our caller wants
* to do some cleaning up.
*/
return FALSE;
}
return TRUE;
Allow either old-style (pre-tvbuff) or new-style (tvbuffified) dissectors to be registered as dissectors for particular ports, registered as heuristic dissectors, and registered as dissectors for conversations, and have routines to be used both by old-style and new-style dissectors to call registered dissectors. Have the code that calls those dissectors translate the arguments as necessary. (For conversation dissectors, replace "find_conversation_dissector()", which just returns a pointer to the dissector, with "old_try_conversation_dissector()" and "try_conversation_dissector()", which actually call the dissector, so that there's a single place at which we can do that translation. Also make "dissector_lookup()" static and, instead of calling it and, if it returns a non-null pointer, calling that dissector, just use "old_dissector_try_port()" or "dissector_try_port()", for the same reason.) This allows some dissectors that took old-style arguments and immediately translated them to new-style arguments to just take new-style arguments; make them do so. It also allows some new-style dissectors not to have to translate arguments before calling routines to look up and call dissectors; make them not do so. Get rid of checks for too-short frames in new-style dissectors - the tvbuff code does those checks for you. Give the routines to register old-style dissectors, and to call dissectors from old-style dissectors, names beginning with "old_", with the routines for new-style dissectors not having the "old_". Update the dissectors that use those routines appropriately. Rename "dissect_data()" to "old_dissect_data()", and "dissect_data_tvb()" to "dissect_data()". svn path=/trunk/; revision=2218
2000-08-07 03:21:25 +00:00
}
return FALSE;
}
/** A helper function that calls find_conversation() using data from pinfo
* The frame number and addresses are taken from pinfo.
*/
conversation_t *
find_conversation_pinfo(packet_info *pinfo, const guint options)
{
conversation_t *conv=NULL;
DPRINT(("called for frame #%d: %s:%d -> %s:%d (ptype=%d)",
pinfo->num, address_to_str(wmem_packet_scope(), &pinfo->src), pinfo->srcport,
address_to_str(wmem_packet_scope(), &pinfo->dst), pinfo->destport, pinfo->ptype));
DINDENT();
/* Have we seen this conversation before? */
if (pinfo->use_endpoint) {
DISSECTOR_ASSERT(pinfo->conv_endpoint);
if((conv = find_conversation(pinfo->num, &pinfo->conv_endpoint->addr1, &pinfo->conv_endpoint->addr2,
pinfo->conv_endpoint->etype, pinfo->conv_endpoint->port1,
pinfo->conv_endpoint->port2, options)) != NULL) {
DPRINT(("found previous conversation for frame #%d (last_frame=%d)",
pinfo->num, conv->last_frame));
if (pinfo->num > conv->last_frame) {
conv->last_frame = pinfo->num;
}
}
} else {
if((conv = find_conversation(pinfo->num, &pinfo->src, &pinfo->dst,
conversation_pt_to_endpoint_type(pinfo->ptype), pinfo->srcport,
pinfo->destport, options)) != NULL) {
DPRINT(("found previous conversation for frame #%d (last_frame=%d)",
pinfo->num, conv->last_frame));
if (pinfo->num > conv->last_frame) {
conv->last_frame = pinfo->num;
}
}
}
return conv;
}
/* A helper function that calls find_conversation() and, if a conversation is
* not found, calls conversation_new().
* The frame number and addresses are taken from pinfo.
* No options are used, though we could extend this API to include an options
* parameter.
*/
conversation_t *
find_or_create_conversation(packet_info *pinfo)
{
conversation_t *conv=NULL;
/* Have we seen this conversation before? */
if((conv = find_conversation_pinfo(pinfo, 0)) == NULL) {
/* No, this is a new conversation. */
DPRINT(("did not find previous conversation for frame #%d",
pinfo->num));
DINDENT();
conv = conversation_new(pinfo->num, &pinfo->src,
&pinfo->dst, conversation_pt_to_endpoint_type(pinfo->ptype),
pinfo->srcport, pinfo->destport, 0);
DENDENT();
}
DENDENT();
return conv;
}
void conversation_create_endpoint(struct _packet_info *pinfo, address* addr1, address* addr2,
endpoint_type etype, guint32 port1, guint32 port2, const guint options)
{
pinfo->conv_endpoint = wmem_new0(pinfo->pool, struct endpoint);
pinfo->use_endpoint = TRUE;
if (addr1 != NULL)
copy_address_wmem(pinfo->pool, &pinfo->conv_endpoint->addr1, addr1);
if (addr2 != NULL)
copy_address_wmem(pinfo->pool, &pinfo->conv_endpoint->addr2, addr2);
pinfo->conv_endpoint->etype = etype;
pinfo->conv_endpoint->port1 = port1;
pinfo->conv_endpoint->port2 = port2;
pinfo->conv_endpoint->options = options;
}
wmem_map_t *
get_conversation_hashtable_exact(void)
{
return conversation_hashtable_exact;
}
wmem_map_t *
get_conversation_hashtable_no_addr2(void)
{
return conversation_hashtable_no_addr2;
}
wmem_map_t *
get_conversation_hashtable_no_port2(void)
{
return conversation_hashtable_no_port2;
}
wmem_map_t *
get_conversation_hashtable_no_addr2_or_port2(void)
{
return conversation_hashtable_no_addr2_or_port2;
}
address*
conversation_key_addr1(const conversation_key_t key)
{
return &key->addr1;
}
address*
conversation_key_addr2(const conversation_key_t key)
{
return &key->addr2;
}
guint32
conversation_key_port1(const conversation_key_t key)
{
return key->port1;
}
guint32
conversation_key_port2(const conversation_key_t key)
{
return key->port2;
}
WS_DLL_PUBLIC
endpoint_type conversation_pt_to_endpoint_type(port_type pt)
{
switch (pt)
{
case PT_NONE:
return ENDPOINT_NONE;
case PT_SCTP:
return ENDPOINT_SCTP;
case PT_TCP:
return ENDPOINT_TCP;
case PT_UDP:
return ENDPOINT_UDP;
case PT_DCCP:
return ENDPOINT_DCCP;
case PT_IPX:
return ENDPOINT_IPX;
case PT_DDP:
return ENDPOINT_DDP;
case PT_IDP:
return ENDPOINT_IDP;
case PT_USB:
return ENDPOINT_USB;
case PT_I2C:
return ENDPOINT_I2C;
case PT_IBQP:
return ENDPOINT_IBQP;
case PT_BLUETOOTH:
return ENDPOINT_BLUETOOTH;
}
DISSECTOR_ASSERT(FALSE);
return ENDPOINT_NONE;
}
gchar*
conversation_get_html_hash(const conversation_key_t key)
{
gchar *hash, *addr1, *addr2;
addr1 = address_to_str(NULL, &key->addr1);
addr2 = address_to_str(NULL, &key->addr2);
hash = wmem_strdup_printf(NULL, "<tr><td>%s</td><td>%d</td><td>%s</td><td>%d</td></tr>\n",
addr1, key->port1, addr2, key->port2);
wmem_free(NULL, addr1);
wmem_free(NULL, addr2);
return hash;
}
/*
* Editor modelines - http://www.wireshark.org/tools/modelines.html
*
* Local variables:
* c-basic-offset: 8
* tab-width: 8
* indent-tabs-mode: t
* End:
*
* vi: set shiftwidth=8 tabstop=8 noexpandtab:
* :indentSize=8:tabSize=8:noTabs=false:
*/