osmith
/
wireshark
forked from osmocom/wireshark
1
0
Fork 0
Code Pull Requests Releases Wiki Activity
wireshark/wiretap/libpcap.c

464 lines
14 KiB
C
Raw Normal View History

Add support to wiretap for reading Sun "snoop" capture files. That requires that, in the packet-reading loop, we pass to the callback routine the offset in the file of a packet's data, because we can no longer compute that offset by subtracting the size of the captured packet data from the offset in the file after the data was read - "snoop" may stick padding in after the packet data to align packet headers on 4-byte boundaries. Doing that required that we arrange that we do that for "libpcap" capture files as well; the cleanest way to do that was to write our own code for reading "libpcap" capture files, rather than using the "libpcap" code to do it. Make "wtap_dispatch_cb()" and "pcap_dispatch_cb()" static to "file.c", as they're not used elsewhere. If we're using wiretap, don't define in "file.h" stuff used only when we're not using wiretap. Update the wiretap README to reflect Gilbert's and my recent changes. Clean up some memory leaks in "wiretap/lanalyzer.c" and "wiretap/ngsniffer.c", where the capture-file-format-specific data wasn't freed if the open failed. svn path=/trunk/; revision=91
1998-11-15 05:29:17 +00:00
/* libpcap.c
*
Add support for capture files written by programs linked with Alexey Kuznetsov's modified "libpcap" *as long as you have the ss990915 or later patch*; the 990417 patch, alas, changes the per-packet header but *doesn't* change the magic number, so you can't just look at the magic number to see that it's Not Standard Libpcap. (Even more unfortunately, Red Hat appears to have picked up *that* patch for Red Hat 6.1; I've filed bug 6773 with Bugzilla on their site - hopefully, if I'm not misremembering the RH 6.1 code I've seen, and they really *did* pick up the older patch, they'll fix it ASAP to use the new magic number, and will make updates available.) svn path=/trunk/; revision=986
1999-11-06 08:42:01 +00:00
* $Id: libpcap.c,v 1.22 1999/11/06 08:42:00 guy Exp $
Add support to wiretap for reading Sun "snoop" capture files. That requires that, in the packet-reading loop, we pass to the callback routine the offset in the file of a packet's data, because we can no longer compute that offset by subtracting the size of the captured packet data from the offset in the file after the data was read - "snoop" may stick padding in after the packet data to align packet headers on 4-byte boundaries. Doing that required that we arrange that we do that for "libpcap" capture files as well; the cleanest way to do that was to write our own code for reading "libpcap" capture files, rather than using the "libpcap" code to do it. Make "wtap_dispatch_cb()" and "pcap_dispatch_cb()" static to "file.c", as they're not used elsewhere. If we're using wiretap, don't define in "file.h" stuff used only when we're not using wiretap. Update the wiretap README to reflect Gilbert's and my recent changes. Clean up some memory leaks in "wiretap/lanalyzer.c" and "wiretap/ngsniffer.c", where the capture-file-format-specific data wasn't freed if the open failed. svn path=/trunk/; revision=91
1998-11-15 05:29:17 +00:00
*
* Wiretap Library
* Copyright (c) 1998 by Gilbert Ramirez <gram@verdict.uthscsa.edu>
*
* 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.
*
* 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.
*
* 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., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.
*
*/
Added support for compiling on win32 with Visual C and 'nmake'. It compiles, but does not link. Perhaps someone who understands the MS tools can help out. I made it link a few months ago, but with different version of glib/gtk+. I can't remember how I made it link. Most of the compatibility issues were resolved with adding #ifdef HAVE_UNISTD_H the the source code. Please be sure to add this to all future code. svn path=/trunk/; revision=359
1999-07-13 02:53:26 +00:00
#ifdef HAVE_CONFIG_H
#include "config.h"
#endif
Add to Wiretap the ability to write capture files; for now, it can only write them in "libpcap" format, but the mechanism can have other formats added. When creating the temporary file for a capture, use "create_tempfile()", to close a security hole opened by the fact that "tempnam()" creates a temporary file, but doesn't open it, and we open the file with the name it gives us - somebody could remove the file and plant a link to some file, and, if as may well be the case when Ethereal is capturing packets, it's running as "root", that means we write a capture on top of that file.... (The aforementioned changes to Wiretap let you open a capture file for writing given an file descriptor, "fdopen()"-style, which this change requires.) svn path=/trunk/; revision=509
1999-08-18 04:17:38 +00:00
#include <stdlib.h>
#include <errno.h>
Add support to wiretap for reading Sun "snoop" capture files. That requires that, in the packet-reading loop, we pass to the callback routine the offset in the file of a packet's data, because we can no longer compute that offset by subtracting the size of the captured packet data from the offset in the file after the data was read - "snoop" may stick padding in after the packet data to align packet headers on 4-byte boundaries. Doing that required that we arrange that we do that for "libpcap" capture files as well; the cleanest way to do that was to write our own code for reading "libpcap" capture files, rather than using the "libpcap" code to do it. Make "wtap_dispatch_cb()" and "pcap_dispatch_cb()" static to "file.c", as they're not used elsewhere. If we're using wiretap, don't define in "file.h" stuff used only when we're not using wiretap. Update the wiretap README to reflect Gilbert's and my recent changes. Clean up some memory leaks in "wiretap/lanalyzer.c" and "wiretap/ngsniffer.c", where the capture-file-format-specific data wasn't freed if the open failed. svn path=/trunk/; revision=91
1998-11-15 05:29:17 +00:00
#include "wtap.h"
Fix it so that it builds with "--disable-zlib". The "fh" member of a "wtap" structure points to something constructed from the "fd" member of that structure, so that closing the stream referred to by "fh" also closes the underlying file descriptor; get rid of an unnecessary close of "wth->fd". svn path=/trunk/; revision=720
1999-09-24 05:49:53 +00:00
#include "file.h"
Added display filters to wiretap. svn path=/trunk/; revision=198
1999-03-01 18:57:07 +00:00
#include "buffer.h"
Add support to wiretap for reading Sun "snoop" capture files. That requires that, in the packet-reading loop, we pass to the callback routine the offset in the file of a packet's data, because we can no longer compute that offset by subtracting the size of the captured packet data from the offset in the file after the data was read - "snoop" may stick padding in after the packet data to align packet headers on 4-byte boundaries. Doing that required that we arrange that we do that for "libpcap" capture files as well; the cleanest way to do that was to write our own code for reading "libpcap" capture files, rather than using the "libpcap" code to do it. Make "wtap_dispatch_cb()" and "pcap_dispatch_cb()" static to "file.c", as they're not used elsewhere. If we're using wiretap, don't define in "file.h" stuff used only when we're not using wiretap. Update the wiretap README to reflect Gilbert's and my recent changes. Clean up some memory leaks in "wiretap/lanalyzer.c" and "wiretap/ngsniffer.c", where the capture-file-format-specific data wasn't freed if the open failed. svn path=/trunk/; revision=91
1998-11-15 05:29:17 +00:00
#include "libpcap.h"
/* See source to the "libpcap" library for information on the "libpcap"
file format. */
/* Magic numbers in "libpcap" files.
"libpcap" file records are written in the byte order of the host that
writes them, and the reader is expected to fix this up.
PCAP_MAGIC is the magic number, in host byte order; PCAP_SWAPPED_MAGIC
Add support for capture files written by programs linked with Alexey Kuznetsov's modified "libpcap" *as long as you have the ss990915 or later patch*; the 990417 patch, alas, changes the per-packet header but *doesn't* change the magic number, so you can't just look at the magic number to see that it's Not Standard Libpcap. (Even more unfortunately, Red Hat appears to have picked up *that* patch for Red Hat 6.1; I've filed bug 6773 with Bugzilla on their site - hopefully, if I'm not misremembering the RH 6.1 code I've seen, and they really *did* pick up the older patch, they'll fix it ASAP to use the new magic number, and will make updates available.) svn path=/trunk/; revision=986
1999-11-06 08:42:01 +00:00
is a byte-swapped version of that.
PCAP_MUTANT_MAGIC is for Alexey Kuznetsov's modified "libpcap"
format, as generated on Linux systems that have a "libpcap" with
his patches, at
http://ftp.sunet.se/pub/os/Linux/ip-routing/lbl-tools/
applied; PCAP_SWAPPED_MUTANT_MAGIC is the byte-swapped version. */
#define PCAP_MAGIC 0xa1b2c3d4
#define PCAP_SWAPPED_MAGIC 0xd4c3b2a1
#define PCAP_MUTANT_MAGIC 0xa1b2cd34
#define PCAP_SWAPPED_MUTANT_MAGIC 0x34cdb2a1
Add support to wiretap for reading Sun "snoop" capture files. That requires that, in the packet-reading loop, we pass to the callback routine the offset in the file of a packet's data, because we can no longer compute that offset by subtracting the size of the captured packet data from the offset in the file after the data was read - "snoop" may stick padding in after the packet data to align packet headers on 4-byte boundaries. Doing that required that we arrange that we do that for "libpcap" capture files as well; the cleanest way to do that was to write our own code for reading "libpcap" capture files, rather than using the "libpcap" code to do it. Make "wtap_dispatch_cb()" and "pcap_dispatch_cb()" static to "file.c", as they're not used elsewhere. If we're using wiretap, don't define in "file.h" stuff used only when we're not using wiretap. Update the wiretap README to reflect Gilbert's and my recent changes. Clean up some memory leaks in "wiretap/lanalyzer.c" and "wiretap/ngsniffer.c", where the capture-file-format-specific data wasn't freed if the open failed. svn path=/trunk/; revision=91
1998-11-15 05:29:17 +00:00
/* Macros to byte-swap 32-bit and 16-bit quantities. */
#define BSWAP32(x) \
((((x)&0xFF000000)>>24) | \
(((x)&0x00FF0000)>>8) | \
(((x)&0x0000FF00)<<8) | \
(((x)&0x000000FF)<<24))
#define BSWAP16(x) \
((((x)&0xFF00)>>8) | \
(((x)&0x00FF)<<8))
Add a new Wiretap encapsulation type WTAP_ENCAP_FDDI_BITSWAPPED, meaning "FDDI with the MAC addresses bit-swapped"; whether the MAC addresses are bit-swapped is a property of the machine on which the capture was taken, not of the machine on which the capture is being read - right now, none of the capture file formats we read indicate whether FDDI MAC addresses are bit-swapped, but this does let us treat non-"libpcap" captures as being bit-swapped or not bit-swapped independent of the machine on which they're being read (and of the machine on which they were captured, but I have the impression they're bit-swapped on most platforms), and allows us to, if, as, and when we implement packet capture in Wiretap, mark packets in a capture file written in Wiretap-native format based on the machine on which they are captured (assuming the rule "Ultrix, Alpha, and BSD/OS are the only platforms that don't bit-swap", or some other compile-time rule, gets the right answer, or that some platform has drivers that can tell us whether the addresses are bit-swapped). (NOTE: if, for any of the capture file formats used only on one platform, FDDI MAC addresses aren't bit-swapped, the code to read that capture file format should be fixed to flag them as not bit-swapped.) Use the encapsulation type to decide whether to bit-swap addresses in "dissect_fddi()". svn path=/trunk/; revision=557
1999-08-24 03:19:34 +00:00
/* On some systems, the FDDI MAC addresses are bit-swapped. */
Fix, from Bert Driehuis, to the stuff that checks for BSD/OS. svn path=/trunk/; revision=620
1999-08-31 22:36:20 +00:00
#if !defined(ultrix) && !defined(__alpha) && !defined(__bsdi__)
Add a new Wiretap encapsulation type WTAP_ENCAP_FDDI_BITSWAPPED, meaning "FDDI with the MAC addresses bit-swapped"; whether the MAC addresses are bit-swapped is a property of the machine on which the capture was taken, not of the machine on which the capture is being read - right now, none of the capture file formats we read indicate whether FDDI MAC addresses are bit-swapped, but this does let us treat non-"libpcap" captures as being bit-swapped or not bit-swapped independent of the machine on which they're being read (and of the machine on which they were captured, but I have the impression they're bit-swapped on most platforms), and allows us to, if, as, and when we implement packet capture in Wiretap, mark packets in a capture file written in Wiretap-native format based on the machine on which they are captured (assuming the rule "Ultrix, Alpha, and BSD/OS are the only platforms that don't bit-swap", or some other compile-time rule, gets the right answer, or that some platform has drivers that can tell us whether the addresses are bit-swapped). (NOTE: if, for any of the capture file formats used only on one platform, FDDI MAC addresses aren't bit-swapped, the code to read that capture file format should be fixed to flag them as not bit-swapped.) Use the encapsulation type to decide whether to bit-swap addresses in "dissect_fddi()". svn path=/trunk/; revision=557
1999-08-24 03:19:34 +00:00
#define BIT_SWAPPED_MAC_ADDRS
#endif
Add support to wiretap for reading Sun "snoop" capture files. That requires that, in the packet-reading loop, we pass to the callback routine the offset in the file of a packet's data, because we can no longer compute that offset by subtracting the size of the captured packet data from the offset in the file after the data was read - "snoop" may stick padding in after the packet data to align packet headers on 4-byte boundaries. Doing that required that we arrange that we do that for "libpcap" capture files as well; the cleanest way to do that was to write our own code for reading "libpcap" capture files, rather than using the "libpcap" code to do it. Make "wtap_dispatch_cb()" and "pcap_dispatch_cb()" static to "file.c", as they're not used elsewhere. If we're using wiretap, don't define in "file.h" stuff used only when we're not using wiretap. Update the wiretap README to reflect Gilbert's and my recent changes. Clean up some memory leaks in "wiretap/lanalyzer.c" and "wiretap/ngsniffer.c", where the capture-file-format-specific data wasn't freed if the open failed. svn path=/trunk/; revision=91
1998-11-15 05:29:17 +00:00
/* "libpcap" file header (minus magic number). */
struct pcap_hdr {
guint16 version_major; /* major version number */
guint16 version_minor; /* minor version number */
gint32 thiszone; /* GMT to local correction */
guint32 sigfigs; /* accuracy of timestamps */
guint32 snaplen; /* max length of captured packets, in octets */
guint32 network; /* data link type */
};
/* "libpcap" record header. */
struct pcaprec_hdr {
guint32 ts_sec; /* timestamp seconds */
guint32 ts_usec; /* timestamp microseconds */
Add to Wiretap the ability to write capture files; for now, it can only write them in "libpcap" format, but the mechanism can have other formats added. When creating the temporary file for a capture, use "create_tempfile()", to close a security hole opened by the fact that "tempnam()" creates a temporary file, but doesn't open it, and we open the file with the name it gives us - somebody could remove the file and plant a link to some file, and, if as may well be the case when Ethereal is capturing packets, it's running as "root", that means we write a capture on top of that file.... (The aforementioned changes to Wiretap let you open a capture file for writing given an file descriptor, "fdopen()"-style, which this change requires.) svn path=/trunk/; revision=509
1999-08-18 04:17:38 +00:00
guint32 incl_len; /* number of octets of packet saved in file */
Add support to wiretap for reading Sun "snoop" capture files. That requires that, in the packet-reading loop, we pass to the callback routine the offset in the file of a packet's data, because we can no longer compute that offset by subtracting the size of the captured packet data from the offset in the file after the data was read - "snoop" may stick padding in after the packet data to align packet headers on 4-byte boundaries. Doing that required that we arrange that we do that for "libpcap" capture files as well; the cleanest way to do that was to write our own code for reading "libpcap" capture files, rather than using the "libpcap" code to do it. Make "wtap_dispatch_cb()" and "pcap_dispatch_cb()" static to "file.c", as they're not used elsewhere. If we're using wiretap, don't define in "file.h" stuff used only when we're not using wiretap. Update the wiretap README to reflect Gilbert's and my recent changes. Clean up some memory leaks in "wiretap/lanalyzer.c" and "wiretap/ngsniffer.c", where the capture-file-format-specific data wasn't freed if the open failed. svn path=/trunk/; revision=91
1998-11-15 05:29:17 +00:00
guint32 orig_len; /* actual length of packet */
};
Add support for capture files written by programs linked with Alexey Kuznetsov's modified "libpcap" *as long as you have the ss990915 or later patch*; the 990417 patch, alas, changes the per-packet header but *doesn't* change the magic number, so you can't just look at the magic number to see that it's Not Standard Libpcap. (Even more unfortunately, Red Hat appears to have picked up *that* patch for Red Hat 6.1; I've filed bug 6773 with Bugzilla on their site - hopefully, if I'm not misremembering the RH 6.1 code I've seen, and they really *did* pick up the older patch, they'll fix it ASAP to use the new magic number, and will make updates available.) svn path=/trunk/; revision=986
1999-11-06 08:42:01 +00:00
/* "libpcap" record header for Alexey's patched version. */
struct pcaprec_mutant_hdr {
struct pcaprec_hdr hdr; /* the regular header */
guint32 ifindex; /* index, in *capturing* machine's list of
interfaces, of the interface on which this
packet came in. */
guint16 protocol; /* Ethernet packet type */
guint8 pkt_type; /* broadcast/multicast/etc. indication */
};
Have the per-capture-file-type open routines "wtap_open_offline()" calls return 1 on success, -1 if they got an error, and 0 if the file isn't of the type that file is checking for, and supply an error code if they return -1; have "wtap_open_offline()" use that error code. Also, have the per-capture-file-type open routines treat errors accessing the file as errors, and return -1, rather than just returning 0 so that we try another file type. Have the per-capture-file-type read routines "wtap_loop()" calls return -1 and supply an error code on error (and not, as they did in some cases, call "g_error()" and abort), and have "wtap_loop()", if the read routine returned an error, return FALSE (and pass an error-code-pointer argument onto the read routines, so they fill it in), and return TRUE on success. Add some new error codes for them to return. Now that "wtap_loop()" can return a success/failure indication and an error code, in "read_cap_file()" put up a message box if we get an error reading the file, and return the error code. Handle the additional errors we can get when opening a capture file. If the attempt to open a capture file succeeds, but the attempt to read it fails, don't treat that as a complete failure - we may have managed to read some of the capture file, and we should display what we managed to read. svn path=/trunk/; revision=516
1999-08-19 05:31:38 +00:00
static int libpcap_read(wtap *wth, int *err);
Add to Wiretap the ability to write capture files; for now, it can only write them in "libpcap" format, but the mechanism can have other formats added. When creating the temporary file for a capture, use "create_tempfile()", to close a security hole opened by the fact that "tempnam()" creates a temporary file, but doesn't open it, and we open the file with the name it gives us - somebody could remove the file and plant a link to some file, and, if as may well be the case when Ethereal is capturing packets, it's running as "root", that means we write a capture on top of that file.... (The aforementioned changes to Wiretap let you open a capture file for writing given an file descriptor, "fdopen()"-style, which this change requires.) svn path=/trunk/; revision=509
1999-08-18 04:17:38 +00:00
static int libpcap_dump(wtap_dumper *wdh, const struct wtap_pkthdr *phdr,
Make "wtap_dump()" and "wtap_dump_close()" return error codes, and check for errors when closing a file to which we've written packets (we don't bother checking if we're giving up on a capture). Add some more error checks in Wiretap. Make a single list of all Wiretap error codes, giving them all different values (some can be returned by more than one routine, so they shouldn't be per-routine). svn path=/trunk/; revision=510
1999-08-18 04:41:20 +00:00
const u_char *pd, int *err);
static int libpcap_dump_close(wtap_dumper *wdh, int *err);
Add to Wiretap the ability to write capture files; for now, it can only write them in "libpcap" format, but the mechanism can have other formats added. When creating the temporary file for a capture, use "create_tempfile()", to close a security hole opened by the fact that "tempnam()" creates a temporary file, but doesn't open it, and we open the file with the name it gives us - somebody could remove the file and plant a link to some file, and, if as may well be the case when Ethereal is capturing packets, it's running as "root", that means we write a capture on top of that file.... (The aforementioned changes to Wiretap let you open a capture file for writing given an file descriptor, "fdopen()"-style, which this change requires.) svn path=/trunk/; revision=509
1999-08-18 04:17:38 +00:00
DLT_NULL, from "libpcap", means different things on different platforms and in different capture files; throw in some heuristics to try to figure out whether the 4-byte header is: 1) PPP-over-HDLC (some version of ISDN4BSD?); 2) big-endian AF_ value (BSD on big-endian platforms); 3) little-endian AF_ value (BSD on little-endian platforms); 4) two octets of 0 followed by an Ethernet type (Linux, at least on little-endian platforms, as mutated by "libpcap"). Make a separate Wiretap encapsulation type, WTAP_ENCAP_NULL, corresponding to DLT_NULL. Have the PPP code dissect the frame if it's PPP-over-HDLC, and have "ethertype()" dissect the Ethernet type and the rest of the packet if it's a Linux-style header; dissect it ourselves only if it's an AF_ value. Have Wiretap impose a maximum packet size of 65535 bytes, so that it fails more gracefully when handed a corrupt "libpcap" capture file (other capture file formats with more than a 16-bit capture length field, if any, will have that check added later), and put that size in "wtap.h" and have Ethereal use it as its notion of a maximum packet size. Have Ethereal put up a "this file appears to be damaged or corrupt" message box if Wiretap returns a WTAP_ERR_BAD_RECORD error when opening or reading a capture file. Include loopback interfaces in the list of interfaces offered by the "Capture" dialog box, but put them at the end of the list so that it doesn't default to a loopback interface unless there are no other interfaces. Also, don't require that an interface in the list have an IP address associated with it, and only put one entry in the list for a given interface (SIOCGIFCONF returns one entry per interface *address*, not per *interface* - and even if you were to use only IP addresses, an interface could conceivably have more than one IP address). Exclusively use Wiretap encapsulation types internally, even when capturing; don't use DLT_ types. svn path=/trunk/; revision=540
1999-08-22 00:47:56 +00:00
/*
Modify the comments to show how much of a mess the different DLT_ values on different flavors of BSD are - even worse than I thought, now that I've seen the BSD/OS <net/bpf.h>. svn path=/trunk/; revision=552
1999-08-22 19:08:40 +00:00
* XXX - this is a bit of a mess. OpenBSD, and perhaps NetBSD, and
* BSD/OS have different DLT_ codes from FreeBSD (and from the LBL
* BPF code), and, at least in some cases, from each other.
Get rid of some cruft left in by previous checkins as placeholders. Get rid of WTAP_ENCAP_NONE; replace it with WTAP_ENCAP_UNKNOWN, which means "I can't handle that file, it's using an encapsulation I don't support". Check for encapsulations we don't support, and return an error (as is already done in "libpcap.c"). Check for too-large packet sizes, and return an error (as is already done in "libpcap.c"). Print unsigned quantities in Wiretap messages with "%u", not "%d". svn path=/trunk/; revision=544
1999-08-22 02:29:40 +00:00
* For now, we simply treat those type values with different
* meanings on different platforms, except for DLT_RAW, as "unknown";
* this means you won't be able to capture from a network using those
DLT_NULL, from "libpcap", means different things on different platforms and in different capture files; throw in some heuristics to try to figure out whether the 4-byte header is: 1) PPP-over-HDLC (some version of ISDN4BSD?); 2) big-endian AF_ value (BSD on big-endian platforms); 3) little-endian AF_ value (BSD on little-endian platforms); 4) two octets of 0 followed by an Ethernet type (Linux, at least on little-endian platforms, as mutated by "libpcap"). Make a separate Wiretap encapsulation type, WTAP_ENCAP_NULL, corresponding to DLT_NULL. Have the PPP code dissect the frame if it's PPP-over-HDLC, and have "ethertype()" dissect the Ethernet type and the rest of the packet if it's a Linux-style header; dissect it ourselves only if it's an AF_ value. Have Wiretap impose a maximum packet size of 65535 bytes, so that it fails more gracefully when handed a corrupt "libpcap" capture file (other capture file formats with more than a 16-bit capture length field, if any, will have that check added later), and put that size in "wtap.h" and have Ethereal use it as its notion of a maximum packet size. Have Ethereal put up a "this file appears to be damaged or corrupt" message box if Wiretap returns a WTAP_ERR_BAD_RECORD error when opening or reading a capture file. Include loopback interfaces in the list of interfaces offered by the "Capture" dialog box, but put them at the end of the list so that it doesn't default to a loopback interface unless there are no other interfaces. Also, don't require that an interface in the list have an IP address associated with it, and only put one entry in the list for a given interface (SIOCGIFCONF returns one entry per interface *address*, not per *interface* - and even if you were to use only IP addresses, an interface could conceivably have more than one IP address). Exclusively use Wiretap encapsulation types internally, even when capturing; don't use DLT_ types. svn path=/trunk/; revision=540
1999-08-22 00:47:56 +00:00
* types in Ethereal (and that capturing from the loopback interface
Get rid of some cruft left in by previous checkins as placeholders. Get rid of WTAP_ENCAP_NONE; replace it with WTAP_ENCAP_UNKNOWN, which means "I can't handle that file, it's using an encapsulation I don't support". Check for encapsulations we don't support, and return an error (as is already done in "libpcap.c"). Check for too-large packet sizes, and return an error (as is already done in "libpcap.c"). Print unsigned quantities in Wiretap messages with "%u", not "%d". svn path=/trunk/; revision=544
1999-08-22 02:29:40 +00:00
* won't necessarily work right on OpenBSD, either, as it uses
* DLT_LOOP, which is the same as DLT_RAW on other platforms).
DLT_NULL, from "libpcap", means different things on different platforms and in different capture files; throw in some heuristics to try to figure out whether the 4-byte header is: 1) PPP-over-HDLC (some version of ISDN4BSD?); 2) big-endian AF_ value (BSD on big-endian platforms); 3) little-endian AF_ value (BSD on little-endian platforms); 4) two octets of 0 followed by an Ethernet type (Linux, at least on little-endian platforms, as mutated by "libpcap"). Make a separate Wiretap encapsulation type, WTAP_ENCAP_NULL, corresponding to DLT_NULL. Have the PPP code dissect the frame if it's PPP-over-HDLC, and have "ethertype()" dissect the Ethernet type and the rest of the packet if it's a Linux-style header; dissect it ourselves only if it's an AF_ value. Have Wiretap impose a maximum packet size of 65535 bytes, so that it fails more gracefully when handed a corrupt "libpcap" capture file (other capture file formats with more than a 16-bit capture length field, if any, will have that check added later), and put that size in "wtap.h" and have Ethereal use it as its notion of a maximum packet size. Have Ethereal put up a "this file appears to be damaged or corrupt" message box if Wiretap returns a WTAP_ERR_BAD_RECORD error when opening or reading a capture file. Include loopback interfaces in the list of interfaces offered by the "Capture" dialog box, but put them at the end of the list so that it doesn't default to a loopback interface unless there are no other interfaces. Also, don't require that an interface in the list have an IP address associated with it, and only put one entry in the list for a given interface (SIOCGIFCONF returns one entry per interface *address*, not per *interface* - and even if you were to use only IP addresses, an interface could conceivably have more than one IP address). Exclusively use Wiretap encapsulation types internally, even when capturing; don't use DLT_ types. svn path=/trunk/; revision=540
1999-08-22 00:47:56 +00:00
*
* Does anybody know what BSD/OS uses as DLT_ types for SLIP and
* PPP? The LBL code, and the OpenBSD code, appear to disagree....
*
* Nothing in FreeBSD appears to use DLT_RAW, so it's not clear what
Get rid of some cruft left in by previous checkins as placeholders. Get rid of WTAP_ENCAP_NONE; replace it with WTAP_ENCAP_UNKNOWN, which means "I can't handle that file, it's using an encapsulation I don't support". Check for encapsulations we don't support, and return an error (as is already done in "libpcap.c"). Check for too-large packet sizes, and return an error (as is already done in "libpcap.c"). Print unsigned quantities in Wiretap messages with "%u", not "%d". svn path=/trunk/; revision=544
1999-08-22 02:29:40 +00:00
* link-layer header or fake header appears for DLT_RAW. If it's
* completely unused, or if it behaves the same way OpenBSD DLT_LOOP
* behaves, i.e. it puts an address family in *network* byte order
* (as opposed to the *host* byte order that DLT_NULL uses on FreeBSD),
* then we should just make it WTAP_ENCAP_NULL, which we treat in
* such a fashion as to cause it to work with DLT_LOOP headers.
DLT_NULL, from "libpcap", means different things on different platforms and in different capture files; throw in some heuristics to try to figure out whether the 4-byte header is: 1) PPP-over-HDLC (some version of ISDN4BSD?); 2) big-endian AF_ value (BSD on big-endian platforms); 3) little-endian AF_ value (BSD on little-endian platforms); 4) two octets of 0 followed by an Ethernet type (Linux, at least on little-endian platforms, as mutated by "libpcap"). Make a separate Wiretap encapsulation type, WTAP_ENCAP_NULL, corresponding to DLT_NULL. Have the PPP code dissect the frame if it's PPP-over-HDLC, and have "ethertype()" dissect the Ethernet type and the rest of the packet if it's a Linux-style header; dissect it ourselves only if it's an AF_ value. Have Wiretap impose a maximum packet size of 65535 bytes, so that it fails more gracefully when handed a corrupt "libpcap" capture file (other capture file formats with more than a 16-bit capture length field, if any, will have that check added later), and put that size in "wtap.h" and have Ethereal use it as its notion of a maximum packet size. Have Ethereal put up a "this file appears to be damaged or corrupt" message box if Wiretap returns a WTAP_ERR_BAD_RECORD error when opening or reading a capture file. Include loopback interfaces in the list of interfaces offered by the "Capture" dialog box, but put them at the end of the list so that it doesn't default to a loopback interface unless there are no other interfaces. Also, don't require that an interface in the list have an IP address associated with it, and only put one entry in the list for a given interface (SIOCGIFCONF returns one entry per interface *address*, not per *interface* - and even if you were to use only IP addresses, an interface could conceivably have more than one IP address). Exclusively use Wiretap encapsulation types internally, even when capturing; don't use DLT_ types. svn path=/trunk/; revision=540
1999-08-22 00:47:56 +00:00
*/
Add to Wiretap the ability to write capture files; for now, it can only write them in "libpcap" format, but the mechanism can have other formats added. When creating the temporary file for a capture, use "create_tempfile()", to close a security hole opened by the fact that "tempnam()" creates a temporary file, but doesn't open it, and we open the file with the name it gives us - somebody could remove the file and plant a link to some file, and, if as may well be the case when Ethereal is capturing packets, it's running as "root", that means we write a capture on top of that file.... (The aforementioned changes to Wiretap let you open a capture file for writing given an file descriptor, "fdopen()"-style, which this change requires.) svn path=/trunk/; revision=509
1999-08-18 04:17:38 +00:00
static const int pcap_encap[] = {
DLT_NULL, from "libpcap", means different things on different platforms and in different capture files; throw in some heuristics to try to figure out whether the 4-byte header is: 1) PPP-over-HDLC (some version of ISDN4BSD?); 2) big-endian AF_ value (BSD on big-endian platforms); 3) little-endian AF_ value (BSD on little-endian platforms); 4) two octets of 0 followed by an Ethernet type (Linux, at least on little-endian platforms, as mutated by "libpcap"). Make a separate Wiretap encapsulation type, WTAP_ENCAP_NULL, corresponding to DLT_NULL. Have the PPP code dissect the frame if it's PPP-over-HDLC, and have "ethertype()" dissect the Ethernet type and the rest of the packet if it's a Linux-style header; dissect it ourselves only if it's an AF_ value. Have Wiretap impose a maximum packet size of 65535 bytes, so that it fails more gracefully when handed a corrupt "libpcap" capture file (other capture file formats with more than a 16-bit capture length field, if any, will have that check added later), and put that size in "wtap.h" and have Ethereal use it as its notion of a maximum packet size. Have Ethereal put up a "this file appears to be damaged or corrupt" message box if Wiretap returns a WTAP_ERR_BAD_RECORD error when opening or reading a capture file. Include loopback interfaces in the list of interfaces offered by the "Capture" dialog box, but put them at the end of the list so that it doesn't default to a loopback interface unless there are no other interfaces. Also, don't require that an interface in the list have an IP address associated with it, and only put one entry in the list for a given interface (SIOCGIFCONF returns one entry per interface *address*, not per *interface* - and even if you were to use only IP addresses, an interface could conceivably have more than one IP address). Exclusively use Wiretap encapsulation types internally, even when capturing; don't use DLT_ types. svn path=/trunk/; revision=540
1999-08-22 00:47:56 +00:00
WTAP_ENCAP_NULL, /* null encapsulation */
Add to Wiretap the ability to write capture files; for now, it can only write them in "libpcap" format, but the mechanism can have other formats added. When creating the temporary file for a capture, use "create_tempfile()", to close a security hole opened by the fact that "tempnam()" creates a temporary file, but doesn't open it, and we open the file with the name it gives us - somebody could remove the file and plant a link to some file, and, if as may well be the case when Ethereal is capturing packets, it's running as "root", that means we write a capture on top of that file.... (The aforementioned changes to Wiretap let you open a capture file for writing given an file descriptor, "fdopen()"-style, which this change requires.) svn path=/trunk/; revision=509
1999-08-18 04:17:38 +00:00
WTAP_ENCAP_ETHERNET,
DLT_NULL, from "libpcap", means different things on different platforms and in different capture files; throw in some heuristics to try to figure out whether the 4-byte header is: 1) PPP-over-HDLC (some version of ISDN4BSD?); 2) big-endian AF_ value (BSD on big-endian platforms); 3) little-endian AF_ value (BSD on little-endian platforms); 4) two octets of 0 followed by an Ethernet type (Linux, at least on little-endian platforms, as mutated by "libpcap"). Make a separate Wiretap encapsulation type, WTAP_ENCAP_NULL, corresponding to DLT_NULL. Have the PPP code dissect the frame if it's PPP-over-HDLC, and have "ethertype()" dissect the Ethernet type and the rest of the packet if it's a Linux-style header; dissect it ourselves only if it's an AF_ value. Have Wiretap impose a maximum packet size of 65535 bytes, so that it fails more gracefully when handed a corrupt "libpcap" capture file (other capture file formats with more than a 16-bit capture length field, if any, will have that check added later), and put that size in "wtap.h" and have Ethereal use it as its notion of a maximum packet size. Have Ethereal put up a "this file appears to be damaged or corrupt" message box if Wiretap returns a WTAP_ERR_BAD_RECORD error when opening or reading a capture file. Include loopback interfaces in the list of interfaces offered by the "Capture" dialog box, but put them at the end of the list so that it doesn't default to a loopback interface unless there are no other interfaces. Also, don't require that an interface in the list have an IP address associated with it, and only put one entry in the list for a given interface (SIOCGIFCONF returns one entry per interface *address*, not per *interface* - and even if you were to use only IP addresses, an interface could conceivably have more than one IP address). Exclusively use Wiretap encapsulation types internally, even when capturing; don't use DLT_ types. svn path=/trunk/; revision=540
1999-08-22 00:47:56 +00:00
WTAP_ENCAP_UNKNOWN, /* 3Mb experimental Ethernet */
WTAP_ENCAP_UNKNOWN, /* Amateur Radio AX.25 */
WTAP_ENCAP_UNKNOWN, /* Proteon ProNET Token Ring */
WTAP_ENCAP_UNKNOWN, /* Chaos */
Add to Wiretap the ability to write capture files; for now, it can only write them in "libpcap" format, but the mechanism can have other formats added. When creating the temporary file for a capture, use "create_tempfile()", to close a security hole opened by the fact that "tempnam()" creates a temporary file, but doesn't open it, and we open the file with the name it gives us - somebody could remove the file and plant a link to some file, and, if as may well be the case when Ethereal is capturing packets, it's running as "root", that means we write a capture on top of that file.... (The aforementioned changes to Wiretap let you open a capture file for writing given an file descriptor, "fdopen()"-style, which this change requires.) svn path=/trunk/; revision=509
1999-08-18 04:17:38 +00:00
WTAP_ENCAP_TR, /* IEEE 802 Networks - assume token ring */
WTAP_ENCAP_ARCNET,
WTAP_ENCAP_SLIP,
WTAP_ENCAP_PPP,
Add a new Wiretap encapsulation type WTAP_ENCAP_FDDI_BITSWAPPED, meaning "FDDI with the MAC addresses bit-swapped"; whether the MAC addresses are bit-swapped is a property of the machine on which the capture was taken, not of the machine on which the capture is being read - right now, none of the capture file formats we read indicate whether FDDI MAC addresses are bit-swapped, but this does let us treat non-"libpcap" captures as being bit-swapped or not bit-swapped independent of the machine on which they're being read (and of the machine on which they were captured, but I have the impression they're bit-swapped on most platforms), and allows us to, if, as, and when we implement packet capture in Wiretap, mark packets in a capture file written in Wiretap-native format based on the machine on which they are captured (assuming the rule "Ultrix, Alpha, and BSD/OS are the only platforms that don't bit-swap", or some other compile-time rule, gets the right answer, or that some platform has drivers that can tell us whether the addresses are bit-swapped). (NOTE: if, for any of the capture file formats used only on one platform, FDDI MAC addresses aren't bit-swapped, the code to read that capture file format should be fixed to flag them as not bit-swapped.) Use the encapsulation type to decide whether to bit-swap addresses in "dissect_fddi()". svn path=/trunk/; revision=557
1999-08-24 03:19:34 +00:00
#ifdef BIT_SWAPPED_MAC_ADDRS
WTAP_ENCAP_FDDI_BITSWAPPED,
#else
Add to Wiretap the ability to write capture files; for now, it can only write them in "libpcap" format, but the mechanism can have other formats added. When creating the temporary file for a capture, use "create_tempfile()", to close a security hole opened by the fact that "tempnam()" creates a temporary file, but doesn't open it, and we open the file with the name it gives us - somebody could remove the file and plant a link to some file, and, if as may well be the case when Ethereal is capturing packets, it's running as "root", that means we write a capture on top of that file.... (The aforementioned changes to Wiretap let you open a capture file for writing given an file descriptor, "fdopen()"-style, which this change requires.) svn path=/trunk/; revision=509
1999-08-18 04:17:38 +00:00
WTAP_ENCAP_FDDI,
Add a new Wiretap encapsulation type WTAP_ENCAP_FDDI_BITSWAPPED, meaning "FDDI with the MAC addresses bit-swapped"; whether the MAC addresses are bit-swapped is a property of the machine on which the capture was taken, not of the machine on which the capture is being read - right now, none of the capture file formats we read indicate whether FDDI MAC addresses are bit-swapped, but this does let us treat non-"libpcap" captures as being bit-swapped or not bit-swapped independent of the machine on which they're being read (and of the machine on which they were captured, but I have the impression they're bit-swapped on most platforms), and allows us to, if, as, and when we implement packet capture in Wiretap, mark packets in a capture file written in Wiretap-native format based on the machine on which they are captured (assuming the rule "Ultrix, Alpha, and BSD/OS are the only platforms that don't bit-swap", or some other compile-time rule, gets the right answer, or that some platform has drivers that can tell us whether the addresses are bit-swapped). (NOTE: if, for any of the capture file formats used only on one platform, FDDI MAC addresses aren't bit-swapped, the code to read that capture file format should be fixed to flag them as not bit-swapped.) Use the encapsulation type to decide whether to bit-swap addresses in "dissect_fddi()". svn path=/trunk/; revision=557
1999-08-24 03:19:34 +00:00
#endif
Modify the comments to show how much of a mess the different DLT_ values on different flavors of BSD are - even worse than I thought, now that I've seen the BSD/OS <net/bpf.h>. svn path=/trunk/; revision=552
1999-08-22 19:08:40 +00:00
WTAP_ENCAP_ATM_RFC1483, /* or, on BSD/OS, Frame Relay */
WTAP_ENCAP_RAW_IP, /* or, on OpenBSD, DLT_LOOP, and on BSD/OS,
Cisco HDLC */
WTAP_ENCAP_UNKNOWN, /* In LBL BPF and FreeBSD, BSD/OS SLIP;
on OpenBSD, DLT_ENC; on BSD/OS,
DLT_ATM_RFC1483 */
WTAP_ENCAP_UNKNOWN, /* In LBL BPF and FreeBSD, BSD/OS PPP;
on OpenBSD and BSD/OS, DLT_RAW */
WTAP_ENCAP_UNKNOWN, /* In OpenBSD and BSD/OS, BSD/OS SLIP,
but the BSD/OS header says "internal
to libpcap", whatever that means */
WTAP_ENCAP_UNKNOWN, /* In OpenBSD and BSD/OS, BSD/OS PPP,
but the BSD/OS header says "internal
to libpcap", whatever that means */
DLT_NULL, from "libpcap", means different things on different platforms and in different capture files; throw in some heuristics to try to figure out whether the 4-byte header is: 1) PPP-over-HDLC (some version of ISDN4BSD?); 2) big-endian AF_ value (BSD on big-endian platforms); 3) little-endian AF_ value (BSD on little-endian platforms); 4) two octets of 0 followed by an Ethernet type (Linux, at least on little-endian platforms, as mutated by "libpcap"). Make a separate Wiretap encapsulation type, WTAP_ENCAP_NULL, corresponding to DLT_NULL. Have the PPP code dissect the frame if it's PPP-over-HDLC, and have "ethertype()" dissect the Ethernet type and the rest of the packet if it's a Linux-style header; dissect it ourselves only if it's an AF_ value. Have Wiretap impose a maximum packet size of 65535 bytes, so that it fails more gracefully when handed a corrupt "libpcap" capture file (other capture file formats with more than a 16-bit capture length field, if any, will have that check added later), and put that size in "wtap.h" and have Ethereal use it as its notion of a maximum packet size. Have Ethereal put up a "this file appears to be damaged or corrupt" message box if Wiretap returns a WTAP_ERR_BAD_RECORD error when opening or reading a capture file. Include loopback interfaces in the list of interfaces offered by the "Capture" dialog box, but put them at the end of the list so that it doesn't default to a loopback interface unless there are no other interfaces. Also, don't require that an interface in the list have an IP address associated with it, and only put one entry in the list for a given interface (SIOCGIFCONF returns one entry per interface *address*, not per *interface* - and even if you were to use only IP addresses, an interface could conceivably have more than one IP address). Exclusively use Wiretap encapsulation types internally, even when capturing; don't use DLT_ types. svn path=/trunk/; revision=540
1999-08-22 00:47:56 +00:00
WTAP_ENCAP_UNKNOWN,
WTAP_ENCAP_UNKNOWN,
Add to Wiretap the ability to write capture files; for now, it can only write them in "libpcap" format, but the mechanism can have other formats added. When creating the temporary file for a capture, use "create_tempfile()", to close a security hole opened by the fact that "tempnam()" creates a temporary file, but doesn't open it, and we open the file with the name it gives us - somebody could remove the file and plant a link to some file, and, if as may well be the case when Ethereal is capturing packets, it's running as "root", that means we write a capture on top of that file.... (The aforementioned changes to Wiretap let you open a capture file for writing given an file descriptor, "fdopen()"-style, which this change requires.) svn path=/trunk/; revision=509
1999-08-18 04:17:38 +00:00
WTAP_ENCAP_LINUX_ATM_CLIP
};
#define NUM_PCAP_ENCAPS (sizeof pcap_encap / sizeof pcap_encap[0])
Have the per-capture-file-type open routines "wtap_open_offline()" calls return 1 on success, -1 if they got an error, and 0 if the file isn't of the type that file is checking for, and supply an error code if they return -1; have "wtap_open_offline()" use that error code. Also, have the per-capture-file-type open routines treat errors accessing the file as errors, and return -1, rather than just returning 0 so that we try another file type. Have the per-capture-file-type read routines "wtap_loop()" calls return -1 and supply an error code on error (and not, as they did in some cases, call "g_error()" and abort), and have "wtap_loop()", if the read routine returned an error, return FALSE (and pass an error-code-pointer argument onto the read routines, so they fill it in), and return TRUE on success. Add some new error codes for them to return. Now that "wtap_loop()" can return a success/failure indication and an error code, in "read_cap_file()" put up a message box if we get an error reading the file, and return the error code. Handle the additional errors we can get when opening a capture file. If the attempt to open a capture file succeeds, but the attempt to read it fails, don't treat that as a complete failure - we may have managed to read some of the capture file, and we should display what we managed to read. svn path=/trunk/; revision=516
1999-08-19 05:31:38 +00:00
int libpcap_open(wtap *wth, int *err)
Add support to wiretap for reading Sun "snoop" capture files. That requires that, in the packet-reading loop, we pass to the callback routine the offset in the file of a packet's data, because we can no longer compute that offset by subtracting the size of the captured packet data from the offset in the file after the data was read - "snoop" may stick padding in after the packet data to align packet headers on 4-byte boundaries. Doing that required that we arrange that we do that for "libpcap" capture files as well; the cleanest way to do that was to write our own code for reading "libpcap" capture files, rather than using the "libpcap" code to do it. Make "wtap_dispatch_cb()" and "pcap_dispatch_cb()" static to "file.c", as they're not used elsewhere. If we're using wiretap, don't define in "file.h" stuff used only when we're not using wiretap. Update the wiretap README to reflect Gilbert's and my recent changes. Clean up some memory leaks in "wiretap/lanalyzer.c" and "wiretap/ngsniffer.c", where the capture-file-format-specific data wasn't freed if the open failed. svn path=/trunk/; revision=91
1998-11-15 05:29:17 +00:00
{
int bytes_read;
guint32 magic;
struct pcap_hdr hdr;
Add support for capture files written by programs linked with Alexey Kuznetsov's modified "libpcap" *as long as you have the ss990915 or later patch*; the 990417 patch, alas, changes the per-packet header but *doesn't* change the magic number, so you can't just look at the magic number to see that it's Not Standard Libpcap. (Even more unfortunately, Red Hat appears to have picked up *that* patch for Red Hat 6.1; I've filed bug 6773 with Bugzilla on their site - hopefully, if I'm not misremembering the RH 6.1 code I've seen, and they really *did* pick up the older patch, they'll fix it ASAP to use the new magic number, and will make updates available.) svn path=/trunk/; revision=986
1999-11-06 08:42:01 +00:00
gboolean byte_swapped = FALSE;
gboolean mutant = FALSE;
Add support to wiretap for reading Sun "snoop" capture files. That requires that, in the packet-reading loop, we pass to the callback routine the offset in the file of a packet's data, because we can no longer compute that offset by subtracting the size of the captured packet data from the offset in the file after the data was read - "snoop" may stick padding in after the packet data to align packet headers on 4-byte boundaries. Doing that required that we arrange that we do that for "libpcap" capture files as well; the cleanest way to do that was to write our own code for reading "libpcap" capture files, rather than using the "libpcap" code to do it. Make "wtap_dispatch_cb()" and "pcap_dispatch_cb()" static to "file.c", as they're not used elsewhere. If we're using wiretap, don't define in "file.h" stuff used only when we're not using wiretap. Update the wiretap README to reflect Gilbert's and my recent changes. Clean up some memory leaks in "wiretap/lanalyzer.c" and "wiretap/ngsniffer.c", where the capture-file-format-specific data wasn't freed if the open failed. svn path=/trunk/; revision=91
1998-11-15 05:29:17 +00:00
/* Read in the number that should be at the start of a "libpcap" file */
This commit contains support for reading capture files compressed using gzip. The zLib library is used for this purpose. If zLib is not available (or it's use is disabled by the --disable-zlib option to configure), you can still compile Ethereal but it will be unable to read compressed capture files. IMPORTANT: Now all file accesses to capture files should be done through special macros. Specifically, for any use of the following functions on capture files, replace them. The arguments for the right-side functions are exactly the same as for the original stdio functions. fopen file_open fdopen filed_open fread file_read fwrite file_write fseek file_seek fclose file_close ferror file_error svn path=/trunk/; revision=695
1999-09-22 01:26:50 +00:00
file_seek(wth->fh, 0, SEEK_SET);
Keep in the "wtap" structure the current offset into the file being read, and maintain it ourselves as we read through the file, rather than calling "ftell()" for every packet we read - "ftell()" may involve an "lseek()" call, which could add a noticeable CPU overhead when reading a large file. svn path=/trunk/; revision=596
1999-08-28 01:19:45 +00:00
wth->data_offset = 0;
Have the per-capture-file-type open routines "wtap_open_offline()" calls return 1 on success, -1 if they got an error, and 0 if the file isn't of the type that file is checking for, and supply an error code if they return -1; have "wtap_open_offline()" use that error code. Also, have the per-capture-file-type open routines treat errors accessing the file as errors, and return -1, rather than just returning 0 so that we try another file type. Have the per-capture-file-type read routines "wtap_loop()" calls return -1 and supply an error code on error (and not, as they did in some cases, call "g_error()" and abort), and have "wtap_loop()", if the read routine returned an error, return FALSE (and pass an error-code-pointer argument onto the read routines, so they fill it in), and return TRUE on success. Add some new error codes for them to return. Now that "wtap_loop()" can return a success/failure indication and an error code, in "read_cap_file()" put up a message box if we get an error reading the file, and return the error code. Handle the additional errors we can get when opening a capture file. If the attempt to open a capture file succeeds, but the attempt to read it fails, don't treat that as a complete failure - we may have managed to read some of the capture file, and we should display what we managed to read. svn path=/trunk/; revision=516
1999-08-19 05:31:38 +00:00
errno = WTAP_ERR_CANT_READ;
This commit contains support for reading capture files compressed using gzip. The zLib library is used for this purpose. If zLib is not available (or it's use is disabled by the --disable-zlib option to configure), you can still compile Ethereal but it will be unable to read compressed capture files. IMPORTANT: Now all file accesses to capture files should be done through special macros. Specifically, for any use of the following functions on capture files, replace them. The arguments for the right-side functions are exactly the same as for the original stdio functions. fopen file_open fdopen filed_open fread file_read fwrite file_write fseek file_seek fclose file_close ferror file_error svn path=/trunk/; revision=695
1999-09-22 01:26:50 +00:00
bytes_read = file_read(&magic, 1, sizeof magic, wth->fh);
Add support to wiretap for reading Sun "snoop" capture files. That requires that, in the packet-reading loop, we pass to the callback routine the offset in the file of a packet's data, because we can no longer compute that offset by subtracting the size of the captured packet data from the offset in the file after the data was read - "snoop" may stick padding in after the packet data to align packet headers on 4-byte boundaries. Doing that required that we arrange that we do that for "libpcap" capture files as well; the cleanest way to do that was to write our own code for reading "libpcap" capture files, rather than using the "libpcap" code to do it. Make "wtap_dispatch_cb()" and "pcap_dispatch_cb()" static to "file.c", as they're not used elsewhere. If we're using wiretap, don't define in "file.h" stuff used only when we're not using wiretap. Update the wiretap README to reflect Gilbert's and my recent changes. Clean up some memory leaks in "wiretap/lanalyzer.c" and "wiretap/ngsniffer.c", where the capture-file-format-specific data wasn't freed if the open failed. svn path=/trunk/; revision=91
1998-11-15 05:29:17 +00:00
if (bytes_read != sizeof magic) {
Better handle errors from zlib: Assign a range of Wiretap errors for zlib errors, and have "wtap_strerror()" use "zError()" to get an error message for them. Have the internal "file_error()" routine return 0 for no error and a Wiretap error code for an error. svn path=/trunk/; revision=769
1999-10-05 07:06:08 +00:00
*err = file_error(wth->fh);
if (*err != 0)
Have the per-capture-file-type open routines "wtap_open_offline()" calls return 1 on success, -1 if they got an error, and 0 if the file isn't of the type that file is checking for, and supply an error code if they return -1; have "wtap_open_offline()" use that error code. Also, have the per-capture-file-type open routines treat errors accessing the file as errors, and return -1, rather than just returning 0 so that we try another file type. Have the per-capture-file-type read routines "wtap_loop()" calls return -1 and supply an error code on error (and not, as they did in some cases, call "g_error()" and abort), and have "wtap_loop()", if the read routine returned an error, return FALSE (and pass an error-code-pointer argument onto the read routines, so they fill it in), and return TRUE on success. Add some new error codes for them to return. Now that "wtap_loop()" can return a success/failure indication and an error code, in "read_cap_file()" put up a message box if we get an error reading the file, and return the error code. Handle the additional errors we can get when opening a capture file. If the attempt to open a capture file succeeds, but the attempt to read it fails, don't treat that as a complete failure - we may have managed to read some of the capture file, and we should display what we managed to read. svn path=/trunk/; revision=516
1999-08-19 05:31:38 +00:00
return -1;
return 0;
Add support to wiretap for reading Sun "snoop" capture files. That requires that, in the packet-reading loop, we pass to the callback routine the offset in the file of a packet's data, because we can no longer compute that offset by subtracting the size of the captured packet data from the offset in the file after the data was read - "snoop" may stick padding in after the packet data to align packet headers on 4-byte boundaries. Doing that required that we arrange that we do that for "libpcap" capture files as well; the cleanest way to do that was to write our own code for reading "libpcap" capture files, rather than using the "libpcap" code to do it. Make "wtap_dispatch_cb()" and "pcap_dispatch_cb()" static to "file.c", as they're not used elsewhere. If we're using wiretap, don't define in "file.h" stuff used only when we're not using wiretap. Update the wiretap README to reflect Gilbert's and my recent changes. Clean up some memory leaks in "wiretap/lanalyzer.c" and "wiretap/ngsniffer.c", where the capture-file-format-specific data wasn't freed if the open failed. svn path=/trunk/; revision=91
1998-11-15 05:29:17 +00:00
}
Keep in the "wtap" structure the current offset into the file being read, and maintain it ourselves as we read through the file, rather than calling "ftell()" for every packet we read - "ftell()" may involve an "lseek()" call, which could add a noticeable CPU overhead when reading a large file. svn path=/trunk/; revision=596
1999-08-28 01:19:45 +00:00
wth->data_offset += sizeof magic;
Add support to wiretap for reading Sun "snoop" capture files. That requires that, in the packet-reading loop, we pass to the callback routine the offset in the file of a packet's data, because we can no longer compute that offset by subtracting the size of the captured packet data from the offset in the file after the data was read - "snoop" may stick padding in after the packet data to align packet headers on 4-byte boundaries. Doing that required that we arrange that we do that for "libpcap" capture files as well; the cleanest way to do that was to write our own code for reading "libpcap" capture files, rather than using the "libpcap" code to do it. Make "wtap_dispatch_cb()" and "pcap_dispatch_cb()" static to "file.c", as they're not used elsewhere. If we're using wiretap, don't define in "file.h" stuff used only when we're not using wiretap. Update the wiretap README to reflect Gilbert's and my recent changes. Clean up some memory leaks in "wiretap/lanalyzer.c" and "wiretap/ngsniffer.c", where the capture-file-format-specific data wasn't freed if the open failed. svn path=/trunk/; revision=91
1998-11-15 05:29:17 +00:00
Add support for capture files written by programs linked with Alexey Kuznetsov's modified "libpcap" *as long as you have the ss990915 or later patch*; the 990417 patch, alas, changes the per-packet header but *doesn't* change the magic number, so you can't just look at the magic number to see that it's Not Standard Libpcap. (Even more unfortunately, Red Hat appears to have picked up *that* patch for Red Hat 6.1; I've filed bug 6773 with Bugzilla on their site - hopefully, if I'm not misremembering the RH 6.1 code I've seen, and they really *did* pick up the older patch, they'll fix it ASAP to use the new magic number, and will make updates available.) svn path=/trunk/; revision=986
1999-11-06 08:42:01 +00:00
switch (magic) {
case PCAP_MAGIC:
/* Host that wrote it has our byte order. */
byte_swapped = FALSE;
mutant = FALSE;
break;
case PCAP_MUTANT_MAGIC:
/* Host that wrote it has our byte order, but was running
a program using the patched "libpcap". */
byte_swapped = FALSE;
mutant = TRUE;
break;
case PCAP_SWAPPED_MAGIC:
Add support to wiretap for reading Sun "snoop" capture files. That requires that, in the packet-reading loop, we pass to the callback routine the offset in the file of a packet's data, because we can no longer compute that offset by subtracting the size of the captured packet data from the offset in the file after the data was read - "snoop" may stick padding in after the packet data to align packet headers on 4-byte boundaries. Doing that required that we arrange that we do that for "libpcap" capture files as well; the cleanest way to do that was to write our own code for reading "libpcap" capture files, rather than using the "libpcap" code to do it. Make "wtap_dispatch_cb()" and "pcap_dispatch_cb()" static to "file.c", as they're not used elsewhere. If we're using wiretap, don't define in "file.h" stuff used only when we're not using wiretap. Update the wiretap README to reflect Gilbert's and my recent changes. Clean up some memory leaks in "wiretap/lanalyzer.c" and "wiretap/ngsniffer.c", where the capture-file-format-specific data wasn't freed if the open failed. svn path=/trunk/; revision=91
1998-11-15 05:29:17 +00:00
/* Host that wrote it has a byte order opposite to ours. */
Add support for capture files written by programs linked with Alexey Kuznetsov's modified "libpcap" *as long as you have the ss990915 or later patch*; the 990417 patch, alas, changes the per-packet header but *doesn't* change the magic number, so you can't just look at the magic number to see that it's Not Standard Libpcap. (Even more unfortunately, Red Hat appears to have picked up *that* patch for Red Hat 6.1; I've filed bug 6773 with Bugzilla on their site - hopefully, if I'm not misremembering the RH 6.1 code I've seen, and they really *did* pick up the older patch, they'll fix it ASAP to use the new magic number, and will make updates available.) svn path=/trunk/; revision=986
1999-11-06 08:42:01 +00:00
byte_swapped = TRUE;
mutant = FALSE;
break;
case PCAP_SWAPPED_MUTANT_MAGIC:
/* Host that wrote it out has a byte order opposite to
ours, and was running a program using the patched
"libpcap". */
byte_swapped = TRUE;
mutant = TRUE;
break;
default:
/* Not a "libpcap" type we know about. */
Have the per-capture-file-type open routines "wtap_open_offline()" calls return 1 on success, -1 if they got an error, and 0 if the file isn't of the type that file is checking for, and supply an error code if they return -1; have "wtap_open_offline()" use that error code. Also, have the per-capture-file-type open routines treat errors accessing the file as errors, and return -1, rather than just returning 0 so that we try another file type. Have the per-capture-file-type read routines "wtap_loop()" calls return -1 and supply an error code on error (and not, as they did in some cases, call "g_error()" and abort), and have "wtap_loop()", if the read routine returned an error, return FALSE (and pass an error-code-pointer argument onto the read routines, so they fill it in), and return TRUE on success. Add some new error codes for them to return. Now that "wtap_loop()" can return a success/failure indication and an error code, in "read_cap_file()" put up a message box if we get an error reading the file, and return the error code. Handle the additional errors we can get when opening a capture file. If the attempt to open a capture file succeeds, but the attempt to read it fails, don't treat that as a complete failure - we may have managed to read some of the capture file, and we should display what we managed to read. svn path=/trunk/; revision=516
1999-08-19 05:31:38 +00:00
return 0;
Add support to wiretap for reading Sun "snoop" capture files. That requires that, in the packet-reading loop, we pass to the callback routine the offset in the file of a packet's data, because we can no longer compute that offset by subtracting the size of the captured packet data from the offset in the file after the data was read - "snoop" may stick padding in after the packet data to align packet headers on 4-byte boundaries. Doing that required that we arrange that we do that for "libpcap" capture files as well; the cleanest way to do that was to write our own code for reading "libpcap" capture files, rather than using the "libpcap" code to do it. Make "wtap_dispatch_cb()" and "pcap_dispatch_cb()" static to "file.c", as they're not used elsewhere. If we're using wiretap, don't define in "file.h" stuff used only when we're not using wiretap. Update the wiretap README to reflect Gilbert's and my recent changes. Clean up some memory leaks in "wiretap/lanalyzer.c" and "wiretap/ngsniffer.c", where the capture-file-format-specific data wasn't freed if the open failed. svn path=/trunk/; revision=91
1998-11-15 05:29:17 +00:00
}
/* Read the rest of the header. */
Have the per-capture-file-type open routines "wtap_open_offline()" calls return 1 on success, -1 if they got an error, and 0 if the file isn't of the type that file is checking for, and supply an error code if they return -1; have "wtap_open_offline()" use that error code. Also, have the per-capture-file-type open routines treat errors accessing the file as errors, and return -1, rather than just returning 0 so that we try another file type. Have the per-capture-file-type read routines "wtap_loop()" calls return -1 and supply an error code on error (and not, as they did in some cases, call "g_error()" and abort), and have "wtap_loop()", if the read routine returned an error, return FALSE (and pass an error-code-pointer argument onto the read routines, so they fill it in), and return TRUE on success. Add some new error codes for them to return. Now that "wtap_loop()" can return a success/failure indication and an error code, in "read_cap_file()" put up a message box if we get an error reading the file, and return the error code. Handle the additional errors we can get when opening a capture file. If the attempt to open a capture file succeeds, but the attempt to read it fails, don't treat that as a complete failure - we may have managed to read some of the capture file, and we should display what we managed to read. svn path=/trunk/; revision=516
1999-08-19 05:31:38 +00:00
errno = WTAP_ERR_CANT_READ;
This commit contains support for reading capture files compressed using gzip. The zLib library is used for this purpose. If zLib is not available (or it's use is disabled by the --disable-zlib option to configure), you can still compile Ethereal but it will be unable to read compressed capture files. IMPORTANT: Now all file accesses to capture files should be done through special macros. Specifically, for any use of the following functions on capture files, replace them. The arguments for the right-side functions are exactly the same as for the original stdio functions. fopen file_open fdopen filed_open fread file_read fwrite file_write fseek file_seek fclose file_close ferror file_error svn path=/trunk/; revision=695
1999-09-22 01:26:50 +00:00
bytes_read = file_read(&hdr, 1, sizeof hdr, wth->fh);
Add support to wiretap for reading Sun "snoop" capture files. That requires that, in the packet-reading loop, we pass to the callback routine the offset in the file of a packet's data, because we can no longer compute that offset by subtracting the size of the captured packet data from the offset in the file after the data was read - "snoop" may stick padding in after the packet data to align packet headers on 4-byte boundaries. Doing that required that we arrange that we do that for "libpcap" capture files as well; the cleanest way to do that was to write our own code for reading "libpcap" capture files, rather than using the "libpcap" code to do it. Make "wtap_dispatch_cb()" and "pcap_dispatch_cb()" static to "file.c", as they're not used elsewhere. If we're using wiretap, don't define in "file.h" stuff used only when we're not using wiretap. Update the wiretap README to reflect Gilbert's and my recent changes. Clean up some memory leaks in "wiretap/lanalyzer.c" and "wiretap/ngsniffer.c", where the capture-file-format-specific data wasn't freed if the open failed. svn path=/trunk/; revision=91
1998-11-15 05:29:17 +00:00
if (bytes_read != sizeof hdr) {
Better handle errors from zlib: Assign a range of Wiretap errors for zlib errors, and have "wtap_strerror()" use "zError()" to get an error message for them. Have the internal "file_error()" routine return 0 for no error and a Wiretap error code for an error. svn path=/trunk/; revision=769
1999-10-05 07:06:08 +00:00
*err = file_error(wth->fh);
if (*err != 0)
Have the per-capture-file-type open routines "wtap_open_offline()" calls return 1 on success, -1 if they got an error, and 0 if the file isn't of the type that file is checking for, and supply an error code if they return -1; have "wtap_open_offline()" use that error code. Also, have the per-capture-file-type open routines treat errors accessing the file as errors, and return -1, rather than just returning 0 so that we try another file type. Have the per-capture-file-type read routines "wtap_loop()" calls return -1 and supply an error code on error (and not, as they did in some cases, call "g_error()" and abort), and have "wtap_loop()", if the read routine returned an error, return FALSE (and pass an error-code-pointer argument onto the read routines, so they fill it in), and return TRUE on success. Add some new error codes for them to return. Now that "wtap_loop()" can return a success/failure indication and an error code, in "read_cap_file()" put up a message box if we get an error reading the file, and return the error code. Handle the additional errors we can get when opening a capture file. If the attempt to open a capture file succeeds, but the attempt to read it fails, don't treat that as a complete failure - we may have managed to read some of the capture file, and we should display what we managed to read. svn path=/trunk/; revision=516
1999-08-19 05:31:38 +00:00
return -1;
return 0;
Add support to wiretap for reading Sun "snoop" capture files. That requires that, in the packet-reading loop, we pass to the callback routine the offset in the file of a packet's data, because we can no longer compute that offset by subtracting the size of the captured packet data from the offset in the file after the data was read - "snoop" may stick padding in after the packet data to align packet headers on 4-byte boundaries. Doing that required that we arrange that we do that for "libpcap" capture files as well; the cleanest way to do that was to write our own code for reading "libpcap" capture files, rather than using the "libpcap" code to do it. Make "wtap_dispatch_cb()" and "pcap_dispatch_cb()" static to "file.c", as they're not used elsewhere. If we're using wiretap, don't define in "file.h" stuff used only when we're not using wiretap. Update the wiretap README to reflect Gilbert's and my recent changes. Clean up some memory leaks in "wiretap/lanalyzer.c" and "wiretap/ngsniffer.c", where the capture-file-format-specific data wasn't freed if the open failed. svn path=/trunk/; revision=91
1998-11-15 05:29:17 +00:00
}
Keep in the "wtap" structure the current offset into the file being read, and maintain it ourselves as we read through the file, rather than calling "ftell()" for every packet we read - "ftell()" may involve an "lseek()" call, which could add a noticeable CPU overhead when reading a large file. svn path=/trunk/; revision=596
1999-08-28 01:19:45 +00:00
wth->data_offset += sizeof hdr;
Add support to wiretap for reading Sun "snoop" capture files. That requires that, in the packet-reading loop, we pass to the callback routine the offset in the file of a packet's data, because we can no longer compute that offset by subtracting the size of the captured packet data from the offset in the file after the data was read - "snoop" may stick padding in after the packet data to align packet headers on 4-byte boundaries. Doing that required that we arrange that we do that for "libpcap" capture files as well; the cleanest way to do that was to write our own code for reading "libpcap" capture files, rather than using the "libpcap" code to do it. Make "wtap_dispatch_cb()" and "pcap_dispatch_cb()" static to "file.c", as they're not used elsewhere. If we're using wiretap, don't define in "file.h" stuff used only when we're not using wiretap. Update the wiretap README to reflect Gilbert's and my recent changes. Clean up some memory leaks in "wiretap/lanalyzer.c" and "wiretap/ngsniffer.c", where the capture-file-format-specific data wasn't freed if the open failed. svn path=/trunk/; revision=91
1998-11-15 05:29:17 +00:00
if (byte_swapped) {
/* Byte-swap the header fields about which we care. */
hdr.version_major = BSWAP16(hdr.version_major);
hdr.version_minor = BSWAP16(hdr.version_minor);
hdr.snaplen = BSWAP32(hdr.snaplen);
hdr.network = BSWAP32(hdr.network);
}
if (hdr.version_major < 2) {
/* We only support version 2.0 and later. */
Get rid of some cruft left in by previous checkins as placeholders. Get rid of WTAP_ENCAP_NONE; replace it with WTAP_ENCAP_UNKNOWN, which means "I can't handle that file, it's using an encapsulation I don't support". Check for encapsulations we don't support, and return an error (as is already done in "libpcap.c"). Check for too-large packet sizes, and return an error (as is already done in "libpcap.c"). Print unsigned quantities in Wiretap messages with "%u", not "%d". svn path=/trunk/; revision=544
1999-08-22 02:29:40 +00:00
g_message("pcap: major version %u unsupported",
Have the per-capture-file-type open routines "wtap_open_offline()" calls return 1 on success, -1 if they got an error, and 0 if the file isn't of the type that file is checking for, and supply an error code if they return -1; have "wtap_open_offline()" use that error code. Also, have the per-capture-file-type open routines treat errors accessing the file as errors, and return -1, rather than just returning 0 so that we try another file type. Have the per-capture-file-type read routines "wtap_loop()" calls return -1 and supply an error code on error (and not, as they did in some cases, call "g_error()" and abort), and have "wtap_loop()", if the read routine returned an error, return FALSE (and pass an error-code-pointer argument onto the read routines, so they fill it in), and return TRUE on success. Add some new error codes for them to return. Now that "wtap_loop()" can return a success/failure indication and an error code, in "read_cap_file()" put up a message box if we get an error reading the file, and return the error code. Handle the additional errors we can get when opening a capture file. If the attempt to open a capture file succeeds, but the attempt to read it fails, don't treat that as a complete failure - we may have managed to read some of the capture file, and we should display what we managed to read. svn path=/trunk/; revision=516
1999-08-19 05:31:38 +00:00
hdr.version_major);
*err = WTAP_ERR_UNSUPPORTED;
return -1;
Add support to wiretap for reading Sun "snoop" capture files. That requires that, in the packet-reading loop, we pass to the callback routine the offset in the file of a packet's data, because we can no longer compute that offset by subtracting the size of the captured packet data from the offset in the file after the data was read - "snoop" may stick padding in after the packet data to align packet headers on 4-byte boundaries. Doing that required that we arrange that we do that for "libpcap" capture files as well; the cleanest way to do that was to write our own code for reading "libpcap" capture files, rather than using the "libpcap" code to do it. Make "wtap_dispatch_cb()" and "pcap_dispatch_cb()" static to "file.c", as they're not used elsewhere. If we're using wiretap, don't define in "file.h" stuff used only when we're not using wiretap. Update the wiretap README to reflect Gilbert's and my recent changes. Clean up some memory leaks in "wiretap/lanalyzer.c" and "wiretap/ngsniffer.c", where the capture-file-format-specific data wasn't freed if the open failed. svn path=/trunk/; revision=91
1998-11-15 05:29:17 +00:00
}
DLT_NULL, from "libpcap", means different things on different platforms and in different capture files; throw in some heuristics to try to figure out whether the 4-byte header is: 1) PPP-over-HDLC (some version of ISDN4BSD?); 2) big-endian AF_ value (BSD on big-endian platforms); 3) little-endian AF_ value (BSD on little-endian platforms); 4) two octets of 0 followed by an Ethernet type (Linux, at least on little-endian platforms, as mutated by "libpcap"). Make a separate Wiretap encapsulation type, WTAP_ENCAP_NULL, corresponding to DLT_NULL. Have the PPP code dissect the frame if it's PPP-over-HDLC, and have "ethertype()" dissect the Ethernet type and the rest of the packet if it's a Linux-style header; dissect it ourselves only if it's an AF_ value. Have Wiretap impose a maximum packet size of 65535 bytes, so that it fails more gracefully when handed a corrupt "libpcap" capture file (other capture file formats with more than a 16-bit capture length field, if any, will have that check added later), and put that size in "wtap.h" and have Ethereal use it as its notion of a maximum packet size. Have Ethereal put up a "this file appears to be damaged or corrupt" message box if Wiretap returns a WTAP_ERR_BAD_RECORD error when opening or reading a capture file. Include loopback interfaces in the list of interfaces offered by the "Capture" dialog box, but put them at the end of the list so that it doesn't default to a loopback interface unless there are no other interfaces. Also, don't require that an interface in the list have an IP address associated with it, and only put one entry in the list for a given interface (SIOCGIFCONF returns one entry per interface *address*, not per *interface* - and even if you were to use only IP addresses, an interface could conceivably have more than one IP address). Exclusively use Wiretap encapsulation types internally, even when capturing; don't use DLT_ types. svn path=/trunk/; revision=540
1999-08-22 00:47:56 +00:00
if (hdr.network >= NUM_PCAP_ENCAPS
|| pcap_encap[hdr.network] == WTAP_ENCAP_UNKNOWN) {
Get rid of some cruft left in by previous checkins as placeholders. Get rid of WTAP_ENCAP_NONE; replace it with WTAP_ENCAP_UNKNOWN, which means "I can't handle that file, it's using an encapsulation I don't support". Check for encapsulations we don't support, and return an error (as is already done in "libpcap.c"). Check for too-large packet sizes, and return an error (as is already done in "libpcap.c"). Print unsigned quantities in Wiretap messages with "%u", not "%d". svn path=/trunk/; revision=544
1999-08-22 02:29:40 +00:00
g_message("pcap: network type %u unknown or unsupported",
hdr.network);
Have the per-capture-file-type open routines "wtap_open_offline()" calls return 1 on success, -1 if they got an error, and 0 if the file isn't of the type that file is checking for, and supply an error code if they return -1; have "wtap_open_offline()" use that error code. Also, have the per-capture-file-type open routines treat errors accessing the file as errors, and return -1, rather than just returning 0 so that we try another file type. Have the per-capture-file-type read routines "wtap_loop()" calls return -1 and supply an error code on error (and not, as they did in some cases, call "g_error()" and abort), and have "wtap_loop()", if the read routine returned an error, return FALSE (and pass an error-code-pointer argument onto the read routines, so they fill it in), and return TRUE on success. Add some new error codes for them to return. Now that "wtap_loop()" can return a success/failure indication and an error code, in "read_cap_file()" put up a message box if we get an error reading the file, and return the error code. Handle the additional errors we can get when opening a capture file. If the attempt to open a capture file succeeds, but the attempt to read it fails, don't treat that as a complete failure - we may have managed to read some of the capture file, and we should display what we managed to read. svn path=/trunk/; revision=516
1999-08-19 05:31:38 +00:00
*err = WTAP_ERR_UNSUPPORTED;
return -1;
Add support to wiretap for reading Sun "snoop" capture files. That requires that, in the packet-reading loop, we pass to the callback routine the offset in the file of a packet's data, because we can no longer compute that offset by subtracting the size of the captured packet data from the offset in the file after the data was read - "snoop" may stick padding in after the packet data to align packet headers on 4-byte boundaries. Doing that required that we arrange that we do that for "libpcap" capture files as well; the cleanest way to do that was to write our own code for reading "libpcap" capture files, rather than using the "libpcap" code to do it. Make "wtap_dispatch_cb()" and "pcap_dispatch_cb()" static to "file.c", as they're not used elsewhere. If we're using wiretap, don't define in "file.h" stuff used only when we're not using wiretap. Update the wiretap README to reflect Gilbert's and my recent changes. Clean up some memory leaks in "wiretap/lanalyzer.c" and "wiretap/ngsniffer.c", where the capture-file-format-specific data wasn't freed if the open failed. svn path=/trunk/; revision=91
1998-11-15 05:29:17 +00:00
}
/* This is a libpcap file */
Have the per-capture-file-type open routines "wtap_open_offline()" calls return 1 on success, -1 if they got an error, and 0 if the file isn't of the type that file is checking for, and supply an error code if they return -1; have "wtap_open_offline()" use that error code. Also, have the per-capture-file-type open routines treat errors accessing the file as errors, and return -1, rather than just returning 0 so that we try another file type. Have the per-capture-file-type read routines "wtap_loop()" calls return -1 and supply an error code on error (and not, as they did in some cases, call "g_error()" and abort), and have "wtap_loop()", if the read routine returned an error, return FALSE (and pass an error-code-pointer argument onto the read routines, so they fill it in), and return TRUE on success. Add some new error codes for them to return. Now that "wtap_loop()" can return a success/failure indication and an error code, in "read_cap_file()" put up a message box if we get an error reading the file, and return the error code. Handle the additional errors we can get when opening a capture file. If the attempt to open a capture file succeeds, but the attempt to read it fails, don't treat that as a complete failure - we may have managed to read some of the capture file, and we should display what we managed to read. svn path=/trunk/; revision=516
1999-08-19 05:31:38 +00:00
wth->file_type = WTAP_FILE_PCAP;
Add support to wiretap for reading Sun "snoop" capture files. That requires that, in the packet-reading loop, we pass to the callback routine the offset in the file of a packet's data, because we can no longer compute that offset by subtracting the size of the captured packet data from the offset in the file after the data was read - "snoop" may stick padding in after the packet data to align packet headers on 4-byte boundaries. Doing that required that we arrange that we do that for "libpcap" capture files as well; the cleanest way to do that was to write our own code for reading "libpcap" capture files, rather than using the "libpcap" code to do it. Make "wtap_dispatch_cb()" and "pcap_dispatch_cb()" static to "file.c", as they're not used elsewhere. If we're using wiretap, don't define in "file.h" stuff used only when we're not using wiretap. Update the wiretap README to reflect Gilbert's and my recent changes. Clean up some memory leaks in "wiretap/lanalyzer.c" and "wiretap/ngsniffer.c", where the capture-file-format-specific data wasn't freed if the open failed. svn path=/trunk/; revision=91
1998-11-15 05:29:17 +00:00
wth->capture.pcap = g_malloc(sizeof(libpcap_t));
wth->capture.pcap->byte_swapped = byte_swapped;
Add support for capture files written by programs linked with Alexey Kuznetsov's modified "libpcap" *as long as you have the ss990915 or later patch*; the 990417 patch, alas, changes the per-packet header but *doesn't* change the magic number, so you can't just look at the magic number to see that it's Not Standard Libpcap. (Even more unfortunately, Red Hat appears to have picked up *that* patch for Red Hat 6.1; I've filed bug 6773 with Bugzilla on their site - hopefully, if I'm not misremembering the RH 6.1 code I've seen, and they really *did* pick up the older patch, they'll fix it ASAP to use the new magic number, and will make updates available.) svn path=/trunk/; revision=986
1999-11-06 08:42:01 +00:00
wth->capture.pcap->mutant = mutant;
Add support to wiretap for reading Sun "snoop" capture files. That requires that, in the packet-reading loop, we pass to the callback routine the offset in the file of a packet's data, because we can no longer compute that offset by subtracting the size of the captured packet data from the offset in the file after the data was read - "snoop" may stick padding in after the packet data to align packet headers on 4-byte boundaries. Doing that required that we arrange that we do that for "libpcap" capture files as well; the cleanest way to do that was to write our own code for reading "libpcap" capture files, rather than using the "libpcap" code to do it. Make "wtap_dispatch_cb()" and "pcap_dispatch_cb()" static to "file.c", as they're not used elsewhere. If we're using wiretap, don't define in "file.h" stuff used only when we're not using wiretap. Update the wiretap README to reflect Gilbert's and my recent changes. Clean up some memory leaks in "wiretap/lanalyzer.c" and "wiretap/ngsniffer.c", where the capture-file-format-specific data wasn't freed if the open failed. svn path=/trunk/; revision=91
1998-11-15 05:29:17 +00:00
wth->capture.pcap->version_major = hdr.version_major;
wth->capture.pcap->version_minor = hdr.version_minor;
wth->subtype_read = libpcap_read;
Added display filters to wiretap. svn path=/trunk/; revision=198
1999-03-01 18:57:07 +00:00
wth->file_encap = pcap_encap[hdr.network];
Add support to wiretap for reading Sun "snoop" capture files. That requires that, in the packet-reading loop, we pass to the callback routine the offset in the file of a packet's data, because we can no longer compute that offset by subtracting the size of the captured packet data from the offset in the file after the data was read - "snoop" may stick padding in after the packet data to align packet headers on 4-byte boundaries. Doing that required that we arrange that we do that for "libpcap" capture files as well; the cleanest way to do that was to write our own code for reading "libpcap" capture files, rather than using the "libpcap" code to do it. Make "wtap_dispatch_cb()" and "pcap_dispatch_cb()" static to "file.c", as they're not used elsewhere. If we're using wiretap, don't define in "file.h" stuff used only when we're not using wiretap. Update the wiretap README to reflect Gilbert's and my recent changes. Clean up some memory leaks in "wiretap/lanalyzer.c" and "wiretap/ngsniffer.c", where the capture-file-format-specific data wasn't freed if the open failed. svn path=/trunk/; revision=91
1998-11-15 05:29:17 +00:00
wth->snapshot_length = hdr.snaplen;
Have the per-capture-file-type open routines "wtap_open_offline()" calls return 1 on success, -1 if they got an error, and 0 if the file isn't of the type that file is checking for, and supply an error code if they return -1; have "wtap_open_offline()" use that error code. Also, have the per-capture-file-type open routines treat errors accessing the file as errors, and return -1, rather than just returning 0 so that we try another file type. Have the per-capture-file-type read routines "wtap_loop()" calls return -1 and supply an error code on error (and not, as they did in some cases, call "g_error()" and abort), and have "wtap_loop()", if the read routine returned an error, return FALSE (and pass an error-code-pointer argument onto the read routines, so they fill it in), and return TRUE on success. Add some new error codes for them to return. Now that "wtap_loop()" can return a success/failure indication and an error code, in "read_cap_file()" put up a message box if we get an error reading the file, and return the error code. Handle the additional errors we can get when opening a capture file. If the attempt to open a capture file succeeds, but the attempt to read it fails, don't treat that as a complete failure - we may have managed to read some of the capture file, and we should display what we managed to read. svn path=/trunk/; revision=516
1999-08-19 05:31:38 +00:00
return 1;
Add support to wiretap for reading Sun "snoop" capture files. That requires that, in the packet-reading loop, we pass to the callback routine the offset in the file of a packet's data, because we can no longer compute that offset by subtracting the size of the captured packet data from the offset in the file after the data was read - "snoop" may stick padding in after the packet data to align packet headers on 4-byte boundaries. Doing that required that we arrange that we do that for "libpcap" capture files as well; the cleanest way to do that was to write our own code for reading "libpcap" capture files, rather than using the "libpcap" code to do it. Make "wtap_dispatch_cb()" and "pcap_dispatch_cb()" static to "file.c", as they're not used elsewhere. If we're using wiretap, don't define in "file.h" stuff used only when we're not using wiretap. Update the wiretap README to reflect Gilbert's and my recent changes. Clean up some memory leaks in "wiretap/lanalyzer.c" and "wiretap/ngsniffer.c", where the capture-file-format-specific data wasn't freed if the open failed. svn path=/trunk/; revision=91
1998-11-15 05:29:17 +00:00
}
/* Read the next packet */
Have the per-capture-file-type open routines "wtap_open_offline()" calls return 1 on success, -1 if they got an error, and 0 if the file isn't of the type that file is checking for, and supply an error code if they return -1; have "wtap_open_offline()" use that error code. Also, have the per-capture-file-type open routines treat errors accessing the file as errors, and return -1, rather than just returning 0 so that we try another file type. Have the per-capture-file-type read routines "wtap_loop()" calls return -1 and supply an error code on error (and not, as they did in some cases, call "g_error()" and abort), and have "wtap_loop()", if the read routine returned an error, return FALSE (and pass an error-code-pointer argument onto the read routines, so they fill it in), and return TRUE on success. Add some new error codes for them to return. Now that "wtap_loop()" can return a success/failure indication and an error code, in "read_cap_file()" put up a message box if we get an error reading the file, and return the error code. Handle the additional errors we can get when opening a capture file. If the attempt to open a capture file succeeds, but the attempt to read it fails, don't treat that as a complete failure - we may have managed to read some of the capture file, and we should display what we managed to read. svn path=/trunk/; revision=516
1999-08-19 05:31:38 +00:00
static int libpcap_read(wtap *wth, int *err)
Add support to wiretap for reading Sun "snoop" capture files. That requires that, in the packet-reading loop, we pass to the callback routine the offset in the file of a packet's data, because we can no longer compute that offset by subtracting the size of the captured packet data from the offset in the file after the data was read - "snoop" may stick padding in after the packet data to align packet headers on 4-byte boundaries. Doing that required that we arrange that we do that for "libpcap" capture files as well; the cleanest way to do that was to write our own code for reading "libpcap" capture files, rather than using the "libpcap" code to do it. Make "wtap_dispatch_cb()" and "pcap_dispatch_cb()" static to "file.c", as they're not used elsewhere. If we're using wiretap, don't define in "file.h" stuff used only when we're not using wiretap. Update the wiretap README to reflect Gilbert's and my recent changes. Clean up some memory leaks in "wiretap/lanalyzer.c" and "wiretap/ngsniffer.c", where the capture-file-format-specific data wasn't freed if the open failed. svn path=/trunk/; revision=91
1998-11-15 05:29:17 +00:00
{
DLT_NULL, from "libpcap", means different things on different platforms and in different capture files; throw in some heuristics to try to figure out whether the 4-byte header is: 1) PPP-over-HDLC (some version of ISDN4BSD?); 2) big-endian AF_ value (BSD on big-endian platforms); 3) little-endian AF_ value (BSD on little-endian platforms); 4) two octets of 0 followed by an Ethernet type (Linux, at least on little-endian platforms, as mutated by "libpcap"). Make a separate Wiretap encapsulation type, WTAP_ENCAP_NULL, corresponding to DLT_NULL. Have the PPP code dissect the frame if it's PPP-over-HDLC, and have "ethertype()" dissect the Ethernet type and the rest of the packet if it's a Linux-style header; dissect it ourselves only if it's an AF_ value. Have Wiretap impose a maximum packet size of 65535 bytes, so that it fails more gracefully when handed a corrupt "libpcap" capture file (other capture file formats with more than a 16-bit capture length field, if any, will have that check added later), and put that size in "wtap.h" and have Ethereal use it as its notion of a maximum packet size. Have Ethereal put up a "this file appears to be damaged or corrupt" message box if Wiretap returns a WTAP_ERR_BAD_RECORD error when opening or reading a capture file. Include loopback interfaces in the list of interfaces offered by the "Capture" dialog box, but put them at the end of the list so that it doesn't default to a loopback interface unless there are no other interfaces. Also, don't require that an interface in the list have an IP address associated with it, and only put one entry in the list for a given interface (SIOCGIFCONF returns one entry per interface *address*, not per *interface* - and even if you were to use only IP addresses, an interface could conceivably have more than one IP address). Exclusively use Wiretap encapsulation types internally, even when capturing; don't use DLT_ types. svn path=/trunk/; revision=540
1999-08-22 00:47:56 +00:00
guint packet_size;
Add support for capture files written by programs linked with Alexey Kuznetsov's modified "libpcap" *as long as you have the ss990915 or later patch*; the 990417 patch, alas, changes the per-packet header but *doesn't* change the magic number, so you can't just look at the magic number to see that it's Not Standard Libpcap. (Even more unfortunately, Red Hat appears to have picked up *that* patch for Red Hat 6.1; I've filed bug 6773 with Bugzilla on their site - hopefully, if I'm not misremembering the RH 6.1 code I've seen, and they really *did* pick up the older patch, they'll fix it ASAP to use the new magic number, and will make updates available.) svn path=/trunk/; revision=986
1999-11-06 08:42:01 +00:00
int bytes_to_read, bytes_read;
struct pcaprec_mutant_hdr hdr;
Add support to wiretap for reading Sun "snoop" capture files. That requires that, in the packet-reading loop, we pass to the callback routine the offset in the file of a packet's data, because we can no longer compute that offset by subtracting the size of the captured packet data from the offset in the file after the data was read - "snoop" may stick padding in after the packet data to align packet headers on 4-byte boundaries. Doing that required that we arrange that we do that for "libpcap" capture files as well; the cleanest way to do that was to write our own code for reading "libpcap" capture files, rather than using the "libpcap" code to do it. Make "wtap_dispatch_cb()" and "pcap_dispatch_cb()" static to "file.c", as they're not used elsewhere. If we're using wiretap, don't define in "file.h" stuff used only when we're not using wiretap. Update the wiretap README to reflect Gilbert's and my recent changes. Clean up some memory leaks in "wiretap/lanalyzer.c" and "wiretap/ngsniffer.c", where the capture-file-format-specific data wasn't freed if the open failed. svn path=/trunk/; revision=91
1998-11-15 05:29:17 +00:00
int data_offset;
/* Read record header. */
Have the per-capture-file-type open routines "wtap_open_offline()" calls return 1 on success, -1 if they got an error, and 0 if the file isn't of the type that file is checking for, and supply an error code if they return -1; have "wtap_open_offline()" use that error code. Also, have the per-capture-file-type open routines treat errors accessing the file as errors, and return -1, rather than just returning 0 so that we try another file type. Have the per-capture-file-type read routines "wtap_loop()" calls return -1 and supply an error code on error (and not, as they did in some cases, call "g_error()" and abort), and have "wtap_loop()", if the read routine returned an error, return FALSE (and pass an error-code-pointer argument onto the read routines, so they fill it in), and return TRUE on success. Add some new error codes for them to return. Now that "wtap_loop()" can return a success/failure indication and an error code, in "read_cap_file()" put up a message box if we get an error reading the file, and return the error code. Handle the additional errors we can get when opening a capture file. If the attempt to open a capture file succeeds, but the attempt to read it fails, don't treat that as a complete failure - we may have managed to read some of the capture file, and we should display what we managed to read. svn path=/trunk/; revision=516
1999-08-19 05:31:38 +00:00
errno = WTAP_ERR_CANT_READ;
Add support for capture files written by programs linked with Alexey Kuznetsov's modified "libpcap" *as long as you have the ss990915 or later patch*; the 990417 patch, alas, changes the per-packet header but *doesn't* change the magic number, so you can't just look at the magic number to see that it's Not Standard Libpcap. (Even more unfortunately, Red Hat appears to have picked up *that* patch for Red Hat 6.1; I've filed bug 6773 with Bugzilla on their site - hopefully, if I'm not misremembering the RH 6.1 code I've seen, and they really *did* pick up the older patch, they'll fix it ASAP to use the new magic number, and will make updates available.) svn path=/trunk/; revision=986
1999-11-06 08:42:01 +00:00
bytes_to_read = wth->capture.pcap->mutant ? sizeof hdr : sizeof hdr.hdr;
bytes_read = file_read(&hdr, 1, bytes_to_read, wth->fh);
if (bytes_read != bytes_to_read) {
Better handle errors from zlib: Assign a range of Wiretap errors for zlib errors, and have "wtap_strerror()" use "zError()" to get an error message for them. Have the internal "file_error()" routine return 0 for no error and a Wiretap error code for an error. svn path=/trunk/; revision=769
1999-10-05 07:06:08 +00:00
*err = file_error(wth->fh);
if (*err != 0)
Have the per-capture-file-type open routines "wtap_open_offline()" calls return 1 on success, -1 if they got an error, and 0 if the file isn't of the type that file is checking for, and supply an error code if they return -1; have "wtap_open_offline()" use that error code. Also, have the per-capture-file-type open routines treat errors accessing the file as errors, and return -1, rather than just returning 0 so that we try another file type. Have the per-capture-file-type read routines "wtap_loop()" calls return -1 and supply an error code on error (and not, as they did in some cases, call "g_error()" and abort), and have "wtap_loop()", if the read routine returned an error, return FALSE (and pass an error-code-pointer argument onto the read routines, so they fill it in), and return TRUE on success. Add some new error codes for them to return. Now that "wtap_loop()" can return a success/failure indication and an error code, in "read_cap_file()" put up a message box if we get an error reading the file, and return the error code. Handle the additional errors we can get when opening a capture file. If the attempt to open a capture file succeeds, but the attempt to read it fails, don't treat that as a complete failure - we may have managed to read some of the capture file, and we should display what we managed to read. svn path=/trunk/; revision=516
1999-08-19 05:31:38 +00:00
return -1;
Add support to wiretap for reading Sun "snoop" capture files. That requires that, in the packet-reading loop, we pass to the callback routine the offset in the file of a packet's data, because we can no longer compute that offset by subtracting the size of the captured packet data from the offset in the file after the data was read - "snoop" may stick padding in after the packet data to align packet headers on 4-byte boundaries. Doing that required that we arrange that we do that for "libpcap" capture files as well; the cleanest way to do that was to write our own code for reading "libpcap" capture files, rather than using the "libpcap" code to do it. Make "wtap_dispatch_cb()" and "pcap_dispatch_cb()" static to "file.c", as they're not used elsewhere. If we're using wiretap, don't define in "file.h" stuff used only when we're not using wiretap. Update the wiretap README to reflect Gilbert's and my recent changes. Clean up some memory leaks in "wiretap/lanalyzer.c" and "wiretap/ngsniffer.c", where the capture-file-format-specific data wasn't freed if the open failed. svn path=/trunk/; revision=91
1998-11-15 05:29:17 +00:00
if (bytes_read != 0) {
Have the per-capture-file-type open routines "wtap_open_offline()" calls return 1 on success, -1 if they got an error, and 0 if the file isn't of the type that file is checking for, and supply an error code if they return -1; have "wtap_open_offline()" use that error code. Also, have the per-capture-file-type open routines treat errors accessing the file as errors, and return -1, rather than just returning 0 so that we try another file type. Have the per-capture-file-type read routines "wtap_loop()" calls return -1 and supply an error code on error (and not, as they did in some cases, call "g_error()" and abort), and have "wtap_loop()", if the read routine returned an error, return FALSE (and pass an error-code-pointer argument onto the read routines, so they fill it in), and return TRUE on success. Add some new error codes for them to return. Now that "wtap_loop()" can return a success/failure indication and an error code, in "read_cap_file()" put up a message box if we get an error reading the file, and return the error code. Handle the additional errors we can get when opening a capture file. If the attempt to open a capture file succeeds, but the attempt to read it fails, don't treat that as a complete failure - we may have managed to read some of the capture file, and we should display what we managed to read. svn path=/trunk/; revision=516
1999-08-19 05:31:38 +00:00
*err = WTAP_ERR_SHORT_READ;
Add support to wiretap for reading Sun "snoop" capture files. That requires that, in the packet-reading loop, we pass to the callback routine the offset in the file of a packet's data, because we can no longer compute that offset by subtracting the size of the captured packet data from the offset in the file after the data was read - "snoop" may stick padding in after the packet data to align packet headers on 4-byte boundaries. Doing that required that we arrange that we do that for "libpcap" capture files as well; the cleanest way to do that was to write our own code for reading "libpcap" capture files, rather than using the "libpcap" code to do it. Make "wtap_dispatch_cb()" and "pcap_dispatch_cb()" static to "file.c", as they're not used elsewhere. If we're using wiretap, don't define in "file.h" stuff used only when we're not using wiretap. Update the wiretap README to reflect Gilbert's and my recent changes. Clean up some memory leaks in "wiretap/lanalyzer.c" and "wiretap/ngsniffer.c", where the capture-file-format-specific data wasn't freed if the open failed. svn path=/trunk/; revision=91
1998-11-15 05:29:17 +00:00
return -1;
}
return 0;
}
Add support for capture files written by programs linked with Alexey Kuznetsov's modified "libpcap" *as long as you have the ss990915 or later patch*; the 990417 patch, alas, changes the per-packet header but *doesn't* change the magic number, so you can't just look at the magic number to see that it's Not Standard Libpcap. (Even more unfortunately, Red Hat appears to have picked up *that* patch for Red Hat 6.1; I've filed bug 6773 with Bugzilla on their site - hopefully, if I'm not misremembering the RH 6.1 code I've seen, and they really *did* pick up the older patch, they'll fix it ASAP to use the new magic number, and will make updates available.) svn path=/trunk/; revision=986
1999-11-06 08:42:01 +00:00
wth->data_offset += bytes_read;
Add support to wiretap for reading Sun "snoop" capture files. That requires that, in the packet-reading loop, we pass to the callback routine the offset in the file of a packet's data, because we can no longer compute that offset by subtracting the size of the captured packet data from the offset in the file after the data was read - "snoop" may stick padding in after the packet data to align packet headers on 4-byte boundaries. Doing that required that we arrange that we do that for "libpcap" capture files as well; the cleanest way to do that was to write our own code for reading "libpcap" capture files, rather than using the "libpcap" code to do it. Make "wtap_dispatch_cb()" and "pcap_dispatch_cb()" static to "file.c", as they're not used elsewhere. If we're using wiretap, don't define in "file.h" stuff used only when we're not using wiretap. Update the wiretap README to reflect Gilbert's and my recent changes. Clean up some memory leaks in "wiretap/lanalyzer.c" and "wiretap/ngsniffer.c", where the capture-file-format-specific data wasn't freed if the open failed. svn path=/trunk/; revision=91
1998-11-15 05:29:17 +00:00
if (wth->capture.pcap->byte_swapped) {
/* Byte-swap the record header fields. */
Add support for capture files written by programs linked with Alexey Kuznetsov's modified "libpcap" *as long as you have the ss990915 or later patch*; the 990417 patch, alas, changes the per-packet header but *doesn't* change the magic number, so you can't just look at the magic number to see that it's Not Standard Libpcap. (Even more unfortunately, Red Hat appears to have picked up *that* patch for Red Hat 6.1; I've filed bug 6773 with Bugzilla on their site - hopefully, if I'm not misremembering the RH 6.1 code I've seen, and they really *did* pick up the older patch, they'll fix it ASAP to use the new magic number, and will make updates available.) svn path=/trunk/; revision=986
1999-11-06 08:42:01 +00:00
hdr.hdr.ts_sec = BSWAP32(hdr.hdr.ts_sec);
hdr.hdr.ts_usec = BSWAP32(hdr.hdr.ts_usec);
hdr.hdr.incl_len = BSWAP32(hdr.hdr.incl_len);
hdr.hdr.orig_len = BSWAP32(hdr.hdr.orig_len);
Add support to wiretap for reading Sun "snoop" capture files. That requires that, in the packet-reading loop, we pass to the callback routine the offset in the file of a packet's data, because we can no longer compute that offset by subtracting the size of the captured packet data from the offset in the file after the data was read - "snoop" may stick padding in after the packet data to align packet headers on 4-byte boundaries. Doing that required that we arrange that we do that for "libpcap" capture files as well; the cleanest way to do that was to write our own code for reading "libpcap" capture files, rather than using the "libpcap" code to do it. Make "wtap_dispatch_cb()" and "pcap_dispatch_cb()" static to "file.c", as they're not used elsewhere. If we're using wiretap, don't define in "file.h" stuff used only when we're not using wiretap. Update the wiretap README to reflect Gilbert's and my recent changes. Clean up some memory leaks in "wiretap/lanalyzer.c" and "wiretap/ngsniffer.c", where the capture-file-format-specific data wasn't freed if the open failed. svn path=/trunk/; revision=91
1998-11-15 05:29:17 +00:00
}
/* In file format version 2.3, the "incl_len" and "orig_len" fields
were swapped, in order to match the BPF header layout.
Unfortunately, some files were, according to a comment in the
"libpcap" source, written with version 2.3 in their headers
but without the interchanged fields, so if "incl_len" is
greater than "orig_len" - which would make no sense - we
assume that we need to swap them. */
if (wth->capture.pcap->version_major == 2 &&
(wth->capture.pcap->version_minor < 3 ||
(wth->capture.pcap->version_minor == 3 &&
Add support for capture files written by programs linked with Alexey Kuznetsov's modified "libpcap" *as long as you have the ss990915 or later patch*; the 990417 patch, alas, changes the per-packet header but *doesn't* change the magic number, so you can't just look at the magic number to see that it's Not Standard Libpcap. (Even more unfortunately, Red Hat appears to have picked up *that* patch for Red Hat 6.1; I've filed bug 6773 with Bugzilla on their site - hopefully, if I'm not misremembering the RH 6.1 code I've seen, and they really *did* pick up the older patch, they'll fix it ASAP to use the new magic number, and will make updates available.) svn path=/trunk/; revision=986
1999-11-06 08:42:01 +00:00
hdr.hdr.incl_len > hdr.hdr.orig_len))) {
Add support to wiretap for reading Sun "snoop" capture files. That requires that, in the packet-reading loop, we pass to the callback routine the offset in the file of a packet's data, because we can no longer compute that offset by subtracting the size of the captured packet data from the offset in the file after the data was read - "snoop" may stick padding in after the packet data to align packet headers on 4-byte boundaries. Doing that required that we arrange that we do that for "libpcap" capture files as well; the cleanest way to do that was to write our own code for reading "libpcap" capture files, rather than using the "libpcap" code to do it. Make "wtap_dispatch_cb()" and "pcap_dispatch_cb()" static to "file.c", as they're not used elsewhere. If we're using wiretap, don't define in "file.h" stuff used only when we're not using wiretap. Update the wiretap README to reflect Gilbert's and my recent changes. Clean up some memory leaks in "wiretap/lanalyzer.c" and "wiretap/ngsniffer.c", where the capture-file-format-specific data wasn't freed if the open failed. svn path=/trunk/; revision=91
1998-11-15 05:29:17 +00:00
guint32 temp;
Add support for capture files written by programs linked with Alexey Kuznetsov's modified "libpcap" *as long as you have the ss990915 or later patch*; the 990417 patch, alas, changes the per-packet header but *doesn't* change the magic number, so you can't just look at the magic number to see that it's Not Standard Libpcap. (Even more unfortunately, Red Hat appears to have picked up *that* patch for Red Hat 6.1; I've filed bug 6773 with Bugzilla on their site - hopefully, if I'm not misremembering the RH 6.1 code I've seen, and they really *did* pick up the older patch, they'll fix it ASAP to use the new magic number, and will make updates available.) svn path=/trunk/; revision=986
1999-11-06 08:42:01 +00:00
temp = hdr.hdr.orig_len;
hdr.hdr.orig_len = hdr.hdr.incl_len;
hdr.hdr.incl_len = temp;
Add support to wiretap for reading Sun "snoop" capture files. That requires that, in the packet-reading loop, we pass to the callback routine the offset in the file of a packet's data, because we can no longer compute that offset by subtracting the size of the captured packet data from the offset in the file after the data was read - "snoop" may stick padding in after the packet data to align packet headers on 4-byte boundaries. Doing that required that we arrange that we do that for "libpcap" capture files as well; the cleanest way to do that was to write our own code for reading "libpcap" capture files, rather than using the "libpcap" code to do it. Make "wtap_dispatch_cb()" and "pcap_dispatch_cb()" static to "file.c", as they're not used elsewhere. If we're using wiretap, don't define in "file.h" stuff used only when we're not using wiretap. Update the wiretap README to reflect Gilbert's and my recent changes. Clean up some memory leaks in "wiretap/lanalyzer.c" and "wiretap/ngsniffer.c", where the capture-file-format-specific data wasn't freed if the open failed. svn path=/trunk/; revision=91
1998-11-15 05:29:17 +00:00
}
Add support for capture files written by programs linked with Alexey Kuznetsov's modified "libpcap" *as long as you have the ss990915 or later patch*; the 990417 patch, alas, changes the per-packet header but *doesn't* change the magic number, so you can't just look at the magic number to see that it's Not Standard Libpcap. (Even more unfortunately, Red Hat appears to have picked up *that* patch for Red Hat 6.1; I've filed bug 6773 with Bugzilla on their site - hopefully, if I'm not misremembering the RH 6.1 code I've seen, and they really *did* pick up the older patch, they'll fix it ASAP to use the new magic number, and will make updates available.) svn path=/trunk/; revision=986
1999-11-06 08:42:01 +00:00
packet_size = hdr.hdr.incl_len;
DLT_NULL, from "libpcap", means different things on different platforms and in different capture files; throw in some heuristics to try to figure out whether the 4-byte header is: 1) PPP-over-HDLC (some version of ISDN4BSD?); 2) big-endian AF_ value (BSD on big-endian platforms); 3) little-endian AF_ value (BSD on little-endian platforms); 4) two octets of 0 followed by an Ethernet type (Linux, at least on little-endian platforms, as mutated by "libpcap"). Make a separate Wiretap encapsulation type, WTAP_ENCAP_NULL, corresponding to DLT_NULL. Have the PPP code dissect the frame if it's PPP-over-HDLC, and have "ethertype()" dissect the Ethernet type and the rest of the packet if it's a Linux-style header; dissect it ourselves only if it's an AF_ value. Have Wiretap impose a maximum packet size of 65535 bytes, so that it fails more gracefully when handed a corrupt "libpcap" capture file (other capture file formats with more than a 16-bit capture length field, if any, will have that check added later), and put that size in "wtap.h" and have Ethereal use it as its notion of a maximum packet size. Have Ethereal put up a "this file appears to be damaged or corrupt" message box if Wiretap returns a WTAP_ERR_BAD_RECORD error when opening or reading a capture file. Include loopback interfaces in the list of interfaces offered by the "Capture" dialog box, but put them at the end of the list so that it doesn't default to a loopback interface unless there are no other interfaces. Also, don't require that an interface in the list have an IP address associated with it, and only put one entry in the list for a given interface (SIOCGIFCONF returns one entry per interface *address*, not per *interface* - and even if you were to use only IP addresses, an interface could conceivably have more than one IP address). Exclusively use Wiretap encapsulation types internally, even when capturing; don't use DLT_ types. svn path=/trunk/; revision=540
1999-08-22 00:47:56 +00:00
if (packet_size > WTAP_MAX_PACKET_SIZE) {
/*
* Probably a corrupt capture file; don't blow up trying
* to allocate space for an immensely-large packet.
*/
g_message("pcap: File has %u-byte packet, bigger than maximum of %u",
packet_size, WTAP_MAX_PACKET_SIZE);
*err = WTAP_ERR_BAD_RECORD;
return -1;
}
Added display filters to wiretap. svn path=/trunk/; revision=198
1999-03-01 18:57:07 +00:00
buffer_assure_space(wth->frame_buffer, packet_size);
Keep in the "wtap" structure the current offset into the file being read, and maintain it ourselves as we read through the file, rather than calling "ftell()" for every packet we read - "ftell()" may involve an "lseek()" call, which could add a noticeable CPU overhead when reading a large file. svn path=/trunk/; revision=596
1999-08-28 01:19:45 +00:00
data_offset = wth->data_offset;
Have the per-capture-file-type open routines "wtap_open_offline()" calls return 1 on success, -1 if they got an error, and 0 if the file isn't of the type that file is checking for, and supply an error code if they return -1; have "wtap_open_offline()" use that error code. Also, have the per-capture-file-type open routines treat errors accessing the file as errors, and return -1, rather than just returning 0 so that we try another file type. Have the per-capture-file-type read routines "wtap_loop()" calls return -1 and supply an error code on error (and not, as they did in some cases, call "g_error()" and abort), and have "wtap_loop()", if the read routine returned an error, return FALSE (and pass an error-code-pointer argument onto the read routines, so they fill it in), and return TRUE on success. Add some new error codes for them to return. Now that "wtap_loop()" can return a success/failure indication and an error code, in "read_cap_file()" put up a message box if we get an error reading the file, and return the error code. Handle the additional errors we can get when opening a capture file. If the attempt to open a capture file succeeds, but the attempt to read it fails, don't treat that as a complete failure - we may have managed to read some of the capture file, and we should display what we managed to read. svn path=/trunk/; revision=516
1999-08-19 05:31:38 +00:00
errno = WTAP_ERR_CANT_READ;
This commit contains support for reading capture files compressed using gzip. The zLib library is used for this purpose. If zLib is not available (or it's use is disabled by the --disable-zlib option to configure), you can still compile Ethereal but it will be unable to read compressed capture files. IMPORTANT: Now all file accesses to capture files should be done through special macros. Specifically, for any use of the following functions on capture files, replace them. The arguments for the right-side functions are exactly the same as for the original stdio functions. fopen file_open fdopen filed_open fread file_read fwrite file_write fseek file_seek fclose file_close ferror file_error svn path=/trunk/; revision=695
1999-09-22 01:26:50 +00:00
bytes_read = file_read(buffer_start_ptr(wth->frame_buffer), 1,
Add support to wiretap for reading Sun "snoop" capture files. That requires that, in the packet-reading loop, we pass to the callback routine the offset in the file of a packet's data, because we can no longer compute that offset by subtracting the size of the captured packet data from the offset in the file after the data was read - "snoop" may stick padding in after the packet data to align packet headers on 4-byte boundaries. Doing that required that we arrange that we do that for "libpcap" capture files as well; the cleanest way to do that was to write our own code for reading "libpcap" capture files, rather than using the "libpcap" code to do it. Make "wtap_dispatch_cb()" and "pcap_dispatch_cb()" static to "file.c", as they're not used elsewhere. If we're using wiretap, don't define in "file.h" stuff used only when we're not using wiretap. Update the wiretap README to reflect Gilbert's and my recent changes. Clean up some memory leaks in "wiretap/lanalyzer.c" and "wiretap/ngsniffer.c", where the capture-file-format-specific data wasn't freed if the open failed. svn path=/trunk/; revision=91
1998-11-15 05:29:17 +00:00
packet_size, wth->fh);
if (bytes_read != packet_size) {
Better handle errors from zlib: Assign a range of Wiretap errors for zlib errors, and have "wtap_strerror()" use "zError()" to get an error message for them. Have the internal "file_error()" routine return 0 for no error and a Wiretap error code for an error. svn path=/trunk/; revision=769
1999-10-05 07:06:08 +00:00
*err = file_error(wth->fh);
if (*err == 0)
Have the per-capture-file-type open routines "wtap_open_offline()" calls return 1 on success, -1 if they got an error, and 0 if the file isn't of the type that file is checking for, and supply an error code if they return -1; have "wtap_open_offline()" use that error code. Also, have the per-capture-file-type open routines treat errors accessing the file as errors, and return -1, rather than just returning 0 so that we try another file type. Have the per-capture-file-type read routines "wtap_loop()" calls return -1 and supply an error code on error (and not, as they did in some cases, call "g_error()" and abort), and have "wtap_loop()", if the read routine returned an error, return FALSE (and pass an error-code-pointer argument onto the read routines, so they fill it in), and return TRUE on success. Add some new error codes for them to return. Now that "wtap_loop()" can return a success/failure indication and an error code, in "read_cap_file()" put up a message box if we get an error reading the file, and return the error code. Handle the additional errors we can get when opening a capture file. If the attempt to open a capture file succeeds, but the attempt to read it fails, don't treat that as a complete failure - we may have managed to read some of the capture file, and we should display what we managed to read. svn path=/trunk/; revision=516
1999-08-19 05:31:38 +00:00
*err = WTAP_ERR_SHORT_READ;
Add support to wiretap for reading Sun "snoop" capture files. That requires that, in the packet-reading loop, we pass to the callback routine the offset in the file of a packet's data, because we can no longer compute that offset by subtracting the size of the captured packet data from the offset in the file after the data was read - "snoop" may stick padding in after the packet data to align packet headers on 4-byte boundaries. Doing that required that we arrange that we do that for "libpcap" capture files as well; the cleanest way to do that was to write our own code for reading "libpcap" capture files, rather than using the "libpcap" code to do it. Make "wtap_dispatch_cb()" and "pcap_dispatch_cb()" static to "file.c", as they're not used elsewhere. If we're using wiretap, don't define in "file.h" stuff used only when we're not using wiretap. Update the wiretap README to reflect Gilbert's and my recent changes. Clean up some memory leaks in "wiretap/lanalyzer.c" and "wiretap/ngsniffer.c", where the capture-file-format-specific data wasn't freed if the open failed. svn path=/trunk/; revision=91
1998-11-15 05:29:17 +00:00
return -1;
}
Keep in the "wtap" structure the current offset into the file being read, and maintain it ourselves as we read through the file, rather than calling "ftell()" for every packet we read - "ftell()" may involve an "lseek()" call, which could add a noticeable CPU overhead when reading a large file. svn path=/trunk/; revision=596
1999-08-28 01:19:45 +00:00
wth->data_offset += packet_size;
Add support to wiretap for reading Sun "snoop" capture files. That requires that, in the packet-reading loop, we pass to the callback routine the offset in the file of a packet's data, because we can no longer compute that offset by subtracting the size of the captured packet data from the offset in the file after the data was read - "snoop" may stick padding in after the packet data to align packet headers on 4-byte boundaries. Doing that required that we arrange that we do that for "libpcap" capture files as well; the cleanest way to do that was to write our own code for reading "libpcap" capture files, rather than using the "libpcap" code to do it. Make "wtap_dispatch_cb()" and "pcap_dispatch_cb()" static to "file.c", as they're not used elsewhere. If we're using wiretap, don't define in "file.h" stuff used only when we're not using wiretap. Update the wiretap README to reflect Gilbert's and my recent changes. Clean up some memory leaks in "wiretap/lanalyzer.c" and "wiretap/ngsniffer.c", where the capture-file-format-specific data wasn't freed if the open failed. svn path=/trunk/; revision=91
1998-11-15 05:29:17 +00:00
Add support for capture files written by programs linked with Alexey Kuznetsov's modified "libpcap" *as long as you have the ss990915 or later patch*; the 990417 patch, alas, changes the per-packet header but *doesn't* change the magic number, so you can't just look at the magic number to see that it's Not Standard Libpcap. (Even more unfortunately, Red Hat appears to have picked up *that* patch for Red Hat 6.1; I've filed bug 6773 with Bugzilla on their site - hopefully, if I'm not misremembering the RH 6.1 code I've seen, and they really *did* pick up the older patch, they'll fix it ASAP to use the new magic number, and will make updates available.) svn path=/trunk/; revision=986
1999-11-06 08:42:01 +00:00
wth->phdr.ts.tv_sec = hdr.hdr.ts_sec;
wth->phdr.ts.tv_usec = hdr.hdr.ts_usec;
Add support to wiretap for reading Sun "snoop" capture files. That requires that, in the packet-reading loop, we pass to the callback routine the offset in the file of a packet's data, because we can no longer compute that offset by subtracting the size of the captured packet data from the offset in the file after the data was read - "snoop" may stick padding in after the packet data to align packet headers on 4-byte boundaries. Doing that required that we arrange that we do that for "libpcap" capture files as well; the cleanest way to do that was to write our own code for reading "libpcap" capture files, rather than using the "libpcap" code to do it. Make "wtap_dispatch_cb()" and "pcap_dispatch_cb()" static to "file.c", as they're not used elsewhere. If we're using wiretap, don't define in "file.h" stuff used only when we're not using wiretap. Update the wiretap README to reflect Gilbert's and my recent changes. Clean up some memory leaks in "wiretap/lanalyzer.c" and "wiretap/ngsniffer.c", where the capture-file-format-specific data wasn't freed if the open failed. svn path=/trunk/; revision=91
1998-11-15 05:29:17 +00:00
wth->phdr.caplen = packet_size;
Add support for capture files written by programs linked with Alexey Kuznetsov's modified "libpcap" *as long as you have the ss990915 or later patch*; the 990417 patch, alas, changes the per-packet header but *doesn't* change the magic number, so you can't just look at the magic number to see that it's Not Standard Libpcap. (Even more unfortunately, Red Hat appears to have picked up *that* patch for Red Hat 6.1; I've filed bug 6773 with Bugzilla on their site - hopefully, if I'm not misremembering the RH 6.1 code I've seen, and they really *did* pick up the older patch, they'll fix it ASAP to use the new magic number, and will make updates available.) svn path=/trunk/; revision=986
1999-11-06 08:42:01 +00:00
wth->phdr.len = hdr.hdr.orig_len;
Added display filters to wiretap. svn path=/trunk/; revision=198
1999-03-01 18:57:07 +00:00
wth->phdr.pkt_encap = wth->file_encap;
Add support to wiretap for reading Sun "snoop" capture files. That requires that, in the packet-reading loop, we pass to the callback routine the offset in the file of a packet's data, because we can no longer compute that offset by subtracting the size of the captured packet data from the offset in the file after the data was read - "snoop" may stick padding in after the packet data to align packet headers on 4-byte boundaries. Doing that required that we arrange that we do that for "libpcap" capture files as well; the cleanest way to do that was to write our own code for reading "libpcap" capture files, rather than using the "libpcap" code to do it. Make "wtap_dispatch_cb()" and "pcap_dispatch_cb()" static to "file.c", as they're not used elsewhere. If we're using wiretap, don't define in "file.h" stuff used only when we're not using wiretap. Update the wiretap README to reflect Gilbert's and my recent changes. Clean up some memory leaks in "wiretap/lanalyzer.c" and "wiretap/ngsniffer.c", where the capture-file-format-specific data wasn't freed if the open failed. svn path=/trunk/; revision=91
1998-11-15 05:29:17 +00:00
return data_offset;
}
Add to Wiretap the ability to write capture files; for now, it can only write them in "libpcap" format, but the mechanism can have other formats added. When creating the temporary file for a capture, use "create_tempfile()", to close a security hole opened by the fact that "tempnam()" creates a temporary file, but doesn't open it, and we open the file with the name it gives us - somebody could remove the file and plant a link to some file, and, if as may well be the case when Ethereal is capturing packets, it's running as "root", that means we write a capture on top of that file.... (The aforementioned changes to Wiretap let you open a capture file for writing given an file descriptor, "fdopen()"-style, which this change requires.) svn path=/trunk/; revision=509
1999-08-18 04:17:38 +00:00
int wtap_pcap_encap_to_wtap_encap(int encap)
{
if (encap < 0 || encap >= NUM_PCAP_ENCAPS)
"wtap_pcap_encap_to_wtap_encap()" shouldn't return a file type if it can't translate the encapsulation type, it should return an encapsulation type; we add a new one, WTAP_ENCAP_UNKNOWN. and have it return that. Have "capture()" handle "wtap_pcap_encap_to_wtap_encap()" returning that encapsulation type (if it happens, we need to add a new Wiretap encapsulation type to handle the new "libpcap" encapsulation type). svn path=/trunk/; revision=513
1999-08-18 17:08:47 +00:00
return WTAP_ENCAP_UNKNOWN;
Add to Wiretap the ability to write capture files; for now, it can only write them in "libpcap" format, but the mechanism can have other formats added. When creating the temporary file for a capture, use "create_tempfile()", to close a security hole opened by the fact that "tempnam()" creates a temporary file, but doesn't open it, and we open the file with the name it gives us - somebody could remove the file and plant a link to some file, and, if as may well be the case when Ethereal is capturing packets, it's running as "root", that means we write a capture on top of that file.... (The aforementioned changes to Wiretap let you open a capture file for writing given an file descriptor, "fdopen()"-style, which this change requires.) svn path=/trunk/; revision=509
1999-08-18 04:17:38 +00:00
return pcap_encap[encap];
}
/* Returns 1 on success, 0 on failure; sets "*err" to an error code on
failure */
int libpcap_dump_open(wtap_dumper *wdh, int *err)
{
static const guint32 pcap_magic = PCAP_MAGIC;
struct pcap_hdr file_hdr;
static const int wtap_encap[] = {
Fix the pcap-encapsulation-to-wiretap-encapsulation mapping table. Have the code that opens "libpcap" files for writing check to make sure that the Wiretap encapsulation can be written to a "libpcap" file, and return -1 and supply a new WTAP_ERR_UNSUPPORTED_ENCAP error code if it can't. Handle that new error code in "wtap_strerror()". svn path=/trunk/; revision=546
1999-08-22 03:50:31 +00:00
-1, /* WTAP_ENCAP_UNKNOWN -> unsupported */
Get rid of some cruft left in by previous checkins as placeholders. Get rid of WTAP_ENCAP_NONE; replace it with WTAP_ENCAP_UNKNOWN, which means "I can't handle that file, it's using an encapsulation I don't support". Check for encapsulations we don't support, and return an error (as is already done in "libpcap.c"). Check for too-large packet sizes, and return an error (as is already done in "libpcap.c"). Print unsigned quantities in Wiretap messages with "%u", not "%d". svn path=/trunk/; revision=544
1999-08-22 02:29:40 +00:00
1, /* WTAP_ENCAP_ETHERNET -> DLT_EN10MB */
6, /* WTAP_ENCAP_TR -> DLT_IEEE802 */
8, /* WTAP_ENCAP_SLIP -> DLT_SLIP */
9, /* WTAP_ENCAP_PPP -> DLT_PPP */
10, /* WTAP_ENCAP_FDDI -> DLT_FDDI */
Add a new Wiretap encapsulation type WTAP_ENCAP_FDDI_BITSWAPPED, meaning "FDDI with the MAC addresses bit-swapped"; whether the MAC addresses are bit-swapped is a property of the machine on which the capture was taken, not of the machine on which the capture is being read - right now, none of the capture file formats we read indicate whether FDDI MAC addresses are bit-swapped, but this does let us treat non-"libpcap" captures as being bit-swapped or not bit-swapped independent of the machine on which they're being read (and of the machine on which they were captured, but I have the impression they're bit-swapped on most platforms), and allows us to, if, as, and when we implement packet capture in Wiretap, mark packets in a capture file written in Wiretap-native format based on the machine on which they are captured (assuming the rule "Ultrix, Alpha, and BSD/OS are the only platforms that don't bit-swap", or some other compile-time rule, gets the right answer, or that some platform has drivers that can tell us whether the addresses are bit-swapped). (NOTE: if, for any of the capture file formats used only on one platform, FDDI MAC addresses aren't bit-swapped, the code to read that capture file format should be fixed to flag them as not bit-swapped.) Use the encapsulation type to decide whether to bit-swap addresses in "dissect_fddi()". svn path=/trunk/; revision=557
1999-08-24 03:19:34 +00:00
10, /* WTAP_ENCAP_FDDI_BITSWAPPED -> DLT_FDDI */
Get rid of some cruft left in by previous checkins as placeholders. Get rid of WTAP_ENCAP_NONE; replace it with WTAP_ENCAP_UNKNOWN, which means "I can't handle that file, it's using an encapsulation I don't support". Check for encapsulations we don't support, and return an error (as is already done in "libpcap.c"). Check for too-large packet sizes, and return an error (as is already done in "libpcap.c"). Print unsigned quantities in Wiretap messages with "%u", not "%d". svn path=/trunk/; revision=544
1999-08-22 02:29:40 +00:00
12, /* WTAP_ENCAP_RAW_IP -> DLT_RAW */
7, /* WTAP_ENCAP_ARCNET -> DLT_ARCNET */
11, /* WTAP_ENCAP_ATM_RFC1483 -> DLT_ATM_RFC1483 */
19, /* WTAP_ENCAP_LINUX_ATM_CLIP */
Fix the pcap-encapsulation-to-wiretap-encapsulation mapping table. Have the code that opens "libpcap" files for writing check to make sure that the Wiretap encapsulation can be written to a "libpcap" file, and return -1 and supply a new WTAP_ERR_UNSUPPORTED_ENCAP error code if it can't. Handle that new error code in "wtap_strerror()". svn path=/trunk/; revision=546
1999-08-22 03:50:31 +00:00
-1, /* WTAP_ENCAP_LAPB -> unsupported*/
-1, /* WTAP_ENCAP_ATM_SNIFFER -> unsupported */
Get rid of some cruft left in by previous checkins as placeholders. Get rid of WTAP_ENCAP_NONE; replace it with WTAP_ENCAP_UNKNOWN, which means "I can't handle that file, it's using an encapsulation I don't support". Check for encapsulations we don't support, and return an error (as is already done in "libpcap.c"). Check for too-large packet sizes, and return an error (as is already done in "libpcap.c"). Print unsigned quantities in Wiretap messages with "%u", not "%d". svn path=/trunk/; revision=544
1999-08-22 02:29:40 +00:00
0 /* WTAP_ENCAP_NULL -> DLT_NULL */
Add to Wiretap the ability to write capture files; for now, it can only write them in "libpcap" format, but the mechanism can have other formats added. When creating the temporary file for a capture, use "create_tempfile()", to close a security hole opened by the fact that "tempnam()" creates a temporary file, but doesn't open it, and we open the file with the name it gives us - somebody could remove the file and plant a link to some file, and, if as may well be the case when Ethereal is capturing packets, it's running as "root", that means we write a capture on top of that file.... (The aforementioned changes to Wiretap let you open a capture file for writing given an file descriptor, "fdopen()"-style, which this change requires.) svn path=/trunk/; revision=509
1999-08-18 04:17:38 +00:00
};
Fix the pcap-encapsulation-to-wiretap-encapsulation mapping table. Have the code that opens "libpcap" files for writing check to make sure that the Wiretap encapsulation can be written to a "libpcap" file, and return -1 and supply a new WTAP_ERR_UNSUPPORTED_ENCAP error code if it can't. Handle that new error code in "wtap_strerror()". svn path=/trunk/; revision=546
1999-08-22 03:50:31 +00:00
#define NUM_WTAP_ENCAPS (sizeof wtap_encap / sizeof wtap_encap[0])
Add to Wiretap the ability to write capture files; for now, it can only write them in "libpcap" format, but the mechanism can have other formats added. When creating the temporary file for a capture, use "create_tempfile()", to close a security hole opened by the fact that "tempnam()" creates a temporary file, but doesn't open it, and we open the file with the name it gives us - somebody could remove the file and plant a link to some file, and, if as may well be the case when Ethereal is capturing packets, it's running as "root", that means we write a capture on top of that file.... (The aforementioned changes to Wiretap let you open a capture file for writing given an file descriptor, "fdopen()"-style, which this change requires.) svn path=/trunk/; revision=509
1999-08-18 04:17:38 +00:00
int nwritten;
/* Per-packet encapsulations aren't supported. */
if (wdh->encap == WTAP_ENCAP_PER_PACKET) {
*err = WTAP_ERR_ENCAP_PER_PACKET_UNSUPPORTED;
return 0;
}
Fix the pcap-encapsulation-to-wiretap-encapsulation mapping table. Have the code that opens "libpcap" files for writing check to make sure that the Wiretap encapsulation can be written to a "libpcap" file, and return -1 and supply a new WTAP_ERR_UNSUPPORTED_ENCAP error code if it can't. Handle that new error code in "wtap_strerror()". svn path=/trunk/; revision=546
1999-08-22 03:50:31 +00:00
if (wdh->encap < 0 || wdh->encap >= NUM_WTAP_ENCAPS
|| wtap_encap[wdh->encap] == -1) {
*err = WTAP_ERR_UNSUPPORTED_ENCAP;
return 0;
}
Add to Wiretap the ability to write capture files; for now, it can only write them in "libpcap" format, but the mechanism can have other formats added. When creating the temporary file for a capture, use "create_tempfile()", to close a security hole opened by the fact that "tempnam()" creates a temporary file, but doesn't open it, and we open the file with the name it gives us - somebody could remove the file and plant a link to some file, and, if as may well be the case when Ethereal is capturing packets, it's running as "root", that means we write a capture on top of that file.... (The aforementioned changes to Wiretap let you open a capture file for writing given an file descriptor, "fdopen()"-style, which this change requires.) svn path=/trunk/; revision=509
1999-08-18 04:17:38 +00:00
/* This is a libpcap file */
wdh->subtype_write = libpcap_dump;
wdh->subtype_close = libpcap_dump_close;
/* Write the file header. */
When reading a capture file, we can detect whether it's compressed or not, so it's OK to use "zlib" to read capture files, as it handles uncompressed files correctly. When *writing* capture files, however, we can't detect automatically whether the user wanted to write the file out as a compressed file or not, so we should *NOT* use "zlib" until we add a flag to the API specifying whether to write the file out as a compressed file or not. Furthermore, the code in Ethereal that implements the "-S" flag depends on being able to get the "FILE *" for a capture file being written, so that it can "fflush()" it. svn path=/trunk/; revision=703
1999-09-23 05:00:59 +00:00
nwritten = fwrite(&pcap_magic, 1, sizeof pcap_magic, wdh->fh);
Add to Wiretap the ability to write capture files; for now, it can only write them in "libpcap" format, but the mechanism can have other formats added. When creating the temporary file for a capture, use "create_tempfile()", to close a security hole opened by the fact that "tempnam()" creates a temporary file, but doesn't open it, and we open the file with the name it gives us - somebody could remove the file and plant a link to some file, and, if as may well be the case when Ethereal is capturing packets, it's running as "root", that means we write a capture on top of that file.... (The aforementioned changes to Wiretap let you open a capture file for writing given an file descriptor, "fdopen()"-style, which this change requires.) svn path=/trunk/; revision=509
1999-08-18 04:17:38 +00:00
if (nwritten != sizeof pcap_magic) {
if (nwritten < 0)
*err = errno;
else
*err = WTAP_ERR_SHORT_WRITE;
return 0;
}
/* current "libpcap" format is 2.4 */
file_hdr.version_major = 2;
file_hdr.version_minor = 4;
file_hdr.thiszone = 0; /* XXX - current offset? */
file_hdr.sigfigs = 0; /* unknown, but also apparently unused */
file_hdr.snaplen = wdh->snaplen;
file_hdr.network = wtap_encap[wdh->encap];
When reading a capture file, we can detect whether it's compressed or not, so it's OK to use "zlib" to read capture files, as it handles uncompressed files correctly. When *writing* capture files, however, we can't detect automatically whether the user wanted to write the file out as a compressed file or not, so we should *NOT* use "zlib" until we add a flag to the API specifying whether to write the file out as a compressed file or not. Furthermore, the code in Ethereal that implements the "-S" flag depends on being able to get the "FILE *" for a capture file being written, so that it can "fflush()" it. svn path=/trunk/; revision=703
1999-09-23 05:00:59 +00:00
nwritten = fwrite(&file_hdr, 1, sizeof file_hdr, wdh->fh);
Add to Wiretap the ability to write capture files; for now, it can only write them in "libpcap" format, but the mechanism can have other formats added. When creating the temporary file for a capture, use "create_tempfile()", to close a security hole opened by the fact that "tempnam()" creates a temporary file, but doesn't open it, and we open the file with the name it gives us - somebody could remove the file and plant a link to some file, and, if as may well be the case when Ethereal is capturing packets, it's running as "root", that means we write a capture on top of that file.... (The aforementioned changes to Wiretap let you open a capture file for writing given an file descriptor, "fdopen()"-style, which this change requires.) svn path=/trunk/; revision=509
1999-08-18 04:17:38 +00:00
if (nwritten != sizeof file_hdr) {
if (nwritten < 0)
*err = errno;
else
*err = WTAP_ERR_SHORT_WRITE;
return 0;
}
return 1;
}
/* Write a record for a packet to a dump file.
Returns 1 on success, 0 on failure. */
static int libpcap_dump(wtap_dumper *wdh, const struct wtap_pkthdr *phdr,
Make "wtap_dump()" and "wtap_dump_close()" return error codes, and check for errors when closing a file to which we've written packets (we don't bother checking if we're giving up on a capture). Add some more error checks in Wiretap. Make a single list of all Wiretap error codes, giving them all different values (some can be returned by more than one routine, so they shouldn't be per-routine). svn path=/trunk/; revision=510
1999-08-18 04:41:20 +00:00
const u_char *pd, int *err)
Add to Wiretap the ability to write capture files; for now, it can only write them in "libpcap" format, but the mechanism can have other formats added. When creating the temporary file for a capture, use "create_tempfile()", to close a security hole opened by the fact that "tempnam()" creates a temporary file, but doesn't open it, and we open the file with the name it gives us - somebody could remove the file and plant a link to some file, and, if as may well be the case when Ethereal is capturing packets, it's running as "root", that means we write a capture on top of that file.... (The aforementioned changes to Wiretap let you open a capture file for writing given an file descriptor, "fdopen()"-style, which this change requires.) svn path=/trunk/; revision=509
1999-08-18 04:17:38 +00:00
{
struct pcaprec_hdr rec_hdr;
int nwritten;
rec_hdr.ts_sec = phdr->ts.tv_sec;
rec_hdr.ts_usec = phdr->ts.tv_usec;
rec_hdr.incl_len = phdr->caplen;
rec_hdr.orig_len = phdr->len;
When reading a capture file, we can detect whether it's compressed or not, so it's OK to use "zlib" to read capture files, as it handles uncompressed files correctly. When *writing* capture files, however, we can't detect automatically whether the user wanted to write the file out as a compressed file or not, so we should *NOT* use "zlib" until we add a flag to the API specifying whether to write the file out as a compressed file or not. Furthermore, the code in Ethereal that implements the "-S" flag depends on being able to get the "FILE *" for a capture file being written, so that it can "fflush()" it. svn path=/trunk/; revision=703
1999-09-23 05:00:59 +00:00
nwritten = fwrite(&rec_hdr, 1, sizeof rec_hdr, wdh->fh);
Make "wtap_dump()" and "wtap_dump_close()" return error codes, and check for errors when closing a file to which we've written packets (we don't bother checking if we're giving up on a capture). Add some more error checks in Wiretap. Make a single list of all Wiretap error codes, giving them all different values (some can be returned by more than one routine, so they shouldn't be per-routine). svn path=/trunk/; revision=510
1999-08-18 04:41:20 +00:00
if (nwritten != sizeof rec_hdr) {
if (nwritten < 0)
*err = errno;
else
*err = WTAP_ERR_SHORT_WRITE;
return 0;
}
When reading a capture file, we can detect whether it's compressed or not, so it's OK to use "zlib" to read capture files, as it handles uncompressed files correctly. When *writing* capture files, however, we can't detect automatically whether the user wanted to write the file out as a compressed file or not, so we should *NOT* use "zlib" until we add a flag to the API specifying whether to write the file out as a compressed file or not. Furthermore, the code in Ethereal that implements the "-S" flag depends on being able to get the "FILE *" for a capture file being written, so that it can "fflush()" it. svn path=/trunk/; revision=703
1999-09-23 05:00:59 +00:00
nwritten = fwrite(pd, 1, phdr->caplen, wdh->fh);
Make "wtap_dump()" and "wtap_dump_close()" return error codes, and check for errors when closing a file to which we've written packets (we don't bother checking if we're giving up on a capture). Add some more error checks in Wiretap. Make a single list of all Wiretap error codes, giving them all different values (some can be returned by more than one routine, so they shouldn't be per-routine). svn path=/trunk/; revision=510
1999-08-18 04:41:20 +00:00
if (nwritten != phdr->caplen) {
if (nwritten < 0)
*err = errno;
else
*err = WTAP_ERR_SHORT_WRITE;
return 0;
}
Add to Wiretap the ability to write capture files; for now, it can only write them in "libpcap" format, but the mechanism can have other formats added. When creating the temporary file for a capture, use "create_tempfile()", to close a security hole opened by the fact that "tempnam()" creates a temporary file, but doesn't open it, and we open the file with the name it gives us - somebody could remove the file and plant a link to some file, and, if as may well be the case when Ethereal is capturing packets, it's running as "root", that means we write a capture on top of that file.... (The aforementioned changes to Wiretap let you open a capture file for writing given an file descriptor, "fdopen()"-style, which this change requires.) svn path=/trunk/; revision=509
1999-08-18 04:17:38 +00:00
return 1;
}
Make "wtap_dump()" and "wtap_dump_close()" return error codes, and check for errors when closing a file to which we've written packets (we don't bother checking if we're giving up on a capture). Add some more error checks in Wiretap. Make a single list of all Wiretap error codes, giving them all different values (some can be returned by more than one routine, so they shouldn't be per-routine). svn path=/trunk/; revision=510
1999-08-18 04:41:20 +00:00
/* Finish writing to a dump file.
Add to Wiretap the ability to write capture files; for now, it can only write them in "libpcap" format, but the mechanism can have other formats added. When creating the temporary file for a capture, use "create_tempfile()", to close a security hole opened by the fact that "tempnam()" creates a temporary file, but doesn't open it, and we open the file with the name it gives us - somebody could remove the file and plant a link to some file, and, if as may well be the case when Ethereal is capturing packets, it's running as "root", that means we write a capture on top of that file.... (The aforementioned changes to Wiretap let you open a capture file for writing given an file descriptor, "fdopen()"-style, which this change requires.) svn path=/trunk/; revision=509
1999-08-18 04:17:38 +00:00
Returns 1 on success, 0 on failure. */
Make "wtap_dump()" and "wtap_dump_close()" return error codes, and check for errors when closing a file to which we've written packets (we don't bother checking if we're giving up on a capture). Add some more error checks in Wiretap. Make a single list of all Wiretap error codes, giving them all different values (some can be returned by more than one routine, so they shouldn't be per-routine). svn path=/trunk/; revision=510
1999-08-18 04:41:20 +00:00
static int libpcap_dump_close(wtap_dumper *wdh, int *err)
Add to Wiretap the ability to write capture files; for now, it can only write them in "libpcap" format, but the mechanism can have other formats added. When creating the temporary file for a capture, use "create_tempfile()", to close a security hole opened by the fact that "tempnam()" creates a temporary file, but doesn't open it, and we open the file with the name it gives us - somebody could remove the file and plant a link to some file, and, if as may well be the case when Ethereal is capturing packets, it's running as "root", that means we write a capture on top of that file.... (The aforementioned changes to Wiretap let you open a capture file for writing given an file descriptor, "fdopen()"-style, which this change requires.) svn path=/trunk/; revision=509
1999-08-18 04:17:38 +00:00
{
Make "wtap_dump()" and "wtap_dump_close()" return error codes, and check for errors when closing a file to which we've written packets (we don't bother checking if we're giving up on a capture). Add some more error checks in Wiretap. Make a single list of all Wiretap error codes, giving them all different values (some can be returned by more than one routine, so they shouldn't be per-routine). svn path=/trunk/; revision=510
1999-08-18 04:41:20 +00:00
/* Nothing to do here. */
Add to Wiretap the ability to write capture files; for now, it can only write them in "libpcap" format, but the mechanism can have other formats added. When creating the temporary file for a capture, use "create_tempfile()", to close a security hole opened by the fact that "tempnam()" creates a temporary file, but doesn't open it, and we open the file with the name it gives us - somebody could remove the file and plant a link to some file, and, if as may well be the case when Ethereal is capturing packets, it's running as "root", that means we write a capture on top of that file.... (The aforementioned changes to Wiretap let you open a capture file for writing given an file descriptor, "fdopen()"-style, which this change requires.) svn path=/trunk/; revision=509
1999-08-18 04:17:38 +00:00
return 1;
}