[prev in list] [next in list] [prev in thread] [next in thread]
List: ethereal-users
Subject: Re: [ethereal-users] Problems with different libpcap versions
From: Guy Harris <gharris () flashcom ! net>
Date: 2000-08-22 6:16:15
[Download RAW message or body]
On Mon, Aug 21, 2000 at 04:09:42PM +0200, Christian Fiedler wrote:
> Hi!
>
> In wanted to upgrade ethereal (from 0.8.8 to 0.8.11) and libpcap
> (from the "official" version 0.4 to 0.5.2 from tcpdump.org) simultanously.
> The problem was I got a message when trying to sniff the traffic on my
> ppp0-device. It said: "The type of network is not supported by ethereal".
> I tried downgrading to ethereal 0.8.8 and experienced the same problems.
>
> tcpdump was working fine all the time.
>
> After various attempts with the two different versions of libpcap and
> ethereal my box runs fine now with libpcap 0.4 and ethereal 0.8.11.
> For uninstalling ethereal I did "make uninstall" and removed
> /usr/local/lib/ethereal completely.
> For uninstalling libpcap I simply overwrote one version with the other.
> My config: Slackware 7.0, Kernel 2.2.16
>
> Is this a known problem?
It is now. :-)
It's a consequence of an initial attempt, in libpcap 0.5.2, to try to
cope with the hellish mess of incompatible network type values in
different versions of libpcap (see recent mail messages from me with
"ATM" in the subject line ranting about said hellish mess).
As a consequence of that change, Linux PPP devices use a network type
value of 100, rather than 9, when built with libpcap 0.5.2.
A change to Ethereal I'm working on to address the problem brought up in
the thread with "ATM" in the subject line should also fix this problem.
If you want to try libpcap 0.5.2 again, try replacing "capture.c" and
"wiretap/libpcap.c" in the Ethereal source tree (I infer from the "make
uninstall" that you built Ethereal from source) and rebuilding, after
installing libpcap 0.5.2.
(Note to anybody else who picks up these changes:
1) I make no guarantee that they'll work with an Ethereal source
tree other than an 0.8.11 tree or a subsequent CVS tree;
2) they will probably not compile if your system doesn't have
libpcap installed - Wiretap, unlike Ethereal, doesn't
currently check whether you have <pcap.h>, but my change
requires it; if I check it in, I'll also check in changes to
check for it and not include <pcap.h> if it's not there, but
those changes would probably be a pain for people to drop
into their 0.8.11 source tree and try.)
["capture.c" (text/plain)]
/* capture.c
* Routines for packet capture windows
*
* $Id: capture.c,v 1.122 2000/08/22 06:04:40 guy Exp $
*
* Ethereal - Network traffic analyzer
* By Gerald Combs <gerald@zing.org>
* Copyright 1998 Gerald Combs
*
*
* 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.
*/
#ifdef HAVE_CONFIG_H
# include "config.h"
#endif
#ifdef HAVE_LIBPCAP
#ifdef HAVE_SYS_TYPES_H
# include <sys/types.h>
#endif
#ifdef HAVE_SYS_STAT_H
# include <sys/stat.h>
#endif
#ifdef HAVE_SYS_WAIT_H
# include <sys/wait.h>
#endif
#ifdef HAVE_IO_H
#include <io.h>
#endif
#include <gtk/gtk.h>
#include <stdlib.h>
#include <stdio.h>
#include <ctype.h>
#include <string.h>
#ifdef HAVE_FCNTL_H
#include <fcntl.h>
#endif
#ifdef HAVE_UNISTD_H
#include <unistd.h>
#endif
#include <time.h>
#ifdef HAVE_SYS_SOCKET_H
#include <sys/socket.h>
#endif
#ifdef HAVE_SYS_IOCTL_H
#include <sys/ioctl.h>
#endif
#ifdef HAVE_NET_IF_H
#include <net/if.h>
#endif
#include <signal.h>
#include <errno.h>
#ifdef NEED_SNPRINTF_H
# include "snprintf.h"
#endif
#ifndef lib_pcap_h
#include <pcap.h>
#endif
#ifdef _WIN32
#include <process.h> /* For spawning child process */
#endif
#include "gtk/main.h"
#include "gtk/gtkglobals.h"
#include "packet.h"
#include "file.h"
#include "capture.h"
#include "util.h"
#include "simple_dialog.h"
#include "prefs.h"
#include "globals.h"
#include "wiretap/libpcap.h"
#include "wiretap/wtap-int.h"
#include "packet-clip.h"
#include "packet-eth.h"
#include "packet-fddi.h"
#include "packet-null.h"
#include "packet-ppp.h"
#include "packet-raw.h"
#include "packet-tr.h"
int sync_mode; /* fork a child to do the capture, and sync between them */
static int sync_pipe[2]; /* used to sync father */
enum PIPES { READ, WRITE }; /* Constants 0 and 1 for READ and WRITE */
int quit_after_cap; /* Makes a "capture only mode". Implies -k */
gboolean capture_child; /* if this is the child for "-S" */
static int fork_child; /* In parent, process ID of child */
static guint cap_input_id;
#ifdef _WIN32
static guint cap_timer_id;
static int cap_timer_cb(gpointer); /* Win32 kludge to check for pipe input */
#endif
static void cap_file_input_cb(gpointer, gint, GdkInputCondition);
static void capture_delete_cb(GtkWidget *, GdkEvent *, gpointer);
static void capture_stop_cb(GtkWidget *, gpointer);
static void capture_pcap_cb(u_char *, const struct pcap_pkthdr *,
const u_char *);
static void send_errmsg_to_parent(const char *);
static float pct(gint, gint);
typedef struct _loop_data {
gint go;
gint max;
gint linktype;
gint sync_packets;
gboolean from_pipe; /* TRUE if we are capturing data from a pipe */
gboolean modified; /* TRUE if data in the pipe uses modified pcap headers */
gboolean byte_swapped; /* TRUE if data in the pipe is byte swapped */
packet_counts counts;
wtap_dumper *pdh;
} loop_data;
#ifndef _WIN32
static void adjust_header(loop_data *, struct pcap_hdr *, struct pcaprec_hdr *);
static int pipe_open_live(char *, struct pcap_hdr *, loop_data *, char *);
static int pipe_dispatch(int, loop_data *, struct pcap_hdr *);
#endif
/* Win32 needs the O_BINARY flag for open() */
#ifndef O_BINARY
#define O_BINARY 0
#endif
#ifdef _WIN32
/* Win32 needs a handle to the child capture process */
int child_process;
#endif
/* Open a specified file, or create a temporary file, and start a capture
to the file in question. */
void
do_capture(char *capfile_name)
{
char tmpname[128+1];
gboolean is_tempfile;
u_char c;
int i;
guint byte_count;
char *msg;
int err;
int capture_succeeded;
if (capfile_name != NULL) {
/* Try to open/create the specified file for use as a capture buffer. */
cfile.save_file_fd = open(capfile_name, O_RDWR|O_BINARY|O_TRUNC|O_CREAT, 0600);
is_tempfile = FALSE;
} else {
/* Choose a random name for the capture buffer */
cfile.save_file_fd = create_tempfile(tmpname, sizeof tmpname, "ether");
capfile_name = g_strdup(tmpname);
is_tempfile = TRUE;
}
if (cfile.save_file_fd == -1) {
if (is_tempfile) {
simple_dialog(ESD_TYPE_CRIT, NULL,
"The temporary file to which the capture would be saved (\"%s\")"
"could not be opened: %s.", capfile_name, strerror(errno));
} else {
simple_dialog(ESD_TYPE_CRIT, NULL,
file_open_error_message(errno, TRUE), capfile_name);
}
return;
}
close_cap_file(&cfile, info_bar);
g_assert(cfile.save_file == NULL);
cfile.save_file = capfile_name;
if (sync_mode) { /* do the capture in a child process */
char ssnap[24];
char scount[24]; /* need a constant for len of numbers */
char save_file_fd[24];
char errmsg[1024+1];
int error;
#ifdef _WIN32
char sync_pipe_fd[24];
char *filterstring;
#endif
sprintf(ssnap,"%d",cfile.snap); /* in lieu of itoa */
sprintf(scount,"%d",cfile.count);
sprintf(save_file_fd,"%d",cfile.save_file_fd);
#ifdef _WIN32
/* Create a pipe for the child process */
if(_pipe(sync_pipe, 512, O_BINARY) < 0) {
/* Couldn't create the pipe between parent and child. */
error = errno;
unlink(cfile.save_file);
g_free(cfile.save_file);
cfile.save_file = NULL;
simple_dialog(ESD_TYPE_CRIT, NULL, "Couldn't create sync pipe: %s",
strerror(error));
return;
}
/* Convert pipe write handle to a string and pass to child */
itoa(sync_pipe[WRITE], sync_pipe_fd, 10);
/* Convert filter string to a quote delimited string */
filterstring = g_new(char, strlen(cfile.cfilter) + 3);
sprintf(filterstring, "\"%s\"", cfile.cfilter);
filterstring[strlen(cfile.cfilter) + 2] = 0;
/* Spawn process */
fork_child = spawnlp(_P_NOWAIT, ethereal_path, CHILD_NAME, "-i", cfile.iface,
"-w", cfile.save_file, "-W", save_file_fd,
"-c", scount, "-s", ssnap,
"-Z", sync_pipe_fd,
strlen(cfile.cfilter) == 0 ? (const char *)NULL : "-f",
strlen(cfile.cfilter) == 0 ? (const char *)NULL : filterstring,
(const char *)NULL);
g_free(filterstring);
/* Keep a copy for later evaluation by _cwait() */
child_process = fork_child;
#else
signal(SIGCHLD, SIG_IGN);
if (pipe(sync_pipe) < 0) {
/* Couldn't create the pipe between parent and child. */
error = errno;
unlink(cfile.save_file);
g_free(cfile.save_file);
cfile.save_file = NULL;
simple_dialog(ESD_TYPE_CRIT, NULL, "Couldn't create sync pipe: %s",
strerror(error));
return;
}
if ((fork_child = fork()) == 0) {
/*
* Child process - run Ethereal with the right arguments to make
* it just pop up the live capture dialog box and capture with
* the specified capture parameters, writing to the specified file.
*
* args: -i interface specification
* -w file to write
* -W file descriptor to write
* -c count to capture
* -s snaplen
* -m / -b fonts
* -f "filter expression"
*/
close(1);
dup(sync_pipe[WRITE]);
close(sync_pipe[READ]);
execlp(ethereal_path, CHILD_NAME, "-i", cfile.iface,
"-w", cfile.save_file, "-W", save_file_fd,
"-c", scount, "-s", ssnap,
"-m", prefs.gui_font_name,
(cfile.cfilter == NULL)? 0 : "-f",
(cfile.cfilter == NULL)? 0 : cfile.cfilter,
(const char *)NULL);
snprintf(errmsg, sizeof errmsg, "Couldn't run %s in child process: %s",
ethereal_path, strerror(errno));
send_errmsg_to_parent(errmsg);
/* Exit with "_exit()", so that we don't close the connection
to the X server (and cause stuff buffered up by our parent but
not yet sent to be sent, as that stuff should only be sent by
our parent). */
_exit(2);
}
#endif
/* Parent process - read messages from the child process over the
sync pipe. */
/* Close the write side of the pipe, so that only the child has it
open, and thus it completely closes, and thus returns to us
an EOF indication, if the child closes it (either deliberately
or by exiting abnormally). */
close(sync_pipe[WRITE]);
/* Close the save file FD, as we won't be using it - we'll be opening
it and reading the save file through Wiretap. */
close(cfile.save_file_fd);
if (fork_child == -1) {
/* We couldn't even create the child process. */
error = errno;
close(sync_pipe[READ]);
unlink(cfile.save_file);
g_free(cfile.save_file);
cfile.save_file = NULL;
simple_dialog(ESD_TYPE_CRIT, NULL, "Couldn't create child process: %s",
strerror(error));
return;
}
/* Read a byte count from "sync_pipe[READ]", terminated with a
colon; if the count is 0, the child process created the
capture file and we should start reading from it, otherwise
the capture couldn't start and the count is a count of bytes
of error message, and we should display the message. */
byte_count = 0;
for (;;) {
i = read(sync_pipe[READ], &c, 1);
if (i == 0) {
/* EOF - the child process died.
Close the read side of the sync pipe, remove the capture file,
and report the failure.
XXX - reap the child process and report the status in detail. */
close(sync_pipe[READ]);
unlink(cfile.save_file);
g_free(cfile.save_file);
cfile.save_file = NULL;
simple_dialog(ESD_TYPE_WARN, NULL, "Capture child process died");
return;
}
if (c == ';')
break;
if (!isdigit(c)) {
/* Child process handed us crap.
Close the read side of the sync pipe, remove the capture file,
and report the failure. */
close(sync_pipe[READ]);
unlink(cfile.save_file);
g_free(cfile.save_file);
cfile.save_file = NULL;
simple_dialog(ESD_TYPE_WARN, NULL,
"Capture child process sent us a bad message");
return;
}
byte_count = byte_count*10 + c - '0';
}
if (byte_count == 0) {
/* Success. Open the capture file, and set up to read it. */
err = start_tail_cap_file(cfile.save_file, is_tempfile, &cfile);
if (err == 0) {
/* We were able to open and set up to read the capture file;
arrange that our callback be called whenever it's possible
to read from the sync pipe, so that it's called when
the child process wants to tell us something. */
#ifdef _WIN32
/* Tricky to use pipes in win9x, as no concept of wait. NT can
do this but that doesn't cover all win32 platforms. GTK can do
this but doesn't seem to work over processes. Attempt to do
something similar here, start a timer and check for data on every
timeout. */
cap_timer_id = gtk_timeout_add(1000, cap_timer_cb, NULL);
#else
cap_input_id = gtk_input_add_full(sync_pipe[READ],
GDK_INPUT_READ|GDK_INPUT_EXCEPTION,
cap_file_input_cb,
NULL,
(gpointer) &cfile,
NULL);
#endif
} else {
/* We weren't able to open the capture file; complain, and
close the sync pipe. */
simple_dialog(ESD_TYPE_CRIT, NULL,
file_open_error_message(err, FALSE), cfile.save_file);
/* Close the sync pipe. */
close(sync_pipe[READ]);
/* Don't unlink the save file - leave it around, for debugging
purposes. */
g_free(cfile.save_file);
cfile.save_file = NULL;
}
} else {
/* Failure - the child process sent us a message indicating
what the problem was. */
msg = g_malloc(byte_count + 1);
if (msg == NULL) {
simple_dialog(ESD_TYPE_WARN, NULL,
"Capture child process failed, but its error message was too big.");
} else {
i = read(sync_pipe[READ], msg, byte_count);
if (i < 0) {
simple_dialog(ESD_TYPE_WARN, NULL,
"Capture child process failed: Error %s reading its error message.",
strerror(errno));
} else if (i == 0) {
simple_dialog(ESD_TYPE_WARN, NULL,
"Capture child process failed: EOF reading its error message.");
} else
simple_dialog(ESD_TYPE_WARN, NULL, msg);
g_free(msg);
/* Close the sync pipe. */
close(sync_pipe[READ]);
/* Get rid of the save file - the capture never started. */
unlink(cfile.save_file);
g_free(cfile.save_file);
cfile.save_file = NULL;
}
}
} else {
/* Not sync mode. */
capture_succeeded = capture();
if (quit_after_cap) {
/* DON'T unlink the save file. Presumably someone wants it. */
gtk_exit(0);
}
if (capture_succeeded) {
/* Capture succeeded; read in the capture file. */
if ((err = open_cap_file(cfile.save_file, is_tempfile, &cfile)) == 0) {
/* Set the read filter to NULL. */
cfile.rfcode = NULL;
switch (read_cap_file(&cfile, &err)) {
case READ_SUCCESS:
case READ_ERROR:
/* Just because we got an error, that doesn't mean we were unable
to read any of the file; we handle what we could get from the
file. */
break;
case READ_ABORTED:
/* Exit by leaving the main loop, so that any quit functions
we registered get called. */
gtk_main_quit();
return;
}
}
}
/* We're not doing a capture any more, so we don't have a save
file. */
g_free(cfile.save_file);
cfile.save_file = NULL;
}
}
#ifdef _WIN32
/* The timer has expired, see if there's stuff to read from the pipe,
if so call the cap_file_input_cb */
static gint
cap_timer_cb(gpointer data)
{
HANDLE handle;
DWORD avail = 0;
gboolean result, result1;
DWORD childstatus;
/* Oddly enough although Named pipes don't work on win9x,
PeekNamedPipe does !!! */
handle = (HANDLE) _get_osfhandle (sync_pipe[READ]);
result = PeekNamedPipe(handle, NULL, 0, NULL, &avail, NULL);
/* Get the child process exit status */
result1 = GetExitCodeProcess((HANDLE)child_process, &childstatus);
/* If the Peek returned an error, or there are bytes to be read
or the childwatcher thread has terminated then call the normal
callback */
if (!result || avail > 0 || childstatus != STILL_ACTIVE) {
/* avoid reentrancy problems and stack overflow */
gtk_timeout_remove(cap_timer_id);
/* And call the real handler */
cap_file_input_cb((gpointer) &cfile, 0, 0);
/* Return false so that the timer is not run again */
return FALSE;
}
else {
/* No data so let timer run again */
return TRUE;
}
}
#endif
/* There's stuff to read from the sync pipe, meaning the child has sent
us a message, or the sync pipe has closed, meaning the child has
closed it (perhaps because it exited). */
static void
cap_file_input_cb(gpointer data, gint source, GdkInputCondition condition)
{
capture_file *cf = (capture_file *)data;
char buffer[256+1], *p = buffer, *q = buffer;
int nread;
int to_read = 0;
gboolean exit_loop = FALSE;
int err;
int wstatus;
int wsignal;
char *msg;
char *sigmsg;
char sigmsg_buf[6+1+3+1];
char *coredumped;
#ifndef _WIN32
/* avoid reentrancy problems and stack overflow */
gtk_input_remove(cap_input_id);
#endif
if ((nread = read(sync_pipe[READ], buffer, 256)) <= 0) {
/* The child has closed the sync pipe, meaning it's not going to be
capturing any more packets. Pick up its exit status, and
complain if it died of a signal. */
#ifdef _WIN32
/* XXX - analyze the wait stuatus and display more information
in the dialog box? */
if (_cwait(&wstatus, child_process, _WAIT_CHILD) == -1) {
simple_dialog(ESD_TYPE_WARN, NULL, "Child capture process stopped unexpectedly");
}
#else
if (wait(&wstatus) != -1) {
/* XXX - are there any platforms on which we can run that *don't*
support POSIX.1's <sys/wait.h> and macros therein? */
wsignal = wstatus & 0177;
coredumped = "";
if (wstatus == 0177) {
/* It stopped, rather than exiting. "Should not happen." */
msg = "stopped";
wsignal = (wstatus >> 8) & 0xFF;
} else {
msg = "terminated";
if (wstatus & 0200)
coredumped = " - core dumped";
}
if (wsignal != 0) {
switch (wsignal) {
case SIGHUP:
sigmsg = "Hangup";
break;
case SIGINT:
sigmsg = "Interrupted";
break;
case SIGQUIT:
sigmsg = "Quit";
break;
case SIGILL:
sigmsg = "Illegal instruction";
break;
case SIGTRAP:
sigmsg = "Trace trap";
break;
case SIGABRT:
sigmsg = "Abort";
break;
case SIGFPE:
sigmsg = "Arithmetic exception";
break;
case SIGKILL:
sigmsg = "Killed";
break;
case SIGBUS:
sigmsg = "Bus error";
break;
case SIGSEGV:
sigmsg = "Segmentation violation";
break;
/* http://metalab.unc.edu/pub/Linux/docs/HOWTO/GCC-HOWTO
Linux is POSIX compliant. These are not POSIX-defined signals ---
ISO/IEC 9945-1:1990 (IEEE Std 1003.1-1990), paragraph B.3.3.1.1 sez:
``The signals SIGBUS, SIGEMT, SIGIOT, SIGTRAP, and SIGSYS
were omitted from POSIX.1 because their behavior is
implementation dependent and could not be adequately catego-
rized. Conforming implementations may deliver these sig-
nals, but must document the circumstances under which they
are delivered and note any restrictions concerning their
delivery.''
*/
#ifdef SIGSYS
case SIGSYS:
sigmsg = "Bad system call";
break;
#endif
case SIGPIPE:
sigmsg = "Broken pipe";
break;
case SIGALRM:
sigmsg = "Alarm clock";
break;
case SIGTERM:
sigmsg = "Terminated";
break;
default:
sprintf(sigmsg_buf, "Signal %d", wsignal);
sigmsg = sigmsg_buf;
break;
}
simple_dialog(ESD_TYPE_WARN, NULL,
"Child capture process %s: %s%s", msg, sigmsg, coredumped);
}
}
#endif
/* Read what remains of the capture file, and finish the capture.
XXX - do something if this fails? */
switch (finish_tail_cap_file(cf, &err)) {
case READ_SUCCESS:
case READ_ERROR:
/* Just because we got an error, that doesn't mean we were unable
to read any of the file; we handle what we could get from the
file. */
break;
case READ_ABORTED:
/* Exit by leaving the main loop, so that any quit functions
we registered get called. */
gtk_main_quit();
return;
}
/* We're not doing a capture any more, so we don't have a save
file. */
g_free(cf->save_file);
cf->save_file = NULL;
return;
}
buffer[nread] = '\0';
while(!exit_loop) {
/* look for (possibly multiple) '*' */
switch (*q) {
case '*' :
to_read += atoi(p);
p = q + 1;
q++;
break;
case '\0' :
/* XXX should handle the case of a pipe full (i.e. no star found) */
exit_loop = TRUE;
break;
default :
q++;
break;
}
}
/* Read from the capture file the number of records the child told us
it added.
XXX - do something if this fails? */
switch (continue_tail_cap_file(cf, to_read, &err)) {
case READ_SUCCESS:
case READ_ERROR:
/* Just because we got an error, that doesn't mean we were unable
to read any of the file; we handle what we could get from the
file.
XXX - abort on a read error? */
break;
case READ_ABORTED:
/* Kill the child capture process; the user wants to exit, and we
shouldn't just leave it running. */
#ifdef _WIN32
/* XXX - kill it. */
#else
kill(fork_child, SIGTERM); /* SIGTERM so it can clean up if necessary */
#endif
break;
}
/* restore pipe handler */
#ifdef _WIN32
cap_timer_id = gtk_timeout_add(1000, cap_timer_cb, NULL);
#else
cap_input_id = gtk_input_add_full (sync_pipe[READ],
GDK_INPUT_READ|GDK_INPUT_EXCEPTION,
cap_file_input_cb,
NULL,
(gpointer) cf,
NULL);
#endif
}
/*
* Timeout, in milliseconds, for reads from the stream of captured packets.
*/
#define CAP_READ_TIMEOUT 250
#ifndef _WIN32
/* Take carre of byte order in the libpcap headers read from pipes.
* (function taken from wiretap/libpcap.c) */
static void
adjust_header(loop_data *ld, struct pcap_hdr *hdr, struct pcaprec_hdr *rechdr)
{
if (ld->byte_swapped) {
/* Byte-swap the record header fields. */
rechdr->ts_sec = BSWAP32(rechdr->ts_sec);
rechdr->ts_usec = BSWAP32(rechdr->ts_usec);
rechdr->incl_len = BSWAP32(rechdr->incl_len);
rechdr->orig_len = BSWAP32(rechdr->orig_len);
}
/* 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 (hdr->version_major == 2 &&
(hdr->version_minor < 3 ||
(hdr->version_minor == 3 && rechdr->incl_len > rechdr->orig_len))) {
guint32 temp;
temp = rechdr->orig_len;
rechdr->orig_len = rechdr->incl_len;
rechdr->incl_len = temp;
}
}
/* Mimic pcap_open_live() for pipe captures
* We check if "pipename" is "-" (stdin) or a FIFO, open it, and read the
* header.
* N.B. : we can't read the libpcap formats used in RedHat 6.1 or SuSE 6.3
* because we can't seek on pipes (see wiretap/libpcap.c for details) */
static int
pipe_open_live(char *pipename, struct pcap_hdr *hdr, loop_data *ld, char *ebuf)
{
struct stat pipe_stat;
int fd;
guint32 magic;
int bytes_read, b;
if (strcmp(pipename, "-") == 0) fd = 0; /* read from stdin */
else if (stat(pipename, &pipe_stat) == 0 && S_ISFIFO(pipe_stat.st_mode)) {
if ((fd = open(pipename, O_RDONLY)) == -1) return -1;
} else return -1;
ld->from_pipe = TRUE;
/* read the pcap header */
if (read(fd, &magic, sizeof magic) != sizeof magic) {
close(fd);
return -1;
}
switch (magic) {
case PCAP_MAGIC:
/* Host that wrote it has our byte order, and was running
a program using either standard or ss990417 libpcap. */
ld->byte_swapped = FALSE;
ld->modified = FALSE;
break;
case PCAP_MODIFIED_MAGIC:
/* Host that wrote it has our byte order, but was running
a program using either ss990915 or ss991029 libpcap. */
ld->byte_swapped = FALSE;
ld->modified = TRUE;
break;
case PCAP_SWAPPED_MAGIC:
/* Host that wrote it has a byte order opposite to ours,
and was running a program using either standard or
ss990417 libpcap. */
ld->byte_swapped = TRUE;
ld->modified = FALSE;
break;
case PCAP_SWAPPED_MODIFIED_MAGIC:
/* Host that wrote it out has a byte order opposite to
ours, and was running a program using either ss990915
or ss991029 libpcap. */
ld->byte_swapped = TRUE;
ld->modified = TRUE;
break;
default:
/* Not a "libpcap" type we know about. */
close(fd);
return -1;
}
/* Read the rest of the header */
bytes_read = read(fd, hdr, sizeof(struct pcap_hdr));
if (bytes_read <= 0) {
close(fd);
return -1;
}
while (bytes_read < sizeof(struct pcap_hdr))
{
b = read(fd, ((char *)&hdr)+bytes_read, sizeof(struct pcap_hdr) - bytes_read);
if (b <= 0) {
close(fd);
return -1;
}
bytes_read += b;
}
if (ld->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) {
close(fd);
return -1;
}
return fd;
}
/* We read one record from the pipe, take care of byte order in the record
* header, write the record in the capture file, and update capture statistics. */
static int
pipe_dispatch(int fd, loop_data *ld, struct pcap_hdr *hdr)
{
struct wtap_pkthdr whdr;
struct pcaprec_modified_hdr rechdr;
int bytes_to_read, bytes_read, b;
u_char pd[WTAP_MAX_PACKET_SIZE];
int err;
/* read the record header */
bytes_to_read = ld->modified ? sizeof rechdr : sizeof rechdr.hdr;
bytes_read = read(fd, &rechdr, bytes_to_read);
if (bytes_read <= 0) {
close(fd);
ld->go = FALSE;
return 0;
}
while (bytes_read < bytes_to_read)
{
b = read(fd, ((char *)&rechdr)+bytes_read, bytes_to_read - bytes_read);
if (b <= 0) {
close(fd);
ld->go = FALSE;
return 0;
}
bytes_read += b;
}
/* take care of byte order */
adjust_header(ld, hdr, &rechdr.hdr);
if (rechdr.hdr.incl_len > WTAP_MAX_PACKET_SIZE) {
close(fd);
ld->go = FALSE;
return 0;
}
/* read the packet data */
bytes_read = read(fd, pd, rechdr.hdr.incl_len);
if (bytes_read <= 0) {
close(fd);
ld->go = FALSE;
return 0;
}
while (bytes_read < rechdr.hdr.incl_len)
{
b = read(fd, pd+bytes_read, rechdr.hdr.incl_len - bytes_read);
if (b <= 0) {
close(fd);
ld->go = FALSE;
return 0;
}
bytes_read += b;
}
/* dump the packet data to the capture file */
whdr.ts.tv_sec = rechdr.hdr.ts_sec;
whdr.ts.tv_usec = rechdr.hdr.ts_usec;
whdr.caplen = rechdr.hdr.incl_len;
whdr.len = rechdr.hdr.orig_len;
whdr.pkt_encap = ld->linktype;
wtap_dump(ld->pdh, &whdr, NULL, pd, &err);
/* Set the initial payload to the packet length, and the initial
captured payload to the capture length (other protocols may
reduce them if their headers say they're less). */
pi.len = whdr.len;
pi.captured_len = whdr.caplen;
/* update capture statistics */
switch (ld->linktype) {
case WTAP_ENCAP_ETHERNET:
capture_eth(pd, 0, &ld->counts);
break;
case WTAP_ENCAP_FDDI:
case WTAP_ENCAP_FDDI_BITSWAPPED:
capture_fddi(pd, &ld->counts);
break;
case WTAP_ENCAP_TR:
capture_tr(pd, 0, &ld->counts);
break;
case WTAP_ENCAP_NULL:
capture_null(pd, &ld->counts);
break;
case WTAP_ENCAP_PPP:
capture_ppp(pd, 0, &ld->counts);
break;
case WTAP_ENCAP_RAW_IP:
capture_raw(pd, &ld->counts);
break;
case WTAP_ENCAP_LINUX_ATM_CLIP:
capture_clip(pd, &ld->counts);
break;
/* XXX - FreeBSD may append 4-byte ATM pseudo-header to DLT_ATM_RFC1483,
with LLC header following; we should implement it at some
point. */
}
return 1;
}
#endif
/* Do the low-level work of a capture.
Returns TRUE if it succeeds, FALSE otherwise. */
int
capture(void)
{
GtkWidget *cap_w, *main_vb, *count_lb, *sctp_lb, *tcp_lb, *udp_lb, *icmp_lb,
*ospf_lb, *gre_lb, *netbios_lb, *ipx_lb, *vines_lb, *other_lb, *stop_bt;
pcap_t *pch;
int pcap_encap;
int snaplen;
gchar err_str[PCAP_ERRBUF_SIZE], label_str[64];
loop_data ld;
bpf_u_int32 netnum, netmask;
time_t upd_time, cur_time;
int err, inpkts;
char errmsg[4096+1];
#ifndef _WIN32
static const char ppamsg[] = "can't find PPA for ";
char *libpcap_warn;
#endif
fd_set set1;
struct timeval timeout;
#ifdef linux
int pcap_fd = 0;
#endif
#ifdef _WIN32
WORD wVersionRequested;
WSADATA wsaData;
#endif
#ifndef _WIN32
int pipe_fd = -1;
struct pcap_hdr hdr;
#endif
/* Initialize Windows Socket if we are in a WIN32 OS
This needs to be done before querying the interface for network/netmask */
#ifdef _WIN32
wVersionRequested = MAKEWORD( 1, 1 );
err = WSAStartup( wVersionRequested, &wsaData );
if (err!=0) {
snprintf(errmsg, sizeof errmsg,
"Couldn't initialize Windows Sockets.");
pch=NULL;
goto error;
}
#endif
ld.go = TRUE;
ld.counts.total = 0;
ld.max = cfile.count;
ld.linktype = WTAP_ENCAP_UNKNOWN;
ld.from_pipe = FALSE;
ld.sync_packets = 0;
ld.counts.sctp = 0;
ld.counts.tcp = 0;
ld.counts.udp = 0;
ld.counts.icmp = 0;
ld.counts.ospf = 0;
ld.counts.gre = 0;
ld.counts.ipx = 0;
ld.counts.netbios = 0;
ld.counts.vines = 0;
ld.counts.other = 0;
ld.pdh = NULL;
/* Open the network interface to capture from it. */
pch = pcap_open_live(cfile.iface, cfile.snap, 1, CAP_READ_TIMEOUT, err_str);
if (pch == NULL) {
#ifdef _WIN32
/* Well, we couldn't start the capture.
If this is a child process that does the capturing in sync
mode or fork mode, it shouldn't do any UI stuff until we pop up the
capture-progress window, and, since we couldn't start the
capture, we haven't popped it up. */
if (!capture_child) {
while (gtk_events_pending()) gtk_main_iteration();
}
/* On Win32 OSes, the capture devices are probably available to all
users; don't warn about permissions problems.
Do, however, warn that Token Ring and PPP devices aren't supported. */
snprintf(errmsg, sizeof errmsg,
"The capture session could not be initiated (%s).\n"
"Please check that you have the proper interface specified.\n"
"\n"
"Note that the driver Ethereal uses for packet capture on Windows\n"
"doesn't support capturing on Token Ring interfaces, and doesn't\n"
"support capturing on PPP/WAN interfaces in Windows NT/2000.\n",
err_str);
goto error;
#else
/* try to open cfile.iface as a pipe */
pipe_fd = pipe_open_live(cfile.iface, &hdr, &ld, err_str);
if (pipe_fd == -1) {
/* Well, we couldn't start the capture.
If this is a child process that does the capturing in sync
mode or fork mode, it shouldn't do any UI stuff until we pop up the
capture-progress window, and, since we couldn't start the
capture, we haven't popped it up. */
if (!capture_child) {
while (gtk_events_pending()) gtk_main_iteration();
}
/* If we got a "can't find PPA for XXX" message, warn the user (who
is running Ethereal on HP-UX) that they don't have a version
of libpcap patched to properly handle HP-UX (the patched version
says "can't find /dev/dlpi PPA for XXX" rather than "can't find
PPA for XXX"). */
if (strncmp(err_str, ppamsg, sizeof ppamsg - 1) == 0)
libpcap_warn =
"\n\n"
"You are running Ethereal with a version of the libpcap library\n"
"that doesn't handle HP-UX network devices well; this means that\n"
"Ethereal may not be able to capture packets.\n"
"\n"
"To fix this, you will need to download the source to Ethereal\n"
"from ethereal.zing.org if you have not already done so, read\n"
"the instructions in the \"README.hpux\" file in the source\n"
"distribution, download the source to libpcap if you have not\n"
"already done so, patch libpcap as per the instructions, rebuild\n"
"and install libpcap, and then build Ethereal (if you have already\n"
"built Ethereal from source, do a \"make distclean\" and re-run\n"
"configure before building).";
else
libpcap_warn = "";
snprintf(errmsg, sizeof errmsg,
"The capture session could not be initiated (%s).\n"
"Please check to make sure you have sufficient permissions, and that\n"
"you have the proper interface or pipe specified.%s", err_str,
libpcap_warn);
goto error;
}
#endif
}
/* capture filters only work on real interfaces */
if (cfile.cfilter && !ld.from_pipe) {
/* A capture filter was specified; set it up. */
if (pcap_lookupnet (cfile.iface, &netnum, &netmask, err_str) < 0) {
snprintf(errmsg, sizeof errmsg,
"Can't use filter: Couldn't obtain netmask info (%s).", err_str);
goto error;
}
if (pcap_compile(pch, &cfile.fcode, cfile.cfilter, 1, netmask) < 0) {
snprintf(errmsg, sizeof errmsg, "Unable to parse filter string (%s).",
pcap_geterr(pch));
goto error;
}
if (pcap_setfilter(pch, &cfile.fcode) < 0) {
snprintf(errmsg, sizeof errmsg, "Can't install filter (%s).",
pcap_geterr(pch));
goto error;
}
}
/* Set up to write to the capture file. */
#ifndef _WIN32
if (ld.from_pipe) {
pcap_encap = hdr.network;
snaplen = hdr.snaplen;
} else
#endif
{
pcap_encap = pcap_datalink(pch);
snaplen = pcap_snapshot(pch);
}
ld.linktype = wtap_pcap_encap_to_wtap_encap(pcap_encap);
if (ld.linktype == WTAP_ENCAP_UNKNOWN) {
snprintf(errmsg, sizeof errmsg,
"The network you're capturing from is of a type"
" that Ethereal doesn't support (data link type %d).", pcap_encap);
goto error;
}
ld.pdh = wtap_dump_fdopen(cfile.save_file_fd, WTAP_FILE_PCAP,
ld.linktype, snaplen, &err);
if (ld.pdh == NULL) {
/* We couldn't set up to write to the capture file. */
switch (err) {
case WTAP_ERR_CANT_OPEN:
strcpy(errmsg, "The file to which the capture would be saved"
" couldn't be created for some unknown reason.");
break;
case WTAP_ERR_SHORT_WRITE:
strcpy(errmsg, "A full header couldn't be written to the file"
" to which the capture would be saved.");
break;
default:
if (err < 0) {
snprintf(errmsg, sizeof(errmsg),
"The file to which the capture would be"
" saved (\"%s\") could not be opened: Error %d.",
cfile.save_file, err);
} else {
snprintf(errmsg, sizeof(errmsg),
"The file to which the capture would be"
" saved (\"%s\") could not be opened: %s.",
cfile.save_file, strerror(err));
}
break;
}
goto error;
}
/* XXX - capture SIGTERM and close the capture, in case we're on a
Linux 2.0[.x] system and you have to explicitly close the capture
stream in order to turn promiscuous mode off? We need to do that
in other places as well - and I don't think that works all the
time in any case, due to libpcap bugs. */
if (capture_child) {
/* Well, we should be able to start capturing.
This is the child process for a sync mode capture, so sync out
the capture file, so the header makes it to the file system,
and send a "capture started successfully and capture file created"
message to our parent so that they'll open the capture file and
update its windows to indicate that we have a live capture in
progress. */
fflush(wtap_dump_file(ld.pdh));
write(1, "0;", 2);
}
cap_w = gtk_window_new(GTK_WINDOW_TOPLEVEL);
gtk_window_set_title(GTK_WINDOW(cap_w), "Ethereal: Capture");
gtk_window_set_modal(GTK_WINDOW(cap_w), TRUE);
/* Container for capture display widgets */
main_vb = gtk_vbox_new(FALSE, 1);
gtk_container_border_width(GTK_CONTAINER(main_vb), 5);
gtk_container_add(GTK_CONTAINER(cap_w), main_vb);
gtk_widget_show(main_vb);
count_lb = gtk_label_new("Count: 0");
gtk_box_pack_start(GTK_BOX(main_vb), count_lb, FALSE, FALSE, 3);
gtk_widget_show(count_lb);
sctp_lb = gtk_label_new("SCTP: 0 (0.0%)");
gtk_box_pack_start(GTK_BOX(main_vb), sctp_lb, FALSE, FALSE, 3);
gtk_widget_show(sctp_lb);
tcp_lb = gtk_label_new("TCP: 0 (0.0%)");
gtk_box_pack_start(GTK_BOX(main_vb), tcp_lb, FALSE, FALSE, 3);
gtk_widget_show(tcp_lb);
udp_lb = gtk_label_new("UDP: 0 (0.0%)");
gtk_box_pack_start(GTK_BOX(main_vb), udp_lb, FALSE, FALSE, 3);
gtk_widget_show(udp_lb);
icmp_lb = gtk_label_new("ICMP: 0 (0.0%)");
gtk_box_pack_start(GTK_BOX(main_vb), icmp_lb, FALSE, FALSE, 3);
gtk_widget_show(icmp_lb);
ospf_lb = gtk_label_new("OSPF: 0 (0.0%)");
gtk_box_pack_start(GTK_BOX(main_vb), ospf_lb, FALSE, FALSE, 3);
gtk_widget_show(ospf_lb);
gre_lb = gtk_label_new("GRE: 0 (0.0%)");
gtk_box_pack_start(GTK_BOX(main_vb), gre_lb, FALSE, FALSE, 3);
gtk_widget_show(gre_lb);
netbios_lb = gtk_label_new("NetBIOS: 0 (0.0%)");
gtk_box_pack_start(GTK_BOX(main_vb), netbios_lb, FALSE, FALSE, 3);
gtk_widget_show(netbios_lb);
ipx_lb = gtk_label_new("IPX: 0 (0.0%)");
gtk_box_pack_start(GTK_BOX(main_vb), ipx_lb, FALSE, FALSE, 3);
gtk_widget_show(ipx_lb);
vines_lb = gtk_label_new("VINES: 0 (0.0%)");
gtk_box_pack_start(GTK_BOX(main_vb), vines_lb, FALSE, FALSE, 3);
gtk_widget_show(vines_lb);
other_lb = gtk_label_new("Other: 0 (0.0%)");
gtk_box_pack_start(GTK_BOX(main_vb), other_lb, FALSE, FALSE, 3);
gtk_widget_show(other_lb);
/* allow user to either click a stop button, or the close button on
the window to stop a capture in progress. */
stop_bt = gtk_button_new_with_label ("Stop");
gtk_signal_connect(GTK_OBJECT(stop_bt), "clicked",
GTK_SIGNAL_FUNC(capture_stop_cb), (gpointer) &ld);
gtk_signal_connect(GTK_OBJECT(cap_w), "delete_event",
GTK_SIGNAL_FUNC(capture_delete_cb), (gpointer) &ld);
gtk_box_pack_end(GTK_BOX(main_vb), stop_bt, FALSE, FALSE, 3);
GTK_WIDGET_SET_FLAGS(stop_bt, GTK_CAN_DEFAULT);
gtk_widget_grab_default(stop_bt);
GTK_WIDGET_SET_FLAGS(stop_bt, GTK_CAN_DEFAULT);
gtk_widget_grab_default(stop_bt);
gtk_widget_show(stop_bt);
gtk_widget_show(cap_w);
upd_time = time(NULL);
#ifdef linux
if (!ld.from_pipe) pcap_fd = pcap_fileno(pch);
#endif
while (ld.go) {
while (gtk_events_pending()) gtk_main_iteration();
#ifndef _WIN32
if (ld.from_pipe) {
FD_ZERO(&set1);
FD_SET(pipe_fd, &set1);
timeout.tv_sec = 0;
timeout.tv_usec = CAP_READ_TIMEOUT*1000;
if (select(pipe_fd+1, &set1, NULL, NULL, &timeout) != 0) {
/*
* "select()" says we can read from the pipe without blocking; go for
* it. We are not sure we can read a whole record, but at least the
* begninning of one. pipe_dispatch() will block reading the whole
* record.
*/
inpkts = pipe_dispatch(pipe_fd, &ld, &hdr);
} else
inpkts = 0;
}
else
#endif
{
#ifdef linux
/*
* Sigh. The semantics of the read timeout argument to
* "pcap_open_live()" aren't particularly well specified by
* the "pcap" man page - at least with the BSD BPF code, the
* intent appears to be, at least in part, a way of cutting
* down the number of reads done on a capture, by blocking
* until the buffer fills or a timer expires - and the Linux
* libpcap doesn't actually support it, so we can't use it
* to break out of the "pcap_dispatch()" every 1/4 of a second
* or so.
*
* Thus, on Linux, we do a "select()" on the file descriptor for the
* capture, with a timeout of CAP_READ_TIMEOUT milliseconds, or
* CAP_READ_TIMEOUT*1000 microseconds.
*/
FD_ZERO(&set1);
FD_SET(pcap_fd, &set1);
timeout.tv_sec = 0;
timeout.tv_usec = CAP_READ_TIMEOUT*1000;
if (select(pcap_fd+1, &set1, NULL, NULL, &timeout) != 0) {
/*
* "select()" says we can read from it without blocking; go for
* it.
*/
inpkts = pcap_dispatch(pch, 1, capture_pcap_cb, (u_char *) &ld);
} else
inpkts = 0;
#else
inpkts = pcap_dispatch(pch, 1, capture_pcap_cb, (u_char *) &ld);
#endif
}
if (inpkts > 0)
ld.sync_packets += inpkts;
/* Only update once a second so as not to overload slow displays */
cur_time = time(NULL);
if (cur_time > upd_time) {
upd_time = cur_time;
sprintf(label_str, "Count: %d", ld.counts.total);
gtk_label_set(GTK_LABEL(count_lb), label_str);
sprintf(label_str, "SCTP: %d (%.1f%%)", ld.counts.sctp,
pct(ld.counts.sctp, ld.counts.total));
gtk_label_set(GTK_LABEL(sctp_lb), label_str);
sprintf(label_str, "TCP: %d (%.1f%%)", ld.counts.tcp,
pct(ld.counts.tcp, ld.counts.total));
gtk_label_set(GTK_LABEL(tcp_lb), label_str);
sprintf(label_str, "UDP: %d (%.1f%%)", ld.counts.udp,
pct(ld.counts.udp, ld.counts.total));
gtk_label_set(GTK_LABEL(udp_lb), label_str);
sprintf(label_str, "ICMP: %d (%.1f%%)", ld.counts.icmp,
pct(ld.counts.icmp, ld.counts.total));
gtk_label_set(GTK_LABEL(icmp_lb), label_str);
sprintf(label_str, "OSPF: %d (%.1f%%)", ld.counts.ospf,
pct(ld.counts.ospf, ld.counts.total));
gtk_label_set(GTK_LABEL(ospf_lb), label_str);
sprintf(label_str, "GRE: %d (%.1f%%)", ld.counts.gre,
pct(ld.counts.gre, ld.counts.total));
gtk_label_set(GTK_LABEL(gre_lb), label_str);
sprintf(label_str, "NetBIOS: %d (%.1f%%)", ld.counts.netbios,
pct(ld.counts.netbios, ld.counts.total));
gtk_label_set(GTK_LABEL(netbios_lb), label_str);
sprintf(label_str, "IPX: %d (%.1f%%)", ld.counts.ipx,
pct(ld.counts.ipx, ld.counts.total));
gtk_label_set(GTK_LABEL(ipx_lb), label_str);
sprintf(label_str, "VINES: %d (%.1f%%)", ld.counts.vines,
pct(ld.counts.vines, ld.counts.total));
gtk_label_set(GTK_LABEL(vines_lb), label_str);
sprintf(label_str, "Other: %d (%.1f%%)", ld.counts.other,
pct(ld.counts.other, ld.counts.total));
gtk_label_set(GTK_LABEL(other_lb), label_str);
/* do sync here, too */
fflush(wtap_dump_file(ld.pdh));
if (capture_child && ld.sync_packets) {
/* This is the child process for a sync mode capture, so send
our parent a message saying we've written out "ld.sync_packets"
packets to the capture file. */
char tmp[20];
sprintf(tmp, "%d*", ld.sync_packets);
write(1, tmp, strlen(tmp));
ld.sync_packets = 0;
}
}
}
if (!wtap_dump_close(ld.pdh, &err)) {
/* XXX - in fork mode, this may not pop up, or, if it does,
it may disappear as soon as we exit.
We should have the parent process, while it's reading
the packet count update messages, catch error messages
and pop up a message box if it sees one. */
switch (err) {
case WTAP_ERR_CANT_CLOSE:
simple_dialog(ESD_TYPE_WARN, NULL,
"The file to which the capture was being saved"
" couldn't be closed for some unknown reason.");
break;
case WTAP_ERR_SHORT_WRITE:
simple_dialog(ESD_TYPE_WARN, NULL,
"Not all the data could be written to the file"
" to which the capture was being saved.");
break;
default:
simple_dialog(ESD_TYPE_WARN, NULL,
"The file to which the capture was being"
" saved (\"%s\") could not be closed: %s.",
cfile.save_file, wtap_strerror(err));
break;
}
}
#ifndef _WIN32
if (ld.from_pipe)
close(pipe_fd);
else
#endif
pcap_close(pch);
gtk_grab_remove(GTK_WIDGET(cap_w));
gtk_widget_destroy(GTK_WIDGET(cap_w));
return TRUE;
error:
/* We can't use the save file, and we have no wtap_dump stream
to close in order to close it, so close the FD directly. */
close(cfile.save_file_fd);
/* We couldn't even start the capture, so get rid of the capture
file. */
unlink(cfile.save_file); /* silently ignore error */
g_free(cfile.save_file);
cfile.save_file = NULL;
if (capture_child) {
/* This is the child process for a sync mode capture.
Send the error message to our parent, so they can display a
dialog box containing it. */
send_errmsg_to_parent(errmsg);
} else {
/* Display the dialog box ourselves; there's no parent. */
simple_dialog(ESD_TYPE_CRIT, NULL, errmsg);
}
if (pch != NULL && !ld.from_pipe)
pcap_close(pch);
return FALSE;
}
static void
send_errmsg_to_parent(const char *errmsg)
{
int msglen = strlen(errmsg);
char lenbuf[10+1+1];
sprintf(lenbuf, "%u;", msglen);
write(1, lenbuf, strlen(lenbuf));
write(1, errmsg, msglen);
}
static float
pct(gint num, gint denom) {
if (denom) {
return (float) num * 100.0 / (float) denom;
} else {
return 0.0;
}
}
static void
capture_delete_cb(GtkWidget *w, GdkEvent *event, gpointer data) {
capture_stop_cb(NULL, data);
}
static void
capture_stop_cb(GtkWidget *w, gpointer data) {
loop_data *ld = (loop_data *) data;
ld->go = FALSE;
}
static void
capture_pcap_cb(u_char *user, const struct pcap_pkthdr *phdr,
const u_char *pd) {
struct wtap_pkthdr whdr;
loop_data *ld = (loop_data *) user;
int err;
if ((++ld->counts.total >= ld->max) && (ld->max > 0))
{
ld->go = FALSE;
}
if (ld->pdh) {
/* "phdr->ts" may not necessarily be a "struct timeval" - it may
be a "struct bpf_timeval", with member sizes wired to 32
bits - and we may go that way ourselves in the future, so
copy the members individually. */
whdr.ts.tv_sec = phdr->ts.tv_sec;
whdr.ts.tv_usec = phdr->ts.tv_usec;
whdr.caplen = phdr->caplen;
whdr.len = phdr->len;
whdr.pkt_encap = ld->linktype;
/* XXX - do something if this fails */
wtap_dump(ld->pdh, &whdr, NULL, pd, &err);
}
/* Set the initial payload to the packet length, and the initial
captured payload to the capture length (other protocols may
reduce them if their headers say they're less). */
pi.len = phdr->len;
pi.captured_len = phdr->caplen;
switch (ld->linktype) {
case WTAP_ENCAP_ETHERNET:
capture_eth(pd, 0, &ld->counts);
break;
case WTAP_ENCAP_FDDI:
case WTAP_ENCAP_FDDI_BITSWAPPED:
capture_fddi(pd, &ld->counts);
break;
case WTAP_ENCAP_TR:
capture_tr(pd, 0, &ld->counts);
break;
case WTAP_ENCAP_NULL:
capture_null(pd, &ld->counts);
break;
case WTAP_ENCAP_PPP:
capture_ppp(pd, 0, &ld->counts);
break;
case WTAP_ENCAP_RAW_IP:
capture_raw(pd, &ld->counts);
break;
case WTAP_ENCAP_LINUX_ATM_CLIP:
capture_clip(pd, &ld->counts);
break;
/* XXX - FreeBSD may append 4-byte ATM pseudo-header to DLT_ATM_RFC1483,
with LLC header following; we should implement it at some
point. */
}
}
#endif /* HAVE_LIBPCAP */
["libpcap.c" (text/plain)]
/* libpcap.c
*
* $Id: libpcap.c,v 1.37 2000/07/30 16:54:11 oabad Exp $
*
* Wiretap Library
* Copyright (c) 1998 by Gilbert Ramirez <gram@xiexie.org>
*
* 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.
*
*/
#ifdef HAVE_CONFIG_H
#include "config.h"
#endif
#include <stdlib.h>
#include <errno.h>
#include "wtap-int.h"
#include "file_wrappers.h"
#include "buffer.h"
#include "libpcap.h"
/* See source to the "libpcap" library for information on the "libpcap"
file format. */
/* On some systems, the FDDI MAC addresses are bit-swapped. */
#if !defined(ultrix) && !defined(__alpha) && !defined(__bsdi__)
#define BIT_SWAPPED_MAC_ADDRS
#endif
static int libpcap_read(wtap *wth, int *err);
static void adjust_header(wtap *wth, struct pcaprec_hdr *hdr);
static void libpcap_close(wtap *wth);
static gboolean libpcap_dump(wtap_dumper *wdh, const struct wtap_pkthdr *phdr,
const union wtap_pseudo_header *pseudo_header, const u_char *pd, int *err);
/*
* 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.
*
* 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
* 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.
*
* We cope with this with a scheme that, if the machine on which Wiretap
* is built has <pcap.h>, uses the values from that header, so that we
* use the interpretation of DLT_ values that the libpcap on that machine
* uses.
*/
/*
* Step 1: include this platform's <pcap.h>, if it has one, to find its
* its DLT_ values, so that we interpret those DLT_ values that differ
* between platforms in the way they'd be interpreted on this platform's
* libpcap.
*/
#include <pcap.h>
/*
* Step 2: for those DLT_ values defined by the standard LBL 0.4 libpcap,
* and for other types that are (probably) the same on all platforms
* that define them, if the value isn't defined (either because this
* machine has no <pcap.h> or because its <pcap.h> doesn't define it),
* define it ourselves, with the LBL value (yes, this could mean we
* misinterpret that type on a platform that defines it differently, but
* So It Goes).
*/
#ifndef DLT_NULL
#define DLT_NULL 0
#endif
#ifndef DLT_EN10MB
#define DLT_EN10MB 1
#endif
#ifndef DLT_EN3MB
#define DLT_EN3MB 2
#endif
#ifndef DLT_AX25
#define DLT_AX25 3
#endif
#ifndef DLT_PRONET
#define DLT_PRONET 4
#endif
#ifndef DLT_CHAOS
#define DLT_CHAOS 5
#endif
#ifndef DLT_IEEE802
#define DLT_IEEE802 6
#endif
#ifndef DLT_ARCNET
#define DLT_ARCNET 7
#endif
#ifndef DLT_SLIP
#define DLT_SLIP 8
#endif
#ifndef DLT_PPP
#define DLT_PPP 9
#endif
#ifndef DLT_FDDI
#define DLT_FDDI 10
#endif
#ifndef DLT_ATM_RFC1483
#define DLT_ATM_RFC1483 11
#endif
#ifndef DLT_RAW
#define DLT_RAW 12
#endif
#ifndef DLT_SLIP_BSDOS
#define DLT_SLIP_BSDOS 13
#endif
#ifndef DLT_PPP_BSDOS
#define DLT_PPP_BSDOS 14
#endif
#ifndef DLT_ATM_CLIP
#define DLT_ATM_CLIP 19
#endif
/*
* Step 3: for each of the types that's defined, put an entry in this
* table that maps its value to a corresponding WTAP_ENCAP_ value, if
* there is one.
*/
static const struct {
int dlt_value;
int wtap_encap_value;
} pcap_to_wtap_map[] = {
{ DLT_NULL, WTAP_ENCAP_NULL }, /* null encapsulation */
{ DLT_EN10MB, WTAP_ENCAP_ETHERNET },
{ DLT_IEEE802, WTAP_ENCAP_TR }, /* IEEE 802 Networks - assume token ring */
{ DLT_ARCNET, WTAP_ENCAP_ARCNET },
{ DLT_SLIP, WTAP_ENCAP_SLIP },
{ DLT_PPP, WTAP_ENCAP_PPP },
#ifdef BIT_SWAPPED_MAC_ADDRS
{ DLT_FDDI, WTAP_ENCAP_FDDI_BITSWAPPED },
#else
{ DLT_FDDI, WTAP_ENCAP_FDDI },
#endif
{ DLT_ATM_RFC1483, WTAP_ENCAP_ATM_RFC1483 },
{ DLT_RAW, WTAP_ENCAP_RAW_IP },
#ifdef DLT_CIP
{ DLT_CIP, WTAP_ENCAP_LINUX_ATM_CLIP },
#endif
#ifdef DLT_LOOP
{ DLT_LOOP, WTAP_ENCAP_NULL }, /* OpenBSD loopback */
#endif
{ DLT_ATM_CLIP, WTAP_ENCAP_LINUX_ATM_CLIP }
};
#define NUM_PCAP_ENCAPS (sizeof pcap_to_wtap_map / sizeof pcap_to_wtap_map[0])
int libpcap_open(wtap *wth, int *err)
{
int bytes_read;
guint32 magic;
struct pcap_hdr hdr;
gboolean byte_swapped;
gboolean modified;
int file_encap;
struct pcaprec_modified_hdr first_rec_hdr;
struct pcaprec_modified_hdr second_rec_hdr;
int hdr_len;
/* Read in the number that should be at the start of a "libpcap" file */
file_seek(wth->fh, 0, SEEK_SET);
wth->data_offset = 0;
errno = WTAP_ERR_CANT_READ;
bytes_read = file_read(&magic, 1, sizeof magic, wth->fh);
if (bytes_read != sizeof magic) {
*err = file_error(wth->fh);
if (*err != 0)
return -1;
return 0;
}
wth->data_offset += sizeof magic;
switch (magic) {
case PCAP_MAGIC:
/* Host that wrote it has our byte order, and was running
a program using either standard or ss990417 libpcap. */
byte_swapped = FALSE;
modified = FALSE;
break;
case PCAP_MODIFIED_MAGIC:
/* Host that wrote it has our byte order, and was running
a program using either ss990915 or ss991029 libpcap. */
byte_swapped = FALSE;
modified = TRUE;
break;
case PCAP_SWAPPED_MAGIC:
/* Host that wrote it has a byte order opposite to ours,
and was running a program using either standard or
ss990417 libpcap. */
byte_swapped = TRUE;
modified = FALSE;
break;
case PCAP_SWAPPED_MODIFIED_MAGIC:
/* Host that wrote it out has a byte order opposite to
ours, and was running a program using either ss990915
or ss991029 libpcap. */
byte_swapped = TRUE;
modified = TRUE;
break;
default:
/* Not a "libpcap" type we know about. */
return 0;
}
/* Read the rest of the header. */
errno = WTAP_ERR_CANT_READ;
bytes_read = file_read(&hdr, 1, sizeof hdr, wth->fh);
if (bytes_read != sizeof hdr) {
*err = file_error(wth->fh);
if (*err != 0)
return -1;
return 0;
}
wth->data_offset += sizeof hdr;
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. */
g_message("pcap: major version %u unsupported",
hdr.version_major);
*err = WTAP_ERR_UNSUPPORTED;
return -1;
}
file_encap = wtap_pcap_encap_to_wtap_encap(hdr.network);
if (file_encap == WTAP_ENCAP_UNKNOWN) {
g_message("pcap: network type %u unknown or unsupported",
hdr.network);
*err = WTAP_ERR_UNSUPPORTED_ENCAP;
return -1;
}
/* This is a libpcap file */
wth->file_type = modified ? WTAP_FILE_PCAP_SS991029 : WTAP_FILE_PCAP;
wth->capture.pcap = g_malloc(sizeof(libpcap_t));
wth->capture.pcap->byte_swapped = byte_swapped;
wth->capture.pcap->version_major = hdr.version_major;
wth->capture.pcap->version_minor = hdr.version_minor;
wth->subtype_read = libpcap_read;
wth->subtype_seek_read = wtap_def_seek_read;
wth->subtype_close = libpcap_close;
wth->file_encap = file_encap;
wth->snapshot_length = hdr.snaplen;
/*
* Yes. Let's look at the header for the first record,
* and see if, interpreting it as a standard header (if the
* magic number was standard) or a modified header (if the
* magic number was modified), the position where it says the
* header for the *second* record is contains a corrupted header.
*
* If so, then:
*
* If this file had the standard magic number, it may be
* an ss990417 capture file - in that version of Alexey's
* patch, the packet header format was changed but the
* magic number wasn't, and, alas, Red Hat appear to have
* picked up that version of the patch for RH 6.1, meaning
* RH 6.1 has a tcpdump that writes out files that can't
* be read by any software that expects non-modified headers
* if the magic number isn't the modified magic number (e.g.,
* any normal version of tcpdump, and Ethereal if we don't
* do this gross heuristic).
*
* If this file had the modified magic number, it may be
* an ss990915 capture file - in that version of Alexey's
* patch, the magic number was changed, but the record
* header had some extra fields, and, alas, SuSE appear
* to have picked up that version of the patch for SuSE
* 6.3, meaning that programs expecting the standard per-
* packet header in captures with the modified magic number
* can't read dumps from its tcpdump.
*/
hdr_len = modified ? sizeof (struct pcaprec_modified_hdr) :
sizeof (struct pcaprec_hdr);
bytes_read = file_read(&first_rec_hdr, 1, hdr_len, wth->fh);
if (bytes_read != hdr_len) {
*err = file_error(wth->fh);
if (*err != 0)
return -1; /* failed to read it */
/*
* Short read - assume the file isn't modified,
* and put the seek pointer back. The attempt
* to read the first packet will presumably get
* the same short read.
*/
goto give_up;
}
adjust_header(wth, &first_rec_hdr.hdr);
if (first_rec_hdr.hdr.incl_len > WTAP_MAX_PACKET_SIZE) {
/*
* The first record is bogus, so this is probably
* a corrupt file. Assume the file has the
* expected header type, and put the seek pointer
* back. The attempt to read the first packet will
* probably get the same bogus length.
*/
goto give_up;
}
file_seek(wth->fh,
wth->data_offset + hdr_len + first_rec_hdr.hdr.incl_len, SEEK_SET);
bytes_read = file_read(&second_rec_hdr, 1, hdr_len, wth->fh);
/*
* OK, does the next packet's header look sane?
*/
if (bytes_read != hdr_len) {
*err = file_error(wth->fh);
if (*err != 0)
return -1; /* failed to read it */
/*
* Short read - assume the file has the expected
* header type, and put the seek pointer back. The
* attempt to read the second packet will presumably get
* the same short read error.
*/
goto give_up;
}
adjust_header(wth, &second_rec_hdr.hdr);
if (second_rec_hdr.hdr.incl_len > WTAP_MAX_PACKET_SIZE) {
/*
* Oh, dear. Maybe it's a Capture File
* From Hell, and what looks like the
* "header" of the next packet is actually
* random junk from the middle of a packet.
* Try treating it as having the other type for
* the magic number it had; if that doesn't work,
* it probably *is* a corrupt file.
*/
wth->file_type = modified ? WTAP_FILE_PCAP_SS990915 :
WTAP_FILE_PCAP_SS990417;
}
give_up:
/*
* Restore the seek pointer.
*/
file_seek(wth->fh, wth->data_offset, SEEK_SET);
return 1;
}
/* Read the next packet */
static int libpcap_read(wtap *wth, int *err)
{
guint packet_size;
int bytes_to_read, bytes_read;
struct pcaprec_ss990915_hdr hdr;
int data_offset;
/* Read record header. */
errno = WTAP_ERR_CANT_READ;
switch (wth->file_type) {
case WTAP_FILE_PCAP:
bytes_to_read = sizeof (struct pcaprec_hdr);
break;
case WTAP_FILE_PCAP_SS990417:
case WTAP_FILE_PCAP_SS991029:
bytes_to_read = sizeof (struct pcaprec_modified_hdr);
break;
case WTAP_FILE_PCAP_SS990915:
bytes_to_read = sizeof (struct pcaprec_ss990915_hdr);
break;
default:
g_assert_not_reached();
bytes_to_read = 0;
}
bytes_read = file_read(&hdr, 1, bytes_to_read, wth->fh);
if (bytes_read != bytes_to_read) {
*err = file_error(wth->fh);
if (*err != 0)
return -1;
if (bytes_read != 0) {
*err = WTAP_ERR_SHORT_READ;
return -1;
}
return 0;
}
wth->data_offset += bytes_read;
adjust_header(wth, &hdr.hdr);
packet_size = hdr.hdr.incl_len;
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;
}
buffer_assure_space(wth->frame_buffer, packet_size);
data_offset = wth->data_offset;
errno = WTAP_ERR_CANT_READ;
bytes_read = file_read(buffer_start_ptr(wth->frame_buffer), 1,
packet_size, wth->fh);
if (bytes_read != packet_size) {
*err = file_error(wth->fh);
if (*err == 0)
*err = WTAP_ERR_SHORT_READ;
return -1;
}
wth->data_offset += packet_size;
wth->phdr.ts.tv_sec = hdr.hdr.ts_sec;
wth->phdr.ts.tv_usec = hdr.hdr.ts_usec;
wth->phdr.caplen = packet_size;
wth->phdr.len = hdr.hdr.orig_len;
wth->phdr.pkt_encap = wth->file_encap;
return data_offset;
}
static void
adjust_header(wtap *wth, struct pcaprec_hdr *hdr)
{
if (wth->capture.pcap->byte_swapped) {
/* Byte-swap the record header fields. */
hdr->ts_sec = BSWAP32(hdr->ts_sec);
hdr->ts_usec = BSWAP32(hdr->ts_usec);
hdr->incl_len = BSWAP32(hdr->incl_len);
hdr->orig_len = BSWAP32(hdr->orig_len);
}
/* 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 &&
hdr->incl_len > hdr->orig_len))) {
guint32 temp;
temp = hdr->orig_len;
hdr->orig_len = hdr->incl_len;
hdr->incl_len = temp;
}
}
static void
libpcap_close(wtap *wth)
{
g_free(wth->capture.pcap);
}
int wtap_pcap_encap_to_wtap_encap(int encap)
{
int i;
for (i = 0; i < NUM_PCAP_ENCAPS; i++) {
if (pcap_to_wtap_map[i].dlt_value == encap)
return pcap_to_wtap_map[i].wtap_encap_value;
}
return WTAP_ENCAP_UNKNOWN;
}
static int wtap_wtap_encap_to_pcap_encap(int encap)
{
int i;
/*
* Special-case WTAP_ENCAP_FDDI and WTAP_ENCAP_FDDI_BITSWAPPED;
* both of them get mapped to DLT_FDDI (even though that may
* mean that the bit order in the FDDI MAC addresses is wrong;
* so it goes - libpcap format doesn't record the byte order,
* so that's not fixable).
*/
if (encap == WTAP_ENCAP_FDDI || encap == WTAP_ENCAP_FDDI_BITSWAPPED)
return DLT_FDDI;
for (i = 0; i < NUM_PCAP_ENCAPS; i++) {
if (pcap_to_wtap_map[i].wtap_encap_value == encap)
return pcap_to_wtap_map[i].dlt_value;
}
return -1;
}
/* Returns 0 if we could write the specified encapsulation type,
an error indication otherwise. */
int libpcap_dump_can_write_encap(int filetype, int encap)
{
/* Per-packet encapsulations aren't supported. */
if (encap == WTAP_ENCAP_PER_PACKET)
return WTAP_ERR_ENCAP_PER_PACKET_UNSUPPORTED;
if (wtap_wtap_encap_to_pcap_encap(encap) == -1)
return WTAP_ERR_UNSUPPORTED_ENCAP;
return 0;
}
/* Returns TRUE on success, FALSE on failure; sets "*err" to an error code on
failure */
gboolean libpcap_dump_open(wtap_dumper *wdh, int *err)
{
guint32 magic;
struct pcap_hdr file_hdr;
int nwritten;
/* This is a libpcap file */
wdh->subtype_write = libpcap_dump;
wdh->subtype_close = NULL;
/* Write the file header. */
switch (wdh->file_type) {
case WTAP_FILE_PCAP:
case WTAP_FILE_PCAP_SS990417: /* modified, but with the old magic, sigh */
magic = PCAP_MAGIC;
break;
case WTAP_FILE_PCAP_SS990915: /* new magic, extra crap */
case WTAP_FILE_PCAP_SS991029:
magic = PCAP_MODIFIED_MAGIC;
break;
default:
/* We should never get h
[prev in list] [next in list] [prev in thread] [next in thread]
Configure |
About |
News |
Add a list |
Sponsored by KoreLogic