]> The Tcpdump Group git mirrors - tcpdump/log
tcpdump
19 years agoadd support for AS-PIC cookies
hannes [Thu, 12 May 2005 08:42:46 +0000 (08:42 +0000)] 
add support for AS-PIC cookies

19 years agoadd support for DLT_JUNIPER_PPPOE_ATM printer
hannes [Thu, 12 May 2005 07:23:04 +0000 (07:23 +0000)] 
add support for DLT_JUNIPER_PPPOE_ATM printer
replace the JUNIPER_ enums with real DLT_JUNIPER types
  (does this break environments where an old libpcap is present ?)

19 years agodo not dependend on eflag setting wether to copy cookie data or not - e.g. the atm...
hannes [Tue, 10 May 2005 10:47:47 +0000 (10:47 +0000)] 
do not dependend on eflag setting wether to copy cookie data or not - e.g. the atm printer needs to access the cookies

19 years ago-make the PPP printer multiline (behind vflag)
hannes [Sun, 8 May 2005 21:01:25 +0000 (21:01 +0000)] 
-make the PPP printer multiline (behind vflag)
-make more use of tok2str()
-improve unknown option hexprinting

19 years agoFix up the bounds checking for NetBIOS-over-TCP. Among other things,
guy [Sun, 8 May 2005 20:01:09 +0000 (20:01 +0000)] 
Fix up the bounds checking for NetBIOS-over-TCP.  Among other things,
this means that we don't go past the end of the captured data when
dumping raw packet data and that we distinguish between "this packet was
cut off by the snapshot length" and "this packet is continued in a
subsequent TCP segment".

19 years agoFix the bounds check for the chunk.
guy [Fri, 6 May 2005 10:53:20 +0000 (10:53 +0000)] 
Fix the bounds check for the chunk.

19 years agoBounds-check the individual components of a SAP reply.
guy [Fri, 6 May 2005 08:27:00 +0000 (08:27 +0000)] 
Bounds-check the individual components of a SAP reply.

Make "fn_printzp()" return 0 if we don't run past the end of the packet
and we don't find any padding NULs.

19 years agoAdd an "fn_printzp()" routine for printing null-padded strings (strings
guy [Fri, 6 May 2005 07:57:17 +0000 (07:57 +0000)] 
Add an "fn_printzp()" routine for printing null-padded strings (strings
with a maximum length, where a string shorter than that length is padded
with NULs), as "fn_print()" won't handle the maximum length *and* the
snapshot length and "fn_printn()" won't stop on a null string.  Use it
where appropriate.

Always pass "snapend" to "fn_print()" and "fn_printn()" if they're
passed a pointer into the packet data; only pass NULL if they're being
handed a pointer into a buffer that's not part of the packet data.

Always check the return value of "fn_print()", "fn_printn()", and
"fn_printzp()" if they're passed "snapend", and do the appropriate
string termination and "packet truncated" indication if they return 1.

20 years agoDo more bounds checking.
guy [Fri, 6 May 2005 04:19:39 +0000 (04:19 +0000)] 
Do more bounds checking.

20 years agoUse TCHECK2() rather than "if(!TTEST()) goto trunc".
guy [Fri, 6 May 2005 02:53:41 +0000 (02:53 +0000)] 
Use TCHECK2() rather than "if(!TTEST()) goto trunc".

Make some length and type values unsigned, as they can't be negative.
Don't check for them being negative.

Check for a TLV length less than the TLV header length.

20 years agoAdd a bunch of bounds checks.
guy [Fri, 6 May 2005 02:16:53 +0000 (02:16 +0000)] 
Add a bunch of bounds checks.

Don't bother copying the route header - we're using the EXTRACT_ macros,
so we don't have to worry about alignment.

Use the length field from the packet, not just the length of the
containing packet, and check to make sure the containing packet is long
enough to contain it.

Add a bunch of checks against the packet length.

In "default_print()" calls, cut the length off at the captured data
length.  In other calls, don't - let the bounds checks handle caplen <
length.

20 years agoAdd more bounds checks, and check for bogus chunk lengths (too short).
guy [Thu, 5 May 2005 23:08:59 +0000 (23:08 +0000)] 
Add more bounds checks, and check for bogus chunk lengths (too short).

20 years agoAdd bounds checking and other sanity checking.
guy [Thu, 5 May 2005 22:49:34 +0000 (22:49 +0000)] 
Add bounds checking and other sanity checking.

If "decode_prefix4()" returns -1, quit any loop using it to return the
length of an item.

20 years agoAlways check for "smb_fdata()" returning null - for example, check it on
guy [Thu, 5 May 2005 22:30:40 +0000 (22:30 +0000)] 
Always check for "smb_fdata()" returning null - for example, check it on
every iteration of a loop.

20 years agoAdded some new files to the cygnus and MSVC makefiles. The cignus makefile
risso [Wed, 4 May 2005 19:22:50 +0000 (19:22 +0000)] 
Added some new files to the cygnus and MSVC makefiles. The cignus makefile
still doesn't work because of the missing support for getnameinfo() into the
official release of cygwin.

20 years agoAdded again some #define's for MSVC compilers that were wrongly removed.
risso [Wed, 4 May 2005 19:20:20 +0000 (19:20 +0000)] 
Added again some #define's for MSVC compilers that were wrongly removed.

20 years agoadd support for DLT_JUNIPER_PPPOE, fix zero-length cookies
hannes [Tue, 3 May 2005 20:39:25 +0000 (20:39 +0000)] 
add support for DLT_JUNIPER_PPPOE, fix zero-length cookies

20 years agoadd boundary, infinite loop checks
hannes [Tue, 3 May 2005 08:23:07 +0000 (08:23 +0000)] 
add boundary, infinite loop checks

20 years agochange vendor private codepoint
hannes [Tue, 3 May 2005 08:12:31 +0000 (08:12 +0000)] 
change vendor private codepoint

20 years agoDocument "radio[M:N]".
guy [Mon, 2 May 2005 21:27:34 +0000 (21:27 +0000)] 
Document "radio[M:N]".

20 years agoauthentication present flag is 0x04 and not 0x40
hannes [Thu, 28 Apr 2005 09:28:47 +0000 (09:28 +0000)] 
authentication present flag is 0x04 and not 0x40

20 years agodo proper masking and shifting of the flag bits
hannes [Thu, 28 Apr 2005 07:31:02 +0000 (07:31 +0000)] 
do proper masking and shifting of the flag bits

20 years agoboundary check for authentication header
hannes [Wed, 27 Apr 2005 23:15:27 +0000 (23:15 +0000)] 
boundary check for authentication header

20 years agoadd support for BFDv1 as per draft-ietf-bfd-base-02
hannes [Wed, 27 Apr 2005 23:01:38 +0000 (23:01 +0000)] 
add support for BFDv1 as per draft-ietf-bfd-base-02

20 years agoGet rid of the multiple-include protection in our RPC headers, so we
guy [Wed, 27 Apr 2005 21:44:06 +0000 (21:44 +0000)] 
Get rid of the multiple-include protection in our RPC headers, so we
include them even if we also happen to have included system RPC headers,
as we want *our* versions of the structures (which use types with
explicit sizes for structure members, as we use them for on-the-wire
data).

Change pmap_prot.h to prefix all #defines and structure names with
"SUNRPC_" or "sunrpc_", to avoid collisions with any definitions and
structure names from the system header files.

20 years agoadditional check against zero length messages
hannes [Wed, 27 Apr 2005 18:56:51 +0000 (18:56 +0000)] 
additional check against zero length messages

20 years agoadd tracefiles for infinite loop testing
hannes [Wed, 27 Apr 2005 18:53:12 +0000 (18:53 +0000)] 
add tracefiles for infinite loop testing

20 years agoadd testfile for a normal sized slarp frame
hannes [Wed, 27 Apr 2005 14:50:45 +0000 (14:50 +0000)] 
add testfile for a normal sized slarp frame

20 years agomake sure that there are the full four timestamp bytes available
hannes [Wed, 27 Apr 2005 14:35:56 +0000 (14:35 +0000)] 
make sure that there are the full four timestamp bytes available

20 years agoadd testfile for a truncated slarp frame
hannes [Wed, 27 Apr 2005 14:34:25 +0000 (14:34 +0000)] 
add testfile for a truncated slarp frame

20 years agoFrom Alexander Dupuy: don't define our own [hn]to[nh][sl] macros if
guy [Wed, 27 Apr 2005 01:13:46 +0000 (01:13 +0000)] 
From Alexander Dupuy: don't define our own [hn]to[nh][sl] macros if
__ntohl is defined; __ntohl is defined in some OSes as a special asm
function that does the same sort of fast byte swapping, and that gets in
the way of our so defining it.

20 years agoAccording to RFC 2516, use of the End-Of-List tag is not required, and
guy [Tue, 26 Apr 2005 19:48:56 +0000 (19:48 +0000)] 
According to RFC 2516, use of the End-Of-List tag is not required, and
at least one PPPoE implementation doesn't put that tag into the tag
list, so we can't ignore the payload length.

20 years agoCall the maximum length #define SLARP_MAX_LEN rather than SLARP_LEN, to
guy [Tue, 26 Apr 2005 19:08:45 +0000 (19:08 +0000)] 
Call the maximum length #define SLARP_MAX_LEN rather than SLARP_LEN, to
better reflect what it is.

20 years agodifferentiate between SLARP_LEN and SLARP_MIN_LEN - otherwise we do print the trailin...
hannes [Tue, 26 Apr 2005 09:42:44 +0000 (09:42 +0000)] 
differentiate between SLARP_LEN and SLARP_MIN_LEN - otherwise we do print the trailing-junk message for all short frames (and we should it print for frames > 18

20 years agoMake "snap_print()" handle fetching and printing the OUI.
guy [Tue, 26 Apr 2005 07:27:15 +0000 (07:27 +0000)] 
Make "snap_print()" handle fetching and printing the OUI.

Don't copy the LLC header to a "struct llc", just construct the
individual fields.

Fetch the control field early in the dissection process, and check to
make sure we have at least 4 bytes of LLC header if it's not a U frame.
Fetch both bytes of the control field for I and S frames, and display
all 4 hex digits of it.

Only dissect UI frames as BPDUs, and pass the BPDU dissector a pointer
to the beginning of the BPDU, not the beginning of the LLC header - not
all BPDUs are encapsulated in LLC headers.

Only dissect UI frames as IP packets.

Do bounds checking in "snap_print()".

20 years agoDon't print ", " before the length if we're printing the link-layer
guy [Tue, 26 Apr 2005 07:14:25 +0000 (07:14 +0000)] 
Don't print ", " before the length if we're printing the link-layer
header, as the ", " isn't necessary, and looks ugly, in that case.

Use TCHECK() and TCHECK2() for bounds checking, and do more bounds
checking.

20 years agoDon't print the LLC header stuff if it's an 802.3 IPX frame - there
guy [Tue, 26 Apr 2005 03:52:00 +0000 (03:52 +0000)] 
Don't print the LLC header stuff if it's an 802.3 IPX frame - there
isn't an LLC header on those frames.

20 years agoDon't run past the end of the data when printing unknown data.
guy [Tue, 26 Apr 2005 03:44:36 +0000 (03:44 +0000)] 
Don't run past the end of the data when printing unknown data.

20 years agoProperly terminate a "struct tok" list.
guy [Tue, 26 Apr 2005 03:38:45 +0000 (03:38 +0000)] 
Properly terminate a "struct tok" list.

Add more "struct tok" tables for SNAP PIDs.

Don't print the source and/or destination addresses and SAPs for
undecoded packets if we printed them already with the "-e" flag.

20 years agoDon't show the packet type as an Ethernet type if it's not one.
guy [Tue, 26 Apr 2005 00:16:43 +0000 (00:16 +0000)] 
Don't show the packet type as an Ethernet type if it's not one.

20 years agoAdd some missing braces.
guy [Mon, 25 Apr 2005 19:28:56 +0000 (19:28 +0000)] 
Add some missing braces.

20 years agoDereference the *right* function pointer.
guy [Mon, 25 Apr 2005 19:17:50 +0000 (19:17 +0000)] 
Dereference the *right* function pointer.

20 years agoFrom Albert Chin: IBM's C compiler, at least for AIX 4.3.3, doesn't
guy [Mon, 25 Apr 2005 18:55:01 +0000 (18:55 +0000)] 
From Albert Chin: IBM's C compiler, at least for AIX 4.3.3, doesn't
allow a trailing comma on the last member of an enum.

20 years agoAdd a bunch of error checking.
guy [Mon, 25 Apr 2005 18:51:52 +0000 (18:51 +0000)] 
Add a bunch of error checking.

20 years agoHave separate "struct tok" PID-to-name tables for different OUIs; don't
guy [Mon, 25 Apr 2005 17:57:15 +0000 (17:57 +0000)] 
Have separate "struct tok" PID-to-name tables for different OUIs; don't

20 years agoGet rid of some extra blanks.
guy [Mon, 25 Apr 2005 17:30:16 +0000 (17:30 +0000)] 
Get rid of some extra blanks.

20 years agoBase all interpretion of IE contents on the code set.
guy [Mon, 25 Apr 2005 16:29:18 +0000 (16:29 +0000)] 
Base all interpretion of IE contents on the code set.

Get rid of an extra blank.

20 years agoDon't dump core if a null "struct tok" array pointer is passed to
guy [Mon, 25 Apr 2005 16:15:07 +0000 (16:15 +0000)] 
Don't dump core if a null "struct tok" array pointer is passed to
"tok2strbuf()".

20 years agobugfix: prevent infinite loop in ERO processing, print ERO subtype length
hannes [Mon, 25 Apr 2005 13:19:34 +0000 (13:19 +0000)] 
bugfix: prevent infinite loop in ERO processing, print ERO subtype length

20 years agocheck against min. TLV length / get if logic right
hannes [Mon, 25 Apr 2005 10:43:24 +0000 (10:43 +0000)] 
check against min. TLV length / get if logic right

20 years agobugfix: do not attempt to decode zero-length TLVs, only attempt to decode IS-REACH...
hannes [Mon, 25 Apr 2005 09:11:04 +0000 (09:11 +0000)] 
bugfix: do not attempt to decode zero-length TLVs, only attempt to decode IS-REACH (varlength) TLVs if the TLV length indicates that there is at least a single byte on the wire

20 years agoDon't exit just because an NSAP had a length of 0 or a too-long length;
guy [Mon, 25 Apr 2005 08:43:05 +0000 (08:43 +0000)] 
Don't exit just because an NSAP had a length of 0 or a too-long length;
just report it as an illegal length.

Check for some zero-length or too-short TLVs not already checked for.

20 years agoSplit a line to match the way it's split in libpcap's configure.in.
guy [Sun, 24 Apr 2005 01:36:19 +0000 (01:36 +0000)] 
Split a line to match the way it's split in libpcap's configure.in.

20 years agoAlbert Chin's patch to libpcap's configure.in:
guy [Sun, 24 Apr 2005 01:26:27 +0000 (01:26 +0000)] 
Albert Chin's patch to libpcap's configure.in:

AC_CHECK_DECLS will define the HAVE_DECL_xxx variable whether it
succeeds or fails, so you can't erase a "no" result and try
again with a different header.  Use AC_CHECK_DECL, which doesn't
define HAVE_DECL_xxx, and then explicitly define it based on
whether we found a declaration or not.

is required for tcpdump's configure.in as well.

20 years agoGive a URL for the Zephyr protocol spec.
guy [Thu, 21 Apr 2005 06:51:24 +0000 (06:51 +0000)] 
Give a URL for the Zephyr protocol spec.

20 years agoSquelch some unsigned vs. signed warnings.
guy [Thu, 21 Apr 2005 06:44:57 +0000 (06:44 +0000)] 
Squelch some unsigned vs. signed warnings.

20 years agoSquelch a signed vs. unsigned warning.
guy [Thu, 21 Apr 2005 06:36:05 +0000 (06:36 +0000)] 
Squelch a signed vs. unsigned warning.

20 years agoSquelch some GCC 4.0 signed-vs-unsigned warnings.
guy [Thu, 21 Apr 2005 04:09:58 +0000 (04:09 +0000)] 
Squelch some GCC 4.0 signed-vs-unsigned warnings.

20 years agoMake the string pointer in "struct juniper_cookie_table" just a "char
guy [Thu, 21 Apr 2005 04:04:01 +0000 (04:04 +0000)] 
Make the string pointer in "struct juniper_cookie_table" just a "char
*", so as to avoid GCC 4.0 warnings when initializing it from a string.

20 years agoDeclare "juniper_mfr_print()".
guy [Thu, 21 Apr 2005 03:59:25 +0000 (03:59 +0000)] 
Declare "juniper_mfr_print()".

20 years agoAC_TRY_COMPILE works only for code that fits inside "main()"; the test
guy [Thu, 21 Apr 2005 03:51:19 +0000 (03:51 +0000)] 
AC_TRY_COMPILE works only for code that fits inside "main()"; the test
for __attribute__ doesn't fit inside "main()" with GCC 4.0, as it
defines a function.

20 years agoPut back AC_LBL_C_INLINE; it turns out that wasn't the cause of the
guy [Thu, 21 Apr 2005 02:22:38 +0000 (02:22 +0000)] 
Put back AC_LBL_C_INLINE; it turns out that wasn't the cause of the
inline problem, and AC_LBL_C_INLINE checks for a case that some versions
of the HP C compiler don't handle, and only uses inline if that case
succeeds.

20 years agoGet rid of an unused variable.
guy [Wed, 20 Apr 2005 22:33:21 +0000 (22:33 +0000)] 
Get rid of an unused variable.

20 years agoGet rid of, or use, unused variables.
guy [Wed, 20 Apr 2005 22:31:09 +0000 (22:31 +0000)] 
Get rid of, or use, unused variables.

20 years agoGet rid of an unused variable.
guy [Wed, 20 Apr 2005 22:21:16 +0000 (22:21 +0000)] 
Get rid of an unused variable.

20 years agoGet rid of unused variables.
guy [Wed, 20 Apr 2005 22:19:06 +0000 (22:19 +0000)] 
Get rid of unused variables.

20 years agoCheck, and use, the length from the PAP header.
guy [Wed, 20 Apr 2005 22:15:34 +0000 (22:15 +0000)] 
Check, and use, the length from the PAP header.

20 years agoGet rid of an unused variable. (It should be possible to tell the type
guy [Wed, 20 Apr 2005 22:08:44 +0000 (22:08 +0000)] 
Get rid of an unused variable.  (It should be possible to tell the type
of address from the way it's printed.)

20 years agoAdd a comment asking about the address and mask length fields in
guy [Wed, 20 Apr 2005 22:05:27 +0000 (22:05 +0000)] 
Add a comment asking about the address and mask length fields in
join/prune requests.

20 years agoGet rid of an unused variable.
guy [Wed, 20 Apr 2005 21:55:32 +0000 (21:55 +0000)] 
Get rid of an unused variable.

20 years agoAttempt to squelch a warning from the SGI C compiler.
guy [Wed, 20 Apr 2005 21:52:31 +0000 (21:52 +0000)] 
Attempt to squelch a warning from the SGI C compiler.

20 years agoIf the length is specified, check to make sure it doesn't go past the
guy [Wed, 20 Apr 2005 21:50:16 +0000 (21:50 +0000)] 
If the length is specified, check to make sure it doesn't go past the
end of the packet, and isn't shorter than the header length.

Control messages have to have lengths.

20 years agoGet rid of an unused variable.
guy [Wed, 20 Apr 2005 21:36:27 +0000 (21:36 +0000)] 
Get rid of an unused variable.

20 years agoGet rid of an unused variable.
guy [Wed, 20 Apr 2005 21:02:15 +0000 (21:02 +0000)] 
Get rid of an unused variable.

20 years agoIf we get an ELT other than a bit label in labellen(), log it.
guy [Wed, 20 Apr 2005 20:59:00 +0000 (20:59 +0000)] 
If we get an ELT other than a bit label in labellen(), log it.

20 years agoGet rid of an unused variable.
guy [Wed, 20 Apr 2005 20:53:39 +0000 (20:53 +0000)] 
Get rid of an unused variable.

#if out some other unused stuff, and note how it *should* be used.

20 years agoCheck for a negative "advance" value, which indicates an error.
guy [Wed, 20 Apr 2005 20:46:05 +0000 (20:46 +0000)] 
Check for a negative "advance" value, which indicates an error.

Don't make the last statement in a loop a break statement, as that means
the body of the loop will be executed at most once.

When looping over the data in a TCP segment containing BGP packets, loop
until we reach the first of the end of the captured data or the end of
the segment, not just the end of the captured data.

20 years agoGet rid of an unused variable (the length from the BGP header is passed
guy [Wed, 20 Apr 2005 20:22:33 +0000 (20:22 +0000)] 
Get rid of an unused variable (the length from the BGP header is passed
as the "length" argument).

20 years agoAttempt to squelch a warning from the SGI C compiler.
guy [Wed, 20 Apr 2005 19:32:41 +0000 (19:32 +0000)] 
Attempt to squelch a warning from the SGI C compiler.

20 years agorefactor the juniper_parse header printer, add juniper FRF.16 dissector
hannes [Wed, 20 Apr 2005 12:45:06 +0000 (12:45 +0000)] 
refactor the juniper_parse header printer, add juniper FRF.16 dissector

20 years agoSome systems (*cough*Irix*cough) not only don't declare
guy [Wed, 20 Apr 2005 11:20:51 +0000 (11:20 +0000)] 
Some systems (*cough*Irix*cough) not only don't declare
"ether_ntohost()" but don't even declare "struct ether_addr"; on those
systems, declare "struct ether_addr".

20 years agoIf we don't have ether_ntohost(), don't bother declaring it merely
guy [Wed, 20 Apr 2005 10:51:02 +0000 (10:51 +0000)] 
If we don't have ether_ntohost(), don't bother declaring it merely
because it's not declared in a system header file, as the reason it's
not declared is that it doesn't exist.

20 years agoPut in some missing brackets.
guy [Wed, 20 Apr 2005 10:31:16 +0000 (10:31 +0000)] 
Put in some missing brackets.

20 years agoMove some statements before the case's "break;" statement, so that
guy [Wed, 20 Apr 2005 10:30:11 +0000 (10:30 +0000)] 
Move some statements before the case's "break;" statement, so that
they're executed; if they shouldn't be executed, they shouldn't be in
the code.

20 years ago"ipaddr_string()" takes, as an argument, a pointer to the bytes of an
guy [Wed, 20 Apr 2005 10:19:23 +0000 (10:19 +0000)] 
"ipaddr_string()" takes, as an argument, a pointer to the bytes of an
IPv4 address, not an extracted IPv4 address value.

20 years agoIRIX defines IPV6_VERSION in <netinet/in.h>, and we don't use it, so
guy [Wed, 20 Apr 2005 10:04:37 +0000 (10:04 +0000)] 
IRIX defines IPV6_VERSION in <netinet/in.h>, and we don't use it, so
remove the definition from ip6.h, to get rid of compiler warnings.

20 years agoFrom Albert Chin:
guy [Wed, 20 Apr 2005 09:51:03 +0000 (09:51 +0000)] 
From Albert Chin:

  1. aclocal.m4 patch required by Tru64 UNIX which has
     inttypes.h in a non-standard location so gcc cannot find
     the PRI types. So, use <sys/bitypes.h> to get u_int#_t
     types for help detecting %llx, etc.
  2. Help gcc on aix find ether_ntohost declaration.

20 years agoFrom Albert Chin:
guy [Wed, 20 Apr 2005 09:46:42 +0000 (09:46 +0000)] 
From Albert Chin:

  1. On AIX, AC_LBL_C_INLINE detected the compiler supported
     the inline keyword which is wrong. AC_C_INLINE from
     autoconf-2.59 worked.
  2. AC_CHECK_TYPE from autoconf-2.5x is no longer broken.
     Replaced AC_LBL_CHECK_TYPE with it, mainly to use
     <sys/bitypes.h> for Tru64 UNIX where some of the u_int#_t
     types are defined.
  3. Tru64 UNIX 4.0D doesn't support %llx; however, it does support %lx.
  4. Added <stdint.h> to interface.h for int#_t types on
     Tru64 UNIX 4.0D (required for missing/snprintf.c).
  5. Reworked includes in tcpdump-stdinc.h for int#_t types.

20 years agoFrom Albert Chin: IBM's AIX C compiler complains about "&buf[sizeof
guy [Wed, 20 Apr 2005 02:59:49 +0000 (02:59 +0000)] 
From Albert Chin: IBM's AIX C compiler complains about "&buf[sizeof
buf]", but doesn't complain about the equivalent "buf + sizeof(buf)".
(C89 allows a pointer into an array to point one element past the end of
the array, so both should, at least as I understand it, be legal, and we
decrement the resulting pointer before we use it, so both should be
safe.)

20 years agoDeclare "gndo" as "extern", so that it's only defined in tcpdump.c; on
guy [Tue, 19 Apr 2005 20:33:37 +0000 (20:33 +0000)] 
Declare "gndo" as "extern", so that it's only defined in tcpdump.c; on
some platforms, you get linker errors or, at least, warnings from
multiple definitions.

20 years agodon't call bittok2str() twice in a printf() statement, fix typo (cc_values vs cv_values)
hannes [Tue, 19 Apr 2005 20:19:50 +0000 (20:19 +0000)] 
don't call bittok2str() twice in a printf() statement, fix typo (cc_values vs cv_values)

20 years ago-bugfix: attempt to print ldp messages as long there is something on the wire
hannes [Tue, 19 Apr 2005 19:42:08 +0000 (19:42 +0000)] 
-bugfix: attempt to print ldp messages as long there is something on the wire
-improve support (add interface parameter TLV parser) for the Martini FEC TLV

20 years agobackout redundant Sequence # printing
hannes [Tue, 19 Apr 2005 12:49:21 +0000 (12:49 +0000)] 
backout redundant Sequence # printing

20 years agobugfix: don't forget to increment the tlv header size, misc. display cosmetics
hannes [Tue, 19 Apr 2005 12:18:28 +0000 (12:18 +0000)] 
bugfix: don't forget to increment the tlv header size, misc. display cosmetics

20 years agoGet rid of zero-length arrays in structures - they're not standard C89
guy [Tue, 19 Apr 2005 08:07:03 +0000 (08:07 +0000)] 
Get rid of zero-length arrays in structures - they're not standard C89
and not all C compilers support them.

The minimum length of a SLARP packet is 14 bytes, not 18 bytes.  SLARP
requests don't have any data, apparently.

20 years agoDocument the "portrange" keyword.
guy [Tue, 19 Apr 2005 04:40:30 +0000 (04:40 +0000)] 
Document the "portrange" keyword.

Update the date stamp.

20 years agoMake it clearer that a filter "ip" or "proto ip" or... refers to IPv4,
guy [Mon, 18 Apr 2005 01:38:30 +0000 (01:38 +0000)] 
Make it clearer that a filter "ip" or "proto ip" or... refers to IPv4,
not IPv6.

20 years agoFrom Noritoshi Demizu <[email protected]>: don't print "sack"
guy [Mon, 18 Apr 2005 00:23:05 +0000 (00:23 +0000)] 
From Noritoshi Demizu <[email protected]>: don't print "sack"
twice.

20 years agoBoost size of buffers so that they can handle a 32-bit integer; from
guy [Mon, 18 Apr 2005 00:08:02 +0000 (00:08 +0000)] 
Boost size of buffers so that they can handle a 32-bit integer; from
OpenBSD.

20 years agoPropagate
guy [Sun, 17 Apr 2005 01:38:26 +0000 (01:38 +0000)] 
Propagate

date: 2005/04/07 01:27:52;  author: mcr;  state: Exp;  lines: +5 -2
adjust #ifdef to work with NetBSD file as well, which
has slightly different multiple-include protection (alas)

from the main branch.