#include winsock2.h pulls in about 90 distinct .h files
and about 140 total .h files.
Currently winsock2.h is (mostly unnecessarily) included
for each dissector via packet.h/wtap.h.
This patch removes #include winsock2.h from wtap.h and
then includes winsock2.h (or windows.h) in the
few specific places required.
With this patch, my Windows Wireshark build takes
about 30% less time.
svn path=/trunk/; revision=26535
capinfos and dumpcap don't need to depend on libwireshark nor directly pull
in those modules). Because capinfos and editcap were only being linked with
privileges.c if we had plugins, this allows those programs to be linked when
someone is compiling --without-plugins.
svn path=/trunk/; revision=25640
A few changes from me:
- make use of nstime_set_unset and nstime_is_unset i.s.o. extra variable first_pass
- change 'if' to 'while' to allow intervals with no packets
- remove 'unused' variable current_pkt_ts
svn path=/trunk/; revision=25499
plugins should already be linked against libwireshark.
Don't link capinfos and editcap against libwireshark: they only needed to be
because the plugins were linked against libwireshark (see rev 24123 and the
ensuing discussion on -dev).
capinfos and editcap: don't complain if plugins fail to load: dissector
plugins should fail to load because they need libwireshark. I am assuming
here that wiretap plugins don't need libwireshark (I've never seen such a
plugin but LEGO's code and comments suggest this is the case).
(The goal of this checkin is to stop linking capinfos and editcap against
libwireshark while still allowing wiretap plugins. Since we don't have any
such plugins in the tree I do somewhat doubt the need for all this but I don't
want to be the one to remove the functionality.)
svn path=/trunk/; revision=24650
previously called. This prevents the function from always returning TRUE in
programs that hadn't called get_credential_info().
Call get_credential_info() in the programs that should have been.
svn path=/trunk/; revision=24648
meta information that is found at the beginning of
the data - this lets me fuzztest my .out files
properly.
Also make some whitespace more consistent.
svn path=/trunk/; revision=24401
"time_t" is not guaranteed to be an "unsigned long"; when printing it
with %lu, cast it to "unsigned long".
The "secs" field of a wtap_nstime, however, *is* defined to be a time_t;
there's no need to cast it to time_t.
svn path=/trunk/; revision=23036
configure is run with "--without-plugins"
cc1: warnings being treated as errors
about_dlg.c: In function ‘about_wireshark_cb’:
about_dlg.c:426: warning: unused variable ‘plugins_page’
make[2]: *** [about_dlg.o] Error 1
editcap.c: In function ‘main’:
editcap.c:663: error: ‘check_ts’ undeclared (first use in this function)
editcap.c:663: error: (Each undeclared identifier is reported only once
editcap.c:663: error: for each function it appears in.)
make[2]: *** [editcap.o] Error 1
svn path=/trunk/; revision=22761
epan/filesystem.c
have get_plugin_dir() calling init_plugin_dir() if necessary
epan/epan.c and epan/report_err.c
move the report_failure family into the new report_err.c file, have epan_init() calling the initializer
epan/plugins.h and epan/proto.c
do not have init_plugins() calling the proto_reg functions instead do it in init_proto()
gtk/main.c and tshark.c
init_plugin_dir() has become suprefluous
capinfos.c and editcap.c
load the wiretap plugins
Makefiles
do what's needed to build withe the above changes.
svn path=/trunk/; revision=21935
a patch to avoid the warning "implicit declaration of function
'strptime'" in editcap.c
glib.h is included just after the define __USE_XOPEN and include <time.h>
svn path=/trunk/; revision=20455
option explanation is more detailed now, I've added the option parameters to the description
added version information to the usage output
instead of using the usage page to display the available file and encapsulation types (which makes the usage page almost unreadable), use empty options -F or -T to print the available types. I've used optopt for this, it seems to be portable that way ...
svn path=/trunk/; revision=16991
- Editcap
Mikko Tiihonen filed bug 379 including a patch for editcap. This wasn't picked up so far. I've ported the patch to svn 16820 and included a documentation patch.
-packet-ieee80211.c
Radek Vokal of RedHat filed a bug found by Vladimir Kondratiev of Intel in the 802.11 dissector. Radek provided a sample capture and Vladimir a oneliner patch. I've ported the patch to svn 16820 and tested it against the provided capture. Works well.
-From Kan Sasaki
A patch for packet-ospf.c is attached:
- Fix the handling of the DN-bit of options field.
- Add a new function dissect_ospf_bitfield() to dissect a bitfield
such as options, flags. The following functions are merged by
using this function.
- dissect_ospf_lls_extended_options()
- dissect_ospf_dbd()
- dissect_ospf_options()
- dissect_ospf_v3_prefix_options()
- dissect the flags and prefix-options bitfield.
- lldp Bugfix Bug 596 LLDP TIA Network Policy Decode is not correct
- Camel make it possible to dissect based on OID.
svn path=/trunk/; revision=16822
currently limited to Ethereal and all the variants of libpcap filetypes only.
We might want to add output compression support to the other tools as well (tethereal, mergecap, ...).
We might also want to add support for the other filetypes, but this is only possible if the filetype functions doesn't use special output operations like fseek.
One bug is still left: if the input and output filetypes while saving are the same, Ethereal currently optimizes this by simply copy the binary file instead of using wiretap (so it will be faster but it will ignore the compress setting).
Don't know a good workaround for this, as I don't know a way to find out if the input file is currently compressed or not. One idea might be to use a heuristic on the filesize (compared to the packet size summmary). Another workaround I see is to remove this optimization, which is of course not the way I like to do it ...
svn path=/trunk/; revision=15804
Jeff Morriss