them and running all taps on them, but not reconstructing the packet
list. Use that in the IO-stat tap rather than "redissect_packet()"; the
latter does more work and redraws the display, neither of which are
necessary.
Call the filter callback when the Calc field is changed, to redraw the
graphs; that change also fixes things so that it's called when the
Filter field is changed.
Rename the "filter_button" member of an io_stat_graph_t to
"filter_field", as it's not the "Filter:" button, it's the text field
containing the filter expression.
svn path=/trunk/; revision=9659
"gtk_iostat_reset()" call it, and have calls to
"gtk_iostat_reset(&io->graphs[0])" just call "io_stat_reset(io)".
Add a routine "io_stat_draw()" to draw an io_stat_t; have
"gtk_iostat_draw()" call it, and have calls to
"gtk_iostat_draw(&io->graphs[0])" just call "io_stat_draw(io)".
Add a routine "io_stat_redraw()" to set the "needs_redraw" flag on an
io_stat_t and call "io_stat_draw()" on it, in order to force a redraw.
Use that in place of set flag/io_stat_draw pairs.
Add a routine "enable_graph()" to set an io_stat_graph_t's display flag
to TRUE and register a tap listener for it, and add a routine
"disable_graph()" to, if an io_stat_graph_t's display flag is set, clear
it, remove its tap listener, and turn its display button off. Use
"disable_graph()" to disable graphs, rather than just turning the
display flag off, so that we ensure that the invariant "tap listener
present iff graph enabled" is preserved - the code to destroy an io-stat
window only removes tap listeners for enabled graphs, so it depends on
that invariant being preserved.
Check the validity of a display filter by compiling it, not by going
through the whole sequence of registering a tap listener.
svn path=/trunk/; revision=9656
DISSECTOR_SUPPORT_{SRC,OBJECTS}.
Add some additional files, required by dissectors, to those lists.
Extract the stuff to get version information strings for libraries and
the OS, which is *not* needed by dissectors, from "util.c", which
contains routines that *are* needed by dissectors, and put it into a
separate file.
Make "dftest" link only with the dissector support stuff, not with all
of the Ethereal common files.
svn path=/trunk/; revision=9645
the color filter's colors" bug: keep a list of removed color filters, so
there's no problem if we delete a color filter for which there are one or
more packets that have its colors. Only get rid of the lists when they
ar no longer needed.
svn path=/trunk/; revision=9613
to add functionality or improve performance, although, until we make it
work on GTK+ 1.3[.x] and 2.x, we shouldn't count on the performance
improvements, or make its API different from that of the GtkClist API
(other than names).
Move all the code that knows about the packet list into
gtk/packet_list.c, so that the GtkClist vs. EthClist stuff is
encapsulated inside it.
svn path=/trunk/; revision=9608
filter's colors" bug, by storing the pointer to the matching color_filter_t
structure in the frame_data structure.
Replace "frame" and "frames" by "packet" and "packets" in many places.
svn path=/trunk/; revision=9607
Now all we need is some enterprising individual to write the small amount of
preferences code to actually allow it to be changed etc, but I have to go to
Costco, so, later...
svn path=/trunk/; revision=9463
Windows - the problem is that GTK+ 1.3[.x] and later assume strings
handed to them are UTF-8 strings, not, for example, ISO 8859/x strings.
In packet-radius.c, re-define "isprint()" rather than #ifdeffing its use
(the old code was also incorrectly treating 0x7f as a printable).
svn path=/trunk/; revision=9436
currently-selected packet.
Use "cfile.count" as the total number of packets in the capture.
Use "plurality" to display "packet" vs. "packets".
svn path=/trunk/; revision=9423
Fix NaN when a capture and/or a filter don't have any matching packets
so the average packet size is obtained by dividing by zero.
svn path=/trunk/; revision=9369
use the common display filter dialog infrastructure in the MGCP
service response time tap;
add common infrastructure for updating the titles of tap dialogs
when the capture file name changes.
svn path=/trunk/; revision=9366
Do the same checks for negative and >255 version numbers in the
Tethereal DCE RPC statistics tap that are done in the Ethereal tap.
In the Ethereal tap, do those checks before looking up the protocol name.
svn path=/trunk/; revision=9290
"opened capture files" and "display filter used",
the settings will be saved in the file "recent" in the users config path
svn path=/trunk/; revision=9275
captures, as it has to compute the width of an auto-resizing column in
every row. Just pick fixed widths for the columns (and tune the width
of the "Protocol" column so that it's not narrower than the column
title).
svn path=/trunk/; revision=9219
pointers to the first *and* last child, in the "proto_node" structure
itself. That saves us one level of indirection and memory allocation,
and lets us append to a tree by appending to the last child directly,
rather than having to scan through the list of siblings of the first
child to find the end of that list.
svn path=/trunk/; revision=9171
last columns, if any, with that format, and use that to speed up
processing of columns with a particular format and checking whether
we're displaying a column with a particular format.
svn path=/trunk/; revision=9147
structure, rather than separately allocating "fvalue_t"s and having the
"field_info" structure point to them - this appears to speed up protocol
tree construction a bit.
svn path=/trunk/; revision=9146
"construct_args_t" is for use with filter dialogs, and the members other
than the title apply only to filter dialogs.
Have "select_file_cb()" actually use the title supplied to it.
svn path=/trunk/; revision=9125
rename it to select_file_cb to reflect its function.
While this cleans things up a bit, I am still not happy because now
filter_prefs.h must be included before file_dlg.h just to get
construct_args_t.
svn path=/trunk/; revision=9119
data to a file. This allows the user to select some stuff, and analyse it with
external tools, and is very useful for quickly prototying dissectors etc.
This works by retrieving the info that is needed to define where the selected
region is.
It puts up a dialog box that asks for the file to save in.
However, it is an ugly hack, because it reuses print_file_cb, while
print_file_cb should be moved into file_dlg.c.
It also needs to have some warning dialogs put up in error cases.
Finally, it would be good to be able to select a region in the byte_view,
which you can do with click and drag, and then have a menu item to save the
selected bytes.
svn path=/trunk/; revision=9116
This function is also very small, so small that teh overhead for the actual function call and return is likely to be a significant part
of its execution time.
change it into a macro and make it thus slightly faster by eliminating the function call overhead.
svn path=/trunk/; revision=9083
- added a pointer to raw rtp data to _rtp_info that can be used by
taps;
- RTP packets are passed to the tap queue only if they are not
error packets (so that you don't need to filter out ICMP
packets)
- use that pointer in rtp_stream, so it handles packets with
padding, and should handle RTP packets fragmented across
lower-level packets
- moved rtp_stream from tap sources to normal files
(prevents on-start-up registration of the rtp_stream tap
listener)
- rtp_stream tap gets registered/unregistered with the "RTP
Streams" dialog box
i.e. the tap is registered as long as the dialog box is open.
Alternatively, it is de-/registered on demand if RTP Analysis is
called directly on a packet.
- rtp_stream tap listener no longer uses a filter in dissection
` and does not need to have a proto tree being built.
(performance increase)
- fixed: RTP Streams list will get updated in real time if the
dialog box is open while a redissection takes place.
svn path=/trunk/; revision=9051
DISPLAY, and CLIENTNAME (in that order). If any of them are set, create
a capture filter that excludes their traffic and set it as the default.
The longer filters should be efficient without being overly long; they
may need some tweaking.
svn path=/trunk/; revision=8994
variables to be initialized to non-constant values (C89 says that "All
the expressions in an initializer for an object that has static storage
duration or in an initializer list for an object that has aggregate or
union type shall be constant expressions"; presumably the intent of the
former was to avoid run-time initialization and of the latter was to let
the initialization be done by copying from a compile-time-created blob
of memory), so we have to initialize "info->counts" by hand.
svn path=/trunk/; revision=8984
Make "proto_is_protocol_enabled()" and "proto_get_protocol_short_name()"
take a "protocol_t *" as an argument, so they don't have to look up the
"protocol_t" - this will probably speed them up considerably, and
they're called on almost every dissector handoff.
Get rid of a number of "proto_is_protocol_enabled()" calls that aren't
necessary (dissectors called through handles, including those called
through dissector tables, or called as heuristic dissectors, aren't even
called if their protocol isn't enabled).
Change some direct dissector calls to go through handles.
svn path=/trunk/; revision=8979
If we failed to dissect the GSS-SPNEGO blob it probably means that the segment
is somewhere in the middle of an LDAP PDU.
Just bail out and stop dissecting the PDU instead of aborting ethereal completely
using g_assert() since this is not really a pathological error, its just something that can and will happen normally.
svn path=/trunk/; revision=8925
open.
The toolbar equivalent should *not* be available if we have an "Update
list of packets in real time" capture running.
The toolbar "Save" button should not be available if we don't have an
unsaved capture file.
svn path=/trunk/; revision=8902
somewhat; this fixes a bug wherein we were assuming all modules
necessarily had preference pages (they won't, if the only preferences
for the module are obsolete preferences).
svn path=/trunk/; revision=8880
the interface isn't in the list of known (local) interfaces - that way
we don't try to get entries if the user's in the middle of changing the
"Interface:" text, or if the interface is an rpcap: URL (getting the
link-layer header type list for a remote interface is currently
impossible, and even just getting its *default* interface type could
hang for a long time if the remote machine isn't responding).
Free the link-layer header type list when we're done with it.
Label the option menu for that list "Link-layer header type", as it
doesn't control the data link type of the interface in the only case I
know of where it's settable, namely 802.11 interfaces - it just controls
whether the packets you get from the interface have a fake Ethernet
header or a real 802.11 header.
svn path=/trunk/; revision=8867
need to call it if you already have a pointer to the first entry in the
list, which is what "a pointer to the list" is.
svn path=/trunk/; revision=8866
- fix missing detection of first packet when writing payload.
- fix bug of erroneous handling of confort noise when writing payload.
- fix bug of possible endless silence insertion on first packet when
writing payload.
svn path=/trunk/; revision=8846
"Edit -> Display Filters..." menu item, as per Ulf Lamping - there's
apparently some problem wherein &args doesn't get passed properly to the
callback.
svn path=/trunk/; revision=8819
a button that brings up the Create Filter dialog box. While it works, the
problem is that it also acts as an Enter keypress as far as the start_stat
button is concerned.
Probably needs a small fix.
svn path=/trunk/; revision=8782
return "<no file>" if there is no file loaded yet instead of crashing in io-stat
io-stat only print the label for the top tick on the y axis to make it look less cluttered
svn path=/trunk/; revision=8781
"selected_tree_row_enabled()" routine, enable it by default, so that
tap windows can be popped up even if you have no capture file.
Assorted code cleanups.
svn path=/trunk/; revision=8740
- return FALSE immediately if the text entry is empty ;
- return FALSE at the end of the function so that the signal is sent to
the entry. It avoids a Gtk-ERROR (and an abort) :
file: gtkentry.c: line 4338 (blink_cb): assertion failed:
GTK_WIDGET_HAS_FOCUS(entry))
svn path=/trunk/; revision=8735
style.
Make the style text arrays static, as nobody uses them outside prefs.c.
Use FALSE and TRUE for the values for the Boolean controlling the
highlighting style.
Note that we're now using stock icons in the toolbar in GTK+ 2.x.
Put back the resizing of elements in the top-level container, at least
for GTK+ 1.2[.x]; otherwise, the toolbar's height never gets smaller,
even if you change the style in such a way as to reduce the height of
the elements (icons+text -> icons or text, icons -> text).
Make some routines and variables not used outside gtk/toolbar.c static.
svn path=/trunk/; revision=8720
- use GTK1/2 compatibility macros [GS]ET_OBJECT_DATA where needed
- add a set_toolbar_object_data() function which associates the display
filter entry (from the main window) with the E_DFILTER_TE_KEY for the
open and reload buttons (it is needed by the open and reload
callbacks). The function is called in create_main_window()
- reindent
svn path=/trunk/; revision=8718
of the filter text entry when reloading the file, and:
1) that doesn't work with the toolbar "reload" button (the
widget passed in for that button doesn't have a
E_DFILTER_TE_KEY data item pointing to the text entry);
2) that causes the Tools > Summary dialog box to report what
you've typed in that box, not the filter that's actually in
effect (i.e., it causes "cfile.dfilter" to reflect what's
been typed, not what's been applied);
so don't bother doing so. That also means that the "/File/Reload" menu
item doesn't need a E_DFILTER_TE_KEY data item, so don't give it one.
svn path=/trunk/; revision=8713
Ulf Lamping