https://bugs.wireshark.org/bugzilla/show_bug.cgi?id=5187
Bill Meier <wmeier@xxxxxxxxxxx> changed:
What |Removed |Added
----------------------------------------------------------------------------
Status|NEW |RESOLVED
Resolution| |FIXED
--- Comment #5 from Bill Meier <wmeier@xxxxxxxxxxx> 2011-03-10 18:27:14 EST ---
For the record:
> a note: for the fourth time, it had enough data that it opened a dialog window
> indicating the sort progress. the program crashed immediately after the sort
> was completed and the dialog went away
In this case (when a dialog window appears during the sort) the crash was
caused by a "race" problem. The following pseudo-code reflects the actual
Wireshark code and demonstrates the problem.
sort:
max_row = list_nrows;
for row = 1 to max_row {
columnize_list_row(row); // fill in fields req'd by sort
update_dialog_window(); // processes waiting events (like:
>> // "interrupt" event processed which does the following
>> append_new_row_to_list();
>> list_nrows += 1;
}
sort_list(list_nrows); // fails since appended row(s) not columnized
---------
> #1 begin capturing data
> #2 in the capture window, click on the No., Time, Source, Destination,
> Protocol, or Info headers until crash.
> for me, step 2 took 1 click the first time, 2 clicks the second time,
> 15 clicks the 3rd time,
In these cases (no dialog window ?) I imagine there is a similar "race
condition" as above (altho I don't immediately see the exact mechanism).
============
In any case: SVN #34126 was committed (9/23/10) to disable sorting while a
capture is in progress thus preventing the above scenarios.
It might be nice to find a better solution so that sorts could be allowed while
capturing. See Bug #4273.
--
Configure bugmail: https://bugs.wireshark.org/bugzilla/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.
