diff options
| author | Peter Wu <peter@lekensteyn.nl> | 2017-06-12 16:21:29 +0200 |
|---|---|---|
| committer | Peter Wu <peter@lekensteyn.nl> | 2017-06-17 14:22:56 +0000 |
| commit | 553b1e0215ad5826906aafbe0e38c1e44cf9e1c3 (patch) | |
| tree | 3cf0b3d5d3e668785146f64b099bcce10d792ae8 /ui | |
| parent | 800a856fb4784e31347e15a8ec825d2daeb62350 (diff) | |
Do not process UI events while reading from live capture
For at least Qt, the main_window_update callback is not necessary to
make the stop button work. When restarting a live capture during a
flood (via Ctrl-R), this callback actually results in an infinite loop
in MainWindow::captureStop since the capture state never changes from
FILE_READ_IN_PROGRESS.
Remove this callback to ensure that the problematic
pipeActivated / sync_pipe_input_cb / capture_input_new_packets /
main_window_update / ... / on_actionCaptureRestart_triggered /
testCaptureFileClose / captureStop sequence is avoided.
Even though captureStop invokes capture_stop, I guess that this does not
change the state because the pipeActivated callback is already active.
Bug: 10917
Change-Id: I6ca4fa946963928b7bc8a53ca14f9a9a3a35eaa7
Reviewed-on: https://code.wireshark.org/review/22097
Reviewed-by: Jaap Keuter <jaap.keuter@xs4all.nl>
Reviewed-by: Peter Wu <peter@lekensteyn.nl>
Diffstat (limited to 'ui')
| -rw-r--r-- | ui/capture.c | 6 |
1 files changed, 0 insertions, 6 deletions
diff --git a/ui/capture.c b/ui/capture.c index f0615e82fa..67282f70a3 100644 --- a/ui/capture.c +++ b/ui/capture.c @@ -396,12 +396,6 @@ capture_input_new_packets(capture_session *cap_session, int to_read) capture_callback_invoke(capture_cb_capture_fixed_continue, cap_session); } - /* update the main window so we get events (e.g. from the stop toolbar button) */ - /* This causes a hang on Windows (see bug 7305). Do we need this on any platform? */ -#ifndef _WIN32 - main_window_update(); -#endif - if(capture_opts->show_info) capture_info_new_packets(to_read, cap_session->cap_data_info); } |