| Message ID | ea6cc1a5-40d3-444b-b9a4-2b460acb08f8@app.fastmail.com |
|---|---|
| State | Changes Requested |
| Headers | show |
| Series | Change logging level for CTRL-EVENT-SIGNAL-CHANGE to MSG_DEBUG | expand |
On Thu, Nov 28, 2024 at 12:18:37AM +0900, Kan-Ru Chen wrote: > The default logging level for the CTRL-EVENT-SIGNAL-CHANGE message > may be repeated many times and fill the log file or journal. > > For example https://bugzilla.redhat.com/show_bug.cgi?id=2309148 and > the first few results from searching CTRL-EVENT-SIGNAL-CHANGE on the > web contain various complaints and workarounds. > > Changing the logging level to MSG_DEBUG would allow the message to > be still useful for debugging. This event is not really for debugging purposes, but for providing it to any upper layer component that is listening to event message in the wpa_supplicant control interface. > diff --git a/wpa_supplicant/events.c b/wpa_supplicant/events.c > @@ -6755,7 +6755,7 @@ void wpa_supplicant_event(void *ctx, enum wpa_event_type event, > case EVENT_SIGNAL_CHANGE: > - wpa_msg(wpa_s, MSG_INFO, WPA_EVENT_SIGNAL_CHANGE > + wpa_msg(wpa_s, MSG_DEBUG, WPA_EVENT_SIGNAL_CHANGE > "above=%d signal=%d noise=%d txrate=%lu", This would break existing uses for the wpa_supplicant control interface. I'm not completely sure how those system log messages get collected, but a more appropriate change might be to replace this wpa_msg() call with a call to wpa_msg_ctrl() with the exact same set of parameters. That would remove the print from stdout and hopefully from the undesired system log mechanisms.
On Mon, Dec 23, 2024, at 7:21 AM, Jouni Malinen wrote: > On Thu, Nov 28, 2024 at 12:18:37AM +0900, Kan-Ru Chen wrote: >> The default logging level for the CTRL-EVENT-SIGNAL-CHANGE message >> may be repeated many times and fill the log file or journal. >> >> For example https://bugzilla.redhat.com/show_bug.cgi?id=2309148 and >> the first few results from searching CTRL-EVENT-SIGNAL-CHANGE on the >> web contain various complaints and workarounds. >> >> Changing the logging level to MSG_DEBUG would allow the message to >> be still useful for debugging. > > This event is not really for debugging purposes, but for providing it to > any upper layer component that is listening to event message in the > wpa_supplicant control interface. > >> diff --git a/wpa_supplicant/events.c b/wpa_supplicant/events.c >> @@ -6755,7 +6755,7 @@ void wpa_supplicant_event(void *ctx, enum wpa_event_type event, >> case EVENT_SIGNAL_CHANGE: >> - wpa_msg(wpa_s, MSG_INFO, WPA_EVENT_SIGNAL_CHANGE >> + wpa_msg(wpa_s, MSG_DEBUG, WPA_EVENT_SIGNAL_CHANGE >> "above=%d signal=%d noise=%d txrate=%lu", > > This would break existing uses for the wpa_supplicant control interface. > I'm not completely sure how those system log messages get collected, but > a more appropriate change might be to replace this wpa_msg() call with a > call to wpa_msg_ctrl() with the exact same set of parameters. That would > remove the print from stdout and hopefully from the undesired system log > mechanisms. Thanks for pointing me to wpa_msg_ctrl(). The comment for wpa_msg_ctrl() also mentioned it can be used for frequent events that do not need to be sent to syslog. I'll send an updated patch. Kan-Ru
diff --git a/wpa_supplicant/events.c b/wpa_supplicant/events.c index a7c56f771..c411f5faf 100644 --- a/wpa_supplicant/events.c +++ b/wpa_supplicant/events.c @@ -6755,7 +6755,7 @@ void wpa_supplicant_event(void *ctx, enum wpa_event_type event, data->eapol_rx.encrypted); break; case EVENT_SIGNAL_CHANGE: - wpa_msg(wpa_s, MSG_INFO, WPA_EVENT_SIGNAL_CHANGE + wpa_msg(wpa_s, MSG_DEBUG, WPA_EVENT_SIGNAL_CHANGE "above=%d signal=%d noise=%d txrate=%lu", data->signal_change.above_threshold, data->signal_change.data.signal,
The default logging level for the CTRL-EVENT-SIGNAL-CHANGE message may be repeated many times and fill the log file or journal. For example https://bugzilla.redhat.com/show_bug.cgi?id=2309148 and the first few results from searching CTRL-EVENT-SIGNAL-CHANGE on the web contain various complaints and workarounds. Changing the logging level to MSG_DEBUG would allow the message to be still useful for debugging. Signed-off-by: Kan-Ru Chen <kanru@kanru.info> --- wpa_supplicant/events.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)