http://defect.opensolaris.org/bz/show_bug.cgi?id=9796



--- Comment #14 from Anurag S. Maskey <Anurag.Maskey at Sun.COM> 2009-08-31 
17:39:47 UTC ---
(In reply to comment #12)
> In the GUI, we determine that a wifi network is connected if at any point it
> had the "connected" flag set in the wifi_info from either a scan event or a
> wlan connect event, but also we check the state of the ncu, and if it's not
> online/active then we assume it's not really connected. But this is possibly
> flawed too given the information in other comments.

The link ncu will change to the online/up state once we've connected to a wifi
network.  

> So, I guess the question is what events can we expect to be seeing in the 
> cases
> mentioned, i.e. when a connection is made and a disconnect done. Since it 
> seems
> we're getting out of sync here.

When a wifi network is selected, a CONNECTION_REPORT event is sent detailing
whether the connection succeeded or not.  The GUI can use this to determine
which wifi network it is connected with.  A SCAN_REPORT is also sent with when
nwamd updates its representation of what it is connected to.  If a wifi network
is disconnected, the signal monitoring thread notices the disconnect and sends
a CONNECTION_REPORT.
The "connected" flag in nwam_wlan struct that is passed in these events says
whether we are connected to that wifi network or not.

> Similarly, at what point is it 100% that we've actually disconnected from the
> network, or looking it another way, what states/aux_states can we end up yet
> still be connected (thus inferring that if we're not in one of these, we're
> disconnected).

When the link ncu is online/up, we've definitely connected to a wifi network. 
When the link ncu is offline/down, offline/conditions not met, disabled, we
haven't connected to a wifi network.  The offline* and online* states are
transitioning so the next state change event should provide the actual state of
the conneciton.

-- 
Configure bugmail: http://defect.opensolaris.org/bz/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the QA contact for the bug.

Reply via email to