On Fri, 26 Mar 2010 15:27:28 -0700
Erik Nordmark <Erik.Nordmark at Sun.COM> wrote:
> On 03/25/10 11:21 PM, Michael Hunter wrote:
> > On Fri, 26 Mar 2010 13:37:54 +0800
> > Lin Ma<Lin.Ma at Sun.COM> wrote:
> >
> >> According to the code of nwam_event_wait, only msgrcv will return the
> >> error code EINVAL or E2BIG which are translated to NWAM_INVALID_ARG. I
> >> think this bug need be investigated from libnwam side.
> >
> > Hmmm, Looks like we are willing to send more then we are willing to
> > receive. nwamd thought the max number of wlans was 64 and libnwam
> > thought it was 32. Erik try the attached libnwam.
>
> Too late - I already left the IETF.
>
> Next attempt to test in that environment will be in July.
Bummer.
>
> I don't know if you guys have access to APs that can be configured to
> advertise multiple essids - and if that can be used to setup > 32 and >
> 64 essids.
I don't know if we do, but I think simnet could do that. At one point
it was limited to 10 APs, but that shouldn't be hard to change.
>
> What happens when we have more than 64 essids and this fix?
That was why I asked what is just below your signature :) The best way
to deal with this would be to not have an arbitrary limit. It makes
the ipc slightly easier if there is no limit. But not so much easier I
want to define what happens at the limit.
Michael
>
> Erik
>
> > We should also do some testing at the limits and make sure we do
> > something reasonable when there are more then we want to deal with.
> >
> > FWIW the GUI shouldn't spin when it gets back an error.
> >
> > Michael
> >
> >>
> >> Thanks,
> >> lin
> >>
> >> On 03/26/10 10:20 AM, Lin Ma wrote:
> >>> I'm now looking http://defect.opensolaris.org/bz/show_bug.cgi?id=15329
> >>> Maybe they are the same bug. Anyway, please send me the core file. I
> >>> will take a look.
> >>>
> >>> Thanks,
> >>> lin
> >>>
> >>>>> Also, the --debug generates some information about the persistent
> >>>>> popups. The relevant parts seem to be
> >>>>> ** (nwam-manager.orig:11564): DEBUG: ESSID ietf-sec life state 1
> >>>>> ** (nwam-manager.orig:11564): DEBUG: - 39: -- ESSID ietf-sec BSSID
> >>>>> 0:21:1c:ff:c2:a0
> >>>>> ** (nwam-manager.orig:11564): DEBUG: ESSID ietf-sec life state 1
> >>>>> ** (nwam-manager.orig:11564): DEBUG: - 40: -- ESSID ietf-sec BSSID
> >>>>> 0:24:f7:bd:60:80
> >>>>> ** (nwam-manager.orig:11564): DEBUG: ESSID ietf-sec life state 1
> >>>>> ** (nwam-manager.orig:11564): DEBUG: Event wait error: argument is
> >>>>> invalid
> >>>>> ** (nwam-manager.orig:11564): DEBUG: Attempting to reopen connection
> >>>>> to daemon
> >>>>> ** (nwam-manager.orig:11564): DEBUG: nwamui_daemon_nwam_disconnect:
> >>>>> Closing connection to nwam daemon
> >>>>> ** (nwam-manager.orig:11564): DEBUG: nwamui_daemon_set_status:
> >>>>> Daemon status changed to ERROR
> >>>>> ** (nwam-manager.orig:11564): DEBUG: daemon_status_changed:
> >>>>> StatusIcon (Debug): show self now!
> >>>>> ** (nwam-manager.orig:11564): DEBUG: nwam_notification_msg_new:
> >>>>> NOTIFICATION_MESSAGE: Summary = 'Automatic network configuration
> >>>>> daemon is unavailable.' ; Body = 'For further information please run
> >>>>> "svcs -xv nwam" in a terminal.' ; Action = 'NULL' ; Label = 'NULL'
> >>>>>
> >>>>> The "arguemnt is inavlid" looks like a potential internal problem.
> >>>>>
> >>>>> Who wants to look at the debug output?
> >>>
> >>> _______________________________________________
> >>> nwam-dev mailing list
> >>> nwam-dev at opensolaris.org
> >>> http://mail.opensolaris.org/mailman/listinfo/nwam-dev
> >>
> >> _______________________________________________
> >> nwam-dev mailing list
> >> nwam-dev at opensolaris.org
> >> http://mail.opensolaris.org/mailman/listinfo/nwam-dev
> >>
> >>
> >> _______________________________________________
> >> nwam-dev mailing list
> >> nwam-dev at opensolaris.org
> >> http://mail.opensolaris.org/mailman/listinfo/nwam-dev
>
> _______________________________________________
> nwam-dev mailing list
> nwam-dev at opensolaris.org
> http://mail.opensolaris.org/mailman/listinfo/nwam-dev
