Skip site navigation (1)Skip section navigation (2)
Date:      Sun, 15 Jul 2007 23:13:46 +1200
From:      Andrew Thompson <thompsa@FreeBSD.org>
To:        Matteo Riondato <matteo@freebsd.org>, src-committers@FreeBSD.org, cvs-src@FreeBSD.org, cvs-all@FreeBSD.org
Subject:   Re: cvs commit: src/sys/dev/if_ndis if_ndis.c if_ndisvar.h
Message-ID:  <20070715111346.GJ95956@heff.fud.org.nz>
In-Reply-To: <20070713230833.GA2642@krapfengeist.dei.unipd.it>
References:  <200707120254.l6C2s5Yg041022@repoman.freebsd.org> <20070713230833.GA2642@krapfengeist.dei.unipd.it>

next in thread | previous in thread | raw e-mail | index | archive | help
On Sat, Jul 14, 2007 at 01:08:33AM +0200, Matteo Riondato wrote:
> On Thu, Jul 12, 2007 at 02:54:05AM +0000, Andrew Thompson wrote:
> > thompsa     2007-07-12 02:54:05 UTC
> > 
> >   FreeBSD src repository
> > 
> >   Modified files:
> >     sys/dev/if_ndis      if_ndis.c if_ndisvar.h 
> >   Log:
> >   Improve the net80211 handling within ndis
> >    - use net80211 for scanning and pass the results back to the scan cache
> >    - use ieee80211_init_channels to fill our channel list
> >    - fix up state transitions
> >    - depreciate the old wicontrol ioctls
> >    - add some debugging lines (#define NDIS_DEBUG)
> 
> I wonder whether this commit can be guilty of my wireless connection
> (which uses ndis) no longer working. I rebuild my system today and when
> I rebooted my machine can no longer connect to the network. "ifconfig
> ndis0 scan" seems not to work any more. 
> I will be pleased to give more debug information, if you tell me what
> you need. 

You can turn on net80211 debugging my setting sysctl
net.wlan.0.debug=0xffffffff (or use wlandebug). Kick off a scan and see
if anything is returned.

The ndis code uses a highly suspicious 2 second delay between issuing the
scan to the card and reading the results. It has always been like this
but the changes I have recently done put this in the same execution path.
Before it would return to the userland context and then read the
results, this probably gave the card a little more time to respond.

As a test you could bump up the wait time on the tsleep on line 3510 of
if_ndis.c


cheers,
Andrew




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20070715111346.GJ95956>