On Mon, Nov 26, 2007 at 05:44:31PM +0100, Quentin Garnier wrote:
> On Mon, Nov 26, 2007 at 10:24:27AM -0600, David Young wrote:
> > On Mon, Nov 26, 2007 at 05:16:57PM +0100, Quentin Garnier wrote:
> > > On Mon, Nov 26, 2007 at 10:13:05AM -0600, David Young wrote:
> > > > Crypto works differently on wi in hostap mode than otherwise: as you
> > > > mention below, Tx WEP is broken, Rx WEP works, and net80211 is not
> > > > especially accommodating of that asymmetry.
> > > 
> > > Yes but my point here is that it seems to me that the RX bug should
> > > happen in modes other than hostap.  I don't see anything specific to
> > > hostap in that code path.
> > 
> > I am guessing that in other modes, the NIC decapsulates WEP frames after
> > decrypting, and it also strips the WEP flag.
> 
> But then _F_DROPUNENC would fire just the same way, wouldn't it?

You're right.  However, wi contains a grotty workaround.
See wi_mend_flags().

Dave

-- 
David Young             OJC Technologies
dyoung@ojctech.com      Urbana, IL * (217) 278-3933 ext 24