Re: [Acx100-devel] [PATCH] acxsm: merge from acx 0.3.32

[Denis V. <vd...@il...>]

On Saturday 04 February 2006 00:34, Carlos Mart=C3=ADn wrote:
> On Friday 03 February 2006 13:14, Denis Vlasenko wrote:
> > Standalone acx driver had several fixes since
> > acxsm fork, this patch merges them:
> > - initial support for new TNETW1450 USB chip
> > - support for firmware 2.3.1.31
> >=20
> > Also we had one report that acxsm is actually working.
> > That's quite unexpected.
>=20
> For me it doesn't work with WEP, though it does discover two APs which me=
ans=20
> non-encrypted send/receive works.
>=20
> Is there any actual reason why we want to do host-side WEP encryption or =
is it=20
> just for testing? I'll see if I can get it to work card-side in the next =
few=20
> days.

Implementing as much of 802.11 as possible on the host-side means that
we invest development time in code which will be usable with vast number
of different cards. Also we don't need to discover and work around
firmware bugs (multiply this by number of different hardware vendors
and then multiply this again by number ofdifferent firmware versions).

Example:

                        /* BTW acx111 firmware would not send probe respons=
es
                        ** if probe request does not have all basic rates f=
lagged
                        ** by 0x80! Thus firmware does not conform to 802.1=
1,
                        ** it should ignore 0x80 bit in ratevector from STA.
                        ** We can 'fix' it by not using this template and
                        ** sending probe responses by hand. TODO --vda */

When host side WEP/AES/etc crypto will be working 100% ok, we can
implement/debug device-assisted one.
=2D-
vda