[PATCH 7/8] nl80211 driver: use monitor rather than management interface

[Johannes Berg]

> > +	/*
> > +	 * ugh, guess what, the generic code sets one of the version
> > +	 * bits to request tx callback
> > +	 */
> > +	hdr->frame_control &= ~host_to_le16(BIT(1));
> > +	return sendmsg(drv->monitor_sock, &msg, flags);
> 
> How is the TX callback/status requested with the radiotap version? It
> looks like there is a flag (IEEE80211_RADIOTAP_TX_FLAGS) for figuring
> out status for frames that were reported, but it was not immediately
> obvious to me when that information would be provided back to userspace
> for injected frames. Is this enabled for all injected frames by default
> or there something in the radiotap header that requests this?

Currently, it's implicitly done for all frames, even those that came
over the 802.3 interfaces, i.e. whenever we get a status notification we
show it on the monitor interfaces with the appropriate status. That will
need to change though.

> > --- /dev/null	1970-01-01 00:00:00.000000000 +0000
> > +++ hostap/hostapd/radiotap.c	2007-12-24 12:10:59.000000000 +0100
> > @@ -0,0 +1,290 @@
> > +/*
> > + * Radiotap parser
> > + *
> > + * Copyright 2007		Andy Green <andy at warmcat.com>
> > + *
> > + * This program is free software; you can redistribute it and/or modify
> > + * it under the terms of the GNU General Public License version 2 as
> > + * published by the Free Software Foundation.
> > + *
> > + * Alternatively, this software may be distributed under the terms of BSD
> > + * license.
> 
> Hmm.. This is based on code from Linux kernel (net/wireless/radiotap.c).
> Right? And that file does not seem to indicate anything special about
> license, so I would assume it is GPLv2. Has Andy licensed this code
> under dual-GPL/BSD license?

Yes, I asked him to license it under dual-BPL/BSD (which sort of makes
the GPL moot). CC'ed again, Andy please confirm.

> > --- /dev/null	1970-01-01 00:00:00.000000000 +0000
> > +++ hostap/hostapd/radiotap.h	2007-12-24 12:10:59.000000000 +0100
> > @@ -0,0 +1,242 @@
> > +/* $FreeBSD: src/sys/net80211/ieee80211_radiotap.h,v 1.5 2005/01/22 20:12:05 sam Exp $ */
> > +/* $NetBSD: ieee80211_radiotap.h,v 1.11 2005/06/22 06:16:02 dyoung Exp $ */
> > +
> > +/*-
> > + * Copyright (c) 2003, 2004 David Young.  All rights reserved.
> 
> ..
> 
> Do we need to distribute this file with hostapd? Is there not a standard
> user space header for defining radiotap headers? This file has slightly
> different license terms from rest of the files and as such, adds extra
> complexity on that area.

Oh, I didn't realise it had slightly different terms. We actually do
need the file unless we want to rewrite it, but I'm sure we can ask
David Young to adjust the terms to something we can agree with, after
all, rewriting it would mostly be mechanical and now even be possible
with the info I put on radiotap.org. I'll look into the exact terms and
ask David Young.

johannes
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 828 bytes
Desc: This is a digitally signed message part
Url : http://lists.shmoo.com/pipermail/hostap/attachments/20071225/c25dad63/attachment.pgp