I made this patch to test RTS/CTS operation in drivers, do you think
it's useful apart that and should be merged?
---
Documentation/networking/mac80211-injection.txt | 6 ++++++
net/mac80211/tx.c | 10 ++++++++++
2 files changed, 16 insertions(+)
--- everything.orig/net/mac80211/tx.c 2008-02-19 18:41:18.401563530 +0100
+++ everything/net/mac80211/tx.c 2008-02-19 18:41:21.631563150 +0100
@@ -840,6 +840,7 @@ __ieee80211_parse_tx_radiotap(struct iee
struct ieee80211_supported_band *sband;
int ret = ieee80211_radiotap_iterator_init(&iterator, rthdr, skb->len);
struct ieee80211_tx_control *control = tx->u.tx.control;
+ u16 val16;
sband = tx->local->hw.wiphy->bands[tx->local->hw.conf.channel->band];
@@ -922,6 +923,15 @@ __ieee80211_parse_tx_radiotap(struct iee
tx->flags |= IEEE80211_TXRXD_FRAGMENTED;
break;
+ case IEEE80211_RADIOTAP_TX_FLAGS:
+ val16 = le16_to_cpu(get_unaligned((__le16 *) iterator.this_arg));
+
+ if (val16 & IEEE80211_RADIOTAP_F_TX_RTS)
+ control->flags |= IEEE80211_TXCTL_USE_RTS_CTS;
+ else if (val16 & IEEE80211_RADIOTAP_F_TX_CTS)
+ control->flags |= IEEE80211_TXCTL_USE_CTS_PROTECT;
+ break;
+
/*
* Please update the file
* Documentation/networking/mac80211-injection.txt
--- everything.orig/Documentation/networking/mac80211-injection.txt 2008-02-18 15:29:41.068223795 +0100
+++ everything/Documentation/networking/mac80211-injection.txt 2008-02-19 18:41:21.661561252 +0100
@@ -40,6 +40,12 @@ radiotap headers and used to control inj
this flag is only reliable when software
fragmentation is enabled)
+ * IEEE80211_RADIOTAP_TX_FLAGS
+
+ IEEE80211_RADIOTAP_F_TX_RTS: use RTS/CTS handshake
+ IEEE80211_RADIOTAP_F_TX_CTS: use CTS-to-self protection
+
+
The injection code can also skip all other currently defined radiotap fields
facilitating replay of captured radiotap headers directly.
Somebody in the thread at some point said:
> I made this patch to test RTS/CTS operation in drivers, do you think
> it's useful apart that and should be merged?
> + * IEEE80211_RADIOTAP_TX_FLAGS
> +
> + IEEE80211_RADIOTAP_F_TX_RTS: use RTS/CTS handshake
> + IEEE80211_RADIOTAP_F_TX_CTS: use CTS-to-self protection
> +
> +
> The injection code can also skip all other currently defined radiotap fields
> facilitating replay of captured radiotap headers directly.
Hi Johannes -
Sure the footprint is nothing to speak of and it just exposes flags
there already supported using radiotap flags already existing.
I didn't think of a use for it but if the original flag has any use then
no doubt it can be used by the injection path by some enterprising soul.
(Sorry for late reply, travelling)
-Andy
On Thu, 2008-02-21 at 11:11 -0500, John W. Linville wrote:
> On Thu, Feb 21, 2008 at 01:17:14AM +0100, Johannes Berg wrote:
> > I made this patch to test RTS/CTS operation in drivers, do you think
> > it's useful apart that and should be merged?
>
> I don't see why not...what would be the objection?
You can't specify "Don't send RTS/CTS" because not specifying either
flag means "automatic".
johannes
On Thu, Feb 21, 2008 at 01:17:14AM +0100, Johannes Berg wrote:
> I made this patch to test RTS/CTS operation in drivers, do you think
> it's useful apart that and should be merged?
I don't see why not...what would be the objection?
--
John W. Linville
[email protected]