Hi, all
This is based on the patch submitted by Simon [1]. An early test shows
that it could improve the UDP throughput of legacy STA from 16Mbps to
21Mbps. But not so significant in 11n, probably around 1 to 2Mbps
based on iperf measurement.
However, I am not too sure whether this is allowed in mesh setup and
perhaps someone can comment on this.
Anyway, feel free to test out and let me know.
Thanks
Regards,
Chun Yeow
diff --git a/net/mac80211/mesh.c b/net/mac80211/mesh.c
index 1bf03f9..ab1924d 100644
--- a/net/mac80211/mesh.c
+++ b/net/mac80211/mesh.c
@@ -607,6 +607,13 @@ void ieee80211_start_mesh(struct
ieee80211_sub_if_data *sdata)
{
struct ieee80211_if_mesh *ifmsh = &sdata->u.mesh;
struct ieee80211_local *local = sdata->local;
+ u32 changed;
+
+ changed = BSS_CHANGED_BEACON |
+ BSS_CHANGED_BEACON_ENABLED |
+ BSS_CHANGED_HT |
+ BSS_CHANGED_BASIC_RATES |
+ BSS_CHANGED_BEACON_INT;
local->fif_other_bss++;
/* mesh ifaces must set allmulti to forward mcast traffic */
@@ -628,11 +635,13 @@ void ieee80211_start_mesh(struct
ieee80211_sub_if_data *sdata)
sdata->vif.bss_conf.basic_rates =
ieee80211_mandatory_rates(sdata->local,
ieee80211_get_sdata_band(sdata));
- ieee80211_bss_info_change_notify(sdata, BSS_CHANGED_BEACON |
- BSS_CHANGED_BEACON_ENABLED |
- BSS_CHANGED_HT |
- BSS_CHANGED_BASIC_RATES |
- BSS_CHANGED_BEACON_INT);
+
+ if (sdata->local->oper_channel->band == IEEE80211_BAND_5GHZ) {
+ sdata->vif.bss_conf.use_short_slot = true;
+ changed |= BSS_CHANGED_ERP_SLOT;
+ }
+
+ ieee80211_bss_info_change_notify(sdata, changed);
netif_carrier_on(sdata->dev);
}
[1] http://www.spinics.net/lists/linux-wireless/msg100158.html
_______________________________________________
Devel mailing list
[email protected]
http://lists.open80211s.org/cgi-bin/mailman/listinfo/devel
