There was a problem with mesh not enabling beacons in select
drivers (notably libertas_tf, though others are likely affected),
due to the addition of the SDATA_STATE_RUNNING flag at the end
of ieee80211_do_open. The problem is mesh gets started and
wants to send the beacon-enable message to the driver before
SDATA_STATE_RUNNING gets set, but ieee80211_bss_info_change_notify
has the following check before enabling beacons:
!ieee80211_sdata_running(sdata)
This patch adds an exception for mesh to the sdata running check,
to allow mesh to start beacons.
I don't know if this is the best solution, perhaps moving the
SDATA_STATE_RUNNING set, or a driver modification is in order,
but this seems the easiest and least dammaging way to do this.
For reference the SDATA_STATE_RUNNING stuff was added in commit
34d4bc4d about two weeks ago.
Signed-off-by: Steve deRosier <[email protected]>
---
net/mac80211/main.c | 4 +++-
1 files changed, 3 insertions(+), 1 deletions(-)
diff --git a/net/mac80211/main.c b/net/mac80211/main.c
index 4935b84..d1ad7fa 100644
--- a/net/mac80211/main.c
+++ b/net/mac80211/main.c
@@ -221,7 +221,9 @@ void ieee80211_bss_info_change_notify(struct ieee80211_sub_if_data *sdata,
}
if (changed & BSS_CHANGED_BEACON_ENABLED) {
- if (local->quiescing || !ieee80211_sdata_running(sdata) ||
+ if (local->quiescing ||
+ (!ieee80211_sdata_running(sdata) &&
+ sdata->vif.type != NL80211_IFTYPE_MESH_POINT) ||
test_bit(SCAN_SW_SCANNING, &local->scanning)) {
sdata->vif.bss_conf.enable_beacon = false;
} else {
--
1.7.0