summaryrefslogtreecommitdiffstats
path: root/net/mac80211/rx.c
diff options
context:
space:
mode:
authorMichal Kazior <michal.kazior@tieto.com>2014-09-02 14:05:10 +0200
committerJohannes Berg <johannes.berg@intel.com>2014-09-03 13:40:38 +0200
commit4549cf2b1803d29cfd019f7bfeaa784f8f9c558f (patch)
treebf39255e299b752cea30ce8437b9a685b189e5d4 /net/mac80211/rx.c
parentnet: rfkill: gpio: Add more Broadcom bluetooth ACPI IDs (diff)
downloadlinux-4549cf2b1803d29cfd019f7bfeaa784f8f9c558f.tar.xz
linux-4549cf2b1803d29cfd019f7bfeaa784f8f9c558f.zip
mac80211: fix offloaded BA session traffic after hw restart
When starting an offloaded BA session it is unknown what starting sequence number should be used. Using last_seq worked in most cases except after hw restart. When hw restart is requested last_seq is (rightfully so) kept unmodified. This ended up with BA sessions being restarted with an aribtrary BA window values resulting in dropped frames until sequence numbers caught up. Instead of last_seq pick seqno of a first Rxed frame of a given BA session. This fixes stalled traffic after hw restart with offloaded BA sessions (currently only ath10k). Signed-off-by: Michal Kazior <michal.kazior@tieto.com> Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Diffstat (limited to 'net/mac80211/rx.c')
-rw-r--r--net/mac80211/rx.c10
1 files changed, 10 insertions, 0 deletions
diff --git a/net/mac80211/rx.c b/net/mac80211/rx.c
index a8d862f9183c..41eb12c87240 100644
--- a/net/mac80211/rx.c
+++ b/net/mac80211/rx.c
@@ -835,6 +835,16 @@ static bool ieee80211_sta_manage_reorder_buf(struct ieee80211_sub_if_data *sdata
spin_lock(&tid_agg_rx->reorder_lock);
+ /*
+ * Offloaded BA sessions have no known starting sequence number so pick
+ * one from first Rxed frame for this tid after BA was started.
+ */
+ if (unlikely(tid_agg_rx->auto_seq)) {
+ tid_agg_rx->auto_seq = false;
+ tid_agg_rx->ssn = mpdu_seq_num;
+ tid_agg_rx->head_seq_num = mpdu_seq_num;
+ }
+
buf_size = tid_agg_rx->buf_size;
head_seq_num = tid_agg_rx->head_seq_num;