ath10k: Fix bug in max. VHT A-MPDU size
authorSujith Manoharan <c_manoha@qca.qualcomm.com>
Tue, 8 Oct 2013 02:51:57 +0000 (19:51 -0700)
committerKalle Valo <kvalo@qca.qualcomm.com>
Tue, 8 Oct 2013 19:10:22 +0000 (22:10 +0300)
For VHT peers, the maximum A-MPDU size has to be calculated
from the VHT capabilities element and not the HT-cap. The formula
is the same, but a higher value is used in VHT, allowing larger
aggregates to be transmitted.

The patch contains a workaround for some Netgear/Linksys APs that
report Rx A-MPDU factor incorrectly.

Tested-by: Kalle Valo <kvalo@qca.qualcomm.com>
Signed-off-by: Sujith Manoharan <c_manoha@qca.qualcomm.com>
Signed-off-by: Michal Kazior <michal.kazior@tieto.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
drivers/net/wireless/ath/ath10k/mac.c

index 4b7c949..049eca2 100644 (file)
@@ -1033,14 +1033,27 @@ static void ath10k_peer_assoc_h_vht(struct ath10k *ar,
                                    struct wmi_peer_assoc_complete_arg *arg)
 {
        const struct ieee80211_sta_vht_cap *vht_cap = &sta->vht_cap;
+       u8 ampdu_factor;
 
        if (!vht_cap->vht_supported)
                return;
 
        arg->peer_flags |= WMI_PEER_VHT;
-
        arg->peer_vht_caps = vht_cap->cap;
 
+
+       ampdu_factor = (vht_cap->cap &
+                       IEEE80211_VHT_CAP_MAX_A_MPDU_LENGTH_EXPONENT_MASK) >>
+                      IEEE80211_VHT_CAP_MAX_A_MPDU_LENGTH_EXPONENT_SHIFT;
+
+       /* Workaround: Some Netgear/Linksys 11ac APs set Rx A-MPDU factor to
+        * zero in VHT IE. Using it would result in degraded throughput.
+        * arg->peer_max_mpdu at this point contains HT max_mpdu so keep
+        * it if VHT max_mpdu is smaller. */
+       arg->peer_max_mpdu = max(arg->peer_max_mpdu,
+                                (1U << (IEEE80211_HT_MAX_AMPDU_FACTOR +
+                                       ampdu_factor)) - 1);
+
        if (sta->bandwidth == IEEE80211_STA_RX_BW_80)
                arg->peer_flags |= WMI_PEER_80MHZ;