wifi: mac80211: add a per-PHY AQL limit to improve fairness

In order to maintain fairness, the amount of queueing needs to be limited
beyond the simple per-station AQL budget, otherwise the driver can simply
repeatedly do scheduling rounds until all queues that have not used their
AQL budget become eligble.

To be conservative, use the high AQL limit for the first txq and add half
of the low AQL for each subsequent queue.
Signed-off-by: Felix Fietkau <nbd@nbd.name>
Acked-by: Toke Høiland-Jørgensen <toke@toke.dk>
Link: https://lore.kernel.org/r/20220625212411.36675-5-nbd@nbd.nameSigned-off-by: Johannes Berg <johannes.berg@intel.com>

net/mac80211/ieee80211_i.h

@@ -1239,6 +1239,7 @@ struct ieee80211_local {
 	u32 aql_txq_limit_high[IEEE80211_NUM_ACS];
 	u32 aql_threshold;
 	atomic_t aql_total_pending_airtime;
+	atomic_t aql_ac_pending_airtime[IEEE80211_NUM_ACS];
 
 	const struct ieee80211_ops *ops;
 

net/mac80211/main.c

@@ -789,6 +789,7 @@ struct ieee80211_hw *ieee80211_alloc_hw_nm(size_t priv_data_len,
 		local->aql_txq_limit_low[i] = IEEE80211_DEFAULT_AQL_TXQ_LIMIT_L;
 		local->aql_txq_limit_high[i] =
 			IEEE80211_DEFAULT_AQL_TXQ_LIMIT_H;
+		atomic_set(&local->aql_ac_pending_airtime[i], 0);
 	}
 
 	local->airtime_flags = AIRTIME_USE_TX | AIRTIME_USE_RX;

net/mac80211/sta_info.c

@@ -2075,6 +2075,7 @@ void ieee80211_sta_update_pending_airtime(struct ieee80211_local *local,
 				   &sta->airtime[ac].aql_tx_pending);
 
 		atomic_add(tx_airtime, &local->aql_total_pending_airtime);
+		atomic_add(tx_airtime, &local->aql_ac_pending_airtime[ac]);
 		return;
 	}
 
@@ -2086,14 +2087,17 @@ void ieee80211_sta_update_pending_airtime(struct ieee80211_local *local,
 				       tx_pending, 0);
 	}
 
+	atomic_sub(tx_airtime, &local->aql_total_pending_airtime);
 	tx_pending = atomic_sub_return(tx_airtime,
-				       &local->aql_total_pending_airtime);
+				       &local->aql_ac_pending_airtime[ac]);
 	if (WARN_ONCE(tx_pending < 0,
 		      "Device %s AC %d pending airtime underflow: %u, %u",
 		      wiphy_name(local->hw.wiphy), ac, tx_pending,
-		      tx_airtime))
-		atomic_cmpxchg(&local->aql_total_pending_airtime,
+		      tx_airtime)) {
+		atomic_cmpxchg(&local->aql_ac_pending_airtime[ac],
 			       tx_pending, 0);
+		atomic_sub(tx_pending, &local->aql_total_pending_airtime);
+	}
 }
 
 int sta_info_move_state(struct sta_info *sta,

net/mac80211/tx.c

@@ -3846,6 +3846,9 @@ struct ieee80211_txq *ieee80211_next_txq(struct ieee80211_hw *hw, u8 ac)
 
 	spin_lock_bh(&local->active_txq_lock[ac]);
 
+	if (!local->schedule_round[ac])
+		goto out;
+
  begin:
 	txqi = list_first_entry_or_null(&local->active_txqs[ac],
 					struct txq_info,
@@ -3967,6 +3970,25 @@ bool ieee80211_txq_airtime_check(struct ieee80211_hw *hw,
 }
 EXPORT_SYMBOL(ieee80211_txq_airtime_check);
 
+static bool
+ieee80211_txq_schedule_airtime_check(struct ieee80211_local *local, u8 ac)
+{
+	unsigned int num_txq = 0;
+	struct txq_info *txq;
+	u32 aql_limit;
+
+	if (!wiphy_ext_feature_isset(local->hw.wiphy, NL80211_EXT_FEATURE_AQL))
+		return true;
+
+	list_for_each_entry(txq, &local->active_txqs[ac], schedule_order)
+		num_txq++;
+
+	aql_limit = (num_txq - 1) * local->aql_txq_limit_low[ac] / 2 +
+		    local->aql_txq_limit_high[ac];
+
+	return atomic_read(&local->aql_ac_pending_airtime[ac]) < aql_limit;
+}
+
 bool ieee80211_txq_may_transmit(struct ieee80211_hw *hw,
 				struct ieee80211_txq *txq)
 {
@@ -3983,6 +4005,9 @@ bool ieee80211_txq_may_transmit(struct ieee80211_hw *hw,
 	if (list_empty(&txqi->schedule_order))
 		goto out;
 
+	if (!ieee80211_txq_schedule_airtime_check(local, ac))
+		goto out;
+
 	list_for_each_entry_safe(iter, tmp, &local->active_txqs[ac],
 				 schedule_order) {
 		if (iter == txqi)
@@ -4022,7 +4047,15 @@ void ieee80211_txq_schedule_start(struct ieee80211_hw *hw, u8 ac)
 	struct ieee80211_local *local = hw_to_local(hw);
 
 	spin_lock_bh(&local->active_txq_lock[ac]);
-	local->schedule_round[ac]++;
+
+	if (ieee80211_txq_schedule_airtime_check(local, ac)) {
+		local->schedule_round[ac]++;
+		if (!local->schedule_round[ac])
+			local->schedule_round[ac]++;
+	} else {
+		local->schedule_round[ac] = 0;
+	}
+
 	spin_unlock_bh(&local->active_txq_lock[ac]);
 }
 EXPORT_SYMBOL(ieee80211_txq_schedule_start);