iwlwifi: mvm: support a000 SCD queue configuration

a000 devices queue management is going to change significantly.
We will have 512 queues. Those queues will be assigned number
by the firmware and not by the driver.

In addition, due to SN offload having TX queue shared between TIDs
is impossible

Also, the ADD_STA command no longer updates queues status.
The only point of changing queue in the SCD queue config API.

From driver perspective we have here a new design:

Queue sharing and inactivity checks are disabled.

Once this is done, the only paths that call scd_queue_cfg command
are paths that alloc and release TX queues - which will make future
accommodation to queue number assignment by FW easier.
Since allocating 512 queues statically is not advisable, transport
will allocate the queue on demand, fill the command with DRAM data
and send it. This is reflected in the new transport API.

Signed-off-by: Sara Sharon <sara.sharon@intel.com>
Signed-off-by: Luca Coelho <luciano.coelho@intel.com>

1 files changed, 57 insertions, 5 deletions

diff --git a/drivers/net/wireless/intel/iwlwifi/mvm/sta.c b/drivers/net/wireless/intel/iwlwifi/mvm/sta.c
index fb031197266e..e51b74cd5b9f 100644
--- a/drivers/net/wireless/intel/iwlwifi/mvm/sta.c
+++ b/drivers/net/wireless/intel/iwlwifi/mvm/sta.c
@@ -127,11 +127,17 @@ int iwl_mvm_sta_send_to_fw(struct iwl_mvm *mvm, struct ieee80211_sta *sta,
 	u32 agg_size = 0, mpdu_dens = 0;
 
 	if (!update || (flags & STA_MODIFY_QUEUES)) {
-		add_sta_cmd.tfd_queue_msk = cpu_to_le32(mvm_sta->tfd_queue_msk);
 		memcpy(&add_sta_cmd.addr, sta->addr, ETH_ALEN);
 
-		if (flags & STA_MODIFY_QUEUES)
-			add_sta_cmd.modify_mask |= STA_MODIFY_QUEUES;
+		if (!iwl_mvm_has_new_tx_api(mvm)) {
+			add_sta_cmd.tfd_queue_msk =
+				cpu_to_le32(mvm_sta->tfd_queue_msk);
+
+			if (flags & STA_MODIFY_QUEUES)
+				add_sta_cmd.modify_mask |= STA_MODIFY_QUEUES;
+		} else {
+			WARN_ON(flags & STA_MODIFY_QUEUES);
+		}
 	}
 
 	switch (sta->bandwidth) {
@@ -337,6 +343,9 @@ static int iwl_mvm_invalidate_sta_queue(struct iwl_mvm *mvm, int queue,
 	u8 sta_id;
 	int ret;
 
+	if (WARN_ON(iwl_mvm_has_new_tx_api(mvm)))
+		return -EINVAL;
+
 	spin_lock_bh(&mvm->queue_info_lock);
 	sta_id = mvm->queue_info[queue].ra_sta_id;
 	spin_unlock_bh(&mvm->queue_info_lock);
@@ -387,6 +396,9 @@ static int iwl_mvm_get_queue_agg_tids(struct iwl_mvm *mvm, int queue)
 
 	lockdep_assert_held(&mvm->mutex);
 
+	if (WARN_ON(iwl_mvm_has_new_tx_api(mvm)))
+		return -EINVAL;
+
 	spin_lock_bh(&mvm->queue_info_lock);
 	sta_id = mvm->queue_info[queue].ra_sta_id;
 	tid_bitmap = mvm->queue_info[queue].tid_bitmap;
@@ -426,6 +438,9 @@ static int iwl_mvm_remove_sta_queue_marking(struct iwl_mvm *mvm, int queue)
 
 	lockdep_assert_held(&mvm->mutex);
 
+	if (WARN_ON(iwl_mvm_has_new_tx_api(mvm)))
+		return -EINVAL;
+
 	spin_lock_bh(&mvm->queue_info_lock);
 	sta_id = mvm->queue_info[queue].ra_sta_id;
 	tid_bitmap = mvm->queue_info[queue].tid_bitmap;
@@ -468,6 +483,9 @@ static int iwl_mvm_free_inactive_queue(struct iwl_mvm *mvm, int queue,
 
 	lockdep_assert_held(&mvm->mutex);
 
+	if (WARN_ON(iwl_mvm_has_new_tx_api(mvm)))
+		return -EINVAL;
+
 	spin_lock_bh(&mvm->queue_info_lock);
 	txq_curr_ac = mvm->queue_info[queue].mac80211_ac;
 	sta_id = mvm->queue_info[queue].ra_sta_id;
@@ -512,6 +530,8 @@ static int iwl_mvm_get_shared_queue(struct iwl_mvm *mvm,
 	int i;
 
 	lockdep_assert_held(&mvm->queue_info_lock);
+	if (WARN_ON(iwl_mvm_has_new_tx_api(mvm)))
+		return -EINVAL;
 
 	memset(&ac_to_queue, IEEE80211_INVAL_HW_QUEUE, sizeof(ac_to_queue));
 
@@ -596,6 +616,9 @@ int iwl_mvm_scd_queue_redirect(struct iwl_mvm *mvm, int queue, int tid,
 	unsigned long mq;
 	int ret;
 
+	if (WARN_ON(iwl_mvm_has_new_tx_api(mvm)))
+		return -EINVAL;
+
 	/*
 	 * If the AC is lower than current one - FIFO needs to be redirected to
 	 * the lowest one of the streams in the queue. Check if this is needed
@@ -757,6 +780,15 @@ static int iwl_mvm_sta_alloc_queue(struct iwl_mvm *mvm,
 
 	/* No free queue - we'll have to share */
 	if (queue <= 0) {
+		/* This shouldn't happen in new HW - we have 512 queues */
+		if (WARN(iwl_mvm_has_new_tx_api(mvm),
+			 "No available queues for tid %d on sta_id %d\n",
+			 tid, cfg.sta_id)) {
+			spin_unlock_bh(&mvm->queue_info_lock);
+
+			return queue;
+		}
+
 		queue = iwl_mvm_get_shared_queue(mvm, tfd_queue_mask, ac);
 		if (queue > 0) {
 			shared_queue = true;
@@ -841,6 +873,9 @@ static int iwl_mvm_sta_alloc_queue(struct iwl_mvm *mvm,
 		mvmsta->reserved_queue = IEEE80211_INVAL_HW_QUEUE;
 	spin_unlock_bh(&mvmsta->lock);
 
+	if (iwl_mvm_has_new_tx_api(mvm))
+		return 0;
+
 	if (!shared_queue) {
 		ret = iwl_mvm_sta_send_to_fw(mvm, sta, true, STA_MODIFY_QUEUES);
 		if (ret)
@@ -880,6 +915,9 @@ static void iwl_mvm_change_queue_owner(struct iwl_mvm *mvm, int queue)
 
 	lockdep_assert_held(&mvm->mutex);
 
+	if (WARN_ON(iwl_mvm_has_new_tx_api(mvm)))
+		return;
+
 	spin_lock_bh(&mvm->queue_info_lock);
 	tid_bitmap = mvm->queue_info[queue].tid_bitmap;
 	spin_unlock_bh(&mvm->queue_info_lock);
@@ -917,6 +955,10 @@ static void iwl_mvm_unshare_queue(struct iwl_mvm *mvm, int queue)
 	int ssn;
 	int ret = true;
 
+	/* queue sharing is disabled on new TX path */
+	if (WARN_ON(iwl_mvm_has_new_tx_api(mvm)))
+		return;
+
 	lockdep_assert_held(&mvm->mutex);
 
 	spin_lock_bh(&mvm->queue_info_lock);
@@ -1675,7 +1717,8 @@ static int iwl_mvm_add_int_sta_common(struct iwl_mvm *mvm,
 	cmd.mac_id_n_color = cpu_to_le32(FW_CMD_ID_AND_COLOR(mac_id,
 							     color));
 
-	cmd.tfd_queue_msk = cpu_to_le32(sta->tfd_queue_msk);
+	if (!iwl_mvm_has_new_tx_api(mvm))
+		cmd.tfd_queue_msk = cpu_to_le32(sta->tfd_queue_msk);
 	cmd.tid_disable_tx = cpu_to_le16(0xffff);
 
 	if (addr)
@@ -2293,7 +2336,9 @@ int iwl_mvm_sta_tx_agg(struct iwl_mvm *mvm, struct ieee80211_sta *sta,
 	cmd.mac_id_n_color = cpu_to_le32(mvm_sta->mac_id_n_color);
 	cmd.sta_id = mvm_sta->sta_id;
 	cmd.add_modify = STA_MODE_MODIFY;
-	cmd.modify_mask = STA_MODIFY_QUEUES | STA_MODIFY_TID_DISABLE_TX;
+	if (!iwl_mvm_has_new_tx_api(mvm))
+		cmd.modify_mask = STA_MODIFY_QUEUES;
+	cmd.modify_mask |= STA_MODIFY_TID_DISABLE_TX;
 	cmd.tfd_queue_msk = cpu_to_le32(mvm_sta->tfd_queue_msk);
 	cmd.tid_disable_tx = cpu_to_le16(mvm_sta->tid_disable_agg);
 
@@ -2494,6 +2539,13 @@ int iwl_mvm_sta_tx_agg_oper(struct iwl_mvm *mvm, struct ieee80211_vif *vif,
 		 */
 		if (!alloc_queue && buf_size < mvmsta->max_agg_bufsize) {
 			/*
+			 * On new TX API rs and BA manager are offloaded.
+			 * For now though, just don't support being reconfigured
+			 */
+			if (iwl_mvm_has_new_tx_api(mvm))
+				return -ENOTSUPP;
+
+			/*
 			 * If reconfiguring an existing queue, it first must be
 			 * drained
 			 */