[PATCH ath-next 0/4] wifi: ath11k: add support for Tx Power insertion

[PATCH ath-next 0/4] wifi: ath11k: add support for Tx Power insertion

[Aditya Kumar Singh]

For certain action frames like the TPC Report IE in the spectrum management
TPC Report action frame, and in the Radio Measurement Link Measurement
Report action frame there is a requirement to fill in the current
and max Tx power of the device in the packet.

Add support to populate these fields in the relevant packets. Advertise
this capability from the driver using the feature flag
NL80211_FEATURE_TX_POWER_INSERTION.

In software-encrypted cases such as PMF, skip insertion since the packets
are already encrypted and cannot be modified.

Patch Overview:
Patches 1 and 2 serve as preparatory groundwork.
Patches 3 and 4 implement the core changes described.

For context, this ath11k implementation is adapted from prior ath12k work [1].

Reference:
[1] https://lore.kernel.org/linux-wireless/20250630-support-for-tx-power-insertion-v1-0-77f45484d5bb@oss.qualcomm.com/

---
Aditya Kumar Singh (4):
      wifi: ath11k: relocate some Tx power related functions in mac.c
      wifi: ath11k: wrap ath11k_mac_op_get_txpower() with lock-aware internal helper
      wifi: ath11k: add support for Tx Power insertion in RRM action frame
      wifi: ath11k: advertise NL80211_FEATURE_TX_POWER_INSERTION

 drivers/net/wireless/ath/ath11k/mac.c | 447 ++++++++++++++++++++++++----------
 1 file changed, 312 insertions(+), 135 deletions(-)
---
base-commit: d5ce93f136fbee2b7afbe221f34ca881036f8de3
change-id: 20251008-add_tx_power_insertion_support-e5225e924bfd

[PATCH ath-next 1/4] wifi: ath11k: relocate some Tx power related functions in mac.c

[Aditya Kumar Singh]

A forthcoming change necessitates that these functions be defined prior to
their usage. Therefore, relocate them now as a preparatory step for the
upcoming modifications.

Compile tested only.

Signed-off-by: Aditya Kumar Singh <aditya.kumar.singh@oss.qualcomm.com>
---
 drivers/net/wireless/ath/ath11k/mac.c | 270 +++++++++++++++++-----------------
 1 file changed, 135 insertions(+), 135 deletions(-)

diff --git a/drivers/net/wireless/ath/ath11k/mac.c b/drivers/net/wireless/ath/ath11k/mac.c
index 0e41b5a91d66..a7bd0ddf5fdf 100644
--- a/drivers/net/wireless/ath/ath11k/mac.c
+++ b/drivers/net/wireless/ath/ath11k/mac.c
@@ -4028,6 +4028,141 @@ static int ath11k_start_scan(struct ath11k *ar,
 	return 0;
 }
 
+static void ath11k_mac_fw_stats_reset(struct ath11k *ar)
+{
+	spin_lock_bh(&ar->data_lock);
+	ath11k_fw_stats_pdevs_free(&ar->fw_stats.pdevs);
+	ath11k_fw_stats_vdevs_free(&ar->fw_stats.vdevs);
+	ar->fw_stats.num_vdev_recvd = 0;
+	ar->fw_stats.num_bcn_recvd = 0;
+	spin_unlock_bh(&ar->data_lock);
+}
+
+int ath11k_mac_fw_stats_request(struct ath11k *ar,
+				struct stats_request_params *req_param)
+{
+	struct ath11k_base *ab = ar->ab;
+	unsigned long time_left;
+	int ret;
+
+	lockdep_assert_held(&ar->conf_mutex);
+
+	ath11k_mac_fw_stats_reset(ar);
+
+	reinit_completion(&ar->fw_stats_complete);
+	reinit_completion(&ar->fw_stats_done);
+
+	ret = ath11k_wmi_send_stats_request_cmd(ar, req_param);
+
+	if (ret) {
+		ath11k_warn(ab, "could not request fw stats (%d)\n",
+			    ret);
+		return ret;
+	}
+
+	time_left = wait_for_completion_timeout(&ar->fw_stats_complete, 1 * HZ);
+	if (!time_left)
+		return -ETIMEDOUT;
+
+	/* FW stats can get split when exceeding the stats data buffer limit.
+	 * In that case, since there is no end marking for the back-to-back
+	 * received 'update stats' event, we keep a 3 seconds timeout in case,
+	 * fw_stats_done is not marked yet
+	 */
+	time_left = wait_for_completion_timeout(&ar->fw_stats_done, 3 * HZ);
+	if (!time_left)
+		return -ETIMEDOUT;
+
+	return 0;
+}
+
+static int ath11k_mac_get_fw_stats(struct ath11k *ar, u32 pdev_id,
+				   u32 vdev_id, u32 stats_id)
+{
+	struct ath11k_base *ab = ar->ab;
+	struct stats_request_params req_param;
+	int ret;
+
+	lockdep_assert_held(&ar->conf_mutex);
+
+	if (ar->state != ATH11K_STATE_ON)
+		return -ENETDOWN;
+
+	req_param.pdev_id = pdev_id;
+	req_param.vdev_id = vdev_id;
+	req_param.stats_id = stats_id;
+
+	ret = ath11k_mac_fw_stats_request(ar, &req_param);
+	if (ret)
+		ath11k_warn(ab, "failed to request fw stats: %d\n", ret);
+
+	ath11k_dbg(ab, ATH11K_DBG_WMI,
+		   "debug get fw stat pdev id %d vdev id %d stats id 0x%x\n",
+		   pdev_id, vdev_id, stats_id);
+
+	return ret;
+}
+
+static int ath11k_mac_op_get_txpower(struct ieee80211_hw *hw,
+				     struct ieee80211_vif *vif,
+				     unsigned int link_id,
+				     int *dbm)
+{
+	struct ath11k *ar = hw->priv;
+	struct ath11k_base *ab = ar->ab;
+	struct ath11k_fw_stats_pdev *pdev;
+	int ret;
+
+	/* Final Tx power is minimum of Target Power, CTL power, Regulatory
+	 * Power, PSD EIRP Power. We just know the Regulatory power from the
+	 * regulatory rules obtained. FW knows all these power and sets the min
+	 * of these. Hence, we request the FW pdev stats in which FW reports
+	 * the minimum of all vdev's channel Tx power.
+	 */
+	mutex_lock(&ar->conf_mutex);
+
+	/* Firmware doesn't provide Tx power during CAC hence no need to fetch
+	 * the stats.
+	 */
+	if (test_bit(ATH11K_CAC_RUNNING, &ar->dev_flags)) {
+		mutex_unlock(&ar->conf_mutex);
+		return -EAGAIN;
+	}
+
+	ret = ath11k_mac_get_fw_stats(ar, ar->pdev->pdev_id, 0,
+				      WMI_REQUEST_PDEV_STAT);
+	if (ret) {
+		ath11k_warn(ab, "failed to request fw pdev stats: %d\n", ret);
+		goto err_fallback;
+	}
+
+	spin_lock_bh(&ar->data_lock);
+	pdev = list_first_entry_or_null(&ar->fw_stats.pdevs,
+					struct ath11k_fw_stats_pdev, list);
+	if (!pdev) {
+		spin_unlock_bh(&ar->data_lock);
+		goto err_fallback;
+	}
+
+	/* tx power is set as 2 units per dBm in FW. */
+	*dbm = pdev->chan_tx_power / 2;
+
+	spin_unlock_bh(&ar->data_lock);
+	mutex_unlock(&ar->conf_mutex);
+
+	ath11k_dbg(ar->ab, ATH11K_DBG_MAC, "txpower from firmware %d, reported %d dBm\n",
+		   pdev->chan_tx_power, *dbm);
+	return 0;
+
+err_fallback:
+	mutex_unlock(&ar->conf_mutex);
+	/* We didn't get txpower from FW. Hence, relying on vif->bss_conf.txpower */
+	*dbm = vif->bss_conf.txpower;
+	ath11k_dbg(ar->ab, ATH11K_DBG_MAC, "txpower from firmware NaN, reported %d dBm\n",
+		   *dbm);
+	return 0;
+}
+
 static int ath11k_mac_op_hw_scan(struct ieee80211_hw *hw,
 				 struct ieee80211_vif *vif,
 				 struct ieee80211_scan_request *hw_req)
@@ -9079,81 +9214,6 @@ static void ath11k_mac_put_chain_rssi(struct station_info *sinfo,
 	}
 }
 
-static void ath11k_mac_fw_stats_reset(struct ath11k *ar)
-{
-	spin_lock_bh(&ar->data_lock);
-	ath11k_fw_stats_pdevs_free(&ar->fw_stats.pdevs);
-	ath11k_fw_stats_vdevs_free(&ar->fw_stats.vdevs);
-	ar->fw_stats.num_vdev_recvd = 0;
-	ar->fw_stats.num_bcn_recvd = 0;
-	spin_unlock_bh(&ar->data_lock);
-}
-
-int ath11k_mac_fw_stats_request(struct ath11k *ar,
-				struct stats_request_params *req_param)
-{
-	struct ath11k_base *ab = ar->ab;
-	unsigned long time_left;
-	int ret;
-
-	lockdep_assert_held(&ar->conf_mutex);
-
-	ath11k_mac_fw_stats_reset(ar);
-
-	reinit_completion(&ar->fw_stats_complete);
-	reinit_completion(&ar->fw_stats_done);
-
-	ret = ath11k_wmi_send_stats_request_cmd(ar, req_param);
-
-	if (ret) {
-		ath11k_warn(ab, "could not request fw stats (%d)\n",
-			    ret);
-		return ret;
-	}
-
-	time_left = wait_for_completion_timeout(&ar->fw_stats_complete, 1 * HZ);
-	if (!time_left)
-		return -ETIMEDOUT;
-
-	/* FW stats can get split when exceeding the stats data buffer limit.
-	 * In that case, since there is no end marking for the back-to-back
-	 * received 'update stats' event, we keep a 3 seconds timeout in case,
-	 * fw_stats_done is not marked yet
-	 */
-	time_left = wait_for_completion_timeout(&ar->fw_stats_done, 3 * HZ);
-	if (!time_left)
-		return -ETIMEDOUT;
-
-	return 0;
-}
-
-static int ath11k_mac_get_fw_stats(struct ath11k *ar, u32 pdev_id,
-				   u32 vdev_id, u32 stats_id)
-{
-	struct ath11k_base *ab = ar->ab;
-	struct stats_request_params req_param;
-	int ret;
-
-	lockdep_assert_held(&ar->conf_mutex);
-
-	if (ar->state != ATH11K_STATE_ON)
-		return -ENETDOWN;
-
-	req_param.pdev_id = pdev_id;
-	req_param.vdev_id = vdev_id;
-	req_param.stats_id = stats_id;
-
-	ret = ath11k_mac_fw_stats_request(ar, &req_param);
-	if (ret)
-		ath11k_warn(ab, "failed to request fw stats: %d\n", ret);
-
-	ath11k_dbg(ab, ATH11K_DBG_WMI,
-		   "debug get fw stat pdev id %d vdev id %d stats id 0x%x\n",
-		   pdev_id, vdev_id, stats_id);
-
-	return ret;
-}
-
 static void ath11k_mac_op_sta_statistics(struct ieee80211_hw *hw,
 					 struct ieee80211_vif *vif,
 					 struct ieee80211_sta *sta,
@@ -9539,66 +9599,6 @@ static int ath11k_mac_op_remain_on_channel(struct ieee80211_hw *hw,
 	return ret;
 }
 
-static int ath11k_mac_op_get_txpower(struct ieee80211_hw *hw,
-				     struct ieee80211_vif *vif,
-				     unsigned int link_id,
-				     int *dbm)
-{
-	struct ath11k *ar = hw->priv;
-	struct ath11k_base *ab = ar->ab;
-	struct ath11k_fw_stats_pdev *pdev;
-	int ret;
-
-	/* Final Tx power is minimum of Target Power, CTL power, Regulatory
-	 * Power, PSD EIRP Power. We just know the Regulatory power from the
-	 * regulatory rules obtained. FW knows all these power and sets the min
-	 * of these. Hence, we request the FW pdev stats in which FW reports
-	 * the minimum of all vdev's channel Tx power.
-	 */
-	mutex_lock(&ar->conf_mutex);
-
-	/* Firmware doesn't provide Tx power during CAC hence no need to fetch
-	 * the stats.
-	 */
-	if (test_bit(ATH11K_CAC_RUNNING, &ar->dev_flags)) {
-		mutex_unlock(&ar->conf_mutex);
-		return -EAGAIN;
-	}
-
-	ret = ath11k_mac_get_fw_stats(ar, ar->pdev->pdev_id, 0,
-				      WMI_REQUEST_PDEV_STAT);
-	if (ret) {
-		ath11k_warn(ab, "failed to request fw pdev stats: %d\n", ret);
-		goto err_fallback;
-	}
-
-	spin_lock_bh(&ar->data_lock);
-	pdev = list_first_entry_or_null(&ar->fw_stats.pdevs,
-					struct ath11k_fw_stats_pdev, list);
-	if (!pdev) {
-		spin_unlock_bh(&ar->data_lock);
-		goto err_fallback;
-	}
-
-	/* tx power is set as 2 units per dBm in FW. */
-	*dbm = pdev->chan_tx_power / 2;
-
-	spin_unlock_bh(&ar->data_lock);
-	mutex_unlock(&ar->conf_mutex);
-
-	ath11k_dbg(ar->ab, ATH11K_DBG_MAC, "txpower from firmware %d, reported %d dBm\n",
-		   pdev->chan_tx_power, *dbm);
-	return 0;
-
-err_fallback:
-	mutex_unlock(&ar->conf_mutex);
-	/* We didn't get txpower from FW. Hence, relying on vif->bss_conf.txpower */
-	*dbm = vif->bss_conf.txpower;
-	ath11k_dbg(ar->ab, ATH11K_DBG_MAC, "txpower from firmware NaN, reported %d dBm\n",
-		   *dbm);
-	return 0;
-}
-
 static int ath11k_mac_station_add(struct ath11k *ar,
 				  struct ieee80211_vif *vif,
 				  struct ieee80211_sta *sta)

-- 
2.34.1

[PATCH ath-next 2/4] wifi: ath11k: wrap ath11k_mac_op_get_txpower() with lock-aware internal helper

[Aditya Kumar Singh]

Refactor ath11k_mac_op_get_txpower() by introducing a new internal function
ath11k_mac_handle_get_txpower(), which assumes the caller holds the
appropriate lock. This prepares the codebase for future change where the
internal function may be invoked directly with the lock already acquired,
improving modularity and lock handling consistency.

No functional change intended.

Tested-on: QCN9074 hw1.0 PCI WLAN.HK.2.9.0.1-02146-QCAHKSWPL_SILICONZ-1

Signed-off-by: Aditya Kumar Singh <aditya.kumar.singh@oss.qualcomm.com>
---
 drivers/net/wireless/ath/ath11k/mac.c | 31 ++++++++++++++++++++-----------
 1 file changed, 20 insertions(+), 11 deletions(-)

diff --git a/drivers/net/wireless/ath/ath11k/mac.c b/drivers/net/wireless/ath/ath11k/mac.c
index a7bd0ddf5fdf..cd6e98ee1232 100644
--- a/drivers/net/wireless/ath/ath11k/mac.c
+++ b/drivers/net/wireless/ath/ath11k/mac.c
@@ -4103,12 +4103,10 @@ static int ath11k_mac_get_fw_stats(struct ath11k *ar, u32 pdev_id,
 	return ret;
 }
 
-static int ath11k_mac_op_get_txpower(struct ieee80211_hw *hw,
-				     struct ieee80211_vif *vif,
-				     unsigned int link_id,
-				     int *dbm)
+static int ath11k_mac_handle_get_txpower(struct ath11k *ar,
+					 struct ieee80211_vif *vif,
+					 int *dbm)
 {
-	struct ath11k *ar = hw->priv;
 	struct ath11k_base *ab = ar->ab;
 	struct ath11k_fw_stats_pdev *pdev;
 	int ret;
@@ -4119,15 +4117,13 @@ static int ath11k_mac_op_get_txpower(struct ieee80211_hw *hw,
 	 * of these. Hence, we request the FW pdev stats in which FW reports
 	 * the minimum of all vdev's channel Tx power.
 	 */
-	mutex_lock(&ar->conf_mutex);
+	lockdep_assert_held(&ar->conf_mutex);
 
 	/* Firmware doesn't provide Tx power during CAC hence no need to fetch
 	 * the stats.
 	 */
-	if (test_bit(ATH11K_CAC_RUNNING, &ar->dev_flags)) {
-		mutex_unlock(&ar->conf_mutex);
+	if (test_bit(ATH11K_CAC_RUNNING, &ar->dev_flags))
 		return -EAGAIN;
-	}
 
 	ret = ath11k_mac_get_fw_stats(ar, ar->pdev->pdev_id, 0,
 				      WMI_REQUEST_PDEV_STAT);
@@ -4148,14 +4144,12 @@ static int ath11k_mac_op_get_txpower(struct ieee80211_hw *hw,
 	*dbm = pdev->chan_tx_power / 2;
 
 	spin_unlock_bh(&ar->data_lock);
-	mutex_unlock(&ar->conf_mutex);
 
 	ath11k_dbg(ar->ab, ATH11K_DBG_MAC, "txpower from firmware %d, reported %d dBm\n",
 		   pdev->chan_tx_power, *dbm);
 	return 0;
 
 err_fallback:
-	mutex_unlock(&ar->conf_mutex);
 	/* We didn't get txpower from FW. Hence, relying on vif->bss_conf.txpower */
 	*dbm = vif->bss_conf.txpower;
 	ath11k_dbg(ar->ab, ATH11K_DBG_MAC, "txpower from firmware NaN, reported %d dBm\n",
@@ -4163,6 +4157,21 @@ static int ath11k_mac_op_get_txpower(struct ieee80211_hw *hw,
 	return 0;
 }
 
+static int ath11k_mac_op_get_txpower(struct ieee80211_hw *hw,
+				     struct ieee80211_vif *vif,
+				     unsigned int link_id,
+				     int *dbm)
+{
+	struct ath11k *ar = hw->priv;
+	int ret;
+
+	mutex_lock(&ar->conf_mutex);
+	ret = ath11k_mac_handle_get_txpower(ar, vif, dbm);
+	mutex_unlock(&ar->conf_mutex);
+
+	return ret;
+}
+
 static int ath11k_mac_op_hw_scan(struct ieee80211_hw *hw,
 				 struct ieee80211_vif *vif,
 				 struct ieee80211_scan_request *hw_req)

-- 
2.34.1

[PATCH ath-next 3/4] wifi: ath11k: add support for Tx Power insertion in RRM action frame

[Aditya Kumar Singh]

For certain action frames like the TPC Report IE in the spectrum management
TPC Report action frame, and in the Radio Measurement Link Measurement
Report action frame there is a requirement to fill in the current
and max Tx power of the device in the packet.

Add support to populate these fields in the relevant packets.

In software-encrypted cases such as PMF, skip insertion since the packets
are already encrypted and cannot be modified.

Tested-on: QCN9074 hw1.0 PCI WLAN.HK.2.9.0.1-02146-QCAHKSWPL_SILICONZ-1

Signed-off-by: Aditya Kumar Singh <aditya.kumar.singh@oss.qualcomm.com>
---
 drivers/net/wireless/ath/ath11k/mac.c | 166 ++++++++++++++++++++++++++++++++++
 1 file changed, 166 insertions(+)

diff --git a/drivers/net/wireless/ath/ath11k/mac.c b/drivers/net/wireless/ath/ath11k/mac.c
index cd6e98ee1232..773ea1d4b358 100644
--- a/drivers/net/wireless/ath/ath11k/mac.c
+++ b/drivers/net/wireless/ath/ath11k/mac.c
@@ -6251,6 +6251,159 @@ static void ath11k_mgmt_over_wmi_tx_purge(struct ath11k *ar)
 		ath11k_mgmt_over_wmi_tx_drop(ar, skb);
 }
 
+static int ath11k_mac_mgmt_action_frame_fill_elem_data(struct ath11k_vif *arvif,
+						       struct sk_buff *skb)
+{
+	struct ieee80211_hdr *hdr = (struct ieee80211_hdr *)skb->data;
+	u8 category, *buf, iv_len, action_code, dialog_token;
+	int cur_tx_power, max_tx_power;
+	struct ath11k *ar = arvif->ar;
+	struct cfg80211_chan_def def;
+	struct ath11k_skb_cb *skb_cb;
+	struct ieee80211_mgmt *mgmt;
+	unsigned int remaining_len;
+	bool has_protected;
+
+	lockdep_assert_held(&ar->conf_mutex);
+
+	/* make sure category field is present */
+	if (skb->len < IEEE80211_MIN_ACTION_SIZE)
+		return -EINVAL;
+
+	remaining_len = skb->len - IEEE80211_MIN_ACTION_SIZE;
+	has_protected = ieee80211_has_protected(hdr->frame_control);
+
+	/* In case of SW crypto and hdr protected (PMF), packet will already be encrypted,
+	 * we can't put in data in this case
+	 */
+	if (test_bit(ATH11K_FLAG_HW_CRYPTO_DISABLED, &ar->ab->dev_flags) &&
+	    has_protected)
+		return 0;
+
+	mgmt = (struct ieee80211_mgmt *)hdr;
+	buf = (u8 *)&mgmt->u.action;
+
+	/* FCTL_PROTECTED frame might have extra space added for HDR_LEN. Offset that
+	 * many bytes if it is there
+	 */
+	if (has_protected) {
+		skb_cb = ATH11K_SKB_CB(skb);
+
+		switch (skb_cb->cipher) {
+		/* Cipher suite having flag %IEEE80211_KEY_FLAG_GENERATE_IV_MGMT set in
+		 * key needs to be processed. See ath11k_install_key()
+		 */
+		case WLAN_CIPHER_SUITE_CCMP:
+		case WLAN_CIPHER_SUITE_CCMP_256:
+		case WLAN_CIPHER_SUITE_GCMP:
+		case WLAN_CIPHER_SUITE_GCMP_256:
+			iv_len = IEEE80211_CCMP_HDR_LEN;
+			break;
+		case WLAN_CIPHER_SUITE_TKIP:
+			iv_len = 0;
+			break;
+		default:
+			return -EINVAL;
+		}
+
+		if (remaining_len < iv_len)
+			return -EINVAL;
+
+		buf += iv_len;
+		remaining_len -= iv_len;
+	}
+
+	category = *buf++;
+	/* category code is already taken care in %IEEE80211_MIN_ACTION_SIZE hence
+	 * no need to adjust remaining_len
+	 */
+
+	switch (category) {
+	case WLAN_CATEGORY_RADIO_MEASUREMENT:
+		/* need action code and dialog token */
+		if (remaining_len < 2)
+			return -EINVAL;
+
+		/* Packet Format:
+		 *	Action Code | Dialog Token | Variable Len (based on Action Code)
+		 */
+		action_code = *buf++;
+		dialog_token = *buf++;
+		remaining_len -= 2;
+
+		if (ath11k_mac_vif_chan(arvif->vif, &def))
+			return -ENOENT;
+
+		cur_tx_power = arvif->vif->bss_conf.txpower;
+		max_tx_power = min(def.chan->max_reg_power, (int)ar->max_tx_power / 2);
+		ath11k_mac_handle_get_txpower(ar, arvif->vif, &cur_tx_power);
+
+		switch (action_code) {
+		case WLAN_RM_ACTION_LINK_MEASUREMENT_REQUEST:
+			/* need variable fields to be present in len */
+			if (remaining_len < 2)
+				return -EINVAL;
+
+			/* Variable length format as defined in IEEE 802.11-2024,
+			 * Figure 9-1187-Link Measurement Request frame Action field
+			 * format.
+			 *	Transmit Power | Max Tx Power
+			 * We fill both of these.
+			 */
+			*buf++ = cur_tx_power;
+			*buf = max_tx_power;
+
+			ath11k_dbg(ar->ab, ATH11K_DBG_MAC,
+				   "RRM: Link Measurement Req dialog_token %u cur_tx_power %d max_tx_power %d\n",
+				   dialog_token, cur_tx_power, max_tx_power);
+			break;
+		case WLAN_RM_ACTION_LINK_MEASUREMENT_REPORT:
+			/* need variable fields to be present in len */
+			if (remaining_len < 3)
+				return -EINVAL;
+
+			/* Variable length format as defined in IEEE 802.11-2024,
+			 * Figure 9-1188-Link Measurement Report frame Action field format
+			 *	TPC Report | Variable Fields
+			 *
+			 * TPC Report Format:
+			 *	Element ID | Len | Tx Power | Link Margin
+			 *
+			 * We fill Tx power in the TPC Report (2nd index)
+			 */
+			buf[2] = cur_tx_power;
+
+			/* TODO: At present, Link margin data is not present so can't
+			 * really fill it now. Once it is available, it can be added
+			 * here
+			 */
+			ath11k_dbg(ar->ab, ATH11K_DBG_MAC,
+				   "RRM: Link Measurement Report dialog_token %u cur_tx_power %d\n",
+				   dialog_token, cur_tx_power);
+			break;
+		default:
+			return -EINVAL;
+		}
+		break;
+	default:
+		/* nothing to fill */
+		return 0;
+	}
+
+	return 0;
+}
+
+static int ath11k_mac_mgmt_frame_fill_elem_data(struct ath11k_vif *arvif,
+						struct sk_buff *skb)
+{
+	struct ieee80211_hdr *hdr = (struct ieee80211_hdr *)skb->data;
+
+	if (!ieee80211_is_action(hdr->frame_control))
+		return 0;
+
+	return ath11k_mac_mgmt_action_frame_fill_elem_data(arvif, skb);
+}
+
 static void ath11k_mgmt_over_wmi_tx_work(struct work_struct *work)
 {
 	struct ath11k *ar = container_of(work, struct ath11k, wmi_mgmt_tx_work);
@@ -6270,6 +6423,19 @@ static void ath11k_mgmt_over_wmi_tx_work(struct work_struct *work)
 		arvif = ath11k_vif_to_arvif(skb_cb->vif);
 		mutex_lock(&ar->conf_mutex);
 		if (ar->allocated_vdev_map & (1LL << arvif->vdev_id)) {
+			/* Fill in the data which is required to be filled by the driver
+			 * For example: Max Tx power in Link Measurement Request/Report
+			 */
+			ret = ath11k_mac_mgmt_frame_fill_elem_data(arvif, skb);
+			if (ret) {
+				/* If we couldn't fill the data due to any reason,
+				 * let's not discard transmitting the packet.
+				 */
+				ath11k_dbg(ar->ab, ATH11K_DBG_MAC,
+					   "Failed to fill the required data for the mgmt packet err %d\n",
+					   ret);
+			}
+
 			ret = ath11k_mac_mgmt_tx_wmi(ar, arvif, skb);
 			if (ret) {
 				ath11k_warn(ar->ab, "failed to tx mgmt frame, vdev_id %d :%d\n",

-- 
2.34.1

[PATCH ath-next 4/4] wifi: ath11k: advertise NL80211_FEATURE_TX_POWER_INSERTION

[Aditya Kumar Singh]

Now that driver is capable of inserting Tx power, advertise the support
for the same to upper layers.

Tested-on: QCN9074 hw1.0 PCI WLAN.HK.2.9.0.1-02146-QCAHKSWPL_SILICONZ-1

Signed-off-by: Aditya Kumar Singh <aditya.kumar.singh@oss.qualcomm.com>
---
 drivers/net/wireless/ath/ath11k/mac.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/drivers/net/wireless/ath/ath11k/mac.c b/drivers/net/wireless/ath/ath11k/mac.c
index 773ea1d4b358..23b02e505099 100644
--- a/drivers/net/wireless/ath/ath11k/mac.c
+++ b/drivers/net/wireless/ath/ath11k/mac.c
@@ -10543,6 +10543,8 @@ static int __ath11k_mac_register(struct ath11k *ar)
 	ar->hw->wiphy->features |= NL80211_FEATURE_AP_MODE_CHAN_WIDTH_CHANGE |
 				   NL80211_FEATURE_AP_SCAN;
 
+	ar->hw->wiphy->features |= NL80211_FEATURE_TX_POWER_INSERTION;
+
 	ar->max_num_stations = TARGET_NUM_STATIONS(ab);
 	ar->max_num_peers = TARGET_NUM_PEERS_PDEV(ab);
 

-- 
2.34.1

Re: [PATCH ath-next 0/4] wifi: ath11k: add support for Tx Power insertion

[Baochen Qiang]

On 10/17/2025 12:07 PM, Aditya Kumar Singh wrote:
> For certain action frames like the TPC Report IE in the spectrum management
> TPC Report action frame, and in the Radio Measurement Link Measurement
> Report action frame there is a requirement to fill in the current
> and max Tx power of the device in the packet.
> 
> Add support to populate these fields in the relevant packets. Advertise
> this capability from the driver using the feature flag
> NL80211_FEATURE_TX_POWER_INSERTION.
> 
> In software-encrypted cases such as PMF, skip insertion since the packets
> are already encrypted and cannot be modified.
> 
> Patch Overview:
> Patches 1 and 2 serve as preparatory groundwork.
> Patches 3 and 4 implement the core changes described.
> 
> For context, this ath11k implementation is adapted from prior ath12k work [1].
> 
> Reference:
> [1] https://lore.kernel.org/linux-wireless/20250630-support-for-tx-power-insertion-v1-0-77f45484d5bb@oss.qualcomm.com/
> 
> ---
> Aditya Kumar Singh (4):
>       wifi: ath11k: relocate some Tx power related functions in mac.c
>       wifi: ath11k: wrap ath11k_mac_op_get_txpower() with lock-aware internal helper
>       wifi: ath11k: add support for Tx Power insertion in RRM action frame
>       wifi: ath11k: advertise NL80211_FEATURE_TX_POWER_INSERTION
> 
>  drivers/net/wireless/ath/ath11k/mac.c | 447 ++++++++++++++++++++++++----------
>  1 file changed, 312 insertions(+), 135 deletions(-)
> ---
> base-commit: d5ce93f136fbee2b7afbe221f34ca881036f8de3
> change-id: 20251008-add_tx_power_insertion_support-e5225e924bfd
> 
> 

Reviewed-by: Baochen Qiang <baochen.qiang@oss.qualcomm.com>

Re: [PATCH ath-next 0/4] wifi: ath11k: add support for Tx Power insertion

[Vasanthakumar Thiagarajan]

On 10/17/2025 9:37 AM, Aditya Kumar Singh wrote:
> For certain action frames like the TPC Report IE in the spectrum management
> TPC Report action frame, and in the Radio Measurement Link Measurement
> Report action frame there is a requirement to fill in the current
> and max Tx power of the device in the packet.
> 
> Add support to populate these fields in the relevant packets. Advertise
> this capability from the driver using the feature flag
> NL80211_FEATURE_TX_POWER_INSERTION.
> 
> In software-encrypted cases such as PMF, skip insertion since the packets
> are already encrypted and cannot be modified.
> 
> Patch Overview:
> Patches 1 and 2 serve as preparatory groundwork.
> Patches 3 and 4 implement the core changes described.
> 
> For context, this ath11k implementation is adapted from prior ath12k work [1].
> 
> Reference:
> [1] https://lore.kernel.org/linux-wireless/20250630-support-for-tx-power-insertion-v1-0-77f45484d5bb@oss.qualcomm.com/
> 
> ---
> Aditya Kumar Singh (4):
>        wifi: ath11k: relocate some Tx power related functions in mac.c
>        wifi: ath11k: wrap ath11k_mac_op_get_txpower() with lock-aware internal helper
>        wifi: ath11k: add support for Tx Power insertion in RRM action frame
>        wifi: ath11k: advertise NL80211_FEATURE_TX_POWER_INSERTION
> 
>   drivers/net/wireless/ath/ath11k/mac.c | 447 ++++++++++++++++++++++++----------
>   1 file changed, 312 insertions(+), 135 deletions(-)

Reviewed-by: Vasanthakumar Thiagarajan <vasanthakumar.thiagarajan@oss.qualcomm.com>

Re: [PATCH ath-next 0/4] wifi: ath11k: add support for Tx Power insertion

[Jeff Johnson]

On Fri, 17 Oct 2025 09:37:56 +0530, Aditya Kumar Singh wrote:
> For certain action frames like the TPC Report IE in the spectrum management
> TPC Report action frame, and in the Radio Measurement Link Measurement
> Report action frame there is a requirement to fill in the current
> and max Tx power of the device in the packet.
> 
> Add support to populate these fields in the relevant packets. Advertise
> this capability from the driver using the feature flag
> NL80211_FEATURE_TX_POWER_INSERTION.
> 
> [...]

Applied, thanks!

[1/4] wifi: ath11k: relocate some Tx power related functions in mac.c
      commit: 66887282233d281cd9109dabfdad5d86b709acc0
[2/4] wifi: ath11k: wrap ath11k_mac_op_get_txpower() with lock-aware internal helper
      commit: c243d5e44f6ecbb29bf55b82e6dd92bca4fde0b1
[3/4] wifi: ath11k: add support for Tx Power insertion in RRM action frame
      commit: 722015690f52d046ae609e1b90cd3f018644d93d
[4/4] wifi: ath11k: advertise NL80211_FEATURE_TX_POWER_INSERTION
      commit: 998c68e96c03f10dec19b65279ade9d4000d1ae9

Best regards,
-- 
Jeff Johnson <jeff.johnson@oss.qualcomm.com>