public inbox for linux-wireless@vger.kernel.org
 help / color / mirror / Atom feed
* [PATCH v3 00/11] wifi: ath12k: some fixes and clean up for monitor mode
@ 2024-10-16  6:56 Kang Yang
  2024-10-16  6:56 ` [PATCH v3 01/11] wifi: ath12k: remove unused variable monitor_present Kang Yang
                   ` (10 more replies)
  0 siblings, 11 replies; 21+ messages in thread
From: Kang Yang @ 2024-10-16  6:56 UTC (permalink / raw)
  To: ath12k; +Cc: linux-wireless, quic_kangyang

This patch set does some fixes and clean up for monitor mode.

v3: rebase on tag: ath/main(ath-202410111606).
v2: rebase on tag: ath-202410072115.

Kang Yang (11):
  wifi: ath12k: remove unused variable monitor_present
  wifi: ath12k: optimize storage size for struct ath12k
  wifi: ath12k: fix struct hal_rx_ppdu_end_user_stats
  wifi: ath12k: fix struct hal_rx_ppdu_start
  wifi: ath12k: fix struct hal_rx_phyrx_rssi_legacy_info
  wifi: ath12k: fix struct hal_rx_mpdu_start
  wifi: ath12k: properly handling the state variables of monitor mode
  wifi: ath12k: delete NSS and TX power setting for monitor vdev
  wifi: ath12k: use tail MSDU to get MSDU information
  wifi: ath12k: fix A-MSDU indication in monitor mode
  wifi: ath12k: delete mon reap timer

 drivers/net/wireless/ath/ath12k/core.c   |   5 ++
 drivers/net/wireless/ath/ath12k/core.h   |  23 +++--
 drivers/net/wireless/ath/ath12k/dp.c     |  25 ------
 drivers/net/wireless/ath/ath12k/dp_mon.c | 107 +++++++++++++----------
 drivers/net/wireless/ath/ath12k/hal_rx.h |  53 ++++++-----
 drivers/net/wireless/ath/ath12k/mac.c    |  24 +++--
 6 files changed, 115 insertions(+), 122 deletions(-)


base-commit: 69eabe24843f238e79a6dbbd2b3fcc8eef39d6b8
-- 
2.34.1


^ permalink raw reply	[flat|nested] 21+ messages in thread

* [PATCH v3 01/11] wifi: ath12k: remove unused variable monitor_present
  2024-10-16  6:56 [PATCH v3 00/11] wifi: ath12k: some fixes and clean up for monitor mode Kang Yang
@ 2024-10-16  6:56 ` Kang Yang
  2024-10-16  6:56 ` [PATCH v3 02/11] wifi: ath12k: optimize storage size for struct ath12k Kang Yang
                   ` (9 subsequent siblings)
  10 siblings, 0 replies; 21+ messages in thread
From: Kang Yang @ 2024-10-16  6:56 UTC (permalink / raw)
  To: ath12k; +Cc: linux-wireless, quic_kangyang

Variable monitor_present is never used. So delete it.

Tested-on: WCN7850 hw2.0 PCI WLAN.HMT.1.0.c5-00481-QCAHMTSWPL_V1.0_V2.0_SILICONZ-3

Signed-off-by: Kang Yang <quic_kangyang@quicinc.com>
---
 drivers/net/wireless/ath/ath12k/core.h | 1 -
 1 file changed, 1 deletion(-)

diff --git a/drivers/net/wireless/ath/ath12k/core.h b/drivers/net/wireless/ath/ath12k/core.h
index ebfc1e370acc..cbcace4671cb 100644
--- a/drivers/net/wireless/ath/ath12k/core.h
+++ b/drivers/net/wireless/ath/ath12k/core.h
@@ -561,7 +561,6 @@ struct ath12k {
 	u32 chan_tx_pwr;
 	u32 num_stations;
 	u32 max_num_stations;
-	bool monitor_present;
 
 	/* protects the radio specific data like debug stats, ppdu_stats_info stats,
 	 * vdev_stop_status info, scan data, ath12k_sta info, ath12k_vif info,
-- 
2.34.1


^ permalink raw reply related	[flat|nested] 21+ messages in thread

* [PATCH v3 02/11] wifi: ath12k: optimize storage size for struct ath12k
  2024-10-16  6:56 [PATCH v3 00/11] wifi: ath12k: some fixes and clean up for monitor mode Kang Yang
  2024-10-16  6:56 ` [PATCH v3 01/11] wifi: ath12k: remove unused variable monitor_present Kang Yang
@ 2024-10-16  6:56 ` Kang Yang
  2024-10-16  6:56 ` [PATCH v3 03/11] wifi: ath12k: fix struct hal_rx_ppdu_end_user_stats Kang Yang
                   ` (8 subsequent siblings)
  10 siblings, 0 replies; 21+ messages in thread
From: Kang Yang @ 2024-10-16  6:56 UTC (permalink / raw)
  To: ath12k; +Cc: linux-wireless, quic_kangyang

Optimize storage size for struct ath12k by clean up bool variables.

Tested-on: WCN7850 hw2.0 PCI WLAN.HMT.1.0.c5-00481-QCAHMTSWPL_V1.0_V2.0_SILICONZ-3

Signed-off-by: Kang Yang <quic_kangyang@quicinc.com>
---
 drivers/net/wireless/ath/ath12k/core.h | 20 ++++++++++----------
 1 file changed, 10 insertions(+), 10 deletions(-)

diff --git a/drivers/net/wireless/ath/ath12k/core.h b/drivers/net/wireless/ath/ath12k/core.h
index cbcace4671cb..0c39b37d6a47 100644
--- a/drivers/net/wireless/ath/ath12k/core.h
+++ b/drivers/net/wireless/ath/ath12k/core.h
@@ -531,17 +531,16 @@ struct ath12k {
 	u32 ht_cap_info;
 	u32 vht_cap_info;
 	struct ath12k_he ar_he;
-	bool supports_6ghz;
 	struct {
 		struct completion started;
 		struct completion completed;
 		struct completion on_channel;
 		struct delayed_work timeout;
 		enum ath12k_scan_state state;
-		bool is_roc;
 		int vdev_id;
 		int roc_freq;
-		bool roc_notify;
+		bool is_roc:1;
+		bool roc_notify:1;
 	} scan;
 
 	struct {
@@ -618,7 +617,6 @@ struct ath12k {
 	 * However there are deviations from this rule. This flag is used to
 	 * avoid reporting garbage data.
 	 */
-	bool ch_info_can_report_survey;
 	struct survey_info survey[ATH12K_NUM_CHANS];
 	struct completion bss_survey_done;
 
@@ -629,7 +627,6 @@ struct ath12k {
 
 	struct ath12k_wow wow;
 	struct completion target_suspend;
-	bool target_suspend_ack;
 	struct ath12k_per_peer_tx_stats peer_tx_stats;
 	struct list_head ppdu_stats_info;
 	u32 ppdu_stat_list_depth;
@@ -641,16 +638,19 @@ struct ath12k {
 	struct ath12k_debug debug;
 #endif
 
-	bool dfs_block_radar_events;
-	bool monitor_conf_enabled;
-	bool monitor_vdev_created;
-	bool monitor_started;
 	int monitor_vdev_id;
 
 	u32 freq_low;
 	u32 freq_high;
 
-	bool nlo_enabled;
+	bool supports_6ghz:1;
+	bool ch_info_can_report_survey:1;
+	bool target_suspend_ack:1;
+	bool dfs_block_radar_events:1;
+	bool monitor_conf_enabled:1;
+	bool monitor_vdev_created:1;
+	bool monitor_started:1;
+	bool nlo_enabled:1;
 };
 
 struct ath12k_hw {
-- 
2.34.1


^ permalink raw reply related	[flat|nested] 21+ messages in thread

* [PATCH v3 03/11] wifi: ath12k: fix struct hal_rx_ppdu_end_user_stats
  2024-10-16  6:56 [PATCH v3 00/11] wifi: ath12k: some fixes and clean up for monitor mode Kang Yang
  2024-10-16  6:56 ` [PATCH v3 01/11] wifi: ath12k: remove unused variable monitor_present Kang Yang
  2024-10-16  6:56 ` [PATCH v3 02/11] wifi: ath12k: optimize storage size for struct ath12k Kang Yang
@ 2024-10-16  6:56 ` Kang Yang
  2024-10-21  8:45   ` Kalle Valo
  2024-10-16  6:56 ` [PATCH v3 04/11] wifi: ath12k: fix struct hal_rx_ppdu_start Kang Yang
                   ` (7 subsequent siblings)
  10 siblings, 1 reply; 21+ messages in thread
From: Kang Yang @ 2024-10-16  6:56 UTC (permalink / raw)
  To: ath12k; +Cc: linux-wireless, quic_kangyang

Current struct hal_rx_ppdu_end_user_stats in hal_rx.h is not matched
with hardware descriptor definition.

So update this structure and related code.

Tested-on: WCN7850 hw2.0 PCI WLAN.HMT.1.0.c5-00481-QCAHMTSWPL_V1.0_V2.0_SILICONZ-3

Signed-off-by: Kang Yang <quic_kangyang@quicinc.com>
---
 drivers/net/wireless/ath/ath12k/dp_mon.c | 10 +++++-----
 drivers/net/wireless/ath/ath12k/hal_rx.h | 22 +++++++++++-----------
 2 files changed, 16 insertions(+), 16 deletions(-)

diff --git a/drivers/net/wireless/ath/ath12k/dp_mon.c b/drivers/net/wireless/ath/ath12k/dp_mon.c
index 5c6749bc4039..8d0d3c849c90 100644
--- a/drivers/net/wireless/ath/ath12k/dp_mon.c
+++ b/drivers/net/wireless/ath/ath12k/dp_mon.c
@@ -21,20 +21,20 @@ static void ath12k_dp_mon_rx_handle_ofdma_info(void *rx_tlv,
 		__le32_to_cpu(ppdu_end_user->usr_resp_ref_ext);
 }
 
-static void
+static inline void
 ath12k_dp_mon_rx_populate_byte_count(const struct hal_rx_ppdu_end_user_stats *stats,
 				     void *ppduinfo,
 				     struct hal_rx_user_status *rx_user_status)
 {
-	u32 mpdu_ok_byte_count = __le32_to_cpu(stats->mpdu_ok_cnt);
-	u32 mpdu_err_byte_count = __le32_to_cpu(stats->mpdu_err_cnt);
+	u32 mpdu_ok_byte_count = __le32_to_cpu(stats->info7);
+	u32 mpdu_err_byte_count = __le32_to_cpu(stats->info8);
 
 	rx_user_status->mpdu_ok_byte_count =
 		u32_get_bits(mpdu_ok_byte_count,
-			     HAL_RX_PPDU_END_USER_STATS_MPDU_DELIM_OK_BYTE_COUNT);
+			     HAL_RX_PPDU_END_USER_STATS_INFO7_MPDU_OK_BYTE_COUNT);
 	rx_user_status->mpdu_err_byte_count =
 		u32_get_bits(mpdu_err_byte_count,
-			     HAL_RX_PPDU_END_USER_STATS_MPDU_DELIM_ERR_BYTE_COUNT);
+			     HAL_RX_PPDU_END_USER_STATS_INFO8_MPDU_ERR_BYTE_COUNT);
 }
 
 static void
diff --git a/drivers/net/wireless/ath/ath12k/hal_rx.h b/drivers/net/wireless/ath/ath12k/hal_rx.h
index 095216eabc01..837ba4adba88 100644
--- a/drivers/net/wireless/ath/ath12k/hal_rx.h
+++ b/drivers/net/wireless/ath/ath12k/hal_rx.h
@@ -1,7 +1,7 @@
 /* SPDX-License-Identifier: BSD-3-Clause-Clear */
 /*
  * Copyright (c) 2018-2021 The Linux Foundation. All rights reserved.
- * Copyright (c) 2021-2023 Qualcomm Innovation Center, Inc. All rights reserved.
+ * Copyright (c) 2021-2024 Qualcomm Innovation Center, Inc. All rights reserved.
  */
 
 #ifndef ATH12K_HAL_RX_H
@@ -240,13 +240,13 @@ struct hal_rx_ppdu_start {
 	__le32 ppdu_start_ts;
 } __packed;
 
-#define HAL_RX_PPDU_END_USER_STATS_INFO0_MPDU_CNT_FCS_ERR	GENMASK(25, 16)
+#define HAL_RX_PPDU_END_USER_STATS_INFO0_MPDU_CNT_FCS_ERR	GENMASK(26, 16)
 
-#define HAL_RX_PPDU_END_USER_STATS_INFO1_MPDU_CNT_FCS_OK	GENMASK(8, 0)
-#define HAL_RX_PPDU_END_USER_STATS_INFO1_FC_VALID		BIT(9)
-#define HAL_RX_PPDU_END_USER_STATS_INFO1_QOS_CTRL_VALID		BIT(10)
-#define HAL_RX_PPDU_END_USER_STATS_INFO1_HT_CTRL_VALID		BIT(11)
-#define HAL_RX_PPDU_END_USER_STATS_INFO1_PKT_TYPE		GENMASK(23, 20)
+#define HAL_RX_PPDU_END_USER_STATS_INFO1_MPDU_CNT_FCS_OK	GENMASK(10, 0)
+#define HAL_RX_PPDU_END_USER_STATS_INFO1_FC_VALID		BIT(11)
+#define HAL_RX_PPDU_END_USER_STATS_INFO1_QOS_CTRL_VALID		BIT(12)
+#define HAL_RX_PPDU_END_USER_STATS_INFO1_HT_CTRL_VALID		BIT(13)
+#define HAL_RX_PPDU_END_USER_STATS_INFO1_PKT_TYPE              GENMASK(24, 21)
 
 #define HAL_RX_PPDU_END_USER_STATS_INFO2_AST_INDEX		GENMASK(15, 0)
 #define HAL_RX_PPDU_END_USER_STATS_INFO2_FRAME_CTRL		GENMASK(31, 16)
@@ -262,8 +262,8 @@ struct hal_rx_ppdu_start {
 #define HAL_RX_PPDU_END_USER_STATS_INFO6_TID_BITMAP		GENMASK(15, 0)
 #define HAL_RX_PPDU_END_USER_STATS_INFO6_TID_EOSP_BITMAP	GENMASK(31, 16)
 
-#define HAL_RX_PPDU_END_USER_STATS_MPDU_DELIM_OK_BYTE_COUNT	GENMASK(24, 0)
-#define HAL_RX_PPDU_END_USER_STATS_MPDU_DELIM_ERR_BYTE_COUNT	GENMASK(24, 0)
+#define HAL_RX_PPDU_END_USER_STATS_INFO7_MPDU_OK_BYTE_COUNT    GENMASK(24, 0)
+#define HAL_RX_PPDU_END_USER_STATS_INFO8_MPDU_ERR_BYTE_COUNT   GENMASK(24, 0)
 
 struct hal_rx_ppdu_end_user_stats {
 	__le32 rsvd0[2];
@@ -278,9 +278,9 @@ struct hal_rx_ppdu_end_user_stats {
 	__le32 usr_resp_ref;
 	__le32 info6;
 	__le32 rsvd3[4];
-	__le32 mpdu_ok_cnt;
+	__le32 info7;
 	__le32 rsvd4;
-	__le32 mpdu_err_cnt;
+	__le32 info8;
 	__le32 rsvd5[2];
 	__le32 usr_resp_ref_ext;
 	__le32 rsvd6;
-- 
2.34.1


^ permalink raw reply related	[flat|nested] 21+ messages in thread

* [PATCH v3 04/11] wifi: ath12k: fix struct hal_rx_ppdu_start
  2024-10-16  6:56 [PATCH v3 00/11] wifi: ath12k: some fixes and clean up for monitor mode Kang Yang
                   ` (2 preceding siblings ...)
  2024-10-16  6:56 ` [PATCH v3 03/11] wifi: ath12k: fix struct hal_rx_ppdu_end_user_stats Kang Yang
@ 2024-10-16  6:56 ` Kang Yang
  2024-10-16  6:56 ` [PATCH v3 05/11] wifi: ath12k: fix struct hal_rx_phyrx_rssi_legacy_info Kang Yang
                   ` (6 subsequent siblings)
  10 siblings, 0 replies; 21+ messages in thread
From: Kang Yang @ 2024-10-16  6:56 UTC (permalink / raw)
  To: ath12k; +Cc: linux-wireless, quic_kangyang

Current struct hal_rx_ppdu_start in hal_rx.h is not matched with
hardware descriptor definition.

So update this structure and related code.

Tested-on: WCN7850 hw2.0 PCI WLAN.HMT.1.0.c5-00481-QCAHMTSWPL_V1.0_V2.0_SILICONZ-3

Signed-off-by: Kang Yang <quic_kangyang@quicinc.com>
---
 drivers/net/wireless/ath/ath12k/dp_mon.c | 16 ++++++++++++----
 drivers/net/wireless/ath/ath12k/hal_rx.h | 11 ++++++++---
 2 files changed, 20 insertions(+), 7 deletions(-)

diff --git a/drivers/net/wireless/ath/ath12k/dp_mon.c b/drivers/net/wireless/ath/ath12k/dp_mon.c
index 8d0d3c849c90..d40da3a059c6 100644
--- a/drivers/net/wireless/ath/ath12k/dp_mon.c
+++ b/drivers/net/wireless/ath/ath12k/dp_mon.c
@@ -593,12 +593,20 @@ ath12k_dp_mon_rx_parse_status_tlv(struct ath12k_base *ab,
 		struct hal_rx_ppdu_start *ppdu_start =
 			(struct hal_rx_ppdu_start *)tlv_data;
 
+		u64 ppdu_ts = __le32_to_cpu(ppdu_start->ppdu_start_ts_31_0) |
+			(u64)(__le32_to_cpu(ppdu_start->ppdu_start_ts_63_32)) << 32;
+
 		info[0] = __le32_to_cpu(ppdu_start->info0);
 
-		ppdu_info->ppdu_id =
-			u32_get_bits(info[0], HAL_RX_PPDU_START_INFO0_PPDU_ID);
-		ppdu_info->chan_num = __le32_to_cpu(ppdu_start->chan_num);
-		ppdu_info->ppdu_ts = __le32_to_cpu(ppdu_start->ppdu_start_ts);
+		ppdu_info->ppdu_id = u32_get_bits(info[0],
+						  HAL_RX_PPDU_START_INFO0_PPDU_ID);
+
+		info[1] = __le32_to_cpu(ppdu_start->info1);
+		ppdu_info->chan_num = u32_get_bits(info[1],
+						   HAL_RX_PPDU_START_INFO1_CHAN_NUM);
+		ppdu_info->freq = u32_get_bits(info[1],
+					       HAL_RX_PPDU_START_INFO1_CHAN_FREQ);
+		ppdu_info->ppdu_ts = ppdu_ts;
 
 		if (ppdu_info->ppdu_id != ppdu_info->last_ppdu_id) {
 			ppdu_info->last_ppdu_id = ppdu_info->ppdu_id;
diff --git a/drivers/net/wireless/ath/ath12k/hal_rx.h b/drivers/net/wireless/ath/ath12k/hal_rx.h
index 837ba4adba88..6ab33d5f1b2a 100644
--- a/drivers/net/wireless/ath/ath12k/hal_rx.h
+++ b/drivers/net/wireless/ath/ath12k/hal_rx.h
@@ -156,6 +156,7 @@ struct hal_rx_mon_ppdu_info {
 	u32 preamble_type;
 	u32 mpdu_len;
 	u16 chan_num;
+	u16 freq;
 	u16 tcp_msdu_count;
 	u16 tcp_ack_msdu_count;
 	u16 udp_msdu_count;
@@ -232,12 +233,16 @@ struct hal_rx_mon_ppdu_info {
 	u8 medium_prot_type;
 };
 
-#define HAL_RX_PPDU_START_INFO0_PPDU_ID		GENMASK(15, 0)
+#define HAL_RX_PPDU_START_INFO0_PPDU_ID			GENMASK(15, 0)
+#define HAL_RX_PPDU_START_INFO1_CHAN_NUM		GENMASK(15, 0)
+#define HAL_RX_PPDU_START_INFO1_CHAN_FREQ		GENMASK(31, 16)
 
 struct hal_rx_ppdu_start {
 	__le32 info0;
-	__le32 chan_num;
-	__le32 ppdu_start_ts;
+	__le32 info1;
+	__le32 ppdu_start_ts_31_0;
+	__le32 ppdu_start_ts_63_32;
+	__le32 rsvd[2];
 } __packed;
 
 #define HAL_RX_PPDU_END_USER_STATS_INFO0_MPDU_CNT_FCS_ERR	GENMASK(26, 16)
-- 
2.34.1


^ permalink raw reply related	[flat|nested] 21+ messages in thread

* [PATCH v3 05/11] wifi: ath12k: fix struct hal_rx_phyrx_rssi_legacy_info
  2024-10-16  6:56 [PATCH v3 00/11] wifi: ath12k: some fixes and clean up for monitor mode Kang Yang
                   ` (3 preceding siblings ...)
  2024-10-16  6:56 ` [PATCH v3 04/11] wifi: ath12k: fix struct hal_rx_ppdu_start Kang Yang
@ 2024-10-16  6:56 ` Kang Yang
  2024-10-16  6:56 ` [PATCH v3 06/11] wifi: ath12k: fix struct hal_rx_mpdu_start Kang Yang
                   ` (5 subsequent siblings)
  10 siblings, 0 replies; 21+ messages in thread
From: Kang Yang @ 2024-10-16  6:56 UTC (permalink / raw)
  To: ath12k; +Cc: linux-wireless, quic_kangyang

Current struct hal_rx_phyrx_rssi_legacy_info in hal_rx.h is not matched
with hardware descriptor definition.

So update this structure and related code.

Tested-on: WCN7850 hw2.0 PCI WLAN.HMT.1.0.c5-00481-QCAHMTSWPL_V1.0_V2.0_SILICONZ-3

Signed-off-by: Kang Yang <quic_kangyang@quicinc.com>
---
 drivers/net/wireless/ath/ath12k/dp_mon.c | 25 ++++++------------------
 drivers/net/wireless/ath/ath12k/hal_rx.h |  9 ++++++---
 2 files changed, 12 insertions(+), 22 deletions(-)

diff --git a/drivers/net/wireless/ath/ath12k/dp_mon.c b/drivers/net/wireless/ath/ath12k/dp_mon.c
index d40da3a059c6..486f95593aa6 100644
--- a/drivers/net/wireless/ath/ath12k/dp_mon.c
+++ b/drivers/net/wireless/ath/ath12k/dp_mon.c
@@ -734,33 +734,20 @@ ath12k_dp_mon_rx_parse_status_tlv(struct ath12k_base *ab,
 	case HAL_PHYRX_RSSI_LEGACY: {
 		struct hal_rx_phyrx_rssi_legacy_info *rssi =
 			(struct hal_rx_phyrx_rssi_legacy_info *)tlv_data;
-		u32 reception_type = 0;
-		u32 rssi_legacy_info = __le32_to_cpu(rssi->rsvd[0]);
 
 		info[0] = __le32_to_cpu(rssi->info0);
+		info[1] = __le32_to_cpu(rssi->info1);
 
 		/* TODO: Please note that the combined rssi will not be accurate
 		 * in MU case. Rssi in MU needs to be retrieved from
 		 * PHYRX_OTHER_RECEIVE_INFO TLV.
 		 */
 		ppdu_info->rssi_comb =
-			u32_get_bits(info[0],
-				     HAL_RX_PHYRX_RSSI_LEGACY_INFO_INFO0_RSSI_COMB);
-		reception_type =
-			u32_get_bits(rssi_legacy_info,
-				     HAL_RX_PHYRX_RSSI_LEGACY_INFO_RSVD1_RECEPTION);
-
-		switch (reception_type) {
-		case HAL_RECEPTION_TYPE_ULOFMDA:
-			ppdu_info->reception_type = HAL_RX_RECEPTION_TYPE_MU_OFDMA;
-			break;
-		case HAL_RECEPTION_TYPE_ULMIMO:
-			ppdu_info->reception_type = HAL_RX_RECEPTION_TYPE_MU_MIMO;
-			break;
-		default:
-			ppdu_info->reception_type = HAL_RX_RECEPTION_TYPE_SU;
-			break;
-		}
+			u32_get_bits(info[1],
+				     HAL_RX_PHYRX_RSSI_LEGACY_INFO_INFO1_RSSI_COMB);
+
+		ppdu_info->bw = u32_get_bits(info[0],
+					     HAL_RX_PHYRX_RSSI_LEGACY_INFO_INFO0_RX_BW);
 		break;
 	}
 	case HAL_RXPCU_PPDU_END_INFO: {
diff --git a/drivers/net/wireless/ath/ath12k/hal_rx.h b/drivers/net/wireless/ath/ath12k/hal_rx.h
index 6ab33d5f1b2a..c69df852ae69 100644
--- a/drivers/net/wireless/ath/ath12k/hal_rx.h
+++ b/drivers/net/wireless/ath/ath12k/hal_rx.h
@@ -441,12 +441,15 @@ enum hal_rx_ul_reception_type {
 	HAL_RECEPTION_TYPE_FRAMELESS
 };
 
-#define HAL_RX_PHYRX_RSSI_LEGACY_INFO_INFO0_RSSI_COMB	GENMASK(15, 8)
-#define HAL_RX_PHYRX_RSSI_LEGACY_INFO_RSVD1_RECEPTION   GENMASK(3, 0)
+#define HAL_RX_PHYRX_RSSI_LEGACY_INFO_INFO0_RECEPTION	GENMASK(3, 0)
+#define HAL_RX_PHYRX_RSSI_LEGACY_INFO_INFO0_RX_BW	GENMASK(7, 5)
+#define HAL_RX_PHYRX_RSSI_LEGACY_INFO_INFO1_RSSI_COMB	GENMASK(15, 8)
 
 struct hal_rx_phyrx_rssi_legacy_info {
-	__le32 rsvd[35];
 	__le32 info0;
+	__le32 rsvd0[39];
+	__le32 info1;
+	__le32 rsvd1;
 } __packed;
 
 #define HAL_RX_MPDU_START_INFO0_PPDU_ID	GENMASK(31, 16)
-- 
2.34.1


^ permalink raw reply related	[flat|nested] 21+ messages in thread

* [PATCH v3 06/11] wifi: ath12k: fix struct hal_rx_mpdu_start
  2024-10-16  6:56 [PATCH v3 00/11] wifi: ath12k: some fixes and clean up for monitor mode Kang Yang
                   ` (4 preceding siblings ...)
  2024-10-16  6:56 ` [PATCH v3 05/11] wifi: ath12k: fix struct hal_rx_phyrx_rssi_legacy_info Kang Yang
@ 2024-10-16  6:56 ` Kang Yang
  2024-10-16  6:56 ` [PATCH v3 07/11] wifi: ath12k: properly handling the state variables of monitor mode Kang Yang
                   ` (4 subsequent siblings)
  10 siblings, 0 replies; 21+ messages in thread
From: Kang Yang @ 2024-10-16  6:56 UTC (permalink / raw)
  To: ath12k; +Cc: linux-wireless, quic_kangyang

Current struct hal_rx_mpdu_start in hal_rx.h is not matched with
hardware descriptor definition.

So update this structure and related code.

Tested-on: WCN7850 hw2.0 PCI WLAN.HMT.1.0.c5-00481-QCAHMTSWPL_V1.0_V2.0_SILICONZ-3

Signed-off-by: Kang Yang <quic_kangyang@quicinc.com>
---
 drivers/net/wireless/ath/ath12k/hal_rx.h | 11 ++++++-----
 1 file changed, 6 insertions(+), 5 deletions(-)

diff --git a/drivers/net/wireless/ath/ath12k/hal_rx.h b/drivers/net/wireless/ath/ath12k/hal_rx.h
index c69df852ae69..2de7b0eba9f2 100644
--- a/drivers/net/wireless/ath/ath12k/hal_rx.h
+++ b/drivers/net/wireless/ath/ath12k/hal_rx.h
@@ -452,15 +452,16 @@ struct hal_rx_phyrx_rssi_legacy_info {
 	__le32 rsvd1;
 } __packed;
 
-#define HAL_RX_MPDU_START_INFO0_PPDU_ID	GENMASK(31, 16)
-#define HAL_RX_MPDU_START_INFO1_PEERID	GENMASK(31, 16)
-#define HAL_RX_MPDU_START_INFO2_MPDU_LEN GENMASK(13, 0)
+#define HAL_RX_MPDU_START_INFO0_PPDU_ID			GENMASK(31, 16)
+#define HAL_RX_MPDU_START_INFO1_PEERID			GENMASK(31, 16)
+#define HAL_RX_MPDU_START_INFO2_MPDU_LEN		GENMASK(13, 0)
 struct hal_rx_mpdu_start {
+	__le32 rsvd0[9];
 	__le32 info0;
 	__le32 info1;
-	__le32 rsvd1[11];
+	__le32 rsvd1[2];
 	__le32 info2;
-	__le32 rsvd2[9];
+	__le32 rsvd2[16];
 } __packed;
 
 #define HAL_RX_PPDU_END_DURATION	GENMASK(23, 0)
-- 
2.34.1


^ permalink raw reply related	[flat|nested] 21+ messages in thread

* [PATCH v3 07/11] wifi: ath12k: properly handling the state variables of monitor mode
  2024-10-16  6:56 [PATCH v3 00/11] wifi: ath12k: some fixes and clean up for monitor mode Kang Yang
                   ` (5 preceding siblings ...)
  2024-10-16  6:56 ` [PATCH v3 06/11] wifi: ath12k: fix struct hal_rx_mpdu_start Kang Yang
@ 2024-10-16  6:56 ` Kang Yang
  2024-10-16  6:56 ` [PATCH v3 08/11] wifi: ath12k: delete NSS and TX power setting for monitor vdev Kang Yang
                   ` (3 subsequent siblings)
  10 siblings, 0 replies; 21+ messages in thread
From: Kang Yang @ 2024-10-16  6:56 UTC (permalink / raw)
  To: ath12k; +Cc: linux-wireless, quic_kangyang

Now monitor mode has some status variables, such as monitor_started,
monitor_vdev_created. They are not always handled correctly.

For monitor_started, it should be true when driver works in monitor
mode, and should be false at other times.

For monitor_vdev_id, it should be -1 when driver doesn't work in monitor
mode.

For monitor_vdev_created, it should be true after monitor vdev is
created successfully and should be false at other times.

For monitor_conf_enabled, it should be true when mac80211 sets
IEEE80211_CONF_MONITOR, and should be false at other times.

Handle those state variables according to above descriptions.

Tested-on: WCN7850 hw2.0 PCI WLAN.HMT.1.0.c5-00481-QCAHMTSWPL_V1.0_V2.0_SILICONZ-3

Signed-off-by: Kang Yang <quic_kangyang@quicinc.com>
---
 drivers/net/wireless/ath/ath12k/core.c |  5 +++++
 drivers/net/wireless/ath/ath12k/mac.c  | 10 ++++++++++
 2 files changed, 15 insertions(+)

diff --git a/drivers/net/wireless/ath/ath12k/core.c b/drivers/net/wireless/ath/ath12k/core.c
index 9cd485ed42ab..419599c483d2 100644
--- a/drivers/net/wireless/ath/ath12k/core.c
+++ b/drivers/net/wireless/ath/ath12k/core.c
@@ -1060,6 +1060,11 @@ static void ath12k_core_pre_reconfigure_recovery(struct ath12k_base *ab)
 				     ath12k_mac_tx_mgmt_pending_free, ar);
 			idr_destroy(&ar->txmgmt_idr);
 			wake_up(&ar->txmgmt_empty_waitq);
+
+			ar->monitor_vdev_id = -1;
+			ar->monitor_started = false;
+			ar->monitor_vdev_created = false;
+			ar->monitor_conf_enabled = false;
 		}
 	}
 
diff --git a/drivers/net/wireless/ath/ath12k/mac.c b/drivers/net/wireless/ath/ath12k/mac.c
index d4b438e4b7bf..a145cb217cbd 100644
--- a/drivers/net/wireless/ath/ath12k/mac.c
+++ b/drivers/net/wireless/ath/ath12k/mac.c
@@ -6547,6 +6547,9 @@ static int ath12k_mac_vdev_create(struct ath12k *ar, struct ieee80211_vif *vif)
 			goto err_peer_del;
 		}
 		break;
+	case WMI_VDEV_TYPE_MONITOR:
+		ar->monitor_vdev_created = true;
+		break;
 	default:
 		break;
 	}
@@ -6596,6 +6599,8 @@ static int ath12k_mac_vdev_create(struct ath12k *ar, struct ieee80211_vif *vif)
 err_vdev_del:
 	ath12k_wmi_vdev_delete(ar, arvif->vdev_id);
 	ar->num_created_vdevs--;
+	ar->monitor_vdev_created = false;
+	ar->monitor_vdev_id = -1;
 	arvif->is_created = false;
 	arvif->ar = NULL;
 	ar->allocated_vdev_map &= ~(1LL << arvif->vdev_id);
@@ -9262,6 +9267,11 @@ static void ath12k_mac_setup(struct ath12k *ar)
 
 	INIT_WORK(&ar->wmi_mgmt_tx_work, ath12k_mgmt_over_wmi_tx_work);
 	skb_queue_head_init(&ar->wmi_mgmt_tx_queue);
+
+	ar->monitor_vdev_id = -1;
+	ar->monitor_started = false;
+	ar->monitor_vdev_created = false;
+	ar->monitor_conf_enabled = false;
 }
 
 int ath12k_mac_register(struct ath12k_base *ab)
-- 
2.34.1


^ permalink raw reply related	[flat|nested] 21+ messages in thread

* [PATCH v3 08/11] wifi: ath12k: delete NSS and TX power setting for monitor vdev
  2024-10-16  6:56 [PATCH v3 00/11] wifi: ath12k: some fixes and clean up for monitor mode Kang Yang
                   ` (6 preceding siblings ...)
  2024-10-16  6:56 ` [PATCH v3 07/11] wifi: ath12k: properly handling the state variables of monitor mode Kang Yang
@ 2024-10-16  6:56 ` Kang Yang
  2024-10-21  8:46   ` Kalle Valo
  2024-10-16  6:56 ` [PATCH v3 09/11] wifi: ath12k: use tail MSDU to get MSDU information Kang Yang
                   ` (2 subsequent siblings)
  10 siblings, 1 reply; 21+ messages in thread
From: Kang Yang @ 2024-10-16  6:56 UTC (permalink / raw)
  To: ath12k; +Cc: linux-wireless, quic_kangyang

ath12k_mac_monitor_vdev_create() will set NSS and TX power for monitor
vdev. But this is not needed for monitor vdev.

So delete them.

Tested-on: WCN7850 hw2.0 PCI WLAN.HMT.1.0.c5-00481-QCAHMTSWPL_V1.0_V2.0_SILICONZ-3
Tested-on: QCN9274 hw2.0 PCI WLAN.WBE.1.0.1-00029-QCAHKSWPL_SILICONZ-1

Signed-off-by: Kang Yang <quic_kangyang@quicinc.com>
---
 drivers/net/wireless/ath/ath12k/mac.c | 14 --------------
 1 file changed, 14 deletions(-)

diff --git a/drivers/net/wireless/ath/ath12k/mac.c b/drivers/net/wireless/ath/ath12k/mac.c
index a145cb217cbd..e4412f7a35b4 100644
--- a/drivers/net/wireless/ath/ath12k/mac.c
+++ b/drivers/net/wireless/ath/ath12k/mac.c
@@ -1068,7 +1068,6 @@ static int ath12k_mac_monitor_vdev_create(struct ath12k *ar)
 	struct ath12k_wmi_vdev_create_arg arg = {};
 	int bit, ret;
 	u8 tmp_addr[6];
-	u16 nss;
 
 	lockdep_assert_wiphy(ath12k_ar_to_hw(ar)->wiphy);
 
@@ -1108,19 +1107,6 @@ static int ath12k_mac_monitor_vdev_create(struct ath12k *ar)
 		return ret;
 	}
 
-	nss = hweight32(ar->cfg_tx_chainmask) ? : 1;
-	ret = ath12k_wmi_vdev_set_param_cmd(ar, ar->monitor_vdev_id,
-					    WMI_VDEV_PARAM_NSS, nss);
-	if (ret) {
-		ath12k_warn(ar->ab, "failed to set vdev %d chainmask 0x%x, nss %d :%d\n",
-			    ar->monitor_vdev_id, ar->cfg_tx_chainmask, nss, ret);
-		return ret;
-	}
-
-	ret = ath12k_mac_txpower_recalc(ar);
-	if (ret)
-		return ret;
-
 	ar->allocated_vdev_map |= 1LL << ar->monitor_vdev_id;
 	ar->ab->free_vdev_map &= ~(1LL << ar->monitor_vdev_id);
 	ar->num_created_vdevs++;
-- 
2.34.1


^ permalink raw reply related	[flat|nested] 21+ messages in thread

* [PATCH v3 09/11] wifi: ath12k: use tail MSDU to get MSDU information
  2024-10-16  6:56 [PATCH v3 00/11] wifi: ath12k: some fixes and clean up for monitor mode Kang Yang
                   ` (7 preceding siblings ...)
  2024-10-16  6:56 ` [PATCH v3 08/11] wifi: ath12k: delete NSS and TX power setting for monitor vdev Kang Yang
@ 2024-10-16  6:56 ` Kang Yang
  2024-10-16  6:56 ` [PATCH v3 10/11] wifi: ath12k: fix A-MSDU indication in monitor mode Kang Yang
  2024-10-16  6:57 ` [PATCH v3 11/11] wifi: ath12k: delete mon reap timer Kang Yang
  10 siblings, 0 replies; 21+ messages in thread
From: Kang Yang @ 2024-10-16  6:56 UTC (permalink / raw)
  To: ath12k; +Cc: linux-wireless, quic_kangyang

When receiving a packet in monitor mode, hardware will fill status
information msdu_end for MSDUs. Then host can use these status
information to get specific information about this packet, such as
l3 offset, bandwidth, mcs, packet type.

If this packet is composed of multiple MSDUs, then hardware will only
fill status information for the last(tail) MSDU. At this time,
MSDU information is uncertain for others MSDUs. But current code use
the first MSDU in monitor mode, so will get the wrong information.

Fix it by getting msdu information from last(tail) msdu.

Tested-on: WCN7850 hw2.0 PCI WLAN.HMT.1.0.c5-00481-QCAHMTSWPL_V1.0_V2.0_SILICONZ-3

Signed-off-by: Kang Yang <quic_kangyang@quicinc.com>
---
 drivers/net/wireless/ath/ath12k/dp_mon.c | 41 +++++++++++++-----------
 1 file changed, 23 insertions(+), 18 deletions(-)

diff --git a/drivers/net/wireless/ath/ath12k/dp_mon.c b/drivers/net/wireless/ath/ath12k/dp_mon.c
index 486f95593aa6..ed963ea2cb8f 100644
--- a/drivers/net/wireless/ath/ath12k/dp_mon.c
+++ b/drivers/net/wireless/ath/ath12k/dp_mon.c
@@ -855,24 +855,26 @@ ath12k_dp_mon_rx_parse_status_tlv(struct ath12k_base *ab,
 	return HAL_RX_MON_STATUS_PPDU_NOT_DONE;
 }
 
-static void ath12k_dp_mon_rx_msdus_set_payload(struct ath12k *ar, struct sk_buff *msdu)
+static void ath12k_dp_mon_rx_msdus_set_payload(struct ath12k *ar,
+					       struct sk_buff *head_msdu,
+					       struct sk_buff *tail_msdu)
 {
 	u32 rx_pkt_offset, l2_hdr_offset;
 
 	rx_pkt_offset = ar->ab->hal.hal_desc_sz;
-	l2_hdr_offset = ath12k_dp_rx_h_l3pad(ar->ab,
-					     (struct hal_rx_desc *)msdu->data);
-	skb_pull(msdu, rx_pkt_offset + l2_hdr_offset);
+	l2_hdr_offset =
+		ath12k_dp_rx_h_l3pad(ar->ab, (struct hal_rx_desc *)tail_msdu->data);
+	skb_pull(head_msdu, rx_pkt_offset + l2_hdr_offset);
 }
 
 static struct sk_buff *
-ath12k_dp_mon_rx_merg_msdus(struct ath12k *ar,
-			    u32 mac_id, struct sk_buff *head_msdu,
+ath12k_dp_mon_rx_merg_msdus(struct ath12k *ar, u32 mac_id,
+			    struct sk_buff *head_msdu, struct sk_buff *tail_msdu,
 			    struct ieee80211_rx_status *rxs, bool *fcs_err)
 {
 	struct ath12k_base *ab = ar->ab;
 	struct sk_buff *msdu, *mpdu_buf, *prev_buf;
-	struct hal_rx_desc *rx_desc;
+	struct hal_rx_desc *rx_desc, *tail_rx_desc;
 	u8 *hdr_desc, *dest, decap_format;
 	struct ieee80211_hdr_3addr *wh;
 	u32 err_bitmap;
@@ -883,23 +885,24 @@ ath12k_dp_mon_rx_merg_msdus(struct ath12k *ar,
 		goto err_merge_fail;
 
 	rx_desc = (struct hal_rx_desc *)head_msdu->data;
-	err_bitmap = ath12k_dp_rx_h_mpdu_err(ab, rx_desc);
+	tail_rx_desc = (struct hal_rx_desc *)tail_msdu->data;
 
+	err_bitmap = ath12k_dp_rx_h_mpdu_err(ab, tail_rx_desc);
 	if (err_bitmap & HAL_RX_MPDU_ERR_FCS)
 		*fcs_err = true;
 
-	decap_format = ath12k_dp_rx_h_decap_type(ab, rx_desc);
+	decap_format = ath12k_dp_rx_h_decap_type(ab, tail_rx_desc);
 
-	ath12k_dp_rx_h_ppdu(ar, rx_desc, rxs);
+	ath12k_dp_rx_h_ppdu(ar, tail_rx_desc, rxs);
 
 	if (decap_format == DP_RX_DECAP_TYPE_RAW) {
-		ath12k_dp_mon_rx_msdus_set_payload(ar, head_msdu);
+		ath12k_dp_mon_rx_msdus_set_payload(ar, head_msdu, tail_msdu);
 
 		prev_buf = head_msdu;
 		msdu = head_msdu->next;
 
 		while (msdu) {
-			ath12k_dp_mon_rx_msdus_set_payload(ar, msdu);
+			ath12k_dp_mon_rx_msdus_set_payload(ar, msdu, tail_msdu);
 
 			prev_buf = msdu;
 			msdu = msdu->next;
@@ -924,7 +927,7 @@ ath12k_dp_mon_rx_merg_msdus(struct ath12k *ar,
 		msdu = head_msdu;
 
 		while (msdu) {
-			ath12k_dp_mon_rx_msdus_set_payload(ar, msdu);
+			ath12k_dp_mon_rx_msdus_set_payload(ar, msdu, tail_msdu);
 			if (qos_pkt) {
 				dest = skb_push(msdu, sizeof(__le16));
 				if (!dest)
@@ -1130,7 +1133,7 @@ static void ath12k_dp_mon_rx_deliver_msdu(struct ath12k *ar, struct napi_struct
 }
 
 static int ath12k_dp_mon_rx_deliver(struct ath12k *ar, u32 mac_id,
-				    struct sk_buff *head_msdu,
+				    struct sk_buff *head_msdu, struct sk_buff *tail_msdu,
 				    struct hal_rx_mon_ppdu_info *ppduinfo,
 				    struct napi_struct *napi)
 {
@@ -1139,7 +1142,8 @@ static int ath12k_dp_mon_rx_deliver(struct ath12k *ar, u32 mac_id,
 	struct ieee80211_rx_status *rxs = &dp->rx_status;
 	bool fcs_err = false;
 
-	mon_skb = ath12k_dp_mon_rx_merg_msdus(ar, mac_id, head_msdu,
+	mon_skb = ath12k_dp_mon_rx_merg_msdus(ar, mac_id,
+					      head_msdu, tail_msdu,
 					      rxs, &fcs_err);
 	if (!mon_skb)
 		goto mon_deliver_fail;
@@ -1247,7 +1251,7 @@ ath12k_dp_mon_rx_parse_mon_status(struct ath12k *ar,
 
 		if (head_msdu && tail_msdu) {
 			ath12k_dp_mon_rx_deliver(ar, mac_id, head_msdu,
-						 ppdu_info, napi);
+						 tail_msdu, ppdu_info, napi);
 		}
 
 		kfree(mon_mpdu);
@@ -1943,15 +1947,16 @@ ath12k_dp_mon_tx_process_ppdu_info(struct ath12k *ar, int mac_id,
 				   struct dp_mon_tx_ppdu_info *tx_ppdu_info)
 {
 	struct dp_mon_mpdu *tmp, *mon_mpdu;
-	struct sk_buff *head_msdu;
+	struct sk_buff *head_msdu, *tail_msdu;
 
 	list_for_each_entry_safe(mon_mpdu, tmp,
 				 &tx_ppdu_info->dp_tx_mon_mpdu_list, list) {
 		list_del(&mon_mpdu->list);
 		head_msdu = mon_mpdu->head;
+		tail_msdu = mon_mpdu->tail;
 
 		if (head_msdu)
-			ath12k_dp_mon_rx_deliver(ar, mac_id, head_msdu,
+			ath12k_dp_mon_rx_deliver(ar, mac_id, head_msdu, tail_msdu,
 						 &tx_ppdu_info->rx_status, napi);
 
 		kfree(mon_mpdu);
-- 
2.34.1


^ permalink raw reply related	[flat|nested] 21+ messages in thread

* [PATCH v3 10/11] wifi: ath12k: fix A-MSDU indication in monitor mode
  2024-10-16  6:56 [PATCH v3 00/11] wifi: ath12k: some fixes and clean up for monitor mode Kang Yang
                   ` (8 preceding siblings ...)
  2024-10-16  6:56 ` [PATCH v3 09/11] wifi: ath12k: use tail MSDU to get MSDU information Kang Yang
@ 2024-10-16  6:56 ` Kang Yang
  2024-10-16  6:57 ` [PATCH v3 11/11] wifi: ath12k: delete mon reap timer Kang Yang
  10 siblings, 0 replies; 21+ messages in thread
From: Kang Yang @ 2024-10-16  6:56 UTC (permalink / raw)
  To: ath12k; +Cc: linux-wireless, quic_kangyang

In monitor mode, host will reap MSDUs and then the MSDUs with same PPDU
ID will be aggregated into an A-MSDU in ath12k_dp_mon_rx_merg_msdus().

However, ath12k_dp_mon_rx_merg_msdus() doesn't calculate the total
length of the A-MSDU. This causes Wireshark to display the A-MSDU
incorrectly.

Therefore, need to add calculation of the A-MSDU length.

Tested-on: WCN7850 hw2.0 PCI WLAN.HMT.1.0.c5-00481-QCAHMTSWPL_V1.0_V2.0_SILICONZ-3

Signed-off-by: Kang Yang <quic_kangyang@quicinc.com>
---
 drivers/net/wireless/ath/ath12k/dp_mon.c | 15 +++++++++++++--
 1 file changed, 13 insertions(+), 2 deletions(-)

diff --git a/drivers/net/wireless/ath/ath12k/dp_mon.c b/drivers/net/wireless/ath/ath12k/dp_mon.c
index ed963ea2cb8f..1105c1a0da19 100644
--- a/drivers/net/wireless/ath/ath12k/dp_mon.c
+++ b/drivers/net/wireless/ath/ath12k/dp_mon.c
@@ -873,11 +873,11 @@ ath12k_dp_mon_rx_merg_msdus(struct ath12k *ar, u32 mac_id,
 			    struct ieee80211_rx_status *rxs, bool *fcs_err)
 {
 	struct ath12k_base *ab = ar->ab;
-	struct sk_buff *msdu, *mpdu_buf, *prev_buf;
+	struct sk_buff *msdu, *mpdu_buf, *prev_buf, *head_frag_list;
 	struct hal_rx_desc *rx_desc, *tail_rx_desc;
 	u8 *hdr_desc, *dest, decap_format;
 	struct ieee80211_hdr_3addr *wh;
-	u32 err_bitmap;
+	u32 err_bitmap, frag_list_sum_len = 0;
 
 	mpdu_buf = NULL;
 
@@ -900,10 +900,15 @@ ath12k_dp_mon_rx_merg_msdus(struct ath12k *ar, u32 mac_id,
 
 		prev_buf = head_msdu;
 		msdu = head_msdu->next;
+		head_frag_list = NULL;
 
 		while (msdu) {
 			ath12k_dp_mon_rx_msdus_set_payload(ar, msdu, tail_msdu);
 
+			if (!head_frag_list)
+				head_frag_list = msdu;
+
+			frag_list_sum_len += msdu->len;
 			prev_buf = msdu;
 			msdu = msdu->next;
 		}
@@ -911,6 +916,12 @@ ath12k_dp_mon_rx_merg_msdus(struct ath12k *ar, u32 mac_id,
 		prev_buf->next = NULL;
 
 		skb_trim(prev_buf, prev_buf->len - HAL_RX_FCS_LEN);
+		if (head_frag_list) {
+			skb_shinfo(head_msdu)->frag_list = head_frag_list;
+			head_msdu->data_len = frag_list_sum_len;
+			head_msdu->len += head_msdu->data_len;
+			head_msdu->next = NULL;
+		}
 	} else if (decap_format == DP_RX_DECAP_TYPE_NATIVE_WIFI) {
 		u8 qos_pkt = 0;
 
-- 
2.34.1


^ permalink raw reply related	[flat|nested] 21+ messages in thread

* [PATCH v3 11/11] wifi: ath12k: delete mon reap timer
  2024-10-16  6:56 [PATCH v3 00/11] wifi: ath12k: some fixes and clean up for monitor mode Kang Yang
                   ` (9 preceding siblings ...)
  2024-10-16  6:56 ` [PATCH v3 10/11] wifi: ath12k: fix A-MSDU indication in monitor mode Kang Yang
@ 2024-10-16  6:57 ` Kang Yang
  2024-10-21  8:47   ` Kalle Valo
  10 siblings, 1 reply; 21+ messages in thread
From: Kang Yang @ 2024-10-16  6:57 UTC (permalink / raw)
  To: ath12k; +Cc: linux-wireless, quic_kangyang

Currently mon reap timer is not used, and it is not needed anymore.

So remove related code.

Tested-on: WCN7850 hw2.0 PCI WLAN.HMT.1.0.c5-00481-QCAHMTSWPL_V1.0_V2.0_SILICONZ-3

Signed-off-by: Kang Yang <quic_kangyang@quicinc.com>
---
 drivers/net/wireless/ath/ath12k/core.h |  2 --
 drivers/net/wireless/ath/ath12k/dp.c   | 25 -------------------------
 2 files changed, 27 deletions(-)

diff --git a/drivers/net/wireless/ath/ath12k/core.h b/drivers/net/wireless/ath/ath12k/core.h
index 0c39b37d6a47..87240fb57c73 100644
--- a/drivers/net/wireless/ath/ath12k/core.h
+++ b/drivers/net/wireless/ath/ath12k/core.h
@@ -891,8 +891,6 @@ struct ath12k_base {
 	struct ath12k_dbring_cap *db_caps;
 	u32 num_db_cap;
 
-	struct timer_list mon_reap_timer;
-
 	struct completion htc_suspend;
 
 	u64 fw_soc_drop_count;
diff --git a/drivers/net/wireless/ath/ath12k/dp.c b/drivers/net/wireless/ath/ath12k/dp.c
index 2ab2a7d45be9..099a1fcd88b4 100644
--- a/drivers/net/wireless/ath/ath12k/dp.c
+++ b/drivers/net/wireless/ath/ath12k/dp.c
@@ -974,8 +974,6 @@ void ath12k_dp_pdev_free(struct ath12k_base *ab)
 {
 	int i;
 
-	del_timer_sync(&ab->mon_reap_timer);
-
 	for (i = 0; i < ab->num_radios; i++)
 		ath12k_dp_rx_pdev_free(ab, i);
 }
@@ -1020,27 +1018,6 @@ void ath12k_dp_hal_rx_desc_init(struct ath12k_base *ab)
 		ab->hal_rx_ops->rx_desc_get_desc_size();
 }
 
-static void ath12k_dp_service_mon_ring(struct timer_list *t)
-{
-	struct ath12k_base *ab = from_timer(ab, t, mon_reap_timer);
-	int i;
-
-	for (i = 0; i < ab->hw_params->num_rxdma_per_pdev; i++)
-		ath12k_dp_mon_process_ring(ab, i, NULL, DP_MON_SERVICE_BUDGET,
-					   ATH12K_DP_RX_MONITOR_MODE);
-
-	mod_timer(&ab->mon_reap_timer, jiffies +
-		  msecs_to_jiffies(ATH12K_MON_TIMER_INTERVAL));
-}
-
-static void ath12k_dp_mon_reap_timer_init(struct ath12k_base *ab)
-{
-	if (ab->hw_params->rxdma1_enable)
-		return;
-
-	timer_setup(&ab->mon_reap_timer, ath12k_dp_service_mon_ring, 0);
-}
-
 int ath12k_dp_pdev_alloc(struct ath12k_base *ab)
 {
 	struct ath12k *ar;
@@ -1051,8 +1028,6 @@ int ath12k_dp_pdev_alloc(struct ath12k_base *ab)
 	if (ret)
 		goto out;
 
-	ath12k_dp_mon_reap_timer_init(ab);
-
 	/* TODO: Per-pdev rx ring unlike tx ring which is mapped to different AC's */
 	for (i = 0; i < ab->num_radios; i++) {
 		ar = ab->pdevs[i].ar;
-- 
2.34.1


^ permalink raw reply related	[flat|nested] 21+ messages in thread

* Re: [PATCH v3 03/11] wifi: ath12k: fix struct hal_rx_ppdu_end_user_stats
  2024-10-16  6:56 ` [PATCH v3 03/11] wifi: ath12k: fix struct hal_rx_ppdu_end_user_stats Kang Yang
@ 2024-10-21  8:45   ` Kalle Valo
  2024-10-21  8:50     ` Kalle Valo
  0 siblings, 1 reply; 21+ messages in thread
From: Kalle Valo @ 2024-10-21  8:45 UTC (permalink / raw)
  To: Kang Yang; +Cc: ath12k, linux-wireless

Kang Yang <quic_kangyang@quicinc.com> writes:

> Current struct hal_rx_ppdu_end_user_stats in hal_rx.h is not matched
> with hardware descriptor definition.
>
> So update this structure and related code.
>
> Tested-on: WCN7850 hw2.0 PCI WLAN.HMT.1.0.c5-00481-QCAHMTSWPL_V1.0_V2.0_SILICONZ-3
>
> Signed-off-by: Kang Yang <quic_kangyang@quicinc.com>

What about QCN9274? The commit message doesn't mention anything about
that.

-- 
https://patchwork.kernel.org/project/linux-wireless/list/

https://wireless.wiki.kernel.org/en/developers/documentation/submittingpatches

^ permalink raw reply	[flat|nested] 21+ messages in thread

* Re: [PATCH v3 08/11] wifi: ath12k: delete NSS and TX power setting for monitor vdev
  2024-10-16  6:56 ` [PATCH v3 08/11] wifi: ath12k: delete NSS and TX power setting for monitor vdev Kang Yang
@ 2024-10-21  8:46   ` Kalle Valo
  2024-10-21  9:21     ` Kang Yang
  0 siblings, 1 reply; 21+ messages in thread
From: Kalle Valo @ 2024-10-21  8:46 UTC (permalink / raw)
  To: Kang Yang; +Cc: ath12k, linux-wireless

Kang Yang <quic_kangyang@quicinc.com> writes:

> ath12k_mac_monitor_vdev_create() will set NSS and TX power for monitor
> vdev. But this is not needed for monitor vdev.
>
> So delete them.
>
> Tested-on: WCN7850 hw2.0 PCI WLAN.HMT.1.0.c5-00481-QCAHMTSWPL_V1.0_V2.0_SILICONZ-3
> Tested-on: QCN9274 hw2.0 PCI WLAN.WBE.1.0.1-00029-QCAHKSWPL_SILICONZ-1
>
> Signed-off-by: Kang Yang <quic_kangyang@quicinc.com>

Does this fix any known issues? Or is just cleanup?

-- 
https://patchwork.kernel.org/project/linux-wireless/list/

https://wireless.wiki.kernel.org/en/developers/documentation/submittingpatches

^ permalink raw reply	[flat|nested] 21+ messages in thread

* Re: [PATCH v3 11/11] wifi: ath12k: delete mon reap timer
  2024-10-16  6:57 ` [PATCH v3 11/11] wifi: ath12k: delete mon reap timer Kang Yang
@ 2024-10-21  8:47   ` Kalle Valo
  2024-10-21  9:20     ` Kang Yang
  0 siblings, 1 reply; 21+ messages in thread
From: Kalle Valo @ 2024-10-21  8:47 UTC (permalink / raw)
  To: Kang Yang; +Cc: ath12k, linux-wireless

Kang Yang <quic_kangyang@quicinc.com> writes:

> Currently mon reap timer is not used, and it is not needed anymore.
>
> So remove related code.
>
> Tested-on: WCN7850 hw2.0 PCI WLAN.HMT.1.0.c5-00481-QCAHMTSWPL_V1.0_V2.0_SILICONZ-3
>
> Signed-off-by: Kang Yang <quic_kangyang@quicinc.com>

Same question, does it fix a known issue or is just cleanup? Always
answer to the question 'why?'.

-- 
https://patchwork.kernel.org/project/linux-wireless/list/

https://wireless.wiki.kernel.org/en/developers/documentation/submittingpatches

^ permalink raw reply	[flat|nested] 21+ messages in thread

* Re: [PATCH v3 03/11] wifi: ath12k: fix struct hal_rx_ppdu_end_user_stats
  2024-10-21  8:45   ` Kalle Valo
@ 2024-10-21  8:50     ` Kalle Valo
  2024-10-21  9:32       ` Kang Yang
  0 siblings, 1 reply; 21+ messages in thread
From: Kalle Valo @ 2024-10-21  8:50 UTC (permalink / raw)
  To: Kang Yang; +Cc: ath12k, linux-wireless

Kalle Valo <kvalo@kernel.org> writes:

> Kang Yang <quic_kangyang@quicinc.com> writes:
>
>> Current struct hal_rx_ppdu_end_user_stats in hal_rx.h is not matched
>> with hardware descriptor definition.
>>
>> So update this structure and related code.
>>
>> Tested-on: WCN7850 hw2.0 PCI WLAN.HMT.1.0.c5-00481-QCAHMTSWPL_V1.0_V2.0_SILICONZ-3
>>
>> Signed-off-by: Kang Yang <quic_kangyang@quicinc.com>
>
> What about QCN9274? The commit message doesn't mention anything about
> that.

I have the same question for patches 3-6. Also I would like to
understand the backwards compatibility with older WCN7850 firmwares.

-- 
https://patchwork.kernel.org/project/linux-wireless/list/

https://wireless.wiki.kernel.org/en/developers/documentation/submittingpatches

^ permalink raw reply	[flat|nested] 21+ messages in thread

* Re: [PATCH v3 11/11] wifi: ath12k: delete mon reap timer
  2024-10-21  8:47   ` Kalle Valo
@ 2024-10-21  9:20     ` Kang Yang
  0 siblings, 0 replies; 21+ messages in thread
From: Kang Yang @ 2024-10-21  9:20 UTC (permalink / raw)
  To: Kalle Valo; +Cc: ath12k, linux-wireless



On 10/21/2024 4:47 PM, Kalle Valo wrote:
> Kang Yang <quic_kangyang@quicinc.com> writes:
> 
>> Currently mon reap timer is not used, and it is not needed anymore.
>>
>> So remove related code.
>>
>> Tested-on: WCN7850 hw2.0 PCI WLAN.HMT.1.0.c5-00481-QCAHMTSWPL_V1.0_V2.0_SILICONZ-3
>>
>> Signed-off-by: Kang Yang <quic_kangyang@quicinc.com>
> 
> Same question, does it fix a known issue or is just cleanup? Always
> answer to the question 'why?'.

It is a clean up. Ath11k uses mon reap timer to implement rx pktlog. But 
ath12k doesn't support it yet. If we want to support, reuse the 
interrupt of monitor status ring is better. Because:
1. monitor mode also need the configuration of monitor status ring.
2. the rx pktlog is highly dependent on the implementation of monitor mode.

> 


^ permalink raw reply	[flat|nested] 21+ messages in thread

* Re: [PATCH v3 08/11] wifi: ath12k: delete NSS and TX power setting for monitor vdev
  2024-10-21  8:46   ` Kalle Valo
@ 2024-10-21  9:21     ` Kang Yang
  2024-10-28  7:13       ` Kalle Valo
  0 siblings, 1 reply; 21+ messages in thread
From: Kang Yang @ 2024-10-21  9:21 UTC (permalink / raw)
  To: Kalle Valo; +Cc: ath12k, linux-wireless



On 10/21/2024 4:46 PM, Kalle Valo wrote:
> Kang Yang <quic_kangyang@quicinc.com> writes:
> 
>> ath12k_mac_monitor_vdev_create() will set NSS and TX power for monitor
>> vdev. But this is not needed for monitor vdev.
>>
>> So delete them.
>>
>> Tested-on: WCN7850 hw2.0 PCI WLAN.HMT.1.0.c5-00481-QCAHMTSWPL_V1.0_V2.0_SILICONZ-3
>> Tested-on: QCN9274 hw2.0 PCI WLAN.WBE.1.0.1-00029-QCAHKSWPL_SILICONZ-1
>>
>> Signed-off-by: Kang Yang <quic_kangyang@quicinc.com>
> 
> Does this fix any known issues? Or is just cleanup?

Clean up

> 


^ permalink raw reply	[flat|nested] 21+ messages in thread

* Re: [PATCH v3 03/11] wifi: ath12k: fix struct hal_rx_ppdu_end_user_stats
  2024-10-21  8:50     ` Kalle Valo
@ 2024-10-21  9:32       ` Kang Yang
  2024-10-28  7:16         ` Kalle Valo
  0 siblings, 1 reply; 21+ messages in thread
From: Kang Yang @ 2024-10-21  9:32 UTC (permalink / raw)
  To: Kalle Valo; +Cc: ath12k, linux-wireless



On 10/21/2024 4:50 PM, Kalle Valo wrote:
> Kalle Valo <kvalo@kernel.org> writes:
> 
>> Kang Yang <quic_kangyang@quicinc.com> writes:
>>
>>> Current struct hal_rx_ppdu_end_user_stats in hal_rx.h is not matched
>>> with hardware descriptor definition.
>>>
>>> So update this structure and related code.
>>>
>>> Tested-on: WCN7850 hw2.0 PCI WLAN.HMT.1.0.c5-00481-QCAHMTSWPL_V1.0_V2.0_SILICONZ-3
>>>
>>> Signed-off-by: Kang Yang <quic_kangyang@quicinc.com>
>>
>> What about QCN9274? The commit message doesn't mention anything about
>> that.
> 
> I have the same question for patches 3-6. Also I would like to
> understand the backwards compatibility with older WCN7850 firmwares.

For QCN9274, we reuse the same struct for them.
WIN team already checked them during internal review.

The struct definitions in patch 3-6 are hardware descriptor definition, 
won't be affected by firmware and host. They are fixed.

> 


^ permalink raw reply	[flat|nested] 21+ messages in thread

* Re: [PATCH v3 08/11] wifi: ath12k: delete NSS and TX power setting for monitor vdev
  2024-10-21  9:21     ` Kang Yang
@ 2024-10-28  7:13       ` Kalle Valo
  0 siblings, 0 replies; 21+ messages in thread
From: Kalle Valo @ 2024-10-28  7:13 UTC (permalink / raw)
  To: Kang Yang; +Cc: ath12k, linux-wireless

Kang Yang <quic_kangyang@quicinc.com> writes:

> On 10/21/2024 4:46 PM, Kalle Valo wrote:
>> Kang Yang <quic_kangyang@quicinc.com> writes:
>> 
>>> ath12k_mac_monitor_vdev_create() will set NSS and TX power for monitor
>>> vdev. But this is not needed for monitor vdev.
>>>
>>> So delete them.
>>>
>>> Tested-on: WCN7850 hw2.0 PCI WLAN.HMT.1.0.c5-00481-QCAHMTSWPL_V1.0_V2.0_SILICONZ-3
>>> Tested-on: QCN9274 hw2.0 PCI WLAN.WBE.1.0.1-00029-QCAHKSWPL_SILICONZ-1
>>>
>>> Signed-off-by: Kang Yang <quic_kangyang@quicinc.com>
>> Does this fix any known issues? Or is just cleanup?
>
> Clean up

Ok, please always mention this in the commit message. In essence you
need to answer to the question 'why are we doing this?'.

-- 
https://patchwork.kernel.org/project/linux-wireless/list/

https://wireless.wiki.kernel.org/en/developers/documentation/submittingpatches

^ permalink raw reply	[flat|nested] 21+ messages in thread

* Re: [PATCH v3 03/11] wifi: ath12k: fix struct hal_rx_ppdu_end_user_stats
  2024-10-21  9:32       ` Kang Yang
@ 2024-10-28  7:16         ` Kalle Valo
  0 siblings, 0 replies; 21+ messages in thread
From: Kalle Valo @ 2024-10-28  7:16 UTC (permalink / raw)
  To: Kang Yang; +Cc: ath12k, linux-wireless

Kang Yang <quic_kangyang@quicinc.com> writes:

> On 10/21/2024 4:50 PM, Kalle Valo wrote:
>> Kalle Valo <kvalo@kernel.org> writes:
>> 
>>> Kang Yang <quic_kangyang@quicinc.com> writes:
>>>
>>>> Current struct hal_rx_ppdu_end_user_stats in hal_rx.h is not matched
>>>> with hardware descriptor definition.
>>>>
>>>> So update this structure and related code.
>>>>
>>>> Tested-on: WCN7850 hw2.0 PCI WLAN.HMT.1.0.c5-00481-QCAHMTSWPL_V1.0_V2.0_SILICONZ-3
>>>>
>>>> Signed-off-by: Kang Yang <quic_kangyang@quicinc.com>
>>>
>>> What about QCN9274? The commit message doesn't mention anything about
>>> that.
>> I have the same question for patches 3-6. Also I would like to
>> understand the backwards compatibility with older WCN7850 firmwares.
>
> For QCN9274, we reuse the same struct for them.
> WIN team already checked them during internal review.
>
> The struct definitions in patch 3-6 are hardware descriptor
> definition, won't be affected by firmware and host. They are fixed.

Please always include this information to the commit message so that we
know that you have done the research properly and the patch is not
breaking other hardware. ath12k is not only about WCN7850.

-- 
https://patchwork.kernel.org/project/linux-wireless/list/

https://wireless.wiki.kernel.org/en/developers/documentation/submittingpatches

^ permalink raw reply	[flat|nested] 21+ messages in thread

end of thread, other threads:[~2024-10-28  7:16 UTC | newest]

Thread overview: 21+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2024-10-16  6:56 [PATCH v3 00/11] wifi: ath12k: some fixes and clean up for monitor mode Kang Yang
2024-10-16  6:56 ` [PATCH v3 01/11] wifi: ath12k: remove unused variable monitor_present Kang Yang
2024-10-16  6:56 ` [PATCH v3 02/11] wifi: ath12k: optimize storage size for struct ath12k Kang Yang
2024-10-16  6:56 ` [PATCH v3 03/11] wifi: ath12k: fix struct hal_rx_ppdu_end_user_stats Kang Yang
2024-10-21  8:45   ` Kalle Valo
2024-10-21  8:50     ` Kalle Valo
2024-10-21  9:32       ` Kang Yang
2024-10-28  7:16         ` Kalle Valo
2024-10-16  6:56 ` [PATCH v3 04/11] wifi: ath12k: fix struct hal_rx_ppdu_start Kang Yang
2024-10-16  6:56 ` [PATCH v3 05/11] wifi: ath12k: fix struct hal_rx_phyrx_rssi_legacy_info Kang Yang
2024-10-16  6:56 ` [PATCH v3 06/11] wifi: ath12k: fix struct hal_rx_mpdu_start Kang Yang
2024-10-16  6:56 ` [PATCH v3 07/11] wifi: ath12k: properly handling the state variables of monitor mode Kang Yang
2024-10-16  6:56 ` [PATCH v3 08/11] wifi: ath12k: delete NSS and TX power setting for monitor vdev Kang Yang
2024-10-21  8:46   ` Kalle Valo
2024-10-21  9:21     ` Kang Yang
2024-10-28  7:13       ` Kalle Valo
2024-10-16  6:56 ` [PATCH v3 09/11] wifi: ath12k: use tail MSDU to get MSDU information Kang Yang
2024-10-16  6:56 ` [PATCH v3 10/11] wifi: ath12k: fix A-MSDU indication in monitor mode Kang Yang
2024-10-16  6:57 ` [PATCH v3 11/11] wifi: ath12k: delete mon reap timer Kang Yang
2024-10-21  8:47   ` Kalle Valo
2024-10-21  9:20     ` Kang Yang

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox