ath10k: Fixes during subsystem recovery

This patch series includes some fixes when the device is in recovery mode, i.e. when the firmware goes down.

- Pausing TX queues when FW goes down
- Removed unwanted/extra error logging in pkt TX path
- Skipping wait for FW response for delete cmds
- Handling the -ESHUTDOWN error code in case of SSR.

Rakesh Pillai (2):
ath10k: Pause the tx queues when firmware is down
ath10k: Skip wait for delete response if firmware is down

 drivers/net/wireless/ath/ath10k/core.h |  1 +
 drivers/net/wireless/ath/ath10k/mac.c  | 36 ++++++++++++++++++++++------------
 drivers/net/wireless/ath/ath10k/snoc.c |  3 +++
 3 files changed, 28 insertions(+), 12 deletions(-)

2.7.4


_______________________________________________
ath10k mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/ath10k

When the FW goes down, the transmission of packets cannot be done. Currently we are not pausing the tx queues, hence there will be too many attempts to trasmit packets and each of them resulting in failure and thereby logging unwanted failure messages onto the console, as shown below

[   60.842192] ath10k_snoc 18800000.wifi: failed to transmit packet, dropping: -108
[   60.850181] ath10k_snoc 18800000.wifi: failed to submit frame: -108
[   60.857058] ath10k_snoc 18800000.wifi: failed to push frame: -108
[   60.936217] ath10k_snoc 18800000.wifi: failed to transmit packet, dropping: -108
[   60.944496] ath10k_snoc 18800000.wifi: failed to submit frame: -108
[   60.951265] ath10k_snoc 18800000.wifi: failed to push frame: -108
[   61.976589] ath10k_snoc 18800000.wifi: failed to transmit packet, dropping: -108

Stop the tx queues when the wlan firmware is down and also remove the unwanted duplicate/extra logging in the packet transmission path.

Tested-on: WCN3990 hw1.0 SNOC WLAN.HL.3.1-01040-QCAHLSWMTPLZ-1

---
 drivers/net/wireless/ath/ath10k/core.h |  1 +
 drivers/net/wireless/ath/ath10k/mac.c  | 18 ++++++++++--------
 drivers/net/wireless/ath/ath10k/snoc.c |  3 +++
 3 files changed, 14 insertions(+), 8 deletions(-)

index 62b1502..040053a 100644
@@ -900,6 +900,7 @@ static inline const char *ath10k_scan_state_str(enum ath10k_scan_state state)
 
 enum ath10k_tx_pause_reason {
 	ATH10K_TX_PAUSE_Q_FULL,
+	ATH10K_TX_PAUSE_FW_DOWN,
 	ATH10K_TX_PAUSE_MAX,
 };
 

index 3e38962..dc7befc 100644
@@ -3348,6 +3348,7 @@ void ath10k_mac_tx_lock(struct ath10k *ar, int reason)
 	ar->tx_paused |= BIT(reason);
 	ieee80211_stop_queues(ar->hw);
 }
+EXPORT_SYMBOL(ath10k_mac_tx_lock);
 
 static void ath10k_mac_tx_unlock_iter(void *data, u8 *mac,
 				      struct ieee80211_vif *vif)

@@ -3378,6 +3379,7 @@ void ath10k_mac_tx_unlock(struct ath10k *ar, int reason)
 
 	ieee80211_wake_queue(ar->hw, ar->hw->offchannel_tx_hw_queue);
 }
+EXPORT_SYMBOL(ath10k_mac_tx_unlock);
 
 void ath10k_mac_vif_tx_lock(struct ath10k_vif *arvif, int reason)
 {

@@ -3748,11 +3750,8 @@ static int ath10k_mac_tx_submit(struct ath10k *ar,
 		break;
 	}
 
-	if (ret) {
-		ath10k_warn(ar, "failed to transmit packet, dropping: %d\n",
-			    ret);
+	if (ret)
 		ieee80211_free_txskb(ar->hw, skb);
-	}
 
 	return ret;
 }

@@ -3806,10 +3805,8 @@ static int ath10k_mac_tx(struct ath10k *ar,
 	}
 
 	ret = ath10k_mac_tx_submit(ar, txmode, txpath, skb);
-	if (ret) {
-		ath10k_warn(ar, "failed to submit frame: %d\n", ret);
+	if (ret)
 		return ret;
-	}
 
 	return 0;
 }

@@ -4441,7 +4438,12 @@ static void ath10k_mac_op_tx(struct ieee80211_hw *hw,
 
 	ret = ath10k_mac_tx(ar, vif, txmode, txpath, skb, false);
 	if (ret) {
-		ath10k_warn(ar, "failed to transmit frame: %d\n", ret);
+		if (ret == -ESHUTDOWN)
+			ath10k_dbg(ar, ATH10K_DBG_MAC, "failed to transmit frame: %d\n",
+				   ret);
+		else
+			ath10k_warn(ar, "failed to transmit frame: %d\n", ret);
+
 		if (is_htt) {
 			spin_lock_bh(&ar->htt.tx_lock);
 			ath10k_htt_tx_dec_pending(htt);

index 407a074..8440e64 100644
@@ -20,6 +20,7 @@
 #include "hif.h"
 #include "htc.h"
 #include "snoc.h"
+#include "mac.h"
 
 #define ATH10K_SNOC_RX_POST_RETRY_MS 50
 #define CE_POLL_PIPE 4

@@ -1296,6 +1297,7 @@ int ath10k_snoc_fw_indication(struct ath10k *ar, u64 type)
 	switch (type) {
 	case ATH10K_QMI_EVENT_FW_READY_IND:
 		if (test_bit(ATH10K_SNOC_FLAG_REGISTERED, &ar_snoc->flags)) {
+			ath10k_mac_tx_unlock(ar, ATH10K_TX_PAUSE_FW_DOWN);
 			queue_work(ar->workqueue, &ar->restart_work);
 			break;
 		}

@@ -1313,6 +1315,7 @@ int ath10k_snoc_fw_indication(struct ath10k *ar, u64 type)
 	case ATH10K_QMI_EVENT_FW_DOWN_IND:
 		set_bit(ATH10K_SNOC_FLAG_RECOVERY, &ar_snoc->flags);
 		set_bit(ATH10K_FLAG_CRASH_FLUSH, &ar->dev_flags);
+		ath10k_mac_tx_lock(ar, ATH10K_TX_PAUSE_FW_DOWN);
 		break;
 	default:
 		ath10k_err(ar, "invalid fw indication: %llx\n", type);

2.7.4


_______________________________________________
ath10k mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/ath10k

Currently the driver waits for response from the firmware for all the delete cmds, eg: vdev_delete, peer delete. If the firmware is down, these wait will always timeout and return an error.

Also during subsytems recovery, any attempt to send a WMI cmd to the FW will return the -ESHUTDOWN status, which when returned to mac80211, can cause unnecessary warnings to be printed on to the console, as shown below

[ 2559.529565] Call trace:
[ 2559.532214]  __sta_info_destroy_part2+0x160/0x168 [mac80211]
[ 2559.538157]  __sta_info_flush+0x124/0x180 [mac80211]
[ 2559.543402]  ieee80211_set_disassoc+0x130/0x2c0 [mac80211]
[ 2559.549172]  ieee80211_mgd_deauth+0x238/0x25c [mac80211]
[ 2559.554764]  ieee80211_deauth+0x24/0x30 [mac80211]
[ 2559.559860]  cfg80211_mlme_deauth+0x258/0x2b0 [cfg80211]
[ 2559.565446]  nl80211_deauthenticate+0xe4/0x110 [cfg80211]
[ 2559.571064]  genl_rcv_msg+0x3a0/0x440
[ 2559.574888]  netlink_rcv_skb+0xb4/0x11c
[ 2559.578877]  genl_rcv+0x34/0x48
[ 2559.582162]  netlink_unicast+0x14c/0x1e4
[ 2559.586235]  netlink_sendmsg+0x2f0/0x360
[ 2559.590317]  sock_sendmsg+0x44/0x5c
[ 2559.593951]  ____sys_sendmsg+0x1c8/0x290
[ 2559.598029]  ___sys_sendmsg+0xa8/0xfc
[ 2559.601840]  __sys_sendmsg+0x8c/0xd0
[ 2559.605572]  __arm64_compat_sys_sendmsg+0x2c/0x38
[ 2559.610468]  el0_svc_common+0xa8/0x160
[ 2559.614372]  el0_svc_compat_handler+0x2c/0x38
[ 2559.618905]  el0_svc_compat+0x8/0x10

Skip the wait for delete response from the firmware if the firmware is down. Also return success to the mac80211 calls when the peer delete cmd fails with return status -ESHUTDOWN.

Tested-on: WCN3990 hw1.0 SNOC WLAN.HL.3.1-01040-QCAHLSWMTPLZ-1

---
 drivers/net/wireless/ath/ath10k/mac.c | 18 ++++++++++++++----
 1 file changed, 14 insertions(+), 4 deletions(-)

index dc7befc..6882fcc 100644
@@ -701,7 +701,8 @@ static void ath10k_wait_for_peer_delete_done(struct ath10k *ar, u32 vdev_id,
 	unsigned long time_left;
 	int ret;
 
-	if (test_bit(WMI_SERVICE_SYNC_DELETE_CMDS, ar->wmi.svc_map)) {
+	if (test_bit(WMI_SERVICE_SYNC_DELETE_CMDS, ar->wmi.svc_map) &&
+	    !test_bit(ATH10K_FLAG_CRASH_FLUSH, &ar->dev_flags)) {
 		ret = ath10k_wait_for_peer_deleted(ar, vdev_id, addr);
 		if (ret) {
 			ath10k_warn(ar, "failed wait for peer deleted");

@@ -841,7 +842,8 @@ static int ath10k_peer_delete(struct ath10k *ar, u32 vdev_id, const u8 *addr)
 	if (ret)
 		return ret;
 
-	if (test_bit(WMI_SERVICE_SYNC_DELETE_CMDS, ar->wmi.svc_map)) {
+	if (test_bit(WMI_SERVICE_SYNC_DELETE_CMDS, ar->wmi.svc_map) &&
+	    !test_bit(ATH10K_FLAG_CRASH_FLUSH, &ar->dev_flags)) {
 		unsigned long time_left;
 
 		time_left = wait_for_completion_timeout

@@ -5673,7 +5675,8 @@ static void ath10k_remove_interface(struct ieee80211_hw *hw,
 		ath10k_warn(ar, "failed to delete WMI vdev %i: %d\n",
 			    arvif->vdev_id, ret);
 
-	if (test_bit(WMI_SERVICE_SYNC_DELETE_CMDS, ar->wmi.svc_map)) {
+	if (test_bit(WMI_SERVICE_SYNC_DELETE_CMDS, ar->wmi.svc_map) &&
+	    !test_bit(ATH10K_FLAG_CRASH_FLUSH, &ar->dev_flags)) {
 		time_left = wait_for_completion_timeout(&ar->vdev_delete_done,
 							ATH10K_VDEV_DELETE_TIMEOUT_HZ);
 		if (time_left == 0) {

@@ -6110,6 +6113,11 @@ static int ath10k_hw_scan(struct ieee80211_hw *hw,
 		goto exit;
 	}
 
+	if (test_bit(ATH10K_FLAG_CRASH_FLUSH, &ar->dev_flags)) {
+		ret = -ESHUTDOWN;
+		goto exit;
+	}
+
 	spin_lock_bh(&ar->data_lock);
 	switch (ar->scan.state) {
 	case ATH10K_SCAN_IDLE:

@@ -6758,7 +6766,9 @@ static int ath10k_sta_state(struct ieee80211_hw *hw,
 		}
 
 		ret = ath10k_peer_delete(ar, arvif->vdev_id, sta->addr);
-		if (ret)
+		if (ret == -ESHUTDOWN)
+			ret = 0;
+		else if (ret)
 			ath10k_warn(ar, "failed to delete peer %pM for vdev %d: %i\n",
 				    sta->addr, arvif->vdev_id, ret);
 

2.7.4


_______________________________________________
ath10k mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/ath10k

Hi Kalle,
I see that this patch series is in Deferred state. Is there something missing or blocking this ?

_______________________________________________
ath10k mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/ath10k

Yeah, time to review it ;) I still have quite a lot of patches in deferred state but I'm trying to go through it, albeit slowly.

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

_______________________________________________
ath10k mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/ath10k

This has been tested only on WCN3990. How do I know that this doesn't break other hardware families?

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

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

" ath10k: Pause the tx queues when firmware is down"
This is done only for SNOC (wcn3990) target and does not affect other targets.

" ath10k: Skip wait for delete response if firmware is down" The skip for wmi responses is done based on the flag "ATH10K_FLAG_CRASH_FLUSH", which is generic for all targets. Since if the FW is down, there wont be any response from the FW for any target.

_______________________________________________
ath10k mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/ath10k

Hi Kalle,
I see that these patches are in "deferred" state. Is there any info/change needed in this patch ?

Thanks,
Rakesh Pillai.

_______________________________________________
ath10k mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/ath10k