diff mbox series

[iwl-net,v1] igc: Fix TX Hang issue when QBV Gate is close

Message ID 20230531075907.26251-1-muhammad.husaini.zulkifli@intel.com
State Changes Requested
Headers show
Series [iwl-net,v1] igc: Fix TX Hang issue when QBV Gate is close | expand

Commit Message

Zulkifli, Muhammad Husaini May 31, 2023, 7:59 a.m. UTC
If a user schedules a Gate Control List (GCL) to close one of
the QBV gates while also transmitting a packet to that closed gate,
TX Hang will be happen. HW would not drop any packet when the gate
is close and keep queueing up in HW TX FIFO until the gate is re-open.
This patch implement the solution to drop the packet for the closed
gate.

This patch will additionally include a reset adapter to perform
SW initialization for each 1st Gate Control List (GCL) to avoid hang.
This is due to the HW design, where changing to TSN transmit mode
requires SW initialization. Intel Discrete I225/6 transmit mode
cannot be changed when in dynamic mode according to Software User
Manual Section 7.5.2.1. Subsequent Gate Control List (GCL) operations
will proceed without a reset, as they already in TSN Mode.

Step to reproduce:

DUT:
1) Configure GCL List with certain gate close.
2) Transmit the packet to close gate.

Fixes: ec50a9d437f0 ("igc: Add support for taprio offloading")
Co-developed-by: Tan Tee Min <tee.min.tan@linux.intel.com>
Signed-off-by: Tan Tee Min <tee.min.tan@linux.intel.com>
Tested-by: Chwee Lin Choong <chwee.lin.choong@intel.com>
Signed-off-by: Muhammad Husaini Zulkifli <muhammad.husaini.zulkifli@intel.com>
---
 drivers/net/ethernet/intel/igc/igc.h      |  6 +++
 drivers/net/ethernet/intel/igc/igc_main.c | 57 +++++++++++++++++++++--
 drivers/net/ethernet/intel/igc/igc_tsn.c  | 41 ++++++++++------
 3 files changed, 87 insertions(+), 17 deletions(-)

Comments

Tony Nguyen June 1, 2023, 8:37 p.m. UTC | #1
On 5/31/2023 12:59 AM, Muhammad Husaini Zulkifli wrote:
> If a user schedules a Gate Control List (GCL) to close one of
> the QBV gates while also transmitting a packet to that closed gate,
> TX Hang will be happen. HW would not drop any packet when the gate
> is close and keep queueing up in HW TX FIFO until the gate is re-open.
> This patch implement the solution to drop the packet for the closed
> gate.
> 
> This patch will additionally include a reset adapter to perform
> SW initialization for each 1st Gate Control List (GCL) to avoid hang.
> This is due to the HW design, where changing to TSN transmit mode
> requires SW initialization. Intel Discrete I225/6 transmit mode
> cannot be changed when in dynamic mode according to Software User
> Manual Section 7.5.2.1. Subsequent Gate Control List (GCL) operations
> will proceed without a reset, as they already in TSN Mode.
> 
> Step to reproduce:
> 
> DUT:
> 1) Configure GCL List with certain gate close.
> 2) Transmit the packet to close gate.

This patch doesn't apply.

> Fixes: ec50a9d437f0 ("igc: Add support for taprio offloading")
> Co-developed-by: Tan Tee Min <tee.min.tan@linux.intel.com>
> Signed-off-by: Tan Tee Min <tee.min.tan@linux.intel.com>
> Tested-by: Chwee Lin Choong <chwee.lin.choong@intel.com>
> Signed-off-by: Muhammad Husaini Zulkifli <muhammad.husaini.zulkifli@intel.com>
diff mbox series

Patch

diff --git a/drivers/net/ethernet/intel/igc/igc.h b/drivers/net/ethernet/intel/igc/igc.h
index cb5751fab03c9..866b41a92cf8a 100644
--- a/drivers/net/ethernet/intel/igc/igc.h
+++ b/drivers/net/ethernet/intel/igc/igc.h
@@ -14,6 +14,7 @@ 
 #include <linux/timecounter.h>
 #include <linux/net_tstamp.h>
 #include <linux/bitfield.h>
+#include <linux/hrtimer.h>
 
 #include "igc_hw.h"
 
@@ -101,6 +102,8 @@  struct igc_ring {
 	u32 start_time;
 	u32 end_time;
 	u32 max_sdu;
+	bool oper_gate_closed;
+	bool admin_gate_closed;
 
 	/* CBS parameters */
 	bool cbs_enable;                /* indicates if CBS is enabled */
@@ -160,6 +163,7 @@  struct igc_adapter {
 	struct timer_list watchdog_timer;
 	struct timer_list dma_err_timer;
 	struct timer_list phy_info_timer;
+	struct hrtimer hrtimer;
 
 	u32 wol;
 	u32 en_mng_pt;
@@ -189,6 +193,8 @@  struct igc_adapter {
 	ktime_t cycle_time;
 	bool qbv_enable;
 	u32 qbv_config_change_errors;
+	bool qbv_transition;
+	int qbv_count;
 
 	/* OS defined structs */
 	struct pci_dev *pdev;
diff --git a/drivers/net/ethernet/intel/igc/igc_main.c b/drivers/net/ethernet/intel/igc/igc_main.c
index 774c80073579e..8ab6ffc2f1005 100644
--- a/drivers/net/ethernet/intel/igc/igc_main.c
+++ b/drivers/net/ethernet/intel/igc/igc_main.c
@@ -1521,6 +1521,9 @@  static netdev_tx_t igc_xmit_frame_ring(struct sk_buff *skb,
 	u8 hdr_len = 0;
 	int tso = 0;
 
+	if (adapter->qbv_transition || tx_ring->oper_gate_closed)
+		goto out_drop;
+
 	/* need: 1 descriptor per page * PAGE_SIZE/IGC_MAX_DATA_PER_TXD,
 	 *	+ 1 desc for skb_headlen/IGC_MAX_DATA_PER_TXD,
 	 *	+ 2 desc gap to keep tail from touching head,
@@ -3012,7 +3015,8 @@  static bool igc_clean_tx_irq(struct igc_q_vector *q_vector, int napi_budget)
 		    (adapter->tx_timeout_factor * HZ)) &&
 		    !(rd32(IGC_STATUS) & IGC_STATUS_TXOFF) &&
 		    (rd32(IGC_TDH(tx_ring->reg_idx)) !=
-		     readl(tx_ring->tail))) {
+		    readl(tx_ring->tail)) &&
+		    !tx_ring->oper_gate_closed) {
 			/* detected Tx unit hang */
 			netdev_err(tx_ring->netdev,
 				   "Detected Tx Unit Hang\n"
@@ -6102,6 +6106,8 @@  static int igc_tsn_clear_schedule(struct igc_adapter *adapter)
 	adapter->base_time = 0;
 	adapter->cycle_time = NSEC_PER_SEC;
 	adapter->qbv_config_change_errors = 0;
+	adapter->qbv_transition = false;
+	adapter->qbv_count = 0;
 
 	for (i = 0; i < adapter->num_tx_queues; i++) {
 		struct igc_ring *ring = adapter->tx_ring[i];
@@ -6109,6 +6115,8 @@  static int igc_tsn_clear_schedule(struct igc_adapter *adapter)
 		ring->start_time = 0;
 		ring->end_time = NSEC_PER_SEC;
 		ring->max_sdu = 0;
+		ring->oper_gate_closed = false;
+		ring->admin_gate_closed = false;
 	}
 
 	return 0;
@@ -6120,6 +6128,7 @@  static int igc_save_qbv_schedule(struct igc_adapter *adapter,
 	bool queue_configured[IGC_MAX_TX_QUEUES] = { };
 	struct igc_hw *hw = &adapter->hw;
 	u32 start_time = 0, end_time = 0;
+	struct timespec64 now;
 	size_t n;
 	int i;
 
@@ -6140,6 +6149,8 @@  static int igc_save_qbv_schedule(struct igc_adapter *adapter,
 	adapter->cycle_time = qopt->cycle_time;
 	adapter->base_time = qopt->base_time;
 
+	igc_ptp_read(adapter, &now);
+
 	for (n = 0; n < qopt->num_entries; n++) {
 		struct tc_taprio_sched_entry *e = &qopt->entries[n];
 
@@ -6174,7 +6185,10 @@  static int igc_save_qbv_schedule(struct igc_adapter *adapter,
 				ring->start_time = start_time;
 			ring->end_time = end_time;
 
-			queue_configured[i] = true;
+			if (ring->start_time >= adapter->cycle_time)
+				queue_configured[i] = false;
+			else
+				queue_configured[i] = true;
 		}
 
 		start_time += e->interval;
@@ -6184,8 +6198,20 @@  static int igc_save_qbv_schedule(struct igc_adapter *adapter,
 	 * If not, set the start and end time to be end time.
 	 */
 	for (i = 0; i < adapter->num_tx_queues; i++) {
+		struct igc_ring *ring = adapter->tx_ring[i];
+
+		if (!is_base_time_past(qopt->base_time, &now)) {
+			ring->admin_gate_closed = false;
+		} else {
+			ring->oper_gate_closed = false;
+			ring->admin_gate_closed = false;
+		}
+
 		if (!queue_configured[i]) {
-			struct igc_ring *ring = adapter->tx_ring[i];
+			if (!is_base_time_past(qopt->base_time, &now))
+				ring->admin_gate_closed = true;
+			else
+				ring->oper_gate_closed = true;
 
 			ring->start_time = end_time;
 			ring->end_time = end_time;
@@ -6566,6 +6592,27 @@  static const struct xdp_metadata_ops igc_xdp_metadata_ops = {
 	.xmo_rx_timestamp		= igc_xdp_rx_timestamp,
 };
 
+static enum hrtimer_restart igc_qbv_scheduling_timer(struct hrtimer *timer)
+{
+	struct igc_adapter *adapter = container_of(timer, struct igc_adapter,
+						   hrtimer);
+	int i;
+
+	adapter->qbv_transition = true;
+	for (i = 0; i < adapter->num_tx_queues; i++) {
+		struct igc_ring *tx_ring = adapter->tx_ring[i];
+
+		if (tx_ring->admin_gate_closed) {
+			tx_ring->admin_gate_closed = false;
+			tx_ring->oper_gate_closed = true;
+		} else {
+			tx_ring->oper_gate_closed = false;
+		}
+	}
+	adapter->qbv_transition = false;
+	return HRTIMER_NORESTART;
+}
+
 /**
  * igc_probe - Device Initialization Routine
  * @pdev: PCI device information struct
@@ -6744,6 +6791,9 @@  static int igc_probe(struct pci_dev *pdev,
 	INIT_WORK(&adapter->reset_task, igc_reset_task);
 	INIT_WORK(&adapter->watchdog_task, igc_watchdog_task);
 
+	hrtimer_init(&adapter->hrtimer, CLOCK_MONOTONIC, HRTIMER_MODE_REL);
+	adapter->hrtimer.function = &igc_qbv_scheduling_timer;
+
 	/* Initialize link properties that are user-changeable */
 	adapter->fc_autoneg = true;
 	hw->mac.autoneg = true;
@@ -6847,6 +6897,7 @@  static void igc_remove(struct pci_dev *pdev)
 
 	cancel_work_sync(&adapter->reset_task);
 	cancel_work_sync(&adapter->watchdog_task);
+	hrtimer_cancel(&adapter->hrtimer);
 
 	/* Release control of h/w to f/w.  If f/w is AMT enabled, this
 	 * would have already happened in close and is redundant.
diff --git a/drivers/net/ethernet/intel/igc/igc_tsn.c b/drivers/net/ethernet/intel/igc/igc_tsn.c
index 6b299b83e7ef2..81770955029dc 100644
--- a/drivers/net/ethernet/intel/igc/igc_tsn.c
+++ b/drivers/net/ethernet/intel/igc/igc_tsn.c
@@ -114,7 +114,6 @@  static int igc_tsn_disable_offload(struct igc_adapter *adapter)
 static int igc_tsn_enable_offload(struct igc_adapter *adapter)
 {
 	struct igc_hw *hw = &adapter->hw;
-	bool tsn_mode_reconfig = false;
 	u32 tqavctrl, baset_l, baset_h;
 	u32 sec, nsec, cycle;
 	ktime_t base_time, systim;
@@ -228,11 +227,10 @@  static int igc_tsn_enable_offload(struct igc_adapter *adapter)
 
 	tqavctrl = rd32(IGC_TQAVCTRL) & ~IGC_TQAVCTRL_FUTSCDDIS;
 
-	if (tqavctrl & IGC_TQAVCTRL_TRANSMIT_MODE_TSN)
-		tsn_mode_reconfig = true;
-
 	tqavctrl |= IGC_TQAVCTRL_TRANSMIT_MODE_TSN | IGC_TQAVCTRL_ENHANCED_QAV;
 
+	adapter->qbv_count++;
+
 	cycle = adapter->cycle_time;
 	base_time = adapter->base_time;
 
@@ -250,17 +248,28 @@  static int igc_tsn_enable_offload(struct igc_adapter *adapter)
 		 */
 		if ((rd32(IGC_BASET_H) || rd32(IGC_BASET_L)) &&
 		    (adapter->tc_setup_type == TC_SETUP_QDISC_TAPRIO) &&
-		    tsn_mode_reconfig)
+		    (adapter->qbv_count > 1))
 			adapter->qbv_config_change_errors++;
 	} else {
-		/* According to datasheet section 7.5.2.9.3.3, FutScdDis bit
-		 * has to be configured before the cycle time and base time.
-		 * Tx won't hang if there is a GCL is already running,
-		 * so in this case we don't need to set FutScdDis.
-		 */
-		if (igc_is_device_id_i226(hw) &&
-		    !(rd32(IGC_BASET_H) || rd32(IGC_BASET_L)))
-			tqavctrl |= IGC_TQAVCTRL_FUTSCDDIS;
+		if (igc_is_device_id_i226(hw)) {
+			ktime_t adjust_time, expires_time;
+
+		       /* According to datasheet section 7.5.2.9.3.3, FutScdDis bit
+			* has to be configured before the cycle time and base time.
+			* Tx won't hang if there is a GCL is already running,
+			* so in this case we don't need to set FutScdDis.
+			*/
+			if (!(rd32(IGC_BASET_H) || rd32(IGC_BASET_L)))
+				tqavctrl |= IGC_TQAVCTRL_FUTSCDDIS;
+
+			nsec = rd32(IGC_SYSTIML);
+			sec = rd32(IGC_SYSTIMH);
+			systim = ktime_set(sec, nsec);
+
+			adjust_time = adapter->base_time;
+			expires_time = ktime_sub_ns(adjust_time, systim);
+			hrtimer_start(&adapter->hrtimer, expires_time, HRTIMER_MODE_REL);
+		}
 	}
 
 	wr32(IGC_TQAVCTRL, tqavctrl);
@@ -306,7 +315,11 @@  int igc_tsn_offload_apply(struct igc_adapter *adapter)
 {
 	struct igc_hw *hw = &adapter->hw;
 
-	if (netif_running(adapter->netdev) && igc_is_device_id_i225(hw)) {
+	/* Per I225/6 HW Design Section 7.5.2.1, transmit mode
+	 * cannot be change dynamically. Require reset the adapter.
+	 */
+	if (netif_running(adapter->netdev) &&
+	    (igc_is_device_id_i225(hw) || !adapter->qbv_count)) {
 		schedule_work(&adapter->reset_task);
 		return 0;
 	}