net: wwan: t7xx: Add NAPI support

Replace the work queue based RX flow with a NAPI implementation
Remove rx_thread and dpmaif_rxq_work.
Enable GRO on RX path.
Introduce dummy network device. its responsibility is
    - Binds one NAPI object for each DL HW queue and acts as
      the agent of all those network devices.
    - Use NAPI object to poll DL packets.
    - Helps to dispatch each packet to the network interface.

Signed-off-by: Haijun Liu <[email protected]>
Co-developed-by: Sreehari Kancharla <[email protected]>
Signed-off-by: Sreehari Kancharla <[email protected]>
Signed-off-by: Chandrashekar Devegowda <[email protected]>
Acked-by: Ricardo Martinez <[email protected]>
Acked-by: M Chetan Kumar <[email protected]>
Reviewed-by: Ilpo Järvinen <[email protected]>
Signed-off-by: David S. Miller <[email protected]>

Author: Haijun Liu

drivers/net/wwan/t7xx/t7xx_hif_dpmaif.h

@@ -20,6 +20,7 @@
 
 #include <linux/bitmap.h>
 #include <linux/mm_types.h>
+#include <linux/netdevice.h>
 #include <linux/sched.h>
 #include <linux/skbuff.h>
 #include <linux/spinlock.h>
@@ -109,20 +110,14 @@ struct dpmaif_rx_queue {
 	struct dpmaif_bat_request *bat_req;
 	struct dpmaif_bat_request *bat_frag;
 
-	wait_queue_head_t	rx_wq;
-	struct task_struct	*rx_thread;
-	struct sk_buff_head	skb_list;
-	unsigned int		skb_list_max_len;
-
-	struct workqueue_struct	*worker;
-	struct work_struct	dpmaif_rxq_work;
-
 	atomic_t		rx_processing;
 
 	struct dpmaif_ctrl	*dpmaif_ctrl;
 	unsigned int		expect_pit_seq;
 	unsigned int		pit_remain_release_cnt;
 	struct dpmaif_cur_rx_skb_info rx_data_info;
+	struct napi_struct	napi;
+	bool			sleep_lock_pending;
 };
 
 struct dpmaif_tx_queue {
@@ -168,7 +163,8 @@ enum dpmaif_txq_state {
 struct dpmaif_callbacks {
 	void (*state_notify)(struct t7xx_pci_dev *t7xx_dev,
 			     enum dpmaif_txq_state state, int txq_number);
-	void (*recv_skb)(struct t7xx_pci_dev *t7xx_dev, struct sk_buff *skb);
+	void (*recv_skb)(struct t7xx_ccmni_ctrl *ccmni_ctlb, struct sk_buff *skb,
+			 struct napi_struct *napi);
 };
 
 struct dpmaif_ctrl {

drivers/net/wwan/t7xx/t7xx_hif_dpmaif_rx.c

@@ -45,6 +45,7 @@
 #include "t7xx_dpmaif.h"
 #include "t7xx_hif_dpmaif.h"
 #include "t7xx_hif_dpmaif_rx.h"
+#include "t7xx_netdev.h"
 #include "t7xx_pci.h"
 
 #define DPMAIF_BAT_COUNT		8192
@@ -76,43 +77,6 @@ static unsigned int t7xx_normal_pit_bid(const struct dpmaif_pit *pit_info)
 	return value;
 }
 
-static int t7xx_dpmaif_net_rx_push_thread(void *arg)
-{
-	struct dpmaif_rx_queue *q = arg;
-	struct dpmaif_ctrl *hif_ctrl;
-	struct dpmaif_callbacks *cb;
-
-	hif_ctrl = q->dpmaif_ctrl;
-	cb = hif_ctrl->callbacks;
-
-	while (!kthread_should_stop()) {
-		struct sk_buff *skb;
-		unsigned long flags;
-
-		if (skb_queue_empty(&q->skb_list)) {
-			if (wait_event_interruptible(q->rx_wq,
-						     !skb_queue_empty(&q->skb_list) ||
-						     kthread_should_stop()))
-				continue;
-
-			if (kthread_should_stop())
-				break;
-		}
-
-		spin_lock_irqsave(&q->skb_list.lock, flags);
-		skb = __skb_dequeue(&q->skb_list);
-		spin_unlock_irqrestore(&q->skb_list.lock, flags);
-
-		if (!skb)
-			continue;
-
-		cb->recv_skb(hif_ctrl->t7xx_dev, skb);
-		cond_resched();
-	}
-
-	return 0;
-}
-
 static int t7xx_dpmaif_update_bat_wr_idx(struct dpmaif_ctrl *dpmaif_ctrl,
 					 const unsigned int q_num, const unsigned int bat_cnt)
 {
@@ -726,21 +690,10 @@ static int t7xx_dpmaifq_rx_notify_hw(struct dpmaif_rx_queue *rxq)
 	return ret;
 }
 
-static void t7xx_dpmaif_rx_skb_enqueue(struct dpmaif_rx_queue *rxq, struct sk_buff *skb)
-{
-	unsigned long flags;
-
-	spin_lock_irqsave(&rxq->skb_list.lock, flags);
-	if (rxq->skb_list.qlen < rxq->skb_list_max_len)
-		__skb_queue_tail(&rxq->skb_list, skb);
-	else
-		dev_kfree_skb_any(skb);
-	spin_unlock_irqrestore(&rxq->skb_list.lock, flags);
-}
-
 static void t7xx_dpmaif_rx_skb(struct dpmaif_rx_queue *rxq,
 			       struct dpmaif_cur_rx_skb_info *skb_info)
 {
+	struct dpmaif_ctrl *dpmaif_ctrl = rxq->dpmaif_ctrl;
 	struct sk_buff *skb = skb_info->cur_skb;
 	struct t7xx_skb_cb *skb_cb;
 	u8 netif_id;
@@ -758,11 +711,11 @@ static void t7xx_dpmaif_rx_skb(struct dpmaif_rx_queue *rxq,
 	skb_cb = T7XX_SKB_CB(skb);
 	skb_cb->netif_idx = netif_id;
 	skb_cb->rx_pkt_type = skb_info->pkt_type;
-	t7xx_dpmaif_rx_skb_enqueue(rxq, skb);
+	dpmaif_ctrl->callbacks->recv_skb(dpmaif_ctrl->t7xx_dev->ccmni_ctlb, skb, &rxq->napi);
 }
 
 static int t7xx_dpmaif_rx_start(struct dpmaif_rx_queue *rxq, const unsigned int pit_cnt,
-				const unsigned long timeout)
+				const unsigned int budget, int *once_more)
 {
 	unsigned int cur_pit, pit_len, rx_cnt, recv_skb_cnt = 0;
 	struct device *dev = rxq->dpmaif_ctrl->dev;
@@ -777,13 +730,14 @@ static int t7xx_dpmaif_rx_start(struct dpmaif_rx_queue *rxq, const unsigned int
 		struct dpmaif_pit *pkt_info;
 		u32 val;
 
-		if (!skb_info->msg_pit_received && time_after_eq(jiffies, timeout))
+		if (!skb_info->msg_pit_received && recv_skb_cnt >= budget)
 			break;
 
 		pkt_info = (struct dpmaif_pit *)rxq->pit_base + cur_pit;
 		if (t7xx_dpmaif_check_pit_seq(rxq, pkt_info)) {
 			dev_err_ratelimited(dev, "RXQ%u checks PIT SEQ fail\n", rxq->index);
-			return -EAGAIN;
+			*once_more = 1;
+			return recv_skb_cnt;
 		}
 
 		val = FIELD_GET(PD_PIT_PACKET_TYPE, le32_to_cpu(pkt_info->header));
@@ -817,12 +771,7 @@ static int t7xx_dpmaif_rx_start(struct dpmaif_rx_queue *rxq, const unsigned int
 				}
 
 				memset(skb_info, 0, sizeof(*skb_info));
-
 				recv_skb_cnt++;
-				if (!(recv_skb_cnt & DPMAIF_RX_PUSH_THRESHOLD_MASK)) {
-					wake_up_all(&rxq->rx_wq);
-					recv_skb_cnt = 0;
-				}
 			}
 		}
 
@@ -837,16 +786,13 @@ static int t7xx_dpmaif_rx_start(struct dpmaif_rx_queue *rxq, const unsigned int
 		}
 	}
 
-	if (recv_skb_cnt)
-		wake_up_all(&rxq->rx_wq);
-
 	if (!ret)
 		ret = t7xx_dpmaifq_rx_notify_hw(rxq);
 
 	if (ret)
 		return ret;
 
-	return rx_cnt;
+	return recv_skb_cnt;
 }
 
 static unsigned int t7xx_dpmaifq_poll_pit(struct dpmaif_rx_queue *rxq)
@@ -863,76 +809,83 @@ static unsigned int t7xx_dpmaifq_poll_pit(struct dpmaif_rx_queue *rxq)
 	return pit_cnt;
 }
 
-static int t7xx_dpmaif_rx_data_collect(struct dpmaif_ctrl *dpmaif_ctrl,
-				       const unsigned int q_num, const unsigned int budget)
+static int t7xx_dpmaif_napi_rx_data_collect(struct dpmaif_ctrl *dpmaif_ctrl,
+					    const unsigned int q_num,
+					    const unsigned int budget, int *once_more)
 {
 	struct dpmaif_rx_queue *rxq = &dpmaif_ctrl->rxq[q_num];
-	unsigned long time_limit;
 	unsigned int cnt;
+	int ret = 0;
 
-	time_limit = jiffies + msecs_to_jiffies(DPMAIF_WQ_TIME_LIMIT_MS);
-
-	while ((cnt = t7xx_dpmaifq_poll_pit(rxq))) {
-		unsigned int rd_cnt;
-		int real_cnt;
-
-		rd_cnt = min(cnt, budget);
-
-		real_cnt = t7xx_dpmaif_rx_start(rxq, rd_cnt, time_limit);
-		if (real_cnt < 0)
-			return real_cnt;
-
-		if (real_cnt < cnt)
-			return -EAGAIN;
-	}
-
-	return 0;
-}
+	cnt = t7xx_dpmaifq_poll_pit(rxq);
+	if (!cnt)
+		return ret;
 
-static void t7xx_dpmaif_do_rx(struct dpmaif_ctrl *dpmaif_ctrl, struct dpmaif_rx_queue *rxq)
-{
-	struct dpmaif_hw_info *hw_info = &dpmaif_ctrl->hw_info;
-	int ret;
+	ret = t7xx_dpmaif_rx_start(rxq, cnt, budget, once_more);
+	if (ret < 0)
+		dev_err(dpmaif_ctrl->dev, "dlq%u rx ERR:%d\n", rxq->index, ret);
 
-	ret = t7xx_dpmaif_rx_data_collect(dpmaif_ctrl, rxq->index, rxq->budget);
-	if (ret < 0) {
-		/* Try one more time */
-		queue_work(rxq->worker, &rxq->dpmaif_rxq_work);
-		t7xx_dpmaif_clr_ip_busy_sts(hw_info);
-	} else {
-		t7xx_dpmaif_clr_ip_busy_sts(hw_info);
-		t7xx_dpmaif_dlq_unmask_rx_done(hw_info, rxq->index);
-	}
+	return ret;
 }
 
-static void t7xx_dpmaif_rxq_work(struct work_struct *work)
+int t7xx_dpmaif_napi_rx_poll(struct napi_struct *napi, const int budget)
 {
-	struct dpmaif_rx_queue *rxq = container_of(work, struct dpmaif_rx_queue, dpmaif_rxq_work);
-	struct dpmaif_ctrl *dpmaif_ctrl = rxq->dpmaif_ctrl;
-	int ret;
+	struct dpmaif_rx_queue *rxq = container_of(napi, struct dpmaif_rx_queue, napi);
+	struct t7xx_pci_dev *t7xx_dev = rxq->dpmaif_ctrl->t7xx_dev;
+	int ret, once_more = 0, work_done = 0;
 
 	atomic_set(&rxq->rx_processing, 1);
 	/* Ensure rx_processing is changed to 1 before actually begin RX flow */
 	smp_mb();
 
 	if (!rxq->que_started) {
 		atomic_set(&rxq->rx_processing, 0);
-		dev_err(dpmaif_ctrl->dev, "Work RXQ: %d has not been started\n", rxq->index);
-		return;
+		dev_err(rxq->dpmaif_ctrl->dev, "Work RXQ: %d has not been started\n", rxq->index);
+		return work_done;
 	}
 
-	ret = pm_runtime_resume_and_get(dpmaif_ctrl->dev);
-	if (ret < 0 && ret != -EACCES)
-		return;
+	if (!rxq->sleep_lock_pending) {
+		pm_runtime_get_noresume(rxq->dpmaif_ctrl->dev);
+		t7xx_pci_disable_sleep(t7xx_dev);
+	}
 
-	t7xx_pci_disable_sleep(dpmaif_ctrl->t7xx_dev);
-	if (t7xx_pci_sleep_disable_complete(dpmaif_ctrl->t7xx_dev))
-		t7xx_dpmaif_do_rx(dpmaif_ctrl, rxq);
+	ret = try_wait_for_completion(&t7xx_dev->sleep_lock_acquire);
+	if (!ret) {
+		napi_complete_done(napi, work_done);
+		rxq->sleep_lock_pending = true;
+		napi_reschedule(napi);
+		return work_done;
+	}
 
-	t7xx_pci_enable_sleep(dpmaif_ctrl->t7xx_dev);
-	pm_runtime_mark_last_busy(dpmaif_ctrl->dev);
-	pm_runtime_put_autosuspend(dpmaif_ctrl->dev);
+	rxq->sleep_lock_pending = false;
+	while (work_done < budget) {
+		int each_budget = budget - work_done;
+		int rx_cnt = t7xx_dpmaif_napi_rx_data_collect(rxq->dpmaif_ctrl, rxq->index,
+							      each_budget, &once_more);
+		if (rx_cnt > 0)
+			work_done += rx_cnt;
+		else
+			break;
+	}
+
+	if (once_more) {
+		napi_gro_flush(napi, false);
+		work_done = budget;
+		t7xx_dpmaif_clr_ip_busy_sts(&rxq->dpmaif_ctrl->hw_info);
+	} else if (work_done < budget) {
+		napi_complete_done(napi, work_done);
+		t7xx_dpmaif_clr_ip_busy_sts(&rxq->dpmaif_ctrl->hw_info);
+		t7xx_dpmaif_dlq_unmask_rx_done(&rxq->dpmaif_ctrl->hw_info, rxq->index);
+	} else {
+		t7xx_dpmaif_clr_ip_busy_sts(&rxq->dpmaif_ctrl->hw_info);
+	}
+
+	t7xx_pci_enable_sleep(rxq->dpmaif_ctrl->t7xx_dev);
+	pm_runtime_mark_last_busy(rxq->dpmaif_ctrl->dev);
+	pm_runtime_put_noidle(rxq->dpmaif_ctrl->dev);
 	atomic_set(&rxq->rx_processing, 0);
+
+	return work_done;
 }
 
 void t7xx_dpmaif_irq_rx_done(struct dpmaif_ctrl *dpmaif_ctrl, const unsigned int que_mask)
@@ -947,7 +900,7 @@ void t7xx_dpmaif_irq_rx_done(struct dpmaif_ctrl *dpmaif_ctrl, const unsigned int
 	}
 
 	rxq = &dpmaif_ctrl->rxq[qno];
-	queue_work(rxq->worker, &rxq->dpmaif_rxq_work);
+	napi_schedule(&rxq->napi);
 }
 
 static void t7xx_dpmaif_base_free(const struct dpmaif_ctrl *dpmaif_ctrl,
@@ -1082,50 +1035,14 @@ int t7xx_dpmaif_rxq_init(struct dpmaif_rx_queue *queue)
 	int ret;
 
 	ret = t7xx_dpmaif_rx_alloc(queue);
-	if (ret < 0) {
+	if (ret < 0)
 		dev_err(queue->dpmaif_ctrl->dev, "Failed to allocate RX buffers: %d\n", ret);
-		return ret;
-	}
-
-	INIT_WORK(&queue->dpmaif_rxq_work, t7xx_dpmaif_rxq_work);
-
-	queue->worker = alloc_workqueue("dpmaif_rx%d_worker",
-					WQ_UNBOUND | WQ_MEM_RECLAIM | WQ_HIGHPRI, 1, queue->index);
-	if (!queue->worker) {
-		ret = -ENOMEM;
-		goto err_free_rx_buffer;
-	}
-
-	init_waitqueue_head(&queue->rx_wq);
-	skb_queue_head_init(&queue->skb_list);
-	queue->skb_list_max_len = queue->bat_req->pkt_buf_sz;
-	queue->rx_thread = kthread_run(t7xx_dpmaif_net_rx_push_thread,
-				       queue, "dpmaif_rx%d_push", queue->index);
-
-	ret = PTR_ERR_OR_ZERO(queue->rx_thread);
-	if (ret)
-		goto err_free_workqueue;
-
-	return 0;
-
-err_free_workqueue:
-	destroy_workqueue(queue->worker);
-
-err_free_rx_buffer:
-	t7xx_dpmaif_rx_buf_free(queue);
 
 	return ret;
 }
 
 void t7xx_dpmaif_rxq_free(struct dpmaif_rx_queue *queue)
 {
-	if (queue->worker)
-		destroy_workqueue(queue->worker);
-
-	if (queue->rx_thread)
-		kthread_stop(queue->rx_thread);
-
-	skb_queue_purge(&queue->skb_list);
 	t7xx_dpmaif_rx_buf_free(queue);
 }
 
@@ -1188,8 +1105,6 @@ void t7xx_dpmaif_rx_stop(struct dpmaif_ctrl *dpmaif_ctrl)
 		struct dpmaif_rx_queue *rxq = &dpmaif_ctrl->rxq[i];
 		int timeout, value;
 
-		flush_work(&rxq->dpmaif_rxq_work);
-
 		timeout = readx_poll_timeout_atomic(atomic_read, &rxq->rx_processing, value,
 						    !value, 0, DPMAIF_CHECK_INIT_TIMEOUT_US);
 		if (timeout)
@@ -1205,7 +1120,6 @@ static void t7xx_dpmaif_stop_rxq(struct dpmaif_rx_queue *rxq)
 {
 	int cnt, j = 0;
 
-	flush_work(&rxq->dpmaif_rxq_work);
 	rxq->que_started = false;
 
 	do {

drivers/net/wwan/t7xx/t7xx_hif_dpmaif_rx.h

@@ -112,5 +112,6 @@ int t7xx_dpmaif_bat_alloc(const struct dpmaif_ctrl *dpmaif_ctrl, struct dpmaif_b
 			  const enum bat_type buf_type);
 void t7xx_dpmaif_bat_free(const struct dpmaif_ctrl *dpmaif_ctrl,
 			  struct dpmaif_bat_request *bat_req);
+int t7xx_dpmaif_napi_rx_poll(struct napi_struct *napi, const int budget);
 
 #endif /* __T7XX_HIF_DPMA_RX_H__ */