[PATCH 1/4] mt76: mt76x02: use napi polling for tx cleanup

[PATCH 1/4] mt76: mt76x02: use napi polling for tx cleanup

[Felix Fietkau]

This allows tx scheduling and tx cleanup to run concurrently

Signed-off-by: Felix Fietkau <nbd@nbd.name>
---
 drivers/net/wireless/mediatek/mt76/mt76x02.h  |  2 +-
 .../net/wireless/mediatek/mt76/mt76x02_mmio.c | 49 ++++++++++++++-----
 2 files changed, 37 insertions(+), 14 deletions(-)

diff --git a/drivers/net/wireless/mediatek/mt76/mt76x02.h b/drivers/net/wireless/mediatek/mt76/mt76x02.h
index dfd3a4f1a624..cd37f44580ba 100644
--- a/drivers/net/wireless/mediatek/mt76/mt76x02.h
+++ b/drivers/net/wireless/mediatek/mt76/mt76x02.h
@@ -89,7 +89,7 @@ struct mt76x02_dev {
 
 	struct sk_buff *rx_head;
 
-	struct tasklet_struct tx_tasklet;
+	struct napi_struct tx_napi;
 	struct tasklet_struct pre_tbtt_tasklet;
 	struct delayed_work cal_work;
 	struct delayed_work wdt_work;
diff --git a/drivers/net/wireless/mediatek/mt76/mt76x02_mmio.c b/drivers/net/wireless/mediatek/mt76/mt76x02_mmio.c
index 644706ab2893..87e14af7a93b 100644
--- a/drivers/net/wireless/mediatek/mt76/mt76x02_mmio.c
+++ b/drivers/net/wireless/mediatek/mt76/mt76x02_mmio.c
@@ -154,18 +154,32 @@ static void mt76x02_process_tx_status_fifo(struct mt76x02_dev *dev)
 static void mt76x02_tx_tasklet(unsigned long data)
 {
 	struct mt76x02_dev *dev = (struct mt76x02_dev *)data;
-	int i;
 
+	mt76x02_mac_poll_tx_status(dev, false);
 	mt76x02_process_tx_status_fifo(dev);
 
+	mt76_txq_schedule_all(&dev->mt76);
+}
+
+int mt76x02_poll_tx(struct napi_struct *napi, int budget)
+{
+	struct mt76x02_dev *dev = container_of(napi, struct mt76x02_dev, tx_napi);
+	int i;
+
+	mt76x02_mac_poll_tx_status(dev, false);
+
 	for (i = MT_TXQ_MCU; i >= 0; i--)
 		mt76_queue_tx_cleanup(dev, i, false);
 
-	mt76x02_mac_poll_tx_status(dev, false);
+	if (napi_complete_done(napi, 0))
+		mt76x02_irq_enable(dev, MT_INT_TX_DONE_ALL);
 
-	mt76_txq_schedule_all(&dev->mt76);
+	for (i = MT_TXQ_MCU; i >= 0; i--)
+		mt76_queue_tx_cleanup(dev, i, false);
 
-	mt76x02_irq_enable(dev, MT_INT_TX_DONE_ALL);
+	tasklet_schedule(&dev->mt76.tx_tasklet);
+
+	return 0;
 }
 
 int mt76x02_dma_init(struct mt76x02_dev *dev)
@@ -223,7 +237,15 @@ int mt76x02_dma_init(struct mt76x02_dev *dev)
 	if (ret)
 		return ret;
 
-	return mt76_init_queues(dev);
+	ret = mt76_init_queues(dev);
+	if (ret)
+		return ret;
+
+	netif_tx_napi_add(&dev->mt76.napi_dev, &dev->tx_napi, mt76x02_poll_tx,
+			  NAPI_POLL_WEIGHT);
+	napi_enable(&dev->tx_napi);
+
+	return 0;
 }
 EXPORT_SYMBOL_GPL(mt76x02_dma_init);
 
@@ -251,11 +273,6 @@ irqreturn_t mt76x02_irq_handler(int irq, void *dev_instance)
 
 	intr &= dev->mt76.mmio.irqmask;
 
-	if (intr & MT_INT_TX_DONE_ALL) {
-		mt76x02_irq_disable(dev, MT_INT_TX_DONE_ALL);
-		tasklet_schedule(&dev->mt76.tx_tasklet);
-	}
-
 	if (intr & MT_INT_RX_DONE(0)) {
 		mt76x02_irq_disable(dev, MT_INT_RX_DONE(0));
 		napi_schedule(&dev->mt76.napi[0]);
@@ -277,9 +294,12 @@ irqreturn_t mt76x02_irq_handler(int irq, void *dev_instance)
 			mt76_queue_kick(dev, dev->mt76.q_tx[MT_TXQ_PSD].q);
 	}
 
-	if (intr & MT_INT_TX_STAT) {
+	if (intr & MT_INT_TX_STAT)
 		mt76x02_mac_poll_tx_status(dev, true);
-		tasklet_schedule(&dev->mt76.tx_tasklet);
+
+	if (intr & (MT_INT_TX_STAT | MT_INT_TX_DONE_ALL)) {
+		mt76x02_irq_disable(dev, MT_INT_TX_DONE_ALL);
+		napi_schedule(&dev->tx_napi);
 	}
 
 	if (intr & MT_INT_GPTIMER) {
@@ -310,6 +330,7 @@ static void mt76x02_dma_enable(struct mt76x02_dev *dev)
 void mt76x02_dma_cleanup(struct mt76x02_dev *dev)
 {
 	tasklet_kill(&dev->mt76.tx_tasklet);
+	netif_napi_del(&dev->tx_napi);
 	mt76_dma_cleanup(&dev->mt76);
 }
 EXPORT_SYMBOL_GPL(mt76x02_dma_cleanup);
@@ -429,6 +450,7 @@ static void mt76x02_watchdog_reset(struct mt76x02_dev *dev)
 
 	tasklet_disable(&dev->pre_tbtt_tasklet);
 	tasklet_disable(&dev->mt76.tx_tasklet);
+	napi_disable(&dev->tx_napi);
 
 	for (i = 0; i < ARRAY_SIZE(dev->mt76.napi); i++)
 		napi_disable(&dev->mt76.napi[i]);
@@ -482,7 +504,8 @@ static void mt76x02_watchdog_reset(struct mt76x02_dev *dev)
 	clear_bit(MT76_RESET, &dev->mt76.state);
 
 	tasklet_enable(&dev->mt76.tx_tasklet);
-	tasklet_schedule(&dev->mt76.tx_tasklet);
+	napi_enable(&dev->tx_napi);
+	napi_schedule(&dev->tx_napi);
 
 	tasklet_enable(&dev->pre_tbtt_tasklet);
 
-- 
2.17.0

[PATCH 2/4] mt76: mt76x02: remove irqsave/restore in locking for tx status fifo

[Felix Fietkau]

Use a separate lock and spin_trylock to avoid disabling interrupts.
Should improve performance and latency

Signed-off-by: Felix Fietkau <nbd@nbd.name>
---
 drivers/net/wireless/mediatek/mt76/mt76x02.h      | 1 +
 drivers/net/wireless/mediatek/mt76/mt76x02_mac.c  | 7 ++++---
 drivers/net/wireless/mediatek/mt76/mt76x02_mmio.c | 1 +
 3 files changed, 6 insertions(+), 3 deletions(-)

diff --git a/drivers/net/wireless/mediatek/mt76/mt76x02.h b/drivers/net/wireless/mediatek/mt76/mt76x02.h
index cd37f44580ba..9f103c2506db 100644
--- a/drivers/net/wireless/mediatek/mt76/mt76x02.h
+++ b/drivers/net/wireless/mediatek/mt76/mt76x02.h
@@ -86,6 +86,7 @@ struct mt76x02_dev {
 
 	u8 txdone_seq;
 	DECLARE_KFIFO_PTR(txstatus_fifo, struct mt76x02_tx_status);
+	spinlock_t txstatus_fifo_lock;
 
 	struct sk_buff *rx_head;
 
diff --git a/drivers/net/wireless/mediatek/mt76/mt76x02_mac.c b/drivers/net/wireless/mediatek/mt76/mt76x02_mac.c
index 28851060aa0f..c1f041e1a279 100644
--- a/drivers/net/wireless/mediatek/mt76/mt76x02_mac.c
+++ b/drivers/net/wireless/mediatek/mt76/mt76x02_mac.c
@@ -739,7 +739,6 @@ int mt76x02_mac_process_rx(struct mt76x02_dev *dev, struct sk_buff *skb,
 void mt76x02_mac_poll_tx_status(struct mt76x02_dev *dev, bool irq)
 {
 	struct mt76x02_tx_status stat = {};
-	unsigned long flags;
 	u8 update = 1;
 	bool ret;
 
@@ -749,9 +748,11 @@ void mt76x02_mac_poll_tx_status(struct mt76x02_dev *dev, bool irq)
 	trace_mac_txstat_poll(dev);
 
 	while (!irq || !kfifo_is_full(&dev->txstatus_fifo)) {
-		spin_lock_irqsave(&dev->mt76.mmio.irq_lock, flags);
+		if (!spin_trylock(&dev->txstatus_fifo_lock))
+			break;
+
 		ret = mt76x02_mac_load_tx_status(dev, &stat);
-		spin_unlock_irqrestore(&dev->mt76.mmio.irq_lock, flags);
+		spin_unlock(&dev->txstatus_fifo_lock);
 
 		if (!ret)
 			break;
diff --git a/drivers/net/wireless/mediatek/mt76/mt76x02_mmio.c b/drivers/net/wireless/mediatek/mt76/mt76x02_mmio.c
index 87e14af7a93b..5bc1b901f897 100644
--- a/drivers/net/wireless/mediatek/mt76/mt76x02_mmio.c
+++ b/drivers/net/wireless/mediatek/mt76/mt76x02_mmio.c
@@ -201,6 +201,7 @@ int mt76x02_dma_init(struct mt76x02_dev *dev)
 	tasklet_init(&dev->pre_tbtt_tasklet, mt76x02_pre_tbtt_tasklet,
 		     (unsigned long)dev);
 
+	spin_lock_init(&dev->txstatus_fifo_lock);
 	kfifo_init(&dev->txstatus_fifo, status_fifo, fifo_size);
 
 	mt76_dma_attach(&dev->mt76);
-- 
2.17.0

[PATCH 3/4] mt76: mt7603: fix initialization of max rx length

[Felix Fietkau]

The previous version only accidentally disabled A-MSDU deaggregation by
using the wrong mask for rx length configuration, which left previous
length value in place.
Fix the length and initialize the register completely to keep A-MSDU
de-aggregation remaining disabled

Signed-off-by: Felix Fietkau <nbd@nbd.name>
---
 drivers/net/wireless/mediatek/mt76/mt7603/init.c | 3 ++-
 drivers/net/wireless/mediatek/mt76/mt7603/regs.h | 4 ++++
 2 files changed, 6 insertions(+), 1 deletion(-)

diff --git a/drivers/net/wireless/mediatek/mt76/mt7603/init.c b/drivers/net/wireless/mediatek/mt76/mt7603/init.c
index d394839f1bd8..0d347ac6dfa6 100644
--- a/drivers/net/wireless/mediatek/mt76/mt7603/init.c
+++ b/drivers/net/wireless/mediatek/mt76/mt7603/init.c
@@ -167,7 +167,8 @@ mt7603_mac_init(struct mt7603_dev *dev)
 		FIELD_PREP(MT_AGG_RETRY_CONTROL_BAR_LIMIT, 1) |
 		FIELD_PREP(MT_AGG_RETRY_CONTROL_RTS_LIMIT, 15));
 
-	mt76_rmw(dev, MT_DMA_DCR0, ~0xfffc, 4096);
+	mt76_wr(dev, MT_DMA_DCR0, MT_DMA_DCR0_RX_VEC_DROP |
+		FIELD_PREP(MT_DMA_DCR0_MAX_RX_LEN, 4096));
 
 	mt76_rmw(dev, MT_DMA_VCFR0, BIT(0), BIT(13));
 	mt76_rmw(dev, MT_DMA_TMCFR0, BIT(0) | BIT(1), BIT(13));
diff --git a/drivers/net/wireless/mediatek/mt76/mt7603/regs.h b/drivers/net/wireless/mediatek/mt76/mt7603/regs.h
index da6827ae6cee..9d257d5c309d 100644
--- a/drivers/net/wireless/mediatek/mt76/mt7603/regs.h
+++ b/drivers/net/wireless/mediatek/mt76/mt7603/regs.h
@@ -233,6 +233,10 @@
 #define MT_WF_DMA(ofs)			(MT_WF_DMA_BASE + (ofs))
 
 #define MT_DMA_DCR0			MT_WF_DMA(0x000)
+#define MT_DMA_DCR0_MAX_RX_LEN		GENMASK(15, 0)
+#define MT_DMA_DCR0_DAMSDU		BIT(16)
+#define MT_DMA_DCR0_RX_VEC_DROP		BIT(17)
+
 #define MT_DMA_DCR1			MT_WF_DMA(0x004)
 
 #define MT_DMA_FQCR0			MT_WF_DMA(0x008)
-- 
2.17.0

[PATCH 4/4] mt76: mt7615: use sizeof instead of sizeof_field

[Felix Fietkau]

It is simpler in this case

Signed-off-by: Felix Fietkau <nbd@nbd.name>
---
 drivers/net/wireless/mediatek/mt76/mt7615/mcu.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)

diff --git a/drivers/net/wireless/mediatek/mt76/mt7615/mcu.c b/drivers/net/wireless/mediatek/mt76/mt7615/mcu.c
index b09540654b09..ea67c6022fe6 100644
--- a/drivers/net/wireless/mediatek/mt76/mt7615/mcu.c
+++ b/drivers/net/wireless/mediatek/mt76/mt7615/mcu.c
@@ -88,8 +88,7 @@ static int __mt7615_mcu_msg_send(struct mt7615_dev *dev, struct sk_buff *skb,
 	      FIELD_PREP(MT_TXD1_PKT_FMT, pkt_fmt);
 	txd[1] = cpu_to_le32(val);
 
-	mcu_txd->len = cpu_to_le16(skb->len -
-				   sizeof_field(struct mt7615_mcu_txd, txd));
+	mcu_txd->len = cpu_to_le16(skb->len - sizeof(mcu_txd->txd));
 	mcu_txd->pq_id = cpu_to_le16(MCU_PQ_ID(MT_TX_PORT_IDX_MCU, q_idx));
 	mcu_txd->pkt_type = MCU_PKT_ID;
 	mcu_txd->seq = seq;
-- 
2.17.0