net: lan966x: Add support for XDP_TX

Extend lan966x XDP support with the action XDP_TX. In this case when the
received buffer needs to execute XDP_TX, the buffer will be moved to the
TX buffers. So a new RX buffer will be allocated.
When the TX finish with the frame, it would give back the buffer to the
page pool.
Signed-off-by: Horatiu Vultur <horatiu.vultur@microchip.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

drivers/net/ethernet/microchip/lan966x/lan966x_fdma.c

@@ -410,12 +410,17 @@ static void lan966x_fdma_tx_clear_buf(struct lan966x *lan966x, int weight)
 		dcb_buf->dev->stats.tx_bytes += dcb_buf->len;
 
 		dcb_buf->used = false;
+		if (dcb_buf->use_skb) {
 			dma_unmap_single(lan966x->dev,
 					 dcb_buf->dma_addr,
 					 dcb_buf->len,
 					 DMA_TO_DEVICE);
+
 			if (!dcb_buf->ptp)
-			dev_kfree_skb_any(dcb_buf->skb);
+				napi_consume_skb(dcb_buf->data.skb, weight);
+		} else {
+			xdp_return_frame_rx_napi(dcb_buf->data.xdpf);
+		}
 
 		clear = true;
 	}
@@ -548,6 +553,9 @@ static int lan966x_fdma_napi_poll(struct napi_struct *napi, int weight)
 			lan966x_fdma_rx_free_page(rx);
 			lan966x_fdma_rx_advance_dcb(rx);
 			goto allocate_new;
+		case FDMA_TX:
+			lan966x_fdma_rx_advance_dcb(rx);
+			continue;
 		case FDMA_DROP:
 			lan966x_fdma_rx_free_page(rx);
 			lan966x_fdma_rx_advance_dcb(rx);
@@ -669,6 +677,62 @@ static void lan966x_fdma_tx_start(struct lan966x_tx *tx, int next_to_use)
 	tx->last_in_use = next_to_use;
 }
 
+int lan966x_fdma_xmit_xdpf(struct lan966x_port *port,
+			   struct xdp_frame *xdpf,
+			   struct page *page)
+{
+	struct lan966x *lan966x = port->lan966x;
+	struct lan966x_tx_dcb_buf *next_dcb_buf;
+	struct lan966x_tx *tx = &lan966x->tx;
+	dma_addr_t dma_addr;
+	int next_to_use;
+	__be32 *ifh;
+	int ret = 0;
+
+	spin_lock(&lan966x->tx_lock);
+
+	/* Get next index */
+	next_to_use = lan966x_fdma_get_next_dcb(tx);
+	if (next_to_use < 0) {
+		netif_stop_queue(port->dev);
+		ret = NETDEV_TX_BUSY;
+		goto out;
+	}
+
+	/* Generate new IFH */
+	ifh = page_address(page) + XDP_PACKET_HEADROOM;
+	memset(ifh, 0x0, sizeof(__be32) * IFH_LEN);
+	lan966x_ifh_set_bypass(ifh, 1);
+	lan966x_ifh_set_port(ifh, BIT_ULL(port->chip_port));
+
+	dma_addr = page_pool_get_dma_addr(page);
+	dma_sync_single_for_device(lan966x->dev, dma_addr + XDP_PACKET_HEADROOM,
+				   xdpf->len + IFH_LEN_BYTES,
+				   DMA_TO_DEVICE);
+
+	/* Setup next dcb */
+	lan966x_fdma_tx_setup_dcb(tx, next_to_use, xdpf->len + IFH_LEN_BYTES,
+				  dma_addr + XDP_PACKET_HEADROOM);
+
+	/* Fill up the buffer */
+	next_dcb_buf = &tx->dcbs_buf[next_to_use];
+	next_dcb_buf->use_skb = false;
+	next_dcb_buf->data.xdpf = xdpf;
+	next_dcb_buf->len = xdpf->len + IFH_LEN_BYTES;
+	next_dcb_buf->dma_addr = dma_addr;
+	next_dcb_buf->used = true;
+	next_dcb_buf->ptp = false;
+	next_dcb_buf->dev = port->dev;
+
+	/* Start the transmission */
+	lan966x_fdma_tx_start(tx, next_to_use);
+
+out:
+	spin_unlock(&lan966x->tx_lock);
+
+	return ret;
+}
+
 int lan966x_fdma_xmit(struct sk_buff *skb, __be32 *ifh, struct net_device *dev)
 {
 	struct lan966x_port *port = netdev_priv(dev);
@@ -724,7 +788,8 @@ int lan966x_fdma_xmit(struct sk_buff *skb, __be32 *ifh, struct net_device *dev)
 
 	/* Fill up the buffer */
 	next_dcb_buf = &tx->dcbs_buf[next_to_use];
-	next_dcb_buf->skb = skb;
+	next_dcb_buf->use_skb = true;
+	next_dcb_buf->data.skb = skb;
 	next_dcb_buf->len = skb->len;
 	next_dcb_buf->dma_addr = dma_addr;
 	next_dcb_buf->used = true;

drivers/net/ethernet/microchip/lan966x/lan966x_main.c

@@ -302,13 +302,13 @@ static int lan966x_port_ifh_xmit(struct sk_buff *skb,
 	return NETDEV_TX_BUSY;
 }
 
-static void lan966x_ifh_set_bypass(void *ifh, u64 bypass)
+void lan966x_ifh_set_bypass(void *ifh, u64 bypass)
 {
 	packing(ifh, &bypass, IFH_POS_BYPASS + IFH_WID_BYPASS - 1,
 		IFH_POS_BYPASS, IFH_LEN * 4, PACK, 0);
 }
 
-static void lan966x_ifh_set_port(void *ifh, u64 bypass)
+void lan966x_ifh_set_port(void *ifh, u64 bypass)
 {
 	packing(ifh, &bypass, IFH_POS_DSTS + IFH_WID_DSTS - 1,
 		IFH_POS_DSTS, IFH_LEN * 4, PACK, 0);

drivers/net/ethernet/microchip/lan966x/lan966x_main.h

@@ -105,11 +105,13 @@ enum macaccess_entry_type {
  * FDMA_PASS, frame is valid and can be used
  * FDMA_ERROR, something went wrong, stop getting more frames
  * FDMA_DROP, frame is dropped, but continue to get more frames
+ * FDMA_TX, frame is given to TX, but continue to get more frames
  */
 enum lan966x_fdma_action {
 	FDMA_PASS = 0,
 	FDMA_ERROR,
 	FDMA_DROP,
+	FDMA_TX,
 };
 
 struct lan966x_port;
@@ -175,10 +177,14 @@ struct lan966x_rx {
 struct lan966x_tx_dcb_buf {
 	dma_addr_t dma_addr;
 	struct net_device *dev;
+	union {
 		struct sk_buff *skb;
+		struct xdp_frame *xdpf;
+	} data;
 	u32 len;
 	u32 used : 1;
 	u32 ptp : 1;
+	u32 use_skb : 1;
 };
 
 struct lan966x_tx {
@@ -360,6 +366,8 @@ bool lan966x_hw_offload(struct lan966x *lan966x, u32 port, struct sk_buff *skb);
 
 void lan966x_ifh_get_src_port(void *ifh, u64 *src_port);
 void lan966x_ifh_get_timestamp(void *ifh, u64 *timestamp);
+void lan966x_ifh_set_bypass(void *ifh, u64 bypass);
+void lan966x_ifh_set_port(void *ifh, u64 bypass);
 
 void lan966x_stats_get(struct net_device *dev,
 		       struct rtnl_link_stats64 *stats);
@@ -460,6 +468,9 @@ u32 lan966x_ptp_get_period_ps(void);
 int lan966x_ptp_gettime64(struct ptp_clock_info *ptp, struct timespec64 *ts);
 
 int lan966x_fdma_xmit(struct sk_buff *skb, __be32 *ifh, struct net_device *dev);
+int lan966x_fdma_xmit_xdpf(struct lan966x_port *port,
+			   struct xdp_frame *frame,
+			   struct page *page);
 int lan966x_fdma_change_mtu(struct lan966x *lan966x);
 void lan966x_fdma_netdev_init(struct lan966x *lan966x, struct net_device *dev);
 void lan966x_fdma_netdev_deinit(struct lan966x *lan966x, struct net_device *dev);

drivers/net/ethernet/microchip/lan966x/lan966x_xdp.c

@@ -54,6 +54,7 @@ int lan966x_xdp_run(struct lan966x_port *port, struct page *page, u32 data_len)
 {
 	struct bpf_prog *xdp_prog = port->xdp_prog;
 	struct lan966x *lan966x = port->lan966x;
+	struct xdp_frame *xdpf;
 	struct xdp_buff xdp;
 	u32 act;
 
@@ -66,6 +67,13 @@ int lan966x_xdp_run(struct lan966x_port *port, struct page *page, u32 data_len)
 	switch (act) {
 	case XDP_PASS:
 		return FDMA_PASS;
+	case XDP_TX:
+		xdpf = xdp_convert_buff_to_frame(&xdp);
+		if (!xdpf)
+			return FDMA_DROP;
+
+		return lan966x_fdma_xmit_xdpf(port, xdpf, page) ?
+		       FDMA_DROP : FDMA_TX;
 	default:
 		bpf_warn_invalid_xdp_action(port->dev, xdp_prog, act);
 		fallthrough;