1 files changed, 263 insertions, 193 deletions

diff --git a/drivers/net/ethernet/ti/am65-cpsw-nuss.c b/drivers/net/ethernet/ti/am65-cpsw-nuss.c
index 81d9f21086ec..cbe99017cbfa 100644
--- a/drivers/net/ethernet/ti/am65-cpsw-nuss.c
+++ b/drivers/net/ethernet/ti/am65-cpsw-nuss.c
@@ -138,7 +138,7 @@
 	 AM65_CPSW_PN_TS_CTL_RX_ANX_F_EN)
 
 #define AM65_CPSW_ALE_AGEOUT_DEFAULT	30
-/* Number of TX/RX descriptors */
+/* Number of TX/RX descriptors per channel/flow */
 #define AM65_CPSW_MAX_TX_DESC	500
 #define AM65_CPSW_MAX_RX_DESC	500
 
@@ -150,18 +150,20 @@
 			 NETIF_MSG_RX_ERR | NETIF_MSG_TX_ERR)
 
 #define AM65_CPSW_DEFAULT_TX_CHNS	8
+#define AM65_CPSW_DEFAULT_RX_CHN_FLOWS	1
 
 /* CPPI streaming packet interface */
 #define AM65_CPSW_CPPI_TX_FLOW_ID  0x3FFF
 #define AM65_CPSW_CPPI_TX_PKT_TYPE 0x7
 
 /* XDP */
-#define AM65_CPSW_XDP_CONSUMED 2
-#define AM65_CPSW_XDP_REDIRECT 1
+#define AM65_CPSW_XDP_CONSUMED BIT(1)
+#define AM65_CPSW_XDP_REDIRECT BIT(0)
 #define AM65_CPSW_XDP_PASS     0
 
 /* Include headroom compatible with both skb and xdpf */
-#define AM65_CPSW_HEADROOM (max(NET_SKB_PAD, XDP_PACKET_HEADROOM) + NET_IP_ALIGN)
+#define AM65_CPSW_HEADROOM_NA (max(NET_SKB_PAD, XDP_PACKET_HEADROOM) + NET_IP_ALIGN)
+#define AM65_CPSW_HEADROOM ALIGN(AM65_CPSW_HEADROOM_NA, sizeof(long))
 
 static void am65_cpsw_port_set_sl_mac(struct am65_cpsw_port *slave,
 				      const u8 *dev_addr)
@@ -330,7 +332,7 @@ static void am65_cpsw_nuss_ndo_host_tx_timeout(struct net_device *ndev,
 }
 
 static int am65_cpsw_nuss_rx_push(struct am65_cpsw_common *common,
-				  struct page *page)
+				  struct page *page, u32 flow_idx)
 {
 	struct am65_cpsw_rx_chn *rx_chn = &common->rx_chns;
 	struct cppi5_host_desc_t *desc_rx;
@@ -363,7 +365,8 @@ static int am65_cpsw_nuss_rx_push(struct am65_cpsw_common *common,
 	swdata = cppi5_hdesc_get_swdata(desc_rx);
 	*((void **)swdata) = page_address(page);
 
-	return k3_udma_glue_push_rx_chn(rx_chn->rx_chn, 0, desc_rx, desc_dma);
+	return k3_udma_glue_push_rx_chn(rx_chn->rx_chn, flow_idx,
+					desc_rx, desc_dma);
 }
 
 void am65_cpsw_nuss_set_p0_ptype(struct am65_cpsw_common *common)
@@ -398,22 +401,27 @@ static void am65_cpsw_init_port_emac_ale(struct am65_cpsw_port *port);
 static void am65_cpsw_destroy_xdp_rxqs(struct am65_cpsw_common *common)
 {
 	struct am65_cpsw_rx_chn *rx_chn = &common->rx_chns;
+	struct am65_cpsw_rx_flow *flow;
 	struct xdp_rxq_info *rxq;
-	int i;
+	int id, port;
 
-	for (i = 0; i < common->port_num; i++) {
-		if (!common->ports[i].ndev)
-			continue;
+	for (id = 0; id < common->rx_ch_num_flows; id++) {
+		flow = &rx_chn->flows[id];
 
-		rxq = &common->ports[i].xdp_rxq;
+		for (port = 0; port < common->port_num; port++) {
+			if (!common->ports[port].ndev)
+				continue;
 
-		if (xdp_rxq_info_is_reg(rxq))
-			xdp_rxq_info_unreg(rxq);
-	}
+			rxq = &common->ports[port].xdp_rxq[id];
+
+			if (xdp_rxq_info_is_reg(rxq))
+				xdp_rxq_info_unreg(rxq);
+		}
 
-	if (rx_chn->page_pool) {
-		page_pool_destroy(rx_chn->page_pool);
-		rx_chn->page_pool = NULL;
+		if (flow->page_pool) {
+			page_pool_destroy(flow->page_pool);
+			flow->page_pool = NULL;
+		}
 	}
 }
 
@@ -427,31 +435,44 @@ static int am65_cpsw_create_xdp_rxqs(struct am65_cpsw_common *common)
 		.nid = dev_to_node(common->dev),
 		.dev = common->dev,
 		.dma_dir = DMA_BIDIRECTIONAL,
-		.napi = &common->napi_rx,
+		/* .napi set dynamically */
 	};
+	struct am65_cpsw_rx_flow *flow;
 	struct xdp_rxq_info *rxq;
 	struct page_pool *pool;
-	int i, ret;
-
-	pool = page_pool_create(&pp_params);
-	if (IS_ERR(pool))
-		return PTR_ERR(pool);
+	int id, port, ret;
+
+	for (id = 0; id < common->rx_ch_num_flows; id++) {
+		flow = &rx_chn->flows[id];
+		pp_params.napi = &flow->napi_rx;
+		pool = page_pool_create(&pp_params);
+		if (IS_ERR(pool)) {
+			ret = PTR_ERR(pool);
+			goto err;
+		}
 
-	rx_chn->page_pool = pool;
+		flow->page_pool = pool;
 
-	for (i = 0; i < common->port_num; i++) {
-		if (!common->ports[i].ndev)
-			continue;
+		/* using same page pool is allowed as no running rx handlers
+		 * simultaneously for both ndevs
+		 */
+		for (port = 0; port < common->port_num; port++) {
+			if (!common->ports[port].ndev)
+				continue;
 
-		rxq = &common->ports[i].xdp_rxq;
+			rxq = &common->ports[port].xdp_rxq[id];
 
-		ret = xdp_rxq_info_reg(rxq, common->ports[i].ndev, i, 0);
-		if (ret)
-			goto err;
+			ret = xdp_rxq_info_reg(rxq, common->ports[port].ndev,
+					       id, flow->napi_rx.napi_id);
+			if (ret)
+				goto err;
 
-		ret = xdp_rxq_info_reg_mem_model(rxq, MEM_TYPE_PAGE_POOL, pool);
-		if (ret)
-			goto err;
+			ret = xdp_rxq_info_reg_mem_model(rxq,
+							 MEM_TYPE_PAGE_POOL,
+							 pool);
+			if (ret)
+				goto err;
+		}
 	}
 
 	return 0;
@@ -496,25 +517,27 @@ static enum am65_cpsw_tx_buf_type am65_cpsw_nuss_buf_type(struct am65_cpsw_tx_ch
 								       desc_idx);
 }
 
-static inline void am65_cpsw_put_page(struct am65_cpsw_rx_chn *rx_chn,
+static inline void am65_cpsw_put_page(struct am65_cpsw_rx_flow *flow,
 				      struct page *page,
 				      bool allow_direct,
 				      int desc_idx)
 {
-	page_pool_put_full_page(rx_chn->page_pool, page, allow_direct);
-	rx_chn->pages[desc_idx] = NULL;
+	page_pool_put_full_page(flow->page_pool, page, allow_direct);
+	flow->pages[desc_idx] = NULL;
 }
 
 static void am65_cpsw_nuss_rx_cleanup(void *data, dma_addr_t desc_dma)
 {
-	struct am65_cpsw_rx_chn *rx_chn = data;
+	struct am65_cpsw_rx_flow *flow = data;
 	struct cppi5_host_desc_t *desc_rx;
+	struct am65_cpsw_rx_chn *rx_chn;
 	dma_addr_t buf_dma;
 	u32 buf_dma_len;
 	void *page_addr;
 	void **swdata;
 	int desc_idx;
 
+	rx_chn = &flow->common->rx_chns;
 	desc_rx = k3_cppi_desc_pool_dma2virt(rx_chn->desc_pool, desc_dma);
 	swdata = cppi5_hdesc_get_swdata(desc_rx);
 	page_addr = *swdata;
@@ -525,7 +548,7 @@ static void am65_cpsw_nuss_rx_cleanup(void *data, dma_addr_t desc_dma)
 
 	desc_idx = am65_cpsw_nuss_desc_idx(rx_chn->desc_pool, desc_rx,
 					   rx_chn->dsize_log2);
-	am65_cpsw_put_page(rx_chn, virt_to_page(page_addr), false, desc_idx);
+	am65_cpsw_put_page(flow, virt_to_page(page_addr), false, desc_idx);
 }
 
 static void am65_cpsw_nuss_xmit_free(struct am65_cpsw_tx_chn *tx_chn,
@@ -601,7 +624,8 @@ static int am65_cpsw_nuss_common_open(struct am65_cpsw_common *common)
 	struct am65_cpsw_host *host_p = am65_common_get_host(common);
 	struct am65_cpsw_rx_chn *rx_chn = &common->rx_chns;
 	struct am65_cpsw_tx_chn *tx_chn = common->tx_chns;
-	int port_idx, i, ret, tx;
+	int port_idx, i, ret, tx, flow_idx;
+	struct am65_cpsw_rx_flow *flow;
 	u32 val, port_mask;
 	struct page *page;
 
@@ -669,27 +693,26 @@ static int am65_cpsw_nuss_common_open(struct am65_cpsw_common *common)
 		return ret;
 	}
 
-	for (i = 0; i < rx_chn->descs_num; i++) {
-		page = page_pool_dev_alloc_pages(rx_chn->page_pool);
-		if (!page) {
-			ret = -ENOMEM;
-			if (i)
+	for (flow_idx = 0; flow_idx < common->rx_ch_num_flows; flow_idx++) {
+		flow = &rx_chn->flows[flow_idx];
+		for (i = 0; i < AM65_CPSW_MAX_RX_DESC; i++) {
+			page = page_pool_dev_alloc_pages(flow->page_pool);
+			if (!page) {
+				dev_err(common->dev, "cannot allocate page in flow %d\n",
+					flow_idx);
+				ret = -ENOMEM;
 				goto fail_rx;
+			}
+			flow->pages[i] = page;
 
-			return ret;
-		}
-		rx_chn->pages[i] = page;
-
-		ret = am65_cpsw_nuss_rx_push(common, page);
-		if (ret < 0) {
-			dev_err(common->dev,
-				"cannot submit page to channel rx: %d\n",
-				ret);
-			am65_cpsw_put_page(rx_chn, page, false, i);
-			if (i)
+			ret = am65_cpsw_nuss_rx_push(common, page, flow_idx);
+			if (ret < 0) {
+				dev_err(common->dev,
+					"cannot submit page to rx channel flow %d, error %d\n",
+					flow_idx, ret);
+				am65_cpsw_put_page(flow, page, false, i);
 				goto fail_rx;
-
-			return ret;
+			}
 		}
 	}
 
@@ -699,6 +722,14 @@ static int am65_cpsw_nuss_common_open(struct am65_cpsw_common *common)
 		goto fail_rx;
 	}
 
+	for (i = 0; i < common->rx_ch_num_flows ; i++) {
+		napi_enable(&rx_chn->flows[i].napi_rx);
+		if (rx_chn->flows[i].irq_disabled) {
+			rx_chn->flows[i].irq_disabled = false;
+			enable_irq(rx_chn->flows[i].irq);
+		}
+	}
+
 	for (tx = 0; tx < common->tx_ch_num; tx++) {
 		ret = k3_udma_glue_enable_tx_chn(tx_chn[tx].tx_chn);
 		if (ret) {
@@ -710,12 +741,6 @@ static int am65_cpsw_nuss_common_open(struct am65_cpsw_common *common)
 		napi_enable(&tx_chn[tx].napi_tx);
 	}
 
-	napi_enable(&common->napi_rx);
-	if (common->rx_irq_disabled) {
-		common->rx_irq_disabled = false;
-		enable_irq(rx_chn->irq);
-	}
-
 	dev_dbg(common->dev, "cpsw_nuss started\n");
 	return 0;
 
@@ -726,11 +751,24 @@ fail_tx:
 		tx--;
 	}
 
+	for (flow_idx = 0; i < common->rx_ch_num_flows; flow_idx++) {
+		flow = &rx_chn->flows[flow_idx];
+		if (!flow->irq_disabled) {
+			disable_irq(flow->irq);
+			flow->irq_disabled = true;
+		}
+		napi_disable(&flow->napi_rx);
+	}
+
 	k3_udma_glue_disable_rx_chn(rx_chn->rx_chn);
 
 fail_rx:
-	k3_udma_glue_reset_rx_chn(rx_chn->rx_chn, 0, rx_chn,
-				  am65_cpsw_nuss_rx_cleanup, 0);
+	for (i = 0; i < common->rx_ch_num_flows; i--)
+		k3_udma_glue_reset_rx_chn(rx_chn->rx_chn, i, &rx_chn->flows[i],
+					  am65_cpsw_nuss_rx_cleanup, 0);
+
+	am65_cpsw_destroy_xdp_rxqs(common);
+
 	return ret;
 }
 
@@ -779,12 +817,12 @@ static int am65_cpsw_nuss_common_stop(struct am65_cpsw_common *common)
 			dev_err(common->dev, "rx teardown timeout\n");
 	}
 
-	napi_disable(&common->napi_rx);
-	hrtimer_cancel(&common->rx_hrtimer);
-
-	for (i = 0; i < AM65_CPSW_MAX_RX_FLOWS; i++)
-		k3_udma_glue_reset_rx_chn(rx_chn->rx_chn, i, rx_chn,
-					  am65_cpsw_nuss_rx_cleanup, !!i);
+	for (i = 0; i < common->rx_ch_num_flows; i++) {
+		napi_disable(&rx_chn->flows[i].napi_rx);
+		hrtimer_cancel(&rx_chn->flows[i].rx_hrtimer);
+		k3_udma_glue_reset_rx_chn(rx_chn->rx_chn, i, &rx_chn->flows[i],
+					  am65_cpsw_nuss_rx_cleanup, 0);
+	}
 
 	k3_udma_glue_disable_rx_chn(rx_chn->rx_chn);
 
@@ -793,10 +831,6 @@ static int am65_cpsw_nuss_common_stop(struct am65_cpsw_common *common)
 	writel(0, common->cpsw_base + AM65_CPSW_REG_CTL);
 	writel(0, common->cpsw_base + AM65_CPSW_REG_STAT_PORT_EN);
 
-	for (i = 0; i < rx_chn->descs_num; i++) {
-		if (rx_chn->pages[i])
-			am65_cpsw_put_page(rx_chn, rx_chn->pages[i], false, i);
-	}
 	am65_cpsw_destroy_xdp_rxqs(common);
 
 	dev_dbg(common->dev, "cpsw_nuss stopped\n");
@@ -867,7 +901,7 @@ static int am65_cpsw_nuss_ndo_slave_open(struct net_device *ndev)
 		goto runtime_put;
 	}
 
-	ret = netif_set_real_num_rx_queues(ndev, AM65_CPSW_MAX_RX_QUEUES);
+	ret = netif_set_real_num_rx_queues(ndev, common->rx_ch_num_flows);
 	if (ret) {
 		dev_err(common->dev, "cannot set real number of rx queues\n");
 		goto runtime_put;
@@ -933,7 +967,7 @@ static int am65_cpsw_xdp_tx_frame(struct net_device *ndev,
 	host_desc = k3_cppi_desc_pool_alloc(tx_chn->desc_pool);
 	if (unlikely(!host_desc)) {
 		ndev->stats.tx_dropped++;
-		return -ENOMEM;
+		return AM65_CPSW_XDP_CONSUMED;	/* drop */
 	}
 
 	am65_cpsw_nuss_set_buf_type(tx_chn, host_desc, buf_type);
@@ -942,7 +976,7 @@ static int am65_cpsw_xdp_tx_frame(struct net_device *ndev,
 				 pkt_len, DMA_TO_DEVICE);
 	if (unlikely(dma_mapping_error(tx_chn->dma_dev, dma_buf))) {
 		ndev->stats.tx_dropped++;
-		ret = -ENOMEM;
+		ret = AM65_CPSW_XDP_CONSUMED;	/* drop */
 		goto pool_free;
 	}
 
@@ -977,6 +1011,7 @@ static int am65_cpsw_xdp_tx_frame(struct net_device *ndev,
 		/* Inform BQL */
 		netdev_tx_completed_queue(netif_txq, 1, pkt_len);
 		ndev->stats.tx_errors++;
+		ret = AM65_CPSW_XDP_CONSUMED; /* drop */
 		goto dma_unmap;
 	}
 
@@ -990,13 +1025,15 @@ pool_free:
 	return ret;
 }
 
-static int am65_cpsw_run_xdp(struct am65_cpsw_common *common,
+static int am65_cpsw_run_xdp(struct am65_cpsw_rx_flow *flow,
 			     struct am65_cpsw_port *port,
 			     struct xdp_buff *xdp,
 			     int desc_idx, int cpu, int *len)
 {
-	struct am65_cpsw_rx_chn *rx_chn = &common->rx_chns;
+	struct am65_cpsw_common *common = flow->common;
+	struct am65_cpsw_ndev_priv *ndev_priv;
 	struct net_device *ndev = port->ndev;
+	struct am65_cpsw_ndev_stats *stats;
 	int ret = AM65_CPSW_XDP_CONSUMED;
 	struct am65_cpsw_tx_chn *tx_chn;
 	struct netdev_queue *netif_txq;
@@ -1004,6 +1041,7 @@ static int am65_cpsw_run_xdp(struct am65_cpsw_common *common,
 	struct bpf_prog *prog;
 	struct page *page;
 	u32 act;
+	int err;
 
 	prog = READ_ONCE(port->xdp_prog);
 	if (!prog)
@@ -1013,41 +1051,49 @@ static int am65_cpsw_run_xdp(struct am65_cpsw_common *common,
 	/* XDP prog might have changed packet data and boundaries */
 	*len = xdp->data_end - xdp->data;
 
+	ndev_priv = netdev_priv(ndev);
+	stats = this_cpu_ptr(ndev_priv->stats);
+
 	switch (act) {
 	case XDP_PASS:
 		ret = AM65_CPSW_XDP_PASS;
 		goto out;
 	case XDP_TX:
-		tx_chn = &common->tx_chns[cpu % AM65_CPSW_MAX_TX_QUEUES];
+		tx_chn = &common->tx_chns[cpu % AM65_CPSW_MAX_QUEUES];
 		netif_txq = netdev_get_tx_queue(ndev, tx_chn->id);
 
 		xdpf = xdp_convert_buff_to_frame(xdp);
 		if (unlikely(!xdpf))
-			break;
+			goto drop;
 
 		__netif_tx_lock(netif_txq, cpu);
-		ret = am65_cpsw_xdp_tx_frame(ndev, tx_chn, xdpf,
+		err = am65_cpsw_xdp_tx_frame(ndev, tx_chn, xdpf,
 					     AM65_CPSW_TX_BUF_TYPE_XDP_TX);
 		__netif_tx_unlock(netif_txq);
-		if (ret)
-			break;
+		if (err)
+			goto drop;
 
-		ndev->stats.rx_bytes += *len;
-		ndev->stats.rx_packets++;
+		u64_stats_update_begin(&stats->syncp);
+		stats->rx_bytes += *len;
+		stats->rx_packets++;
+		u64_stats_update_end(&stats->syncp);
 		ret = AM65_CPSW_XDP_CONSUMED;
 		goto out;
 	case XDP_REDIRECT:
 		if (unlikely(xdp_do_redirect(ndev, xdp, prog)))
-			break;
+			goto drop;
 
-		ndev->stats.rx_bytes += *len;
-		ndev->stats.rx_packets++;
+		u64_stats_update_begin(&stats->syncp);
+		stats->rx_bytes += *len;
+		stats->rx_packets++;
+		u64_stats_update_end(&stats->syncp);
 		ret = AM65_CPSW_XDP_REDIRECT;
 		goto out;
 	default:
 		bpf_warn_invalid_xdp_action(ndev, prog, act);
 		fallthrough;
 	case XDP_ABORTED:
+drop:
 		trace_xdp_exception(ndev, prog, act);
 		fallthrough;
 	case XDP_DROP:
@@ -1055,7 +1101,7 @@ static int am65_cpsw_run_xdp(struct am65_cpsw_common *common,
 	}
 
 	page = virt_to_head_page(xdp->data);
-	am65_cpsw_put_page(rx_chn, page, true, desc_idx);
+	am65_cpsw_put_page(flow, page, true, desc_idx);
 
 out:
 	return ret;
@@ -1094,11 +1140,12 @@ static void am65_cpsw_nuss_rx_csum(struct sk_buff *skb, u32 csum_info)
 	}
 }
 
-static int am65_cpsw_nuss_rx_packets(struct am65_cpsw_common *common,
-				     u32 flow_idx, int cpu)
+static int am65_cpsw_nuss_rx_packets(struct am65_cpsw_rx_flow *flow,
+				     int cpu, int *xdp_state)
 {
-	struct am65_cpsw_rx_chn *rx_chn = &common->rx_chns;
+	struct am65_cpsw_rx_chn *rx_chn = &flow->common->rx_chns;
 	u32 buf_dma_len, pkt_len, port_id = 0, csum_info;
+	struct am65_cpsw_common *common = flow->common;
 	struct am65_cpsw_ndev_priv *ndev_priv;
 	struct am65_cpsw_ndev_stats *stats;
 	struct cppi5_host_desc_t *desc_rx;
@@ -1108,12 +1155,14 @@ static int am65_cpsw_nuss_rx_packets(struct am65_cpsw_common *common,
 	struct am65_cpsw_port *port;
 	int headroom, desc_idx, ret;
 	struct net_device *ndev;
+	u32 flow_idx = flow->id;
 	struct sk_buff *skb;
 	struct xdp_buff	xdp;
 	void *page_addr;
 	void **swdata;
 	u32 *psdata;
 
+	*xdp_state = AM65_CPSW_XDP_PASS;
 	ret = k3_udma_glue_pop_rx_chn(rx_chn->rx_chn, flow_idx, &desc_dma);
 	if (ret) {
 		if (ret != -ENODATA)
@@ -1161,15 +1210,13 @@ static int am65_cpsw_nuss_rx_packets(struct am65_cpsw_common *common,
 	}
 
 	if (port->xdp_prog) {
-		xdp_init_buff(&xdp, AM65_CPSW_MAX_PACKET_SIZE, &port->xdp_rxq);
-
-		xdp_prepare_buff(&xdp, page_addr, skb_headroom(skb),
+		xdp_init_buff(&xdp, PAGE_SIZE, &port->xdp_rxq[flow->id]);
+		xdp_prepare_buff(&xdp, page_addr, AM65_CPSW_HEADROOM,
 				 pkt_len, false);
-
-		ret = am65_cpsw_run_xdp(common, port, &xdp, desc_idx,
-					cpu, &pkt_len);
-		if (ret != AM65_CPSW_XDP_PASS)
-			return ret;
+		*xdp_state = am65_cpsw_run_xdp(flow, port, &xdp, desc_idx,
+					       cpu, &pkt_len);
+		if (*xdp_state != AM65_CPSW_XDP_PASS)
+			goto allocate;
 
 		/* Compute additional headroom to be reserved */
 		headroom = (xdp.data - xdp.data_hard_start) - skb_headroom(skb);
@@ -1184,7 +1231,7 @@ static int am65_cpsw_nuss_rx_packets(struct am65_cpsw_common *common,
 	skb_mark_for_recycle(skb);
 	skb->protocol = eth_type_trans(skb, ndev);
 	am65_cpsw_nuss_rx_csum(skb, csum_info);
-	napi_gro_receive(&common->napi_rx, skb);
+	napi_gro_receive(&flow->napi_rx, skb);
 
 	stats = this_cpu_ptr(ndev_priv->stats);
 
@@ -1193,21 +1240,25 @@ static int am65_cpsw_nuss_rx_packets(struct am65_cpsw_common *common,
 	stats->rx_bytes += pkt_len;
 	u64_stats_update_end(&stats->syncp);
 
-	new_page = page_pool_dev_alloc_pages(rx_chn->page_pool);
-	if (unlikely(!new_page))
+allocate:
+	new_page = page_pool_dev_alloc_pages(flow->page_pool);
+	if (unlikely(!new_page)) {
+		dev_err(dev, "page alloc failed\n");
 		return -ENOMEM;
-	rx_chn->pages[desc_idx] = new_page;
+	}
+
+	flow->pages[desc_idx] = new_page;
 
 	if (netif_dormant(ndev)) {
-		am65_cpsw_put_page(rx_chn, new_page, true, desc_idx);
+		am65_cpsw_put_page(flow, new_page, true, desc_idx);
 		ndev->stats.rx_dropped++;
 		return 0;
 	}
 
 requeue:
-	ret = am65_cpsw_nuss_rx_push(common, new_page);
+	ret = am65_cpsw_nuss_rx_push(common, new_page, flow_idx);
 	if (WARN_ON(ret < 0)) {
-		am65_cpsw_put_page(rx_chn, new_page, true, desc_idx);
+		am65_cpsw_put_page(flow, new_page, true, desc_idx);
 		ndev->stats.rx_errors++;
 		ndev->stats.rx_dropped++;
 	}
@@ -1217,54 +1268,48 @@ requeue:
 
 static enum hrtimer_restart am65_cpsw_nuss_rx_timer_callback(struct hrtimer *timer)
 {
-	struct am65_cpsw_common *common =
-			container_of(timer, struct am65_cpsw_common, rx_hrtimer);
+	struct am65_cpsw_rx_flow *flow = container_of(timer,
+						      struct am65_cpsw_rx_flow,
+						      rx_hrtimer);
 
-	enable_irq(common->rx_chns.irq);
+	enable_irq(flow->irq);
 	return HRTIMER_NORESTART;
 }
 
 static int am65_cpsw_nuss_rx_poll(struct napi_struct *napi_rx, int budget)
 {
-	struct am65_cpsw_common *common = am65_cpsw_napi_to_common(napi_rx);
-	int flow = AM65_CPSW_MAX_RX_FLOWS;
+	struct am65_cpsw_rx_flow *flow = am65_cpsw_napi_to_rx_flow(napi_rx);
+	struct am65_cpsw_common *common = flow->common;
 	int cpu = smp_processor_id();
-	bool xdp_redirect = false;
+	int xdp_state_or = 0;
 	int cur_budget, ret;
+	int xdp_state;
 	int num_rx = 0;
 
-	/* process every flow */
-	while (flow--) {
-		cur_budget = budget - num_rx;
-
-		while (cur_budget--) {
-			ret = am65_cpsw_nuss_rx_packets(common, flow, cpu);
-			if (ret) {
-				if (ret == AM65_CPSW_XDP_REDIRECT)
-					xdp_redirect = true;
-				break;
-			}
-			num_rx++;
-		}
-
-		if (num_rx >= budget)
+	/* process only this flow */
+	cur_budget = budget;
+	while (cur_budget--) {
+		ret = am65_cpsw_nuss_rx_packets(flow, cpu, &xdp_state);
+		xdp_state_or |= xdp_state;
+		if (ret)
 			break;
+		num_rx++;
 	}
 
-	if (xdp_redirect)
+	if (xdp_state_or & AM65_CPSW_XDP_REDIRECT)
 		xdp_do_flush();
 
 	dev_dbg(common->dev, "%s num_rx:%d %d\n", __func__, num_rx, budget);
 
 	if (num_rx < budget && napi_complete_done(napi_rx, num_rx)) {
-		if (common->rx_irq_disabled) {
-			common->rx_irq_disabled = false;
-			if (unlikely(common->rx_pace_timeout)) {
-				hrtimer_start(&common->rx_hrtimer,
-					      ns_to_ktime(common->rx_pace_timeout),
+		if (flow->irq_disabled) {
+			flow->irq_disabled = false;
+			if (unlikely(flow->rx_pace_timeout)) {
+				hrtimer_start(&flow->rx_hrtimer,
+					      ns_to_ktime(flow->rx_pace_timeout),
 					      HRTIMER_MODE_REL_PINNED);
 			} else {
-				enable_irq(common->rx_chns.irq);
+				enable_irq(flow->irq);
 			}
 		}
 	}
@@ -1512,11 +1557,11 @@ static int am65_cpsw_nuss_tx_poll(struct napi_struct *napi_tx, int budget)
 
 static irqreturn_t am65_cpsw_nuss_rx_irq(int irq, void *dev_id)
 {
-	struct am65_cpsw_common *common = dev_id;
+	struct am65_cpsw_rx_flow *flow = dev_id;
 
-	common->rx_irq_disabled = true;
+	flow->irq_disabled = true;
 	disable_irq_nosync(irq);
-	napi_schedule(&common->napi_rx);
+	napi_schedule(&flow->napi_rx);
 
 	return IRQ_HANDLED;
 }
@@ -1918,12 +1963,13 @@ static int am65_cpsw_ndo_bpf(struct net_device *ndev, struct netdev_bpf *bpf)
 static int am65_cpsw_ndo_xdp_xmit(struct net_device *ndev, int n,
 				  struct xdp_frame **frames, u32 flags)
 {
+	struct am65_cpsw_common *common = am65_ndev_to_common(ndev);
 	struct am65_cpsw_tx_chn *tx_chn;
 	struct netdev_queue *netif_txq;
 	int cpu = smp_processor_id();
 	int i, nxmit = 0;
 
-	tx_chn = &am65_ndev_to_common(ndev)->tx_chns[cpu % AM65_CPSW_MAX_TX_QUEUES];
+	tx_chn = &common->tx_chns[cpu % common->tx_ch_num];
 	netif_txq = netdev_get_tx_queue(ndev, tx_chn->id);
 
 	__netif_tx_lock(netif_txq, cpu);
@@ -2160,7 +2206,7 @@ static void am65_cpsw_nuss_free_tx_chns(void *data)
 	}
 }
 
-void am65_cpsw_nuss_remove_tx_chns(struct am65_cpsw_common *common)
+static void am65_cpsw_nuss_remove_tx_chns(struct am65_cpsw_common *common)
 {
 	struct device *dev = common->dev;
 	int i;
@@ -2175,15 +2221,9 @@ void am65_cpsw_nuss_remove_tx_chns(struct am65_cpsw_common *common)
 			devm_free_irq(dev, tx_chn->irq, tx_chn);
 
 		netif_napi_del(&tx_chn->napi_tx);
-
-		if (!IS_ERR_OR_NULL(tx_chn->desc_pool))
-			k3_cppi_desc_pool_destroy(tx_chn->desc_pool);
-
-		if (!IS_ERR_OR_NULL(tx_chn->tx_chn))
-			k3_udma_glue_release_tx_chn(tx_chn->tx_chn);
-
-		memset(tx_chn, 0, sizeof(*tx_chn));
 	}
+
+	am65_cpsw_nuss_free_tx_chns(common);
 }
 
 static int am65_cpsw_nuss_ndev_add_tx_napi(struct am65_cpsw_common *common)
@@ -2315,19 +2355,22 @@ static void am65_cpsw_nuss_free_rx_chns(void *data)
 		k3_udma_glue_release_rx_chn(rx_chn->rx_chn);
 }
 
-static void am65_cpsw_nuss_remove_rx_chns(void *data)
+static void am65_cpsw_nuss_remove_rx_chns(struct am65_cpsw_common *common)
 {
-	struct am65_cpsw_common *common = data;
 	struct device *dev = common->dev;
 	struct am65_cpsw_rx_chn *rx_chn;
+	struct am65_cpsw_rx_flow *flows;
+	int i;
 
 	rx_chn = &common->rx_chns;
+	flows = rx_chn->flows;
 	devm_remove_action(dev, am65_cpsw_nuss_free_rx_chns, common);
 
-	if (!(rx_chn->irq < 0))
-		devm_free_irq(dev, rx_chn->irq, common);
-
-	netif_napi_del(&common->napi_rx);
+	for (i = 0; i < common->rx_ch_num_flows; i++) {
+		if (!(flows[i].irq < 0))
+			devm_free_irq(dev, flows[i].irq, &flows[i]);
+		netif_napi_del(&flows[i].napi_rx);
+	}
 
 	am65_cpsw_nuss_free_rx_chns(common);
 
@@ -2340,6 +2383,7 @@ static int am65_cpsw_nuss_init_rx_chns(struct am65_cpsw_common *common)
 	struct k3_udma_glue_rx_channel_cfg rx_cfg = { 0 };
 	u32  max_desc_num = AM65_CPSW_MAX_RX_DESC;
 	struct device *dev = common->dev;
+	struct am65_cpsw_rx_flow *flow;
 	u32 hdesc_size, hdesc_size_out;
 	u32 fdqring_id;
 	int i, ret = 0;
@@ -2348,12 +2392,21 @@ static int am65_cpsw_nuss_init_rx_chns(struct am65_cpsw_common *common)
 					   AM65_CPSW_NAV_SW_DATA_SIZE);
 
 	rx_cfg.swdata_size = AM65_CPSW_NAV_SW_DATA_SIZE;
-	rx_cfg.flow_id_num = AM65_CPSW_MAX_RX_FLOWS;
+	rx_cfg.flow_id_num = common->rx_ch_num_flows;
 	rx_cfg.flow_id_base = common->rx_flow_id_base;
 
 	/* init all flows */
 	rx_chn->dev = dev;
-	rx_chn->descs_num = max_desc_num;
+	rx_chn->descs_num = max_desc_num * rx_cfg.flow_id_num;
+
+	for (i = 0; i < common->rx_ch_num_flows; i++) {
+		flow = &rx_chn->flows[i];
+		flow->page_pool = NULL;
+		flow->pages = devm_kcalloc(dev, AM65_CPSW_MAX_RX_DESC,
+					   sizeof(*flow->pages), GFP_KERNEL);
+		if (!flow->pages)
+			return -ENOMEM;
+	}
 
 	rx_chn->rx_chn = k3_udma_glue_request_rx_chn(dev, "rx", &rx_cfg);
 	if (IS_ERR(rx_chn->rx_chn)) {
@@ -2376,13 +2429,6 @@ static int am65_cpsw_nuss_init_rx_chns(struct am65_cpsw_common *common)
 	rx_chn->dsize_log2 = __fls(hdesc_size_out);
 	WARN_ON(hdesc_size_out != (1 << rx_chn->dsize_log2));
 
-	rx_chn->page_pool = NULL;
-
-	rx_chn->pages = devm_kcalloc(dev, rx_chn->descs_num,
-				     sizeof(*rx_chn->pages), GFP_KERNEL);
-	if (!rx_chn->pages)
-		return -ENOMEM;
-
 	common->rx_flow_id_base =
 			k3_udma_glue_rx_get_flow_id_base(rx_chn->rx_chn);
 	dev_info(dev, "set new flow-id-base %u\n", common->rx_flow_id_base);
@@ -2406,6 +2452,10 @@ static int am65_cpsw_nuss_init_rx_chns(struct am65_cpsw_common *common)
 				K3_UDMA_GLUE_SRC_TAG_LO_USE_REMOTE_SRC_TAG,
 		};
 
+		flow = &rx_chn->flows[i];
+		flow->id = i;
+		flow->common = common;
+
 		rx_flow_cfg.ring_rxfdq0_id = fdqring_id;
 		rx_flow_cfg.rx_cfg.size = max_desc_num;
 		rx_flow_cfg.rxfdq_cfg.size = max_desc_num;
@@ -2422,30 +2472,37 @@ static int am65_cpsw_nuss_init_rx_chns(struct am65_cpsw_common *common)
 				k3_udma_glue_rx_flow_get_fdq_id(rx_chn->rx_chn,
 								i);
 
-		rx_chn->irq = k3_udma_glue_rx_get_irq(rx_chn->rx_chn, i);
-
-		if (rx_chn->irq < 0) {
+		flow->irq = k3_udma_glue_rx_get_irq(rx_chn->rx_chn, i);
+		if (flow->irq <= 0) {
 			dev_err(dev, "Failed to get rx dma irq %d\n",
-				rx_chn->irq);
-			ret = rx_chn->irq;
+				flow->irq);
+			ret = flow->irq;
 			goto err;
 		}
-	}
 
-	netif_napi_add(common->dma_ndev, &common->napi_rx,
-		       am65_cpsw_nuss_rx_poll);
-	hrtimer_init(&common->rx_hrtimer, CLOCK_MONOTONIC, HRTIMER_MODE_REL_PINNED);
-	common->rx_hrtimer.function = &am65_cpsw_nuss_rx_timer_callback;
-
-	ret = devm_request_irq(dev, rx_chn->irq,
-			       am65_cpsw_nuss_rx_irq,
-			       IRQF_TRIGGER_HIGH, dev_name(dev), common);
-	if (ret) {
-		dev_err(dev, "failure requesting rx irq %u, %d\n",
-			rx_chn->irq, ret);
-		goto err;
+		snprintf(flow->name,
+			 sizeof(flow->name), "%s-rx%d",
+			 dev_name(dev), i);
+		netif_napi_add(common->dma_ndev, &flow->napi_rx,
+			       am65_cpsw_nuss_rx_poll);
+		hrtimer_init(&flow->rx_hrtimer, CLOCK_MONOTONIC,
+			     HRTIMER_MODE_REL_PINNED);
+		flow->rx_hrtimer.function = &am65_cpsw_nuss_rx_timer_callback;
+
+		ret = devm_request_irq(dev, flow->irq,
+				       am65_cpsw_nuss_rx_irq,
+				       IRQF_TRIGGER_HIGH,
+				       flow->name, flow);
+		if (ret) {
+			dev_err(dev, "failure requesting rx %d irq %u, %d\n",
+				i, flow->irq, ret);
+			goto err;
+		}
 	}
 
+	/* setup classifier to route priorities to flows */
+	cpsw_ale_classifier_setup_default(common->ale, common->rx_ch_num_flows);
+
 err:
 	i = devm_add_action(dev, am65_cpsw_nuss_free_rx_chns, common);
 	if (i) {
@@ -2689,8 +2746,8 @@ am65_cpsw_nuss_init_port_ndev(struct am65_cpsw_common *common, u32 port_idx)
 	/* alloc netdev */
 	port->ndev = devm_alloc_etherdev_mqs(common->dev,
 					     sizeof(struct am65_cpsw_ndev_priv),
-					     AM65_CPSW_MAX_TX_QUEUES,
-					     AM65_CPSW_MAX_RX_QUEUES);
+					     AM65_CPSW_MAX_QUEUES,
+					     AM65_CPSW_MAX_QUEUES);
 	if (!port->ndev) {
 		dev_err(dev, "error allocating slave net_device %u\n",
 			port->port_id);
@@ -2761,7 +2818,7 @@ am65_cpsw_nuss_init_port_ndev(struct am65_cpsw_common *common, u32 port_idx)
 	}
 
 	phylink = phylink_create(&port->slave.phylink_config,
-				 of_node_to_fwnode(port->slave.port_np),
+				 of_fwnode_handle(port->slave.port_np),
 				 port->slave.phy_if,
 				 &am65_cpsw_phylink_mac_ops);
 	if (IS_ERR(phylink))
@@ -3287,9 +3344,10 @@ static int am65_cpsw_nuss_register_ndevs(struct am65_cpsw_common *common)
 		k3_udma_glue_disable_tx_chn(tx_chan[i].tx_chn);
 	}
 
-	for (i = 0; i < AM65_CPSW_MAX_RX_FLOWS; i++)
-		k3_udma_glue_reset_rx_chn(rx_chan->rx_chn, i, rx_chan,
-					  am65_cpsw_nuss_rx_cleanup, !!i);
+	for (i = 0; i < common->rx_ch_num_flows; i++)
+		k3_udma_glue_reset_rx_chn(rx_chan->rx_chn, i,
+					  &rx_chan->flows[i],
+					  am65_cpsw_nuss_rx_cleanup, 0);
 
 	k3_udma_glue_disable_rx_chn(rx_chan->rx_chn);
 
@@ -3330,12 +3388,21 @@ err_cleanup_ndev:
 	return ret;
 }
 
-int am65_cpsw_nuss_update_tx_chns(struct am65_cpsw_common *common, int num_tx)
+int am65_cpsw_nuss_update_tx_rx_chns(struct am65_cpsw_common *common,
+				     int num_tx, int num_rx)
 {
 	int ret;
 
+	am65_cpsw_nuss_remove_tx_chns(common);
+	am65_cpsw_nuss_remove_rx_chns(common);
+
 	common->tx_ch_num = num_tx;
+	common->rx_ch_num_flows = num_rx;
 	ret = am65_cpsw_nuss_init_tx_chns(common);
+	if (ret)
+		return ret;
+
+	ret = am65_cpsw_nuss_init_rx_chns(common);
 
 	return ret;
 }
@@ -3465,6 +3532,7 @@ static int am65_cpsw_nuss_probe(struct platform_device *pdev)
 	common->rx_flow_id_base = -1;
 	init_completion(&common->tdown_complete);
 	common->tx_ch_num = AM65_CPSW_DEFAULT_TX_CHNS;
+	common->rx_ch_num_flows = AM65_CPSW_DEFAULT_RX_CHN_FLOWS;
 	common->pf_p0_rx_ptype_rrobin = false;
 	common->default_vlan = 1;
 
@@ -3656,8 +3724,10 @@ static int am65_cpsw_nuss_resume(struct device *dev)
 		return ret;
 
 	/* If RX IRQ was disabled before suspend, keep it disabled */
-	if (common->rx_irq_disabled)
-		disable_irq(common->rx_chns.irq);
+	for (i = 0; i < common->rx_ch_num_flows; i++) {
+		if (common->rx_chns.flows[i].irq_disabled)
+			disable_irq(common->rx_chns.flows[i].irq);
+	}
 
 	am65_cpts_resume(common->cpts);