5 files changed, 423 insertions, 96 deletions

diff --git a/drivers/net/ethernet/freescale/dpaa2/Makefile b/drivers/net/ethernet/freescale/dpaa2/Makefile
index 2f424e0a8225..d1e78cdd512f 100644
--- a/drivers/net/ethernet/freescale/dpaa2/Makefile
+++ b/drivers/net/ethernet/freescale/dpaa2/Makefile
@@ -7,6 +7,7 @@ obj-$(CONFIG_FSL_DPAA2_ETH)		+= fsl-dpaa2-eth.o
 obj-$(CONFIG_FSL_DPAA2_PTP_CLOCK)	+= fsl-dpaa2-ptp.o
 
 fsl-dpaa2-eth-objs	:= dpaa2-eth.o dpaa2-ethtool.o dpni.o
+fsl-dpaa2-eth-${CONFIG_DEBUG_FS} += dpaa2-eth-debugfs.o
 fsl-dpaa2-ptp-objs	:= dpaa2-ptp.o dprtc.o
 
 # Needed by the tracing framework
diff --git a/drivers/net/ethernet/freescale/dpaa2/dpaa2-eth-debugfs.c b/drivers/net/ethernet/freescale/dpaa2/dpaa2-eth-debugfs.c
new file mode 100644
index 000000000000..a027f4a9d0cc
--- /dev/null
+++ b/drivers/net/ethernet/freescale/dpaa2/dpaa2-eth-debugfs.c
@@ -0,0 +1,237 @@
+// SPDX-License-Identifier: (GPL-2.0+ OR BSD-3-Clause)
+/* Copyright 2015 Freescale Semiconductor Inc.
+ * Copyright 2018-2019 NXP
+ */
+#include <linux/module.h>
+#include <linux/debugfs.h>
+#include "dpaa2-eth.h"
+#include "dpaa2-eth-debugfs.h"
+
+#define DPAA2_ETH_DBG_ROOT "dpaa2-eth"
+
+static struct dentry *dpaa2_dbg_root;
+
+static int dpaa2_dbg_cpu_show(struct seq_file *file, void *offset)
+{
+	struct dpaa2_eth_priv *priv = (struct dpaa2_eth_priv *)file->private;
+	struct rtnl_link_stats64 *stats;
+	struct dpaa2_eth_drv_stats *extras;
+	int i;
+
+	seq_printf(file, "Per-CPU stats for %s\n", priv->net_dev->name);
+	seq_printf(file, "%s%16s%16s%16s%16s%16s%16s%16s%16s%16s\n",
+		   "CPU", "Rx", "Rx Err", "Rx SG", "Tx", "Tx Err", "Tx conf",
+		   "Tx SG", "Tx realloc", "Enq busy");
+
+	for_each_online_cpu(i) {
+		stats = per_cpu_ptr(priv->percpu_stats, i);
+		extras = per_cpu_ptr(priv->percpu_extras, i);
+		seq_printf(file, "%3d%16llu%16llu%16llu%16llu%16llu%16llu%16llu%16llu%16llu\n",
+			   i,
+			   stats->rx_packets,
+			   stats->rx_errors,
+			   extras->rx_sg_frames,
+			   stats->tx_packets,
+			   stats->tx_errors,
+			   extras->tx_conf_frames,
+			   extras->tx_sg_frames,
+			   extras->tx_reallocs,
+			   extras->tx_portal_busy);
+	}
+
+	return 0;
+}
+
+static int dpaa2_dbg_cpu_open(struct inode *inode, struct file *file)
+{
+	int err;
+	struct dpaa2_eth_priv *priv = (struct dpaa2_eth_priv *)inode->i_private;
+
+	err = single_open(file, dpaa2_dbg_cpu_show, priv);
+	if (err < 0)
+		netdev_err(priv->net_dev, "single_open() failed\n");
+
+	return err;
+}
+
+static const struct file_operations dpaa2_dbg_cpu_ops = {
+	.open = dpaa2_dbg_cpu_open,
+	.read = seq_read,
+	.llseek = seq_lseek,
+	.release = single_release,
+};
+
+static char *fq_type_to_str(struct dpaa2_eth_fq *fq)
+{
+	switch (fq->type) {
+	case DPAA2_RX_FQ:
+		return "Rx";
+	case DPAA2_TX_CONF_FQ:
+		return "Tx conf";
+	default:
+		return "N/A";
+	}
+}
+
+static int dpaa2_dbg_fqs_show(struct seq_file *file, void *offset)
+{
+	struct dpaa2_eth_priv *priv = (struct dpaa2_eth_priv *)file->private;
+	struct dpaa2_eth_fq *fq;
+	u32 fcnt, bcnt;
+	int i, err;
+
+	seq_printf(file, "FQ stats for %s:\n", priv->net_dev->name);
+	seq_printf(file, "%s%16s%16s%16s%16s\n",
+		   "VFQID", "CPU", "Type", "Frames", "Pending frames");
+
+	for (i = 0; i <  priv->num_fqs; i++) {
+		fq = &priv->fq[i];
+		err = dpaa2_io_query_fq_count(NULL, fq->fqid, &fcnt, &bcnt);
+		if (err)
+			fcnt = 0;
+
+		seq_printf(file, "%5d%16d%16s%16llu%16u\n",
+			   fq->fqid,
+			   fq->target_cpu,
+			   fq_type_to_str(fq),
+			   fq->stats.frames,
+			   fcnt);
+	}
+
+	return 0;
+}
+
+static int dpaa2_dbg_fqs_open(struct inode *inode, struct file *file)
+{
+	int err;
+	struct dpaa2_eth_priv *priv = (struct dpaa2_eth_priv *)inode->i_private;
+
+	err = single_open(file, dpaa2_dbg_fqs_show, priv);
+	if (err < 0)
+		netdev_err(priv->net_dev, "single_open() failed\n");
+
+	return err;
+}
+
+static const struct file_operations dpaa2_dbg_fq_ops = {
+	.open = dpaa2_dbg_fqs_open,
+	.read = seq_read,
+	.llseek = seq_lseek,
+	.release = single_release,
+};
+
+static int dpaa2_dbg_ch_show(struct seq_file *file, void *offset)
+{
+	struct dpaa2_eth_priv *priv = (struct dpaa2_eth_priv *)file->private;
+	struct dpaa2_eth_channel *ch;
+	int i;
+
+	seq_printf(file, "Channel stats for %s:\n", priv->net_dev->name);
+	seq_printf(file, "%s%16s%16s%16s%16s\n",
+		   "CHID", "CPU", "Deq busy", "CDANs", "Buf count");
+
+	for (i = 0; i < priv->num_channels; i++) {
+		ch = priv->channel[i];
+		seq_printf(file, "%4d%16d%16llu%16llu%16d\n",
+			   ch->ch_id,
+			   ch->nctx.desired_cpu,
+			   ch->stats.dequeue_portal_busy,
+			   ch->stats.cdan,
+			   ch->buf_count);
+	}
+
+	return 0;
+}
+
+static int dpaa2_dbg_ch_open(struct inode *inode, struct file *file)
+{
+	int err;
+	struct dpaa2_eth_priv *priv = (struct dpaa2_eth_priv *)inode->i_private;
+
+	err = single_open(file, dpaa2_dbg_ch_show, priv);
+	if (err < 0)
+		netdev_err(priv->net_dev, "single_open() failed\n");
+
+	return err;
+}
+
+static const struct file_operations dpaa2_dbg_ch_ops = {
+	.open = dpaa2_dbg_ch_open,
+	.read = seq_read,
+	.llseek = seq_lseek,
+	.release = single_release,
+};
+
+void dpaa2_dbg_add(struct dpaa2_eth_priv *priv)
+{
+	if (!dpaa2_dbg_root)
+		return;
+
+	/* Create a directory for the interface */
+	priv->dbg.dir = debugfs_create_dir(priv->net_dev->name,
+					   dpaa2_dbg_root);
+	if (!priv->dbg.dir) {
+		netdev_err(priv->net_dev, "debugfs_create_dir() failed\n");
+		return;
+	}
+
+	/* per-cpu stats file */
+	priv->dbg.cpu_stats = debugfs_create_file("cpu_stats", 0444,
+						  priv->dbg.dir, priv,
+						  &dpaa2_dbg_cpu_ops);
+	if (!priv->dbg.cpu_stats) {
+		netdev_err(priv->net_dev, "debugfs_create_file() failed\n");
+		goto err_cpu_stats;
+	}
+
+	/* per-fq stats file */
+	priv->dbg.fq_stats = debugfs_create_file("fq_stats", 0444,
+						 priv->dbg.dir, priv,
+						 &dpaa2_dbg_fq_ops);
+	if (!priv->dbg.fq_stats) {
+		netdev_err(priv->net_dev, "debugfs_create_file() failed\n");
+		goto err_fq_stats;
+	}
+
+	/* per-fq stats file */
+	priv->dbg.ch_stats = debugfs_create_file("ch_stats", 0444,
+						 priv->dbg.dir, priv,
+						 &dpaa2_dbg_ch_ops);
+	if (!priv->dbg.fq_stats) {
+		netdev_err(priv->net_dev, "debugfs_create_file() failed\n");
+		goto err_ch_stats;
+	}
+
+	return;
+
+err_ch_stats:
+	debugfs_remove(priv->dbg.fq_stats);
+err_fq_stats:
+	debugfs_remove(priv->dbg.cpu_stats);
+err_cpu_stats:
+	debugfs_remove(priv->dbg.dir);
+}
+
+void dpaa2_dbg_remove(struct dpaa2_eth_priv *priv)
+{
+	debugfs_remove(priv->dbg.fq_stats);
+	debugfs_remove(priv->dbg.ch_stats);
+	debugfs_remove(priv->dbg.cpu_stats);
+	debugfs_remove(priv->dbg.dir);
+}
+
+void dpaa2_eth_dbg_init(void)
+{
+	dpaa2_dbg_root = debugfs_create_dir(DPAA2_ETH_DBG_ROOT, NULL);
+	if (!dpaa2_dbg_root) {
+		pr_err("DPAA2-ETH: debugfs create failed\n");
+		return;
+	}
+
+	pr_debug("DPAA2-ETH: debugfs created\n");
+}
+
+void dpaa2_eth_dbg_exit(void)
+{
+	debugfs_remove(dpaa2_dbg_root);
+}
diff --git a/drivers/net/ethernet/freescale/dpaa2/dpaa2-eth-debugfs.h b/drivers/net/ethernet/freescale/dpaa2/dpaa2-eth-debugfs.h
new file mode 100644
index 000000000000..4f63de997a26
--- /dev/null
+++ b/drivers/net/ethernet/freescale/dpaa2/dpaa2-eth-debugfs.h
@@ -0,0 +1,31 @@
+/* SPDX-License-Identifier: (GPL-2.0+ OR BSD-3-Clause) */
+/* Copyright 2015 Freescale Semiconductor Inc.
+ * Copyright 2018-2019 NXP
+ */
+#ifndef DPAA2_ETH_DEBUGFS_H
+#define DPAA2_ETH_DEBUGFS_H
+
+#include <linux/dcache.h>
+
+struct dpaa2_eth_priv;
+
+struct dpaa2_debugfs {
+	struct dentry *dir;
+	struct dentry *fq_stats;
+	struct dentry *ch_stats;
+	struct dentry *cpu_stats;
+};
+
+#ifdef CONFIG_DEBUG_FS
+void dpaa2_eth_dbg_init(void);
+void dpaa2_eth_dbg_exit(void);
+void dpaa2_dbg_add(struct dpaa2_eth_priv *priv);
+void dpaa2_dbg_remove(struct dpaa2_eth_priv *priv);
+#else
+static inline void dpaa2_eth_dbg_init(void) {}
+static inline void dpaa2_eth_dbg_exit(void) {}
+static inline void dpaa2_dbg_add(struct dpaa2_eth_priv *priv) {}
+static inline void dpaa2_dbg_remove(struct dpaa2_eth_priv *priv) {}
+#endif /* CONFIG_DEBUG_FS */
+
+#endif /* DPAA2_ETH_DEBUGFS_H */
diff --git a/drivers/net/ethernet/freescale/dpaa2/dpaa2-eth.c b/drivers/net/ethernet/freescale/dpaa2/dpaa2-eth.c
index 1ca9a18139ec..87777b09f5e0 100644
--- a/drivers/net/ethernet/freescale/dpaa2/dpaa2-eth.c
+++ b/drivers/net/ethernet/freescale/dpaa2/dpaa2-eth.c
@@ -86,16 +86,16 @@ static void free_rx_fd(struct dpaa2_eth_priv *priv,
 	for (i = 1; i < DPAA2_ETH_MAX_SG_ENTRIES; i++) {
 		addr = dpaa2_sg_get_addr(&sgt[i]);
 		sg_vaddr = dpaa2_iova_to_virt(priv->iommu_domain, addr);
-		dma_unmap_single(dev, addr, DPAA2_ETH_RX_BUF_SIZE,
-				 DMA_BIDIRECTIONAL);
+		dma_unmap_page(dev, addr, DPAA2_ETH_RX_BUF_SIZE,
+			       DMA_BIDIRECTIONAL);
 
-		skb_free_frag(sg_vaddr);
+		free_pages((unsigned long)sg_vaddr, 0);
 		if (dpaa2_sg_is_final(&sgt[i]))
 			break;
 	}
 
 free_buf:
-	skb_free_frag(vaddr);
+	free_pages((unsigned long)vaddr, 0);
 }
 
 /* Build a linear skb based on a single-buffer frame descriptor */
@@ -109,7 +109,7 @@ static struct sk_buff *build_linear_skb(struct dpaa2_eth_channel *ch,
 
 	ch->buf_count--;
 
-	skb = build_skb(fd_vaddr, DPAA2_ETH_SKB_SIZE);
+	skb = build_skb(fd_vaddr, DPAA2_ETH_RX_BUF_RAW_SIZE);
 	if (unlikely(!skb))
 		return NULL;
 
@@ -144,19 +144,19 @@ static struct sk_buff *build_frag_skb(struct dpaa2_eth_priv *priv,
 		/* Get the address and length from the S/G entry */
 		sg_addr = dpaa2_sg_get_addr(sge);
 		sg_vaddr = dpaa2_iova_to_virt(priv->iommu_domain, sg_addr);
-		dma_unmap_single(dev, sg_addr, DPAA2_ETH_RX_BUF_SIZE,
-				 DMA_BIDIRECTIONAL);
+		dma_unmap_page(dev, sg_addr, DPAA2_ETH_RX_BUF_SIZE,
+			       DMA_BIDIRECTIONAL);
 
 		sg_length = dpaa2_sg_get_len(sge);
 
 		if (i == 0) {
 			/* We build the skb around the first data buffer */
-			skb = build_skb(sg_vaddr, DPAA2_ETH_SKB_SIZE);
+			skb = build_skb(sg_vaddr, DPAA2_ETH_RX_BUF_RAW_SIZE);
 			if (unlikely(!skb)) {
 				/* Free the first SG entry now, since we already
 				 * unmapped it and obtained the virtual address
 				 */
-				skb_free_frag(sg_vaddr);
+				free_pages((unsigned long)sg_vaddr, 0);
 
 				/* We still need to subtract the buffers used
 				 * by this FD from our software counter
@@ -211,9 +211,9 @@ static void free_bufs(struct dpaa2_eth_priv *priv, u64 *buf_array, int count)
 
 	for (i = 0; i < count; i++) {
 		vaddr = dpaa2_iova_to_virt(priv->iommu_domain, buf_array[i]);
-		dma_unmap_single(dev, buf_array[i], DPAA2_ETH_RX_BUF_SIZE,
-				 DMA_BIDIRECTIONAL);
-		skb_free_frag(vaddr);
+		dma_unmap_page(dev, buf_array[i], DPAA2_ETH_RX_BUF_SIZE,
+			       DMA_BIDIRECTIONAL);
+		free_pages((unsigned long)vaddr, 0);
 	}
 }
 
@@ -264,9 +264,7 @@ static int xdp_enqueue(struct dpaa2_eth_priv *priv, struct dpaa2_fd *fd,
 
 	fq = &priv->fq[queue_id];
 	for (i = 0; i < DPAA2_ETH_ENQUEUE_RETRIES; i++) {
-		err = dpaa2_io_service_enqueue_qd(fq->channel->dpio,
-						  priv->tx_qdid, 0,
-						  fq->tx_qdbin, fd);
+		err = priv->enqueue(priv, fq, fd, 0);
 		if (err != -EBUSY)
 			break;
 	}
@@ -378,16 +376,16 @@ static void dpaa2_eth_rx(struct dpaa2_eth_priv *priv,
 			return;
 		}
 
-		dma_unmap_single(dev, addr, DPAA2_ETH_RX_BUF_SIZE,
-				 DMA_BIDIRECTIONAL);
+		dma_unmap_page(dev, addr, DPAA2_ETH_RX_BUF_SIZE,
+			       DMA_BIDIRECTIONAL);
 		skb = build_linear_skb(ch, fd, vaddr);
 	} else if (fd_format == dpaa2_fd_sg) {
 		WARN_ON(priv->xdp_prog);
 
-		dma_unmap_single(dev, addr, DPAA2_ETH_RX_BUF_SIZE,
-				 DMA_BIDIRECTIONAL);
+		dma_unmap_page(dev, addr, DPAA2_ETH_RX_BUF_SIZE,
+			       DMA_BIDIRECTIONAL);
 		skb = build_frag_skb(priv, ch, buf_data);
-		skb_free_frag(vaddr);
+		free_pages((unsigned long)vaddr, 0);
 		percpu_extras->rx_sg_frames++;
 		percpu_extras->rx_sg_bytes += dpaa2_fd_get_len(fd);
 	} else {
@@ -657,7 +655,7 @@ static int build_single_fd(struct dpaa2_eth_priv *priv,
  * dpaa2_eth_tx().
  */
 static void free_tx_fd(const struct dpaa2_eth_priv *priv,
-		       const struct dpaa2_fd *fd)
+		       const struct dpaa2_fd *fd, bool in_napi)
 {
 	struct device *dev = priv->net_dev->dev.parent;
 	dma_addr_t fd_addr;
@@ -712,7 +710,7 @@ static void free_tx_fd(const struct dpaa2_eth_priv *priv,
 		skb_free_frag(skbh);
 
 	/* Move on with skb release */
-	dev_kfree_skb(skb);
+	napi_consume_skb(skb, in_napi);
 }
 
 static netdev_tx_t dpaa2_eth_tx(struct sk_buff *skb, struct net_device *net_dev)
@@ -785,9 +783,7 @@ static netdev_tx_t dpaa2_eth_tx(struct sk_buff *skb, struct net_device *net_dev)
 	queue_mapping = skb_get_queue_mapping(skb);
 	fq = &priv->fq[queue_mapping];
 	for (i = 0; i < DPAA2_ETH_ENQUEUE_RETRIES; i++) {
-		err = dpaa2_io_service_enqueue_qd(fq->channel->dpio,
-						  priv->tx_qdid, 0,
-						  fq->tx_qdbin, &fd);
+		err = priv->enqueue(priv, fq, &fd, 0);
 		if (err != -EBUSY)
 			break;
 	}
@@ -795,7 +791,7 @@ static netdev_tx_t dpaa2_eth_tx(struct sk_buff *skb, struct net_device *net_dev)
 	if (unlikely(err < 0)) {
 		percpu_stats->tx_errors++;
 		/* Clean up everything, including freeing the skb */
-		free_tx_fd(priv, &fd);
+		free_tx_fd(priv, &fd, false);
 	} else {
 		fd_len = dpaa2_fd_get_len(&fd);
 		percpu_stats->tx_packets++;
@@ -837,7 +833,7 @@ static void dpaa2_eth_tx_conf(struct dpaa2_eth_priv *priv,
 
 	/* Check frame errors in the FD field */
 	fd_errors = dpaa2_fd_get_ctrl(fd) & DPAA2_FD_TX_ERR_MASK;
-	free_tx_fd(priv, fd);
+	free_tx_fd(priv, fd, true);
 
 	if (likely(!fd_errors))
 		return;
@@ -903,7 +899,7 @@ static int add_bufs(struct dpaa2_eth_priv *priv,
 {
 	struct device *dev = priv->net_dev->dev.parent;
 	u64 buf_array[DPAA2_ETH_BUFS_PER_CMD];
-	void *buf;
+	struct page *page;
 	dma_addr_t addr;
 	int i, err;
 
@@ -911,14 +907,16 @@ static int add_bufs(struct dpaa2_eth_priv *priv,
 		/* Allocate buffer visible to WRIOP + skb shared info +
 		 * alignment padding
 		 */
-		buf = napi_alloc_frag(dpaa2_eth_buf_raw_size(priv));
-		if (unlikely(!buf))
+		/* allocate one page for each Rx buffer. WRIOP sees
+		 * the entire page except for a tailroom reserved for
+		 * skb shared info
+		 */
+		page = dev_alloc_pages(0);
+		if (!page)
 			goto err_alloc;
 
-		buf = PTR_ALIGN(buf, priv->rx_buf_align);
-
-		addr = dma_map_single(dev, buf, DPAA2_ETH_RX_BUF_SIZE,
-				      DMA_BIDIRECTIONAL);
+		addr = dma_map_page(dev, page, 0, DPAA2_ETH_RX_BUF_SIZE,
+				    DMA_BIDIRECTIONAL);
 		if (unlikely(dma_mapping_error(dev, addr)))
 			goto err_map;
 
@@ -926,7 +924,7 @@ static int add_bufs(struct dpaa2_eth_priv *priv,
 
 		/* tracing point */
 		trace_dpaa2_eth_buf_seed(priv->net_dev,
-					 buf, dpaa2_eth_buf_raw_size(priv),
+					 page, DPAA2_ETH_RX_BUF_RAW_SIZE,
 					 addr, DPAA2_ETH_RX_BUF_SIZE,
 					 bpid);
 	}
@@ -948,7 +946,7 @@ release_bufs:
 	return i;
 
 err_map:
-	skb_free_frag(buf);
+	__free_pages(page, 0);
 err_alloc:
 	/* If we managed to allocate at least some buffers,
 	 * release them to hardware
@@ -1243,34 +1241,36 @@ enable_err:
 	return err;
 }
 
-/* The DPIO store must be empty when we call this,
- * at the end of every NAPI cycle.
- */
-static u32 drain_channel(struct dpaa2_eth_channel *ch)
+/* Total number of in-flight frames on ingress queues */
+static u32 ingress_fq_count(struct dpaa2_eth_priv *priv)
 {
-	u32 drained = 0, total = 0;
+	struct dpaa2_eth_fq *fq;
+	u32 fcnt = 0, bcnt = 0, total = 0;
+	int i, err;
 
-	do {
-		pull_channel(ch);
-		drained = consume_frames(ch, NULL);
-		total += drained;
-	} while (drained);
+	for (i = 0; i < priv->num_fqs; i++) {
+		fq = &priv->fq[i];
+		err = dpaa2_io_query_fq_count(NULL, fq->fqid, &fcnt, &bcnt);
+		if (err) {
+			netdev_warn(priv->net_dev, "query_fq_count failed");
+			break;
+		}
+		total += fcnt;
+	}
 
 	return total;
 }
 
-static u32 drain_ingress_frames(struct dpaa2_eth_priv *priv)
+static void wait_for_fq_empty(struct dpaa2_eth_priv *priv)
 {
-	struct dpaa2_eth_channel *ch;
-	int i;
-	u32 drained = 0;
-
-	for (i = 0; i < priv->num_channels; i++) {
-		ch = priv->channel[i];
-		drained += drain_channel(ch);
-	}
+	int retries = 10;
+	u32 pending;
 
-	return drained;
+	do {
+		pending = ingress_fq_count(priv);
+		if (pending)
+			msleep(100);
+	} while (pending && --retries);
 }
 
 static int dpaa2_eth_stop(struct net_device *net_dev)
@@ -1278,14 +1278,22 @@ static int dpaa2_eth_stop(struct net_device *net_dev)
 	struct dpaa2_eth_priv *priv = netdev_priv(net_dev);
 	int dpni_enabled = 0;
 	int retries = 10;
-	u32 drained;
 
 	netif_tx_stop_all_queues(net_dev);
 	netif_carrier_off(net_dev);
 
-	/* Loop while dpni_disable() attempts to drain the egress FQs
-	 * and confirm them back to us.
+	/* On dpni_disable(), the MC firmware will:
+	 * - stop MAC Rx and wait for all Rx frames to be enqueued to software
+	 * - cut off WRIOP dequeues from egress FQs and wait until transmission
+	 * of all in flight Tx frames is finished (and corresponding Tx conf
+	 * frames are enqueued back to software)
+	 *
+	 * Before calling dpni_disable(), we wait for all Tx frames to arrive
+	 * on WRIOP. After it finishes, wait until all remaining frames on Rx
+	 * and Tx conf queues are consumed on NAPI poll.
 	 */
+	msleep(500);
+
 	do {
 		dpni_disable(priv->mc_io, 0, priv->mc_token);
 		dpni_is_enabled(priv->mc_io, 0, priv->mc_token, &dpni_enabled);
@@ -1300,19 +1308,9 @@ static int dpaa2_eth_stop(struct net_device *net_dev)
 		 */
 	}
 
-	/* Wait for NAPI to complete on every core and disable it.
-	 * In particular, this will also prevent NAPI from being rescheduled if
-	 * a new CDAN is serviced, effectively discarding the CDAN. We therefore
-	 * don't even need to disarm the channels, except perhaps for the case
-	 * of a huge coalescing value.
-	 */
+	wait_for_fq_empty(priv);
 	disable_ch_napi(priv);
 
-	 /* Manually drain the Rx and TxConf queues */
-	drained = drain_ingress_frames(priv);
-	if (drained)
-		netdev_dbg(net_dev, "Drained %d frames.\n", drained);
-
 	/* Empty the buffer pool */
 	drain_pool(priv);
 
@@ -2134,6 +2132,7 @@ static int set_buffer_layout(struct dpaa2_eth_priv *priv)
 {
 	struct device *dev = priv->net_dev->dev.parent;
 	struct dpni_buffer_layout buf_layout = {0};
+	u16 rx_buf_align;
 	int err;
 
 	/* We need to check for WRIOP version 1.0.0, but depending on the MC
@@ -2142,9 +2141,9 @@ static int set_buffer_layout(struct dpaa2_eth_priv *priv)
 	 */
 	if (priv->dpni_attrs.wriop_version == DPAA2_WRIOP_VERSION(0, 0, 0) ||
 	    priv->dpni_attrs.wriop_version == DPAA2_WRIOP_VERSION(1, 0, 0))
-		priv->rx_buf_align = DPAA2_ETH_RX_BUF_ALIGN_REV1;
+		rx_buf_align = DPAA2_ETH_RX_BUF_ALIGN_REV1;
 	else
-		priv->rx_buf_align = DPAA2_ETH_RX_BUF_ALIGN;
+		rx_buf_align = DPAA2_ETH_RX_BUF_ALIGN;
 
 	/* tx buffer */
 	buf_layout.private_data_size = DPAA2_ETH_SWA_SIZE;
@@ -2184,7 +2183,7 @@ static int set_buffer_layout(struct dpaa2_eth_priv *priv)
 	/* rx buffer */
 	buf_layout.pass_frame_status = true;
 	buf_layout.pass_parser_result = true;
-	buf_layout.data_align = priv->rx_buf_align;
+	buf_layout.data_align = rx_buf_align;
 	buf_layout.data_head_room = dpaa2_eth_rx_head_room(priv);
 	buf_layout.private_data_size = 0;
 	buf_layout.options = DPNI_BUF_LAYOUT_OPT_PARSER_RESULT |
@@ -2202,6 +2201,36 @@ static int set_buffer_layout(struct dpaa2_eth_priv *priv)
 	return 0;
 }
 
+#define DPNI_ENQUEUE_FQID_VER_MAJOR	7
+#define DPNI_ENQUEUE_FQID_VER_MINOR	9
+
+static inline int dpaa2_eth_enqueue_qd(struct dpaa2_eth_priv *priv,
+				       struct dpaa2_eth_fq *fq,
+				       struct dpaa2_fd *fd, u8 prio)
+{
+	return dpaa2_io_service_enqueue_qd(fq->channel->dpio,
+					   priv->tx_qdid, prio,
+					   fq->tx_qdbin, fd);
+}
+
+static inline int dpaa2_eth_enqueue_fq(struct dpaa2_eth_priv *priv,
+				       struct dpaa2_eth_fq *fq,
+				       struct dpaa2_fd *fd,
+				       u8 prio __always_unused)
+{
+	return dpaa2_io_service_enqueue_fq(fq->channel->dpio,
+					   fq->tx_fqid, fd);
+}
+
+static void set_enqueue_mode(struct dpaa2_eth_priv *priv)
+{
+	if (dpaa2_eth_cmp_dpni_ver(priv, DPNI_ENQUEUE_FQID_VER_MAJOR,
+				   DPNI_ENQUEUE_FQID_VER_MINOR) < 0)
+		priv->enqueue = dpaa2_eth_enqueue_qd;
+	else
+		priv->enqueue = dpaa2_eth_enqueue_fq;
+}
+
 /* Configure the DPNI object this interface is associated with */
 static int setup_dpni(struct fsl_mc_device *ls_dev)
 {
@@ -2255,6 +2284,8 @@ static int setup_dpni(struct fsl_mc_device *ls_dev)
 	if (err)
 		goto close;
 
+	set_enqueue_mode(priv);
+
 	priv->cls_rules = devm_kzalloc(dev, sizeof(struct dpaa2_eth_cls_rule) *
 				       dpaa2_eth_fs_count(priv), GFP_KERNEL);
 	if (!priv->cls_rules)
@@ -2339,6 +2370,7 @@ static int setup_tx_flow(struct dpaa2_eth_priv *priv,
 	}
 
 	fq->tx_qdbin = qid.qdbin;
+	fq->tx_fqid = qid.fqid;
 
 	err = dpni_get_queue(priv->mc_io, 0, priv->mc_token,
 			     DPNI_QUEUE_TX_CONFIRM, 0, fq->flowid,
@@ -3083,6 +3115,10 @@ static int dpaa2_eth_probe(struct fsl_mc_device *dpni_dev)
 		goto err_netdev_reg;
 	}
 
+#ifdef CONFIG_DEBUG_FS
+	dpaa2_dbg_add(priv);
+#endif
+
 	dev_info(dev, "Probed interface %s\n", net_dev->name);
 	return 0;
 
@@ -3126,6 +3162,9 @@ static int dpaa2_eth_remove(struct fsl_mc_device *ls_dev)
 	net_dev = dev_get_drvdata(dev);
 	priv = netdev_priv(net_dev);
 
+#ifdef CONFIG_DEBUG_FS
+	dpaa2_dbg_remove(priv);
+#endif
 	unregister_netdev(net_dev);
 
 	if (priv->do_link_poll)
@@ -3170,4 +3209,25 @@ static struct fsl_mc_driver dpaa2_eth_driver = {
 	.match_id_table = dpaa2_eth_match_id_table
 };
 
-module_fsl_mc_driver(dpaa2_eth_driver);
+static int __init dpaa2_eth_driver_init(void)
+{
+	int err;
+
+	dpaa2_eth_dbg_init();
+	err = fsl_mc_driver_register(&dpaa2_eth_driver);
+	if (err) {
+		dpaa2_eth_dbg_exit();
+		return err;
+	}
+
+	return 0;
+}
+
+static void __exit dpaa2_eth_driver_exit(void)
+{
+	dpaa2_eth_dbg_exit();
+	fsl_mc_driver_unregister(&dpaa2_eth_driver);
+}
+
+module_init(dpaa2_eth_driver_init);
+module_exit(dpaa2_eth_driver_exit);
diff --git a/drivers/net/ethernet/freescale/dpaa2/dpaa2-eth.h b/drivers/net/ethernet/freescale/dpaa2/dpaa2-eth.h
index 69c965de192b..9510928b7cca 100644
--- a/drivers/net/ethernet/freescale/dpaa2/dpaa2-eth.h
+++ b/drivers/net/ethernet/freescale/dpaa2/dpaa2-eth.h
@@ -16,6 +16,7 @@
 #include "dpni-cmd.h"
 
 #include "dpaa2-eth-trace.h"
+#include "dpaa2-eth-debugfs.h"
 
 #define DPAA2_WRIOP_VERSION(x, y, z) ((x) << 10 | (y) << 5 | (z) << 0)
 
@@ -52,7 +53,8 @@
  */
 #define DPAA2_ETH_MAX_FRAMES_PER_QUEUE	(DPAA2_ETH_TAILDROP_THRESH / 64)
 #define DPAA2_ETH_NUM_BUFS		(DPAA2_ETH_MAX_FRAMES_PER_QUEUE + 256)
-#define DPAA2_ETH_REFILL_THRESH		DPAA2_ETH_MAX_FRAMES_PER_QUEUE
+#define DPAA2_ETH_REFILL_THRESH \
+	(DPAA2_ETH_NUM_BUFS - DPAA2_ETH_BUFS_PER_CMD)
 
 /* Maximum number of buffers that can be acquired/released through a single
  * QBMan command
@@ -62,9 +64,11 @@
 /* Hardware requires alignment for ingress/egress buffer addresses */
 #define DPAA2_ETH_TX_BUF_ALIGN		64
 
-#define DPAA2_ETH_RX_BUF_SIZE		2048
-#define DPAA2_ETH_SKB_SIZE \
-	(DPAA2_ETH_RX_BUF_SIZE + SKB_DATA_ALIGN(sizeof(struct skb_shared_info)))
+#define DPAA2_ETH_RX_BUF_RAW_SIZE	PAGE_SIZE
+#define DPAA2_ETH_RX_BUF_TAILROOM \
+	SKB_DATA_ALIGN(sizeof(struct skb_shared_info))
+#define DPAA2_ETH_RX_BUF_SIZE \
+	(DPAA2_ETH_RX_BUF_RAW_SIZE - DPAA2_ETH_RX_BUF_TAILROOM)
 
 /* Hardware annotation area in RX/TX buffers */
 #define DPAA2_ETH_RX_HWA_SIZE		64
@@ -273,6 +277,7 @@ struct dpaa2_eth_priv;
 struct dpaa2_eth_fq {
 	u32 fqid;
 	u32 tx_qdbin;
+	u32 tx_fqid;
 	u16 flowid;
 	int target_cpu;
 	u32 dq_frames;
@@ -325,6 +330,9 @@ struct dpaa2_eth_priv {
 
 	u8 num_fqs;
 	struct dpaa2_eth_fq fq[DPAA2_ETH_MAX_QUEUES];
+	int (*enqueue)(struct dpaa2_eth_priv *priv,
+		       struct dpaa2_eth_fq *fq,
+		       struct dpaa2_fd *fd, u8 prio);
 
 	u8 num_channels;
 	struct dpaa2_eth_channel *channel[DPAA2_ETH_MAX_DPCONS];
@@ -342,7 +350,6 @@ struct dpaa2_eth_priv {
 	bool rx_tstamp; /* Rx timestamping enabled */
 
 	u16 tx_qdid;
-	u16 rx_buf_align;
 	struct fsl_mc_io *mc_io;
 	/* Cores which have an affine DPIO/DPCON.
 	 * This is the cpu set on which Rx and Tx conf frames are processed
@@ -365,6 +372,9 @@ struct dpaa2_eth_priv {
 	struct dpaa2_eth_cls_rule *cls_rules;
 	u8 rx_cls_enabled;
 	struct bpf_prog *xdp_prog;
+#ifdef CONFIG_DEBUG_FS
+	struct dpaa2_debugfs dbg;
+#endif
 };
 
 #define DPAA2_RXH_SUPPORTED	(RXH_L2DA | RXH_VLAN | RXH_L3_PROTO \
@@ -405,20 +415,15 @@ static inline int dpaa2_eth_cmp_dpni_ver(struct dpaa2_eth_priv *priv,
 #define dpaa2_eth_fs_count(priv)        \
 	((priv)->dpni_attrs.fs_entries)
 
+/* We have exactly one {Rx, Tx conf} queue per channel */
+#define dpaa2_eth_queue_count(priv)     \
+	((priv)->num_channels)
+
 enum dpaa2_eth_rx_dist {
 	DPAA2_ETH_RX_DIST_HASH,
 	DPAA2_ETH_RX_DIST_CLS
 };
 
-/* Hardware only sees DPAA2_ETH_RX_BUF_SIZE, but the skb built around
- * the buffer also needs space for its shared info struct, and we need
- * to allocate enough to accommodate hardware alignment restrictions
- */
-static inline unsigned int dpaa2_eth_buf_raw_size(struct dpaa2_eth_priv *priv)
-{
-	return DPAA2_ETH_SKB_SIZE + priv->rx_buf_align;
-}
-
 static inline
 unsigned int dpaa2_eth_needed_headroom(struct dpaa2_eth_priv *priv,
 				       struct sk_buff *skb)
@@ -443,14 +448,7 @@ unsigned int dpaa2_eth_needed_headroom(struct dpaa2_eth_priv *priv,
  */
 static inline unsigned int dpaa2_eth_rx_head_room(struct dpaa2_eth_priv *priv)
 {
-	return priv->tx_data_offset + DPAA2_ETH_TX_BUF_ALIGN -
-	       DPAA2_ETH_RX_HWA_SIZE;
-}
-
-/* We have exactly one {Rx, Tx conf} queue per channel */
-static int dpaa2_eth_queue_count(struct dpaa2_eth_priv *priv)
-{
-	return priv->num_channels;
+	return priv->tx_data_offset - DPAA2_ETH_RX_HWA_SIZE;
 }
 
 int dpaa2_eth_set_hash(struct net_device *net_dev, u64 flags);