1 files changed, 320 insertions, 176 deletions

diff --git a/drivers/net/ethernet/broadcom/bgmac.c b/drivers/net/ethernet/broadcom/bgmac.c
index 0469f72c6e7e..5cb93d1f50a4 100644
--- a/drivers/net/ethernet/broadcom/bgmac.c
+++ b/drivers/net/ethernet/broadcom/bgmac.c
@@ -14,6 +14,7 @@
 #include <linux/etherdevice.h>
 #include <linux/mii.h>
 #include <linux/phy.h>
+#include <linux/phy_fixed.h>
 #include <linux/interrupt.h>
 #include <linux/dma-mapping.h>
 #include <bcm47xx_nvram.h>
@@ -114,54 +115,89 @@ static void bgmac_dma_tx_enable(struct bgmac *bgmac,
 	bgmac_write(bgmac, ring->mmio_base + BGMAC_DMA_TX_CTL, ctl);
 }
 
+static void
+bgmac_dma_tx_add_buf(struct bgmac *bgmac, struct bgmac_dma_ring *ring,
+		     int i, int len, u32 ctl0)
+{
+	struct bgmac_slot_info *slot;
+	struct bgmac_dma_desc *dma_desc;
+	u32 ctl1;
+
+	if (i == BGMAC_TX_RING_SLOTS - 1)
+		ctl0 |= BGMAC_DESC_CTL0_EOT;
+
+	ctl1 = len & BGMAC_DESC_CTL1_LEN;
+
+	slot = &ring->slots[i];
+	dma_desc = &ring->cpu_base[i];
+	dma_desc->addr_low = cpu_to_le32(lower_32_bits(slot->dma_addr));
+	dma_desc->addr_high = cpu_to_le32(upper_32_bits(slot->dma_addr));
+	dma_desc->ctl0 = cpu_to_le32(ctl0);
+	dma_desc->ctl1 = cpu_to_le32(ctl1);
+}
+
 static netdev_tx_t bgmac_dma_tx_add(struct bgmac *bgmac,
 				    struct bgmac_dma_ring *ring,
 				    struct sk_buff *skb)
 {
 	struct device *dma_dev = bgmac->core->dma_dev;
 	struct net_device *net_dev = bgmac->net_dev;
-	struct bgmac_dma_desc *dma_desc;
-	struct bgmac_slot_info *slot;
-	u32 ctl0, ctl1;
-	int free_slots;
+	int index = ring->end % BGMAC_TX_RING_SLOTS;
+	struct bgmac_slot_info *slot = &ring->slots[index];
+	int nr_frags;
+	u32 flags;
+	int i;
 
 	if (skb->len > BGMAC_DESC_CTL1_LEN) {
 		bgmac_err(bgmac, "Too long skb (%d)\n", skb->len);
-		goto err_stop_drop;
+		goto err_drop;
 	}
 
-	if (ring->start <= ring->end)
-		free_slots = ring->start - ring->end + BGMAC_TX_RING_SLOTS;
-	else
-		free_slots = ring->start - ring->end;
-	if (free_slots == 1) {
+	if (skb->ip_summed == CHECKSUM_PARTIAL)
+		skb_checksum_help(skb);
+
+	nr_frags = skb_shinfo(skb)->nr_frags;
+
+	/* ring->end - ring->start will return the number of valid slots,
+	 * even when ring->end overflows
+	 */
+	if (ring->end - ring->start + nr_frags + 1 >= BGMAC_TX_RING_SLOTS) {
 		bgmac_err(bgmac, "TX ring is full, queue should be stopped!\n");
 		netif_stop_queue(net_dev);
 		return NETDEV_TX_BUSY;
 	}
 
-	slot = &ring->slots[ring->end];
-	slot->skb = skb;
-	slot->dma_addr = dma_map_single(dma_dev, skb->data, skb->len,
+	slot->dma_addr = dma_map_single(dma_dev, skb->data, skb_headlen(skb),
 					DMA_TO_DEVICE);
-	if (dma_mapping_error(dma_dev, slot->dma_addr)) {
-		bgmac_err(bgmac, "Mapping error of skb on ring 0x%X\n",
-			  ring->mmio_base);
-		goto err_stop_drop;
-	}
+	if (unlikely(dma_mapping_error(dma_dev, slot->dma_addr)))
+		goto err_dma_head;
 
-	ctl0 = BGMAC_DESC_CTL0_IOC | BGMAC_DESC_CTL0_SOF | BGMAC_DESC_CTL0_EOF;
-	if (ring->end == ring->num_slots - 1)
-		ctl0 |= BGMAC_DESC_CTL0_EOT;
-	ctl1 = skb->len & BGMAC_DESC_CTL1_LEN;
+	flags = BGMAC_DESC_CTL0_SOF;
+	if (!nr_frags)
+		flags |= BGMAC_DESC_CTL0_EOF | BGMAC_DESC_CTL0_IOC;
 
-	dma_desc = ring->cpu_base;
-	dma_desc += ring->end;
-	dma_desc->addr_low = cpu_to_le32(lower_32_bits(slot->dma_addr));
-	dma_desc->addr_high = cpu_to_le32(upper_32_bits(slot->dma_addr));
-	dma_desc->ctl0 = cpu_to_le32(ctl0);
-	dma_desc->ctl1 = cpu_to_le32(ctl1);
+	bgmac_dma_tx_add_buf(bgmac, ring, index, skb_headlen(skb), flags);
+	flags = 0;
+
+	for (i = 0; i < nr_frags; i++) {
+		struct skb_frag_struct *frag = &skb_shinfo(skb)->frags[i];
+		int len = skb_frag_size(frag);
+
+		index = (index + 1) % BGMAC_TX_RING_SLOTS;
+		slot = &ring->slots[index];
+		slot->dma_addr = skb_frag_dma_map(dma_dev, frag, 0,
+						  len, DMA_TO_DEVICE);
+		if (unlikely(dma_mapping_error(dma_dev, slot->dma_addr)))
+			goto err_dma;
+
+		if (i == nr_frags - 1)
+			flags |= BGMAC_DESC_CTL0_EOF | BGMAC_DESC_CTL0_IOC;
+
+		bgmac_dma_tx_add_buf(bgmac, ring, index, len, flags);
+	}
 
+	slot->skb = skb;
+	ring->end += nr_frags + 1;
 	netdev_sent_queue(net_dev, skb->len);
 
 	wmb();
@@ -169,20 +205,34 @@ static netdev_tx_t bgmac_dma_tx_add(struct bgmac *bgmac,
 	/* Increase ring->end to point empty slot. We tell hardware the first
 	 * slot it should *not* read.
 	 */
-	if (++ring->end >= BGMAC_TX_RING_SLOTS)
-		ring->end = 0;
 	bgmac_write(bgmac, ring->mmio_base + BGMAC_DMA_TX_INDEX,
 		    ring->index_base +
-		    ring->end * sizeof(struct bgmac_dma_desc));
+		    (ring->end % BGMAC_TX_RING_SLOTS) *
+		    sizeof(struct bgmac_dma_desc));
 
-	/* Always keep one slot free to allow detecting bugged calls. */
-	if (--free_slots == 1)
+	if (ring->end - ring->start >= BGMAC_TX_RING_SLOTS - 8)
 		netif_stop_queue(net_dev);
 
 	return NETDEV_TX_OK;
 
-err_stop_drop:
-	netif_stop_queue(net_dev);
+err_dma:
+	dma_unmap_single(dma_dev, slot->dma_addr, skb_headlen(skb),
+			 DMA_TO_DEVICE);
+
+	while (i > 0) {
+		int index = (ring->end + i) % BGMAC_TX_RING_SLOTS;
+		struct bgmac_slot_info *slot = &ring->slots[index];
+		u32 ctl1 = le32_to_cpu(ring->cpu_base[index].ctl1);
+		int len = ctl1 & BGMAC_DESC_CTL1_LEN;
+
+		dma_unmap_page(dma_dev, slot->dma_addr, len, DMA_TO_DEVICE);
+	}
+
+err_dma_head:
+	bgmac_err(bgmac, "Mapping error of skb on ring 0x%X\n",
+		  ring->mmio_base);
+
+err_drop:
 	dev_kfree_skb(skb);
 	return NETDEV_TX_OK;
 }
@@ -202,34 +252,45 @@ static void bgmac_dma_tx_free(struct bgmac *bgmac, struct bgmac_dma_ring *ring)
 	empty_slot &= BGMAC_DMA_TX_STATDPTR;
 	empty_slot /= sizeof(struct bgmac_dma_desc);
 
-	while (ring->start != empty_slot) {
-		struct bgmac_slot_info *slot = &ring->slots[ring->start];
+	while (ring->start != ring->end) {
+		int slot_idx = ring->start % BGMAC_TX_RING_SLOTS;
+		struct bgmac_slot_info *slot = &ring->slots[slot_idx];
+		u32 ctl1;
+		int len;
 
-		if (slot->skb) {
+		if (slot_idx == empty_slot)
+			break;
+
+		ctl1 = le32_to_cpu(ring->cpu_base[slot_idx].ctl1);
+		len = ctl1 & BGMAC_DESC_CTL1_LEN;
+		if (ctl1 & BGMAC_DESC_CTL0_SOF)
 			/* Unmap no longer used buffer */
-			dma_unmap_single(dma_dev, slot->dma_addr,
-					 slot->skb->len, DMA_TO_DEVICE);
-			slot->dma_addr = 0;
+			dma_unmap_single(dma_dev, slot->dma_addr, len,
+					 DMA_TO_DEVICE);
+		else
+			dma_unmap_page(dma_dev, slot->dma_addr, len,
+				       DMA_TO_DEVICE);
 
+		if (slot->skb) {
 			bytes_compl += slot->skb->len;
 			pkts_compl++;
 
 			/* Free memory! :) */
 			dev_kfree_skb(slot->skb);
 			slot->skb = NULL;
-		} else {
-			bgmac_err(bgmac, "Hardware reported transmission for empty TX ring slot %d! End of ring: %d\n",
-				  ring->start, ring->end);
 		}
 
-		if (++ring->start >= BGMAC_TX_RING_SLOTS)
-			ring->start = 0;
+		slot->dma_addr = 0;
+		ring->start++;
 		freed = true;
 	}
 
+	if (!pkts_compl)
+		return;
+
 	netdev_completed_queue(bgmac->net_dev, pkts_compl, bytes_compl);
 
-	if (freed && netif_queue_stopped(bgmac->net_dev))
+	if (netif_queue_stopped(bgmac->net_dev))
 		netif_wake_queue(bgmac->net_dev);
 }
 
@@ -275,43 +336,53 @@ static int bgmac_dma_rx_skb_for_slot(struct bgmac *bgmac,
 				     struct bgmac_slot_info *slot)
 {
 	struct device *dma_dev = bgmac->core->dma_dev;
-	struct sk_buff *skb;
 	dma_addr_t dma_addr;
 	struct bgmac_rx_header *rx;
+	void *buf;
 
 	/* Alloc skb */
-	skb = netdev_alloc_skb(bgmac->net_dev, BGMAC_RX_BUF_SIZE);
-	if (!skb)
+	buf = netdev_alloc_frag(BGMAC_RX_ALLOC_SIZE);
+	if (!buf)
 		return -ENOMEM;
 
 	/* Poison - if everything goes fine, hardware will overwrite it */
-	rx = (struct bgmac_rx_header *)skb->data;
+	rx = buf + BGMAC_RX_BUF_OFFSET;
 	rx->len = cpu_to_le16(0xdead);
 	rx->flags = cpu_to_le16(0xbeef);
 
 	/* Map skb for the DMA */
-	dma_addr = dma_map_single(dma_dev, skb->data,
+	dma_addr = dma_map_single(dma_dev, buf + BGMAC_RX_BUF_OFFSET,
 				  BGMAC_RX_BUF_SIZE, DMA_FROM_DEVICE);
 	if (dma_mapping_error(dma_dev, dma_addr)) {
 		bgmac_err(bgmac, "DMA mapping error\n");
-		dev_kfree_skb(skb);
+		put_page(virt_to_head_page(buf));
 		return -ENOMEM;
 	}
 
 	/* Update the slot */
-	slot->skb = skb;
+	slot->buf = buf;
 	slot->dma_addr = dma_addr;
 
 	return 0;
 }
 
+static void bgmac_dma_rx_update_index(struct bgmac *bgmac,
+				      struct bgmac_dma_ring *ring)
+{
+	dma_wmb();
+
+	bgmac_write(bgmac, ring->mmio_base + BGMAC_DMA_RX_INDEX,
+		    ring->index_base +
+		    ring->end * sizeof(struct bgmac_dma_desc));
+}
+
 static void bgmac_dma_rx_setup_desc(struct bgmac *bgmac,
 				    struct bgmac_dma_ring *ring, int desc_idx)
 {
 	struct bgmac_dma_desc *dma_desc = ring->cpu_base + desc_idx;
 	u32 ctl0 = 0, ctl1 = 0;
 
-	if (desc_idx == ring->num_slots - 1)
+	if (desc_idx == BGMAC_RX_RING_SLOTS - 1)
 		ctl0 |= BGMAC_DESC_CTL0_EOT;
 	ctl1 |= BGMAC_RX_BUF_SIZE & BGMAC_DESC_CTL1_LEN;
 	/* Is there any BGMAC device that requires extension? */
@@ -323,6 +394,21 @@ static void bgmac_dma_rx_setup_desc(struct bgmac *bgmac,
 	dma_desc->addr_high = cpu_to_le32(upper_32_bits(ring->slots[desc_idx].dma_addr));
 	dma_desc->ctl0 = cpu_to_le32(ctl0);
 	dma_desc->ctl1 = cpu_to_le32(ctl1);
+
+	ring->end = desc_idx;
+}
+
+static void bgmac_dma_rx_poison_buf(struct device *dma_dev,
+				    struct bgmac_slot_info *slot)
+{
+	struct bgmac_rx_header *rx = slot->buf + BGMAC_RX_BUF_OFFSET;
+
+	dma_sync_single_for_cpu(dma_dev, slot->dma_addr, BGMAC_RX_BUF_SIZE,
+				DMA_FROM_DEVICE);
+	rx->len = cpu_to_le16(0xdead);
+	rx->flags = cpu_to_le16(0xbeef);
+	dma_sync_single_for_device(dma_dev, slot->dma_addr, BGMAC_RX_BUF_SIZE,
+				   DMA_FROM_DEVICE);
 }
 
 static int bgmac_dma_rx_read(struct bgmac *bgmac, struct bgmac_dma_ring *ring,
@@ -337,70 +423,62 @@ static int bgmac_dma_rx_read(struct bgmac *bgmac, struct bgmac_dma_ring *ring,
 	end_slot &= BGMAC_DMA_RX_STATDPTR;
 	end_slot /= sizeof(struct bgmac_dma_desc);
 
-	ring->end = end_slot;
-
-	while (ring->start != ring->end) {
+	while (ring->start != end_slot) {
 		struct device *dma_dev = bgmac->core->dma_dev;
 		struct bgmac_slot_info *slot = &ring->slots[ring->start];
-		struct sk_buff *skb = slot->skb;
-		struct bgmac_rx_header *rx;
+		struct bgmac_rx_header *rx = slot->buf + BGMAC_RX_BUF_OFFSET;
+		struct sk_buff *skb;
+		void *buf = slot->buf;
+		dma_addr_t dma_addr = slot->dma_addr;
 		u16 len, flags;
 
-		/* Unmap buffer to make it accessible to the CPU */
-		dma_sync_single_for_cpu(dma_dev, slot->dma_addr,
-					BGMAC_RX_BUF_SIZE, DMA_FROM_DEVICE);
+		do {
+			/* Prepare new skb as replacement */
+			if (bgmac_dma_rx_skb_for_slot(bgmac, slot)) {
+				bgmac_dma_rx_poison_buf(dma_dev, slot);
+				break;
+			}
 
-		/* Get info from the header */
-		rx = (struct bgmac_rx_header *)skb->data;
-		len = le16_to_cpu(rx->len);
-		flags = le16_to_cpu(rx->flags);
+			/* Unmap buffer to make it accessible to the CPU */
+			dma_unmap_single(dma_dev, dma_addr,
+					 BGMAC_RX_BUF_SIZE, DMA_FROM_DEVICE);
 
-		do {
-			dma_addr_t old_dma_addr = slot->dma_addr;
-			int err;
+			/* Get info from the header */
+			len = le16_to_cpu(rx->len);
+			flags = le16_to_cpu(rx->flags);
 
 			/* Check for poison and drop or pass the packet */
 			if (len == 0xdead && flags == 0xbeef) {
 				bgmac_err(bgmac, "Found poisoned packet at slot %d, DMA issue!\n",
 					  ring->start);
-				dma_sync_single_for_device(dma_dev,
-							   slot->dma_addr,
-							   BGMAC_RX_BUF_SIZE,
-							   DMA_FROM_DEVICE);
+				put_page(virt_to_head_page(buf));
 				break;
 			}
 
-			/* Omit CRC. */
-			len -= ETH_FCS_LEN;
-
-			/* Prepare new skb as replacement */
-			err = bgmac_dma_rx_skb_for_slot(bgmac, slot);
-			if (err) {
-				/* Poison the old skb */
-				rx->len = cpu_to_le16(0xdead);
-				rx->flags = cpu_to_le16(0xbeef);
-
-				dma_sync_single_for_device(dma_dev,
-							   slot->dma_addr,
-							   BGMAC_RX_BUF_SIZE,
-							   DMA_FROM_DEVICE);
+			if (len > BGMAC_RX_ALLOC_SIZE) {
+				bgmac_err(bgmac, "Found oversized packet at slot %d, DMA issue!\n",
+					  ring->start);
+				put_page(virt_to_head_page(buf));
 				break;
 			}
-			bgmac_dma_rx_setup_desc(bgmac, ring, ring->start);
 
-			/* Unmap old skb, we'll pass it to the netfif */
-			dma_unmap_single(dma_dev, old_dma_addr,
-					 BGMAC_RX_BUF_SIZE, DMA_FROM_DEVICE);
+			/* Omit CRC. */
+			len -= ETH_FCS_LEN;
 
-			skb_put(skb, BGMAC_RX_FRAME_OFFSET + len);
-			skb_pull(skb, BGMAC_RX_FRAME_OFFSET);
+			skb = build_skb(buf, BGMAC_RX_ALLOC_SIZE);
+			skb_put(skb, BGMAC_RX_FRAME_OFFSET +
+				BGMAC_RX_BUF_OFFSET + len);
+			skb_pull(skb, BGMAC_RX_FRAME_OFFSET +
+				 BGMAC_RX_BUF_OFFSET);
 
 			skb_checksum_none_assert(skb);
 			skb->protocol = eth_type_trans(skb, bgmac->net_dev);
-			netif_receive_skb(skb);
+			napi_gro_receive(&bgmac->napi, skb);
 			handled++;
 		} while (0);
 
+		bgmac_dma_rx_setup_desc(bgmac, ring, ring->start);
+
 		if (++ring->start >= BGMAC_RX_RING_SLOTS)
 			ring->start = 0;
 
@@ -408,6 +486,8 @@ static int bgmac_dma_rx_read(struct bgmac *bgmac, struct bgmac_dma_ring *ring,
 			break;
 	}
 
+	bgmac_dma_rx_update_index(bgmac, ring);
+
 	return handled;
 }
 
@@ -433,40 +513,90 @@ static bool bgmac_dma_unaligned(struct bgmac *bgmac,
 	return false;
 }
 
-static void bgmac_dma_ring_free(struct bgmac *bgmac,
-				struct bgmac_dma_ring *ring)
+static void bgmac_dma_tx_ring_free(struct bgmac *bgmac,
+				   struct bgmac_dma_ring *ring)
 {
 	struct device *dma_dev = bgmac->core->dma_dev;
+	struct bgmac_dma_desc *dma_desc = ring->cpu_base;
 	struct bgmac_slot_info *slot;
-	int size;
 	int i;
 
-	for (i = 0; i < ring->num_slots; i++) {
+	for (i = 0; i < BGMAC_TX_RING_SLOTS; i++) {
+		int len = dma_desc[i].ctl1 & BGMAC_DESC_CTL1_LEN;
+
 		slot = &ring->slots[i];
-		if (slot->skb) {
-			if (slot->dma_addr)
-				dma_unmap_single(dma_dev, slot->dma_addr,
-						 slot->skb->len, DMA_TO_DEVICE);
-			dev_kfree_skb(slot->skb);
-		}
+		dev_kfree_skb(slot->skb);
+
+		if (!slot->dma_addr)
+			continue;
+
+		if (slot->skb)
+			dma_unmap_single(dma_dev, slot->dma_addr,
+					 len, DMA_TO_DEVICE);
+		else
+			dma_unmap_page(dma_dev, slot->dma_addr,
+				       len, DMA_TO_DEVICE);
 	}
+}
 
-	if (ring->cpu_base) {
-		/* Free ring of descriptors */
-		size = ring->num_slots * sizeof(struct bgmac_dma_desc);
-		dma_free_coherent(dma_dev, size, ring->cpu_base,
-				  ring->dma_base);
+static void bgmac_dma_rx_ring_free(struct bgmac *bgmac,
+				   struct bgmac_dma_ring *ring)
+{
+	struct device *dma_dev = bgmac->core->dma_dev;
+	struct bgmac_slot_info *slot;
+	int i;
+
+	for (i = 0; i < BGMAC_RX_RING_SLOTS; i++) {
+		slot = &ring->slots[i];
+		if (!slot->dma_addr)
+			continue;
+
+		dma_unmap_single(dma_dev, slot->dma_addr,
+				 BGMAC_RX_BUF_SIZE,
+				 DMA_FROM_DEVICE);
+		put_page(virt_to_head_page(slot->buf));
+		slot->dma_addr = 0;
 	}
 }
 
+static void bgmac_dma_ring_desc_free(struct bgmac *bgmac,
+				     struct bgmac_dma_ring *ring,
+				     int num_slots)
+{
+	struct device *dma_dev = bgmac->core->dma_dev;
+	int size;
+
+	if (!ring->cpu_base)
+	    return;
+
+	/* Free ring of descriptors */
+	size = num_slots * sizeof(struct bgmac_dma_desc);
+	dma_free_coherent(dma_dev, size, ring->cpu_base,
+			  ring->dma_base);
+}
+
+static void bgmac_dma_cleanup(struct bgmac *bgmac)
+{
+	int i;
+
+	for (i = 0; i < BGMAC_MAX_TX_RINGS; i++)
+		bgmac_dma_tx_ring_free(bgmac, &bgmac->tx_ring[i]);
+
+	for (i = 0; i < BGMAC_MAX_RX_RINGS; i++)
+		bgmac_dma_rx_ring_free(bgmac, &bgmac->rx_ring[i]);
+}
+
 static void bgmac_dma_free(struct bgmac *bgmac)
 {
 	int i;
 
 	for (i = 0; i < BGMAC_MAX_TX_RINGS; i++)
-		bgmac_dma_ring_free(bgmac, &bgmac->tx_ring[i]);
+		bgmac_dma_ring_desc_free(bgmac, &bgmac->tx_ring[i],
+					 BGMAC_TX_RING_SLOTS);
+
 	for (i = 0; i < BGMAC_MAX_RX_RINGS; i++)
-		bgmac_dma_ring_free(bgmac, &bgmac->rx_ring[i]);
+		bgmac_dma_ring_desc_free(bgmac, &bgmac->rx_ring[i],
+					 BGMAC_RX_RING_SLOTS);
 }
 
 static int bgmac_dma_alloc(struct bgmac *bgmac)
@@ -489,11 +619,10 @@ static int bgmac_dma_alloc(struct bgmac *bgmac)
 
 	for (i = 0; i < BGMAC_MAX_TX_RINGS; i++) {
 		ring = &bgmac->tx_ring[i];
-		ring->num_slots = BGMAC_TX_RING_SLOTS;
 		ring->mmio_base = ring_base[i];
 
 		/* Alloc ring of descriptors */
-		size = ring->num_slots * sizeof(struct bgmac_dma_desc);
+		size = BGMAC_TX_RING_SLOTS * sizeof(struct bgmac_dma_desc);
 		ring->cpu_base = dma_zalloc_coherent(dma_dev, size,
 						     &ring->dma_base,
 						     GFP_KERNEL);
@@ -514,14 +643,11 @@ static int bgmac_dma_alloc(struct bgmac *bgmac)
 	}
 
 	for (i = 0; i < BGMAC_MAX_RX_RINGS; i++) {
-		int j;
-
 		ring = &bgmac->rx_ring[i];
-		ring->num_slots = BGMAC_RX_RING_SLOTS;
 		ring->mmio_base = ring_base[i];
 
 		/* Alloc ring of descriptors */
-		size = ring->num_slots * sizeof(struct bgmac_dma_desc);
+		size = BGMAC_RX_RING_SLOTS * sizeof(struct bgmac_dma_desc);
 		ring->cpu_base = dma_zalloc_coherent(dma_dev, size,
 						     &ring->dma_base,
 						     GFP_KERNEL);
@@ -538,15 +664,6 @@ static int bgmac_dma_alloc(struct bgmac *bgmac)
 			ring->index_base = lower_32_bits(ring->dma_base);
 		else
 			ring->index_base = 0;
-
-		/* Alloc RX slots */
-		for (j = 0; j < ring->num_slots; j++) {
-			err = bgmac_dma_rx_skb_for_slot(bgmac, &ring->slots[j]);
-			if (err) {
-				bgmac_err(bgmac, "Can't allocate skb for slot in RX ring\n");
-				goto err_dma_free;
-			}
-		}
 	}
 
 	return 0;
@@ -556,10 +673,10 @@ err_dma_free:
 	return -ENOMEM;
 }
 
-static void bgmac_dma_init(struct bgmac *bgmac)
+static int bgmac_dma_init(struct bgmac *bgmac)
 {
 	struct bgmac_dma_ring *ring;
-	int i;
+	int i, err;
 
 	for (i = 0; i < BGMAC_MAX_TX_RINGS; i++) {
 		ring = &bgmac->tx_ring[i];
@@ -591,16 +708,24 @@ static void bgmac_dma_init(struct bgmac *bgmac)
 		if (ring->unaligned)
 			bgmac_dma_rx_enable(bgmac, ring);
 
-		for (j = 0; j < ring->num_slots; j++)
-			bgmac_dma_rx_setup_desc(bgmac, ring, j);
-
-		bgmac_write(bgmac, ring->mmio_base + BGMAC_DMA_RX_INDEX,
-			    ring->index_base +
-			    ring->num_slots * sizeof(struct bgmac_dma_desc));
-
 		ring->start = 0;
 		ring->end = 0;
+		for (j = 0; j < BGMAC_RX_RING_SLOTS; j++) {
+			err = bgmac_dma_rx_skb_for_slot(bgmac, &ring->slots[j]);
+			if (err)
+				goto error;
+
+			bgmac_dma_rx_setup_desc(bgmac, ring, j);
+		}
+
+		bgmac_dma_rx_update_index(bgmac, ring);
 	}
+
+	return 0;
+
+error:
+	bgmac_dma_cleanup(bgmac);
+	return err;
 }
 
 /**************************************************
@@ -1008,8 +1133,6 @@ static void bgmac_chip_reset(struct bgmac *bgmac)
 	bgmac_phy_init(bgmac);
 
 	netdev_reset_queue(bgmac->net_dev);
-
-	bgmac->int_status = 0;
 }
 
 static void bgmac_chip_intrs_on(struct bgmac *bgmac)
@@ -1078,11 +1201,8 @@ static void bgmac_enable(struct bgmac *bgmac)
 }
 
 /* http://bcm-v4.sipsolutions.net/mac-gbit/gmac/chipinit */
-static void bgmac_chip_init(struct bgmac *bgmac, bool full_init)
+static void bgmac_chip_init(struct bgmac *bgmac)
 {
-	struct bgmac_dma_ring *ring;
-	int i;
-
 	/* 1 interrupt per received frame */
 	bgmac_write(bgmac, BGMAC_INT_RECV_LAZY, 1 << BGMAC_IRL_FC_SHIFT);
 
@@ -1100,16 +1220,7 @@ static void bgmac_chip_init(struct bgmac *bgmac, bool full_init)
 
 	bgmac_write(bgmac, BGMAC_RXMAX_LENGTH, 32 + ETHER_MAX_LEN);
 
-	if (full_init) {
-		bgmac_dma_init(bgmac);
-		if (1) /* FIXME: is there any case we don't want IRQs? */
-			bgmac_chip_intrs_on(bgmac);
-	} else {
-		for (i = 0; i < BGMAC_MAX_RX_RINGS; i++) {
-			ring = &bgmac->rx_ring[i];
-			bgmac_dma_rx_enable(bgmac, ring);
-		}
-	}
+	bgmac_chip_intrs_on(bgmac);
 
 	bgmac_enable(bgmac);
 }
@@ -1124,14 +1235,13 @@ static irqreturn_t bgmac_interrupt(int irq, void *dev_id)
 	if (!int_status)
 		return IRQ_NONE;
 
-	/* Ack */
-	bgmac_write(bgmac, BGMAC_INT_STATUS, int_status);
+	int_status &= ~(BGMAC_IS_TX0 | BGMAC_IS_RX);
+	if (int_status)
+		bgmac_err(bgmac, "Unknown IRQs: 0x%08X\n", int_status);
 
 	/* Disable new interrupts until handling existing ones */
 	bgmac_chip_intrs_off(bgmac);
 
-	bgmac->int_status = int_status;
-
 	napi_schedule(&bgmac->napi);
 
 	return IRQ_HANDLED;
@@ -1140,25 +1250,17 @@ static irqreturn_t bgmac_interrupt(int irq, void *dev_id)
 static int bgmac_poll(struct napi_struct *napi, int weight)
 {
 	struct bgmac *bgmac = container_of(napi, struct bgmac, napi);
-	struct bgmac_dma_ring *ring;
 	int handled = 0;
 
-	if (bgmac->int_status & BGMAC_IS_TX0) {
-		ring = &bgmac->tx_ring[0];
-		bgmac_dma_tx_free(bgmac, ring);
-		bgmac->int_status &= ~BGMAC_IS_TX0;
-	}
+	/* Ack */
+	bgmac_write(bgmac, BGMAC_INT_STATUS, ~0);
 
-	if (bgmac->int_status & BGMAC_IS_RX) {
-		ring = &bgmac->rx_ring[0];
-		handled += bgmac_dma_rx_read(bgmac, ring, weight);
-		bgmac->int_status &= ~BGMAC_IS_RX;
-	}
+	bgmac_dma_tx_free(bgmac, &bgmac->tx_ring[0]);
+	handled += bgmac_dma_rx_read(bgmac, &bgmac->rx_ring[0], weight);
 
-	if (bgmac->int_status) {
-		bgmac_err(bgmac, "Unknown IRQs: 0x%08X\n", bgmac->int_status);
-		bgmac->int_status = 0;
-	}
+	/* Poll again if more events arrived in the meantime */
+	if (bgmac_read(bgmac, BGMAC_INT_STATUS) & (BGMAC_IS_TX0 | BGMAC_IS_RX))
+		return handled;
 
 	if (handled < weight) {
 		napi_complete(napi);
@@ -1178,23 +1280,27 @@ static int bgmac_open(struct net_device *net_dev)
 	int err = 0;
 
 	bgmac_chip_reset(bgmac);
+
+	err = bgmac_dma_init(bgmac);
+	if (err)
+		return err;
+
 	/* Specs say about reclaiming rings here, but we do that in DMA init */
-	bgmac_chip_init(bgmac, true);
+	bgmac_chip_init(bgmac);
 
 	err = request_irq(bgmac->core->irq, bgmac_interrupt, IRQF_SHARED,
 			  KBUILD_MODNAME, net_dev);
 	if (err < 0) {
 		bgmac_err(bgmac, "IRQ request error: %d!\n", err);
-		goto err_out;
+		bgmac_dma_cleanup(bgmac);
+		return err;
 	}
 	napi_enable(&bgmac->napi);
 
 	phy_start(bgmac->phy_dev);
 
 	netif_carrier_on(net_dev);
-
-err_out:
-	return err;
+	return 0;
 }
 
 static int bgmac_stop(struct net_device *net_dev)
@@ -1210,6 +1316,7 @@ static int bgmac_stop(struct net_device *net_dev)
 	free_irq(bgmac->core->irq, net_dev);
 
 	bgmac_chip_reset(bgmac);
+	bgmac_dma_cleanup(bgmac);
 
 	return 0;
 }
@@ -1330,13 +1437,46 @@ static void bgmac_adjust_link(struct net_device *net_dev)
 	}
 }
 
+static int bgmac_fixed_phy_register(struct bgmac *bgmac)
+{
+	struct fixed_phy_status fphy_status = {
+		.link = 1,
+		.speed = SPEED_1000,
+		.duplex = DUPLEX_FULL,
+	};
+	struct phy_device *phy_dev;
+	int err;
+
+	phy_dev = fixed_phy_register(PHY_POLL, &fphy_status, NULL);
+	if (!phy_dev || IS_ERR(phy_dev)) {
+		bgmac_err(bgmac, "Failed to register fixed PHY device\n");
+		return -ENODEV;
+	}
+
+	err = phy_connect_direct(bgmac->net_dev, phy_dev, bgmac_adjust_link,
+				 PHY_INTERFACE_MODE_MII);
+	if (err) {
+		bgmac_err(bgmac, "Connecting PHY failed\n");
+		return err;
+	}
+
+	bgmac->phy_dev = phy_dev;
+
+	return err;
+}
+
 static int bgmac_mii_register(struct bgmac *bgmac)
 {
+	struct bcma_chipinfo *ci = &bgmac->core->bus->chipinfo;
 	struct mii_bus *mii_bus;
 	struct phy_device *phy_dev;
 	char bus_id[MII_BUS_ID_SIZE + 3];
 	int i, err = 0;
 
+	if (ci->id == BCMA_CHIP_ID_BCM4707 ||
+	    ci->id == BCMA_CHIP_ID_BCM53018)
+		return bgmac_fixed_phy_register(bgmac);
+
 	mii_bus = mdiobus_alloc();
 	if (!mii_bus)
 		return -ENOMEM;
@@ -1517,6 +1657,10 @@ static int bgmac_probe(struct bcma_device *core)
 		goto err_dma_free;
 	}
 
+	net_dev->features = NETIF_F_SG | NETIF_F_IP_CSUM | NETIF_F_IPV6_CSUM;
+	net_dev->hw_features = net_dev->features;
+	net_dev->vlan_features = net_dev->features;
+
 	err = register_netdev(bgmac->net_dev);
 	if (err) {
 		bgmac_err(bgmac, "Cannot register net device\n");