Introduce xdp tx mb support for i40e and ixgbe

index c4a4954aa317..0dd5b2687609 100644
@@ -8571,57 +8571,84 @@ static u16 ixgbe_select_queue(struct net_device *dev, struct sk_buff *skb,
 int ixgbe_xmit_xdp_ring(struct ixgbe_ring *ring,
 			struct xdp_frame *xdpf)
 {
-	struct ixgbe_tx_buffer *tx_buffer;
-	union ixgbe_adv_tx_desc *tx_desc;
-	u32 len, cmd_type;
-	dma_addr_t dma;
-	u16 i;
+	struct skb_shared_info *sinfo = xdp_get_shared_info_from_frame(xdpf);
+	u8 nr_frags = unlikely(xdp_frame_has_frags(xdpf)) ? sinfo->nr_frags : 0;
+	u16 i = 0, index = ring->next_to_use;
+	struct ixgbe_tx_buffer *tx_head = &ring->tx_buffer_info[index];
+	struct ixgbe_tx_buffer *tx_buff = tx_head;
+	union ixgbe_adv_tx_desc *tx_desc = IXGBE_TX_DESC(ring, index);
+	u32 cmd_type, len = xdpf->len;
+	void *data = xdpf->data;
+
+	if (unlikely(ixgbe_desc_unused(ring) < 1 + nr_frags))
+		return IXGBE_XDP_CONSUMED;
 
-	len = xdpf->len;
+	tx_head->bytecount = xdp_get_frame_len(xdpf);
+	tx_head->gso_segs = 1;
+	tx_head->xdpf = xdpf;
 
-	if (unlikely(!ixgbe_desc_unused(ring)))
-		return IXGBE_XDP_CONSUMED;
+	tx_desc->read.olinfo_status =
+		cpu_to_le32(tx_head->bytecount << IXGBE_ADVTXD_PAYLEN_SHIFT);
 
-	dma = dma_map_single(ring->dev, xdpf->data, len, DMA_TO_DEVICE);
-	if (dma_mapping_error(ring->dev, dma))
-		return IXGBE_XDP_CONSUMED;
+	for (;;) {
+		dma_addr_t dma;
 
-	/* record the location of the first descriptor for this packet */
-	tx_buffer = &ring->tx_buffer_info[ring->next_to_use];
-	tx_buffer->bytecount = len;
-	tx_buffer->gso_segs = 1;
-	tx_buffer->protocol = 0;
+		dma = dma_map_single(ring->dev, data, len, DMA_TO_DEVICE);
+		if (dma_mapping_error(ring->dev, dma))
+			goto unmap;
+
+		dma_unmap_len_set(tx_buff, len, len);
+		dma_unmap_addr_set(tx_buff, dma, dma);
 
-	i = ring->next_to_use;
-	tx_desc = IXGBE_TX_DESC(ring, i);
+		cmd_type = IXGBE_ADVTXD_DTYP_DATA | IXGBE_ADVTXD_DCMD_DEXT |
+			   IXGBE_ADVTXD_DCMD_IFCS | len;
+		tx_desc->read.cmd_type_len = cpu_to_le32(cmd_type);
+		tx_desc->read.buffer_addr = cpu_to_le64(dma);
+		tx_buff->protocol = 0;
 
-	dma_unmap_len_set(tx_buffer, len, len);
-	dma_unmap_addr_set(tx_buffer, dma, dma);
-	tx_buffer->xdpf = xdpf;
+		if (++index == ring->count)
+			index = 0;
 
-	tx_desc->read.buffer_addr = cpu_to_le64(dma);
+		if (i == nr_frags)
+			break;
+
+		tx_buff = &ring->tx_buffer_info[index];
+		tx_desc = IXGBE_TX_DESC(ring, index);
+
+		tx_desc->read.olinfo_status = 0;
 
+		data = skb_frag_address(&sinfo->frags[i]);
+		len = skb_frag_size(&sinfo->frags[i]);
+		i++;
+	}
 	/* put descriptor type bits */
-	cmd_type = IXGBE_ADVTXD_DTYP_DATA |
-		   IXGBE_ADVTXD_DCMD_DEXT |
-		   IXGBE_ADVTXD_DCMD_IFCS;
-	cmd_type |= len | IXGBE_TXD_CMD;
-	tx_desc->read.cmd_type_len = cpu_to_le32(cmd_type);
-	tx_desc->read.olinfo_status =
-		cpu_to_le32(len << IXGBE_ADVTXD_PAYLEN_SHIFT);
+	tx_desc->read.cmd_type_len |= cpu_to_le32(IXGBE_TXD_CMD);
 
 	/* Avoid any potential race with xdp_xmit and cleanup */
 	smp_wmb();
 
-	/* set next_to_watch value indicating a packet is present */
-	i++;
-	if (i == ring->count)
-		i = 0;
-
-	tx_buffer->next_to_watch = tx_desc;
-	ring->next_to_use = i;
+	tx_head->next_to_watch = tx_desc;
+	ring->next_to_use = index;
 
 	return IXGBE_XDP_TX;
+
+unmap:
+	for (;;) {
+		tx_buff = &ring->tx_buffer_info[index];
+		if (dma_unmap_len(tx_buff, len))
+			dma_unmap_page(ring->dev, dma_unmap_addr(tx_buff, dma),
+				       dma_unmap_len(tx_buff, len),
+				       DMA_TO_DEVICE);
+		dma_unmap_len_set(tx_buff, len, 0);
+		if (tx_buff == tx_head)
+			break;
+
+		if (!index)
+			index += ring->count;
+		index--;
+	}
+
+	return IXGBE_XDP_CONSUMED;
 }
 
 netdev_tx_t ixgbe_xmit_frame_ring(struct sk_buff *skb,

index 0eae5858f2fe..e9a772db869c 100644
@@ -3688,35 +3688,55 @@ u16 i40e_lan_select_queue(struct net_device *netdev,
 static int i40e_xmit_xdp_ring(struct xdp_frame *xdpf,
 			      struct i40e_ring *xdp_ring)
 {
-	u16 i = xdp_ring->next_to_use;
-	struct i40e_tx_buffer *tx_bi;
-	struct i40e_tx_desc *tx_desc;
+	struct skb_shared_info *sinfo = xdp_get_shared_info_from_frame(xdpf);
+	u8 nr_frags = unlikely(xdp_frame_has_frags(xdpf)) ? sinfo->nr_frags : 0;
+	u16 i = 0, index = xdp_ring->next_to_use;
+	struct i40e_tx_buffer *tx_head = &xdp_ring->tx_bi[index];
+	struct i40e_tx_buffer *tx_bi = tx_head;
+	struct i40e_tx_desc *tx_desc = I40E_TX_DESC(xdp_ring, index);
 	void *data = xdpf->data;
 	u32 size = xdpf->len;
-	dma_addr_t dma;
 
-	if (!unlikely(I40E_DESC_UNUSED(xdp_ring))) {
+	if (unlikely(I40E_DESC_UNUSED(xdp_ring) < 1 + nr_frags)) {
 		xdp_ring->tx_stats.tx_busy++;
 		return I40E_XDP_CONSUMED;
 	}
-	dma = dma_map_single(xdp_ring->dev, data, size, DMA_TO_DEVICE);
-	if (dma_mapping_error(xdp_ring->dev, dma))
-		return I40E_XDP_CONSUMED;
 
-	tx_bi = &xdp_ring->tx_bi[i];
-	tx_bi->bytecount = size;
-	tx_bi->gso_segs = 1;
-	tx_bi->xdpf = xdpf;
+	tx_head->bytecount = xdp_get_frame_len(xdpf);
+	tx_head->gso_segs = 1;
+	tx_head->xdpf = xdpf;
 
-	/* record length, and DMA address */
-	dma_unmap_len_set(tx_bi, len, size);
-	dma_unmap_addr_set(tx_bi, dma, dma);
+	for (;;) {
+		dma_addr_t dma;
 
-	tx_desc = I40E_TX_DESC(xdp_ring, i);
-	tx_desc->buffer_addr = cpu_to_le64(dma);
-	tx_desc->cmd_type_offset_bsz = build_ctob(I40E_TX_DESC_CMD_ICRC
-						  | I40E_TXD_CMD,
-						  0, size, 0);
+		dma = dma_map_single(xdp_ring->dev, data, size, DMA_TO_DEVICE);
+		if (dma_mapping_error(xdp_ring->dev, dma))
+			goto unmap;
+
+		/* record length, and DMA address */
+		dma_unmap_len_set(tx_bi, len, size);
+		dma_unmap_addr_set(tx_bi, dma, dma);
+
+		tx_desc->buffer_addr = cpu_to_le64(dma);
+		tx_desc->cmd_type_offset_bsz =
+			build_ctob(I40E_TX_DESC_CMD_ICRC, 0, size, 0);
+
+		if (++index == xdp_ring->count)
+			index = 0;
+
+		if (i == nr_frags)
+			break;
+
+		tx_bi = &xdp_ring->tx_bi[index];
+		tx_desc = I40E_TX_DESC(xdp_ring, index);
+
+		data = skb_frag_address(&sinfo->frags[i]);
+		size = skb_frag_size(&sinfo->frags[i]);
+		i++;
+	}
+
+	tx_desc->cmd_type_offset_bsz |=
+		cpu_to_le64(I40E_TXD_CMD << I40E_TXD_QW1_CMD_SHIFT);
 
 	/* Make certain all of the status bits have been updated
 	 * before next_to_watch is written.

@@ -3724,14 +3744,30 @@ static int i40e_xmit_xdp_ring(struct xdp_frame *xdpf,
 	smp_wmb();
 
 	xdp_ring->xdp_tx_active++;
-	i++;
-	if (i == xdp_ring->count)
-		i = 0;
 
-	tx_bi->next_to_watch = tx_desc;
-	xdp_ring->next_to_use = i;
+	tx_head->next_to_watch = tx_desc;
+	xdp_ring->next_to_use = index;
 
 	return I40E_XDP_TX;
+
+unmap:
+	for (;;) {
+		tx_bi = &xdp_ring->tx_bi[index];
+		if (dma_unmap_len(tx_bi, len))
+			dma_unmap_page(xdp_ring->dev,
+				       dma_unmap_addr(tx_bi, dma),
+				       dma_unmap_len(tx_bi, len),
+				       DMA_TO_DEVICE);
+		dma_unmap_len_set(tx_bi, len, 0);
+		if (tx_bi == tx_head)
+			break;
+
+		if (!index)
+			index += xdp_ring->count;
+		index--;
+	}
+
+	return I40E_XDP_CONSUMED;
 }
 
 /**