diff --git a/drivers/net/ethernet/amazon/ena/ena_netdev.c b/drivers/net/ethernet/amazon/ena/ena_netdev.c
index 71c8cac6e44e..7e40d14682f7 100644
--- a/drivers/net/ethernet/amazon/ena/ena_netdev.c
+++ b/drivers/net/ethernet/amazon/ena/ena_netdev.c
@@ -2236,7 +2236,7 @@ static netdev_tx_t ena_start_xmit(struct sk_buff *skb, struct net_device *dev)
 		}
 	}
 
-	if (netif_xmit_stopped(txq) || !skb->xmit_more) {
+	if (netif_xmit_stopped(txq) || !netdev_xmit_more()) {
 		/* trigger the dma engine. ena_com_write_sq_doorbell()
 		 * has a mb
 		 */
diff --git a/drivers/net/ethernet/amd/xgbe/xgbe-dev.c b/drivers/net/ethernet/amd/xgbe/xgbe-dev.c
index 4666084eda16..d5fd49dd25f3 100644
--- a/drivers/net/ethernet/amd/xgbe/xgbe-dev.c
+++ b/drivers/net/ethernet/amd/xgbe/xgbe-dev.c
@@ -1887,7 +1887,7 @@ static void xgbe_dev_xmit(struct xgbe_channel *channel)
 	smp_wmb();
 
 	ring->cur = cur_index + 1;
-	if (!packet->skb->xmit_more ||
+	if (!netdev_xmit_more() ||
 	    netif_xmit_stopped(netdev_get_tx_queue(pdata->netdev,
 						   channel->queue_index)))
 		xgbe_tx_start_xmit(channel, ring);
diff --git a/drivers/net/ethernet/aurora/nb8800.c b/drivers/net/ethernet/aurora/nb8800.c
index 6f56276015a4..f62deeb6e941 100644
--- a/drivers/net/ethernet/aurora/nb8800.c
+++ b/drivers/net/ethernet/aurora/nb8800.c
@@ -404,6 +404,7 @@ static int nb8800_xmit(struct sk_buff *skb, struct net_device *dev)
 	unsigned int dma_len;
 	unsigned int align;
 	unsigned int next;
+	bool xmit_more;
 
 	if (atomic_read(&priv->tx_free) <= NB8800_DESC_LOW) {
 		netif_stop_queue(dev);
@@ -423,9 +424,10 @@ static int nb8800_xmit(struct sk_buff *skb, struct net_device *dev)
 		return NETDEV_TX_OK;
 	}
 
+	xmit_more = netdev_xmit_more();
 	if (atomic_dec_return(&priv->tx_free) <= NB8800_DESC_LOW) {
 		netif_stop_queue(dev);
-		skb->xmit_more = 0;
+		xmit_more = false;
 	}
 
 	next = priv->tx_next;
@@ -450,7 +452,7 @@ static int nb8800_xmit(struct sk_buff *skb, struct net_device *dev)
 	desc->n_addr = priv->tx_bufs[next].dma_desc;
 	desc->config = DESC_BTS(2) | DESC_DS | DESC_EOF | dma_len;
 
-	if (!skb->xmit_more)
+	if (!xmit_more)
 		desc->config |= DESC_EOC;
 
 	txb->skb = skb;
@@ -468,7 +470,7 @@ static int nb8800_xmit(struct sk_buff *skb, struct net_device *dev)
 
 	priv->tx_next = next;
 
-	if (!skb->xmit_more) {
+	if (!xmit_more) {
 		smp_wmb();
 		priv->tx_chain->ready = true;
 		priv->tx_chain = NULL;
diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt.c b/drivers/net/ethernet/broadcom/bnxt/bnxt.c
index 35e34e23ba33..d22691403d28 100644
--- a/drivers/net/ethernet/broadcom/bnxt/bnxt.c
+++ b/drivers/net/ethernet/broadcom/bnxt/bnxt.c
@@ -551,7 +551,7 @@ normal_tx:
 	prod = NEXT_TX(prod);
 	txr->tx_prod = prod;
 
-	if (!skb->xmit_more || netif_xmit_stopped(txq))
+	if (!netdev_xmit_more() || netif_xmit_stopped(txq))
 		bnxt_db_write(bp, &txr->tx_db, prod);
 
 tx_done:
@@ -559,7 +559,7 @@ tx_done:
 	mmiowb();
 
 	if (unlikely(bnxt_tx_avail(bp, txr) <= MAX_SKB_FRAGS + 1)) {
-		if (skb->xmit_more && !tx_buf->is_push)
+		if (netdev_xmit_more() && !tx_buf->is_push)
 			bnxt_db_write(bp, &txr->tx_db, prod);
 
 		netif_tx_stop_queue(txq);
diff --git a/drivers/net/ethernet/broadcom/genet/bcmgenet.c b/drivers/net/ethernet/broadcom/genet/bcmgenet.c
index 983245c0867c..4fd973571e4c 100644
--- a/drivers/net/ethernet/broadcom/genet/bcmgenet.c
+++ b/drivers/net/ethernet/broadcom/genet/bcmgenet.c
@@ -1665,7 +1665,7 @@ static netdev_tx_t bcmgenet_xmit(struct sk_buff *skb, struct net_device *dev)
 	if (ring->free_bds <= (MAX_SKB_FRAGS + 1))
 		netif_tx_stop_queue(txq);
 
-	if (!skb->xmit_more || netif_xmit_stopped(txq))
+	if (!netdev_xmit_more() || netif_xmit_stopped(txq))
 		/* Packets are ready, update producer index */
 		bcmgenet_tdma_ring_writel(priv, ring->index,
 					  ring->prod_index, TDMA_PROD_INDEX);
diff --git a/drivers/net/ethernet/broadcom/tg3.c b/drivers/net/ethernet/broadcom/tg3.c
index 328373e0578f..45ccadee02af 100644
--- a/drivers/net/ethernet/broadcom/tg3.c
+++ b/drivers/net/ethernet/broadcom/tg3.c
@@ -8156,7 +8156,7 @@ static netdev_tx_t tg3_start_xmit(struct sk_buff *skb, struct net_device *dev)
 			netif_tx_wake_queue(txq);
 	}
 
-	if (!skb->xmit_more || netif_xmit_stopped(txq)) {
+	if (!netdev_xmit_more() || netif_xmit_stopped(txq)) {
 		/* Packets are ready, update Tx producer idx on card. */
 		tw32_tx_mbox(tnapi->prodmbox, entry);
 		mmiowb();
diff --git a/drivers/net/ethernet/cavium/liquidio/lio_main.c b/drivers/net/ethernet/cavium/liquidio/lio_main.c
index fb6f813cff65..eab805579f96 100644
--- a/drivers/net/ethernet/cavium/liquidio/lio_main.c
+++ b/drivers/net/ethernet/cavium/liquidio/lio_main.c
@@ -2522,7 +2522,7 @@ static netdev_tx_t liquidio_xmit(struct sk_buff *skb, struct net_device *netdev)
 		irh->vlan = skb_vlan_tag_get(skb) & 0xfff;
 	}
 
-	xmit_more = skb->xmit_more;
+	xmit_more = netdev_xmit_more();
 
 	if (unlikely(cmdsetup.s.timestamp))
 		status = send_nic_timestamp_pkt(oct, &ndata, finfo, xmit_more);
diff --git a/drivers/net/ethernet/cavium/liquidio/lio_vf_main.c b/drivers/net/ethernet/cavium/liquidio/lio_vf_main.c
index 54b245797d2e..db0b90555acb 100644
--- a/drivers/net/ethernet/cavium/liquidio/lio_vf_main.c
+++ b/drivers/net/ethernet/cavium/liquidio/lio_vf_main.c
@@ -1585,7 +1585,7 @@ static netdev_tx_t liquidio_xmit(struct sk_buff *skb, struct net_device *netdev)
 		irh->vlan = skb_vlan_tag_get(skb) & VLAN_VID_MASK;
 	}
 
-	xmit_more = skb->xmit_more;
+	xmit_more = netdev_xmit_more();
 
 	if (unlikely(cmdsetup.s.timestamp))
 		status = send_nic_timestamp_pkt(oct, &ndata, finfo, xmit_more);
diff --git a/drivers/net/ethernet/cisco/enic/enic_main.c b/drivers/net/ethernet/cisco/enic/enic_main.c
index 733d9172425b..acb2856936d2 100644
--- a/drivers/net/ethernet/cisco/enic/enic_main.c
+++ b/drivers/net/ethernet/cisco/enic/enic_main.c
@@ -897,7 +897,7 @@ static netdev_tx_t enic_hard_start_xmit(struct sk_buff *skb,
 	if (vnic_wq_desc_avail(wq) < MAX_SKB_FRAGS + ENIC_DESC_MAX_SPLITS)
 		netif_tx_stop_queue(txq);
 	skb_tx_timestamp(skb);
-	if (!skb->xmit_more || netif_xmit_stopped(txq))
+	if (!netdev_xmit_more() || netif_xmit_stopped(txq))
 		vnic_wq_doorbell(wq);
 
 	spin_unlock(&enic->wq_lock[txq_map]);
diff --git a/drivers/net/ethernet/emulex/benet/be_main.c b/drivers/net/ethernet/emulex/benet/be_main.c
index 3c7c04406a2b..e2f9fbced174 100644
--- a/drivers/net/ethernet/emulex/benet/be_main.c
+++ b/drivers/net/ethernet/emulex/benet/be_main.c
@@ -1376,7 +1376,7 @@ static netdev_tx_t be_xmit(struct sk_buff *skb, struct net_device *netdev)
 	u16 q_idx = skb_get_queue_mapping(skb);
 	struct be_tx_obj *txo = &adapter->tx_obj[q_idx];
 	struct be_wrb_params wrb_params = { 0 };
-	bool flush = !skb->xmit_more;
+	bool flush = !netdev_xmit_more();
 	u16 wrb_cnt;
 
 	skb = be_xmit_workarounds(adapter, skb, &wrb_params);
diff --git a/drivers/net/ethernet/huawei/hinic/hinic_tx.c b/drivers/net/ethernet/huawei/hinic/hinic_tx.c
index e17bf33eba0c..0fbe8046824b 100644
--- a/drivers/net/ethernet/huawei/hinic/hinic_tx.c
+++ b/drivers/net/ethernet/huawei/hinic/hinic_tx.c
@@ -518,7 +518,7 @@ process_sq_wqe:
 
 flush_skbs:
 	netdev_txq = netdev_get_tx_queue(netdev, q_id);
-	if ((!skb->xmit_more) || (netif_xmit_stopped(netdev_txq)))
+	if ((!netdev_xmit_more()) || (netif_xmit_stopped(netdev_txq)))
 		hinic_sq_write_db(txq->sq, prod_idx, wqe_size, 0);
 
 	return err;
diff --git a/drivers/net/ethernet/intel/e1000/e1000_main.c b/drivers/net/ethernet/intel/e1000/e1000_main.c
index a7c76732849f..6f72ab139fd9 100644
--- a/drivers/net/ethernet/intel/e1000/e1000_main.c
+++ b/drivers/net/ethernet/intel/e1000/e1000_main.c
@@ -3267,7 +3267,7 @@ static netdev_tx_t e1000_xmit_frame(struct sk_buff *skb,
 		/* Make sure there is space in the ring for the next send. */
 		e1000_maybe_stop_tx(netdev, tx_ring, desc_needed);
 
-		if (!skb->xmit_more ||
+		if (!netdev_xmit_more() ||
 		    netif_xmit_stopped(netdev_get_tx_queue(netdev, 0))) {
 			writel(tx_ring->next_to_use, hw->hw_addr + tx_ring->tdt);
 			/* we need this if more than one processor can write to
diff --git a/drivers/net/ethernet/intel/e1000e/netdev.c b/drivers/net/ethernet/intel/e1000e/netdev.c
index 745c1242a2d9..a8fa4a1628f5 100644
--- a/drivers/net/ethernet/intel/e1000e/netdev.c
+++ b/drivers/net/ethernet/intel/e1000e/netdev.c
@@ -5897,7 +5897,7 @@ static netdev_tx_t e1000_xmit_frame(struct sk_buff *skb,
 				     DIV_ROUND_UP(PAGE_SIZE,
 						  adapter->tx_fifo_limit) + 2));
 
-		if (!skb->xmit_more ||
+		if (!netdev_xmit_more() ||
 		    netif_xmit_stopped(netdev_get_tx_queue(netdev, 0))) {
 			if (adapter->flags2 & FLAG2_PCIM2PCI_ARBITER_WA)
 				e1000e_update_tdt_wa(tx_ring,
diff --git a/drivers/net/ethernet/intel/fm10k/fm10k_main.c b/drivers/net/ethernet/intel/fm10k/fm10k_main.c
index 5a0419421511..e2fa112bed9a 100644
--- a/drivers/net/ethernet/intel/fm10k/fm10k_main.c
+++ b/drivers/net/ethernet/intel/fm10k/fm10k_main.c
@@ -1035,7 +1035,7 @@ static void fm10k_tx_map(struct fm10k_ring *tx_ring,
 	fm10k_maybe_stop_tx(tx_ring, DESC_NEEDED);
 
 	/* notify HW of packet */
-	if (netif_xmit_stopped(txring_txq(tx_ring)) || !skb->xmit_more) {
+	if (netif_xmit_stopped(txring_txq(tx_ring)) || !netdev_xmit_more()) {
 		writel(i, tx_ring->tail);
 
 		/* we need this if more than one processor can write to our tail
diff --git a/drivers/net/ethernet/intel/i40e/i40e_txrx.c b/drivers/net/ethernet/intel/i40e/i40e_txrx.c
index 6c97667d20ef..1a95223c9f99 100644
--- a/drivers/net/ethernet/intel/i40e/i40e_txrx.c
+++ b/drivers/net/ethernet/intel/i40e/i40e_txrx.c
@@ -3469,7 +3469,7 @@ static inline int i40e_tx_map(struct i40e_ring *tx_ring, struct sk_buff *skb,
 	first->next_to_watch = tx_desc;
 
 	/* notify HW of packet */
-	if (netif_xmit_stopped(txring_txq(tx_ring)) || !skb->xmit_more) {
+	if (netif_xmit_stopped(txring_txq(tx_ring)) || !netdev_xmit_more()) {
 		writel(i, tx_ring->tail);
 
 		/* we need this if more than one processor can write to our tail
diff --git a/drivers/net/ethernet/intel/iavf/iavf_txrx.c b/drivers/net/ethernet/intel/iavf/iavf_txrx.c
index 9b4d7cec2e18..b64187753ad6 100644
--- a/drivers/net/ethernet/intel/iavf/iavf_txrx.c
+++ b/drivers/net/ethernet/intel/iavf/iavf_txrx.c
@@ -2358,7 +2358,7 @@ static inline void iavf_tx_map(struct iavf_ring *tx_ring, struct sk_buff *skb,
 	first->next_to_watch = tx_desc;
 
 	/* notify HW of packet */
-	if (netif_xmit_stopped(txring_txq(tx_ring)) || !skb->xmit_more) {
+	if (netif_xmit_stopped(txring_txq(tx_ring)) || !netdev_xmit_more()) {
 		writel(i, tx_ring->tail);
 
 		/* we need this if more than one processor can write to our tail
diff --git a/drivers/net/ethernet/intel/ice/ice_txrx.c b/drivers/net/ethernet/intel/ice/ice_txrx.c
index f2462799154a..a6f7b7feaf3c 100644
--- a/drivers/net/ethernet/intel/ice/ice_txrx.c
+++ b/drivers/net/ethernet/intel/ice/ice_txrx.c
@@ -1646,7 +1646,7 @@ ice_tx_map(struct ice_ring *tx_ring, struct ice_tx_buf *first,
 	ice_maybe_stop_tx(tx_ring, DESC_NEEDED);
 
 	/* notify HW of packet */
-	if (netif_xmit_stopped(txring_txq(tx_ring)) || !skb->xmit_more) {
+	if (netif_xmit_stopped(txring_txq(tx_ring)) || !netdev_xmit_more()) {
 		writel(i, tx_ring->tail);
 
 		/* we need this if more than one processor can write to our tail
diff --git a/drivers/net/ethernet/intel/igb/igb_main.c b/drivers/net/ethernet/intel/igb/igb_main.c
index bea7175d171b..32d61d5a2706 100644
--- a/drivers/net/ethernet/intel/igb/igb_main.c
+++ b/drivers/net/ethernet/intel/igb/igb_main.c
@@ -6029,7 +6029,7 @@ static int igb_tx_map(struct igb_ring *tx_ring,
 	/* Make sure there is space in the ring for the next send. */
 	igb_maybe_stop_tx(tx_ring, DESC_NEEDED);
 
-	if (netif_xmit_stopped(txring_txq(tx_ring)) || !skb->xmit_more) {
+	if (netif_xmit_stopped(txring_txq(tx_ring)) || !netdev_xmit_more()) {
 		writel(i, tx_ring->tail);
 
 		/* we need this if more than one processor can write to our tail
diff --git a/drivers/net/ethernet/intel/igc/igc_main.c b/drivers/net/ethernet/intel/igc/igc_main.c
index a883b3f357e7..f79728381e8a 100644
--- a/drivers/net/ethernet/intel/igc/igc_main.c
+++ b/drivers/net/ethernet/intel/igc/igc_main.c
@@ -939,7 +939,7 @@ static int igc_tx_map(struct igc_ring *tx_ring,
 	/* Make sure there is space in the ring for the next send. */
 	igc_maybe_stop_tx(tx_ring, DESC_NEEDED);
 
-	if (netif_xmit_stopped(txring_txq(tx_ring)) || !skb->xmit_more) {
+	if (netif_xmit_stopped(txring_txq(tx_ring)) || !netdev_xmit_more()) {
 		writel(i, tx_ring->tail);
 
 		/* we need this if more than one processor can write to our tail
diff --git a/drivers/net/ethernet/intel/ixgbe/ixgbe_main.c b/drivers/net/ethernet/intel/ixgbe/ixgbe_main.c
index 16c728984164..60cec3540dd7 100644
--- a/drivers/net/ethernet/intel/ixgbe/ixgbe_main.c
+++ b/drivers/net/ethernet/intel/ixgbe/ixgbe_main.c
@@ -8297,7 +8297,7 @@ static int ixgbe_tx_map(struct ixgbe_ring *tx_ring,
 
 	ixgbe_maybe_stop_tx(tx_ring, DESC_NEEDED);
 
-	if (netif_xmit_stopped(txring_txq(tx_ring)) || !skb->xmit_more) {
+	if (netif_xmit_stopped(txring_txq(tx_ring)) || !netdev_xmit_more()) {
 		writel(i, tx_ring->tail);
 
 		/* we need this if more than one processor can write to our tail
diff --git a/drivers/net/ethernet/marvell/mvneta.c b/drivers/net/ethernet/marvell/mvneta.c
index a944be3c57b1..bb68737dce56 100644
--- a/drivers/net/ethernet/marvell/mvneta.c
+++ b/drivers/net/ethernet/marvell/mvneta.c
@@ -2467,7 +2467,7 @@ out:
 		if (txq->count >= txq->tx_stop_threshold)
 			netif_tx_stop_queue(nq);
 
-		if (!skb->xmit_more || netif_xmit_stopped(nq) ||
+		if (!netdev_xmit_more() || netif_xmit_stopped(nq) ||
 		    txq->pending + frags > MVNETA_TXQ_DEC_SENT_MASK)
 			mvneta_txq_pend_desc_add(pp, txq, frags);
 		else
diff --git a/drivers/net/ethernet/mediatek/mtk_eth_soc.c b/drivers/net/ethernet/mediatek/mtk_eth_soc.c
index 549d36497b8c..53abe925ecb1 100644
--- a/drivers/net/ethernet/mediatek/mtk_eth_soc.c
+++ b/drivers/net/ethernet/mediatek/mtk_eth_soc.c
@@ -767,7 +767,8 @@ static int mtk_tx_map(struct sk_buff *skb, struct net_device *dev,
 	 */
 	wmb();
 
-	if (netif_xmit_stopped(netdev_get_tx_queue(dev, 0)) || !skb->xmit_more)
+	if (netif_xmit_stopped(netdev_get_tx_queue(dev, 0)) ||
+	    !netdev_xmit_more())
 		mtk_w32(eth, txd->txd2, MTK_QTX_CTX_PTR);
 
 	return 0;
diff --git a/drivers/net/ethernet/mellanox/mlx4/en_tx.c b/drivers/net/ethernet/mellanox/mlx4/en_tx.c
index fba54fb06e18..36a92b19e613 100644
--- a/drivers/net/ethernet/mellanox/mlx4/en_tx.c
+++ b/drivers/net/ethernet/mellanox/mlx4/en_tx.c
@@ -1042,7 +1042,7 @@ netdev_tx_t mlx4_en_xmit(struct sk_buff *skb, struct net_device *dev)
 
 	send_doorbell = __netdev_tx_sent_queue(ring->tx_queue,
 					       tx_info->nr_bytes,
-					       skb->xmit_more);
+					       netdev_xmit_more());
 
 	real_size = (real_size / 16) & 0x3f;
 
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en.h b/drivers/net/ethernet/mellanox/mlx5/core/en.h
index 9e71cf03369c..3e1ea8b42c77 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en.h
@@ -772,7 +772,7 @@ u16 mlx5e_select_queue(struct net_device *dev, struct sk_buff *skb,
 		       struct net_device *sb_dev);
 netdev_tx_t mlx5e_xmit(struct sk_buff *skb, struct net_device *dev);
 netdev_tx_t mlx5e_sq_xmit(struct mlx5e_txqsq *sq, struct sk_buff *skb,
-			  struct mlx5e_tx_wqe *wqe, u16 pi);
+			  struct mlx5e_tx_wqe *wqe, u16 pi, bool xmit_more);
 
 void mlx5e_completion_event(struct mlx5_core_cq *mcq);
 void mlx5e_cq_error_event(struct mlx5_core_cq *mcq, enum mlx5_event event);
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_accel/tls_rxtx.c b/drivers/net/ethernet/mellanox/mlx5/core/en_accel/tls_rxtx.c
index be137d4a9169..439bf5953885 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_accel/tls_rxtx.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_accel/tls_rxtx.c
@@ -181,7 +181,6 @@ static void mlx5e_tls_complete_sync_skb(struct sk_buff *skb,
 	 */
 	nskb->ip_summed = CHECKSUM_PARTIAL;
 
-	nskb->xmit_more = 1;
 	nskb->queue_mapping = skb->queue_mapping;
 }
 
@@ -248,7 +247,7 @@ mlx5e_tls_handle_ooo(struct mlx5e_tls_offload_context_tx *context,
 	sq->stats->tls_resync_bytes += nskb->len;
 	mlx5e_tls_complete_sync_skb(skb, nskb, tcp_seq, headln,
 				    cpu_to_be64(info.rcd_sn));
-	mlx5e_sq_xmit(sq, nskb, *wqe, *pi);
+	mlx5e_sq_xmit(sq, nskb, *wqe, *pi, true);
 	mlx5e_sq_fetch_wqe(sq, wqe, pi);
 	return skb;
 
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_tx.c b/drivers/net/ethernet/mellanox/mlx5/core/en_tx.c
index 41e2a01d3713..40f3f98aa279 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_tx.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_tx.c
@@ -297,7 +297,8 @@ static inline void mlx5e_fill_sq_frag_edge(struct mlx5e_txqsq *sq,
 static inline void
 mlx5e_txwqe_complete(struct mlx5e_txqsq *sq, struct sk_buff *skb,
 		     u8 opcode, u16 ds_cnt, u8 num_wqebbs, u32 num_bytes, u8 num_dma,
-		     struct mlx5e_tx_wqe_info *wi, struct mlx5_wqe_ctrl_seg *cseg)
+		     struct mlx5e_tx_wqe_info *wi, struct mlx5_wqe_ctrl_seg *cseg,
+		     bool xmit_more)
 {
 	struct mlx5_wq_cyc *wq = &sq->wq;
 
@@ -320,14 +321,14 @@ mlx5e_txwqe_complete(struct mlx5e_txqsq *sq, struct sk_buff *skb,
 		sq->stats->stopped++;
 	}
 
-	if (!skb->xmit_more || netif_xmit_stopped(sq->txq))
+	if (!xmit_more || netif_xmit_stopped(sq->txq))
 		mlx5e_notify_hw(wq, sq->pc, sq->uar_map, cseg);
 }
 
 #define INL_HDR_START_SZ (sizeof(((struct mlx5_wqe_eth_seg *)NULL)->inline_hdr.start))
 
 netdev_tx_t mlx5e_sq_xmit(struct mlx5e_txqsq *sq, struct sk_buff *skb,
-			  struct mlx5e_tx_wqe *wqe, u16 pi)
+			  struct mlx5e_tx_wqe *wqe, u16 pi, bool xmit_more)
 {
 	struct mlx5_wq_cyc *wq = &sq->wq;
 	struct mlx5_wqe_ctrl_seg *cseg;
@@ -360,7 +361,7 @@ netdev_tx_t mlx5e_sq_xmit(struct mlx5e_txqsq *sq, struct sk_buff *skb,
 	}
 
 	stats->bytes     += num_bytes;
-	stats->xmit_more += skb->xmit_more;
+	stats->xmit_more += netdev_xmit_more();
 
 	headlen = skb->len - ihs - skb->data_len;
 	ds_cnt += !!headlen;
@@ -423,7 +424,7 @@ netdev_tx_t mlx5e_sq_xmit(struct mlx5e_txqsq *sq, struct sk_buff *skb,
 		goto err_drop;
 
 	mlx5e_txwqe_complete(sq, skb, opcode, ds_cnt, num_wqebbs, num_bytes,
-			     num_dma, wi, cseg);
+			     num_dma, wi, cseg, xmit_more);
 
 	return NETDEV_TX_OK;
 
@@ -449,7 +450,7 @@ netdev_tx_t mlx5e_xmit(struct sk_buff *skb, struct net_device *dev)
 	if (unlikely(!skb))
 		return NETDEV_TX_OK;
 
-	return mlx5e_sq_xmit(sq, skb, wqe, pi);
+	return mlx5e_sq_xmit(sq, skb, wqe, pi, netdev_xmit_more());
 }
 
 static void mlx5e_dump_error_cqe(struct mlx5e_txqsq *sq,
@@ -659,7 +660,7 @@ netdev_tx_t mlx5i_sq_xmit(struct mlx5e_txqsq *sq, struct sk_buff *skb,
 	}
 
 	stats->bytes     += num_bytes;
-	stats->xmit_more += skb->xmit_more;
+	stats->xmit_more += netdev_xmit_more();
 
 	headlen = skb->len - ihs - skb->data_len;
 	ds_cnt += !!headlen;
@@ -704,7 +705,7 @@ netdev_tx_t mlx5i_sq_xmit(struct mlx5e_txqsq *sq, struct sk_buff *skb,
 		goto err_drop;
 
 	mlx5e_txwqe_complete(sq, skb, opcode, ds_cnt, num_wqebbs, num_bytes,
-			     num_dma, wi, cseg);
+			     num_dma, wi, cseg, false);
 
 	return NETDEV_TX_OK;
 
diff --git a/drivers/net/ethernet/netronome/nfp/nfp_net_common.c b/drivers/net/ethernet/netronome/nfp/nfp_net_common.c
index 99200b5dac76..961cd5e7bf2b 100644
--- a/drivers/net/ethernet/netronome/nfp/nfp_net_common.c
+++ b/drivers/net/ethernet/netronome/nfp/nfp_net_common.c
@@ -909,7 +909,7 @@ static int nfp_net_tx(struct sk_buff *skb, struct net_device *netdev)
 		nfp_net_tx_ring_stop(nd_q, tx_ring);
 
 	tx_ring->wr_ptr_add += nr_frags + 1;
-	if (__netdev_tx_sent_queue(nd_q, txbuf->real_len, skb->xmit_more))
+	if (__netdev_tx_sent_queue(nd_q, txbuf->real_len, netdev_xmit_more()))
 		nfp_net_tx_xmit_more_flush(tx_ring);
 
 	return NETDEV_TX_OK;
diff --git a/drivers/net/ethernet/qlogic/qede/qede_fp.c b/drivers/net/ethernet/qlogic/qede/qede_fp.c
index c342b07e3a93..954015d2011a 100644
--- a/drivers/net/ethernet/qlogic/qede/qede_fp.c
+++ b/drivers/net/ethernet/qlogic/qede/qede_fp.c
@@ -1665,12 +1665,12 @@ netdev_tx_t qede_start_xmit(struct sk_buff *skb, struct net_device *ndev)
 	txq->tx_db.data.bd_prod =
 		cpu_to_le16(qed_chain_get_prod_idx(&txq->tx_pbl));
 
-	if (!skb->xmit_more || netif_xmit_stopped(netdev_txq))
+	if (!netdev_xmit_more() || netif_xmit_stopped(netdev_txq))
 		qede_update_tx_producer(txq);
 
 	if (unlikely(qed_chain_get_elem_left(&txq->tx_pbl)
 		      < (MAX_SKB_FRAGS + 1))) {
-		if (skb->xmit_more)
+		if (netdev_xmit_more())
 			qede_update_tx_producer(txq);
 
 		netif_tx_stop_queue(netdev_txq);
diff --git a/drivers/net/ethernet/rdc/r6040.c b/drivers/net/ethernet/rdc/r6040.c
index 04aa592f35c3..ad335bca3273 100644
--- a/drivers/net/ethernet/rdc/r6040.c
+++ b/drivers/net/ethernet/rdc/r6040.c
@@ -840,7 +840,7 @@ static netdev_tx_t r6040_start_xmit(struct sk_buff *skb,
 	skb_tx_timestamp(skb);
 
 	/* Trigger the MAC to check the TX descriptor */
-	if (!skb->xmit_more || netif_queue_stopped(dev))
+	if (!netdev_xmit_more() || netif_queue_stopped(dev))
 		iowrite16(TM2TX, ioaddr + MTPR);
 	lp->tx_insert_ptr = descptr->vndescp;
 
diff --git a/drivers/net/ethernet/sfc/falcon/tx.c b/drivers/net/ethernet/sfc/falcon/tx.c
index 3409bbf5b19f..c5059f456f37 100644
--- a/drivers/net/ethernet/sfc/falcon/tx.c
+++ b/drivers/net/ethernet/sfc/falcon/tx.c
@@ -321,7 +321,7 @@ netdev_tx_t ef4_enqueue_skb(struct ef4_tx_queue *tx_queue, struct sk_buff *skb)
 	netdev_tx_sent_queue(tx_queue->core_txq, skb_len);
 
 	/* Pass off to hardware */
-	if (!skb->xmit_more || netif_xmit_stopped(tx_queue->core_txq)) {
+	if (!netdev_xmit_more() || netif_xmit_stopped(tx_queue->core_txq)) {
 		struct ef4_tx_queue *txq2 = ef4_tx_queue_partner(tx_queue);
 
 		/* There could be packets left on the partner queue if those
@@ -333,7 +333,7 @@ netdev_tx_t ef4_enqueue_skb(struct ef4_tx_queue *tx_queue, struct sk_buff *skb)
 
 		ef4_nic_push_buffers(tx_queue);
 	} else {
-		tx_queue->xmit_more_available = skb->xmit_more;
+		tx_queue->xmit_more_available = netdev_xmit_more();
 	}
 
 	tx_queue->tx_packets++;
diff --git a/drivers/net/ethernet/sfc/tx.c b/drivers/net/ethernet/sfc/tx.c
index 06c8f282263f..e182055ec2eb 100644
--- a/drivers/net/ethernet/sfc/tx.c
+++ b/drivers/net/ethernet/sfc/tx.c
@@ -478,8 +478,6 @@ static int efx_tx_tso_fallback(struct efx_tx_queue *tx_queue,
 		next = skb->next;
 		skb->next = NULL;
 
-		if (next)
-			skb->xmit_more = true;
 		efx_enqueue_skb(tx_queue, skb);
 		skb = next;
 	}
@@ -506,7 +504,7 @@ static int efx_tx_tso_fallback(struct efx_tx_queue *tx_queue,
 netdev_tx_t efx_enqueue_skb(struct efx_tx_queue *tx_queue, struct sk_buff *skb)
 {
 	unsigned int old_insert_count = tx_queue->insert_count;
-	bool xmit_more = skb->xmit_more;
+	bool xmit_more = netdev_xmit_more();
 	bool data_mapped = false;
 	unsigned int segments;
 	unsigned int skb_len;
@@ -533,7 +531,7 @@ netdev_tx_t efx_enqueue_skb(struct efx_tx_queue *tx_queue, struct sk_buff *skb)
 		if (rc)
 			goto err;
 #ifdef EFX_USE_PIO
-	} else if (skb_len <= efx_piobuf_size && !skb->xmit_more &&
+	} else if (skb_len <= efx_piobuf_size && !xmit_more &&
 		   efx_nic_may_tx_pio(tx_queue)) {
 		/* Use PIO for short packets with an empty queue. */
 		if (efx_enqueue_skb_pio(tx_queue, skb))
@@ -559,8 +557,8 @@ netdev_tx_t efx_enqueue_skb(struct efx_tx_queue *tx_queue, struct sk_buff *skb)
 	if (__netdev_tx_sent_queue(tx_queue->core_txq, skb_len, xmit_more)) {
 		struct efx_tx_queue *txq2 = efx_tx_queue_partner(tx_queue);
 
-		/* There could be packets left on the partner queue if those
-		 * SKBs had skb->xmit_more set. If we do not push those they
+		/* There could be packets left on the partner queue if
+		 * xmit_more was set. If we do not push those they
 		 * could be left for a long time and cause a netdev watchdog.
 		 */
 		if (txq2->xmit_more_available)
@@ -568,7 +566,7 @@ netdev_tx_t efx_enqueue_skb(struct efx_tx_queue *tx_queue, struct sk_buff *skb)
 
 		efx_nic_push_buffers(tx_queue);
 	} else {
-		tx_queue->xmit_more_available = skb->xmit_more;
+		tx_queue->xmit_more_available = xmit_more;
 	}
 
 	if (segments) {
diff --git a/drivers/net/ethernet/synopsys/dwc-xlgmac-hw.c b/drivers/net/ethernet/synopsys/dwc-xlgmac-hw.c
index 99d86e39ff54..bf6c1c6779ff 100644
--- a/drivers/net/ethernet/synopsys/dwc-xlgmac-hw.c
+++ b/drivers/net/ethernet/synopsys/dwc-xlgmac-hw.c
@@ -995,7 +995,7 @@ static void xlgmac_dev_xmit(struct xlgmac_channel *channel)
 	smp_wmb();
 
 	ring->cur = cur_index + 1;
-	if (!pkt_info->skb->xmit_more ||
+	if (!netdev_xmit_more() ||
 	    netif_xmit_stopped(netdev_get_tx_queue(pdata->netdev,
 						   channel->queue_index)))
 		xlgmac_tx_start_xmit(channel, ring);
diff --git a/drivers/net/hyperv/netvsc.c b/drivers/net/hyperv/netvsc.c
index 813d195bbd57..9a022539d305 100644
--- a/drivers/net/hyperv/netvsc.c
+++ b/drivers/net/hyperv/netvsc.c
@@ -964,7 +964,7 @@ int netvsc_send(struct net_device *ndev,
 	/* Keep aggregating only if stack says more data is coming
 	 * and not doing mixed modes send and not flow blocked
 	 */
-	xmit_more = skb->xmit_more &&
+	xmit_more = netdev_xmit_more() &&
 		!packet->cp_partial &&
 		!netif_xmit_stopped(netdev_get_tx_queue(ndev, packet->q_idx));
 
diff --git a/drivers/net/virtio_net.c b/drivers/net/virtio_net.c
index 1b03c4b6ebff..ba246fc475ae 100644
--- a/drivers/net/virtio_net.c
+++ b/drivers/net/virtio_net.c
@@ -1568,7 +1568,7 @@ static netdev_tx_t start_xmit(struct sk_buff *skb, struct net_device *dev)
 	struct send_queue *sq = &vi->sq[qnum];
 	int err;
 	struct netdev_queue *txq = netdev_get_tx_queue(dev, qnum);
-	bool kick = !skb->xmit_more;
+	bool kick = !netdev_xmit_more();
 	bool use_napi = sq->napi.weight;
 
 	/* Free up any pending old buffers before queueing new ones. */
diff --git a/drivers/staging/mt7621-eth/mtk_eth_soc.c b/drivers/staging/mt7621-eth/mtk_eth_soc.c
index 6027b19f7bc2..02a8584b3d1d 100644
--- a/drivers/staging/mt7621-eth/mtk_eth_soc.c
+++ b/drivers/staging/mt7621-eth/mtk_eth_soc.c
@@ -741,7 +741,8 @@ static int mtk_pdma_tx_map(struct sk_buff *skb, struct net_device *dev,
 	wmb();
 	atomic_set(&ring->tx_free_count, mtk_pdma_empty_txd(ring));
 
-	if (netif_xmit_stopped(netdev_get_tx_queue(dev, 0)) || !skb->xmit_more)
+	if (netif_xmit_stopped(netdev_get_tx_queue(dev, 0)) ||
+	    !netdev_xmit_more())
 		mtk_reg_w32(eth, ring->tx_next_idx, MTK_REG_TX_CTX_IDX0);
 
 	return 0;
@@ -935,7 +936,8 @@ static int mtk_qdma_tx_map(struct sk_buff *skb, struct net_device *dev,
 	 */
 	wmb();
 
-	if (netif_xmit_stopped(netdev_get_tx_queue(dev, 0)) || !skb->xmit_more)
+	if (netif_xmit_stopped(netdev_get_tx_queue(dev, 0)) ||
+	    !netdev_xmit_more())
 		mtk_w32(eth, txd->txd2, MTK_QTX_CTX_PTR);
 
 	return 0;
diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h
index 78f5ec4ebf64..eb9f05e0863d 100644
--- a/include/linux/netdevice.h
+++ b/include/linux/netdevice.h
@@ -2659,14 +2659,6 @@ void netdev_freemem(struct net_device *dev);
 void synchronize_net(void);
 int init_dummy_netdev(struct net_device *dev);
 
-DECLARE_PER_CPU(int, xmit_recursion);
-#define XMIT_RECURSION_LIMIT	10
-
-static inline int dev_recursion_level(void)
-{
-	return this_cpu_read(xmit_recursion);
-}
-
 struct net_device *dev_get_by_index(struct net *net, int ifindex);
 struct net_device *__dev_get_by_index(struct net *net, int ifindex);
 struct net_device *dev_get_by_index_rcu(struct net *net, int ifindex);
@@ -3015,6 +3007,11 @@ struct softnet_data {
 #ifdef CONFIG_XFRM_OFFLOAD
 	struct sk_buff_head	xfrm_backlog;
 #endif
+	/* written and read only by owning cpu: */
+	struct {
+		u16 recursion;
+		u8  more;
+	} xmit;
 #ifdef CONFIG_RPS
 	/* input_queue_head should be written by cpu owning this struct,
 	 * and only read by other cpus. Worth using a cache line.
@@ -3050,6 +3047,28 @@ static inline void input_queue_tail_incr_save(struct softnet_data *sd,
 
 DECLARE_PER_CPU_ALIGNED(struct softnet_data, softnet_data);
 
+static inline int dev_recursion_level(void)
+{
+	return __this_cpu_read(softnet_data.xmit.recursion);
+}
+
+#define XMIT_RECURSION_LIMIT	10
+static inline bool dev_xmit_recursion(void)
+{
+	return unlikely(__this_cpu_read(softnet_data.xmit.recursion) >
+			XMIT_RECURSION_LIMIT);
+}
+
+static inline void dev_xmit_recursion_inc(void)
+{
+	__this_cpu_inc(softnet_data.xmit.recursion);
+}
+
+static inline void dev_xmit_recursion_dec(void)
+{
+	__this_cpu_dec(softnet_data.xmit.recursion);
+}
+
 void __netif_schedule(struct Qdisc *q);
 void netif_schedule_queue(struct netdev_queue *txq);
 
@@ -4405,10 +4424,15 @@ static inline netdev_tx_t __netdev_start_xmit(const struct net_device_ops *ops,
 					      struct sk_buff *skb, struct net_device *dev,
 					      bool more)
 {
-	skb->xmit_more = more ? 1 : 0;
+	__this_cpu_write(softnet_data.xmit.more, more);
 	return ops->ndo_start_xmit(skb, dev);
 }
 
+static inline bool netdev_xmit_more(void)
+{
+	return __this_cpu_read(softnet_data.xmit.more);
+}
+
 static inline netdev_tx_t netdev_start_xmit(struct sk_buff *skb, struct net_device *dev,
 					    struct netdev_queue *txq, bool more)
 {
diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h
index 9027a8c4219f..69b5538adcea 100644
--- a/include/linux/skbuff.h
+++ b/include/linux/skbuff.h
@@ -657,7 +657,6 @@ typedef unsigned char *sk_buff_data_t;
  *	@tc_index: Traffic control index
  *	@hash: the packet hash
  *	@queue_mapping: Queue mapping for multiqueue devices
- *	@xmit_more: More SKBs are pending for this queue
  *	@pfmemalloc: skbuff was allocated from PFMEMALLOC reserves
  *	@active_extensions: active extensions (skb_ext_id types)
  *	@ndisc_nodetype: router type (from link layer)
@@ -764,7 +763,6 @@ struct sk_buff {
 				fclone:2,
 				peeked:1,
 				head_frag:1,
-				xmit_more:1,
 				pfmemalloc:1;
 #ifdef CONFIG_SKB_EXTENSIONS
 	__u8			active_extensions;
diff --git a/net/core/dev.c b/net/core/dev.c
index 9823b7713f79..d5b1315218d3 100644
--- a/net/core/dev.c
+++ b/net/core/dev.c
@@ -3566,9 +3566,6 @@ static void skb_update_prio(struct sk_buff *skb)
 #define skb_update_prio(skb)
 #endif
 
-DEFINE_PER_CPU(int, xmit_recursion);
-EXPORT_SYMBOL(xmit_recursion);
-
 /**
  *	dev_loopback_xmit - loop back @skb
  *	@net: network namespace this loopback is happening in
@@ -3857,8 +3854,7 @@ static int __dev_queue_xmit(struct sk_buff *skb, struct net_device *sb_dev)
 		int cpu = smp_processor_id(); /* ok because BHs are off */
 
 		if (txq->xmit_lock_owner != cpu) {
-			if (unlikely(__this_cpu_read(xmit_recursion) >
-				     XMIT_RECURSION_LIMIT))
+			if (dev_xmit_recursion())
 				goto recursion_alert;
 
 			skb = validate_xmit_skb(skb, dev, &again);
@@ -3868,9 +3864,9 @@ static int __dev_queue_xmit(struct sk_buff *skb, struct net_device *sb_dev)
 			HARD_TX_LOCK(dev, txq, cpu);
 
 			if (!netif_xmit_stopped(txq)) {
-				__this_cpu_inc(xmit_recursion);
+				dev_xmit_recursion_inc();
 				skb = dev_hard_start_xmit(skb, dev, txq, &rc);
-				__this_cpu_dec(xmit_recursion);
+				dev_xmit_recursion_dec();
 				if (dev_xmit_complete(rc)) {
 					HARD_TX_UNLOCK(dev, txq);
 					goto out;
diff --git a/net/core/filter.c b/net/core/filter.c
index 4a8455757507..cdaafa3322db 100644
--- a/net/core/filter.c
+++ b/net/core/filter.c
@@ -2016,7 +2016,7 @@ static inline int __bpf_tx_skb(struct net_device *dev, struct sk_buff *skb)
 {
 	int ret;
 
-	if (unlikely(__this_cpu_read(xmit_recursion) > XMIT_RECURSION_LIMIT)) {
+	if (dev_xmit_recursion()) {
 		net_crit_ratelimited("bpf: recursion limit reached on datapath, buggy bpf program?\n");
 		kfree_skb(skb);
 		return -ENETDOWN;
@@ -2024,9 +2024,9 @@ static inline int __bpf_tx_skb(struct net_device *dev, struct sk_buff *skb)
 
 	skb->dev = dev;
 
-	__this_cpu_inc(xmit_recursion);
+	dev_xmit_recursion_inc();
 	ret = dev_queue_xmit(skb);
-	__this_cpu_dec(xmit_recursion);
+	dev_xmit_recursion_dec();
 
 	return ret;
 }