Merge git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net

Documentation/networking/ip-sysctl.rst

···
 
 	default: initial size of receive buffer used by TCP sockets.
 	This value overrides net.core.rmem_default used by other protocols.
-	Default: 87380 bytes. This value results in window of 65535 with
-	default setting of tcp_adv_win_scale and tcp_app_win:0 and a bit
-	less for default tcp_app_win. See below about these variables.
+	Default: 131072 bytes.
+	This value results in initial window of 65535.
 
 	max: maximal size of receive buffer allowed for automatically
 	selected receiver buffers for TCP socket. This value does not override
 	net.core.rmem_max.  Calling setsockopt() with SO_RCVBUF disables
 	automatic tuning of that socket's receive buffer size, in which
 	case this value is ignored.
-	Default: between 87380B and 6MB, depending on RAM size.
+	Default: between 131072 and 6MB, depending on RAM size.
 
 tcp_sack - BOOLEAN
 	Enable select acknowledgments (SACKS).

Documentation/networking/netdev-FAQ.rst

···
 Posting as one thread is discouraged because it confuses patchwork
 (as of patchwork 2.2.2).
 
+Can I reproduce the checks from patchwork on my local machine?
+--------------------------------------------------------------
+
+Checks in patchwork are mostly simple wrappers around existing kernel
+scripts, the sources are available at:
+
+https://github.com/kuba-moo/nipa/tree/master/tests
+
+Running all the builds and checks locally is a pain, can I post my patches and have the patchwork bot validate them?
+--------------------------------------------------------------------------------------------------------------------
+
+No, you must ensure that your patches are ready by testing them locally
+before posting to the mailing list. The patchwork build bot instance
+gets overloaded very easily and netdev@vger really doesn't need more
+traffic if we can help it.
+
 Any other tips to help ensure my net/net-next patch gets OK'd?
 --------------------------------------------------------------
 Attention to detail.  Re-read your own work as if you were the

drivers/atm/idt77252.c

···
 	struct sk_buff *skb;
 
 	printk("%s: at %p\n", __func__, idt77252_init);
-
-	if (sizeof(skb->cb) < sizeof(struct atm_skb_data) +
-			      sizeof(struct idt77252_skb_prv)) {
-		printk(KERN_ERR "%s: skb->cb is too small (%lu < %lu)\n",
-		       __func__, (unsigned long) sizeof(skb->cb),
-		       (unsigned long) sizeof(struct atm_skb_data) +
-				       sizeof(struct idt77252_skb_prv));
-		return -EIO;
-	}
-
+	BUILD_BUG_ON(sizeof(skb->cb) < sizeof(struct idt77252_skb_prv) + sizeof(struct atm_skb_data));
 	return pci_register_driver(&idt77252_driver);
 }
 

drivers/atm/idt77252.h

···
 	struct scqe	tbd;	/* Transmit Buffer Descriptor */
 	dma_addr_t	paddr;	/* DMA handle */
 	u32		pool;	/* sb_pool handle */
-};
+} __packed;
 
 #define IDT77252_PRV_TBD(skb)	\
 	(((struct idt77252_skb_prv *)(ATM_SKB(skb)+1))->tbd)

drivers/net/ethernet/amd/xgbe/xgbe-common.h

···
 #define MDIO_PMA_10GBR_FECCTRL		0x00ab
 #endif
 
+#ifndef MDIO_PMA_RX_CTRL1
+#define MDIO_PMA_RX_CTRL1		0x8051
+#endif
+
 #ifndef MDIO_PCS_DIG_CTRL
 #define MDIO_PCS_DIG_CTRL		0x8000
+#endif
+
+#ifndef MDIO_PCS_DIGITAL_STAT
+#define MDIO_PCS_DIGITAL_STAT		0x8010
 #endif
 
 #ifndef MDIO_AN_XNP
···
 #define XGBE_KR_TRAINING_ENABLE		BIT(1)
 
 #define XGBE_PCS_CL37_BP		BIT(12)
+#define XGBE_PCS_PSEQ_STATE_MASK	0x1c
+#define XGBE_PCS_PSEQ_STATE_POWER_GOOD	0x10
 
 #define XGBE_AN_CL37_INT_CMPLT		BIT(0)
 #define XGBE_AN_CL37_INT_MASK		0x01
···
 #define XGBE_PMA_CDR_TRACK_EN_MASK	0x01
 #define XGBE_PMA_CDR_TRACK_EN_OFF	0x00
 #define XGBE_PMA_CDR_TRACK_EN_ON	0x01
+
+#define XGBE_PMA_RX_RST_0_MASK		BIT(4)
+#define XGBE_PMA_RX_RST_0_RESET_ON	0x10
+#define XGBE_PMA_RX_RST_0_RESET_OFF	0x00
 
 /* Bit setting and getting macros
  *  The get macro will extract the current bit field value from within

drivers/net/ethernet/amd/xgbe/xgbe-drv.c

···
 		return;
 
 	netif_tx_stop_all_queues(netdev);
+	netif_carrier_off(pdata->netdev);
 
 	xgbe_stop_timers(pdata);
 	flush_workqueue(pdata->dev_workqueue);

drivers/net/ethernet/amd/xgbe/xgbe-mdio.c

···
 							     &an_restart);
 	if (an_restart) {
 		xgbe_phy_config_aneg(pdata);
-		return;
+		goto adjust_link;
 	}
 
 	if (pdata->phy.link) {
···
 	pdata->phy_if.phy_impl.stop(pdata);
 
 	pdata->phy.link = 0;
-	netif_carrier_off(pdata->netdev);
 
 	xgbe_phy_adjust_link(pdata);
 }

drivers/net/ethernet/amd/xgbe/xgbe-phy-v2.c

···
 	if ((phy_id & 0xfffffff0) != 0x03625d10)
 		return false;
 
+	/* Reset PHY - wait for self-clearing reset bit to clear */
+	genphy_soft_reset(phy_data->phydev);
+
 	/* Disable RGMII mode */
 	phy_write(phy_data->phydev, 0x18, 0x7007);
 	reg = phy_read(phy_data->phydev, 0x18);
···
 	xgbe_phy_put_comm_ownership(pdata);
 }
 
+static void xgbe_phy_rx_reset(struct xgbe_prv_data *pdata)
+{
+	int reg;
+
+	reg = XMDIO_READ_BITS(pdata, MDIO_MMD_PCS, MDIO_PCS_DIGITAL_STAT,
+			      XGBE_PCS_PSEQ_STATE_MASK);
+	if (reg == XGBE_PCS_PSEQ_STATE_POWER_GOOD) {
+		/* Mailbox command timed out, reset of RX block is required.
+		 * This can be done by asseting the reset bit and wait for
+		 * its compeletion.
+		 */
+		XMDIO_WRITE_BITS(pdata, MDIO_MMD_PMAPMD, MDIO_PMA_RX_CTRL1,
+				 XGBE_PMA_RX_RST_0_MASK, XGBE_PMA_RX_RST_0_RESET_ON);
+		ndelay(20);
+		XMDIO_WRITE_BITS(pdata, MDIO_MMD_PMAPMD, MDIO_PMA_RX_CTRL1,
+				 XGBE_PMA_RX_RST_0_MASK, XGBE_PMA_RX_RST_0_RESET_OFF);
+		usleep_range(40, 50);
+		netif_err(pdata, link, pdata->netdev, "firmware mailbox reset performed\n");
+	}
+}
+
 static void xgbe_phy_perform_ratechange(struct xgbe_prv_data *pdata,
 					unsigned int cmd, unsigned int sub_cmd)
 {
···
 	unsigned int wait;
 
 	/* Log if a previous command did not complete */
-	if (XP_IOREAD_BITS(pdata, XP_DRIVER_INT_RO, STATUS))
+	if (XP_IOREAD_BITS(pdata, XP_DRIVER_INT_RO, STATUS)) {
 		netif_dbg(pdata, link, pdata->netdev,
 			  "firmware mailbox not ready for command\n");
+		xgbe_phy_rx_reset(pdata);
+	}
 
 	/* Construct the command */
 	XP_SET_BITS(s0, XP_DRIVER_SCRATCH_0, COMMAND, cmd);
···
 
 	netif_dbg(pdata, link, pdata->netdev,
 		  "firmware mailbox command did not complete\n");
+
+	/* Reset on error */
+	xgbe_phy_rx_reset(pdata);
 }
 
 static void xgbe_phy_rrc(struct xgbe_prv_data *pdata)
···
 	reg = XMDIO_READ(pdata, MDIO_MMD_PCS, MDIO_STAT1);
 	if (reg & MDIO_STAT1_LSTATUS)
 		return 1;
+
+	if (pdata->phy.autoneg == AUTONEG_ENABLE &&
+	    phy_data->port_mode == XGBE_PORT_MODE_BACKPLANE) {
+		if (!test_bit(XGBE_LINK_INIT, &pdata->dev_state)) {
+			netif_carrier_off(pdata->netdev);
+			*an_restart = 1;
+		}
+	}
 
 	/* No link, attempt a receiver reset cycle */
 	if (phy_data->rrc_count++ > XGBE_RRC_FREQUENCY) {

drivers/net/ethernet/aquantia/atlantic/aq_main.c

···
 		goto err_exit;
 
 	err = aq_nic_start(aq_nic);
-	if (err < 0)
+	if (err < 0) {
+		aq_nic_stop(aq_nic);
 		goto err_exit;
+	}
 
 err_exit:
 	if (err < 0)

drivers/net/ethernet/broadcom/bnxt/bnxt.c

···
 			txr->dev_state = BNXT_DEV_STATE_CLOSING;
 		}
 	}
+	/* Drop carrier first to prevent TX timeout */
+	netif_carrier_off(bp->dev);
 	/* Stop all TX queues */
 	netif_tx_disable(bp->dev);
-	netif_carrier_off(bp->dev);
 }
 
 void bnxt_tx_enable(struct bnxt *bp)

drivers/net/ethernet/broadcom/bnxt/bnxt_devlink.c

···
 	if (BNXT_PF(bp) && !bnxt_hwrm_get_nvm_cfg_ver(bp, &nvm_cfg_ver)) {
 		u32 ver = nvm_cfg_ver.vu32;
 
-		sprintf(buf, "%X.%X.%X", (ver >> 16) & 0xF, (ver >> 8) & 0xF,
-			ver & 0xF);
+		sprintf(buf, "%d.%d.%d", (ver >> 16) & 0xf, (ver >> 8) & 0xf,
+			ver & 0xf);
 		rc = bnxt_dl_info_put(bp, req, BNXT_VERSION_STORED,
 				      DEVLINK_INFO_VERSION_GENERIC_FW_PSID,
 				      buf);

drivers/net/ethernet/chelsio/cxgb4/cxgb4_uld.h

···
 #define MAX_ULD_QSETS 16
 #define MAX_ULD_NPORTS 4
 
+/* ulp_mem_io + ulptx_idata + payload + padding */
+#define MAX_IMM_ULPTX_WR_LEN (32 + 8 + 256 + 8)
+
 /* CPL message priority levels */
 enum {
 	CPL_PRIORITY_DATA     = 0,  /* data messages */

drivers/net/ethernet/chelsio/cxgb4/sge.c

···
  *	@skb: the packet
  *
  *	Returns true if a packet can be sent as an offload WR with immediate
- *	data.  We currently use the same limit as for Ethernet packets.
+ *	data.
+ *	FW_OFLD_TX_DATA_WR limits the payload to 255 bytes due to 8-bit field.
+ *      However, FW_ULPTX_WR commands have a 256 byte immediate only
+ *      payload limit.
  */
 static inline int is_ofld_imm(const struct sk_buff *skb)
 {
 	struct work_request_hdr *req = (struct work_request_hdr *)skb->data;
 	unsigned long opcode = FW_WR_OP_G(ntohl(req->wr_hi));
 
-	if (opcode == FW_CRYPTO_LOOKASIDE_WR)
+	if (unlikely(opcode == FW_ULPTX_WR))
+		return skb->len <= MAX_IMM_ULPTX_WR_LEN;
+	else if (opcode == FW_CRYPTO_LOOKASIDE_WR)
 		return skb->len <= SGE_MAX_WR_LEN;
 	else
-		return skb->len <= MAX_IMM_TX_PKT_LEN;
+		return skb->len <= MAX_IMM_OFLD_TX_DATA_WR_LEN;
 }
 
 /**

drivers/net/ethernet/chelsio/inline_crypto/chtls/chtls_cm.h

···
 #define MIN_RCV_WND (24 * 1024U)
 #define LOOPBACK(x)     (((x) & htonl(0xff000000)) == htonl(0x7f000000))
 
-/* ulp_mem_io + ulptx_idata + payload + padding */
-#define MAX_IMM_ULPTX_WR_LEN (32 + 8 + 256 + 8)
-
 /* for TX: a skb must have a headroom of at least TX_HEADER_LEN bytes */
 #define TX_HEADER_LEN \
 	(sizeof(struct fw_ofld_tx_data_wr) + sizeof(struct sge_opaque_hdr))

drivers/net/ethernet/freescale/dpaa2/dpaa2-eth.c

···
 		xdp.frame_sz = DPAA2_ETH_RX_BUF_RAW_SIZE;
 
 		err = xdp_do_redirect(priv->net_dev, &xdp, xdp_prog);
-		if (unlikely(err))
+		if (unlikely(err)) {
+			addr = dma_map_page(priv->net_dev->dev.parent,
+					    virt_to_page(vaddr), 0,
+					    priv->rx_buf_size, DMA_BIDIRECTIONAL);
+			if (unlikely(dma_mapping_error(priv->net_dev->dev.parent, addr))) {
+				free_pages((unsigned long)vaddr, 0);
+			} else {
+				ch->buf_count++;
+				dpaa2_eth_xdp_release_buf(priv, ch, addr);
+			}
 			ch->stats.xdp_drop++;
-		else
+		} else {
 			ch->stats.xdp_redirect++;
+		}
 		break;
 	}
 

drivers/net/ethernet/freescale/enetc/Kconfig

···
 
 config FSL_ENETC_MDIO
 	tristate "ENETC MDIO driver"
-	depends on PCI
+	depends on PCI && MDIO_DEVRES && MDIO_BUS
 	help
 	  This driver supports NXP ENETC Central MDIO controller as a PCIe
 	  physical function (PF) device.

drivers/net/ethernet/freescale/enetc/enetc_pf.c

···
 	struct enetc_ndev_priv *priv;
 
 	priv = netdev_priv(si->ndev);
-	enetc_phylink_destroy(priv);
-	enetc_mdiobus_destroy(pf);
 
 	if (pf->num_vfs)
 		enetc_sriov_configure(pdev, 0);
 
 	unregister_netdev(si->ndev);
+
+	enetc_phylink_destroy(priv);
+	enetc_mdiobus_destroy(pf);
 
 	enetc_free_msix(priv);
 

drivers/net/ethernet/ibm/ibmvnic.c

···
 	if (!ltb->buff)
 		return;
 
+	/* VIOS automatically unmaps the long term buffer at remote
+	 * end for the following resets:
+	 * FAILOVER, MOBILITY, TIMEOUT.
+	 */
 	if (adapter->reset_reason != VNIC_RESET_FAILOVER &&
-	    adapter->reset_reason != VNIC_RESET_MOBILITY)
+	    adapter->reset_reason != VNIC_RESET_MOBILITY &&
+	    adapter->reset_reason != VNIC_RESET_TIMEOUT)
 		send_request_unmap(adapter, ltb->map_id);
 	dma_free_coherent(dev, ltb->size, ltb->buff, ltb->addr);
 }
···
 
 	adapter->state = VNIC_CLOSING;
 	rc = set_link_state(adapter, IBMVNIC_LOGICAL_LNK_DN);
-	if (rc)
-		return rc;
 	adapter->state = VNIC_CLOSED;
-	return 0;
+	return rc;
 }
 
 static int ibmvnic_close(struct net_device *netdev)
···
 	} else {
 		skb_copy_from_linear_data(skb, dst, skb->len);
 	}
+
+	/* post changes to long_term_buff *dst before VIOS accessing it */
+	dma_wmb();
 
 	tx_pool->consumer_index =
 	    (tx_pool->consumer_index + 1) % tx_pool->num_buffers;
···
 	unsigned long flags;
 	int ret;
 
-	/* If failover is pending don't schedule any other reset.
+	/*
+	 * If failover is pending don't schedule any other reset.
 	 * Instead let the failover complete. If there is already a
 	 * a failover reset scheduled, we will detect and drop the
 	 * duplicate reset when walking the ->rwi_list below.
···
 		goto err;
 	}
 
-	spin_lock_irqsave(&adapter->rwi_lock, flags);
-
 	list_for_each(entry, &adapter->rwi_list) {
 		tmp = list_entry(entry, struct ibmvnic_rwi, list);
 		if (tmp->reset_reason == reason) {
 			netdev_dbg(netdev, "Skipping matching reset, reason=%d\n",
 				   reason);
-			spin_unlock_irqrestore(&adapter->rwi_lock, flags);
 			ret = EBUSY;
 			goto err;
 		}
···
 
 	rwi = kzalloc(sizeof(*rwi), GFP_ATOMIC);
 	if (!rwi) {
-		spin_unlock_irqrestore(&adapter->rwi_lock, flags);
-		ibmvnic_close(netdev);
 		ret = ENOMEM;
 		goto err;
 	}
···
 	}
 	rwi->reset_reason = reason;
 	list_add_tail(&rwi->list, &adapter->rwi_list);
-	spin_unlock_irqrestore(&adapter->rwi_lock, flags);
 	netdev_dbg(adapter->netdev, "Scheduling reset (reason %d)\n", reason);
 	schedule_work(&adapter->ibmvnic_reset);
 
-	return 0;
+	ret = 0;
 err:
+	/* ibmvnic_close() below can block, so drop the lock first */
+	spin_unlock_irqrestore(&adapter->rwi_lock, flags);
+
+	if (ret == ENOMEM)
+		ibmvnic_close(netdev);
+
 	return -ret;
 }
 
···
 		offset = be16_to_cpu(next->rx_comp.off_frame_data);
 		flags = next->rx_comp.flags;
 		skb = rx_buff->skb;
+		/* load long_term_buff before copying to skb */
+		dma_rmb();
 		skb_copy_to_linear_data(skb, rx_buff->data + offset,
 					length);
 
···
 	unsigned long flags;
 
 	spin_lock_irqsave(&adapter->state_lock, flags);
+
+	/* If ibmvnic_reset() is scheduling a reset, wait for it to
+	 * finish. Then, set the state to REMOVING to prevent it from
+	 * scheduling any more work and to have reset functions ignore
+	 * any resets that have already been scheduled. Drop the lock
+	 * after setting state, so __ibmvnic_reset() which is called
+	 * from the flush_work() below, can make progress.
+	 */
+	spin_lock_irqsave(&adapter->rwi_lock, flags);
 	adapter->state = VNIC_REMOVING;
+	spin_unlock_irqrestore(&adapter->rwi_lock, flags);
+
 	spin_unlock_irqrestore(&adapter->state_lock, flags);
 
 	flush_work(&adapter->ibmvnic_reset);

drivers/net/ethernet/ibm/ibmvnic.h

···
 #define IBMVNIC_BUFFS_PER_POOL	100
 #define IBMVNIC_MAX_QUEUES	16
 #define IBMVNIC_MAX_QUEUE_SZ   4096
-#define IBMVNIC_MAX_IND_DESCS  128
+#define IBMVNIC_MAX_IND_DESCS  16
 #define IBMVNIC_IND_ARR_SZ	(IBMVNIC_MAX_IND_DESCS * 32)
 
 #define IBMVNIC_TSO_BUF_SZ	65536
···
 	/* Used for serializatin of state field */
 	spinlock_t state_lock;
 	enum ibmvnic_reset_reason reset_reason;
+	/* when taking both state and rwi locks, take state lock first */
+	spinlock_t rwi_lock;
 	struct list_head rwi_list;
 	/* Used for serialization of rwi_list */
 	spinlock_t rwi_lock;

drivers/net/ethernet/mellanox/mlx5/core/devlink.c

···
 		return -EOPNOTSUPP;
 	}
 
+	if (mlx5_lag_is_active(dev)) {
+		NL_SET_ERR_MSG_MOD(extack, "reload is unsupported in Lag mode\n");
+		return -EOPNOTSUPP;
+	}
+
 	switch (action) {
 	case DEVLINK_RELOAD_ACTION_DRIVER_REINIT:
 		mlx5_unload_one(dev, false);
···
 
 	if (new_state && !MLX5_CAP_GEN(dev, roce)) {
 		NL_SET_ERR_MSG_MOD(extack, "Device doesn't support RoCE");
+		return -EOPNOTSUPP;
+	}
+	if (mlx5_core_is_mp_slave(dev) || mlx5_lag_is_active(dev)) {
+		NL_SET_ERR_MSG_MOD(extack, "Multi port slave/Lag device can't configure RoCE");
 		return -EOPNOTSUPP;
 	}
 

drivers/net/ethernet/mellanox/mlx5/core/en/tc_ct.c

···
 #include <net/flow_offload.h>
 #include <net/netfilter/nf_flow_table.h>
 #include <linux/workqueue.h>
+#include <linux/refcount.h>
 #include <linux/xarray.h>
 
 #include "lib/fs_chains.h"
···
 	struct mlx5_flow_table *ct_nat;
 	struct mlx5_flow_table *post_ct;
 	struct mutex control_lock; /* guards parallel adds/dels */
-	struct mutex shared_counter_lock;
 	struct mapping_ctx *zone_mapping;
 	struct mapping_ctx *labels_mapping;
 	enum mlx5_flow_namespace_type ns_type;
 	struct mlx5_fs_chains *chains;
+	spinlock_t ht_lock; /* protects ft entries */
 };
 
 struct mlx5_ct_flow {
···
 	bool is_shared;
 };
 
+enum {
+	MLX5_CT_ENTRY_FLAG_VALID,
+};
+
 struct mlx5_ct_entry {
 	struct rhash_head node;
 	struct rhash_head tuple_node;
···
 	struct mlx5_ct_tuple tuple;
 	struct mlx5_ct_tuple tuple_nat;
 	struct mlx5_ct_zone_rule zone_rules[2];
+
+	struct mlx5_tc_ct_priv *ct_priv;
+	struct work_struct work;
+
+	refcount_t refcnt;
+	unsigned long flags;
 };
 
 static const struct rhashtable_params cts_ht_params = {
···
 	return err;
 }
 
+static bool
+mlx5_tc_ct_entry_valid(struct mlx5_ct_entry *entry)
+{
+	return test_bit(MLX5_CT_ENTRY_FLAG_VALID, &entry->flags);
+}
+
+static struct mlx5_ct_entry *
+mlx5_tc_ct_entry_get(struct mlx5_tc_ct_priv *ct_priv, struct mlx5_ct_tuple *tuple)
+{
+	struct mlx5_ct_entry *entry;
+
+	entry = rhashtable_lookup_fast(&ct_priv->ct_tuples_ht, tuple,
+				       tuples_ht_params);
+	if (entry && mlx5_tc_ct_entry_valid(entry) &&
+	    refcount_inc_not_zero(&entry->refcnt)) {
+		return entry;
+	} else if (!entry) {
+		entry = rhashtable_lookup_fast(&ct_priv->ct_tuples_nat_ht,
+					       tuple, tuples_nat_ht_params);
+		if (entry && mlx5_tc_ct_entry_valid(entry) &&
+		    refcount_inc_not_zero(&entry->refcnt))
+			return entry;
+	}
+
+	return entry ? ERR_PTR(-EINVAL) : NULL;
+}
+
+static void mlx5_tc_ct_entry_remove_from_tuples(struct mlx5_ct_entry *entry)
+{
+	struct mlx5_tc_ct_priv *ct_priv = entry->ct_priv;
+
+	rhashtable_remove_fast(&ct_priv->ct_tuples_nat_ht,
+			       &entry->tuple_nat_node,
+			       tuples_nat_ht_params);
+	rhashtable_remove_fast(&ct_priv->ct_tuples_ht, &entry->tuple_node,
+			       tuples_ht_params);
+}
+
+static void mlx5_tc_ct_entry_del(struct mlx5_ct_entry *entry)
+{
+	struct mlx5_tc_ct_priv *ct_priv = entry->ct_priv;
+
+	mlx5_tc_ct_entry_del_rules(ct_priv, entry);
+
+	spin_lock_bh(&ct_priv->ht_lock);
+	mlx5_tc_ct_entry_remove_from_tuples(entry);
+	spin_unlock_bh(&ct_priv->ht_lock);
+
+	mlx5_tc_ct_counter_put(ct_priv, entry);
+	kfree(entry);
+}
+
+static void
+mlx5_tc_ct_entry_put(struct mlx5_ct_entry *entry)
+{
+	if (!refcount_dec_and_test(&entry->refcnt))
+		return;
+
+	mlx5_tc_ct_entry_del(entry);
+}
+
+static void mlx5_tc_ct_entry_del_work(struct work_struct *work)
+{
+	struct mlx5_ct_entry *entry = container_of(work, struct mlx5_ct_entry, work);
+
+	mlx5_tc_ct_entry_del(entry);
+}
+
+static void
+__mlx5_tc_ct_entry_put(struct mlx5_ct_entry *entry)
+{
+	struct mlx5e_priv *priv;
+
+	if (!refcount_dec_and_test(&entry->refcnt))
+		return;
+
+	priv = netdev_priv(entry->ct_priv->netdev);
+	INIT_WORK(&entry->work, mlx5_tc_ct_entry_del_work);
+	queue_work(priv->wq, &entry->work);
+}
+
 static struct mlx5_ct_counter *
 mlx5_tc_ct_counter_create(struct mlx5_tc_ct_priv *ct_priv)
 {
···
 	}
 
 	/* Use the same counter as the reverse direction */
-	mutex_lock(&ct_priv->shared_counter_lock);
-	rev_entry = rhashtable_lookup_fast(&ct_priv->ct_tuples_ht, &rev_tuple,
-					   tuples_ht_params);
-	if (rev_entry) {
-		if (refcount_inc_not_zero(&rev_entry->counter->refcount)) {
-			mutex_unlock(&ct_priv->shared_counter_lock);
-			return rev_entry->counter;
-		}
+	spin_lock_bh(&ct_priv->ht_lock);
+	rev_entry = mlx5_tc_ct_entry_get(ct_priv, &rev_tuple);
+
+	if (IS_ERR(rev_entry)) {
+		spin_unlock_bh(&ct_priv->ht_lock);
+		goto create_counter;
 	}
-	mutex_unlock(&ct_priv->shared_counter_lock);
+
+	if (rev_entry && refcount_inc_not_zero(&rev_entry->counter->refcount)) {
+		ct_dbg("Using shared counter entry=0x%p rev=0x%p\n", entry, rev_entry);
+		shared_counter = rev_entry->counter;
+		spin_unlock_bh(&ct_priv->ht_lock);
+
+		mlx5_tc_ct_entry_put(rev_entry);
+		return shared_counter;
+	}
+
+	spin_unlock_bh(&ct_priv->ht_lock);
+
+create_counter:
 
 	shared_counter = mlx5_tc_ct_counter_create(ct_priv);
 	if (IS_ERR(shared_counter))
···
 	if (!meta_action)
 		return -EOPNOTSUPP;
 
-	entry = rhashtable_lookup_fast(&ft->ct_entries_ht, &cookie,
-				       cts_ht_params);
-	if (entry)
-		return 0;
+	spin_lock_bh(&ct_priv->ht_lock);
+	entry = rhashtable_lookup_fast(&ft->ct_entries_ht, &cookie, cts_ht_params);
+	if (entry && refcount_inc_not_zero(&entry->refcnt)) {
+		spin_unlock_bh(&ct_priv->ht_lock);
+		mlx5_tc_ct_entry_put(entry);
+		return -EEXIST;
+	}
+	spin_unlock_bh(&ct_priv->ht_lock);
 
 	entry = kzalloc(sizeof(*entry), GFP_KERNEL);
 	if (!entry)
···
 	entry->tuple.zone = ft->zone;
 	entry->cookie = flow->cookie;
 	entry->restore_cookie = meta_action->ct_metadata.cookie;
+	refcount_set(&entry->refcnt, 2);
+	entry->ct_priv = ct_priv;
 
 	err = mlx5_tc_ct_rule_to_tuple(&entry->tuple, flow_rule);
 	if (err)
···
 	if (err)
 		goto err_set;
 
-	err = rhashtable_insert_fast(&ct_priv->ct_tuples_ht,
-				     &entry->tuple_node,
-				     tuples_ht_params);
+	spin_lock_bh(&ct_priv->ht_lock);
+
+	err = rhashtable_lookup_insert_fast(&ft->ct_entries_ht, &entry->node,
+					    cts_ht_params);
+	if (err)
+		goto err_entries;
+
+	err = rhashtable_lookup_insert_fast(&ct_priv->ct_tuples_ht,
+					    &entry->tuple_node,
+					    tuples_ht_params);
 	if (err)
 		goto err_tuple;
 
 	if (memcmp(&entry->tuple, &entry->tuple_nat, sizeof(entry->tuple))) {
-		err = rhashtable_insert_fast(&ct_priv->ct_tuples_nat_ht,
-					     &entry->tuple_nat_node,
-					     tuples_nat_ht_params);
+		err = rhashtable_lookup_insert_fast(&ct_priv->ct_tuples_nat_ht,
+						    &entry->tuple_nat_node,
+						    tuples_nat_ht_params);
 		if (err)
 			goto err_tuple_nat;
 	}
+	spin_unlock_bh(&ct_priv->ht_lock);
 
 	err = mlx5_tc_ct_entry_add_rules(ct_priv, flow_rule, entry,
 					 ft->zone_restore_id);
 	if (err)
 		goto err_rules;
 
-	err = rhashtable_insert_fast(&ft->ct_entries_ht, &entry->node,
-				     cts_ht_params);
-	if (err)
-		goto err_insert;
+	set_bit(MLX5_CT_ENTRY_FLAG_VALID, &entry->flags);
+	mlx5_tc_ct_entry_put(entry); /* this function reference */
 
 	return 0;
 
-err_insert:
-	mlx5_tc_ct_entry_del_rules(ct_priv, entry);
 err_rules:
+	spin_lock_bh(&ct_priv->ht_lock);
 	if (mlx5_tc_ct_entry_has_nat(entry))
 		rhashtable_remove_fast(&ct_priv->ct_tuples_nat_ht,
 				       &entry->tuple_nat_node, tuples_nat_ht_params);
···
 			       &entry->tuple_node,
 			       tuples_ht_params);
 err_tuple:
+	rhashtable_remove_fast(&ft->ct_entries_ht,
+			       &entry->node,
+			       cts_ht_params);
+err_entries:
+	spin_unlock_bh(&ct_priv->ht_lock);
 err_set:
 	kfree(entry);
-	netdev_warn(ct_priv->netdev,
-		    "Failed to offload ct entry, err: %d\n", err);
+	if (err != -EEXIST)
+		netdev_warn(ct_priv->netdev, "Failed to offload ct entry, err: %d\n", err);
 	return err;
-}
-
-static void
-mlx5_tc_ct_del_ft_entry(struct mlx5_tc_ct_priv *ct_priv,
-			struct mlx5_ct_entry *entry)
-{
-	mlx5_tc_ct_entry_del_rules(ct_priv, entry);
-	mutex_lock(&ct_priv->shared_counter_lock);
-	if (mlx5_tc_ct_entry_has_nat(entry))
-		rhashtable_remove_fast(&ct_priv->ct_tuples_nat_ht,
-				       &entry->tuple_nat_node,
-				       tuples_nat_ht_params);
-	rhashtable_remove_fast(&ct_priv->ct_tuples_ht, &entry->tuple_node,
-			       tuples_ht_params);
-	mutex_unlock(&ct_priv->shared_counter_lock);
-	mlx5_tc_ct_counter_put(ct_priv, entry);
-
 }
 
 static int
 mlx5_tc_ct_block_flow_offload_del(struct mlx5_ct_ft *ft,
 				  struct flow_cls_offload *flow)
 {
+	struct mlx5_tc_ct_priv *ct_priv = ft->ct_priv;
 	unsigned long cookie = flow->cookie;
 	struct mlx5_ct_entry *entry;
 
-	entry = rhashtable_lookup_fast(&ft->ct_entries_ht, &cookie,
-				       cts_ht_params);
-	if (!entry)
+	spin_lock_bh(&ct_priv->ht_lock);
+	entry = rhashtable_lookup_fast(&ft->ct_entries_ht, &cookie, cts_ht_params);
+	if (!entry) {
+		spin_unlock_bh(&ct_priv->ht_lock);
 		return -ENOENT;
+	}
 
-	mlx5_tc_ct_del_ft_entry(ft->ct_priv, entry);
-	WARN_ON(rhashtable_remove_fast(&ft->ct_entries_ht,
-				       &entry->node,
-				       cts_ht_params));
-	kfree(entry);
+	if (!mlx5_tc_ct_entry_valid(entry)) {
+		spin_unlock_bh(&ct_priv->ht_lock);
+		return -EINVAL;
+	}
+
+	rhashtable_remove_fast(&ft->ct_entries_ht, &entry->node, cts_ht_params);
+	mlx5_tc_ct_entry_remove_from_tuples(entry);
+	spin_unlock_bh(&ct_priv->ht_lock);
+
+	mlx5_tc_ct_entry_put(entry);
 
 	return 0;
 }
···
 mlx5_tc_ct_block_flow_offload_stats(struct mlx5_ct_ft *ft,
 				    struct flow_cls_offload *f)
 {
+	struct mlx5_tc_ct_priv *ct_priv = ft->ct_priv;
 	unsigned long cookie = f->cookie;
 	struct mlx5_ct_entry *entry;
 	u64 lastuse, packets, bytes;
 
-	entry = rhashtable_lookup_fast(&ft->ct_entries_ht, &cookie,
-				       cts_ht_params);
-	if (!entry)
+	spin_lock_bh(&ct_priv->ht_lock);
+	entry = rhashtable_lookup_fast(&ft->ct_entries_ht, &cookie, cts_ht_params);
+	if (!entry) {
+		spin_unlock_bh(&ct_priv->ht_lock);
 		return -ENOENT;
+	}
+
+	if (!mlx5_tc_ct_entry_valid(entry) || !refcount_inc_not_zero(&entry->refcnt)) {
+		spin_unlock_bh(&ct_priv->ht_lock);
+		return -EINVAL;
+	}
+
+	spin_unlock_bh(&ct_priv->ht_lock);
 
 	mlx5_fc_query_cached(entry->counter->counter, &bytes, &packets, &lastuse);
 	flow_stats_update(&f->stats, bytes, packets, 0, lastuse,
 			  FLOW_ACTION_HW_STATS_DELAYED);
 
+	mlx5_tc_ct_entry_put(entry);
 	return 0;
 }
 
···
 static void
 mlx5_tc_ct_flush_ft_entry(void *ptr, void *arg)
 {
-	struct mlx5_tc_ct_priv *ct_priv = arg;
 	struct mlx5_ct_entry *entry = ptr;
 
-	mlx5_tc_ct_del_ft_entry(ct_priv, entry);
-	kfree(entry);
+	mlx5_tc_ct_entry_put(entry);
 }
 
 static void
···
 		goto err_mapping_labels;
 	}
 
+	spin_lock_init(&ct_priv->ht_lock);
 	ct_priv->ns_type = ns_type;
 	ct_priv->chains = chains;
 	ct_priv->netdev = priv->netdev;
···
 
 	idr_init(&ct_priv->fte_ids);
 	mutex_init(&ct_priv->control_lock);
-	mutex_init(&ct_priv->shared_counter_lock);
 	rhashtable_init(&ct_priv->zone_ht, &zone_params);
 	rhashtable_init(&ct_priv->ct_tuples_ht, &tuples_ht_params);
 	rhashtable_init(&ct_priv->ct_tuples_nat_ht, &tuples_nat_ht_params);
···
 	rhashtable_destroy(&ct_priv->ct_tuples_nat_ht);
 	rhashtable_destroy(&ct_priv->zone_ht);
 	mutex_destroy(&ct_priv->control_lock);
-	mutex_destroy(&ct_priv->shared_counter_lock);
 	idr_destroy(&ct_priv->fte_ids);
 	kfree(ct_priv);
 }
···
 	if (!mlx5_tc_ct_skb_to_tuple(skb, &tuple, zone))
 		return false;
 
-	entry = rhashtable_lookup_fast(&ct_priv->ct_tuples_ht, &tuple,
-				       tuples_ht_params);
-	if (!entry)
-		entry = rhashtable_lookup_fast(&ct_priv->ct_tuples_nat_ht,
-					       &tuple, tuples_nat_ht_params);
-	if (!entry)
+	spin_lock(&ct_priv->ht_lock);
+
+	entry = mlx5_tc_ct_entry_get(ct_priv, &tuple);
+	if (!entry) {
+		spin_unlock(&ct_priv->ht_lock);
 		return false;
+	}
+
+	if (IS_ERR(entry)) {
+		spin_unlock(&ct_priv->ht_lock);
+		return false;
+	}
+	spin_unlock(&ct_priv->ht_lock);
 
 	tcf_ct_flow_table_restore_skb(skb, entry->restore_cookie);
+	__mlx5_tc_ct_entry_put(entry);
+
 	return true;
 }

drivers/net/ethernet/mellanox/mlx5/core/en/xdp.h

···
 
 	clear_bit(MLX5E_STATE_XDP_TX_ENABLED, &priv->state);
 	/* Let other device's napi(s) and XSK wakeups see our new state. */
-	synchronize_rcu();
+	synchronize_net();
 }
 
 static inline bool mlx5e_xdp_tx_is_enabled(struct mlx5e_priv *priv)

drivers/net/ethernet/mellanox/mlx5/core/en/xsk/setup.c

···
 void mlx5e_close_xsk(struct mlx5e_channel *c)
 {
 	clear_bit(MLX5E_CHANNEL_STATE_XSK, c->state);
-	synchronize_rcu(); /* Sync with the XSK wakeup and with NAPI. */
+	synchronize_net(); /* Sync with the XSK wakeup and with NAPI. */
 
 	mlx5e_close_rq(&c->xskrq);
 	mlx5e_close_cq(&c->xskrq.cq);

drivers/net/ethernet/mellanox/mlx5/core/en_accel/en_accel.h

···
 #endif
 
 #if IS_ENABLED(CONFIG_GENEVE)
-	if (skb->encapsulation)
+	if (skb->encapsulation && skb->ip_summed == CHECKSUM_PARTIAL)
 		mlx5e_tx_tunnel_accel(skb, eseg, ihs);
 #endif
 

drivers/net/ethernet/mellanox/mlx5/core/en_accel/ktls_rx.c

···
 	struct mlx5e_ktls_rx_resync_ctx resync;
 };
 
+static bool mlx5e_ktls_priv_rx_put(struct mlx5e_ktls_offload_context_rx *priv_rx)
+{
+	if (!refcount_dec_and_test(&priv_rx->resync.refcnt))
+		return false;
+
+	kfree(priv_rx);
+	return true;
+}
+
+static void mlx5e_ktls_priv_rx_get(struct mlx5e_ktls_offload_context_rx *priv_rx)
+{
+	refcount_inc(&priv_rx->resync.refcnt);
+}
+
 static int mlx5e_ktls_create_tir(struct mlx5_core_dev *mdev, u32 *tirn, u32 rqtn)
 {
 	int err, inlen;
···
 	priv_rx = container_of(resync, struct mlx5e_ktls_offload_context_rx, resync);
 
 	if (unlikely(test_bit(MLX5E_PRIV_RX_FLAG_DELETING, priv_rx->flags))) {
-		refcount_dec(&resync->refcnt);
+		mlx5e_ktls_priv_rx_put(priv_rx);
 		return;
 	}
 
···
 	sq = &c->async_icosq;
 
 	if (resync_post_get_progress_params(sq, priv_rx))
-		refcount_dec(&resync->refcnt);
+		mlx5e_ktls_priv_rx_put(priv_rx);
 }
 
 static void resync_init(struct mlx5e_ktls_rx_resync_ctx *resync,
···
 	return err;
 }
 
-/* Function is called with elevated refcount, it decreases it. */
+/* Function can be called with the refcount being either elevated or not.
+ * It decreases the refcount and may free the kTLS priv context.
+ * Refcount is not elevated only if tls_dev_del has been called, but GET_PSV was
+ * already in flight.
+ */
 void mlx5e_ktls_handle_get_psv_completion(struct mlx5e_icosq_wqe_info *wi,
 					  struct mlx5e_icosq *sq)
 {
···
 	tls_offload_rx_resync_async_request_end(priv_rx->sk, cpu_to_be32(hw_seq));
 	priv_rx->stats->tls_resync_req_end++;
 out:
-	refcount_dec(&resync->refcnt);
+	mlx5e_ktls_priv_rx_put(priv_rx);
 	dma_unmap_single(dev, buf->dma_addr, PROGRESS_PARAMS_PADDED_SIZE, DMA_FROM_DEVICE);
 	kfree(buf);
 }
···
 		return false;
 
 	resync = &priv_rx->resync;
-	refcount_inc(&resync->refcnt);
+	mlx5e_ktls_priv_rx_get(priv_rx);
 	if (unlikely(!queue_work(resync->priv->tls->rx_wq, &resync->work)))
-		refcount_dec(&resync->refcnt);
+		mlx5e_ktls_priv_rx_put(priv_rx);
 
 	return true;
 }
···
 	return err;
 }
 
-/* Elevated refcount on the resync object means there are
- * outstanding operations (uncompleted GET_PSV WQEs) that
- * will read the resync / priv_rx objects once completed.
- * Wait for them to avoid use-after-free.
- */
-static void wait_for_resync(struct net_device *netdev,
-			    struct mlx5e_ktls_rx_resync_ctx *resync)
-{
-#define MLX5E_KTLS_RX_RESYNC_TIMEOUT 20000 /* msecs */
-	unsigned long exp_time = jiffies + msecs_to_jiffies(MLX5E_KTLS_RX_RESYNC_TIMEOUT);
-	unsigned int refcnt;
-
-	do {
-		refcnt = refcount_read(&resync->refcnt);
-		if (refcnt == 1)
-			return;
-
-		msleep(20);
-	} while (time_before(jiffies, exp_time));
-
-	netdev_warn(netdev,
-		    "Failed waiting for kTLS RX resync refcnt to be released (%u).\n",
-		    refcnt);
-}
-
 void mlx5e_ktls_del_rx(struct net_device *netdev, struct tls_context *tls_ctx)
 {
 	struct mlx5e_ktls_offload_context_rx *priv_rx;
···
 	priv_rx = mlx5e_get_ktls_rx_priv_ctx(tls_ctx);
 	set_bit(MLX5E_PRIV_RX_FLAG_DELETING, priv_rx->flags);
 	mlx5e_set_ktls_rx_priv_ctx(tls_ctx, NULL);
-	synchronize_rcu(); /* Sync with NAPI */
+	synchronize_net(); /* Sync with NAPI */
 	if (!cancel_work_sync(&priv_rx->rule.work))
 		/* completion is needed, as the priv_rx in the add flow
 		 * is maintained on the wqe info (wi), not on the socket.
···
 		wait_for_completion(&priv_rx->add_ctx);
 	resync = &priv_rx->resync;
 	if (cancel_work_sync(&resync->work))
-		refcount_dec(&resync->refcnt);
-	wait_for_resync(netdev, resync);
+		mlx5e_ktls_priv_rx_put(priv_rx);
 
 	priv_rx->stats->tls_del++;
 	if (priv_rx->rule.rule)
···
 
 	mlx5_core_destroy_tir(mdev, priv_rx->tirn);
 	mlx5_ktls_destroy_key(mdev, priv_rx->key_id);
-	kfree(priv_rx);
+	/* priv_rx should normally be freed here, but if there is an outstanding
+	 * GET_PSV, deallocation will be delayed until the CQE for GET_PSV is
+	 * processed.
+	 */
+	mlx5e_ktls_priv_rx_put(priv_rx);
 }

drivers/net/ethernet/mellanox/mlx5/core/en_ethtool.c

···
 #define MLX5E_MAX_COAL_FRAMES		MLX5_MAX_CQ_COUNT
 
 static void
-mlx5e_set_priv_channels_coalesce(struct mlx5e_priv *priv, struct ethtool_coalesce *coal)
+mlx5e_set_priv_channels_tx_coalesce(struct mlx5e_priv *priv, struct ethtool_coalesce *coal)
 {
 	struct mlx5_core_dev *mdev = priv->mdev;
 	int tc;
···
 						coal->tx_coalesce_usecs,
 						coal->tx_max_coalesced_frames);
 		}
+	}
+}
+
+static void
+mlx5e_set_priv_channels_rx_coalesce(struct mlx5e_priv *priv, struct ethtool_coalesce *coal)
+{
+	struct mlx5_core_dev *mdev = priv->mdev;
+	int i;
+
+	for (i = 0; i < priv->channels.num; ++i) {
+		struct mlx5e_channel *c = priv->channels.c[i];
 
 		mlx5_core_modify_cq_moderation(mdev, &c->rq.cq.mcq,
 					       coal->rx_coalesce_usecs,
···
 	tx_moder->pkts    = coal->tx_max_coalesced_frames;
 	new_channels.params.tx_dim_enabled = !!coal->use_adaptive_tx_coalesce;
 
-	if (!test_bit(MLX5E_STATE_OPENED, &priv->state)) {
-		priv->channels.params = new_channels.params;
-		goto out;
-	}
-	/* we are opened */
-
 	reset_rx = !!coal->use_adaptive_rx_coalesce != priv->channels.params.rx_dim_enabled;
 	reset_tx = !!coal->use_adaptive_tx_coalesce != priv->channels.params.tx_dim_enabled;
-
-	if (!reset_rx && !reset_tx) {
-		mlx5e_set_priv_channels_coalesce(priv, coal);
-		priv->channels.params = new_channels.params;
-		goto out;
-	}
 
 	if (reset_rx) {
 		u8 mode = MLX5E_GET_PFLAG(&new_channels.params,
···
 					  MLX5E_PFLAG_TX_CQE_BASED_MODER);
 
 		mlx5e_reset_tx_moderation(&new_channels.params, mode);
+	}
+
+	if (!test_bit(MLX5E_STATE_OPENED, &priv->state)) {
+		priv->channels.params = new_channels.params;
+		goto out;
+	}
+
+	if (!reset_rx && !reset_tx) {
+		if (!coal->use_adaptive_rx_coalesce)
+			mlx5e_set_priv_channels_rx_coalesce(priv, coal);
+		if (!coal->use_adaptive_tx_coalesce)
+			mlx5e_set_priv_channels_tx_coalesce(priv, coal);
+		priv->channels.params = new_channels.params;
+		goto out;
 	}
 
 	err = mlx5e_safe_switch_channels(priv, &new_channels, NULL, NULL);

drivers/net/ethernet/mellanox/mlx5/core/en_main.c

···
 #include "en/ptp.h"
 #include "qos.h"
 #include "en/trap.h"
+#include "fpga/ipsec.h"
 
 bool mlx5e_check_fragmented_striding_rq_cap(struct mlx5_core_dev *mdev)
 {
···
 	if (!mlx5e_check_fragmented_striding_rq_cap(mdev))
 		return false;
 
-	if (MLX5_IPSEC_DEV(mdev))
+	if (mlx5_fpga_is_ipsec_device(mdev))
 		return false;
 
 	if (params->xdp_prog) {
···
 void mlx5e_deactivate_rq(struct mlx5e_rq *rq)
 {
 	clear_bit(MLX5E_RQ_STATE_ENABLED, &rq->state);
-	synchronize_rcu(); /* Sync with NAPI to prevent mlx5e_post_rx_wqes. */
+	synchronize_net(); /* Sync with NAPI to prevent mlx5e_post_rx_wqes. */
 }
 
 void mlx5e_close_rq(struct mlx5e_rq *rq)
···
 	struct mlx5_wq_cyc *wq = &sq->wq;
 
 	clear_bit(MLX5E_SQ_STATE_ENABLED, &sq->state);
-	synchronize_rcu(); /* Sync with NAPI to prevent netif_tx_wake_queue. */
+	synchronize_net(); /* Sync with NAPI to prevent netif_tx_wake_queue. */
 
 	mlx5e_tx_disable_queue(sq->txq);
 
···
 void mlx5e_deactivate_icosq(struct mlx5e_icosq *icosq)
 {
 	clear_bit(MLX5E_SQ_STATE_ENABLED, &icosq->state);
-	synchronize_rcu(); /* Sync with NAPI. */
+	synchronize_net(); /* Sync with NAPI. */
 }
 
 void mlx5e_close_icosq(struct mlx5e_icosq *sq)
···
 	struct mlx5e_channel *c = sq->channel;
 
 	clear_bit(MLX5E_SQ_STATE_ENABLED, &sq->state);
-	synchronize_rcu(); /* Sync with NAPI. */
+	synchronize_net(); /* Sync with NAPI. */
 
 	mlx5e_destroy_sq(c->mdev, sq->sqn);
 	mlx5e_free_xdpsq_descs(sq);
···
 
 	mlx5e_build_create_cq_param(&ccp, c);
 
-	err = mlx5e_open_cq(c->priv, icocq_moder, &cparam->icosq.cqp, &ccp,
+	err = mlx5e_open_cq(c->priv, icocq_moder, &cparam->async_icosq.cqp, &ccp,
 			    &c->async_icosq.cq);
 	if (err)
 		return err;
 
-	err = mlx5e_open_cq(c->priv, icocq_moder, &cparam->async_icosq.cqp, &ccp,
+	err = mlx5e_open_cq(c->priv, icocq_moder, &cparam->icosq.cqp, &ccp,
 			    &c->icosq.cq);
 	if (err)
 		goto err_close_async_icosq_cq;
···
 	u32 buf_size = 0;
 	int i;
 
+<<<<<<< HEAD
 	if (MLX5_IPSEC_DEV(mdev))
+=======
+#ifdef CONFIG_MLX5_EN_IPSEC
+	if (mlx5_fpga_is_ipsec_device(mdev))
+>>>>>>> 3af409ca278d4a8d50e91f9f7c4c33b175645cf3
 		byte_count += MLX5E_METADATA_ETHER_LEN;
 
 	if (mlx5e_rx_is_linear_skb(params, xsk)) {
···
 		return -EINVAL;
 	}
 
-	if (MLX5_IPSEC_DEV(priv->mdev)) {
-		netdev_warn(netdev, "can't set XDP with IPSec offload\n");
+	if (mlx5_fpga_is_ipsec_device(priv->mdev)) {
+		netdev_warn(netdev,
+			    "XDP is not available on Innova cards with IPsec support\n");
 		return -EINVAL;
 	}
 

drivers/net/ethernet/mellanox/mlx5/core/en_rx.c

···
 		rq->dealloc_wqe = mlx5e_dealloc_rx_mpwqe;
 
 		rq->handle_rx_cqe = priv->profile->rx_handlers->handle_rx_cqe_mpwqe;
-		if (MLX5_IPSEC_DEV(mdev)) {
-			netdev_err(netdev, "MPWQE RQ with IPSec offload not supported\n");
+		if (mlx5_fpga_is_ipsec_device(mdev)) {
+			netdev_err(netdev, "MPWQE RQ with Innova IPSec offload not supported\n");
 			return -EINVAL;
 		}
 		if (!rq->handle_rx_cqe) {

drivers/net/ethernet/mellanox/mlx5/core/en_tc.c

···
 	 */
 	if (rate) {
 		rate = (rate * BITS_PER_BYTE) + 500000;
-		rate_mbps = max_t(u32, do_div(rate, 1000000), 1);
+		rate_mbps = max_t(u64, do_div(rate, 1000000), 1);
 	}
 
 	err = mlx5_esw_modify_vport_rate(esw, vport_num, rate_mbps);

drivers/net/ethernet/mellanox/mlx5/core/fpga/ipsec.c

···
 	struct ida halloc;
 };
 
-static bool mlx5_fpga_is_ipsec_device(struct mlx5_core_dev *mdev)
+bool mlx5_fpga_is_ipsec_device(struct mlx5_core_dev *mdev)
 {
 	if (!mdev->fpga || !MLX5_CAP_GEN(mdev, fpga))
 		return false;

drivers/net/ethernet/mellanox/mlx5/core/fpga/ipsec.h

···
 const struct mlx5_flow_cmds *
 mlx5_fs_cmd_get_default_ipsec_fpga_cmds(enum fs_flow_table_type type);
 void mlx5_fpga_ipsec_build_fs_cmds(void);
+bool mlx5_fpga_is_ipsec_device(struct mlx5_core_dev *mdev);
 #else
 static inline
 const struct mlx5_accel_ipsec_ops *mlx5_fpga_ipsec_ops(struct mlx5_core_dev *mdev)
···
 }
 
 static inline void mlx5_fpga_ipsec_build_fs_cmds(void) {};
+static inline bool mlx5_fpga_is_ipsec_device(struct mlx5_core_dev *mdev) { return false; }
 
 #endif /* CONFIG_MLX5_FPGA_IPSEC */
 #endif	/* __MLX5_FPGA_IPSEC_H__ */

drivers/net/ethernet/mellanox/mlx5/core/health.c

···
 	return true;
 }
 
+static void enter_error_state(struct mlx5_core_dev *dev, bool force)
+{
+	if (mlx5_health_check_fatal_sensors(dev) || force) { /* protected state setting */
+		dev->state = MLX5_DEVICE_STATE_INTERNAL_ERROR;
+		mlx5_cmd_flush(dev);
+	}
+
+	mlx5_notifier_call_chain(dev->priv.events, MLX5_DEV_EVENT_SYS_ERROR, (void *)1);
+}
+
 void mlx5_enter_error_state(struct mlx5_core_dev *dev, bool force)
 {
 	bool err_detected = false;
···
 		goto unlock;
 	}
 
-	if (mlx5_health_check_fatal_sensors(dev) || force) { /* protected state setting */
-		dev->state = MLX5_DEVICE_STATE_INTERNAL_ERROR;
-		mlx5_cmd_flush(dev);
-	}
-
-	mlx5_notifier_call_chain(dev->priv.events, MLX5_DEV_EVENT_SYS_ERROR, (void *)1);
+	enter_error_state(dev, force);
 unlock:
 	mutex_unlock(&dev->intf_state_mutex);
 }
···
 	priv = container_of(health, struct mlx5_priv, health);
 	dev = container_of(priv, struct mlx5_core_dev, priv);
 
-	mlx5_enter_error_state(dev, false);
+	enter_error_state(dev, false);
 	if (IS_ERR_OR_NULL(health->fw_fatal_reporter)) {
 		if (mlx5_health_try_recover(dev))
 			mlx5_core_err(dev, "health recovery failed\n");
···
 		mlx5_core_err(dev, "Fatal error %u detected\n", fatal_error);
 		dev->priv.health.fatal_error = fatal_error;
 		print_health_info(dev);
+		dev->state = MLX5_DEVICE_STATE_INTERNAL_ERROR;
 		mlx5_trigger_health_work(dev);
-		goto out;
+		return;
 	}
 
 	count = ioread32be(health->health_counter);

drivers/net/ethernet/mellanox/mlx5/core/main.c

···
 		dev_err(&pdev->dev, "mlx5_crdump_enable failed with error code %d\n", err);
 
 	pci_save_state(pdev);
-	devlink_reload_enable(devlink);
+	if (!mlx5_core_is_mp_slave(dev))
+		devlink_reload_enable(devlink);
 	return 0;
 
 err_load_one:

drivers/net/ethernet/realtek/r8169_main.c

···
 		phy_speed_down(tp->phydev, false);
 		rtl_wol_enable_rx(tp);
 	}
+
+	switch (tp->mac_version) {
+	case RTL_GIGA_MAC_VER_25 ... RTL_GIGA_MAC_VER_26:
+	case RTL_GIGA_MAC_VER_29 ... RTL_GIGA_MAC_VER_30:
+	case RTL_GIGA_MAC_VER_32 ... RTL_GIGA_MAC_VER_33:
+	case RTL_GIGA_MAC_VER_37:
+	case RTL_GIGA_MAC_VER_39:
+	case RTL_GIGA_MAC_VER_43:
+	case RTL_GIGA_MAC_VER_44:
+	case RTL_GIGA_MAC_VER_45:
+	case RTL_GIGA_MAC_VER_46:
+	case RTL_GIGA_MAC_VER_47:
+	case RTL_GIGA_MAC_VER_48:
+	case RTL_GIGA_MAC_VER_50 ... RTL_GIGA_MAC_VER_63:
+		RTL_W8(tp, PMCH, RTL_R8(tp, PMCH) & ~0x80);
+		break;
+	case RTL_GIGA_MAC_VER_40:
+	case RTL_GIGA_MAC_VER_41:
+	case RTL_GIGA_MAC_VER_49:
+		rtl_eri_clear_bits(tp, 0x1a8, 0xfc000000);
+		RTL_W8(tp, PMCH, RTL_R8(tp, PMCH) & ~0x80);
+		break;
+	default:
+		break;
+	}
 }
 
 static void rtl_init_rxcfg(struct rtl8169_private *tp)

drivers/net/ethernet/xilinx/xilinx_axienet_main.c

···
 	lp->options = XAE_OPTION_DEFAULTS;
 	lp->rx_bd_num = RX_BD_NUM_DEFAULT;
 	lp->tx_bd_num = TX_BD_NUM_DEFAULT;
+
+	lp->clk = devm_clk_get_optional(&pdev->dev, NULL);
+	if (IS_ERR(lp->clk)) {
+		ret = PTR_ERR(lp->clk);
+		goto free_netdev;
+	}
+	ret = clk_prepare_enable(lp->clk);
+	if (ret) {
+		dev_err(&pdev->dev, "Unable to enable clock: %d\n", ret);
+		goto free_netdev;
+	}
+
 	/* Map device registers */
 	ethres = platform_get_resource(pdev, IORESOURCE_MEM, 0);
 	lp->regs = devm_ioremap_resource(&pdev->dev, ethres);
···
 
 	lp->phy_node = of_parse_phandle(pdev->dev.of_node, "phy-handle", 0);
 	if (lp->phy_node) {
-		lp->clk = devm_clk_get(&pdev->dev, NULL);
-		if (IS_ERR(lp->clk)) {
-			dev_warn(&pdev->dev, "Failed to get clock: %ld\n",
-				 PTR_ERR(lp->clk));
-			lp->clk = NULL;
-		} else {
-			ret = clk_prepare_enable(lp->clk);
-			if (ret) {
-				dev_err(&pdev->dev, "Unable to enable clock: %d\n",
-					ret);
-				goto free_netdev;
-			}
-		}
-
 		ret = axienet_mdio_setup(lp);
 		if (ret)
 			dev_warn(&pdev->dev,

drivers/net/ipa/ipa_main.c

···
 		return -EINVAL;
 
 	for (i = 0; i < data->resource_src_count; i++)
-		ipa_resource_config_src(ipa, data->resource_src);
+		ipa_resource_config_src(ipa, &data->resource_src[i]);
 
 	for (i = 0; i < data->resource_dst_count; i++)
-		ipa_resource_config_dst(ipa, data->resource_dst);
+		ipa_resource_config_dst(ipa, &data->resource_dst[i]);
 
 	return 0;
 }

drivers/net/phy/phy_device.c

···
 
 	phydev->suspended_by_mdio_bus = 0;
 
-	ret = phy_resume(phydev);
+	ret = phy_init_hw(phydev);
 	if (ret < 0)
 		return ret;
 
+	ret = phy_resume(phydev);
+	if (ret < 0)
+		return ret;
 no_resume:
 	if (phydev->attached_dev && phydev->adjust_link)
 		phy_start_machine(phydev);
···
 	return 0;
 }
 
-static int mdio_bus_phy_restore(struct device *dev)
-{
-	struct phy_device *phydev = to_phy_device(dev);
-	struct net_device *netdev = phydev->attached_dev;
-	int ret;
-
-	if (!netdev)
-		return 0;
-
-	ret = phy_init_hw(phydev);
-	if (ret < 0)
-		return ret;
-
-	if (phydev->attached_dev && phydev->adjust_link)
-		phy_start_machine(phydev);
-
-	return 0;
-}
-
-static const struct dev_pm_ops mdio_bus_phy_pm_ops = {
-	.suspend = mdio_bus_phy_suspend,
-	.resume = mdio_bus_phy_resume,
-	.freeze = mdio_bus_phy_suspend,
-	.thaw = mdio_bus_phy_resume,
-	.restore = mdio_bus_phy_restore,
-};
-
-#define MDIO_BUS_PHY_PM_OPS (&mdio_bus_phy_pm_ops)
-
-#else
-
-#define MDIO_BUS_PHY_PM_OPS NULL
-
+static SIMPLE_DEV_PM_OPS(mdio_bus_phy_pm_ops, mdio_bus_phy_suspend,
+			 mdio_bus_phy_resume);
 #endif /* CONFIG_PM */
 
 /**
···
 	.name = "PHY",
 	.groups = phy_dev_groups,
 	.release = phy_device_release,
-	.pm = MDIO_BUS_PHY_PM_OPS,
+	.pm = pm_ptr(&mdio_bus_phy_pm_ops),
 };
 
 static int phy_request_driver_module(struct phy_device *dev, u32 phy_id)
···
 	if (ret < 0)
 		return ret;
 
-	if (phydev->drv->config_init)
+	if (phydev->drv->config_init) {
 		ret = phydev->drv->config_init(phydev);
+		if (ret < 0)
+			return ret;
+	}
 
-	return ret;
+	if (phydev->drv->config_intr) {
+		ret = phydev->drv->config_intr(phydev);
+		if (ret < 0)
+			return ret;
+	}
+
+	return 0;
 }
 EXPORT_SYMBOL(phy_init_hw);
 

drivers/net/wan/lmc/lmc_main.c

···
 	spin_lock_init(&sc->lmc_lock);
 	pci_set_master(pdev);
 
-	printk(KERN_INFO "%s: detected at %lx, irq %d\n", dev->name,
+	printk(KERN_INFO "hdlc: detected at %lx, irq %d\n",
 	       dev->base_addr, dev->irq);
 
 	err = register_hdlc_device(dev);
···
         break;
     default:
 	printk(KERN_WARNING "%s: LMC UNKNOWN CARD!\n", dev->name);
+	unregister_hdlc_device(dev);
+	return -EIO;
         break;
     }
 

drivers/net/wireless/broadcom/b43/phy_n.c

···
 
 	for (i = 0; i < 4; i++) {
 		if (dev->phy.rev >= 3)
-			table[i] = coef[i];
+			coef[i] = table[i];
 		else
 			coef[i] = 0;
 	}

drivers/target/iscsi/cxgbit/cxgbit_target.c

···
 	if (likely(cxgbit_skcb_flags(skb) & SKCBF_TX_ISO))
 		length += sizeof(struct cpl_tx_data_iso);
 
-#define MAX_IMM_TX_PKT_LEN	256
-	return length <= MAX_IMM_TX_PKT_LEN;
+	return length <= MAX_IMM_OFLD_TX_DATA_WR_LEN;
 }
 
 /*

include/linux/atmdev.h

···
 	struct atm_vcc	*vcc;		/* ATM VCC */
 	unsigned long	atm_options;	/* ATM layer options */
 	unsigned int	acct_truesize;  /* truesize accounted to vcc */
-};
+} __packed;
 
 #define VCC_HTABLE_SIZE 32
 

include/net/act_api.h

···
 			      struct nlattr *est, struct tc_action **a,
 			      const struct tc_action_ops *ops, int bind,
 			      u32 flags);
+void tcf_idr_insert_many(struct tc_action *actions[]);
 void tcf_idr_cleanup(struct tc_action_net *tn, u32 index);
 int tcf_idr_check_alloc(struct tc_action_net *tn, u32 *index,
 			struct tc_action **a, int bind);

include/uapi/linux/pkt_cls.h

···
 	TCA_FLOWER_KEY_CT_FLAGS_TRACKED = 1 << 3, /* Conntrack has occurred. */
 	TCA_FLOWER_KEY_CT_FLAGS_INVALID = 1 << 4, /* Conntrack is invalid. */
 	TCA_FLOWER_KEY_CT_FLAGS_REPLY = 1 << 5, /* Packet is in the reply direction. */
+	__TCA_FLOWER_KEY_CT_FLAGS_MAX,
 };
 
 enum {

kernel/bpf/verifier.c

···
 			bool isdiv = BPF_OP(insn->code) == BPF_DIV;
 			struct bpf_insn *patchlet;
 			struct bpf_insn chk_and_div[] = {
-				/* Rx div 0 -> 0 */
+				/* [R,W]x div 0 -> 0 */
 				BPF_RAW_INSN((is64 ? BPF_JMP : BPF_JMP32) |
 					     BPF_JNE | BPF_K, insn->src_reg,
 					     0, 2, 0),
···
 				*insn,
 			};
 			struct bpf_insn chk_and_mod[] = {
-				/* Rx mod 0 -> Rx */
+				/* [R,W]x mod 0 -> [R,W]x */
 				BPF_RAW_INSN((is64 ? BPF_JMP : BPF_JMP32) |
 					     BPF_JEQ | BPF_K, insn->src_reg,
-					     0, 1, 0),
+					     0, 1 + (is64 ? 0 : 1), 0),
 				*insn,
+				BPF_JMP_IMM(BPF_JA, 0, 0, 1),
+				BPF_MOV32_REG(insn->dst_reg, insn->dst_reg),
 			};
 
 			patchlet = isdiv ? chk_and_div : chk_and_mod;
 			cnt = isdiv ? ARRAY_SIZE(chk_and_div) :
-				      ARRAY_SIZE(chk_and_mod);
+				      ARRAY_SIZE(chk_and_mod) - (is64 ? 2 : 0);
 
 			new_prog = bpf_patch_insn_data(env, i + delta, patchlet, cnt);
 			if (!new_prog)

net/appletalk/ddp.c

···
 	struct sk_buff *skb;
 	struct net_device *dev;
 	struct ddpehdr *ddp;
-	int size;
-	struct atalk_route *rt;
+	int size, hard_header_len;
+	struct atalk_route *rt, *rt_lo = NULL;
 	int err;
 
 	if (flags & ~(MSG_DONTWAIT|MSG_CMSG_COMPAT))
···
 	SOCK_DEBUG(sk, "SK %p: Size needed %d, device %s\n",
 			sk, size, dev->name);
 
-	size += dev->hard_header_len;
+	hard_header_len = dev->hard_header_len;
+	/* Leave room for loopback hardware header if necessary */
+	if (usat->sat_addr.s_node == ATADDR_BCAST &&
+	    (dev->flags & IFF_LOOPBACK || !(rt->flags & RTF_GATEWAY))) {
+		struct atalk_addr at_lo;
+
+		at_lo.s_node = 0;
+		at_lo.s_net  = 0;
+
+		rt_lo = atrtr_find(&at_lo);
+
+		if (rt_lo && rt_lo->dev->hard_header_len > hard_header_len)
+			hard_header_len = rt_lo->dev->hard_header_len;
+	}
+
+	size += hard_header_len;
 	release_sock(sk);
 	skb = sock_alloc_send_skb(sk, size, (flags & MSG_DONTWAIT), &err);
 	lock_sock(sk);
···
 		goto out;
 
 	skb_reserve(skb, ddp_dl->header_length);
-	skb_reserve(skb, dev->hard_header_len);
+	skb_reserve(skb, hard_header_len);
 	skb->dev = dev;
 
 	SOCK_DEBUG(sk, "SK %p: Begin build.\n", sk);
···
 		/* loop back */
 		skb_orphan(skb);
 		if (ddp->deh_dnode == ATADDR_BCAST) {
-			struct atalk_addr at_lo;
-
-			at_lo.s_node = 0;
-			at_lo.s_net  = 0;
-
-			rt = atrtr_find(&at_lo);
-			if (!rt) {
+			if (!rt_lo) {
 				kfree_skb(skb);
 				err = -ENETUNREACH;
 				goto out;
 			}
-			dev = rt->dev;
+			dev = rt_lo->dev;
 			skb->dev = dev;
 		}
 		ddp_dl->request(ddp_dl, skb, dev->dev_addr);

net/caif/chnl_net.c

···
 	else
 		skb->ip_summed = CHECKSUM_NONE;
 
-	if (in_interrupt())
-		netif_rx(skb);
-	else
-		netif_rx_ni(skb);
+	netif_rx_any_context(skb);
 
 	/* Update statistics. */
 	priv->netdev->stats.rx_packets++;

net/core/flow_dissector.c

···
 			key_control->addr_type = FLOW_DISSECTOR_KEY_IPV4_ADDRS;
 		}
 
+		__skb_flow_dissect_ipv4(skb, flow_dissector,
+					target_container, data, iph);
+
 		if (ip_is_fragment(iph)) {
 			key_control->flags |= FLOW_DIS_IS_FRAGMENT;
 
···
 				}
 			}
 		}
-
-		__skb_flow_dissect_ipv4(skb, flow_dissector,
-					target_container, data, iph);
 
 		break;
 	}

net/mptcp/options.c

···
 {
 	struct mptcp_subflow_context *subflow = mptcp_subflow_ctx(sk);
 	struct mptcp_sock *msk = mptcp_sk(subflow->conn);
+	u64 snd_data_fin_enable, ack_seq;
 	unsigned int dss_size = 0;
-	u64 snd_data_fin_enable;
 	struct mptcp_ext *mpext;
 	unsigned int ack_size;
 	bool ret = false;
···
 		return ret;
 	}
 
+	ack_seq = READ_ONCE(msk->ack_seq);
 	if (READ_ONCE(msk->use_64bit_ack)) {
 		ack_size = TCPOLEN_MPTCP_DSS_ACK64;
-		opts->ext_copy.data_ack = READ_ONCE(msk->ack_seq);
+		opts->ext_copy.data_ack = ack_seq;
 		opts->ext_copy.ack64 = 1;
 	} else {
 		ack_size = TCPOLEN_MPTCP_DSS_ACK32;
-		opts->ext_copy.data_ack32 = (uint32_t)READ_ONCE(msk->ack_seq);
+		opts->ext_copy.data_ack32 = (uint32_t)ack_seq;
 		opts->ext_copy.ack64 = 0;
 	}
 	opts->ext_copy.use_ack = 1;
···
 		msk->wnd_end = new_wnd_end;
 
 	/* this assumes mptcp_incoming_options() is invoked after tcp_ack() */
-	if (after64(msk->wnd_end, READ_ONCE(msk->snd_nxt)) &&
-	    sk_stream_memory_free(ssk))
+	if (after64(msk->wnd_end, READ_ONCE(msk->snd_nxt)))
 		__mptcp_check_push(sk, ssk);
 
 	if (after64(new_snd_una, old_snd_una)) {

net/mptcp/protocol.c

···
 
 	/* Look for an acknowledged DATA_FIN */
 	if (mptcp_pending_data_fin_ack(sk)) {
-		mptcp_stop_timer(sk);
-
 		WRITE_ONCE(msk->snd_data_fin_enable, 0);
 
 		switch (sk->sk_state) {
···
 static void mptcp_cleanup_rbuf(struct mptcp_sock *msk)
 {
 	struct sock *ack_hint = READ_ONCE(msk->ack_hint);
+	int old_space = READ_ONCE(msk->old_wspace);
 	struct mptcp_subflow_context *subflow;
+	struct sock *sk = (struct sock *)msk;
+	bool cleanup;
+
+	/* this is a simple superset of what tcp_cleanup_rbuf() implements
+	 * so that we don't have to acquire the ssk socket lock most of the time
+	 * to do actually nothing
+	 */
+	cleanup = __mptcp_space(sk) - old_space >= max(0, old_space);
+	if (!cleanup)
+		return;
 
 	/* if the hinted ssk is still active, try to use it */
 	if (likely(ack_hint)) {
···
 		mptcp_set_timeout(sk, ssk);
 		tcp_push(ssk, 0, info.mss_now, tcp_sk(ssk)->nonagle,
 			 info.size_goal);
+		if (!mptcp_timer_pending(sk))
+			mptcp_reset_timer(sk);
+
 		if (msk->snd_data_fin_enable &&
 		    msk->snd_nxt + 1 == msk->write_seq)
 			mptcp_schedule_work(sk);
···
 	skb_queue_splice_tail_init(&sk->sk_receive_queue, &msk->receive_queue);
 }
 
-static bool __mptcp_move_skbs(struct mptcp_sock *msk, unsigned int rcv)
+static bool __mptcp_move_skbs(struct mptcp_sock *msk)
 {
 	struct sock *sk = (struct sock *)msk;
 	unsigned int moved = 0;
···
 
 		slowpath = lock_sock_fast(ssk);
 		mptcp_data_lock(sk);
+		__mptcp_update_rmem(sk);
 		done = __mptcp_move_skbs_from_subflow(msk, ssk, &moved);
 		mptcp_data_unlock(sk);
-		if (moved && rcv) {
-			WRITE_ONCE(msk->rmem_pending, min(rcv, moved));
-			tcp_cleanup_rbuf(ssk, 1);
-			WRITE_ONCE(msk->rmem_pending, 0);
-		}
+		tcp_cleanup_rbuf(ssk, moved);
 		unlock_sock_fast(ssk, slowpath);
 	} while (!done);
 
···
 		ret |= __mptcp_ofo_queue(msk);
 		__mptcp_splice_receive_queue(sk);
 		mptcp_data_unlock(sk);
+		mptcp_cleanup_rbuf(msk);
 	}
 	if (ret)
 		mptcp_check_data_fin((struct sock *)msk);
···
 	target = sock_rcvlowat(sk, flags & MSG_WAITALL, len);
 
 	while (copied < len) {
-		int bytes_read, old_space;
+		int bytes_read;
 
 		bytes_read = __mptcp_recvmsg_mskq(msk, msg, len - copied);
 		if (unlikely(bytes_read < 0)) {
···
 
 		copied += bytes_read;
 
-		if (skb_queue_empty(&msk->receive_queue) &&
-		    __mptcp_move_skbs(msk, len - copied))
-			continue;
-
 		/* be sure to advertise window change */
-		old_space = READ_ONCE(msk->old_wspace);
-		if ((tcp_space(sk) - old_space) >= old_space)
-			mptcp_cleanup_rbuf(msk);
+		mptcp_cleanup_rbuf(msk);
+
+		if (skb_queue_empty(&msk->receive_queue) && __mptcp_move_skbs(msk))
+			continue;
 
 		/* only the master socket status is relevant here. The exit
 		 * conditions mirror closely tcp_recvmsg()
···
 				/* race breaker: the shutdown could be after the
 				 * previous receive queue check
 				 */
-				if (__mptcp_move_skbs(msk, len - copied))
+				if (__mptcp_move_skbs(msk))
 					continue;
 				break;
 			}
···
 		/* .. race-breaker: ssk might have gotten new data
 		 * after last __mptcp_move_skbs() returned false.
 		 */
-		if (unlikely(__mptcp_move_skbs(msk, 0)))
+		if (unlikely(__mptcp_move_skbs(msk)))
 			set_bit(MPTCP_DATA_READY, &msk->flags);
 	} else if (unlikely(!test_bit(MPTCP_DATA_READY, &msk->flags))) {
 		/* data to read but mptcp_wait_data() cleared DATA_READY */
···
 	if (!test_and_clear_bit(MPTCP_WORK_RTX, &msk->flags))
 		goto unlock;
 
+	__mptcp_clean_una(sk);
 	dfrag = mptcp_rtx_head(sk);
 	if (!dfrag)
 		goto unlock;
···
 		mptcp_push_pending(sk, 0);
 		spin_lock_bh(&sk->sk_lock.slock);
 	}
+	if (test_and_clear_bit(MPTCP_ERROR_REPORT, &mptcp_sk(sk)->flags))
+		__mptcp_error_report(sk);
 
 	/* clear any wmem reservation and errors */
 	__mptcp_update_wmem(sk);
···
 	struct sock *sk = (struct sock *)msk;
 
 	if (unlikely(sk->sk_shutdown & SEND_SHUTDOWN))
-		return 0;
+		return EPOLLOUT | EPOLLWRNORM;
 
 	if (sk_stream_is_writeable(sk))
 		return EPOLLOUT | EPOLLWRNORM;
···
 		mask |= mptcp_check_readable(msk);
 		mask |= mptcp_check_writeable(msk);
 	}
+	if (sk->sk_shutdown == SHUTDOWN_MASK || state == TCP_CLOSE)
+		mask |= EPOLLHUP;
 	if (sk->sk_shutdown & RCV_SHUTDOWN)
 		mask |= EPOLLIN | EPOLLRDNORM | EPOLLRDHUP;
+
+	/* This barrier is coupled with smp_wmb() in tcp_reset() */
+	smp_rmb();
+	if (sk->sk_err)
+		mask |= EPOLLERR;
 
 	return mask;
 }

net/mptcp/protocol.h

···
 #define MPTCP_WORK_CLOSE_SUBFLOW 5
 #define MPTCP_PUSH_PENDING	6
 #define MPTCP_CLEAN_UNA		7
+#define MPTCP_ERROR_REPORT	8
 
 static inline bool before64(__u64 seq1, __u64 seq2)
 {
···
 	u64		wnd_end;
 	unsigned long	timer_ival;
 	u32		token;
-	int		rmem_pending;
 	int		rmem_released;
 	unsigned long	flags;
 	bool		can_ack;
···
 
 static inline int __mptcp_space(const struct sock *sk)
 {
-	return tcp_space(sk) + READ_ONCE(mptcp_sk(sk)->rmem_pending);
+	return tcp_space(sk) + READ_ONCE(mptcp_sk(sk)->rmem_released);
 }
 
 static inline struct mptcp_data_frag *mptcp_send_head(const struct sock *sk)
···
 	return list_last_entry(&msk->rtx_queue, struct mptcp_data_frag, list);
 }
 
-static inline struct mptcp_data_frag *mptcp_rtx_tail(const struct sock *sk)
-{
-	struct mptcp_sock *msk = mptcp_sk(sk);
-
-	if (!before64(msk->snd_nxt, READ_ONCE(msk->snd_una)))
-		return NULL;
-
-	return list_last_entry(&msk->rtx_queue, struct mptcp_data_frag, list);
-}
-
 static inline struct mptcp_data_frag *mptcp_rtx_head(const struct sock *sk)
 {
 	struct mptcp_sock *msk = mptcp_sk(sk);
+
+	if (msk->snd_una == READ_ONCE(msk->snd_nxt))
+		return NULL;
 
 	return list_first_entry_or_null(&msk->rtx_queue, struct mptcp_data_frag, list);
 }
···
 	void	(*tcp_data_ready)(struct sock *sk);
 	void	(*tcp_state_change)(struct sock *sk);
 	void	(*tcp_write_space)(struct sock *sk);
+	void	(*tcp_error_report)(struct sock *sk);
 
 	struct	rcu_head rcu;
 };
···
 	sk->sk_data_ready = ctx->tcp_data_ready;
 	sk->sk_state_change = ctx->tcp_state_change;
 	sk->sk_write_space = ctx->tcp_write_space;
+	sk->sk_error_report = ctx->tcp_error_report;
 
 	inet_csk(sk)->icsk_af_ops = ctx->icsk_af_ops;
 }
···
 bool mptcp_schedule_work(struct sock *sk);
 void __mptcp_check_push(struct sock *sk, struct sock *ssk);
 void __mptcp_data_acked(struct sock *sk);
+void __mptcp_error_report(struct sock *sk);
 void mptcp_subflow_eof(struct sock *sk);
 bool mptcp_update_rcv_data_fin(struct mptcp_sock *msk, u64 data_fin_seq, bool use_64bit);
 void __mptcp_flush_join_list(struct mptcp_sock *msk);

net/mptcp/subflow.c

···
 	return msk;
 }
 
-static int __subflow_init_req(struct request_sock *req, const struct sock *sk_listener)
+static void subflow_init_req(struct request_sock *req, const struct sock *sk_listener)
 {
 	struct mptcp_subflow_request_sock *subflow_req = mptcp_subflow_rsk(req);
 
···
 	subflow_req->mp_join = 0;
 	subflow_req->msk = NULL;
 	mptcp_token_init_request(req);
-
-#ifdef CONFIG_TCP_MD5SIG
-	/* no MPTCP if MD5SIG is enabled on this socket or we may run out of
-	 * TCP option space.
-	 */
-	if (rcu_access_pointer(tcp_sk(sk_listener)->md5sig_info))
-		return -EINVAL;
-#endif
-
-	return 0;
 }
 
 static bool subflow_use_different_sport(struct mptcp_sock *msk, const struct sock *sk)
···
  * Returns an error code if a JOIN has failed and a TCP reset
  * should be sent.
  */
-static int subflow_init_req(struct request_sock *req,
-			    const struct sock *sk_listener,
-			    struct sk_buff *skb)
+static int subflow_check_req(struct request_sock *req,
+			     const struct sock *sk_listener,
+			     struct sk_buff *skb)
 {
 	struct mptcp_subflow_context *listener = mptcp_subflow_ctx(sk_listener);
 	struct mptcp_subflow_request_sock *subflow_req = mptcp_subflow_rsk(req);
 	struct mptcp_options_received mp_opt;
-	int ret;
 
 	pr_debug("subflow_req=%p, listener=%p", subflow_req, listener);
 
-	ret = __subflow_init_req(req, sk_listener);
-	if (ret)
-		return 0;
+#ifdef CONFIG_TCP_MD5SIG
+	/* no MPTCP if MD5SIG is enabled on this socket or we may run out of
+	 * TCP option space.
+	 */
+	if (rcu_access_pointer(tcp_sk(sk_listener)->md5sig_info))
+		return -EINVAL;
+#endif
 
 	mptcp_get_options(skb, &mp_opt);
 
···
 	struct mptcp_options_received mp_opt;
 	int err;
 
-	err = __subflow_init_req(req, sk_listener);
-	if (err)
-		return err;
-
+	subflow_init_req(req, sk_listener);
 	mptcp_get_options(skb, &mp_opt);
 
 	if (mp_opt.mp_capable && mp_opt.mp_join)
···
 	int err;
 
 	tcp_rsk(req)->is_mptcp = 1;
+	subflow_init_req(req, sk);
 
 	dst = tcp_request_sock_ipv4_ops.route_req(sk, skb, fl, req);
 	if (!dst)
 		return NULL;
 
-	err = subflow_init_req(req, sk, skb);
+	err = subflow_check_req(req, sk, skb);
 	if (err == 0)
 		return dst;
 
···
 	int err;
 
 	tcp_rsk(req)->is_mptcp = 1;
+	subflow_init_req(req, sk);
 
 	dst = tcp_request_sock_ipv6_ops.route_req(sk, skb, fl, req);
 	if (!dst)
 		return NULL;
 
-	err = subflow_init_req(req, sk, skb);
+	err = subflow_check_req(req, sk, skb);
 	if (err == 0)
 		return dst;
 
···
 	mptcp_write_space(sk);
 }
 
+void __mptcp_error_report(struct sock *sk)
+{
+	struct mptcp_subflow_context *subflow;
+	struct mptcp_sock *msk = mptcp_sk(sk);
+
+	mptcp_for_each_subflow(msk, subflow) {
+		struct sock *ssk = mptcp_subflow_tcp_sock(subflow);
+		int err = sock_error(ssk);
+
+		if (!err)
+			continue;
+
+		/* only propagate errors on fallen-back sockets or
+		 * on MPC connect
+		 */
+		if (sk->sk_state != TCP_SYN_SENT && !__mptcp_check_fallback(msk))
+			continue;
+
+		inet_sk_state_store(sk, inet_sk_state_load(ssk));
+		sk->sk_err = -err;
+
+		/* This barrier is coupled with smp_rmb() in mptcp_poll() */
+		smp_wmb();
+		sk->sk_error_report(sk);
+		break;
+	}
+}
+
+static void subflow_error_report(struct sock *ssk)
+{
+	struct sock *sk = mptcp_subflow_ctx(ssk)->conn;
+
+	mptcp_data_lock(sk);
+	if (!sock_owned_by_user(sk))
+		__mptcp_error_report(sk);
+	else
+		set_bit(MPTCP_ERROR_REPORT,  &mptcp_sk(sk)->flags);
+	mptcp_data_unlock(sk);
+}
+
 static struct inet_connection_sock_af_ops *
 subflow_default_af_ops(struct sock *sk)
 {
···
 	ctx->tcp_data_ready = sk->sk_data_ready;
 	ctx->tcp_state_change = sk->sk_state_change;
 	ctx->tcp_write_space = sk->sk_write_space;
+	ctx->tcp_error_report = sk->sk_error_report;
 	sk->sk_data_ready = subflow_data_ready;
 	sk->sk_write_space = subflow_write_space;
 	sk->sk_state_change = subflow_state_change;
+	sk->sk_error_report = subflow_error_report;
 out:
 	return err;
 }
···
 	new_ctx->tcp_data_ready = old_ctx->tcp_data_ready;
 	new_ctx->tcp_state_change = old_ctx->tcp_state_change;
 	new_ctx->tcp_write_space = old_ctx->tcp_write_space;
+	new_ctx->tcp_error_report = old_ctx->tcp_error_report;
 	new_ctx->rel_write_seq = 1;
 	new_ctx->tcp_sock = newsk;
 

net/sched/act_api.c

···
 	[TCA_ACT_HW_STATS]	= NLA_POLICY_BITFIELD32(TCA_ACT_HW_STATS_ANY),
 };
 
-static void tcf_idr_insert_many(struct tc_action *actions[])
+void tcf_idr_insert_many(struct tc_action *actions[])
 {
 	int i;
 

net/sched/cls_api.c

···
 			act->type = exts->type = TCA_OLD_COMPAT;
 			exts->actions[0] = act;
 			exts->nr_actions = 1;
+			tcf_idr_insert_many(exts->actions);
 		} else if (exts->action && tb[exts->action]) {
 			int err;
 

net/sched/cls_flower.c

···
 
 #include <uapi/linux/netfilter/nf_conntrack_common.h>
 
+#define TCA_FLOWER_KEY_CT_FLAGS_MAX \
+		((__TCA_FLOWER_KEY_CT_FLAGS_MAX - 1) << 1)
+#define TCA_FLOWER_KEY_CT_FLAGS_MASK \
+		(TCA_FLOWER_KEY_CT_FLAGS_MAX - 1)
+
 struct fl_flow_key {
 	struct flow_dissector_key_meta meta;
 	struct flow_dissector_key_control control;
···
 	[TCA_FLOWER_KEY_ENC_IP_TTL_MASK] = { .type = NLA_U8 },
 	[TCA_FLOWER_KEY_ENC_OPTS]	= { .type = NLA_NESTED },
 	[TCA_FLOWER_KEY_ENC_OPTS_MASK]	= { .type = NLA_NESTED },
-	[TCA_FLOWER_KEY_CT_STATE]	= { .type = NLA_U16 },
-	[TCA_FLOWER_KEY_CT_STATE_MASK]	= { .type = NLA_U16 },
+	[TCA_FLOWER_KEY_CT_STATE]	=
+		NLA_POLICY_MASK(NLA_U16, TCA_FLOWER_KEY_CT_FLAGS_MASK),
+	[TCA_FLOWER_KEY_CT_STATE_MASK]	=
+		NLA_POLICY_MASK(NLA_U16, TCA_FLOWER_KEY_CT_FLAGS_MASK),
 	[TCA_FLOWER_KEY_CT_ZONE]	= { .type = NLA_U16 },
 	[TCA_FLOWER_KEY_CT_ZONE_MASK]	= { .type = NLA_U16 },
 	[TCA_FLOWER_KEY_CT_MARK]	= { .type = NLA_U32 },
···
 	return 0;
 }
 
+static int fl_validate_ct_state(u16 state, struct nlattr *tb,
+				struct netlink_ext_ack *extack)
+{
+	if (state && !(state & TCA_FLOWER_KEY_CT_FLAGS_TRACKED)) {
+		NL_SET_ERR_MSG_ATTR(extack, tb,
+				    "no trk, so no other flag can be set");
+		return -EINVAL;
+	}
+
+	if (state & TCA_FLOWER_KEY_CT_FLAGS_NEW &&
+	    state & TCA_FLOWER_KEY_CT_FLAGS_ESTABLISHED) {
+		NL_SET_ERR_MSG_ATTR(extack, tb,
+				    "new and est are mutually exclusive");
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
 static int fl_set_key_ct(struct nlattr **tb,
 			 struct flow_dissector_key_ct *key,
 			 struct flow_dissector_key_ct *mask,
 			 struct netlink_ext_ack *extack)
 {
 	if (tb[TCA_FLOWER_KEY_CT_STATE]) {
+		int err;
+
 		if (!IS_ENABLED(CONFIG_NF_CONNTRACK)) {
 			NL_SET_ERR_MSG(extack, "Conntrack isn't enabled");
 			return -EOPNOTSUPP;
···
 		fl_set_key_val(tb, &key->ct_state, TCA_FLOWER_KEY_CT_STATE,
 			       &mask->ct_state, TCA_FLOWER_KEY_CT_STATE_MASK,
 			       sizeof(key->ct_state));
+
+		err = fl_validate_ct_state(mask->ct_state,
+					   tb[TCA_FLOWER_KEY_CT_STATE_MASK],
+					   extack);
+		if (err)
+			return err;
+
 	}
 	if (tb[TCA_FLOWER_KEY_CT_ZONE]) {
 		if (!IS_ENABLED(CONFIG_NF_CONNTRACK_ZONES)) {

tools/testing/selftests/bpf/test_xdp_redirect.sh

···
-#!/bin/sh
+#!/bin/bash
 # Create 2 namespaces with two veth peers, and
 # forward packets in-between using generic XDP
 #
···
 	ip link set dev veth1 $xdpmode obj test_xdp_redirect.o sec redirect_to_222 &> /dev/null
 	ip link set dev veth2 $xdpmode obj test_xdp_redirect.o sec redirect_to_111 &> /dev/null
 
-	ip netns exec ns1 ping -c 1 10.1.1.22 &> /dev/null
-	local ret1=$?
-	ip netns exec ns2 ping -c 1 10.1.1.11 &> /dev/null
-	local ret2=$?
-
-	if [ $ret1 -eq 0 -a $ret2 -eq 0 ]; then
+	if ip netns exec ns1 ping -c 1 10.1.1.22 &> /dev/null &&
+	   ip netns exec ns2 ping -c 1 10.1.1.11 &> /dev/null; then
 		echo "selftests: test_xdp_redirect $xdpmode [PASS]";
 	else
 		ret=1

tools/testing/selftests/net/forwarding/tc_flower.sh

···
 
 ALL_TESTS="match_dst_mac_test match_src_mac_test match_dst_ip_test \
 	match_src_ip_test match_ip_flags_test match_pcp_test match_vlan_test \
-	match_ip_tos_test match_indev_test match_mpls_label_test \
+	match_ip_tos_test match_indev_testmatch_ip_ttl_test match_mpls_label_test \
 	match_mpls_tc_test match_mpls_bos_test match_mpls_ttl_test \
 	match_mpls_lse_test"
 NUM_NETIFS=2
···
 	tc filter del dev $h2 ingress protocol ip pref 1 handle 101 flower
 
 	log_test "ip_tos match ($tcflags)"
+}
+
+match_ip_ttl_test()
+{
+	RET=0
+
+	tc filter add dev $h2 ingress protocol ip pref 1 handle 101 flower \
+		$tcflags dst_ip 192.0.2.2 ip_ttl 63 action drop
+	tc filter add dev $h2 ingress protocol ip pref 2 handle 102 flower \
+		$tcflags dst_ip 192.0.2.2 action drop
+
+	$MZ $h1 -c 1 -p 64 -a $h1mac -b $h2mac -A 192.0.2.1 -B 192.0.2.2 \
+		-t ip "ttl=63" -q
+
+	$MZ $h1 -c 1 -p 64 -a $h1mac -b $h2mac -A 192.0.2.1 -B 192.0.2.2 \
+		-t ip "ttl=63,mf,frag=256" -q
+
+	tc_check_packets "dev $h2 ingress" 102 1
+	check_fail $? "Matched on the wrong filter (no check on ttl)"
+
+	tc_check_packets "dev $h2 ingress" 101 2
+	check_err $? "Did not match on correct filter (ttl=63)"
+
+	$MZ $h1 -c 1 -p 64 -a $h1mac -b $h2mac -A 192.0.2.1 -B 192.0.2.2 \
+		-t ip "ttl=255" -q
+
+	tc_check_packets "dev $h2 ingress" 101 3
+	check_fail $? "Matched on a wrong filter (ttl=63)"
+
+	tc_check_packets "dev $h2 ingress" 102 1
+	check_err $? "Did not match on correct filter (no check on ttl)"
+
+	tc filter del dev $h2 ingress protocol ip pref 2 handle 102 flower
+	tc filter del dev $h2 ingress protocol ip pref 1 handle 101 flower
+
+	log_test "ip_ttl match ($tcflags)"
 }
 
 match_indev_test()