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

Pull networking fixes from David Miller:

 1) Limit xt_hashlimit hash table size to avoid OOM or hung tasks, from
    Cong Wang.

 2) Fix deadlock in xsk by publishing global consumer pointers when NAPI
    is finished, from Magnus Karlsson.

 3) Set table field properly to RT_TABLE_COMPAT when necessary, from
    Jethro Beekman.

 4) NLA_STRING attributes are not necessary NULL terminated, deal wiht
    that in IFLA_ALT_IFNAME. From Eric Dumazet.

 5) Fix checksum handling in atlantic driver, from Dmitry Bezrukov.

 6) Handle mtu==0 devices properly in wireguard, from Jason A.
    Donenfeld.

 7) Fix several lockdep warnings in bonding, from Taehee Yoo.

 8) Fix cls_flower port blocking, from Jason Baron.

 9) Sanitize internal map names in libbpf, from Toke Høiland-Jørgensen.

10) Fix RDMA race in qede driver, from Michal Kalderon.

11) Fix several false lockdep warnings by adding conditions to
    list_for_each_entry_rcu(), from Madhuparna Bhowmik.

12) Fix sleep in atomic in mlx5 driver, from Huy Nguyen.

13) Fix potential deadlock in bpf_map_do_batch(), from Yonghong Song.

14) Hey, variables declared in switch statement before any case
    statements are not initialized. I learn something every day. Get
    rids of this stuff in several parts of the networking, from Kees
    Cook.

* git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net: (99 commits)
  bnxt_en: Issue PCIe FLR in kdump kernel to cleanup pending DMAs.
  bnxt_en: Improve device shutdown method.
  net: netlink: cap max groups which will be considered in netlink_bind()
  net: thunderx: workaround BGX TX Underflow issue
  ionic: fix fw_status read
  net: disable BRIDGE_NETFILTER by default
  net: macb: Properly handle phylink on at91rm9200
  s390/qeth: fix off-by-one in RX copybreak check
  s390/qeth: don't warn for napi with 0 budget
  s390/qeth: vnicc Fix EOPNOTSUPP precedence
  openvswitch: Distribute switch variables for initialization
  net: ip6_gre: Distribute switch variables for initialization
  net: core: Distribute switch variables for initialization
  udp: rehash on disconnect
  net/tls: Fix to avoid gettig invalid tls record
  bpf: Fix a potential deadlock with bpf_map_do_batch
  bpf: Do not grab the bucket spinlock by default on htab batch ops
  ice: Wait for VF to be reset/ready before configuration
  ice: Don't tell the OS that link is going down
  ice: Don't reject odd values of usecs set by user
  ...

4 files changed, 115 insertions, 80 deletions

diff --git a/drivers/net/ethernet/intel/ice/ice_ethtool.c b/drivers/net/ethernet/intel/ice/ice_ethtool.c
index b002ab4e5838..77c412a7e7a4 100644
--- a/drivers/net/ethernet/intel/ice/ice_ethtool.c
+++ b/drivers/net/ethernet/intel/ice/ice_ethtool.c
@@ -2936,13 +2936,6 @@ ice_set_pauseparam(struct net_device *netdev, struct ethtool_pauseparam *pause)
 	else
 		return -EINVAL;
 
-	/* Tell the OS link is going down, the link will go back up when fw
-	 * says it is ready asynchronously
-	 */
-	ice_print_link_msg(vsi, false);
-	netif_carrier_off(netdev);
-	netif_tx_stop_all_queues(netdev);
-
 	/* Set the FC mode and only restart AN if link is up */
 	status = ice_set_fc(pi, &aq_failures, link_up);
 
@@ -3489,21 +3482,13 @@ ice_set_rc_coalesce(enum ice_container_type c_type, struct ethtool_coalesce *ec,
 		return -EINVAL;
 	}
 
-	/* hardware only supports an ITR granularity of 2us */
-	if (coalesce_usecs % 2 != 0) {
-		netdev_info(vsi->netdev, "Invalid value, %s-usecs must be even\n",
-			    c_type_str);
-		return -EINVAL;
-	}
-
 	if (use_adaptive_coalesce) {
 		rc->itr_setting |= ICE_ITR_DYNAMIC;
 	} else {
-		/* store user facing value how it was set */
+		/* save the user set usecs */
 		rc->itr_setting = coalesce_usecs;
-		/* set to static and convert to value HW understands */
-		rc->target_itr =
-			ITR_TO_REG(ITR_REG_ALIGN(rc->itr_setting));
+		/* device ITR granularity is in 2 usec increments */
+		rc->target_itr = ITR_REG_ALIGN(rc->itr_setting);
 	}
 
 	return 0;
@@ -3597,6 +3582,30 @@ ice_is_coalesce_param_invalid(struct net_device *netdev,
 }
 
 /**
+ * ice_print_if_odd_usecs - print message if user tries to set odd [tx|rx]-usecs
+ * @netdev: netdev used for print
+ * @itr_setting: previous user setting
+ * @use_adaptive_coalesce: if adaptive coalesce is enabled or being enabled
+ * @coalesce_usecs: requested value of [tx|rx]-usecs
+ * @c_type_str: either "rx" or "tx" to match user set field of [tx|rx]-usecs
+ */
+static void
+ice_print_if_odd_usecs(struct net_device *netdev, u16 itr_setting,
+		       u32 use_adaptive_coalesce, u32 coalesce_usecs,
+		       const char *c_type_str)
+{
+	if (use_adaptive_coalesce)
+		return;
+
+	itr_setting = ITR_TO_REG(itr_setting);
+
+	if (itr_setting != coalesce_usecs && (coalesce_usecs % 2))
+		netdev_info(netdev, "User set %s-usecs to %d, device only supports even values. Rounding down and attempting to set %s-usecs to %d\n",
+			    c_type_str, coalesce_usecs, c_type_str,
+			    ITR_REG_ALIGN(coalesce_usecs));
+}
+
+/**
  * __ice_set_coalesce - set ITR/INTRL values for the device
  * @netdev: pointer to the netdev associated with this query
  * @ec: ethtool structure to fill with driver's coalesce settings
@@ -3616,8 +3625,19 @@ __ice_set_coalesce(struct net_device *netdev, struct ethtool_coalesce *ec,
 		return -EINVAL;
 
 	if (q_num < 0) {
+		struct ice_q_vector *q_vector = vsi->q_vectors[0];
 		int v_idx;
 
+		if (q_vector) {
+			ice_print_if_odd_usecs(netdev, q_vector->rx.itr_setting,
+					       ec->use_adaptive_rx_coalesce,
+					       ec->rx_coalesce_usecs, "rx");
+
+			ice_print_if_odd_usecs(netdev, q_vector->tx.itr_setting,
+					       ec->use_adaptive_tx_coalesce,
+					       ec->tx_coalesce_usecs, "tx");
+		}
+
 		ice_for_each_q_vector(vsi, v_idx) {
 			/* In some cases if DCB is configured the num_[rx|tx]q
 			 * can be less than vsi->num_q_vectors. This check
diff --git a/drivers/net/ethernet/intel/ice/ice_txrx.h b/drivers/net/ethernet/intel/ice/ice_txrx.h
index 14a1bf445889..7ee00a128663 100644
--- a/drivers/net/ethernet/intel/ice/ice_txrx.h
+++ b/drivers/net/ethernet/intel/ice/ice_txrx.h
@@ -222,7 +222,7 @@ enum ice_rx_dtype {
 #define ICE_ITR_GRAN_S		1	/* ITR granularity is always 2us */
 #define ICE_ITR_GRAN_US		BIT(ICE_ITR_GRAN_S)
 #define ICE_ITR_MASK		0x1FFE	/* ITR register value alignment mask */
-#define ITR_REG_ALIGN(setting)	__ALIGN_MASK(setting, ~ICE_ITR_MASK)
+#define ITR_REG_ALIGN(setting)	((setting) & ICE_ITR_MASK)
 
 #define ICE_ITR_ADAPTIVE_MIN_INC	0x0002
 #define ICE_ITR_ADAPTIVE_MIN_USECS	0x0002
diff --git a/drivers/net/ethernet/intel/ice/ice_virtchnl_pf.c b/drivers/net/ethernet/intel/ice/ice_virtchnl_pf.c
index 262714d5f54a..75c70d432c72 100644
--- a/drivers/net/ethernet/intel/ice/ice_virtchnl_pf.c
+++ b/drivers/net/ethernet/intel/ice/ice_virtchnl_pf.c
@@ -1874,6 +1874,48 @@ error_param:
 }
 
 /**
+ * ice_wait_on_vf_reset - poll to make sure a given VF is ready after reset
+ * @vf: The VF being resseting
+ *
+ * The max poll time is about ~800ms, which is about the maximum time it takes
+ * for a VF to be reset and/or a VF driver to be removed.
+ */
+static void ice_wait_on_vf_reset(struct ice_vf *vf)
+{
+	int i;
+
+	for (i = 0; i < ICE_MAX_VF_RESET_TRIES; i++) {
+		if (test_bit(ICE_VF_STATE_INIT, vf->vf_states))
+			break;
+		msleep(ICE_MAX_VF_RESET_SLEEP_MS);
+	}
+}
+
+/**
+ * ice_check_vf_ready_for_cfg - check if VF is ready to be configured/queried
+ * @vf: VF to check if it's ready to be configured/queried
+ *
+ * The purpose of this function is to make sure the VF is not in reset, not
+ * disabled, and initialized so it can be configured and/or queried by a host
+ * administrator.
+ */
+static int ice_check_vf_ready_for_cfg(struct ice_vf *vf)
+{
+	struct ice_pf *pf;
+
+	ice_wait_on_vf_reset(vf);
+
+	if (ice_is_vf_disabled(vf))
+		return -EINVAL;
+
+	pf = vf->pf;
+	if (ice_check_vf_init(pf, vf))
+		return -EBUSY;
+
+	return 0;
+}
+
+/**
  * ice_set_vf_spoofchk
  * @netdev: network interface device structure
  * @vf_id: VF identifier
@@ -1890,16 +1932,16 @@ int ice_set_vf_spoofchk(struct net_device *netdev, int vf_id, bool ena)
 	enum ice_status status;
 	struct device *dev;
 	struct ice_vf *vf;
-	int ret = 0;
+	int ret;
 
 	dev = ice_pf_to_dev(pf);
 	if (ice_validate_vf_id(pf, vf_id))
 		return -EINVAL;
 
 	vf = &pf->vf[vf_id];
-
-	if (ice_check_vf_init(pf, vf))
-		return -EBUSY;
+	ret = ice_check_vf_ready_for_cfg(vf);
+	if (ret)
+		return ret;
 
 	vf_vsi = pf->vsi[vf->lan_vsi_idx];
 	if (!vf_vsi) {
@@ -2696,7 +2738,7 @@ ice_set_vf_port_vlan(struct net_device *netdev, int vf_id, u16 vlan_id, u8 qos,
 	struct ice_vsi *vsi;
 	struct device *dev;
 	struct ice_vf *vf;
-	int ret = 0;
+	int ret;
 
 	dev = ice_pf_to_dev(pf);
 	if (ice_validate_vf_id(pf, vf_id))
@@ -2714,13 +2756,15 @@ ice_set_vf_port_vlan(struct net_device *netdev, int vf_id, u16 vlan_id, u8 qos,
 
 	vf = &pf->vf[vf_id];
 	vsi = pf->vsi[vf->lan_vsi_idx];
-	if (ice_check_vf_init(pf, vf))
-		return -EBUSY;
+
+	ret = ice_check_vf_ready_for_cfg(vf);
+	if (ret)
+		return ret;
 
 	if (le16_to_cpu(vsi->info.pvid) == vlanprio) {
 		/* duplicate request, so just return success */
 		dev_dbg(dev, "Duplicate pvid %d request\n", vlanprio);
-		return ret;
+		return 0;
 	}
 
 	/* If PVID, then remove all filters on the old VLAN */
@@ -2731,7 +2775,7 @@ ice_set_vf_port_vlan(struct net_device *netdev, int vf_id, u16 vlan_id, u8 qos,
 	if (vlan_id || qos) {
 		ret = ice_vsi_manage_pvid(vsi, vlanprio, true);
 		if (ret)
-			goto error_set_pvid;
+			return ret;
 	} else {
 		ice_vsi_manage_pvid(vsi, 0, false);
 		vsi->info.pvid = 0;
@@ -2744,7 +2788,7 @@ ice_set_vf_port_vlan(struct net_device *netdev, int vf_id, u16 vlan_id, u8 qos,
 		/* add new VLAN filter for each MAC */
 		ret = ice_vsi_add_vlan(vsi, vlan_id);
 		if (ret)
-			goto error_set_pvid;
+			return ret;
 	}
 
 	/* The Port VLAN needs to be saved across resets the same as the
@@ -2752,8 +2796,7 @@ ice_set_vf_port_vlan(struct net_device *netdev, int vf_id, u16 vlan_id, u8 qos,
 	 */
 	vf->port_vlan_id = le16_to_cpu(vsi->info.pvid);
 
-error_set_pvid:
-	return ret;
+	return 0;
 }
 
 /**
@@ -3237,23 +3280,6 @@ ice_get_vf_cfg(struct net_device *netdev, int vf_id, struct ifla_vf_info *ivi)
 }
 
 /**
- * ice_wait_on_vf_reset
- * @vf: The VF being resseting
- *
- * Poll to make sure a given VF is ready after reset
- */
-static void ice_wait_on_vf_reset(struct ice_vf *vf)
-{
-	int i;
-
-	for (i = 0; i < ICE_MAX_VF_RESET_WAIT; i++) {
-		if (test_bit(ICE_VF_STATE_INIT, vf->vf_states))
-			break;
-		msleep(20);
-	}
-}
-
-/**
  * ice_set_vf_mac
  * @netdev: network interface device structure
  * @vf_id: VF identifier
@@ -3265,29 +3291,21 @@ int ice_set_vf_mac(struct net_device *netdev, int vf_id, u8 *mac)
 {
 	struct ice_pf *pf = ice_netdev_to_pf(netdev);
 	struct ice_vf *vf;
-	int ret = 0;
+	int ret;
 
 	if (ice_validate_vf_id(pf, vf_id))
 		return -EINVAL;
 
-	vf = &pf->vf[vf_id];
-	/* Don't set MAC on disabled VF */
-	if (ice_is_vf_disabled(vf))
-		return -EINVAL;
-
-	/* In case VF is in reset mode, wait until it is completed. Depending
-	 * on factors like queue disabling routine, this could take ~250ms
-	 */
-	ice_wait_on_vf_reset(vf);
-
-	if (ice_check_vf_init(pf, vf))
-		return -EBUSY;
-
 	if (is_zero_ether_addr(mac) || is_multicast_ether_addr(mac)) {
 		netdev_err(netdev, "%pM not a valid unicast address\n", mac);
 		return -EINVAL;
 	}
 
+	vf = &pf->vf[vf_id];
+	ret = ice_check_vf_ready_for_cfg(vf);
+	if (ret)
+		return ret;
+
 	/* copy MAC into dflt_lan_addr and trigger a VF reset. The reset
 	 * flow will use the updated dflt_lan_addr and add a MAC filter
 	 * using ice_add_mac. Also set pf_set_mac to indicate that the PF has
@@ -3299,7 +3317,7 @@ int ice_set_vf_mac(struct net_device *netdev, int vf_id, u8 *mac)
 		    vf_id, mac);
 
 	ice_vc_reset_vf(vf);
-	return ret;
+	return 0;
 }
 
 /**
@@ -3314,22 +3332,15 @@ int ice_set_vf_trust(struct net_device *netdev, int vf_id, bool trusted)
 {
 	struct ice_pf *pf = ice_netdev_to_pf(netdev);
 	struct ice_vf *vf;
+	int ret;
 
 	if (ice_validate_vf_id(pf, vf_id))
 		return -EINVAL;
 
 	vf = &pf->vf[vf_id];
-	/* Don't set Trusted Mode on disabled VF */
-	if (ice_is_vf_disabled(vf))
-		return -EINVAL;
-
-	/* In case VF is in reset mode, wait until it is completed. Depending
-	 * on factors like queue disabling routine, this could take ~250ms
-	 */
-	ice_wait_on_vf_reset(vf);
-
-	if (ice_check_vf_init(pf, vf))
-		return -EBUSY;
+	ret = ice_check_vf_ready_for_cfg(vf);
+	if (ret)
+		return ret;
 
 	/* Check if already trusted */
 	if (trusted == vf->trusted)
@@ -3355,13 +3366,15 @@ int ice_set_vf_link_state(struct net_device *netdev, int vf_id, int link_state)
 {
 	struct ice_pf *pf = ice_netdev_to_pf(netdev);
 	struct ice_vf *vf;
+	int ret;
 
 	if (ice_validate_vf_id(pf, vf_id))
 		return -EINVAL;
 
 	vf = &pf->vf[vf_id];
-	if (ice_check_vf_init(pf, vf))
-		return -EBUSY;
+	ret = ice_check_vf_ready_for_cfg(vf);
+	if (ret)
+		return ret;
 
 	switch (link_state) {
 	case IFLA_VF_LINK_STATE_AUTO:
@@ -3397,14 +3410,15 @@ int ice_get_vf_stats(struct net_device *netdev, int vf_id,
 	struct ice_eth_stats *stats;
 	struct ice_vsi *vsi;
 	struct ice_vf *vf;
+	int ret;
 
 	if (ice_validate_vf_id(pf, vf_id))
 		return -EINVAL;
 
 	vf = &pf->vf[vf_id];
-
-	if (ice_check_vf_init(pf, vf))
-		return -EBUSY;
+	ret = ice_check_vf_ready_for_cfg(vf);
+	if (ret)
+		return ret;
 
 	vsi = pf->vsi[vf->lan_vsi_idx];
 	if (!vsi)
diff --git a/drivers/net/ethernet/intel/ice/ice_virtchnl_pf.h b/drivers/net/ethernet/intel/ice/ice_virtchnl_pf.h
index 4647d636ed36..ac67982751df 100644
--- a/drivers/net/ethernet/intel/ice/ice_virtchnl_pf.h
+++ b/drivers/net/ethernet/intel/ice/ice_virtchnl_pf.h
@@ -38,7 +38,8 @@
 #define ICE_MAX_POLICY_INTR_PER_VF	33
 #define ICE_MIN_INTR_PER_VF		(ICE_MIN_QS_PER_VF + 1)
 #define ICE_DFLT_INTR_PER_VF		(ICE_DFLT_QS_PER_VF + 1)
-#define ICE_MAX_VF_RESET_WAIT		15
+#define ICE_MAX_VF_RESET_TRIES		40
+#define ICE_MAX_VF_RESET_SLEEP_MS	20
 
 #define ice_for_each_vf(pf, i) \
 	for ((i) = 0; (i) < (pf)->num_alloc_vfs; (i)++)