Merge branch 'nfp-flower-updates-and-netconsole'

Jakub Kicinski says:

====================
nfp: flower updates and netconsole

This set contains assorted updates to driver base and flower.
First patch is a follow up to a fix to calculating counters which
went into net.  For ethtool counters we should also make sure
they are visible even after ring reconfiguration.  Next patch
is a safety measure in case we are running on a machine with a
broken BIOS we should fail the probe when risk of incorrect
operation is too high.  The next two patches add netpoll support
and make use of napi_consume_skb().  Last we populate bus info
on all representors.

Pieter adds support for offload of the checksum action in flower.

John follows up to another fix he's done in net, we set TTL
values on tunnels to stack's default, now Johns does a full
route lookup to get a more precise information, he populates
ToS field as well.  Last but not least he follows up on Jiri's
request to enable LAG offload in case the team driver is used
and then hash function is unknown.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>

Author: davem330

drivers/net/ethernet/netronome/nfp/flower/action.c

@@ -34,6 +34,7 @@
 #include <linux/bitfield.h>
 #include <net/pkt_cls.h>
 #include <net/switchdev.h>
+#include <net/tc_act/tc_csum.h>
 #include <net/tc_act/tc_gact.h>
 #include <net/tc_act/tc_mirred.h>
 #include <net/tc_act/tc_pedit.h>
@@ -44,6 +45,8 @@
 #include "main.h"
 #include "../nfp_net_repr.h"
 
+#define NFP_FL_SUPPORTED_IPV4_UDP_TUN_FLAGS	(TUNNEL_CSUM | TUNNEL_KEY)
+
 static void nfp_fl_pop_vlan(struct nfp_fl_pop_vlan *pop_vlan)
 {
 	size_t act_size = sizeof(struct nfp_fl_pop_vlan);
@@ -235,9 +238,12 @@ nfp_fl_set_ipv4_udp_tun(struct nfp_fl_set_ipv4_udp_tun *set_tun,
 	size_t act_size = sizeof(struct nfp_fl_set_ipv4_udp_tun);
 	struct ip_tunnel_info *ip_tun = tcf_tunnel_info(action);
 	u32 tmp_set_ip_tun_type_index = 0;
+	struct flowi4 flow = {};
 	/* Currently support one pre-tunnel so index is always 0. */
 	int pretun_idx = 0;
+	struct rtable *rt;
 	struct net *net;
+	int err;
 
 	if (ip_tun->options_len)
 		return -EOPNOTSUPP;
@@ -254,7 +260,28 @@ nfp_fl_set_ipv4_udp_tun(struct nfp_fl_set_ipv4_udp_tun *set_tun,
 
 	set_tun->tun_type_index = cpu_to_be32(tmp_set_ip_tun_type_index);
 	set_tun->tun_id = ip_tun->key.tun_id;
-	set_tun->ttl = net->ipv4.sysctl_ip_default_ttl;
+
+	/* Do a route lookup to determine ttl - if fails then use default.
+	 * Note that CONFIG_INET is a requirement of CONFIG_NET_SWITCHDEV so
+	 * must be defined here.
+	 */
+	flow.daddr = ip_tun->key.u.ipv4.dst;
+	flow.flowi4_proto = IPPROTO_UDP;
+	rt = ip_route_output_key(net, &flow);
+	err = PTR_ERR_OR_ZERO(rt);
+	if (!err) {
+		set_tun->ttl = ip4_dst_hoplimit(&rt->dst);
+		ip_rt_put(rt);
+	} else {
+		set_tun->ttl = net->ipv4.sysctl_ip_default_ttl;
+	}
+
+	set_tun->tos = ip_tun->key.tos;
+
+	if (!(ip_tun->key.tun_flags & TUNNEL_KEY) ||
+	    ip_tun->key.tun_flags & ~NFP_FL_SUPPORTED_IPV4_UDP_TUN_FLAGS)
+		return -EOPNOTSUPP;
+	set_tun->tun_flags = ip_tun->key.tun_flags;
 
 	/* Complete pre_tunnel action. */
 	pre_tun->ipv4_dst = ip_tun->key.u.ipv4.dst;
@@ -398,8 +425,27 @@ nfp_fl_set_tport(const struct tc_action *action, int idx, u32 off,
 	return 0;
 }
 
+static u32 nfp_fl_csum_l4_to_flag(u8 ip_proto)
+{
+	switch (ip_proto) {
+	case 0:
+		/* Filter doesn't force proto match,
+		 * both TCP and UDP will be updated if encountered
+		 */
+		return TCA_CSUM_UPDATE_FLAG_TCP | TCA_CSUM_UPDATE_FLAG_UDP;
+	case IPPROTO_TCP:
+		return TCA_CSUM_UPDATE_FLAG_TCP;
+	case IPPROTO_UDP:
+		return TCA_CSUM_UPDATE_FLAG_UDP;
+	default:
+		/* All other protocols will be ignored by FW */
+		return 0;
+	}
+}
+
 static int
-nfp_fl_pedit(const struct tc_action *action, char *nfp_action, int *a_len)
+nfp_fl_pedit(const struct tc_action *action, struct tc_cls_flower_offload *flow,
+	     char *nfp_action, int *a_len, u32 *csum_updated)
 {
 	struct nfp_fl_set_ipv6_addr set_ip6_dst, set_ip6_src;
 	struct nfp_fl_set_ip4_addrs set_ip_addr;
@@ -409,6 +455,7 @@ nfp_fl_pedit(const struct tc_action *action, char *nfp_action, int *a_len)
 	int idx, nkeys, err;
 	size_t act_size;
 	u32 offset, cmd;
+	u8 ip_proto = 0;
 
 	memset(&set_ip6_dst, 0, sizeof(set_ip6_dst));
 	memset(&set_ip6_src, 0, sizeof(set_ip6_src));
@@ -451,6 +498,15 @@ nfp_fl_pedit(const struct tc_action *action, char *nfp_action, int *a_len)
 			return err;
 	}
 
+	if (dissector_uses_key(flow->dissector, FLOW_DISSECTOR_KEY_BASIC)) {
+		struct flow_dissector_key_basic *basic;
+
+		basic = skb_flow_dissector_target(flow->dissector,
+						  FLOW_DISSECTOR_KEY_BASIC,
+						  flow->key);
+		ip_proto = basic->ip_proto;
+	}
+
 	if (set_eth.head.len_lw) {
 		act_size = sizeof(set_eth);
 		memcpy(nfp_action, &set_eth, act_size);
@@ -459,6 +515,10 @@ nfp_fl_pedit(const struct tc_action *action, char *nfp_action, int *a_len)
 		act_size = sizeof(set_ip_addr);
 		memcpy(nfp_action, &set_ip_addr, act_size);
 		*a_len += act_size;
+
+		/* Hardware will automatically fix IPv4 and TCP/UDP checksum. */
+		*csum_updated |= TCA_CSUM_UPDATE_FLAG_IPV4HDR |
+				nfp_fl_csum_l4_to_flag(ip_proto);
 	} else if (set_ip6_dst.head.len_lw && set_ip6_src.head.len_lw) {
 		/* TC compiles set src and dst IPv6 address as a single action,
 		 * the hardware requires this to be 2 separate actions.
@@ -471,18 +531,30 @@ nfp_fl_pedit(const struct tc_action *action, char *nfp_action, int *a_len)
 		memcpy(&nfp_action[sizeof(set_ip6_src)], &set_ip6_dst,
 		       act_size);
 		*a_len += act_size;
+
+		/* Hardware will automatically fix TCP/UDP checksum. */
+		*csum_updated |= nfp_fl_csum_l4_to_flag(ip_proto);
 	} else if (set_ip6_dst.head.len_lw) {
 		act_size = sizeof(set_ip6_dst);
 		memcpy(nfp_action, &set_ip6_dst, act_size);
 		*a_len += act_size;
+
+		/* Hardware will automatically fix TCP/UDP checksum. */
+		*csum_updated |= nfp_fl_csum_l4_to_flag(ip_proto);
 	} else if (set_ip6_src.head.len_lw) {
 		act_size = sizeof(set_ip6_src);
 		memcpy(nfp_action, &set_ip6_src, act_size);
 		*a_len += act_size;
+
+		/* Hardware will automatically fix TCP/UDP checksum. */
+		*csum_updated |= nfp_fl_csum_l4_to_flag(ip_proto);
 	} else if (set_tport.head.len_lw) {
 		act_size = sizeof(set_tport);
 		memcpy(nfp_action, &set_tport, act_size);
 		*a_len += act_size;
+
+		/* Hardware will automatically fix TCP/UDP checksum. */
+		*csum_updated |= nfp_fl_csum_l4_to_flag(ip_proto);
 	}
 
 	return 0;
@@ -493,12 +565,18 @@ nfp_flower_output_action(struct nfp_app *app, const struct tc_action *a,
 			 struct nfp_fl_payload *nfp_fl, int *a_len,
 			 struct net_device *netdev, bool last,
 			 enum nfp_flower_tun_type *tun_type, int *tun_out_cnt,
-			 int *out_cnt)
+			 int *out_cnt, u32 *csum_updated)
 {
 	struct nfp_flower_priv *priv = app->priv;
 	struct nfp_fl_output *output;
 	int err, prelag_size;
 
+	/* If csum_updated has not been reset by now, it means HW will
+	 * incorrectly update csums when they are not requested.
+	 */
+	if (*csum_updated)
+		return -EOPNOTSUPP;
+
 	if (*a_len + sizeof(struct nfp_fl_output) > NFP_FL_MAX_A_SIZ)
 		return -EOPNOTSUPP;
 
@@ -529,10 +607,11 @@ nfp_flower_output_action(struct nfp_app *app, const struct tc_action *a,
 
 static int
 nfp_flower_loop_action(struct nfp_app *app, const struct tc_action *a,
+		       struct tc_cls_flower_offload *flow,
 		       struct nfp_fl_payload *nfp_fl, int *a_len,
 		       struct net_device *netdev,
 		       enum nfp_flower_tun_type *tun_type, int *tun_out_cnt,
-		       int *out_cnt)
+		       int *out_cnt, u32 *csum_updated)
 {
 	struct nfp_fl_set_ipv4_udp_tun *set_tun;
 	struct nfp_fl_pre_tunnel *pre_tun;
@@ -545,14 +624,14 @@ nfp_flower_loop_action(struct nfp_app *app, const struct tc_action *a,
 	} else if (is_tcf_mirred_egress_redirect(a)) {
 		err = nfp_flower_output_action(app, a, nfp_fl, a_len, netdev,
 					       true, tun_type, tun_out_cnt,
-					       out_cnt);
+					       out_cnt, csum_updated);
 		if (err)
 			return err;
 
 	} else if (is_tcf_mirred_egress_mirror(a)) {
 		err = nfp_flower_output_action(app, a, nfp_fl, a_len, netdev,
 					       false, tun_type, tun_out_cnt,
-					       out_cnt);
+					       out_cnt, csum_updated);
 		if (err)
 			return err;
 
@@ -602,8 +681,17 @@ nfp_flower_loop_action(struct nfp_app *app, const struct tc_action *a,
 		/* Tunnel decap is handled by default so accept action. */
 		return 0;
 	} else if (is_tcf_pedit(a)) {
-		if (nfp_fl_pedit(a, &nfp_fl->action_data[*a_len], a_len))
+		if (nfp_fl_pedit(a, flow, &nfp_fl->action_data[*a_len],
+				 a_len, csum_updated))
+			return -EOPNOTSUPP;
+	} else if (is_tcf_csum(a)) {
+		/* csum action requests recalc of something we have not fixed */
+		if (tcf_csum_update_flags(a) & ~*csum_updated)
 			return -EOPNOTSUPP;
+		/* If we will correctly fix the csum we can remove it from the
+		 * csum update list. Which will later be used to check support.
+		 */
+		*csum_updated &= ~tcf_csum_update_flags(a);
 	} else {
 		/* Currently we do not handle any other actions. */
 		return -EOPNOTSUPP;
@@ -620,6 +708,7 @@ int nfp_flower_compile_action(struct nfp_app *app,
 	int act_len, act_cnt, err, tun_out_cnt, out_cnt;
 	enum nfp_flower_tun_type tun_type;
 	const struct tc_action *a;
+	u32 csum_updated = 0;
 	LIST_HEAD(actions);
 
 	memset(nfp_flow->action_data, 0, NFP_FL_MAX_A_SIZ);
@@ -632,8 +721,9 @@ int nfp_flower_compile_action(struct nfp_app *app,
 
 	tcf_exts_to_list(flow->exts, &actions);
 	list_for_each_entry(a, &actions, list) {
-		err = nfp_flower_loop_action(app, a, nfp_flow, &act_len, netdev,
-					     &tun_type, &tun_out_cnt, &out_cnt);
+		err = nfp_flower_loop_action(app, a, flow, nfp_flow, &act_len,
+					     netdev, &tun_type, &tun_out_cnt,
+					     &out_cnt, &csum_updated);
 		if (err)
 			return err;
 		act_cnt++;

drivers/net/ethernet/netronome/nfp/flower/cmsg.h

@@ -203,9 +203,9 @@ struct nfp_fl_set_ipv4_udp_tun {
 	__be16 reserved;
 	__be64 tun_id __packed;
 	__be32 tun_type_index;
-	__be16 reserved2;
+	__be16 tun_flags;
 	u8 ttl;
-	u8 reserved3;
+	u8 tos;
 	__be32 extra[2];
 };
 

drivers/net/ethernet/netronome/nfp/flower/lag_conf.c

@@ -564,8 +564,9 @@ nfp_fl_lag_changeupper_event(struct nfp_fl_lag *lag,
 	if (lag_upper_info &&
 	    lag_upper_info->tx_type != NETDEV_LAG_TX_TYPE_ACTIVEBACKUP &&
 	    (lag_upper_info->tx_type != NETDEV_LAG_TX_TYPE_HASH ||
-	    (lag_upper_info->hash_type != NETDEV_LAG_HASH_L34 &&
-	    lag_upper_info->hash_type != NETDEV_LAG_HASH_E34))) {
+	     (lag_upper_info->hash_type != NETDEV_LAG_HASH_L34 &&
+	      lag_upper_info->hash_type != NETDEV_LAG_HASH_E34 &&
+	      lag_upper_info->hash_type != NETDEV_LAG_HASH_UNKNOWN))) {
 		can_offload = false;
 		nfp_flower_cmsg_warn(priv->app,
 				     "Unable to offload tx_type %u hash %u\n",

drivers/net/ethernet/netronome/nfp/nfp_net_common.c

@@ -945,11 +945,12 @@ static int nfp_net_tx(struct sk_buff *skb, struct net_device *netdev)
 
 /**
  * nfp_net_tx_complete() - Handled completed TX packets
- * @tx_ring:   TX ring structure
+ * @tx_ring:	TX ring structure
+ * @budget:	NAPI budget (only used as bool to determine if in NAPI context)
  *
  * Return: Number of completed TX descriptors
  */
-static void nfp_net_tx_complete(struct nfp_net_tx_ring *tx_ring)
+static void nfp_net_tx_complete(struct nfp_net_tx_ring *tx_ring, int budget)
 {
 	struct nfp_net_r_vector *r_vec = tx_ring->r_vec;
 	struct nfp_net_dp *dp = &r_vec->nfp_net->dp;
@@ -999,7 +1000,7 @@ static void nfp_net_tx_complete(struct nfp_net_tx_ring *tx_ring)
 
 		/* check for last gather fragment */
 		if (fidx == nr_frags - 1)
-			dev_consume_skb_any(skb);
+			napi_consume_skb(skb, budget);
 
 		tx_ring->txbufs[idx].dma_addr = 0;
 		tx_ring->txbufs[idx].skb = NULL;
@@ -1828,7 +1829,7 @@ static int nfp_net_poll(struct napi_struct *napi, int budget)
 	unsigned int pkts_polled = 0;
 
 	if (r_vec->tx_ring)
-		nfp_net_tx_complete(r_vec->tx_ring);
+		nfp_net_tx_complete(r_vec->tx_ring, budget);
 	if (r_vec->rx_ring)
 		pkts_polled = nfp_net_rx(r_vec->rx_ring, budget);
 
@@ -2062,7 +2063,7 @@ static void nfp_ctrl_poll(unsigned long arg)
 	struct nfp_net_r_vector *r_vec = (void *)arg;
 
 	spin_lock_bh(&r_vec->lock);
-	nfp_net_tx_complete(r_vec->tx_ring);
+	nfp_net_tx_complete(r_vec->tx_ring, 0);
 	__nfp_ctrl_tx_queued(r_vec);
 	spin_unlock_bh(&r_vec->lock);
 
@@ -3115,6 +3116,21 @@ nfp_net_vlan_rx_kill_vid(struct net_device *netdev, __be16 proto, u16 vid)
 	return nfp_net_reconfig_mbox(nn, NFP_NET_CFG_MBOX_CMD_CTAG_FILTER_KILL);
 }
 
+#ifdef CONFIG_NET_POLL_CONTROLLER
+static void nfp_net_netpoll(struct net_device *netdev)
+{
+	struct nfp_net *nn = netdev_priv(netdev);
+	int i;
+
+	/* nfp_net's NAPIs are statically allocated so even if there is a race
+	 * with reconfig path this will simply try to schedule some disabled
+	 * NAPI instances.
+	 */
+	for (i = 0; i < nn->dp.num_stack_tx_rings; i++)
+		napi_schedule_irqoff(&nn->r_vecs[i].napi);
+}
+#endif
+
 static void nfp_net_stat64(struct net_device *netdev,
 			   struct rtnl_link_stats64 *stats)
 {
@@ -3482,6 +3498,9 @@ const struct net_device_ops nfp_net_netdev_ops = {
 	.ndo_get_stats64	= nfp_net_stat64,
 	.ndo_vlan_rx_add_vid	= nfp_net_vlan_rx_add_vid,
 	.ndo_vlan_rx_kill_vid	= nfp_net_vlan_rx_kill_vid,
+#ifdef CONFIG_NET_POLL_CONTROLLER
+	.ndo_poll_controller	= nfp_net_netpoll,
+#endif
 	.ndo_set_vf_mac         = nfp_app_set_vf_mac,
 	.ndo_set_vf_vlan        = nfp_app_set_vf_vlan,
 	.ndo_set_vf_spoofchk    = nfp_app_set_vf_spoofchk,