Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-next

Pull networking changes from David S Miller:

 1) Remove the ipv4 routing cache.  Now lookups go directly into the FIB
    trie and use prebuilt routes cached there.

    No more garbage collection, no more rDOS attacks on the routing
    cache.  Instead we now get predictable and consistent performance,
    no matter what the pattern of traffic we service.

    This has been almost 2 years in the making.  Special thanks to
    Julian Anastasov, Eric Dumazet, Steffen Klassert, and others who
    have helped along the way.

    I'm sure that with a change of this magnitude there will be some
    kind of fallout, but such things ought the be simple to fix at this
    point.  Luckily I'm not European so I'll be around all of August to
    fix things :-)

    The major stages of this work here are each fronted by a forced
    merge commit whose commit message contains a top-level description
    of the motivations and implementation issues.

 2) Pre-demux of established ipv4 TCP sockets, saves a route demux on
    input.

 3) TCP SYN/ACK performance tweaks from Eric Dumazet.

 4) Add namespace support for netfilter L4 conntrack helpers, from Gao
    Feng.

 5) Add config mechanism for Energy Efficient Ethernet to ethtool, from
    Yuval Mintz.

 6) Remove quadratic behavior from /proc/net/unix, from Eric Dumazet.

 7) Support for connection tracker helpers in userspace, from Pablo
    Neira Ayuso.

 8) Allow userspace driven TX load balancing functions in TEAM driver,
    from Jiri Pirko.

 9) Kill off NLMSG_PUT and RTA_PUT macros, more gross stuff with
    embedded gotos.

10) TCP Small Queues, essentially minimize the amount of TCP data queued
    up in the packet scheduler layer.  Whereas the existing BQL (Byte
    Queue Limits) limits the pkt_sched --> netdevice queuing levels,
    this controls the TCP --> pkt_sched queueing levels.

    From Eric Dumazet.

11) Reduce the number of get_page/put_page ops done on SKB fragments,
    from Alexander Duyck.

12) Implement protection against blind resets in TCP (RFC 5961), from
    Eric Dumazet.

13) Support the client side of TCP Fast Open, basically the ability to
    send data in the SYN exchange, from Yuchung Cheng.

    Basically, the sender queues up data with a sendmsg() call using
    MSG_FASTOPEN, then they do the connect() which emits the queued up
    fastopen data.

14) Avoid all the problems we get into in TCP when timers or PMTU events
    hit a locked socket.  The TCP Small Queues changes added a
    tcp_release_cb() that allows us to queue work up to the
    release_sock() caller, and that's what we use here too.  From Eric
    Dumazet.

15) Zero copy on TX support for TUN driver, from Michael S. Tsirkin.

* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-next: (1870 commits)
  genetlink: define lockdep_genl_is_held() when CONFIG_LOCKDEP
  r8169: revert "add byte queue limit support".
  ipv4: Change rt->rt_iif encoding.
  net: Make skb->skb_iif always track skb->dev
  ipv4: Prepare for change of rt->rt_iif encoding.
  ipv4: Remove all RTCF_DIRECTSRC handliing.
  ipv4: Really ignore ICMP address requests/replies.
  decnet: Don't set RTCF_DIRECTSRC.
  net/ipv4/ip_vti.c: Fix __rcu warnings detected by sparse.
  ipv4: Remove redundant assignment
  rds: set correct msg_namelen
  openvswitch: potential NULL deref in sample()
  tcp: dont drop MTU reduction indications
  bnx2x: Add new 57840 device IDs
  tcp: avoid oops in tcp_metrics and reset tcpm_stamp
  niu: Change niu_rbr_fill() to use unlikely() to check niu_rbr_add_page() return value
  niu: Fix to check for dma mapping errors.
  net: Fix references to out-of-scope variables in put_cmsg_compat()
  net: ethernet: davinci_emac: add pm_runtime support
  net: ethernet: davinci_emac: Remove unnecessary #include
  ...

1 files changed, 115 insertions, 58 deletions

diff --git a/net/mac80211/mesh_hwmp.c b/net/mac80211/mesh_hwmp.c
index 9b59658e8650..494bc39f61a4 100644
--- a/net/mac80211/mesh_hwmp.c
+++ b/net/mac80211/mesh_hwmp.c
@@ -13,13 +13,6 @@
 #include "wme.h"
 #include "mesh.h"
 
-#ifdef CONFIG_MAC80211_VERBOSE_MHWMP_DEBUG
-#define mhwmp_dbg(fmt, args...) \
-	printk(KERN_DEBUG "Mesh HWMP (%s): " fmt "\n", sdata->name, ##args)
-#else
-#define mhwmp_dbg(fmt, args...)   do { (void)(0); } while (0)
-#endif
-
 #define TEST_FRAME_LEN	8192
 #define MAX_METRIC	0xffffffff
 #define ARITH_SHIFT	8
@@ -98,6 +91,8 @@ static inline u32 u16_field_get(u8 *preq_elem, int offset, bool ae)
 #define max_preq_retries(s) (s->u.mesh.mshcfg.dot11MeshHWMPmaxPREQretries)
 #define disc_timeout_jiff(s) \
 	msecs_to_jiffies(sdata->u.mesh.mshcfg.min_discovery_timeout)
+#define root_path_confirmation_jiffies(s) \
+	msecs_to_jiffies(sdata->u.mesh.mshcfg.dot11MeshHWMPconfirmationInterval)
 
 enum mpath_frame_type {
 	MPATH_PREQ = 0,
@@ -142,19 +137,19 @@ static int mesh_path_sel_frame_tx(enum mpath_frame_type action, u8 flags,
 
 	switch (action) {
 	case MPATH_PREQ:
-		mhwmp_dbg("sending PREQ to %pM", target);
+		mhwmp_dbg(sdata, "sending PREQ to %pM\n", target);
 		ie_len = 37;
 		pos = skb_put(skb, 2 + ie_len);
 		*pos++ = WLAN_EID_PREQ;
 		break;
 	case MPATH_PREP:
-		mhwmp_dbg("sending PREP to %pM", target);
+		mhwmp_dbg(sdata, "sending PREP to %pM\n", target);
 		ie_len = 31;
 		pos = skb_put(skb, 2 + ie_len);
 		*pos++ = WLAN_EID_PREP;
 		break;
 	case MPATH_RANN:
-		mhwmp_dbg("sending RANN from %pM", orig_addr);
+		mhwmp_dbg(sdata, "sending RANN from %pM\n", orig_addr);
 		ie_len = sizeof(struct ieee80211_rann_ie);
 		pos = skb_put(skb, 2 + ie_len);
 		*pos++ = WLAN_EID_RANN;
@@ -303,7 +298,7 @@ int mesh_path_error_tx(u8 ttl, u8 *target, __le32 target_sn,
 }
 
 void ieee80211s_update_metric(struct ieee80211_local *local,
-		struct sta_info *stainfo, struct sk_buff *skb)
+		struct sta_info *sta, struct sk_buff *skb)
 {
 	struct ieee80211_tx_info *txinfo = IEEE80211_SKB_CB(skb);
 	struct ieee80211_hdr *hdr = (struct ieee80211_hdr *) skb->data;
@@ -315,15 +310,14 @@ void ieee80211s_update_metric(struct ieee80211_local *local,
 	failed = !(txinfo->flags & IEEE80211_TX_STAT_ACK);
 
 	/* moving average, scaled to 100 */
-	stainfo->fail_avg = ((80 * stainfo->fail_avg + 5) / 100 + 20 * failed);
-	if (stainfo->fail_avg > 95)
-		mesh_plink_broken(stainfo);
+	sta->fail_avg = ((80 * sta->fail_avg + 5) / 100 + 20 * failed);
+	if (sta->fail_avg > 95)
+		mesh_plink_broken(sta);
 }
 
 static u32 airtime_link_metric_get(struct ieee80211_local *local,
 				   struct sta_info *sta)
 {
-	struct ieee80211_supported_band *sband;
 	struct rate_info rinfo;
 	/* This should be adjusted for each device */
 	int device_constant = 1 << ARITH_SHIFT;
@@ -333,8 +327,6 @@ static u32 airtime_link_metric_get(struct ieee80211_local *local,
 	u32 tx_time, estimated_retx;
 	u64 result;
 
-	sband = local->hw.wiphy->bands[local->hw.conf.channel->band];
-
 	if (sta->fail_avg >= 100)
 		return MAX_METRIC;
 
@@ -519,10 +511,11 @@ static void hwmp_preq_frame_process(struct ieee80211_sub_if_data *sdata,
 	struct mesh_path *mpath = NULL;
 	u8 *target_addr, *orig_addr;
 	const u8 *da;
-	u8 target_flags, ttl;
-	u32 orig_sn, target_sn, lifetime;
+	u8 target_flags, ttl, flags;
+	u32 orig_sn, target_sn, lifetime, orig_metric;
 	bool reply = false;
 	bool forward = true;
+	bool root_is_gate;
 
 	/* Update target SN, if present */
 	target_addr = PREQ_IE_TARGET_ADDR(preq_elem);
@@ -530,11 +523,15 @@ static void hwmp_preq_frame_process(struct ieee80211_sub_if_data *sdata,
 	target_sn = PREQ_IE_TARGET_SN(preq_elem);
 	orig_sn = PREQ_IE_ORIG_SN(preq_elem);
 	target_flags = PREQ_IE_TARGET_F(preq_elem);
+	orig_metric = metric;
+	/* Proactive PREQ gate announcements */
+	flags = PREQ_IE_FLAGS(preq_elem);
+	root_is_gate = !!(flags & RANN_FLAG_IS_GATE);
 
-	mhwmp_dbg("received PREQ from %pM", orig_addr);
+	mhwmp_dbg(sdata, "received PREQ from %pM\n", orig_addr);
 
 	if (ether_addr_equal(target_addr, sdata->vif.addr)) {
-		mhwmp_dbg("PREQ is for us");
+		mhwmp_dbg(sdata, "PREQ is for us\n");
 		forward = false;
 		reply = true;
 		metric = 0;
@@ -544,6 +541,22 @@ static void hwmp_preq_frame_process(struct ieee80211_sub_if_data *sdata,
 			target_sn = ++ifmsh->sn;
 			ifmsh->last_sn_update = jiffies;
 		}
+	} else if (is_broadcast_ether_addr(target_addr) &&
+		   (target_flags & IEEE80211_PREQ_TO_FLAG)) {
+		rcu_read_lock();
+		mpath = mesh_path_lookup(orig_addr, sdata);
+		if (mpath) {
+			if (flags & IEEE80211_PREQ_PROACTIVE_PREP_FLAG) {
+				reply = true;
+				target_addr = sdata->vif.addr;
+				target_sn = ++ifmsh->sn;
+				metric = 0;
+				ifmsh->last_sn_update = jiffies;
+			}
+			if (root_is_gate)
+				mesh_path_add_gate(mpath);
+		}
+		rcu_read_unlock();
 	} else {
 		rcu_read_lock();
 		mpath = mesh_path_lookup(target_addr, sdata);
@@ -570,19 +583,20 @@ static void hwmp_preq_frame_process(struct ieee80211_sub_if_data *sdata,
 		lifetime = PREQ_IE_LIFETIME(preq_elem);
 		ttl = ifmsh->mshcfg.element_ttl;
 		if (ttl != 0) {
-			mhwmp_dbg("replying to the PREQ");
+			mhwmp_dbg(sdata, "replying to the PREQ\n");
 			mesh_path_sel_frame_tx(MPATH_PREP, 0, orig_addr,
 				cpu_to_le32(orig_sn), 0, target_addr,
 				cpu_to_le32(target_sn), mgmt->sa, 0, ttl,
 				cpu_to_le32(lifetime), cpu_to_le32(metric),
 				0, sdata);
-		} else
+		} else {
 			ifmsh->mshstats.dropped_frames_ttl++;
+		}
 	}
 
 	if (forward && ifmsh->mshcfg.dot11MeshForwarding) {
 		u32 preq_id;
-		u8 hopcount, flags;
+		u8 hopcount;
 
 		ttl = PREQ_IE_TTL(preq_elem);
 		lifetime = PREQ_IE_LIFETIME(preq_elem);
@@ -590,13 +604,19 @@ static void hwmp_preq_frame_process(struct ieee80211_sub_if_data *sdata,
 			ifmsh->mshstats.dropped_frames_ttl++;
 			return;
 		}
-		mhwmp_dbg("forwarding the PREQ from %pM", orig_addr);
+		mhwmp_dbg(sdata, "forwarding the PREQ from %pM\n", orig_addr);
 		--ttl;
-		flags = PREQ_IE_FLAGS(preq_elem);
 		preq_id = PREQ_IE_PREQ_ID(preq_elem);
 		hopcount = PREQ_IE_HOPCOUNT(preq_elem) + 1;
 		da = (mpath && mpath->is_root) ?
 			mpath->rann_snd_addr : broadcast_addr;
+
+		if (flags & IEEE80211_PREQ_PROACTIVE_PREP_FLAG) {
+			target_addr = PREQ_IE_TARGET_ADDR(preq_elem);
+			target_sn = PREQ_IE_TARGET_SN(preq_elem);
+			metric = orig_metric;
+		}
+
 		mesh_path_sel_frame_tx(MPATH_PREQ, flags, orig_addr,
 				cpu_to_le32(orig_sn), target_flags, target_addr,
 				cpu_to_le32(target_sn), da,
@@ -631,7 +651,8 @@ static void hwmp_prep_frame_process(struct ieee80211_sub_if_data *sdata,
 	u8 next_hop[ETH_ALEN];
 	u32 target_sn, orig_sn, lifetime;
 
-	mhwmp_dbg("received PREP from %pM", PREP_IE_ORIG_ADDR(prep_elem));
+	mhwmp_dbg(sdata, "received PREP from %pM\n",
+		  PREP_IE_ORIG_ADDR(prep_elem));
 
 	orig_addr = PREP_IE_ORIG_ADDR(prep_elem);
 	if (ether_addr_equal(orig_addr, sdata->vif.addr))
@@ -744,11 +765,6 @@ static void hwmp_rann_frame_process(struct ieee80211_sub_if_data *sdata,
 	bool root_is_gate;
 
 	ttl = rann->rann_ttl;
-	if (ttl <= 1) {
-		ifmsh->mshstats.dropped_frames_ttl++;
-		return;
-	}
-	ttl--;
 	flags = rann->rann_flags;
 	root_is_gate = !!(flags & RANN_FLAG_IS_GATE);
 	orig_addr = rann->rann_addr;
@@ -762,8 +778,9 @@ static void hwmp_rann_frame_process(struct ieee80211_sub_if_data *sdata,
 	if (ether_addr_equal(orig_addr, sdata->vif.addr))
 		return;
 
-	mhwmp_dbg("received RANN from %pM via neighbour %pM (is_gate=%d)",
-			orig_addr, mgmt->sa, root_is_gate);
+	mhwmp_dbg(sdata,
+		  "received RANN from %pM via neighbour %pM (is_gate=%d)\n",
+		  orig_addr, mgmt->sa, root_is_gate);
 
 	rcu_read_lock();
 	sta = sta_info_get(sdata, mgmt->sa);
@@ -785,34 +802,50 @@ static void hwmp_rann_frame_process(struct ieee80211_sub_if_data *sdata,
 		}
 	}
 
+	if (!(SN_LT(mpath->sn, orig_sn)) &&
+	    !(mpath->sn == orig_sn && metric < mpath->rann_metric)) {
+		rcu_read_unlock();
+		return;
+	}
+
 	if ((!(mpath->flags & (MESH_PATH_ACTIVE | MESH_PATH_RESOLVING)) ||
-	     time_after(jiffies, mpath->exp_time - 1*HZ)) &&
-	     !(mpath->flags & MESH_PATH_FIXED)) {
-		mhwmp_dbg("%s time to refresh root mpath %pM", sdata->name,
-							       orig_addr);
+	     (time_after(jiffies, mpath->last_preq_to_root +
+				  root_path_confirmation_jiffies(sdata)) ||
+	     time_before(jiffies, mpath->last_preq_to_root))) &&
+	     !(mpath->flags & MESH_PATH_FIXED) && (ttl != 0)) {
+		mhwmp_dbg(sdata,
+			  "time to refresh root mpath %pM\n",
+			  orig_addr);
 		mesh_queue_preq(mpath, PREQ_Q_F_START | PREQ_Q_F_REFRESH);
+		mpath->last_preq_to_root = jiffies;
 	}
 
-	if ((SN_LT(mpath->sn, orig_sn) || (mpath->sn == orig_sn &&
-	   metric < mpath->rann_metric)) && ifmsh->mshcfg.dot11MeshForwarding) {
+	mpath->sn = orig_sn;
+	mpath->rann_metric = metric + metric_txsta;
+	mpath->is_root = true;
+	/* Recording RANNs sender address to send individually
+	 * addressed PREQs destined for root mesh STA */
+	memcpy(mpath->rann_snd_addr, mgmt->sa, ETH_ALEN);
+
+	if (root_is_gate)
+		mesh_path_add_gate(mpath);
+
+	if (ttl <= 1) {
+		ifmsh->mshstats.dropped_frames_ttl++;
+		rcu_read_unlock();
+		return;
+	}
+	ttl--;
+
+	if (ifmsh->mshcfg.dot11MeshForwarding) {
 		mesh_path_sel_frame_tx(MPATH_RANN, flags, orig_addr,
 				       cpu_to_le32(orig_sn),
 				       0, NULL, 0, broadcast_addr,
 				       hopcount, ttl, cpu_to_le32(interval),
 				       cpu_to_le32(metric + metric_txsta),
 				       0, sdata);
-		mpath->sn = orig_sn;
-		mpath->rann_metric = metric + metric_txsta;
-		/* Recording RANNs sender address to send individually
-		 * addressed PREQs destined for root mesh STA */
-		memcpy(mpath->rann_snd_addr, mgmt->sa, ETH_ALEN);
 	}
 
-	mpath->is_root = true;
-
-	if (root_is_gate)
-		mesh_path_add_gate(mpath);
-
 	rcu_read_unlock();
 }
 
@@ -889,7 +922,7 @@ static void mesh_queue_preq(struct mesh_path *mpath, u8 flags)
 
 	preq_node = kmalloc(sizeof(struct mesh_preq_queue), GFP_ATOMIC);
 	if (!preq_node) {
-		mhwmp_dbg("could not allocate PREQ node");
+		mhwmp_dbg(sdata, "could not allocate PREQ node\n");
 		return;
 	}
 
@@ -898,7 +931,7 @@ static void mesh_queue_preq(struct mesh_path *mpath, u8 flags)
 		spin_unlock_bh(&ifmsh->mesh_preq_queue_lock);
 		kfree(preq_node);
 		if (printk_ratelimit())
-			mhwmp_dbg("PREQ node queue full");
+			mhwmp_dbg(sdata, "PREQ node queue full\n");
 		return;
 	}
 
@@ -1021,12 +1054,15 @@ enddiscovery:
 	kfree(preq_node);
 }
 
-/* mesh_nexthop_resolve - lookup next hop for given skb and start path
- * discovery if no forwarding information is found.
+/**
+ * mesh_nexthop_resolve - lookup next hop; conditionally start path discovery
  *
  * @skb: 802.11 frame to be sent
  * @sdata: network subif the frame will be sent through
  *
+ * Lookup next hop for given skb and start path discovery if no
+ * forwarding information is found.
+ *
  * Returns: 0 if the next hop was found and -ENOENT if the frame was queued.
  * skb is freeed here if no mpath could be allocated.
  */
@@ -1146,7 +1182,7 @@ void mesh_path_timer(unsigned long data)
 		if (!mpath->is_gate && mesh_gate_num(sdata) > 0) {
 			ret = mesh_path_send_to_gates(mpath);
 			if (ret)
-				mhwmp_dbg("no gate was reachable");
+				mhwmp_dbg(sdata, "no gate was reachable\n");
 		} else
 			mesh_path_flush_pending(mpath);
 	}
@@ -1157,13 +1193,34 @@ mesh_path_tx_root_frame(struct ieee80211_sub_if_data *sdata)
 {
 	struct ieee80211_if_mesh *ifmsh = &sdata->u.mesh;
 	u32 interval = ifmsh->mshcfg.dot11MeshHWMPRannInterval;
-	u8 flags;
+	u8 flags, target_flags = 0;
 
 	flags = (ifmsh->mshcfg.dot11MeshGateAnnouncementProtocol)
 			? RANN_FLAG_IS_GATE : 0;
-	mesh_path_sel_frame_tx(MPATH_RANN, flags, sdata->vif.addr,
+
+	switch (ifmsh->mshcfg.dot11MeshHWMPRootMode) {
+	case IEEE80211_PROACTIVE_RANN:
+		mesh_path_sel_frame_tx(MPATH_RANN, flags, sdata->vif.addr,
 			       cpu_to_le32(++ifmsh->sn),
 			       0, NULL, 0, broadcast_addr,
-			       0, sdata->u.mesh.mshcfg.element_ttl,
+			       0, ifmsh->mshcfg.element_ttl,
 			       cpu_to_le32(interval), 0, 0, sdata);
+		break;
+	case IEEE80211_PROACTIVE_PREQ_WITH_PREP:
+		flags |= IEEE80211_PREQ_PROACTIVE_PREP_FLAG;
+	case IEEE80211_PROACTIVE_PREQ_NO_PREP:
+		interval = ifmsh->mshcfg.dot11MeshHWMPactivePathToRootTimeout;
+		target_flags |= IEEE80211_PREQ_TO_FLAG |
+				IEEE80211_PREQ_USN_FLAG;
+		mesh_path_sel_frame_tx(MPATH_PREQ, flags, sdata->vif.addr,
+				cpu_to_le32(++ifmsh->sn), target_flags,
+				(u8 *) broadcast_addr, 0, broadcast_addr,
+				0, ifmsh->mshcfg.element_ttl,
+				cpu_to_le32(interval),
+				0, cpu_to_le32(ifmsh->preq_id++), sdata);
+		break;
+	default:
+		mhwmp_dbg(sdata, "Proactive mechanism not supported\n");
+		return;
+	}
 }