wifi: mac80211: mesh fast xmit support

Previously, fast xmit only worked on interface types where initially a
sta lookup is performed, and a cached header can be attached to the sta,
requiring only some fields to be updated at runtime.

This technique is not directly applicable for a mesh device type due
to the dynamic nature of the topology and protocol. There are more
addresses that need to be filled, and there is an extra header with a
dynamic length based on the addressing mode.

Change the code to cache entries contain a copy of the mesh subframe header +
bridge tunnel header, as well as an embedded struct ieee80211_fast_tx, which
contains the information for building the 802.11 header.

Add a mesh specific early fast xmit call, which looks up a cached entry and
adds only the mesh subframe header, before passing it over to the generic
fast xmit code.

To ensure the changes in network are reflected in these cached headers,
flush affected cached entries on path changes, as well as other conditions
that currently trigger a fast xmit check in other modes (key changes etc.)

This code is loosely based on a previous implementation by:
Sriram R <quic_srirrama@quicinc.com>

Cc: Sriram R <quic_srirrama@quicinc.com>
Signed-off-by: Ryder Lee <ryder.lee@mediatek.com>
Signed-off-by: Felix Fietkau <nbd@nbd.name>
Link: https://lore.kernel.org/r/20230314095956.62085-4-nbd@nbd.name
Signed-off-by: Johannes Berg <johannes.berg@intel.com>

1 files changed, 282 insertions, 0 deletions

diff --git a/net/mac80211/mesh_pathtbl.c b/net/mac80211/mesh_pathtbl.c
index 3b81e6df3f34..d32e304eeb4b 100644
--- a/net/mac80211/mesh_pathtbl.c
+++ b/net/mac80211/mesh_pathtbl.c
@@ -14,6 +14,7 @@
 #include "wme.h"
 #include "ieee80211_i.h"
 #include "mesh.h"
+#include <linux/rhashtable.h>
 
 static void mesh_path_free_rcu(struct mesh_table *tbl, struct mesh_path *mpath);
 
@@ -32,6 +33,41 @@ static const struct rhashtable_params mesh_rht_params = {
 	.hashfn = mesh_table_hash,
 };
 
+static const struct rhashtable_params fast_tx_rht_params = {
+	.nelem_hint = 10,
+	.automatic_shrinking = true,
+	.key_len = ETH_ALEN,
+	.key_offset = offsetof(struct ieee80211_mesh_fast_tx, addr_key),
+	.head_offset = offsetof(struct ieee80211_mesh_fast_tx, rhash),
+	.hashfn = mesh_table_hash,
+};
+
+static void __mesh_fast_tx_entry_free(void *ptr, void *tblptr)
+{
+	struct ieee80211_mesh_fast_tx *entry = ptr;
+
+	kfree_rcu(entry, fast_tx.rcu_head);
+}
+
+static void mesh_fast_tx_deinit(struct ieee80211_sub_if_data *sdata)
+{
+	struct mesh_tx_cache *cache;
+
+	cache = &sdata->u.mesh.tx_cache;
+	rhashtable_free_and_destroy(&cache->rht,
+				    __mesh_fast_tx_entry_free, NULL);
+}
+
+static void mesh_fast_tx_init(struct ieee80211_sub_if_data *sdata)
+{
+	struct mesh_tx_cache *cache;
+
+	cache = &sdata->u.mesh.tx_cache;
+	rhashtable_init(&cache->rht, &fast_tx_rht_params);
+	INIT_HLIST_HEAD(&cache->walk_head);
+	spin_lock_init(&cache->walk_lock);
+}
+
 static inline bool mpath_expired(struct mesh_path *mpath)
 {
 	return (mpath->flags & MESH_PATH_ACTIVE) &&
@@ -381,6 +417,243 @@ struct mesh_path *mesh_path_new(struct ieee80211_sub_if_data *sdata,
 	return new_mpath;
 }
 
+static void mesh_fast_tx_entry_free(struct mesh_tx_cache *cache,
+				    struct ieee80211_mesh_fast_tx *entry)
+{
+	hlist_del_rcu(&entry->walk_list);
+	rhashtable_remove_fast(&cache->rht, &entry->rhash, fast_tx_rht_params);
+	kfree_rcu(entry, fast_tx.rcu_head);
+}
+
+struct ieee80211_mesh_fast_tx *
+mesh_fast_tx_get(struct ieee80211_sub_if_data *sdata, const u8 *addr)
+{
+	struct ieee80211_mesh_fast_tx *entry;
+	struct mesh_tx_cache *cache;
+
+	cache = &sdata->u.mesh.tx_cache;
+	entry = rhashtable_lookup(&cache->rht, addr, fast_tx_rht_params);
+	if (!entry)
+		return NULL;
+
+	if (!(entry->mpath->flags & MESH_PATH_ACTIVE) ||
+	    mpath_expired(entry->mpath)) {
+		spin_lock_bh(&cache->walk_lock);
+		entry = rhashtable_lookup(&cache->rht, addr, fast_tx_rht_params);
+		if (entry)
+		    mesh_fast_tx_entry_free(cache, entry);
+		spin_unlock_bh(&cache->walk_lock);
+		return NULL;
+	}
+
+	mesh_path_refresh(sdata, entry->mpath, NULL);
+	if (entry->mppath)
+		entry->mppath->exp_time = jiffies;
+	entry->timestamp = jiffies;
+
+	return entry;
+}
+
+void mesh_fast_tx_cache(struct ieee80211_sub_if_data *sdata,
+			struct sk_buff *skb, struct mesh_path *mpath)
+{
+	struct ieee80211_hdr *hdr = (struct ieee80211_hdr *)skb->data;
+	struct ieee80211_tx_info *info = IEEE80211_SKB_CB(skb);
+	struct ieee80211_mesh_fast_tx *entry, *prev;
+	struct ieee80211_mesh_fast_tx build = {};
+	struct ieee80211s_hdr *meshhdr;
+	struct mesh_tx_cache *cache;
+	struct ieee80211_key *key;
+	struct mesh_path *mppath;
+	struct sta_info *sta;
+	u8 *qc;
+
+	if (sdata->noack_map ||
+	    !ieee80211_is_data_qos(hdr->frame_control))
+		return;
+
+	build.fast_tx.hdr_len = ieee80211_hdrlen(hdr->frame_control);
+	meshhdr = (struct ieee80211s_hdr *)(skb->data + build.fast_tx.hdr_len);
+	build.hdrlen = ieee80211_get_mesh_hdrlen(meshhdr);
+
+	cache = &sdata->u.mesh.tx_cache;
+	if (atomic_read(&cache->rht.nelems) >= MESH_FAST_TX_CACHE_MAX_SIZE)
+		return;
+
+	sta = rcu_dereference(mpath->next_hop);
+	if (!sta)
+		return;
+
+	if ((meshhdr->flags & MESH_FLAGS_AE) == MESH_FLAGS_AE_A5_A6) {
+		/* This is required to keep the mppath alive */
+		mppath = mpp_path_lookup(sdata, meshhdr->eaddr1);
+		if (!mppath)
+			return;
+		build.mppath = mppath;
+	} else if (ieee80211_has_a4(hdr->frame_control)) {
+		mppath = mpath;
+	} else {
+		return;
+	}
+
+	/* rate limit, in case fast xmit can't be enabled */
+	if (mppath->fast_tx_check == jiffies)
+		return;
+
+	mppath->fast_tx_check = jiffies;
+
+	/*
+	 * Same use of the sta lock as in ieee80211_check_fast_xmit, in order
+	 * to protect against concurrent sta key updates.
+	 */
+	spin_lock_bh(&sta->lock);
+	key = rcu_access_pointer(sta->ptk[sta->ptk_idx]);
+	if (!key)
+		key = rcu_access_pointer(sdata->default_unicast_key);
+	build.fast_tx.key = key;
+
+	if (key) {
+		bool gen_iv, iv_spc;
+
+		gen_iv = key->conf.flags & IEEE80211_KEY_FLAG_GENERATE_IV;
+		iv_spc = key->conf.flags & IEEE80211_KEY_FLAG_PUT_IV_SPACE;
+
+		if (!(key->flags & KEY_FLAG_UPLOADED_TO_HARDWARE) ||
+		    (key->flags & KEY_FLAG_TAINTED))
+			goto unlock_sta;
+
+		switch (key->conf.cipher) {
+		case WLAN_CIPHER_SUITE_CCMP:
+		case WLAN_CIPHER_SUITE_CCMP_256:
+			if (gen_iv)
+				build.fast_tx.pn_offs = build.fast_tx.hdr_len;
+			if (gen_iv || iv_spc)
+				build.fast_tx.hdr_len += IEEE80211_CCMP_HDR_LEN;
+			break;
+		case WLAN_CIPHER_SUITE_GCMP:
+		case WLAN_CIPHER_SUITE_GCMP_256:
+			if (gen_iv)
+				build.fast_tx.pn_offs = build.fast_tx.hdr_len;
+			if (gen_iv || iv_spc)
+				build.fast_tx.hdr_len += IEEE80211_GCMP_HDR_LEN;
+			break;
+		default:
+			goto unlock_sta;
+		}
+	}
+
+	memcpy(build.addr_key, mppath->dst, ETH_ALEN);
+	build.timestamp = jiffies;
+	build.fast_tx.band = info->band;
+	build.fast_tx.da_offs = offsetof(struct ieee80211_hdr, addr3);
+	build.fast_tx.sa_offs = offsetof(struct ieee80211_hdr, addr4);
+	build.mpath = mpath;
+	memcpy(build.hdr, meshhdr, build.hdrlen);
+	memcpy(build.hdr + build.hdrlen, rfc1042_header, sizeof(rfc1042_header));
+	build.hdrlen += sizeof(rfc1042_header);
+	memcpy(build.fast_tx.hdr, hdr, build.fast_tx.hdr_len);
+
+	hdr = (struct ieee80211_hdr *)build.fast_tx.hdr;
+	if (build.fast_tx.key)
+		hdr->frame_control |= cpu_to_le16(IEEE80211_FCTL_PROTECTED);
+
+	qc = ieee80211_get_qos_ctl(hdr);
+	qc[1] |= IEEE80211_QOS_CTL_MESH_CONTROL_PRESENT >> 8;
+
+	entry = kmemdup(&build, sizeof(build), GFP_ATOMIC);
+	if (!entry)
+		goto unlock_sta;
+
+	spin_lock(&cache->walk_lock);
+	prev = rhashtable_lookup_get_insert_fast(&cache->rht,
+						 &entry->rhash,
+						 fast_tx_rht_params);
+	if (unlikely(IS_ERR(prev))) {
+		kfree(entry);
+		goto unlock_cache;
+	}
+
+	/*
+	 * replace any previous entry in the hash table, in case we're
+	 * replacing it with a different type (e.g. mpath -> mpp)
+	 */
+	if (unlikely(prev)) {
+		rhashtable_replace_fast(&cache->rht, &prev->rhash,
+					&entry->rhash, fast_tx_rht_params);
+		hlist_del_rcu(&prev->walk_list);
+		kfree_rcu(prev, fast_tx.rcu_head);
+	}
+
+	hlist_add_head(&entry->walk_list, &cache->walk_head);
+
+unlock_cache:
+	spin_unlock(&cache->walk_lock);
+unlock_sta:
+	spin_unlock_bh(&sta->lock);
+}
+
+void mesh_fast_tx_gc(struct ieee80211_sub_if_data *sdata)
+{
+	unsigned long timeout = msecs_to_jiffies(MESH_FAST_TX_CACHE_TIMEOUT);
+	struct mesh_tx_cache *cache;
+	struct ieee80211_mesh_fast_tx *entry;
+	struct hlist_node *n;
+
+	cache = &sdata->u.mesh.tx_cache;
+	if (atomic_read(&cache->rht.nelems) < MESH_FAST_TX_CACHE_THRESHOLD_SIZE)
+		return;
+
+	spin_lock_bh(&cache->walk_lock);
+	hlist_for_each_entry_safe(entry, n, &cache->walk_head, walk_list)
+		if (!time_is_after_jiffies(entry->timestamp + timeout))
+			mesh_fast_tx_entry_free(cache, entry);
+	spin_unlock_bh(&cache->walk_lock);
+}
+
+void mesh_fast_tx_flush_mpath(struct mesh_path *mpath)
+{
+	struct ieee80211_sub_if_data *sdata = mpath->sdata;
+	struct mesh_tx_cache *cache = &sdata->u.mesh.tx_cache;
+	struct ieee80211_mesh_fast_tx *entry;
+	struct hlist_node *n;
+
+	cache = &sdata->u.mesh.tx_cache;
+	spin_lock_bh(&cache->walk_lock);
+	hlist_for_each_entry_safe(entry, n, &cache->walk_head, walk_list)
+		if (entry->mpath == mpath)
+			mesh_fast_tx_entry_free(cache, entry);
+	spin_unlock_bh(&cache->walk_lock);
+}
+
+void mesh_fast_tx_flush_sta(struct ieee80211_sub_if_data *sdata,
+			    struct sta_info *sta)
+{
+	struct mesh_tx_cache *cache = &sdata->u.mesh.tx_cache;
+	struct ieee80211_mesh_fast_tx *entry;
+	struct hlist_node *n;
+
+	cache = &sdata->u.mesh.tx_cache;
+	spin_lock_bh(&cache->walk_lock);
+	hlist_for_each_entry_safe(entry, n, &cache->walk_head, walk_list)
+		if (rcu_access_pointer(entry->mpath->next_hop) == sta)
+			mesh_fast_tx_entry_free(cache, entry);
+	spin_unlock_bh(&cache->walk_lock);
+}
+
+void mesh_fast_tx_flush_addr(struct ieee80211_sub_if_data *sdata,
+			     const u8 *addr)
+{
+	struct mesh_tx_cache *cache = &sdata->u.mesh.tx_cache;
+	struct ieee80211_mesh_fast_tx *entry;
+
+	cache = &sdata->u.mesh.tx_cache;
+	spin_lock_bh(&cache->walk_lock);
+	entry = rhashtable_lookup(&cache->rht, addr, fast_tx_rht_params);
+	if (entry)
+		mesh_fast_tx_entry_free(cache, entry);
+	spin_unlock_bh(&cache->walk_lock);
+}
+
 /**
  * mesh_path_add - allocate and add a new path to the mesh path table
  * @dst: destination address of the path (ETH_ALEN length)
@@ -464,6 +737,8 @@ int mpp_path_add(struct ieee80211_sub_if_data *sdata,
 
 	if (ret)
 		kfree(new_mpath);
+	else
+		mesh_fast_tx_flush_addr(sdata, dst);
 
 	sdata->u.mesh.mpp_paths_generation++;
 	return ret;
@@ -523,6 +798,10 @@ static void __mesh_path_del(struct mesh_table *tbl, struct mesh_path *mpath)
 {
 	hlist_del_rcu(&mpath->walk_list);
 	rhashtable_remove_fast(&tbl->rhead, &mpath->rhash, mesh_rht_params);
+	if (tbl == &mpath->sdata->u.mesh.mpp_paths)
+		mesh_fast_tx_flush_addr(mpath->sdata, mpath->dst);
+	else
+		mesh_fast_tx_flush_mpath(mpath);
 	mesh_path_free_rcu(tbl, mpath);
 }
 
@@ -747,6 +1026,7 @@ void mesh_path_fix_nexthop(struct mesh_path *mpath, struct sta_info *next_hop)
 	mpath->exp_time = 0;
 	mpath->flags = MESH_PATH_FIXED | MESH_PATH_SN_VALID;
 	mesh_path_activate(mpath);
+	mesh_fast_tx_flush_mpath(mpath);
 	spin_unlock_bh(&mpath->state_lock);
 	ewma_mesh_fail_avg_init(&next_hop->mesh->fail_avg);
 	/* init it at a low value - 0 start is tricky */
@@ -758,6 +1038,7 @@ void mesh_pathtbl_init(struct ieee80211_sub_if_data *sdata)
 {
 	mesh_table_init(&sdata->u.mesh.mesh_paths);
 	mesh_table_init(&sdata->u.mesh.mpp_paths);
+	mesh_fast_tx_init(sdata);
 }
 
 static
@@ -785,6 +1066,7 @@ void mesh_path_expire(struct ieee80211_sub_if_data *sdata)
 
 void mesh_pathtbl_unregister(struct ieee80211_sub_if_data *sdata)
 {
+	mesh_fast_tx_deinit(sdata);
 	mesh_table_free(&sdata->u.mesh.mesh_paths);
 	mesh_table_free(&sdata->u.mesh.mpp_paths);
 }