summaryrefslogtreecommitdiff
path: root/net/openvswitch/datapath.c
diff options
context:
space:
mode:
Diffstat (limited to 'net/openvswitch/datapath.c')
-rw-r--r--net/openvswitch/datapath.c783
1 files changed, 585 insertions, 198 deletions
diff --git a/net/openvswitch/datapath.c b/net/openvswitch/datapath.c
index 6679e96ab1dc..d5b6e2002bc1 100644
--- a/net/openvswitch/datapath.c
+++ b/net/openvswitch/datapath.c
@@ -1,19 +1,6 @@
+// SPDX-License-Identifier: GPL-2.0-only
/*
* Copyright (c) 2007-2014 Nicira, Inc.
- *
- * This program is free software; you can redistribute it and/or
- * modify it under the terms of version 2 of the GNU General Public
- * License as published by the Free Software Foundation.
- *
- * This program is distributed in the hope that it will be useful, but
- * WITHOUT ANY WARRANTY; without even the implied warranty of
- * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
- * General Public License for more details.
- *
- * You should have received a copy of the GNU General Public License
- * along with this program; if not, write to the Free Software
- * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA
- * 02110-1301, USA
*/
#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
@@ -28,7 +15,6 @@
#include <linux/delay.h>
#include <linux/time.h>
#include <linux/etherdevice.h>
-#include <linux/genetlink.h>
#include <linux/kernel.h>
#include <linux/kthread.h>
#include <linux/mutex.h>
@@ -48,14 +34,18 @@
#include <linux/rculist.h>
#include <linux/dmi.h>
#include <net/genetlink.h>
+#include <net/gso.h>
#include <net/net_namespace.h>
#include <net/netns/generic.h>
+#include <net/pkt_cls.h>
#include "datapath.h"
+#include "drop.h"
#include "flow.h"
#include "flow_table.h"
#include "flow_netlink.h"
#include "meter.h"
+#include "openvswitch_trace.h"
#include "vport-internal_dev.h"
#include "vport-netdev.h"
@@ -143,6 +133,10 @@ static int queue_userspace_packet(struct datapath *dp, struct sk_buff *,
const struct dp_upcall_info *,
uint32_t cutlen);
+static void ovs_dp_masks_rebalance(struct work_struct *work);
+
+static int ovs_dp_set_upcall_portids(struct datapath *, const struct nlattr *);
+
/* Must be called with rcu_read_lock or ovs_mutex. */
const char *ovs_dp_name(const struct datapath *dp)
{
@@ -176,6 +170,7 @@ static void destroy_dp_rcu(struct rcu_head *rcu)
free_percpu(dp->stats_percpu);
kfree(dp->ports);
ovs_meters_exit(dp);
+ kfree(rcu_dereference_raw(dp->upcall_portids));
kfree(dp);
}
@@ -192,7 +187,8 @@ struct vport *ovs_lookup_vport(const struct datapath *dp, u16 port_no)
struct hlist_head *head;
head = vport_hash_bucket(dp, port_no);
- hlist_for_each_entry_rcu(vport, head, dp_hash_node) {
+ hlist_for_each_entry_rcu(vport, head, dp_hash_node,
+ lockdep_ovsl_is_held()) {
if (vport->port_no == port_no)
return vport;
}
@@ -214,6 +210,26 @@ static struct vport *new_vport(const struct vport_parms *parms)
return vport;
}
+static void ovs_vport_update_upcall_stats(struct sk_buff *skb,
+ const struct dp_upcall_info *upcall_info,
+ bool upcall_result)
+{
+ struct vport *p = OVS_CB(skb)->input_vport;
+ struct vport_upcall_stats_percpu *stats;
+
+ if (upcall_info->cmd != OVS_PACKET_CMD_MISS &&
+ upcall_info->cmd != OVS_PACKET_CMD_ACTION)
+ return;
+
+ stats = this_cpu_ptr(p->upcall_stats);
+ u64_stats_update_begin(&stats->syncp);
+ if (upcall_result)
+ u64_stats_inc(&stats->n_success);
+ else
+ u64_stats_inc(&stats->n_fail);
+ u64_stats_update_end(&stats->syncp);
+}
+
void ovs_dp_detach_port(struct vport *p)
{
ASSERT_OVSL();
@@ -228,38 +244,76 @@ void ovs_dp_detach_port(struct vport *p)
/* Must be called with rcu_read_lock. */
void ovs_dp_process_packet(struct sk_buff *skb, struct sw_flow_key *key)
{
+ struct ovs_pcpu_storage *ovs_pcpu = this_cpu_ptr(ovs_pcpu_storage);
const struct vport *p = OVS_CB(skb)->input_vport;
struct datapath *dp = p->dp;
struct sw_flow *flow;
struct sw_flow_actions *sf_acts;
struct dp_stats_percpu *stats;
+ bool ovs_pcpu_locked = false;
u64 *stats_counter;
u32 n_mask_hit;
+ u32 n_cache_hit;
+ int error;
stats = this_cpu_ptr(dp->stats_percpu);
/* Look up flow. */
- flow = ovs_flow_tbl_lookup_stats(&dp->table, key, &n_mask_hit);
+ flow = ovs_flow_tbl_lookup_stats(&dp->table, key, skb_get_hash(skb),
+ &n_mask_hit, &n_cache_hit);
if (unlikely(!flow)) {
struct dp_upcall_info upcall;
- int error;
memset(&upcall, 0, sizeof(upcall));
upcall.cmd = OVS_PACKET_CMD_MISS;
- upcall.portid = ovs_vport_find_upcall_portid(p, skb);
+
+ if (OVS_CB(skb)->upcall_pid)
+ upcall.portid = OVS_CB(skb)->upcall_pid;
+ else if (dp->user_features & OVS_DP_F_DISPATCH_UPCALL_PER_CPU)
+ upcall.portid =
+ ovs_dp_get_upcall_portid(dp, smp_processor_id());
+ else
+ upcall.portid = ovs_vport_find_upcall_portid(p, skb);
+
upcall.mru = OVS_CB(skb)->mru;
error = ovs_dp_upcall(dp, skb, key, &upcall, 0);
- if (unlikely(error))
- kfree_skb(skb);
- else
+ switch (error) {
+ case 0:
+ case -EAGAIN:
+ case -ERESTARTSYS:
+ case -EINTR:
consume_skb(skb);
+ break;
+ default:
+ kfree_skb(skb);
+ break;
+ }
stats_counter = &stats->n_missed;
goto out;
}
ovs_flow_stats_update(flow, key->tp.flags, skb);
sf_acts = rcu_dereference(flow->sf_acts);
- ovs_execute_actions(dp, skb, sf_acts, key);
+ /* This path can be invoked recursively: Use the current task to
+ * identify recursive invocation - the lock must be acquired only once.
+ * Even with disabled bottom halves this can be preempted on PREEMPT_RT.
+ * Limit the locking to RT to avoid assigning `owner' if it can be
+ * avoided.
+ */
+ if (IS_ENABLED(CONFIG_PREEMPT_RT) && ovs_pcpu->owner != current) {
+ local_lock_nested_bh(&ovs_pcpu_storage->bh_lock);
+ ovs_pcpu->owner = current;
+ ovs_pcpu_locked = true;
+ }
+
+ error = ovs_execute_actions(dp, skb, sf_acts, key);
+ if (unlikely(error))
+ net_dbg_ratelimited("ovs: action execution error on datapath %s: %d\n",
+ ovs_dp_name(dp), error);
+ if (ovs_pcpu_locked) {
+ ovs_pcpu->owner = NULL;
+ local_unlock_nested_bh(&ovs_pcpu_storage->bh_lock);
+ }
stats_counter = &stats->n_hit;
@@ -268,6 +322,7 @@ out:
u64_stats_update_begin(&stats->syncp);
(*stats_counter)++;
stats->n_mask_hit += n_mask_hit;
+ stats->n_cache_hit += n_cache_hit;
u64_stats_update_end(&stats->syncp);
}
@@ -279,6 +334,9 @@ int ovs_dp_upcall(struct datapath *dp, struct sk_buff *skb,
struct dp_stats_percpu *stats;
int err;
+ if (trace_ovs_dp_upcall_enabled())
+ trace_ovs_dp_upcall(dp, skb, key, upcall_info);
+
if (upcall_info->portid == 0) {
err = -ENOTCONN;
goto err;
@@ -288,6 +346,8 @@ int ovs_dp_upcall(struct datapath *dp, struct sk_buff *skb,
err = queue_userspace_packet(dp, skb, key, upcall_info, cutlen);
else
err = queue_gso_packets(dp, skb, key, upcall_info, cutlen);
+
+ ovs_vport_update_upcall_stats(skb, upcall_info, !err);
if (err)
goto err;
@@ -306,14 +366,14 @@ err:
static int queue_gso_packets(struct datapath *dp, struct sk_buff *skb,
const struct sw_flow_key *key,
const struct dp_upcall_info *upcall_info,
- uint32_t cutlen)
+ uint32_t cutlen)
{
unsigned int gso_type = skb_shinfo(skb)->gso_type;
struct sw_flow_key later_key;
struct sk_buff *segs, *nskb;
int err;
- BUILD_BUG_ON(sizeof(*OVS_CB(skb)) > SKB_SGO_CB_OFFSET);
+ BUILD_BUG_ON(sizeof(*OVS_CB(skb)) > SKB_GSO_CB_OFFSET);
segs = __skb_gso_segment(skb, NETIF_F_SG, false);
if (IS_ERR(segs))
return PTR_ERR(segs);
@@ -330,8 +390,7 @@ static int queue_gso_packets(struct datapath *dp, struct sk_buff *skb,
}
/* Queue all of the segments. */
- skb = segs;
- do {
+ skb_list_walk_safe(segs, skb, nskb) {
if (gso_type & SKB_GSO_UDP && skb != segs)
key = &later_key;
@@ -339,17 +398,15 @@ static int queue_gso_packets(struct datapath *dp, struct sk_buff *skb,
if (err)
break;
- } while ((skb = skb->next));
+ }
/* Free all of the segments. */
- skb = segs;
- do {
- nskb = skb->next;
+ skb_list_walk_safe(segs, skb, nskb) {
if (err)
kfree_skb(skb);
else
consume_skb(skb);
- } while ((skb = nskb));
+ }
return err;
}
@@ -359,7 +416,8 @@ static size_t upcall_msg_size(const struct dp_upcall_info *upcall_info,
size_t size = NLMSG_ALIGN(sizeof(struct ovs_header))
+ nla_total_size(hdrlen) /* OVS_PACKET_ATTR_PACKET */
+ nla_total_size(ovs_key_attr_size()) /* OVS_PACKET_ATTR_KEY */
- + nla_total_size(sizeof(unsigned int)); /* OVS_PACKET_ATTR_LEN */
+ + nla_total_size(sizeof(unsigned int)) /* OVS_PACKET_ATTR_LEN */
+ + nla_total_size(sizeof(u64)); /* OVS_PACKET_ATTR_HASH */
/* OVS_PACKET_ATTR_USERDATA */
if (upcall_info->userdata)
@@ -402,6 +460,7 @@ static int queue_userspace_packet(struct datapath *dp, struct sk_buff *skb,
size_t len;
unsigned int hlen;
int err, dp_ifindex;
+ u64 hash;
dp_ifindex = get_dpifindex(dp);
if (!dp_ifindex)
@@ -448,10 +507,15 @@ static int queue_userspace_packet(struct datapath *dp, struct sk_buff *skb,
upcall = genlmsg_put(user_skb, 0, 0, &dp_packet_genl_family,
0, upcall_info->cmd);
+ if (!upcall) {
+ err = -EINVAL;
+ goto out;
+ }
upcall->dp_ifindex = dp_ifindex;
err = ovs_nla_put_key(key, key, OVS_PACKET_ATTR_KEY, false, user_skb);
- BUG_ON(err);
+ if (err)
+ goto out;
if (upcall_info->userdata)
__nla_put(user_skb, OVS_PACKET_ATTR_USERDATA,
@@ -459,15 +523,26 @@ static int queue_userspace_packet(struct datapath *dp, struct sk_buff *skb,
nla_data(upcall_info->userdata));
if (upcall_info->egress_tun_info) {
- nla = nla_nest_start(user_skb, OVS_PACKET_ATTR_EGRESS_TUN_KEY);
+ nla = nla_nest_start_noflag(user_skb,
+ OVS_PACKET_ATTR_EGRESS_TUN_KEY);
+ if (!nla) {
+ err = -EMSGSIZE;
+ goto out;
+ }
err = ovs_nla_put_tunnel_info(user_skb,
upcall_info->egress_tun_info);
- BUG_ON(err);
+ if (err)
+ goto out;
+
nla_nest_end(user_skb, nla);
}
if (upcall_info->actions_len) {
- nla = nla_nest_start(user_skb, OVS_PACKET_ATTR_ACTIONS);
+ nla = nla_nest_start_noflag(user_skb, OVS_PACKET_ATTR_ACTIONS);
+ if (!nla) {
+ err = -EMSGSIZE;
+ goto out;
+ }
err = ovs_nla_put_actions(upcall_info->actions,
upcall_info->actions_len,
user_skb);
@@ -478,23 +553,30 @@ static int queue_userspace_packet(struct datapath *dp, struct sk_buff *skb,
}
/* Add OVS_PACKET_ATTR_MRU */
- if (upcall_info->mru) {
- if (nla_put_u16(user_skb, OVS_PACKET_ATTR_MRU,
- upcall_info->mru)) {
- err = -ENOBUFS;
- goto out;
- }
- pad_packet(dp, user_skb);
+ if (upcall_info->mru &&
+ nla_put_u16(user_skb, OVS_PACKET_ATTR_MRU, upcall_info->mru)) {
+ err = -ENOBUFS;
+ goto out;
}
/* Add OVS_PACKET_ATTR_LEN when packet is truncated */
- if (cutlen > 0) {
- if (nla_put_u32(user_skb, OVS_PACKET_ATTR_LEN,
- skb->len)) {
- err = -ENOBUFS;
- goto out;
- }
- pad_packet(dp, user_skb);
+ if (cutlen > 0 &&
+ nla_put_u32(user_skb, OVS_PACKET_ATTR_LEN, skb->len)) {
+ err = -ENOBUFS;
+ goto out;
+ }
+
+ /* Add OVS_PACKET_ATTR_HASH */
+ hash = skb_get_hash_raw(skb);
+ if (skb->sw_hash)
+ hash |= OVS_PACKET_HASH_SW_BIT;
+
+ if (skb->l4_hash)
+ hash |= OVS_PACKET_HASH_L4_BIT;
+
+ if (nla_put(user_skb, OVS_PACKET_ATTR_HASH, sizeof (u64), &hash)) {
+ err = -ENOBUFS;
+ goto out;
}
/* Only reserve room for attribute header, packet data is added
@@ -519,14 +601,15 @@ static int queue_userspace_packet(struct datapath *dp, struct sk_buff *skb,
out:
if (err)
skb_tx_error(skb);
- kfree_skb(user_skb);
- kfree_skb(nskb);
+ consume_skb(user_skb);
+ consume_skb(nskb);
+
return err;
}
static int ovs_packet_cmd_execute(struct sk_buff *skb, struct genl_info *info)
{
- struct ovs_header *ovs_header = info->userhdr;
+ struct ovs_header *ovs_header = genl_info_userhdr(info);
struct net *net = sock_net(skb->sk);
struct nlattr **a = info->attrs;
struct sw_flow_actions *acts;
@@ -536,6 +619,7 @@ static int ovs_packet_cmd_execute(struct sk_buff *skb, struct genl_info *info)
struct datapath *dp;
struct vport *input_vport;
u16 mru = 0;
+ u64 hash;
int len;
int err;
bool log = !a[OVS_PACKET_ATTR_PROBE];
@@ -561,6 +645,17 @@ static int ovs_packet_cmd_execute(struct sk_buff *skb, struct genl_info *info)
}
OVS_CB(packet)->mru = mru;
+ if (a[OVS_PACKET_ATTR_HASH]) {
+ hash = nla_get_u64(a[OVS_PACKET_ATTR_HASH]);
+
+ __skb_set_hash(packet, hash & 0xFFFFFFFFULL,
+ !!(hash & OVS_PACKET_HASH_SW_BIT),
+ !!(hash & OVS_PACKET_HASH_L4_BIT));
+ }
+
+ OVS_CB(packet)->upcall_pid =
+ nla_get_u32_default(a[OVS_PACKET_ATTR_UPCALL_PID], 0);
+
/* Build an sw_flow for sending this packet. */
flow = ovs_flow_alloc();
err = PTR_ERR(flow);
@@ -599,7 +694,13 @@ static int ovs_packet_cmd_execute(struct sk_buff *skb, struct genl_info *info)
sf_acts = rcu_dereference(flow->sf_acts);
local_bh_disable();
+ local_lock_nested_bh(&ovs_pcpu_storage->bh_lock);
+ if (IS_ENABLED(CONFIG_PREEMPT_RT))
+ this_cpu_write(ovs_pcpu_storage->owner, current);
err = ovs_execute_actions(dp, packet, sf_acts, &flow->key);
+ if (IS_ENABLED(CONFIG_PREEMPT_RT))
+ this_cpu_write(ovs_pcpu_storage->owner, NULL);
+ local_unlock_nested_bh(&ovs_pcpu_storage->bh_lock);
local_bh_enable();
rcu_read_unlock();
@@ -622,12 +723,14 @@ static const struct nla_policy packet_policy[OVS_PACKET_ATTR_MAX + 1] = {
[OVS_PACKET_ATTR_ACTIONS] = { .type = NLA_NESTED },
[OVS_PACKET_ATTR_PROBE] = { .type = NLA_FLAG },
[OVS_PACKET_ATTR_MRU] = { .type = NLA_U16 },
+ [OVS_PACKET_ATTR_HASH] = { .type = NLA_U64 },
+ [OVS_PACKET_ATTR_UPCALL_PID] = { .type = NLA_U32 },
};
-static const struct genl_ops dp_packet_genl_ops[] = {
+static const struct genl_small_ops dp_packet_genl_ops[] = {
{ .cmd = OVS_PACKET_CMD_EXECUTE,
+ .validate = GENL_DONT_VALIDATE_STRICT | GENL_DONT_VALIDATE_DUMP,
.flags = GENL_UNS_ADMIN_PERM, /* Requires CAP_NET_ADMIN privilege. */
- .policy = packet_policy,
.doit = ovs_packet_cmd_execute
}
};
@@ -637,10 +740,12 @@ static struct genl_family dp_packet_genl_family __ro_after_init = {
.name = OVS_PACKET_FAMILY,
.version = OVS_PACKET_VERSION,
.maxattr = OVS_PACKET_ATTR_MAX,
+ .policy = packet_policy,
.netnsok = true,
.parallel_ops = true,
- .ops = dp_packet_genl_ops,
- .n_ops = ARRAY_SIZE(dp_packet_genl_ops),
+ .small_ops = dp_packet_genl_ops,
+ .n_small_ops = ARRAY_SIZE(dp_packet_genl_ops),
+ .resv_start_op = OVS_PACKET_CMD_EXECUTE + 1,
.module = THIS_MODULE,
};
@@ -664,14 +769,15 @@ static void get_dp_stats(const struct datapath *dp, struct ovs_dp_stats *stats,
percpu_stats = per_cpu_ptr(dp->stats_percpu, i);
do {
- start = u64_stats_fetch_begin_irq(&percpu_stats->syncp);
+ start = u64_stats_fetch_begin(&percpu_stats->syncp);
local_stats = *percpu_stats;
- } while (u64_stats_fetch_retry_irq(&percpu_stats->syncp, start));
+ } while (u64_stats_fetch_retry(&percpu_stats->syncp, start));
stats->n_hit += local_stats.n_hit;
stats->n_missed += local_stats.n_missed;
stats->n_lost += local_stats.n_lost;
mega_stats->n_mask_hit += local_stats.n_mask_hit;
+ mega_stats->n_cache_hit += local_stats.n_cache_hit;
}
}
@@ -697,9 +803,13 @@ static size_t ovs_flow_cmd_msg_size(const struct sw_flow_actions *acts,
{
size_t len = NLMSG_ALIGN(sizeof(struct ovs_header));
- /* OVS_FLOW_ATTR_UFID */
+ /* OVS_FLOW_ATTR_UFID, or unmasked flow key as fallback
+ * see ovs_nla_put_identifier()
+ */
if (sfid && ovs_identifier_is_ufid(sfid))
len += nla_total_size(sfid->ufid_len);
+ else
+ len += nla_total_size(ovs_key_attr_size());
/* OVS_FLOW_ATTR_KEY */
if (!sfid || should_fill_key(sfid, ufid_flags))
@@ -764,7 +874,7 @@ static int ovs_flow_cmd_fill_actions(const struct sw_flow *flow,
* This can only fail for dump operations because the skb is always
* properly sized for single flows.
*/
- start = nla_nest_start(skb, OVS_FLOW_ATTR_ACTIONS);
+ start = nla_nest_start_noflag(skb, OVS_FLOW_ATTR_ACTIONS);
if (start) {
const struct sw_flow_actions *sf_acts;
@@ -875,7 +985,10 @@ static struct sk_buff *ovs_flow_cmd_build_info(const struct sw_flow *flow,
retval = ovs_flow_cmd_fill_info(flow, dp_ifindex, skb,
info->snd_portid, info->snd_seq, 0,
cmd, ufid_flags);
- BUG_ON(retval < 0);
+ if (WARN_ON_ONCE(retval < 0)) {
+ kfree_skb(skb);
+ skb = ERR_PTR(retval);
+ }
return skb;
}
@@ -883,11 +996,12 @@ static int ovs_flow_cmd_new(struct sk_buff *skb, struct genl_info *info)
{
struct net *net = sock_net(skb->sk);
struct nlattr **a = info->attrs;
- struct ovs_header *ovs_header = info->userhdr;
+ struct ovs_header *ovs_header = genl_info_userhdr(info);
struct sw_flow *flow = NULL, *new_flow;
struct sw_flow_mask mask;
struct sk_buff *reply;
struct datapath *dp;
+ struct sw_flow_key *key;
struct sw_flow_actions *acts;
struct sw_flow_match match;
u32 ufid_flags = ovs_nla_get_ufid_flags(a[OVS_FLOW_ATTR_UFID_FLAGS]);
@@ -915,30 +1029,32 @@ static int ovs_flow_cmd_new(struct sk_buff *skb, struct genl_info *info)
}
/* Extract key. */
- ovs_match_init(&match, &new_flow->key, false, &mask);
+ key = kzalloc(sizeof(*key), GFP_KERNEL);
+ if (!key) {
+ error = -ENOMEM;
+ goto err_kfree_flow;
+ }
+
+ ovs_match_init(&match, key, false, &mask);
error = ovs_nla_get_match(net, &match, a[OVS_FLOW_ATTR_KEY],
a[OVS_FLOW_ATTR_MASK], log);
if (error)
- goto err_kfree_flow;
+ goto err_kfree_key;
+
+ ovs_flow_mask_key(&new_flow->key, key, true, &mask);
/* Extract flow identifier. */
error = ovs_nla_get_identifier(&new_flow->id, a[OVS_FLOW_ATTR_UFID],
- &new_flow->key, log);
+ key, log);
if (error)
- goto err_kfree_flow;
-
- /* unmasked key is needed to match when ufid is not used. */
- if (ovs_identifier_is_key(&new_flow->id))
- match.key = new_flow->id.unmasked_key;
-
- ovs_flow_mask_key(&new_flow->key, &new_flow->key, true, &mask);
+ goto err_kfree_key;
/* Validate actions. */
error = ovs_nla_copy_actions(net, a[OVS_FLOW_ATTR_ACTIONS],
&new_flow->key, &acts, log);
if (error) {
OVS_NLERR(log, "Flow actions may not be safe on all matching packets.");
- goto err_kfree_flow;
+ goto err_kfree_key;
}
reply = ovs_flow_cmd_alloc_info(acts, &new_flow->id, info, false,
@@ -959,7 +1075,7 @@ static int ovs_flow_cmd_new(struct sk_buff *skb, struct genl_info *info)
if (ovs_identifier_is_ufid(&new_flow->id))
flow = ovs_flow_tbl_lookup_ufid(&dp->table, &new_flow->id);
if (!flow)
- flow = ovs_flow_tbl_lookup(&dp->table, &new_flow->key);
+ flow = ovs_flow_tbl_lookup(&dp->table, key);
if (likely(!flow)) {
rcu_assign_pointer(new_flow->sf_acts, acts);
@@ -1029,6 +1145,8 @@ static int ovs_flow_cmd_new(struct sk_buff *skb, struct genl_info *info)
if (reply)
ovs_notify(&dp_flow_genl_family, reply, info);
+
+ kfree(key);
return 0;
err_unlock_ovs:
@@ -1036,6 +1154,8 @@ err_unlock_ovs:
kfree_skb(reply);
err_kfree_acts:
ovs_nla_free_flow_actions(acts);
+err_kfree_key:
+ kfree(key);
err_kfree_flow:
ovs_flow_free(new_flow, false);
error:
@@ -1043,11 +1163,12 @@ error:
}
/* Factor out action copy to avoid "Wframe-larger-than=1024" warning. */
-static struct sw_flow_actions *get_flow_actions(struct net *net,
- const struct nlattr *a,
- const struct sw_flow_key *key,
- const struct sw_flow_mask *mask,
- bool log)
+static noinline_for_stack
+struct sw_flow_actions *get_flow_actions(struct net *net,
+ const struct nlattr *a,
+ const struct sw_flow_key *key,
+ const struct sw_flow_mask *mask,
+ bool log)
{
struct sw_flow_actions *acts;
struct sw_flow_key masked_key;
@@ -1077,12 +1198,13 @@ static struct sw_flow_actions *get_flow_actions(struct net *net,
* we should not to return match object with dangling reference
* to mask.
* */
-static int ovs_nla_init_match_and_action(struct net *net,
- struct sw_flow_match *match,
- struct sw_flow_key *key,
- struct nlattr **a,
- struct sw_flow_actions **acts,
- bool log)
+static noinline_for_stack int
+ovs_nla_init_match_and_action(struct net *net,
+ struct sw_flow_match *match,
+ struct sw_flow_key *key,
+ struct nlattr **a,
+ struct sw_flow_actions **acts,
+ bool log)
{
struct sw_flow_mask mask;
int error = 0;
@@ -1121,7 +1243,7 @@ static int ovs_flow_cmd_set(struct sk_buff *skb, struct genl_info *info)
{
struct net *net = sock_net(skb->sk);
struct nlattr **a = info->attrs;
- struct ovs_header *ovs_header = info->userhdr;
+ struct ovs_header *ovs_header = genl_info_userhdr(info);
struct sw_flow_key key;
struct sw_flow *flow;
struct sk_buff *reply = NULL;
@@ -1222,7 +1344,7 @@ error:
static int ovs_flow_cmd_get(struct sk_buff *skb, struct genl_info *info)
{
struct nlattr **a = info->attrs;
- struct ovs_header *ovs_header = info->userhdr;
+ struct ovs_header *ovs_header = genl_info_userhdr(info);
struct net *net = sock_net(skb->sk);
struct sw_flow_key key;
struct sk_buff *reply;
@@ -1281,7 +1403,7 @@ unlock:
static int ovs_flow_cmd_del(struct sk_buff *skb, struct genl_info *info)
{
struct nlattr **a = info->attrs;
- struct ovs_header *ovs_header = info->userhdr;
+ struct ovs_header *ovs_header = genl_info_userhdr(info);
struct net *net = sock_net(skb->sk);
struct sw_flow_key key;
struct sk_buff *reply;
@@ -1330,7 +1452,7 @@ static int ovs_flow_cmd_del(struct sk_buff *skb, struct genl_info *info)
reply = ovs_flow_cmd_alloc_info((const struct sw_flow_actions __force *) flow->sf_acts,
&flow->id, info, false, ufid_flags);
if (likely(reply)) {
- if (likely(!IS_ERR(reply))) {
+ if (!IS_ERR(reply)) {
rcu_read_lock(); /*To keep RCU checker happy. */
err = ovs_flow_cmd_fill_info(flow, ovs_header->dp_ifindex,
reply, info->snd_portid,
@@ -1338,14 +1460,19 @@ static int ovs_flow_cmd_del(struct sk_buff *skb, struct genl_info *info)
OVS_FLOW_CMD_DEL,
ufid_flags);
rcu_read_unlock();
- BUG_ON(err < 0);
+ if (WARN_ON_ONCE(err < 0)) {
+ kfree_skb(reply);
+ goto out_free;
+ }
ovs_notify(&dp_flow_genl_family, reply, info);
} else {
- netlink_set_err(sock_net(skb->sk)->genl_sock, 0, 0, PTR_ERR(reply));
+ netlink_set_err(sock_net(skb->sk)->genl_sock, 0, 0,
+ PTR_ERR(reply));
}
}
+out_free:
ovs_flow_free(flow, true);
return 0;
unlock:
@@ -1362,8 +1489,8 @@ static int ovs_flow_cmd_dump(struct sk_buff *skb, struct netlink_callback *cb)
u32 ufid_flags;
int err;
- err = genlmsg_parse(cb->nlh, &dp_flow_genl_family, a,
- OVS_FLOW_ATTR_MAX, flow_policy, NULL);
+ err = genlmsg_parse_deprecated(cb->nlh, &dp_flow_genl_family, a,
+ OVS_FLOW_ATTR_MAX, flow_policy, NULL);
if (err)
return err;
ufid_flags = ovs_nla_get_ufid_flags(a[OVS_FLOW_ATTR_UFID_FLAGS]);
@@ -1409,26 +1536,26 @@ static const struct nla_policy flow_policy[OVS_FLOW_ATTR_MAX + 1] = {
[OVS_FLOW_ATTR_UFID_FLAGS] = { .type = NLA_U32 },
};
-static const struct genl_ops dp_flow_genl_ops[] = {
+static const struct genl_small_ops dp_flow_genl_ops[] = {
{ .cmd = OVS_FLOW_CMD_NEW,
+ .validate = GENL_DONT_VALIDATE_STRICT | GENL_DONT_VALIDATE_DUMP,
.flags = GENL_UNS_ADMIN_PERM, /* Requires CAP_NET_ADMIN privilege. */
- .policy = flow_policy,
.doit = ovs_flow_cmd_new
},
{ .cmd = OVS_FLOW_CMD_DEL,
+ .validate = GENL_DONT_VALIDATE_STRICT | GENL_DONT_VALIDATE_DUMP,
.flags = GENL_UNS_ADMIN_PERM, /* Requires CAP_NET_ADMIN privilege. */
- .policy = flow_policy,
.doit = ovs_flow_cmd_del
},
{ .cmd = OVS_FLOW_CMD_GET,
+ .validate = GENL_DONT_VALIDATE_STRICT | GENL_DONT_VALIDATE_DUMP,
.flags = 0, /* OK for unprivileged users. */
- .policy = flow_policy,
.doit = ovs_flow_cmd_get,
.dumpit = ovs_flow_cmd_dump
},
{ .cmd = OVS_FLOW_CMD_SET,
+ .validate = GENL_DONT_VALIDATE_STRICT | GENL_DONT_VALIDATE_DUMP,
.flags = GENL_UNS_ADMIN_PERM, /* Requires CAP_NET_ADMIN privilege. */
- .policy = flow_policy,
.doit = ovs_flow_cmd_set,
},
};
@@ -1438,10 +1565,12 @@ static struct genl_family dp_flow_genl_family __ro_after_init = {
.name = OVS_FLOW_FAMILY,
.version = OVS_FLOW_VERSION,
.maxattr = OVS_FLOW_ATTR_MAX,
+ .policy = flow_policy,
.netnsok = true,
.parallel_ops = true,
- .ops = dp_flow_genl_ops,
- .n_ops = ARRAY_SIZE(dp_flow_genl_ops),
+ .small_ops = dp_flow_genl_ops,
+ .n_small_ops = ARRAY_SIZE(dp_flow_genl_ops),
+ .resv_start_op = OVS_FLOW_CMD_SET + 1,
.mcgrps = &ovs_dp_flow_multicast_group,
.n_mcgrps = 1,
.module = THIS_MODULE,
@@ -1455,6 +1584,8 @@ static size_t ovs_dp_cmd_msg_size(void)
msgsize += nla_total_size_64bit(sizeof(struct ovs_dp_stats));
msgsize += nla_total_size_64bit(sizeof(struct ovs_dp_megaflow_stats));
msgsize += nla_total_size(sizeof(u32)); /* OVS_DP_ATTR_USER_FEATURES */
+ msgsize += nla_total_size(sizeof(u32)); /* OVS_DP_ATTR_MASKS_CACHE_SIZE */
+ msgsize += nla_total_size(sizeof(u32) * nr_cpu_ids); /* OVS_DP_ATTR_PER_CPU_PIDS */
return msgsize;
}
@@ -1466,10 +1597,11 @@ static int ovs_dp_cmd_fill_info(struct datapath *dp, struct sk_buff *skb,
struct ovs_header *ovs_header;
struct ovs_dp_stats dp_stats;
struct ovs_dp_megaflow_stats dp_megaflow_stats;
- int err;
+ struct dp_nlsk_pids *pids = ovsl_dereference(dp->upcall_portids);
+ int err, pids_len;
ovs_header = genlmsg_put(skb, portid, seq, &dp_datapath_genl_family,
- flags, cmd);
+ flags, cmd);
if (!ovs_header)
goto error;
@@ -1492,6 +1624,16 @@ static int ovs_dp_cmd_fill_info(struct datapath *dp, struct sk_buff *skb,
if (nla_put_u32(skb, OVS_DP_ATTR_USER_FEATURES, dp->user_features))
goto nla_put_failure;
+ if (nla_put_u32(skb, OVS_DP_ATTR_MASKS_CACHE_SIZE,
+ ovs_flow_tbl_masks_cache_size(&dp->table)))
+ goto nla_put_failure;
+
+ if (dp->user_features & OVS_DP_F_DISPATCH_UPCALL_PER_CPU && pids) {
+ pids_len = min(pids->n_pids, nr_cpu_ids) * sizeof(u32);
+ if (nla_put(skb, OVS_DP_ATTR_PER_CPU_PIDS, pids_len, &pids->pids))
+ goto nla_put_failure;
+ }
+
genlmsg_end(skb, ovs_header);
return 0;
@@ -1524,22 +1666,147 @@ static struct datapath *lookup_datapath(struct net *net,
return dp ? dp : ERR_PTR(-ENODEV);
}
-static void ovs_dp_reset_user_features(struct sk_buff *skb, struct genl_info *info)
+static void ovs_dp_reset_user_features(struct sk_buff *skb,
+ struct genl_info *info)
{
struct datapath *dp;
- dp = lookup_datapath(sock_net(skb->sk), info->userhdr, info->attrs);
+ dp = lookup_datapath(sock_net(skb->sk), genl_info_userhdr(info),
+ info->attrs);
if (IS_ERR(dp))
return;
- WARN(dp->user_features, "Dropping previously announced user features\n");
+ pr_warn("%s: Dropping previously announced user features\n",
+ ovs_dp_name(dp));
dp->user_features = 0;
}
-static void ovs_dp_change(struct datapath *dp, struct nlattr *a[])
+static int ovs_dp_set_upcall_portids(struct datapath *dp,
+ const struct nlattr *ids)
+{
+ struct dp_nlsk_pids *old, *dp_nlsk_pids;
+
+ if (!nla_len(ids) || nla_len(ids) % sizeof(u32))
+ return -EINVAL;
+
+ old = ovsl_dereference(dp->upcall_portids);
+
+ dp_nlsk_pids = kmalloc(sizeof(*dp_nlsk_pids) + nla_len(ids),
+ GFP_KERNEL);
+ if (!dp_nlsk_pids)
+ return -ENOMEM;
+
+ dp_nlsk_pids->n_pids = nla_len(ids) / sizeof(u32);
+ nla_memcpy(dp_nlsk_pids->pids, ids, nla_len(ids));
+
+ rcu_assign_pointer(dp->upcall_portids, dp_nlsk_pids);
+
+ kfree_rcu(old, rcu);
+
+ return 0;
+}
+
+u32 ovs_dp_get_upcall_portid(const struct datapath *dp, uint32_t cpu_id)
+{
+ struct dp_nlsk_pids *dp_nlsk_pids;
+
+ dp_nlsk_pids = rcu_dereference(dp->upcall_portids);
+
+ if (dp_nlsk_pids) {
+ if (cpu_id < dp_nlsk_pids->n_pids) {
+ return dp_nlsk_pids->pids[cpu_id];
+ } else if (dp_nlsk_pids->n_pids > 0 &&
+ cpu_id >= dp_nlsk_pids->n_pids) {
+ /* If the number of netlink PIDs is mismatched with
+ * the number of CPUs as seen by the kernel, log this
+ * and send the upcall to an arbitrary socket (0) in
+ * order to not drop packets
+ */
+ pr_info_ratelimited("cpu_id mismatch with handler threads");
+ return dp_nlsk_pids->pids[cpu_id %
+ dp_nlsk_pids->n_pids];
+ } else {
+ return 0;
+ }
+ } else {
+ return 0;
+ }
+}
+
+static int ovs_dp_change(struct datapath *dp, struct nlattr *a[])
+{
+ u32 user_features = 0, old_features = dp->user_features;
+ int err;
+
+ if (a[OVS_DP_ATTR_USER_FEATURES]) {
+ user_features = nla_get_u32(a[OVS_DP_ATTR_USER_FEATURES]);
+
+ if (user_features & ~(OVS_DP_F_VPORT_PIDS |
+ OVS_DP_F_UNALIGNED |
+ OVS_DP_F_TC_RECIRC_SHARING |
+ OVS_DP_F_DISPATCH_UPCALL_PER_CPU))
+ return -EOPNOTSUPP;
+
+#if !IS_ENABLED(CONFIG_NET_TC_SKB_EXT)
+ if (user_features & OVS_DP_F_TC_RECIRC_SHARING)
+ return -EOPNOTSUPP;
+#endif
+ }
+
+ if (a[OVS_DP_ATTR_MASKS_CACHE_SIZE]) {
+ int err;
+ u32 cache_size;
+
+ cache_size = nla_get_u32(a[OVS_DP_ATTR_MASKS_CACHE_SIZE]);
+ err = ovs_flow_tbl_masks_cache_resize(&dp->table, cache_size);
+ if (err)
+ return err;
+ }
+
+ dp->user_features = user_features;
+
+ if (dp->user_features & OVS_DP_F_DISPATCH_UPCALL_PER_CPU &&
+ a[OVS_DP_ATTR_PER_CPU_PIDS]) {
+ /* Upcall Netlink Port IDs have been updated */
+ err = ovs_dp_set_upcall_portids(dp,
+ a[OVS_DP_ATTR_PER_CPU_PIDS]);
+ if (err)
+ return err;
+ }
+
+ if ((dp->user_features & OVS_DP_F_TC_RECIRC_SHARING) &&
+ !(old_features & OVS_DP_F_TC_RECIRC_SHARING))
+ tc_skb_ext_tc_enable();
+ else if (!(dp->user_features & OVS_DP_F_TC_RECIRC_SHARING) &&
+ (old_features & OVS_DP_F_TC_RECIRC_SHARING))
+ tc_skb_ext_tc_disable();
+
+ return 0;
+}
+
+static int ovs_dp_stats_init(struct datapath *dp)
+{
+ dp->stats_percpu = netdev_alloc_pcpu_stats(struct dp_stats_percpu);
+ if (!dp->stats_percpu)
+ return -ENOMEM;
+
+ return 0;
+}
+
+static int ovs_dp_vport_init(struct datapath *dp)
{
- if (a[OVS_DP_ATTR_USER_FEATURES])
- dp->user_features = nla_get_u32(a[OVS_DP_ATTR_USER_FEATURES]);
+ int i;
+
+ dp->ports = kmalloc_array(DP_VPORT_HASH_BUCKETS,
+ sizeof(struct hlist_head),
+ GFP_KERNEL);
+ if (!dp->ports)
+ return -ENOMEM;
+
+ for (i = 0; i < DP_VPORT_HASH_BUCKETS; i++)
+ INIT_HLIST_HEAD(&dp->ports[i]);
+
+ return 0;
}
static int ovs_dp_cmd_new(struct sk_buff *skb, struct genl_info *info)
@@ -1550,7 +1817,7 @@ static int ovs_dp_cmd_new(struct sk_buff *skb, struct genl_info *info)
struct datapath *dp;
struct vport *vport;
struct ovs_net *ovs_net;
- int err, i;
+ int err;
err = -EINVAL;
if (!a[OVS_DP_ATTR_NAME] || !a[OVS_DP_ATTR_UPCALL_PID])
@@ -1563,35 +1830,26 @@ static int ovs_dp_cmd_new(struct sk_buff *skb, struct genl_info *info)
err = -ENOMEM;
dp = kzalloc(sizeof(*dp), GFP_KERNEL);
if (dp == NULL)
- goto err_free_reply;
+ goto err_destroy_reply;
ovs_dp_set_net(dp, sock_net(skb->sk));
/* Allocate table. */
err = ovs_flow_tbl_init(&dp->table);
if (err)
- goto err_free_dp;
+ goto err_destroy_dp;
- dp->stats_percpu = netdev_alloc_pcpu_stats(struct dp_stats_percpu);
- if (!dp->stats_percpu) {
- err = -ENOMEM;
+ err = ovs_dp_stats_init(dp);
+ if (err)
goto err_destroy_table;
- }
- dp->ports = kmalloc_array(DP_VPORT_HASH_BUCKETS,
- sizeof(struct hlist_head),
- GFP_KERNEL);
- if (!dp->ports) {
- err = -ENOMEM;
- goto err_destroy_percpu;
- }
-
- for (i = 0; i < DP_VPORT_HASH_BUCKETS; i++)
- INIT_HLIST_HEAD(&dp->ports[i]);
+ err = ovs_dp_vport_init(dp);
+ if (err)
+ goto err_destroy_stats;
err = ovs_meters_init(dp);
if (err)
- goto err_destroy_ports_array;
+ goto err_destroy_ports;
/* Set up our datapath device. */
parms.name = nla_data(a[OVS_DP_ATTR_NAME]);
@@ -1600,12 +1858,15 @@ static int ovs_dp_cmd_new(struct sk_buff *skb, struct genl_info *info)
parms.dp = dp;
parms.port_no = OVSP_LOCAL;
parms.upcall_portids = a[OVS_DP_ATTR_UPCALL_PID];
-
- ovs_dp_change(dp, a);
+ parms.desired_ifindex = nla_get_s32_default(a[OVS_DP_ATTR_IFINDEX], 0);
/* So far only local changes have been made, now need the lock. */
ovs_lock();
+ err = ovs_dp_change(dp, a);
+ if (err)
+ goto err_unlock_and_destroy_meters;
+
vport = new_vport(&parms);
if (IS_ERR(vport)) {
err = PTR_ERR(vport);
@@ -1621,7 +1882,7 @@ static int ovs_dp_cmd_new(struct sk_buff *skb, struct genl_info *info)
ovs_dp_reset_user_features(skb, info);
}
- goto err_destroy_meters;
+ goto err_destroy_portids;
}
err = ovs_dp_cmd_fill_info(dp, reply, info->snd_portid,
@@ -1636,18 +1897,20 @@ static int ovs_dp_cmd_new(struct sk_buff *skb, struct genl_info *info)
ovs_notify(&dp_datapath_genl_family, reply, info);
return 0;
-err_destroy_meters:
+err_destroy_portids:
+ kfree(rcu_dereference_raw(dp->upcall_portids));
+err_unlock_and_destroy_meters:
ovs_unlock();
ovs_meters_exit(dp);
-err_destroy_ports_array:
+err_destroy_ports:
kfree(dp->ports);
-err_destroy_percpu:
+err_destroy_stats:
free_percpu(dp->stats_percpu);
err_destroy_table:
ovs_flow_tbl_destroy(&dp->table);
-err_free_dp:
+err_destroy_dp:
kfree(dp);
-err_free_reply:
+err_destroy_reply:
kfree_skb(reply);
err:
return err;
@@ -1656,8 +1919,12 @@ err:
/* Called with ovs_mutex. */
static void __dp_destroy(struct datapath *dp)
{
+ struct flow_table *table = &dp->table;
int i;
+ if (dp->user_features & OVS_DP_F_TC_RECIRC_SHARING)
+ tc_skb_ext_tc_disable();
+
for (i = 0; i < DP_VPORT_HASH_BUCKETS; i++) {
struct vport *vport;
struct hlist_node *n;
@@ -1674,7 +1941,14 @@ static void __dp_destroy(struct datapath *dp)
*/
ovs_dp_detach_port(ovs_vport_ovsl(dp, OVSP_LOCAL));
- /* RCU destroy the flow table */
+ /* Flush sw_flow in the tables. RCU cb only releases resource
+ * such as dp, ports and tables. That may avoid some issues
+ * such as RCU usage warning.
+ */
+ table_instance_flow_flush(table, ovsl_dereference(table->ti),
+ ovsl_dereference(table->ufid_ti));
+
+ /* RCU destroy the ports, meters and flow tables. */
call_rcu(&dp->rcu, destroy_dp_rcu);
}
@@ -1689,7 +1963,8 @@ static int ovs_dp_cmd_del(struct sk_buff *skb, struct genl_info *info)
return -ENOMEM;
ovs_lock();
- dp = lookup_datapath(sock_net(skb->sk), info->userhdr, info->attrs);
+ dp = lookup_datapath(sock_net(skb->sk), genl_info_userhdr(info),
+ info->attrs);
err = PTR_ERR(dp);
if (IS_ERR(dp))
goto err_unlock_free;
@@ -1722,12 +1997,15 @@ static int ovs_dp_cmd_set(struct sk_buff *skb, struct genl_info *info)
return -ENOMEM;
ovs_lock();
- dp = lookup_datapath(sock_net(skb->sk), info->userhdr, info->attrs);
+ dp = lookup_datapath(sock_net(skb->sk), genl_info_userhdr(info),
+ info->attrs);
err = PTR_ERR(dp);
if (IS_ERR(dp))
goto err_unlock_free;
- ovs_dp_change(dp, info->attrs);
+ err = ovs_dp_change(dp, info->attrs);
+ if (err)
+ goto err_unlock_free;
err = ovs_dp_cmd_fill_info(dp, reply, info->snd_portid,
info->snd_seq, 0, OVS_DP_CMD_SET);
@@ -1755,7 +2033,8 @@ static int ovs_dp_cmd_get(struct sk_buff *skb, struct genl_info *info)
return -ENOMEM;
ovs_lock();
- dp = lookup_datapath(sock_net(skb->sk), info->userhdr, info->attrs);
+ dp = lookup_datapath(sock_net(skb->sk), genl_info_userhdr(info),
+ info->attrs);
if (IS_ERR(dp)) {
err = PTR_ERR(dp);
goto err_unlock_free;
@@ -1800,28 +2079,31 @@ static const struct nla_policy datapath_policy[OVS_DP_ATTR_MAX + 1] = {
[OVS_DP_ATTR_NAME] = { .type = NLA_NUL_STRING, .len = IFNAMSIZ - 1 },
[OVS_DP_ATTR_UPCALL_PID] = { .type = NLA_U32 },
[OVS_DP_ATTR_USER_FEATURES] = { .type = NLA_U32 },
+ [OVS_DP_ATTR_MASKS_CACHE_SIZE] = NLA_POLICY_RANGE(NLA_U32, 0,
+ PCPU_MIN_UNIT_SIZE / sizeof(struct mask_cache_entry)),
+ [OVS_DP_ATTR_IFINDEX] = NLA_POLICY_MIN(NLA_S32, 0),
};
-static const struct genl_ops dp_datapath_genl_ops[] = {
+static const struct genl_small_ops dp_datapath_genl_ops[] = {
{ .cmd = OVS_DP_CMD_NEW,
+ .validate = GENL_DONT_VALIDATE_STRICT | GENL_DONT_VALIDATE_DUMP,
.flags = GENL_UNS_ADMIN_PERM, /* Requires CAP_NET_ADMIN privilege. */
- .policy = datapath_policy,
.doit = ovs_dp_cmd_new
},
{ .cmd = OVS_DP_CMD_DEL,
+ .validate = GENL_DONT_VALIDATE_STRICT | GENL_DONT_VALIDATE_DUMP,
.flags = GENL_UNS_ADMIN_PERM, /* Requires CAP_NET_ADMIN privilege. */
- .policy = datapath_policy,
.doit = ovs_dp_cmd_del
},
{ .cmd = OVS_DP_CMD_GET,
+ .validate = GENL_DONT_VALIDATE_STRICT | GENL_DONT_VALIDATE_DUMP,
.flags = 0, /* OK for unprivileged users. */
- .policy = datapath_policy,
.doit = ovs_dp_cmd_get,
.dumpit = ovs_dp_cmd_dump
},
{ .cmd = OVS_DP_CMD_SET,
+ .validate = GENL_DONT_VALIDATE_STRICT | GENL_DONT_VALIDATE_DUMP,
.flags = GENL_UNS_ADMIN_PERM, /* Requires CAP_NET_ADMIN privilege. */
- .policy = datapath_policy,
.doit = ovs_dp_cmd_set,
},
};
@@ -1831,10 +2113,12 @@ static struct genl_family dp_datapath_genl_family __ro_after_init = {
.name = OVS_DATAPATH_FAMILY,
.version = OVS_DATAPATH_VERSION,
.maxattr = OVS_DP_ATTR_MAX,
+ .policy = datapath_policy,
.netnsok = true,
.parallel_ops = true,
- .ops = dp_datapath_genl_ops,
- .n_ops = ARRAY_SIZE(dp_datapath_genl_ops),
+ .small_ops = dp_datapath_genl_ops,
+ .n_small_ops = ARRAY_SIZE(dp_datapath_genl_ops),
+ .resv_start_op = OVS_DP_CMD_SET + 1,
.mcgrps = &ovs_dp_datapath_multicast_group,
.n_mcgrps = 1,
.module = THIS_MODULE,
@@ -1843,10 +2127,11 @@ static struct genl_family dp_datapath_genl_family __ro_after_init = {
/* Called with ovs_mutex or RCU read lock. */
static int ovs_vport_cmd_fill_info(struct vport *vport, struct sk_buff *skb,
struct net *net, u32 portid, u32 seq,
- u32 flags, u8 cmd)
+ u32 flags, u8 cmd, gfp_t gfp)
{
struct ovs_header *ovs_header;
struct ovs_vport_stats vport_stats;
+ struct net *net_vport;
int err;
ovs_header = genlmsg_put(skb, portid, seq, &dp_vport_genl_family,
@@ -1863,12 +2148,15 @@ static int ovs_vport_cmd_fill_info(struct vport *vport, struct sk_buff *skb,
nla_put_u32(skb, OVS_VPORT_ATTR_IFINDEX, vport->dev->ifindex))
goto nla_put_failure;
- if (!net_eq(net, dev_net(vport->dev))) {
- int id = peernet2id_alloc(net, dev_net(vport->dev));
+ rcu_read_lock();
+ net_vport = dev_net_rcu(vport->dev);
+ if (!net_eq(net, net_vport)) {
+ int id = peernet2id_alloc(net, net_vport, GFP_ATOMIC);
if (nla_put_s32(skb, OVS_VPORT_ATTR_NETNSID, id))
- goto nla_put_failure;
+ goto nla_put_failure_unlock;
}
+ rcu_read_unlock();
ovs_vport_get_stats(vport, &vport_stats);
if (nla_put_64bit(skb, OVS_VPORT_ATTR_STATS,
@@ -1876,6 +2164,9 @@ static int ovs_vport_cmd_fill_info(struct vport *vport, struct sk_buff *skb,
OVS_VPORT_ATTR_PAD))
goto nla_put_failure;
+ if (ovs_vport_get_upcall_stats(vport, skb))
+ goto nla_put_failure;
+
if (ovs_vport_get_upcall_portids(vport, skb))
goto nla_put_failure;
@@ -1886,6 +2177,8 @@ static int ovs_vport_cmd_fill_info(struct vport *vport, struct sk_buff *skb,
genlmsg_end(skb, ovs_header);
return 0;
+nla_put_failure_unlock:
+ rcu_read_unlock();
nla_put_failure:
err = -EMSGSIZE;
error:
@@ -1905,11 +2198,12 @@ struct sk_buff *ovs_vport_cmd_build_info(struct vport *vport, struct net *net,
struct sk_buff *skb;
int retval;
- skb = nlmsg_new(NLMSG_DEFAULT_SIZE, GFP_ATOMIC);
+ skb = nlmsg_new(NLMSG_DEFAULT_SIZE, GFP_KERNEL);
if (!skb)
return ERR_PTR(-ENOMEM);
- retval = ovs_vport_cmd_fill_info(vport, skb, net, portid, seq, 0, cmd);
+ retval = ovs_vport_cmd_fill_info(vport, skb, net, portid, seq, 0, cmd,
+ GFP_KERNEL);
BUG_ON(retval < 0);
return skb;
@@ -1952,16 +2246,16 @@ static struct vport *lookup_vport(struct net *net,
}
-/* Called with ovs_mutex */
-static void update_headroom(struct datapath *dp)
+static unsigned int ovs_get_max_headroom(struct datapath *dp)
{
- unsigned dev_headroom, max_headroom = 0;
+ unsigned int dev_headroom, max_headroom = 0;
struct net_device *dev;
struct vport *vport;
int i;
for (i = 0; i < DP_VPORT_HASH_BUCKETS; i++) {
- hlist_for_each_entry_rcu(vport, &dp->ports[i], dp_hash_node) {
+ hlist_for_each_entry_rcu(vport, &dp->ports[i], dp_hash_node,
+ lockdep_ovsl_is_held()) {
dev = vport->dev;
dev_headroom = netdev_get_fwd_headroom(dev);
if (dev_headroom > max_headroom)
@@ -1969,31 +2263,45 @@ static void update_headroom(struct datapath *dp)
}
}
- dp->max_headroom = max_headroom;
- for (i = 0; i < DP_VPORT_HASH_BUCKETS; i++)
- hlist_for_each_entry_rcu(vport, &dp->ports[i], dp_hash_node)
- netdev_set_rx_headroom(vport->dev, max_headroom);
+ return max_headroom;
+}
+
+/* Called with ovs_mutex */
+static void ovs_update_headroom(struct datapath *dp, unsigned int new_headroom)
+{
+ struct vport *vport;
+ int i;
+
+ dp->max_headroom = new_headroom;
+ for (i = 0; i < DP_VPORT_HASH_BUCKETS; i++) {
+ hlist_for_each_entry_rcu(vport, &dp->ports[i], dp_hash_node,
+ lockdep_ovsl_is_held())
+ netdev_set_rx_headroom(vport->dev, new_headroom);
+ }
}
static int ovs_vport_cmd_new(struct sk_buff *skb, struct genl_info *info)
{
struct nlattr **a = info->attrs;
- struct ovs_header *ovs_header = info->userhdr;
+ struct ovs_header *ovs_header = genl_info_userhdr(info);
struct vport_parms parms;
struct sk_buff *reply;
struct vport *vport;
struct datapath *dp;
+ unsigned int new_headroom;
u32 port_no;
int err;
if (!a[OVS_VPORT_ATTR_NAME] || !a[OVS_VPORT_ATTR_TYPE] ||
!a[OVS_VPORT_ATTR_UPCALL_PID])
return -EINVAL;
- if (a[OVS_VPORT_ATTR_IFINDEX])
+
+ parms.type = nla_get_u32(a[OVS_VPORT_ATTR_TYPE]);
+
+ if (a[OVS_VPORT_ATTR_IFINDEX] && parms.type != OVS_VPORT_TYPE_INTERNAL)
return -EOPNOTSUPP;
- port_no = a[OVS_VPORT_ATTR_PORT_NO]
- ? nla_get_u32(a[OVS_VPORT_ATTR_PORT_NO]) : 0;
+ port_no = nla_get_u32_default(a[OVS_VPORT_ATTR_PORT_NO], 0);
if (port_no >= DP_MAX_PORTS)
return -EFBIG;
@@ -2026,11 +2334,12 @@ restart:
}
parms.name = nla_data(a[OVS_VPORT_ATTR_NAME]);
- parms.type = nla_get_u32(a[OVS_VPORT_ATTR_TYPE]);
parms.options = a[OVS_VPORT_ATTR_OPTIONS];
parms.dp = dp;
parms.port_no = port_no;
parms.upcall_portids = a[OVS_VPORT_ATTR_UPCALL_PID];
+ parms.desired_ifindex = nla_get_s32_default(a[OVS_VPORT_ATTR_IFINDEX],
+ 0);
vport = new_vport(&parms);
err = PTR_ERR(vport);
@@ -2042,10 +2351,12 @@ restart:
err = ovs_vport_cmd_fill_info(vport, reply, genl_info_net(info),
info->snd_portid, info->snd_seq, 0,
- OVS_VPORT_CMD_NEW);
+ OVS_VPORT_CMD_NEW, GFP_KERNEL);
- if (netdev_get_fwd_headroom(vport->dev) > dp->max_headroom)
- update_headroom(dp);
+ new_headroom = netdev_get_fwd_headroom(vport->dev);
+
+ if (new_headroom > dp->max_headroom)
+ ovs_update_headroom(dp, new_headroom);
else
netdev_set_rx_headroom(vport->dev, dp->max_headroom);
@@ -2073,7 +2384,7 @@ static int ovs_vport_cmd_set(struct sk_buff *skb, struct genl_info *info)
return -ENOMEM;
ovs_lock();
- vport = lookup_vport(sock_net(skb->sk), info->userhdr, a);
+ vport = lookup_vport(sock_net(skb->sk), genl_info_userhdr(info), a);
err = PTR_ERR(vport);
if (IS_ERR(vport))
goto exit_unlock_free;
@@ -2101,7 +2412,7 @@ static int ovs_vport_cmd_set(struct sk_buff *skb, struct genl_info *info)
err = ovs_vport_cmd_fill_info(vport, reply, genl_info_net(info),
info->snd_portid, info->snd_seq, 0,
- OVS_VPORT_CMD_SET);
+ OVS_VPORT_CMD_SET, GFP_KERNEL);
BUG_ON(err < 0);
ovs_unlock();
@@ -2116,11 +2427,12 @@ exit_unlock_free:
static int ovs_vport_cmd_del(struct sk_buff *skb, struct genl_info *info)
{
- bool must_update_headroom = false;
+ bool update_headroom = false;
struct nlattr **a = info->attrs;
struct sk_buff *reply;
struct datapath *dp;
struct vport *vport;
+ unsigned int new_headroom;
int err;
reply = ovs_vport_cmd_alloc_info();
@@ -2128,7 +2440,7 @@ static int ovs_vport_cmd_del(struct sk_buff *skb, struct genl_info *info)
return -ENOMEM;
ovs_lock();
- vport = lookup_vport(sock_net(skb->sk), info->userhdr, a);
+ vport = lookup_vport(sock_net(skb->sk), genl_info_userhdr(info), a);
err = PTR_ERR(vport);
if (IS_ERR(vport))
goto exit_unlock_free;
@@ -2140,18 +2452,23 @@ static int ovs_vport_cmd_del(struct sk_buff *skb, struct genl_info *info)
err = ovs_vport_cmd_fill_info(vport, reply, genl_info_net(info),
info->snd_portid, info->snd_seq, 0,
- OVS_VPORT_CMD_DEL);
+ OVS_VPORT_CMD_DEL, GFP_KERNEL);
BUG_ON(err < 0);
/* the vport deletion may trigger dp headroom update */
dp = vport->dp;
if (netdev_get_fwd_headroom(vport->dev) == dp->max_headroom)
- must_update_headroom = true;
+ update_headroom = true;
+
netdev_reset_rx_headroom(vport->dev);
ovs_dp_detach_port(vport);
- if (must_update_headroom)
- update_headroom(dp);
+ if (update_headroom) {
+ new_headroom = ovs_get_max_headroom(dp);
+
+ if (new_headroom < dp->max_headroom)
+ ovs_update_headroom(dp, new_headroom);
+ }
ovs_unlock();
ovs_notify(&dp_vport_genl_family, reply, info);
@@ -2166,7 +2483,7 @@ exit_unlock_free:
static int ovs_vport_cmd_get(struct sk_buff *skb, struct genl_info *info)
{
struct nlattr **a = info->attrs;
- struct ovs_header *ovs_header = info->userhdr;
+ struct ovs_header *ovs_header = genl_info_userhdr(info);
struct sk_buff *reply;
struct vport *vport;
int err;
@@ -2182,7 +2499,7 @@ static int ovs_vport_cmd_get(struct sk_buff *skb, struct genl_info *info)
goto exit_unlock_free;
err = ovs_vport_cmd_fill_info(vport, reply, genl_info_net(info),
info->snd_portid, info->snd_seq, 0,
- OVS_VPORT_CMD_GET);
+ OVS_VPORT_CMD_GET, GFP_ATOMIC);
BUG_ON(err < 0);
rcu_read_unlock();
@@ -2218,7 +2535,8 @@ static int ovs_vport_cmd_dump(struct sk_buff *skb, struct netlink_callback *cb)
NETLINK_CB(cb->skb).portid,
cb->nlh->nlmsg_seq,
NLM_F_MULTI,
- OVS_VPORT_CMD_GET) < 0)
+ OVS_VPORT_CMD_GET,
+ GFP_ATOMIC) < 0)
goto out;
j++;
@@ -2234,37 +2552,55 @@ out:
return skb->len;
}
+static void ovs_dp_masks_rebalance(struct work_struct *work)
+{
+ struct ovs_net *ovs_net = container_of(work, struct ovs_net,
+ masks_rebalance.work);
+ struct datapath *dp;
+
+ ovs_lock();
+
+ list_for_each_entry(dp, &ovs_net->dps, list_node)
+ ovs_flow_masks_rebalance(&dp->table);
+
+ ovs_unlock();
+
+ schedule_delayed_work(&ovs_net->masks_rebalance,
+ msecs_to_jiffies(DP_MASKS_REBALANCE_INTERVAL));
+}
+
static const struct nla_policy vport_policy[OVS_VPORT_ATTR_MAX + 1] = {
[OVS_VPORT_ATTR_NAME] = { .type = NLA_NUL_STRING, .len = IFNAMSIZ - 1 },
[OVS_VPORT_ATTR_STATS] = { .len = sizeof(struct ovs_vport_stats) },
[OVS_VPORT_ATTR_PORT_NO] = { .type = NLA_U32 },
[OVS_VPORT_ATTR_TYPE] = { .type = NLA_U32 },
- [OVS_VPORT_ATTR_UPCALL_PID] = { .type = NLA_U32 },
+ [OVS_VPORT_ATTR_UPCALL_PID] = { .type = NLA_UNSPEC },
[OVS_VPORT_ATTR_OPTIONS] = { .type = NLA_NESTED },
- [OVS_VPORT_ATTR_IFINDEX] = { .type = NLA_U32 },
+ [OVS_VPORT_ATTR_IFINDEX] = NLA_POLICY_MIN(NLA_S32, 0),
[OVS_VPORT_ATTR_NETNSID] = { .type = NLA_S32 },
+ [OVS_VPORT_ATTR_UPCALL_STATS] = { .type = NLA_NESTED },
};
-static const struct genl_ops dp_vport_genl_ops[] = {
+static const struct genl_small_ops dp_vport_genl_ops[] = {
{ .cmd = OVS_VPORT_CMD_NEW,
+ .validate = GENL_DONT_VALIDATE_STRICT | GENL_DONT_VALIDATE_DUMP,
.flags = GENL_UNS_ADMIN_PERM, /* Requires CAP_NET_ADMIN privilege. */
- .policy = vport_policy,
.doit = ovs_vport_cmd_new
},
{ .cmd = OVS_VPORT_CMD_DEL,
+ .validate = GENL_DONT_VALIDATE_STRICT | GENL_DONT_VALIDATE_DUMP,
.flags = GENL_UNS_ADMIN_PERM, /* Requires CAP_NET_ADMIN privilege. */
- .policy = vport_policy,
.doit = ovs_vport_cmd_del
},
{ .cmd = OVS_VPORT_CMD_GET,
+ .validate = GENL_DONT_VALIDATE_STRICT | GENL_DONT_VALIDATE_DUMP,
.flags = 0, /* OK for unprivileged users. */
- .policy = vport_policy,
.doit = ovs_vport_cmd_get,
.dumpit = ovs_vport_cmd_dump
},
{ .cmd = OVS_VPORT_CMD_SET,
+ .validate = GENL_DONT_VALIDATE_STRICT | GENL_DONT_VALIDATE_DUMP,
.flags = GENL_UNS_ADMIN_PERM, /* Requires CAP_NET_ADMIN privilege. */
- .policy = vport_policy,
.doit = ovs_vport_cmd_set,
},
};
@@ -2274,10 +2610,12 @@ struct genl_family dp_vport_genl_family __ro_after_init = {
.name = OVS_VPORT_FAMILY,
.version = OVS_VPORT_VERSION,
.maxattr = OVS_VPORT_ATTR_MAX,
+ .policy = vport_policy,
.netnsok = true,
.parallel_ops = true,
- .ops = dp_vport_genl_ops,
- .n_ops = ARRAY_SIZE(dp_vport_genl_ops),
+ .small_ops = dp_vport_genl_ops,
+ .n_small_ops = ARRAY_SIZE(dp_vport_genl_ops),
+ .resv_start_op = OVS_VPORT_CMD_SET + 1,
.mcgrps = &ovs_dp_vport_multicast_group,
.n_mcgrps = 1,
.module = THIS_MODULE,
@@ -2324,10 +2662,19 @@ error:
static int __net_init ovs_init_net(struct net *net)
{
struct ovs_net *ovs_net = net_generic(net, ovs_net_id);
+ int err;
INIT_LIST_HEAD(&ovs_net->dps);
INIT_WORK(&ovs_net->dp_notify_work, ovs_dp_notify_wq);
- return ovs_ct_init(net);
+ INIT_DELAYED_WORK(&ovs_net->masks_rebalance, ovs_dp_masks_rebalance);
+
+ err = ovs_ct_init(net);
+ if (err)
+ return err;
+
+ schedule_delayed_work(&ovs_net->masks_rebalance,
+ msecs_to_jiffies(DP_MASKS_REBALANCE_INTERVAL));
+ return 0;
}
static void __net_exit list_vports_from_net(struct net *net, struct net *dnet,
@@ -2361,8 +2708,10 @@ static void __net_exit ovs_exit_net(struct net *dnet)
struct net *net;
LIST_HEAD(head);
- ovs_ct_exit(dnet);
ovs_lock();
+
+ ovs_ct_exit(dnet);
+
list_for_each_entry_safe(dp, dp_next, &ovs_net->dps, list_node)
__dp_destroy(dp);
@@ -2379,6 +2728,7 @@ static void __net_exit ovs_exit_net(struct net *dnet)
ovs_unlock();
+ cancel_delayed_work_sync(&ovs_net->masks_rebalance);
cancel_work_sync(&ovs_net->dp_notify_work);
}
@@ -2389,21 +2739,55 @@ static struct pernet_operations ovs_net_ops = {
.size = sizeof(struct ovs_net),
};
+static const char * const ovs_drop_reasons[] = {
+#define S(x) [(x) & ~SKB_DROP_REASON_SUBSYS_MASK] = (#x),
+ OVS_DROP_REASONS(S)
+#undef S
+};
+
+static struct drop_reason_list drop_reason_list_ovs = {
+ .reasons = ovs_drop_reasons,
+ .n_reasons = ARRAY_SIZE(ovs_drop_reasons),
+};
+
+static int __init ovs_alloc_percpu_storage(void)
+{
+ unsigned int cpu;
+
+ ovs_pcpu_storage = alloc_percpu(*ovs_pcpu_storage);
+ if (!ovs_pcpu_storage)
+ return -ENOMEM;
+
+ for_each_possible_cpu(cpu) {
+ struct ovs_pcpu_storage *ovs_pcpu;
+
+ ovs_pcpu = per_cpu_ptr(ovs_pcpu_storage, cpu);
+ local_lock_init(&ovs_pcpu->bh_lock);
+ }
+ return 0;
+}
+
+static void ovs_free_percpu_storage(void)
+{
+ free_percpu(ovs_pcpu_storage);
+}
+
static int __init dp_init(void)
{
int err;
- BUILD_BUG_ON(sizeof(struct ovs_skb_cb) > FIELD_SIZEOF(struct sk_buff, cb));
+ BUILD_BUG_ON(sizeof(struct ovs_skb_cb) >
+ sizeof_field(struct sk_buff, cb));
pr_info("Open vSwitch switching datapath\n");
- err = action_fifos_init();
+ err = ovs_alloc_percpu_storage();
if (err)
goto error;
err = ovs_internal_dev_rtnl_link_register();
if (err)
- goto error_action_fifos_exit;
+ goto error;
err = ovs_flow_init();
if (err)
@@ -2429,6 +2813,9 @@ static int __init dp_init(void)
if (err < 0)
goto error_unreg_netdev;
+ drop_reasons_register_subsys(SKB_DROP_REASON_SUBSYS_OPENVSWITCH,
+ &drop_reason_list_ovs);
+
return 0;
error_unreg_netdev:
@@ -2443,9 +2830,8 @@ error_flow_exit:
ovs_flow_exit();
error_unreg_rtnl_link:
ovs_internal_dev_rtnl_link_unregister();
-error_action_fifos_exit:
- action_fifos_exit();
error:
+ ovs_free_percpu_storage();
return err;
}
@@ -2455,11 +2841,12 @@ static void dp_cleanup(void)
ovs_netdev_exit();
unregister_netdevice_notifier(&ovs_dp_device_notifier);
unregister_pernet_device(&ovs_net_ops);
+ drop_reasons_unregister_subsys(SKB_DROP_REASON_SUBSYS_OPENVSWITCH);
rcu_barrier();
ovs_vport_exit();
ovs_flow_exit();
ovs_internal_dev_rtnl_link_unregister();
- action_fifos_exit();
+ ovs_free_percpu_storage();
}
module_init(dp_init);
generated by cgit (git 2.34.1) at 2025-12-22 09:24:00 +0000