X-Git-Url: http://git.onelab.eu/?a=blobdiff_plain;f=datapath%2Fvport.c;h=dcb7616531ff51d0a5d23f8f118b7eb0487be512;hb=ec9f40dce11c7e81bc41d42e3bbfaaf8287165ce;hp=ad5a10e7adefdb34b6ad31b089e465d78ac412ab;hpb=e9141eec249f97e547c6e47205b61ee98d9b52e9;p=sliver-openvswitch.git diff --git a/datapath/vport.c b/datapath/vport.c index ad5a10e7a..dcb761653 100644 --- a/datapath/vport.c +++ b/datapath/vport.c @@ -1,15 +1,25 @@ /* - * Copyright (c) 2010, 2011 Nicira Networks. - * Distributed under the terms of the GNU GPL version 2. + * Copyright (c) 2007-2012 Nicira, Inc. * - * Significant portions of this file may be copied from parts of the Linux - * kernel, by Linus Torvalds and others. + * 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 */ -#include #include #include #include +#include #include #include #include @@ -18,138 +28,82 @@ #include #include #include +#include +#include "datapath.h" #include "vport.h" #include "vport-internal_dev.h" /* List of statically compiled vport implementations. Don't forget to also * add yours to the list at the bottom of vport.h. */ -static const struct vport_ops *base_vport_ops_list[] = { - &netdev_vport_ops, - &internal_vport_ops, - &patch_vport_ops, - &gre_vport_ops, +static const struct vport_ops *vport_ops_list[] = { + &ovs_netdev_vport_ops, + &ovs_internal_vport_ops, + &ovs_gre_vport_ops, + &ovs_gre64_vport_ops, #if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,26) - &capwap_vport_ops, + &ovs_vxlan_vport_ops, + &ovs_lisp_vport_ops, #endif }; -static const struct vport_ops **vport_ops_list; -static int n_vport_types; - -/* Protected by RCU read lock for reading, RTNL lock for writing. */ +/* Protected by RCU read lock for reading, ovs_mutex for writing. */ static struct hlist_head *dev_table; #define VPORT_HASH_BUCKETS 1024 /** - * vport_init - initialize vport subsystem + * ovs_vport_init - initialize vport subsystem * - * Called at module load time to initialize the vport subsystem and any - * compiled in vport types. + * Called at module load time to initialize the vport subsystem. */ -int vport_init(void) +int ovs_vport_init(void) { - int err; - int i; - dev_table = kzalloc(VPORT_HASH_BUCKETS * sizeof(struct hlist_head), GFP_KERNEL); - if (!dev_table) { - err = -ENOMEM; - goto error; - } - - vport_ops_list = kmalloc(ARRAY_SIZE(base_vport_ops_list) * - sizeof(struct vport_ops *), GFP_KERNEL); - if (!vport_ops_list) { - err = -ENOMEM; - goto error_dev_table; - } - - for (i = 0; i < ARRAY_SIZE(base_vport_ops_list); i++) { - const struct vport_ops *new_ops = base_vport_ops_list[i]; - - if (new_ops->init) - err = new_ops->init(); - else - err = 0; - - if (!err) - vport_ops_list[n_vport_types++] = new_ops; - else if (new_ops->flags & VPORT_F_REQUIRED) { - vport_exit(); - goto error; - } - } + if (!dev_table) + return -ENOMEM; return 0; - -error_dev_table: - kfree(dev_table); -error: - return err; } /** - * vport_exit - shutdown vport subsystem + * ovs_vport_exit - shutdown vport subsystem * - * Called at module exit time to shutdown the vport subsystem and any - * initialized vport types. + * Called at module exit time to shutdown the vport subsystem. */ -void vport_exit(void) +void ovs_vport_exit(void) { - int i; - - for (i = 0; i < n_vport_types; i++) { - if (vport_ops_list[i]->exit) - vport_ops_list[i]->exit(); - } - - kfree(vport_ops_list); kfree(dev_table); } -static struct hlist_head *hash_bucket(const char *name) +static struct hlist_head *hash_bucket(struct net *net, const char *name) { - unsigned int hash = full_name_hash(name, strlen(name)); + unsigned int hash = jhash(name, strlen(name), (unsigned long) net); return &dev_table[hash & (VPORT_HASH_BUCKETS - 1)]; } /** - * vport_locate - find a port that has already been created + * ovs_vport_locate - find a port that has already been created * * @name: name of port to find * - * Must be called with RTNL or RCU read lock. + * Must be called with ovs or RCU read lock. */ -struct vport *vport_locate(const char *name) +struct vport *ovs_vport_locate(struct net *net, const char *name) { - struct hlist_head *bucket = hash_bucket(name); + struct hlist_head *bucket = hash_bucket(net, name); struct vport *vport; - struct hlist_node *node; - hlist_for_each_entry_rcu(vport, node, bucket, hash_node) - if (!strcmp(name, vport_get_name(vport))) + hlist_for_each_entry_rcu(vport, bucket, hash_node) + if (!strcmp(name, vport->ops->get_name(vport)) && + net_eq(ovs_dp_get_net(vport->dp), net)) return vport; return NULL; } -static void release_vport(struct kobject *kobj) -{ - struct vport *p = container_of(kobj, struct vport, kobj); - kfree(p); -} - -static struct kobj_type brport_ktype = { -#ifdef CONFIG_SYSFS - .sysfs_ops = &brport_sysfs_ops, -#endif - .release = release_vport -}; - /** - * vport_alloc - allocate and initialize new vport + * ovs_vport_alloc - allocate and initialize new vport * * @priv_size: Size of private data area to allocate. * @ops: vport device ops @@ -157,9 +111,10 @@ static struct kobj_type brport_ktype = { * Allocate and initialize a new vport defined by @ops. The vport will contain * a private data area of size @priv_size that can be accessed using * vport_priv(). vports that are no longer needed should be released with - * vport_free(). + * ovs_vport_free(). */ -struct vport *vport_alloc(int priv_size, const struct vport_ops *ops, const struct vport_parms *parms) +struct vport *ovs_vport_alloc(int priv_size, const struct vport_ops *ops, + const struct vport_parms *parms) { struct vport *vport; size_t alloc_size; @@ -176,17 +131,15 @@ struct vport *vport_alloc(int priv_size, const struct vport_ops *ops, const stru vport->dp = parms->dp; vport->port_no = parms->port_no; - vport->upcall_pid = parms->upcall_pid; + vport->upcall_portid = parms->upcall_portid; vport->ops = ops; + INIT_HLIST_NODE(&vport->dp_hash_node); - /* Initialize kobject for bridge. This will be added as - * /sys/class/net//brport later, if sysfs is enabled. */ - vport->kobj.kset = NULL; - kobject_init(&vport->kobj, &brport_ktype); - - vport->percpu_stats = alloc_percpu(struct vport_percpu_stats); - if (!vport->percpu_stats) + vport->percpu_stats = alloc_percpu(struct pcpu_tstats); + if (!vport->percpu_stats) { + kfree(vport); return ERR_PTR(-ENOMEM); + } spin_lock_init(&vport->stats_lock); @@ -194,48 +147,48 @@ struct vport *vport_alloc(int priv_size, const struct vport_ops *ops, const stru } /** - * vport_free - uninitialize and free vport + * ovs_vport_free - uninitialize and free vport * * @vport: vport to free * - * Frees a vport allocated with vport_alloc() when it is no longer needed. + * Frees a vport allocated with ovs_vport_alloc() when it is no longer needed. * * The caller must ensure that an RCU grace period has passed since the last * time @vport was in a datapath. */ -void vport_free(struct vport *vport) +void ovs_vport_free(struct vport *vport) { free_percpu(vport->percpu_stats); - - kobject_put(&vport->kobj); + kfree(vport); } /** - * vport_add - add vport device (for kernel callers) + * ovs_vport_add - add vport device (for kernel callers) * * @parms: Information about new vport. * * Creates a new vport with the specified configuration (which is dependent on - * device type). RTNL lock must be held. + * device type). ovs_mutex must be held. */ -struct vport *vport_add(const struct vport_parms *parms) +struct vport *ovs_vport_add(const struct vport_parms *parms) { struct vport *vport; int err = 0; int i; - ASSERT_RTNL(); - - for (i = 0; i < n_vport_types; i++) { + for (i = 0; i < ARRAY_SIZE(vport_ops_list); i++) { if (vport_ops_list[i]->type == parms->type) { + struct hlist_head *bucket; + vport = vport_ops_list[i]->create(parms); if (IS_ERR(vport)) { err = PTR_ERR(vport); goto out; } - hlist_add_head_rcu(&vport->hash_node, - hash_bucket(vport_get_name(vport))); + bucket = hash_bucket(ovs_dp_get_net(vport->dp), + vport->ops->get_name(vport)); + hlist_add_head_rcu(&vport->hash_node, bucket); return vport; } } @@ -247,65 +200,39 @@ out: } /** - * vport_set_options - modify existing vport device (for kernel callers) + * ovs_vport_set_options - modify existing vport device (for kernel callers) * * @vport: vport to modify. * @port: New configuration. * * Modifies an existing device with the specified configuration (which is - * dependent on device type). RTNL lock must be held. + * dependent on device type). ovs_mutex must be held. */ -int vport_set_options(struct vport *vport, struct nlattr *options) +int ovs_vport_set_options(struct vport *vport, struct nlattr *options) { - ASSERT_RTNL(); - if (!vport->ops->set_options) return -EOPNOTSUPP; return vport->ops->set_options(vport, options); } /** - * vport_del - delete existing vport device + * ovs_vport_del - delete existing vport device * * @vport: vport to delete. * * Detaches @vport from its datapath and destroys it. It is possible to fail - * for reasons such as lack of memory. RTNL lock must be held. + * for reasons such as lack of memory. ovs_mutex must be held. */ -void vport_del(struct vport *vport) +void ovs_vport_del(struct vport *vport) { - ASSERT_RTNL(); + ASSERT_OVSL(); hlist_del_rcu(&vport->hash_node); - vport->ops->destroy(vport); } /** - * vport_set_addr - set device Ethernet address (for kernel callers) - * - * @vport: vport on which to set Ethernet address. - * @addr: New address. - * - * Sets the Ethernet address of the given device. Some devices may not support - * setting the Ethernet address, in which case the result will always be - * -EOPNOTSUPP. RTNL lock must be held. - */ -int vport_set_addr(struct vport *vport, const unsigned char *addr) -{ - ASSERT_RTNL(); - - if (!is_valid_ether_addr(addr)) - return -EADDRNOTAVAIL; - - if (vport->ops->set_addr) - return vport->ops->set_addr(vport, addr); - else - return -EOPNOTSUPP; -} - -/** - * vport_set_stats - sets offset device stats + * ovs_vport_set_stats - sets offset device stats * * @vport: vport on which to set stats * @stats: stats to set @@ -315,83 +242,26 @@ int vport_set_addr(struct vport *vport, const unsigned char *addr) * support setting the stats, in which case the result will always be * -EOPNOTSUPP. * - * Must be called with RTNL lock. + * Must be called with ovs_mutex. */ -void vport_set_stats(struct vport *vport, struct ovs_vport_stats *stats) +void ovs_vport_set_stats(struct vport *vport, struct ovs_vport_stats *stats) { - ASSERT_RTNL(); - spin_lock_bh(&vport->stats_lock); vport->offset_stats = *stats; spin_unlock_bh(&vport->stats_lock); } /** - * vport_get_name - retrieve device name - * - * @vport: vport from which to retrieve the name. - * - * Retrieves the name of the given device. Either RTNL lock or rcu_read_lock - * must be held for the entire duration that the name is in use. - */ -const char *vport_get_name(const struct vport *vport) -{ - return vport->ops->get_name(vport); -} - -/** - * vport_get_type - retrieve device type - * - * @vport: vport from which to retrieve the type. - * - * Retrieves the type of the given device. - */ -enum ovs_vport_type vport_get_type(const struct vport *vport) -{ - return vport->ops->type; -} - -/** - * vport_get_addr - retrieve device Ethernet address (for kernel callers) - * - * @vport: vport from which to retrieve the Ethernet address. - * - * Retrieves the Ethernet address of the given device. Either RTNL lock or - * rcu_read_lock must be held for the entire duration that the Ethernet address - * is in use. - */ -const unsigned char *vport_get_addr(const struct vport *vport) -{ - return vport->ops->get_addr(vport); -} - -/** - * vport_get_kobj - retrieve associated kobj - * - * @vport: vport from which to retrieve the associated kobj - * - * Retrieves the associated kobj or null if no kobj. The returned kobj is - * valid for as long as the vport exists. - */ -struct kobject *vport_get_kobj(const struct vport *vport) -{ - if (vport->ops->get_kobj) - return vport->ops->get_kobj(vport); - else - return NULL; -} - -/** - * vport_get_stats - retrieve device stats + * ovs_vport_get_stats - retrieve device stats * * @vport: vport from which to retrieve the stats * @stats: location to store stats * * Retrieves transmit, receive, and error stats for the given device. * - * Must be called with RTNL lock or rcu_read_lock. + * Must be called with ovs_mutex or rcu_read_lock. */ -void vport_get_stats(struct vport *vport, struct ovs_vport_stats *stats) +void ovs_vport_get_stats(struct vport *vport, struct ovs_vport_stats *stats) { int i; @@ -418,16 +288,16 @@ void vport_get_stats(struct vport *vport, struct ovs_vport_stats *stats) spin_unlock_bh(&vport->stats_lock); for_each_possible_cpu(i) { - const struct vport_percpu_stats *percpu_stats; - struct vport_percpu_stats local_stats; - unsigned seqcount; + const struct pcpu_tstats *percpu_stats; + struct pcpu_tstats local_stats; + unsigned int start; percpu_stats = per_cpu_ptr(vport->percpu_stats, i); do { - seqcount = read_seqcount_begin(&percpu_stats->seqlock); + start = u64_stats_fetch_begin_bh(&percpu_stats->syncp); local_stats = *percpu_stats; - } while (read_seqcount_retry(&percpu_stats->seqlock, seqcount)); + } while (u64_stats_fetch_retry_bh(&percpu_stats->syncp, start)); stats->rx_bytes += local_stats.rx_bytes; stats->rx_packets += local_stats.rx_packets; @@ -437,84 +307,7 @@ void vport_get_stats(struct vport *vport, struct ovs_vport_stats *stats) } /** - * vport_get_flags - retrieve device flags - * - * @vport: vport from which to retrieve the flags - * - * Retrieves the flags of the given device. - * - * Must be called with RTNL lock or rcu_read_lock. - */ -unsigned vport_get_flags(const struct vport *vport) -{ - return vport->ops->get_dev_flags(vport); -} - -/** - * vport_get_flags - check whether device is running - * - * @vport: vport on which to check status. - * - * Checks whether the given device is running. - * - * Must be called with RTNL lock or rcu_read_lock. - */ -int vport_is_running(const struct vport *vport) -{ - return vport->ops->is_running(vport); -} - -/** - * vport_get_flags - retrieve device operating state - * - * @vport: vport from which to check status - * - * Retrieves the RFC2863 operstate of the given device. - * - * Must be called with RTNL lock or rcu_read_lock. - */ -unsigned char vport_get_operstate(const struct vport *vport) -{ - return vport->ops->get_operstate(vport); -} - -/** - * vport_get_ifindex - retrieve device system interface index - * - * @vport: vport from which to retrieve index - * - * Retrieves the system interface index of the given device or 0 if - * the device does not have one (in the case of virtual ports). - * Returns a negative index on error. - * - * Must be called with RTNL lock or rcu_read_lock. - */ -int vport_get_ifindex(const struct vport *vport) -{ - if (vport->ops->get_ifindex) - return vport->ops->get_ifindex(vport); - else - return 0; -} - -/** - * vport_get_mtu - retrieve device MTU - * - * @vport: vport from which to retrieve MTU - * - * Retrieves the MTU of the given device. Returns 0 if @vport does not have an - * MTU (as e.g. some tunnels do not). Either RTNL lock or rcu_read_lock must - * be held. - */ -int vport_get_mtu(const struct vport *vport) -{ - if (!vport->ops->get_mtu) - return 0; - return vport->ops->get_mtu(vport); -} - -/** - * vport_get_options - retrieve device options + * ovs_vport_get_options - retrieve device options * * @vport: vport from which to retrieve the options. * @skb: sk_buff where options should be appended. @@ -527,22 +320,24 @@ int vport_get_mtu(const struct vport *vport) * negative error code if a real error occurred. If an error occurs, @skb is * left unmodified. * - * Must be called with RTNL lock or rcu_read_lock. + * Must be called with ovs_mutex or rcu_read_lock. */ -int vport_get_options(const struct vport *vport, struct sk_buff *skb) +int ovs_vport_get_options(const struct vport *vport, struct sk_buff *skb) { struct nlattr *nla; + int err; + + if (!vport->ops->get_options) + return 0; nla = nla_nest_start(skb, OVS_VPORT_ATTR_OPTIONS); if (!nla) return -EMSGSIZE; - if (vport->ops->get_options) { - int err = vport->ops->get_options(vport, skb); - if (err) { - nla_nest_cancel(skb, nla); - return err; - } + err = vport->ops->get_options(vport, skb); + if (err) { + nla_nest_cancel(skb, nla); + return err; } nla_nest_end(skb, nla); @@ -550,7 +345,7 @@ int vport_get_options(const struct vport *vport, struct sk_buff *skb) } /** - * vport_receive - pass up received packet to the datapath for processing + * ovs_vport_receive - pass up received packet to the datapath for processing * * @vport: vport that received the packet * @skb: skb that was received @@ -559,60 +354,62 @@ int vport_get_options(const struct vport *vport, struct sk_buff *skb) * skb->data should point to the Ethernet header. The caller must have already * called compute_ip_summed() to initialize the checksumming fields. */ -void vport_receive(struct vport *vport, struct sk_buff *skb) +void ovs_vport_receive(struct vport *vport, struct sk_buff *skb, + struct ovs_key_ipv4_tunnel *tun_key) { - struct vport_percpu_stats *stats; + struct pcpu_tstats *stats; - stats = per_cpu_ptr(vport->percpu_stats, smp_processor_id()); - - write_seqcount_begin(&stats->seqlock); + stats = this_cpu_ptr(vport->percpu_stats); + u64_stats_update_begin(&stats->syncp); stats->rx_packets++; stats->rx_bytes += skb->len; - write_seqcount_end(&stats->seqlock); - - if (!(vport->ops->flags & VPORT_F_FLOW)) - OVS_CB(skb)->flow = NULL; + u64_stats_update_end(&stats->syncp); - if (!(vport->ops->flags & VPORT_F_TUN_ID)) - OVS_CB(skb)->tun_id = 0; - - dp_process_received_packet(vport, skb); + OVS_CB(skb)->tun_key = tun_key; + ovs_dp_process_received_packet(vport, skb); } /** - * vport_send - send a packet on a device + * ovs_vport_send - send a packet on a device * * @vport: vport on which to send the packet * @skb: skb to send * - * Sends the given packet and returns the length of data sent. Either RTNL + * Sends the given packet and returns the length of data sent. Either ovs * lock or rcu_read_lock must be held. */ -int vport_send(struct vport *vport, struct sk_buff *skb) +int ovs_vport_send(struct vport *vport, struct sk_buff *skb) { - struct vport_percpu_stats *stats; int sent = vport->ops->send(vport, skb); - stats = per_cpu_ptr(vport->percpu_stats, smp_processor_id()); + if (likely(sent > 0)) { + struct pcpu_tstats *stats; + + stats = this_cpu_ptr(vport->percpu_stats); - write_seqcount_begin(&stats->seqlock); - stats->tx_packets++; - stats->tx_bytes += sent; - write_seqcount_end(&stats->seqlock); + u64_stats_update_begin(&stats->syncp); + stats->tx_packets++; + stats->tx_bytes += sent; + u64_stats_update_end(&stats->syncp); + } else if (sent < 0) { + ovs_vport_record_error(vport, VPORT_E_TX_ERROR); + kfree_skb(skb); + } else + ovs_vport_record_error(vport, VPORT_E_TX_DROPPED); return sent; } /** - * vport_record_error - indicate device error to generic stats layer + * ovs_vport_record_error - indicate device error to generic stats layer * * @vport: vport that encountered the error * @err_type: one of enum vport_err_type types to indicate the error type * * If using the vport generic stats layer indicate that an error of the given - * type has occured. + * type has occurred. */ -void vport_record_error(struct vport *vport, enum vport_err_type err_type) +void ovs_vport_record_error(struct vport *vport, enum vport_err_type err_type) { spin_lock(&vport->stats_lock); @@ -632,7 +429,22 @@ void vport_record_error(struct vport *vport, enum vport_err_type err_type) case VPORT_E_TX_ERROR: vport->err_stats.tx_errors++; break; - }; + } spin_unlock(&vport->stats_lock); } + +static void free_vport_rcu(struct rcu_head *rcu) +{ + struct vport *vport = container_of(rcu, struct vport, rcu); + + ovs_vport_free(vport); +} + +void ovs_vport_deferred_free(struct vport *vport) +{ + if (!vport) + return; + + call_rcu(&vport->rcu, free_vport_rcu); +}