X-Git-Url: http://git.onelab.eu/?a=blobdiff_plain;f=datapath%2Fvport.c;h=8ef96f73e1fc6ea7e77cb1881d5617e0be93ab85;hb=9df30404d9f806daa6f9851412b2a295f3540e75;hp=38c71476e4f81b33b2aeca191a5269b3d541adb4;hpb=780e620781c5237ab6c26d5edbc7edb5b4953c70;p=sliver-openvswitch.git diff --git a/datapath/vport.c b/datapath/vport.c index 38c71476e..8ef96f73e 100644 --- a/datapath/vport.c +++ b/datapath/vport.c @@ -1,94 +1,55 @@ /* - * Copyright (c) 2010 Nicira Networks. + * Copyright (c) 2010, 2011 Nicira Networks. * Distributed under the terms of the GNU GPL version 2. * * Significant portions of this file may be copied from parts of the Linux * kernel, by Linus Torvalds and others. */ +#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt + #include #include #include +#include #include #include #include #include +#include #include #include +#include #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 struct vport_ops *base_vport_ops_list[] = { +static const struct vport_ops *base_vport_ops_list[] = { &netdev_vport_ops, &internal_vport_ops, &patch_vport_ops, &gre_vport_ops, +#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,26) + &capwap_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. */ static struct hlist_head *dev_table; #define VPORT_HASH_BUCKETS 1024 -/* Both RTNL lock and vport_mutex need to be held when updating dev_table. - * - * If you use vport_locate and then perform some operations, you need to hold - * one of these locks if you don't want the vport to be deleted out from under - * you. - * - * If you get a reference to a vport through a dp_port, it is protected - * by RCU and you need to hold rcu_read_lock instead when reading. - * - * If multiple locks are taken, the hierarchy is: - * 1. RTNL - * 2. DP - * 3. vport - */ -static DEFINE_MUTEX(vport_mutex); - -/** - * vport_lock - acquire vport lock - * - * Acquire global vport lock. See above comment about locking requirements - * and specific function definitions. May sleep. - */ -void -vport_lock(void) -{ - mutex_lock(&vport_mutex); -} - -/** - * vport_unlock - release vport lock - * - * Release lock acquired with vport_lock. - */ -void -vport_unlock(void) -{ - mutex_unlock(&vport_mutex); -} - -#define ASSERT_VPORT() do { \ - if (unlikely(!mutex_is_locked(&vport_mutex))) { \ - printk(KERN_ERR "openvswitch: vport lock not held at %s (%d)\n", \ - __FILE__, __LINE__); \ - dump_stack(); \ - } \ -} while(0) - /** * vport_init - initialize vport subsystem * * Called at module load time to initialize the vport subsystem and any * compiled in vport types. */ -int -vport_init(void) +int vport_init(void) { int err; int i; @@ -108,7 +69,7 @@ vport_init(void) } for (i = 0; i < ARRAY_SIZE(base_vport_ops_list); i++) { - struct vport_ops *new_ops = base_vport_ops_list[i]; + const struct vport_ops *new_ops = base_vport_ops_list[i]; if (new_ops->init) err = new_ops->init(); @@ -131,40 +92,16 @@ error: return err; } -static void -vport_del_all(void) -{ - int i; - - rtnl_lock(); - vport_lock(); - - for (i = 0; i < VPORT_HASH_BUCKETS; i++) { - struct hlist_head *bucket = &dev_table[i]; - struct vport *vport; - struct hlist_node *node, *next; - - hlist_for_each_entry_safe(vport, node, next, bucket, hash_node) - vport_del(vport); - } - - vport_unlock(); - rtnl_unlock(); -} - /** * vport_exit - shutdown vport subsystem * * Called at module exit time to shutdown the vport subsystem and any * initialized vport types. */ -void -vport_exit(void) +void vport_exit(void) { int i; - vport_del_all(); - for (i = 0; i < n_vport_types; i++) { if (vport_ops_list[i]->exit) vport_ops_list[i]->exit(); @@ -174,441 +111,7 @@ vport_exit(void) kfree(dev_table); } -static int -do_vport_add(struct odp_vport_add *vport_config) -{ - struct vport *vport; - int err = 0; - - vport_config->port_type[VPORT_TYPE_SIZE - 1] = '\0'; - vport_config->devname[IFNAMSIZ - 1] = '\0'; - - rtnl_lock(); - - vport = vport_locate(vport_config->devname); - if (vport) { - err = -EEXIST; - goto out; - } - - vport_lock(); - vport = vport_add(vport_config->devname, vport_config->port_type, - vport_config->config); - vport_unlock(); - - if (IS_ERR(vport)) - err = PTR_ERR(vport); - -out: - rtnl_unlock(); - return err; -} - -/** - * vport_user_add - add vport device (for userspace callers) - * - * @uvport_config: New port configuration. - * - * Creates a new vport with the specified configuration (which is dependent - * on device type). This function is for userspace callers and assumes no - * locks are held. - */ -int -vport_user_add(const struct odp_vport_add __user *uvport_config) -{ - struct odp_vport_add vport_config; - - if (copy_from_user(&vport_config, uvport_config, sizeof(struct odp_vport_add))) - return -EFAULT; - - return do_vport_add(&vport_config); -} - -#ifdef CONFIG_COMPAT -int -compat_vport_user_add(struct compat_odp_vport_add *ucompat) -{ - struct compat_odp_vport_add compat; - struct odp_vport_add vport_config; - - if (copy_from_user(&compat, ucompat, sizeof(struct compat_odp_vport_add))) - return -EFAULT; - - memcpy(vport_config.port_type, compat.port_type, VPORT_TYPE_SIZE); - memcpy(vport_config.devname, compat.devname, IFNAMSIZ); - vport_config.config = compat_ptr(compat.config); - - return do_vport_add(&vport_config); -} -#endif - -static int -do_vport_mod(struct odp_vport_mod *vport_config) -{ - struct vport *vport; - int err; - - vport_config->devname[IFNAMSIZ - 1] = '\0'; - - rtnl_lock(); - - vport = vport_locate(vport_config->devname); - if (!vport) { - err = -ENODEV; - goto out; - } - - vport_lock(); - err = vport_mod(vport, vport_config->config); - vport_unlock(); - -out: - rtnl_unlock(); - return err; -} - -/** - * vport_user_mod - modify existing vport device (for userspace callers) - * - * @uvport_config: New configuration for vport - * - * Modifies an existing device with the specified configuration (which is - * dependent on device type). This function is for userspace callers and - * assumes no locks are held. - */ -int -vport_user_mod(const struct odp_vport_mod __user *uvport_config) -{ - struct odp_vport_mod vport_config; - - if (copy_from_user(&vport_config, uvport_config, sizeof(struct odp_vport_mod))) - return -EFAULT; - - return do_vport_mod(&vport_config); -} - -#ifdef CONFIG_COMPAT -int -compat_vport_user_mod(struct compat_odp_vport_mod *ucompat) -{ - struct compat_odp_vport_mod compat; - struct odp_vport_mod vport_config; - - if (copy_from_user(&compat, ucompat, sizeof(struct compat_odp_vport_mod))) - return -EFAULT; - - memcpy(vport_config.devname, compat.devname, IFNAMSIZ); - vport_config.config = compat_ptr(compat.config); - - return do_vport_mod(&vport_config); -} -#endif - -/** - * vport_user_del - delete existing vport device (for userspace callers) - * - * @udevname: Name of device to delete - * - * Deletes the specified device. Detaches the device from a datapath first - * if it is attached. Deleting the device will fail if it does not exist or it - * is the datapath local port. It is also possible to fail for less obvious - * reasons, such as lack of memory. This function is for userspace callers and - * assumes no locks are held. - */ -int -vport_user_del(const char __user *udevname) -{ - char devname[IFNAMSIZ]; - struct vport *vport; - struct dp_port *dp_port; - int err = 0; - int retval; - - retval = strncpy_from_user(devname, udevname, IFNAMSIZ); - if (retval < 0) - return -EFAULT; - else if (retval >= IFNAMSIZ) - return -ENAMETOOLONG; - - rtnl_lock(); - - vport = vport_locate(devname); - if (!vport) { - err = -ENODEV; - goto out; - } - - dp_port = vport_get_dp_port(vport); - if (dp_port) { - struct datapath *dp = dp_port->dp; - - mutex_lock(&dp->mutex); - - if (!strcmp(dp_name(dp), devname)) { - err = -EINVAL; - goto dp_port_out; - } - - err = dp_detach_port(dp_port, 0); - -dp_port_out: - mutex_unlock(&dp->mutex); - - if (err) - goto out; - } - - vport_lock(); - err = vport_del(vport); - vport_unlock(); - -out: - rtnl_unlock(); - return err; -} - -/** - * vport_user_stats_get - retrieve device stats (for userspace callers) - * - * @ustats_req: Stats request parameters. - * - * Retrieves transmit, receive, and error stats for the given device. This - * function is for userspace callers and assumes no locks are held. - */ -int -vport_user_stats_get(struct odp_vport_stats_req __user *ustats_req) -{ - struct odp_vport_stats_req stats_req; - struct vport *vport; - int err; - - if (copy_from_user(&stats_req, ustats_req, sizeof(struct odp_vport_stats_req))) - return -EFAULT; - - stats_req.devname[IFNAMSIZ - 1] = '\0'; - - vport_lock(); - - vport = vport_locate(stats_req.devname); - if (!vport) { - err = -ENODEV; - goto out; - } - - err = vport_get_stats(vport, &stats_req.stats); - -out: - vport_unlock(); - - if (!err) - if (copy_to_user(ustats_req, &stats_req, sizeof(struct odp_vport_stats_req))) - err = -EFAULT; - - return err; -} - -/** - * vport_user_stats_set - sets offset device stats (for userspace callers) - * - * @ustats_req: Stats set parameters. - * - * Provides a set of transmit, receive, and error stats to be added as an - * offset to the collect data when stats are retreived. Some devices may not - * support setting the stats, in which case the result will always be - * -EOPNOTSUPP. This function is for userspace callers and assumes no locks - * are held. - */ -int -vport_user_stats_set(struct odp_vport_stats_req __user *ustats_req) -{ - struct odp_vport_stats_req stats_req; - struct vport *vport; - int err; - - if (copy_from_user(&stats_req, ustats_req, sizeof(struct odp_vport_stats_req))) - return -EFAULT; - - stats_req.devname[IFNAMSIZ - 1] = '\0'; - - rtnl_lock(); - vport_lock(); - - vport = vport_locate(stats_req.devname); - if (!vport) { - err = -ENODEV; - goto out; - } - - err = vport_set_stats(vport, &stats_req.stats); - -out: - vport_unlock(); - rtnl_unlock(); - return err; -} - - -/** - * vport_user_ether_get - retrieve device Ethernet address (for userspace callers) - * - * @uvport_ether: Ethernet address request parameters. - * - * Retrieves the Ethernet address of the given device. This function is for - * userspace callers and assumes no locks are held. - */ -int -vport_user_ether_get(struct odp_vport_ether __user *uvport_ether) -{ - struct odp_vport_ether vport_ether; - struct vport *vport; - int err = 0; - - if (copy_from_user(&vport_ether, uvport_ether, sizeof(struct odp_vport_ether))) - return -EFAULT; - - vport_ether.devname[IFNAMSIZ - 1] = '\0'; - - vport_lock(); - - vport = vport_locate(vport_ether.devname); - if (!vport) { - err = -ENODEV; - goto out; - } - - rcu_read_lock(); - memcpy(vport_ether.ether_addr, vport_get_addr(vport), ETH_ALEN); - rcu_read_unlock(); - -out: - vport_unlock(); - - if (!err) - if (copy_to_user(uvport_ether, &vport_ether, sizeof(struct odp_vport_ether))) - err = -EFAULT; - - return err; -} - -/** - * vport_user_ether_set - set device Ethernet address (for userspace callers) - * - * @uvport_ether: Ethernet address request parameters. - * - * 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. This function is for userspace callers and assumes no locks - * are held. - */ -int -vport_user_ether_set(struct odp_vport_ether __user *uvport_ether) -{ - struct odp_vport_ether vport_ether; - struct vport *vport; - int err; - - if (copy_from_user(&vport_ether, uvport_ether, sizeof(struct odp_vport_ether))) - return -EFAULT; - - vport_ether.devname[IFNAMSIZ - 1] = '\0'; - - rtnl_lock(); - vport_lock(); - - vport = vport_locate(vport_ether.devname); - if (!vport) { - err = -ENODEV; - goto out; - } - - err = vport_set_addr(vport, vport_ether.ether_addr); - -out: - vport_unlock(); - rtnl_unlock(); - return err; -} - -/** - * vport_user_mtu_get - retrieve device MTU (for userspace callers) - * - * @uvport_mtu: MTU request parameters. - * - * Retrieves the MTU of the given device. This function is for userspace - * callers and assumes no locks are held. - */ -int -vport_user_mtu_get(struct odp_vport_mtu __user *uvport_mtu) -{ - struct odp_vport_mtu vport_mtu; - struct vport *vport; - int err = 0; - - if (copy_from_user(&vport_mtu, uvport_mtu, sizeof(struct odp_vport_mtu))) - return -EFAULT; - - vport_mtu.devname[IFNAMSIZ - 1] = '\0'; - - vport_lock(); - - vport = vport_locate(vport_mtu.devname); - if (!vport) { - err = -ENODEV; - goto out; - } - - vport_mtu.mtu = vport_get_mtu(vport); - -out: - vport_unlock(); - - if (!err) - if (copy_to_user(uvport_mtu, &vport_mtu, sizeof(struct odp_vport_mtu))) - err = -EFAULT; - - return err; -} - -/** - * vport_user_mtu_set - set device MTU (for userspace callers) - * - * @uvport_mtu: MTU request parameters. - * - * Sets the MTU of the given device. Some devices may not support setting the - * MTU, in which case the result will always be -EOPNOTSUPP. This function is - * for userspace callers and assumes no locks are held. - */ -int -vport_user_mtu_set(struct odp_vport_mtu __user *uvport_mtu) -{ - struct odp_vport_mtu vport_mtu; - struct vport *vport; - int err; - - if (copy_from_user(&vport_mtu, uvport_mtu, sizeof(struct odp_vport_mtu))) - return -EFAULT; - - vport_mtu.devname[IFNAMSIZ - 1] = '\0'; - - rtnl_lock(); - vport_lock(); - - vport = vport_locate(vport_mtu.devname); - if (!vport) { - err = -ENODEV; - goto out; - } - - err = vport_set_mtu(vport, vport_mtu.mtu); - -out: - vport_unlock(); - rtnl_unlock(); - return err; -} - -static struct hlist_head * -hash_bucket(const char *name) +static struct hlist_head *hash_bucket(const char *name) { unsigned int hash = full_name_hash(name, strlen(name)); return &dev_table[hash & (VPORT_HASH_BUCKETS - 1)]; @@ -619,46 +122,33 @@ hash_bucket(const char *name) * * @name: name of port to find * - * Either RTNL or vport lock must be acquired before calling this function - * and held while using the found port. See the locking comments at the - * top of the file. + * Must be called with RTNL or RCU read lock. */ -struct vport * -vport_locate(const char *name) +struct vport *vport_locate(const char *name) { struct hlist_head *bucket = hash_bucket(name); struct vport *vport; struct hlist_node *node; - if (unlikely(!mutex_is_locked(&vport_mutex) && !rtnl_is_locked())) { - printk(KERN_ERR "openvswitch: neither RTNL nor vport lock held in vport_locate\n"); - dump_stack(); - } - - rcu_read_lock(); - - hlist_for_each_entry(vport, node, bucket, hash_node) + hlist_for_each_entry_rcu(vport, node, bucket, hash_node) if (!strcmp(name, vport_get_name(vport))) - goto out; - - vport = NULL; + return vport; -out: - rcu_read_unlock(); - return vport; + return NULL; } -static void -register_vport(struct vport *vport) +static void release_vport(struct kobject *kobj) { - hlist_add_head(&vport->hash_node, hash_bucket(vport_get_name(vport))); + struct vport *p = container_of(kobj, struct vport, kobj); + kfree(p); } -static void -unregister_vport(struct vport *vport) -{ - hlist_del(&vport->hash_node); -} +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 @@ -671,8 +161,7 @@ unregister_vport(struct vport *vport) * vport_priv(). vports that are no longer needed should be released with * vport_free(). */ -struct vport * -vport_alloc(int priv_size, const struct vport_ops *ops) +struct vport *vport_alloc(int priv_size, const struct vport_ops *ops, const struct vport_parms *parms) { struct vport *vport; size_t alloc_size; @@ -687,8 +176,16 @@ vport_alloc(int priv_size, const struct vport_ops *ops) if (!vport) return ERR_PTR(-ENOMEM); + vport->dp = parms->dp; + vport->port_no = parms->port_no; + atomic_set(&vport->sflow_pool, 0); vport->ops = ops; + /* 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); + if (vport->ops->flags & VPORT_F_GEN_STATS) { vport->percpu_stats = alloc_percpu(struct vport_percpu_stats); if (!vport->percpu_stats) @@ -706,46 +203,44 @@ vport_alloc(int priv_size, const struct vport_ops *ops) * @vport: vport to free * * Frees a vport allocated with 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 vport_free(struct vport *vport) { if (vport->ops->flags & VPORT_F_GEN_STATS) free_percpu(vport->percpu_stats); - kfree(vport); + kobject_put(&vport->kobj); } /** * vport_add - add vport device (for kernel callers) * - * @name: Name of new device. - * @type: Type of new device (to be matched against types in registered vport - * ops). - * @config: Device type specific configuration. Userspace pointer. + * @parms: Information about new vport. * - * Creates a new vport with the specified configuration (which is dependent - * on device type). Both RTNL and vport locks must be held. + * Creates a new vport with the specified configuration (which is dependent on + * device type) and attaches it to a datapath. RTNL lock must be held. */ -struct vport * -vport_add(const char *name, const char *type, const void __user *config) +struct vport *vport_add(const struct vport_parms *parms) { struct vport *vport; int err = 0; int i; ASSERT_RTNL(); - ASSERT_VPORT(); for (i = 0; i < n_vport_types; i++) { - if (!strcmp(vport_ops_list[i]->type, type)) { - vport = vport_ops_list[i]->create(name, config); + if (vport_ops_list[i]->type == parms->type) { + vport = vport_ops_list[i]->create(parms); if (IS_ERR(vport)) { err = PTR_ERR(vport); goto out; } - register_vport(vport); + hlist_add_head_rcu(&vport->hash_node, + hash_bucket(vport_get_name(vport))); return vport; } } @@ -757,111 +252,40 @@ out: } /** - * vport_mod - modify existing vport device (for kernel callers) + * vport_set_options - modify existing vport device (for kernel callers) * * @vport: vport to modify. - * @config: Device type specific configuration. Userspace pointer. + * @port: New configuration. * * Modifies an existing device with the specified configuration (which is - * dependent on device type). Both RTNL and vport locks must be held. + * dependent on device type). RTNL lock must be held. */ -int -vport_mod(struct vport *vport, const void __user *config) +int vport_set_options(struct vport *vport, struct nlattr *options) { ASSERT_RTNL(); - ASSERT_VPORT(); - if (vport->ops->modify) - return vport->ops->modify(vport, config); - else + if (!vport->ops->set_options) return -EOPNOTSUPP; + return vport->ops->set_options(vport, options); } /** - * vport_del - delete existing vport device (for kernel callers) + * vport_del - delete existing vport device * * @vport: vport to delete. * - * Deletes the specified device. The device must not be currently attached to - * a datapath. It is possible to fail for reasons such as lack of memory. - * Both RTNL and vport locks must be held. + * 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. */ -int -vport_del(struct vport *vport) +int vport_del(struct vport *vport) { ASSERT_RTNL(); - ASSERT_VPORT(); - BUG_ON(vport_get_dp_port(vport)); - unregister_vport(vport); + hlist_del_rcu(&vport->hash_node); return vport->ops->destroy(vport); } -/** - * vport_attach - attach a vport to a datapath - * - * @vport: vport to attach. - * @dp_port: Datapath port to attach the vport to. - * - * Attaches a vport to a specific datapath so that packets may be exchanged. - * Both ports must be currently unattached. @dp_port must be successfully - * attached to a vport before it is connected to a datapath and must not be - * modified while connected. RTNL lock and the appropriate DP mutex must be held. - */ -int -vport_attach(struct vport *vport, struct dp_port *dp_port) -{ - ASSERT_RTNL(); - - if (dp_port->vport) - return -EBUSY; - - if (vport_get_dp_port(vport)) - return -EBUSY; - - if (vport->ops->attach) { - int err; - - err = vport->ops->attach(vport); - if (err) - return err; - } - - dp_port->vport = vport; - rcu_assign_pointer(vport->dp_port, dp_port); - - return 0; -} - -/** - * vport_detach - detach a vport from a datapath - * - * @vport: vport to detach. - * - * Detaches a vport from a datapath. May fail for a variety of reasons, - * including lack of memory. RTNL lock and the appropriate DP mutex must be held. - */ -int -vport_detach(struct vport *vport) -{ - struct dp_port *dp_port; - - ASSERT_RTNL(); - - dp_port = vport_get_dp_port(vport); - if (!dp_port) - return -EINVAL; - - dp_port->vport = NULL; - rcu_assign_pointer(vport->dp_port, NULL); - - if (vport->ops->detach) - return vport->ops->detach(vport); - else - return 0; -} - /** * vport_set_mtu - set device MTU (for kernel callers) * @@ -872,8 +296,7 @@ vport_detach(struct vport *vport) * MTU, in which case the result will always be -EOPNOTSUPP. RTNL lock must * be held. */ -int -vport_set_mtu(struct vport *vport, int mtu) +int vport_set_mtu(struct vport *vport, int mtu) { ASSERT_RTNL(); @@ -885,12 +308,8 @@ vport_set_mtu(struct vport *vport, int mtu) ret = vport->ops->set_mtu(vport, mtu); - if (!ret && !is_internal_vport(vport)) { - struct dp_port *dp_port = vport_get_dp_port(vport); - - if (dp_port) - set_internal_devs_mtu(dp_port->dp); - } + if (!ret && !is_internal_vport(vport)) + set_internal_devs_mtu(vport->dp); return ret; } else @@ -907,8 +326,7 @@ vport_set_mtu(struct vport *vport, int mtu) * 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) +int vport_set_addr(struct vport *vport, const unsigned char *addr) { ASSERT_RTNL(); @@ -922,7 +340,7 @@ vport_set_addr(struct vport *vport, const unsigned char *addr) } /** - * vport_set_stats - sets offset device stats (for kernel callers) + * vport_set_stats - sets offset device stats * * @vport: vport on which to set stats * @stats: stats to set @@ -930,22 +348,21 @@ vport_set_addr(struct vport *vport, const unsigned char *addr) * Provides a set of transmit, receive, and error stats to be added as an * offset to the collect data when stats are retreived. Some devices may not * support setting the stats, in which case the result will always be - * -EOPNOTSUPP. RTNL lock must be held. + * -EOPNOTSUPP. + * + * Must be called with RTNL lock. */ -int -vport_set_stats(struct vport *vport, struct odp_vport_stats *stats) +int vport_set_stats(struct vport *vport, struct rtnl_link_stats64 *stats) { ASSERT_RTNL(); if (vport->ops->flags & VPORT_F_GEN_STATS) { spin_lock_bh(&vport->stats_lock); - memcpy(&vport->offset_stats, stats, sizeof(struct odp_vport_stats)); + vport->offset_stats = *stats; spin_unlock_bh(&vport->stats_lock); return 0; - } else if (vport->ops->set_stats) - return vport->ops->set_stats(vport, stats); - else + } else return -EOPNOTSUPP; } @@ -957,8 +374,7 @@ vport_set_stats(struct vport *vport, struct odp_vport_stats *stats) * 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) +const char *vport_get_name(const struct vport *vport) { return vport->ops->get_name(vport); } @@ -968,11 +384,9 @@ vport_get_name(const struct vport *vport) * * @vport: vport from which to retrieve the type. * - * Retrieves the type of the given device. Either RTNL lock or rcu_read_lock - * must be held for the entire duration that the type is in use. + * Retrieves the type of the given device. */ -const char * -vport_get_type(const struct vport *vport) +enum odp_vport_type vport_get_type(const struct vport *vport) { return vport->ops->type; } @@ -986,27 +400,11 @@ vport_get_type(const struct vport *vport) * 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) +const unsigned char *vport_get_addr(const struct vport *vport) { return vport->ops->get_addr(vport); } -/** - * vport_get_dp_port - retrieve attached datapath port - * - * @vport: vport from which to retrieve the datapath port. - * - * Retrieves the attached datapath port or null if not attached. Either RTNL - * lock or rcu_read_lock must be held for the entire duration that the datapath - * port is being accessed. - */ -struct dp_port * -vport_get_dp_port(const struct vport *vport) -{ - return rcu_dereference(vport->dp_port); -} - /** * vport_get_kobj - retrieve associated kobj * @@ -1015,8 +413,7 @@ vport_get_dp_port(const struct vport *vport) * 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) +struct kobject *vport_get_kobj(const struct vport *vport) { if (vport->ops->get_kobj) return vport->ops->get_kobj(vport); @@ -1024,90 +421,100 @@ vport_get_kobj(const struct vport *vport) return NULL; } +static int vport_call_get_stats(struct vport *vport, struct rtnl_link_stats64 *stats) +{ + int err; + + rcu_read_lock(); + err = vport->ops->get_stats(vport, stats); + rcu_read_unlock(); + + return err; +} + /** - * vport_get_stats - retrieve device stats (for kernel callers) + * 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. */ -int -vport_get_stats(struct vport *vport, struct odp_vport_stats *stats) +int vport_get_stats(struct vport *vport, struct rtnl_link_stats64 *stats) { - struct odp_vport_stats dev_stats; - struct odp_vport_stats *dev_statsp = NULL; - int err; + int i; - if (vport->ops->get_stats) { - if (vport->ops->flags & VPORT_F_GEN_STATS) - dev_statsp = &dev_stats; - else - dev_statsp = stats; + if (!(vport->ops->flags & VPORT_F_GEN_STATS)) + return vport_call_get_stats(vport, stats); - rcu_read_lock(); - err = vport->ops->get_stats(vport, dev_statsp); - rcu_read_unlock(); + /* We potentially have 3 sources of stats that need to be + * combined: those we have collected (split into err_stats and + * percpu_stats), offset_stats from set_stats(), and device + * error stats from get_stats() (for errors that happen + * downstream and therefore aren't reported through our + * vport_record_error() function). */ - if (err) - goto out; - } + spin_lock_bh(&vport->stats_lock); - if (vport->ops->flags & VPORT_F_GEN_STATS) { - int i; + *stats = vport->offset_stats; - /* We potentially have 3 sources of stats that need to be - * combined: those we have collected (split into err_stats and - * percpu_stats), offset_stats from set_stats(), and device - * error stats from get_stats() (for errors that happen - * downstream and therefore aren't reported through our - * vport_record_error() function). */ + stats->rx_errors += vport->err_stats.rx_errors; + stats->tx_errors += vport->err_stats.tx_errors; + stats->tx_dropped += vport->err_stats.tx_dropped; + stats->rx_dropped += vport->err_stats.rx_dropped; - spin_lock_bh(&vport->stats_lock); + spin_unlock_bh(&vport->stats_lock); - memcpy(stats, &vport->offset_stats, sizeof(struct odp_vport_stats)); + if (vport->ops->get_stats) { + struct rtnl_link_stats64 dev_stats; + int err; - stats->rx_errors += vport->err_stats.rx_errors - + vport->err_stats.rx_frame_err - + vport->err_stats.rx_over_err - + vport->err_stats.rx_crc_err; - stats->tx_errors += vport->err_stats.tx_errors; - stats->tx_dropped += vport->err_stats.tx_dropped; - stats->rx_dropped += vport->err_stats.rx_dropped; - stats->rx_over_err += vport->err_stats.rx_over_err; - stats->rx_crc_err += vport->err_stats.rx_crc_err; - stats->rx_frame_err += vport->err_stats.rx_frame_err; - stats->collisions += vport->err_stats.collisions; + err = vport_call_get_stats(vport, &dev_stats); + if (err) + return err; - spin_unlock_bh(&vport->stats_lock); + stats->rx_errors += dev_stats.rx_errors; + stats->tx_errors += dev_stats.tx_errors; + stats->rx_dropped += dev_stats.rx_dropped; + stats->tx_dropped += dev_stats.tx_dropped; + stats->multicast += dev_stats.multicast; + stats->collisions += dev_stats.collisions; + stats->rx_length_errors += dev_stats.rx_length_errors; + stats->rx_over_errors += dev_stats.rx_over_errors; + stats->rx_crc_errors += dev_stats.rx_crc_errors; + stats->rx_frame_errors += dev_stats.rx_frame_errors; + stats->rx_fifo_errors += dev_stats.rx_fifo_errors; + stats->rx_missed_errors += dev_stats.rx_missed_errors; + stats->tx_aborted_errors += dev_stats.tx_aborted_errors; + stats->tx_carrier_errors += dev_stats.tx_carrier_errors; + stats->tx_fifo_errors += dev_stats.tx_fifo_errors; + stats->tx_heartbeat_errors += dev_stats.tx_heartbeat_errors; + stats->tx_window_errors += dev_stats.tx_window_errors; + stats->rx_compressed += dev_stats.rx_compressed; + stats->tx_compressed += dev_stats.tx_compressed; + } - if (dev_statsp) { - stats->rx_errors += dev_statsp->rx_errors; - stats->tx_errors += dev_statsp->tx_errors; - stats->rx_dropped += dev_statsp->rx_dropped; - stats->tx_dropped += dev_statsp->tx_dropped; - stats->rx_over_err += dev_statsp->rx_over_err; - stats->rx_crc_err += dev_statsp->rx_crc_err; - stats->rx_frame_err += dev_statsp->rx_frame_err; - stats->collisions += dev_statsp->collisions; - } + for_each_possible_cpu(i) { + const struct vport_percpu_stats *percpu_stats; + struct vport_percpu_stats local_stats; + unsigned seqcount; - for_each_possible_cpu(i) { - const struct vport_percpu_stats *percpu_stats; + percpu_stats = per_cpu_ptr(vport->percpu_stats, i); - percpu_stats = per_cpu_ptr(vport->percpu_stats, i); - stats->rx_bytes += percpu_stats->rx_bytes; - stats->rx_packets += percpu_stats->rx_packets; - stats->tx_bytes += percpu_stats->tx_bytes; - stats->tx_packets += percpu_stats->tx_packets; - } + do { + seqcount = read_seqcount_begin(&percpu_stats->seqlock); + local_stats = *percpu_stats; + } while (read_seqcount_retry(&percpu_stats->seqlock, seqcount)); - err = 0; - } else - err = -EOPNOTSUPP; + stats->rx_bytes += local_stats.rx_bytes; + stats->rx_packets += local_stats.rx_packets; + stats->tx_bytes += local_stats.tx_bytes; + stats->tx_packets += local_stats.tx_packets; + } -out: - return err; + return 0; } /** @@ -1115,11 +522,11 @@ out: * * @vport: vport from which to retrieve the flags * - * Retrieves the flags of the given device. Either RTNL lock or rcu_read_lock - * must be held. + * 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) +unsigned vport_get_flags(const struct vport *vport) { return vport->ops->get_dev_flags(vport); } @@ -1129,11 +536,11 @@ vport_get_flags(const struct vport *vport) * * @vport: vport on which to check status. * - * Checks whether the given device is running. Either RTNL lock or - * rcu_read_lock must be held. + * 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) +int vport_is_running(const struct vport *vport) { return vport->ops->is_running(vport); } @@ -1143,11 +550,11 @@ vport_is_running(const struct vport *vport) * * @vport: vport from which to check status * - * Retrieves the RFC2863 operstate of the given device. Either RTNL lock or - * rcu_read_lock must be held. + * 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) +unsigned char vport_get_operstate(const struct vport *vport) { return vport->ops->get_operstate(vport); } @@ -1157,26 +564,18 @@ vport_get_operstate(const struct vport *vport) * * @vport: vport from which to retrieve index * - * Retrieves the system interface index of the given device. Not all devices - * will have system indexes, in which case the index of the datapath local - * port is returned. Returns a negative index on error. Either RTNL lock or - * rcu_read_lock must be held. + * 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) +int vport_get_ifindex(const struct vport *vport) { - const struct dp_port *dp_port; - if (vport->ops->get_ifindex) return vport->ops->get_ifindex(vport); - - /* If we don't actually have an ifindex, use the local port's. - * Userspace doesn't check it anyways. */ - dp_port = vport_get_dp_port(vport); - if (!dp_port) - return -EAGAIN; - - return vport_get_ifindex(dp_port->dp->ports[ODPP_LOCAL]->vport); + else + return 0; } /** @@ -1187,11 +586,11 @@ vport_get_ifindex(const struct vport *vport) * Retrieves the system link index of the given device. The link is the index * of the interface on which the packet will actually be sent. In most cases * this is the same as the ifindex but may be different for tunnel devices. - * Returns a negative index on error. Either RTNL lock or rcu_read_lock must - * be held. + * Returns a negative index on error. + * + * Must be called with RTNL lock or rcu_read_lock. */ -int -vport_get_iflink(const struct vport *vport) +int vport_get_iflink(const struct vport *vport) { if (vport->ops->get_iflink) return vport->ops->get_iflink(vport); @@ -1202,19 +601,57 @@ vport_get_iflink(const struct vport *vport) } /** - * vport_get_mtu - retrieve device MTU (for kernel callers) + * vport_get_mtu - retrieve device MTU * * @vport: vport from which to retrieve MTU * - * Retrieves the MTU of the given device. Either RTNL lock or rcu_read_lock - * must be held. + * 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) +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 + * + * @vport: vport from which to retrieve the options. + * @skb: sk_buff where options should be appended. + * + * Retrieves the configuration of the given device, appending an + * %ODP_VPORT_ATTR_OPTIONS attribute that in turn contains nested + * vport-specific attributes to @skb. + * + * Returns 0 if successful, -EMSGSIZE if @skb has insufficient room, or another + * 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. + */ +int vport_get_options(const struct vport *vport, struct sk_buff *skb) +{ + struct nlattr *nla; + + nla = nla_nest_start(skb, ODP_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; + } + } + + nla_nest_end(skb, nla); + return 0; +} + /** * vport_receive - pass up received packet to the datapath for processing * @@ -1225,34 +662,39 @@ vport_get_mtu(const struct vport *vport) * 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 vport_receive(struct vport *vport, struct sk_buff *skb) { - struct dp_port *dp_port = vport_get_dp_port(vport); - - if (!dp_port) { - vport_record_error(vport, VPORT_E_RX_DROPPED); - kfree_skb(skb); - - return; - } - if (vport->ops->flags & VPORT_F_GEN_STATS) { struct vport_percpu_stats *stats; local_bh_disable(); - stats = per_cpu_ptr(vport->percpu_stats, smp_processor_id()); + + write_seqcount_begin(&stats->seqlock); stats->rx_packets++; stats->rx_bytes += skb->len; + write_seqcount_end(&stats->seqlock); local_bh_enable(); } + if (!(vport->ops->flags & VPORT_F_FLOW)) + OVS_CB(skb)->flow = NULL; + if (!(vport->ops->flags & VPORT_F_TUN_ID)) OVS_CB(skb)->tun_id = 0; - dp_process_received_packet(dp_port, skb); + dp_process_received_packet(vport, skb); +} + +static inline unsigned packet_length(const struct sk_buff *skb) +{ + unsigned length = skb->len - ETH_HLEN; + + if (skb->protocol == htons(ETH_P_8021Q)) + length -= VLAN_HLEN; + + return length; } /** @@ -1264,26 +706,41 @@ vport_receive(struct vport *vport, struct sk_buff *skb) * Sends the given packet and returns the length of data sent. Either RTNL * lock or rcu_read_lock must be held. */ -int -vport_send(struct vport *vport, struct sk_buff *skb) +int vport_send(struct vport *vport, struct sk_buff *skb) { + int mtu; int sent; + mtu = vport_get_mtu(vport); + if (mtu && unlikely(packet_length(skb) > mtu && !skb_is_gso(skb))) { + if (net_ratelimit()) + pr_warn("%s: dropped over-mtu packet: %d > %d\n", + dp_name(vport->dp), packet_length(skb), mtu); + goto error; + } + sent = vport->ops->send(vport, skb); if (vport->ops->flags & VPORT_F_GEN_STATS && sent > 0) { struct vport_percpu_stats *stats; local_bh_disable(); - stats = per_cpu_ptr(vport->percpu_stats, smp_processor_id()); + + write_seqcount_begin(&stats->seqlock); stats->tx_packets++; stats->tx_bytes += sent; + write_seqcount_end(&stats->seqlock); local_bh_enable(); } return sent; + +error: + kfree_skb(skb); + vport_record_error(vport, VPORT_E_TX_DROPPED); + return 0; } /** @@ -1295,8 +752,7 @@ vport_send(struct vport *vport, struct sk_buff *skb) * If using the vport generic stats layer indicate that an error of the given * type has occured. */ -void -vport_record_error(struct vport *vport, enum vport_err_type err_type) +void vport_record_error(struct vport *vport, enum vport_err_type err_type) { if (vport->ops->flags & VPORT_F_GEN_STATS) { @@ -1311,18 +767,6 @@ vport_record_error(struct vport *vport, enum vport_err_type err_type) vport->err_stats.rx_errors++; break; - case VPORT_E_RX_FRAME: - vport->err_stats.rx_frame_err++; - break; - - case VPORT_E_RX_OVER: - vport->err_stats.rx_over_err++; - break; - - case VPORT_E_RX_CRC: - vport->err_stats.rx_crc_err++; - break; - case VPORT_E_TX_DROPPED: vport->err_stats.tx_dropped++; break; @@ -1330,10 +774,6 @@ vport_record_error(struct vport *vport, enum vport_err_type err_type) case VPORT_E_TX_ERROR: vport->err_stats.tx_errors++; break; - - case VPORT_E_COLLISION: - vport->err_stats.collisions++; - break; }; spin_unlock_bh(&vport->stats_lock);