diff options
author | Arvid Brodin <arvid.brodin@alten.se> | 2014-07-04 23:37:27 +0200 |
---|---|---|
committer | David S. Miller <davem@davemloft.net> | 2014-07-08 11:35:31 -0700 |
commit | 51f3c605318b056ac5deb9079bbef2a976558827 (patch) | |
tree | 892ae91481ed018ce84f70a9b1c49ee0972d9406 /net/hsr/hsr_device.c | |
parent | e9aae56ea43ef4a32527b9d86c1f6b5eebfbd223 (diff) |
net/hsr: Move slave init to hsr_slave.c.
Also try to prevent some possible slave dereference race conditions. This is
finalized in the next patch, which abandons the slave array in favour of
a list_head list and list RCU.
Signed-off-by: Arvid Brodin <arvid.brodin@alten.se>
Signed-off-by: David S. Miller <davem@davemloft.net>
Diffstat (limited to 'net/hsr/hsr_device.c')
-rw-r--r-- | net/hsr/hsr_device.c | 192 |
1 files changed, 55 insertions, 137 deletions
diff --git a/net/hsr/hsr_device.c b/net/hsr/hsr_device.c index 8936b3c2bb84..1f8869cb70ae 100644 --- a/net/hsr/hsr_device.c +++ b/net/hsr/hsr_device.c @@ -15,7 +15,6 @@ #include <linux/netdevice.h> #include <linux/skbuff.h> #include <linux/etherdevice.h> -#include <linux/if_arp.h> #include <linux/rtnetlink.h> #include <linux/pkt_sched.h> #include "hsr_device.h" @@ -108,23 +107,27 @@ void hsr_check_carrier_and_operstate(struct hsr_priv *hsr) hsr_check_announce(hsr->dev, old_operstate); } - int hsr_get_max_mtu(struct hsr_priv *hsr) { - int mtu_max; - - if (hsr->slave[0] && hsr->slave[1]) - mtu_max = min(hsr->slave[0]->mtu, hsr->slave[1]->mtu); - else if (hsr->slave[0]) - mtu_max = hsr->slave[0]->mtu; - else if (hsr->slave[1]) - mtu_max = hsr->slave[1]->mtu; - else - mtu_max = HSR_HLEN; - + unsigned int mtu_max; + struct net_device *slave; + + mtu_max = ETH_DATA_LEN; + rcu_read_lock(); + slave = hsr->slave[0]; + if (slave) + mtu_max = min(slave->mtu, mtu_max); + slave = hsr->slave[1]; + if (slave) + mtu_max = min(slave->mtu, mtu_max); + rcu_read_unlock(); + + if (mtu_max < HSR_HLEN) + return 0; return mtu_max - HSR_HLEN; } + static int hsr_dev_change_mtu(struct net_device *dev, int new_mtu) { struct hsr_priv *hsr; @@ -145,18 +148,20 @@ static int hsr_dev_change_mtu(struct net_device *dev, int new_mtu) static int hsr_dev_open(struct net_device *dev) { struct hsr_priv *hsr; + struct net_device *slave; int i; char *slave_name; hsr = netdev_priv(dev); for (i = 0; i < HSR_MAX_SLAVE; i++) { - if (hsr->slave[i]) - slave_name = hsr->slave[i]->name; + slave = hsr->slave[i]; + if (slave) + slave_name = slave->name; else slave_name = "null"; - if (!is_slave_up(hsr->slave[i])) + if (!is_slave_up(slave)) netdev_warn(dev, "Slave %c (%s) is not up; please bring it up to get a working HSR network\n", 'A' + i, slave_name); } @@ -223,6 +228,8 @@ static int slave_xmit(struct sk_buff *skb, struct hsr_priv *hsr, hsr_ethhdr = (struct hsr_ethhdr *) skb->data; skb->dev = hsr->slave[dev_idx]; + if (unlikely(!skb->dev)) + return NET_XMIT_DROP; hsr_addr_subst_dest(hsr, &hsr_ethhdr->ethhdr, dev_idx); @@ -252,14 +259,8 @@ static int hsr_dev_xmit(struct sk_buff *skb, struct net_device *dev) } skb2 = pskb_copy(skb, GFP_ATOMIC); - - res1 = NET_XMIT_DROP; - if (likely(hsr->slave[HSR_DEV_SLAVE_A])) - res1 = slave_xmit(skb, hsr, HSR_DEV_SLAVE_A); - - res2 = NET_XMIT_DROP; - if (likely(skb2 && hsr->slave[HSR_DEV_SLAVE_B])) - res2 = slave_xmit(skb2, hsr, HSR_DEV_SLAVE_B); + res1 = slave_xmit(skb, hsr, HSR_DEV_SLAVE_A); + res2 = slave_xmit(skb2, hsr, HSR_DEV_SLAVE_B); if (likely(res1 == NET_XMIT_SUCCESS || res1 == NET_XMIT_CN || res2 == NET_XMIT_SUCCESS || res2 == NET_XMIT_CN)) { @@ -406,28 +407,10 @@ static void hsr_announce(unsigned long data) static void restore_slaves(struct net_device *hsr_dev) { struct hsr_priv *hsr; - int i; - int res; hsr = netdev_priv(hsr_dev); - - rtnl_lock(); - - for (i = 0; i < HSR_MAX_SLAVE; i++) { - if (!hsr->slave[i]) - continue; - res = dev_set_promiscuity(hsr->slave[i], -1); - if (res) - netdev_info(hsr_dev, - "Cannot restore slave promiscuity (%s, %d)\n", - hsr->slave[i]->name, res); - - if (hsr->slave[i]->rx_handler == hsr_handle_frame) - netdev_rx_handler_unregister(hsr->slave[i]); - } - - - rtnl_unlock(); + hsr_del_slave(hsr, 1); + hsr_del_slave(hsr, 0); } static void reclaim_hsr_dev(struct rcu_head *rh) @@ -483,38 +466,6 @@ bool is_hsr_master(struct net_device *dev) return (dev->netdev_ops->ndo_start_xmit == hsr_dev_xmit); } -static int check_slave_ok(struct net_device *dev) -{ - /* Don't allow HSR on non-ethernet like devices */ - if ((dev->flags & IFF_LOOPBACK) || (dev->type != ARPHRD_ETHER) || - (dev->addr_len != ETH_ALEN)) { - netdev_info(dev, "Cannot use loopback or non-ethernet device as HSR slave.\n"); - return -EINVAL; - } - - /* Don't allow enslaving hsr devices */ - if (is_hsr_master(dev)) { - netdev_info(dev, "Cannot create trees of HSR devices.\n"); - return -EINVAL; - } - - if (is_hsr_slave(dev)) { - netdev_info(dev, "This device is already a HSR slave.\n"); - return -EINVAL; - } - - if (dev->priv_flags & IFF_802_1Q_VLAN) { - netdev_info(dev, "HSR on top of VLAN is not yet supported in this driver.\n"); - return -EINVAL; - } - - /* HSR over bonded devices has not been tested, but I'm not sure it - * won't work... - */ - - return 0; -} - /* Default multicast address for HSR Supervision frames */ static const unsigned char def_multicast_addr[ETH_ALEN] __aligned(2) = { @@ -525,15 +476,30 @@ int hsr_dev_finalize(struct net_device *hsr_dev, struct net_device *slave[2], unsigned char multicast_spec) { struct hsr_priv *hsr; - int i; int res; hsr = netdev_priv(hsr_dev); hsr->dev = hsr_dev; + hsr->slave[0] = NULL; + hsr->slave[1] = NULL; INIT_LIST_HEAD(&hsr->node_db); INIT_LIST_HEAD(&hsr->self_node_db); - for (i = 0; i < HSR_MAX_SLAVE; i++) - hsr->slave[i] = slave[i]; + + ether_addr_copy(hsr_dev->dev_addr, slave[0]->dev_addr); + + /* Make sure we recognize frames from ourselves in hsr_rcv() */ + res = hsr_create_self_node(&hsr->self_node_db, hsr_dev->dev_addr, + slave[1]->dev_addr); + if (res < 0) + return res; + + hsr_dev->features = slave[0]->features & slave[1]->features; + /* Prevent recursive tx locking */ + hsr_dev->features |= NETIF_F_LLTX; + /* VLAN on top of HSR needs testing and probably some work on + * hsr_header_create() etc. + */ + hsr_dev->features |= NETIF_F_VLAN_CHALLENGED; spin_lock_init(&hsr->seqnr_lock); /* Overflow soon to find bugs easier: */ @@ -560,65 +526,21 @@ int hsr_dev_finalize(struct net_device *hsr_dev, struct net_device *slave[2], * IFF_HSR_MASTER/SLAVE? */ - for (i = 0; i < HSR_MAX_SLAVE; i++) { - res = check_slave_ok(slave[i]); - if (res) - return res; - } - - hsr_dev->features = slave[0]->features & slave[1]->features; - /* Prevent recursive tx locking */ - hsr_dev->features |= NETIF_F_LLTX; - /* VLAN on top of HSR needs testing and probably some work on - * hsr_header_create() etc. - */ - hsr_dev->features |= NETIF_F_VLAN_CHALLENGED; - - /* Set hsr_dev's MAC address to that of mac_slave1 */ - ether_addr_copy(hsr_dev->dev_addr, hsr->slave[0]->dev_addr); - - /* Set required header length */ - for (i = 0; i < HSR_MAX_SLAVE; i++) { - if (slave[i]->hard_header_len + HSR_HLEN > - hsr_dev->hard_header_len) - hsr_dev->hard_header_len = - slave[i]->hard_header_len + HSR_HLEN; - } - - /* MTU */ - for (i = 0; i < HSR_MAX_SLAVE; i++) - if (slave[i]->mtu - HSR_HLEN < hsr_dev->mtu) - hsr_dev->mtu = slave[i]->mtu - HSR_HLEN; - /* Make sure the 1st call to netif_carrier_on() gets through */ netif_carrier_off(hsr_dev); - /* Promiscuity */ - for (i = 0; i < HSR_MAX_SLAVE; i++) { - res = dev_set_promiscuity(slave[i], 1); - if (res) { - netdev_info(hsr_dev, "Cannot set slave promiscuity (%s, %d)\n", - slave[i]->name, res); - goto fail; - } - } - - for (i = 0; i < HSR_MAX_SLAVE; i++) { - res = netdev_rx_handler_register(slave[i], hsr_handle_frame, - hsr); - if (res) - goto fail; - } - - /* Make sure we recognize frames from ourselves in hsr_rcv() */ - res = hsr_create_self_node(&hsr->self_node_db, hsr_dev->dev_addr, - hsr->slave[1]->dev_addr); - if (res < 0) - goto fail; - res = register_netdevice(hsr_dev); if (res) - goto fail; + return res; + + res = hsr_add_slave(hsr, slave[0], 0); + if (res) + return res; + res = hsr_add_slave(hsr, slave[1], 1); + if (res) { + hsr_del_slave(hsr, 0); + return res; + } hsr->prune_timer.expires = jiffies + msecs_to_jiffies(PRUNE_PERIOD); add_timer(&hsr->prune_timer); @@ -626,8 +548,4 @@ int hsr_dev_finalize(struct net_device *hsr_dev, struct net_device *slave[2], register_hsr_master(hsr); return 0; - -fail: - restore_slaves(hsr_dev); - return res; } |